chapter 1 - ResearchSpace@Auckland - The University of Auckland
chapter 1 - ResearchSpace@Auckland - The University of Auckland
chapter 1 - ResearchSpace@Auckland - The University of Auckland
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Copyright Statement<br />
<strong>The</strong> digital copy <strong>of</strong> this thesis is protected by the Copyright Act 1994 (New<br />
Zealand). This thesis may be consulted by you, provided you comply with the<br />
provisions <strong>of</strong> the Act and the following conditions <strong>of</strong> use:<br />
• Any use you make <strong>of</strong> these documents or images must be for<br />
research or private study purposes only, and you may not make<br />
them available to any other person.<br />
• Authors control the copyright <strong>of</strong> their thesis. You will recognise the<br />
author's right to be identified as the author <strong>of</strong> this thesis, and due<br />
acknowledgement will be made to the author where appropriate.<br />
• You will obtain the author's permission before publishing any<br />
material from their thesis.<br />
To request permissions please use the Feedback form on our webpage.<br />
http://researchspace.auckland.ac.nz/feedback<br />
General copyright and disclaimer<br />
In addition to the above conditions, authors give their consent for the digital<br />
copy <strong>of</strong> their work to be used subject to the conditions specified on the Library<br />
<strong>The</strong>sis Consent Form
READING LAPITA IN NEAR OCEANIA:<br />
INTERTIDAL AND SHALLOW-WATER POTTERY<br />
SCATTERS, ROVIANA LAGOON, NEW GEORGIA,<br />
SOLOMON ISLANDS<br />
MATTHEW WALTER FELGATE<br />
A thesis submitted in partial fulfilment <strong>of</strong> the requirements for the degree <strong>of</strong><br />
Doctor <strong>of</strong> Philosophy in Anthropology, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>, 2003
ABSTRACT<br />
Lapita is the name given by archaeologists to a material culture complex distributed from<br />
Papua New Guinea to Samoa about 3000 years ago, which marks major economic changes<br />
in Near Oceania and the first settlement by humans <strong>of</strong> Remote Oceania. Those parts <strong>of</strong><br />
Solomon Islands that lie in Near Oceania, together with Bougainville, comprise a large gap<br />
in the recorded distribution <strong>of</strong> Lapita, which the current research seeks to explain. At<br />
Roviana Lagoon, centrally located in this gap, scatters <strong>of</strong> pottery, stone artefacts, and other<br />
stone items are found in shallow water in this sheltered, landlocked lagoon, initially thought<br />
to be late derivatives <strong>of</strong> Lapita. This research seeks method and theory to aid in the<br />
interpretation <strong>of</strong> this type <strong>of</strong> archaeological record.<br />
Intensive littoral survey discovered a wider chronological range <strong>of</strong> pottery styles<br />
than had previously been recorded, including materials attributable directly to the Lapita<br />
material culture complex. A study <strong>of</strong> vessel brokenness and completeness enabled sample<br />
evaluation, estimation <strong>of</strong> a parent population from which the sample derived, assessment<br />
<strong>of</strong> the state <strong>of</strong> preservation <strong>of</strong> the sample, and systematic choice <strong>of</strong> unit <strong>of</strong> quantification.<br />
Studies <strong>of</strong> wave exposure <strong>of</strong> collection sites and taphonomic evidence from sherds<br />
concluded that the cultural formation process <strong>of</strong> these sites was stilt house settlement (as<br />
found elsewhere in Near Oceania for Lapita) over deeper water than today. Falling relative<br />
sea levels and consequent increasing effects <strong>of</strong> swash-zone processes have resulted in high<br />
archaeological visibility and poor state <strong>of</strong> preservation at Roviana Lagoon.<br />
Analysis <strong>of</strong> ceramic and lithic variability and spatial analysis allowed the<br />
construction <strong>of</strong> a provisional chronology in need <strong>of</strong> further testing. Indications are that<br />
there is good potential to construct a robust, high-resolution ceramic chronology by<br />
focussing on carefully controlled surface collection from this sort <strong>of</strong> location, ceramic<br />
seriation and testing/calibration using direct dating by AMS radiocarbon and<br />
<strong>The</strong>rmoluminescence.<br />
Data on preservation and archaeological visibility <strong>of</strong> stilt house settlements along<br />
a sheltered emerging coastline allows preservation and visibility for this type <strong>of</strong> settlement<br />
to be modeled elsewhere. When such a model is applied to other areas <strong>of</strong> the Lapita gap,<br />
which are predominantly either less favourable for preservation or less favourable for<br />
archaeological visibility, the gap in the distribution <strong>of</strong> Lapita can be seen to be an area <strong>of</strong><br />
low probability <strong>of</strong> detection by archaeologists, meaning there is currently no evidence for<br />
absence <strong>of</strong> settlement in the past, and good reason to think that Lapita was continuously<br />
distributed across Near Oceania as a network <strong>of</strong> stilt village settlement. This finding<br />
highlights the need for explicit models <strong>of</strong> probability <strong>of</strong> detection to discover or read the<br />
Lapita archaeological record.<br />
Keywords: pottery; Lapita; formation processes; surface archaeology; tidal archaeology;<br />
Oceania
ACKNOWLEDGEMENTS<br />
Primary supervisor, Dr. Peter Sheppard, got the New Georgia Archaeological Survey <strong>of</strong>f<br />
the ground (and into the sea), an achievement that made this research possible. As a<br />
supervisor he gave me a lot <strong>of</strong> rope, as well as the occasional well-timed tug back to the<br />
topic, and has always been available when needed. Peter was instrumental in obtaining<br />
institutional financial support in the form <strong>of</strong> a <strong>University</strong> <strong>of</strong> <strong>Auckland</strong> Doctoral<br />
Scholarship, and a research stipend from the Marsden Fund. Field expenses and substantial<br />
analytical expenses were funded primarily from grants to the NGAS by the <strong>University</strong> <strong>of</strong><br />
<strong>Auckland</strong>, National Geographic Research and the Marsden Fund. Peter has also been a<br />
steady companion and mentor in the field and at all stages <strong>of</strong> the research, and he and<br />
Debbie have made my family and I welcome in their home on numerous occasions.<br />
This research in effect continues an academic program begun in the 1970s by<br />
Pr<strong>of</strong>essor Roger C. Green, as the Southeast Solomons Culture History Project, which<br />
eventually brought Peter Sheppard to this part <strong>of</strong> the world to work on Lapita lithics from<br />
Solomon Islands. It was Roger Green who first focused my attention on Oceanic<br />
prehistory, initially with an ample shirt emblazoned with pacific canoes, and later through<br />
an intense series <strong>of</strong> lectures on Oceanic Prehistory, capped by his last 3-week fieldschool<br />
before retirement from active fieldwork in the mid 1980s. Roger predicted in 1978 that<br />
Lapita would be found at New Georgia, and like Peter Sheppard, has been a mentor<br />
throughout. He also organized and personally funded petrographic analysis <strong>of</strong> ceramic<br />
tempers, and made his extensive personal library available.<br />
Dr. Simon Holdaway generously accepted the task <strong>of</strong> secondary supervision in midresearch<br />
when the rules changed, and has, through suggestions in the lab, comments on<br />
drafts, and by convening an archaeology theory reading group, contributed substantially<br />
to the outcome.<br />
Material and moral support from family were essential to successful completion <strong>of</strong><br />
the research, especially from my parents, who largely financed the final years, and elder<br />
brother, who was always ready to help pack up the house and store incredible<br />
accumulations <strong>of</strong> household effects every time we decamped to the Solomons, and is still<br />
doing so.<br />
In Solomon Islands, support and encouragement from Mr Lawrence Foanoata,<br />
Director <strong>of</strong> the National Museum has been constant and essential. Permissions from the<br />
Research committee <strong>of</strong> the Ministry <strong>of</strong> Education, and from the Western Province<br />
Administration were key to the success <strong>of</strong> the research. Mr John Keopo <strong>of</strong> the National<br />
Museum was instrumental in the smooth running <strong>of</strong> the first field season. Rhys and<br />
Margaret Richards provided hospitality in Honiara on more than one occasion, and Rhys<br />
was instrumental as the New Zealand High Comissioner in obtaining a residence permit<br />
during extended fieldwork.<br />
II
Mr Kenneth Roga, Western Province Field Archaeologist, has been a primary architect <strong>of</strong><br />
fieldwork throughout. Unflagging energy in the field, his encyclopaedic network <strong>of</strong> wantok<br />
which has ensured a warm welcome on so many occasions, commitment to archaeology<br />
despite difficult political circumstances, and a range <strong>of</strong> talents too numerous to mention<br />
have been invaluable. A rock-steady figure at the helm <strong>of</strong> a canoe over many long journeys<br />
through sometimes dangerous seas, his and Janet Roga’s hospitality in Gizo on so many<br />
occasions is deeply appreciated.<br />
At Roviana Lagoon, permission for research was received from the late Chief<br />
Johnathan Roni. Chief Joseph Kama <strong>of</strong> Kaliquongu and Chief Nathan Kera <strong>of</strong> Saikile each<br />
consented to research in areas under their jurisdiction. Thanks are also due to Mr. Solomon<br />
Roni and his family, especially all at Miho in Sasavele: Sam Roni and Waring, Martha<br />
Roni, Aggie Roni and Abel Kae, my father-in-law Sale Maebule and late mother-in-law<br />
Mabelo Roni, Sitiveni, my wife Noseduri, and many <strong>of</strong> the next generation too numerous<br />
to mention. Thanks also are due to members <strong>of</strong> the Elelo community who helped out in so<br />
many ways during our extended stay, especially Hetti Lanni, Barrie and Selina Ford and<br />
George and Visa Sapolo. <strong>The</strong> late Mr Phillip Lanni generously assisted with his time during<br />
a visit to Gharanga, despite ill health. In Munda, thanks are due to David Kera and family,<br />
Dave and Mariana Cook, and Trevor and Zahi Cumberland, for assistance with<br />
communications and travel on many occasions. In Saikile, John Kororo and family abetted<br />
us on our search for pottery at Mbaraulu, and Sae Oka was a prime mover in the survey<br />
for pottery on Ndora Island.<br />
During analysis and writeup at <strong>Auckland</strong> many individuals contributed their skills:<br />
Dilys Johns and Dr Rod Wallace supplied technical and conservation advice, Hamish Mc<br />
Donald and Tim Mackrell took all the artefact photographs (or at least all the good ones),<br />
Joan Lawrence illustrated the pottery and lithic artefacts, Dr Simon Bickler collaborated<br />
on the problem <strong>of</strong> sample interpretation, and wrote a simulation program to estimate<br />
parent population <strong>of</strong> vessels. Dr Robin Parker <strong>of</strong> the Geology Department provided thin<br />
section descriptions <strong>of</strong> lithics. Barry Curnham, also in Geology, showed me how to make<br />
epoxy-impregnated ceramic thin sections. Peter Sheppard examined chert thin sections. Bill<br />
Dickinson <strong>of</strong> the <strong>University</strong> <strong>of</strong> Arizona analysed and reported on ceramic thin sections, and<br />
collaborated on a paper writing up the results. Jim Feathers <strong>of</strong> the Anthropology<br />
Department, <strong>University</strong> <strong>of</strong> Washington, did the thermoluminescence analysis and was<br />
generous with his time in discussing progress and results. Dr Christine Prior <strong>of</strong> the Rafter<br />
Radiocarbon Lab organized AMS dating <strong>of</strong> samples from Honiavasa and Hoghoi, including<br />
some careful experimentation with the Honiavasa sample. Pr<strong>of</strong>essor Jim Allen gave<br />
permission to cite the 1984 and 1985 Lapita Homeland Project Field Reports.<br />
Discussions with Dr Stuart Bedford, Moira Doherty, Dr Simon Best and Dr Simon Bickler,<br />
contributed to the research and identified several errors. Stuart, Simon Best and Moira<br />
provided useful comments on parts <strong>of</strong> the manuscript. Stuart and Carolyn made me<br />
welcome in their home during the final stages <strong>of</strong> writing.<br />
III
CHAPTER 1:<br />
CHAPTER 2:<br />
Contents:<br />
RESEARCH QUESTIONS AND METHODOLOGY ................ 1<br />
Introduction: ................................................. 1<br />
<strong>The</strong> Research Region: .......................................... 6<br />
<strong>The</strong> Roviana Early Ceramic Archaeological Record: ............... 14<br />
Research Questions: .......................................... 15<br />
Approaches to Ceramic Classification and Analysis: ................ 21<br />
Quantification: ............................................... 34<br />
Seriation <strong>The</strong>ory : a Review: ................................... 43<br />
Form and Function: a Review : ................................. 57<br />
Chapter summary and conclusions: .............................. 70<br />
A REVIEW OF CONSTRUCTIONS OF LAPITA TEMPORAL<br />
VARIABILITY .............................................. 77<br />
Introduction: ................................................ 77<br />
Green’s 1978 Lapita Ceramic Series: ............................ 79<br />
Anson’s Early Far Western Lapita: .............................. 85<br />
Mussau: ..................................................... 88<br />
<strong>The</strong> “Changing” Face <strong>of</strong> Lapita: ................................ 89<br />
Summerhayes, West New Britain, Anir, and a Three-stage Lapita Ceramic<br />
Series: ................................................ 93<br />
<strong>The</strong> Watom Lapita Series: .....................................103<br />
Wahome’s Seriation <strong>of</strong> Admiralties Pottery Assemblages: ...........107<br />
Specht on Buka: ..............................................109<br />
Wickler on Buka: .............................................111<br />
Vanuatu: ....................................................116<br />
New Caledonia: ..............................................119<br />
<strong>The</strong> View from the East: Fiji and Tonga: .........................123
Lapita Temporal Variability -Conclusions: ...................... 125<br />
CHAPTER 3:<br />
CHAPTER 4:<br />
CHAPTER 5:<br />
SCALE AND METHOD OF FIELD SURVEY REQUIRED TO GENERATE<br />
A SAMPLE OF LAPITA SITES FOR SERIATION IN THE NEW<br />
GEORGIA REGION ......................................... 133<br />
Introduction: ............................................... 133<br />
Review <strong>of</strong> Near-oceanic Survey Methods and Results: ............. 139<br />
Sampling <strong>The</strong>ory and the Roviana/Kaliquongu Survey Regions: ..... 158<br />
Survey Methods: ............................................ 163<br />
Results: .................................................... 188<br />
Discussion and Conclusions: the Two Surveys. .................... 190<br />
CERAMICS: UNITS OF DESCRIPTION ....................... 193<br />
Introduction: ............................................... 193<br />
Database Structure: .......................................... 195<br />
Thickness, Form and Decoration <strong>of</strong> Various Vessel Parts Represented on<br />
Sherds: .............................................. 202<br />
Vessel Form: ............................................... 205<br />
Decoration: ................................................. 219<br />
Transforming Relational Data into a Flat Table: .................. 242<br />
Chapter Summary and Conclusions: ............................ 244<br />
SAMPLE EVALUATION AND QUANTIFYING SAMPLE SIZE. ... 247<br />
Introduction ................................................ 247<br />
Methods for Establishing Vessel Brokenness and Completeness: ..... 248<br />
Simulation Approach: ........................................ 253<br />
Statistical Approaches: ....................................... 257<br />
Simulation Results: .......................................... 259<br />
Statistical Results: ........................................... 262
Discussion: ..................................................263<br />
CHAPTER 6:<br />
CHAPTER 7:<br />
CHAPTER 8:<br />
Summary and Conclusions: ....................................263<br />
WAVE EXPOSURE ..........................................269<br />
Introduction: ................................................269<br />
Review: .....................................................270<br />
Roviana Intertidal Ceramics and Waves: .........................272<br />
Solomon Islands Tides: ........................................277<br />
Winds: ......................................................280<br />
Measuring Fetch: .............................................282<br />
Results: .....................................................283<br />
Discussion: ..................................................283<br />
Chapter Summary and Conclusions: ......................286<br />
SITE/ASSEMBLAGE FORMATION PROCESSES (SHERD<br />
TAPHONOMY). .............................................289<br />
Introduction: ................................................289<br />
Review: Middle Range <strong>The</strong>ory for the Identification <strong>of</strong> Formation Processes:<br />
......................................................291<br />
Models <strong>of</strong> Cultural Formation Process: ...........................298<br />
Models <strong>of</strong> Post-depositional Formation Process: ...................299<br />
Summary <strong>of</strong> Models: ..........................................301<br />
Qualitative Evidence From Samples: .............................302<br />
Quantitative Evidence: Size/Density Effects and the Structure and Context<br />
<strong>of</strong> the Deposit: ..........................................305<br />
Sea Level History: ............................................312<br />
Chapter Summary and Conclusions: .............................313
VESSEL FORM VARIABILITY AND VESSEL FUNCTION ....... 319<br />
CHAPTER 9:<br />
Introduction: ............................................... 319<br />
Form: Initial Nominal Classification: ........................... 322<br />
Metric Variability Within Form 6: ............................. 331<br />
Inferring Vessel Form from Body Sherds: ....................... 339<br />
Vessel Function from Evidence for Use-Alteration: ............... 346<br />
Secondary Smoothing <strong>of</strong> Vessel Interiors, and Interior Neck Form: . . 347<br />
Shoulder Form Classes <strong>of</strong> Form 6 Vessels: ...................... 351<br />
Sherd Restriction Factor: ..................................... 353<br />
Chapter Summary: .......................................... 355<br />
CERAMIC DECORATIVE CLASSIFICATION AND<br />
DECORATION/FORM VARIABILITY: ....................... 359<br />
Introduction: ............................................... 359<br />
Classification <strong>of</strong> Linear Motifs: ................................ 360<br />
Part Representation in the Potsherd Sample: ..................... 361<br />
Covariation Between Vessel Part and Decoration: ................. 364<br />
Summary <strong>of</strong> Decorative Structure Across the Vessel: .............. 371<br />
Interpretation <strong>of</strong> Decorative Co-variation by Vessel Part: .......... 373<br />
Attribute Frequencies: ....................................... 374<br />
Variability <strong>of</strong> Impressed Lips: ................................. 375<br />
Decorative Attributes and Vessel Form: ......................... 383<br />
Chapter Summary: .......................................... 391<br />
CHAPTER 10:<br />
LITHICS ................................................... 395<br />
Introduction: ............................................... 395<br />
Review: .................................................... 398<br />
Roviana Lithic Artefacts: ..................................... 401<br />
Chert Flakes/Fragments: ..................................... 411<br />
Analysis <strong>of</strong> Hoghoi Water-rounded and Fractured Volcanic Manuports:
......................................................412<br />
Chapter Summary: ...........................................417<br />
CHAPTER 11:<br />
INTRASITE SPATIAL STRUCTURE OF CERAMIC AND LITHIC<br />
VARIABILITY: ..............................................421<br />
Introduction: ................................................421<br />
Selection <strong>of</strong> Sites for Intrasite Spatial Analysis: ....................423<br />
Objectives: ..................................................423<br />
Method .....................................................425<br />
Zangana: ....................................................428<br />
Hoghoi: .....................................................438<br />
Honiavasa: ..................................................443<br />
Chapter Summary and Conclusions. .............................447<br />
CHAPTER 12:<br />
CHRONOLOGY .............................................451<br />
Introduction: ................................................451<br />
AMS Radiocarbon Dates on Potsherds: ..........................453<br />
<strong>The</strong>rmoluminescence (TL) Dates from Quartz-Calcite Sherds: .......458<br />
Roviana TL Data: ............................................463<br />
Seriation: ....................................................465<br />
Discussion <strong>of</strong> Correspondence Analyses: the Alternatives: ...........478<br />
Conclusions: Integrating 14 C, TL and Seriation: ...................480<br />
CHAPTER 13:<br />
SUMMARY AND CONCLUSIONS .............................483<br />
Introduction: ................................................483<br />
Summary: ...................................................484<br />
External Comparisons: ........................................498<br />
<strong>The</strong> Lapita Gap as an Area <strong>of</strong> Low Probability <strong>of</strong> Detection : ........503<br />
Intertidal-Zone and Shallow-Water Archaeology: ..................504
REFERENCES .................................................... 509
List <strong>of</strong> Tables:<br />
Table 1: Sampling, assemblage richness, and the Jaccard coefficient (p=present, a=absent).<br />
............................................................ 52<br />
Table 2: Reef/ Santa Cruz motif counts as given in Anson 1983. .............. 80<br />
Table 3: Sherd counts and MNI assembled from various tables in Parker (1981). . 81<br />
Table 4: Relative proportions <strong>of</strong> dentate and incised, recalculated from data presented by<br />
Donovan (1973). .............................................. 82<br />
Table 5: Summary <strong>of</strong> a review <strong>of</strong> survey methods and Lapita results. ...........155<br />
Table 6: Site densities by period for Roviana and Kaliquongu surveys. .........188<br />
Table 7: Structure <strong>of</strong> master record for each sherd. .........................196<br />
Table 8: Classes <strong>of</strong> mineral identified at 10x magnification in reflected light. ....199<br />
Table 9: Examples <strong>of</strong> descriptive syntax for tempers. .......................200<br />
Table 10: Temper groupings after Dickinson 2000. .........................201<br />
Table 11: Data structure <strong>of</strong> table <strong>of</strong> thicknesses for each part <strong>of</strong> the sherd. .......204<br />
Table 12: Data structure for the table <strong>of</strong> records <strong>of</strong> sherd form attributes by vessel part.<br />
............................................................206<br />
Table 13: Data structure for decoration records.............................220<br />
Table 14: Decorative techniques. .......................................221<br />
Table 15: Decoration pattern definitions..................................229<br />
Table 16: Decorative elements..........................................240<br />
Table 17: Data structure for flat table <strong>of</strong> summary data <strong>of</strong> sherd properties. ......245<br />
Table 18: Variation in lip brokenness between sites. ........................250<br />
Table 19: Selection <strong>of</strong> sample for estimating vessel completeness. .............252<br />
Table 20: Breakage population estimate for the combined Roviana highly decorated lip<br />
sample using the statistic <strong>of</strong> Chao 1984. ............................262<br />
Table 21: Fetch measurements for collection sites as an indicator <strong>of</strong> wave exposure.<br />
............................................................283<br />
Table 22: Total sherd count and average sherd area by collection site, resulting from<br />
combined effects <strong>of</strong> collection intensity and wave exposure. ............286<br />
Table 23: Ratio <strong>of</strong> lip-rim sherds to body sherds as a measure <strong>of</strong> collector effect.
........................................................... 309<br />
Table 24: Ratios <strong>of</strong> plain to decorated sherds, controlled by vessel part (lips, rims, necks<br />
and shoulders. ............................................... 310<br />
Table 25: Body sherd to neck sherd ratio as an index <strong>of</strong> sherd skipping. ........ 311<br />
Table 26: Counts <strong>of</strong> vessel form classes. ................................ 336<br />
Table 27: <strong>The</strong> attribute “even/not even” by site. .......................... 348<br />
Table 28: Relative abundance <strong>of</strong> “hard neck” interior pr<strong>of</strong>iles to “not hard”. .... 350<br />
Table 29: Relative abundance <strong>of</strong> even interiors with hard neck interior pr<strong>of</strong>iles to even<br />
interiors without hard pr<strong>of</strong>ile. ................................... 350<br />
Table 30: Sherd restriction factor comparison <strong>of</strong> site samples. ............... 353<br />
Table 31: Preservational bias <strong>of</strong> vessel parts as an explanation for differences in sherd<br />
restriction factor (data has the form average thickness (count) standard deviation).<br />
........................................................... 354<br />
Table 32: Part representation, all sherds including necked sherds but excluding carinated<br />
and/or inverted-rim sherds. ..................................... 362<br />
Table 33: Part representation for inverted-rim sherds (no neck corner point expected from<br />
morphology <strong>of</strong> sherd). ......................................... 362<br />
Table 34: Carinated sherds (excluding carinated sherds with inverted rims). .... 363<br />
Table 35: Form strength variation for different pottery styles and the effect on part<br />
representation. ............................................... 363<br />
Table 36: Lip decoration and rim decoration.............................. 365<br />
Table 37: Lip decoration and neck decoration............................. 366<br />
Table 38: Cross-tabulation <strong>of</strong> lip decoration with shoulder decoration. ......... 367<br />
Table 39: Cross-tabulation <strong>of</strong> rim decoration and neck decoration. ............ 368<br />
Table 40: Cross-tabulation <strong>of</strong> rim decoration and shoulder decoration. ......... 369<br />
Table 41: Cross-tabulation <strong>of</strong> neck decoration and shoulder decoration. ........ 370<br />
Table 42: Counts <strong>of</strong> occurrences <strong>of</strong> decorative attributes. ................... 375<br />
Table 43: Petrographic descriptions <strong>of</strong> lithic artefacts. ...................... 407<br />
Table 44: Petrographic classification <strong>of</strong> manuports collected at Hoghoi. ....... 414<br />
Table 45: Size comparisons between petrographic classes <strong>of</strong> lithic manuports. . . 415<br />
Table 46: Radiocarbon determinations from pottery (calibrated using OxCal 3.5, Stuiver<br />
et al. 1998 atmospheric data). ................................... 454
Table 47: <strong>The</strong>rmoluminescence dating results. Precision shown is at confidence limits <strong>of</strong><br />
one standard deviation. .........................................464<br />
Table 48: Definitions <strong>of</strong> attribute codes used in seriation tables and plots. .......465<br />
Table 49: Relative contribution <strong>of</strong> first and second components <strong>of</strong> CA using all attributes<br />
and forms. ...................................................468<br />
Table 50: CA diagnostics by attribute, all attributes included. .................468<br />
Table 51: CA diagnostics by site, all attributes included (*=inertia outlier). ......470<br />
Table 52: Eigenvalues and contributions to intertia <strong>of</strong> CA components, for data excluding<br />
form-correlated attributes. .......................................473<br />
Table 53: CA diagnostic table for attributes, form-correlated attributes omitted (*=intertia<br />
outlier). ......................................................473<br />
Table 54: CA diagnostics by site, form-correlated attributes omitted (*=inertia outlier).<br />
............................................................475<br />
Table 55: CA diagnostics for attributes, omitting Honiavasa data and form-correlated<br />
attributes. ....................................................477<br />
Table 56: CA diagnostics by site, omitting Honiavasa data and form-correlated attributes.<br />
............................................................477<br />
Table 57: CA eigenvalues and component contributions to inertia omitting Honiavasa data<br />
and form-correlated attributes ....................................477<br />
Table 58: Summary <strong>of</strong> variability .......................................479
List <strong>of</strong> Figures:<br />
Figure 1: Near Oceania, Remote Oceania and Solomon Islands, showing the location <strong>of</strong><br />
the New Georgia Group. ......................................... 3<br />
Figure 2: Map <strong>of</strong> New Georgia Group showing principal geological formations (after<br />
Coulson, Dunkerly, Hughes and Ridgeway 1987). ..................... 9<br />
Figure 3: <strong>The</strong> author providing scale for solution notches in Plio-Pleistocene limestone<br />
cliff near Saikile passage, Roviana Lagoon (photograph courtesy <strong>of</strong> Peter<br />
Sheppard) .................................................... 14<br />
Figure 4: Paniavile at low tide, looking north, with several inhabited islets and the New<br />
Georgia mainland in the distance. ................................. 16<br />
Figure 5: Processes <strong>of</strong> formation <strong>of</strong> the Archaeological record (After Felgate and Bickler<br />
n.d., adapted from De Boer 1983): inferring a breakage population from an<br />
archaeological sample........................................... 44<br />
Figure 6: Roviana and Kaliquongu surveys as samples <strong>of</strong> the New Georgia lagoon and<br />
barrier island system. ...........................................158<br />
Figure 7: Slab-built carinated vessels from Honiavasa initially assigned to the Lapita<br />
period (this assignment is examined in more detail in Chapters 8 and 9): the solid<br />
vertical line in HV.2.464 represents an estimated location <strong>of</strong> the central vertical<br />
avis (CVA) <strong>of</strong> the pot. ..........................................167<br />
Figure 8: Carinated vessels from Honiavasa, showing double-line markes on some design<br />
zones, bands <strong>of</strong> nubbins at the neck in two cases, and a band <strong>of</strong> fingernail<br />
impression in one case (top). .....................................168<br />
Figure 9: HV.4.202 has an incised motif laid out in double lines; HV.1.314 is dentate-<br />
stamped, with similar design structure and a double carination; HV.2.341 is<br />
carinated, with the design laid out in applied fillets, bounded horizontally by<br />
decorated lap joins between slabs; HV.2.297 and HV.4.379 have bands <strong>of</strong> single<br />
fingernail impressions and nubbins at the neck respectively. ............169<br />
Figure 10: Incised rims with opposed-pinch fingernail impressed band at the neck,<br />
diagnostic <strong>of</strong> the Miho subgroup <strong>of</strong> Post-Lapita styles. ................171<br />
Figure 11: Miho-style post-Lapita sherds. ................................172<br />
Figure 12: Miho-style post-Lapita sherds with incised shoulders. ..............172
Figure 13: Miho-style post-Lapita sherds with CVA measurements shown (MH290 was<br />
measured at to locations on the pr<strong>of</strong>ile; at the interior <strong>of</strong> the neck orifice and the<br />
exterior shoulder). ............................................ 173<br />
Figure 14: Miho-style post-Lapita sherds; dashed CVA lines are measurements based on<br />
non-circular (uneven) curvature at the pr<strong>of</strong>ile locations indicated by arrows.<br />
........................................................... 174<br />
Figure 15: Gharanga-style post-Lapita sherds. (Gharanga is a short-rim subgroup <strong>of</strong><br />
Gharanga-Kopo which may have multiple bands <strong>of</strong> opposed-pinch fingernail<br />
impression on the shoulder. ..................................... 175<br />
Figure 16: Gharanga-style post-Lapita sherds. ............................ 175<br />
Figure 17: Gharanga/Kopo style post-Lapita sherds (top) and a large Gharanga-style<br />
sherd (bottom) (four measurement points used as arrowed to estimate CVA.<br />
........................................................... 176<br />
Figure 18: Intermediate between Gharanga and Kopo styles: all post-Lapita. .... 177<br />
Figure 19: Kopo-style post-Lapita rim sherds. ............................ 177<br />
Figure 20: Another large Gharanga-style post-Lapita sherd from an unrestricted vessel<br />
form........................................................ 178<br />
Figure 21: Gharanga-style small-orifice post-Lapita sherd showing rolled rim and thin wall<br />
common to this style. .......................................... 178<br />
Figure 22: Kopo-Style post-Lapita sherd (taller, less everted rim than Gharanga style)<br />
from a large-orifice vessel, with bands <strong>of</strong> impressions along both inner and outer<br />
edges <strong>of</strong> the lip. .............................................. 178<br />
Figure 23: Less decorated variant <strong>of</strong> Gharanga-Kopo post-Lapita style, without a strong<br />
corner point in vertical section at the neck. ......................... 179<br />
Figure 24: Large-orifice Gharanga/Kopo-style sherd with deformation <strong>of</strong> the lip into a<br />
wave pattern. ................................................ 179<br />
Figure 25: Large Gharanga-style post-Lapita sherd with typical decoration, including a<br />
band <strong>of</strong> impressions along the inner edge <strong>of</strong> the lip. .................. 179<br />
Figure 26: Gharanga/Kopo-style post-Lapita vessels: most are weakly restricted at the<br />
neck, with short, heavily everted rims. Punctate band at the neck is the most<br />
common decoration in this group, while multiple bands <strong>of</strong> fingernail pinch are<br />
common on the short-rim examples. One sherd (MH.33) had exotic quartz-calcite
hybrid temper (see Chapter 4). ...................................180<br />
Figure 27: Locations mentioned in the text in relation to Roviana and Kaliquongu surveys.<br />
............................................................183<br />
Figure 28: Kaliquongu survey transects: the unfilled symbols represent sites discovered<br />
by informant-prospection during the Roviana survey. .................189<br />
Figure 29: Data structure for each sherd record; each sherd can have many records in the<br />
detail tables pertaining to the various parts <strong>of</strong> the vessel represented. .....195<br />
Figure 30: Major vessel form variants showing part terminology; L=lip, R=rim, N=neck,<br />
S=shoulder, C=carination, B=body; inverted or unrestricted vessels have no neck,<br />
while for restricted vessels with everted rims (the first seven) the only distinction<br />
in neck types is between the double neck (top centre) and the single neck<br />
(including all unlabelled). .......................................203<br />
Figure 31: Lip form variants showing database codes .......................207<br />
Figure 33: Measurement <strong>of</strong> rim depth, rim Vcurve, neck angle and neck Vcurve at<br />
different levels <strong>of</strong> brokenness. ....................................212<br />
Figure 34: Shoulder form measurement. ..................................213<br />
Figure 35: Derivation <strong>of</strong> conical/cannister ( C ) and spheroidal (G) sherds from various<br />
body forms. ..................................................216<br />
Figure 36: Form codes for vessel interiors. ...............................217<br />
Figure 37: Possible conical base sherds from robust vessels. ..................218<br />
Figure 38: Examples <strong>of</strong> applied decoration. ...............................223<br />
Figure 39: Examples <strong>of</strong> applied decoration. ...............................223<br />
Figure 40: Applied decoration on compound rims. .........................224<br />
Figure 41: Examples <strong>of</strong> deformation <strong>of</strong> the lip into a discontinuous band. .......224<br />
Figure 42: Horizontal deformation <strong>of</strong> the lip into a continuous band in a wave pattern.<br />
............................................................225<br />
Figure 43: Examples <strong>of</strong> discontinuous deformation. ........................225<br />
Figure 44: Excision <strong>of</strong> outer lip to form a band <strong>of</strong> notches. ...................226<br />
Figure 45: Compound rims from Nusa Roviana. NR.34 has excised lines forming the<br />
triangle pattern on the upper rim. ..................................226<br />
Figure 46: Impressions on the top <strong>of</strong> the lip; and spatula impressions on the neck, thought<br />
to indicate forming <strong>of</strong> the neck using a tool. .........................227
Figure 47: Excision by rotation: one end <strong>of</strong> a small twig or rod has been poked into the<br />
clay and the other end moved in a circle, leaving a conical hole. ....... 227<br />
Figure 48: Perforation (upper hole). .................................... 227<br />
Figure 49: Examples <strong>of</strong> wavy stamping and an example <strong>of</strong> dentate-stamping. . . . 228<br />
Figure 50: Applied decoration (in combination with incised decoration) with detachment<br />
scars indicating a v-pattern, and also possible attached disc. ........... 228<br />
Figure 51: A band <strong>of</strong> fingernail impressions (opposed pinch), each <strong>of</strong> which is oriented<br />
diagonal to the CVA rather than vertically, or parallel to the CVA. ...... 230<br />
Figure 52: Deformation, perforation, and the “bnd” incomplete pattern example.<br />
........................................................... 230<br />
Figure 53: “cf” (Crow’s foot) pattern on the vessel rim above a band <strong>of</strong> pinching.<br />
........................................................... 230<br />
Figure 54: Examples <strong>of</strong> lip impression in pattern “bpi” (band parallel inner-edge <strong>of</strong> lip).<br />
........................................................... 231<br />
Figure 55: Examples <strong>of</strong> lip impressions laid out in pattern “bpo” (band parallel outer-<br />
edge). ...................................................... 231<br />
Figure 56: Fragmented examples with lip impressions assigned to “band parallel outer”<br />
pattern. ..................................................... 232<br />
Figure 57: Examples <strong>of</strong> lip impressions/incision laid out in patterns “bot” (band opposing<br />
top), “bdt” (band diagonal top) and “bpt” (band parallel top), which were regarded<br />
as equivalent in analysis due to non-exclusive nature <strong>of</strong> these descriptions.<br />
........................................................... 232<br />
Figure 58: Pattern expressed using linear arrangements <strong>of</strong> fingernail pinching. . . 233<br />
Figure 59: Miscellaneous incised patterns: MH360 is middle row, left-hand column.<br />
GW258 is bottom-right. ........................................ 233<br />
Figure 60: Examples <strong>of</strong> applied nubbins and some bounded incised patterns (MH259 is an<br />
unbounded pattern). ........................................... 234<br />
Figure 61: Decorated sherds with quartz-calcite hybrid granitic temper. ........ 234<br />
Figure 62: Unbounded incised patterns. ................................. 235<br />
Figure 63: Thin incised rims from Hoghoi. .............................. 235<br />
Figure 64: Unbounded incised pattern on the shoulder from Paniavile. ......... 235<br />
Figure 65: An example <strong>of</strong> “chv” pattern, a band <strong>of</strong> unbounded linear triangles filled by
alternating fields <strong>of</strong> parallel lines. .................................236<br />
Figure 66: Example <strong>of</strong> “chv” pattern on tall rim (pinching at neck). ............236<br />
Figure 67 Curvilinear incised patterns. ...................................237<br />
Figure 68: Crow’s foot mark, probably post-deposition scratching. ............237<br />
Figure 69: Example <strong>of</strong> cross-hatch pattern “ct3”............................237<br />
Figure 70: Cross hatch pattern “cvh”. ...................................238<br />
Figure 71: Example <strong>of</strong> pattern “gm5". ...................................238<br />
Figure 72: Sole example <strong>of</strong> pattern “rl1", a double-line arrangement <strong>of</strong> single repeated<br />
fingernail impressions. .........................................238<br />
Figure 73 Roviana vessel completeness against sampling fraction, sampling fraction<br />
against vessel representation fraction; random assignment <strong>of</strong> shreds to vessels;<br />
mean EVE <strong>of</strong> 5.78%. ..........................................260<br />
Figure 74: Roviana vessel completeness against vessel representation fraction; random<br />
assignment <strong>of</strong> sherds to vessels; mean EVE <strong>of</strong> 5.78%. .................261<br />
Figure 76: Effect <strong>of</strong> sherd thickness on sherd strength (controlled for temper variation by<br />
using placered volcanic tempered sherds only). ......................306<br />
Figure 77: Body sherd size for the various temper classes. ...................306<br />
Figure 78: Histogram <strong>of</strong> body sherd size classes by site. .....................308<br />
Figure 79: Initial classification <strong>of</strong> vessel forms. ............................321<br />
Figure 80: Examples <strong>of</strong> unrestricted Form 2 variants: Form 2a (top); Form 2b (upper<br />
middle); Form 2c (lower middle); and Form 2d (bottom). ..............325<br />
Figure 81: Additional Form 2a Sherds (top 6); the two decorated gambrelled vessels from<br />
Nusa Roviana are a short-rim variant <strong>of</strong> Form 2a; the lower sherd is transitional<br />
between Form 2a and Form 2d, being unrestricted. ...................326<br />
Figure 82: Form 3 inverted restricted vessels. .............................329<br />
Figure 83: Form 3 or Form 4 (top left) and another example <strong>of</strong> a short-rim variant <strong>of</strong><br />
Form 2a (top right); the sole example <strong>of</strong> Form 3 with loop handle(s) (2 nd to top);<br />
the only confirmed Form 4 carinated bowl, in exotic temper (2 nd to bottom) and a<br />
large base sherd or frying pan, Form 5 (bottom). .....................330<br />
Figure 84: External neck radius, all sites, size intervals 10mm.................331<br />
Figure 85: All sites, sherds >25cm 2 , neck Hcurve intervals 10mm..............332<br />
Figure 86: Form 6 “Neck Hcurve” variation; the two top sherds are too small to get a
hand into (Form 6a), while the two lower sherds have head-sized or larger orifices<br />
(Form 6b). .................................................. 334<br />
Figure 87: Relationship between rim angle and rim height for two sites, all measurements<br />
shown. ..................................................... 337<br />
Figure 88: Rim angle and rim depth, all sites, filtered so that EVE is greater than 9%, to<br />
reduce the effects <strong>of</strong> measurement error. .......................... 338<br />
Figure 89:Body sherd curvature measurements, showing spheroidal sherds (diagonal<br />
alignment) and other canister/conical forms (e.g. s-shaped pattern <strong>of</strong> Honiavasa<br />
sherd measurements). ......................................... 340<br />
Figure 90: Comparison <strong>of</strong> body sherd curvature measurements by site. ........ 342<br />
Figure 91: Robust base sherds and cannister-shaped large body sherd (the latter having<br />
exotic quartz-calcite temper). ................................... 344<br />
Figure 92: Curvature <strong>of</strong> robust body sherds (thicker than 14mm). ............. 345<br />
Figure 93: Hard interior neck pr<strong>of</strong>ile and even interior body/shoulder pr<strong>of</strong>ile (top); a<br />
s<strong>of</strong>ter neck interior pr<strong>of</strong>ile (middle) and uneven interior pr<strong>of</strong>ile (bottom). . 349<br />
Figure 94: Hard shoulder variants <strong>of</strong> Form 6 (top and middle) as distinct from the more<br />
common s<strong>of</strong>t shoulder form (bottom). ............................. 352<br />
Figure 95: Tall, fragile everted excurvate rims. ........................... 356<br />
Figure 96: Lip impression, single band on outer edge <strong>of</strong> lip, labeled by mark section: u=u-<br />
shaped, v=v-shaped, o=oblique v, w=w-shaped, s=flat-bottomed groove. . 379<br />
Figure 97: Lip impression, single band on inner edge <strong>of</strong> lip, labeled by mark section: u=u-<br />
shaped section, etc.. ........................................... 379<br />
Figure 98: Lip impression, single band on top face <strong>of</strong> lip, labeled by mark section: u=u-<br />
shaped, etc.. ................................................. 380<br />
Figure 99: Lip impression, bands on both edges <strong>of</strong> the lip. Labeled by mark section: u=u-<br />
shaped, etc.. ................................................. 380<br />
Figure 100: Calculation <strong>of</strong> lip orientation angle. .......................... 382<br />
Figure 101: Lip orientation and location <strong>of</strong> impressions..................... 383<br />
Figure 102: Rim depth by decorative class. .............................. 384<br />
Figure 103: Rim depth and rim angle <strong>of</strong> undecorated lip-rim-neck sherds. ...... 387<br />
Figure 104: Location <strong>of</strong> bands <strong>of</strong> lip impression in relation to rim form variability.<br />
........................................................... 388
Figure 105: Neck thickness comparison <strong>of</strong> Gharanga/Kopo and Miho styles/types.<br />
............................................................389<br />
Figure 106: Neck Vcurve for Gharanga/Kopo decoration, Miho Decoration, and plain<br />
rim-plain neck-plain shoulder sherds. ..............................390<br />
Figure 107: Planilateral sectioned adze from Hoghoi initial surface collection (top), plano-<br />
convex-sectioned type V adze from Zangana (middle), and planilateral adze<br />
fragment from Zangana (bottom). .................................402<br />
Figure 108: Images <strong>of</strong> the adzes shown in the preceding illustration. ...........403<br />
Figure 109: Green “type VI” or “type VIII” triangular-section adze fragment from Miho<br />
(top); butt-end <strong>of</strong> a plano-lateral-sectioned adze from Zangana South (middle) and<br />
a fragment <strong>of</strong> a trapezoidal-sectioned Green “type IV” adze from Zangana.<br />
............................................................404<br />
Figure 110: Images <strong>of</strong> adzes illustrated on previous page. ....................405<br />
Figure 111: Canarium hammerstone from H5 ceramic findspot (top) (a similar artefact<br />
was found at Hoghoi); waisted sandstone slab from Zangana (middle); and chert<br />
flakes from Hoghoi (bottom). ....................................406<br />
Figure 112: Artefacts photographed from Oka collection and reported to be from the<br />
Paniavile site: shell and stone adzes (top); stone adzes (middle) and waisted<br />
tools/weapons and a pineapple club (bottom). ........................409<br />
Figure 113: Waisted axes photographed from the Lanni collection, courtesy <strong>of</strong> the late<br />
Mr. Phillip Lanni, found in the vicinity <strong>of</strong> Gharanga Stream. ............410<br />
Figure 114: Un-ground adze preform photographed from Lanni collection courtesy <strong>of</strong> the<br />
late Mr. Phillip Lanni, reportedly found at the Gharanga site. ...........411<br />
Figure 115: Water-rounded lithic manuports, cortex complete. ................416<br />
Figure 116: Size distribution <strong>of</strong> fractured lithic manuports, either with some cortex or<br />
without. .....................................................416<br />
Figure 117: Intertidal collection units at Zangana: numbers are values in “unit” column in<br />
table “Flat.db” appended on CD. Units without numbers are those which yielded<br />
no sherds. ....................................................427<br />
Figure 118: Spatial distribution <strong>of</strong> the sherd sample at Zangana................428<br />
Figure 119: Across-shore size sorting at Zangana...........................429<br />
Figure 120: Possible linear settlement patterning in the distribution <strong>of</strong> large sherds at
Zangana south. ............................................... 431<br />
Figure 121: Initial point-provenanced collection <strong>of</strong> decorated sherds at Zangana<br />
(subsequent collection transects shown for spatial reference). .......... 432<br />
Figure 122: Lip deformation into a wave present in both Zangana-North and Zangana-<br />
South. ...................................................... 433<br />
Figure 123: Bands <strong>of</strong> punctation were restricted to Zangana-North. ........... 434<br />
Figure 124: Unbounded linear incision or necks banded with pinching at Zangana.<br />
........................................................... 435<br />
Figure 125: Distribution <strong>of</strong> temper classes at Zangana. ..................... 436<br />
Figure 126: Detail <strong>of</strong> distribution <strong>of</strong> temper classes at Zangana-North.......... 437<br />
Figure 127: Distribution <strong>of</strong> the sherd sample at Hoghoi. .................... 438<br />
Figure 128: Lack <strong>of</strong> sherd size variation at Hoghoi, except at 35-40m (n=4). .... 439<br />
Figure 129: Larger average manuport mass from 25m to 60m at Hoghoi. ....... 440<br />
Figure 130: Large unfractured stones and small fractured stones concentrated between<br />
25m and 75m at Hoghoi, with small rounded stones more widely distributed.<br />
........................................................... 440<br />
Figure 131: Size-sorting <strong>of</strong> manuport petrographic classes at Hoghoi, suggestive <strong>of</strong><br />
different size-procurement patterns by source. ...................... 441<br />
Figure 132: Distribution <strong>of</strong> potsherd temper classes at Hoghoi. .............. 442<br />
Figure 133: Distribution <strong>of</strong> the sherd sample at Honiavasa................... 444<br />
Figure 134: Effects <strong>of</strong> wave refraction (and collection intensity?) on sherd size at<br />
Honiavasa: the western margin is exposed to waves from Honiavasa channel,<br />
which expend their energy in a swash zone at about the centre <strong>of</strong> the site at low<br />
tide. ....................................................... 445<br />
Figure 135: Distribution <strong>of</strong> pottery tempers at Honiavasa.................... 445<br />
Figure 136: Co-joining sherds from deeper western margin <strong>of</strong> Honiavasa.. ..... 446<br />
Figure 137: Distribution <strong>of</strong> pottery decorative attributes at Honiavasa.......... 447<br />
Figure 138: A method <strong>of</strong> identifying sub-fossil organic inclusions? (Image supplied by<br />
Rafter Radiocarbon Lab) ....................................... 455<br />
Figure 139: Calibration <strong>of</strong> radiocarbon determination from a charcoal inclusion in a sherd<br />
from Paniavile................................................ 455<br />
Figure 140: Calibration <strong>of</strong> a radiocarbon determination from smoke-derived carbon on a
sherd from Hoghoi. ............................................456<br />
Figure 141: AMS sample taken from blackened sherd at far left, found on the surface <strong>of</strong><br />
unit 12 at Hoghoi. <strong>The</strong> sherd to the right, found in a subsurface test <strong>of</strong> unit 15 at<br />
Hoghoi, may be from the same vessel, and also has a sooted surface. .....457<br />
Figure 142: Vessel with surface sooting from Hoghoi dated by AMS radiocarbon.<br />
............................................................458<br />
Figure 143: Attributes used in seriations, groupings explained in <strong>chapter</strong> conclusions.<br />
............................................................467<br />
Figure 144: Correspondence plot (attributes) using all attributes and forms. .....469<br />
Figure 145: Correspondence plot (sites) using all attributes and forms. .........471<br />
Figure 146: Correspondence plot (attributes) excluding form-correlated attributes<br />
............................................................474<br />
Figure 147: Correspondence plot (sites) excluding form-correlated attributes. ....475<br />
Figure 148: Correspondence plot (sites), Honiavasa sample and form-correllated attributes<br />
omitted from data-set. ..........................................478
CHAPTER 1:<br />
RESEARCH QUESTIONS AND METHODOLOGY<br />
Introduction:<br />
I went to Roviana Lagoon as a student in a research team looking for Lapita pottery,<br />
among other things. We were faced with a conundrum, in that anything that looked<br />
vaguely like Lapita was in the sea. How to proceed? It became clear that this pattern,<br />
beyond being a nuisance, posed key research questions fundamental to archaeological<br />
interpretation in the region for the Lapita period.<br />
Lapita:<br />
Lapita pottery is a component <strong>of</strong> an archaeological horizon-style found from the Bismarck<br />
Archipelago to Samoa, dating to approximately 1300BC-800BC. Lapita pottery marks the<br />
first human colonization <strong>of</strong> Remote Oceania, that part <strong>of</strong> the Pacific that cannot be<br />
reached other than by making lengthy ocean crossings (Green 1991b) indicating that this<br />
pottery style was correlated in Remote Oceania with a maritime colonizing cultural<br />
adaptation and a period <strong>of</strong> rapid expansion (Green 1991a). Near-Oceania has by contrast<br />
been occupied for about 30 000 years, at least as far to the southeast as Buka, to the north<br />
<strong>of</strong> Bougainville (Allen et al. 1989, Wickler 1995, 2001, Wickler & Spriggs 1988), thus<br />
requiring a more complex model <strong>of</strong> local formation <strong>of</strong> the maritime colonizing adaptation,<br />
incorporating intrusive Island-Southeast-Asian Neolithic elements, local innovations, and<br />
integration with pre-existing indigenous populations (Green 1991a).<br />
1
Lapita and Solomon Islands: <strong>The</strong> Lapita Gap:<br />
Establishing the distribution <strong>of</strong> the Lapita Pottery horizon has been a major goal <strong>of</strong> Pacific<br />
archaeology since the 1950s. <strong>The</strong> Near Oceanic Solomon Islands (the Solomon Islands<br />
including Bougainville and excluding Ontong Java, Te Motu Province and Rennell-<br />
Bellona) (Figure 1) comprise a major and puzzling gap in the recorded distribution <strong>of</strong><br />
early-Lapita pottery sites (Green 1978, Kirch 1997:53, Kirch & Hunt 1988, Roe 1992,<br />
1993, Spriggs 1997: 128). Ultimately at issue is whether the distribution <strong>of</strong> early-Lapita<br />
pottery was continuous or discontinuous in Near Oceania. This issue has significant<br />
implications for our ideas about what Lapita represents. A continuous distribution <strong>of</strong><br />
early-Lapita pottery across this region would favour a model <strong>of</strong> the largely indigenous<br />
development <strong>of</strong> Lapita pottery, or at least raise the probability <strong>of</strong> integration into local<br />
populations <strong>of</strong> any migrants from Island Southeast Asia at an early stage in the<br />
development <strong>of</strong> the Lapita cultural complex. A discontinuous distribution would favour<br />
an avoidance model <strong>of</strong> Lapita colonization (e.g. Sergeantson & Gao 1995: 169), where<br />
Lapita represents “foreign” intrusion, by an expansionist society living at the fringes <strong>of</strong><br />
an already occupied hostile and/or malarious Near-Oceanic Solomon Islands. In this sort<br />
<strong>of</strong> model Lapita expansion “colonies” are seen as confined mostly to <strong>of</strong>fshore islands in<br />
the Bismarck Archipelago, and remain culturally and genetically relatively distinct from<br />
earlier occupants <strong>of</strong> Near Oceania for an extended period; generally bypassing the bulk<br />
<strong>of</strong> the Near Oceanic Solomon Islands in favour <strong>of</strong> a previously unoccupied or otherwise<br />
healthier Remote Oceania.<br />
2
Figure 1: Near Oceania, Remote Oceania and Solomon Islands, showing the location <strong>of</strong> the New Georgia Group.
Opinion amongst archaeologists with an interest in Lapita is divided as to the reasons for<br />
the gap in the distribution <strong>of</strong> early Lapita sites. Two possible explanations commonly<br />
discussed previously are (a) that the gap is purely an artefact <strong>of</strong> insufficient survey in this<br />
region (Green 1978, Spriggs 1997:128) and (b) that the gap directly reflects a<br />
discontinuous distribution <strong>of</strong> Lapita in this region or even complete avoidance <strong>of</strong> this<br />
region by Lapita peoples in the past (Gorecki 1992, Roe 1992, 1993:185, Sheppard et al.<br />
1999). Some take an equivocal position in relation to these possibilities (e.g. Kirch<br />
1997:53) (but see also Kirch 1997:283). An additional possibility (c), less <strong>of</strong>ten considered<br />
is that tectonic instability has reduced coastal site visibility/preservation in the Near-<br />
Oceanic Solomon Islands (Dickinson & Green 1998, Kirch & Hunt 1988:18). To these a<br />
fourth possibility is added here (d): that the primarily terrestrial archaeological survey<br />
which has occurred in the Solomon Islands has failed to locate Lapita sites due to a pattern<br />
in the past in this region <strong>of</strong> Lapita settlements located exclusively over the intertidal zone,<br />
comprising stilt houses or small artificial islets. Implicit in this proposition is the hypothesis,<br />
related to (c) above, that archaeological deposits resulting from intertidal-zone settlement<br />
in tropical waters are easily destroyed or hidden by even slight changes in sea-level, and<br />
are subject to destruction by wave action along most coastlines. A case is made for this<br />
fourth possibility in this thesis.<br />
<strong>The</strong> Lapita Ceramic Series:<br />
While Lapita as a defined stylistic horizon has its uses, a longitudinal (temporal) view <strong>of</strong><br />
Lapita as an evolving ceramic series with regional expressions is a major research objective<br />
for many. Although considerable effort has been expended to construct and compare<br />
regional chronologies for the Lapita period (Anson 1983, 1986, 2000, Best 1984, 2002,<br />
Green 1978, 1979, Sand 2000, Specht 1969, Spriggs 1990, Summerhayes 2000a, 2002,<br />
Wickler 1995, 2001) the temporal dimension <strong>of</strong> the Lapita horizon and its aftermath<br />
remains poorly defined in many areas, and this is especially true for the Western Province<br />
4
<strong>of</strong> Solomon Islands. A major focus <strong>of</strong> any work concerned with the Lapita period must<br />
therefore be understanding ceramic variability, including the definition <strong>of</strong> temporal<br />
variability. <strong>The</strong> reliability and resolution <strong>of</strong> such constructs <strong>of</strong> ceramic variability are also<br />
key issues. <strong>The</strong>se subjects are treated in more detail in Chapter 2.<br />
Archaeology <strong>of</strong> the Intertidal Zone and Subtidal Reef Flat:<br />
<strong>The</strong> first questions that tend to arise in conversations about an archaeological record<br />
written mainly from the intertidal-zone concern cultural formation processes: whether the<br />
materials in the sea got there as a result <strong>of</strong> discard from settlement over water, or whether<br />
they were re-deposited in the sea, either as a result <strong>of</strong> human agency or as a result <strong>of</strong><br />
erosion or subsidence. A further question contingent on the answer to these is whether<br />
cultural or activity patterning is retained in the spatial structure <strong>of</strong> intertidal scatters, or<br />
whether purely geomorphological processes such as sediment size-sorting are evidenced.<br />
An understanding <strong>of</strong> both cultural and natural formation processes <strong>of</strong> intertidal<br />
sites is critical to interpretation <strong>of</strong> the Lapita gap. Models formulated in this regard,<br />
especially those concerning the state <strong>of</strong> preservation <strong>of</strong> sites in relation to degree <strong>of</strong> wave<br />
exposure, have fundamental implications for all archaeologists working in Near Oceania<br />
with an interest in the Lapita period.<br />
A related question on which higher level cultural or behavioural inferences depend<br />
is “what sort <strong>of</strong> samples are obtainable from these sites?” If we wish to obtain a saturated<br />
sample <strong>of</strong> production diversity, for example (as might be used in a seriation analysis), we<br />
might want to know whether these site-samples are representative <strong>of</strong> production in the<br />
past, or whether they are subject to some systematic or predictable biases. What is the<br />
sample size in terms <strong>of</strong> vessels, as represented by sherds? Is there evidence <strong>of</strong> historical<br />
and heritable continuity linking all site samples, or are there discontinuities in the overall<br />
sample in either <strong>of</strong> these respects? How can sample size be quantified from a collection<br />
<strong>of</strong> sherds?<br />
5
It is in the nature <strong>of</strong> surface archaeology as opposed to stratigraphic excavation to<br />
go out rather than down, in search <strong>of</strong> horizontally structured variability rather than<br />
superposition-structured variability (although intertidal-zone materials are, strictly<br />
speaking, benthic-collected rather than surface-collected, the terrestrial terms surface-<br />
collection and surface archaeology will be widely used in this thesis, to avoid unnecessary<br />
jargon). <strong>The</strong> benefits <strong>of</strong> going out rather than down are manifold when the aim is to make<br />
the region the primary unit <strong>of</strong> analysis (as advocated by Binford 1964): samples from large<br />
areas are easily obtainable, in turn allowing more sites to be collected within the resources<br />
available, which in turn permits a closer approach to the ideal <strong>of</strong> a saturated sample,<br />
particularly if people in the past tended to move periodically rather than stay in one place.<br />
When doing surface archaeology in the sea, dates by stratigraphic association are<br />
not to be had, requiring direct dating <strong>of</strong> materials adhering to artefacts or incorporated<br />
within artefacts. While this has its difficulties, in other ways it is no bad thing, since dating<br />
by stratigraphic association can be problematic (Feathers 1997), particularly for small<br />
carbonized fragments potentially affected by turbation. In stratigraphic excavation there<br />
is no remedy in collecting large numbers <strong>of</strong> samples from chronostratigraphic units/strata:<br />
the problem is only worsened, with no way to know whether a date should be accepted as<br />
associated with the materials in the unit or rejected unless its age is outrageously<br />
incongruous. A number <strong>of</strong> direct dates <strong>of</strong> various sorts were obtained in the present study,<br />
which is encouraging, suggesting the prospects <strong>of</strong> surface archaeology are much better<br />
than would have been imagined a few decades ago.<br />
<strong>The</strong> Research Region:<br />
<strong>The</strong> New Georgia Group is situated in the northwest <strong>of</strong> the Near Oceanic Solomon<br />
Islands, in the Western Province. At a latitude <strong>of</strong> eight degrees south, climate at low<br />
6
altitude is torrid-equatorial in the summer months, relieved on the coast by sea-breezes<br />
during the day, while the southeast trade-wind blows during the winter months, bringing<br />
comparatively mild showery weather. It is rare for cyclones to track directly through the<br />
Western Province (Bath & Deguara 2003), but strong winds are <strong>of</strong>ten experienced during<br />
the late summer when cyclones typically pass to the south in the region <strong>of</strong> Rennel-Bellona.<br />
A fuller discussion <strong>of</strong> winds, tides and climate is given in Chapter 6.<br />
Most <strong>of</strong> the New Georgia Group is covered in lowland primary rainforest, which<br />
has been the focus <strong>of</strong> extensive log extraction in recent decades. Inland areas were<br />
apparently intensively settled in past centuries and irrigation terraces can still be seen. <strong>The</strong><br />
modern population <strong>of</strong> around 20 000 is concentrated in coastal villages <strong>of</strong> a few hundred<br />
people, and in the main towns <strong>of</strong> Gizo, Munda and Noro. Coastal areas with good<br />
gardening soils (alluvium or uplifted backreef sediments) have a mosaic <strong>of</strong> gardens and<br />
regenerating bush. Various forms <strong>of</strong> indigenous plantation and subsistence arboriculture<br />
occupy significant tracts <strong>of</strong> land near the coast. A commercial plantation timber company<br />
operates on Kolombangara. Most local transport is by outboard-powered canoe, except<br />
within local centres or in logging areas.<br />
Geology, Geomorphology and Related Human Geography:<br />
Geological exploration has been comprehensive, beginning in the 1880s, with general<br />
descriptions and with the details filled in by the Solomon Islands Geological Survey from<br />
the 1950s till independence. Interests first in bauxite prospecting and more recently in gold<br />
prospecting have seen many prospecting-licence surveys in the 1970s and 1980s.<br />
Additional field survey and a synthesis <strong>of</strong> existing information were undertaken by the<br />
British Geological Survey between 1976 and 1983 (Dunkley 1986).<br />
<strong>The</strong> New Georgia Group was formed during an episode <strong>of</strong> volcanic activity which<br />
began in the upper-Miocene and continues today (Dunkley 1986:7), and there is rapid and<br />
substantial uplift <strong>of</strong> a forearc or outer arc region comprising Ranongga, Tetepare and the<br />
7
southern part <strong>of</strong> Rendova (Figure 2), beneath which very young, warm and ductile<br />
buoyant lithosphere <strong>of</strong> the spreading Woodlark basin is being subducted. A notable feature<br />
<strong>of</strong> this process relevant to the archaeology is the marked reduction in seismicity relative<br />
to Bougainville or Guadalcanal (Dunkley 1986:9). although a serious earthquake caused<br />
shoreline changes on Vella Lavella in 1959 (Stoddart 1969b:378).<br />
<strong>The</strong> group comprises a large number <strong>of</strong> islands trending NW-SE over a distance<br />
<strong>of</strong> 230km with a total landmass <strong>of</strong> 5,060 km 2 (Dunkley 1986:abstract), in an area <strong>of</strong> land<br />
and sea totaling 15,000 km 2 . <strong>The</strong> group is within 24-hour paddling distance (50km) <strong>of</strong> the<br />
main Choiseul/Ysabel chain, 50km distant across the New Georgia Sound to the northeast.<br />
<strong>The</strong> group is exposed to the Solomon Sea and Coral Sea to the southwest, across which<br />
lie the Louisiade Archipelago and the Queensland coast, at distances <strong>of</strong> 540km and<br />
1600km respectively, on what would have been an ideal return-voyaging course in relation<br />
to the Southeast trade-wind (Irwin 1992).<br />
<strong>The</strong> main volcanic chain (Vella Lavella, Kolombangara, New Georgia, Vangunu,<br />
Nggatokae and Northern Rendova) comprises a series <strong>of</strong> remnant composite volcanic<br />
cones <strong>of</strong> basaltic composition, with various small intrusive complexes, fringed by an<br />
intricate system <strong>of</strong> uplifted Pleistocene coral reefs and lagoons (Dunkley 1986: 12). <strong>The</strong><br />
largest <strong>of</strong> the volcanic cones is Kolombangara, rising to 1760m, while others are in an<br />
advanced state <strong>of</strong> erosion, exposing central sub-volcanic intrusion complexes. Upon<br />
examining charts in 1928, W.M Davis noted that “....none <strong>of</strong> the elevated reefs <strong>of</strong> the <strong>of</strong><br />
the Solomon Islands is more remarkable than the emerged barrier reef which skirts...New<br />
Georgia” (Davis 1928:397-398), and this complex was the subject <strong>of</strong> a more detailed field<br />
study by the Royal Society Expedition to the British Solomon Islands in the mid 1960s<br />
(Stoddart 1969a). A more recent project explicates a detailed sea-level history for this<br />
formation (Mann et al. 1998).<br />
8
9<br />
Figure 2: Map <strong>of</strong> New Georgia Group showing principal geological formations (after Coulson, Dunkerly, Hughes and Ridgeway 1987).
<strong>The</strong> Roviana Formation:<br />
<strong>The</strong> archaeological materials that form the data <strong>of</strong> this thesis are all found on intertidal or<br />
shallow-water reef flats. “<strong>The</strong> differential uplift <strong>of</strong> the New Georgia barrier reefs has led<br />
to the development <strong>of</strong> a wide series <strong>of</strong> reef shores, from vertical and overhanging cliffs to<br />
horizontal intertidal flats. (Stoddart 1969b:371).” <strong>The</strong> height <strong>of</strong> the raised-reef chains<br />
around New Georgia vary along their length and with proximity to the mainland, with local<br />
variations in uplift due to tilting and displacement along faults (Dunkley 1986:36). Of the<br />
upraised reefs forming lagoons around New Georgia, only the barrier reefs <strong>of</strong> the Roviana<br />
Lagoon are suitable for gardening, with extensive areas <strong>of</strong> Pleistocene upraised backreef<br />
or lagoonal sediments (see geological map and Dunkley 1986:37). <strong>The</strong>se may have<br />
originated as pools formed by double-barrier reef systems, which are an unusual<br />
characteristic <strong>of</strong> the New Georgia barrier system, and which testify to a complex interplay<br />
between tectonic subsidence followed by uplift and fluctuating relative sea-levels (Mann<br />
et al. 1998).<br />
<strong>The</strong>re are extensive areas <strong>of</strong> uplifted backreef sediments in some parts <strong>of</strong> the New<br />
Georgia mainland also, which are elevated to a height <strong>of</strong> 190m above sea level in the<br />
Munda and Kazukuru areas, unlike the barrier-reef chain. It is more usual elsewhere<br />
around New Georgia for the tops <strong>of</strong> the upraised barrier reef to be bare <strong>of</strong> soil, with<br />
solution erosion providing a jagged micro topography, on which only trees will grow. This<br />
seems to be the case for all <strong>of</strong> the upraised barriers around Marovo, Ngerrasi and<br />
VonaVona lagoons. <strong>The</strong> raised reefs <strong>of</strong> the Roviana formation are typically indurated and<br />
recrystallized, with karstic features resulting from partial dissolution. Ages are uppermost<br />
Pliocene in some cases on fossil evidence (Dunkley 1986:38)<br />
Deltaic fans <strong>of</strong> volcanic alluvium are found at the mouths <strong>of</strong> the larger rivers that<br />
drain the New Georgia landmass, the higher well-drained central areas <strong>of</strong> which provide<br />
favoured garden areas to the modern population. <strong>The</strong> actively prograding seaward margins<br />
10
<strong>of</strong> these deltas are s<strong>of</strong>t mud, and are not used at present for agriculture, other than Sago<br />
plantation for ro<strong>of</strong>ing and walling material. Gardens are also located at the mouths <strong>of</strong><br />
smaller streams along the mainland shore, where the shore is more accessible by canoe, and<br />
where swamp-taro may be grown in pits dug inland <strong>of</strong> the shoreline.<br />
In Roviana lagoon, extensive areas <strong>of</strong> coastal saline flats with fresher uplifted<br />
sandy/coralline lagoonal sediments containing fragmented Acropora corals are cyclically<br />
hoed, and planted in sago, coconuts and bananas. <strong>The</strong>se flats are roughly 30cm above the<br />
high tide mark, and may be inundated by saltwater during storm surges or other extreme<br />
tides. <strong>The</strong>se are most likely evidence for a recent high-stand <strong>of</strong> sea-levels (late Holocene<br />
sea level changes are quite well documented by Mann et al. 1998). Pandanus grows thickly<br />
in untended areas. Calophyllum species prosper on this flat and are an important modern<br />
economic resource, while a number <strong>of</strong> tree species used in manufacture <strong>of</strong> canoe parts and<br />
fishing spears are harvested from this flat also. Modern housing is located mostly on the<br />
uplifted backreef sediments, among gardens, but some kitchens and canoe houses are<br />
located on the lower flat adjacent to canoe wharves. Crocodiles may traverse this lower<br />
flat at night in search <strong>of</strong> penned livestock or unwary dogs or small humans. As this flat was<br />
probably inundated in the pre-1000 bp period, lagoon shorelines would have been quite<br />
steep in many places compared to their present-day conformation, making canoe<br />
landing/storage difficult in many areas, which may have been an important influence on<br />
settlement type in the Lapita period.<br />
<strong>The</strong>re are many sand keys to the west <strong>of</strong> Nusa Roviana, where tectonic uplift is less<br />
than elsewhere in the Roviana lagoon, probably as a result <strong>of</strong> faulting/tilting <strong>of</strong> the Munda-<br />
Noro reefal formation, and also to the southeast <strong>of</strong> Vangunu, at the southern margin <strong>of</strong><br />
Marovo Lagoon. <strong>The</strong>se sand keys, while picturesque in a tropical travel-brochure sort <strong>of</strong><br />
way, are sometimes swept by waves during storms, and are not favoured for settlement by<br />
local people.<br />
11
Archaeological survey for the present thesis was confined to Roviana lagoon,<br />
between Munda and Kalena bay, with intensive intertidal survey restricted to the<br />
Kaliquongu region. While there is some variation in the height above sea level <strong>of</strong> the<br />
uplifted Pleistocene reef through the survey region from Nusa Roviana to Kalena Bay, the<br />
uplifted backreef sediments and lower coastal flats are common throughout the survey<br />
region, although there are localized cliffed shores also in some places. This constancy<br />
suggests a common recent relative sea-level history to the entire survey region, borne out<br />
by the research <strong>of</strong> Mann et al. (1998). Elsewhere around New Georgia, outside the survey<br />
frame <strong>of</strong> this research, recent sea-level history is more varied.<br />
Regarding long-term (post-Pliocene) differences in uplift, to the east <strong>of</strong> the survey<br />
region, between Kalena bay and Viru Harbour, the situation differs, with uplift obviously<br />
greater, and backreef/lagoon sediments extending from the reef to the mainland, with no<br />
modern lagoon remaining. To the West <strong>of</strong> the survey region, uplift is less, with the<br />
Pleistocene reef barely emerged, forming the base <strong>of</strong> the sand keys that shelter the Munda<br />
coastline. West <strong>of</strong> the Honiavasa passage the lagoon-side geomorphological record <strong>of</strong> the<br />
uplifted barrier chain is complicated by extensive road works during the Second World<br />
War (WWII), which in places substantially modified the shore pr<strong>of</strong>ile to create roads and<br />
industrial installations (Bethlehem Island and Zangana Point are examples <strong>of</strong> major<br />
earthmoving during WWII).<br />
<strong>The</strong> outer-arc region has more sedimentary rocks and low-grade metamorphic<br />
rocks than the main volcanic chain, where these are virtually absent other than as a small<br />
formation on Kolombangara. <strong>The</strong>se sedimentary rocks and tuffs are <strong>of</strong> interest due to the<br />
presence in intertidal archaeological scatters <strong>of</strong> numerous abrader fragments and adze<br />
fragments that are either sedimentary or possibly metamorphosed sedimentary materials<br />
(see Chapter 10). <strong>The</strong> outer-arc region is also <strong>of</strong> interest due to higher rates <strong>of</strong> tectonic<br />
uplift, that have outstripped the Holocene marine transgression by hundreds <strong>of</strong> metres.<br />
12
<strong>The</strong> sea-level history <strong>of</strong> the outer-arc coastlines thus contrasts markedly with that <strong>of</strong> the<br />
New Georgia uplifted fringing reef and comparison <strong>of</strong> the archaeological record there with<br />
Roviana Lagoon is desirable, although this has not yet been undertaken. <strong>The</strong> lower two-<br />
thirds <strong>of</strong> the Kiorosi section on Tetepare is <strong>of</strong> Early Pleistocene age, beginning about 1.65<br />
Ma. at the base <strong>of</strong> the exposure, and the upper third is <strong>of</strong> late-Pleistocene age, postdating<br />
0.27Ma. (Dunkley 1986:28), and has undergone exceptionally high rates <strong>of</strong> uplift in the<br />
Late Pleistocene. It is not clear to what extent remnant uplifted shorelines are exposed on<br />
the outer arc, as massive slumping has occurred in places along the exposed southern coast<br />
<strong>of</strong> Rendova/Tetepare, although Plio-Pleistocene raised fringing reefs are well preserved as<br />
a series <strong>of</strong> terraces to the southwest <strong>of</strong> the Rendova volcanic cone.<br />
Summary <strong>of</strong> Geology:<br />
In summary the geology <strong>of</strong> the New Georgia Group presents a complex tapestry <strong>of</strong> rock<br />
types, structural geology/plate tectonics and sea-level change, dominated from the<br />
archaeologist’s point <strong>of</strong> view by raised reef formations <strong>of</strong> considerable antiquity. <strong>The</strong>re are<br />
some remaining uncertainties about the timing and magnitude <strong>of</strong> Holocene changes in<br />
relative sea-level, but there is no evidence for significant change in the overall<br />
conformation <strong>of</strong> the Roviana Lagoon during the late Holocene, in terms <strong>of</strong> exposure to<br />
ocean storm waves at least. <strong>The</strong> Roviana Lagoon existed in approximately its present form<br />
since the Holocene marine transgression, but inundation <strong>of</strong> the coastal flat during a recent<br />
high-stand would have made many <strong>of</strong> the shorelines more difficult to occupy and land<br />
canoes on. <strong>The</strong>re seems to be no difference across the survey area in this respect, other<br />
than the down-faulted or down-warped sand keys west <strong>of</strong> Nusa Roviana, which would<br />
have been near-submerged during a 1m high-stand, making them even more exposed to<br />
storm surge than at present and probably undesirable for any form <strong>of</strong> settlement. Some<br />
cliffed lagoon-side coasts in the Aroroso Passage area that do not have<br />
13
Figure 3: <strong>The</strong> author providing scale for solution notches in Plio-Pleistocene limestone<br />
cliff near Saikile passage, Roviana Lagoon (photograph courtesy <strong>of</strong> Peter Sheppard)<br />
the slightly-raised shoreline flat are examples <strong>of</strong> local topography unsuitable for the growth<br />
<strong>of</strong> coral, along which the more usual series <strong>of</strong> shore terraces has not formed (Figure 3).<br />
<strong>The</strong> Roviana Early Ceramic Archaeological Record:<br />
Reeve reported four archaeological sites in the Western province, most notably the<br />
Paniavile intertidal site in Roviana Lagoon (Figure 4),<br />
“Containing a distinctive and possibly Lapita-related style <strong>of</strong> decorated<br />
pottery.... While the absence <strong>of</strong> dentate-stamping makes it impossible to<br />
consider the Paniavile material as belonging within the classic Lapita<br />
tradition, its decorative techniques and its wide range <strong>of</strong> vessel<br />
shapes...suggest that it represents a related tradition, possibly derived or<br />
descended from Lapita” (Reeve 1989)<br />
14
Reeve suggested that these ceramics closely post-date Lapita, and further suggested<br />
similarities to a number <strong>of</strong> ceramic assemblages from elsewhere in Melanesia. Reeve<br />
suggested that the Paniavile site in its design system is more similar to late Lapita material<br />
at Watom (Anson 1983, Green & Anson 1991). This suggestion is examined in the course<br />
<strong>of</strong> this thesis, and the Paniavile pottery is found to be quite dissimilar to Watom material<br />
(see Chapter 13), although his statement, “It may be that the Paniavile ceramics in some<br />
way represent a link between the late Lapita material... and the Mangaasi/Sohano ceramic<br />
traditions.” (Reeve 1989) is closer to the conclusions drawn in Chapter 13.<br />
Research Questions:<br />
Diverse specific research questions arise in relation to these materials in the sea. <strong>The</strong><br />
overall question being, “How to proceed?”, leads to a series <strong>of</strong> “How to proceed”<br />
questions and answers, rather than a Roviana prehistory.<br />
How can the formation <strong>of</strong> the Roviana intertidal-zone and reef-flat record be<br />
researched? What scale, method and intensity <strong>of</strong> survey is required to obtain a useful<br />
sample set? What cultural and post-depositional formation processes can be inferred to<br />
have produced the record? Do the Roviana materials show heritable continuity from<br />
Lapita? Were settlements over water or on land? Can we see settlements in the distribution<br />
<strong>of</strong> materials, or are artefact/manuport density distributions a product <strong>of</strong> post-depositional<br />
factors? What systematic biases are there in the record? How should the record be written<br />
(what collection/recording methods are appropriate; how much detail is needed)? How<br />
fragile are these sites ( for example, in what state <strong>of</strong> preservation are they on discovery,<br />
and what is the impact <strong>of</strong> surface collection)?<br />
How can we infer the distribution <strong>of</strong> Lapita-phase pottery use in the past in Near Oceania?<br />
What state <strong>of</strong> preservation are the Roviana materials in, and what are the taphonomic<br />
processes affecting them? Can we model preservation <strong>of</strong> similar sites elsewhere using<br />
these data?<br />
15
Figure 4: Paniavile at low tide, looking north, with several inhabited islets and the New<br />
Georgia mainland in the distance.<br />
How might a ceramic chronology be constructed? What classificatory units fit the<br />
sample and the period? How should the properties <strong>of</strong> samples be quantified? How can the<br />
various dimensions <strong>of</strong> time, style, space and function be defined and identified or controlled<br />
for? Are the available samples adequate for the task? If not, how might a better sample be<br />
obtained? What are the prospects for obtaining direct dates, and what methodological<br />
development research is needed in this respect?<br />
What can we infer about social interaction from transport <strong>of</strong> lithic raw materials and<br />
pottery temper, and how might these results be extended in future? What are the priorities<br />
for sampling <strong>of</strong> geological source rocks and sediments in future fieldwork? What are the<br />
implications <strong>of</strong> the recovered materials suite for future field collection strategies?<br />
Archaeological theory and method through which these questions could be addressed is<br />
reviewed below. <strong>The</strong> review is structured into four parts: (a) approaches to classification<br />
16
in ceramic studies generally, with some reference to Oceania; (b) a review <strong>of</strong> quantification<br />
methods and methods <strong>of</strong> sample evaluation, (c) a review <strong>of</strong> seriation method and theory;<br />
and (d) review <strong>of</strong> the links between form and function. Other topics, survey and sherd<br />
taphonomy, are reviewed in detail in their own <strong>chapter</strong>s.<br />
Fundamental to the construction <strong>of</strong> temporal variability are: survey method and<br />
extent, ceramic description, classification, quantification <strong>of</strong> class membership, exploratory<br />
analysis <strong>of</strong> variability, the disentanglement <strong>of</strong> style and function, and seriation theory and<br />
method. This is especially the case where stratigraphy and superposition do not supply<br />
clear temporal information. In Oceania we tend to dutifully search out “undisturbed”<br />
deposits and avoid “disturbed deposits” within what could be called in its more extreme<br />
manifestations a “bedded strata” theory <strong>of</strong> archaeological formation, searching for stratified<br />
sequences <strong>of</strong> (especially ceramic) change. <strong>The</strong>se sorts <strong>of</strong> temporal constructions are tested<br />
against other formation theories in Chapter 2, and in the main found to be insecure when<br />
viewed from the perspectives <strong>of</strong> contemporaneous spatial diversity, settlement pattern<br />
instability, turbation, erosion, re-deposition, postdepositional stratification and sherd<br />
taphonomy, these factors acting in various combinations.<br />
Conceptions <strong>of</strong> change and classificatory systematics <strong>of</strong> Lapita studies tend to<br />
favour, or result in, coarse periodization, the temporal resolution <strong>of</strong> which is ill-matched<br />
to some <strong>of</strong> the suggested social explanations <strong>of</strong> Lapita, leading to a proliferation <strong>of</strong><br />
speculative and untested social theory. Thus some <strong>of</strong> the temporal constructions reviewed<br />
in Chapter 2 are phrased in terms <strong>of</strong> Lapita and post-Lapita; early middle and late Lapita;<br />
Early Eastern Lapita and Late Eastern Lapita, Early motifs and Late motifs, utilitarian<br />
vessel forms and non-utilitarian. Similarly, coarse approaches to time <strong>of</strong>ten involve<br />
conception <strong>of</strong> complex variables like site occupation span in terms <strong>of</strong> atemporal states,<br />
where site assemblages are “slices in time”, readable as Lapita in its early state, or Lapita<br />
in its late state.<br />
17
Two things we do know are that a style <strong>of</strong> pottery we call Lapita was widespread<br />
in Oceania about three thousand years ago and that it is not around in the same style today.<br />
Lapita has been defined in various ways, as a culture (Bellwood 1978:244), as a cultural<br />
complex and as a ceramic series (Golson 1971, Green 1992). Regardless <strong>of</strong> which<br />
definition one chooses, the weight <strong>of</strong> evidence available to us that there was a marked<br />
similarity in material culture across occupied Oceania around the first millennium BC is<br />
overwhelming to the point where this can be considered to be a fact. We also know that<br />
many non-Lapita styles <strong>of</strong> pottery have been in use in the intervening millennia, (not<br />
worrying too much about where the cut-<strong>of</strong>f point between Lapita and non-Lapita is as this<br />
is a matter <strong>of</strong> definition: non-Lapita because these do not fit the various definitions, not<br />
because they lack heritable continuity with Lapita), and can expect, except under an<br />
extreme bedded-strata formation theory, that the archaeological record will present us with<br />
a vast array <strong>of</strong> varying mixtures <strong>of</strong> various temporal and spatial components <strong>of</strong> these styles<br />
in various stratification and superposition arrangements. To suggest on the basis <strong>of</strong> these<br />
facts that we know the direction <strong>of</strong> ceramic change is to treat the record as though the<br />
direction <strong>of</strong> change is constant. Such a doctrine <strong>of</strong> gradualism seems implicit in the<br />
conclusions <strong>of</strong> some <strong>of</strong> the studies reviewed in Chapter 2, and encourages complacence<br />
regarding the resolution with which ceramic change has thus far been defined.<br />
It seems unlikely for example, if aims are pitched at these coarse levels <strong>of</strong><br />
resolution, that the challenges posed by post-processual critiques or by evolutionary<br />
archaeology to produce historical accounts and/or to model individual and political agency<br />
can be met. We condemn ourselves to explanation only at the broadest timescale, in which<br />
the material effects <strong>of</strong> individuals and smaller groups over short timescales are invisible.<br />
For Lapita, the corpus <strong>of</strong> short-term events is more limited that in many other regions and<br />
time periods around the world: Lapita burials, for example are rare or unknown; but the<br />
stuff <strong>of</strong> this thesis, the manufacture and discard <strong>of</strong> pots, comprises short term events<br />
18
structured by the actions <strong>of</strong> individuals, and rich in stylistic data which can be used to tell<br />
time, and thus there must be information to hand in the patterning <strong>of</strong> these events across<br />
the landscape that pertains to evolutionary historicist as well as humanist orientations.<br />
This information cannot be accessed though, without a fine-grained ceramic<br />
chronology. Social explanations require a measure <strong>of</strong> social time or risk reading temporal<br />
or other sorts <strong>of</strong> variability as social, or vice versa. Ceramic descriptive and classificatory<br />
units must be formulated to capture variability that is salient to the desired level <strong>of</strong><br />
analytical resolution, and yet the fineness <strong>of</strong> one’s classification is constrained in reality by<br />
the properties <strong>of</strong> that sample in terms <strong>of</strong> historical continuity and sample size. Samples<br />
without large time-gaps or space-gaps are needed if research questions are to be pitched<br />
at a humanist or social-evolutionary level. Also, the more complete the sample, the more<br />
finely it can be split into classificatory units, which also affects the resolution <strong>of</strong> seriation<br />
chronologies. Sample evaluation is thus a crucial element in establishing the type <strong>of</strong><br />
questions that can be asked <strong>of</strong> sample variability.<br />
Seriation continues to be one <strong>of</strong> the backbones <strong>of</strong> archaeological dating in the<br />
radiocarbon era, to a greater extent than was initially expected (O'Brien & Lyman<br />
2000:226), although this is not always apparent in Pacific archaeology. Over the decades,<br />
in the work <strong>of</strong> Green, Parker, Sheppard and Green, Irwin, Specht, Wickler, Egl<strong>of</strong>f,<br />
Wahome and Best reviewed in Chapter 2, surface scatters <strong>of</strong> ceramics are valued, and<br />
attempts are made to seriate these in some cases. Seriation method and theory have<br />
undergone many reincarnations through the Culture Historical era and the period <strong>of</strong> the<br />
“New Archaeology” and behavioural archaeology, and over the last decade, in which new<br />
analytical techniques have become widely available.<br />
An emerging concern with formation theory is apparent in recent decades in world<br />
archaeology, which has complicated the interpretation <strong>of</strong> archaeological variability,<br />
particularly for ceramics, and challenged our methods <strong>of</strong> identifying temporal, stylistic and<br />
19
functional variability in the archaeological record. Regional historical examples are given<br />
<strong>of</strong> Oceanic seriations in Chapter 2, which are compared methodologically with current<br />
techniques, methods and theory.<br />
A seductive notion for those using many <strong>of</strong> the newer techniques <strong>of</strong> multivariate<br />
exploratory analysis for seriation is that the theoretical dimensions <strong>of</strong> sample-size, space,<br />
time, style and function equate to summary dimensions <strong>of</strong> variability in the data<br />
(Summerhayes 2000a, Wahome 1999). Seriation is a temporal construct, and is largely a<br />
search for temporal variability, regardless <strong>of</strong> the seriation technique used. A theme <strong>of</strong> the<br />
review in Chapter 2 is to assess the degree to which researchers have sought out temporal,<br />
as opposed to other sorts <strong>of</strong> variability, with which to construct their seriations.<br />
Furthermore, form and function are not equivalent, nor are style and decoration. <strong>The</strong>re is<br />
a tendency also for archaeologists, when faced with multiple summary dimensions <strong>of</strong><br />
variability, to ascribe each <strong>of</strong> these uncritically to one <strong>of</strong> the above. Thus in a Principle<br />
Components Analysis there is a tendency to ascribe one summary dimension to function,<br />
one to sample size, and so on, <strong>of</strong>ten with little or no explicit justification. If purely<br />
temporal variability is input into such a multivariate analysis, we will still get multiple<br />
summary dimensions <strong>of</strong> variability. Some attributes may serve to differentiate some<br />
temporal units, other stylistic variants will serve to discriminate others, as has been<br />
recognized for many decades from simpler methods <strong>of</strong> seriation.<br />
Techniques like Correspondence Analysis, which provided diagnostic statistics for<br />
the purpose, allow a clear understanding <strong>of</strong> the origins <strong>of</strong> summary components <strong>of</strong><br />
variability, to a greater degree than some <strong>of</strong> the early computer multivariate methods that<br />
used similarity matrices. <strong>The</strong> era <strong>of</strong> similarity matrix clustering approaches is largely past,<br />
as such analyses are difficult to evaluate, but it will be seen that we have a substantial<br />
residue <strong>of</strong> Lapita studies <strong>of</strong> this sort, which require careful reading.<br />
<strong>The</strong> conclusions reached in the review in Chapter 2 regarding the significance <strong>of</strong><br />
20
the main constructions <strong>of</strong> Lapita and post-Lapita variability, and regarding formation <strong>of</strong><br />
the Lapita record in general terms, can be summarized in the suggestion that Lapita<br />
chronology in Near-Oceania is not yet well understood at any but the coarsest level <strong>of</strong><br />
temporal resolution, that <strong>of</strong> the culture historical horizon, and that a major methodological<br />
and theoretical re-evaluation is required in order to achieve better resolution than this.<br />
More <strong>of</strong> the same sort <strong>of</strong> data is not enough, the type <strong>of</strong> data and what we do with it must<br />
change. I thus take issue with Spriggs’ characterization <strong>of</strong> Near-Oceanic archaeology, and<br />
Lapita archaeology in particular, as being in a data-led pioneering phase (Spriggs 1993).<br />
It will remain in that phase only if we continue to approach the record as though we are<br />
in that phase. Gosden’s call for “concerted frameworks” for analysis (Gosden 1991a), is<br />
extended here to a call for more emphasis on sampling and formation frameworks, applied<br />
not just to individual site samples, but to analysis <strong>of</strong> regional archaeological landscapes.<br />
Approaches to Ceramic Classification and Analysis:<br />
Introduction:<br />
Typological approaches to the archaeological record and the construction <strong>of</strong> chronology<br />
have their origins in nineteenth-century Europe, with developments like the Montelian<br />
synthesis <strong>of</strong> European prehistory in the 1880s, with full-fledged development <strong>of</strong> a culture<br />
historical approach and the notion <strong>of</strong> “archaeological cultures” apparent by the publication<br />
<strong>of</strong> Childe’s “Dawn <strong>of</strong> European Civilization” in 1925 (Trigger 1989:163-169). Culture-<br />
historical approaches in the Americas involved typological method that attained its<br />
archaeological zenith in the work <strong>of</strong> Willey and Phillips (1958).<br />
Rice advocates what has more recently been called a reflexive approach to<br />
classification and typology (Rice 1982:49), citing Brew’s 1949 statement “We need more<br />
rather than fewer classifications, different classifications, always new classifications, to<br />
21
meet new needs”. Rice identifies the perennial issues as<br />
• Are types real or created, (essentialist vs. materialist approaches, see Dunnell<br />
1986)<br />
• Is it better to lump or split,<br />
• What and how many attributes should be used in classifications?<br />
• What kinds <strong>of</strong> types are there (historical, analytical, cultural)<br />
• What is the difference between a grouping, a classification, and a taxonomy?<br />
Beginning with the last question, Rice suggests grouping is phenomenological, groups<br />
consist <strong>of</strong> members, while classes are defined by criteria. Groups are historical or localized<br />
events, classes are ahistorical. Where attributes are <strong>of</strong> equal weight, they create unordered<br />
classes or groups, whereas where attributes or features are differentially weighted they<br />
create hierarchical classifications, or taxonomic classifications, like the type-variety system<br />
or the Linnaean system.<br />
In the type-variety system <strong>of</strong> hierarchical taxonomy, “ware” was a broad, high-level<br />
unit <strong>of</strong> synthesis and comparison, defined, for example, by such attributes as commonalities<br />
<strong>of</strong> paste and surface treatment (Rice 1982:50). Wares were composed <strong>of</strong> groups and<br />
groups were composed <strong>of</strong> types. Types were the basic American unit <strong>of</strong> description,<br />
frequently ascribed a cultural meaning as a ceramic idea <strong>of</strong> a society. Types were formed<br />
by combining varieties, the initial sorting units.<br />
Types vs Attributes:<br />
Shepard pointed out the principal limitations <strong>of</strong> types is that these are time-sampling and<br />
space-sampling-dependent, decrying the 1950s paradigm that types are in some measure<br />
cultural entities (Shepard 1963:308). She proposed the use <strong>of</strong> technical features, such as<br />
tempering material, as providing simple criteria for delimiting types, and this influence<br />
22
can be seen strongly in Oceania in Specht’s analysis <strong>of</strong> Buka ceramic variation (Specht<br />
1969). Where materials exhibit continuous gradation, these do not form separate types, but<br />
where one class <strong>of</strong> tempering material does not grade into another, this provides a useful<br />
criterion for classification.<br />
She regarded 1950s types as excessively coarse-grained and proposed “design<br />
styles” as a finer classification, defined in terms <strong>of</strong> elements and motifs, and falling within<br />
a technological tradition, this being a cluster <strong>of</strong> associated or interdependent techniques<br />
(Shepard 1963:320). Again the influence on Specht’s approach is clear. She noted Colton’s<br />
proposition from the 1950s that a “series” consisted <strong>of</strong> a temporal sequence <strong>of</strong> types, as<br />
opposed to a “ceramic group” which was a contemporary grouping <strong>of</strong> pottery types within<br />
a site <strong>of</strong> short duration and limited area, and an “index ware” which was a group <strong>of</strong><br />
functional types peculiar to a particular prehistoric tribe.<br />
Classification and the New Archaeology:<br />
In critique <strong>of</strong> the type-variety system Dunnell wrote<br />
“If classifications <strong>of</strong> any kind are to be devices useful in constructing<br />
explanations, if they are done for something other than amusement, they<br />
must be capable <strong>of</strong> evaluation, susceptible to change. In short, they must<br />
be hypotheses about the ordering <strong>of</strong> data for a specific problem....To<br />
expect that the same set <strong>of</strong> classes defined by criteria relevant to use will<br />
prove the most useful for chronology is foolish....Brew’s admonitions for<br />
more classes, not fewer, is today just as true as when it was<br />
written....Classifications need not be taken for granted. <strong>The</strong>y must suit their<br />
problem or they are useless (Dunnell 1971).”<br />
Groupings were not seen by Dunnell as an appropriate means <strong>of</strong> constructing classes, and<br />
were regarded rather as a means <strong>of</strong> generating and testing hypotheses about classes.<br />
<strong>The</strong> “New Archaeology” was a radiocarbon-dating-era development (Trigger<br />
1989:294-295) heralded by Binford as a dramatic break from the methodologies <strong>of</strong> the<br />
past (Trigger 1989:295). <strong>The</strong> New Archaeology was marked by a shift in objectives<br />
beyond chronology <strong>of</strong> archaeological “cultures” (the latter strongly identified with ethnic<br />
23
groups by the later end <strong>of</strong> the culture-historical period). Grouping rather than classification<br />
was the preferred method <strong>of</strong> analysis <strong>of</strong> variability (Graves 1998), and variation was<br />
increasingly interpreted using anthropological models. Chronology was no longer the major<br />
goal <strong>of</strong> ceramic analysis, and there was frequently an assumption <strong>of</strong> contemporaneity for<br />
pottery variety within a settlement. Ceramic design variability could therefore inform on<br />
small-scale prehistoric organizational properties.<br />
Graves’ historical account <strong>of</strong> ceramic method and theory in the American<br />
Southwest makes some points which are relevant to the Pacific and Lapita studies during<br />
the period <strong>of</strong> the New Archaeology. Graves describes this general situation in the<br />
American southwest as atemporality to a fault. Where previously culture history had<br />
employed relatively coarse-grained temporal-typological analysis <strong>of</strong> stylistic and to some<br />
extent functional variation across regions, for the early “New Archaeologists”,<br />
understanding prehistoric society through patterning in the site came to the fore. Graves<br />
identifies this phase <strong>of</strong> archaeology in the southwest with a lack <strong>of</strong> explicit consideration<br />
<strong>of</strong> site formation processes, although substantial progress was made in this regard through<br />
the 1980s.<br />
I argue in Chapter 2 that there is an “atemporality” <strong>of</strong> this sort in some analyses<br />
<strong>of</strong> Lapita variability <strong>of</strong> that era (e.g. Green 1978), which is ultimately founded on a C14-<br />
based slice in time view <strong>of</strong> site assemblages, and that this approach to chronology has<br />
become widespread in Oceanic archaeology in subsequent decades. Unlike Specht’s<br />
approach to definition <strong>of</strong> a cultural sequence for Buka, chronology was no longer the<br />
primary objective <strong>of</strong> ceramic analysis, as chronology could be derived from other,<br />
ostensibly more scientific and objective information (radiocarbon dating). Instead, ceramic<br />
variability was used to identify activity patterning or to measure the similarity <strong>of</strong> sites to<br />
infer interaction (Green 1978) rather than to tell time. Although Green more recently<br />
returned to ceramic variability for a chronology when the accumulating 14 C evidence<br />
24
egan to look less clear-cut (Green 1991c) the extent to which the understanding <strong>of</strong><br />
ceramic variability was initially predicated on a “slice-in-time” view <strong>of</strong> sites is difficult to<br />
appreciate without careful reading <strong>of</strong> Donovan’s thesis (Donovan 1973) and Green’s<br />
working paper (Green 1978).<br />
Seriation continued in widespread use around the world through the period <strong>of</strong> the<br />
New Archaeology, but with a difference. Typological classification approaches had been<br />
largely swept away by attribute-based grouping analyses (e.g. Le Blanc 1975). Orton<br />
identifies the work <strong>of</strong> Shepard as a nodal point in this transition, where in addition to<br />
typological construction <strong>of</strong> chronology, a preoccupation with the geological origins <strong>of</strong><br />
ceramic materials emerged as an indicator <strong>of</strong> trade/exchange, and where physical properties<br />
<strong>of</strong> vessels were used to infer technological character (Orton et al. 1993:13). Shepard’s<br />
design attributes were however, far removed from the decorative attribute combination<br />
approach common in Pacific archaeology since the 1970s (Best 1984, Frost 1974, Green<br />
1978, Irwin 1972, 1985), and more akin to Mead’s structural approach (Mead 1975) in<br />
that she felt we should “insist that the meaning <strong>of</strong> design should be sought wherever<br />
possible (Shepard 1963:259)”, so Orton’s review, while pertinent to Pacific analyses <strong>of</strong><br />
technological attributes, glosses over some <strong>of</strong> the details <strong>of</strong> the transition to attribute-based<br />
analyses <strong>of</strong> design systems.<br />
Shepard attributed the notion <strong>of</strong> elements being the basic, irreducible units <strong>of</strong><br />
design to Chapman (Shepard 1963:266-267), while motifs were more varied, complex and<br />
distinctive than elements, and more appropriate units <strong>of</strong> design analysis than elements for<br />
some pottery. She noted that element/motif analysis was convenient for “strictly geometric<br />
designs”. She pointed out that element analysis might be helpful in the circumstance<br />
where such designs were in a state <strong>of</strong> extreme fragmentation, in which only a small part <strong>of</strong><br />
the design appears on the fragment, but cautioned regarding higher-level decisions<br />
about which sort <strong>of</strong> analysis was appropriate to a given set <strong>of</strong> designs, that “It would be<br />
25
unfortunate if dependence on sherds should dictate methods <strong>of</strong> analysis”<br />
This is the essence <strong>of</strong> two critiques by whole-design proponents levelled at the<br />
Mead system (Mead 1975), and also the Anson system (Anson 1983, 1986, 1987) to a<br />
lesser extent (Best 2002, Spriggs 1990), that is, that these seek to analyse the complex<br />
Lapita face designs at the level <strong>of</strong> the sherd by using systematics suited to strictly<br />
geometric designs. Shepard’s comments are particularly relevant to the transition from<br />
Lapita to non-Lapita, as this is widely thought by the whole-design proponents to involve<br />
a transition from complex pictorial representations (ill adapted to element-motif<br />
systematics) to strictly geometric patterns (to which such systematics are well adapted).<br />
<strong>The</strong> historical process by which this mismatch has come about is that Mead’s initial<br />
influential formulation was for a sample <strong>of</strong> mainly geometric decoration from Yanuca in<br />
Fiji (Mead 1975:37), while subsequent sampling <strong>of</strong> Lapita in Near-Oceania has recovered<br />
complex designs with greater frequency. <strong>The</strong> Mead system <strong>of</strong> analysis has in many cases<br />
been extended to cover new complex motifs without much change to the underlying<br />
element-motif-based systematics.<br />
Shepard’s propositions regarding ceramic attributes became mainstream by the<br />
1970s, by which time some were using the term “micro-seriation”. Le Blanc, for example,<br />
considered that types in any form limited the temporal resolution <strong>of</strong> seriation, as one could<br />
only define a limited number <strong>of</strong> types, beyond which declining sample size limited their<br />
utility (Le Blanc 1975:24). Too few types, it was argued, limit temporal resolution <strong>of</strong> the<br />
seriation. This is compounded by the difficulty <strong>of</strong> assigning many sherds to type, even<br />
though such sherds may display attributes.<br />
Element/motif attribute analyses possibly make better use <strong>of</strong> the variability present<br />
in small samples, but are subject to at least as much need for either independent or<br />
intuitive selection <strong>of</strong> attributes <strong>of</strong> chronological saliency, if analytical noise or<br />
contemporaneous functional variation is not to overwhelm the chronological signal. For<br />
26
attribute-based analyses, there has been very little general discussion in recent years <strong>of</strong> how<br />
this step <strong>of</strong> the seriation process might be achieved.<br />
In spite <strong>of</strong> the dominance <strong>of</strong> attribute-based analyses, arguments for a structural<br />
approach to attribute patterning continued to be put forward during the processual era:<br />
“in pottery studies, it is not the attributes (whether technological or<br />
decorative) that carry information, but the way these attributes are<br />
patterned on a particular vessel shape. By focusing on a potsherd rather<br />
than a whole vessel, the basic unit <strong>of</strong> behaviour (the vessel shape) and the<br />
structure <strong>of</strong> the other cultural units on it are missed. Ascertaining<br />
behavioural structure from potsherds, rather the structure on the vessel<br />
shape, and attempting to derive culturally meaningful information from<br />
them is like shredding a dictionary and trying to reconstruct its organization<br />
without reconstructing the pages.” (Arnold 1985:5).”<br />
Ceramic Attribute-combination Grouping Approaches in Oceania:<br />
<strong>The</strong> first example <strong>of</strong> a computer-aided multivariate attribute grouping approach to ceramic<br />
analysis in Oceania, marking the impact <strong>of</strong> the New Archaeology on Pacific ceramic<br />
classification was the work <strong>of</strong> Everett Frost in Fiji (Frost 1974). It is important to note that<br />
the aims in these early computer studies were largely chronological, rather than to do with<br />
interaction or within-site activity patterning. Frost used Sokal and Sneath’s numerical<br />
taxonomy approach, developed for taxonomy in biology (Sokal & Sneath 1963) to classify<br />
pottery assemblages. This has historical origins (in archaeology) in non-computer<br />
numerical grouping approaches (Rouse 1960, Spaulding 1953), and marks a fundamental<br />
departure (in terms <strong>of</strong> systematics) from previous approaches to ceramic chronological<br />
analysis in Oceanic archaeology (e.g. Specht 1969). A plethora <strong>of</strong> innovations in the use<br />
<strong>of</strong> computer algorithms to sort similarity matrices had emerged in archaeology in the 1960s<br />
and early ‘70s, (Marquardt 1978:270-273), <strong>of</strong> which Frost’s work is a fairly typical<br />
application for the times.<br />
Aspects <strong>of</strong> Frost’s approach, particularly Frost’s choice and coding for the<br />
computer <strong>of</strong> ceramic sherd attributes, rapidly became semi-standardized for non-Lapita<br />
27
ceramics. This standardizing effect can be seen in a series <strong>of</strong> analyses in the years<br />
subsequent to Frost’s Fijian thesis research. Frost’s analytical scheme was implemented by<br />
Irwin for his study <strong>of</strong> the Shortland Islands sequence, with some lumping <strong>of</strong> attributes,<br />
while Irwin’s PhD research on Mailu used a similar scheme (Irwin 1972:78-79, 1985:115-<br />
116), and others followed suit (Best 1984, Egl<strong>of</strong>f 1971).<br />
While similarity grouping methods have varied more in recent studies, the method<br />
<strong>of</strong> coding ceramic attributes and attribute combinations has in some studies maintained this<br />
phylogenetic similarity (see for example Bickler 1998, Clark 1999:54, Wahome 1999:36-<br />
46). Irwin provides a detailed explanation <strong>of</strong> how this was generally done. Each<br />
combination <strong>of</strong> decorative characteristics <strong>of</strong> multiple parts <strong>of</strong> the vessel was counted, and<br />
the frequency <strong>of</strong> these combinations by site-assemblage was counted (omitting measured<br />
form variables, which were analyzed separately as an “independent test” <strong>of</strong> the other<br />
results) (Irwin 1972:77-88). In Irwin’s case this initially resulted in 300 combinations,<br />
which were further reduced by combining some rim shape classes and excluding some data<br />
from sherds too small to yield anything other than “nugatory” data. Some decorative<br />
attributes were disregarded because they were seen to co-occur with others which were<br />
retained. <strong>The</strong> number <strong>of</strong> attribute combinations was thus reduced to 103.<br />
Variation in brokenness between assemblages is potentially problematic when<br />
using this method <strong>of</strong> ceramic description, although perhaps this is what Irwin meant by<br />
“nugatory”. Two identical sherds could potentially count as three entirely different<br />
combinations if one <strong>of</strong> the sherds is broken apart. In this sort <strong>of</strong> quantification tight<br />
control for the vessel parts represented by sherds would need to be maintained to negate<br />
this. Best achieved such control by recording different parts <strong>of</strong> the vessel as separate<br />
records in his data table (Best 1984), confining his attribute list to technological attributes<br />
rather than design motifs, but here the disadvantage is the loss <strong>of</strong> information from large<br />
sherds concerning the structure <strong>of</strong> decoration across the vessel (see also Wahome<br />
28
1999:100). Thus depending on the level <strong>of</strong> control for part representation, this approach<br />
can be a structural approach to whole vessels where attribute combinations are reminiscent<br />
<strong>of</strong> types, or a part-attribute approach, akin to Arnold’s “shredded dictionary”.<br />
Other Systems <strong>of</strong> Decorative Classification:<br />
A less subjective, but potentially less behaviourally salient approach is to avoid selection<br />
<strong>of</strong> attribute combinations altogether, and simply count the occurrence <strong>of</strong> all individual<br />
attributes (e.g. Anson 1983: motif inventory appendix) (see also Wahome 1999:101-103),<br />
rather than trying to capture patterned covariance-by-part in the descriptive scheme. This<br />
approach has the flaw that the types <strong>of</strong> attributes used are unknown. Which attributes are<br />
chronologically sensitive and which vary with other factors than time? This general<br />
approach has been applied by Wickler and Summerhayes, but in both cases some form <strong>of</strong><br />
control for vessel location was maintained: in Wickler’s case this was at the level <strong>of</strong> motif<br />
counts by location, while in Summerhayes’ case broader decorative techniques, rather than<br />
individual motifs, were coded by decorative location. <strong>The</strong> difficulty here is that the larger<br />
view <strong>of</strong> patterning <strong>of</strong> decorative attributes across the vessel is lost, but one advantage <strong>of</strong><br />
such an approach is that data can be obtained from quite small sherds (in Summerhayes’<br />
case patterning <strong>of</strong> decoration across the vessel was recorded by illustrating “minimum<br />
number <strong>of</strong> vessels” (MNV) sherd sets or families).<br />
Another potential problem with the Anson/Summerhayes motif analysis is that<br />
some motifs are identifiable from very small sherds, while others are larger. This provides<br />
an additional mechanism by which motif frequencies or presence-absence may be biased<br />
by differences in brokenness between assemblages, particularly for large incised motifs, a<br />
problem noted in relation to Reef-Santa Cruz incised Lapita motifs (Donovan 1973:15).<br />
Other approaches to descriptive classification <strong>of</strong> Lapita decoration include Mead’s<br />
formulaic system <strong>of</strong> design classification (Mead 1975) (see Specht 1977 for a critique),<br />
29
and Siorat’s structurist system, whereby variants <strong>of</strong> a design are classified by reference to<br />
clues as to their sequence <strong>of</strong> execution (Siorat 1990). <strong>The</strong>re are some differences between<br />
these broadly similar approaches in how the distribution <strong>of</strong> decoration across the vessel is<br />
coded. Mead’s “design zones” are not very specific in some cases, are defined in relation<br />
to a limited range <strong>of</strong> vessel forms and are not primary to the classification <strong>of</strong> design<br />
variability into motifs. Parker, using the Mead system, noted which locations on the<br />
vessels each motif occurred on (Parker 1981), but these records were not reported on a<br />
sherd-by-sherd basis, and frequency data <strong>of</strong> decorative locations were not given. Siorat<br />
was more specific than Mead on the sequence <strong>of</strong> application <strong>of</strong> decoration, and the<br />
location <strong>of</strong> decoration, and noted the difficulties <strong>of</strong> applying such a system to very<br />
fragmented collections.<br />
Green’s synthesis <strong>of</strong> Lapita variability (Green 1978) serves as a study in contrast<br />
to the Frost/Irwin approach to description <strong>of</strong> decorative variability, although analytical<br />
methods were very much in the Frost/Irwin attribute multivariate grouping mode, and also<br />
included Irwin-style application <strong>of</strong> Renfrew and Sterud’s close-proximity analysis. One <strong>of</strong><br />
Green’s aims was to show that there was a Lapita ceramic series divisible into a western<br />
adaptation and a derivative, but isolated, Eastern Lapita. He thus wished to identify spatial<br />
and temporal components to the Lapita horizon. His construction <strong>of</strong> a Lapita Ceramic<br />
series using the Mead system <strong>of</strong> decorative classification was made to test whether ceramic<br />
variability was consistent with his theory <strong>of</strong> an isolated Eastern Lapita region.<br />
Green’s motifs were those identified by contributors to the Mead system <strong>of</strong> design<br />
classification (Donovan 1973, Mead 1975), and were a fundamental departure in ceramic<br />
decorative description from the Irwin/Frost attribute combination approach. A<br />
consequence <strong>of</strong> the schism between the Frost/Irwin approach to ceramic description and<br />
the Mead/Donovan approach was that different periods within the Oceania regional<br />
30
chronology began to be analyzed using these separate systems <strong>of</strong> ceramic design<br />
description. Frost/Irwin attribute descriptions (with modifications) were generally applied<br />
to post-Lapita ceramics (Best 1984, Clark 1999, Egl<strong>of</strong>f 1971, Wahome 1999), while<br />
Lapita-phase ceramics were analyzed using the Mead-Donovan system, or its “splitting”<br />
derivative, the Anson system <strong>of</strong> motif inventory (Anson 1983). Another Mead/Donovan<br />
derivative, (having an extra hierarchical level <strong>of</strong> motif classification), was added by Sharp<br />
(Sharp 1988, Sharp 1991), allowing a bit more <strong>of</strong> a lumping approach, always beneficial<br />
for sample sizes, but this work was never completed.<br />
Another structurist approach to ceramic attribute description, in this case to post-<br />
Lapita decoration, for which variations in brokenness between contexts can be controlled,<br />
is seen in a transformation <strong>of</strong> the Irwin/Frost data structure (Clark 1999:65-66, Appendix<br />
2). For the purpose <strong>of</strong> his multidimensional scaling analysis (MDS) <strong>of</strong> variation, Clark re-<br />
organized the attributes “surface modification position” and “surface modification” (which<br />
had followed the Irwin/Frost/Best structure in initial recording) into three attributes,<br />
“Decoration- Lip”, “Decoration-Rim” and “Decoration-Body” (a similar data structure<br />
can be found in Shennan 1997:6). In this scheme control for brokenness can be maintained<br />
more easily than in the Frost/Irwin scheme, by coding absence <strong>of</strong> a part from a sherd into<br />
this data structure as a filter attribute, or by limiting the analysis to a subset <strong>of</strong> sherds for<br />
which part representation is either complete or equal (as is also possible, but more difficult,<br />
in the Frost/Irwin system by excluding some “nugatory” combinations). <strong>The</strong> trade<strong>of</strong>f is<br />
proliferation <strong>of</strong> data fields or attributes, as each vessel part needs its own set <strong>of</strong> attributes,<br />
which can lead to an unwieldly table unless a relational database is used for recording.<br />
A potential problem with such an approach is the saliency <strong>of</strong> the vessel part<br />
classification used. If the conception by past potters <strong>of</strong> design zones/vessel parts is <strong>of</strong> a<br />
different level <strong>of</strong> precision to that <strong>of</strong> the analyst, some analytical “noise” can result.<br />
31
Examples <strong>of</strong> this will be given in the Roviana ceramic analysis in this thesis (see Chapter<br />
9). <strong>The</strong> solution to this problem lies in conducting exploratory data analysis, seeking to<br />
understand patterning in the data, covariance between decorative attributes and location<br />
on the vessel. If the analyst’s definition <strong>of</strong> vessel parts is too finely split, there should be<br />
a tendency for some decorative attributes to occur on adjacent vessel parts, and these part-<br />
attribute combinations can be lumped to boost sample size. Iterative re-classification <strong>of</strong><br />
both decoration and vessel parts is needed to capture structure in the data, <strong>of</strong> the layout<br />
<strong>of</strong> decoration across the vessel.<br />
It can be seen above that decorative classification for different periods <strong>of</strong> the<br />
Oceanic sequence has developed along different lines. Lapita and post-Lapita/non-Lapita<br />
studies have diverged in their units <strong>of</strong> description/classification, more so than in methods<br />
<strong>of</strong> analysis. One reason for this is the reduction in complexity <strong>of</strong> decorative technique and<br />
decorative patterning from Lapita to non-Lapita. Lapita technique comprises a set <strong>of</strong><br />
decorative tools; roulettes, dentate-stamps, circular stamps and carving blades. Post-<br />
Lapita, (or perhaps it would be better to say non-Lapita,) these specialist tools are absent,<br />
leaving widely available general tools such as fingernails, shell edges, sticks or leaf midribs,<br />
and simple blades or fingernails for incision (comb incision and carved paddles are<br />
exceptions to this trend). <strong>The</strong> potters’ construction tools may not have changed much<br />
(paddle and anvil technique seems to be ubiquitous in all periods, if predominant post<br />
Lapita), but the decorative toolkit became less specialized. In addition to this change to the<br />
decorative toolkit, the complex pictorial and geometric designs <strong>of</strong> the Lapita phase give<br />
way to simpler designs <strong>of</strong> a more geometric type, some <strong>of</strong> which may be found anywhere<br />
from the European Neolithic (e.g. Beaker pottery) to the late-prehistoric pottery <strong>of</strong> the<br />
Amazon Basin (e.g. Arqueologicas 1970). It is also suggested above that many analytical<br />
schemes are sensitive to differences in the level <strong>of</strong> brokenness or fragmentation <strong>of</strong> pottery<br />
assemblages.<br />
32
How then might we classify Lapita and post-Lapita pottery from Oceania to avoid<br />
making a leap in systematics in mid-analysis? What is needed is a system <strong>of</strong> design<br />
classification sensitive to temporal variability, but insensitive to variation in the level <strong>of</strong><br />
fragmentation or brokenness, and salient to the Lapita/post-Lapita transition. An example<br />
<strong>of</strong> such a design-classification approach can be seen in Wickler’s analytical classification<br />
<strong>of</strong> incised decoration into bounded incised versus unbounded incised (using a modification<br />
<strong>of</strong> Mead’s general zone marker concept), although Wickler does not state the advantages<br />
<strong>of</strong> his classification. By making a decision that this simple distinction was salient to his<br />
chronological objectives, he subsumed a range <strong>of</strong> designs within a simple binary<br />
categorization, where Mead, faced with the same sherds, would, in attempting to describe<br />
how each design was put together, have come up with a whole list <strong>of</strong> motifs and their<br />
all<strong>of</strong>orms (elaborations <strong>of</strong> the same underlying motif structure), and Anson would have<br />
been in danger <strong>of</strong> splitting even further, classifying at the level <strong>of</strong> these variants.<br />
Latitudinal bands <strong>of</strong> decoration (general zone markers in Mead’s terminology) are<br />
usually located near corner points, like carinations, necks and lips. Typically carinations<br />
and necks preserve well due to their form strength, and even in very broken assemblages<br />
one can gain a good idea <strong>of</strong> the relative frequencies <strong>of</strong> these categories. Latitudinal general<br />
zone markers are therefore identifiable to type and location on the vessel over a range <strong>of</strong><br />
brokenness in some cases, since lips, necks and carinations are identifiable even when<br />
sherds are quite small. Where larger sherds suggest these bands are continuous around the<br />
latitude <strong>of</strong> the vessel and comprising repeats <strong>of</strong> closely spaced elements, there is no danger<br />
<strong>of</strong> their identity changing with changes in brokenness, provided the spacing <strong>of</strong> elements<br />
is substantially closer than the minimum dimension <strong>of</strong> sherds. For more widely spaced<br />
bands <strong>of</strong> elements, such as when an element is only repeated three or four times around<br />
the band <strong>of</strong> latitude, a high level <strong>of</strong> brokenness (small sherd size) would result in a<br />
mixture <strong>of</strong> plain and decorated sherds from what was originally a single decorated vessel,<br />
33
so the spacing <strong>of</strong> elements in relation to sherd size must be borne in mind when using<br />
banded decoration in this manner, and the abundance <strong>of</strong> plain lip sherds would be<br />
misleading.<br />
Quantification:<br />
If similarity between pottery samples is quantified using counts <strong>of</strong> the abundance <strong>of</strong><br />
attributes, what measures <strong>of</strong> quantity are appropriate to one’s particular samples? This<br />
subject has been thoroughly reviewed elsewhere (Orton 1982, 1993, Orton & Tyers 1991),<br />
but the last word on the subject has yet to be written, and this fundamental problem can<br />
only be partially avoided by resorting to occurrence seriation (discussed below), as sample<br />
richness is related to sample size for unsaturated samples, and sample size is a<br />
quantification problem.<br />
Units <strong>of</strong> Quantification and Sample Brokenness:<br />
Orton (1993) identifies one <strong>of</strong> the principal issues in pottery quantification as whether the<br />
measure used will bias the relative abundance <strong>of</strong> types/attributes. A complicating factor<br />
is that the attributes chosen for study themselves affect which measures will bias the<br />
relative abundance. A good example <strong>of</strong> this is the relative abundance <strong>of</strong> plain versus<br />
decorated pottery. If we have a hypothetical assemblage <strong>of</strong> ten globular pots, <strong>of</strong> which five<br />
vessels each have a different band <strong>of</strong> decoration somewhere on the vessel, while five pots<br />
are plain, quantification <strong>of</strong> the attributes is simple, until we smash the vessels. <strong>The</strong><br />
complete vessels can be counted if unbroken, yielding 50% decorated. Unfortunately, if<br />
using sherd count as a measure <strong>of</strong> quantity <strong>of</strong> the attributes decorated and plain, the more<br />
we smash them, the plainer the assemblage gets, because the undecorated areas <strong>of</strong> the<br />
decorated pots are counted as plain observations. Clearly simple sherd count is the wrong<br />
34
measure <strong>of</strong> quantity for these particular attributes.<br />
If the smashed assemblage were quantified as the logical minimum number <strong>of</strong><br />
vessels (MNV) represented, the five different bands <strong>of</strong> decoration would result in a count<br />
from the pile <strong>of</strong> sherds <strong>of</strong> at least five decorated vessels, while sherds from the five<br />
undecorated vessels would be counted as a single vessel, generating an assemblage <strong>of</strong> 83%<br />
decorated pots. Clearly this would be a complete mismatch between decorative attributes<br />
and unit <strong>of</strong> quantification. <strong>The</strong>re are measures that would work for this hypothetical<br />
example, for example, the sherds could be quantified using sherd area, if we had some prior<br />
knowledge <strong>of</strong> the extent <strong>of</strong> decoration on vessels and the surface area <strong>of</strong> vessels, we could<br />
work out that each <strong>of</strong> the summed areas <strong>of</strong> the five decorative types corresponded to a<br />
single vessel, and that there was enough sherdage in total for ten vessels, thus by<br />
subtraction we might arrive at the correct answer, 50% <strong>of</strong> vessels decorated. Assuming,<br />
that is, that the recovered sample includes all sherds from each pot.<br />
<strong>The</strong> difficulty and need for prior knowledge could be substantially alleviated<br />
though, by forgetting about the relative abundance <strong>of</strong> undecorated sherds, and<br />
concentrating on the relative abundances <strong>of</strong> the five different types <strong>of</strong> decoration. If the<br />
hypothetical situation is complicated slightly though, by having one <strong>of</strong> the decorative types<br />
extend over a greater portion <strong>of</strong> the vessel than another decorative type, then the MNV<br />
measure <strong>of</strong> relative abundance will be better than the sherd count. Unfortunately, if there<br />
were five pots <strong>of</strong> one decorative type and one <strong>of</strong> each <strong>of</strong> the other four types, this would<br />
not work. In this circumstance sherd count might be giving a better approximation <strong>of</strong> the<br />
relative abundance <strong>of</strong> the vessel decorative types. <strong>The</strong> point is that quantification is a tricky<br />
business, even if one is not so naive as to wish to seriate using the relative abundance <strong>of</strong><br />
plain and decorated pottery, quantified by sherd count.<br />
Units <strong>of</strong> Quantification and Vessel Completeness:<br />
35
In the critical review <strong>of</strong> the Lapita temporal constructs in Chapter 2, quantification is an<br />
issue that arises continually, both in relation to relative abundances <strong>of</strong> decorative attributes,<br />
and in relation to quantification <strong>of</strong> sample size. Sample size(n) in the statistical sense refers<br />
to the number <strong>of</strong> observations comprising the sample. For pottery we must ask,<br />
observations <strong>of</strong> what? Of pots or <strong>of</strong> pieces <strong>of</strong> pots? <strong>The</strong> difficulty with pieces <strong>of</strong> pots is<br />
that the observations are clearly not independent if decoration occurs in repeated elements<br />
or motifs across the pot. This is why sherd count is a biased measure if the extent <strong>of</strong><br />
decoration across the pot is uneven, and why MNV is not if there is only one pot per<br />
decorative type. Pots themselves may not be independent observations <strong>of</strong> behavioural<br />
variability if production is highly standardized, thus if one potter turns out fifty identical<br />
pots <strong>of</strong> one type, while another turns out five identical pots <strong>of</strong> another, and we are<br />
interested in the question <strong>of</strong> how variable pottery production was between producers, the<br />
number <strong>of</strong> observations in the sample is two, not fifty-five.<br />
In the review that follows I try to make the point that occurrence seriations <strong>of</strong><br />
broken pottery assemblages are mis-applications <strong>of</strong> a technique, because sample size is so<br />
difficult to quantify, and yet sample richness is so <strong>of</strong>ten related to sample size. How can<br />
we tell which decorative attributes are common in some sites, and thus robust indicators<br />
<strong>of</strong> presence or absence, and which are not? Can this be done by counting the pieces or by<br />
some other measure such as MNV? I have shown above that both <strong>of</strong> these measures can<br />
be wildly misleading in some circumstances. A more fundamental question that might in<br />
many circumstances allow an assessment <strong>of</strong> sample size, is to ask “what sort <strong>of</strong> sample is<br />
this?” Can the sherds in the sample be refitted to form complete vessels, or do none <strong>of</strong><br />
them seem to match? If they don’t match, is this because edges are rounded from some<br />
taphonomic process, or is it because the adjacent sherds are absent from the sample? Even<br />
if none <strong>of</strong> the sherds has matching edges, the question <strong>of</strong> whether the sherds are likely to<br />
be from the same pot or not can still be asked. Thus one <strong>of</strong> the primary questions in<br />
36
ceramic sample evaluation is “what is the vessel completeness <strong>of</strong> the sample?”<br />
Analysis <strong>of</strong> vessel completeness can give an indication <strong>of</strong> which measures <strong>of</strong> sample<br />
size and/or relative abundance are appropriate to the specific sample. At one extreme, if<br />
there are indications that most sherds are singletons, i.e. other sherds from the same pot<br />
are not present, then sherd count and vessel count are the same, sherd count observations<br />
are likely to be independent in most cases, and can be used as a unit <strong>of</strong> measurement for<br />
relative abundances and sample size. At the other extreme, where vessel completeness is<br />
high for all vessels in the sample, a vessel-based count might be the most appropriate<br />
measure <strong>of</strong> relative abundances, and would certainly be a much better indication <strong>of</strong> sample<br />
size for occurrence seriation than sherd count. Vessel completeness information is<br />
therefore an important tool in choosing an appropriate method <strong>of</strong> sample size<br />
quantification, although vessel completeness is not always easy or even possible to<br />
measure. Where it is impossible to measure vessel completeness, sample size is difficult to<br />
quantify other than by resort to sum <strong>of</strong> some measure <strong>of</strong> estimated vessel equivalent, to<br />
calculate a minimum number <strong>of</strong> vessels.<br />
Sherd Assemblages as Samples:<br />
<strong>The</strong> development <strong>of</strong> relatively labour-intensive sherds-as-vessels approaches to<br />
quantification is in part a recognition <strong>of</strong> the difficulty <strong>of</strong> comparing assemblage<br />
compositions, unless they happen to have the same levels <strong>of</strong> completeness and brokenness<br />
(Orton 1993: 169-170), and vessel completeness cannot be measured other than by the<br />
grouping <strong>of</strong> fragments into vessel classes. If the measured or estimated levels <strong>of</strong><br />
brokenness and/or completeness differ between assemblages, we must conclude that the<br />
assemblages are not equivalent samples <strong>of</strong> the material discarded by people in the past.<br />
How numerically large is our sample <strong>of</strong> potting behaviour <strong>of</strong> the past? Can the<br />
sample be read directly as representing people's potting behaviour, or is it a sample <strong>of</strong> a<br />
37
sample <strong>of</strong> a sample, subject to a number <strong>of</strong> transformations in composition? In either case,<br />
what sort <strong>of</strong> sample (representative or biased)? What is the sampling fraction? (Again,<br />
implicit in this question is the problem <strong>of</strong> how and what we are counting).<br />
Several Pacific archaeologists have touched on these subjects over the years.<br />
Specht provided a cogent discussion (Specht 1969: 70-73), Egl<strong>of</strong>f published on the subject<br />
(Egl<strong>of</strong>f 1973) independently inventing a new measure <strong>of</strong> quantity that has subsequently<br />
been widely used. Parker was careful to provide at least two different measures <strong>of</strong> sample<br />
size in her analysis <strong>of</strong> Reef-Santa-Cruz pottery (Parker 1981). Summerhayes has given the<br />
matter some thought, particularly in regard to the effect <strong>of</strong> variation in the extent <strong>of</strong><br />
decoration across the pot on the ratio <strong>of</strong> plain to decorated sherd count (Summerhayes<br />
2000c, 2002), and Clark has noted some <strong>of</strong> the biases inherent in the MNV count (Clark<br />
1999). On the world stage, the subject has been thoroughly reviewed for pottery in general<br />
(Orton 1993), amid some pessimism that we will ever posses absolute measures or<br />
estimates <strong>of</strong> vessel populations (see for example Orton 2000).<br />
<strong>The</strong> analysis <strong>of</strong> brokenness and completeness in Chapter 5 derives from an<br />
unpublished manuscript (Felgate & Bickler n.d), a methodological paper on estimating<br />
parent vessel populations from sherd samples. This approach is a fundamental departure<br />
(enabled in part by properties <strong>of</strong> samples obtained from large-area surface collections) from<br />
the approaches usually taken in Pacific archaeology, which implicitly regard the recovered<br />
archaeological sample as the target population, by regarding the measured properties <strong>of</strong><br />
the sample as representative in an unspecified way <strong>of</strong> the properties <strong>of</strong> people’s lives in the<br />
past. We need to be more specific about the nature <strong>of</strong> that representation if our behavioural<br />
inferences based on sherd assemblages are to be robust.<br />
An estimate <strong>of</strong> the total quantity <strong>of</strong> pottery represented by the recovered sample<br />
from the Roviana intertidal-zone sites is useful firstly as it gives some idea <strong>of</strong> the potential<br />
for bias in the sample. Seriation using decorative attribute abundances will be affected if<br />
38
the relative abundance <strong>of</strong> an attribute is linked to the degree and type <strong>of</strong> postdepositional<br />
sampling. Comparison <strong>of</strong> estimated breakage population and sample assemblage size will<br />
allow some idea <strong>of</strong> sampling fraction (both in terms <strong>of</strong> the bulk <strong>of</strong> sherdage and the<br />
number <strong>of</strong> vessels), and the potential for bias. This subject will be explored further in<br />
Chapter 7, where differences in part-representation for different vessel styles will be<br />
investigated.<br />
Accumulations research (estimating the intensity/duration <strong>of</strong> occupation through<br />
the quantity <strong>of</strong> pottery in the breakage population) is possible where an estimate <strong>of</strong> the<br />
breakage population is available (see Varien & Mills 1997, Varien & Potter 1997). If an<br />
independent chronology is available that allows total duration and historical continuity <strong>of</strong><br />
occupation to be modeled, then intensity <strong>of</strong> occupation can be the focus <strong>of</strong> study. <strong>The</strong><br />
results <strong>of</strong> the analysis in this <strong>chapter</strong> will be examined from this perspective in Chapter<br />
13.<br />
Knowing how much is missing from the sample informs indirectly on<br />
postdepositional formation processes. In this thesis evidence is presented that suggests the<br />
vast majority <strong>of</strong> the sherdage discarded at the Roviana intertidal-zone sites is no longer<br />
extant or recognizable as potsherds. <strong>The</strong> extent <strong>of</strong> this loss <strong>of</strong> material is estimated and<br />
explanations for this loss are explored in Chapter 7.<br />
Interpretation <strong>of</strong> survey results requires an assessment <strong>of</strong> the probability <strong>of</strong><br />
detection <strong>of</strong> evidence: when is absence <strong>of</strong> evidence acceptable evidence <strong>of</strong> absence? <strong>The</strong><br />
sherd sampling fraction and the vessel sampling fraction are also ways <strong>of</strong> characterizing<br />
the state <strong>of</strong> preservation <strong>of</strong> the extant deposit, which says something about site<br />
detectability. To what extent are they preserved in particular circumstances? Under what<br />
general conditions <strong>of</strong> wave exposure etc. do we see these sites preserved? Developing<br />
answers to these questions is crucial to interpreting the archaeological record <strong>of</strong> the wider<br />
region. If almost all <strong>of</strong> the breakage population has vanished even in the sheltered setting<br />
39
<strong>of</strong> the Roviana Lagoon, this has fundamental implications for the interpretation <strong>of</strong> the<br />
archaeological record elsewhere in Near Oceania. Just how fragile is this site type?<br />
Sample Formation <strong>The</strong>ory : the Breakage Population:<br />
<strong>The</strong> central theoretical concept put forward (following Felgate & Bickler n.d) to provide<br />
a definition <strong>of</strong> a target population is the breakage population, a theoretical stage in the<br />
formation <strong>of</strong> an archaeological sample where vessels are broken in some unspecified<br />
manner, but are also complete, i.e. all the fragments are present (Felgate & Bickler n.d).<br />
While it is clear that an archaeological sample <strong>of</strong> vessel sherds is unlikely to have ever been<br />
all in such a state at the same time, this does not negate the utility <strong>of</strong> the concept as a<br />
target population for estimation procedures (Figure 5) (see also De Boer 1983). Points<br />
to note are that the “behavioural assemblage” is regarded as equivalent or closely related<br />
to Schiffer's “systemic context” (Schiffer 1995a) rather than being a life assemblage in the<br />
palaeoecological sense <strong>of</strong> a fossilized tableau <strong>of</strong> organisms in life position (Brenchley &<br />
Harper 1998:69), such as might be seen in parts <strong>of</strong> Pompeii.<br />
Felgate and Bickler introduced the term “breakage population” as roughly<br />
analogous to DeBoer's use <strong>of</strong> the palaeoecological term “death assemblage” and draw a<br />
distinction between the breakage population and DeBoer's “discard assemblage” or<br />
Schiffer’s “discard flow”. <strong>The</strong> breakage population is defined as a theoretical time-<br />
accumulated set <strong>of</strong> the broken, but complete, pottery vessels that form the parent vessel<br />
population <strong>of</strong> a sample <strong>of</strong> potsherds from an archaeological context. This definition<br />
assumes that the use-life <strong>of</strong> a vessel ends upon breakage, but is unaffected by re-use <strong>of</strong><br />
some portions <strong>of</strong> the vessel as a secondary artefact or recycling into grog temper. DeBoer<br />
(1983) sometimes refers to a “total discard assemblage” which would seem to be<br />
40
effectively equivalent to the breakage population.<br />
Use-life seems to be the most important factor linking composition <strong>of</strong> behavioural<br />
assemblages with breakage populations (David & Wilson 1999, de Barros 1982, Mills<br />
1989, Shott 1989b, 1996). <strong>The</strong> relationship between behavioural assemblages and discard<br />
assemblages is more complex than that between behavioural assemblages and breakage<br />
populations, involving a number <strong>of</strong> factors in addition to use life, such as discard practices<br />
and re-use <strong>of</strong> fragments. By separating breakage and discard, a parent population is<br />
modeled in the formation <strong>of</strong> the archaeological record in which artefacts can exist beyond<br />
their use-life, but in a complete state.<br />
<strong>The</strong> composition <strong>of</strong> this breakage population has been structured by use-life, but<br />
with vessel completeness at 100%, unaffected by any re-use <strong>of</strong> sherds, etc. No statement<br />
regarding the degree <strong>of</strong> brokenness <strong>of</strong> this population is implied here beyond the<br />
expectation that the vessel is cracked beyond use or broken into any number <strong>of</strong> sherds.<br />
Comparison <strong>of</strong> average vessel completeness in the archaeologist’s sample with average<br />
vessel completeness in the breakage population (the latter is by definition 100%) is the key<br />
to estimating the breakage population.<br />
It is worth reiterating that the target population in the estimation approach <strong>of</strong><br />
Felgate and Bickler is the breakage population, rather than any <strong>of</strong> the other “stages” in the<br />
model <strong>of</strong> formation given in Figure 5. <strong>The</strong> estimate obtained does not inform directly on<br />
the extent <strong>of</strong> the archaeologist's sampling <strong>of</strong> the extant deposit, or on taphonomic<br />
transforms, nor does such an estimate provide information on use-life factors as these<br />
transform a behavioural assemblage into a breakage population.<br />
To show how vessel completeness can be used to estimate a breakage population,<br />
41
Felgate and Bickler turn to an explanation <strong>of</strong> similar estimation in faunal analysis. Orton<br />
explains:<br />
“<strong>The</strong> formula was based on the idea <strong>of</strong> matching pairs <strong>of</strong> bone<br />
elements; in a population, there are equal numbers <strong>of</strong> left and right<br />
elements. As the population is reduced by sampling effects, some<br />
left and some right elements are removed, leaving unmatched left<br />
elements, unmatched right elements, and matching pairs. <strong>The</strong><br />
smaller the sampling fraction, the smaller the proportion <strong>of</strong><br />
matching pairs is likely to be. <strong>The</strong> numbers <strong>of</strong> unmatched left and<br />
right elements, and <strong>of</strong> matching pairs, can thus be used to estimate<br />
the original size <strong>of</strong> the population, given assumptions about the<br />
nature <strong>of</strong> the taphonomic process (Krantz 1968:215). (Orton<br />
2000:54).<br />
Pots do not break naturally into matching pairs, unlike skeletons, but the model <strong>of</strong> sample<br />
formation described by Orton can be expanded to any archaeological fragmented materials<br />
in which essential classes exist, like collections <strong>of</strong> potsherds, in which essential vessel<br />
classes must exist (even though these may be difficult or impossible to identify in many<br />
cases). In general terms, the model <strong>of</strong> sample formation described by Orton can be<br />
rephrased as the problem <strong>of</strong> estimating the number <strong>of</strong> species or classes in a population,<br />
based on the properties <strong>of</strong> the sample.<br />
This is not just an archaeological problem; nonparametric statistical methods exist<br />
for solutions applied to a wide range <strong>of</strong> such estimation problems. Felgate and Bickler refer<br />
to the vast literature on capture-recapture studies where capturing and recapturing <strong>of</strong><br />
tagged animals can be used to estimate populations <strong>of</strong> animal species in a landscape (see<br />
Seber 1986 for a review). A review <strong>of</strong> the related problem <strong>of</strong> the estimation <strong>of</strong> number <strong>of</strong><br />
species/classes in a population (Bunge & Fitzpatrick 1993) noted a range <strong>of</strong> working<br />
solutions to the “number <strong>of</strong> species” problem, and noted prospects for improvements in<br />
this regard in the future (see also Solow 1994 for a Bayesian approach).<br />
<strong>The</strong> question <strong>of</strong> the types <strong>of</strong> bias likely to be present in samples becomes important<br />
in the context <strong>of</strong> samples comprising mostly singleton sherds. This is one <strong>of</strong> the questions<br />
42
investigated in Chapter 7. <strong>The</strong> other major question raised by the analysis in Chapter 5,<br />
also addressed in Chapter 7, is, “What happened to the pottery that didn’t make it into our<br />
sample?”<br />
For seriation, if it can be shown through an analysis <strong>of</strong> vessel completeness that<br />
most decorated sherds are likely to be singletons, or independent observations <strong>of</strong> the nature<br />
<strong>of</strong> vessels in the breakage population, then sherd count is a useful measure <strong>of</strong> sample size<br />
and relative abundance <strong>of</strong> attributes. An argument will be put forward in Chapter 5 to<br />
suggest that this is the case for the Roviana early ceramic samples from the sea. <strong>The</strong> same<br />
argument is applicable to the calculation <strong>of</strong> sample sizes. If decorated sherds are mostly<br />
independent observation <strong>of</strong> vessel properties, then sherd counts <strong>of</strong> decorative and form<br />
attributes are good indications <strong>of</strong> the number <strong>of</strong> vessels represented in the samples. This<br />
is important for seriation analysis, in which, as will be detailed in the following section,<br />
attribute sample sizes weight the analysis and affect the outcome.<br />
Seriation <strong>The</strong>ory : a Review:<br />
Seriation is a term in use principally in the fields <strong>of</strong> psychology and archaeology. In<br />
psychology it refers to the cognitive ability to order objects based on similarity, but in<br />
archaeology the meaning is more specific. When archaeologists talk about seriation they<br />
usually mean (or should mean) the creation <strong>of</strong> a temporal series based on similarity and<br />
sometimes evolutionary hypotheses.<br />
Types <strong>of</strong> Seriation:<br />
A recent taxonomy <strong>of</strong> seriation techniques divides seriation into similarity approaches, and<br />
evolutionary approaches, the latter based on a rule <strong>of</strong> evolutionary development.<br />
43
Figure 5: Processes <strong>of</strong> formation <strong>of</strong> the Archaeological record (After<br />
Felgate and Bickler n.d., adapted from De Boer 1983): inferring a<br />
breakage population from an archaeological sample.<br />
Similarity approaches are further subdivided into frequency seriation, occurrence seriation<br />
and phyletic seriation (O'Brien & Lyman 2000:64). All <strong>of</strong> these approaches have been<br />
applied to Lapita, in various mixtures, with little explicit awareness or discussion <strong>of</strong> the<br />
differences between them.<br />
Evolutionary seriation is one in which a constant direction <strong>of</strong> change or an<br />
44
evolutionary rule <strong>of</strong> development is specified. Summerhayes’ statement, for example, that<br />
his stratigraphic data identifies the direction <strong>of</strong> change (Summerhayes 2000a:151) has an<br />
implicit evolutionary component.<br />
Seriation involves firstly description <strong>of</strong> ceramic assemblages in terms <strong>of</strong> criteria<br />
(types or attributes) that are thought to vary over time (Cowgill 1972, Duff 1996, Dunnell<br />
1970, Lyman et al. 1998, Marquardt 1978, O'Brien & Lyman 2000, Shennan 1997:342).<br />
If a randomly selected set <strong>of</strong> descriptive attributes is used rather than a set selected as<br />
representing temporal change, an “...inseparable hodgepodge...” <strong>of</strong> geographic, functional<br />
and temporal variation might result (Dunnell 1970:310). (See also Braun 1983:113 on the<br />
confusion <strong>of</strong> functional and decorative variation, and the additional factor <strong>of</strong> the<br />
relationship between vessel size and decoration). While form/function/technological<br />
attributes may vary over time (Rice 1982:48), these should only be used in seriation if<br />
independent evidence is available to show that this was the case. Also, pottery production<br />
styles can be expected to vary across space. Braun also notes the need, given different use<br />
and breakage rates in different contexts, to seriate comparable contexts <strong>of</strong> deposition, and<br />
the need to control for vessel part in treating decorative variability (Braun 1983).<br />
Approaches to the identification <strong>of</strong> chronological variability range from intuitive<br />
examination in combination sometimes with the direct historical method <strong>of</strong> working back<br />
from the recent past as in Kroeber’s original formulation (O'Brien & Lyman 2000:111-<br />
114), to sophisticated computer techniques <strong>of</strong> data exploration. A simple method,<br />
sometimes used by archaeologists in the culture historical era, was to control for functional<br />
variation by performing seriation on a single vessel form, and to control for spatial<br />
variation by limiting the seriation to sites from a small region, regarding the remaining<br />
variation as temporal variation in style.<br />
Various people have pointed out that seriation is best carried out on sites that<br />
represent a short occupation, or on sites with similar occupation spans (O'Brien & Lyman<br />
2000:117). Marquardt (1978) takes a more relaxed stance on this issue than Dunnell<br />
45
(1970). Ideally one would like large samples from short, well-preserved occupations, and<br />
these may well be conditions which are met in the archaeological record in some parts <strong>of</strong><br />
the world. This has yet to be securely demonstrated in the Oceanic Lapita examples<br />
reviewed in Chapter 2.<br />
In traditional graphical frequency seriation, assemblages are graphed as relative<br />
frequencies summing to 100%. Bars representing the relative proportions <strong>of</strong> selected types<br />
or attributes are sorted until unimodal “battleship curves” are produced, usually interpreted<br />
as representing the waxing and waning popularity through time <strong>of</strong> items meeting the<br />
classification criteria. (Marquardt 1978:261). This “popularity principle” is <strong>of</strong>ten<br />
incorrectly stated as an assumption <strong>of</strong> the technique. Other interpretations are possible,<br />
relating to deposition similarity rather than directly to production similarity and popularity<br />
(Goldmann 1971). Le Blanc notes in regard to this “popularity principle”:<br />
“Even if... changes were for all practical purposes instantaneous, the<br />
archaeological record <strong>of</strong> these changes will almost invariably show them as<br />
gradual with periods <strong>of</strong> mixed technology or style (Le Blanc 1975:23).”<br />
McNutt shows that while temporal ordering <strong>of</strong> sites using graphical type frequency<br />
seriation may or may not produce a correct ordering, to interpret battleship curves as<br />
popularity curves is to misinterpret the seriation schema as reflecting absolute abundance<br />
<strong>of</strong> types over time. (McNutt 1973:51, 58). McNutt considers frequent errors <strong>of</strong> this sort<br />
to be due to casual use <strong>of</strong> the word “frequency”(relative abundance might be a better<br />
term).<br />
Dunnell points out that the extent to which types or attributes form such battleship<br />
curves can be used in dialogue with reclassification to develop a classificatory scheme<br />
(Dunnell 1970:309). Selecting attributes which form battleship curves and produce a neat<br />
seriation is key in Dunell’s view to identifying those aspects <strong>of</strong> variability which are<br />
stylistic.<br />
<strong>The</strong> most widespread seriation technique in the 70s and 80s was to construct an<br />
assemblage similarity matrix using attribute/type counts (converted to relative frequencies<br />
46
summing to 100% for all units), and the Robinson coefficient <strong>of</strong> similarity. It is important<br />
to note that there is no control or weighting for assemblage sample size built into the<br />
resulting similarity matrix, and that, as for graphical frequency seriation, the resulting<br />
pattern <strong>of</strong> site similarity may contain spurious observations from mixed contexts or small<br />
samples. <strong>The</strong>se may be less obvious than in graphical frequency seriation (where the<br />
<strong>of</strong>fending units would be unlikely to fit well anywhere in the series). Fortunately methods<br />
exist for evaluating the goodness <strong>of</strong> fit <strong>of</strong> a matrix sort using this sort <strong>of</strong> data, and seriation<br />
using Robinson coefficient <strong>of</strong> similarity, followed by matrix sorting or cluster analysis has<br />
been successful in many instances.<br />
Lack <strong>of</strong> fit in the sort order can be as a result <strong>of</strong> inadequate sample size, or<br />
representation <strong>of</strong> an anomalous occupation span, possibly <strong>of</strong> material from temporally<br />
separate occupations, but not necessarily so (de Barros 1982, Dunnell 1970: 312). <strong>The</strong><br />
trouble in Oceania is that usually, due to poor sampling, one must put up with a certain<br />
jaggedness in the seriation or wind up with no seriation at all, the corpus <strong>of</strong> sites for the<br />
Lapita period at least being too small to allow fussiness in most cases, when dealing with<br />
Lapita sites from a small area.<br />
McNutt considered that both graphical type seriation and similarity matrix<br />
approaches such as Robinson matrices should be regarded as techniques that work<br />
sometimes rather than as having any methodological validity, and bemoaned the lack <strong>of</strong><br />
caution in their application by the 1970s in contrast to a commended caution displayed by<br />
earlier practitioners (McNutt 1973:58-60). <strong>The</strong>se comments, although not directed at<br />
Oceania similarity matrix-based seriations, seem pertinent here, as will be argued with<br />
detailed reference to specific examples in Chapter 2.<br />
<strong>The</strong> predominant seriation technique in archaeology today, especially in Europe,<br />
is Correspondence Analysis (CA), a multivariate data-exploratory technique suitable for<br />
presence absence data or counts. When used for counts, some control for sample size is<br />
present, as both attribute/type total sample size and the total count for each unit are used<br />
47
to weight data. Attribute/type covariation is controlled for by calculating summary<br />
variables, which are easily plotted in two or three dimensions. Diagnostic information<br />
showing the contribution <strong>of</strong> various attributes/types and units to the summary variables is<br />
available (unlike cluster analyses <strong>of</strong> similarity matrices). As for graphical frequency<br />
seriation and Robinson seriation, if non-chronological variation is input (e.g. spatial and<br />
functional variation) the summary variables obtained may be strongly influenced by these.<br />
Control for spatial variation is possible as input files can include geographic coordinates,<br />
and diagnostic statistics make it clear to what extent variation across space is structuring<br />
the seriation.<br />
Multivariate ordination approaches such as principle components analysis (PCA),<br />
correspondence analysis(CA), principal coordinates analysis and multidimensional<br />
scaling(MDS) have percolated into Oceanian ceramic analysis over the last decade. PCA<br />
and CA are generally more easily evaluated than cluster analyses (Shennan 1997:253, 297-<br />
298). Ordination techniques in general give an indication <strong>of</strong> the relationships between<br />
ceramic variables and units <strong>of</strong> analysis (e.g. site assemblages). <strong>The</strong>y can suggest whether<br />
there are any summary trends in the data. PCA is generally more suited to numeric data,<br />
while CA is better for counted data, although PCA is sometimes used for this purpose.<br />
Seriation is nowadays almost always done using CA (Shennan, 1997 342), on counts <strong>of</strong><br />
types/attributes in the case <strong>of</strong> ceramic refuse, and on presence-absence <strong>of</strong> features in the<br />
case <strong>of</strong> grave-goods or architecture. CA as available in the Bonn Archaeological S<strong>of</strong>tware<br />
Package (BASP or WinBASP on the internet) can output comprehensive diagnostic<br />
statistics to identify the relative contributions <strong>of</strong> the various variables (including geographic<br />
coordinates <strong>of</strong> the samples) and units to the results.<br />
Principal coordinates analysis and MDS attempt to find the structure in a set <strong>of</strong><br />
proximity measures between objects. Results from these analyses are not as easy to<br />
evaluate as those from PCA or CA, as they operate on a similarity matrix rather than on<br />
raw data or data that has been transformed or standardized in some way (Shennan<br />
48
1997:348-349). A key part <strong>of</strong> the MDS method is that it has the advantage <strong>of</strong> providing<br />
an objective indication <strong>of</strong> the number <strong>of</strong> major summary trends that are significant, through<br />
the measure stress. One advantage <strong>of</strong> retaining the use <strong>of</strong> a similarity matrix is that a mix<br />
<strong>of</strong> metric and nominal variables can be used to construct the matrix, particularly useful if<br />
one has metric data on vessel form or the scale <strong>of</strong> decoration.<br />
Occurrence Seriation and Potsherd Samples:<br />
Occurrence seriation in its graphical form arranges samples so that the occurrence <strong>of</strong><br />
criteria is continuous or as near so as possible. <strong>The</strong> usual interpretation is that this<br />
represents an ordering <strong>of</strong> assemblages in time, assuming that any particular attribute was<br />
continuously discarded through time. To be valid this requires, like frequency seriation,<br />
that the attributes used in the seriation do not have an intermittent occurrence, and assumes<br />
control for contemporaneous variation across space and in relation to vessel/site function.<br />
Green’s seriation <strong>of</strong> Lapita sites (Green 1978) and Graves’ and Cachola-Abad’s<br />
seriation <strong>of</strong> architectural features in Hawaii (Graves & Cachola-Abad 1996) are <strong>of</strong> this<br />
type (O'Brien & Lyman 2000:120-121), but there are fundamental differences between the<br />
data used in these two examples which have serious consequences for ceramic seriation.<br />
A tradition <strong>of</strong> occurrence-based assemblage grouping approaches to ceramic seriation has<br />
been a persistent thread in Oceanic pottery analysis, usually using the Jaccard coefficient<br />
or simple-matching coefficient to create similarity matrices (Anson 1983, Best 1984,<br />
Egl<strong>of</strong>f 1971, Frost 1974, Green 1978, Irwin 1972, 1985, Summerhayes 2000a, Wickler<br />
1995, 2001). In the wider archaeological world occurrence seriation has been largely<br />
avoided by ceramicists since Dempsey and Baumh<strong>of</strong>fs initial proposal (O'Brien & Lyman<br />
2000:119), except as a technique for ordering grave-lots. <strong>The</strong> reasons for this avoidance<br />
are firstly that “chronologies based on the presence-absence <strong>of</strong> types or attributes will<br />
never be as accurate as those based on their relative frequencies (Le Blanc 1975).” Le<br />
49
Blanc noted the applicability <strong>of</strong> presence-absence seriation to grave-lots, where<br />
assemblages represent small intentional sets <strong>of</strong> attributes rather than time-accumulated<br />
discard from settlements. Secondly, and perhaps more importantly, the notion that<br />
occurrence seriation is less sensitive to variation in sample size than frequency seriation is<br />
questionable. Sparse data (a lot <strong>of</strong> types/attributes, each with few members) is where<br />
occurrence seriation shines, hence use with grave lots and architectural features. This<br />
assumes though that the sparse data is a complete representation <strong>of</strong> the unit to be<br />
seriated. This point will be more fully discussed below in relation to some archaeological<br />
seriations in Oceania.<br />
In Oceania, for the Lapita period, we have yet to discover a landscape dotted with<br />
generous numbers <strong>of</strong> Pompeii-like yet easily detectable and accessible ceramic vessel<br />
samples, and must be satisfied with a less stringent set <strong>of</strong> limits within which seriation can<br />
be considered to work. Regional site samples can be expected to be on the small side <strong>of</strong><br />
adequate, with assemblage vessel numbers lower than might be ideal, and durations <strong>of</strong><br />
occupation largely guessed at rather than tightly defined. Compounding these preservation<br />
and mixing problems is the problem <strong>of</strong> sparse data (discussed above), where in some<br />
classificatory schemes, a large number <strong>of</strong> types/attributes exist, each with few occurrences,<br />
in only a subset <strong>of</strong> sites. This sparseness may arise through splitting approaches to<br />
classification/grouping, in combination with the nature <strong>of</strong> ceramic production (lack <strong>of</strong><br />
standardization), and, also, as a result <strong>of</strong> the sampling processes acting to make the<br />
archaeologist’s sample a fraction <strong>of</strong> the pottery discarded at the location in the past in<br />
many cases.<br />
In response to the problem <strong>of</strong> sparse data, an approach commonly taken in<br />
Oceania has been to apply a form <strong>of</strong> occurrence seriation, by recording the<br />
presence/absence <strong>of</strong> attributes in units/sites, and by creating a matrix <strong>of</strong> similarities between<br />
units with a similarity coefficient suited to this type <strong>of</strong> data, and then performing either a<br />
computerized sort or a cluster analysis on the similarity matrix. <strong>The</strong> two similarity<br />
50
coefficients commonly used on this sort <strong>of</strong> data are the simple matching coefficient and the<br />
Jaccard coefficient. Of these, the simple matching coefficient regards negative matches as<br />
significant, and is thus less suited to sparse data, while the Jaccard coefficient disregards<br />
negative matches entirely, and is therefore better for this type <strong>of</strong> data (Shennan 1997:228).<br />
<strong>The</strong> widespread application <strong>of</strong> these techniques in Oceanian archaeology pays<br />
insufficient attention to the sample-richness problem for this type <strong>of</strong> sparse occurrence data<br />
from sherd samples (but see Anson 1987, Green 1978, Kirch 1987a). Whether or not a rare<br />
attribute occurs in an assemblage <strong>of</strong> sherds may well be a sampling effect, even for<br />
presence-absence data. Where sites/units vary greatly in sample size, spurious dissimilarity<br />
may result through size-related differences in sample richness.<br />
Consider a hypothetical case (Table 1), <strong>of</strong> two assemblages are drawn from<br />
identical populations, differing only in sample size. All types are represented in the large<br />
sample, but many are missing from the small. Calculation <strong>of</strong> a Jaccard coefficient <strong>of</strong><br />
similarity would result in the following data (where a=positive matches, b= present in the<br />
small assemblage but not the large, c=present in the large assemblage but not the small):<br />
a=3; b=7; c=0. Jaccard coefficient is calculated as a÷(a+b+c) or 3÷(3+7+0)=0.3, when<br />
ideally these being samples from the same thing should show high similarity (s<br />
approaching 1.0). In this example sample size is structuring the similarity coefficient to a<br />
greater degree than behavioural similarity. It can be seen that although the Jaccard<br />
coefficient is suited to sparse data, it assumes a complete representation <strong>of</strong> behaviour in<br />
the sparse data, and is not suited to assemblages which comprise incomplete samples <strong>of</strong><br />
a behavioural unit. Using a technique adapted to sparse data does not correct for<br />
inadequate sample <strong>of</strong> sites or inadequate (or unsaturated cf. Kintigh 1984) within-site /<br />
unit assemblage size (see also Anson 1983:54, Kirch 1987a, Specht 1977). This example<br />
illustrates a major potential problem with the application <strong>of</strong> the Jaccard coefficient to<br />
ocurrence seriation <strong>of</strong> Oceanic ceramic data, which would not arise with data from grave<br />
51
lots or architectural features.<br />
Phyletic Seriation:<br />
Phyletic seriation assumes heritable continuity, without specifying a rule <strong>of</strong> developmental<br />
direction, by emphasising evolutionary change <strong>of</strong> a chosen attribute over time. Seriation<br />
<strong>of</strong> Lapita face designs for example (Spriggs 1990) involves an assumption <strong>of</strong> the evolution<br />
<strong>of</strong> a single design over time, although Spriggs tended more towards evolutionary seriation<br />
in that instance as the direction <strong>of</strong> change was assumed to be from initial representation<br />
<strong>of</strong> faces to a later stylized geometric remnant. A more recent example <strong>of</strong> phyletic seriation<br />
<strong>of</strong> Lapita designs is more sophisticated, in that it constructs multi-lineal evolutionary series<br />
for three different aspects <strong>of</strong> Lapita design (Ishimura 2002). Phyletic seriation is about how<br />
change occurs, how one attribute becomes another, where frequency or occurrence<br />
seriation use arbitrarily defined attribute classes to freeze-frame change for measurement<br />
purposes. Mead also discussed phyletics <strong>of</strong> motif relationships by speculating that some<br />
motifs were developments <strong>of</strong> or elaborations <strong>of</strong> others (Mead 1975:37)<br />
Table 1: Sampling, assemblage richness, and the Jaccard coefficient (p=present, a=absent).<br />
Attribute 1 2 3 4 5 6 7 8 9 10<br />
Small<br />
sample<br />
Large<br />
Parent<br />
Popn.<br />
p a a p a a p a a a<br />
p p p p p p p p p p<br />
52
Seriation in Oceania:<br />
<strong>The</strong> best examples <strong>of</strong> local regional sequence building in Oceania have at their cores a large<br />
sample <strong>of</strong> collection sites, representing a good percentage <strong>of</strong> the total ceramic diversity<br />
within a small survey region. Additionally, they all use graphical attribute seriations and<br />
stratigraphic superposition to check the computer matrix results. Most, but not all, use<br />
ordered similarity matrices for their seriations, based on attribute frequencies and a<br />
Robinson coefficient <strong>of</strong> similarity.<br />
<strong>The</strong> Frost/Irwin numerical taxonomy approach was typified by a similarity grouping<br />
analytical mode. In analysing his Shortland data, for example, Irwin adopted Clarke’s<br />
proposition (Clarke 1968:512-549) that there is a functional convergence between<br />
occupation sites that imposes some control on functional variation, and that relative<br />
similarity between archaeological assemblages therefore indicates closeness in time (Irwin<br />
1972:83-85) (This assumes invariant site function and complete representation <strong>of</strong> the<br />
properties <strong>of</strong> a breakage population). Thus if a similarity matrix is ordered so that an ideal<br />
pattern emerges, the sites are arranged in order <strong>of</strong> chronological distance. Irwin noted the<br />
possibility <strong>of</strong> contemporaneous spatial variability creating such a pattern in the similarity<br />
matrix, but made no specific mention <strong>of</strong> functional variability, which, as vessel form data<br />
were included, could be seen as problematic. Irwin felt that because site surface collections<br />
were quite large and <strong>of</strong> “not especially long duration” (he had culled them <strong>of</strong> material<br />
thought to be intrusive from different periods prior to analysis) there was little likelihood<br />
<strong>of</strong> anything other than temporal variation emerging in his patterning.<br />
Irwin, citing Frost’s Fiji work, used the simple matching, Jaccard and Robinson<br />
coefficients <strong>of</strong> similarity to perform sorted similarity matrix seriations. <strong>The</strong> first two<br />
coefficients were used with occurrence data, while the Robinson coefficient is a<br />
frequency-based measure <strong>of</strong> similarity. Green (citing Irwin) suggested that occurrence<br />
seriation using the Jaccard coefficient was beneficial in reducing sample size differences<br />
(Green 1978), but this is incorrect: the technique is suitable for the analysis <strong>of</strong> sparse data<br />
53
(discussed in more detail below),which is not the same thing. <strong>The</strong> use <strong>of</strong> these two<br />
different types <strong>of</strong> seriation on the same samples has been widely followed in Pacific<br />
archaeology (Anson 1983, Best 1984, Green 1978, Wickler 2001). <strong>The</strong>se techniques<br />
commonly, and not unexpectedly, produce conflicting ordering <strong>of</strong> sites in the resulting<br />
similarity matrices, requiring a decision as to the “correct” ordering, and encouraging a<br />
variety <strong>of</strong> matrix sorting or clustering techniques, all <strong>of</strong> which can be decidedly divorced<br />
from the original archaeological data, and difficult to evaluate.<br />
Irwin’s study <strong>of</strong> a large number <strong>of</strong> late-prehistoric sites within the local Mailu<br />
region analyzed 28 contextual assemblages, some from excavation zones, some from<br />
surface collections, and constructed a sequence using sorted similarity matrices, graphical<br />
attribute frequency seriation and superposition chronological inference (Irwin 1985:118-<br />
162). Similarly, Egl<strong>of</strong>f made use <strong>of</strong> attribute-frequency sorted similarity matrices <strong>of</strong> a large<br />
number <strong>of</strong> surface sites and some mound excavations in the Collingwood Bay region, and<br />
also made use <strong>of</strong> graphical attribute seriations (Egl<strong>of</strong>f 1979:47, Figs 17-19). Egl<strong>of</strong>f’s<br />
sequence ran from the undated Lapita-like “Group P” ceramics to modern. Best applied<br />
Irwin’s method to construct a ceramic sequence for Lakeba in Fiji running from Lapita to<br />
modern. Both Best and Irwin controlled for sample size in their seriations by omitting<br />
smaller assemblages (under 100 sherds in Irwin’s case), but neither discuss quantification<br />
<strong>of</strong> sample size in any depth, or methods by which mixed or long-occupation assemblages<br />
can be identified, an area in which both Specht and Bedford have pointed out the value <strong>of</strong><br />
stylistically homogenous sampling units (Bedford 2000, Specht 1969). In Bedford’s case<br />
it is arguable that some <strong>of</strong> the lack <strong>of</strong> diversity is related to small sample sizes in the lower<br />
levels <strong>of</strong> rockshelter test pits on Malekula, while Specht was referring to larger samples<br />
surface-collected from abandoned village sites on Buka.<br />
Best, like Irwin, used attribute frequencies to calculate a Robinson matrix <strong>of</strong><br />
similarity, and cluster analysis to create a seriation <strong>of</strong> excavation and surface collection<br />
units (Best 2002:18-20). A particular feature worthy <strong>of</strong> note is the inclusion in his<br />
54
seriation diagram <strong>of</strong> a “dead” column tracking the percent decorated, which shows six<br />
peaks <strong>of</strong> decoration through his series, rather than a unimodal history, illustrating the<br />
dangers <strong>of</strong> using a recurring or fluctuating attribute for seriation. Best shows sixty units<br />
in his Robinson matrix, and thirty in his graphical seriation, with twenty-three attributes<br />
shown, illustrating the desirability <strong>of</strong> having a reasonably large matrix <strong>of</strong> types/attributes<br />
and units for seriation, to lessen the possibility <strong>of</strong> random repeats <strong>of</strong> stylistic attribute<br />
combinations over time. It is important to note that in Best’s case and in Irwin’s, that<br />
stratified sequences were available, which aided the identification <strong>of</strong> temporally sensitive<br />
attributes used in seriations.<br />
Selection <strong>of</strong> Variables in the Absence <strong>of</strong> Stratified Sequences:<br />
How best to structure classificatory systems to meet chronological aims is seldom<br />
discussed in archaeology generally. Some attributes/variables may fluctuate over time and<br />
space (the relative percentages <strong>of</strong> decorated vessels versus undecorated vessels for<br />
example), or may vary in relation to function (vessel form for example) which may in turn<br />
affect decoration (Shepard 1963:260-261). <strong>The</strong> danger in analyses that utilize a large<br />
number <strong>of</strong> variables/attributes in an a-theoretical manner is that the signal they are after<br />
(temporal variation for example) will be obscured by other types <strong>of</strong> variation, such as<br />
functional, spatial or sample-size variation. <strong>The</strong>se problems are lessened in CA and PCA,<br />
where it is relatively easy to discover if spatial or sample-size variables are contributing to<br />
the results, but the golden rule in seriation analysis according to Dunnell is careful selection<br />
<strong>of</strong> types or attributes that vary with time (Dunnell 1970). Some sorts <strong>of</strong> attributes carry<br />
higher risk than others, and can usefully be omitted from analysis when the aim is<br />
chronological. It may be possible as a result <strong>of</strong> such an omission to show, eventually, that<br />
such variables do vary over time, but the converse approach, including all variables, may<br />
result in a mishmash <strong>of</strong> variability that might be difficult to disentangle, particularly for<br />
similarity matrix / clustering approaches in which it may be difficult to ascertain which<br />
55
variables are principally responsible for the series obtained.<br />
Just what is it that one’s units <strong>of</strong> classification are measuring? This is one among<br />
a number <strong>of</strong> reasons why validation <strong>of</strong> cluster analyses <strong>of</strong> similarity scores is problematic.<br />
<strong>The</strong> Mead/Donovan and Siorat approaches have such theory, and seek ideas <strong>of</strong> the<br />
decorators regarding the structure (and sequence in the case <strong>of</strong> Siorat) <strong>of</strong> decoration<br />
(Mead 1975:37, Siorat 1990). In regard to the Mead/Donovan and Siorat approaches, if<br />
past ideas <strong>of</strong> structure is their objective, one must question whether there is likely to be a<br />
strong shift in such structures over time alone, rather than over space or in relation to<br />
functional variability. Clark also provides some discussion <strong>of</strong> the theoretical basis <strong>of</strong> his<br />
attribute selection (Clark 1999:65), and Wickler and Anson are both forthright on the<br />
difficulties <strong>of</strong> separating functional and temporal variability. Summerhayes advances a<br />
specific theory <strong>of</strong> ceramic temporal-functional variability (Summerhayes 2000c),<br />
concluding that function varied over time in the West New Britain data.<br />
<strong>The</strong> review <strong>of</strong> seriation method and theory above sets the stage for a critical review<br />
<strong>of</strong> constructions <strong>of</strong> Lapita temporal variability in Chapter 2, and for the analysis <strong>of</strong><br />
variability and seriation <strong>of</strong> Roviana data in this thesis. Frequency seriation, using attribute<br />
abundances and correspondence analysis, has the potential to yield a high-resolution<br />
chronology if samples are fit for the task and time-sensitive attributes are chosen.<br />
Correspondence analysis negates the argument that occurrence seriation is better for<br />
unsaturated samples, since correspondence analysis is weighted by site/attribute sample<br />
size. <strong>The</strong> quantification issue raises its head here though, as sample size must be assessed<br />
using a measure which is suited to the completeness characteristics <strong>of</strong> the sample.<br />
Occurrence seriation is unlikely to match the chronological resolution <strong>of</strong> frequency<br />
seriation for potsherd samples, being more suited to sparse data such as grave goods or<br />
architectural features, where the number <strong>of</strong> attributes <strong>of</strong> a sampling unit are large and the<br />
membership <strong>of</strong> each type/attribute is small, and where the sparse data is a complete<br />
representation <strong>of</strong> each unit such as a set <strong>of</strong> excavated grave goods, rather than a sample.<br />
56
I have argued above for the importance <strong>of</strong> sample evaluation as a first step in<br />
seriation: sample sizes for sites and units are one side <strong>of</strong> this issue (including the difficult<br />
problem <strong>of</strong> the number <strong>of</strong> independent observations that comprises the sample), but<br />
another difficult area is the assumption <strong>of</strong> heritable and historical continuity in the overall<br />
sample. It is important to try to identify sampling gaps or lack <strong>of</strong> heritable continuity<br />
between units or types. Phyletic analysis has a role to play here. Phyletic seriation is a<br />
mode <strong>of</strong> analysis that can usefully be combined with frequency seriation. When a series is<br />
constructed, phyletic analysis allows the investigation <strong>of</strong> the nature <strong>of</strong> the transformations<br />
that occur to make the series. How does one attribute transform into another? Phyletic<br />
analysis in turn leads on to “why” questions: if the nature <strong>of</strong> change can be understood, the<br />
explanation <strong>of</strong> change can be investigated. In Chapter 12 such a combination <strong>of</strong> frequency<br />
seriation (using CA) and phyletic seriation is attempted.<br />
This review has noted the importance <strong>of</strong> identifying the sorts <strong>of</strong> attributes in the<br />
data (whether functional or stylistic, within a Dunnellian theoretical framework- explained<br />
in the following section). <strong>The</strong> link between site/artefact properties and function is by no<br />
means simple, and further review <strong>of</strong> this topic is warranted.<br />
Form and Function: a Review :<br />
Introduction:<br />
Dunnell turned to evolutionary theory to explain why seriation worked, suggesting that<br />
style and function were fundamentally dichotomous (Dunnell 1978). In this view style is<br />
by definition that which is subject to evolutionary drift, while function is that which is<br />
controlled by selection (Dunnell 2001). Functional variation is thus a poor tool for<br />
seriation, as function may be stable or may fluctuate over time with the appearance <strong>of</strong><br />
analogous similarities (adaptations or evolutionary convergences), where style, by<br />
contrast, varies stochastically over time by definition. In Dunnell’s definition <strong>of</strong> style and<br />
57
function, form can have functional and/or stylistic components, as can decoration.<br />
Separating stylistic variation from functional variation is simple in Dunnell’s view: stylistic<br />
variation will seriate well (in battleship curves, in the traditional graphical form <strong>of</strong><br />
frequency seriation), while functional variation will not. Dunnell felt this explained why<br />
proponents <strong>of</strong> seriation in the culture-historical era had been able to create successful<br />
seriations through trial and error, trialling various classifications in a materialist manner<br />
until the desired battleship pattern (a seriation) was obtained, and successfully isolating<br />
chronological variation by this means (in conjunction with testing by other forms <strong>of</strong><br />
dating).<br />
Dunnell’s “method” (my term) assumes that sample sizes from sites, the number<br />
<strong>of</strong> sites in the sample, and the level <strong>of</strong> comparability <strong>of</strong> site occupation spans will be<br />
sufficient to create an even battleship pattern (with exclusion <strong>of</strong> sites that don’t fit due to<br />
small sample sizes or anomalous occupation span). As stated above, this is seldom the case<br />
in Oceania: neither can the Roviana data support this assumption, as will be seen in<br />
Chapter 12. Given a patchy archaeological sample, other methods <strong>of</strong> distinguishing<br />
functional variation from stylistic, prior to seriation, are needed.<br />
Techno-function refers to utilitarian aspects <strong>of</strong> an artefact’s use, as opposed to<br />
soci<strong>of</strong>unction or ide<strong>of</strong>unction (Skibo 1992:33-34). Skibo makes a further division <strong>of</strong><br />
function into intended function and actual function, with intended function being reflected<br />
in vessel design, and actual function reflected in use alteration. <strong>The</strong>ories <strong>of</strong> vessel<br />
form/function correlation (Rice 1987: 211, Smith 1985) thus <strong>of</strong>fer an opportunity to survey<br />
the total Roviana assemblage from the perspective <strong>of</strong> intended techn<strong>of</strong>unction. If intended<br />
vessel techn<strong>of</strong>unction can be attributed to ceramic forms, then function can be controlled<br />
for to some extent by constructing seriations only within functional classes. <strong>The</strong> situation<br />
is complicated in archaeology by the fragmentary state <strong>of</strong> recovered vessels, from which<br />
forms are inferred.<br />
<strong>The</strong> evolutionary concept <strong>of</strong> exaptation (Gould & Vbra 1982) could usefully be<br />
58
applied to the relationship between pottery form and function, allowing vessels <strong>of</strong><br />
multipurpose forms, and allowing uses to which the form is adaptable, but not adapted. It<br />
might be argued that there is no fixed vessel function, that artefact function is diverse and<br />
unpredictable from form and therefore determination <strong>of</strong> vessel function is an impossibility,<br />
but there are a number <strong>of</strong> factors working in the archaeologist’s favour that generate the<br />
expectation <strong>of</strong> a correlation at least between form and function. Firstly, a substantial<br />
ethnoarchaeological literature has appeared in the last twenty years on vessel use life, and<br />
suggests that frequency <strong>of</strong> use and type <strong>of</strong> use correlate strongly with use life (David 1972,<br />
De Boer 1983, Jensen et al. 1999, Longacre 1991, Mills 1989, Shott 1989b, 1996, Tani<br />
& Longacre 1999, Varien & Mills 1997, Varien & Potter 1997). <strong>The</strong> more frequently a<br />
vessel is used the more likely it is to be broken. Cooking and serving are a frequent<br />
activities, and representative settlement ceramic assemblages are likely to be dominated by<br />
utilitarian cooking and serving vessels (Varien & Mills 1997:147). Serving vessels, used<br />
frequently but not subjected to thermal stress are likely to have longer use life, and be less<br />
common in assemblages than would be the case if they were subjected to thermal stress,<br />
while storage and ritual vessels are likely to have even longer use lives, and be less<br />
common in assemblages, even if present in abundance during settlement occupation in the<br />
past.<br />
This formational bias towards cooking and serving vessels in assemblages (Varien<br />
& Mills 1997:148), is good news for seriation, as this bias creates for the archaeologist a<br />
measure <strong>of</strong> control for function, provided the sample obtained is representative <strong>of</strong> the<br />
settlement site as a whole (likely in the case <strong>of</strong> surface scatters exposed over a wide area,<br />
but more difficult to support in the event that one small excavation square is taken as<br />
representative <strong>of</strong> the whole, or if “site” function varies between sites and not all sites are<br />
“occupations”).<br />
This does not negate the need for ascribing function to sherds. One needs to know<br />
what sorts <strong>of</strong> vessel forms one has in the assemblage to be confident that such a bias does<br />
59
in fact exist in any particular sample, and further, it would be useful to be able to exclude<br />
some observations relating to non-cooking/serving vessels from consideration where these<br />
are identifiable.<br />
A use alteration perspective suggests that cooking pots are arguably more easily<br />
distinguished from other vessel classes in archaeological assemblages based on cracking,<br />
sooting and abrasion resulting directly from their different use environments (Skibo 1992).<br />
This is clearly not the case for the Roviana assemblages, where sooting was noted only on<br />
one vessel, (parts <strong>of</strong> which were found buried in mud), while for the majority <strong>of</strong> sherds the<br />
vessel surface was either clean or ablated, probably a consequence <strong>of</strong> the intertidal, surface<br />
context <strong>of</strong> deposition. For this reason vessel form seemed to be the best available<br />
information about vessel function. Nor has a use alteration perspective been applied to<br />
Lapita pottery in any <strong>of</strong> the studies reviewed in this <strong>chapter</strong>. Sooting, at least, in Oceania,<br />
seems hard to find.<br />
In the absence <strong>of</strong> use-alteration information, how might a classification <strong>of</strong><br />
sherd/vessel form according to vessel function be achieved? A number <strong>of</strong> measures <strong>of</strong> form<br />
properties <strong>of</strong> ceramic vessels relating to vessel function have been suggested or<br />
investigated by archaeologists, and a range <strong>of</strong> technical issues have surfaced in relation to<br />
the measurement <strong>of</strong> vessel form, examples <strong>of</strong> which are given below.<br />
<strong>The</strong> Culture History Approach <strong>of</strong> the 1940s and Onward:<br />
In southwestern American archaeology, <strong>of</strong>ten the distinction is merely between jars, for<br />
cooking and storage, and bowls, being serving and eating vessels (Skibo 1992:36), an<br />
unsatisfactory generalization without much theoretical/methodological rigour.<br />
“<strong>The</strong> notion that use strongly influences pot morphology is venerable, but<br />
few archaeologists or ethnographers have examined it in a uniformitarian<br />
theoretical framework (Smith 1985:257).“<br />
Rice reviews several “inferred use” classificatory schemes, including height-to-width<br />
vessel proportions, vessel contour classifications, and geometric/volume classifications<br />
60
(Rice 1987: 215-222). All <strong>of</strong> these schemes place a heavy burden on sherd size, which<br />
needs to be large to reconstruct forms, although Rice’s own approach, a vessel contour<br />
approach, is less restricted by this factor than either the vessel proportion approach or a<br />
geometric/volume classification.<br />
<strong>The</strong> Methodological Uniformitarians:<br />
As archaeologists we would like to be able to specify a deterministic link between form and<br />
function, but this is not possible. Rye for example states,<br />
“If it can be established that specific materials are usually correlated with<br />
specific functions, the presence <strong>of</strong> these materials can be used to infer<br />
function (Rye 1981)”<br />
Rye goes on to detail ceramic properties appropriate to different functions. He seems to<br />
be suggesting that he is providing a manual or method for the identification/determination<br />
<strong>of</strong> vessel function, and yet the information cannot be used in this way in the absence <strong>of</strong> a<br />
clear causal relationship. For example, the proposition that rounded forms improve thermal<br />
shock resistance is not matched by any statement that rounded forms indicate cooking as<br />
an intended function. <strong>The</strong>re are examples in the ethnoarchaeological literature, too, that<br />
suggest flat-bottom vessels are suited to cooking functions also (e.g De Boer 1983: 109,<br />
Rice 1987:224-226).<br />
A number <strong>of</strong> studies have attempted to develop methodological uniformitarian<br />
predictors <strong>of</strong> use based on ethnographic data (Henrickson 1990, Henrickson & Mcdonald<br />
1985, Smith 1983, 1985). <strong>The</strong>se provide cautionary documentation <strong>of</strong> a complex situation,<br />
where some classes <strong>of</strong> vessel use are relatively easily identified for whole vessels (e.g. dry<br />
storage: Smith) but where identification <strong>of</strong> other functions based on morphology is<br />
problematic, even at the level <strong>of</strong> whole vessels, and more so when dealing with broken<br />
sherd assemblages.<br />
Smith’s concern was solely with morphological indicators <strong>of</strong> use, in contrast to the<br />
61
more common technical and use-alteration approaches. Smith used Binford’s distinction<br />
between technomic, sociotechnic and ideotechnic aspects <strong>of</strong> culture (Binford 1962) to<br />
argue that a narrow focus on technomic aspects <strong>of</strong> ceramic function was possible without<br />
being diverted by this wider debate on function and style. Smith proposed no<br />
comprehensive theory <strong>of</strong> ceramic vessel form: it was noted that many factors besides use<br />
determine form. Smith allowed pots could have multiple uses, and specified also a<br />
distinction between morphological variables <strong>of</strong> whole vessels (e.g. vessel volume, height<br />
to width ratio, etc) and sherd variables, the latter being properties describing vessel size<br />
and shape at a single point, using an assumption <strong>of</strong> vessel symmetry.<br />
Smith’s sample <strong>of</strong> 39 ethnographic pots, for which eight ethnographic use classes<br />
were listed, was drawn from six ethnoarchaeological/ethnographic studies <strong>of</strong> the American<br />
southwest, which raises obvious sampling issues; her subsequent dissertation incorporated<br />
an enlarged and geographically/culturally more diverse sample (Smith 1983:116-140). <strong>The</strong><br />
small sample did not allow detailed correlations <strong>of</strong> different factors <strong>of</strong> use with specific<br />
morphological features. Her use classes were cooking; dry storage; wet storage;<br />
processing; serving <strong>of</strong> food; individual consumption; transport <strong>of</strong> liquid; and washing<br />
within the pot. Morphological data were reconstructed from illustrations.<br />
Smith digitized pot pr<strong>of</strong>iles with the aid <strong>of</strong> a shadow device, and encoded these as<br />
polynomial expressions <strong>of</strong> curves using the GLM regression procedure <strong>of</strong> SAS.<br />
Discriminant analysis was used to evaluate which variables were <strong>of</strong> use in defining<br />
function, and which sherd statistics (sherds statistics used experimentally were derived<br />
from the form data <strong>of</strong> the same vessel sample by a computer simulation <strong>of</strong> breakage).<br />
Predictive success was best when broader functional categories was used, which<br />
cut out instances <strong>of</strong> multiple-use (her short list was cooking, dry storage, wet storage or<br />
transport <strong>of</strong> liquids, and other). A clear distinction could be made between dry-storage<br />
vessels (low exposed surface per volume unit and small mouths, with large volumes<br />
characteristic to a lesser extent) and other vessels <strong>of</strong> all types. Consumption vessels<br />
62
(bowls) were well discriminated (large orifice relative to capacity, large orifice in absolute<br />
terms, and small volume). Processing, serving and washing vessels (other) had a<br />
moderately large ratio <strong>of</strong> exposed surface to capacity and moderately large orifices, with<br />
a slight tendency towards small volumes. Cookpots showed more variation than any other<br />
category, suggesting a category ill-defined or too broad, or that crucial morphological<br />
constraints were not monitored by the variables <strong>of</strong> the study. Cookpots overlapped<br />
significantly with liquid storage and liquid transport samples, with the liquid storage and<br />
transport vessels tending to have small ratios <strong>of</strong> orifice to volume, with these vessels not<br />
displaying the expected provisions for closure (Smith 1985: correlate 9 on p265). This last<br />
finding will be shown to be significant to the Roviana case in Chapter 8.<br />
Smith concluded that food storage and individual eating should be activities with<br />
reasonable archaeological visibility (without considering breakage rates). She noted the<br />
need to develop this method in the direction <strong>of</strong> archaeological practice, where some <strong>of</strong> the<br />
orientation variables are difficult or impossible to measure on broken sherds. Her efforts<br />
in this direction are promising, but preliminary, and are a long way from establishing a<br />
uniformitarian theory. Her results do suggest caution is required in the hasty assignment<br />
<strong>of</strong> globular jars and globular jars only to cooking functions and broader, shallower vessels<br />
to serving functions. Along with her finding that liquid storage/transport vessels had larger<br />
orifices than expected, this suggests that the cooking/serving/water jar distinction <strong>of</strong>ten<br />
seen in ethnographic terminology (e.g. Kirch/Green and Ross, discussed below) need not<br />
have a clear morphological correlation (see Rice 1987: figure 7.14 for a similar point in<br />
relation to cooking pot morphology).<br />
Henricksen and McDonald (Henrickson 1990, Henrickson & Mcdonald 1985)<br />
discriminated between function in a general sense (undefined) and primary function (also<br />
undefined). No distinction was drawn between intended function and actual function. <strong>The</strong><br />
approach to the discrimination <strong>of</strong> function was less systematic than Smith's study:<br />
discriminant analysis was not used. As in Smith's study, samples were small.<br />
63
Morphological correlates <strong>of</strong> function were found to vary according to factors not available<br />
to the archaeologist, for example, the mode <strong>of</strong> transport <strong>of</strong> water transport vessels had a<br />
pr<strong>of</strong>ound influence on form.<br />
Some <strong>of</strong> their statements verge on “ just-so stories” masquerading as explanation<br />
for morphology, for example cooking pots usually had: “A somewhat restricted mouth to<br />
prevent rapid evaporation from boiling foods (Henrickson & Mcdonald 1985:631)”. <strong>The</strong><br />
degree <strong>of</strong> mouth restriction seems unlikely to have any bearing on the rate <strong>of</strong> evaporation<br />
<strong>of</strong> the contents, as given a constant energy supply via the vessel body, evaporation can only<br />
be substantially reduced by pressurizing the contents, (as in a pressure-cooker), with orifice<br />
diameter potentially having only a slight effect on the condensation rate <strong>of</strong> vapour leaving<br />
the vessel. Restricted orifice is more likely to reduce spillage and to increase form strength<br />
<strong>of</strong> the vessel.<br />
<strong>The</strong> case studies presented in the article made heavy use <strong>of</strong> use-alteration to assign<br />
function; problematic if a distinction is drawn between intended and actual function.<br />
Measurement <strong>of</strong> Vessel Form:<br />
On the subject <strong>of</strong> vessel form measurement (orifice measurement and other curvature<br />
measurement on ceramic vessels), Plog noted a number <strong>of</strong> studies suggesting that<br />
measurement error was possibly a significant source <strong>of</strong> ceramic variability in some<br />
instances, and might obscure significant variation in others (Plog 1985). Plog suggested<br />
a dial-gauge technique for curvature measurement rather than curve-fitting using curvature<br />
templates, although he noted a number <strong>of</strong> problems with this, most notably the need to<br />
average a number <strong>of</strong> measurements along a curve to establish curve variability. Curvature<br />
measurements in the Roviana study were made using a custom-made set <strong>of</strong> brass curve<br />
templates, rather than the usual diameter chart (Chapter 4), which had some advantages,<br />
not least <strong>of</strong> which was using curve fitting to determine measurement points at pr<strong>of</strong>ile<br />
corner points such as necks, not easily done using Plog’s dial-gauge method.<br />
64
In the review <strong>of</strong> the Lapita ceramic series in Chapter 2, vessel form is an element<br />
<strong>of</strong> some <strong>of</strong> the chronological propositions critiqued there, and uncertainties in the<br />
measurement <strong>of</strong> vessel form are pointed out. Those criticisms are self-explanatory, and<br />
therefore methods <strong>of</strong> vessel form measurement will not be reviewed in detail here.<br />
Similarly, details <strong>of</strong> measurement methods used in analysis <strong>of</strong> the Roviana ceramics are<br />
given in detail in Chapter 4.<br />
Spatial Variation in Form:<br />
<strong>The</strong>re is some evidence from ethnoarchaeology that vessel form can vary spatially across<br />
quite short distances, with no functional difference observable. Ethnoarchaeology <strong>of</strong><br />
Kalinga pottery identified spatial variation in the form <strong>of</strong> pottery over about 10km within<br />
a single functional class (rice-cooking pot) (Longacre 1991:106-108). <strong>The</strong> variation is<br />
similar to the distinction between s<strong>of</strong>t-shouldered globular vessels and hard-shouldered<br />
globular vessels in the Roviana assemblages detailed in Chapters 4 and 8, so it seems that<br />
there can be substantial production variability for the same functional class within a<br />
smallish region. This brings to mind Smith’s caution that vessel form/function categories<br />
be kept broad.<br />
Oceanic Historical Anthropology and Vessel Function:<br />
Green’s substantive uniformitarian theory <strong>of</strong> Polynesian plainware vessel function, based<br />
on ethnographic observation <strong>of</strong> material culture <strong>of</strong> descendant cultures, suggests the<br />
smaller open post-Lapita bowls from Samoa were used for drinking, serving fluid foods,<br />
kava serving, and holding dyes for decorating bark cloth (Green 1974a:129). Medium-<br />
sized bowls could have been used for preparation <strong>of</strong> barkcloth materials, cooking, and food<br />
preparation. Medium and large bowls were equated with Kava receptacles.<br />
Linguistic terms for small cups and boiling have been reconstructed for Proto-<br />
Polynesian (Kirch & Green 2001:136). Similarly, Ross used lexical reconstruction <strong>of</strong><br />
65
Proto-Oceanic terms to infer a Proto Oceanic taxonomy <strong>of</strong> four functional classes <strong>of</strong> vessel<br />
(cooking pots, water jars, bowls and frying “pans”, but noted that identification <strong>of</strong> the<br />
forms <strong>of</strong> these was a task for archaeologists rather than linguists (Ross 1996). <strong>The</strong><br />
simplicity <strong>of</strong> this scheme conflicts in some respects with expectations for middle-level<br />
societies around the world. <strong>The</strong> various storage, processing, ritual and transport functions<br />
observed ethnographically in middle-level societies in various parts <strong>of</strong> the world<br />
(Henrickson & Mcdonald 1985, Smith 1985) suggest Lapita pottery would have had a<br />
more complex range <strong>of</strong> functional classes <strong>of</strong> vessel.<br />
Ross’s tentative assignment <strong>of</strong> Western Lapita form to function has biconical<br />
rounded-base vessels as cooking vessels, with water jars having a narrower orifice<br />
altogether (Ross 1996:figure 2). <strong>The</strong>se are similar to Summerhayes type V vessels<br />
(Summerhayes 2000a:34), Sigatoka type E (Mead et al. 1975:15, figure 1.5) and<br />
Niuatoputapu vessel form 7 (Kirch 1988a:162 and figure 101d, 101e). This is at odds with<br />
Smith's uniformitarian findings, where there was substantial overlap between water jars<br />
and cookpots, suggesting multipurpose forms were possible in this category. Perhaps the<br />
solution here is to regard small-orifice flasks as liquid transport/storage vessels, and larger<br />
biconical forms as potentially multipurpose forms, which could easily fulfil either a<br />
cooking or liquid-transport function.<br />
It has been suggested that highly decorated Lapita ware might be a high-status<br />
exchange item (Kirch 1988b:161, 1997:148), and that the amount <strong>of</strong> decoration in<br />
assemblages might reflect contemporaneous inter-community social relations (Kirch<br />
1997:147, 143-144). <strong>The</strong>re is something <strong>of</strong> a conflict between this model <strong>of</strong> decorated<br />
Lapita as exchange items and another proposed by Kirch <strong>of</strong> decorated Lapita as a<br />
depiction <strong>of</strong> an ancestor <strong>of</strong> an Austronesian house (Kirch 1997: 188-191). While both<br />
models are consistent with the notion <strong>of</strong> pots being tattooed for display/presentation, pots<br />
as ancestors and pots as exchange items seem mutually exclusive, as house-ancestor<br />
evokes in my mind manufacture by a locally-resident member <strong>of</strong> the descent-group, rather<br />
66
than an exchange item (but this is a subjective interpretation) and these could be seen as<br />
competing models, or models specific to particular vessel types. An ethnoarchaeology <strong>of</strong><br />
the manufacture and exchange <strong>of</strong> material representations <strong>of</strong> Austronesian house-ancestors<br />
might clarify this point. At present my objection is not founded on any good data.<br />
Neither <strong>of</strong> these models <strong>of</strong> the function <strong>of</strong> Lapita pottery have been discussed in<br />
relation to formation theory regarding vessel function and breakage rates. If decorated<br />
Lapita is high-status exchange ware with an occasional serving function, breakage rates<br />
could be expected to be low, whereas if the decorated bowls and dishes were in frequent<br />
use breakage rates would be higher, and frequency in assemblages would reflect this.<br />
Similarly, if pots symbolising house-ancestors were curated in a quiet part <strong>of</strong> the house,<br />
breakage rates could be expected to be extremely low, unless some specific regular<br />
breakage ritual is invoked, and such vessels could be expected to occur only rarely in<br />
assemblages. Use life and discard rated thus provide a test <strong>of</strong> functional interpretations <strong>of</strong><br />
decorated Lapita. High percentages <strong>of</strong> decorated Lapita pottery in assemblages imply a<br />
high relative breakage rate, or perhaps a historically-particular mechanism <strong>of</strong> deposition,<br />
such as abandonment, followed by subsequent breakage <strong>of</strong> a curated behavioural<br />
assemblage <strong>of</strong> pottery (a life-assemblage) (Brenchley & Harper 1998, De Boer 1983). If<br />
we are to model vessel function for Oceanic pottery, we cannot ignore the link between<br />
vessel function, size or frequency <strong>of</strong> use and breakage-rates.<br />
Summerhayes summarized widespread evidence for the loss <strong>of</strong> complex Lapita<br />
forms with dentate decoration, and relatively unchanging persistence <strong>of</strong> simple<br />
undecorated forms in many areas, seeking explanation for these decoration changes in his<br />
Arawe Islands ceramic sequence (Summerhayes 2000c:293-303). Summerhayes saw the<br />
slow rate <strong>of</strong> change <strong>of</strong> the plainwares as related to an “ongoing domestic/utilitarian role”,<br />
with rapid change over time in the level <strong>of</strong> representation <strong>of</strong> dentate-ware in assemblages,<br />
and the shape, motif type and production <strong>of</strong> dentate pottery. I critique Summerhayes'<br />
chronological conclusions in detail in Chapter 2 as being a too-direct reading <strong>of</strong><br />
67
chronology from superposition, with insufficient attention to cultural or natural formation<br />
processes.<br />
<strong>The</strong> high labour-input into dentate decoration was seen by Summerhayes as<br />
suggestive <strong>of</strong> a high-value or prestige “goods”, with the loss <strong>of</strong> this characteristic over time<br />
“equating” to a lessening <strong>of</strong> the social importance <strong>of</strong> these pots. Summerhayes endorsed<br />
Rathje’s contention that the dentate motifs were Austronesian social/ideological signifiers<br />
(Rathje 2000), extending this by suggesting these functioned in an environment <strong>of</strong><br />
continuing interaction between widespread communities, and that loss <strong>of</strong> these social<br />
signifiers indicates changes in the nature <strong>of</strong> interaction between communities. This<br />
explanation is similar to Kirch’s prestige goods explanation, and unlike Kirch’s “house-<br />
ancestor” model, in that the “house-ancestor” pattern is based on modern ethnographic<br />
observation and persists until the present across most <strong>of</strong> the Austronesian world, making<br />
it a poor candidate for a vanished system. Like Kirch, Summerhayes paid no attention to<br />
the link between function, breakage/discard rates, and assemblage composition.<br />
<strong>The</strong>ories <strong>of</strong> the Functional Correlates <strong>of</strong> Rim Form Variation:<br />
Rim form, is a property <strong>of</strong> vessels relatively easily identified from sherds, and worth<br />
considering in more detail from a functional perspective. Rim form incorporates the<br />
attributes <strong>of</strong> the neck (in the case <strong>of</strong> restricted vessels with everted rims), rim and lip,<br />
including the height <strong>of</strong> rim, degree <strong>of</strong> eversion/inversion, rim and neck pr<strong>of</strong>ile curvature,<br />
changes in thickness across the pr<strong>of</strong>ile and neck orifice size. When tested<br />
ethnoarchaeologically, observable rim form variability may bear only a loose relationship<br />
to use (as suggested by Miller 1985:51-74). Nevertheless, if one wishes to avoid forcing<br />
contemporaneous functional rim form variation into a spurious time series, consideration<br />
<strong>of</strong> the possible meanings <strong>of</strong> form variance and some consequent control for rim form<br />
variance in seriation is desirable.<br />
Everted rims can be regarded as multi-functional components <strong>of</strong> vessels. <strong>The</strong>y may<br />
68
add form strength, reduce spillage (Henrickson & Mcdonald 1985:634), allow a tied cover<br />
for storage or fermentation (Henrickson & Mcdonald 1985:632, Smith 1985:263), aid<br />
pouring (Smith 1985:263) and filling, aid stacking (Miller 1985:63, 70, Skibo 1992:77) and<br />
aid lifting when hot by use <strong>of</strong> a torque or tongs (Skibo 1992:66, Smith 1985:263).<br />
Reduction <strong>of</strong> spillage might be best achieved by having a tall vertical or tall inverted rim,<br />
but filling, pouring, stacking and lifting might encourage a more everted form. Heavily<br />
everted and very short rims are probably not the best design for filling, lifting, or<br />
preventing spillage, but may improve functionality for tying covers, pouring and access for<br />
ladling or stirring, and may improve form strength, extending use-life.<br />
In some methods <strong>of</strong> pot manufacture, rim form may be constrained by the plastic<br />
limits <strong>of</strong> the clay-temper mix. Slab-built rims or coil-built rims make little demand on the<br />
plasticity <strong>of</strong> clay during construction. A tall thick, heavily everted rim can be formed from<br />
an arc-shaped slab or by coil building. In contrast, if the entire vessel including the rim is<br />
formed from a single lump <strong>of</strong> clay by the paddle-and-anvil method, there are potential<br />
difficulties in forming a tall and strongly everted rim, as the taller and more everted the rim<br />
the more the clay must be thinned towards the lip, relative to the neck, potentially resulting<br />
in an overly convergent, and thus weak, rim pr<strong>of</strong>ile (thick at the neck, thin at the rim). If<br />
a heavily everted rim is desired for some reason from a pot built by the one-piece method,<br />
the rim may have to be short rather than tall to prevent excessive thinning towards the lip,<br />
resulting in a rolled rim form in the extreme case. If a tall rim is desired in a one-piece pot<br />
(to prevent spillage for example) it may not be practical to have a heavily everted form, as<br />
the lip might split during manufacture or become overly fragile with reduced thickness.<br />
Form and Function: Conclusions:<br />
<strong>The</strong> detailed identification <strong>of</strong> intended use <strong>of</strong> ceramic vessels from archaeological<br />
contexts, relying solely on vessel morphology, is not currently well supported by any<br />
69
established body <strong>of</strong> theory. <strong>The</strong>re are indications from methodological uniformitarian<br />
studies that some broad categories <strong>of</strong> intended use might be identifiable from sherd<br />
morphology in the future, but there are also strong indications from these studies that form<br />
is determined by other contextual factors in addition to intended use, including pottery-<br />
making tradition. While these rare uniformitarian studies are widely cited as providing a<br />
theory <strong>of</strong> the morphological correlates <strong>of</strong> function (for example Sinopoli 1991:84) few<br />
would claim that these correlates are deterministic laws. Some <strong>of</strong> the morphological<br />
correlates <strong>of</strong> function suggested by earlier studies (Ericson et al. 1972, Linton 1944) have<br />
not been borne out by uniformitarian research other than in the most general sense.<br />
This does not, however, negate the virtue <strong>of</strong> controlling for vessel form variation<br />
in seriations, using broad form-categories. While form may vary systematically over time<br />
as a result <strong>of</strong> change in function, in which case a form-seriation may be justified, this<br />
requires chronology to be established by other, independent means, particularly from<br />
Dunnell’s position, that seriation <strong>of</strong> functional variation is a contradiction in terms. Where<br />
the chronology is not known, and seriation is to be the primary means <strong>of</strong> constructing<br />
chronology, control for form is essential, to reduce the possibility <strong>of</strong> incorrectly seriating<br />
contemporaneous or fluctuating vessel functional variation.<br />
Chapter summary and conclusions:<br />
<strong>The</strong> Roviana intertidal archaeological record, investigated as a means <strong>of</strong> assessing whether<br />
Lapita was continuously distributed across Near Oceania or whether there was a huge gap<br />
in the Near-Oceanic Solomon Islands, requires archaeological theory and method<br />
developed for or adapted to the nature and location in the landscape <strong>of</strong> the material under<br />
study. Understanding temporal variability, always a basic task in the archaeology <strong>of</strong><br />
poorly-understood regions, is complicated by the pottery being in the sea. How can this<br />
record be interpreted in behavioural terms? How can a high-resolution chronology be<br />
constructed? How much <strong>of</strong> a sample is needed? How should sample size be quantified?<br />
70
What are the biases present in the samples? Pottery recovered as surface scatters from the<br />
intertidal and reef flat raises a series <strong>of</strong> questions regarding cultural and natural/post-<br />
deposition formation processes, the investigation <strong>of</strong> which can contribute both to<br />
understanding variability in the samples, and to higher-level questions about sample<br />
preservation and visibility, and the implications for the writing <strong>of</strong> Oceanic prehistory.<br />
What is the state <strong>of</strong> preservation <strong>of</strong> the samples? In what taphonomic<br />
circumstances? A summary <strong>of</strong> the physical geography/geology <strong>of</strong> the research region noted<br />
two unusual properties <strong>of</strong> the Roviana landscape, which relate to the preservation <strong>of</strong><br />
surface sites in the sea:<br />
• the Holocene marine transgression has approximately kept pace with tectonic<br />
uplift, as Plio-Pleistocene uplifted barrier reefs are a few metres above sea level<br />
across most <strong>of</strong> the survey region, which is inferred to mean that sea levels have<br />
been relatively stable over recent millennia, falling at least1.5 metres from a high-<br />
stand circa 4500bp, exposing ceramic deposits that were formerly subtidal to<br />
littoral-zone processes, creating lag deposits <strong>of</strong> ceramics, favouring detection by<br />
archaeologists.<br />
• the chain <strong>of</strong> upraised barrier reefs that encircle New Georgia and form a series <strong>of</strong><br />
lagoons, <strong>of</strong> which Roviana is one, shelters the coastline within the lagoons from all<br />
ocean waves, favouring the preservation <strong>of</strong> littoral-zone ceramics, also a factor in<br />
the detection <strong>of</strong> these sites by archaeologists.<br />
<strong>The</strong> extent <strong>of</strong> these effects is investigated in detail in Chapter 5.<br />
Sample acquisition, through the systematic survey and sample collection <strong>of</strong> a<br />
region, is the basis <strong>of</strong> all further analysis. How should the survey region or regions be<br />
chosen, and what methods should be used to survey it/them? How extensive, and how<br />
intensive, should the survey be? How much information is enough for the questions to be<br />
asked <strong>of</strong> the material to be satisfactorarily answered? In Chapter 2 these questions are<br />
investigated, with examples from surveys in Near Oceania reviewed. <strong>The</strong> methods and<br />
71
esults <strong>of</strong> the Roviana surveys are discussed, and some conclusions are drawn regarding<br />
research design.<br />
A review <strong>of</strong> approaches to ceramic classification identified a need for a materialist<br />
approach, where the units <strong>of</strong> classification used derive from an analysis <strong>of</strong> variability <strong>of</strong> the<br />
samples in hand, in a dialectic with the research questions. Temporally sensitive attributes<br />
are needed if a fine-grained chronology is to be constructed, and the identification <strong>of</strong> these<br />
requires a careful analysis <strong>of</strong> the sorts <strong>of</strong> dimensions <strong>of</strong> variability in the data, whether<br />
functional, stylistic, geographic, sample-size-related, or to do with the brokenness or<br />
completeness <strong>of</strong> the sherd samples from which pottery characteristics are inferred. Units<br />
<strong>of</strong> classification that are not sensitive to differences in vessel brokenness are needed, and<br />
units <strong>of</strong> decorative classification need to be identifiable by location on the vessel, to capture<br />
the structure <strong>of</strong> decoration across the vessel, even in broken sherd assemblages. With these<br />
requirements in mind it was concluded that latitudinal bands <strong>of</strong> decoration should be<br />
central in the classification, as these were identifiable by location even for very broken<br />
assemblages, due to location near to the rugged and well-preserved corner points (lips,<br />
necks, carinations) <strong>of</strong> vessels, which also had the advantage <strong>of</strong> boosting sample size.<br />
Chapters 4 and 9 cover this in detail.<br />
Evaluation <strong>of</strong> the recovered ceramic samples is an important step in the study <strong>of</strong><br />
formation processes. Estimating the brokenness and completeness <strong>of</strong> our samples allows<br />
choice <strong>of</strong> appropriate units <strong>of</strong> quantification for evaluating sample size. Comparison <strong>of</strong> the<br />
sample with an inferred breakage population is also possible in many situations, especially<br />
where samples derive from surface sites, or large area excavations. <strong>The</strong> relationship<br />
between the breakage population and the extant deposit can delineate the potential for<br />
taphonomic bias in recovered samples, and provides a method <strong>of</strong> responding to Gosden’s<br />
call for controlled comparisons. Analysis <strong>of</strong> brokenness and completeness is reported in<br />
Chapter 5.<br />
A comparative study <strong>of</strong> wave exposure <strong>of</strong> the Roviana ceramic sites is needed to<br />
72
e able to assess the relative contributions <strong>of</strong> natural and cultural formation processes. In<br />
Chapter 6 the oceanographic setting <strong>of</strong> the sites is described, and an analysis <strong>of</strong> relative<br />
wave exposure is undertaken. <strong>The</strong>se findings form the backdrop to the study <strong>of</strong> other<br />
formation processes. A range <strong>of</strong> formation models are evaluated in Chapter 7 with<br />
reference to the properties <strong>of</strong> sherd assemblages. Spatial analysis as an aid to the<br />
identification <strong>of</strong> formation processes is deferred to Chapter 11, as it is necessary to<br />
formalize units <strong>of</strong> classification <strong>of</strong> materials prior to undertaking spatial analysis.<br />
An extended discussion <strong>of</strong> seriation theory and the need to control seriations for<br />
vessel form variation was given above. In Chapter 8 the details <strong>of</strong> an analysis <strong>of</strong> form<br />
variability are presented, preparatory to seriation analysis. It was also pointed out above<br />
that units <strong>of</strong> classification need to be adapted to the place and period under study, and also<br />
to the range <strong>of</strong> variation present in the sample. For the Roviana materials this means that<br />
the units <strong>of</strong> classification used need to be designed to capture with sensitivity the<br />
transition from a style <strong>of</strong> pottery that is recognizably Lapita to something that is not<br />
Lapita. An analysis <strong>of</strong> variability is needed to translate the initial units <strong>of</strong> description into<br />
units <strong>of</strong> classification appropriate to these aims. Furthermore, units <strong>of</strong> classification need<br />
to be insensitive to differences in vessel brokenness if they are to be useful for<br />
comparisons. Finally, it is beneficial to the analysis <strong>of</strong> relatively plain pottery, in terms <strong>of</strong><br />
the sensitivity to change, if units <strong>of</strong> classification make use <strong>of</strong> the information to be had<br />
from the structure <strong>of</strong> decoration across the vessel. For all <strong>of</strong> these reasons a detailed<br />
exploration <strong>of</strong> ceramic variability is desirable, reported in Chapter 9.<br />
<strong>The</strong> collection sites included lithic items alongside ceramics. While there has been<br />
little discussion <strong>of</strong> these in this introduction, everything written above regarding treating<br />
ceramic sherds as sedimentary particles applies equally to lithic manuports and artefacts.<br />
Analysis <strong>of</strong> lithics in Chapter 10 is made largely from the point <strong>of</strong> view <strong>of</strong><br />
geoarchaeological sourcing <strong>of</strong> raw materials, although a discussion <strong>of</strong> adze forms is<br />
included also. <strong>The</strong> sourcing analysis, which seeks to identify procurement units, is the<br />
73
asis <strong>of</strong> the lithic component <strong>of</strong> analysis <strong>of</strong> intrasite spatial structure in Chapter 11. Spatial<br />
analysis yields information bearing on the identification <strong>of</strong> cultural and natural formation<br />
processes. In Chapter 11 spatial units <strong>of</strong> classification are re-evaluated prior to seriation<br />
analysis. One collection site (Zangana) is split into two spatial units for seriation as a result<br />
<strong>of</strong> the spatial analysis.<br />
<strong>The</strong> extended discussion <strong>of</strong> seriation given above is translated into practical analysis<br />
<strong>of</strong> Roviana ceramic data in Chapter 12. Carbon 14 data and thermoluminescence data are<br />
integrated with seriation data to produce a provisional chronology in need <strong>of</strong> further<br />
independent corroboration and fleshed-out sampling.<br />
Information emerging from these various analytical <strong>chapter</strong>s is pulled together in<br />
Chapter 13. Here the focus returns to the significance <strong>of</strong> the findings regarding Lapita<br />
detectability in Near Oceania. Despite the difficulties <strong>of</strong> constructing a reliable chronology<br />
with the sample in hand, there is also some discussion comparing the Roviana data with<br />
ceramic data from other areas. <strong>The</strong>re is a distinct possibility that some ceramic variants<br />
mixed in with Lapita in other places are strongly separated spatially from Lapita in the<br />
Roviana lagoon early ceramic sample. If this is so, then the approach taken to constructing<br />
chronology, across space rather than by the bedded strata method, is amplifying temporal<br />
resolution beyond what is obtainable from the test-pitting approach <strong>of</strong> the bedded-strata<br />
school.<br />
Ceramic petrographic sourcing has not been discussed in this introduction, and will<br />
not be discussed in any detail in this thesis. Results <strong>of</strong> this sort <strong>of</strong> analysis were interesting<br />
though, but for reasons <strong>of</strong> brevity discussion <strong>of</strong> the implications <strong>of</strong> work either published<br />
elsewhere or appended (appended see Dickinson 2000a, Felgate & Dickinson 2001) is<br />
confined to a short section in Chapter 13, integrating the analysis <strong>of</strong> ceramic tempers with<br />
petrographic analysis <strong>of</strong> lithics.<br />
Before embarking on these analyses <strong>of</strong> variability <strong>of</strong> the Roviana materials though,<br />
it is necessary to review existing literature pertaining to Lapita and the archaeology <strong>of</strong><br />
74
Near Oceania, in the light <strong>of</strong> the methodological review above, to provide ceramic-<br />
chronological context for the Roviana study. <strong>The</strong> particular foci <strong>of</strong> this review in the<br />
following <strong>chapter</strong> will be the temporal resolution <strong>of</strong> existing constructs <strong>of</strong> Lapita temporal<br />
variability, and the level <strong>of</strong> confidence with which these can be accepted as tested<br />
syntheses.<br />
75
CHAPTER 2:<br />
A REVIEW OF CONSTRUCTIONS OF LAPITA<br />
Introduction:<br />
TEMPORAL VARIABILITY<br />
<strong>The</strong> Lapita Ceramic Series (Golson 1971, Green 1978, 1979, 1990, see also Sand 2000)<br />
while not a term in universal use, embodies a concept <strong>of</strong> Lapita ceramics as a time-<br />
transgressive changing system <strong>of</strong> ceramic production/discard comprising regionally varied<br />
sequences within the overall region in which Lapita has been found. Summerhayes<br />
recently suggests a universal series, where there is geographic variation, but where in<br />
some respects ceramics change in parallel through time across the Lapita distribution,<br />
indicating continuing interaction (Summerhayes 2000a, b, 2001, 2002). In this review I<br />
suggest that the security with which a Lapita ceramic series has been defined has been<br />
over-stated in most cases examined, and that formulations <strong>of</strong> a Lapita ceramic series are<br />
mostly working hypotheses rather than confirmed histories <strong>of</strong> ceramic change. This means<br />
that higher-level inferences regarding social processes, such as colonization rates (e.g.<br />
Anderson 2002) are not founded on secure high-resolution chronologies. Getting the<br />
historical facts straight is vital if explanations for change are to be stable and sensible<br />
(Spriggs 2001).<br />
Golson (1971) noted technological continuities between Lapita and post-Lapita<br />
plainware, and in view <strong>of</strong> this evidence for homologous similarity, and also in regard <strong>of</strong><br />
the increasingly varied corpus <strong>of</strong> radiocarbon dates accumulating from Lapita sites,<br />
extended Gifford’s conception <strong>of</strong> Lapita as a widespread pottery time-horizon (Gifford &<br />
Shutler 1956:93-95) to include the notion <strong>of</strong> temporal variability by use <strong>of</strong> the term<br />
tradition, a situation for which Golson suggested the term series as appropriate (Golson<br />
1971:75) (although Gifford did not use the term horizon, he clearly regarded the Watom,<br />
77
Isle De Pins, Site 13 and Sigatoka materials as roughly <strong>of</strong> equal age). Golson’s<br />
understanding <strong>of</strong> “series” was that: “<strong>The</strong> series is made up <strong>of</strong> a set <strong>of</strong> ceramic styles which<br />
are similar and contiguous in either space or time or both (Golson 1971:75)”. Golson’s<br />
“Lapitoid” ceramic series was a hypothetical rather than an “established” detailed<br />
construct, as “...the individual styles that it includes have yet to be isolated, described and<br />
named (Golson 1971:75).”<br />
Green constructed a series for plainware in Samoa, running from circa 800BC<br />
Lapita to Polynesian plainwares by 500BC, and saw parallels in the sequences <strong>of</strong> Fiji and<br />
Tonga (Green 1974b:253), but objected to Golson’s use <strong>of</strong> the term Lapitoid, in spite <strong>of</strong><br />
the evidence for continuity with Lapita, in view <strong>of</strong> the absence <strong>of</strong> definitive (in Green's<br />
definition <strong>of</strong> Lapita) Lapita design motifs on these plain vessels (Green 1974b: 250-251).<br />
He preferred at that time to think <strong>of</strong> Lapita sensu strictu as a style-horizon, characterized<br />
by a wide range <strong>of</strong> elaborate vessel forms, divisible at most into a series comprising Early<br />
Western and Early and Late Eastern Lapita styles, on the basis <strong>of</strong> differences in vessel<br />
shape and the style and frequency <strong>of</strong> decoration (Green 1974b:251). Green saw the demise<br />
<strong>of</strong> Lapita, and then <strong>of</strong> Polynesian plainware, as a functional change, possibly simply<br />
culinary, but also possibly indicating fundamental social changes consequent on settlement<br />
<strong>of</strong> a previously unoccupied island world (Green 1974b:253). Green defined the Lapita<br />
Ceramic Series as<br />
“...assemblages in which various shouldered pots, jars and bowls, as well<br />
as flat-bottomed dishes and plates, occur in association with widely varying<br />
percentages <strong>of</strong> dentate-stamped, notched and incised decoration. <strong>The</strong><br />
decoration consists <strong>of</strong> a catalogue <strong>of</strong> elements and motifs whose<br />
combinations can be listed and compared. In addition, there is a range <strong>of</strong><br />
infrequently decorated bowls <strong>of</strong> simple shapes and varying sizes, plus<br />
several forms <strong>of</strong> rather more frequently decorated sub-globular pots. Site<br />
assemblages <strong>of</strong> Early Western and Early and Late Eastern styles can be<br />
recognized within the series by differences in vessel shape and by the style<br />
and frequency <strong>of</strong> the decoration.” (Green 1974b:251)<br />
78
Green’s 1978 Lapita Ceramic Series:<br />
Green summarized the existing Mead-system analyses as indicating a division into<br />
Western and Eastern Lapita, which both shared a core <strong>of</strong> widespread “early” motifs. <strong>The</strong><br />
Western division was characterized as having a diversity <strong>of</strong> complex motifs, while the<br />
Eastern Lapita division had a relatively poor inventory <strong>of</strong> comparatively simple motifs.<br />
Green defined these motif sets more specifically in his subsequent analysis <strong>of</strong><br />
Lapita motif occurrence (Green 1979), departing substantially from Golson’s and<br />
Specht’s sequence <strong>of</strong> “styles” approach, taking instead an attribute/computer clustering<br />
approach to group similar assemblages in an ordered similarity matrix, very much in the<br />
Irwin mold (Green 1990:37), using Mead-system classificatory units. Although his<br />
primary objective in highlighting differences between Eastern and Western Lapita in this<br />
way was to test his hypothesis that the Fiji/Vanuatu water gap caused an early break in<br />
a unified Lapita interaction system, his paper included a number <strong>of</strong> chronological<br />
statements concerning agreement between his C14 chronology and decorative similarity,<br />
which effectively constitute an occurrence seriation.<br />
At the time Green conceived <strong>of</strong> the sites in his analysis (the Reef-Santa Cruz sites<br />
at least) as slices in time over up to 1000 years <strong>of</strong> Lapita production/discard, and had<br />
pooled C14 ages by site to reflect this view (Green 1976a). Occupation span was not the<br />
focus <strong>of</strong> any extended discussion. <strong>The</strong>se sites are some <strong>of</strong> the best-sampled and dated<br />
Lapita sites on record, and therefore provide an illustration <strong>of</strong> the need for a general shift<br />
in emphasis from the age <strong>of</strong> the “site” to the age <strong>of</strong> the artefact in seeking to construct<br />
high-resolution chronology. This can be done by focusing on the Reef-Santa-Cruz case<br />
where the sampling difficulties <strong>of</strong> a reconnaissance test-pitting approach have been<br />
avoided by systematic surface collection and large area excavations.<br />
Green regarded the closeness <strong>of</strong> sites SZ-8 and Vatcha in the matrix (Green<br />
1978:figure 7) as confirmation <strong>of</strong> the early date <strong>of</strong> both sites, as suggested previously by<br />
the particular archaeologists concerned with those sites (Green/Donovan for SZ-8 and<br />
Frimigacci for Vatcha). He also saw this analysis as reconfirming the close relationship<br />
between RL-2 (RF2) and Watom.<br />
Green presented one further matrix, an analysis <strong>of</strong> the three Reef-Santa Cruz<br />
assemblages (Green 1978:figure 8). He re-stated his 1976 position, that the relative ages<br />
79
Table 2: Reef/ Santa Cruz motif counts as given in Anson 1983.<br />
Site Total Motif count Motif Richness<br />
RF2 841 examples 178 Motifs<br />
RF6 252 examples 79 Motifs<br />
SZ8 627 examples 133 motifs<br />
<strong>of</strong> the sites were known, and saw a high degree <strong>of</strong> motif sharing between these sites as<br />
indicative <strong>of</strong> a strong continuity over an extended period among those motifs restricted<br />
to the Reef-Santa-Cruz area. He saw a different temporal trend in the Reef-Santa Cruz<br />
sites to that noted by Birks and Shaw for the Eastern Lapita sequences, suggesting<br />
instead a trend <strong>of</strong> initial motif efflorescence followed by impoverishment through time,<br />
as evidenced by the motif-poor RF6 sample. Green had applied a correction for sample-<br />
size-related richness using Donovan’s unpublished motif frequencies to assess whether<br />
a motif was absent as a result <strong>of</strong> sampling error or absent-not-present. This assessment<br />
is challenged below, on the basis that allowance for the effect <strong>of</strong> sample-size differences<br />
was probably insufficient.<br />
Despite subsequent revision <strong>of</strong> the C14 chronology (Kirch & Hunt 1988) which<br />
had the three sites <strong>of</strong> indistinguishable age, Green considered that the sequence <strong>of</strong> sites<br />
as outlined by in his (1978) seriation could still be supported on the basis <strong>of</strong> Donovan’s<br />
chronological inferences from the ceramics (Green 1991c). Best has recently taken issue<br />
with this view, suggesting that the diversity <strong>of</strong> motifs in sites is sample-area related (Best<br />
2002:91). Best argued that the chronology <strong>of</strong> the sites should be reversed. Here I present<br />
an alternative view, similar to Best’s in that the “temporal” similarity relations are<br />
suggested to have more to do with sample size than initially thought. I look to Donovan’s<br />
and Parker’s theses for additional information on sample sizes, assemblage brokenness,<br />
and relative frequencies <strong>of</strong> decorative techniques (Table 2, Table 3, Table 4).<br />
Donovan’s hypothesis was that Reef-Santa Cruz Lapita had a decorative<br />
80
Table 3: Sherd counts and MNI assembled from various tables in Parker (1981).<br />
Site Dentate<br />
rims<br />
“richness” that suggested a greater developmental time-depth than had previously been<br />
considered (Donovan 1973:IV). This was consistent with the then C14 chronology, which<br />
had the sites as slices in time over a period <strong>of</strong> many centuries. Assemblages from three<br />
sites were analyzed, RL2 (RF2), thought at that time to date to approximately 900BC,<br />
RL6 (RF6), approximately 600BC and SZ8, undated at that time.<br />
All rims were removed from these assemblages prior to analysis for special study,<br />
and are not included in Donovan’s counts (Donovan 1973:5). <strong>The</strong> rims were later<br />
analyzed by Parker, providing a useful independent quantification <strong>of</strong> relative sample sizes<br />
(Table 3) (Parker 1981). Counts (sherd counts?) <strong>of</strong> decorative techniques were given by<br />
layer by Donovan, but these data cannot be read as sample sizes in the absence <strong>of</strong> any<br />
information on vessel fragmentation, completeness, part representation, or sherd sizes.<br />
<strong>The</strong> relative frequencies <strong>of</strong> decorative techniques were calculated with counts <strong>of</strong> plain<br />
sherds included, which tends to muddle the picture, as plain vs decorated may easily<br />
fluctuate over time rather than follow a steady trajectory, and is sensitive to differences<br />
in brokenness. <strong>The</strong> percentage <strong>of</strong> dentate to incised sherds by site/layer was recalculated<br />
with counts <strong>of</strong> plain sherds excluded (Table 4).<br />
<strong>The</strong>se data do not support any change in relative frequency <strong>of</strong> dentate-stamping<br />
to incised decoration by level in any <strong>of</strong> the three sites (which is not surprising as the sites<br />
had been subject to post-depositional gardening), but more significantly, differences<br />
between<br />
Incised<br />
Rims<br />
81<br />
Impressed<br />
rims<br />
Plain<br />
rims<br />
Total Rim<br />
Sherd<br />
count/<br />
MNI<br />
RF6 43 1 15 2 61/55<br />
SZ8 111 8 10 14 143/?<br />
RF2 848 50 425 100 1423/666
Table 4: Relative proportions <strong>of</strong> dentate and incised, recalculated from data presented by<br />
Donovan (1973).<br />
Dentate<br />
Count<br />
sites are slight. (As presented by Donovan, there were substantial differences in the<br />
relative frequency <strong>of</strong> techniques by site and by layer, but as the recalculation shows, these<br />
have more to do with differences in the percentage <strong>of</strong> plain sherds, which can mean<br />
differences in brokenness between contexts, than with differences in potting behaviour<br />
in the past.)<br />
Donovan thought that marked decorative similarities between the sites were due<br />
to common origin within the same tradition, and strong regionalism (Donovan 1973: 36,<br />
43). <strong>The</strong> possibility that the sites were similar because they were all <strong>of</strong> similar or<br />
overlapping age and that differences arose for reasons other than chronology, was not<br />
considered, because the chronology was thought to be largely known from the C14<br />
evidence.<br />
Dentate % Incised<br />
count<br />
Donovan’s motif counts as given by Anson (Anson counted multiple motifs on a<br />
82<br />
Incised % Total Count<br />
(ds + inc)<br />
Context<br />
40 74 14 26 54 RL2 surface<br />
1647 64 926 36 2573 RL2LayerA<br />
758 66 384 34 1142 RL2LayerB<br />
586 75 199 25 785 RL6LayerA<br />
240 79 65 21 305 RL6LayerB<br />
416 74 143 26 559 SZ8 Surface<br />
789 70 344 30 1133 SZ8levelA<br />
643 66 328 34 971 SZ8levelB<br />
206 67 99 33 305 SZ8levelC<br />
147 67 73 33 220 SZ8levelD<br />
25 71 10 29 35 SZ8levelE
single sherd separately) are the source <strong>of</strong> the data in Table 2 (Anson 1983:175):<br />
<strong>The</strong> low motif count at RF6 is most economically explained as sample-size related,<br />
particularly if the units <strong>of</strong> measurement <strong>of</strong> sample size are reexamined. Decorative<br />
technique abundances are given in Table 3, extracted from various figures in Parker’s<br />
thesis.<br />
Differences in sampling method by site and within sites (Green 1976a: 251-255)<br />
are significant, as Best has pointed out: the RF2 site is the most intensively surface-<br />
collected and Green excavated the largest area here. <strong>The</strong> recovered sample from RF2 is<br />
thus a more complete representation <strong>of</strong> ceramic variability at the site. But even without<br />
making any reference to these data, the ceramic quantities reported by Parker hint at the<br />
same thing. Parker’s rim quantities throw sample sizes for RF6 into stark contrast with<br />
the other two sites. Using sherd count, RF2 has around 12 times as many rims as RF6,<br />
23 times as many using MNI, and 21 times as many counting only those sherds large<br />
enough to measure mouth diameter (providing an unintended size filter). In view <strong>of</strong> this<br />
MNI information on relative sample sizes it seems unlikely that Green’s Jaccard similarity<br />
matrix <strong>of</strong> the three sites is being structured by behavioural variation, and highly likely that<br />
motif presence/absence is strongly structured by sample size. Green’s selection <strong>of</strong> motifs<br />
that were common in some sites to include in the Jaccard similarity matrix is made on the<br />
basis <strong>of</strong> Donovan’s sherd count quantification <strong>of</strong> motif frequency, and is probably an<br />
inadequate compensation for sample-size-related differences in sample richness.<br />
Parker, for vessel forms, like Donovan for motifs, remarks on the similarity <strong>of</strong> the<br />
three sites, also concluding this is “most probably indicating that all belong to a closely<br />
related tradition” (Parker 1981:76). She notes the appearance, though, <strong>of</strong> vessel type 14<br />
(heavily grooved walls) on as many as ten separate vessels at SZ8, absent from other<br />
sites. (<strong>The</strong>se are assumed to be Donovan’s “large, rung-like projections”).<br />
Thus Parker's 1981 conclusion:<br />
83
“It seems that over a period <strong>of</strong> more than a thousand years (1600BC to<br />
400BC) there is in the Western Lapita area little change in the shape <strong>of</strong><br />
pottery vessels, and that we are looking at a very long-lasting and stable<br />
tradition (Parker 1981:115).”<br />
stems more from the earlier slice-in-time view <strong>of</strong> the C14 chronology and the consequent<br />
interpretation <strong>of</strong> Green’s (1978) Jaccard occurrence seriation, than from temporal<br />
behavioural variation in ceramic manufacturing/discard style. An alternative explanation,<br />
where the sites are much closer to each other in age, and potentially contemporaneous or<br />
with overlapping occupation/discard periods, seems a more cautious interpretation <strong>of</strong> the<br />
data, that is, treating Reef-Santa Cruz Lapita as horizon rather than series. This<br />
assessment largely ignores the C14 evidence, which has always tended to suggest that<br />
those RF6 materials which have been dated are younger than dated materials from<br />
SZ8/RF2, and the results <strong>of</strong> current dating research are awaited with interest.<br />
<strong>The</strong> number <strong>of</strong> sites in the Reef-Santa Cruz sample, and their relative occupation<br />
spans are also crucial factors in seriation. Seriation works best when there is a large corpus<br />
<strong>of</strong> sites, and all sites included in the seriation have approximately equal occupation span.<br />
Traditionally this was achieved by rejecting units comprising mixtures from widely different<br />
or excessively long periods, as evidenced by diversity <strong>of</strong> styles. While temporal variation<br />
undoubtedly exists in these three Reef-Santa-Cruz ceramic samples, detailed explication<br />
<strong>of</strong> this variation has yet to be achieved. Best’s call for a re-examination <strong>of</strong> the ceramics<br />
(Best 2002:93) cannot realistically be expected to support a reversal <strong>of</strong> site chronology<br />
as it is unlikely to be possible to construct a robust seriation sequence for Reef-Santa-Cruz<br />
Lapita from a sample <strong>of</strong> only three sites from the Lapita period, four if the Mdailu site is<br />
included (Mccoy & Cleghorn 1988). <strong>The</strong> chances <strong>of</strong> all three being <strong>of</strong> similar duration are<br />
slim, there is a sample-size problem with RF6 at least, and three is simply too few site-<br />
samples to yield a robust seriation. While there are now a number <strong>of</strong> other sites <strong>of</strong> post-<br />
Lapita age recorded in the Reef-Santa Cruz region, these have yet to be seriated, and do<br />
not hold the potential to give a fine-grained chronology <strong>of</strong> Lapita the way<br />
84
a larger sample <strong>of</strong> Lapita sites would. <strong>The</strong> nature <strong>of</strong> this critique needs to be borne in mind<br />
in relation to the Roviana materials from which a seriation is constructed in this thesis: it<br />
suffers from similar problems, with less radiocarbon evidence to test the conclusions.<br />
Difficulties in assessing the relative occupation spans <strong>of</strong> sites are compounded by<br />
differences in the representativeness <strong>of</strong> the ceramic samples between sites and within sites.<br />
Comparison <strong>of</strong> RF2 and RF6 illustrates this. Even the RF2 sample, which must rate as the<br />
most comprehensive and detailed surface collection and area excavation <strong>of</strong> a Lapita site<br />
undertaken to date, is more complete from the southern half <strong>of</strong> the site than the north, due<br />
to more intensive surface collection strategy, a greater degree <strong>of</strong> vertical disturbance to the<br />
south, and concentration <strong>of</strong> excavation in the south. Green felt that the northern area <strong>of</strong><br />
RF 2 was <strong>of</strong> shorter duration than the southern, with deposition <strong>of</strong> ceramics beginning<br />
earlier in the southern area, as evidence by more complex stratigraphy (Green 1976a :255).<br />
This raises a question whether C14 dates from the north <strong>of</strong> RF2, had they been available,<br />
would be similar in age to those from RF6? Certainly some <strong>of</strong> the pottery in the RF2<br />
sample is very similar to some <strong>of</strong> the RF6 pottery, so a simple comparison <strong>of</strong> C14 data<br />
from the RF2 excavations and from RF6 excavations may mislead regarding the stability<br />
<strong>of</strong> the ceramic tradition over time, within the period <strong>of</strong> Lapita production.<br />
Anson’s Early Far Western Lapita:<br />
Anson’s splitting non-classificatory descriptive approach to motif description (Green 1990)<br />
resulted in much sparser data than the Mead/Donovan all<strong>of</strong>orm/motif classification,<br />
potentially creating even greater sampling-related problems for those <strong>of</strong> his analyses that<br />
were motif-occurrence-based. Anson identified nearly 500 different “motifs” in the<br />
analyzed sherd samples, most <strong>of</strong> which occurred only in low frequency, or in a single<br />
instance in many cases. Anson gave “sample sizes” (as motif relative abundances, not as<br />
counts as suggested in Anson 1987) for his sixteen analyzed site assemblages in a lengthy<br />
85
motif appendix, with total motif occurrences (sherd counts?) tabulated by sample in a<br />
subsequent paper (Anson 1986:Table 1). While the overall sherd count <strong>of</strong> motif<br />
occurrences is large for some sites, where sherds-as-vessels analysis has not been<br />
performed we have no way <strong>of</strong> knowing whether these are independent observations. <strong>The</strong><br />
sherd counts and motif occurrences listed by Anson are not necessarily sample sizes in a<br />
behavioural or discard sense.<br />
Using a Robinson coefficient <strong>of</strong> similarity to construct a similarity matrix from<br />
motif relative frequency data for twelve sites from across the Lapita distribution, Anson<br />
found that motifs from Ambitle, Eloaua (Mussau) and Talasea formed a group separate to<br />
the western Lapita sites, in which grouping the Reef-Santa Cruz sites, the Watom sites and<br />
the New Caledonian sites formed separate, and looser sub-groups (Anson 1983:180). On<br />
the basis <strong>of</strong> this grouping and the observations that a similar dichotomy could be seen in<br />
the size and spacing <strong>of</strong> dentate-stamp “teeth”, Anson tentatively defined an “Early Far<br />
Western Lapita”, questioning as he did so whether sample error or other (e.g. functional)<br />
variation could provide a non-temporal explanation for the pattern.<br />
In providing this caveat, Anson sowed the seeds <strong>of</strong> the current critique. <strong>The</strong> spatial<br />
extent <strong>of</strong> the sampling procedure, or the spatial representativeness <strong>of</strong> his samples, is the<br />
first issue. <strong>The</strong> size <strong>of</strong> samples, in terms <strong>of</strong> the number <strong>of</strong> independent behavioural<br />
observations represented, is the second. <strong>The</strong> type <strong>of</strong> behavioural variation that caused<br />
Ambitle, Eloaua and Talasea to cluster separately from the other sites (chrono/stylistic,<br />
functional, geographic) is a third uncertainty.<br />
To compare data from sites in Vanuatu and Fiji for which presence-absence <strong>of</strong><br />
motifs was the only information available, Anson used the MultBET hierarchical cluster<br />
analysis program (he does not give details <strong>of</strong> the similarity coefficient used in this<br />
programme, but it seems likely, like Jaccard’s, to be sensitive to sample size differences in<br />
sparse data, whatever coefficient is used). Kirch has suggested that samples must show<br />
86
diversity saturation (Kintigh 1984) to be comparable (Kirch 1987a:124), but this only<br />
applies to comparing the presence or abundance <strong>of</strong> rare types, while Anson argues the<br />
MultBET clustering results are explicable in terms <strong>of</strong> common types (Anson 1987). This<br />
is supported by similar groupings using the Robinson coefficient <strong>of</strong> similarity with relative<br />
abundance data (Anson 1983:180), and by his sample-size-corrected manually-calculated<br />
motif sharing percentage data (Anson 1987).<br />
As in my critique <strong>of</strong> Green’s 1978 analysis above, the basis on which Anson<br />
discriminated rare motifs from common motifs is questionable, given the impossibility <strong>of</strong><br />
evaluating motif sample sizes in any behavioural sense from sherd counts alone. Similarly,<br />
for the Robinson coefficient results, we cannot know, on the information supplied by<br />
Anson, to what extent the relative motif frequencies as quantified by sherd count are<br />
representative <strong>of</strong> ceramic variability <strong>of</strong> the breakage populations from which they derive.<br />
<strong>The</strong> effects <strong>of</strong> variable occupation span may also be fundamental to the patterning<br />
observed.<br />
<strong>The</strong> implications <strong>of</strong> the spatial/geographic component <strong>of</strong> variation was never fully<br />
discussed. With the exception <strong>of</strong> the three Watom samples, which grouped with the<br />
Western/Eastern Lapita sites, particularly the New Caledonian sites, Anson’s grouping <strong>of</strong><br />
sites largely correlated with geography (Anson 1986:figure 5). <strong>The</strong> similarity between<br />
Watom and western Lapita may be temporal, i.e. Watom and Western/Eastern Lapita is<br />
later than the “far Western” sites; or phyletic ( i.e. either Watom was the origin locality <strong>of</strong><br />
eastwards expansion into the Solomons and Remote Oceania; or Watom represents a back-<br />
migration from the East); or may be reticulate, resulting from some form <strong>of</strong> more<br />
continuous far-west-West-East interaction involving Watom (although the dissimilarity <strong>of</strong><br />
the nearby Ambitle samples to Watom argues against this). Anson in the end plumped for<br />
a temporal explanation, at that time regarding this as a hypothesis in need <strong>of</strong> further<br />
testing.<br />
87
Mussau:<br />
<strong>The</strong> Mussau sites, characterized by Gosden as yielding one <strong>of</strong> the better Lapita ceramic<br />
data-sets <strong>of</strong> the Lapita Homeland Project (Gosden 1991a), cannot be regarded as providing<br />
secure and detailed evidence <strong>of</strong> ceramic change over time until the details <strong>of</strong> the ceramic<br />
analysis are published. Some general review is attempted here, based mostly on preliminary<br />
analyses <strong>of</strong> ceramics (Kirch 1987b, 1988c, 2001). <strong>The</strong> Mussau Lapita Homelands Project<br />
research under Kirch’s leadership located what are claimed to be bedded Lapita deposits<br />
at site ECA area B, documenting a declining frequency <strong>of</strong> dentate-stamping and an<br />
increasing frequency <strong>of</strong> incision with superposition in a stratigraphic/superposition series,<br />
with rim notching reaching a peak in the earlier units (Kirch et al. 1991: Figure 3). <strong>The</strong>re<br />
has recently been a substantial amendment to this preliminary sequence based on C14<br />
evidence, with an early plainware site now understood to predate the dentate Lapita (Kirch<br />
2001:206, 219).<br />
It would be interesting to know whether or not there were sherds from the same<br />
vessels in different zones <strong>of</strong> ECA area B (Summerhayes provides useful information <strong>of</strong><br />
this sort in his Arawe analyses). Zone B1 and B2 were described as highly disturbed by<br />
sand-crab burrowing that was stated to have transported substantial amounts <strong>of</strong> pottery to<br />
the surface (Kirch 2001:86,92), although an argument was made that this had not affected<br />
the waterlogged portion <strong>of</strong> the deposits. Kirch does not theorize the nature and effects <strong>of</strong><br />
bioturbation and swash turbation when the environment <strong>of</strong> deposition was shallow water<br />
in the past. A substantial level <strong>of</strong> burrowing and turbation by marine arthropods,<br />
mammals, flora and fish might be expected to have occurred (Ferrari & Adams 1990);<br />
also in some weather conditions wave processes may have affected the pottery while in<br />
the sea. Regarding the superposition sequence from the upper zones <strong>of</strong> the square,<br />
Gosden’s admonition that sequences need to built from comparable site types (Gosden<br />
1991a) has clearly not been extended to include comparable depositional contexts within<br />
88
an excavation square.<br />
An initial claim for “stratified” evidence from Weisler’s excavation at the EKQ<br />
rockshelter on Elaoua has since been revised. While Kirch now regards the lower<br />
component as “including a late-Lapita component dominated by incised ceramics”(Kirch<br />
2001:214), initial interpretation <strong>of</strong> the ceramic deposit was that it “provided the best<br />
stratified sequence <strong>of</strong> ceramic materials from any Mussau site” (Kirch et al. 1991:151).<br />
Summerhayes, in synthesising the Bismarcks Lapita chronology, incorporated the earlier<br />
view, with a claim for a transition from dentate to incision over time in this deposit<br />
(Summerhayes 2002:27). <strong>The</strong> numerically large sherd sample had an average sherd weight<br />
<strong>of</strong> 1.54 g (Kirch et al. 1991, Weisler 2001), and assessment <strong>of</strong> sample significance is thus<br />
problematic, if even possible. Weisler quantified decorative changes by level using summed<br />
weight by decorative technique, but the total count <strong>of</strong> 26 dentate sherds looks to have been<br />
spread largely at random though the 30 ceramic-bearing levels <strong>of</strong> units one and two (data<br />
for which are illustrated in Weisler 2001:159), and motif analysis has not yet been<br />
published.<br />
<strong>The</strong> “Changing” Face <strong>of</strong> Lapita:<br />
In a departure from either Anson’s Motif inventory approach or Mead’s structural<br />
approach to sherd decoration, Spriggs, reminiscent <strong>of</strong> Specht’s (1968) call for a focus on<br />
whole design, advocated a shift <strong>of</strong> focus to Lapita vessel design in total rather than<br />
quantification <strong>of</strong> decorative motifs represented on sherds (Spriggs 1990). This brings to<br />
mind Shepard’s caution that the brokenness <strong>of</strong> assemblages should not be allowed to<br />
determine units <strong>of</strong> analysis, and that complex pictorial designs are not suited to<br />
element/motif analysis. Spriggs noted that this approach has been severely limited by<br />
generally highly fragmented samples <strong>of</strong> Lapita pottery, but suggested that the Mead<br />
system might not be the most appropriate method <strong>of</strong> analysis for such complex designs,<br />
89
in which he echoed Donovan’s expression <strong>of</strong> the difficulties <strong>of</strong> coding a description <strong>of</strong><br />
complex curvilinear Lapita designs (Donovan 1973:Vol 2: 64, 130-34).<br />
Spriggs suggested that the curvilinear ‘face’ designs formed a coherent<br />
evolutionary chronological series from complex to simple, and that this could be used to<br />
date sites. Spriggs did not discuss the implication that where such a 'series’ was present in<br />
widely-separated sites these would have to have a sufficient level <strong>of</strong> interaction and<br />
convergence in ceramic design (not the same thing as interaction) to have caused parallel<br />
changes in different regions. His analysis is thus founded on an assumed universal Lapita<br />
series rather than derived from detailed regional sequences <strong>of</strong> ceramic change. Some <strong>of</strong> the<br />
principal questions regarding the Lapita ceramic series are to do with whether ceramic<br />
change was synchronous or not across broader regions (Spriggs 2000:355, Summerhayes<br />
2000a:235).<br />
While Mead considered that his structural approach established a phyletic similarity<br />
between all Lapita decoration (as did Green), this is an overstatement for some <strong>of</strong> the<br />
simpler motifs, which could usefully be omitted from analysis as potentially being<br />
analogous similarity rather than homologous. An advantage <strong>of</strong> Spriggs’ approach,<br />
concentrating on the more complex aspects <strong>of</strong> the designs, not stated by Spriggs, is that<br />
potentially analogous similarity <strong>of</strong> some simple designs, widespread throughout the world,<br />
is excluded.<br />
Spriggs’ phyletic approach has recently been elaborated and revised (Ishimura<br />
2002), who, like Spriggs, adopts a view that the overall trend in Lapita decoration, “...the<br />
dynamics <strong>of</strong> chronological change in the design structure....” is one <strong>of</strong> simplification over<br />
time. Mussau is cited as one <strong>of</strong> the regional sequences thought to support this doctrine <strong>of</strong><br />
simplification, which is not wholly congruent with more recent publication <strong>of</strong> Mussau<br />
results which have an early plainware predating Lapita (Kirch 2001:206, 219). While<br />
Ishimura makes no reference to sample-size differences in discussing the presence or<br />
90
absence <strong>of</strong> design variants in site assemblages, his analysis <strong>of</strong> the Santa Cruz sequence,<br />
for example, the occurrence <strong>of</strong> a late variant <strong>of</strong> his “spade” design only in the small RF6<br />
sample but not in the larger RF2 or SZ8 samples is noteworthy. This supports an<br />
interpretation <strong>of</strong> RF6 as including a later date in its occupation span than RF2, unless<br />
absence from RF2 is sampling error.<br />
While Ishimura feels his typological analysis supports a scenario <strong>of</strong> gradual<br />
dispersal, it could also be argued that it supports a scenario <strong>of</strong> rapid dispersal, as some<br />
early variants (Type 3 spade designs for example)are found from Watom to Fiji, showing<br />
no gradual cline, and Type 2 spade designs are only found in the Bismarcks sites, bringing<br />
us back to the uncertainty faced by Anson in trying to interpret essentially the same<br />
dichotomy: is the Type1-Type 2-Type3 evolution <strong>of</strong> spade designs really temporal, or is<br />
this simply geographic variation? Are the Bismarcks designs different because they are<br />
geographically separate or because they are earlier? <strong>The</strong> radiocarbon evidence suggests<br />
the latter, but Type 10 faces occur at Eloaua, RF2 and Site 13 on New Caledonia,<br />
supporting a relatively rapid spread <strong>of</strong> Lapita from West to South within the occupation<br />
span <strong>of</strong> Eloaua, rather than the gradual spread Ishimura suggests. <strong>The</strong> similarity <strong>of</strong> Watom<br />
to New Caledonian Lapita brings to mind processes <strong>of</strong> reticulaton rather than gradual West<br />
to East spread, where potentially Watom is stylistically anomalous in terms <strong>of</strong> its face<br />
designs as a result <strong>of</strong> its geographic location on the stepping-stone route to and from<br />
Remote Oceania. It is not difficult to entertain the possibility <strong>of</strong> a group <strong>of</strong> migrants from<br />
further East becoming established at Watom, for example.<br />
Ishimura demonstrates that a phyletic approach to design seriation is valuable, in<br />
that ideas about the relative occupation spans <strong>of</strong> sites begin to emerge from the lacuna in<br />
which they have languished through the period <strong>of</strong> attribute-based similarity grouping<br />
seriation and heavy reliance on radiocarbon as a primary chronological tool rather than a<br />
confirmatory technique. Ishimura has done us a great favour by bringing focus back to<br />
91
occupation span as a crucial aspect <strong>of</strong> chronology, breaking free from the slice-in-time<br />
view <strong>of</strong> sites so prevalent through the period <strong>of</strong> the New Archaeology. If his phyletic<br />
classification does turn out to be valid, it poses a problem for some <strong>of</strong> the slice-in-time<br />
assumptions, for example a short occupation at RF2 (Sheppard & Green 1991), a site<br />
which along with Eloaua appears on the basis <strong>of</strong> Ishimura’s construct to have a longer<br />
occupation than any other Lapita site from the perspective <strong>of</strong> face designs, while in terms<br />
<strong>of</strong> spade designs it is exceeded in occupation span only by Watom.<br />
While the sampling strategy employed by Green at RF2 is clearly more<br />
comprehensive than the other Lapita samples in Ishimura’s study, which might account for<br />
the richness <strong>of</strong> the RF2 sample, Ishimura’s organization <strong>of</strong> this RF2 material into a phyletic<br />
series spanning Type 3 to Type 14 “face designs” and Type 3 to Type 12 “spade designs”<br />
seems to invoke an alternate explanation, at odds with a slice-in-time view <strong>of</strong> RF2 as a<br />
short occupation.<br />
Whether Ishimura’s Phase 1 (1500-1200BC) is really three centuries in duration,<br />
or whether it is a geographic variant or a slightly earlier variant (perhaps beginning fifty<br />
years earlier?) <strong>of</strong> the more widespread Phase 2 Lapita is not yet known with any certainty.<br />
Ishimura sees his results as supporting Summerhayes’ Early, Middle and Late Lapita, but<br />
one is left wondering how this phyletic analysis gels with Kirch’s recent swing to an early<br />
plainware predating the elaborate designs at Mussau. If his phyletic evolutionary rule from<br />
complex to simple is valid, perhaps the early plainware at Mussau is a functional variant<br />
<strong>of</strong> an even more elaborate set <strong>of</strong> Asiatic designs as yet undiscovered?<br />
A problem with Ishimura’s scheme is that the C14 chronology against which the<br />
phyletic series is tested is <strong>of</strong> greatly different temporal resolution than this phyletic<br />
seriation. <strong>The</strong> C14 data is mostly site-based, while the phyletic data is vessel-based. If ages<br />
<strong>of</strong> “Types” rather than sites in Ishimura’s scheme were available, the prospects for testing<br />
competing theories <strong>of</strong> variability in the ceramic data would be improved.<br />
92
Summerhayes, West New Britain, Anir, and a Three-stage Lapita Ceramic Series:<br />
Summerhayes regarded his west New Britain data as support for Anson’s Early Far-<br />
Western Lapita hypothesis. Like Anson, he saw a temporal dichotomy in the Bismarck<br />
Archipelago data, and extended this to the wider Lapita distribution. Unlike Anson, he<br />
closely identified this dichotomy with form/functional variability within ceramic<br />
assemblages, suggesting that the difference between early and late sites was the loss <strong>of</strong><br />
certain special-function forms, on which particular dentate motifs were found.<br />
His construction <strong>of</strong> a three-phase series for all <strong>of</strong> the Lapita distribution requires<br />
that the New Britain superposition evidence (on which the suggestion that Anson’s theory<br />
was “confirmed” was based) be reviewed in detail. Summerhayes regarded some West<br />
New Britain sites as having lengthy ceramic superposition chronological sequences which<br />
allowed examination <strong>of</strong> the changing nature <strong>of</strong> ceramic production (Summerhayes 2000a:<br />
3-4). He wished to know whether the West New Britain sequence paralleled changes<br />
elsewhere in the Pacific (Summerhayes 2000a:4). Based on his analysis <strong>of</strong> the ceramic<br />
results <strong>of</strong> several West-New Britain excavation conducted during the Lapita Homeland<br />
Project, followed by wider comparisons, he concluded that the stylistic province models<br />
<strong>of</strong> Green (Green 1978, 1979) and Kirch (Kirch 1997) could be replaced with a simple three<br />
stage overall sequence <strong>of</strong> ceramic change, where the far-western, western and eastern<br />
styles become early, middle and late Lapita (Summerhayes 2000a:235). Sites FNY and FOJ<br />
were regarded by Summerhayes as having long sequences, with heavy use made also <strong>of</strong> site<br />
FOH stratigraphy (squares D, E, and F) also, in defining a temporal directional change in<br />
pottery style. Summerhayes’ evidence from these sites will be reviewed in order to evaluate<br />
his Lapita Ceramic Series.<br />
Site FOH:<br />
FOH was located on a sand spit, with the dense pottery concentration below the water<br />
93
table (as at site ECA in the Mussau group). Excavation over four field seasons included<br />
geomorphological investigation (Gosden & Webb 1994, Summerhayes 2000a:21-22). <strong>The</strong><br />
lower levels <strong>of</strong> the site were thought to have originated as the discard from stilt villages<br />
into shallow water, and these lower 45-50cm <strong>of</strong> deposit were excavated in spits forming<br />
Summerhayes’ analytical units A-E for squares D, E and F, with A being the basal spit and<br />
E the uppermost, the latter in the partly-concreted sand layer between the brackish<br />
groundwater and the salt tidal incursion.<br />
Summerhayes reported that sherd counts decreased dramatically up through these<br />
five levels (Summerhayes 2000a:43), and vessel refitting found a high degree <strong>of</strong> vessel<br />
completeness (see illustrations Summerhayes 2000a:67-70) with many joins between these<br />
levels. Summerhayes considered that as only three co-joins from level E were with<br />
underlying units there was little evidence for vertical disturbance (Summerhayes 2000a:22-<br />
23) but I would argue that 100% <strong>of</strong> the joins in level E were to other layers, and that<br />
Summerhayes' conclusion that little disturbance had occurred since deposition is not well<br />
supported by his data. <strong>The</strong> number <strong>of</strong> co-joins to other levels for each level (Summerhayes<br />
2000a: Table 3.1 on p23) seem consistent with the analyzed sample size for each layer<br />
(Summerhayes 2000a: p44 Table 5.1) and a reasonable explanation for the overall pattern<br />
<strong>of</strong> co-joining is that turbation <strong>of</strong> some description has occurred across the four lower<br />
excavation levels. Summerhayes’ claim that pottery assemblages from these levels can be<br />
regarded with confidence as superposed temporal sets is open to question on this basis<br />
alone.<br />
Are there perhaps taphonomic or other sampling explanations for the changes in<br />
decorative technique and motif observed through the levels? <strong>The</strong>re is clearly a different<br />
chemical weathering regime at work in Layer E as evidenced by the presence <strong>of</strong> the<br />
concretion here, and the decorative and stylistic temporal changes suggested by<br />
Summerhayes are furthermore evidenced by a sample <strong>of</strong> only 49 analyzed sherds from this<br />
94
upper concretion level, including18 decorated sherds; thus sample size effects might also<br />
easily account for the “temporal” patterning noted for this part <strong>of</strong> site FOH. This is<br />
particularly so where vessel completeness is this high, as evidenced by Summerhayes'<br />
excellent illustrations <strong>of</strong> vessel families, with many sherds potentially from single vessels.<br />
<strong>The</strong> vessel sample may consequently be much smaller than the sherd count, which may not<br />
provide independent observations in these circumstances.<br />
FOH squares G1 and G2 were situated 15m south <strong>of</strong> squares D, E and F discussed<br />
above, and yielded 2883 sherds, <strong>of</strong> which 156 were analyzed. Stratigraphy seems almost<br />
identical to squares D, E, and F, with layer three and layer four seeming likely to equate<br />
to the previous units A-D <strong>of</strong> squares D, E and F (Summerhayes 2000a:23). Layer 3 yielded<br />
a total <strong>of</strong> eight decorated sherds, while Layer 4 yielded 89 (Summerhayes 2000a:91-92).<br />
Summerhayes argues in Chapter 10 that the assemblage from this square is later than the<br />
assemblage from Squares D, E and F, units A-D, and contemporaneous with unit E (the<br />
sample <strong>of</strong> 18 decorated sherds). As stated above, a chrono-stratigraphic distinction<br />
between any <strong>of</strong> these lower units seems unwarranted on the laudably detailed evidence<br />
presented by Summerhayes, and a stylistic distinction faces the serious and fundamental<br />
questions <strong>of</strong> whether one is seeing activity/functional patterning across a single occupation<br />
unit or chronological change between areas, or whether this is simply sampling error. <strong>The</strong><br />
differences in the frequency <strong>of</strong> dentate-stamp decoration may also relate to<br />
(contemporaneous?) functional differences in vessel form (Summerhayes 2000a:101) rather<br />
than differences in time.<br />
Site FOJ:<br />
Summerhayes’ evidence for stratified stylistic change at FOJ (one line <strong>of</strong> evidence for his<br />
contention that the direction <strong>of</strong> ceramic change is known) is examined below. Units B, C<br />
and A seem broadly uniform in their characteristics, the major difference being that unit<br />
95
A has about six dentate-stamped sherds less than would be expected from the sample size.<br />
In view <strong>of</strong> the difference between these depositional contexts (B/C, D 1n shallow water<br />
and A on land) attributing such differences to temporal change in ceramic production is<br />
problematic.<br />
Gosden and Webb describe unit A as follows:<br />
“<strong>The</strong> bulk <strong>of</strong> sediment accumulation probably occurred during storm<br />
activity, when sea-level is probably elevated.... (Gosden & Webb<br />
1994:41)”, and comprising: “...a unit <strong>of</strong> light coloured sand...roughly 1m<br />
thick...contains relatively small amounts <strong>of</strong> pottery and obsidian, but<br />
moderate amounts <strong>of</strong> shell....Occasional mumu (oven) stones are also<br />
present....<strong>The</strong> reduced percentage <strong>of</strong> artefactual material in this deposit<br />
indicates that it is unlikely that there was still a village located directly on<br />
the site (Gosden & Webb 1994:40-41).”<br />
<strong>The</strong> presence <strong>of</strong> the oven stones was for unspecified reasons considered evidence for<br />
terrestrial use <strong>of</strong> this location (presumably because these were unlikely to remain in<br />
suspension in the water column, but transport by rolling onto the beach ridge as a result<br />
<strong>of</strong> storm-wave action was apparently not considered). Summerhayes states that: “It is<br />
within this layer that occupation occurred on dry land.... (Summerhayes 2000a:24)”,<br />
implying the ceramics from this layer were from a settlement on dry land. Many <strong>of</strong> the<br />
cultural materials in this layer, including pottery, might derive from material originally<br />
deposited in the sea at an earlier time (contemporaneous with the lower units?) and<br />
redeposited by storm waves to form a beach ridge.<br />
Site FNY:<br />
Summerhayes’ analysis <strong>of</strong> the FNY ceramics concluded that dentate-stamp decoration<br />
drops in frequency in the upper <strong>of</strong> three units, and noted trends in vessel form from lower<br />
to top unit (Summerhayes 2000a:125-126). In Chapter 10, this is re-stated as:<br />
“incised and fingernail impressed decoration increasing in the upper units,<br />
although here dentate decoration is always dominant. (151)"<br />
Gosden and Webb interpret the lower two units, both brown clay separated by a white<br />
96
sand lens, as back-swamp deposits during a period <strong>of</strong> Lapita occupation and beach-ridge<br />
formation (Gosden & Webb 1994:46-47, Summerhayes 2000a:25). <strong>The</strong> top unit comprised<br />
black shell midden with Lapita and Late-prehistoric (my emphasis added) pottery mixed<br />
(Gosden & Webb 1994:47). Summerhayes describes this top unit as comprising<br />
“A top black midden layer (1.2m) incorporating post Lapita (my emphasis<br />
added) and Lapita remains (Summerhayes 2000a:25).”<br />
Summerhayes’ use <strong>of</strong> the term “post Lapita”, suggests episodes <strong>of</strong> ceramic<br />
deposition contributing to the upper unit are not widely separated in time. <strong>The</strong> zone<br />
between the upper mixed unit and the lower, purely Lapita units yielded several<br />
radiocarbon determinations which varied between 1290 and 690 cal BP, with a Lapita-age<br />
estimate from an oyster shell lower down the brown clay. (Summerhayes 2000:25). It<br />
seems well attested from the C14 evidence and from the Lapita and late-prehistoric pottery<br />
styles represented that the upper unit represents a mixture <strong>of</strong> materials from widely<br />
different periods, while at least the upper portions <strong>of</strong> the brown clay also comprise mixed<br />
deposits. <strong>The</strong> absence <strong>of</strong> recent styles in the brown clay lower units, and a single Lapita-<br />
age date, do not provide evidence that the lower clay is temporally un-mixed, merely that<br />
the clay formed during a period in which only Lapita age materials were available as either<br />
a natural or cultural sediment supply. If this site comprises variable mixtures <strong>of</strong> styles from<br />
widely separated periods (c.1000 BC Lapita and c.1000 AD late-prehistoric) it should not<br />
be regarded as a superposition sequence <strong>of</strong> change in the Lapita ceramic complex over<br />
time (see for example Summerhayes 2000a:151), and the status <strong>of</strong> this unit in subsequent<br />
seriations (for example Summerhayes 2000a:152 Figure 10.1) should be interpreted with<br />
this in mind.<br />
Summerhayes’ construction <strong>of</strong> the direction <strong>of</strong> ceramic change does not<br />
convincingly link superposition assemblage variability to time, by under-utilizing Gosden<br />
97
and Webb’s study <strong>of</strong> site formation processes, and by comparisons between samples <strong>of</strong><br />
dissimilar depositional type (see Phelan 1997:36-38). Summerhayes’ conclusion that<br />
“the Arawe assemblages largely agree with other Lapita assemblages which<br />
show both a decrease in dentate and an increase in linear incision over<br />
time....”<br />
and that the Arawe assemblages “confirm” the trends noted by others in this regard<br />
(Summerhayes 2000a:151) is not well supported by the evidence presented. Summerhayes<br />
went on to use this directional theory in a series <strong>of</strong> evolutionary seriations, using<br />
decorative technique frequency and motif occurrence.<br />
Summerhayes’ Seriation Using Decorative Technique/ Vessel Form:<br />
Having concluded that he had confirmed the temporal direction <strong>of</strong> ceramic change,<br />
Summerhayes constructed a seriation <strong>of</strong> West New Britain assemblages based on<br />
decorative technique frequency (Summerhayes 2000a:152) Here excavation squares are<br />
treated as single-period samples rather than time-transgressive units as in the previous<br />
stratigraphic analysis, with all layers being included as the site sample. <strong>The</strong>re is no<br />
discussion in relation to the seriation <strong>of</strong> the relative occupation spans <strong>of</strong> these units, or the<br />
degree <strong>of</strong> temporal mixing <strong>of</strong> separate periods, which are both important aspects <strong>of</strong><br />
seriation method as outlined in Chapter 1.<br />
Decorative technique and vessel form were found to be highly correlated. Sites<br />
failed to seriate when controlled for vessel form in his decorative technique seriation<br />
(Summerhayes 2000a:p153 Figure 10.5), leading Summerhayes to conclude that what<br />
changed over time was that some vessel forms dropped out. Alternative explanations for<br />
the failure <strong>of</strong> these sites to seriate when controlled for vessel form is that the classificatory<br />
units <strong>of</strong> decorative technique used were too coarse, or not salient to stylistic variation, or,<br />
more serious, that stylistic change did not occur across his “sequence”. This last<br />
possibility contrasts with Dunnell’s definition <strong>of</strong> style as evolutionary drift. Some form<br />
98
<strong>of</strong> stylistic change must occur within a single vessel form, even in the most conservative<br />
milieu, unless extraordinary measures are taken to prevent change. <strong>The</strong> difficulty is in<br />
identifying that change. It is possible therefore that the “seriation” is simply a similarity<br />
grouping <strong>of</strong> functional variability, that has little or nothing to do with time. <strong>The</strong> most<br />
striking example <strong>of</strong> this is that his seriation separates FOH squares DE&F from FOH<br />
square G, despite convincing stratigraphic similarities between the lower levels <strong>of</strong> these<br />
excavation units and a spatial separation <strong>of</strong> only 15m.<br />
Summerhayes’ Lapita Motif Analysis and Early/Middle/Late Lapita:<br />
Summerhayes went on to construct a motif presence/absence/sharing similarity grouping<br />
for a range <strong>of</strong> Lapita sites from across the Lapita region, using Anson’s motif inventory<br />
as units <strong>of</strong> classification (Summerhayes 2000a:table 10.5). It is important to note that<br />
presence/absence <strong>of</strong> motifs rather than motif frequencies were the base data for his<br />
analyses. Anson’s splitting classification creates a large inventory <strong>of</strong> motifs for which<br />
there is sparse data. Motif sharing may be quite arbitrary in these circumstances (Kirch<br />
1987a).<br />
Summerhayes performed a cluster analysis <strong>of</strong> motif presence/absence, and a<br />
Principal Components Analysis <strong>of</strong> the number <strong>of</strong> shared motifs between sites. He used<br />
Ward’s clustering method (which tends to force outliers into groups and produce tight<br />
clusters in the data or different clusters compared to some other methods such as average<br />
linkage), but does not state what sort <strong>of</strong> similarity information was input, whether number<br />
<strong>of</strong> shared motifs, or some similarity matrix such as Jaccard. This clustering method has<br />
been widely used in archaeology for numeric data such as trace-element concentrations<br />
(Shennan 1997:244) and produced what I see as anomalous groupings with Anson’s motif<br />
presence-absence data. <strong>The</strong> Reef-Santa Cruz sites formed a cluster separate from all other<br />
sites, and Summerhayes states this is a sample-size effect, without discussion <strong>of</strong> the<br />
99
independence <strong>of</strong> observations and how sample size is measured. As shown above, the<br />
Reef/Santa Cruz Lapita sites include quite large samples from RF2 and SZ8, and also a<br />
very small sample, RF6, so sample size should not cause these to cluster separately.<br />
A PCA plot <strong>of</strong> the first and third PCA components agreed with the cluster analysis,<br />
in separating the Bismarcks sites into similar groupings to those obtained in his vessel-<br />
form/decorative technique analysis. As in his decorative technique frequency seriation,<br />
FOH squares fall into separate clusters, Square G being more similar to Yanuca in Fiji<br />
than to squares D, E and F 15 metres away. Summerhayes is not clear about the type and<br />
structure <strong>of</strong> his primary data: he refers to performing “multivariate analysis on the<br />
presence/absence <strong>of</strong> motifs in all sites” (Summerhayes 2000a: 160) but also captions<br />
Figure 10.2 (his PCA plot) “Motif Sharing”, and does not provide the raw data input to<br />
his analyses, rendering his multivariate computer analysis something <strong>of</strong> a black box. PCA<br />
is most appropriate for numeric data (Shennan 1997:298), and Summerhayes may have<br />
done better to use sherd counts or MNV to identify a selection <strong>of</strong> common motifs with<br />
Correspondence Analysis used as a seriation technique on this selection.<br />
Comparisons are made (Summerhayes 2000a: Table 10.5, Fig 10.12, Table 10.7)<br />
without discussion <strong>of</strong> the occupation spans <strong>of</strong> sites or potential functional variability<br />
within sites. If site occupation spans are lengthy or variable, and function varies<br />
substantially across space within sites, assemblages may comprise an inextricable mix <strong>of</strong><br />
temporal and functional information, without any clear temporal patterning identifiable by<br />
PCA or cluster analysis.<br />
Summerhayes’s statement that “...differences in (sample) size are seen in the first<br />
PCA component...(Summerhayes 2000a:160, 161)” suggests that it is indeed motif-<br />
sharing counts being input to the PCA. Eigenvalues <strong>of</strong> the PCA components are not given,<br />
and he chooses to plot the third component against the first, without discussion <strong>of</strong> the<br />
significance <strong>of</strong> either the second component or the third. What is the significance <strong>of</strong> the<br />
100
second hidden component, which must account for more variation in the data than the<br />
third? Furthermore, if the first component is primarily a reflection <strong>of</strong> sample size, why use<br />
it in the plot, or ascribe temporal significance to the clusters thus formed?<br />
<strong>The</strong> results <strong>of</strong> one further analysis, using Anson’s manual motif-sharing similarity<br />
method (Summerhayes 2000a :160-163 and table 10.7), contrast sharply with the previous<br />
multivariate numerical taxonomy results in some respects. FOH square G, more similar to<br />
Yanuca in Fiji than to the adjacent excavation <strong>of</strong> FOH squares D, E, F in his PCA and<br />
cluster results, appears, using the Anson manual method, to be more similar to Squares D,<br />
E and F than to anywhere else, as might be expected from the same layer <strong>of</strong> the same site,<br />
separated by 15m <strong>of</strong> unexcavated ground.<br />
Summerhayes goes on to use these motif analysis results to bolster his argument,<br />
following Anson, that there is a dichotomy in the Lapita assemblages <strong>of</strong> west New Britain,<br />
and that this is present also in the wider Lapita universe, ascribing a temporal dimension<br />
to this dichotomy, crosscutting a regional (spatial) dimension <strong>of</strong> variation.<br />
“Similar changes in the Lapita decorative system occur in the west and<br />
east....the product <strong>of</strong> information exchange which necessitates the<br />
movement <strong>of</strong> people...” (Summerhayes 2000a:233).<br />
In his discussion <strong>of</strong> the motif results, he is clear that the similarities noted between<br />
assemblages relate to the presence or absence <strong>of</strong> particular vessel forms in the recovered<br />
samples, but still does not talk about the extent to which contemporaneous within-site<br />
functional variation, rather than temporal variation, might account for his groupings <strong>of</strong><br />
West New Britain (and other) Lapita assemblages.<br />
While the inventory <strong>of</strong> vessel functions in use in Lapita communities may have<br />
shifted over time, it is also entirely possible that vessel forms/functions were unevenly<br />
distributed across such communities, and that telephone-booth archaeology is hitting<br />
varying mixes <strong>of</strong> the plain stuff and the fancy stuff. FOH in the Arawes may be the spatial<br />
lesson in this regard, as Sheppard and Green have suggested also for RF2 in the Southeast<br />
101
Solomon Islands (Sheppard & Green 1991).<br />
My conclusion is that the Early/Middle/Late Lapita construct is an example <strong>of</strong><br />
Dunnell’s “...inseparable hodgepodge...” <strong>of</strong> geographic, functional and temporal variation<br />
(Dunnell 1970:310), to which could be added uncertainties regarding taphonomic bias,<br />
quantification <strong>of</strong> motifs and the degree to which samples are representative.<br />
Early/Middle Lapita in the Anir Group:<br />
Recent extension <strong>of</strong> the west New Britain research through a program <strong>of</strong> survey and<br />
excavation in the Anir group has augmented the corpus <strong>of</strong> Bismarck Archipelago Lapita<br />
samples, with the most notable assemblages excavated from a single productive testpit at<br />
Malekolon and from a grid <strong>of</strong> 14 productive testpits at Kamgot (Summerhayes 2000b,<br />
2001, 2002). Summerhayes suggests Malekolom TP4 (where the main cultural layer was<br />
located) was deposited on land, near to the then shoreline, while the extensive deposit at<br />
Kamgot was thought to have been deposited either in the water or at the water’s edge<br />
(Summerhayes 2000b). Summerhayes slotted Malekolon and Kamgot sample assemblages<br />
into his early-middle-late Lapita chronology, using the same methods <strong>of</strong> vessel form<br />
classification, frequency <strong>of</strong> dentate-stamping (usefully quantified using both sherd count<br />
and MNI this time) motif-based assemblage grouping, and obsidian source frequency).<br />
Summerhayes' use <strong>of</strong> C14 evidence to test his Early/Middle temporal construct<br />
(Summerhayes 2002) is not an independent confirmation <strong>of</strong> the temporality <strong>of</strong> variation.<br />
While a plot <strong>of</strong> some <strong>of</strong> these dates appears to be convincing evidence <strong>of</strong> the temporal<br />
primacy <strong>of</strong> Summerhayes’ “early Lapita”, his process <strong>of</strong> chronometric hygiene excludes<br />
any dates which don’t agree with his temporal construction. Thus Beta 55323, a date <strong>of</strong><br />
2880 ±70BP from Adwe squares DEF (“early” Lapita) is rejected solely because it does<br />
not fit his temporal hypothesis, and similarly, WK7564, a date <strong>of</strong> 2960-2760 cal BP for the<br />
Kamgot “early” Lapita deposit is rejected “because it is statistically distinguishable<br />
102
from the other four dates (from that context)”.<br />
Summerhayes also rejects Kamgot dates that have a wide confidence interval<br />
(ANU 11193, ANU 11190, ANU 11191). One <strong>of</strong> these dates(ANU 11190: 2750-1530 cal<br />
BP), rather than limiting “the ability to make fine discriminations in the chronology”would<br />
tend to undermine his chronological hypothesis (Kamgot as early Lapita) in spite <strong>of</strong> its<br />
broad confidence interval. Summerhayes’ conclusion that “<strong>The</strong>se new determinations<br />
confirm the chronology for sites in that region that were based on regional comparisons<br />
<strong>of</strong> pottery decoration, form and production.” is thus not unequivocally supported by<br />
current data. Any evidence for occupation span (non-overlapping ages) is ruled out rather<br />
than put to good use to understand what sort <strong>of</strong> temporal mixture the samples comprise.<br />
<strong>The</strong> Watom Lapita Series:<br />
Watom, SAC location, had Lapita deposits preserved with a minimum <strong>of</strong> recent<br />
disturbance under a volcanic ash fall (Green & Anson 2000b:37). Anson concluded that<br />
the stratified deposits containing pottery at SAC locality “enabled the chronological<br />
position <strong>of</strong> unstratified material from SAD and museum collections to be inferred (Anson<br />
2000)”. Best critiqued Green and Anson's temporal conclusions, that Watom Lapita dated<br />
500BC-AD200, suggesting the younger C14 dates obtained postdated the Lapita pottery<br />
recovered (Best 2002:87).<br />
Green and Anson noted that initial Lapita occupation at Watom SAC zone C2<br />
predates their earliest C14 date (Green & Anson 2000b:38), preferring to discard a<br />
Trochus shell date <strong>of</strong> 3490±80BP (calibrated at 1-sigma to 1509-1350BC) in the surface<br />
<strong>of</strong> Zone D as being unrelated to Zone C2 occupations (Green & Anson 2000b:39). Green<br />
and Anson present ample and explicit evidence that the buried Zone C1 paleosol is a<br />
loamy gardening soil incorporating cultural materials (Green & Anson 2000b:42), but this<br />
103
stratigraphy is elsewhere described as “...two cultural occupation layers, C1 and C2....<br />
(Green & Anson 2000b: 35).” Zone C1 was dated using several partly dissolved Tridacna<br />
maxima shells from the base <strong>of</strong> the layer (Green & Anson 2000b:38). C2 was dated using<br />
a stack <strong>of</strong> Tridacna shells in a pit feature unconformably overlain by zone C1. <strong>The</strong>se were<br />
calibrated and interpreted to yield a time span <strong>of</strong> 400BC to about AD100. As Green and<br />
Anson are careful to point out, features associated with Zone C1 could only be identified<br />
as intrusions into the surface <strong>of</strong> C2, confirming a turbation unconformity as the formation<br />
process <strong>of</strong> Zone C1, as do the smaller and more fragmented bones and potsherds from<br />
Zone C1 in comparison to Zone C2.<br />
For Zone C2 it was possible to demonstrate sequences <strong>of</strong> superimposed features.<br />
Fom the section drawings, it looks as though a burial ground was cut in from C1 level,<br />
prior to the formation <strong>of</strong> the C1 unconformity, and which postdates some pottery and<br />
obsidian from C2 (Green & Anson 2000b:45). In view <strong>of</strong> these formation processes,<br />
Anson’s conclusion that “A fair degree <strong>of</strong> motif sharing between layers C1 and C2 at SAC<br />
gives an overall impression <strong>of</strong> continuity, with no evidence <strong>of</strong> a dramatic break....(Anson<br />
2000:133)” ignores the fact that zone C1 clearly incorporates materials mixed in from, and<br />
originating as the upper levels <strong>of</strong> part <strong>of</strong> Zone C2, during the creation <strong>of</strong> the C1 turbation<br />
unconformity. Furthermore, a complex history <strong>of</strong> excavated unconformities is evidenced<br />
in Zone C2 stratigraphy (Green & Anson 2000b:44), which must have acted to bring<br />
materials from the earliest facies <strong>of</strong> C2 into younger stratigraphic features, and into<br />
stratigraphic association with Mopir obsidian, thought to be unavailable in Lapita times<br />
(White & Harris 1997). <strong>The</strong> burial pits cannot have been “...sealed in by layer C1 (Green<br />
& Anson 2000b:45)." given the formation processes so well documented for C1. Materials<br />
constituting C1 must in part derive from the upper levels <strong>of</strong> what was originally C2 prior<br />
to gardening.<br />
Rather than there being no evidence for a dramatic break, an important question<br />
104
is whether there is historical continuity <strong>of</strong> occupation beneath the ash from the time <strong>of</strong><br />
initial Lapita deposition to the time <strong>of</strong> the use <strong>of</strong> the site as a burial ground. As Best points<br />
out, some <strong>of</strong> the zone C1 pottery is not necessarily from a late-Lapita context. <strong>The</strong>re is no<br />
evidence to rule out the possibility that some <strong>of</strong> this pottery could substantially post-date<br />
the Lapita-horizon, while pre-dating the Zone B ash, with an intervening hiatus <strong>of</strong> several<br />
centuries, while those sherds that are clearly Lapita may be incorporated by the well-<br />
evidenced mixing which was the origin <strong>of</strong> C1. How does one tell whether Lapita pottery<br />
in a deposit is mixed with post-Lapita, when post-Lapita sequences are so ill defined as in<br />
the Bismarck Archipelago? Given the clear evidence for excavation into C2 after<br />
deposition <strong>of</strong> Lapita pottery, which involved burial practices not known elsewhere in Near<br />
Oceania for Lapita, it seems reasonable to worry that the stratified sequence <strong>of</strong> “change”<br />
may be blurred by these formation processes. Anson’s analysis <strong>of</strong> ceramic style by layers<br />
is critiqued below in view <strong>of</strong> these formation processes and uncertainties regarding<br />
occupation spans and historical continuity <strong>of</strong> deposition.<br />
Nail-impressed sherds were found in layer C1 with Lapita, but not in Layer C2.<br />
This may simply be a sample size effect (Green & Anson 2000b:77), or a mixing <strong>of</strong><br />
different periods.<br />
Anson found that Lapita motifs from SAC C2 were more similar to SAD and<br />
Meyer’s collection than to SAC C1, although SAD and the Meyer collection are more<br />
similar to each other than to C2 (Anson 2000:132). Anson concludes that SAC Zones C1<br />
and C2 represent temporal stratigraphy, based on a relatively low degree <strong>of</strong> motif sharing.<br />
Given that the 38 motif occurrences listed for “SAC L1" and “SAC L2" in Table 1 are<br />
represented by sparse frequency data, in 29 instances by single sherds per layer, and never<br />
by more than two sherds per layer for the remaining nine sherds, it is not surprising that<br />
these layers have a low degree <strong>of</strong> motif sharing. Contrast this with the combined Meyer<br />
sample where single motifs are represented by up to 16 sherds per layer, (motif 429) and<br />
105
it seems clear that there are major unresolved sampling issues here which have not been<br />
discussed.<br />
Anson, in grouping collections using motif sharing, is attempting to produce an<br />
occurrence seriation, without due regard to the effects <strong>of</strong> sample sizes on motif<br />
occurrences, and also without consideration <strong>of</strong> what his counts mean behaviourally, and<br />
how sample size might be quantified. Even assuming that his method <strong>of</strong> quantification is<br />
salient to the completeness properties <strong>of</strong> the samples, if seven out <strong>of</strong> 16 motif occurrence<br />
at SAC C2 are shared with SAD, and four out <strong>of</strong> 18 motif occurrences in SAC C1 are<br />
shared with SAD, it cannot be concluded with confidence that there is a significant<br />
difference. <strong>The</strong> difference <strong>of</strong> three motifs may simply be due to chance.<br />
Anson’s analysis <strong>of</strong> the complete corpus <strong>of</strong> Watom pottery states that most <strong>of</strong> the<br />
Mussau Lapita sequence is much earlier than Watom Lapita on radiocarbon evidence<br />
(Anson 2000:120, footnotes), but Green and Anson noted that initial Lapita occupation<br />
at Watom SAC zone C2 predates their earliest C14 date (Green & Anson 2000b:38). In<br />
a similar vein as for the Reef-Santa Cruz sites, I suggest that insufficient consideration has<br />
been given to the possibility that the SAC site and the Reber/Rakival area generally have<br />
been the focus <strong>of</strong> Lapita occupation for an extended period, potentially beginning many<br />
centuries earlier than 400BC, and that individual site/excavation unit assemblages may be<br />
palimpsests <strong>of</strong> a variety <strong>of</strong> occupational events and pottery styles within the Lapita period,<br />
with lengthy hiatuses in occupation not ruled out by the evidence. Ishimura’s phyletic<br />
analysis would tend to support this alternate view, given the diversity <strong>of</strong> face and spade<br />
designs from Watom (Ishimura 2002).<br />
<strong>The</strong> pottery styles illustrated (e.g. Green & Anson 2000a: Figure 1) seem to me to<br />
have equivalents in all three analyzed Reef-Santa Cruz assemblages and also in Mussau<br />
and the Arawes. <strong>The</strong> possiblity that initial domestic occupation began substantially earlier<br />
than 400BC, potentially as early as the dated Trochus shell from Zone D (which is<br />
106
consistent with the earlier dates from house-posts at Mussau) is not precluded by Green<br />
and Anson’s detailed excavation data and materials analyses. <strong>The</strong> large numbers <strong>of</strong> Lapita<br />
sherds recovered from Zone C1 are from a chronostraphic unit containing ceramics<br />
potentially manufactured between c. 1400BC and c. 100BC, whereas some <strong>of</strong> the earlier<br />
stratigraphic units in Zone C2 incorporate materials manufactured over an earlier, shorter<br />
period.<br />
Wahome’s Seriation <strong>of</strong> Admiralties Pottery Assemblages:<br />
Wahome’s comparison <strong>of</strong> incised/applied pottery site across Island Melanesia has been<br />
called into question by new data from Vanuatu and by criticism <strong>of</strong> classificatory units used<br />
as being too coarse to be salient to the research question (Bedford 1999:167-189). For the<br />
Admiralties, Wahome came up with a local sequence <strong>of</strong> four periods, from Lapita, through<br />
early post Lapita, late post-Lapita and late prehistoric (Wahome 1999:112-118). This<br />
series conflicts with Ambrose’s assessment <strong>of</strong> historical continuity <strong>of</strong> the Admiralties<br />
sample. According to Ambrose a gap <strong>of</strong> up to seven centuries separates Lapita pottery<br />
from subsequent styles (Ambrose 1991, Bedford 2000:183).<br />
Potential problems with historical continuity aside, Wahome’s approach is<br />
interesting for his use <strong>of</strong> Correspondence Analysis. Wahome used Correspondence<br />
Analysis to seriate fourteen Admiralty Islands pottery samples (Wahome 1999:33-34 and<br />
96-99). Both Lapita-style and later/other pottery variants were included in his attribute and<br />
attribute-combination analyses. Wahome was concerned to identify attributes and attribute<br />
combinations that varied over time. Wahome saw the summary variables in his analysis as<br />
separating functional and temporal variation. He looked also for “reliable sequences”<br />
in these dimensions <strong>of</strong> variability by cross checking against stratigraphic trends in<br />
excavated sites. He had difficulty seriating some units and concluded that some <strong>of</strong> the<br />
division <strong>of</strong> excavated sites into stratigraphic/superposition units was unjustified, due to<br />
107
lack <strong>of</strong> variability between units (Wahome 1999: section 5.3). His seriation became more<br />
coherent when these units were lumped.<br />
His Lapita unit remained distinct from all other units in his analysis (which is<br />
consistent with Ambrose’s suggestion <strong>of</strong> a lack <strong>of</strong> historical continuity). <strong>The</strong> site <strong>of</strong> Mouk<br />
yielded both Lapita and post-Lapita evidence (Wahome 1999:5.3, 5.5.1). Three “early<br />
post-Lapita” sites dating to around 2000BP could be seriated using the frequency <strong>of</strong> shell<br />
impression, with shell impression absent earlier than 1900BP (Wahome 1999:5.4.1,<br />
5.5.2.1). A group <strong>of</strong> ten “Late post-Lapita” sites seriated in the first dimension (66.8%<br />
inertia) on the frequency <strong>of</strong> cross-hatch incisions, rolled-rims, shell impressions, and<br />
combinations <strong>of</strong> these. A second dimension <strong>of</strong> variability (17%) comprised the frequency<br />
<strong>of</strong> deep notching, flat lips on rolled rims, and notching on the plain lips <strong>of</strong> incurving rims.<br />
<strong>The</strong> first dimension was ascribed to time by comparison with stratigraphic relationships<br />
between units, and the second was disregarded. Eight <strong>of</strong> the ten late post-Lapita units<br />
included fingernail and/or fingertip impressions, with one site showing a greater diversity<br />
<strong>of</strong> motifs in these decorative techniques. A “highly diversified range” <strong>of</strong> incised decorations<br />
was recorded in all these units, and punctations were present in seven sites.<br />
His final grouping <strong>of</strong> 19 “Late prehistoric” sites (post-800BP) was differentiated<br />
primarily (inertia <strong>of</strong> 24%) on the frequency <strong>of</strong> brushing, in which respect one site in the<br />
group differed significantly from the others. A number <strong>of</strong> attributes/combinations<br />
contributed to the second dimension (inertia was eighteen percent). He considered that<br />
most <strong>of</strong> this variability correlated with space rather than time or function. <strong>The</strong>re was a<br />
wider range <strong>of</strong> vessel forms in evidence than in the preceding period, with brushing and<br />
perforation more common (Wahome 1999:5.4.3, 5.5.2.3).<br />
Overall, taking into account units <strong>of</strong> classification used, and the use <strong>of</strong> seriation<br />
in combination with stratigraphy and surface sites, Wahome’s approach is similar in<br />
method to the Irwin/Frost school, but with the more up-to-date technique <strong>of</strong><br />
108
correspondence analysis substituted for similarity matrix assemblage grouping. Like Irwin,<br />
Wahome put some effort into identifying temporal variation, primarily through comparison<br />
with stratified superposition chronologies. Wahome’s Lapita and Early post Lapita phases<br />
cannot be considered to form a historically continuous series though, as historical<br />
continuity is poorly evidenced at best, and Wahome makes no argument in support <strong>of</strong><br />
heritable continuity. Thus Lapita remains a horizon rather than a series in the Admiralties.<br />
<strong>The</strong> Buka sequence, to the north <strong>of</strong> Bougainville, was established in its current<br />
incarnation primarily through the pioneering work <strong>of</strong> Specht (Specht 1969), and a further<br />
indirect contribution came out <strong>of</strong> the Lapita Homeland Project on Nissan (Spriggs 1991).<br />
More recently, the Buka area has been substantially researched by Wickler (Wickler<br />
2001).<br />
Specht on Buka:<br />
Specht constructed a detailed sequence for the Buka area based on information from 73<br />
sites (Specht 1969: 211-213), spanning the period from Lapita to the present (Specht<br />
1969: Figure XII-52). Specht did not regard his sequence as continuous, being sure there<br />
were temporal sampling gaps. Specht used the hierarchical construct <strong>of</strong> tradition, style,<br />
substyle and attribute in his ceramic analysis (Gardin 1967, Willey & Phillips 1958),<br />
regarding attributes as irreducible units <strong>of</strong> description, with essential properties as<br />
discrete, mutually exclusive and immutable entities (Specht 1969: 64-65). Although he<br />
implicitly rejected the concept <strong>of</strong> pottery type as hindering the study <strong>of</strong> phyletics,<br />
regarding styles, in a materialist manner, by contrast, as involving time in their definition,<br />
his use <strong>of</strong> phyletic stylistic analysis was limited, as discussed further below, with styles<br />
being largely a finer-grained classification than types. Specht avoided multivariate<br />
computer classification approaches in favour <strong>of</strong> an iterative visual approach to analytical<br />
109
classification. He took a metric approach to vessel form variability, making use <strong>of</strong> vessel<br />
restriction factor, a measure <strong>of</strong> vessel form popular in American archaeology at that time.<br />
Specht included a succinct yet pertinent discussion <strong>of</strong> ceramic quantification, which<br />
anticipated Egl<strong>of</strong>f’s significant contribution in this area, suggesting that the subject had<br />
received some consideration at the ANU generally at that time, perhaps more so then than<br />
more recently (Egl<strong>of</strong>f 1973, Specht 1968:70-71, 2002). He used sherd count as his primary<br />
measure <strong>of</strong> quantity, but counted joining sherds as one. Where there were joins between<br />
layers, the combined sherd was counted in the layer with the most sherds/largest sherd,<br />
providing a rough correction for mixing in some instances.<br />
<strong>The</strong> recent end <strong>of</strong> the sequence was defined using surface-collected material. Here,<br />
Specht again provides a succinct and pertinent methodological discussion, this time <strong>of</strong><br />
seriation method, and suggests that unrestricted cliff-top site locations are more likely to<br />
provide short duration sites than those on the narrow coastal flat constrained by cliffs, the<br />
later being more likely to present mixed assemblages from extended or multiple periods <strong>of</strong><br />
occupation. He supported this theory with a number <strong>of</strong> stylistically homogenous collections<br />
from the cliff tops thought to be “single-period” occupations (Specht 1968:160-161).<br />
Specht placed heavy emphasis on paste classes in his analysis, an approach<br />
followed more recently on materials from the same general region by Summerhayes, and<br />
by Wickler and by Spriggs (Spriggs 1991, Summerhayes 1996, Wickler 2001) and in line<br />
with Shepard’s emphasis on technological attribute analyses (Shepard 1963).This approach<br />
has been partially borne out by Wickler’s subsequent analysis, in that the Lapita-age<br />
pottery largely contrasted with all other groups in this respect, being calcite-tempered,<br />
where all subsequent pottery had a significant terrigenous component, apart from<br />
occasional post-Lapita Sohano-style sherds (Specht 1968:192-194).<br />
Specht constructed a sequence <strong>of</strong> styles and sub-styles, beginning with slab-<br />
110
constructed, shell tempered Buka style, followed by Sohano one-piece paddle and anvil<br />
built style, (considered to have been strip-built, followed by paddle and anvil forming:<br />
Specht 1969:196). Hangan, Malasang, Mararing and Recent styles were considered a<br />
gradual evolution from Sohano, but Specht was equivocal on whether homologous<br />
similarity/heritable continuity and historical continuity could be seen between Buka<br />
(Lapita-age) and Sohano phases (Specht 1968 :195, 214, 216, 230). On the one hand some<br />
early Sohano substyle vessels occurred in crushed shell paste, and use <strong>of</strong> the paddle and<br />
anvil is a linking feature, but abandonment <strong>of</strong> slab construction, and the absence <strong>of</strong> a<br />
Watom-Lapita-style <strong>of</strong> broad rim notching from the large Sohano and later assemblages<br />
suggested a lack <strong>of</strong> heritable continuity.<br />
Some <strong>of</strong> Specht’s complex motifs (e.g. M34) diagnostic <strong>of</strong> particular substyles, are<br />
listed as occurring in other styles also, and this seems to indicate a too-direct reading <strong>of</strong><br />
stratigraphic superposition as recording changes in ceramic production. I think it more<br />
likely that complex motifs like M34 have a relatively specific period <strong>of</strong> production, and that<br />
these occur in other levels as a result <strong>of</strong> the vagaries <strong>of</strong> site formation. Just where this<br />
leaves Specht’s construction <strong>of</strong> a gradual evolution from Sohano to Recent is difficult to<br />
assess (see for example Specht 1969:258)<br />
Wickler on Buka:<br />
Wickler’s recent re-appraisal <strong>of</strong> the Buka/Nissan sequence, which largely reiterated<br />
Specht’s and Spriggs’ ceramic sequences with minor refinement, added substantial<br />
additional Lapita-phase data, resulting largely from inclusion <strong>of</strong> reef-flat locations in his<br />
survey strategy (Wickler 2001:8).<br />
Numerically large (high sherd count) yet quite broken assemblages <strong>of</strong> ceramics<br />
were recovered from Lapita reef sites and terrestrial test excavation <strong>The</strong> average sherd<br />
weight for a sample from site DAF was five grams, although sherds on the outer reef at<br />
111
this site were larger. At site DJQ 7.6% <strong>of</strong> sherds were over 6cm maximum dimension<br />
(Wickler 2001:76).<br />
Attribute description <strong>of</strong> Lapita sherds was made with the aims <strong>of</strong> investigating<br />
assemblage variability, and <strong>of</strong> seriating assemblages using a temporally sensitive subset <strong>of</strong><br />
attributes. Vessel forms were defined primarily on the basis <strong>of</strong> rim attributes (Wickler<br />
2001:77), but the criteria for distinguishing between everted rims <strong>of</strong> restricted vessels and<br />
open bowls are not clear. Wickler’s finding that the Lapita assemblage from site DJQ was<br />
dominated by open bowls (Wickler 2001:90) is difficult to evaluate in the absence <strong>of</strong><br />
illustrated examples <strong>of</strong> sherds that were used to make such identifications. Similarly,<br />
functional distinctions between assemblages based on vessel form (Wickler 2001:122) are<br />
not easily verified. No raw attribute or decorative data at the level <strong>of</strong> the sherd were<br />
presented, nor are examples <strong>of</strong> the sherds identifying some vessel forms illustrated. Some<br />
vessel form categories are very broad (“Form 9" for example). Sherd size was not recorded<br />
on a sherd-by-sherd basis, but for rim sherds an approximation <strong>of</strong> sherd size could be<br />
generated from the attributes “estimated rim diameter” and “rim percentage. <strong>The</strong> lack <strong>of</strong><br />
sherd size information makes identification <strong>of</strong> taphonomic processes and motif difficult.<br />
Lip form data was strikingly patterned by site, with the plain “Form 1" lip dominant<br />
at site DES (on Nissan) (commonly found on “Form 1a bowls” and “Form 4 bowls”).<br />
“Form 5" lip dominated at site DJQ (suggested to be diagnostic <strong>of</strong> “Form 2b” everted<br />
bowls), and “Form 11" lips were dominant in the large DAF sample (diagnostic <strong>of</strong> “Form<br />
9b” necked jars) (Wickler 2001:91). Continuous attribute interval class frequency data<br />
(Wickler 2001:92, Table 4.8) provides useful information on the ceramic sample size and<br />
vessel forms <strong>of</strong> the Lapita sites, with counts <strong>of</strong> neck angle classes, carination angle classes,<br />
and rim percentage classes, but it is impossible to know whether these counts are<br />
independent observations.<br />
112
Macroscopic temper analysis used a system <strong>of</strong> relative abundance ranking (Wickler<br />
2001:98) very similar to the final scheme adopted in Chapter 4 for the Roviana Data.<br />
Ferromagnesian tempers do occur in the Buka Lapita sites in low frequencies, so Specht’s<br />
finding that Lapita-age sherds were exclusively shell tempered was a sampling error. Seven<br />
sherds out <strong>of</strong> the 6519 for which temper class was identified had quartzo-feldspathic<br />
temper dominant, which may turn out to be more significant that initially thought by<br />
Wickler, as sherds <strong>of</strong> this temper type produced arresting petrographic results in the<br />
Roviana case (Dickinson 2000a, Felgate & Dickinson 2001). <strong>The</strong>se sherds were thought<br />
to be local by Wickler, but should be compared with the quartz-calcite hybrid tempers in<br />
the present study.<br />
His starting point in seriating the reef Lapita sites was to use frequency <strong>of</strong><br />
decoration, and frequency <strong>of</strong> dentate decoration, both <strong>of</strong> these characteristics thought to<br />
be indicators <strong>of</strong> early assemblages, based on the work <strong>of</strong> Green, Donovan, Parker, also <strong>of</strong><br />
the Arawe assemblages from the Bismarcks as outlined by Gosden and others, and also as<br />
in Anson’s analysis. Wickler’s seriation thus had an evolutionary component as the<br />
direction <strong>of</strong> change was assumed to be known. This may turn out to be a self-fulfilling<br />
prophesy, in light <strong>of</strong> Kirch’s recent suggestion <strong>of</strong> an early plain assemblage predating<br />
Lapita from Mussau.<br />
Wickler’s attribute frequency seriation <strong>of</strong> Lapita reef sites used an unusual and<br />
potentially problematic set <strong>of</strong> relative frequencies: frequency <strong>of</strong> decorated sherds<br />
(expressed as a percentage <strong>of</strong> the total sherd count), the frequency <strong>of</strong> dentate-stamping<br />
(again expressed as a percentage <strong>of</strong> total decoration?), bounded incision, and unbounded<br />
incision (these are also assumed to be percentages <strong>of</strong> the total number <strong>of</strong> decorated sherds<br />
as the rows in the seriation do not sum to 100). He found a low overall frequency <strong>of</strong><br />
decoration on the beach at DAF, with a high frequency <strong>of</strong> unbounded incised decoration,<br />
and characterized this a late site (300-100BC), while DJQ was regarded as falling within<br />
113
the Lapita period envisaged by Sand, 1000-800BC, with the highest frequency <strong>of</strong><br />
decoration, lots <strong>of</strong> dentate-stamping, and bounded rather than unbounded incision (Wickler<br />
2001:122). As discussed in the methodological reviews <strong>of</strong> quantification and seriation in<br />
Chapter 1, the percentage <strong>of</strong> sherds with decoration is easily biased by differences in<br />
brokenness (and brokenness is clearly different across his seriation units), while percentage<br />
decorated is also an attribute that may not seriate well in the first place due to the<br />
possibility <strong>of</strong> decorative extent and abundance over time.<br />
Wickler was careful to note the difficulty <strong>of</strong> separating functional and<br />
chronological variation, but felt that functional variation could only account in part for the<br />
reef site decorative/form variation. Also, the total number <strong>of</strong> collected samples was five<br />
(three areas from DAF and one each from DES and DJQ), which is on the light side for<br />
assessing the degree <strong>of</strong> temporal mixing. <strong>The</strong> possibility <strong>of</strong> temporal mixing with other<br />
later ceramic styles will be discussed further in Chapter 13.<br />
Wickler’s Lapita design analysis was a hybrid <strong>of</strong> Mead’s and Anson’s analytical<br />
schemes (Wickler 2001:122-127), with an additional feature <strong>of</strong> careful structure in relation<br />
to decorative location and details <strong>of</strong> execution (Wickler 2001: Appendix 1). By using<br />
Mead’s conception <strong>of</strong> design zone, and Mead’s concept <strong>of</strong> General/Restricted Zone<br />
Markers (Mead 1975: 26, Fig 2.11, 2.12), some conflation <strong>of</strong> the categories <strong>of</strong> decorative<br />
technique and decorative layout was introduced, creating proliferation <strong>of</strong> descriptive terms.<br />
<strong>The</strong> careful control for decorative location in his motif frequency analysis (Wickler 2001:<br />
Appendix A) is a welcome innovation to Lapita analysis (Mead’s design zones are<br />
locationally vague by comparison), but structural covariation <strong>of</strong> decoration <strong>of</strong> various<br />
vessel parts was not investigated in any detail or made use <strong>of</strong> in analysis. His motif<br />
occurrence and motif frequency comparisons were very much in the mode <strong>of</strong> Anson.<br />
Decorative classification did not take account <strong>of</strong> the structuring <strong>of</strong> decoration across the<br />
vessel.<br />
114
In reporting his most common Lapita stamped and incised motif frequencies as<br />
Anson motifs 435, 2, 187/188 and 421 (Anson 1986: 189-256, Wickler 2001: 124-125),<br />
the effect <strong>of</strong> sherd size on assignment <strong>of</strong> these very similar motifs is not discussed. In view<br />
<strong>of</strong> the small sherd sizes reported, how does one differentiate between motifs 435, 297, 190,<br />
188, 187 or 434 when these are all made up <strong>of</strong> parallel lines? Illustration <strong>of</strong> examples on<br />
which such identifications were based would have clarified these issues, as would<br />
presentation <strong>of</strong> raw analytical data, including sherd sizes (not recorded individually), vessel<br />
parts represented on the sherd, vessel form ascription and motif ascription. This point is<br />
not entirely critical for comparison with the Roviana materials, as, in his decorative<br />
attribute analysis <strong>of</strong> Lapita, Wickler made a useful distinction between bounded and<br />
unbounded incision, compatible with the present Roviana units <strong>of</strong> classification as<br />
discussed in Chapters 1, 4 and 9.<br />
Wickler’s Robinson Coefficient motif frequency comparison <strong>of</strong> Buka reef sites with<br />
Lapita sites <strong>of</strong> the Reef/Santa Cruz area, Watom and New Caledonia found the reef sites<br />
DJQ and DAF much more similar to each other than to any <strong>of</strong> the other sites, while DES<br />
(Tarmon on Nissan) was about as similar to RF2 and RF6 as to the other sites. <strong>The</strong> Jaccard<br />
coefficient motif sharing analysis using agreement scores gave different results, and<br />
Wickler was inclined to favour the frequency data (Robinson similarity coefficient) over<br />
the occurrence data. DES had a low motif count, so sample error was a possibility there<br />
(Wickler 2001:128-129). This general finding contrasts with Summerhayes’ parallel<br />
changes across the Lapita distribution from early Lapita to middle Lapita, unless the Buka<br />
Lapita is all very late, and thus more regionalized (in which case the presence <strong>of</strong> flat bases,<br />
stands, and some fine needle-like dentate conflicts with the Summerhayes/Anson<br />
hypothesis).<br />
Wickler’s analysis <strong>of</strong> post-Lapita excavated assemblages will not be reviewed, for<br />
reasons <strong>of</strong> brevity, since his findings did not differ from those <strong>of</strong> Specht. Wickler’s<br />
115
finding that there was evidence for heritable continuity between Lapita and Sohano phase<br />
is significant in that it links Lapita and post-Lapita in a tenuous way (as discussed by<br />
Specht, despite his eventual opting for no connection). Also Sohano-phase ceramics are<br />
the only post-Lapita Buka ceramics that bear any resemblance to the Roviana intertidal<br />
ceramics, although late-prehistoric Buka ceramics could be usefully compared to late<br />
Roviana terrestrial plainware (some <strong>of</strong> the Roviana “post-Lapita” styles may be present in<br />
the reef-flat collections <strong>of</strong> Wickler in low frequencies: this will be discussed further in<br />
Chapter 13).<br />
Regarding the nature <strong>of</strong> the transition from Buka (Lapita-Phase) to Sohano<br />
ceramics, he concluded:<br />
“Although handicapped by low sample size, disturbed deposits and a lack<br />
<strong>of</strong> reliable radiocarbon dates, the available evidence indicates a temporal<br />
overlap in the production <strong>of</strong> Buka (Lapita) and Sohano style ceramics, and<br />
a gradual replacement <strong>of</strong> the former by the latter (Wickler 2001 :144). ”<br />
While occasional compositional overlap and a few sherds <strong>of</strong> transitional style hinted at<br />
heritable continuity (Wickler 2001:141) the interpretation <strong>of</strong> this as gradual change, by<br />
default, in the absence <strong>of</strong> evidence to the contrary (Wickler 2001:168) cannot be regarded<br />
as evidence delimiting rates <strong>of</strong> ceramic change, and I take Wickler’s statements to mean<br />
that Buka and Sohano are related traditions, without necessarily accepting his suggestions<br />
<strong>of</strong> a gradual transition or temporal overlap in production, for which there seems to be no<br />
particular evidence in the detail <strong>of</strong> the excavations, where ample evidence for mixing <strong>of</strong><br />
ceramics across levels is presented.<br />
Vanuatu:<br />
This review will focus on the recent work by Bedford, in which earlier works are reviewed.<br />
Although a number <strong>of</strong> surface exposures <strong>of</strong> Lapita pottery have been recorded on Malo,<br />
116
Hedrick, who recorded many <strong>of</strong> these sites, focused on investigating settlement strategy<br />
through site location (Hedrick n.d:225-243), and, despite working with large surface<br />
samples, did not draw any detailed conclusions regarding temporal changes in Lapita<br />
decoration. Galipaud has conducted extensive surveys and limited excavation on Malo,<br />
Santo and Torres (Galipaud 1998), but has yet to advance any detailed ideas on the nature<br />
<strong>of</strong> temporal changes within the Lapita phase.<br />
Bedford in Vanuatu:<br />
A recent extensive reworking <strong>of</strong> the Vanuatu sequence by Bedford, which included<br />
reporting <strong>of</strong> substantial new excavation sequences, resolving a number <strong>of</strong> primary culture-<br />
historical issues, has Lapita as the foundation pottery style (Bedford 2000: 1, 35-37, 49-50,<br />
81, 153-165, 239, ) rather than Mangaasi (Bedford 2000:27). Subdivision or explication<br />
<strong>of</strong> Lapita-phase temporal variation was not attempted by Bedford, due to the paucity <strong>of</strong><br />
Lapita-phase ceramics recovered from reconnaissance testpitting in landscapes <strong>of</strong>ten deeply<br />
covered by tephra. Bedford considered that all Lapita samples were at the late end <strong>of</strong> a<br />
Lapita “Horizon”, with initial Lapita settlement circa 3000BP, after consideration <strong>of</strong> the<br />
radiocarbon evidence and the coarse dentate-stamping on the sherds recovered, which in<br />
the Anson/Summerhayes model indicate late date in a Lapita series. It seems likely that<br />
evidence for settlement dates similar to New Caledonia will turn up eventually.<br />
On Erromango and Efate, Bedford concluded that local sequences were a product<br />
<strong>of</strong> continuous ceramic evolution out <strong>of</strong> Lapita, while on Malakula, a post-Lapita Malua<br />
phase also bore a homologous similarity to Lapita. On Malakula, however, in contrast to<br />
Erromango and Efate, late Chachara bullet shaped pots were regarded as potentially non-<br />
homologous having “no apparent antecedents” among the poorly understood surface<br />
collected wares from the period between Malua and Chachara (Bedford 2000:147, 241).<br />
Accordingly, four regional Vanuatu phyletic series can be constructed from Bedford’s<br />
117
analysis, as follows:<br />
• Erromango: Dentate Lapita >>Ponamla Plain >> Early Ifo excurvate vessels with<br />
Fingernail impression >>Late Ifo incurving rims with fingernail impression and<br />
other decoration>>heavy notched plainware<br />
• Efate: Arapus Lapita-style plainware >> Early Erueti >> Late Erueti >>Early<br />
Mangaasi >> Late Mangaasi (Mangaasi possibly appearing on Malakula in surface<br />
collections also, although this is unconfirmed)<br />
• Malakula 1: Dentate Lapita >> Malua >> ? (small surface collected sample)<br />
• Malakula 2: Chachara >> Naamboi<br />
It is important, in reviewing Bedford’s data and constructing these four phyletic series,<br />
to question their historical completeness. While we know a lot more than we did<br />
previously about ceramic change in Vanuatu, do we know how well the regional ceramic<br />
changes within these four series are sampled? Bedford points out some blank spots in the<br />
Malakula record. Elsewhere, mixing <strong>of</strong> materials across strata may be a factor obscuring<br />
poorly-sampled periods. In Ponamla Area A, for example, age-depth correlation and<br />
“evidence for stylistic change....confirmed the stratigraphic integrity <strong>of</strong> the site” (Bedford<br />
2000:43), but stratigraphic unconformity (Bedford 2000: Figure 3.4) is best read as<br />
indicating that some excavation and possibly erosion <strong>of</strong> Layer 2 and 3a at least has<br />
occurred prior to deposition <strong>of</strong> Layer 1, incorporating materials originally deposited there<br />
into Layer 1, and that “stratigraphic integrity “ should not be read as absence <strong>of</strong> mixing,<br />
which Bedford confirms (Bedford 2002). Also, some <strong>of</strong> the dated Layer 1 material<br />
overlaps at two standard deviations with material from Layer 5, which is not to say that<br />
118
it is mixed to the point where stylistic change cannot be detected, but which should<br />
encourage a cautious approach to the question <strong>of</strong> whether gradual change with height in<br />
the column directly indicates gradual change in ceramic production style over time.<br />
Phyletic blank spaces in the series could be obscured by mixtures from different periods,<br />
especially when plainware phases exist, as on Erromango in the post-Lapita-phase<br />
(Ponamla plain). A degree <strong>of</strong> mixing might be hard to distinguish from the plain component<br />
<strong>of</strong> overlying Early Ifo, other than by vessel form (not possible in the Ponamla plain phase,<br />
in which vessel form seems little different to the succeeding early-Ifo vessel form). Having<br />
the two site sequences to compare has allowed a more complete sequence than would<br />
otherwise be the case, but is there yet more variety to be discovered in additional sites in<br />
the future? If so, how much more? Also, as mentioned above, it seems likely that Lapita<br />
face designs as found on New Caledonia and in the Southeast Solomons and Fiji will turn<br />
up eventually in most areas <strong>of</strong> Vanuatu.<br />
New Caledonia:<br />
<strong>The</strong> New Caledonian Lapita/Post-Lapita ceramic sequence has been elaborated recently<br />
by the work <strong>of</strong> Sand and colleagues, in which dentate-stamped pottery dates favour the<br />
range 1100BC to 800BC (Sand 1998:25-28) (Sand’s table <strong>of</strong> dates shows a broader<br />
absolute spread for Lapita contexts but he appears to have compressed this at an<br />
interpretive level). Sand sees these materials as allowing study <strong>of</strong> the decorative<br />
characteristics <strong>of</strong> the local Lapita motif inventory, with the aim <strong>of</strong> testing Kirch’s<br />
“Southern Lapita “ hypothesis. Sand sees a progressive replacement <strong>of</strong> Lapita ceramic<br />
ware by other incised or applied decorated ceramic traditions, and development from<br />
Lapita <strong>of</strong> the Podtanean paddle-impressed wares and Puen Mangaasi-like incised ware.<br />
Sand considered that some <strong>of</strong> the non-dentate pottery assemblages and non-dentate<br />
components <strong>of</strong> assemblages comprise a “Lapita-associated ceramic-series” within the<br />
119
Kone period (Sand 1999:143). He notes that these pots are found in spatial association<br />
with dentate-stamped Lapita pots, and in sites postdating Lapita. This category comprises<br />
triangular incised “non Lapita”, a range <strong>of</strong> shell-impressed styles, and paddle-impressed<br />
Podtanean pottery. While the presence <strong>of</strong> Podtanean in most Lapita sites, and in the<br />
earliest stratigraphic context at the Vatcha site, are taken by Galipaud to indicate<br />
contemporaneity with Lapita from first settlement, Sand suspects that this style <strong>of</strong> pottery<br />
in many cases represents later habitation than Lapita (Sand 2000:146). Galipaud sees the<br />
absence <strong>of</strong> Lapita in some Podtanean sites as a contemporaneous functional difference,<br />
with Podtanean being a utilitarian Lapita-age vessel type (Galipaud 1996:303). Sand has<br />
late/post Lapita impressed/incised wares in the north differentiating from incised Puen<br />
wares in the south, within his early Kone period (Sand 2000:155).<br />
Sand notes that ceramics excavated are still under analysis, but presents a<br />
preliminary synthesis <strong>of</strong> Kirch’s proposed “Southern Lapita Province”, defined by Kirch<br />
on the presence <strong>of</strong> the “Podtanean” paddle impressed wares in association with dentate-<br />
stamped Lapita ceramics (Kirch 1997:73). Sand proposes a short chronology for Lapita<br />
in this Southern province, beginning not earlier than 1100BC (Sand 1997a), and with<br />
dentate-stamping out <strong>of</strong> use by 750BC, a similar chronology to that recently proposed for<br />
the Eastern area <strong>of</strong> Fiji-western Polynesia. This is substantially shorter than the 1000 year<br />
chronology proposed for Lapita in the Bismarcks (Kirch 1997), although that chronology<br />
is in turn showing signs <strong>of</strong> contracting, with a figure <strong>of</strong> 500 to 600 years suggested more<br />
recently (Spriggs 2001).<br />
Sand saw “Southern Lapita” as having most direct links to the rest <strong>of</strong> the<br />
Melanesian chain, and having few links to less diverse eastern Lapita, which suggests that<br />
“southern” may be a misnomer for the more widespread characteristics, and “Central<br />
Lapita” might be a better term. Sand identified the major presence <strong>of</strong> composite rims on<br />
carinated pots as a Southern Lapita typological differentiation, but also noted a similar<br />
120
presence in Buka sites (Wickler 2001) (see comments above on Wickler’s identification<br />
<strong>of</strong> these as occurring on bowl vessel forms). <strong>The</strong>se composite rims will be shown to be a<br />
feature <strong>of</strong> one <strong>of</strong> the collection samples in the present study, so use <strong>of</strong> these as a defining<br />
characteristic <strong>of</strong> the Southern Lapita Series is significant to external comparison <strong>of</strong> the<br />
Roviana Lapita (see Chapter 13). Sand notes a varied set <strong>of</strong> tooth sizes in dentate-stamping<br />
from single sites, which contrasts with the Anson/Summerhayes “Early Far Western” fine-<br />
dentate hypothesis in the same way as Wickler’s Buka reef site data did. Sand notes that<br />
some parts <strong>of</strong> sites have almost all vessels decorated, while others have a relatively low<br />
percentage. He makes a distinction between geometric and anthropomorphic designs, and<br />
includes the stylized evolutions <strong>of</strong> anthropomorphic designs in the latter category. Of a<br />
range <strong>of</strong> complex geometric motifs, he describes one geometric motif as “characteristic <strong>of</strong><br />
the end <strong>of</strong> the Lapita period”. This motif is not illustrated, but is described as “the imprint<br />
<strong>of</strong> successions <strong>of</strong> straight lines, mostly done with a u-shaped tool”.<br />
Of the anthropomorphic motifs, he regards “double face motifs” as rare and found<br />
only in the earliest levels, and simpler than those found in the Western Lapita province. He<br />
notes that some face motifs (his “Type 3") are found in the Western and Far-Western<br />
provinces also, but not on cylinder stands and open pedestal bowls as in those provinces,<br />
rather on carinated pots, and occasionally plates with feet. In common with Spriggs<br />
(1990), Sand sees an evolution to an eye-nose-eye simplification, with final simplification<br />
being the loss <strong>of</strong> the eye, but does not support this with stratigraphic or other temporal<br />
evidence. In contrast, he states that this “evolved” motif is common from the earliest<br />
levels, indicating a foundation style or rapid local development, but regards these “stylized<br />
faces” as characteristic <strong>of</strong> the “Southern Lapita” dentate inventory. <strong>The</strong> rapid evolution<br />
from the “early” faces to their “later” derivatives in these New Caledonian sites puts the<br />
evolutionary seriations <strong>of</strong> Spriggs, Best and Ishimura under some pressure. Are the<br />
121
“earliest levels” in these sites representing substantial periods <strong>of</strong> time, <strong>of</strong> the order <strong>of</strong> a<br />
century or so, allowing this evolution to occur, or are the supposedly phyletic variants <strong>of</strong><br />
the faces contemporaneous? <strong>The</strong> way to rule out contemporaneity is to either find a<br />
sequence with the variants conveniently arranged in levels (Best does not seem to have<br />
evidence <strong>of</strong> this sort in his recent exposition)(Best 2002) or to seriate a good number <strong>of</strong><br />
large-sample low-diversify assemblages to create a fine-grained chronology, to test the<br />
chronology with some (preferably direct) dates, and to examine the relative positions <strong>of</strong><br />
the variants in the chronology. (Ishimura suggests that Frimiggacci has recorded stratified<br />
evidence for such change, but I have not had the opportunity to examine this).<br />
Non-dentate pots are also characteristic <strong>of</strong> the “earliest levels” (incised, shell-<br />
impressed, plain and paddle impressed). Incised decoration occurs on carinated pots, <strong>of</strong>ten<br />
with a notched rim. Rounded triangles in a frieze are common. <strong>The</strong>se motifs drop out “at<br />
the same time as the dentate-stamped motifs”, and do not occur later, supporting in my<br />
opinion a view <strong>of</strong> Lapita as a stylistic horizon rather than series on current published<br />
evidence. This opinion is given tentatively, in the absence <strong>of</strong> published detailed ceramic<br />
analyses from the recent New Caledonian excavations, but it should be noted that this is<br />
a major contrast to the series proposed for the Bismarcks by Summerhayes. Incised motifs<br />
are said to be strikingly less diverse than those in Western/Far Western Lapita. Shell<br />
impressions are uncommon but are present in all major Lapita sites. Notched lips or<br />
entirely plain pots are mostly carinated with excurvate rim, but some are simpler<br />
uncarinated forms.<br />
Paddle impressions are restricted to carinated forms with outcurved rims (although<br />
Sand’s “carination” as illustrated is a substantially wider-angle affair than that illustrated<br />
for dentate carinations). Paddle-impressed decoration is said to occur on better quality<br />
paste with thinner and harder pots. Paddle-impression occurs in low percentage in “early”<br />
sites; “...their major development takes place after the demise <strong>of</strong> the dentate-stamped and<br />
122
incised Lapita motifs....” Sand does not explicitly discuss the role <strong>of</strong> mixing and issues <strong>of</strong><br />
occupation span in the association between Podtanean and early dentate Lapita but, he<br />
gives examples where the dentate and Podtanean are separate and superposed, e.g. in<br />
rockshelter LWT008 <strong>of</strong> Hnajoisisi <strong>of</strong> Lifou island, while elsewhere intermediate forms are<br />
found. <strong>The</strong>re seems then to be still an open question <strong>of</strong> whether Podtenaean evolved<br />
gradually out <strong>of</strong> Lapita or whether the transition, <strong>of</strong> whatever kind, was relatively sudden,<br />
and blurred in general by mixed deposition and lengthy occupation spans.<br />
<strong>The</strong> View from the East: Fiji and Tonga:<br />
Best constructed an attribute-based Robinson coefficient similarity matrix seriation for the<br />
entire Fijian sequence based on evidence from stratified excavations on Lakeba,<br />
supplemented by other excavated and surface collected assemblages. For the Lapita period<br />
his data comprised the Natunuku and Naigani excavations, and the stratified Lakeba Site<br />
196 (an extensive and rich open site thought to have a stratigraphic sequence <strong>of</strong> extended<br />
duration with some temporal gaps and some mixing). Within the Lapita “style”,<br />
“...the number <strong>of</strong> vessel shapes reduce from 12 to six, carinated forms<br />
vanish, and decoration declines from initial fairly complex designs to simple<br />
arcs and zig-zags along the rims (Best 2002:17).”<br />
<strong>The</strong> chronological trend within the Lapita period noted by Best, the loss <strong>of</strong> carinated<br />
vessel forms, is not in evidence in Best’s data from site 196, where these occur in all levels<br />
<strong>of</strong> the area excavation squares (Best 1984:ppA3-A4). Strictly speaking, some carinated<br />
forms persist in the later levels <strong>of</strong> Lakeba site 196, and it is dentate-decorated sharp<br />
carinations which drop out rapidly from the lowest levels (Best 1984: table <strong>of</strong> attribute<br />
combinations for site 196).<br />
While the Lakeba excavations provide convincing evidence for a shift to arc-<br />
impressed or arc-stamped rims over time, Best’s data on Lapita motifs is not as clear-cut.<br />
A complicating factor making the argument difficult to abstract from the data is the use<br />
123
<strong>of</strong> two separate systems <strong>of</strong> motif description/classification: the Mead system and the<br />
Frost-Irwin system. Arc rims are coded in the Frost-Irwin manner as attribute<br />
combinations, while the Mead-system motif classes are omitted from the seriations.<br />
<strong>The</strong> Mead-system data from excavations at Lakeba site 196 is quantified by sherd<br />
count in an appendix, and does not show clear trends by level, beyond what might be<br />
expected from the motif sample sizes (Best 1984:A21-A22). Best stresses that if the counts<br />
<strong>of</strong> arc rims and plain sherds were added into these data a clearer trend would emerge<br />
(Best, personal communication 2002), but there is no clear motif shift by level using the<br />
Mead motif data alone.<br />
In his recent synthesis which has roulette stamping and “face” motifs early in the<br />
Lapita series, simpler Mead-type motifs incorporating design elements such as DE1 and<br />
DE2 (Best 1984) persisting from early Eastern Lapita to some sort <strong>of</strong> mid-Lapita, and arc-<br />
rims as the last stage <strong>of</strong> Lapita, the supporting data for the first two stages <strong>of</strong> this transition<br />
are seen in the similarity on the one hand <strong>of</strong> Natunuku and Naigani and the Reef/Santa-<br />
Cruz sites (with high frequency <strong>of</strong> “face” sherds and Mead Motif M.33), and on the other<br />
hand the Lakeba open site and most Tongan Lapita sites, being “late” with the complex<br />
anthropomorphic motifs not found or else explicable as heirloom effect.<br />
<strong>The</strong> low frequency <strong>of</strong> “early” motifs at Lakeba site 196 may be a sampling error.<br />
Although 2000 sherds were recovered from the site, from a scatter <strong>of</strong> 17 test pits and two<br />
groups <strong>of</strong> excavation squares each totaling 12m 2 (Best 1984:figure 2.16), the 15000m 2<br />
open site may hide considerable spatial diversity in the large unexcavated areas.<br />
Best’s suggestion in “a View from the East” (2002) that motif M12.2 is common<br />
at Lakeba site 196 and can be traced through a sequence <strong>of</strong> decreasing complexity if<br />
examples are contrasted with sherds from the Reef Islands is a phyletic or evolutionary<br />
seriation that is not supported stratigraphically, as the sherd from square 12 layer A1 <strong>of</strong><br />
site 196 has seven underlying occurrences <strong>of</strong> the same motif, mostly from the basal layers<br />
124
C and B3, none <strong>of</strong> which are illustrated in his “series”. Surely these sherds are relevant to<br />
the proposed series? Picking a phyletic series from the vast area which includes the Reef<br />
Islands and Fiji loses control <strong>of</strong> sources <strong>of</strong> variability, and becomes a re-statement <strong>of</strong> the<br />
definition <strong>of</strong> Lapita, that Lapita simplifies over time and becomes non-Lapita, rather than<br />
being a demonstration or confirmation <strong>of</strong> the way that change happened.<br />
While Best demonstrates convincingly that a specific carination form is present only<br />
in the lowest levels <strong>of</strong> Lakeba, and Best’s recent evolutionary simplification seriation may<br />
well be correct for Fiji, and even for the entire Lapita distribution, like Spriggs'<br />
formulation, it suffers from a circularity <strong>of</strong> reasoning. <strong>The</strong> direction <strong>of</strong> change is known<br />
and is the same everywhere for Lapita, therefore we can pick sherds out <strong>of</strong> sites and order<br />
them in a series, therefore we can order sites in time, and by doing this we can show that<br />
changes occur in concert across the Lapita distribution, therefore interaction is happening,<br />
as similar changes occur everywhere.<br />
Best’s proposal that particular Mead motifs can be used to date Lapita sites might,<br />
like Summerhayes’ model, be attributing the utilitarian component <strong>of</strong> sites to a late date,<br />
and the fancy component to early Lapita. Rounded carinations are present throughout the<br />
Lakeba strata, yet an excavation which finds the rounded carination form but not the sharp<br />
carination form with dentate is liable to be regarded as late (I am not referring here to the<br />
arc-rim post-Lapita decoration, which clearly supersedes Lapita at site 196).<br />
Lapita Temporal Variability -Conclusions:<br />
For West New Britain and Anir I have argued in detail above that Summerhayes’ chrono-<br />
stratigraphic decorative trends are problematic, and that higher-level inferences regarding<br />
Lapita interaction and a universal Lapita series need to accommodate these uncertainties.<br />
This critique was made possible by Summerhayes’ thorough sherds-as-vessels analysis,<br />
125
combined with Gosden and Webb’s geomorphological data. <strong>The</strong>se trends should not be<br />
regarded as confirming a regional West-New Britain gradual transition from dentate<br />
stamped decoration to incised, but should rather continue to have the status <strong>of</strong> hypothesis,<br />
in that they tell us little more than that at first there was Lapita and then there was<br />
something else. A correlation between vessel form and dentate-stamping has been<br />
demonstrated, but whether this varies temporally is less secure.<br />
Summerhayes’ motif-occurrence/motif sharing clustering/PCA comparisons with<br />
the wider Lapita world were dogged by a fundamental mismatch between the systematics<br />
<strong>of</strong> Anson's splitting classificatory approach (which results in sparse data <strong>of</strong> unknown<br />
representativeness), and the motif occurrence grouping methods used. Even if the<br />
dichotomy in the data is real, is the explanation temporal? Summerhayes’ recent<br />
chronological hygiene exercise for the Arawe and Anir dates is not conducted independent<br />
<strong>of</strong> the ceramic temporal hypothesis, and is not independent confirmation <strong>of</strong> the<br />
Early/Middle Lapita chronology in its present form, and is better regarded as a preferred<br />
explanation. Some late dates from “early” contexts were rejected because they did not fit<br />
the percieved pattern <strong>of</strong> temporal ceramic variability..<br />
Watom: the potentially broad age-span <strong>of</strong> materials from the upper C1 SAC<br />
excavation zone means that the recovered ceramics from zone CI and C2 are not <strong>of</strong> an<br />
accurately known age, Also, the spread <strong>of</strong> ages can be expected to differ by level at SAC,<br />
to an extent that suggests Watom does not yet provide a secure well-tested construct <strong>of</strong><br />
Lapita temporal variability against which Roviana variability can be compared.<br />
Mussau: Detailed results <strong>of</strong> ceramic analyses are as yet unpublished, and Kirch’s<br />
claim for stratigraphic evidence for a progression from plain to dentate to incised<br />
decoration is subject to similar C14 and formation-process related criticisms to those<br />
directed at Summerhayes’ Arawe analyses. Formation-based criticisms are principally that<br />
sub-tidal mixing <strong>of</strong> area B zone C may have occurred to some extent prior to burial, and<br />
126
that superposed incised pottery in Zone B may ultimately be derived from functionally or<br />
taphonomically differentiated areas <strong>of</strong> the sub-tidal deposit, cast up on the beach at a later<br />
date. Until detailed information on vessel joins, appropriateness <strong>of</strong> units <strong>of</strong> quantification<br />
etc is available, preliminary ceramic conclusions from Mussau do not provide a detailed<br />
construct <strong>of</strong> temporal variability to compare the Roviana ceramics with.<br />
Admiralties: Wahome’s seriation <strong>of</strong> a large number <strong>of</strong> Admiralties sites, including<br />
excavated and surface-collected units, provides a four-period chronology that on the<br />
surface avoids many <strong>of</strong> the pitfalls <strong>of</strong> seriation by attention to the statistical basics. <strong>The</strong> 700<br />
year gap suggested by Ambrose limits the utility <strong>of</strong> Wahome’s sequence for comparisons<br />
in the present context.<br />
Buka: Specht’s chronology from Buka to later styles, amplified by Wickler to<br />
include more Lapita variability as evidenced by three reef-flat sites (one from Nissan), and<br />
to tenuously link Lapita to Sohano, provided the nearest and most comprehensive ceramic<br />
chronology <strong>of</strong> relevance both temporally, geographically, and in terms <strong>of</strong> formation<br />
process, to the Roviana materials. His seriation <strong>of</strong> a small number <strong>of</strong> ceramic samples<br />
encompasses, in his view, functional variability between assemblages indicated by vessel<br />
form, and also possibly taphonomic variability in that three <strong>of</strong> his assemblages, from<br />
various parts <strong>of</strong> site DAF, represent taphonomic variability between reef edge, reef flat,<br />
and coastal-erosion scatter, the latter from what was regarded as formerly a terrestrial<br />
deposit. DES from Nissan falls into this latter category too, according to Spriggs (1991).<br />
Wickler’s seriation also faces the “number <strong>of</strong> sites” sample size problem (as does the<br />
current Roviana data) and suffers from the use <strong>of</strong> a ratio <strong>of</strong> plain:decorated, a measure<br />
easily biased by differences in brokenness.<br />
Wickler’s motif frequency grouping <strong>of</strong> Buka and other assemblages does not<br />
accord well with Summerhayes’ universal Lapita ceramic series. Despite Wickler’s<br />
hypothesis that the sites represented some time depth, those with reasonable sample sizes<br />
127
were much more similar to each other than to either Watom, the RSZ sites or Ile des Pins,<br />
suggesting a greater degree <strong>of</strong> regionalization than predicted by the Summerhayes model.<br />
Reef / Santa Cruz: use <strong>of</strong> Jaccard coefficient on sparse data <strong>of</strong> unknown<br />
representativeness/sample size (the latter resulting from choice <strong>of</strong> units <strong>of</strong> quantification)<br />
is a flaw in Green’s (1978) seriations. Spriggs, in his temporal hypothesis <strong>of</strong> the changing<br />
face <strong>of</strong> Lapita, suggested that a variety <strong>of</strong> face design types were present at both SZ-8 and<br />
RF-2, as did Ishimura, and this corresponds with my own reading <strong>of</strong> Donovan’s and<br />
Parker’s analyses, in which, in addition to the varied face designs, there seem to be<br />
incised/notched and other geometric linear Lapita variants not discussed by Spriggs, but<br />
potentially <strong>of</strong> chronological significance. Parker understood these sites as sequential slices<br />
in time, based primarily on the then radiocarbon evidence, as did Green in 1978, since<br />
revised (Green 1991c) and currently the subject <strong>of</strong> further dating research by Green and<br />
others. I have suggested that an alternative interpretation <strong>of</strong> these sites as roughly<br />
contemporaneous with variable, potentially overlapping occupation spans is consistent with<br />
the evidence. While C14 dates obtained so far from RF6 are younger than the other sites,<br />
I would think that at least some <strong>of</strong> the material in RF6 is <strong>of</strong> a similar age to SZ8/RF2,<br />
unless it can be demonstrated by C14 that such an overlap in occupation span is unlikely<br />
(the portion <strong>of</strong> the RF2 sample from the northern half <strong>of</strong> that site may be key here). Best’s<br />
suggestion that the site order should be reversed (Best 2002:93) is in my view not<br />
supported by current evidence, and also involves a sequential slice-in-time view <strong>of</strong> the<br />
sites.<br />
<strong>The</strong> evidence from Vanuatu suggests that rather than a Lapita ceramic series,<br />
evidence for ceramic variability does not presently extend beyond the general features <strong>of</strong><br />
the transition from Lapita to post-Lapita, and in this latter period diverse regional<br />
sequences are seen. <strong>The</strong> samples <strong>of</strong> Lapita recovered by recent testpitting reconnaissance<br />
and excavation are too small to allow a fine-grained temporal construction within<br />
128
Bedford’s Lapita phase. It is tempting to see parallels between some Erromango ceramics<br />
and some <strong>of</strong> the Roviana materials but these may be unrelated co-incidence, as will be<br />
discussed further in a Chapter 13 (see also Felgate 2002).<br />
New Caledonia is emerging as a region in which a process <strong>of</strong> Lapita-phase<br />
sequence building is made possible by a large sample <strong>of</strong> Lapita sites, and the only proviso<br />
here is that the descriptive records published to date lack the detail <strong>of</strong>, for example,<br />
Summerhayes’ Arawe sherd analyses. Comparisons can be usefully drawn between New<br />
Caledonian Lapita and Roviana ceramics from Honiavasa, as predicted by Sand’s<br />
comments regarding similarities <strong>of</strong> New Caledonian Lapita to Buka Lapita.<br />
This review <strong>of</strong> the Lapita ceramic series concludes that the regional Lapita ceramic<br />
series from New Britain, Mussau, Buka, and the Reef-Santa-Cruz sites should be<br />
characterized as untested hypotheses rather than securely established chronological<br />
frameworks, and that Lapita still constitutes a ceramic stylistic horizon rather than a series<br />
in the Western/Far Western region at least, for practical purposes <strong>of</strong> external comparisons<br />
<strong>of</strong> the Roviana materials. Publication <strong>of</strong> details <strong>of</strong> the Mussau ceramics would probably<br />
revise this assessment. <strong>The</strong> best samples for external comparisons are the Buka reef-site<br />
samples, since natural formation processes are so similar (wave processes are dominant).<br />
<strong>The</strong> terrestrial samples and buried stilt house samples are problematic in this respect, but<br />
also suffer from chronological uncertainties which make comparisons difficult. Aspects <strong>of</strong><br />
the Roviana ceramics bring to mind characteristics <strong>of</strong> Southern Lapita, and characteristics<br />
<strong>of</strong> Summerhayes’ early Lapita, but decoration is “Late” if Roviana Lapita is slotted into<br />
the Summerhayes model. <strong>The</strong>se issues will be covered in some detail in making external<br />
comparisons in Chapter 13; it is simply noted at this stage that some difficulties arise in<br />
trying to slot the Roviana pottery into any existing “Lapita ceramic series”. <strong>The</strong>se<br />
difficulties may arise from errors in the current formulations <strong>of</strong> those series rather than in<br />
idiosyncratic variability in the Roviana case.<br />
<strong>The</strong> impression given by the authors <strong>of</strong> many <strong>of</strong> these regional sequences (with the<br />
129
exception <strong>of</strong> Anson (1983) and Wickler (2001) is that these are all confirmation <strong>of</strong> an<br />
expected Lapita trend from dentate to incised/applied, and <strong>of</strong> the temporal primacy <strong>of</strong> the<br />
“Far Western” sites (Kirch mentions early plainware at Mussau recently though). <strong>The</strong>re is<br />
a circularity in this reasoning stemming from a tendency to recognize evidence for the null<br />
hypothesis, that Lapita became non-Lapita, as evidence for the details <strong>of</strong> the transition.<br />
I conclude specifically that some attributes <strong>of</strong> pottery samples are particularly likely<br />
to mislead in this regard, for example, using ratio <strong>of</strong> plain to decorated pottery, quantified<br />
by sherd count, in seriations. Smash the pottery up and it gets younger. More robust<br />
properties <strong>of</strong> ceramic samples, such as the number <strong>of</strong> shared motifs, are sensitive to<br />
sample-size differences, particularly when units <strong>of</strong> motif classification result in sparse data,<br />
requiring careful attention to sample evaluation, selection <strong>of</strong> units <strong>of</strong> quantification and<br />
method <strong>of</strong> analysis.<br />
This assessment <strong>of</strong> the nature <strong>of</strong> our current regional samples and the stability <strong>of</strong><br />
existing temporal constructs is not an argument against stratigraphic excavation, nor is it<br />
a condemnation <strong>of</strong> the value <strong>of</strong> ceramic chronology in general, or an attempt to denigrate<br />
the work <strong>of</strong> others. It is an assessment principally <strong>of</strong> the level <strong>of</strong> temporal resolution <strong>of</strong><br />
secure regional chronologies. I argue that while outlines <strong>of</strong> a culture-history are stable now<br />
for several island groups, we are running up against the low temporal resolution <strong>of</strong><br />
reconnaissance sampling and C14 dating-by-association. Research design <strong>of</strong> the Lapita<br />
Homeland Project focused on the identification and excavation <strong>of</strong> stratified sequences, as<br />
has recent work on Vanuatu. <strong>The</strong>re is nothing wrong with that, as long as it is accepted<br />
that it is difficult to assess within-site spatial structure, particularly for the Lapita period.<br />
<strong>The</strong>re is a dearth <strong>of</strong> large, representative samples <strong>of</strong> Lapita from Near Oceania,<br />
which means seriation studies, which are key to increasing the temporal resolution <strong>of</strong><br />
chronologies, are not succeeding as well as they might were better samples available. <strong>The</strong><br />
difficulty <strong>of</strong> obtaining fine-resolution stratified sequences may well be compounded by<br />
Lapita and post-Lapita settlement pattern instability, and low-deposition life ways<br />
130
compared to the archaeology <strong>of</strong> the Near East or American Southwest (post and thatch<br />
in an environment <strong>of</strong> rapid organic weathering rather than mud brick houses in a dry<br />
climate) leading to a dearth <strong>of</strong> well-stratified records <strong>of</strong> change unless there is a regular<br />
natural airfall tephra supply, as in much <strong>of</strong> Vanuatu.<br />
Our archaeological methods have yet to adapt to low-deposition conditions by<br />
focussing on the horizontal component <strong>of</strong> temporal patterning, rather than expecting a<br />
temporal layercake in settlements. Beyond deconstructing Summerhayes’s current<br />
universal Lapita Ceramic Series, this review identifies a need to search out and analyse<br />
archaeological distributions in horizontal space to a greater extent than has been the norm<br />
for Lapita in Near Oceania to date. <strong>The</strong> recovery and seriation <strong>of</strong> a large number <strong>of</strong> good<br />
short-term samples is key to boosting ceramic time-resolution. C14 dating, in this view,<br />
can be used as a confirmatory technique rather than the primary means by which ceramic<br />
chronology is constructed, especially with the rise <strong>of</strong> direct-dating using AMS.<br />
131
132
CHAPTER 3:<br />
SCALE AND METHOD OF FIELD SURVEY<br />
REQUIRED TO GENERATE A SAMPLE OF<br />
LAPITA SITES FOR SERIATION IN THE NEW<br />
Introduction:<br />
GEORGIA REGION<br />
It is traditional to preface the analytical portions <strong>of</strong> PhD theses with a <strong>chapter</strong> on the sites<br />
and their setting. Here, I try rather to take a problem-oriented analytical approach to the<br />
archaeological landscape <strong>of</strong> Near Oceania, in the course <strong>of</strong> which I hope I make clear the<br />
location and nature <strong>of</strong> surface collections used in other analytical <strong>chapter</strong>s. Detailed spatial<br />
analyses <strong>of</strong> sites are provided in Chapter 11.<br />
While the claim could be made that the Roviana intertidal survey data were the<br />
result <strong>of</strong> careful and systematic research design from the outset <strong>of</strong> the project, the prosaic<br />
version <strong>of</strong> events is that research design occurred as a series <strong>of</strong> ideas arising before and<br />
during the fieldwork. This is particularly true <strong>of</strong> the survey design, which began life as a<br />
search for pottery sites primarily through local knowledge (hereafter referred to a the<br />
informant-prospection method), and evolved to a more systematic inspection <strong>of</strong> the<br />
intertidal and subtidal zones once it became apparent that these were the locus <strong>of</strong> early<br />
ceramic evidence. Initial aims were simply site discovery. A concern with site visibility,<br />
survey coverage, site preservation and distributional patterning crystallised in the later<br />
stages <strong>of</strong> fieldwork, during the 1997-8 field seasons, when survey was focused in the<br />
Kaliquongu region, with the aim <strong>of</strong> characterization <strong>of</strong> the archaeological distribution in<br />
the intertidal zone. <strong>The</strong>se two distinct stages or types <strong>of</strong> survey will be referred to as the<br />
133
Roviana survey and the Kaliquongu survey respectively. During writing up, I saw a need<br />
to phrase the information in terms <strong>of</strong> an estimation approach and hypothesis testing, largely<br />
in order to demonstrate the limitations <strong>of</strong> the data in this regard, and to generate<br />
conclusions for future practice.<br />
While the following <strong>chapter</strong> might well be characterized as more and more about<br />
less and less, which in terms <strong>of</strong> the size <strong>of</strong> the region it is, this is not a bad thing; the results<br />
<strong>of</strong> iterative survey at an increasing level <strong>of</strong> detail are informative, and raise fundamental<br />
epistemological issues for the early ceramic period in Near-Oceania. <strong>The</strong> foremost <strong>of</strong> these<br />
concerns sampling at the level <strong>of</strong> regional survey. <strong>The</strong> Roviana Lagoon surveys are<br />
presented below within a theoretical framework <strong>of</strong> sample-surveying. <strong>The</strong> objectives <strong>of</strong> the<br />
surveys are first to test the hypothesis that early-Lapita was present in the New Georgia<br />
raised coral barrier-island/lagoon system in the past, second, to estimate site density for<br />
Lapita and post-Lapita ceramic-lithic intertidal scatters (using a ceramic chronology which<br />
will be justified in detail in succeeding <strong>chapter</strong>s), and third, to draw methodological<br />
conclusions about the scale <strong>of</strong> survey needed to obtain a sufficient site sample for seriation<br />
<strong>of</strong> Lapita sites. <strong>The</strong> characteristic <strong>of</strong> the local archaeological record that allows this<br />
approach is good site visibility on coralline lagoon shorelines in the intertidal zone. <strong>The</strong><br />
level <strong>of</strong> preservation evident in samples, implications for survey and analysis <strong>of</strong> these<br />
results, and what the collection sites represent in terms <strong>of</strong> settlements, etc are the subjects<br />
<strong>of</strong> the following three <strong>chapter</strong>s.<br />
Review <strong>of</strong> Landscape/settlement Pattern Studies in Oceania:<br />
Through the 1970s to 1990s a number <strong>of</strong> settlement-pattern studies were published, which<br />
involved coarse ceramic periodization and an implicit commitment to synonymy <strong>of</strong> “site”<br />
and “settlement”(Anderson 2002, Best 1984, Burley 1994, Hedrick n.d, Irwin 1972, 1985,<br />
Kirch 1988a, Lep<strong>of</strong>sky 1988, Sand 1997b, Swadling 1976). More recently, landscape<br />
approaches have been advocated, in which<br />
“structures <strong>of</strong> reference are made manifest not only through sites, but<br />
134
through broadly distributed scatters <strong>of</strong> material <strong>of</strong> the sort picked up by socalled<br />
“<strong>of</strong>f-site” approaches....understanding the overall distribution <strong>of</strong><br />
materials is dependent in turn on understanding the changes to which the<br />
landscape has been subject (Gosden & Head 1994)”.<br />
Approaches <strong>of</strong> this sort to surface scatters, in which formation processes, archaeological<br />
visibility and site preservation are controlled for have not been applied to the archaeology<br />
<strong>of</strong> Lapita pottery. This aspatial tendency <strong>of</strong> Lapita archaeological method, where the site,<br />
the layer assemblage, or even the test pit tends to be the primary unit <strong>of</strong> analysis, at the<br />
expense <strong>of</strong> regional distributional analysis (see Clark 1999:455 for a recent example)<br />
requires further discussion, as the Kaliquongu survey described below stands in contrast<br />
to this tendency. In near-Oceanic Lapita studies, much effort has been directed at finding<br />
what has been described in another context as,<br />
“<strong>The</strong> optimal solution to a culture-historical puzzle...a single point where<br />
all <strong>of</strong> a region’s typological complexes could be found on top <strong>of</strong> each<br />
other, separated by sterile deposits.” (Wobst 1983:42)<br />
In archaeology elsewhere prior to 1960 this emphasis worked together with a large-site<br />
bias to divert attention from issues <strong>of</strong> sampling and estimation, and limited archaeological<br />
survey to a prospecting method (Wobst 1983), and this characterization would seem to<br />
hold in near-Oceania also, particularly for the Lapita Homeland Project, but also for some<br />
more recent studies (Bedford 2000, Clark 1999).<br />
Wobst noted that seriation studies, in contrast to culture-historical excavation<br />
approaches, led to a search for short-term occupation sites within short distances <strong>of</strong> each<br />
other, which necessarily increased concern for intrasite distributions. Examples <strong>of</strong> this<br />
sort <strong>of</strong> approach to survey can be found in the works <strong>of</strong> Specht and also Irwin (Irwin 1972,<br />
1985, Specht 1969). Wobst criticised culture historical approaches as lacking an explicit<br />
rationale for excavation and survey, and for an inability to evaluate the accuracy,<br />
precision, and consistency <strong>of</strong> their results, and Gosden expressed similar sentiments in<br />
relation to Lapita in the Bismarck archipelago (Gosden 1991a). <strong>The</strong> recent use <strong>of</strong> site<br />
density data by Anderson in the course <strong>of</strong> assessing Lapita mobility suffers a similar fatal<br />
135
flaw, unmitigated by the argument that,<br />
“...the broad pattern <strong>of</strong> current distributional data has remained much the<br />
same for more than 20 years, suggesting that it is not just on the grounds<br />
<strong>of</strong> relative site recording intensity that the density <strong>of</strong> sites on the coasts <strong>of</strong><br />
large islands is lower than on small islands throughout the Lapita range, or<br />
that mainland New Guinea, and perhaps the main Solomon Islands, are<br />
largely bereft <strong>of</strong> Lapita sites (especially early Lapita sites) (Anderson<br />
2002).”<br />
Evidence <strong>of</strong> absence does not accumulate by default, nor does absence <strong>of</strong> evidence<br />
transform into evidence <strong>of</strong> absence with the passage <strong>of</strong> time (see Felgate 2002 for a<br />
discussion <strong>of</strong> the epistemology <strong>of</strong> the Lapita distribution in Near Oceania). It remains to<br />
be demonstrated that the recorded density <strong>of</strong> Lapita sites is a preserved settlement pattern.<br />
I think this is highly unlikely, and showing why this is so is a major component <strong>of</strong> this<br />
thesis.<br />
Early approaches to archaeology as sampling have been described as “...something<br />
<strong>of</strong> a programmatic reaction to culture history rather than well justified applications to<br />
resolve important research questions.”(Wobst 1983:43) Between 1960 and 1975 Binford’s<br />
research design/sampling approach was applied predominantly in regions where there was<br />
excellent archaeological visibility on the surface (Wobst 1983:47). A feature <strong>of</strong> most such<br />
studies was stratified sampling using existing knowledge, and multi-stage survey strategies,<br />
although objectives seldom extended to specifying the wider ramifications for the<br />
distribution beyond the sample region.(Wobst 1983: 47-48). <strong>The</strong> introduction <strong>of</strong><br />
hypothesis-testing survey objectives in the mid 70s (for example Cowgill 1975:261)<br />
marked a further advance in sampling research design, but such developments were still<br />
confined to regions with good archaeological visibility. Wobst notes a trend since 1976 to<br />
extend the new methods to areas with poor surface visibility and a consequent emphasis<br />
on subsurface prospection techniques for such regions.<br />
Seen in the context <strong>of</strong> Wobst’s analysis <strong>of</strong> the development <strong>of</strong> sampling<br />
approaches to regional archaeology, a likely explanation for aspatial approaches to the<br />
study <strong>of</strong> the regional distribution <strong>of</strong> Lapita is the environmental context <strong>of</strong> the Lapita<br />
136
Homeland Project, the major contributor to the Lapita knowledge-base in near-Oceania.<br />
Holocene rhyolitic volcanism and tectonic instability in the Bismarck archipelago has<br />
created a landscape in which few archaeologists would have the temerity to claim the<br />
pristine preservation and visibility <strong>of</strong> Lapita social landscapes. Lapita coastal site<br />
distributions in this dynamic landscape are expected to be subject to a high degree <strong>of</strong><br />
obliteration or burial. Anderson points out that most Lapita sites are coastal, without any<br />
significant discussion <strong>of</strong> the impact <strong>of</strong> coastal geomorphological processes on site<br />
preservation and visibility. It cannot be merely<br />
“...accepted either that the archaeological patterning is approximately<br />
representative, or that proposition may be taken simply as a working<br />
hypothesis (Anderson 2002:16).”<br />
If the record is simply accepted as representative <strong>of</strong> a phenomenon as a working<br />
hypothesis, there can be no assessment <strong>of</strong> the confidence we can have in higher-level<br />
hypotheses built on that data. Resort to a recursive array <strong>of</strong> contingent untested hypotheses<br />
renders archaeology a continual tabula rasa for the idiosyncratic rewriting <strong>of</strong> higher level<br />
theory.<br />
<strong>The</strong> sampling sophistication necessitated by areas with poor preservation and/or<br />
poor surface visibility has been slow to develop in Lapita archaeology, and while it has<br />
been theorized to some extent (Gosden 1991b), the requisite sampling sophistication that<br />
would allow distributional analyses <strong>of</strong> Lapita is largely absent from existing treatments <strong>of</strong><br />
the data and approaches to the archaeology <strong>of</strong> the region and period. Thus the Binfordian<br />
exhortation that archaeology should promote the region to the position <strong>of</strong> the primary unit<br />
<strong>of</strong> archaeological research (Binford 1964) with explicit research design and explicit use <strong>of</strong><br />
sampling theory, has largely fallen on stony (tephra covered?) ground in near-Oceania,<br />
where the site has remained the primary unit <strong>of</strong> analysis in most cases.<br />
A notable exception to this generalization is the landscape approach taken on<br />
Garua Island in the Bismarck Archipelago (Torrence & Stevenson 2000). Here the<br />
approach to the landscape sought explicitly to abandon the coastal bias and occupation-<br />
137
site-prospecting objectives <strong>of</strong> previous work in the Bismarcks. In spite <strong>of</strong> a comprehensive<br />
series <strong>of</strong> C14 and obsidian hydration dates on Garua materials <strong>of</strong> roughly Lapita age, the<br />
study faces difficulties associating ages <strong>of</strong> dated materials with the pottery styles found<br />
with them, and the resulting chronological inferences are potentially spreading the span <strong>of</strong><br />
Lapita production/discard rather than refining a ceramic chronology. <strong>The</strong> Garua ceramic<br />
sample would seem to hold considerable potential for a program <strong>of</strong> direct dating <strong>of</strong> local<br />
pottery production styles. <strong>The</strong> focus on inland locations also initially missed significant<br />
Lapita pottery deposits in the sea nearby, but these project-specific problems do not<br />
detract from the value <strong>of</strong> such landscape-based approaches. In Remote Oceania there has<br />
been some effort put into intensive systematic survey <strong>of</strong> landscapes (e.g. Best 1984, Kirch<br />
1988a, Rogers 1973). In the case <strong>of</strong> Niuatoputapu this approach was favoured by<br />
concentrated effort over a number <strong>of</strong> years by two research teams on an island <strong>of</strong> only<br />
15km 2 .<br />
<strong>The</strong> current study seeks method by which the regional distribution <strong>of</strong> Lapita-age<br />
archaeological materials can be established and interpreted in the Roviana<br />
geomorphological setting. While explicit research design in terms <strong>of</strong> sampling theory was<br />
not a feature <strong>of</strong> the 1996 Roviana survey (Sheppard 1996, Sheppard et al. 1999), and data<br />
collected reflect this, both the Roviana and Kaliquongu surveys can pr<strong>of</strong>itably be phrased<br />
in terms <strong>of</strong> sampling theory when assessing the significance <strong>of</strong> results, since they clearly<br />
are samples <strong>of</strong> some sort.<br />
Intensity <strong>of</strong> Survey:<br />
Wobst found in a study <strong>of</strong> archaeologists that they tended to sample a larger landscape less<br />
intensively regardless <strong>of</strong> available resources when asked to design a survey (Wobst<br />
1983:61). A similar tendency is proposed for Near-Oceanian archaeology, although in this<br />
case resources have clearly been limited. Faced with large terrestrial regions, surveys have<br />
tended to try to cover wide regions (see for example Allen et al. 1984) in an opportunistic<br />
manner, relying largely on informant prospection as a site-discovery technique, although<br />
138
this is not universal (as the example <strong>of</strong> Torrence & Stevenson 2000 shows). <strong>The</strong> following<br />
section reviews the regional survey literature in the light <strong>of</strong> Wobst’s observation.<br />
Review <strong>of</strong> Near-oceanic Survey Methods and Results:<br />
Objectives <strong>of</strong> the Review:<br />
One approach to archaeological sampling is to ask how much work needs to be done to<br />
find out what we want to know (Orton 2000:5). By taking a sampling approach to the<br />
Roviana Lagoon early-ceramic survey, the interest is not so much how much survey is<br />
needed, but, given the amount <strong>of</strong> survey that was done, what is the significance <strong>of</strong> the<br />
results towards answering the major research questions? <strong>The</strong> following review <strong>of</strong> Near-<br />
Oceanic survey objectives, methods and results had these aims:<br />
• to critically assess existing knowledge (pertaining to Lapita site survey method,<br />
recorded site density and site preservation/detectability)<br />
• to arrive at a characterization <strong>of</strong> existing knowledge in terms <strong>of</strong> values<br />
(observations) <strong>of</strong> a variable (site density) measured on objects (survey regions)<br />
To do this requires the reworking <strong>of</strong> existing information to arrive at the values <strong>of</strong> this<br />
variable (site density). This is necessarily imprecise in some instances due to the lack <strong>of</strong><br />
specific detail in reports, the need for a definition <strong>of</strong> what constitutes a site, and the unit<br />
<strong>of</strong> measurement (sites per unit area or per unit coastline length).<br />
Method <strong>of</strong> this Review:<br />
<strong>The</strong> literature on archaeological surveys in near-Oceania was examined for information on<br />
survey objectives, methods and results. <strong>The</strong> coastal extent <strong>of</strong> surveys was reconstructed<br />
as far as possible from the literature available. <strong>The</strong> unit <strong>of</strong> measurement used was site<br />
density per km <strong>of</strong> coastline, which poses a problem <strong>of</strong> scale, since coastlines become more<br />
indented at larger map scales. Distances were calculated using published maps <strong>of</strong> survey<br />
139
egions in conjunction with descriptions <strong>of</strong> surveys, using dividers at a scale <strong>of</strong> either 1cm<br />
=1km or 1cm=5km to “walk” the coastline. This method at the latter scale can be expected<br />
to under represent the length <strong>of</strong> indented coastlines relative to the former. <strong>The</strong> effect <strong>of</strong> this<br />
on the conclusions is discussed. Since almost none <strong>of</strong> the studies supply a definition <strong>of</strong><br />
what constitutes a site, structure in the resulting data needs to be discussed in these terms<br />
too. Ideally, we should seek measures <strong>of</strong> intensity <strong>of</strong> use in behavioural measurement<br />
devices such as estimates <strong>of</strong> discarded ceramic vessel quantities rather than site counts, but<br />
there is no data <strong>of</strong> this sort currently available.<br />
At the broader regional scale <strong>of</strong> the Lapita era in near-Oceania, existing knowledge<br />
includes the archaeological record from early colonial missionaries and <strong>of</strong>ficials (e.g. Father<br />
Otto Meyer), Specht’s work at Watom and Buka in the late 60s, the Southeast Solomons<br />
Culture History project and <strong>of</strong>fshoots, Chikamori’s tours, the Lapita Homeland Project and<br />
descendants (e.g. Kirch on Mussau and Summerhayes on Anir), Terrell’s Bougainville<br />
research (Terrell 1976), Irwin’s Shortlands survey, David Roe’s thesis research on<br />
Guadalcanal (Roe 1993), the Solomon Islands National Museum Site recording scheme<br />
(e.g. Miller & Roe 1982, Reeve 1989) (and equivalents in PNG, which have not been<br />
accessed). Of these, some surveys provide little information <strong>of</strong> relevance, and have been<br />
omitted or only briefly discussed. Others are unpublished or difficult to access, so the data<br />
presented below are incomplete, but are data nonetheless. Recent work on Garua is<br />
omitted from this quantitative review, as “site” densities in that case can be expected to far<br />
exceed those recorded elsewhere due to a combination <strong>of</strong> a well preserved buried<br />
landscape <strong>of</strong> Lapita age and a more intensive programme <strong>of</strong> “<strong>of</strong>f-site” testpitting than has<br />
been conducted elsewhere.<br />
What distributional information regarding site density can be extracted from these<br />
field surveys? While none are prepared to consider the results <strong>of</strong> their surveys as unaltered<br />
and contemporaneous settlement patterns, an approach to this problem adopted here is to<br />
look at the range <strong>of</strong> density values for the Lapita period, assuming that many <strong>of</strong> the lower<br />
site densities will reflect low rates <strong>of</strong> site preservation or visibility (together, these create<br />
140
low probability <strong>of</strong> detection), or low-intensity survey coverage/bad luck, and thus low<br />
detection rate rather than absence <strong>of</strong> sites in the past. This leaves a problem in that some<br />
<strong>of</strong> the lower site densities may be an accurate reflection <strong>of</strong> the intensity <strong>of</strong> Lapita<br />
settlement in the past, but discriminating between absent, never present, and absent, not<br />
found is a problem that has to be temporarily sidestepped given the embryonic state <strong>of</strong><br />
existing knowledge. What the data can tell us is roughly what the maximum recorded site<br />
densities are for coarse periods.<br />
In addition to coastline length, a secondary measure <strong>of</strong> survey region area<br />
employed is simply to calculate the area <strong>of</strong> a rectangle on a map roughly encompassing the<br />
area surveyed, subsuming land, sea, coastline length, arable land, reef resources, and any<br />
other factors likely to be significant for site location frequency within this gross<br />
measurement. Density was calculated by dividing the number <strong>of</strong> Lapita sites found by the<br />
number <strong>of</strong> kilometres <strong>of</strong> coastline surveyed, as far as could be determined from the reports.<br />
<strong>The</strong> definition <strong>of</strong> Lapita sites was largely left as per the report, although in the case <strong>of</strong><br />
Nissan (Spriggs 1991) the Halika phase site (Lapita without pots) was not counted as the<br />
review was limited to early ceramic sites where the characteristics <strong>of</strong> the ceramics was the<br />
primary means <strong>of</strong> assignment to the Lapita period. Survey methods were characterized in<br />
terms <strong>of</strong> the following criteria:<br />
• whether survey was predominantly through the informant-prospection method,<br />
expected to create a late-site bias,<br />
• or whether more systematic searching was conducted, expected to yield results<br />
with less <strong>of</strong> a time-bias;<br />
• whether systematic terrestrial pedestrian coastal searches were made, and/or<br />
whether intertidal survey was conducted.<br />
• A further descriptor was added regarding the intensity <strong>of</strong> survey, high, moderate<br />
or low, which was arrived at largely through a subjective assessment <strong>of</strong> details<br />
supplied in the reports on the amount <strong>of</strong> time spent in the survey region.<br />
141
1984 Lapita Homelands Project Survey:<br />
<strong>The</strong> primary aim <strong>of</strong> the 1984 reconnaissance was to explain the aims and nature <strong>of</strong> the<br />
proposed research and to seek reaction to it. A secondary aim was to locate regions to<br />
form base locations for individual projects in the future: there was no aim to carry out full<br />
and extensive surveys <strong>of</strong> such localities (Allen et al. 1984:1). Nevertheless, site surveys<br />
were carried out in Manus, the Mussau Islands and along both coasts <strong>of</strong> New Ireland,<br />
centred at Manggai/Lemakot, Lasigi and Hilalon on the east coast, and Limpos on the west<br />
coast. Site surveys were carried out also in the Jaquinot Bay area <strong>of</strong> East New Britain<br />
(Allen et al. 1984:3,6-7).<br />
Specht and Allen spent seven days in the Mussau area, searching principally the<br />
small islands <strong>of</strong> Eloaua and Emananus. Fourteen <strong>of</strong> sixteen new sites were recorded on<br />
these small islands. Of the two sites on the main island, one was reported by the provincial<br />
member, while the other was spotted during a brief visit to Lomakuauru village for<br />
purposes <strong>of</strong> consultation. <strong>The</strong>se fourteen sites on the small islands comprised Lapita<br />
middens, other or undiagnostic pottery-bearing middens, and aceramic middens. This seven<br />
day survey <strong>of</strong> roughly thirty square kilometres (land and sea, my calculation) must rate as<br />
one <strong>of</strong> the more intensive in Near-Oceania. An approximate measurement <strong>of</strong> coastline<br />
length from the published map (Kirch’s section in Allen et al. 1984:146) yielded a<br />
measurement <strong>of</strong> thirty-three kilometres, or a recorded Lapita site density <strong>of</strong> 0.09 sites per<br />
km.<br />
<strong>The</strong> 1984 New Ireland survey was more in the nature <strong>of</strong> a tour, with heavy<br />
emphasis on cave sites. <strong>The</strong> Lemau pottery-bearing mounds are assumed to have been<br />
discovered in passing by road, rather than by pedestrian survey, attention being drawn to<br />
these high-visibility features by the favourable location <strong>of</strong> this village on a deepwater bay<br />
(Allen et al. 1984:17-18).<br />
Specht, Ambrose and Allen spent five days in the Pomio-Palamal region <strong>of</strong><br />
Jaquinot Bay, and found the area poor in archaeological resources despite good shelter in<br />
the bay from monsoon-season seas. No pottery sherds were seen in this time. <strong>The</strong><br />
142
possibility <strong>of</strong> poor preservation as a result <strong>of</strong> high local rainfall was noted (5-6000mm per<br />
annum). <strong>The</strong>ir assessment <strong>of</strong> the area as potentially having a low level <strong>of</strong> detectability<br />
means this negative result should be disregarded for the present purposes.<br />
<strong>The</strong> Talasea area was not re-surveyed in 1984, as Specht had surveyed previously<br />
(Specht 1974). Specht and Kamminga had surveyed a 60km stretch <strong>of</strong> coast on the north<br />
coast <strong>of</strong> the Huon peninsula in 1972, without finding Lapita pottery or obsidian. <strong>The</strong><br />
discovery in 1973 at Talasea <strong>of</strong> nine coastal ceramic sites, including three identifiable<br />
Lapita sites was therefore notable at the time, as this seemed to Specht to add weight to<br />
an interpretation <strong>of</strong> an Oceanic distribution <strong>of</strong> Lapita, extending close to, but not to, the<br />
north New Guinea coast.<br />
Specht notes a total <strong>of</strong> 15-20 sites with some dentate sherds, but only three or four<br />
sites with large quantities <strong>of</strong> dentate sherds, in a total coastal survey region <strong>of</strong> 25km,<br />
surveyed by a mix <strong>of</strong> intensive pedestrian cover and informant prospection (the site at<br />
Pasiloke was found by the latter method) (Specht, pers. comm. 2001). Coastal Talasea<br />
Lapita sites were located in some cases in the intertidal (for example the FDK obsidian<br />
source and Lapita findspot (Specht 1981), and in others on the elevated reef shore platform<br />
slightly above sea level, and eroding into the sea. Major disturbance by bulldozers for<br />
roading gravel seems to have been a factor in the discovery <strong>of</strong> at least site FCS (Specht<br />
1974). Four sites in 25km yields a Lapita distribution <strong>of</strong> 0.16 sites per km, or if the larger<br />
total is used, the site density is 20 ÷ 25 = 0.8 sites per km.<br />
<strong>The</strong> 1984 party revisited previously reported <strong>of</strong>fshore sites at Boduna and Garala<br />
islands, and mention water-rolled pottery above the high water mark on islands then<br />
thought to be sinking, suggested either a complex recent sea-level history or storm<br />
deposits relocating materials from now-submerged sites. Another explanation not<br />
considered in the report might be an intertidal distribution <strong>of</strong> settlement in the past. Large,<br />
volcanic-tempered sherds were present in the sea at Boduna and Garala islands, and were<br />
notable for their strength compared to the s<strong>of</strong>t Mussau sherds which in contrast were<br />
exposed on land and calcareous tempered (Allen 1984:19-20, Ambrose & Gosden 1991).<br />
143
Recent collection <strong>of</strong> ceramics from the reef flat on Boduna (Torrence & White 2002,<br />
White et al. 2002), separated from Talasea by a 1km channel, heightens the likelihood <strong>of</strong><br />
an intertidal component <strong>of</strong> the archaeological distribution in the Talasea area, rather like<br />
the situation at Buka, where both terrestrial and intertidal ceramic sites are found.<br />
It has been pointed out that there seems to be ample evidence that Lapita<br />
settlement pattern at Talasea differs from the stilt-house patterning noted by Gosden for<br />
the Arawe Islands (Specht et al. 1991). Specht notes evidence for massive landscape<br />
changes in the Willaumez region, and also suggests that high site density is related to<br />
intensity <strong>of</strong> survey and monitoring over an extended period (Specht, pers. comm. 2001).<br />
Lapita Homelands Project, 1985 Season:<br />
While the 1985 season was reported in more final form in 1991 (Allen & Gosden 1991),<br />
some significant additional information regarding survey is contained in the 1985 field<br />
season report. In relation to work at Bal<strong>of</strong>,<br />
“Explorations for coastal sites were also carried out. Three were found<br />
(ELA, ELB, ELD), each with small obsidian flakes and red, shell tempered<br />
pottery. At ELD two crenulate rims, three sherds with applied nubbins,<br />
some incised decoration and applied strips all link this pottery with that<br />
previously found at Lesu. However, all sites had been very disturbed, and<br />
no undisturbed deposits were found (White et al. 1985)”<br />
No details were given <strong>of</strong> the extent or type <strong>of</strong> survey which located these coastal sites, and<br />
they make no appearance in the 1991 volume, although presumably do exist in the PNG<br />
museum records.<br />
Kirch’s group conducted further survey on the smaller islets near Eloaua, locating<br />
seven new sites, principally aceramic middens (Kirch 1985). Kirch and Gorecki visited<br />
Lavongai, New Hanover, for two days and surveyed beaches, river banks and valleys in<br />
the area (Gorecki 1985). Middens and low mounds were noted along virtually the entire<br />
Lavongai bay, with no pottery found, although the National Museum had previously<br />
received some sherds from there. One <strong>of</strong> the research questions for the 1988 Mussau field<br />
season was to define the chronology and nature <strong>of</strong> the transition from Lapita to post-<br />
144
Lapita, involving excavation <strong>of</strong> an additional aceramic midden site, but no further survey<br />
seems to have occurred by the termination <strong>of</strong> the field phase <strong>of</strong> the project (Kirch 2001:10-<br />
24)<br />
<strong>The</strong>re is considerably more detail on survey method in the Duke <strong>of</strong> Yorks Islands<br />
in Lilley’s contribution to the 1985 field season report (Lilley 1985) than in the subsequent<br />
1991 volume. Intensive investigation <strong>of</strong> the western coasts <strong>of</strong> Duke <strong>of</strong> York, Makada and<br />
Mioko Islands located six Lapita sites. Inland survey transects fostered the opinion that this<br />
was a zone <strong>of</strong> low archaeological potential, and the eastern coastlines were cliffed or<br />
swampy and considered unsuitable for settlement, or possibly having poor archaeological<br />
visibility: this point is not clear. Makada, Duke <strong>of</strong> York and Mioko Islands were surveyed,<br />
predominantly along favourable sections <strong>of</strong> the coastlines for settlement (Lilley 1991). <strong>The</strong><br />
survey comprised a total land and sea area <strong>of</strong> about 75km 2 , yielding a Lapita site density<br />
<strong>of</strong> 0.08 sites/km 2 . This included about 28km <strong>of</strong> coastline, yielding a site density <strong>of</strong> 0.214<br />
sites per km. A subsequent 5-week intensive pedestrian survey <strong>of</strong> 60km <strong>of</strong> coastline (my<br />
measurement) located well-preserved Lapita potttery below the water table at sites SDQ<br />
and SDP, site SEE was found to have a high density <strong>of</strong> Lapita sherds, and SEP had some<br />
Lapita sherds in the lowest Layers (White & Harris 1997). <strong>The</strong> more exclusive definition<br />
<strong>of</strong> what constitutes a site, compared to the earlier survey, means that site dentity decreases<br />
compared to Lilley’s survey, to 67 sites per 1000km, or 0.025 sites per km 2 (I measured<br />
total area, land and sea at 160km 2 ).<br />
In the Arawe Islands, Specht spent five nights on Pililo, and three surface pottery<br />
sites/findspots were recorded, two with Lapita sherds (Specht 1985). On Kumbun, local<br />
residents identified three ceramic sites, Pukwo site with Madang/Sio-Gitua pottery, Iolmo<br />
cave with similar pottery, and Emol beach flat, with similar pottery again. Arawe Island<br />
was visited briefly, and Lapita was immediately discovered together with Madang/Sio<br />
Gitua, and other styles, on the Makegur spit. Specht left at this point, leaving Gosden to<br />
continue the survey (Gosden 1985).<br />
After an unrewarding search for in situ stratified deposits in the Arawe island site,<br />
145
Gosden spent some days excavating in Paligmete village on Pililo island, then “surveyed<br />
the inhabited islands”, revisiting Arawe and Kumbun, and briefly visiting Maklo and<br />
Kauptimete, where an extensive midden with obsidian and a single Lapita sherd was<br />
discovered. A similar site with unspecified pottery was discovered on Maklo. Gosden<br />
describes these surveys as “by no means comprehensive”. Agussak island was visited, and<br />
a pottery site with one possible Lapita sherd was recorded. Seventeen sites (six Lapita)<br />
were recorded on six islands totaling 110 km 2 approximately (land and sea, my calculation<br />
from the details given in the 1985 and 1991 reports).<br />
Gosden finally reported a cluster <strong>of</strong> seven Lapita sites within an area <strong>of</strong> around<br />
100km 2 (land and sea) in the Kauptimete/MakIo/Kumbun/Adwe/Pililo region <strong>of</strong> the Arawe<br />
area, although he was not specific about the extent and method <strong>of</strong> survey. This yielded a<br />
site density <strong>of</strong> roughly 0.07 sites per km 2 (Gosden 1991b). Gosden questioned whether<br />
these sites were contemporaneous or not, and expected other sites would be found in the<br />
future. In spite <strong>of</strong> this caveat, he subsequently suggests the clustered distribution<br />
comprised a social landscape (Gosden & Webb 1994). My coastline measurement yielded<br />
a total <strong>of</strong> 42km for the islands surveyed, or 0.17 sites per square kilometre. It seems sites<br />
with small amounts <strong>of</strong> Lapita pottery are included in these totals.<br />
In the Kandrian district, Specht revisited terrestrial Lapita find spots on Apugi<br />
island and conducted test excavations, without any survey coverage details beyond the<br />
level <strong>of</strong> village names (Specht 1985). During this fieldwork he received insistent reports<br />
<strong>of</strong> pottery on the reef at Kreslo (Specht 1985, 1991) and spent a day visiting the site. His<br />
initial conclusion was that this was a defluved beach deposit but on reflection and in view<br />
<strong>of</strong> results since obtained on Mussau and in the Arawes, Specht noted the difficulty posed<br />
to this explanation by the absence <strong>of</strong> sherds on the adjacent land, and suggested stilt houses<br />
as a possible site-formation process (Specht 1991).<br />
For the seven hundred square kilometres or so <strong>of</strong> the Siassi Islands, a single Lapita<br />
site on Tuam (Lilley 1986: 109&166) gave a rough site density <strong>of</strong> 0.0014 sites/km 2 , or<br />
using my coastline measurement <strong>of</strong> 130km (probably an underestimate due to a crenellate<br />
146
coastline drawn at a smaller scale than the others) 0.008 sites per km. Lilley noted coastline<br />
subsidence and emergence (Lilley 1986:105) and the effects <strong>of</strong> the 1888 Ritter tsunami and<br />
storm action on this coastline. More intensive coastal survey was conducted on the small<br />
islands <strong>of</strong> Tuam and Malai, with a combined coastline <strong>of</strong> around 10km, and along a 10km<br />
stretch <strong>of</strong> the northeast coast. This yielded a much higher return <strong>of</strong> one Lapita site for 10-<br />
20 km <strong>of</strong> coast, depending on how one views the preservation potential <strong>of</strong> the northwest<br />
coast, giving a figure <strong>of</strong> 0.1 to 0.05 Lapita sites per km.<br />
On Nissan, three or four Lapita sites were discovered (Spriggs 1991), depending<br />
on whether Halika-phase sites can be regarded as “Lapita without pots”. Spriggs notes that<br />
“Systematic surveys <strong>of</strong> former village sites were not undertaken, although<br />
coverage was good for the south part <strong>of</strong> Nissan. Most surface sites from<br />
Tanamalit to Nahoi were destroyed by construction <strong>of</strong> a large American<br />
base in WWII, as were some areas <strong>of</strong> Barahun island and Iachtibol.<br />
Nachman had collected pottery from 105 named localities in the vicinity <strong>of</strong><br />
Balil, Siar, Salepen and Poriwan villages and on Sirot island, and so no<br />
resurvey <strong>of</strong> these areas for village sites was undertaken, although many<br />
such sites were noted in passing on Sirot and Northwest Nissan.” (Spriggs<br />
1991:226)<br />
Spriggs’ survey had a strong emphasis on rockshelters, with 45 rockshelters and 21 open<br />
sites recorded (Spriggs 1985). Of these, the reef site at Tarmon village was the only open<br />
Lapita site, the other two being rockshelter deposits with small samples <strong>of</strong> sherds<br />
recovered. <strong>The</strong> atoll covers an area <strong>of</strong> about 100 square kilometres, yielding an open-site<br />
density <strong>of</strong> 0.01 sites per square kilometre, or for an estimated 34 km <strong>of</strong> lagoon-side<br />
shoreline (70km total including ocean-side shoreline) 0.029 sites per km.<br />
Southeast Solomon Islands:<br />
Survey and excavation on Santa Ana (Davenport 1972) located pottery at Feru, leading<br />
Davenport to hypothesize an early, Lapita-related ceramic settlement, but no information<br />
on site distribution can be derived from the report. Survey method is not reported in any<br />
detail, and is assumed by default to comprise a tour <strong>of</strong> villages seeking local information,<br />
with opportunistic examination <strong>of</strong> areas en route, with prospection rather than<br />
147
distributional information as the primary objective.<br />
On Bellona, Poulsen reported a number <strong>of</strong> terrestrial sites sealed by earth mounds,<br />
and demonstrated that the Sikumango mound sealed deposits that contained a sherd that<br />
“looked like Lapita” (Poulsen & Polach 1972). It is clear from Poulsen’s account and<br />
radiocarbon dates that site visibility for this type <strong>of</strong> deposit may bear little relation to the<br />
distribution <strong>of</strong> earth mounds, and therefore the 33km 2 Bellona survey (land and sea) can<br />
only tell us that there was at least one (probable) Lapita site there. Survey methods seem<br />
to have been more intensive than the subsequent reconnaissance efforts <strong>of</strong> others elsewhere<br />
in the Southeast Solomons, with a month spent covering the 9km 2 or so <strong>of</strong> the inland<br />
garden zone and limited coastal area suitable for settlement. Survey seemed principally<br />
aimed at mapping the conspicuous mound and pit features, abandoned pre-WWII<br />
habitation sites, caves in the bush, and freshwater sources on the coast; the distribution <strong>of</strong><br />
which were recorded in detail. <strong>The</strong> results <strong>of</strong> Poulsen’s excavations during the following<br />
month raise the possibility <strong>of</strong> an early Lapita phase <strong>of</strong> low archaeological visibility, that<br />
may bear little distributional relationship beyond general environmental constraints to the<br />
mapped distribution <strong>of</strong> mound sites, which seem to have formed partly through occupation<br />
during the period 480-1290 AD, approximately.<br />
Southeast Solomon Islands Culture History Project:<br />
Green conducted a week-long terrestrial survey <strong>of</strong> the Surville peninsula on Makira (San<br />
Cristobal) in 1970 (Green 1976c), with the aim, “...to determine the types and range in<br />
sites in this area.” <strong>The</strong> survey encompassed an area <strong>of</strong> 250km 2 (my calculation, land and<br />
sea) or about 60km <strong>of</strong> coastline (measured from map presented in Green 1976c), and in<br />
spite <strong>of</strong> good surface visibility <strong>of</strong> artefacts as a result <strong>of</strong> crab burrowing, and a relatively<br />
rich archaeological landscape <strong>of</strong> late-prehistoric coastal terrestrial sites, no early ceramics<br />
were found. Methods are assumed to have been predominantly informant-prospection.<br />
Green regarded the area as favourable for settlement, with good canoe access to a<br />
relatively sheltered harbour, and the complete lack <strong>of</strong> any pottery must have been food for<br />
148
thought, particularly in comparison to the record in the Reef/Santa-Cruz group. Green<br />
noted the dramatic effects <strong>of</strong> 1971 cyclone waves on coastal rockshelters, a comment<br />
which now seems relevant to the question <strong>of</strong> intertidal site preservation on Makira. While<br />
this survey is robust evidence for the absence or rarity <strong>of</strong> terrestrial Lapita pottery sites,<br />
it cannot tell us whether Lapita was deposited in the sea in the past from stilt villages.<br />
A series <strong>of</strong> surveys <strong>of</strong> Ulawa (Hendren 1976, Ward 1976) covered an approximate<br />
area (land and sea) <strong>of</strong> 209km 2 (my calculation) and revealed a coastal terrestrial focus <strong>of</strong><br />
late prehistoric settlement. No details <strong>of</strong> prospection/survey method are reported. Hendren<br />
noted an almost continuous distribution <strong>of</strong> tool and midden scatter in coastal areas. Ward<br />
notes evidence for exposure <strong>of</strong> this coast to cyclone waves, and effects on terrestrial site<br />
preservation. <strong>The</strong> occupation sites excavated by Ward postdated about 700AD. Results<br />
suggest terrestrial pottery sites <strong>of</strong> any period are rare or absent from the extant<br />
archaeological record.<br />
Green’s 1970 reconnaissance survey <strong>of</strong> Uki located sites from the last 500 years<br />
on a series <strong>of</strong> uplifted shorelines (Green 1976b). No pottery was found. In the present<br />
context these results raise the question whether survey methods were sufficiently intense<br />
to locate low-visibility ceramic deposits had these been formerly located in the intertidal<br />
areas presently forming a back-swamp adjacent to the paleoshorelines.<br />
Inland survey <strong>of</strong> Santa Cruz (Nendo) (Yen 1976) located late prehistoric or historic<br />
occupation sites and no pottery.<br />
Green found six Lapita sites over the 78 km 2 (land and sea?) <strong>of</strong> the main Reef<br />
Islands (although in remote-Oceania, included for comparative purposes), and two sites<br />
on the roughly 660 km 2 <strong>of</strong> Santa Cruz. Green reported extensive searching and questioning<br />
<strong>of</strong> local people, without further result (Green 1976a), from which I calculated a range <strong>of</strong><br />
site densities from 0.077 to 0.00303 sites/km 2 . By my measurements, the main Reef<br />
Islands, and Santa Cruz have coastlines <strong>of</strong> 33 and120km (eastern portion as per Green<br />
p249) respectively, giving site densities <strong>of</strong> 0.182 sites per km <strong>of</strong> coast for the Reefs, and<br />
0.012 sites per km for Santa Cruz.<br />
149
A subsequent brief resurvey <strong>of</strong> the western end <strong>of</strong> Santa Cruz, the Graciosa Bay<br />
area (Mccoy & Cleghorn 1988:106) made heavy use <strong>of</strong> informant-prospection. <strong>The</strong> main<br />
emphasis <strong>of</strong> the fieldwork was on area excavations. A two-day survey <strong>of</strong> the coastline <strong>of</strong><br />
To Motu Neo (about 20km from the map published by McCoy and Cleghorn) located an<br />
additional three sites, bringing the To Motu Neo total to five. <strong>The</strong>re is no mention <strong>of</strong><br />
Lapita at the two unexcavated rockshelters, but the Novlao Rockshelter site yielded plain<br />
ceramics and Lapita diagnostic decoration, and C14 dates spanning the Lapita period. <strong>The</strong><br />
intensive 20km coastal survey thus located two Lapita open sites without structural<br />
features, and one Lapita rockshelter/open site with structural features. Site density works<br />
out at 0.15 sites per km <strong>of</strong> coastline for an intensive survey method.<br />
Central Solomon Islands:<br />
Roe’s surveys on Guadalcanal had the following aims:<br />
• to identify the range <strong>of</strong> site types known to local informants as a key to the late<br />
prehistoric/historic settlement patterns<br />
• prospection, to locate archaeological sites with excavation potential<br />
• to search specifically for rock art sites and inland agricultural sites<br />
(Roe 1993: 31)<br />
In forested areas with dissected limestone geology, local informants were the<br />
dominant means <strong>of</strong> prospection for rockshelter sites. In the open grassland <strong>of</strong> the<br />
Northwest coast regional survey, site survey was conducted without prior local<br />
information, but no further information on method was given (Roe 1993:31-32). No<br />
pottery was seen during these surveys, and no sites <strong>of</strong> any kind were found during the<br />
north coast survey. Roe suggested that construction <strong>of</strong> the North Guadalcanal coastal<br />
road, clearance for plantations, and large-scale WWII bulldozing may have been major<br />
factors in creating this archaeological blank but didn’t consider the effects <strong>of</strong> cyclone storm<br />
surge or tsunami on site preservation along this coast.<br />
In a review <strong>of</strong> Solomon Islands prehistory prior to his research, which was<br />
150
considered suggestive <strong>of</strong> a different cultural history than that recorded elsewhere, due to<br />
the absence <strong>of</strong> pottery in surface collections, Roe included a footnote to the effect that<br />
Spriggs’ work on Nissan revealing Lapita occupation, Wickler’s Buka work that found<br />
Lapita on the reef flats, and Reeve’s data on the intertidal ceramics in the Western<br />
Province were all unknown to him at the time <strong>of</strong> fieldwork (Roe 1993:6, 10). What Roe’s<br />
work does tell us is that there is a contrast in the terrestrial record between Guadalcanal<br />
and the north Solomon Islands in the absence <strong>of</strong> late-prehistoric pottery, but absence <strong>of</strong><br />
evidence regarding the intertidal/reef flat record means that this contrast cannot be safely<br />
extended back to the Lapita period given current knowledge. It seems clear enough that<br />
Roe did not set out to find the type <strong>of</strong> reef-flat evidence found on Buka and Nissan, and<br />
that the absence <strong>of</strong> such evidence in his results cannot be taken as evidence for absence.<br />
Roe’s results therefore are limited to negative locational evidence, which in itself is weak<br />
due to recent coastal landscape alterations in the area surveyed.<br />
Cultural Resource Management (CRM) surveys by the National Museum <strong>of</strong> Solomon<br />
Islands:<br />
<strong>The</strong> Solomon Islands National Sites Survey listed among its objectives the assessment <strong>of</strong><br />
archaeological potential <strong>of</strong> each island (Miller & Roe 1982). This is a similar objective to<br />
many <strong>of</strong> the surveys conducted within the Lapita Homeland Project. It appears that the<br />
larger regional surveys were all predominantly reliant on the informant method <strong>of</strong><br />
prospection. Of these only a survey <strong>of</strong> Simbo located pottery deposits. <strong>The</strong>se surveys<br />
included the southeast Choiseul survey, the survey <strong>of</strong> the Bughotu area on Isabel, the<br />
Kwaio area survey on Malaita, a survey <strong>of</strong> Simbo (all prior to 1978); Paripao district,<br />
inland Guadalcanal, Santa Catalina survey, and survey <strong>of</strong> Dorio coastal district on Malaita<br />
(between 1978 and 1980).<br />
Within the National Sites Survey, survey in advance <strong>of</strong> construction or logging has been<br />
more likely to turn up pottery deposits, possibly due partly to more intense and unbiased<br />
151
survey methods. Such surveys include four surveys on Kolombangara, all <strong>of</strong> which located<br />
pottery, coastal plain survey in the Arosi district on Makira, which located no pottery,<br />
survey in the Viru harbour area <strong>of</strong> New Georgia which located a range <strong>of</strong> inland sites but<br />
no pottery (given the results <strong>of</strong> the NGAS project, its absence is anomalous). Survey in<br />
advance <strong>of</strong> mining on Vaghena resulted in test-excavations on pottery-bearing sites. Survey<br />
in advance <strong>of</strong> logging on north New-Georgia recorded significant pottery-bearing sites.<br />
Survey on the north coast <strong>of</strong> Nggela in advance <strong>of</strong> road construction located no pottery<br />
sites, but a range <strong>of</strong> inland sites. Survey on the Shortland islands in advance <strong>of</strong> logging<br />
recorded pottery sites. Miller and Roe concluded that:<br />
“Pottery, previously regarded as being restricted in distribution to the<br />
south-east Solomons, northern Choiseul and the Shortland Islands, appears<br />
to have been in use at some time throughout the greater part <strong>of</strong> the western<br />
Solomons (Miller & Roe 1982).”<br />
None <strong>of</strong> the sites located in these surveys can be described as Lapita, although Reeve<br />
characterized the intertidal Roviana pottery seen later at Paniavile and at other unspecified<br />
locations as possibly Lapita-derived.<br />
<strong>The</strong>se surveys yield no information on the distribution <strong>of</strong> Lapita sites in near-<br />
Oceania, other than some negative evidence against a terrestrial distribution. As for Roe’s<br />
work on Guadalcanal, in view <strong>of</strong> what seem to be many examples <strong>of</strong> intertidal Lapita<br />
settlements across the northern part <strong>of</strong> the region at present, as for many <strong>of</strong> the researchoriented<br />
surveys, these CRM surveys <strong>of</strong>fer no strong evidence for absence <strong>of</strong> Lapita in<br />
southern near-Oceania, but may be evidence for a relative rarity <strong>of</strong> Lapita sites, or less<br />
visible locational patterning than in the Northern region.<br />
A number <strong>of</strong> more recent CRM survey reports are held at the National Museum<br />
and in provincial <strong>of</strong>fices, but were not accessible at the time <strong>of</strong> writing.<br />
Bougainville and the Shortland Islands:<br />
A 1966 survey report (Lampert 1966) was not accessed. In the Buka area, Specht<br />
conducted an extensive coastal survey (Specht 1969), while Wickler restricted his survey<br />
to coastal areas suitable for settlement (Wickler 1995:62-68). Wickler’s survey yielded<br />
three Lapita sites on the reef flats, in some cases extending (as separate sites) on to shore,<br />
152
and at least one other non-cave site (DAI) (Wickler 1995: 714-16). <strong>The</strong> Lapita site density<br />
for the approximately 600km 2 region works out to 0.007 sites/km 2 .<br />
Wickler’s survey aims were to locate sites suitable for excavation rather than to<br />
achieve extensive areal coverage (Wickler 2001:63). Survey <strong>of</strong> the east coast <strong>of</strong> Buka was<br />
by vehicle, with likely locations on raised coral flats, beaches and reef flats examined. In<br />
common with Specht and Terrell (for the Silao Peninsula), Wickler noted the virtually<br />
continuous distribution <strong>of</strong> archaeological materials along the raised coral portions <strong>of</strong> this<br />
shoreline, and an arbitrary classification into sites. It would appear from the detail given<br />
(Wickler 2001:63-68, 215-216) that around 90km was covered in this manner. For the<br />
Lapita period, three early-phase sites were found on reef flats, and two open terrestrial<br />
sites. For Wickler’s more intensive survey <strong>of</strong> favourable locations these results yield a site<br />
density <strong>of</strong> 0.056 Lapita sites per km <strong>of</strong> coastline surveyed (total km rather than coastline<br />
length <strong>of</strong> targeted locations).<br />
A surface collection <strong>of</strong> sherds from Teop was reported in 1964 (Shutler & Shutler<br />
1964). No details <strong>of</strong> this were available to the author. Terrell surveyed four regions on<br />
Bougainville, Silao Area, Teop area and Numa Numa area on the east coast in<br />
north/central Bougainville, and Paubake area <strong>of</strong> south Bougainville around Buin. Terrell<br />
noted the pioneering exploratory nature <strong>of</strong> his surveys, and the extensive rather than<br />
intensive approach (Terrell 1976:218). He wanted to sample extensively to find generalities<br />
rather than the specific sequence <strong>of</strong> a local area, and wanted to include both Austronesian<br />
and Papuan language areas in his sample (Terrell 1976:224-227). Terrell conducted a<br />
second survey at Buin following the initial extensive reconnaissance, undertaken in view<br />
<strong>of</strong> the divergent results obtained for this locality on the initial visit.<br />
Methods included informant-prospection, and can be expected to have yielded<br />
results strongly biased toward the late-prehistoric and historic settlement pattern. Site<br />
survey on the Silao peninsula revealed the northern coast <strong>of</strong> the peninsula to be “one long<br />
archaeological site” for the approximately 14 km <strong>of</strong> coastline surveyed, echoing Specht’s<br />
description. Fifty-two “sites” were abstracted from this near continuous distribution, from<br />
which almost 100 locality collections were made. No sherds in the Buka (Lapita) style<br />
were identified, but the low number <strong>of</strong> these in Specht’s samples was considered by<br />
153
Terrell to make sampling error a sufficient explanation for their absence. <strong>The</strong> pottery all<br />
fell within Specht’s Sohano and later sequence.<br />
In the Teop area, approximately 19km <strong>of</strong> coastline was surveyed, as well as some<br />
inland journeying. Much <strong>of</strong> the coast was either mangrove flats or swampy, and only four<br />
prehistoric pottery sites were recorded on the upraised coral terrestrial coastline, these<br />
being styles from Specht’s sequence, back to at least the Sohano style. With the exception<br />
<strong>of</strong> the Nanava form, thought to have been manufactured to the south near Kieta, these<br />
were locally produced. (Terrell 1976:240). Sherds were scarce in the interior.<br />
In the Numa Numa area, approximately 45km <strong>of</strong> coastline were surveyed, locating<br />
32 sites in spite <strong>of</strong> low visibility caused by dense coastal bush along much <strong>of</strong> this sparsely-<br />
occupied coast. Mararing to recent styles were found (Terrell 1976: 252-257). Terrell<br />
considered these surveys too preliminary and incomplete to assess temporal changes in site<br />
density.<br />
<strong>The</strong> Paubake survey area in the south <strong>of</strong> Bougainville seemed on prior information<br />
to contain a ceramic sequence that differed from Specht’s Buka sequence (Terrell<br />
1976:261-263, Figure 4.7). About 8km <strong>of</strong> coastline was included in the survey. Sixteen<br />
sites were initially recorded, including surface pottery scatters and stone constructions,<br />
with sherds present in smaller numbers than the northern surveys. An additional 44 sites<br />
were visited and mapped during the second survey in this region. Only one <strong>of</strong> these sites<br />
was located on the coast,and it is not reported whether this can be considered to represent<br />
a distributional feature <strong>of</strong> past settlement. No Lapita pottery was seen.<br />
Terrell’s surveys in these four areas are incomplete, in his own estimation, and<br />
cannot be taken as strong evidence for the absence <strong>of</strong> Lapita occupation in the past. All<br />
that can be said with confidence is that nowhere in the areas surveyed is Lapita pottery an<br />
obvious feature <strong>of</strong> the coastal terrestrial archaeological record. We can only speculate as<br />
to whether this is to do with past behaviour patterns that contrast with areas where Lapita<br />
pottery is a more obvious component <strong>of</strong> the record, (this could include more subtle<br />
behavioural variants than avoidance <strong>of</strong> the area by Lapita-using people, e.g. intertidal site<br />
location in the past) or whether site visibility and/or preservation factors can account for<br />
such variability in the Lapita archaeological record. Given Wickler’s survey results, I<br />
would expect that Lapita sites were there, in the sea, 3000 years ago.<br />
154
<strong>The</strong> coastal portions <strong>of</strong> Terrell’s surveys were all terrestrial, and total<br />
approximately 86km <strong>of</strong> coastline on Bougainville, surveyed in a low-intensity manner,<br />
using predominantly informant-prospection and, for much <strong>of</strong> his sample, having low<br />
archaeological visibility due to vegetation cover. All the sites recorded belonged to the last<br />
millennium. As noted above, such survey methods can be expected to create a strong late-<br />
site bias.<br />
Irwin’s survey <strong>of</strong> the Shortland Islands began by surveying<br />
“a large area in order to get some idea <strong>of</strong> the nature and variation <strong>of</strong> the surface evidence<br />
in addition to its distribution.” (Irwin 1972:3)<br />
Table 5: Summary <strong>of</strong> a review <strong>of</strong> survey methods and Lapita results.<br />
SURVEY REGION SURVEY METHOD SURVEY INTENSITY COASTLINE<br />
SURVEYED (KM)<br />
Eloaua/ Emananus<br />
1984<br />
terrestrial<br />
?<br />
Huon Pen. ? high but unfavourable for<br />
settlement (specht pers comm 2001)<br />
155<br />
LAPITA<br />
OPEN SITES<br />
high 33 3 91<br />
60 0 0<br />
DENSITY<br />
(SITES /<br />
1000KM)<br />
Talasea ‘73 terrestrial intertidal high? 25 4-20 160-800<br />
Duke <strong>of</strong> York terrestrial<br />
?<br />
Moderate 28 6 214<br />
Arawe terrestrial moderate 42 7 167<br />
Siassi terrestrial high 10 1 100<br />
Nissan terrestrial<br />
intertidal?<br />
late bias for open sites<br />
moderate 34 1 29<br />
Bellona terrestrial, late bias high - 0 0<br />
Makira terrestrial ? not meas 0 0<br />
Ulawa late bias low 0 0<br />
Uki paleo-shoreline ? 0 0<br />
Santa Cruz terrestrial<br />
inf. prosp.<br />
Reef Islands terrestrial<br />
inf. prosp.<br />
Guadalcanal terrestrial<br />
varies<br />
low to mod. 120 2 12<br />
high 33 6 182<br />
modified coastal landscape ? 0 0<br />
Buka (Wickler) terrestrial intertidal moderate 90 5 56<br />
Bougainville terrestrial<br />
inf. prosp.<br />
Shortland terrestrial<br />
inf. prosp.<br />
Low 86 0 0<br />
targeted 70 0 0<br />
Duke <strong>of</strong> Yorks Terrestrial intensive 60 4 67<br />
To Motu Neo Terrestrial intensive 20 3 150
Survey was concentrated on the east and south coasts <strong>of</strong> Alu Island, and appears to have<br />
comprised a mix <strong>of</strong> informant-prospection and highly targeted terrestrial coverage <strong>of</strong> the<br />
about two-thirds <strong>of</strong> the lengthy Alu coastline, within an overall survey block <strong>of</strong> about<br />
500km 2 . A rough estimate <strong>of</strong> the length <strong>of</strong> coastline covered is 70km, <strong>of</strong> a similar<br />
magnitude to the total coastal portion <strong>of</strong> Terrell’s Bougainville survey. His approach was<br />
highly targeted towards reef-passage locations (Irwin 1972:19), and made use <strong>of</strong><br />
informant-prospection method. Twenty-two sites were recorded, most <strong>of</strong> which yielded<br />
surface pottery collections, which were fitted into a hypothetical post-500AD sequence<br />
comprising two widely-spaced early-period sites, nine closely spaced middle-period sites,<br />
and five closely spaced late-period sites (Irwin 1972:193-196). No Lapita sites were found.<br />
Irwin’s survey seems to provide reasonable evidence that terrestrial Lapita sites were either<br />
rare or absent at the time <strong>of</strong> the survey, and likely to have been so in the past. No strong<br />
evidence for the absence <strong>of</strong> either terrestrial or intertidal Lapita can be claimed though. <strong>The</strong><br />
metric results <strong>of</strong> this analysis <strong>of</strong> survey method and results in Near Oceania and the<br />
Southeast Solomons are summarized in table Table 5.<br />
Discussion <strong>of</strong> Review Results:<br />
Many have commented on the gross patterning evident from these survey results, which<br />
is the absence <strong>of</strong> Lapita pottery in near-Oceania south <strong>of</strong> Buka, and on the Continental<br />
New Guinea coast (Huon Peninsula). <strong>The</strong>se broad-scale discussions (see for example<br />
Anderson 2002) have taken place within a data framework with minimal information on<br />
or discussion <strong>of</strong> site preservation and visibility, despite comments within the individual<br />
156
survey reports on such factors, for example on differential exposure to cyclone and<br />
tsunami waves. With an average return for these 17 surveys <strong>of</strong> 2-3 Lapita sites each,<br />
absence <strong>of</strong> sites through sampling error alone is a distinct possibility, even without<br />
consideration <strong>of</strong> survey coverage, site preservation and site visibility. None <strong>of</strong> the surveys<br />
in this review which returned a negative result for Lapita can be viewed as evidence for<br />
absence, when survey methods and preservation/visibility are taken into account, if one<br />
allows that settlement in the past may have been over water in the Lapita period for the<br />
survey regions which returned negative results. It must also be remembered that there is<br />
very little information about relative sea-level changes in most <strong>of</strong> the surveys.<br />
For Near-Oceania, in regions where Lapita occupation has been demonstrated,<br />
Lapita site density ranged from 29 to 800 sites per 1000km <strong>of</strong> shoreline, with an average<br />
density <strong>of</strong> 112-183 sites per 1000km, depending on the definition <strong>of</strong> a Lapita site. <strong>The</strong><br />
sample <strong>of</strong> surveys includes two remote-Oceanic surveys, and seven near-Oceanic surveys.<br />
Much <strong>of</strong> the information on which these densities are based is uncertain, the distance<br />
surveyed in the Duke <strong>of</strong> Yorks, for example, is something <strong>of</strong> a semi-educated guess. <strong>The</strong><br />
definition <strong>of</strong> what is a Lapita site varies between researchers too, as does survey intensity.<br />
<strong>The</strong>re is full recognition that these figures have a high potential error and bias. That does<br />
not mean they are meaningless though, as they result from a review <strong>of</strong> the best information<br />
we have on the subject, the accumulated work <strong>of</strong> pr<strong>of</strong>essional archaeologists.<br />
<strong>The</strong>re is possibly a hint <strong>of</strong> geographic decay in the data, as the lowest useful<br />
densities were for Nissan and Buka, both to the south <strong>of</strong> the other near-Oceanic data, but<br />
this may simply be a sample-size effect given the low number <strong>of</strong> observations, or may<br />
157
Figure 6: Roviana and Kaliquongu surveys as samples <strong>of</strong> the New Georgia lagoon and<br />
barrier island system.<br />
relate to survey methodology and/or the effects <strong>of</strong> preservation and visibility on probability<br />
<strong>of</strong> detection.<br />
Sampling <strong>The</strong>ory and the Roviana/Kaliquongu Survey Regions:<br />
Orton notes that an archaeological sampling region may be <strong>of</strong> any size, but should<br />
have some geographic or cultural coherence (Orton 2000: Chapter 4). <strong>The</strong> Roviana and<br />
Kaliquongu surveys reported below can be seen as two single building block sample<br />
populations in two de facto regional sampling strategies, but what is the target population<br />
(Orton 2000:18) being sampled? <strong>The</strong>y could be viewed as samples <strong>of</strong> the Lapita/Post<br />
Lapita culture region, which stretched (according to present understandings <strong>of</strong> Lapita)<br />
from the Bismarck Archipelago to Tonga-Samoa. Roviana can also be considered as a<br />
sample <strong>of</strong> the near-Oceanic biogeographic region, in extent from the Bismark Archipelago<br />
158
to Makira/Ulawa. Perhaps the Roviana survey is best considered as a sample <strong>of</strong> the<br />
geographically, geologically and tectonically coherent lagoon circuit <strong>of</strong> New Georgia<br />
(Figure 6). Ultimately the significance <strong>of</strong> the Roviana results need to be considered at<br />
more than one <strong>of</strong> these scales in comparison with other “building block” surveys.<br />
This raises questions <strong>of</strong> comparability, and highlights the importance <strong>of</strong> the details<br />
<strong>of</strong> survey methodology and assessments <strong>of</strong> archaeological visibility and archaeological<br />
preservation for comparative data. In this respect, the Roviana survey is more comparable<br />
to most other regional surveys than the Kaliquongu survey at the broader scales <strong>of</strong><br />
comparison mentioned above. <strong>The</strong> Kaliquongu survey is in some respects comparable to<br />
the recent Buka survey (Wickler 1995) in that Wickler states that examination <strong>of</strong> reef flats<br />
was purposive, but information in that case on site visibility and quantification <strong>of</strong> survey<br />
coverage was not given, limiting comparative utility.<br />
Preservation and Visibility:<br />
In contrast to the Bismark archipelago, the geographic and geological setting <strong>of</strong> the<br />
Roviana/Kaliquongu surveys is an area <strong>of</strong> low seismicity, and despite the presence <strong>of</strong><br />
active undersea volcanism <strong>of</strong>f southern New Georgia, is not blanketed with any airfall<br />
tephras. Recent work explicating sea-level history (Mann et al. 1998) suggests sherd<br />
scatters in the intertidal zone are currently in the process <strong>of</strong> emerging from the sea into the<br />
swash zone, and being destroyed in the process. <strong>The</strong> relatively high visibility <strong>of</strong><br />
archaeological sites in the intertidal zone <strong>of</strong> coralline gravel shorelines and the ease with<br />
which sherd samples can be obtained enable a regional spatial sampling approach along<br />
such shorelines with the aim <strong>of</strong> obtaining intersite distributional information. <strong>The</strong>re are<br />
other s<strong>of</strong>t-sediment shorelines bordering these lagoons for which archaeological visibility<br />
is much reduced. One might expect preserved ceramic/lithic scatters to be obscured to<br />
159
some degree along such s<strong>of</strong>t-sediment shorelines.<br />
Even along the sheltered shorelines <strong>of</strong> the lagoon, archaeological preservation in<br />
zones <strong>of</strong> high intertidal site visibility is less than ideal, as evidenced by estimates <strong>of</strong> the<br />
parent vessel populations from which sherd samples derive (see Chapter 5), and in general<br />
the recovered samples seem to only be a tiny fraction <strong>of</strong> the absolute quantities <strong>of</strong> material<br />
in the minimum breakage population. <strong>The</strong>re is thus great potential for site density in the<br />
zone <strong>of</strong> high visibility to under-represent total site density, although some lithic lag-deposit<br />
scatter should outlast the ceramics (but these are not chronologically diagnostic to the<br />
same extent as the ceramics).<br />
Both the Roviana and Kaliquongu surveys comprise sample regions that are<br />
geographically most directly representative <strong>of</strong> the lagoon systems <strong>of</strong> the New Georgia<br />
Group. While a feature <strong>of</strong> the geology <strong>of</strong> the New Georgia group is very substantial and<br />
rapid uplift <strong>of</strong> the forearc region (Ranongga/Southern Rendova/Tetepare), the bulk <strong>of</strong> the<br />
group, and particularly the main island <strong>of</strong> New Georgia, is more stable, surrounded by an<br />
intricate system <strong>of</strong> elevated Pleistocene-age coral reefs and lagoons (Dunkley 1986, Mann<br />
et al. 1998, Stoddart 1969a, b). As discussed in Chapter 1, this system <strong>of</strong> lagoons,<br />
mainland shorelines and barrier islands forms a near-continuous circuit <strong>of</strong> New Georgia,<br />
and is relatively unvarying in terms <strong>of</strong> the diversity <strong>of</strong> preservational and visibility<br />
zonation over much <strong>of</strong> its length, although uplift is less at Marovo Lagoon. <strong>The</strong> area<br />
covered by these lagoons and elevated barrier islands and the crenulate drowned mainland<br />
shoreline total over 900km 2 (Stoddart 1969a: 387) (15,000km 2 including New Georgia<br />
mainland and the surrounding waters as measured in the review). From the presently<br />
published geological/geomorphological literature it would seem that the 200km 2 Roviana<br />
lagoon is roughly comparable as a settlement/preservation/visibility context to the 700km 2<br />
<strong>of</strong> barrier lagoons along the whole north-east coast <strong>of</strong> New Georgia, comprising Tongavae<br />
160
Lagoon, Ngerrasi Lagoon, Marovo Lagoon, Kolo Lagoon and Kalikolo Lagoon, and also<br />
the Panga Bay area south <strong>of</strong> Vangunu, and can be reasonably regarded as a 2/9 sample <strong>of</strong><br />
this area. <strong>The</strong> Kaliquongu intertidal archaeological survey is best seen as most directly<br />
representative <strong>of</strong> intertidal patterning within this 900km 2 area, and results should only be<br />
extrapolated to other environmental contexts with caution.<br />
In characterizing the archaeological distribution in a sample region, the aim is to<br />
obtain estimates <strong>of</strong> the values <strong>of</strong> variables that can be measured in principle on our target<br />
region (Orton 2000:18). For the Roviana and Kaliquongu surveys aims were to obtain<br />
estimates <strong>of</strong> the ratio variable site density (number <strong>of</strong> sites per km <strong>of</strong> coastline and number<br />
<strong>of</strong> sites per square kilometre) for two coarsely defined periods, Lapita and Post-Lapita (as<br />
measured by ceramic styles).<br />
In summary, features <strong>of</strong> this environment pertinent to the approach taken are:<br />
• reduced wave-exposure through the protection <strong>of</strong> the raised coral barrier<br />
islands, leading to enhanced preservation <strong>of</strong> intertidal sites,<br />
• a low degree <strong>of</strong> sedimentation on the barrier island shorelines, resulting in<br />
high probability <strong>of</strong> site detection in some locations given appropriate<br />
survey methods and reasonable preservation. Additionally, the use <strong>of</strong> a<br />
single sample region for each <strong>of</strong> the two survey methods means the results<br />
do not comprise estimates <strong>of</strong> New Georgia site density in a statistical sense.<br />
One further set <strong>of</strong> geological features worthy <strong>of</strong> particular note is the number and location<br />
<strong>of</strong> deep-water reef passages between the barrier islands. <strong>The</strong>re are around 40 passages in<br />
total but many are shallow, and the deeper <strong>of</strong> these may be favourable settlement<br />
localities, due to navigational advantages and marine biodiversity. Reef passages give<br />
161
access to pelagic resources in the channels surrounding New Georgia, as the outer coasts<br />
<strong>of</strong> barrier islands are (in general) unsuited to settlement (there are modern exceptions, for<br />
example Himo bay in the Saikile district) <strong>The</strong>se ge<strong>of</strong>acts and bi<strong>of</strong>acts are relevant to any<br />
archaeological sampling <strong>of</strong> the lagoon system, as their influence on late-Holocene site<br />
location is likely to be pr<strong>of</strong>ound.<br />
Data Quality:<br />
<strong>The</strong> aim <strong>of</strong> acquiring estimates <strong>of</strong> site density depends not only on good archaeological<br />
data from Roviana Lagoon, but also on a repeatable measure <strong>of</strong> survey area and distance.<br />
Shoreline length is a more objective measure than gross area, as there are a number <strong>of</strong><br />
problems in defining the areal limits <strong>of</strong> survey in an islandic environment, and the land-to-<br />
sea ratio and area-to-coastline ratio will vary from place to place in such environments.<br />
Shoreline length is regarded as a useful predictive unit for the New Georgia lagoon<br />
intertidal target region, with reef-passage count less clear (not all reef passages were<br />
created equal, despite use <strong>of</strong> this measure elsewhere (Felgate 2002).<br />
Probability <strong>of</strong> Detection is a major archaeological sampling issue and source <strong>of</strong> non-<br />
sampling errors (Orton 2000:26). Sites on surveyed reef flats or gravelly sediments were<br />
considered to have an equal probability <strong>of</strong> detection, although variation in the degree <strong>of</strong><br />
preservation could complicate chronological assignment. Sites on s<strong>of</strong>t-sediment shores<br />
were considered to have a lower and more variable probability <strong>of</strong> detection. <strong>The</strong> absolute<br />
values <strong>of</strong> these probabilities can only be guessed at, a problem by no means unique to the<br />
Roviana survey (Orton 2000:39), so that no standard errors can be calculated. For hard-<br />
sediment shorelines the probability <strong>of</strong> detection <strong>of</strong> extant deposits is probably close to<br />
100%, although if taphonomic processes have removed ceramic stylistic information the<br />
162
probability <strong>of</strong> detection by period is less than 100%, while for s<strong>of</strong>t sediments it is probably<br />
near to zero, unless as findspots.<br />
If Lapita does favour reef passage locations, which tend to have hard-sediment<br />
shorelines and high probability <strong>of</strong> detection (assuming sufficient preservation), this might<br />
lead to a higher probability <strong>of</strong> detection for this site type in comparison to more widely<br />
distributed site types.<br />
Survey Methods:<br />
Design <strong>of</strong> the Roviana and Kaliquongu surveys is considered in relation to the 12 stages<br />
<strong>of</strong> sample survey recommended by Orton (Orton 2000:27, 76-100). <strong>The</strong>se are:<br />
1. assimilation <strong>of</strong> existing knowledge<br />
2. objectives <strong>of</strong> the survey<br />
3. population to be sampled<br />
4. data to be collected<br />
5. degree <strong>of</strong> precision required<br />
6. method <strong>of</strong> measurement<br />
7. the frame<br />
8. selection <strong>of</strong> the sample<br />
9. the pre-test<br />
10. organization <strong>of</strong> the fieldwork<br />
11. summary and analysis <strong>of</strong> the data<br />
12. information gained for future surveys<br />
163
Methods <strong>of</strong> the Roviana Survey:<br />
Assimilation <strong>of</strong> existing knowledge:<br />
What do existing surveys tell us about how much work needed to get the information<br />
desired? It has been suggested throughout this <strong>chapter</strong> that pioneering informant-<br />
prospection surveys <strong>of</strong> the type conducted on Bougainville by Terrell and in Roviana<br />
Lagoon by the NGAS have a late-site bias. This suggests that a substantial amount <strong>of</strong> work<br />
would need to be done to fortuitously discover a proportion <strong>of</strong> early sites. Terrell’s 86km<br />
<strong>of</strong> Bougainville coastline surveyed in this manner failed to locate early (Lapita-era) pottery<br />
sites (either because there never were any there or because they were not<br />
preserved/detected by the scale and method <strong>of</strong> survey), as did Irwins’ Shortland survey,<br />
although in the Shortlands case it must be noted that Irwin targeted terrestrial reef passage<br />
locations, which may have improved the probability <strong>of</strong> early site detection. Current<br />
evidence from Buka and Roviana suggests that a) this method will have a low return as a<br />
prospection technique for Lapita sites and b) even large surveys such as Bougainville may<br />
yield no return for the early ceramic period.<br />
<strong>The</strong> Roviana Lagoon has an inner coastline length <strong>of</strong> 180km (estimated by walking<br />
a 1km divider along the mapped coastline at 1:50 000 scale) and therefore is double the<br />
amount <strong>of</strong> coastal survey conducted by Terrell by similar methods on Bougainville.<br />
Additionally, this lagoon shoreline is not exposed to tsunami or cyclone storm swells<br />
(discussed in detail in Chapter 6), although storm surge associated with cyclones passing<br />
to the south through the Bellona area does affect water-levels in the lagoon, flooding the<br />
lower shore platforms on occasion (personal observation, 1998).<br />
164
<strong>The</strong> Roviana survey therefore represents a significant addition to the knowledge-set gained<br />
through informant-prospection and fortuitous observation coastal survey methods.<br />
Objectives <strong>of</strong> the Roviana survey:<br />
<strong>The</strong> stated objectives <strong>of</strong> the Roviana Archaeological survey were to conduct a preliminary<br />
archaeological survey <strong>of</strong> the region to determine the location and type <strong>of</strong> sites present in<br />
this region (Sheppard 1996:4). Clearly these are not density estimation objectives (Orton<br />
2000:79). For the purposes <strong>of</strong> the present thesis these objectives are recast as :<br />
to test the hypothesis <strong>of</strong> Lapita settlement <strong>of</strong> New Georgia in the past, and to gain<br />
if possible an estimate <strong>of</strong> Lapita site density. Further, to estimate the density <strong>of</strong> other early<br />
post-Lapita sites known to be located across the target region.<br />
Population to be sampled by the Roviana archaeological survey:<br />
<strong>The</strong> hypothetical Lapita sites <strong>of</strong> the coastlines <strong>of</strong> the New Georgia lagoon system was the<br />
target population to be sampled. This region comprises an area <strong>of</strong> 900km 2 approximately<br />
(Stoddart 1969a), and is geologically and geographically homogenous and bounded<br />
through relative uniqueness in these respects in the broader region <strong>of</strong> near-Oceania. This<br />
was a surface survey, and for the present purposes, a survey <strong>of</strong> early ceramic sites (the<br />
majority <strong>of</strong> archaeological evidence resulting from the 1996 field seasons was <strong>of</strong> other<br />
types <strong>of</strong> site, but these do not directly relate to the present objectives). This region was<br />
considered as a series <strong>of</strong> environmental zones between which site obtrusiveness,<br />
preservation and visibility could be expected to vary. <strong>The</strong>se were:<br />
165
• the mainland shoreline, consisting for the most part recent alluvium<br />
• the mainland intertidal, consisting for the most part s<strong>of</strong>t sediments with high<br />
preservation potential, low archaeological visibility for the sought-after site types,<br />
and low site obtrusiveness;<br />
• the intertidal zone <strong>of</strong> barrier-island shorelines, predominantly although with major<br />
exceptions sandy to coralline gravel/ coralline rocky shorelines, with low site<br />
obtrusiveness, good archaeological visibility and moderate preservation; the<br />
terrestrial coastlines <strong>of</strong> the barrier islands, the recent emerged coral/reefal detritus<br />
shore platform with for the most part fairly dense vegetation, low site visibility and<br />
moderate preservation potential due to modern and possibly prehistoric gardening<br />
activities.<br />
Data collected:<br />
<strong>The</strong> primary unit <strong>of</strong> measurement was the site, with <strong>of</strong>fsite data recorded as findspots. For<br />
site I use Schiffer’s definition: “a high density area <strong>of</strong> artefacts (Schiffer et al. 1978:2,14).<br />
By findspot I mean the location <strong>of</strong> a group <strong>of</strong> three or less sherds in an area <strong>of</strong> otherwise<br />
low ceramic density. Aceramic lithic scatters are similarly regarded as sites, while aceramic<br />
findspots were not recorded. Late-Lapita sites were initially defined as those which yielded<br />
ceramics <strong>of</strong> multi-part slab vessel construction and Lapita design structure (defined in<br />
detail in Chapters 4, 8,9 and 12) (see Figure 7, Figure 8, Figure 9).<br />
166
Figure 7: Slab-built carinated vessels from Honiavasa initially assigned to the Lapita<br />
period (this assignment is examined in more detail in Chapters 8 and 9): the solid<br />
vertical line in HV.2.464 represents an estimated location <strong>of</strong> the central vertical avis<br />
(CVA) <strong>of</strong> the pot.<br />
167
Figure 8: Carinated vessels from Honiavasa, showing double-line markes on some<br />
design zones, bands <strong>of</strong> nubbins at the neck in two cases, and a band <strong>of</strong> fingernail<br />
impression in one case (top).<br />
168
Figure 9: HV.4.202 has an incised motif laid out in double lines; HV.1.314 is dentatestamped,<br />
with similar design structure and a double carination; HV.2.341 is carinated,<br />
with the design laid out in applied fillets, bounded horizontally by decorated lap joins<br />
between slabs; HV.2.297 and HV.4.379 have bands <strong>of</strong> single fingernail impressions and<br />
nubbins at the neck respectively.<br />
169
Sites with mainly simple one-piece globular forms, together with Miho-style or<br />
Gharanga/Kopo-style decoration and rim forms (decorative styles named for sites in which<br />
they were first collected in quantity, locations shown below) were classified for the coarse<br />
periodization <strong>of</strong> the current <strong>chapter</strong> as post-Lapita. This classification is examined in detail<br />
in other <strong>chapter</strong>s. <strong>The</strong>se styles are defined below as follows:<br />
• Miho style includes a diversity <strong>of</strong> vessel forms, but sharply carinated forms are<br />
virtually absent, and simple one-piece pots generally have a thick neck decorated<br />
most commonly with a band <strong>of</strong> opposed-pinch fingernail impression. Decorative<br />
techniques include also incision, perforation and applique, and incised motifs are<br />
usually made up <strong>of</strong> triangular secondary design zonation within a primary design<br />
zone band, constructed with single lines rather than double. <strong>The</strong>re is no incised<br />
horizontal upper or lower boundary to the primary design zone band. On the rim<br />
the triangles are usually filled by parallel lines, but are unfilled on the shoulder.<br />
Examples are illustrated in Figure 10, Figure 11, Figure 12, Figure 13 and<br />
Figure 14.<br />
• Gharanga/Kopo style can be defined for the present purposes by three diagnostic<br />
attributes, a band <strong>of</strong> punctation in the neck region and/or multiple bands <strong>of</strong><br />
opposed-pinch fingernail impression on the shoulder, and/or a short, heavily<br />
everted rim. Examples are illustrated in Figure 15, Figure 16, Figure 17, Figure<br />
18, Figure 19, Figure 20, Figure 21, Figure 22, Figure 23, Figure 24, Figure<br />
25, and Figure 26.<br />
170
Figure 10: Incised rims with opposed-pinch fingernail impressed band at the neck,<br />
diagnostic <strong>of</strong> the Miho subgroup <strong>of</strong> Post-Lapita styles.<br />
171
Figure 11: Miho-style post-Lapita sherds.<br />
Figure 12: Miho-style post-Lapita sherds with incised shoulders.<br />
172
Figure 13: Miho-style post-Lapita sherds with CVA measurements shown (MH290<br />
was measured at to locations on the pr<strong>of</strong>ile; at the interior <strong>of</strong> the neck orifice and the<br />
exterior shoulder).<br />
173
Figure 14: Miho-style post-Lapita sherds; dashed CVA lines are measurements based<br />
on non-circular (uneven) curvature at the pr<strong>of</strong>ile locations indicated by arrows.<br />
174
Figure 15: Gharanga-style post-Lapita sherds. (Gharanga is a short-rim subgroup<br />
<strong>of</strong> Gharanga-Kopo which may have multiple bands <strong>of</strong> opposed-pinch fingernail<br />
impression on the shoulder.<br />
Figure 16: Gharanga-style post-Lapita sherds.<br />
175
Figure 17: Gharanga/Kopo style post-Lapita sherds (top) and a large Gharanga-style<br />
sherd (bottom) (four measurement points used as arrowed to estimate CVA.<br />
176
Figure 18: Intermediate between Gharanga and Kopo styles: all post-Lapita.<br />
Figure 19: Kopo-style post-Lapita rim sherds.<br />
177
Figure 20: Another large Gharanga-style post-Lapita sherd from an unrestricted vessel<br />
form.<br />
Figure 21: Gharanga-style small-orifice post-Lapita sherd showing rolled rim and thin<br />
wall common to this style.<br />
Figure 22: Kopo-Style post-Lapita sherd (taller, less everted rim than Gharanga style)<br />
from a large-orifice vessel, with bands <strong>of</strong> impressions along both inner and outer edges<br />
<strong>of</strong> the lip.<br />
178
Figure 23: Less decorated variant <strong>of</strong> Gharanga-Kopo post-Lapita style, without a<br />
strong corner point in vertical section at the neck.<br />
Figure 24: Large-orifice Gharanga/Kopo-style sherd with deformation <strong>of</strong> the lip into a<br />
wave pattern.<br />
Figure 25: Large Gharanga-style post-Lapita sherd with typical decoration, including a<br />
band <strong>of</strong> impressions along the inner edge <strong>of</strong> the lip.<br />
179
Figure 26: Gharanga/Kopo-style post-Lapita vessels: most are weakly restricted at the<br />
neck, with short, heavily everted rims. Punctate band at the neck is the most common<br />
decoration in this group, while multiple bands <strong>of</strong> fingernail pinch are common on the<br />
short-rim examples. One sherd (MH.33) had exotic quartz-calcite hybrid temper (see<br />
Chapter 4).<br />
180
Most collection sites yielded a mixture <strong>of</strong> these two periods to a slight extent (this will be<br />
discussed in detail in Chapter 12), but showed predominance <strong>of</strong> one period in the styles<br />
represented.<br />
Degree <strong>of</strong> precision required:<br />
Precision could not be calculated due to use <strong>of</strong> a single sample region.<br />
Method <strong>of</strong> measurement:<br />
Analysis <strong>of</strong> ceramic variability <strong>of</strong> surface collections was used to assign sites to period. Site<br />
locations were recorded on 1:50000 topographic maps. Site visibility for the intertidal zone<br />
was classified as high or low using 1:25000 air photographs (gravelly reef flats and coral<br />
sand flats show pale or white while s<strong>of</strong>t sediments show dark) combined with ground<br />
truthing. Site density was calculated for both the total 180km <strong>of</strong> shoreline and for the<br />
70km <strong>of</strong> high-visibility shoreline<br />
<strong>The</strong> frame:<br />
<strong>The</strong> sampling frame for the Roviana survey was as shown in Figure 6.<br />
Selection <strong>of</strong> the sample:<br />
Survey was conducted on land and in the sea as directed by informants.<br />
<strong>The</strong> pre-test:<br />
No formal pre-test was conducted, other than the informant-prospecting survey reported<br />
by Reeve (Reeve 1989)<br />
Organization <strong>of</strong> the fieldwork:<br />
Sheppard, Felgate, Keopo and Roga spent five weeks in the Kaliquongu area <strong>of</strong> the<br />
Lagoon in January-February <strong>of</strong> 1996, using the informant-prospection method to locate<br />
traditional sites, rockshelters and early ceramic sites and findspots. A further season <strong>of</strong><br />
similar fieldwork was spent in the Saikile area, including participation by Sheppard,<br />
Walter, Felgate, Roga, Jones, Nagaoka and informants, most notably Mr.Sae Oka <strong>of</strong><br />
181
Patmos, Nusahope area. An observer-bias in the results is likely in that Mr.Sae Oka had<br />
developed a particular interest in intertidal pottery scatters and actively located a number<br />
<strong>of</strong> these on the Ndora Island coastline, while in the Kaliquongu area the informants initially<br />
consulted reported only one such locality, at the mouth <strong>of</strong> a mainland stream, Gharanga,<br />
where a landowner (Mr Phillip Lanni) had developed a personal interest in the pottery and<br />
artefacts noticed during gardening/copra activities at that specific locality (the stream is<br />
also used today to obtain water in the dry season).<br />
Summary and analysis <strong>of</strong> the data:<br />
Overall survey results supported the pattern noted previously (Reeve 1989) <strong>of</strong> early<br />
ceramic evidence located exclusively in the intertidal zone. Thin plain ceramics were found<br />
on land, associated with late-prehistoric ritual stone structures. Similar sherds were found<br />
in an inland midden deposit on the mainland dated to 1403-1490AD at one sigma (Site 25<br />
in Sheppard 1996). Only sites with sufficient ceramic evidence to allow coarse temporal<br />
characterization were counted in the analysis. Findspots comprised sites with three or less<br />
sherds in the general locality. No aceramic lithic scatters were noticed during the survey.<br />
Lapita sites were not initially located, although in a subsequent field season the Nusa<br />
Roviana site was located, which had one sherd with dentate-stamping. Results were<br />
initially interpreted (prior to the Kaliquongu survey and the Nusa Roviana dentate find) as<br />
suggesting Lapita was absent in the New Georgia group, and possibly elsewhere in the<br />
near-Oceanic Solomon Islands (Sheppard et al. 1999). A number <strong>of</strong> post-lapita sites were<br />
located, including Hoghoi, Paniavile (previously recorded by Reeve), Ririgomana, Punala,<br />
Humbi Quongu, Rangu, Gharanga, and Kopo. Intertidal ceramic findspots were noted at<br />
Mare point, Kazu, Pikoro, Mbaraulu, and Sasavele (Figure 27).<br />
Information gained for future surveys:<br />
Early ceramics seemed exclusively intertidal, while late ceramics were principally found<br />
associated with terrestrial ritual structures although a broader scatter <strong>of</strong> tiny fragments <strong>of</strong><br />
this thin pottery was noted in garden areas.<br />
182
Munda<br />
Malangari<br />
Island<br />
183<br />
Saikile<br />
Figure 27: Locations mentioned in the text in relation to Roviana and<br />
Kaliquongu surveys.<br />
<strong>The</strong> Kaliquongu Survey:<br />
Objectives <strong>of</strong> the survey:<br />
A primary reason for surveying in 1997-8 in the Kaliquongu region <strong>of</strong> Roviana Lagoon<br />
was to test the hypothesis that Lapita pottery sites were absent. Additional objectives were<br />
to describe the distributions <strong>of</strong> archaeological materials by stylistic period. An important<br />
secondary objective was methodological, and in this regard the Roviana survey is regarded<br />
as an experiment in intertidal-zone/shallow-water survey and sampling. What would the<br />
results <strong>of</strong> such a survey be? What implications would these have for archaeology <strong>of</strong> the<br />
Lapita era in Near-Oceania?
Population to be sampled:<br />
As for the Roviana survey, the Kaliquongu survey region was regarded as geographically<br />
representative <strong>of</strong> the New Georgia lagoonal system. <strong>The</strong> populations to be sampled were<br />
the intertidal/shallow water Lapita sites <strong>of</strong> New Georgia, and the intertidal post-Lapita<br />
early pottery sites.<br />
Data to be collected:<br />
<strong>The</strong> primary unit <strong>of</strong> measurement was the site, with <strong>of</strong>fsite data recorded as findspots. Sites<br />
were defined as high density scatters <strong>of</strong> artefacts, sufficiently numerous to allow a coarse<br />
ceramic periodization. This decision was taken with recording technology in mind: the<br />
relative remoteness <strong>of</strong> the location and lack <strong>of</strong> availability <strong>of</strong> electronic recording aids or<br />
power supplies dictated a reduction in the spatial resolution <strong>of</strong> data recording below what<br />
is fast becoming the norm in distributional or <strong>of</strong>f-site archaeology, i.e. point-provenancing<br />
<strong>of</strong> artefacts. Artefacts were collected rather than documented in situ, both to prevent rare<br />
categories <strong>of</strong> finds being missed due to marine encrustation, and to enable development <strong>of</strong><br />
classificatory systematics in the laboratory. Some sites were collected as single samples,<br />
some as samples from unequal sub-areas, one as a linear transect <strong>of</strong> 5x12m “squares”, and<br />
one as a series <strong>of</strong> 1m transects at 5m intervals, each comprising a linear arrangement <strong>of</strong><br />
5x1m “squares”.<br />
Degree <strong>of</strong> precision required:<br />
In terms <strong>of</strong> the hypothesis that Lapita is/is not present in the Western Solomon Islands, the<br />
184
prime difficulty in specifying the degree <strong>of</strong> precision required arises from unknown<br />
probability <strong>of</strong> detection. <strong>The</strong> existing knowledge used to formulate the expected Lapita site<br />
density was not sufficiently detailed regarding factors contributing to site detectability to<br />
allow any precise estimate <strong>of</strong> the size <strong>of</strong> a survey region required to establish the absence<br />
<strong>of</strong> Lapita. In this regard bigger is better theoretically, but in practice resources were<br />
limited, as was access, imposing limits on the size <strong>of</strong> the sample region that could be<br />
accessed and surveyed intensively in a manner that allowed biases to be assessed. <strong>The</strong><br />
review <strong>of</strong> 17 coastal surveys given above suggests for high-intensity survey Lapita site<br />
density could be around 91 sites per 1000km <strong>of</strong> coastline or upwards, or on average at<br />
least one site per 11km <strong>of</strong> coastline. This suggests that the survey region would need to<br />
be greater than 11km <strong>of</strong> high-detectability coastline favourable to Lapita settlement to<br />
have a reasonable chance <strong>of</strong> locating a Lapita site at the minimum recorded site density for<br />
intensive survey.<br />
Method <strong>of</strong> Measurement:<br />
Site locations were recorded on air photographs or 1:50 000 topographic maps, augmented<br />
on featureless coastlines with compass bearings to prominent landmarks, usually to nearby<br />
islets in the lagoon. Artefacts/sherds and manuports were bagged by intrasite transect or<br />
in the absence <strong>of</strong> transect collection, by site. Transects were set out using fibreglass<br />
measuring tapes and marked out with nylon mon<strong>of</strong>ilament on stakes. Field desalination and<br />
laboratory cleaning were important aspects <strong>of</strong> measurement, particularly <strong>of</strong> sherd<br />
decoration and temper characterization. Site density was calculated in relation to the<br />
185
distance <strong>of</strong> high-visibility shoreline surveyed, and as number <strong>of</strong> sites/findspots/lithic scatters<br />
per square kilometre.<br />
<strong>The</strong> Frame:<br />
<strong>The</strong> frame was as shown in Figure 6, which comprised the intertidal zone <strong>of</strong> the<br />
Kaliquongu region <strong>of</strong> the lagoon, chosen for reasons <strong>of</strong> access and because this appeared<br />
to be a favourable environment for settlement due to presence <strong>of</strong> a deepwater barrier-island<br />
passage and harbour, many small islets, reef shallows and seagrass flats, and a variety <strong>of</strong><br />
barrier-island and mainland resources within easy paddling distance <strong>of</strong> each other (Figure<br />
28). This area encompassed 70km 2 , approximately one third <strong>of</strong> Roviana Lagoon and<br />
approximately 8% <strong>of</strong> the total 900km 2 <strong>of</strong> the New Georgia lagoonal system.<br />
Selection <strong>of</strong> the sample:<br />
<strong>The</strong> sampling transect comprised the firm-sediment walkable sections <strong>of</strong> a natural<br />
landscape unit, the intertidal zone, as shown on Figure 28. <strong>The</strong> sampling fraction was<br />
intended to be 100% <strong>of</strong> the intertidal zone within the survey limits, but in practice some<br />
areas were impassable, and others were not covered in the time available. About 15km <strong>of</strong><br />
high-visibility intertidal was surveyed in this manner, and a further 10km or so <strong>of</strong><br />
mangroves and s<strong>of</strong>t sediments was surveyed where passable.<br />
<strong>The</strong> Pre-test:<br />
<strong>The</strong> 1996 field seasons had effectively provided a pre-test <strong>of</strong> methods. Site location<br />
186
technique was initially predominantly through reports by interested local residents or<br />
through fortuitous discovery during the process <strong>of</strong> traveling to villages by canoe to consult<br />
with communities regarding the archaeological survey. This was regarded as<br />
unsatisfactory, and the determination to take a more systematic sampling approach to the<br />
landscape developed out <strong>of</strong> it. <strong>The</strong> January/August surveys <strong>of</strong> 1996 involved a substantial<br />
terrestrial component, as did later work, which had shown a dominant terrestrial pattern<br />
<strong>of</strong> a thin, largely plain pottery style associated with late-prehistoric archaeological evidence<br />
(Sheppard et al. 1999:319).<br />
Organization <strong>of</strong> the Fieldwork:<br />
Survey was conducted during the season <strong>of</strong> low tides at the low-water spring tides <strong>of</strong> each<br />
month. This had a major beneficial effect on site detectability. Sites were surface-collected<br />
in most cases for laboratory classification <strong>of</strong> site type. Fieldwork assistants and guides<br />
were recruited from Sasavele village for the Kaliquongu survey and surface collections.<br />
Summary and Analysis <strong>of</strong> the Data:<br />
This was as for the Roviana survey, except site visibility was noted qualitatively in the field<br />
during survey rather than assessed from air photographs.<br />
Information Gained for Future Surveys:<br />
Time and cost information, intertidal site detectability under different conditions, intrasite<br />
archaeological potential and the effects on this <strong>of</strong> grab sampling, and the solutions to<br />
187
technical and logistical problems encountered during the survey, should all be <strong>of</strong> interest<br />
to archaeologists interested in the Lapita ceramic series in near-Oceania. This information<br />
will be <strong>of</strong> use to future researchers and is regarded as <strong>of</strong> immediate importance to cultural<br />
resource managers in the region.<br />
Results:<br />
Results <strong>of</strong> the Kaliquongu survey are shown in Figure 28. In tabulating these results<br />
(Table 6), two site types were lumped, aceramic lithic manuport scatters, and lithic<br />
scatters with rare ceramic finds. Two <strong>of</strong> the aceramic scatters recorded during the<br />
Kaliquongu survey may be <strong>of</strong> fluvial rather than anthropogenic origin, so the site density<br />
<strong>of</strong> these site types may be too high.<br />
Table 6: Site densities by period for Roviana and Kaliquongu surveys.<br />
Site Type Roviana<br />
High vis.<br />
(70km)<br />
Roviana<br />
low vis.<br />
(110km)<br />
Roviana<br />
Total<br />
(180km)<br />
188<br />
Kaliqu.<br />
High vis.<br />
(15km)<br />
Lapita 0 0 0 67/1000km<br />
n=1<br />
Post-<br />
Lapita<br />
129/1000km<br />
n=9<br />
9/1000km<br />
n=1<br />
56/1000km<br />
n=10<br />
333/1000km<br />
n=5<br />
aceramic 0/1000km 0 0 267-<br />
400/1000km<br />
n=4-6<br />
findspots 43/1000km<br />
n=3<br />
0 17/1000km<br />
n=3<br />
Kaliqu.<br />
Low vis.<br />
(10km)<br />
0 200/1000km<br />
n=2<br />
Kaliqu<br />
Total<br />
(25km)<br />
0/1000km 40/1000km<br />
n=1<br />
0/1000km 200/1000km<br />
n=5<br />
0 240/1000km<br />
n=6<br />
80/1000km<br />
n=2
Figure 28: Kaliquongu survey transects: the unfilled symbols represent sites discovered<br />
by informant-prospection during the Roviana survey.<br />
189
Discussion and Conclusions: the Two Surveys.<br />
Disregarding the emphasis on intertidal ceramics, and considering the Lapita site density<br />
figures obtained, the results <strong>of</strong> the Roviana survey do not seem significantly different to<br />
those <strong>of</strong> Terrell’s Bougainville survey. Where Terrell surveyed 86 km <strong>of</strong> coastline by the<br />
informant prospection method and found no Lapita, the Roviana Survey covered 180km<br />
<strong>of</strong> Roviana lagoon coastline by similar methods and found a site with a single dentate sherd<br />
(Nusa-Roviana-Zoraka). Neither do the results seem all that significantly different to<br />
results from Bellona, Guadalcanal or Santa Cruz. It must be borne in mind that the results<br />
<strong>of</strong> extensive, low-intensity survey using informant prospection will be highly variable<br />
depending on informant selection, as evidenced by clustered site distribution in the Roviana<br />
survey results (Sae Oka’s site discoveries on Ndora Island for example). Also, factors<br />
such as coastal site preservation and particularly intertidal site preservation, have not<br />
entered in any systematic way into discussions to date. Locating a single Lapita site in a<br />
survey is awkward in that site density calculation clearly suffers from inadequate site<br />
sample size. <strong>The</strong> distances surveyed in the Kaliquongu survey are low, resulting in this<br />
potential source <strong>of</strong> sampling error.<br />
Taking these potential sources <strong>of</strong> error in the Roviana and Kaliquongu survey data<br />
into account, and bearing in mind also the imprecise abstraction <strong>of</strong> the comparative data<br />
from the survey results <strong>of</strong> others, it seems impossible to conclude there is evidence for any<br />
significant difference in terms <strong>of</strong> site density between the Kaliquongu survey and the<br />
general Lapita record elsewhere in Near-Oceania, particularly if one allows for an<br />
190
unknown level <strong>of</strong> site preservation. If, for example, the aceramic lithic scatters at Mbolave<br />
and Elelo were at one time Lapita sites, Kaliquongu Lapita site density would jump to 200<br />
sites per 1000km We would have to have much more extensive data from the New<br />
Georgia Lagoon system encompassing a number <strong>of</strong> sites identifiably within the Lapita<br />
series and much more intensively and systematically surveyed data from other locations in<br />
near-Oceania to demonstrate significant differences in Lapita site density.<br />
<strong>The</strong> implication for the question <strong>of</strong> the presence <strong>of</strong> other Lapita-era sites in the<br />
New Georgia group is clear. A larger survey frame, with careful attention to models <strong>of</strong> site<br />
preservation and visibility, i.e. probability <strong>of</strong> site detection, is likely to locate additional<br />
sites from this general period. What should the sampling fraction be to obtain a sample<br />
suitable for seriation analysis? To answer this we must first ask how many Lapita sites<br />
might be expected from the New Georgia lagoon circuit in total (sampling fraction 100%)?<br />
<strong>The</strong> Kaliquongu survey <strong>of</strong> 70km 2 located one Lapita site, so a best guess for the 900km 2<br />
<strong>of</strong> coastlines and lagoons <strong>of</strong> the New Georgia lagoon circuit on current data would be<br />
around 13 sites using this measure. Similarly, roughly 65 post-Lapita sites might be<br />
expected, and about 78 aceramic sites. A corpus <strong>of</strong> ceramic samples <strong>of</strong> this magnitude<br />
would allow a much better seriation analysis, which together with testing by direct dating<br />
<strong>of</strong> styles could potentially enable a fine-grained chronology <strong>of</strong> the entire period <strong>of</strong> intertidal<br />
site formation. A chronology <strong>of</strong> this sort would allow abandonment <strong>of</strong> the sort <strong>of</strong> coarse<br />
Lapita/Post-Lapita chronological construct currently in use.<br />
It is hoped that this consideration <strong>of</strong> a sampling approach to regional<br />
archaeological questions in the near-Oceanic setting highlights the a-regional and overly<br />
191
culture-historical status quo in Lapita studies (in the published literature at least) which<br />
threatens to stifle the archaeological potential <strong>of</strong> the region by largely dismissing the bulk<br />
<strong>of</strong> the Lapita record, surface or turbated deposits, as “disturbed” and therefore <strong>of</strong> low<br />
archaeological value. It is hoped that this analysis <strong>of</strong> the Roviana and Kaliquongu surveys<br />
will promote re-evaluation <strong>of</strong> the New Georgia early-ceramic surface record as the most<br />
complete witness <strong>of</strong> the history <strong>of</strong> early ceramic variability. Such a re-evaluation <strong>of</strong> survey<br />
objectives beyond a search for “good” (stratified) sites demands as a prerequisite<br />
acceptance <strong>of</strong> the tenet that survey design matters, a simple point, but one that has largely<br />
failed to penetrate Lapita studies.<br />
Uncontrolled direct reading <strong>of</strong> site densities in a comparative manner (as done by<br />
Anderson 2002) is counterproductive when the fundamental inferential issues <strong>of</strong> site<br />
preservation and visibility, or site detectability in relation to survey methods, are simply<br />
ignored in the rush to calculate higher-level behavioural variables like migration rates. <strong>The</strong><br />
poor type and extent <strong>of</strong> our sampling becomes evident when a sample-surveying<br />
perspective is engaged. If we wish to talk about migration rates, or demographic history,<br />
or colonization rates, or the distribution <strong>of</strong> “Lapita”, or when “Lapita” ends, research<br />
design which obtains data matched to the research question is needed, which for any <strong>of</strong><br />
these behavioural questions requires that we concentrate on the difficult formational<br />
questions <strong>of</strong> evidence, such as, “what is the probability <strong>of</strong> a pottery sample being<br />
preserved in a particular location and observed by the survey method?’, rather than easy<br />
questions, like “did we find anything?<br />
192
Introduction:<br />
CHAPTER 4:<br />
CERAMICS: UNITS OF DESCRIPTION<br />
<strong>The</strong> units <strong>of</strong> description employed for ceramic form and decoration were designed to<br />
enable a detailed description <strong>of</strong> ceramic variation at the level <strong>of</strong> the sherd. <strong>The</strong> problem <strong>of</strong><br />
assessing sample formation processes required careful attention to the description <strong>of</strong> sherds<br />
as sedimentary particles. Thus sherd area, mass, thickness and composition were recorded<br />
for all sherds. Description <strong>of</strong> decoration was detailed to the level <strong>of</strong> the layout, size and<br />
spacing <strong>of</strong> decorative elements, in an attempt to capture variation to the level <strong>of</strong> individual<br />
potters if necessary, in the process <strong>of</strong> deciding on units <strong>of</strong> classification (classification is<br />
covered in Chapters 8 and 9).<br />
<strong>The</strong>re is a descriptive and classificatory divide between Lapita and post-Lapita<br />
pottery (discussed in Chapter 1), in that the Frost-Irwin attribute-combination grouping<br />
approach has generally been applied to materials post-dating Lapita, while the Mead<br />
system <strong>of</strong> classification, or its non-classificatory splitting variant, the Anson system <strong>of</strong><br />
description, have dominated Lapita studies. Studying the transition between Lapita and<br />
post-Lapita requires a descriptive system that is capable <strong>of</strong> dealing both with complex<br />
motifs and with simple geometric pattern such as bands <strong>of</strong> repeated elements.<br />
<strong>The</strong> difference between Lapita and post-Lapita is already well-understood: Lapita<br />
has complex decoration including dentate-stamping in a range <strong>of</strong> designs and a diversity<br />
<strong>of</strong> (<strong>of</strong>ten slab constructed) vessel forms, some <strong>of</strong> which are characteristic. Post-Lapita<br />
lacks these distinguishing criteria (as might pre-Lapita). If at the end <strong>of</strong> this thesis I<br />
conclude there was a transition from dentate to incised and simplification in vessel form<br />
I would be merely restating the definition <strong>of</strong> Lapita rather than attempting to understand<br />
the details <strong>of</strong> ceramic change. Description in a context <strong>of</strong> reduced decorative complexity<br />
still needs to be sensitive to the temporal dimension <strong>of</strong> variability. For Roviana pottery<br />
this required that careful attention be paid to decorative location, as the sensitivity <strong>of</strong><br />
193
simple decorative techniques and simple patterns in themselves to change was not expected<br />
to be sufficient to capture fine temporal patterning. <strong>The</strong> structure <strong>of</strong> decoration across the<br />
vessel and in relation to vessel form was therefore <strong>of</strong> interest.<br />
Understanding why change, as expressed in the historical events <strong>of</strong> pot-making,<br />
happened, requires a materialist approach to the analysis <strong>of</strong> variability (Dunnell 1986:153).<br />
An inventory <strong>of</strong> designs as in the motifs <strong>of</strong> the Anson system is not sufficient when the goal<br />
is to understand design change. <strong>The</strong> units <strong>of</strong> description must reconstruct design into<br />
component attributes that are salient to design evolution. <strong>The</strong> process <strong>of</strong> how one design<br />
may have changed into another is <strong>of</strong> interest. Also, we have no reason to expect all<br />
attributes <strong>of</strong> designs to change at the same rate. Some aspect <strong>of</strong> vessel design and<br />
decoration might be highly constrained by functional adaptation or may be changing rapidly<br />
as a result <strong>of</strong> some selective pressure. Also some form or decorative attributes may be free<br />
to change stochastically, i.e. drift, in the sense <strong>of</strong> genetic drift (Dunnell 1978, 2001).<br />
Sufficient sherds in the overall sample were large enough to permit such a<br />
structural approach to the patterned layout <strong>of</strong> elements and the structure <strong>of</strong> this patterning<br />
across the vessel, and some designs were compact and simple enough to allow an<br />
element/motif design analysis. This is not to imply that elements and motifs are always<br />
hierarchically structured: some patterns are constructed as a series <strong>of</strong> repeats <strong>of</strong> an<br />
element, while some designs are more complex, and are just inventoried, being classified<br />
in the next <strong>chapter</strong> according to criteria such as zone markers (Mead 1975:24). Vessel<br />
morphology is an aspect <strong>of</strong> design that, like decoration, may be constrained or directed in<br />
some ways but free to shift or drift in others. Analysis <strong>of</strong> vessel form, using mainly interval<br />
measures like curve template fitting, was designed to capture slight variation, and enable<br />
analysis within the theoretical framework <strong>of</strong> form variability in Chapter 1.<br />
Quantitative analysis <strong>of</strong> vessel design from sherds <strong>of</strong> broken pottery requires<br />
specification <strong>of</strong> a relationship between the pieces and inferred wholes. Some <strong>of</strong> the units<br />
<strong>of</strong> description pertain to this requirement, and are used extensively in Chapter 5 to<br />
ascertain the most appropriate method <strong>of</strong> quantification. In addition to behavioural-<br />
historical and behavioural-explanatory objectives, units <strong>of</strong> description were designed with<br />
194
the objective <strong>of</strong> understanding other non-cultural aspects <strong>of</strong> site/assemblage formation<br />
processes. Accordingly, several measures <strong>of</strong> sherd size and thickness are recorded, in<br />
more detail than usual, for all sherds, allowing detailed taphonomic study in Chapters 5,<br />
6, 7 and 11.<br />
Database Structure:<br />
Figure 29: Data structure for each sherd record; each sherd can have many<br />
records in the detail tables pertaining to the various parts <strong>of</strong> the vessel<br />
represented.<br />
<strong>The</strong> core ceramic relational database used for description <strong>of</strong> sherd attributes consisted <strong>of</strong><br />
a linked arrangement <strong>of</strong> tables, with the master record linked in a one-to-many relationship<br />
to thematic tables about the sherd (Figure 29).<br />
A master table contained data pertaining to the entire sherd. Fields <strong>of</strong> data in the<br />
master record for each sherd were about sherd identity and provenance, area, weight, date<br />
recorded, time recorded, general purpose notes, fabric description code, and Estimated<br />
Vessel Equivalent (EVE) (after Orton 1993), (this is the same measurement as Egl<strong>of</strong>f’s<br />
“Percentage Factor”) (Egl<strong>of</strong>f 1973) (Table 7).<br />
195
Table 7: Structure <strong>of</strong> master record for each sherd.<br />
FIELD PURPOSE ATTRIBUTE /<br />
UNIT<br />
sherd ID identify site, unit<br />
and sherd<br />
196<br />
7-character<br />
alphanumeric<br />
METHOD OF<br />
MEASUREMENT<br />
re-join any postrecovery<br />
sherd<br />
breaks<br />
sherd area measure sherd size cm 2 grid transparency<br />
sherd weight measure sherd size g balance<br />
date recorded quality control date automatic input<br />
time recorded quality control time automatic input<br />
notes general purpose text<br />
fabric description preliminary<br />
variability<br />
special code iterative qualitative<br />
EVE measure sherd size percent rim percent chart<br />
Vessel Family link to other sherds<br />
potentially from<br />
the same vessel<br />
vessel family<br />
identity number<br />
iterative<br />
comparison<br />
(methods detailed<br />
in Chapter 5)<br />
Data pertaining to the various vessel parts represented on the sherd was distributed<br />
across a number <strong>of</strong> tables. A table <strong>of</strong> sherd thickness for each <strong>of</strong> the vessel parts<br />
represented on the sherd was related to the master table via the sherd identity code.<br />
Similarly, a range <strong>of</strong> data about vessel form for each part <strong>of</strong> the vessel represented on the<br />
sherd was related to the master table via the sherd identity code. A fourth table comprised<br />
information about decoration on each part <strong>of</strong> the vessel, again related to the master record<br />
for the sherd via the sherd identity code. <strong>The</strong>se tables are appended on CD (Master.db,<br />
Thickness.db, Form.db and Decoration.db).<br />
Master Record Table: Explanation <strong>of</strong> Data Codes for Each Field:<br />
<strong>The</strong> sherd ID field was a 7-character assignment with the following site prefixes used:<br />
R37 Paniavile (2 nd collection) (R37 had two redundant unit characters <strong>of</strong> theform1a, 2a,
P Paniavile (1 st collection)<br />
C Paniavile (villager collection stored in cave)<br />
A Zangana (1 st collection)<br />
Z Zangana (2 nd Collection)<br />
B Zangana (3 rd collection)<br />
GH Gharanga (1 st collection)<br />
GE Gharanga (2 nd collection- east <strong>of</strong> stream)<br />
GW Gharanga (2 nd collection- west <strong>of</strong> stream)<br />
MH Miho collection<br />
HV Honiavasa collection<br />
HG99 Hoghoi (1 st collection)<br />
HG Hoghoi (2 nd collection)<br />
K Kopo collection<br />
NR Nusa Roviana collection (Zoraka)<br />
etc (so that R371a*** is from the same collection event and<br />
spatial unit as R372a***, R373a*** etc.)<br />
Sherds prefixed P, C, A, B, GH, GE, GW, MH, HG99, K, and NR were from single-unit<br />
collections. For these sherds the remaining digits indicate sherd identity within the unit.<br />
Sherds prefixed Z, HV and HG have a collection unit number following the prefix, as<br />
follows.<br />
Zangana: Zxxxyyy where xxx is a three digit unit identifier and yyy is the sherd<br />
number.<br />
Honiavasa: HV0xyyy where 0x is unit 01-05 and yyy is sherd identity<br />
Hoghoi :HGxxsyy or HGxxayy where ‘xx’ is the unit, ‘yy’ or ‘yyy’ is the sherd<br />
number, ‘s’ means subsurface sample and ‘a’ means additional collection<br />
(a control for additional time spent in collection, to prevent intensity <strong>of</strong><br />
collection bias in the sherd density pattern across the site). <strong>The</strong> usual form<br />
is HGxxyyy. Where multiple fabric readings were taken as a quality control<br />
measure there are additional records in the descriptive database with ‘b’,<br />
‘c’ or ‘a’ added into the unit number or sherd number in place <strong>of</strong> a zero.<br />
197
Sherd Size:<br />
Sherd size was recorded as sherd area, sherd weight and in the case <strong>of</strong> lip and carination<br />
sherds, estimated vessel equivalent (EVE) (Egl<strong>of</strong>f 1973, Orton 1982, 1993, Orton & Tyers<br />
1991). Sherd area was counted using a gridded PVC transparency, which could partially<br />
conform to sherd external curvature. Measurement was made on the exterior <strong>of</strong> sherds, as<br />
a count <strong>of</strong> square centimetres, with an accuracy estimated at ± 10%. Sherd weight was<br />
recorded to the nearest gram using Swedish rounding in general, although initially decimal<br />
places were recorded also. Date and time were input automatically by the database and<br />
were used to correct some temporal inconsistencies in data recording. An additional<br />
measure in this regard was the re-analysis <strong>of</strong> the Paniavile assemblage which had been the<br />
first assemblage analyzed. A “notes” field was used primarily in initial database<br />
development, to record additional information usually coded subsequently in added fields.<br />
Fabric/Temper:<br />
Fabric descriptions were initially made using a point-counting system, which proved too<br />
time-consuming; and also cluster analysis on the results was inconsistent with sherd<br />
groupings established by visual comparisons. This initial approach was discarded. Fabric<br />
grouping <strong>of</strong> sherds was subsequently done using iterative manual sorting and visual<br />
comparison. A corner <strong>of</strong> each sherd was snapped, and most sherds had previously been<br />
acid-etched in the process <strong>of</strong> cleaning marine encrustation. Examination <strong>of</strong> etched sherd<br />
surfaces and the fresh un-etched break allowed reliable discrimination between calcite<br />
fragments, quartz/feldspar fragments, volcanic rock fragments, and Ferromagnesian<br />
mineral fragments, when examined under 10x binocular magnification.<br />
Iterative comparisons were used to develop a fabric type-series, but this also<br />
proved overly time-consuming. An ordinal notational system describing type and size <strong>of</strong><br />
inclusions in declining order <strong>of</strong> abundance was finally used. <strong>The</strong> approach is convergent<br />
with that <strong>of</strong> Wickler (Wickler 2001: 97-99). Major classes <strong>of</strong> grain identifiable under<br />
binocular microscope are listed in Table 8.<br />
198
Table 8: Classes <strong>of</strong> mineral identified at 10x magnification in reflected light.<br />
v ferromagnesian minerals ( opaque ferromagnesian, pyroxene, olivene,<br />
hornblende) black or green<br />
c calcite reef detritus, white/grey, s<strong>of</strong>t, <strong>of</strong>ten partially sintered and easily<br />
acid-etched<br />
q hard translucent mineral grains: quartz or feldspar<br />
t polycrystalline lithic fragments, usually volcanic, black-grey and become<br />
light grey-silver when etched<br />
w white volcanic glass fragments<br />
r red volcanic glass fragments<br />
b black volcanic glass fragments<br />
A microscope eyepiece was fitted with a 10x10 grid <strong>of</strong> 0.5mm squares to aid grain size<br />
estimation. Size descriptor suffixes for the mineral groups were as follows: x (poorly<br />
sorted sand incorporating a range <strong>of</strong> sand-sized grains from 0.1mm to 2mm), l (large), m<br />
(medium), s (small) and f (very fine). <strong>The</strong>se size classes corresponded to >0.5mm, 0.49-<br />
0.25mm, 0.24-0.1mm and
Table 9: Examples <strong>of</strong> descriptive syntax for tempers.<br />
VXCS Volcanic-Calcite hybrid temper comprising dominant Volcanic<br />
ferromagnesian minerals <strong>of</strong> miXed size range with subordinate Calcite<br />
(Small) reefal detritus grains<br />
CSQM Quartz-Calcite hybrid temper comprising Calcite reef detritus (Small)<br />
dominant, with subordinate Quartz/feldspar grains (Medium-sized)<br />
VFSMCF Volcanic particles (very Fine, probably opaques) dominant with a few<br />
Small sized and occasional Medium-sized volcanic particles,<br />
subordinate Calcite grains (very Fine)<br />
established for sites other than Miho and Honiavasa, and for the “P***” sample from<br />
Paniavile, which was etched. <strong>The</strong>se two Groups (1 & 3) while distinct in the sample sent<br />
to Dickinson, form something <strong>of</strong> a continuum, where some sherds are intermediate<br />
between the types, suggesting a more moderate degree <strong>of</strong> placer effect than seen in the<br />
placer volcanic samples. Groups 1 and 3 are therefore regarded as subgroups <strong>of</strong> a single<br />
class.<br />
Group 2, un-placered hornblendic-feldspathic volcanic sand tempers, include sherds<br />
<strong>of</strong> similar unplacered feldspathic composition, but the relative abundance <strong>of</strong> hornblende is<br />
thought to be highly variable within this group. <strong>The</strong> sherds examined by Dickinson and<br />
assigned to this group were all relatively rich in hornblende, and were interpreted as most<br />
likely from Vella Lavella. Many sherds in this group have pyroxene rather than hornblende<br />
as the predominant ferromagnesian mineral, as far as could be established from<br />
examination <strong>of</strong> sherd surfaces under 10x binocular magnification, so this group should be<br />
regarded as comprising a particularly low level <strong>of</strong> placering, consistent with stream sands,<br />
rather than as being entirely exotic. Some sherds within this grouping might be exotic to<br />
the New Georgia Island Group, but many could turn out to be local were further<br />
petrographic analysis carried out.<br />
200
Low-power reflected-light discrimination between Groups 4, 5, and 6 was problematic in<br />
Table 10: Temper groupings after Dickinson 2000.<br />
Group 1 Ferromagnesian Placer Volcanic sand tempers<br />
Group 2 Unplacered Hornblendic-Feldspathic Volcanic Sand Tempers<br />
Group 3 Pyroxenic-Lithic Volcanic Sand Tempers<br />
Group 4 Quartz-Bearing Volcanic Sand Tempers<br />
Group 5 Quartz-Calcite Hybrid Sand Tempers<br />
Group 6 Calcite Tempers (some with possible subordinate fine ferromagnesian<br />
component)<br />
Group 7<br />
(Group 1?)<br />
Placered volcanic with plagioclase-rich microlitic rock fragments<br />
Group 8 Temper not recorded<br />
many cases. While there were many examples <strong>of</strong> a distinctively quartz-calcite hybrid<br />
temper (Group 5), there were other sherds with a greater proportion <strong>of</strong> volcanic fragments,<br />
which tended towards the volcanic-quartz grouping (Group 4). Also, there were quartz-<br />
calcite sherds with very low amounts <strong>of</strong> quartz, difficult to detect in sherd surfaces under<br />
binocular magnification, but which, when analyzed petrographically, turned out to have a<br />
terrigenous fraction <strong>of</strong> similar composition to the quartz-calcite sherds (see Appendix 1:<br />
Table 196-5) It seems clear that varying proportions <strong>of</strong> terrigenous component to calcite<br />
reef detritus component can account for such a variety <strong>of</strong> visual temper characteristics yet<br />
allow derivation <strong>of</strong> these materials from a single coastline (Felgate & Dickinson 2001).<br />
Equally however, it is possible that some <strong>of</strong> the calcite-tempered sherds for which a<br />
quartzo-feldspathic fraction cannot be discerned in the sherd surface could be from a non-<br />
quartzose source area. Further petrographic quality-control is necessary to assess the<br />
situation in this regard between Groups 5 and 6.<br />
Group 7 in the database refers to a fabric in which placered volcanic temper is<br />
dominant, but in which subordinate dark grey microlitic rock fragments occur. Due to<br />
201
constraints on sample size, none <strong>of</strong> these sherds were included in the sample analyzed by<br />
Dickinson. <strong>The</strong>se are provisionally considered to be a less placered variant <strong>of</strong> the<br />
ferromagnesian placered volcanic group (Group 1) with affinities also to pyroxenic-lithic<br />
volcanic sand tempers (Group 3) and thus intermediate between Groups 1 and 3, although<br />
calcite grains were rare or absent in Group 7, suggesting a separate origin.<br />
Group 8 comprises those sherds for which fabric was not recorded, usually due to<br />
difficulties discerning temper type under binocular magnification, most commonly as a<br />
result <strong>of</strong> a dark reduction core which hid darker-coloured mineral grains in reflected light.<br />
Vessel Parts:<br />
Definition <strong>of</strong> vessel part terminology is fundamental to the ceramic data structure (Figure<br />
29, Figure 30). In some instances these vessel part definitions seem more specific than<br />
those <strong>of</strong> potters in the past, as will be argued in Chapter 9. Identification <strong>of</strong> vessel part for<br />
smaller sherds was not always possible, and sometimes a default assignment was used<br />
when a sherd met some criterion. For example all sherds <strong>of</strong> hyperboloid form (Rice<br />
1987:219) other than those from Honiavasa conical-shouldered carinated jars were<br />
classified as rims.<br />
Thickness, Form and Decoration <strong>of</strong> Various Vessel Parts Represented on Sherds:<br />
Three tables containing information on thickness, form and decoration were held in a one-<br />
to-many relationship with the master table via the sherd ID field. This allowed information<br />
pertaining to multiple specific parts <strong>of</strong> the vessel to be recorded separately, yet as a group<br />
for that sherd. By way <strong>of</strong> example, for a sherd where the lip, rim and neck were<br />
represented, the thickness table could contain measurements taken according to a set <strong>of</strong><br />
conventions at each <strong>of</strong> these points on the vessel, allowing thickness trends to be<br />
calculated for that sherd, and also allowing comparison <strong>of</strong> assemblages in terms <strong>of</strong><br />
location-controlled thickness measurements. Similarly, for vessel form, characteristics <strong>of</strong><br />
the lip, rim, neck, shoulder, carination and body could be recorded where present for each<br />
sherd, allowing virtual reconstruction <strong>of</strong> sherd form from the database, and also allowing<br />
202
Figure 30: Major vessel form variants showing part<br />
terminology; L=lip, R=rim, N=neck, S=shoulder,<br />
C=carination, B=body; inverted or unrestricted vessels have<br />
no neck, while for restricted vessels with everted rims (the<br />
first seven) the only distinction in neck types is between the<br />
double neck (top centre) and the single neck (including all<br />
unlabelled).<br />
detailed comparison <strong>of</strong> assemblage variability controlling for position on the vessel.<br />
Decoration was recorded in the same way, i.e. the decorative characteristics <strong>of</strong> each vessel<br />
part represented on the sherd were recorded in some detail. <strong>The</strong>re could be more than one<br />
decoration record for one part <strong>of</strong> a sherd, provided that the decorative technique differed,<br />
as the key for the decoration table was a composite <strong>of</strong> more than one column.<br />
A similar effect could have been achieved using a single flat table <strong>of</strong> data with a<br />
203
large number <strong>of</strong> attributes or variables, but this would have been unwieldly for data entry,<br />
requiring a multi page electronic data entry form to achieve the same level <strong>of</strong> detail as<br />
entered in the relational system.<br />
Thickness: For the table <strong>of</strong> sherd thicknesses by vessel part, the “ID” field linked<br />
to the master table record for that sherd (Table 11). “Part” could be either the lip (l),<br />
rim(r), neck(n), shoulder(s), carination(c), body(b), base(a) or unknown(u). “Part”<br />
comprised a secondary index, to enable only one measurement for each part, and to give<br />
each part measurement a unique identity. Thickness was measured according to the rule<br />
for that part, using digital calipers and measured in mm, using Swedish rounding in most<br />
cases, although some initial measurements recorded thickness to one decimal place.<br />
Table 11: Data structure <strong>of</strong> table <strong>of</strong> thicknesses for<br />
each part <strong>of</strong> the sherd.<br />
Field Name Type Size Key Required Value<br />
Id A 7 Yes<br />
Part A 1 Yes Yes<br />
Thickness N Yes<br />
Thickness at the lip was measured perpendicular to the tangent <strong>of</strong> the interior surface <strong>of</strong><br />
the rim where it meets the lip. Thickness <strong>of</strong> the rim was measured equidistant between the<br />
lip and the point <strong>of</strong> maximum vertical curvature <strong>of</strong> the neck (minimum vertical radius). In<br />
the case <strong>of</strong> broken rims, where the full depth was not present, thickness was measured at<br />
the lowest point. Thickness measurements were an average <strong>of</strong> the reading from both ends<br />
<strong>of</strong> the sherd (in the horizontal dimension) where the sherd could be oriented. For sherds<br />
<strong>of</strong> unknown part, which could not be oriented, thickness was recorded as the average <strong>of</strong><br />
several caliper readings across the sherd.<br />
204
On the data entry form for the sherd, on which the thickness table resided, once a part had<br />
been entered, a thickness for that part was required by the database, and use <strong>of</strong> the<br />
secondary index meant that only one record per part could be entered.<br />
Vessel Form:<br />
Vessel forms were recorded using interval variables as far as possible (the set <strong>of</strong> curve-<br />
fitting templates used provided the intervals). Secondary classification was undertaken<br />
when these observation suggested essential classes in the data as opposed to continuous<br />
variation. While some vessel form descriptive schemes measure the orientation <strong>of</strong> the rim<br />
and the angle between neck and shoulder in some manner (see Irwin 1972:60: inclination<br />
angle), the effects <strong>of</strong> both assemblage brokenness and rim/neck/shoulder curvature on the<br />
measured angle are seldom discussed. Also, necks tend to be treated as though they are<br />
the apex <strong>of</strong> two straight lines, these lines being the vertical sections <strong>of</strong> the rim and<br />
shoulder. In reality, these sections are made up usually <strong>of</strong> a series <strong>of</strong> complex curves rather<br />
than straight lines and corners.<br />
<strong>The</strong> approach taken in this study was to regard these complex curves as reduceable<br />
not to a series <strong>of</strong> straight lines <strong>of</strong> measured length and angles, but to a series <strong>of</strong> simple arcs<br />
<strong>of</strong> circles <strong>of</strong> measurable length, oriented by measurable angles in relation to each other.<br />
In my opinion this approach provides a better approximation <strong>of</strong> the actual shape <strong>of</strong><br />
ceramic vessels in vertical pr<strong>of</strong>ile, and therefore does a better job <strong>of</strong> describing variation<br />
in vessel form. This approach is less sophisticated than the two-curvature method <strong>of</strong><br />
Hagstrum and Hildebrand, where the exterior vertical section can be represented as a<br />
continuously varying curve (Hagstrum & Hildebrand 1990), but at the level <strong>of</strong> sherd<br />
assemblage description there is substantial convergence between my approach and theirs.<br />
Curvature measurements were taken at various parts <strong>of</strong> the vessel as represented by<br />
sherds, and covariation between interval attributes was assessed by means <strong>of</strong> bivariate<br />
plots.<br />
205
Field<br />
Name<br />
Type Size Key _Require<br />
d Value<br />
Id A 7 Yes Yes<br />
Part A 1 Yes Yes<br />
Type A 1<br />
Angle N<br />
Vcurve N<br />
Hcurve N<br />
Sherdfor<br />
m notes<br />
Curvatures in all cases were measured using brass radius templates, manufactured<br />
by the <strong>Auckland</strong> <strong>University</strong> Engineering Workshop for the Anthropology Department.<br />
Internal radii <strong>of</strong> the templates, used to measure the horizontal curvature <strong>of</strong> the exterior <strong>of</strong><br />
the vessel (including lip Hcurve), ranged from 50 to 300 mm in increments <strong>of</strong> 10mm (5mm<br />
increments were manufactured but provided spurious accuracy given the hand-formed,<br />
irregular nature <strong>of</strong> the pottery). <strong>The</strong> brass templates had a width <strong>of</strong> 7mm, so external radii<br />
measurable by the set <strong>of</strong> (concave) curves ranged from 57 to 307 mm, also in 10mm<br />
increments as used.<br />
Data structure for vessel form (Table 12) used “Id” and “part” information as for<br />
the thickness table. How the remaining fields were filled out depended on both the part and<br />
the part form type (the first column is the list <strong>of</strong> data fields which form the columns in the<br />
data table, the other columns detail properties <strong>of</strong> each data field).<br />
Variations in the way different vessel part forms were recorded are listed below<br />
Table 12: Data structure for the table <strong>of</strong> records <strong>of</strong> sherd form attributes by vessel part.<br />
A 240<br />
206<br />
_Min<br />
Value<br />
_Max<br />
Value<br />
Rimdepth N 0.00 200.00
under sub-headings:<br />
Lip Form:<br />
For lips, nine lip “type” options were used (Figure 31), with remaining fields filled out the<br />
same regardless <strong>of</strong> type. Lip angle was measured to the nearest ten degrees using a<br />
goniometer, as the angle between the uppermost tangent to the interior surface <strong>of</strong> the rim,<br />
and the innermost facet <strong>of</strong> the lip, except in the case <strong>of</strong> rounded lips, for which lip angle<br />
was not measurable. Measuring this angle, and another which gave the degree <strong>of</strong> rim<br />
eversion, allowed a reduction in the number <strong>of</strong> lip types recorded below the norm for<br />
Pacific potsherd studies, without loss <strong>of</strong> information on variability.<br />
<strong>The</strong> “Vcurve” field (Vertical curvature) was not used for describing lip form. <strong>The</strong><br />
“Hcurve” field (Horizontal curvature) was measured by curve fitting for 45 lips, these<br />
tending to be the larger sherds where the measurement is less likely to be spurious. Of the<br />
Figure 31: Lip form variants showing<br />
database codes<br />
207
42 sherds in this category for which EVE was recorded, EVE ranged from 3% to 17%,<br />
with an average value <strong>of</strong> 7.2%. A larger number <strong>of</strong> lip sherds were measured in this way<br />
during preparation <strong>of</strong> sherd drawings, where multiple measurement points were used to<br />
establish the best location <strong>of</strong> the central vertical axis (CVA) <strong>of</strong> vessels, but these<br />
measurements, while shown on some sherd illustration figures, were not entered into the<br />
database. In the drawings, irregular Hcurves which did not accurately conform to any <strong>of</strong><br />
the curve templates were shown as a dashed CVA estimate, while solid lines indicated a<br />
good circular fit.<br />
“Sherdform notes” was a general purpose field used particularly in the development<br />
stages <strong>of</strong> database design to note any data that required systematization. “Rim Depth” was<br />
a general purpose field applicable to rims, some types <strong>of</strong> neck, and shoulders, but was not<br />
used for lip data.<br />
Rim Form:<br />
Sherd ID and “part” were used as for lips. Eight rim types were distinguished (Figure 32),<br />
and measurement rules for rims varied slightly with rim type in some cases. Generally<br />
speaking, the rim descriptive terminology <strong>of</strong> Snow was used (Snow & Shutler 1985:20)<br />
which is similar to that used by Irwin (Irwin 1972:55, 1985). Rims which diverge with<br />
height from the CVA <strong>of</strong> the vessel are referred to as everted. Rims which approach the<br />
CVA with increasing height are inverted, while those which do neither are vertical. Snow<br />
and Shutler provide a finer descriptive terminology that subdivides everted rims into<br />
straight everted, excurvate everted, or flared everted, excurvate having a concave exterior<br />
208
vertical contour (hyperboloid form) and flared being convex in that plane. <strong>The</strong> “type” field<br />
was sufficient to encompass all these variants, in addition to some sub-types <strong>of</strong> everted<br />
rims. Eight rim types were defined (Figure 32). Rim pr<strong>of</strong>ile (Snow & Shutler 1985:22)<br />
was not recorded, as thickness measurements for lip, rim and neck captured this variation.<br />
For excurvate rims (x) or straight rims (s), where the lip was present, rim angle<br />
was recorded as the angle between the CVA and the interior tangent (in vertical contour)<br />
at the lip. Where local pr<strong>of</strong>ile concavity was present through thickening <strong>of</strong> the lip, the<br />
goniometer would sit stable on two points, while on convex-pr<strong>of</strong>ile interiors, the<br />
Figure 32: Rim types (database codes shown and<br />
measurement method for rim angles).<br />
209
goniometer would be positioned so that it contacted as a tangent below the lip, and was<br />
then rocked outwards until the point <strong>of</strong> contact reached the top <strong>of</strong> the rim (see “rim slope”<br />
in Smith 1985:263). For flared everted rims (f), the goniometer could be rested stable<br />
across two contact points, one at the neck and one at the rim.<br />
This process was not particularly accurate, but had the advantage <strong>of</strong> being<br />
measurable independent <strong>of</strong> the neck (except in the case <strong>of</strong> flared rims), and required only<br />
that there was enough <strong>of</strong> the lip present to allow the sherd to be oriented, itself a<br />
procedure fraught with error, particularly with uneven hand-formed pottery. This meant<br />
a larger sample <strong>of</strong> measurements could be obtained, as sherds having both the lip and the<br />
neck present were rare. This measurement is similar to the more commonly measured<br />
“orientation angle” (see for example Irwin 1972:60) but not the same, except in the case<br />
<strong>of</strong> flared or straight rims. Where the rim is excurvate this rim angle measurement will be<br />
substantially wider than Irwin’s measure <strong>of</strong> orientation angle, dependent on the amount <strong>of</strong><br />
excurvature as expressed by the Vcurve measurement (see below) and the depth or height<br />
<strong>of</strong> the rim.<br />
For excurvate everted rims (x), brass curvature templates were used to determine<br />
the best expression <strong>of</strong> concavity in the vertical plane (Rim Vcurve, Figure 33), recorded<br />
as radii in millimetres, in increments <strong>of</strong> 10mm. <strong>The</strong> smallest curve template was 47mm,<br />
curves <strong>of</strong> smaller radius were estimated using a ruler or gridded transparency. For Flared<br />
rims (f) the curve measurement indicates the radius <strong>of</strong> concavity in the vertical plane. If<br />
both lip and neck were present, horizontal curvature (Hcurve) was sometimes recorded<br />
210
midway between lip and neck, on the exterior surface. No analytical use was made <strong>of</strong> this<br />
measurement in the present study. Rim depth (Figure 33) was measured in millimetres<br />
with calipers between the outer edge <strong>of</strong> the lip and the topmost point <strong>of</strong> maximum<br />
curvature (minimum radius) <strong>of</strong> the neck, where the Vcurve template would be tangential<br />
to the Vcurve template <strong>of</strong> the neck (see below). Where either the neck or the lip was not<br />
present rim depth was noted as greater than the measured distance to the broken vertical<br />
boundary <strong>of</strong> the rim, and recorded as a “Sherdform note” rather than in the “Rimdepth”<br />
field.<br />
Taken together with the lip and rim thickness measurement, these approximately<br />
reconstruct the form <strong>of</strong> rims, and therefore allowed a more realistic record <strong>of</strong> rim form<br />
variability, more likely to capture subtle variation than previous schemes, while boosting<br />
sample size. As discussed in the preamble to the ceramic analysis, this level <strong>of</strong> detail<br />
seemed desirable in attempting to characterised stylistically depauperate assemblages, and<br />
to analyse variability in detail. For more complex rim forms, hard flare (h), compound(c),<br />
hard compound (k) or modeled (m), measurements were made in the same way, but on the<br />
upper portion <strong>of</strong> the rim only, so that measurements could be obtained for more broken<br />
examples where the neck was not present. In such cases rim depth too was measured from<br />
the outer edge <strong>of</strong> the lip to the first corner point rather than to the neck.<br />
In practice, the advantage <strong>of</strong> being able to measure rim angle independent <strong>of</strong> the<br />
neck was mitigated slightly by the difficulty <strong>of</strong> orienting more broken sherds which did not<br />
include the neck. For some <strong>of</strong> the slab-built Honiavasa pottery though, which tended<br />
211
to have broken at the neck, this system allowed more measurements than it would have<br />
if Irwin’s or Snow’s system had been used.<br />
Neck Form:<br />
<strong>The</strong> vessel neck in this study was defined as the orifice restriction between body and rim<br />
(Figure 30). Unrestricted vessels <strong>of</strong> incurvate or direct form (Snow & Shutler 1985:20)<br />
therefore do not have a neck. <strong>The</strong> double neck vessel form (d) (Figure 30) is the only neck<br />
type with depth, and for which neck depth must be recorded to record the shape <strong>of</strong> the<br />
vessel. In double necks an upper and lower tight radius are separated by a distance, over<br />
which the vertical external section is roughly straight. Other neck types are restricted (r)<br />
or unrestricted (u), meaning there is no neck or a slight inflection (indirectness) in the<br />
external pr<strong>of</strong>ile. <strong>The</strong>se are described only in terms <strong>of</strong> neck type, angle (Figure 33), Vcurve<br />
(Figure 33) and Hcurve. <strong>The</strong> neck seems to have been a rugged portion <strong>of</strong> the<br />
Figure 33: Measurement <strong>of</strong> rim depth, rim Vcurve, neck<br />
angle and neck Vcurve at different levels <strong>of</strong> brokenness.<br />
212
vessel, well represented in the sherd assemblages in most <strong>of</strong> the site samples, and<br />
conveying considerable information on vessel form when thus described.<br />
Neck Hcurve was measured on the external surface <strong>of</strong> the vessel. Where sherds<br />
were smaller than 10cm 2 , or Hcurve was too irregular to fit any <strong>of</strong> the templates well,<br />
measurement was either not taken or disregarded by use <strong>of</strong> a sherd size filter during<br />
analysis. Orifice radii can be obtained by subtracting neck thickness from neck Hcurve.<br />
Neck “Vcurve” was estimated using templates for curves <strong>of</strong> radius $47mm, and<br />
by estimation for smaller curves. <strong>The</strong> neck Vcurve was the largest circle that could fit<br />
vertically and snugly against the neck in the middle <strong>of</strong> the sherd (Figure 33). Minor<br />
irregularities might cause light to show in places between the neck and the template, but<br />
usually this measurement was a straightforward and unambiguous procedure.<br />
<strong>The</strong> neck Vcurve measurement forms the basis for the neck angle measurement<br />
Figure 34: Shoulder form measurement.<br />
213
shown in Figure 33. <strong>The</strong> advantage <strong>of</strong> this approach is that the neck angle is measurable<br />
at the neck, rather than requiring the lip to be present, as is more common (e.g. Irwin<br />
1972:60). This means that for a relatively broken assemblage where few lip-rim-neck<br />
sherds are present, neck angle measurements can still be obtained in quantity, yielding a<br />
valuable increase in the sample size for this type <strong>of</strong> information on vessel form for the<br />
assemblage. Also, this method does not treat the neck as though it is an “inflection point<br />
(emphasis added)” (Wickler 2001:93), but rather as a curve, which is a less arbitrary<br />
determination <strong>of</strong> where the measurement should be taken, particularly where the neck<br />
Vcurve is gentle.<br />
Shoulder Form:<br />
Four shoulder types were defined (Figure 34). For s<strong>of</strong>t shoulders (s), the most numerous<br />
category, shoulder, Vcurve and Hcurve radii were defined as for necks, with one important<br />
difference: the templates used to measure curves were too long to be <strong>of</strong> much use<br />
measuring tight convex curves on shoulders, and Vcurve and shoulder angle were thus not<br />
measured as accurately as in the case <strong>of</strong> necks. S<strong>of</strong>t shoulder angle measurements on small<br />
sherds (
characteristics. Shoulder angle for hard shoulders was effectively the angle between the<br />
tangents to the upper and lower halves <strong>of</strong> the shoulder at the apex <strong>of</strong> the halves. Notes<br />
referring to bevel distance on these sherds are the depth measurement, recorded also in the<br />
Rimdepth field.<br />
For slab-built vessels where a lip-rim-neck slab is joined directly to a shallow body<br />
(Figure 32:“k”), there was no convex (in external pr<strong>of</strong>ile) formed shoulder and these were<br />
therefore classified as shoulderless (l) type, where vcurve measures the concavity <strong>of</strong> the<br />
external pr<strong>of</strong>ile from the lower boundary <strong>of</strong> the neck to the carination/modeling junction<br />
edge. Hcurve was either not measured or disregarded in subsequent analysis.<br />
Where the shoulder was straight in external pr<strong>of</strong>ile or gently convex, and typically<br />
extended, the shoulder type was designated conical ( c ), and was expected to have, or did,<br />
join to the maximum width <strong>of</strong> the vessel. In these cases Vcurve measured the degree <strong>of</strong><br />
convexity <strong>of</strong> this flattish shoulder, and the distance from the lower boundary to the neck<br />
to the beginning <strong>of</strong> the carination was recorded in the “Rimdepth” field.<br />
Carination Form:<br />
Carinations could be either hard (h) (small Vcurve) or s<strong>of</strong>t (s) (larger Vcurve) or double<br />
(d) (there were two examples: see Figure 9 for one <strong>of</strong> these). Carination angle was<br />
measured by the same rules as neck angle or shoulder angle. Carination Hcurve was<br />
measured on the exterior plan <strong>of</strong> the vessel (horizon).<br />
Body Form:<br />
Body sherds were all convex in form by definition, and were classified as either <strong>of</strong> conical<br />
215
type (c), where orientation could be established and where the vertical curvature was slight<br />
in comparison to the horizontal curvature, or <strong>of</strong> globular (spheroidal) type (g) where both<br />
dimensions were similar (Figure 35). Latitudinal rows <strong>of</strong> anvil marks, and horizontal<br />
wiping marks were useful in inferring body sherd orientation in many cases. In cases where<br />
the sherd could not be oriented, maximum and minimum curvature were entered in the<br />
notes field <strong>of</strong> the form table, rather than in Hcurve and Vcurve fields, unless maxima and<br />
minima were the same (spherical form). Where body sherds were smaller than 10cm 2 ,<br />
Vcurve and Hcurve data were regarded as deriving from an unknown vessel part,<br />
irrespective <strong>of</strong> coded vessel part, as these could equally be portions <strong>of</strong> rounded shoulders<br />
rather than body sherds.<br />
For those sherds that could not be oriented, but could be characterized as conical<br />
(C) or globular/spheroidal (G), a number <strong>of</strong> body forms are possible, and some confusion<br />
with conical shoulder forms is possible also. Relative abundance <strong>of</strong> various body sherd<br />
Figure 35: Derivation <strong>of</strong> conical/cannister ( C ) and spheroidal (G) sherds from<br />
various body forms.<br />
216
forms in assemblages must be interpreted with these various possibilities in mind, and may<br />
serve to provide a basis for establishing difference between assemblages, but not similarity.<br />
Interior Form:<br />
It was noticed during analysis that some interiors were smooth, with even body wall<br />
thickness, while others had anvil marks arranged <strong>of</strong>ten in latitudinal circles around the pot.<br />
This was thought to relate to whether or not a final or secondary smoothing had been<br />
performed, using a larger anvil than that used to rough out the body shape. It seemed also<br />
that there was some correlation between this secondary smoothing and a more distinct, or<br />
sharper edge to the interior <strong>of</strong> the neck (orifice). To test whether this<br />
correlation was real, the larger sherds were re-analyzed to record whether the body<br />
interior was even (e) (Figure 36) or uneven, the latter involving the following options:<br />
finger impressions (f), anvil impressions (a) or other (o) specified in the notes field. Also<br />
recorded was whether the interior <strong>of</strong> the neck was sharp or rounded. <strong>The</strong>se data were<br />
subsequently filtered to exclude small sherds from analysis, to preclude the possibility <strong>of</strong><br />
Figure 36: Form codes for vessel interiors.<br />
217
misinterpretation based on too small a sample <strong>of</strong> the vessel.<br />
Unknown Part Form:<br />
<strong>The</strong>se were probably mostly s<strong>of</strong>t shoulders or body sherds, most being nondescript<br />
fragments <strong>of</strong> more-or-less globular form. Where there was some basis for orienting the<br />
sherd, Vcurve and Hcurve were recorded. Descriptive statistics <strong>of</strong> these measurements,<br />
filtered to sherds over 10cm 2 in area were subsequently used in analysis. Some sherds<br />
recorded as rim sherds on the basis <strong>of</strong> their hyperboloid form (concave in one surface axis<br />
and convex in the other) could be parts <strong>of</strong> shoulderless vessels below the neck, but this<br />
seems unlikely, as all decorative evidence points to these being rims. Shoulders <strong>of</strong> that<br />
form were all decorated with bounded incision, except for one vessel from the Paniavile<br />
site which had idiosyncratic decoration, and one plain carination.<br />
Base Form:<br />
Very few base sherds were identified, and those tentatively. No evidence was found for flat<br />
bases, and only tenuous suggestion <strong>of</strong> some more conical forms, and this is regarded<br />
Figure 37: Possible conical base sherds from robust vessels.<br />
218
as indirect evidence for predominantly rounded bases in the entire assemblage. Base sherds<br />
were identified through having a parabolic rather than circular curvature in any dimension<br />
around a point (see the base pr<strong>of</strong>ile <strong>of</strong> the bi-conical form in Figure 35). One or two<br />
sherds have slight flattening and extreme thickness also (see Figure 37 for illustrated<br />
examples).<br />
Decoration:<br />
Objectives <strong>of</strong> the decorative study shifted significantly through the course <strong>of</strong> data recording<br />
and analysis, from an initial poorly-defined aim <strong>of</strong> a detailed social/behavioural analysis,<br />
to a position where understanding production variability was seen as the key to sample<br />
evaluation (allowing the construction <strong>of</strong> vessel families) and to constructing a seriation<br />
chronology. This shift was a consequence <strong>of</strong> the realization that the sherd sample displayed<br />
a low degree <strong>of</strong> vessel completeness for all collection sites, and must have been heavily<br />
sampled from a breakage population by post-discard processes (Felgate 2002). Nowhere<br />
in the analysis is this shift more apparent than in the analytical scheme used to describe<br />
surface decoration <strong>of</strong> the pottery. Initially, I was interested in identifying the work <strong>of</strong><br />
individual potters through examination <strong>of</strong> variation in technical execution, but it became<br />
apparent, as will be discussed, that there was no particular reason to expect a good sample<br />
<strong>of</strong> the work <strong>of</strong> a single potter to have been preserved in the assemblages, and the primary<br />
focus <strong>of</strong> research shifted from the fine details <strong>of</strong> decoration to a slightly broader analysis<br />
<strong>of</strong> production variability. <strong>The</strong>re is substantial data-redundancy in the decorative<br />
information recorded as a result <strong>of</strong> this shift, although some use was made <strong>of</strong> the mark<br />
variability data in testing units <strong>of</strong> decorative classification (particularly for<br />
219
impressed or notched lips).<br />
Data structure for the decoration table are shown in Table 13. Use <strong>of</strong> the fields<br />
“Id”, “Part” and “Type” were as for the form table <strong>of</strong> description, with multiple records<br />
for sherds linked to the master table <strong>of</strong> sherd properties via the “Id” field. <strong>The</strong>se fields, and<br />
Table 13: Data structure for decoration records.<br />
Field Name Type Size Key Required<br />
Value<br />
Id A 7 * *<br />
Part A 1 * *<br />
Type A 1 *<br />
Dec_tech A 3 * *<br />
Pattern A 3<br />
Pattern_rep_dist N<br />
Notes_1 A 240<br />
Element A 3<br />
ERD N<br />
Ms_MarkSection A 1<br />
Ml_MarkLength N<br />
Mw_MarkWidth N<br />
Md_Markdepth N<br />
Hand A 5<br />
Notes A 240<br />
LinDecClass S<br />
LipDecCode S<br />
NeckDecCode S<br />
a fourth, decorative technique (“Dec_tech”) were hierarchically indexed in the database<br />
to allow each vessel part to have more than one decorative technique recorded. In practice<br />
it was found more convenient to have multiple decorative techniques on the same part<br />
recorded as coded combinations rather as separate lines in the table, in which<br />
220
Table 14: Decorative techniques.<br />
Code Description<br />
? Decorative technique uncertain<br />
a Applique (Figure 38, Figure 39, Figure 40)<br />
b Burnishing or polishing (not illustrated: not used in analysis)<br />
d Lip deformation (Figure 41, Figure 42, Figure 43)<br />
e Excision (Figure 44, Figure 45)<br />
f Fingernail impression (subdivided by the “element” field as<br />
detailed below)<br />
g Incision<br />
h Brushing (rare, identified on the presence <strong>of</strong> multiple parallel<br />
bristle marks sharing a common termination)<br />
i Impression not identified as fingernail impression, usually on<br />
lip or carination.<br />
l Linear stamping (rare, but see Figure 12: Z.10.573, which<br />
has linear stamping on the shoulder, in a pattern <strong>of</strong> opposed<br />
lines forming triangles)<br />
m Spatulate tool impression (during vessel forming)?(see Figure<br />
46)<br />
o Other unclassified techniques (see “Notes_1")<br />
p Punctation (Figure 15 )<br />
r Excision by punctation/rotation (see Figure 47)<br />
s Surface smoothing without striation (self slip?)<br />
t Perforation (see Figure 48)<br />
v Anvil impression, exterior (unimportant- one unconfirmed<br />
example)<br />
w Wiping - identified by presence <strong>of</strong> multiple shallow rounded<br />
parallel grooves.<br />
y Wavy stamping or dentate stamping as specified in element<br />
field (Figure 45 and Figure 49)<br />
cases the predominant technique was described in detail in subsequent fields, and sub-<br />
techniques for that vessel part were described in the notes fields <strong>of</strong> the decoration table.<br />
<strong>The</strong> fields <strong>of</strong> decorative information were hierarchically structured from the most general<br />
to the most detailed. Decorative techniques were coded as in Table 14.<br />
221
“Pattern” (Table 13) is roughly equivalent to Anson’s concept <strong>of</strong> motif for more<br />
complex patterns, (Anson 1983:57) or Mead’s concept <strong>of</strong> “all<strong>of</strong>orm” (Mead 1975:23),<br />
being defined as the structure or arrangement <strong>of</strong> decorative elements, and is defined<br />
independent <strong>of</strong> decorative technique. Mead did not separate technique <strong>of</strong> execution and<br />
the pattern in which any technique was applied, thus generating unnecessary proliferation<br />
<strong>of</strong> some units <strong>of</strong> decoration, such as the various “Zone Markers” (see for example Mead<br />
1975:Figure 2.11 page 26, GZ2-GZ4). In this study such a distinction has been maintained<br />
as far as possible, with the objective <strong>of</strong> reducing the number <strong>of</strong> decorative classes.<br />
Pattern entries in the decoration table fell into two classes, formulaic patterns, from<br />
which the pattern could be generated by a formulaic repetition <strong>of</strong> elements, and more<br />
complex patterns described using sherd illustrations. <strong>The</strong> formulaic patterns are mostly<br />
simple latitudinal bands <strong>of</strong> decoration (pattern prefix “b” in the field “patt” in the<br />
decoration data table) running around the vessel parallel to the vessel horizon (with the<br />
central axis <strong>of</strong> symmetry oriented vertically). Pattern definitions are given in Table 15. <strong>The</strong><br />
complex patterns were later classified into two classes based primarily on zone markers,<br />
with a third dustbin class, mostly for pattern fragments.<br />
“Patt_rep_dist” (Pattern repeat distance - Table 13) was a measure (in mm) for simple<br />
formulaic patterns <strong>of</strong> the vertical distance between element repeats (from vertical centre<br />
to vertical centre). Thus in a pattern comprising multiple latitudinal bands <strong>of</strong> opposed pinch<br />
fingernail impression, the pattern repeat distance was a component in specifying the overall<br />
expression <strong>of</strong> the pattern on the sherd.<br />
222
Figure 38: Examples <strong>of</strong> applied decoration.<br />
Figure 39: Examples <strong>of</strong> applied decoration.<br />
223
Figure 40: Applied decoration on compound rims.<br />
Figure 41: Examples <strong>of</strong> deformation <strong>of</strong> the lip into a discontinuous band.<br />
224
Figure 42: Horizontal deformation <strong>of</strong> the lip into a continuous band in a wave pattern.<br />
Figure 43: Examples <strong>of</strong> discontinuous deformation.<br />
225
Figure 44: Excision <strong>of</strong> outer lip to form a band <strong>of</strong> notches.<br />
Figure 45: Compound rims from Nusa Roviana. NR.34 has excised lines forming the<br />
triangle pattern on the upper rim.<br />
226
Figure 46: Impressions on the top <strong>of</strong> the lip; and spatula impressions on the neck,<br />
thought to indicate forming <strong>of</strong> the neck using a tool.<br />
Figure 47: Excision by rotation: one end <strong>of</strong> a small twig or rod has been poked into the<br />
clay and the other end moved in a circle, leaving a conical hole.<br />
Figure 48: Perforation (upper hole).<br />
227
Figure 49: Examples <strong>of</strong> wavy stamping and an example <strong>of</strong> dentate-stamping.<br />
Figure 50: Applied decoration (in combination with incised decoration) with<br />
detachment scars indicating a v-pattern, and also possible attached disc.<br />
228
Table 15: Decoration pattern definitions<br />
“Patt” Description Anson Motif # Count<br />
? incomplete, insufficient to characterize 284<br />
?pi (continuous?) band <strong>of</strong> parallel elements on inner edge <strong>of</strong> part 1<br />
?po (continuous?) band <strong>of</strong> parallel elements on outer edge <strong>of</strong> part 1<br />
?pt (continuous?) band <strong>of</strong> parallel elements on top face <strong>of</strong> part 1<br />
acc accidental marks? 1<br />
am1 (complex applied pattern) see R37.11A.1 in Figure 40 2<br />
as1 (complex applied fillets) see R376a14 in Figure 38 1<br />
as3 see HV.1.3 in Figure 49 3<br />
as? (applied fillet ?) Not illustrated 1<br />
avc (applied v-shaped fillet) see Figure 50 1<br />
b continuous latitudinal band <strong>of</strong> an element (used for circular elements such as punctations or nubbins) 117<br />
bar incomplete, twinned vertical bars made up <strong>of</strong> fingernail impressions 1<br />
bd continuous latitudinal band <strong>of</strong> elements oriented diagonally to horizon (see Figure 51) 16<br />
bdi band, elements diagonally oriented, inner edge <strong>of</strong> lip (not illustrated) 4<br />
bdo band, elements diagonally oriented, outer edge <strong>of</strong> lip (equivalent to “bpi” below) 8<br />
bdt band, elements diagonally oriented, top face <strong>of</strong> lip (equivalent to “bpi” below, see Figure 57) 3<br />
bnd fragment <strong>of</strong> complex pattern? Repeats <strong>of</strong> bent lines, see MH266 (Figure 52) 272?, 273?, 498 sans zone markers? 3<br />
bo band, outer edge <strong>of</strong> lip (as for “bpi” below, but for circular elements) 6<br />
bot band <strong>of</strong> parallel elements in diagonally, symmetrically opposed sets, top face <strong>of</strong> lip (Figure 57) 1<br />
bp band <strong>of</strong> parallel elements, most commonly fingernail pinching but sometimes applied fillets (Figure 7) 124<br />
bpb bands, parallel elements, outer and inner edges <strong>of</strong> lip (Figure 19: Z.89.549, Figure 22, Figure 26: GE266) 16<br />
bpi band <strong>of</strong> parallel elements, inner edge <strong>of</strong> lip (Figure 54) 39<br />
bpo band <strong>of</strong> parallel elements, outer edge <strong>of</strong> lip (Figure 55, Figure 56) 69<br />
bpt band, parallel elements, top face <strong>of</strong> lip (Figure 57) 38<br />
bt band <strong>of</strong> elements on the top face <strong>of</strong> the part (applies to circular elements such as punctation) 4<br />
c10 complex linear pattern: see HV.4.164 (Figure 8) similar 154, 155 4<br />
c11 complex linear pattern: see HV.4.202, HV.2.302, HV.2.297 (Figure 8, Figure 9) 155 3<br />
c12 complex linear pattern: see HV.2.227 (Figure 49) 1<br />
c13 complex linear pattern: see HV.1.314 (Figure 9) 2<br />
c14 complex linear pattern: see HV.02.419 (Figure 58) 155 in fingernail impression? 4<br />
c17 complex linear pattern: see MH 169 (Figure 13: MH169) 4<br />
c18 basic linear pattern: see Z.61.654 (Figure 12) 162 sans zone marker 16<br />
c20 complex linear pattern: see sherd MH360 (Figure 59) similar 163 1<br />
c21 complex linear pattern: see GE 155 (Figure 60) 1<br />
c23 complex linear pattern: see Z.49.751 (Figure 61) 1<br />
c24 complex linear pattern: A11 (Figure 59) 1<br />
c25 complex linear pattern: similar to ch5, see A.2 (Figure 62) 1<br />
cf (crow’s foot) see P205 (Figure 53) compare 197-223 1<br />
ch3 complex linear pattern: see R37.11A.16 (Figure 62) 3<br />
ch5 complex linear pattern: see R37.2A.14 (Figure 64) 157-159 sans zone marker, 307 4<br />
ch6 complex linear pattern: GW.258 (Figure 59) 2<br />
ch7 complex linear pattern: incomplete, see HG.17.8 (Figure 63) 2<br />
ch9 complex linear pattern: see HV.2.124 (?), HV.2.464 (Figure 7), HV.4.175 (Figure 49) 169161 3<br />
ch? complex linear pattern: in triangular pattern category but too small to determine 32<br />
chd complex linear pattern: see MH.185 (Figure 61) 1<br />
chv complex linear pattern: see Z.60.326 (Figure 65) and MH.14 (Figure 66) 187, 188, 190 sans zone markers 45<br />
chz complex linear pattern: see HG.99.061 (Figure 59) 1<br />
cl1 complex curvilinear pattern: R37.7a.21 (Figure 67) 1<br />
cl3 complex curvilinear pattern: see 3711a22 (Figure 67) compare 259 1<br />
cl4 complex curvilinear pattern: see HV.5.214 (Figure 7) 259 with zone marker added? 1<br />
cl5 complex curvilinear pattern: see HV.04.379 (Figure 9) compare 133, 141, 494, 496 1<br />
cl6 complex curvilinear pattern: see GW.159 (Figure 67) 1<br />
cl7 complex curvilinear pattern: see sherd C.14 (Figure 67) 1<br />
cl? curvilinear pattern fragments, not illustrated 3<br />
crw complex linear pattern: see C.101 (Figure 68), post-depositional damage? 1<br />
ct3 (cross-hatch 3) see MH.290 (?), R37.9A.14 (Figure 62), MH.69 (Figure 69) see 230, 237, 7<br />
ctw (cross-hatch/wavy) see MH.290 (Figure 13) 1<br />
cvh (chevron/hatch) see MH.273 (Figure 70) 2<br />
d pattern unknown, element oriented with long axis diagonally oriented to CVA 4<br />
gm3 complex linear geometric pattern: see HV.2.313 (Figure 8) compare 278, 279, 280, 283, 325, 441-448 1<br />
gm4 complex linear geometric pattern: see HV.1.369 (Figure 7) 1<br />
gm5 complex linear geometric pattern: see HG.11.72 (Figure 71) 4<br />
gm6 complex linear pattern: see A.23 (Figure 60) 1<br />
gm7 complex linear pattern: ref R37.7a..13 in data files, faint thin lines, not illustrated 2<br />
gm8 complex linear pattern: See NR.34 (Figure 45) 1<br />
l lateral element, discontinuous, e.g. striation or brush mark 118<br />
lbi transitional between bpt and bpi (one record only) 1<br />
mb multiple stacked continuous latitudinal bands (e.g. Figure 26: HG.21.2) 29<br />
oci discontinuous (widely spaced?) element placed inner edge <strong>of</strong> part 41<br />
oco discontinuous (widely spaced?) element placed outer edge <strong>of</strong> part 2<br />
olb vertical trails <strong>of</strong> paired single impressions (1 example) 1<br />
pc1 complex linear pattern:(?pictorial incised ) see MH44 (Figure 61) 2<br />
pc2 complex linear pattern: see A.24, (Figure 61)<br />
pwc complex linear pattern: see GW055 (Figure 39) 2<br />
rbr complex linear pattern: see R37.11A.2 (Figure 62) 2<br />
rl1 see HV.5.210 (Figure 7) 1<br />
sfl complex linear pattern: see HV.04.185 (Figure 72) 1<br />
wav horizontal wave (created by deformation or impression) (see Figure 42) 96
Figure 51: A band <strong>of</strong> fingernail impressions (opposed pinch), each <strong>of</strong> which is oriented<br />
diagonal to the CVA rather than vertically, or parallel to the CVA.<br />
Figure 52: Deformation, perforation, and the “bnd” incomplete pattern example.<br />
Figure 53: “cf” (Crow’s foot) pattern on the vessel rim above a band <strong>of</strong> pinching.<br />
230
Figure 54: Examples <strong>of</strong> lip impression in pattern “bpi” (band parallel inner-edge <strong>of</strong> lip).<br />
Figure 55: Examples <strong>of</strong> lip impressions laid out in pattern “bpo” (band parallel outeredge).<br />
231
Figure 56: Fragmented examples with lip impressions assigned to “band parallel outer”<br />
pattern.<br />
Figure 57: Examples <strong>of</strong> lip impressions/incision laid out in patterns “bot” (band<br />
opposing top), “bdt” (band diagonal top) and “bpt” (band parallel top), which were<br />
regarded as equivalent in analysis due to non-exclusive nature <strong>of</strong> these descriptions.<br />
232
Figure 58: Pattern expressed using linear arrangements <strong>of</strong> fingernail pinching.<br />
Figure 59: Miscellaneous incised patterns: MH360 is middle row, left-hand column.<br />
GW258 is bottom-right.<br />
233
Figure 60: Examples <strong>of</strong> applied nubbins and some bounded incised patterns (MH259 is<br />
an unbounded pattern).<br />
Figure 61: Decorated sherds with quartz-calcite hybrid granitic temper.<br />
234
Figure 62: Unbounded incised patterns.<br />
Figure 63: Thin incised rims from Hoghoi.<br />
Figure 64: Unbounded incised pattern on the shoulder from Paniavile.<br />
235
Figure 65: An example <strong>of</strong> “chv” pattern, a band <strong>of</strong> unbounded linear triangles filled by<br />
alternating fields <strong>of</strong> parallel lines.<br />
Figure 66: Example <strong>of</strong> “chv” pattern on tall rim (pinching at neck).<br />
236
Figure 67 Curvilinear incised patterns.<br />
Figure 68: Crow’s foot mark, probably post-deposition scratching.<br />
Figure 69: Example <strong>of</strong> cross-hatch pattern “ct3”.<br />
237
Figure 70: Cross hatch pattern “cvh”.<br />
Figure 71: Example <strong>of</strong> pattern “gm5".<br />
Figure 72: Sole example <strong>of</strong> pattern “rl1", a double-line arrangement <strong>of</strong> single repeated<br />
fingernail impressions.<br />
238
For complex patterns expressed graphically, pattern repeat distance was the distance<br />
latitudinally between adjacent homologous points, where the sherd was sufficiently large<br />
to measure this. Pattern repeat distance was not used analytically and is included in case<br />
somebody is interested in this property <strong>of</strong> the assemblages in the future. <strong>The</strong> number <strong>of</strong><br />
vertical repeats <strong>of</strong> patterns comprising stacked multiple bands was recorded along with<br />
miscellaneous other information in the “Notes_1" field <strong>of</strong> “Decoration.db”, and could<br />
easily be systematized into a separate field should anyone so wish in future.<br />
Decorative Elements:<br />
<strong>The</strong> “Element” field specifies that which is repeated in the formulaic patterns: for example,<br />
in the pattern “bp” (lateral band parallel) the pattern consists <strong>of</strong> repeated elements oriented<br />
parallel to each other and perpendicular to the horizon <strong>of</strong> the band, and might be an<br />
opposed-pinch fingernail impression (“opc”), or a deformation <strong>of</strong> the lip into a horizontal<br />
wave (“wav”), or a single impression using a tool or fingernail (“i”). <strong>The</strong> element<br />
represents a single manual decorative operation on the part <strong>of</strong> the potter, each pinch <strong>of</strong> the<br />
fingernails, each impression <strong>of</strong> the tool, each deformation <strong>of</strong> the lip, repeated according to<br />
the pattern. <strong>The</strong> element repeat distance (ERD) specifies the average horizontal spacing,<br />
in mm, between element centres as calculated from two or three measurements, using<br />
calipers, on the sherd. <strong>The</strong>se data were not used in the current analysis, other than for lip<br />
impressions but provide a level <strong>of</strong> descriptive detail that might be <strong>of</strong> use to future analyses,<br />
so were included. Element codes and their meanings are given in Figure 72.<br />
239
Table 16: Decorative elements.<br />
Eleme<br />
nt code<br />
Description count in table<br />
“Decoration”<br />
No element identified (eg surface treatment records) 218<br />
as applied "sausages" or elongate fillets <strong>of</strong> clay 26<br />
ds dentate stamp 7<br />
fi finger/fingertip impression 44<br />
hol hole 82<br />
i linear edge impression not obviously fingernail 141<br />
l linear element (e.g. incision) 239<br />
nub circular nubbin 19<br />
opc opposed pinch fingernail impression 179<br />
pi paired linear impressions 1<br />
psm prismatic or triangular section applied fillet 3<br />
sin single crescentic fingernail impression (not pinched) 27<br />
t applied triangle 1<br />
tfm latitudinal tool forming mark (spatula?) 94<br />
tw? Tool forming marks or striations from wiping? 13<br />
wav wave form, repeated shear-deformation <strong>of</strong> the lip 84<br />
wip wiping or brushing marks 11<br />
ws wavy stamp, probably a shell-edge impression 3<br />
wv3 a closely spaced paired repeat <strong>of</strong> “fi” deformation <strong>of</strong> the lip 8<br />
<strong>The</strong> fields “Mark Length”, “Mark Width” and “Mark Depth” were used to record<br />
dimensions <strong>of</strong> the decorative elements, with maximum depth recorded perpendicular to the<br />
surface <strong>of</strong> the sherd, using a sharpened depth-gauge, in mm, as a positive value regardless<br />
<strong>of</strong> whether decorative element was raised relief, as in the case <strong>of</strong> applied decoration, or<br />
sunken, as in the case <strong>of</strong> incising or punctation.<br />
Measurement conventions for these variables for each element are as given below:<br />
240
Applied fillets (when complete) were measured across their longest and shortest<br />
dimensions for length and width, with depth measured perpendicular to the sherd exterior<br />
surface at a point <strong>of</strong> average depth. dentate-stamp overall length, average width and<br />
average depth were recorded where measurable. Tooth sizes were not recorded but scale<br />
drawings <strong>of</strong> all dentate-stamping are provided (the seven records for dentate-stamping in<br />
the decoration table arise from four sherds with confirmed dentate-stamping and two<br />
sherds with possible dentate-stamping: one sherd yielded two zones <strong>of</strong> dentate-making a<br />
total <strong>of</strong> seven). Fingertip/finger impressions were measured when present as lip<br />
deformations, with length measured around the lip, width measured from the top <strong>of</strong> the lip<br />
to the lowest point <strong>of</strong> the impression. Depth was measured parallel to the top face <strong>of</strong> the<br />
lip, with the base <strong>of</strong> the depth-gauge resting across the length <strong>of</strong> the impression, and the<br />
gauge located at the deepest point. Length <strong>of</strong> holes was measured in their longest<br />
dimension, and width was measured at right angles to the length measurement; depth was<br />
gauged at the deepest point; several readings were averaged where there was variation.<br />
Edge impressions (proportionately deep v-shaped impressions) were measured as<br />
for applied fillets, except for depth, which was measured as for holes. Linear elements were<br />
measured where length was standard for the pattern (e.g. for pattern c18). Nubbins were<br />
measured as for holes, except that depth was measured as for applied fillets, but at<br />
maximum depth. Opposed pinch fingernail impression was measured for length across from<br />
nail-top to nail-top impression at their widest separation, for width at right angles to the<br />
length measurement, and for depth at the deepest point recorded by the depth gauge,<br />
averaging the readings <strong>of</strong> multiple elements where possible. Paired linear impressions<br />
were measured as for linear impressions. Prism applied fillets were measured as for<br />
applied fillets. Rotated tool excision was measured as for finger impression. Single<br />
fingernail impressions were measured as for linear impressions. <strong>The</strong> single example <strong>of</strong> a<br />
pattern made up <strong>of</strong> repeated applied triangles was measured for length across the widest<br />
241
horizontal dimension, and for width in the vertical dimension, and for depth as for other<br />
applique. Wave elements were measured as for finger impressions, and wavelength was<br />
entered as ERD.<br />
Handedness was recorded in many instances but data will not be presented, other<br />
than in the appended database files on CDROM. <strong>The</strong>se data were interesting though as<br />
these provided many clues to vessel orientation during the application <strong>of</strong> decoration<br />
(assuming the potter to be upright) during the process <strong>of</strong> decoration, and also revealed<br />
some variety within the element category <strong>of</strong> opposed pinch fingernail impression.<br />
Additionally, many <strong>of</strong> the deeper fingernail marks were <strong>of</strong> such small size (width) as to<br />
suggest the decorators were women. <strong>The</strong>re is potential for use <strong>of</strong> variation in the shape <strong>of</strong><br />
fingernail impressions as a proxy for genetic variability <strong>of</strong> the potters.<br />
Mark cross-sections across the width <strong>of</strong> the mark (length on the case <strong>of</strong> OPC<br />
fingernail impression) through the depth measurement point were recorded in the “Mark<br />
Section” field as follows: v-section (v), square-section (s), oblique v-section (o), u-shaped<br />
(u), parallel-sided pierced (p), pierced using a conical tool (x), double-v section(w) (as<br />
though mark was incised with a fingernail and fingertip, or with a frayed tool), with “t”<br />
denoting other unclassified cross-sections (specified in “Notes_2").<br />
Transforming Relational Data into a Flat Table:<br />
A flat table with one record (row) per sherd was created from the relational database using<br />
a subset <strong>of</strong> variables pertaining to sherd size, fabric, decoration and form, to provide<br />
summary information on the structure <strong>of</strong> vessel form and decoration as represented by the<br />
sample <strong>of</strong> sherds. (Table 17 shows the data structure and raw data is given in table Flat.db<br />
appended on CD).<br />
Descriptive data for decoration was summarized as a short list <strong>of</strong> decorative<br />
242
attributes, entered into four location fields (decoration-lip = “Part 1", decoration-rim =<br />
“Part 2", decoration-neck = “Part3", decoration-shoulder = “Part 4"). Carination and body<br />
decoration were so highly correlated with upper vessel decoration that these parts were<br />
omitted from the table to prevent multiplicity <strong>of</strong> fields. Attribute value codes used in all<br />
four location-specific decoration fields were as follows :<br />
1. Complex linear motif Class 1 with linear design zone horizontal markers, typically<br />
double-line zone markers<br />
2. Complex linear motif Class2 without linear design zone horizontal markers.<br />
3. Multiple bands <strong>of</strong> opposed pinch fingernail impression (fni)<br />
4. Lateral band <strong>of</strong> punctation<br />
5. Other unclassified decoration not including wiping, brushing or tool-forming marks<br />
6. Plain<br />
(this is a filter attribute)<br />
7. Horizontal finger deformation/impression, discontinuous<br />
8. Wiping, brushing or tool-forming marks (this is a filter attribute as these records<br />
do not constitute decoration in the sense <strong>of</strong> deliberate patterned execution)<br />
9. Staggered opposed inner and outer bands <strong>of</strong> lip impressions forming a wave<br />
pattern without deformation<br />
10. Horizontal finger deformation <strong>of</strong> the lip into a continuous wave<br />
11. Fingernail impression, single band <strong>of</strong> opposed pinching<br />
12. Band <strong>of</strong> impressions along the top <strong>of</strong> the lip<br />
13. Band <strong>of</strong> impressions along the inner edge <strong>of</strong> the lip<br />
14. Opposing bands <strong>of</strong> impressions along both edges <strong>of</strong> the lip, not staggered to form<br />
a wave pattern<br />
15. Band <strong>of</strong> impressions along the outer edge <strong>of</strong> the lip<br />
16. Lateral band <strong>of</strong> applied nubbins<br />
243
Multiple decorative attributes on single vessel part were discarded from the attribute data<br />
for the flat table, but this only affected a small number <strong>of</strong> unusual cases, and does not affect<br />
overall results. Subsequent analyses made use <strong>of</strong> both this flat data structure and the more<br />
comprehensive attributes/variables in the decoration and form tables, through the use <strong>of</strong><br />
multi-table one-to-one and one-to-many database queries.<br />
Parts <strong>of</strong> the vessel not represented by the sherd were identified in “Flat.db” by a<br />
decorative value <strong>of</strong> 99, thus a lip sherd would be coded with a value <strong>of</strong> 99 for each <strong>of</strong> the<br />
missing rim, neck, and shoulder parts. <strong>The</strong> remaining data fields are a transposition <strong>of</strong> basic<br />
variables from the form and thickness tables, separated by vessel part. This draws<br />
information about co-variation by part together into a single table <strong>of</strong> sherd properties for<br />
analysis <strong>of</strong> the structure <strong>of</strong> decoration and form across the vessel.<br />
Chapter Summary and Conclusions:<br />
<strong>The</strong> preceding sections detail the overall data model <strong>of</strong> the ceramic database, and the<br />
contents <strong>of</strong> each <strong>of</strong> the four tables that make up the primary database. A table <strong>of</strong> the vessel<br />
family membership <strong>of</strong> lip and carination sherds, used to estimate a parent breakage<br />
population, is detailed in the next <strong>chapter</strong>. Attributes values were listed and explained in<br />
a series <strong>of</strong> tables and paragraphs pertaining to each attribute, and methods for obtaining<br />
values for variables requiring measured dimensions or counts <strong>of</strong> repeats were explained.<br />
<strong>The</strong> overall objective in designing this database was to try to capture sufficient descriptive<br />
information to be able to draw a picture <strong>of</strong> the vessel as represented by the sherd from the<br />
information recorded, using these rules. Fine structured decorative detail was recorded<br />
formulaically alongside larger, more complex or incomplete patterns that required<br />
illustration.<br />
244
Table 17: Data structure for flat table <strong>of</strong> summary data <strong>of</strong> sherd properties.<br />
Field Name Type Size Key<br />
Id A (alphanumeric) 7 *<br />
Area N (number)<br />
Weight N<br />
Site S (short integer)<br />
Unit (spatial collection S<br />
Code (fabric A 6<br />
Fabric (temper class) S<br />
Form A 2<br />
Sector (EVE N<br />
TYPE S<br />
Decoration Part1 (Lip) S<br />
Decoration Part2 S<br />
Decoration Part3 S<br />
Decoration Part4 S<br />
Lip Type A 1<br />
Lip Angle N<br />
Rim Type A 1<br />
Rim Angle N<br />
Rim Vertical Curve N<br />
Rim depth N<br />
Neck Type A 1<br />
Neck Angle N<br />
NeckVcurve N<br />
Neck Hcurve N<br />
Shoulder Type A 1<br />
Shoulder Angle N<br />
Shoulder Vcurve N<br />
Shoulder Depth N<br />
Carination Type A 1<br />
Carination Angle N<br />
Carination Vcurve N<br />
CariniationHcurve N<br />
Carination notes A 240<br />
Body Type A 1<br />
Body Vcurve N<br />
Body Hcurve N<br />
Lip Thickness N<br />
Rim thickness N<br />
Neck thickness N<br />
Shoulder thickness N<br />
Body thickness N<br />
245
I would like to think that this approach addressed the problem discussed in previous<br />
<strong>chapter</strong>s <strong>of</strong> separate analytical/descriptive schemes for Lapita and for post-Lapita<br />
decorative design, which create an artificial disjuncture in many regional series. This<br />
scheme has been an attempt to bridge that disjuncture, by having a hierarchy <strong>of</strong> descriptive<br />
detail that can capture the broader complexity <strong>of</strong> larger, more complex (and usually<br />
fragmented) designs as well as simpler formulaic repetition <strong>of</strong> basic elements. <strong>The</strong> finer<br />
details <strong>of</strong> technique can be addressed also, down to the physical characteristics <strong>of</strong> potter’s<br />
fingernails, and microscopic differences in technique, measurable by gauge rather than by<br />
eye. This holistic approach to ceramic variation allows the analysis <strong>of</strong> Oceanic pottery to<br />
a level <strong>of</strong> fine detail where variability in potting behaviour from the past can be read<br />
accurately. <strong>The</strong> key, or a major aid at least, to recording at this level <strong>of</strong> detail is a relational<br />
database, where many data themes can be constructed about each part <strong>of</strong> the sherd, while<br />
maintaining the overall relationship <strong>of</strong> these parts to each other and to the whole, without<br />
requiring enormous forms, spreadsheets or tables with hundreds <strong>of</strong> fields <strong>of</strong> information,<br />
most <strong>of</strong> them with missing values due to vessel fragmentation into sherds.<br />
246
CHAPTER 5:<br />
SAMPLE EVALUATION AND QUANTIFYING<br />
Introduction:<br />
SAMPLE SIZE.<br />
As outlined in Chapter 1 the development <strong>of</strong> sherds-as-vessels approaches to ceramic<br />
analysis is recognition <strong>of</strong> the difficulty <strong>of</strong> quantifying sherd samples for comparison unless<br />
they have the same levels <strong>of</strong> brokenness and completeness. Estimation <strong>of</strong> a breakage<br />
population provides a solution to this problem (Felgate & Bickler n.d). Evaluation <strong>of</strong><br />
ceramic sample size and sample representativeness in general continues to be problematic<br />
in Pacific archaeology, where the measured or quantified properties <strong>of</strong> recovered samples<br />
are <strong>of</strong>ten regarded as representative in an unspecified way <strong>of</strong> the properties <strong>of</strong> people’s<br />
lives in the past. In this <strong>chapter</strong> and in subsequent taphonomic and wave-exposure <strong>chapter</strong>s<br />
I try to develop ways to be more specific about the nature <strong>of</strong> that representation, allowing<br />
a more robust behavioral inference from the recovered sherd samples. Objectives in<br />
estimating a breakage population for the Roviana pottery sample are as follows:<br />
• evaluate sample size and significance<br />
• accumulations research: investigate implications for the intensity/duration <strong>of</strong><br />
occupation using the size <strong>of</strong> the estimated breakage population rather than the size<br />
<strong>of</strong> the recovered sample<br />
• estimate the extent <strong>of</strong> post-depositional sherd/pottery attrition between the<br />
breakage population and the extant deposit, as an indication <strong>of</strong> the type/extent <strong>of</strong><br />
sherd-taphonomic process affecting the sites from which the samples were<br />
recovered<br />
247
• characterize the state <strong>of</strong> preservation <strong>of</strong> the deposits (the conditions <strong>of</strong> wave<br />
exposure which are thought to be a key taphonomic factor are explored in Chapter<br />
6). If most <strong>of</strong> the breakage population has vanished even in the sheltered waters<br />
and relative stability <strong>of</strong> sea levels <strong>of</strong> Roviana Lagoon, what are the implications for<br />
the archaeology <strong>of</strong> Lapita in Near Oceania?<br />
• choose a unit <strong>of</strong> quantification: what does the state <strong>of</strong> vessel<br />
brokenness/completeness tell us about how ceramic attributes <strong>of</strong> these sherd<br />
samples should be quantified? If completeness can be shown to be extremely low,<br />
then sherds are independent observations, and sherd counts are likely to be biased<br />
only by differential survival <strong>of</strong> vessel types/parts <strong>of</strong> types, rather than by differences<br />
in brokenness or bias arising from MNI-based counts.<br />
Methods for Establishing Vessel Brokenness and Completeness:<br />
Two characteristics <strong>of</strong> the archaeological sample form the basis <strong>of</strong> an estimate <strong>of</strong> the<br />
breakage population: mean sherd size (measured in EVEs as explained in Chapter 4) as a<br />
measure <strong>of</strong> brokenness (fragmentation); and vessel completeness (Orton et al. 1993:167-<br />
168 and 178). <strong>The</strong> key methodological problem for archaeologists, regardless <strong>of</strong> which<br />
estimation statistic is adopted, is that <strong>of</strong> identifying vessel classes/species/sherd families<br />
in the potsherd sample. Measurement <strong>of</strong> these properties <strong>of</strong> sherd samples has generated<br />
a substantial methodological literature (see for examples Orton 1993, Orton et al.<br />
1993:172, Schiffer 1995c:183-186, Smith 1983:77-79). <strong>The</strong> task <strong>of</strong> identifying sherd<br />
families can be complex, and the prospects <strong>of</strong> a reliable solution being obtained to this<br />
problem vary from sample to sample. This is because pottery is found in an infinite variety<br />
<strong>of</strong> states <strong>of</strong> brokenness and is manufactured with varying degree <strong>of</strong> standardization.<br />
<strong>The</strong> following assumptions were made about the sample:<br />
• <strong>The</strong> sample assemblage is a random sample <strong>of</strong> vessels drawn from a breakage<br />
248
population (this is unlikely to be true, but a good idea <strong>of</strong> the departure from<br />
randomness is obtained in subsequent <strong>chapter</strong>s, and the consequences for this<br />
analysis are discussed)<br />
• Vessel brokenness and vessel completeness can be accurately measured for the<br />
sample.<br />
<strong>The</strong> vessel-randomness <strong>of</strong> a sample and the validity <strong>of</strong> ones measures <strong>of</strong> vessel brokenness<br />
and completeness are central to the level <strong>of</strong> confidence with which the breakage population<br />
estimate can be accepted. It is not a random sample <strong>of</strong> sherd size that is needed.<br />
Taphonomic processes and collection intensity are likely to have structured sherd size, and<br />
as previously discussed, we do not need to concern ourselves with what size that sherds<br />
we do not have were or are. Rather, we must either assume or have reason to believe that<br />
the sherds in the sample assemblage are a random sample <strong>of</strong> vessel membership <strong>of</strong> the<br />
breakage population we wish to estimate.<br />
Collection <strong>of</strong> sherds from broad areas rather than from testpits suggests that the<br />
measured completeness <strong>of</strong> sherd families is unlikely to have been biased by sampling a non-<br />
representative locality in the sites. While test excavation indicated that some pottery<br />
remained buried in some areas <strong>of</strong> some sites, there was no suggestion <strong>of</strong> a buried well-<br />
preserved deposit in any <strong>of</strong> these (Reeve 1989 reports the Paniavile site as such a deposit<br />
but this was not the case by 1996). Turbation by waves and bio turbation are thought to<br />
have created a complex history <strong>of</strong> vertical migration <strong>of</strong> sherds within sediments and some<br />
horizontal movement across the sites, although the lack <strong>of</strong> rolling damage on larger<br />
recovered sherds from the deeper margins <strong>of</strong> sites suggests such horizontal movement was<br />
infrequent.<br />
Lip/rim-based sherd families mostly comprised singleton lip/rim sherds, so spatial<br />
dispersion data could not be obtained from the distribution <strong>of</strong> sherd families.<br />
249
No significant departures from latitudinal circularity were observed among vessel<br />
rims, beyond those attributed to inaccuracy <strong>of</strong> manufacture. Brokenness was measured by<br />
estimating the percentage <strong>of</strong> the circular vessel lip/rim represented by sherds. Lip/rim EVE<br />
was estimated for sherds by curve fitting, using a rim diameter percentage chart marked<br />
in 5% sector intervals, with concentric circles ranging from 20mm radius to 300mm radius,<br />
at intervals <strong>of</strong> 20mm. Where the sherd was too small to estimate lip horizontal radius with<br />
confidence, EVE was estimated using a default radius <strong>of</strong> 140mm, approximating the<br />
average horizontal lip radius recorded for the larger sherds, about which there was little<br />
variation. Sherd sizes ranged up to a maximum lip/rim EVE <strong>of</strong> 25% <strong>of</strong> the vessel<br />
circumference, a recorded minimum <strong>of</strong> 1%, and a recorded average lip and rim EVE for<br />
all sites <strong>of</strong> 5.78% (n=471). Mean EVE by site is shown in Table 18.<br />
Grouping lip/rim sherds into sherd families was the most subjective step in<br />
constructing an estimation <strong>of</strong> the number <strong>of</strong> vessels in a breakage population. Sherds<br />
Table 18: Variation in lip brokenness between sites.<br />
Site Lip sherd count Average EVE (%)<br />
Paniavile 76 4.56<br />
Hoghoi 82 5.16<br />
Miho 60 4.41<br />
Honiavasa 104 7.14<br />
Gharanga 38 5.78<br />
Nusa Roviana 11 7.09<br />
Kopo 3 8.66<br />
Zangana 76 5.93<br />
sorted by site, then sorted further by fabric and by form variation, formed the basis <strong>of</strong> the<br />
primary vessel groupings, which were further subdivided by decoration information. This<br />
required some idea <strong>of</strong> the extent to which fabric and form might vary within a single<br />
250
vessel, and the best information pertaining to these problems was obtained from larger<br />
sherds. Fabric and form attributes were examined at multiple points on such sherds to<br />
develop an assessment <strong>of</strong> the amount <strong>of</strong> within-vessel variation.<br />
No instances were found in which temper <strong>of</strong> different petrographic classes (see<br />
Chapter 4) were incorporated into a single vessel. Grid counts <strong>of</strong> temper density and<br />
mineralogy, under low-powered reflected-light magnification, at multiple points on single<br />
sherds, indicated that there was only slight variation in temper mineralogy (in the relative<br />
proportions <strong>of</strong> mineral grains for example) within vessels, but that temper density could<br />
vary more dramatically within vessels. In sorting lip sherds by fabric, coarse variations in<br />
the relative proportions <strong>of</strong> minerals were used. Colour variation was disregarded, as<br />
surface colour appeared to have been affected by the aerobic regime <strong>of</strong> specific<br />
depositional context, and colour zonation in the break could be expected to vary within a<br />
single vessel as a result <strong>of</strong> firing variability.<br />
Form variation <strong>of</strong> the lip and rim within vessels was examined qualitatively by<br />
looking at larger sherds where a greater proportion <strong>of</strong> the vessel was represented (up to<br />
about 30%). Slight variation in form around the lip/rim circumference were noted, and<br />
small variations <strong>of</strong> this order <strong>of</strong> magnitude were ignored when sorting fabric groups into<br />
vessel groups based on form.<br />
Sherds for which the edges <strong>of</strong> the lip were weathered, making decorative analysis<br />
impossible, were generally small and water-rolled, and had been collected along the<br />
strandline. <strong>The</strong>se were omitted from the analytical sub-sample used to estimate<br />
completeness.<br />
It was noted that few if any <strong>of</strong> the more decorated lip/rim sherds could be matched<br />
to any other, suggesting either that decoration changed around the circumference <strong>of</strong> the<br />
vessel lip/rim, or that vessel completeness for all site assemblages was very low. Some<br />
sherds showed latitudinal decorative discontinuities <strong>of</strong> this type, but these discontinuities<br />
were restricted to a small range <strong>of</strong> incised motifs, omitted from subsequent analyses.<br />
251
This relationship between the amount <strong>of</strong> decorative information on lip and rim<br />
sherds pertaining to vessel membership, and the sherd count for vessel groups, was<br />
systematized by counting the number <strong>of</strong> latitudinal bands <strong>of</strong> repeated decorative elements<br />
that made up the total decoration <strong>of</strong> the lip and rim. <strong>The</strong>se were typically bands <strong>of</strong> repeated<br />
parallel edge-impression along the edges/top <strong>of</strong> the lip, or bands <strong>of</strong> punctation or incised<br />
decoration around the rim. <strong>The</strong> properties <strong>of</strong> the tool used to make the decorative marks<br />
and the average spacing and depth <strong>of</strong> marks were also used to differentiate between vessels<br />
in some cases. Here again, this depended on assessing through examination <strong>of</strong> larger<br />
sherds, the amount <strong>of</strong> variation one might expect in this regard within vessels. For all sites,<br />
this system yielded the data in Table 19 (vessel membership data is appended on CD in<br />
“Vessels.db”).<br />
Table 19: Selection <strong>of</strong> sample for estimating vessel completeness.<br />
Sherd type/<br />
# bands<br />
decoration<br />
Sherd<br />
count<br />
Vessel count Min. # per<br />
vessel<br />
252<br />
Max. # per<br />
vessel<br />
Mean # pieces<br />
per vessel<br />
(completeness<br />
lip/rim (0 bands) 91 53 1 14 1.72<br />
lip/rim (1 band) 264 166 1 15 1.58<br />
lip/rim (2 bands) 68 65 1 2 1.05<br />
lip/rim (3 or more<br />
bands)<br />
lip/rim (2 or more<br />
bands)<br />
28 25 1 2 1.12<br />
96 89-90 1 2 1.07-1.08<br />
Undecorated lips or lips with a single band <strong>of</strong> decoration (usually parallel edge<br />
impressions) yielded up to 15 pieces per vessel. <strong>The</strong> sample <strong>of</strong> more decorated sherds by<br />
contrast yielded lower vessel completeness, with a maximum <strong>of</strong> two lip/rim sherds per<br />
vessel, and an average <strong>of</strong> 1.07-1.08 sherds per vessel, despite similar brokenness. This<br />
lower level <strong>of</strong> completeness is considered to be a more accurate reflection <strong>of</strong> the state <strong>of</strong><br />
the assemblages than the figures for the plain or slightly decorated lips. <strong>The</strong> large vessel<br />
groups for the plain or slightly decorated lips are inferred to be faulty attributions <strong>of</strong>
sherds to vessel groups as a result <strong>of</strong> inadequate information with which to construct<br />
groups. This inference depends on the assumption that stylistic variation between sherds<br />
results from idiosyncratic recombination <strong>of</strong> the potters' repertoire <strong>of</strong> continuous horizontal<br />
bands <strong>of</strong> decoration for each pot constructed, and also on the related assumption that<br />
differences between sherds do not result from changes in decoration around the<br />
circumference <strong>of</strong> the vessel.<br />
Simulation Approach:<br />
<strong>The</strong> number <strong>of</strong> pieces into which a population <strong>of</strong> vessels breaks is variable. <strong>The</strong><br />
probabilities <strong>of</strong> individual vessels being represented in an archaeological sample are thus<br />
unequal. For pottery, the “death” <strong>of</strong> a pot (breakage) generally implies at least two pieces,<br />
although there are exceptions. Moreover, breakage is not necessarily a single event. As<br />
Orton states,<br />
“It is obvious that sherds can either stay the same size or become smaller<br />
(through breakage) over time (Orton et al. 1993:176).”<br />
In the latter case, these would not only become more numerous (increasing probability <strong>of</strong><br />
survival/capture in the archaeologist’s sample), but once below the detection size<br />
threshold, would become less likely to be found (decreasing probability <strong>of</strong> capture in the<br />
archaeologist’s sample). This last factor constrains the minimum size <strong>of</strong> sherds commonly<br />
recovered by archaeologists. An additional size-related factor in survival/detection is that<br />
smaller fragments are potentially more mobile, either within sediments as a result <strong>of</strong><br />
turbation or on the sediment surface due to fluvial or aeolian transport.<br />
While it is possible to envisage either progressive breakage (see Orton 1982 for<br />
a simulation approach to progressive breakage) or a state <strong>of</strong> breakage stasis, it is not<br />
possible to know from the sample assemblage which <strong>of</strong> these situations applied to a<br />
pottery breakage population; these are equifinal in that either progressive breakage or<br />
253
static breakage could produce any found broken assemblage. <strong>The</strong> state <strong>of</strong> brokenness in<br />
which a pottery breakage population existed is usually unknowable from the material<br />
record because the creation <strong>of</strong> a pottery breakage population is usually a diachronic<br />
process, and archaeologists only have access to the deposited/fossil assemblage through<br />
their sample assemblage as a synchronic (or nearly so) observation. A vast array <strong>of</strong> small<br />
breakage and re-breakage events may have occurred (or not) over the period prior to<br />
sample recovery, except in exceptional circumstances, such as where people or natural<br />
formation processes have curated or preserved pots in their death-state upon breakage,<br />
either acting continuously in this way over time, or where by some instantaneous process<br />
a whole assemblage <strong>of</strong> vessels is broken at once and lain undisturbed since (as at Pompeii).<br />
In such circumstances, preservation would be so good that there would be no need for<br />
estimation approaches.<br />
It is the property <strong>of</strong> potsherd samples that whole objects are broken into a number<br />
<strong>of</strong> fragments that allows us to establish vessel representation fraction (the fraction <strong>of</strong> the<br />
breakage population represented in the sample). If an assemblage <strong>of</strong> unbroken pots were<br />
recovered we would have no information on whether others existed at that location in the<br />
past and were carried <strong>of</strong>f in the intervening years. <strong>The</strong> more broken an assemblage, the<br />
more the survival chances <strong>of</strong> each vessel as a fragment improve (to a degree) as do the<br />
chances that some fragment <strong>of</strong> any individual vessel will make it into the archaeologist's<br />
sample.<br />
To reiterate, basic assumptions are:<br />
1: <strong>The</strong> sample assemblage is a random sample <strong>of</strong> vessel membership <strong>of</strong> the<br />
breakage population,<br />
2: vessel brokenness and completeness can be accurately measured from the sherd<br />
sample.<br />
Sample brokenness was defined as the mean number <strong>of</strong> rim sherds per vessel, and was<br />
254
calculated by dividing 100% by the mean EVE <strong>of</strong> sherds in the sample assemblage. As<br />
EVE is an estimate, assemblage brokenness is also an estimate, although no attempt is<br />
made here to calculate precision <strong>of</strong> that estimate. Vessel completeness was calculated as<br />
the estimated mean number <strong>of</strong> pieces per vessel in the archaeological sample, or the<br />
calculated mean number <strong>of</strong> pieces per vessel in any simulated pottery sample<br />
<strong>The</strong> simulation sherd sampling fraction was defined as the number <strong>of</strong> sherds one<br />
would have to remove from a simulated breakage population <strong>of</strong> complete vessels<br />
fragmented as observed in the sample assemblage, to achieve the level <strong>of</strong> completeness<br />
observed. This definition implies that brokenness <strong>of</strong> the assemblage was static in the past<br />
rather than progressive, and means that rather than asking the question 'what fraction <strong>of</strong><br />
an actual population <strong>of</strong> sherds near the time <strong>of</strong> breakage <strong>of</strong> the pottery is present in the<br />
sample?', the question posed in our approach is rather 'If the pottery in this deposit had<br />
existed since initial breakage in the state <strong>of</strong> brokenness observed in the sample, what<br />
fraction <strong>of</strong> the initial sherd population would the sample represent?'<br />
<strong>The</strong> simulation vessel representation fraction is the fraction <strong>of</strong> the simulated<br />
breakage vessel population represented at any given simulated sherd sampling fraction.<br />
While sherd sampling fraction is a stipulation device permitting the calculation <strong>of</strong> vessel<br />
representation in samples, based on the expected number <strong>of</strong> similar-sized pieces in the<br />
whole breakage population, vessel representation fraction seems to be a valid measure <strong>of</strong><br />
the relationship between sample and parent population, and informs as to the fraction <strong>of</strong><br />
the breakage population represented by an archaeological vessel/sherd sample. <strong>The</strong> sherd<br />
sampling fraction range is the range <strong>of</strong> outcomes in the simulation for which vessel<br />
completeness is consistent with the archaeological sample. <strong>The</strong> vessel representation<br />
fraction range is the range <strong>of</strong> simulation vessel representation fractions that are consistent<br />
with the vessel completeness <strong>of</strong> the archaeological sample.<br />
In the simulation, brokenness is treated as static, and is determined by specifying<br />
255
the number <strong>of</strong> vessels and the number <strong>of</strong> sherds in a simulated breakage population. Vessel<br />
completeness in these simulated breakage populations will <strong>of</strong> course always be 100%.<br />
Brokenness is made consistent with that observed in the archaeological sample. <strong>The</strong> goal<br />
<strong>of</strong> the simulation is to establish, for Roviana sample assemblage brokenness, the shapes <strong>of</strong><br />
the curves and the range <strong>of</strong> likely values that relate:<br />
a) simulation sherd sampling fraction to vessel representation fraction, and<br />
b) simulation vessel completeness to simulation sherd sampling fraction.<br />
This is achieved by simulating a large number <strong>of</strong> sherd sample assemblages with a level <strong>of</strong><br />
brokenness consistent with the archaeological sample. Vessel membership <strong>of</strong> sherds in all<br />
samples is assigned randomly, but overall, is consistent with the expected number <strong>of</strong> pieces<br />
per vessel at that sampling fraction. Completeness (mean number <strong>of</strong> sherds per vessel) and<br />
vessel representation fraction are calculated for each sample and are displayed in a plot to<br />
show the relationship between vessel completeness and vessel representation fraction. By<br />
this means a good estimate can be gained <strong>of</strong> the sherd and vessel sampling fractions in the<br />
simulation consistent with the vessel completeness and brokenness estimated for the<br />
archaeological sample.<br />
To reiterate the statement made earlier, this is not a suggestion that the breakage<br />
population had the particular static level <strong>of</strong> brokenness <strong>of</strong> the archaeological sample, and<br />
<strong>of</strong> all the simulated samples. <strong>The</strong> question <strong>of</strong> how many bits someone's pottery broke into<br />
when they dropped it is probably <strong>of</strong> little historical or archaeological interest, and is<br />
usually unknowable. Stipulating a level <strong>of</strong> brokenness consistent with the archaeological<br />
sample assemblages is rather a way <strong>of</strong> calculating, which allows the question 'how much<br />
needs to be taken away from the whole to give it the characteristics <strong>of</strong> the sample? We<br />
also do not really have to be concerned with what size any extant missing sherds are or<br />
were in the total discard assemblage, as ultimately it is vessel representation fraction we<br />
are after as an entry to quantification <strong>of</strong> the breakage population, rather than sherd<br />
256
sampling fraction itself.<br />
Estimating the number <strong>of</strong> vessels in a breakage population requires approaches that<br />
allow a variable number <strong>of</strong> members in classes (for example Chao 1984, Chao & Lee<br />
1992). Accordingly, the simulated samples are created as datasets containing a variable<br />
number <strong>of</strong> sherds per vessel. In the simulated breakage population, sherds are randomly<br />
assigned to vessel membership based on a probability derived from brokenness as measured<br />
by the archaeological sample mean EVE.<br />
Statistical Approaches:<br />
While the simulation approach is useful for elucidating the relationships between<br />
brokenness, completeness, and vessel representation in the sample, a number <strong>of</strong> statistical<br />
algorithms apply to the problem <strong>of</strong> estimating a breakage population from a sherd sample<br />
(estimating the number <strong>of</strong> classes/species in a population). <strong>The</strong>se have the advantage <strong>of</strong><br />
producing formal confidence limits. Of particular relevance to the problem are the capture-<br />
recapture statistics used in ecological research where a sample <strong>of</strong> animals are captured,<br />
tagged and then released back into the population. Once they have redistributed across the<br />
landscape, animals are captured and then released again in a series <strong>of</strong> trials. <strong>The</strong> frequency<br />
<strong>of</strong> capture <strong>of</strong> particular tagged animals is used to generate an estimate <strong>of</strong> the total<br />
population. While there is no attempt here to summarise the vast literature on this topic<br />
and the complexities involved, the following points are particularly relevant to the<br />
archaeological problem.<br />
<strong>The</strong> capture-recapture process works because the probability <strong>of</strong> capturing a tagged<br />
animal from a population tells us something about the population size. Getting one tagged<br />
animal from a single sample does not really tell you anything about an unknown<br />
population size. However, the more tagged animals you get and the more times the same<br />
257
tagged animal is recaptured in successive trials, the more certain you can be about the size<br />
<strong>of</strong> population that it probably derives from. <strong>The</strong>se statistics have been reworked to<br />
estimate the number <strong>of</strong> species in a population, in which form these are directly applicable<br />
to the problem <strong>of</strong> estimating the number <strong>of</strong> pottery vessels represented by a sherd sample.<br />
Here the number <strong>of</strong> ecological capture trials is analogous to the maximum number <strong>of</strong><br />
pieces from a vessel in the archaeologist's sample, with incidence <strong>of</strong> singleton sherds<br />
comprising the sample <strong>of</strong> animal captured in the first trial, the incidence <strong>of</strong> doubletons<br />
(pots represented by two sherds in the recovered sample) being the second capture sample,<br />
the incidence <strong>of</strong> tripletons (pots represented by three sherds) the third, and so on.<br />
This approach does not make explicit use <strong>of</strong> the sherd-size information inherent in<br />
pottery samples, which informs on the probability <strong>of</strong> capture (brokenness), which seems<br />
wasteful <strong>of</strong> information. Such information on the probability <strong>of</strong> capture could potentially<br />
be used to narrow the confidence limits obtainable for a given sample size, or to reduce the<br />
sample size required to obtain meaningful confidence limits.<br />
<strong>The</strong> Chao (1984) estimate was tested experimentally (Felgate & Bickler n.d) using<br />
data simulated from hypothetical breakage populations. Capture-recapture data calculated<br />
from these simulated samples was used to estimate (with confidence limits) the known<br />
breakage population, using the program “EstimateS”(Colwell 1997). <strong>The</strong> estimates from<br />
the Chao 1984 statistic were compared with the known breakage population.<br />
<strong>The</strong> key findings were that:<br />
• Sherd sample fractions <strong>of</strong> 5% or more allow a close estimate <strong>of</strong> the breakage<br />
population (this figure is a product <strong>of</strong> the size <strong>of</strong> the breakage population used in<br />
the trial, and could be improved using a larger population)<br />
• Even at 2% sampling fraction, the result is useful if the breakage population is<br />
large<br />
• Higher mean EVE (less broken assemblages) make the statistic less accurate,<br />
258
equiring a larger sampling fraction (but at the opposite extreme, in very broken<br />
assemblages, although the statistic works well, the data will be impossible to<br />
acquire; if the pottery is too broken the measurement <strong>of</strong> EVE itself is<br />
compromised).<br />
In contrast to the simulation method, the Chao statistic does not require discrete<br />
information on assemblage brokenness. <strong>The</strong> structure <strong>of</strong> the capture-recapture data<br />
provides this information. While the Chao statistic appeared to give good results only if<br />
the sherd sampling fraction is greater than 5% <strong>of</strong> the breakage population, this was a<br />
sample size effect. If the simulated breakage populations used by Felgate and Bickler had<br />
been larger than 10,000 sherds, confidence limits ought to have narrowed at lower<br />
sampling fractions.<br />
Simulation Results:<br />
Using the two-or-more-bands-<strong>of</strong>-decoration approach outlined above, completeness was<br />
estimated to be between 1.07 and 1.08 mean sherds per vessel. Brokenness was calculated<br />
by taking the average EVE <strong>of</strong> all lip sherds in the site regardless <strong>of</strong> the amount <strong>of</strong><br />
decoration present (5.78%). Site samples were combined to boost sample size. <strong>The</strong><br />
question being asked <strong>of</strong> the following simulation is therefore “What is the total breakage<br />
population for these six sites?”<br />
An initial simulated breakage assemblage <strong>of</strong> 10 000 sherds forming 578 vessels was<br />
stipulated, consistent with overall sample Brokenness (5.78% EVE). Initial simulation with<br />
500 sampling fractions from 0% to 100% indicated that vessel completeness <strong>of</strong> 1.08 sherds<br />
per vessel corresponds with simulation sherd sampling fraction <strong>of</strong> somewhere less than 3%<br />
(Figure 73).<br />
259
A second, detailed simulation was run using an enlarged breakage population <strong>of</strong><br />
Figure 73 Roviana vessel completeness against sampling fraction, sampling<br />
fraction against vessel representation fraction; random assignment <strong>of</strong> shreds to<br />
vessels; mean EVE <strong>of</strong> 5.78%.<br />
100,000 sherds (5780 vessels) at 11 sampling fractions between 0% and 2%, iterated 100<br />
times, plotting 1100 experiments in total. Vessel completeness <strong>of</strong> 1.07-1.08 sherds per<br />
vessel correspond on the simulation plot with vessel representation fraction <strong>of</strong> between<br />
8.9% and 17.5% (Figure 74).<br />
260
<strong>The</strong> total archaeological rim/lip sample was 471 sherds. Based on vessel completeness<br />
estimated to be from 1.07 to 1.08 sherds per vessel, this sample was inferred to physically<br />
represent between 436 and 440 vessels. This suggests limits for a total breakage population<br />
for the analyzed sites <strong>of</strong> 3259-4944 at 1.07 completeness, or at 1.08 completeness, 2491-<br />
4152 vessels. <strong>The</strong> combined approximate limits <strong>of</strong> the breakage population for this run <strong>of</strong><br />
the simulation program would be 2491-4944 vessels.<br />
Repeating the simulation two more times yielded vessel representation fractions<br />
(using completeness <strong>of</strong> 1.07-1.08) <strong>of</strong> 9.6%-19%, and 8.8%-18.9%. Combining the three<br />
ranges obtained yields vessel representation fraction range <strong>of</strong> 8.8%-19%, suggesting a<br />
more conservative approximation <strong>of</strong> breakage population confidence limits at 2sd would<br />
be 2295-5000 vessels.<br />
For comparison with the Chao calculation given below, a single completeness<br />
Figure 74: Roviana vessel completeness against vessel representation fraction;<br />
random assignment <strong>of</strong> sherds to vessels; mean EVE <strong>of</strong> 5.78%.<br />
261
value <strong>of</strong> 1.08 was used, yielding a total vessel count <strong>of</strong> 436 in the sample <strong>of</strong> 471 rim<br />
sherds. Vessel representation fraction averaged limits were 10.5%-17.5%, 11.2%-19%,<br />
and 10.6%-18.9%. Combined vessel representation fraction limits at this level <strong>of</strong><br />
completeness were 10.5% to 19%. <strong>The</strong> total sample assemblage <strong>of</strong> 436 vessels (at 1.08<br />
completeness) represents a breakage population <strong>of</strong> 2295-4152 vessels.<br />
Statistical Results:<br />
<strong>The</strong> results <strong>of</strong> analysis <strong>of</strong> the same Roviana data using the Chao (1984) statistic (Table 20)<br />
are similar. <strong>The</strong> results do not coincide, but the ranges overlap, which is encouraging,<br />
especially as the simulation results indicate that the Chao statistic is operating at its limits<br />
<strong>of</strong> usefulness at such small sampling fractions. <strong>The</strong> 89 vessels in the sample yielded parent<br />
population limits at 1s.d. <strong>of</strong> 293.38-716.16 vessels, or vessel representation fraction limits<br />
<strong>of</strong> 0.3034-0.1243. (At 2s.d., i.e. ±422, the breakage population represented by the 89<br />
vessels in the sample would be 83-927, illustrating how at these very small sampling<br />
fractions and limited sample sizes the Chao statistic has insufficient information on the<br />
probability <strong>of</strong> “capture” in the sample to give meaningful confidence limits.) <strong>The</strong> 436<br />
vessels in the total rim/lip sherd sample assemblage, at 1s.d., represent a breakage<br />
population <strong>of</strong> 1437-3508 vessels (compare this to a breakage population <strong>of</strong> 2295-4152<br />
vessels using the same data in the simulation).<br />
Table 20: Breakage population estimate for the combined Roviana highly decorated lip<br />
sample using the statistic <strong>of</strong> Chao 1984.<br />
Vessel count Sherd count Singletons Doubletons Breakage population 1s.d.<br />
89 96 82 7 504.77 211.39<br />
262
Discussion:<br />
In cases where the Chao statistic yields estimates with reasonably tight confidence limits,<br />
it is clearly a more systematic solution to the estimation problem than our simulation<br />
approach as it permits an estimate with confidence limits without the archaeologists<br />
specifically inputting brokenness information, which may itself be subject to error. In the<br />
case <strong>of</strong> the Roviana data the sample size is too small for the structure <strong>of</strong> the multiple<br />
capture data to inform as to the probability <strong>of</strong> capture, leading to an overly large standard<br />
deviation (at two standard deviations the 89 vessels in the sample would represent between<br />
83 and 927 vessels, which is not a very convincing result). <strong>The</strong> Chao estimate does rely on<br />
the structure <strong>of</strong> the capture-recapture data to establish the brokenness <strong>of</strong> the sample. This<br />
structure breaks down at small sampling fractions and small sample sizes. In these<br />
circumstances the simulation approach provides a better estimate than the Chao statistic,<br />
as brokenness is independently established from the EVE characteristics <strong>of</strong> the sample <strong>of</strong><br />
lip/rim sherds.<br />
Summary and Conclusions:<br />
Two methods <strong>of</strong> estimating a breakage population were applied, a simulation approach,<br />
and a statistical approach using the algorithm developed by Chao (1984). Both require that<br />
sherd families be identified in the sample. <strong>The</strong> simulation approach requires a measure <strong>of</strong><br />
completeness (in mean number <strong>of</strong> pieces per vessel) and brokenness (expected number<br />
<strong>of</strong> pieces per vessel) from the archaeological sample. <strong>The</strong> simulation approach proceeds<br />
by simulating many samples <strong>of</strong> a breakage population with a degree <strong>of</strong> brokenness<br />
consistent with the archaeological sample, to develop plots <strong>of</strong> the relationship between<br />
vessel completeness and vessel representation fraction. <strong>The</strong> measured vessel completeness<br />
263
from the archaeological sample is then compared with these plots to discover the<br />
corresponding vessel representation fraction indicated by the simulation. Dividing the<br />
number <strong>of</strong> vessels represented in the archaeological sample by this vessel representation<br />
fraction range yields the likely limits <strong>of</strong> vessels in the breakage population.<br />
<strong>The</strong> statistical approach uses a version <strong>of</strong> the capture-recapture approach for<br />
estimation <strong>of</strong> animal populations, adapted to the problem <strong>of</strong> estimating the number <strong>of</strong><br />
species in a population. In this application to archaeological pottery, the number <strong>of</strong> vessel<br />
sherd families in a breakage population is equivalent to the number <strong>of</strong> species in a<br />
population, and the presence <strong>of</strong> multiple sherds from the same vessel in the archaeological<br />
sample is equivalent to multiple capture events. This approach was tested on a range <strong>of</strong><br />
simulated breakage populations and simulated archaeological samples, and for the sample<br />
sizes tested, found to be accurate for sherd sampling fractions <strong>of</strong> 5% or more. Moreover,<br />
the statistic was more accurate when the vessel assemblage was more broken because <strong>of</strong><br />
the enlarged sherd sample for each vessel, which increases the probability <strong>of</strong> multiple<br />
“captures” at any given sherd sampling fraction.<br />
Drawing a theoretical distinction between a breakage population and a discard<br />
assemblage means that there is no required assumption that all <strong>of</strong> the sherds <strong>of</strong> pottery<br />
were discarded at the site. Sherds discarded or destroyed <strong>of</strong>f site are simply sherds that did<br />
not make it into the sample. Only when seeking a history <strong>of</strong> the pottery outside the sample<br />
assemblage does such transport or destruction become significant.<br />
Both estimation approaches require close attention to the nature <strong>of</strong> the total<br />
sampling processes. A vessel-random selection <strong>of</strong> sherds from the breakage population is<br />
the ideal. In archaeological terms this is most likely the case when sherds from the same<br />
vessel show a high degree <strong>of</strong> dispersion across the site. This suggests the approach is most<br />
suited to analysis <strong>of</strong> surface collections, or collections from large areal excavations, where<br />
dispersion information is attainable. Alternatively, other sampling approaches to<br />
264
archaeological sites such as multiple test excavations across a site might yield adequate<br />
information on the dispersion <strong>of</strong> sherd families.<br />
For both approaches, the accuracy <strong>of</strong> estimates <strong>of</strong> breakage population obtained<br />
depends on the accuracy <strong>of</strong> the estimate <strong>of</strong> vessel completeness. For the simulation<br />
approach an estimate <strong>of</strong> sample brokenness is also required. <strong>The</strong> Roviana data illustrate<br />
some <strong>of</strong> the difficulties in attaining reliable data <strong>of</strong> this sort. This requirement places a<br />
heavy burden on the analyst to develop reliable ways <strong>of</strong> sorting a sherd assemblage into<br />
vessel lots. This is not a simple analytical task and there is vast scope for methodological<br />
development in this area.<br />
For the overall Roviana sample, where vessel completeness is uniformly low, we<br />
can infer from the simulation that around 99% <strong>of</strong> the sherd material <strong>of</strong> the breakage<br />
population has not made it into the sample, and secondly, can appreciate that all is not lost,<br />
and that means exist for estimating the size <strong>of</strong> the total breakage population represented<br />
by the sample collected.<br />
Sample sizes were too small to estimate the composition <strong>of</strong> the breakage<br />
population for individual sites, and an argument is made elsewhere that some production<br />
styles could be better estimated from other vessel parts such as necks or carinations, due<br />
to fragile rim form (elaborated in Chapters 8 and 9).<br />
<strong>The</strong> extent <strong>of</strong> sherd attrition shows that assemblage composition is potentially<br />
biased and is most likely comparable only with other assemblages which have been subject<br />
to a similar taphonomic/sampling regime. It is clear that the vast bulk <strong>of</strong> sherdage deriving<br />
from the breakage population has not made it into the archaeologists sample. <strong>The</strong> count<br />
(assuming breakage stasis) or weight <strong>of</strong> sherds recovered would underestimate the<br />
intensity/duration <strong>of</strong> occupation by a factor <strong>of</strong> 100. <strong>The</strong> estimated number <strong>of</strong> vessels<br />
physically represented in the archaeological sample under-represents the size <strong>of</strong> the<br />
breakage population by a factor <strong>of</strong> between five and twelve.<br />
265
While some pottery was found buried in sediments in some sites, through test<br />
excavation, this could account for only a tiny fraction <strong>of</strong> the missing pottery, and as<br />
sample collection <strong>of</strong> the reef-flat surface scatter was intensive, it is inferred that the bulk<br />
<strong>of</strong> the breakage assemblage pottery had been removed by taphonomic processes.<br />
Observation <strong>of</strong> exotic pyroxene minerals in the local calcite sand <strong>of</strong> one site (Hoghoi)<br />
suggests much <strong>of</strong> the breakage population has dis-aggregated entirely into constituent<br />
mineral grains.<br />
<strong>The</strong> potential for bias in the composition <strong>of</strong> recovered sample assemblages is<br />
therefore inferred to be high for the sites as a group. <strong>The</strong> conclusion for any analyses that<br />
use relative abundance <strong>of</strong> styles or attributes (frequency seriation for example) must be<br />
that external comparisons should be restricted to sites that have undergone similar<br />
taphonomic processes. Direct comparisons should be avoided even with sites that seem<br />
to have similar cultural formation processes, but different taphonomic regimes (for<br />
example the Mussau sites).<br />
For accumulations research, the estimate will under represent some fragile lip<br />
forms, and also cannot represent types or sites that have not made it into the sample at all.<br />
It is entirely possible, indeed likely, that many more than 4000 vessels were broken around<br />
the intertidal zone <strong>of</strong> Roviana lagoon in the past, but the result obtained is sufficient to<br />
confirm the intuitive understanding that most <strong>of</strong> the sherdage must have gone, in order to<br />
produce the observed pattern, and that at least five times as many vessels as the number<br />
observed in the sample were broken here.<br />
How fragile is this site type? <strong>The</strong> degree <strong>of</strong> wave exposure for sites is compared<br />
in Chapter 6 but all sites are broadly similar due to a broadly comparable situation within<br />
the Lagoon. It is clear from the simulation that around 99% <strong>of</strong> the physical quantity <strong>of</strong><br />
ceramics has not been captured in the sample, and as sampling was quite intensive, aiming<br />
at total collection <strong>of</strong> lip sherds, it is clear that site preservation even in the sheltered waters<br />
266
<strong>of</strong> the Lagoon is marginal, and that this type <strong>of</strong> site is almost entirely removed by<br />
taphonomic processes.<br />
This result suggests that aceramic lithic scatters in similar locations may have had<br />
a ceramic component in the past, removed by differential weathering <strong>of</strong> ceramics and<br />
lithics. Had similar sites been located in more wave-exposed settings elsewhere in Near<br />
Oceania, we can be certain that few <strong>of</strong> these would have survived to be detectable as<br />
ceramic scatters by an archaeologist today. <strong>The</strong> poor state <strong>of</strong> preservation <strong>of</strong> the Roviana<br />
intertidal-zone sites suggest settlement patterns, site density etc for this site type cannot<br />
be read directly from survey results, and that preservation and visibility factors ( probability<br />
<strong>of</strong> detection) must be carefully considered to read survey results.<br />
Finally, the low state <strong>of</strong> completeness <strong>of</strong> the Roviana vessels indicates that sherd<br />
counts <strong>of</strong> decorative attributes are likely to be independent in the vast majority <strong>of</strong> cases,<br />
and that sherd count is the best unit for counting decorative attributes (quantifying sample<br />
size) for seriation analysis.<br />
267
268
Introduction:<br />
CHAPTER 6:<br />
WAVE EXPOSURE<br />
Systematic consideration <strong>of</strong> the effects <strong>of</strong> wave exposure on archaeological deposits are<br />
rare or superficial in recent Oceanic archaeology, despite the situation <strong>of</strong> the subject in the<br />
largest wave basin on the planet. In the 1920s the prevailing paradigm in Oceanian<br />
archaeology was that there was no Oceanic subsurface archaeology due to the erosional<br />
effects <strong>of</strong> cyclones, floods, high temperatures and damp conditions (an example <strong>of</strong> this<br />
paradigm can be found in Linton’s introduction to his “Archaeology <strong>of</strong> the Marquesas”<br />
published in 1925 (Linton 1925:3-4), but from the 1930s through to the present the<br />
pendulum <strong>of</strong> scientific opinion has swung to the opposite extreme, with the archaeological<br />
record read relatively directly, and minimal attention has been paid to formation processes<br />
in general and the degree to which wave exposure structures Oceanian archaeological<br />
distributions. A sample surveying approach by contrast demands attention to probability<br />
<strong>of</strong> site detection, including preservation and visibility, as lenses through which the<br />
archaeological record is read more accurately.<br />
In Chapters 7 and 11 the ceramic scatters are viewed as lag deposits, evidenced by<br />
a high degree <strong>of</strong> size sorting. <strong>The</strong> value in constructing this model <strong>of</strong> wave exposure in<br />
the current <strong>chapter</strong> lies not so much in being able to test it against sherd size data, but in<br />
having some means other than the sizes <strong>of</strong> recovered sherds to assess relative exposure.<br />
Sherd size is subject to a collection intensity effect, making highly size-sorted<br />
assemblages difficult to distinguish from poorly sorted assemblages unless scrupulous<br />
control for collection intensity can be maintained (difficult unless one knows the entire<br />
269
history <strong>of</strong> collection <strong>of</strong> the site). Modelling wave-exposure allows a cross-check on size<br />
data, and promotes the view that factors additional to natural transforms are structuring<br />
the taphonomic data in assemblages (in the next <strong>chapter</strong> some conflicts between expected<br />
size distribution from wave exposure and actual size distribution <strong>of</strong> the archaeological<br />
sample will be highlighted which illustrate this).<br />
Review:<br />
Some studies in which wave exposure receives particular mention are reviewed below<br />
(passing reference is made in many studies to wave-laid sediments in excavations, but these<br />
are reviewed elsewhere (Tarlton 1996).<br />
Green’s survey <strong>of</strong> the Surville peninsula on Makira (San Cristobal) in 1970 noted<br />
the dramatic effects <strong>of</strong> 1971 cyclone waves on coastal rockshelters. For Ulawa in the<br />
southeast Solomon Islands (Hendren 1976, Ward 1976) Ward notes evidence for exposure<br />
<strong>of</strong> this coast to cyclone waves, and effects on terrestrial site preservation. <strong>The</strong> occupation<br />
sites excavated by Ward postdated about 700AD. Results suggest terrestrial pottery sites<br />
<strong>of</strong> any period are rare or absent from the extant archaeological record.<br />
Lilley in a study <strong>of</strong> islands in the Vitiaz Strait noted the likely impact <strong>of</strong> the 1888<br />
Ritter tsunami on the north coast <strong>of</strong> Umboi, parts <strong>of</strong> coastal Sakar, western New Britain<br />
and the low-lying Siassi Islands, and coasts <strong>of</strong> Malai and Tuam islands (Lilley 1986:20),<br />
and suggested that low levels <strong>of</strong> site visibility and survival could be expected as a result <strong>of</strong><br />
similar events and also as a result <strong>of</strong> storm action (Lilley 1986:106). Lilley did not make<br />
any sampling-based inferences regarding archaeological distribution in the past, or use this<br />
type <strong>of</strong> information in inferring site formation processes, despite his Lapita-phase evidence<br />
(pottery from the KLK site on Tuam) being,<br />
“...associated with a very restricted quantity and range <strong>of</strong> other cultural<br />
material in a clean beach sand matrix which is devoid <strong>of</strong> structural features<br />
270
such as hearths and postholes (Lilley 1986:455).”<br />
He inferred, like Wickler for Buka, that the Lapita-phase pottery represented, “...probably<br />
only intermittent activity in Siassi generally and on Tuam in particular and second, that<br />
people did not actually live on the KLK site”, but did not regard the lack <strong>of</strong> occupation<br />
features as related to natural formation processes.<br />
<strong>The</strong> effect <strong>of</strong> storm surges and cyclone waves has been studied in some detail for<br />
Tongatapu archaeological sites (Spennemann 1987). Spennemann concluded that waves,<br />
including storm waves, could play an important part in site formation, site visibility, and<br />
site destruction, but gave examples also <strong>of</strong> beach scatters that were the transformed (lag<br />
deposit) remains <strong>of</strong> terrestrial archaeological sites. In his Tongan case study major<br />
erosional effects were limited to exposed sand cays lying <strong>of</strong> the northeast coast <strong>of</strong><br />
Tongatapu. He did not seek to develop any predictive use <strong>of</strong> the information, arguing<br />
instead for continual monitoring and rescue excavation <strong>of</strong> sites subject to coastal erosion.<br />
A general study on the subject <strong>of</strong> tropical cyclones and their effect on Pacific<br />
archaeological sites (Tarlton 1996) considered cyclone frequency for various island groups<br />
in the southwest Pacific and their impact on various island types. <strong>The</strong> study focused<br />
exclusively on terrestrial archaeological sites, while in the Roviana case the evidence<br />
suggests maritime site location in the past. Thus her discussion <strong>of</strong> increases in water level<br />
as having largely negative impacts on site preservation is largely inapplicable to stilt<br />
occupation over water in sheltered locations, where increases in water level during storms<br />
should theoretically protect archaeological deposits from swash processes.<br />
Tarlton also concluded that high islands with indented coastlines were likely to<br />
have sheltered areas relatively unaffected by cyclone waves (Tarlton 1996:108, 114), and<br />
that landforms resulting from cyclone activity or tsunami were readily identifiable in many<br />
instances (Tarlton 1996:118), both conclusions pertinent to discussion <strong>of</strong> Roviana<br />
intertidal formation processes. Tarlton considered that awareness <strong>of</strong> cyclone or other<br />
271
wave-generated landforms would enable an archaeologist to choose locations which were<br />
relatively undisturbed where stratified archaeological deposits might be discovered (Tarlton<br />
1996:123).<br />
Tarlton reviewed archaeological reports <strong>of</strong> cyclone or other wave damage to sites<br />
or landscapes. Among these the most sophisticated distributional inferences seem to be<br />
from Australia, for the Queensland coast (Rowland 1989), and the Torres Strait islands,<br />
where whole coastal landscapes have a lack <strong>of</strong> archaeological evidence thought to relate<br />
to wave damage <strong>of</strong> sites or wave erosion <strong>of</strong> coastlines (Vanderwal 1973:187). A similar<br />
explanation for the distribution <strong>of</strong> Lapita and other early sites in New Caledonia was cited<br />
as a personal communication from Jean-Christophe Galipaud in 1995 (Tarlton 1996:136-<br />
138).<br />
Tarlton provided an extensive section on Kirk’s numerical modeling <strong>of</strong> cyclone<br />
wave exposure for the Rarotongan coastline and suggested that cyclone wave impact on<br />
archaeological sites could be modeled in a similar way using GIS, perhaps in combination<br />
with remote-sensing identification <strong>of</strong> cyclone wave-generated landforms (Tarlton<br />
1996:189-196), but did not extend her discussion to the interpretation <strong>of</strong> regional<br />
archaeological distributions, seeing this approach as principally as an aid to site<br />
prospecting, for locating undisturbed archaeological sites. It is in this section though, that<br />
discussion <strong>of</strong> a predictive modeling approach for a specific coastline is closest to the type<br />
<strong>of</strong> analysis conducted in the present <strong>chapter</strong>. Her emphasis on storm surge models and<br />
wind models was appropriate for exposed coastlines and/or terrestrial sites, but these are<br />
inapplicable in the Roviana case, where collection sites comprise reworked intertidal and<br />
sub-tidal evidence rather than in-situ terrestrial sites.<br />
Roviana Intertidal Ceramics and Waves:<br />
272
Intertidal sites in the Roviana study were all thought to have had broadly comparable post-<br />
deposition natural formation processes, due to location in an almost landlocked lagoon<br />
setting bordered by either mainland high-island shorelines or upraised Plio-Pleistocene<br />
reefs. Discrimination between some <strong>of</strong> the models <strong>of</strong> formation processes constructed in<br />
Chapter 7 require some independent assessment <strong>of</strong> wave exposure variation between sites.<br />
A model <strong>of</strong> wave exposure is presented in this <strong>chapter</strong> which deals with seasonal weather<br />
patterns which can be expected to have prevailed in varying degrees in the past, and a<br />
model <strong>of</strong> wave formation based primarily on the variable “fetch”, which is the distance in<br />
a windward direction between a shore and the nearest land.<br />
Waves coming onto a beach increase in height and steepness until they break. As<br />
waves move into shallow water, circular orbits <strong>of</strong> the water particles become flattened, and<br />
some wave energy will be used in moving sediments to and fro on the sea bed. This is<br />
particularly so for shallow-sloped sea beds, such as the reef flats on which some <strong>of</strong> the<br />
Roviana ceramic sites are located. As a wave breaks, most <strong>of</strong> the energy is dissipated in<br />
the final mixing <strong>of</strong> water, sand and shingle (Open <strong>University</strong> Oceanography Team 1989:27-<br />
29). It is these two “swash zone” effects on sediments which are <strong>of</strong> interest to the shallow-<br />
water/ intertidal archaeologist. <strong>The</strong> generation <strong>of</strong> waves by seismic processes and by wind<br />
will be considered separately below.<br />
Tsunami:<br />
Exposure <strong>of</strong> shorelines within Roviana Lagoon to large waves generated by seismic events<br />
is comparatively low. <strong>The</strong> Blanche Channel, to which the Roviana reef passages open, is<br />
sheltered by Rendova/Tetepare islands, from tsunami originating in the Solomon Sea<br />
(Figure 6). Earthquake epicentres recorded between 1962 and 1967 show only a single<br />
shallow epicentre within the Blanche Channel, the vast majority <strong>of</strong> seismic events having<br />
their epicentres to the north, in the Bougainville area, or well south in the Southeast<br />
273
Solomons (Dunkley 1986:9, Mann et al. 1998). In the event that seismic activity did create<br />
waves in the Blanch channel, the effect within the Lagoon would be negligible due to the<br />
narrowness <strong>of</strong> the reef passages through the uplifted coral barrier islands. It is possible that<br />
refracted waves generated within the Blanche channel by seismic activity could have some<br />
impact on sites like Honiavasa and Miho, located at the inner end <strong>of</strong> a reef passage,<br />
adjacent to deeper water, but only a few small events <strong>of</strong> this type are known <strong>of</strong> , in contrast<br />
to their common occurrence on the Solomon Sea coasts <strong>of</strong> Rendova and elsewhere in the<br />
New Georgia group.<br />
<strong>The</strong> lagoon basin is interrupted with many linear and patch-like shoals, thought to<br />
indicate a submerged karst-like landscape (Stoddart 1969a:394), and generation <strong>of</strong><br />
substantial waves within the lagoon by seismic activity seems unlikely in view <strong>of</strong> the<br />
baffling effect <strong>of</strong> this dissected bathymetry.<br />
Wind-generated Waves:<br />
Waves in this context are defined as the process by which wind kinetic energy is<br />
transferred to and transported across water, as a progressive wave. For an idealised wave,<br />
wave height refers to the vertical distance between trough and peak, which is twice the<br />
wave amplitude, while wavelength is the distance between two successive waves.<br />
Steepness is wave height divided by wave length. Period is the time interval between two<br />
successive analogous points on the wave pattern passing a fixed location. <strong>The</strong> number <strong>of</strong><br />
peaks or troughs passing a fixed point per second is the frequency. Wave speed is related<br />
to wavelength (Open <strong>University</strong> Oceanography Team 1989:7, 10).<br />
On the open ocean, after a period <strong>of</strong> calm weather, when the wind starts to blow,<br />
small, steep waves form as the wind increases, continuing to grow after the wind has<br />
reached maximum speed until they reach a wavelength equal to one third <strong>of</strong> windspeed.<br />
Fetch is the unobstructed distance <strong>of</strong> sea in which the wave can build. Wave size depends<br />
274
on the fetch (Open <strong>University</strong> Oceanography Team 1989:11-12). Fetch imposes the prime<br />
limits on wave size, speed and energy in sheltered waters like Roviana lagoon and the<br />
Blanche Channel, and provides an easily measured proxy for wave exposure when<br />
combined with wind direction.<br />
“<strong>The</strong> factors that are important in increasing the amount <strong>of</strong> energy that<br />
waves obtain are (1) windspeed, (2), the time during which the wind blows<br />
in one direction, and (3) the fetch.....When both maximum fetch and<br />
duration are reached for a given wind velocity, the sea is said to be fully<br />
developed (Thurman 1981:219).”<br />
This is why windspeed and duration are less significant than fetch for the Roviana intertidal<br />
sites, as the limits imposed by short fetch are quickly reached, beyond which waves do not<br />
build. <strong>The</strong> following three anecdotes illustrate the primacy <strong>of</strong> fetch in determining wave<br />
exposure.<br />
During the 1998 cyclone (which tracked through the Rennell/Bellona area)<br />
northerly winds were estimated by the author to have exceeded fifty knots. In these wind<br />
conditions, children were sailing their canoes at respectable speed in the bay near the Miho<br />
site, powered by coconut fronds, across water barely more than rippled, due to the limits<br />
on wave energy imposed by short fetch across that bay. For a sea to become fully<br />
developed in similar windspeed to this would require a fetch <strong>of</strong> about 2500km and a<br />
duration <strong>of</strong> 65 hours (Thurman 1981:219).<br />
By way <strong>of</strong> contrast, moderate tradewind conditions in the Blanche Channel <strong>of</strong><br />
around 25 knots windspeed, <strong>of</strong> extended duration, where the fetch to the southeast is<br />
about 65km, allow a sea to become partially developed, with short and relatively steep<br />
waves hazardous to small craft a fetch <strong>of</strong> about 180km is needed for a sea to become fully<br />
developed for this windspeed).<br />
Wave refraction is well understood in oceanography, allowing oceanographers to<br />
determine regions that are likely to experience high waves due to this effect. Refraction<br />
275
can change wave height and direction, as can shallowing water (Open <strong>University</strong><br />
Oceanography Team 1989:25). Locations such as the inner margins <strong>of</strong> reef passages may<br />
be exposed to refraction waves even though the measured direct fetch is low.<br />
Seagrass, Mangroves and Reefs:<br />
Seagrasses, mangrove forests and reef shallows all serve to mitigate wave exposure, by<br />
effectively reducing fetch at low tide, acting as an energy absorbing barrier to wave<br />
formation and transmission.<br />
Roviana Lagoon nowadays has extensive shallow mixed seagrass meadows,<br />
characterized most visibly by the long strap-like Enhalus acoroides, in which Dugong<br />
dugon and Chelonia mydas graze, and which shelter a pr<strong>of</strong>usion <strong>of</strong> fish, mollusc and<br />
crustacean species. Besides being attractive locations for finding a meal (easily traversed<br />
by paddle canoe, and less easily if reliant on a rotating propellor), these meadows also have<br />
a sheltering effect on marine archaeological deposits, trapping organic and reefal detritus,<br />
and acting as sediment stabilizers (Stewart 1999:581). Seagrasses have the potential to<br />
create stratigraphy by sealing archeological deposits, and also are potentially agents <strong>of</strong><br />
bioturbation through the growth <strong>of</strong> root systems, as are many <strong>of</strong> the creatures that live<br />
within the shelter <strong>of</strong> seagrasses, some <strong>of</strong> which are burrowers and some <strong>of</strong> which pluck<br />
grasses from sediments when grazing.<br />
Seagrass shallows are found principally in the eastern two-thirds <strong>of</strong> the Lagoon,<br />
east <strong>of</strong> Honiavasa, where the lattice <strong>of</strong> shallow reefs nears the surface <strong>of</strong> the water, as a<br />
result <strong>of</strong> a slight tectonic tilting (Mann et al. 1998). West <strong>of</strong> Miho site, the coral heads tend<br />
to be deeper below the surface, as far west as Nusa Roviana, although there is a lattice <strong>of</strong><br />
shallow reefs along the northern (mainland) coastline <strong>of</strong> the lagoon in this region.<br />
Seagrasses are not common west <strong>of</strong> Miho, other than sparse occurrences along the shallow<br />
mainland coast and among the forest <strong>of</strong> islets between Miho and Zangana.<br />
276
Forty-three percent <strong>of</strong> the worlds mangrove species are represented in Solomon<br />
Islands, with 26 species in 13 families (Oreihaka 1997). Mangroves protect many <strong>of</strong> the<br />
Lagoon’s shorelines from wind and waves, having a similar stabilizing role to seagrasses,<br />
and harbouring crustaceans, molluscs and many different fish species, some <strong>of</strong> which are<br />
agents <strong>of</strong> bioturbation. Mangroves may create sedimentary strata post-deposition. At the<br />
Hoghoi site, for example, testpits revealed an orange-brown organic layer beneath the top<br />
10cm <strong>of</strong> sediment, composed almost entirely <strong>of</strong> living (mangrove?) rootlets, with sherds<br />
found both above and below this layer. This could end up looking like bedding planes in<br />
a deposit <strong>of</strong> pottery fragments after the death <strong>of</strong> the mangrove system, which would leave<br />
a thin dark organic band in an excavation section pr<strong>of</strong>ile. Mangroves are found principally<br />
along the mainland coast <strong>of</strong> the lagoon, where s<strong>of</strong>t sediments predominate, and are present<br />
in smaller areas along parts <strong>of</strong> the raised-coral barrier system, where they are found<br />
associated with s<strong>of</strong>ter sediments.<br />
Solomon Islands Tides:<br />
<strong>The</strong> reason tides are significant is that land and sea breezes follow a diurnal cycle, as do<br />
tides in the region for much <strong>of</strong> the year, thus creating a situation in which wind directional<br />
changes or changes in wind strength are in phase with tidal changes for extended periods,<br />
as will be detailed below, leading to differential wave exposure <strong>of</strong> coastlines in some<br />
seasons. Numerous shallow reefs within the lagoon shelter many <strong>of</strong> the shorelines at low<br />
tide (when intertidal scatters are most at risk <strong>of</strong> wave damage) but for those shorelines not<br />
sheltered by reefs within the Lagoon, sea-breeze wave exposure at low tide is thus likely<br />
to have had a more severe effect on the pottery scatters, which are more exposed to<br />
swash-zone energy at low tide.<br />
Lunar-induced tides have a fundamental periodicity <strong>of</strong> 12 hours and 25 minutes,<br />
277
and operate over a cycle <strong>of</strong> 27.3 days governed by the rotation <strong>of</strong> the moon-earth system.<br />
Usually, in most parts <strong>of</strong> the world, for this reason, the time <strong>of</strong> the high tide shifts slightly<br />
each day (the diurnal calendar being based on a 24 hour solar periodicity). Lunar tidal<br />
forces are also influenced by the moon’s declination either side <strong>of</strong> the equatorial plane,<br />
which varies over a cycle <strong>of</strong> 27.2 days, and controls diurnal variation. <strong>The</strong> Moon’s<br />
elliptical orbit also causes variation in tide-producing forces over a 27.5 day cycle<br />
(Macmillan 1966:48-49). Solar tidal forces are <strong>of</strong> smaller magnitude than lunar tidal forces,<br />
with a semi-diurnal period <strong>of</strong> twelve hours.<br />
<strong>The</strong> declination <strong>of</strong> the sun varies over a yearly cycle, and the elliptical orbit <strong>of</strong> the<br />
earth around the sun has a slight effect on a yearly cycle also. Diurnal inequalities (the<br />
tendency for the two tides <strong>of</strong> a 25 hour period to be <strong>of</strong> unequal height) are affected by<br />
declination <strong>of</strong> the sun and moon, and will be greatest at times <strong>of</strong> maximum declination, and<br />
least at times <strong>of</strong> minimal declination (Macmillan 1966:44). This diurnal inequality is<br />
significant for the Roviana case, where for much <strong>of</strong> the year a solar tidal cycle<br />
predominates, as will be explained below. When solar and lunar effects are in phase, either<br />
in conjunction or opposition, large tidal ranges occur, and when they are out <strong>of</strong> phase the<br />
range is small. <strong>The</strong>se are referred to as spring tides and neap tides respectively, changes<br />
which follow a 29.5 day cycle (Macmillan 1966:50).<br />
<strong>The</strong> Pacific ocean appears to favour the occurrence <strong>of</strong> small-range diurnal tides,<br />
and furthermore, the solar contribution sometimes exceeds that <strong>of</strong> the lunar in the Pacific<br />
area,<br />
“...probably due to the resonance <strong>of</strong> the solar disturbing period with the<br />
natural period <strong>of</strong> the body <strong>of</strong> water affected. In consequence, under these<br />
circumstances the diurnal tide due to solar influence dominates that due to<br />
the moon (Macmillan 1966:60).”<br />
Solomon Islands tides seem to be <strong>of</strong> the mixed type, varying seasonally between diurnal<br />
and semi-diurnal with the changing influence <strong>of</strong> solar forces. Tide predictions for Honiara<br />
in year 2000 and 2001 (Sea levels for Honiara were obtained from the National Tidal<br />
278
Facility, <strong>The</strong> Flinders <strong>University</strong> <strong>of</strong> South Australia) show a mixed tide regime where a<br />
diurnal tidal cycle (1 high tide every 24 hours) is dominant for most <strong>of</strong> the year, with high<br />
water occurring near noon in January, February and early march, tending to have a<br />
morning high water in March, and shifting to a semi-diurnal regime in April (two tides in<br />
24 hours), with only small tidal range for this cycle. May sees a return to a diurnal cycle,<br />
but with low tide during the day and high tide in the early hours <strong>of</strong> the morning. June is<br />
dominated by a diurnal cycle with low tide at mid-day and high water at night, and this<br />
pattern continues through July and August, being modified towards a semi-diurnal cycle<br />
in September and October. November sees a return to a stronger diurnal cycle, but with<br />
high water during the day, with the same pattern in December. Although tidal records for<br />
the New Georgia group were not accessed, this pattern is consistent with the author’s own<br />
observation and information from local informants during 1996-1998 at Roviana Lagoon.<br />
This annual pattern shows a strong influence <strong>of</strong> solar declination, dividing the year<br />
into four tidal seasons, with the sun creating a semi-diurnal effect at the equinoxes and a<br />
diurnal regime at the solstices. Within this annual pattern is a monthly cycle created largely<br />
by the changing declination <strong>of</strong> the moon, and the rotation <strong>of</strong> the moon-earth system. <strong>The</strong>re<br />
is a daytime high tide from November to March (October to January according to Aswani<br />
1997:235), and a night time high tide from May through September. <strong>The</strong> timing <strong>of</strong> daily<br />
high-water is following a solar cycle rather than a lunar one (as outlined by Macmillan<br />
1966:60).<br />
<strong>The</strong> daytime solar tide is superior around the summer solstice, and the night-time<br />
tide becomes superior around the time <strong>of</strong> winter solstice. Why this should be in the daytime<br />
through the months <strong>of</strong> northern solar declination, and at night through the months <strong>of</strong><br />
southern declination is explained by a the mechanism <strong>of</strong> diurnal inequality (Macmillan<br />
1966:42), whereby the declination <strong>of</strong> the relevant tide-producing body (in this case the sun)<br />
produces, instead <strong>of</strong> the theoretical equilibrium tide, superior and inferior semi-diurnal<br />
tides (in the Roviana case the inferior tide is completely suppressed when the declination<br />
<strong>of</strong> the moon is low, but is present as a slight tidal oscillation when declination <strong>of</strong> the moon<br />
is high, on a monthly cycle).<br />
<strong>The</strong> broad characteristics <strong>of</strong> this tidal system as explicated above seem likely to be<br />
a pattern <strong>of</strong> antiquity as well as the present, as there seems to be few reasons for this<br />
to have varied over the timescale <strong>of</strong> the past few thousand years. Only at the geological<br />
timescale, where the Pacific Ocean basin undergoes major changes in plate tectonics, can<br />
this resonance with solar tidal forces be expected to change, as a result <strong>of</strong> the changing<br />
279
size and depth <strong>of</strong> the basin, within the current (and longstanding) harmonic theory <strong>of</strong> tides.<br />
Winds:<br />
<strong>The</strong> nearest meteorological station to the Roviana Lagoon is situated in Munda, away from<br />
the coast, surrounded by trees and buildings, and while this location is no doubt convenient<br />
for staff who take the readings, recorded windspeeds and directions cannot be expected<br />
to bear much relationship to winds affecting the Lagoon. This section is therefore written<br />
using a mix <strong>of</strong> data pertaining to the general region acquired from the Pacific Pilot<br />
(Hydrographic-Office 1971: diagrams 8-11), data from personal observation, and data<br />
from Aswani’s thesis (Aswani 1997:238-239). (Author’s note: the airport building was<br />
shifted circa 2001, allowing slightly more exposed location for weather equipment.)<br />
<strong>The</strong> southeast trades (Gevasa) prevail from April/May/June through to about<br />
October, and their strength and duration varies from year to year. Direction is recorded as<br />
blowing from 125 degrees true (Hydrographic-Office 1971:diagram 10). Personal<br />
observation over the course <strong>of</strong> three years suggests the local direction <strong>of</strong> this wind is more<br />
like 135 degrees, as at the Honiavasa reef passage it blows consistently directly from<br />
Tombatuni Island in the Blanche channel. This wind tends to strengthen during the day,<br />
probably due to the local effect <strong>of</strong> the sea breeze, at a time <strong>of</strong> year when tides are low<br />
during the day, so that any pottery scatter at such a depth as to be exposed to the<br />
Southeast, or to refraction waves from the Blanche channel curving around the inner points<br />
<strong>of</strong> the reef passages, can be expected to have undergone a relatively harsh taphonomic<br />
process. Low tides during the day during the season <strong>of</strong> the SE trade mean that any<br />
intertidal pottery sites with a large fetch to the SE (e.g. Gharanga, or the Mbolave lithic<br />
scatter/ceramic findspot) can be expected to undergo a harsh swash-zone taphonomic<br />
regime, unless sheltered by reefs or s<strong>of</strong>t sediments and vegetation.<br />
<strong>The</strong> Northeast trade blows strongly for a period <strong>of</strong> a month or so in late summer,<br />
but locally from the Northwest, and is known by local expatriates as the Northwest<br />
Monsoon (this may be the “Peza” referred to by Aswani, but the period and direction<br />
differ from Aswani’s suggestion that this wind is a westerly. <strong>The</strong> Roviana dictionary lists<br />
Peza as the northwest monsoon (Waterhouse 1926:91). <strong>The</strong> recorded direction <strong>of</strong> origin<br />
<strong>of</strong> the NE trade is about 70 degrees true (Hydrographic-Office 1971: diagram 9). Personal<br />
observation suggests that direction is locally influenced by the New Georgia landmass<br />
280
with its volcanic peaks, tending to blow more from the north to NNW, and tending to<br />
strengthen in the evening and at night with the addition <strong>of</strong> a slight land breeze. Tides at this<br />
time <strong>of</strong> year are predominantly semi-diurnal (two highs and two lows every 24 hours), so<br />
are not in phase with this strengthening <strong>of</strong> the wind in the evening. Sites exposed to<br />
significant Northwesterly fetch in this direction at low tide can be expected to show some<br />
effect <strong>of</strong> this harsher taphonomic regime.<br />
Transitional periods between these seasonal prevailing winds are dominated by a<br />
diurnal cycle <strong>of</strong> light land (night) and sea (day) breezes (Sea breezes blow in from the<br />
Blanche Channel toward the centre <strong>of</strong> New Georgia during the day, and land breezes blow<br />
<strong>of</strong>f the coast towards the Blanche Channel at night). <strong>The</strong>se breezes are most noticeable at<br />
midsummer, in the absence <strong>of</strong> other winds, but the forces generating these winds, the<br />
differential heating and cooling <strong>of</strong> land and sea, act in combination with other winds in the<br />
cooler months also. When tides are high during the day, sites with seaward fetch are<br />
unlikely to be affected as the pottery is well submerged, but sites with landward fetch at<br />
low tide at night (Hoghoi for example) might be expected to be exposed to a build-up <strong>of</strong><br />
small waves through the course <strong>of</strong> the night.<br />
Tropical depressions, tropical storms, and cyclones tend to originate between<br />
Latitude 10 degrees south and Latitude 15 degrees south (Hydrographic-Office 1971:36<br />
and diagram 12). <strong>The</strong>re are an average <strong>of</strong> about two cyclones per year in the region, and<br />
these generally pass to the south <strong>of</strong> the Solomons, and can be expected in most instances<br />
to create strong winds veering (progressing around the compass clockwise) through<br />
northerly as a result although other directions are not ruled out, as some rare cyclones<br />
originate in Solomon Islands. Holocene climate changes may have affected cyclone tracks<br />
and other climatic zonation in the past (Tarlton 1996:18-28), but no information is<br />
available at present on the nature <strong>of</strong> such changes in the period since deposition <strong>of</strong> the<br />
pottery. It seems likely that the presence <strong>of</strong> the high landmasses <strong>of</strong> Choiseul and Ysabel<br />
to the east are moderating influences restricting the formation <strong>of</strong> cyclones in the region,<br />
and that this effect is <strong>of</strong> great geological age.<br />
<strong>The</strong> season for tropical storms is December through April, with most occurring in<br />
January, February and March. As their average track speed is about eight knots, they can<br />
affect an area for many days and nights (Hydrographic-Office 1971:36), so tidal periodicity<br />
is irrelevant to wave exposure during tropical cyclones.<br />
Cyclone storm surge may (but does not necessarily) raise sea levels temporarily,<br />
reducing the effect <strong>of</strong> waves raised during a cyclone on the intertidal pottery scatters,<br />
281
elative to effects <strong>of</strong> lesser winds prevailing at times <strong>of</strong> low tide. Positive tidal surges <strong>of</strong> up<br />
to 4m occur during tropical cyclones (Open <strong>University</strong> Oceanography Team 1989:61), and<br />
in some cases this property <strong>of</strong> tropical storms and cyclones would be expected to reduce<br />
the effect <strong>of</strong> storm generated waves in the lagoon on submerged pottery scatters. Storm<br />
surge is caused primarily by winds driving water ashore and this will depend at a larger<br />
scale on whether wind is onshore or <strong>of</strong>fshore, but within the “radius <strong>of</strong> maximum wind”,<br />
several kilometres from the storm centre, surge response to barometric pressure anomaly<br />
is universal for cyclones (Anthes 1982:5-7, 161), although this effect is slight. Where<br />
cyclones pass to the south <strong>of</strong> the Solomon Islands, as is usually the case, and winds can be<br />
expected to veer through northerly, storm surge would be expected to be slight or absent,<br />
and was observed to be slight during such an event in 1998, but if a cyclone were to take<br />
a more northerly route, storm surge from Easterlies, Southerlies or Westerlies could be<br />
expected in the embayment comprising the Blanche Channel. In terms <strong>of</strong> cyclone events,<br />
the circumstance most likely to affect intertidal pottery scatters would be the more<br />
common northerly-tending cyclone winds with attendant storm surge only slight or absent,<br />
as pottery might be exposed to wave action in the lagoon at low tide in these<br />
circumstances. Sites with significant fetch in a northerly direction (NE through to NW)<br />
could therefore be expected to be more severely impacted by cyclones.<br />
Measuring Fetch:<br />
Fetch for cyclones was measured as the maximum at low tide in a northerly arc between<br />
NW and NE. As the direction <strong>of</strong> the NE trade is uncertain on present information, fetch<br />
for this wind is measured as a maximum between 320 and 350 degrees (consistent with<br />
personal observation) at low tide; and also at 70 degrees at low tide, consistent with the<br />
direction <strong>of</strong> the NE trade shown in the Pacific Pilot. Fetch for the SE trade is measured at<br />
low tide bearing 140 degrees, consistent with personal observation. Fetch was not<br />
measured for the summer sea breeze because it coincides with the diurnal summer-solstice<br />
daytime solar tide, and is unlikely to affect the site areas studied. Fetch in the direction <strong>of</strong><br />
the land breeze was measured as the maximum at low tide 25 degrees either side <strong>of</strong> North.<br />
Fetch was measured using air photographs at a scale <strong>of</strong> 1:33 000.<br />
282
Results:<br />
Fetch under various combinations <strong>of</strong> wind and tide are given in Table 21.<br />
Table 21: Fetch measurements for collection sites as an indicator <strong>of</strong> wave exposure.<br />
Site Cyclone<br />
(NW-NE)<br />
Paniavile 2km<br />
(moderated<br />
by seagrass<br />
and s<strong>of</strong>t<br />
sediments)<br />
Hoghoi 6km<br />
(seagrass)<br />
“NE” trade<br />
(Peza?)<br />
283<br />
SE trade<br />
(Gevasa)<br />
land breeze<br />
(low night)<br />
Maximum<br />
fetch (max<br />
exposure)<br />
0 0
pottery discard is less likely to have survived.<br />
Hoghoi had a wave exposure measurement <strong>of</strong> up to six km to the NNE, with wave<br />
effects moderated due to protection at low tide by an extensive shallow seagrass flat with<br />
numerous small coral heads nearing the surface. If sea levels were higher in the past this<br />
protection would have been absent or reduced, but the ceramics would have been in deeper<br />
water, mitigating swash effects to some extent.<br />
Gharanga could have up to six km exposure in SE tradewind conditions allowing<br />
swash to affect the site at intermediate stages <strong>of</strong> the tide (this effect may have been<br />
extended in the past if sea levels were slightly higher, with the materials in shallow water<br />
rather than exposed at low tide). Gharanga and Kopo are different to the other sites in that<br />
they are situated at the mouth <strong>of</strong> mainland streams. Sherds did not have the cushioning<br />
protection <strong>of</strong> marine growth to the same extent due to reduced growth <strong>of</strong> sponges in the<br />
low-salinity location, and could also be subject to taphonomic effects arising from stream<br />
flooding.<br />
Miho and Honiavasa both yielded maximal wave exposure <strong>of</strong> 6km but Honiavasa<br />
was more exposed to refraction waves in SE tradewind conditions (Miho is better<br />
protected from such waves than Honiavasa due to more extensive shoals adjacent to the<br />
passage and situation in a small embayment, while Honiavasa, by contrast, is located quite<br />
close to the deeper water <strong>of</strong> the passage, and the author noted refracted waves while<br />
collecting there on a rising tide in tradewind conditions. While the seaward portions <strong>of</strong> the<br />
site were in deep enough water to be protected in these conditions, shoaling towards the<br />
centre <strong>of</strong> the site left the latter area exposed to swash, while the eastern end <strong>of</strong> the site was<br />
protected by the central shoal. Sediments across the site support the idea <strong>of</strong> a central area<br />
affected by swash from refraction waves, where ceramics are exposed as large fragments,<br />
and a more sheltered easterly zone where ceramic visibility drops away as sandier<br />
sediments build up. <strong>The</strong> picture could look quite different after a strong northerly though.<br />
284
Nusa Roviana has up to 3km <strong>of</strong> fetch in a NNE direction, and can be expected to<br />
show some signs <strong>of</strong> damage or size sorting in view <strong>of</strong> this relatively exposed location (in<br />
modern times a wharf/breakwater has been constructed here to shelter the landing in<br />
northerly conditions). Sherd surfaces were generally quite damaged in this site with temper<br />
grains slightly pedestalled in many cases (the single dentate sherd was an exception,<br />
retaining a smooth finish).<br />
Unanalysed assemblages from Ririgomana, Punala, Humbi Kongu and Rango were<br />
all in poor condition with sherds showing signs <strong>of</strong> water rolling in many cases. More<br />
recently, large sherds in good condition have been reported from the deeper margins <strong>of</strong><br />
Rango (Rigeo, Pers. comm., 2002). Unlike Paniavile, Ririgomana is free <strong>of</strong> seagrass, and<br />
although the water is shallow generally in the area between Ririgomana and the mainland,<br />
I can envision a chop <strong>of</strong> small waves building up along this shore in a northerly. Punala and<br />
Humbi Kongu were similar in this respect.<br />
A number <strong>of</strong> findspots and lithic scatters on the rocky shore <strong>of</strong> Honiavasa Island<br />
to the east <strong>of</strong> Hoghoi had a lesser degree <strong>of</strong> seagrass shelter than Hoghoi, as attested also<br />
by rockier coral shorelines. <strong>The</strong> low-density scatter <strong>of</strong> sherds, especially lithic manuports,<br />
along this shoreline does not preclude relatively intensive stilt-house settlement in the past,<br />
with most ceramics removed by swash-zone rolling post-deposition. Similarly, the ceramic<br />
findspot at Kazu was on a stony shoreline, and may be a chance survival <strong>of</strong> significant SE<br />
trade exposure. Lithic manuport scatter at Mbolave, where a single sherd was found, could<br />
also be a case <strong>of</strong> a ceramic/lithic scatter transforming into a lithic scatter through time.<br />
<strong>The</strong> fetch data predicts that Honiavasa, Miho, Gharanga and Hoghoi should show<br />
the greatest level <strong>of</strong> sherd damage and size-sorting, closely followed by Nusa Roviana.<br />
When the sheltering effect <strong>of</strong> seagrass shallows is taken into account at Hoghoi, this<br />
285
seems to partly match the size-data from sites (Table 22) in that Honiavasa and Miho had<br />
larger average sherd sizes than other sites (the 25 sherds from Kopo were largest but this<br />
may be a sample size and/or collection intensity effect).<br />
Table 22: Total sherd count and average sherd area by collection site,<br />
resulting from combined effects <strong>of</strong> collection intensity and wave exposure.<br />
Site Sherd Count Average Sherd<br />
Area(cm 2 )<br />
Paniavile 644 14.65<br />
Hoghoi 861 10.94<br />
Miho 382 22.08<br />
Honiavasa 442 28.25<br />
Gharanga 277 17.96<br />
Nusa Roviana 115 15.83<br />
Kopo 25 33.24<br />
Zangana 860 13.59<br />
Chapter Summary and Conclusions:<br />
It was difficult to construct an unequivocal model <strong>of</strong> wave exposure given conflicting<br />
information about the direction <strong>of</strong> prevailing winds, and uncertainty as to the relative<br />
importance <strong>of</strong> the various winds. Personal observation and local anecdotal information<br />
conflict with the information in the 1971 Pacific Pilot, and the latter is probably only<br />
intended as a very general guide for navigational use. Aswani’s recording <strong>of</strong> Peza as a<br />
westerly may be a local effect due to topography at Baraulu Village (Canaan), where he<br />
was mostly based, and a northwesterly direction for this strong wind is preferred here.<br />
Wave exposure for some sites (e.g. Gharanga and Hoghoi) is much higher using<br />
286
Figure 75: Location <strong>of</strong> Honiavasa and Miho collection sites<br />
at the inner ends <strong>of</strong> a long passage through uplifted barrier<br />
reef. Miho is more sheltered from refraction waves than<br />
Honiavasa as it is situated in a small bay oriented to the<br />
NW, while the western edge <strong>of</strong> Honiavasa drops away into<br />
Honiavasa passage.<br />
personally observed wind directions, and any future modeling <strong>of</strong> this sort would benefit<br />
from more formal recording <strong>of</strong> wave exposure at various locations in the various prevailing<br />
conditions.<br />
This analysis provides a caution against a direct reading <strong>of</strong> aceramic scatters as<br />
having a chronological or behavioural significance, suggesting that some <strong>of</strong> these at least<br />
may be the result <strong>of</strong> a harsh taphonomic regime, which has biased the observed assemblage<br />
composition towards relatively robust coarse-grained lithic manuports.<br />
Kopo and Paniavile seemed to be the more sheltered locations, and this seems<br />
consistent with the information initially recorded regarding Paniavile, which was a rich site<br />
prior to a series <strong>of</strong> collecting and fossicking events (Reeve 1989). <strong>The</strong> sample <strong>of</strong> sherds<br />
from Kopo is too small to yield useful taphonomic analysis results, but collection intensity<br />
was fairly high (Sheppard, pers. comm. 2002) suggesting either a harsher taphonomic<br />
regime than predicted by the wave exposure model or a low intensity-duration <strong>of</strong><br />
287
occupation in the past. It seems likely that a number <strong>of</strong> such sites may be well preserved<br />
along the mainland shoreline under s<strong>of</strong>t sediments, in locations sheltered from all the major<br />
wind/fetch combinations (the sort <strong>of</strong> locations which can be expected to be characterized<br />
by a buildup <strong>of</strong> s<strong>of</strong>t sediments). <strong>The</strong> reported ceramic find-spot at Pikoro may be one such<br />
location, although limited test-pitting to the water table revealed a sparse distribution <strong>of</strong><br />
lithic manuports and some small sherd fragments from the adjacent stream banks.<br />
None <strong>of</strong> the ceramic sites had more than 6km maximum fetch, suggesting this may<br />
be an upper limit for site preservation in the water depths in which the sites were found<br />
(intertidal-subtidal). This analysis thus provides a starting point for designing a model <strong>of</strong><br />
site preservation in relation to wave exposure for this type <strong>of</strong> site. <strong>The</strong> model could be<br />
extended by adding data from other areas where similar sites are found at different depths.<br />
This sort <strong>of</strong> modeling process is fundamental to understanding the distribution <strong>of</strong> Lapita<br />
across Near Oceania. <strong>The</strong> Reef-Santa-Cruz terrestrial Lapita sites at the near margin <strong>of</strong><br />
Remote Oceania and terrestrial sites in the Talasea region may mark a shift by some early<br />
Lapita-era people onto land.<br />
288
CHAPTER 7:<br />
SITE/ASSEMBLAGE FORMATION PROCESSES<br />
Introduction:<br />
(SHERD TAPHONOMY).<br />
Determination <strong>of</strong> formation processes (or at least thorough consideration <strong>of</strong> the<br />
possibilities) is widely regarded as an essential stage in making archaeological inferences,<br />
although,<br />
“Many archaeologists simply ignore the subject, perhaps because they are<br />
unconvinced <strong>of</strong> its importance or wary <strong>of</strong> its implications for the validity<br />
<strong>of</strong> their interpretations <strong>of</strong> the past (Shott 1998).”<br />
Interest in archaeological formation processes burgeoned during the 1970s (Schiffer<br />
1995c), however, for archaeological sites located in the sea, development <strong>of</strong> formation<br />
theory and method has lagged by comparison (Stewart 1999). A key question facing the<br />
analyst is whether comparisons between samples, for example, seriation, are invariant<br />
under the transformations brought about by site formation processes (Orton 2000:66). For<br />
the Roviana intertidal sites, another fundamental behavioural question is whether<br />
deposition was in the sea in the past, or whether submergence/coastal erosion has made<br />
intertidal sites out <strong>of</strong> terrestrial sites. Although marine transgression is primarily a<br />
destructive process (Waters 1992:275), it is clear that coastal occupation sites can survive<br />
submergence relatively unscathed in ideal circumstances, these being where sites are<br />
sheltered from wave energy (Flemming 1983) or where submergence (and sometimes<br />
burial in sediments) is rapid (Stewart 1999:572, Waters 1992:277), as in the case <strong>of</strong> Port<br />
Royal (Hamilton & Woodward 1984), and sufficiently deep to evade wave action (Waters<br />
1992:250-251). Comparison <strong>of</strong> Roviana density <strong>of</strong> deposits with the quantities <strong>of</strong> sherdage<br />
289
expected from estimates <strong>of</strong> the breakage population (Felgate 2002, Felgate & Bickler n.d),<br />
as discussed in Chapter 5, suggests that Roviana sherd samples are far from pristine, and<br />
that substantial sherd attrition has occurred during formation (see Chapter 6). <strong>The</strong><br />
possibility exists that the sites originated as terrestrial deposits and have been damaged in<br />
the course <strong>of</strong> becoming intertidal deposits.<br />
I will argue below against submergence or coastal erosion <strong>of</strong> terrestrial sites as an<br />
explanation for the observed archaeological distribution. My arguments in doing so closely<br />
parallel those <strong>of</strong> Reeves, Specht and Wickler for similar sites elsewhere in Near Oceania<br />
(Reeve 1989, Specht 1991, Wickler 2001), and follows closely the argument to this effect<br />
presented previously for Roviana underwater sites (Felgate 2002). Recent paleoshoreline<br />
research is cited as additional support <strong>of</strong> this argument (Mann et al. 1998).<br />
In this <strong>chapter</strong> taphonomic approaches to the identification <strong>of</strong> formation processes<br />
are taken, seeking to determine, from the characteristics <strong>of</strong> artefact assemblages and<br />
individual artefacts, the processes that have acted on these in forming the observed sites.<br />
A series <strong>of</strong> models <strong>of</strong> cultural formation processes are presented below, followed by a<br />
series <strong>of</strong> post-depositional formation process models. A range <strong>of</strong> sherd-based taphonomic<br />
analyses contribute to discriminating between these various formation models. In Chapter<br />
11 a spatial approach to the identification <strong>of</strong> postdepositional process is presented, while<br />
in Chapter 6 wave exposure was modelled for sites, and compared with other taphonomic<br />
information in this <strong>chapter</strong>.<br />
Beyond the identification <strong>of</strong> formation processes, a second major aim in seeking<br />
to understand taphonomic effects on assemblage composition is to assess the likely nature<br />
<strong>of</strong> taphonomic bias in assemblage composition. Assemblage stylistic composition is an<br />
important data source for chronological inference, through seriation analysis (see Chapter<br />
12). While it does not seem possible to fully understand the exact nature <strong>of</strong> assemblage<br />
biases arising through taphonomic processes, a number <strong>of</strong> observations can be made that<br />
inform in this regard, and which aid in the analysis <strong>of</strong> decorative variability.<br />
290
Review: Middle Range <strong>The</strong>ory for the Identification <strong>of</strong> Formation Processes:<br />
A general review <strong>of</strong> suggested methods by which formation processes might be identified<br />
is given here. <strong>The</strong> classic review is that <strong>of</strong> Schiffer (1995c), whose categories <strong>of</strong> evidence<br />
are used here.<br />
Size effects, density, shape, orientation, structure and context <strong>of</strong> the deposit:<br />
<strong>The</strong>re is a substantial literature on the size-sorting effects <strong>of</strong> cultural transformation<br />
processes such as cleanup activities in the past. <strong>The</strong> McKellar principle, for example<br />
states,<br />
“ that small artefacts, especially microartefacts on occupation surfaces<br />
<strong>of</strong>ten indicate primary refuse, whereas in activity areas not habitually<br />
cleaned, such as some lithic workshops, abandoned structures... and vacant<br />
lots..., larger items can accumulate as primary refuse (Schiffer<br />
1995c:175).”<br />
Perhaps more important than such behavioural correlates in the present context are non-<br />
cultural size-sorting effects as governed by the laws <strong>of</strong> hydrology or hydraulic transport,<br />
viewing archaeological items as sedimentary particles (e.g. Shackley 1978). <strong>The</strong>re are<br />
many studies dealing with stream and river environments from this perspective (Waters<br />
1992:321), and also hydrological experimental sedimentological approaches such as flume<br />
experiments (e.g. Day 1980). For coastal sediments, wave processes assume a greater<br />
importance than currents, as waves entrain sediments in a zone extending from the wave<br />
base to the landward margin <strong>of</strong> the swash zone, sediments which are then transported by<br />
beach drift, longshore current, and rip currents (Waters 1992:249-251). <strong>The</strong>se<br />
erosive/transportation processes fluctuate in magnitude during changes in weather, and are<br />
displaced landward and seaward with tidal fluctuations. When such coastal high-energy<br />
processes come into contact with archaeological sediments, the fine fraction <strong>of</strong> the<br />
artifact/sediment matrix becomes suspended in the water column, while a lag deposit <strong>of</strong><br />
291
larger/heavier items may be abraded by suspended sand or reworked by rolling (Waters<br />
1992:270-271).<br />
<strong>The</strong>se processes can create a swash-zone lag deposit by eroding and transporting<br />
terrestrial sediments and artefacts, or can rework materials initially deposited in the water<br />
as primary refuse. Both terrestrial and in-water discard have similar consequences for the<br />
size, shape, density and orientation <strong>of</strong> artefacts in some respects. Artefacts in such settings<br />
should occur in a matrix <strong>of</strong> similarly large or dense sedimentary particles. Some shapes will<br />
transport more easily by benthic rolling than others, while flatter shapes might be more<br />
easily suspended in the water column by turbulence than blockier shapes. <strong>The</strong> same across-<br />
shore size/shape/density-sorting evident in the sedimentary particles should show up in the<br />
size patterning <strong>of</strong> artefacts. Armouring <strong>of</strong> graded sediments may also occur, where a<br />
surface layer <strong>of</strong> larger, less easily transported items protects less-sorted underlying<br />
sediments and artefacts.<br />
To distinguish between erosion <strong>of</strong> a terrestrial deposit into a maritime lag deposit<br />
and primary deposition in the sea, one must look not only at the lag deposit itself, but<br />
landward, in search <strong>of</strong> traces <strong>of</strong> a source <strong>of</strong> the artefacts (remembering also that storm<br />
wash can transport materials from sea to land). This is the line <strong>of</strong> reasoning used by others<br />
in suggesting reef flat lithic/ceramic scatters represent deposition in the sea in the past<br />
(Reeve 1989, Specht 1991, Wickler 2001: 241). Absence <strong>of</strong> adjacent terrestrial erosion<br />
features that might supply a source <strong>of</strong> artefacts would suggest that the materials were<br />
deposited in the sea directly by some means.<br />
In deep water, deposition could occur as a result <strong>of</strong> a shipwreck or capsize, or<br />
funerary/religious practices, or discard from anchored vessels (Stewart 1999). In the<br />
intertidal zone or on the reef flat, one would look shoreward in search <strong>of</strong> an attractant that<br />
would cause a concentration <strong>of</strong> activity in such shallow water, such as a local resource (for<br />
example water or gardening land) that might cause intensive canoe traffic and an<br />
accumulation <strong>of</strong> pottery breakage events and discard. If there is a widespread pattern<br />
292
comprising many instances <strong>of</strong> reef-flat or intertidal site locations, without attendant<br />
deposits on land, and without obvious localized attractant resources, this is more likely a<br />
pattern resulting from settlement. Artifact accumulation quantity/density and assemblage<br />
composition/diversity can allow some further idea <strong>of</strong> site function and occupation intensity<br />
in such circumstances.<br />
Artefact size/density/shape should give clues here, even if the shore deposits have<br />
been reworked by storm wash. In the case <strong>of</strong> primary deposition on land, the terrestrial<br />
artefact deposit should show less size/density sorting than the adjacent marine lag deposit,<br />
as it cannot have been reworked to the same extent as the materials in the sea. In the<br />
circumstance where the artefacts on land are rounded and small, while the artefacts in the<br />
sea are <strong>of</strong> a greater variety <strong>of</strong> sizes and in good condition, one must conclude either that<br />
some harsh taphonomic process such as gardening has reduced sherds on land, or that the<br />
small sherds on land are a wave-transported size fraction that originated in the sea. Even<br />
in the circumstance where sherds on land are not particularly well sorted by size, there is<br />
the possibility that a more random size-selection has been cast up from the sea by large<br />
waves, and that only some lucky sherds still in the sea have been sheltered from the force<br />
<strong>of</strong> the water by microtopography. In such circumstances other geomorphological<br />
characteristics <strong>of</strong> storm wash deposits aught to be identifiable (for example Gosden &<br />
Webb 1994).<br />
Processes affecting artefact orientation and dip are not particularly well understood<br />
as yet. Dip may be particularly useful for identifying primary deposition <strong>of</strong> artefacts<br />
tramped onto a flat surface, and orientation has been used to infer aeolian and fluvial<br />
processes (Schiffer 1995c:178). Marine currents and some propwash effects from treasure-<br />
hunting techniques may leave patterned artefact orientation (Stewart 1999). Strong<br />
currents may occur at times in reef-flat locations near reef passages.<br />
Use Life, Damage and Accretions:<br />
293
Schiffer notes that burials, caches, and floors <strong>of</strong> structures may contain whole pots with<br />
remaining use life (Schiffer 1995c:178). Conversely, refuse deposits will seldom contain<br />
pristine articles. Postdepositional reworking can obscure remaining use life by breaking<br />
pots, but other types <strong>of</strong> use life evidence, such as the degree to which stone adzes have<br />
been re-sharpened in comparison to unused reject preforms, are more robust. Intuitively,<br />
ovenstones are a robust item that may preserve indications <strong>of</strong> remaining use life better than<br />
pottery, which could suggest site abandonment or lack <strong>of</strong> economising behaviour.<br />
Damage such as edge-rounding, surface ablation, spalling and chemical weathering<br />
<strong>of</strong> mineral constituents can give clues as to the use and taphonomic history <strong>of</strong> an artefact<br />
or fragment (see Schiffer 1995c:178-181 for a review). Water transport can cause rapid<br />
rounding <strong>of</strong> sherd edges (Skibo & Schiffer 1987) and abrasion by rolling can be<br />
distinguished (on lithic implements at least) from edge-rounding by smaller waterborne<br />
particles (Shackley 1974). Surface ablation and pedestalling <strong>of</strong> temper grains can result<br />
from water transport <strong>of</strong> sherds, and wetting/drying cycles in a saline environment can be<br />
expected to produce surface spalling or exfoliation (Schiffer 1987). In the extreme<br />
weathering environment <strong>of</strong> the tropical Pacific, calcareous temper grains are known to<br />
sometimes have dissolved after deposition, leaving voids in the fabric <strong>of</strong> sherds (see for<br />
example Dickinson 2000b), but there is no evidence for this occurring in sherds deposited<br />
in the sea. Sherds formerly on land and subsequently deposited in the sea as the result <strong>of</strong><br />
erosion might be expected to display this effect at some level, particularly if it is thought<br />
they were eroded from acid soils. Lack <strong>of</strong> weathering <strong>of</strong> calcareous temper grains may<br />
therefore be indicative in some settings <strong>of</strong> primary deposition in the sea.<br />
It was observed through differential conservation during the project that complete<br />
drying <strong>of</strong> sherds without desalination caused rapid degradation <strong>of</strong> the recovered<br />
assemblage, whereas even rough methods <strong>of</strong> desalination showed a marked improvement<br />
in the condition <strong>of</strong> analysed sherd surfaces and reduction <strong>of</strong> sherd crumbling. <strong>The</strong> absence<br />
<strong>of</strong> salt crystallization damage on sherds might suggest they have spent most <strong>of</strong> their lives<br />
294
in either a terrestrial or an underwater context, rather than in a cyclic wet/dry salty<br />
environment.<br />
Marine accretions have the potential to inform on taphonomic processes and<br />
landscape change. Sherds found on land that have accretion <strong>of</strong> marine organisms (coral<br />
growth for example) must have been in the sea at some time. Similarly, sherds with coral<br />
growth found in shallow water above the coral growth line may be evidence for landward<br />
transport, or for an emerging coastline. <strong>The</strong> converse is not true though, sherds found on<br />
land without such accretions need not have been deposited on land through their entire<br />
history. Abrasion and weathering can remove such accretions, which do not necessarily<br />
form in the first place.<br />
Artifact Diversity:<br />
<strong>The</strong> range <strong>of</strong> types <strong>of</strong> artifact in deposits is sensitive to cultural formation processes,<br />
especially the occupation span <strong>of</strong> settlements, but also to differences in the functions <strong>of</strong><br />
settlements and activity areas (Schiffer 1995c: 182-183). Schiffer’s discussion is weak on<br />
taphonomic effects on artefact diversity, and sampling effects on diversity generally (e.g.<br />
Kintigh 1984, Shott 1989a). Lag deposits, for example, may be size or density sorted,<br />
reducing diversity by favouring chunky types, those that break into large pieces, and<br />
different materials are differentially preserved in different depositional contexts, so there<br />
is always potential for taphonomic variation in diversity. Surface or lag deposits are more<br />
subject to collector effects, which can be expected to reduce diversity rapidly unless items<br />
are highly fragmented (Felgate & Bickler n.d). While Schiffer suggests that great diversity<br />
is found in secondary refuse deposits containing refuse streams from a settlement’s entire<br />
range <strong>of</strong> activities, this assumes a well-preserved deposit. Deposits lacking such diversity<br />
may do so as a result <strong>of</strong> taphonomic processes rather than cultural formation processes.<br />
Shott notes that richness as a measure <strong>of</strong> diversity, as used by Kintigh, fails to<br />
account for the evenness or equitability <strong>of</strong> item frequencies by class (Shott 1989a). Where<br />
295
large numbers <strong>of</strong> some classes <strong>of</strong> artefact, for example ceramics, exist at a location, and<br />
small numbers <strong>of</strong> another class, for example adzes, or where count data by class is sparse<br />
(low average abundance within classes), both richness and other more sophisticated<br />
measures <strong>of</strong> diversity such as the Shannon-Weaver information statistic can be sensitive<br />
to sample size effects. Shott goes on to suggest ways <strong>of</strong> inferring use-lives <strong>of</strong> tools from<br />
relative frequencies in diversity-saturated assemblages, like Schiffer, with minimal attention<br />
to the recognition <strong>of</strong> taphonomic reduction <strong>of</strong> diversity. He seems not to expect a strong<br />
taphonomic bias in the diversity <strong>of</strong> types within a single material class, e.g. lithics. It seems<br />
that a fundamental question in making use <strong>of</strong> diversity measures to identify behavioural or<br />
organizational (e.g. use-life) properties <strong>of</strong> assemblages is the effect <strong>of</strong> taphonomic biases<br />
on the data.<br />
Diversity therefore seems to me to be a poor route into the understanding <strong>of</strong><br />
formation processes for sites in a poor state <strong>of</strong> preservation, as indicated in the Roviana<br />
case (see Chapter 5).<br />
Representation <strong>of</strong> Parts:<br />
Use <strong>of</strong> part representation in ceramic assemblage formation studies is in its infancy<br />
(Schiffer 1995c:186) and preliminary ideas in this regard suffer from a convergence with<br />
inference from part representation in faunal taphonomy. Schiffer, for example, envisages<br />
part selection for recycling by potters as the sort <strong>of</strong> process that might be identifiable<br />
through part representation, rather like the way animal scavenging <strong>of</strong>ten involves selection<br />
<strong>of</strong> skeletal parts. Given that there is no evidence for use or re-use <strong>of</strong> sherds as grog temper<br />
in the Roviana materials, and only minor possible evidence for re-use <strong>of</strong> bases as bowls or<br />
frying pans (see Chapter 8), analysis along these lines would not seem to be a productive<br />
exercise.<br />
Other mechanisms, such as unrecorded collection by archaeologists and others, can<br />
be viewed as taphonomic processes affecting the composition <strong>of</strong> assemblages <strong>of</strong> artefacts<br />
296
(Ruig 1999). While there is as yet no generally accepted theory <strong>of</strong> the effects <strong>of</strong> collecting<br />
on ceramic part representation, some expectations are developed below in relation to the<br />
Roviana samples.<br />
Archaeologists and other destructive collectors are expected to pick up larger items<br />
first, and those with more information or aesthetic value. Whole adzes, more complete<br />
sherd or whole pots, are likely to be the first things collected from sites like the Roviana<br />
intertidal scatters. Next in order <strong>of</strong> preference are smaller items with obvious information<br />
content: decorated sherds, sherds on which the vessel lip, neck or shoulder/carination are<br />
represented, smaller lithic artefacts, or larger fragments <strong>of</strong> lithic artefacts. Some<br />
archaeologists may also recognize the information value <strong>of</strong> exotic lithic manuports such<br />
as sandstone fragments or ovenstones.<br />
Non-archaeologists might selectively remove sherds or lithics for purposes other<br />
than information gathering. Anecdotal evidence suggests two processes <strong>of</strong> this sort,<br />
skipping <strong>of</strong> sherds into deeper water by children, which should affect mainly sherds<br />
lacking complex curvature for hydrodynamic reasons, and collection <strong>of</strong> sherds and lithic<br />
artefacts by villagers following collection by archaeologists, where a perception <strong>of</strong><br />
commercial value has arisen. In this instance collector effects can be expected to closely<br />
mirror the information-value model <strong>of</strong> archaeological reconnaissance collection. A series<br />
<strong>of</strong> models <strong>of</strong> sample characteristics in line with these collection theories will be<br />
constructed below.<br />
Sediments:<br />
<strong>The</strong> sediments comprising the finer fraction <strong>of</strong> the matrix <strong>of</strong> sites may include<br />
archaeosediments (Waters 1992:16-17). Textural properties, mineralogy and chemical<br />
properties <strong>of</strong> sediments may inform on human behaviour (cultural formation processes) or<br />
on postdepositional process that have resulted in the formation <strong>of</strong> sites. For the Roviana<br />
sites, postdepositional sherd attrition as evidenced in Chapter 5 suggests the potential for<br />
297
a disaggregated pottery-derived component within site sediments. This is discussed further<br />
below in relation to models <strong>of</strong> site formation.<br />
This inventory <strong>of</strong> formation theory was used to develop a series <strong>of</strong> models <strong>of</strong><br />
formation and the material correlates <strong>of</strong> these. Models for cultural and natural formation<br />
processes are listed separately below, but there is substantial overlap in the correlates<br />
identified for the models, so data will be considered together, followed by a synthesis and<br />
conclusions.<br />
Models <strong>of</strong> Cultural Formation Process:<br />
<strong>The</strong> investigation <strong>of</strong> cultural formation processes is structured in terms <strong>of</strong> the following<br />
competing models:<br />
• Ceramic/lithic scatters formed as terrestrial discard from coastal settlements,<br />
subsequently inundated by the sea. <strong>The</strong> correlates <strong>of</strong> this model are evidence for<br />
coastal erosion features, poorly sorted artefact deposits on land and possibly sherd<br />
damage consistent with a terrestrial weathering regime, such as the weathering <strong>of</strong><br />
calcite grains.<br />
• Ceramic/lithic scatters formed as secondary accumulations in the sea when<br />
people living in terrestrial settlements dumped sherds and ovenstones into the<br />
sea. <strong>The</strong> correlates <strong>of</strong> this model are evidence for structures and midden on land<br />
from former settlements, or erosional features that suggest obliteration <strong>of</strong> such<br />
evidence.<br />
• Ceramic/lithic scatters are discard from stilt settlements in the<br />
intertidal/subtidal zone. Expected correlates are minimal contemporaneous<br />
evidence for occupation on adjacent land, with any ceramic-lithic evidence on land<br />
showing evidence <strong>of</strong> size or density sorting, and associated with geomorphological<br />
298
stormwash features; we would not expect to see much in the way <strong>of</strong> stone or coral<br />
building materials scattered about if timber posts were the primary means <strong>of</strong><br />
supporting post and thatch structures.<br />
• Ceramic/lithic scatters formed as discard from now-defluved artificial islet<br />
settlement located in the intertidal. This should look similar to stilt houses, and<br />
we might also expect to see a scatter <strong>of</strong> coral stones or boulders that formerly<br />
comprised the sea-walls <strong>of</strong> such structures, although mining for building materials<br />
in later periods might remove such evidence.<br />
Some additional cultural formation models that are not mutually exclusive with the above<br />
list can be advanced:<br />
A. Site abandonment might have a correlate <strong>of</strong> remaining use life <strong>of</strong> adzes and<br />
ovenstones, where dicarded adzes show little sign <strong>of</strong> resharpening, and many large<br />
unfractured ovenstones remain on site, but mining <strong>of</strong> ovenstones and collector<br />
effect on adzes could remove such evidence.<br />
B. Whether terrestrial or over water, permanent settlement should have the dual<br />
correlates <strong>of</strong> diversity <strong>of</strong> artefact types and substantial breakage populations <strong>of</strong> a<br />
diverse range <strong>of</strong> functional types <strong>of</strong> pottery; conversely, short-term and/or low-<br />
intensity use should be evidenced by smaller accumulations and variable artefact<br />
inventories.<br />
C. Specialized site functions should result in low-diversity artifact assemblages, with<br />
preponderance <strong>of</strong> particular vessel forms. Conversely, generalized settlement<br />
should have a diversity <strong>of</strong> special-purpose forms alongside general-purpose forms.<br />
Models <strong>of</strong> Post-depositional Formation Process:<br />
299
Substantial sherd attrition through post-discard processes (as shown to have occurred in<br />
the Roviana cases in Chapter 5) raises the prospect <strong>of</strong> taphonomic sampling bias in the<br />
surviving sherd population. In addition to the cultural formation models outlined in the<br />
previous section, a series <strong>of</strong> models <strong>of</strong> sherd attrition factors will be constructed and<br />
considered in relation to the Roviana data. <strong>The</strong>se, with their correlates, are outlined below.<br />
A. <strong>The</strong> transport model, where the missing sherds were transported by wave action<br />
into the higher-energy environment <strong>of</strong> the strandline and destroyed by mechanical<br />
abrasion. Under this model sherds which were sheltered from wave energy by local<br />
microtopography or sediment armouring form either a random selection <strong>of</strong> sherds,<br />
or a selection structured by sherd transportability. Also, weak sherds should be<br />
recovered, but extremely thin or light sherds should exist principally as small<br />
fragments on the strandline. Sites with a greater wave exposure should show<br />
greater evidence for size-sorting.<br />
B. <strong>The</strong> dis-aggregation model, where poorly-fired sherds or sherds otherwise having<br />
poor fabric strength have disintegrated in place regardless <strong>of</strong> local<br />
microtopography. Under this model, s<strong>of</strong>t or weak sherds should not be present in<br />
assemblages, although porous sherds suffering salt damage (Schiffer 1987:277) on<br />
retrieval may be present. An assessment <strong>of</strong> fabric strength variability (as expressed<br />
by sherd size) in relation to temper types <strong>of</strong> the recovered sherds will be presented<br />
in this <strong>chapter</strong> to try to determine whether much strength variability has survived.<br />
If it has, this favours a transportation model, and if not, a disaggregation model is<br />
favoured. Sherd size in relation to thickness and fabric will be used as a proxy for<br />
sherd strength. No experimental strength tests on sherds were performed. (As will<br />
be argued below in relation to signs <strong>of</strong> weathering, post-depositional reduction in<br />
sherd strength does not seem to have occurred in any <strong>of</strong> the fabric groups during<br />
their time in the sea.)<br />
300
C. <strong>The</strong> collector model, where the ratio <strong>of</strong> plain to decorated diagnostic sherds is<br />
substantially different between sites. One assemblage is known to have been<br />
depleted in the past through collection by archaeologists and villagers, and this site<br />
will be used as a control in identifying the effects <strong>of</strong> sherd collection by humans.<br />
<strong>The</strong> number <strong>of</strong> sherds for which vessel part representation is high can be expected<br />
to vary with assemblage brokenness, but also with the level <strong>of</strong> collecting by humans<br />
in the past. Assemblage brokenness can be compared using body sherd size as a<br />
control, and the ratio <strong>of</strong> eye-catching diagnostic sherds to body sherds can be<br />
developed as an indicator <strong>of</strong> collecting effects in the past. One might however<br />
expect some larger body sherds to be uplifted, which might render this model<br />
equifinal with brokenness through wave exposure or trampling.<br />
D. <strong>The</strong> sherd-skipping model, where anecdotal evidence for sherd skipping by<br />
children (throwing a sherd or flat stone with a spinning action across the water<br />
surface to cause it to perform a series <strong>of</strong> bounces) is tested against assemblage<br />
composition. Assemblages that have been subject to a high degree <strong>of</strong> sherd<br />
skipping should have relatively fewer flattish medium-sized body sherds than would<br />
otherwise be the case.<br />
E. <strong>The</strong> collection intensity model: in this model sherd size varies in relation to<br />
collection intensity. <strong>The</strong> more intensively a site is collected, the more small<br />
inconspicuous or undiagnostic fragments are retained into the sample assemblage.<br />
Summary <strong>of</strong> Models:<br />
<strong>The</strong> consequences <strong>of</strong> the actions <strong>of</strong> these various processes for archaeological sequence-<br />
building objectives, as laid out at the beginning <strong>of</strong> the <strong>chapter</strong>, vary: the transport model<br />
is liable to create bias in favour <strong>of</strong> heavier, thicker pottery, with light calcareous fabrics<br />
301
possibly under-represented, and thin-walled pottery particularly likely to be under-<br />
represented. <strong>The</strong> dis-aggregation model is likely to result in the under-representation <strong>of</strong><br />
non-cooking vessels, which may be lower-fired, and/or less tough through choice <strong>of</strong><br />
temper during manufacture.<br />
This possibility is <strong>of</strong> particular concern for Lapita, since some have suggested that<br />
highly decorated flat-based unrestricted vessels might have a serving function, or that<br />
highly decorated Lapita vessels in general might have a ritual function rather than<br />
utilitarian. If this were the case a harsh taphonomic regime might turn Lapita sites into<br />
post-Lapita sites from the point <strong>of</strong> view <strong>of</strong> the ceramic chronologist. Wickler’s<br />
documentation <strong>of</strong> dentate-stamped pottery in sites <strong>of</strong> this general type on the reef flat at<br />
DAF and DES are reassuring in this regard, suggesting that there is a possibility <strong>of</strong> such<br />
sherds being represented in un-buried intertidal/shallow-water sites (Wickler 2001:Chapter<br />
5). Collector bias might render an assemblage plainer overall than it might have been, but<br />
degradation <strong>of</strong> the overall sample <strong>of</strong> diagnostic sherds rather than analytical bias might be<br />
expected, as the analyst can control for this effect by controlling for sherd<br />
“diagnosticness”. <strong>The</strong> sherd skipping model does not seem to involve any particular<br />
stylistic taphonomic bias, except if decoration is concentrated on the upper portions <strong>of</strong> the<br />
potsherd skipping could affect the overall ratio <strong>of</strong> plain:decorated sherds, but many other<br />
factors could also affect this.<br />
Qualitative Evidence From Samples:<br />
Damage:<br />
Wetting/drying salt crystallization damage was not quantified during analysis, but sherd<br />
condition and the spatial distribution <strong>of</strong> sherd condition was used to infer drying regime<br />
limits for major portions <strong>of</strong> sites in the past. Most sherds at the deeper margins <strong>of</strong> sites<br />
exhibit no sign <strong>of</strong> salt erosion damage or damage arising as a result <strong>of</strong> water transport<br />
302
(surface ablation, pedestalling <strong>of</strong> temper grains, and severe edge rounding). It was inferred<br />
that these have never been exposed to severe, sustained swash zone turbulence (rolling)<br />
or drying. Also, there was no sign <strong>of</strong> postdepositional weathering <strong>of</strong> carbonate temper<br />
grains suggesting the sherds had never been exposed to an acid environment. Some<br />
carbonate grains were partially sintered, with a rind <strong>of</strong> carbonate still lining the sides <strong>of</strong> the<br />
void around a shrunken grain, thought to be an artefact <strong>of</strong> manufacturing firing<br />
temperature, but many sherds had un-sintered carbonate grains which aught to be liable to<br />
solution if the sherds had been in an acid environment. Most sherds had slight edge<br />
rounding, consistent with abrasion by waterborne sand.<br />
Accretions:<br />
No quantitative analysis was conducted, but some general statements are made. Accretions<br />
<strong>of</strong> marine growth were almost universal, except for rolled sherds found on the strandline,<br />
and at the Gharanga stream-mouth site where salinity was probably low. Accretions<br />
included dead hard coral in some cases, live sponges and coralline/green algae in most<br />
cases. In the deeper margins <strong>of</strong> some sites dead coral growth was common on sherds<br />
(Honiavasa site for example) in spite <strong>of</strong> the lack <strong>of</strong> any evidence for active coral growth<br />
nowadays. <strong>The</strong>se accretions may be evidence for a high stand at or post deposition, which<br />
is consistent with current data on sea-level history (Mann et al. 1998). Dating <strong>of</strong> accretions<br />
might be a worthwhile future exercise for palaeo-sea-level studies.<br />
Had wave action transported sherds up the shore in significant quantity one would<br />
expect to find coral accretions on land as evidence. It appeared as though there was a<br />
significant correlation between sherd condition and accretion presence; sherds from deeper<br />
water tended to have coral and sponge accretions and were also larger and in better<br />
condition. No coral accretions were seen on land, but this may be a sample effect, as the<br />
number <strong>of</strong> sherds located on adjacent land was small (one sherd from Honiavasa, two or<br />
three small sherds from Gharanga, none from Hoghoi). It could also be an abrasion effect.<br />
Sherds were seen in the roots <strong>of</strong> a fallen tree a short distance inland from the Honiavasa<br />
site but extensive testpitting in the area did not reveal substantial pottery deposits<br />
(Sheppard, pers. com.).<br />
Artifact Diversity:<br />
Artefact diversity/inventory <strong>of</strong> material is discussed in Chapter 13.<br />
303
Sediments:<br />
In regard to the model <strong>of</strong> cultural formation <strong>of</strong> settlements located on artificial islets in the<br />
past, now reworked, some sites were characterized by a scatter <strong>of</strong> easily transportable<br />
coral blocks <strong>of</strong> the type used locally these days for building wharves and islets. Although<br />
no quantified data were collected that bear directly on this issue, it was noted during<br />
fieldwork that the gravels and coral cobbles <strong>of</strong> some site sediments could be the reworked<br />
remains <strong>of</strong> such structures (especially at Hoghoi). No specific patterned features were<br />
noted. As it is, only the speculative suggestion can be made, and this is a question that<br />
must be left to future research. Cultural formation processes, like sea levels, should not be<br />
expected to have been static over time, and the changing conformation <strong>of</strong> the shoreline<br />
pr<strong>of</strong>ile with falling sea-levels may have enabled or encouraged such cobble structures later<br />
in time, where earlier the depth <strong>of</strong> water over the current reef flat may have precluded islet<br />
construction as too labour intensive.<br />
Micro-examination <strong>of</strong> two sediment samples from Hoghoi suggest that substantial<br />
quantities <strong>of</strong> ferromagnesisn opaque minerals and clinopyroxene were present in one <strong>of</strong> the<br />
samples, and minor quantities <strong>of</strong> fine ferromagnesian minerals and occasional small lathes<br />
<strong>of</strong> pyroxene were present in the other. <strong>The</strong> first sample was collected by the author from<br />
the vicinity <strong>of</strong> the densest accumulations <strong>of</strong> pottery at the site, while the second was<br />
collected from an unknown location in the general vicinity <strong>of</strong> the site (Sheppard pers com).<br />
<strong>The</strong>se results have yet to be confirmed or quantified by thin-section petrography.<br />
Remaining Use Life:<br />
All pottery seen by the team was broken, although anecdotes <strong>of</strong> “whole” pots in deeper<br />
water were heard. Two complete adzes and a number <strong>of</strong> fragments were recovered, and<br />
an unground adze preform in the possession <strong>of</strong> a local landholder was photographed (See<br />
Chapter 10). Comparison <strong>of</strong> the recovered whole adzes and the preform suggested that<br />
both whole adzes found had been heavily re-sharpened, and were too short to have much<br />
remaining use life in their intended function. Ovenstones were systematically collected at<br />
Hoghoi. Many large fragments and some complete water-rounded cobbles were found,<br />
enough to use for cooking, suggesting remaining use life existed for these. <strong>The</strong> presence<br />
<strong>of</strong> these may have resulted from abandonment, or, more likely in view <strong>of</strong> the lack <strong>of</strong> adze<br />
remaining use life, the presence <strong>of</strong> these at Hoghoi may simply indicate a lack <strong>of</strong><br />
economising behaviour, or perhaps discard <strong>of</strong> insufficiently tough stones liable to explode.<br />
<strong>The</strong> stones may have had other, non-cooking functions such as canoe or net anchors.<br />
304
Quantitative Evidence: Size/Density Effects and the Structure and Context <strong>of</strong> the<br />
Deposit:<br />
Sherd Size and Sherd Thickness (Thickness Strength):<br />
To investigate the relationship between thickness and sherd size, while controlling for<br />
fabric variation and form variation, body sherds <strong>of</strong> the most numerous temper class<br />
(placered volcanic temper) were used. It was initially expected that strength, and thus area<br />
<strong>of</strong> recovered sherds would show a positive correlation with thickness, which proved to be<br />
the case (Figure 76).<br />
Control for gross variation in form strength was achieved by selecting out only<br />
those sherds without complex form, that is, body sherds. This means that many <strong>of</strong> the<br />
larger sherds were omitted, due to representation <strong>of</strong> more complex forms likely to confer<br />
sherd strength. Those too small to assign to vessel part were also omitted. Sherd area<br />
seems to plateau above 9mm thickness, with the greater area <strong>of</strong> the 13mm and 15mm<br />
categories possibly being a sample size effect. Thin sherds (less than 8mm thick) decline<br />
in median size roughly proportional to thickness, and sherds less than 4mm thick are rare.<br />
Sherds in the 4-8mm thickness range form the bulk <strong>of</strong> the total sample, while sherds in the<br />
8-12mm thickness range yield consistently larger median sherd sizes. Thickness greater<br />
than 12mm is rare, and seems to yield more variable sherd size, although this might be a<br />
function <strong>of</strong> small sample size. <strong>The</strong>se data seems to support the hypothesis that there may<br />
be a bias against very thin wares in the recovered sample, which are liable to exist only as<br />
small sherds fortuitously preserved rather than transported to the strandline and rolled to<br />
destruction. Large thin sherds are either overly fragile and became broken into small, easily<br />
transported thin sherds, or large thin sherds are more easily transported by water in the first<br />
place, and thus tended to be destroyed by transport to and abrasion in the swash zone.<br />
Temper Class and Sherd Size:<br />
Temper classes are detailed in Chapter 4. Without controlling for thickness, but looking<br />
at just body sherds, sherd size seems only weakly related to temper class (Figure 77),<br />
except in the cases <strong>of</strong> Groups 4 and 7, where sample size was inadequate. Group 5<br />
(Quartz-calcite) median size is slightly smaller.<br />
305
Figure 76: Effect <strong>of</strong> sherd thickness on sherd strength<br />
(controlled for temper variation by using placered volcanic<br />
tempered sherds only).<br />
Figure 77: Body sherd size for the various temper classes.<br />
306
Body Sherd Size Frequency by Site:<br />
<strong>The</strong> sherd skipping model predicts depletion in mid-sized (throw-sized,10-30cm 2 ) body<br />
sherds, but utility <strong>of</strong> this model was limited by the potential for assemblage brokenness to<br />
be equi-finally structured by other postdepositional factors. Analysis was limited to those<br />
sites which retain a reasonable corpus in their collected assemblages <strong>of</strong> body sherds larger<br />
than 30cm 2 . <strong>The</strong> dataset provided 1053 sherds comprising body part <strong>of</strong> the vessel only,<br />
from eight sites, but insufficient larger sherds remained to test the sherd skipping model<br />
using this method. Body sherds in the 20cm 2 size class predominate in the Miho site<br />
(Figure 78), located at a modern village from where anecdotes <strong>of</strong> sherd skipping were<br />
heard, and at Honiavasa, across a narrow channel from the same village, so either there is<br />
something wrong with my theory <strong>of</strong> the correlates <strong>of</strong> this behaviour, or other processes<br />
have structured the assemblages to a greater degree.<br />
Differences between sites in these histograms are largely attributed to collection<br />
intensity: for some sites sherds in the smallest size category (4cm 2 ) are numerous, while<br />
absent in others, a pattern which dominates variability between assemblages. Low-intensity<br />
collections like Honiavasa, Gharanga and Miho have few sherds in the 4mm size class,<br />
while these predominate in the intensively collected sites Zangana, Paniavile and Hoghoi<br />
This is clearly not the whole story though, as despite low collection intensity Honiavasa<br />
yielded many more large (>56cm 2 ) sherds than either Hoghoi, Miho, Gharanga or Zangana.<br />
Representation <strong>of</strong> Parts in the Site Samples:<br />
<strong>The</strong> collector model predicts that assemblages previously collected by archaeologists or<br />
other interested people will be depleted <strong>of</strong> conspicuous large or highly decorated sherds,<br />
particularly those with information pertaining to upper vessel form (lip, rim, neck and<br />
shoulder). Lip/Rim-only sherds and body-only sherds were counted by site. <strong>The</strong>se counts<br />
together with the ratios between them may provide an enumerator <strong>of</strong> previous collection<br />
intensity, operating on the assumption that body sherds have a greater probability <strong>of</strong> being<br />
ignored by a collector (Table 23). Only sherds greater than 15cm 2 were selected, to<br />
control to some extent for assemblage brokenness.<br />
307
Figure 78: Histogram <strong>of</strong> body sherd size classes by site.<br />
308
Table 23: Ratio <strong>of</strong> lip-rim sherds to body sherds as a measure <strong>of</strong> collector effect.<br />
SITE Total >15cm 2 Lip+Rim<br />
count >15cm 2<br />
309<br />
Body count<br />
>15cm 2<br />
Paniavile 184 17 71 0.239<br />
Hoghoi 155 31 66 0.47<br />
Miho 205 39 58 0.672<br />
Honiavasa 319 75 70 1.071<br />
Gharanga 135 20 41 0.479<br />
Nusa Roviana 40 5 9 0.555<br />
Kopo 22 3 10 0.3<br />
Zangana 222 37 56 0.661<br />
ratio <strong>of</strong> LR:B<br />
It seems the ratio <strong>of</strong> lip-rim sherds larger than 15cm 2 to body sherds larger than 10cm 2 may<br />
well be a good measure <strong>of</strong> collector effect, as the most collected site (prior to the present<br />
study) was Paniavile (previously yielding 84 rim sherds, 41 “decorated body sherds” and<br />
an uncounted number <strong>of</strong> plain body sherds in an initial collection event reported by Reeve<br />
1989:50).<br />
What then <strong>of</strong> the other sites? Nusa Roviana and Kopo yielded small samples <strong>of</strong><br />
sherds, and their “collector” ratios are best ignored. Hoghoi and Gharanga both have ratios<br />
nearing 0.5, which provide a weak suggestion that some <strong>of</strong> the more complete sherds may<br />
have been removed from these in the past, prior to initial collection in the course <strong>of</strong> the<br />
present study. Miho and Zangana both yield ratios approaching 0.7, suggesting this is a<br />
reasonable figure for newly-discovered sites collected in a moderately intense fashion (the<br />
bulk <strong>of</strong> these collections being obtained in single days by two and five collectors<br />
respectively). Honiavasa, collected by a lone archaeologist on a rising tide over about four<br />
hours, looks most like a first-hit, low intensity collection <strong>of</strong> large complex-shape sherds,<br />
analogous, perhaps, to the original Reeve/Spriggs Paniavile surface collection in vessel
part representation. <strong>The</strong> major difference between the Paniavile ratio (0.24) and that <strong>of</strong><br />
the most similar sample (Hoghoi, 0.47), reassures to some extent that the major collections<br />
other than Paniavile are not severely biased by previous unreported collection, and alerts<br />
us to the degraded information content <strong>of</strong> the Paniavile sample as used in the current<br />
research.<br />
Frequency <strong>of</strong> Decoration:<br />
This is potentially another way <strong>of</strong> measuring collector effects. While variation in<br />
production in the past might account for different frequency <strong>of</strong> decoration in assemblages,<br />
collector effects could create the same result, so this is something to be aware <strong>of</strong> in<br />
analyzing surface collections. It might be simplest to avoid using frequency <strong>of</strong> decoration<br />
in seriations etc., but as others do this (see Chapters 1&2) it was thought to be worth a<br />
short detour to investigate this topic. For each vessel part, the number <strong>of</strong> plain instances<br />
and the number <strong>of</strong> instances <strong>of</strong> decoration were counted. <strong>The</strong> ratio <strong>of</strong> undecorated to<br />
decorated was calculated (Table 24).<br />
For all vessel parts except shoulders the previously collected Paniavile site scored<br />
within the range <strong>of</strong> the other sites. <strong>The</strong>se data are interpreted as showing that the ratio <strong>of</strong><br />
decorated to undecorated sherds can remain stable through collection events. It should be<br />
Table 24: Ratios <strong>of</strong> plain to decorated sherds, controlled by vessel part (lips, rims,<br />
necks and shoulders.<br />
SITE Plain L:Dec L Plain R:Dec R Plain N:Dec N Plain S:Dec S<br />
1 (Paniavile) 20:59 0.34 110:36 3.1 98:31 3.2 56:26 2.2<br />
2 (Hoghoi) 34:51 0.64 113:25 4.5 134:18 7.4 121:22 5.5<br />
3 (Miho) 14:52 0.24 80:35 2.3 77:30 2.6 84:14 6.0<br />
4 (Honiavasa) 23:80 0.29 154:17 9.1 153:20 7.7 125:23 5.4<br />
5 (Gharanga) 10:24 0.42 36:15 2.4 62:12 5.7 49:21 2.3<br />
8 (Zangana) 36:44 0.82 95:39 2.4 137:36 3.8 114:30 3.8<br />
310
noted in this regard that initial surface collection <strong>of</strong> a full cover sample <strong>of</strong> sherds from<br />
Zangana targeted only decorated sherds, with subsequent 1/5 and full cover collections<br />
targeting all sherds and large sherds respectively. Zangana may thus be slightly enriched<br />
decoratively through the initial collection strategy, but this effect is not large.<br />
Other patterns worthy <strong>of</strong> note in Table 24 are that the Honiavasa site gains a marginal<br />
score in two out <strong>of</strong> four vessel part categories (rim and neck), while lips are more likely<br />
to be decorated than in other sites and rims and necks are more likely to be plain, showing<br />
that overall ratio <strong>of</strong> decorated to plain sherds, without control <strong>of</strong> location on the vessel,<br />
can be something <strong>of</strong> a lottery rather than a salient measure <strong>of</strong> assemblage decorative<br />
variability. Miho, also, for example, has the most decorated lip, rim and neck scores, but<br />
the plainest shoulders. Breakdowns <strong>of</strong> the specifics <strong>of</strong> decoration contributing to these<br />
broad categories <strong>of</strong> data are given in detail in Chapter 9.<br />
Another index <strong>of</strong> sherd skipping was considered, the ratio <strong>of</strong> neck sherds (not a<br />
good skipping shape from personal experience) to body sherds, the latter <strong>of</strong> size range<br />
>12cm 2 ,
epresented would be too sensitive to assemblage brokenness. This index is problematic<br />
as it may relate to overall assemblage vessel form, and/or to collection bias rather than to<br />
postdepositional sherd skipping. <strong>The</strong>se data are presented in Table 25.<br />
<strong>The</strong> weakest <strong>of</strong> these “Skipping indices” is for the Paniavile site, heavily collected<br />
prior to the present study, suggesting again that previous collection may have been biased<br />
against body sherds. <strong>The</strong> lowest (strongest) index was that <strong>of</strong> Honiavasa, despite a<br />
tendency for sherds to break at the neck in this site, probably as a result <strong>of</strong> slab<br />
construction. Zangana is not far separated, suggesting Zangana, at least, might have been<br />
subject to selective removal <strong>of</strong> mid-sized body sherds, particularly since, the intensity <strong>of</strong><br />
collection at Zangana was quite high. An alternative, and more likely possibility is that the<br />
high neck count at Zangana relates to an overall pattern in vessel form where the necked<br />
vessel is more common in production, or more robust and more commonly preserved.<br />
Miho is next strongest in this index, and as Miho is located at the anecdotal skipping<br />
village, and Honiavasa within a short distance, there may be something to it.<br />
Sea Level History:<br />
Radiocarbon ages <strong>of</strong> uplifted corals, combined with an assumed high stand <strong>of</strong> +1m at<br />
5500BP generate mean uplift rates for the New Georgia tectonic block <strong>of</strong> between<br />
0.2mm/yr and 0.9mm/yr, with the three data locations in Roviana Lagoon area suggesting<br />
mean tectonic uplift rates <strong>of</strong> 0.7mm per year (Rereghana Island and Araroso passage) to<br />
0.9mm per year (Nusa Roviana Island) (Mann et al. 1998: Figure 8, Table 1, Table 2.). If<br />
sea levels have been falling an average <strong>of</strong> 0.2mm per year since the late-Holocene high<br />
stand, and the land has being going up at an average 0.7 to 0.9 mm per year in the Roviana<br />
area, then the combined effect over the 2,800 years may be somewhere in the vicinity <strong>of</strong><br />
an effective fall in sea level <strong>of</strong> between 2.5 and 3.0 metres. It is unlikely that anything in<br />
312
nature happens this smoothly, but there is currently no local data for the period between<br />
about 6000BP and 200BP, leaving little alternative but to extrapolate. Data from Ruvi Bay<br />
on Kolombangara is reassuring in this respect, as coral dated there to 2250BP ±60 is one<br />
metre above sea level, and as uplift rates there seem substantially lower (0.1 mm/yr,<br />
calculated from a total Holocene emergence <strong>of</strong> 1.5m) suggesting that if a correction is<br />
applied for the greater uplift at Roviana Lagoon <strong>of</strong> 0.6-0.8mm per year, we could add<br />
1.35-1.8 metres on to this (if the average uplift rates are useful) and that Roviana sea level<br />
(relative) might have been about 2.35-2.8m higher up the shoreline at around 2250BP.<br />
More data from the Roviana area is needed before this figure can be accepted with any<br />
certainty. All paleo-shoreline evidence, however, points to an emergent coastline, with the<br />
artefact scatters in deeper water (possibly buried in s<strong>of</strong>t lagoonal sediments) until fairly<br />
recently. This is consistent with cultural formation processes for the artefact scatters as<br />
discard from anchored vessels, fishing platforms, or more substantial stilt houses. This<br />
conclusion would need to be revised, though, should the gradual sea level model used to<br />
extrapolate between insufficient data points turn out to be incorrect.<br />
Chapter Summary and Conclusions:<br />
This <strong>chapter</strong> was written from the perspective suggested by the results <strong>of</strong> the analysis in<br />
Chapter 5 that most <strong>of</strong> the sherdage discarded in the past has not made it into our sample,<br />
and <strong>of</strong> the missing sherds, most are no longer extant. It has not been possible to provide<br />
a definitive account <strong>of</strong> where the missing pottery went, but a brief synthesis <strong>of</strong> the<br />
analytical results in hand is given below, to clarify which <strong>of</strong> the models set up at the outset<br />
<strong>of</strong> this <strong>chapter</strong> are consistent with the properties <strong>of</strong> the collections.<br />
Cultural Formation Processes:<br />
<strong>The</strong> absence <strong>of</strong> adjacent settlement evidence on land for all intertidal sites argues against<br />
313
models <strong>of</strong> terrestrial settlement. <strong>The</strong> good state <strong>of</strong> preservation <strong>of</strong> Acropora corals on the<br />
adjacent shore platform, and the evidence from sea level studies (Mann et al. 1998)<br />
suggest that this feature is an emerged reef flat which would have been the upper limit <strong>of</strong><br />
coral growth at the time the sites were formed. Any ceramic deposits on this flat would be<br />
highly visible, and would have been deposited into the sea in the Lapita period. <strong>The</strong>se<br />
factors strongly support the conclusion <strong>of</strong> a maritime rather than terrestrial origin.<br />
Additional evidence against terrestrial models is the complete absence <strong>of</strong><br />
weathering <strong>of</strong> carbonate grains, which sometimes occurs in terrestrial deposits in the<br />
tropical Pacific. Reefal carbonate detritus used as tempering material is invariably in<br />
pristine condition except where slight sintering has occurred during firing.<br />
Accretions provide an additional suggestion <strong>of</strong> a high stand <strong>of</strong> sea level at-or-post-<br />
deposition. Lack <strong>of</strong> salt drying damage in the deeper margins <strong>of</strong> sites, and severe damage<br />
<strong>of</strong> this sort to many sherds post-recovery, suggest that gradual submergence <strong>of</strong> terrestrial<br />
sites is unlikely to account for the presence <strong>of</strong> the scatters in the sea.<br />
Some sites could conceivably have been located on piled coral wharves or artificial<br />
islets, now eroded or otherwise reworked. <strong>The</strong> shoreline and intertidal at Hoghoi has many<br />
coral blocks <strong>of</strong> the size used in this sort <strong>of</strong> structure today, but these are common along<br />
extensive stretches <strong>of</strong> the raised-reef barrier island where Hoghoi is located, even where<br />
there is no pottery. Better-preserved buried sites from the same general period in the<br />
Bismarck archipelago (Gosden 1989, 1991b, Gosden et al. 1989, Gosden & Webb 1994,<br />
Kirch 2001), seemed to have post structures, and there is no strong evidence yet for any<br />
different pattern at Roviana Lagoon.<br />
Postdepositional processes are likely to have removed any ceramic evidence for site<br />
abandonment. While there was some remaining use life in “ovenstones” collected from the<br />
Hoghoi site, the two complete adzes recovered were heavily re-sharpened if the<br />
proportions <strong>of</strong> an unused reject preform in a local collection were anything to go by. <strong>The</strong><br />
ovenstones could indicate lack <strong>of</strong> economizing behaviour rather than abandonment.<br />
314
Post-deposition Taphonomic/Formation Processes:<br />
In the disaggregation model outlined above, it was predicted that sherd strength variability<br />
would be low. Analysis <strong>of</strong> sherd size in relation to temper suggests that strength variability<br />
between temper types as reflected by sherd size is low for the assemblages. Whether this<br />
is due to congruence in technical competence using differing materials, or whether the<br />
weaker sherds have already gone is uncertain. Favouring the latter view was the presence<br />
among the stronger sherds <strong>of</strong> infrequent s<strong>of</strong>t sherds or (very rarely) partially dissolved<br />
sherds with solution scarring, suggesting that disaggregation has been a factor in sherd<br />
attrition. Additional support for a disaggregation model was found when sediments from<br />
the Hoghoi site were examined microscopically, and found on examination under low-<br />
power magnification to contain abundant volcanic mineral grains exotic to the calcite<br />
raised coral barrier island local geology. <strong>The</strong>se grains are thought to be temper sands<br />
deriving from disaggregated pottery, but this conclusion has yet to be confirmed<br />
petrographically.<br />
<strong>The</strong> data on sherd thickness indicates that thinner sherds are weaker (break more<br />
easily) which makes the surviving fragments more easily transported by wave action, and<br />
possibly also that large thin sherds have been transported to the strandline and destroyed<br />
(weak fabrics and large thin shapes are selected against by a transport model). This model<br />
was consistent with the frequency distribution by thickness in which small thin sherds were<br />
rare. Large, thick sherds were rare also, which could relate to behavioural factors such as<br />
use life, but may also have resulted from inadequate firing <strong>of</strong> thicker sherds, and thus less<br />
resistance to breakage/transport and to disaggregation.<br />
<strong>The</strong>re is some support in the part representation data for the collector model and<br />
the collection intensity model. <strong>The</strong> collector model is supported by the ratio <strong>of</strong> lip/rim<br />
sherds to body sherds, which showed Paniavile to have been more heavily collected in the<br />
past than other sites (corresponding with events reported in Reeve 1989). Body sherd size<br />
histograms showed correlation with collection intensity for assemblages in the present<br />
study.<br />
315
Sherd skipping behaviour has not left any obvious hole in histograms <strong>of</strong> body sherd<br />
size-frequency, although there was some support for the model in the ratio <strong>of</strong> body sherds<br />
to neck sherds, but underwater archaeology might be a better way to ascertain the extent<br />
<strong>of</strong> recent sherd skipping or sherd throwing into deeper water.<br />
Interpretation:<br />
It seems likely that different processes have predominated at different times. During the<br />
period <strong>of</strong> occupation, stilt houses would have been constructed in water possibly too deep<br />
to stand in at low tide, with sharpened posts driven into lagoonal mud until they reached<br />
the coral underneath. At this stage there was probably a zone <strong>of</strong> shallow-water Acropora<br />
corals to landward <strong>of</strong> some sites, backed by a low cliffed shoreline (Hoghoi, Miho,<br />
Zangana, and Paniavile at least, although this is not immediately apparent at Zangana today<br />
due to landscape alterations during WWII, and possibly Nusa Roviana also), making canoe<br />
landing and access difficult. Early in the history <strong>of</strong> stilt settlements, the pottery is likely to<br />
have been semi-buried in s<strong>of</strong>t lagoon sediments, with disaggregation due to dissolution,<br />
and turbation by plants, grazing animals and burrowing animals and<br />
construction/maintenance being major factors contributing large quantities <strong>of</strong> exotic sands<br />
to the local sediments. Falling sea levels post-abandonment are likely to have brought the<br />
ceramics into the lower margins <strong>of</strong> the swash zone fairly recently, removing organics and<br />
fine sediments to a varying extent depending on the site location, and leaving lag deposits<br />
<strong>of</strong> stronger sherds. Wave transport and anthropogenic processes such as sherd skipping<br />
may have caused gradual or intermittent sherd attrition to a greater extent than other<br />
mechanisms in more recent years. It seems from the Rereghana oyster evidence presented<br />
by Mann et al. that the current sea levels are very recent, and have been at their present<br />
level no more than 400 years, so these latter processes are unlikely to have affected the<br />
sites over the millennia. In the modern era, particularly over the last three decades, interest<br />
in the sites by archaeologists and others has had a major impact on one site (Paniavile),<br />
selectively removing large, decorated, and rim sherds initially, with more intensive<br />
316
collection <strong>of</strong> a number <strong>of</strong> sites during the present research.<br />
Some <strong>of</strong> the most significant taphonomic biases may have occurred during the early<br />
disaggregation phase <strong>of</strong> sherd attrition, where particular functional variants or production<br />
groups may have disappeared almost entirely, due to low firing or incomplete firing <strong>of</strong> the<br />
vessel wall interior. <strong>The</strong>se types <strong>of</strong> biases would have to be taken into account in<br />
comparing intertidal sites with terrestrial sites, as overall effect on assemblage composition<br />
could be large. Evidence will be presented elsewhere (Chapters 8 and 9) that at least one<br />
production style had a fragile tall, thin excurvate lip, and that as a result the style is known<br />
largely through rim/neck/shoulder sherds rather than lip/rim/neck/shoulder sherds.<br />
Transport processes are likely to have biassed preservation towards chunkier types which<br />
broke into large pieces, like Honiavasa carinated sherds, and biases between intertidal sites<br />
can be expected due to slight differences in wave exposure.<br />
Of the taphonomic models constructed in the introduction, it is the collector and<br />
collection intensity models that are most clearly supported by the size data and the part<br />
representation data in this <strong>chapter</strong>. Collector processes seem to have impacted severely on<br />
the information content <strong>of</strong> sites, leaving the Paniavile sample stylistically depauperate in<br />
information on vessel form and decoration.<br />
A methodological conclusion arising from this result is that reconnaissance survey<br />
and recording <strong>of</strong> surface intertidal sites in the future could be formulated in the light <strong>of</strong><br />
these findings, i.e. that the effects <strong>of</strong> casual unprovenanced surface collections on the<br />
information content <strong>of</strong> intertidal surface scatters is drastic. Where possible, in-situ non-<br />
destructive recording is advocated unless detailed spatial recording (point provenancing<br />
<strong>of</strong> all items, including accurate heights in relation to a datum) and field materials<br />
conservation can be carried out. One potential problem with in-situ recording is accretions<br />
<strong>of</strong> marine organisms, which occasionally can hide finer decoration until cleaned in the<br />
laboratory. Partial cleaning in the field might be sufficient. Also, temper studies would<br />
require that snapped samples <strong>of</strong> some or all sherds would need to be collected for analysis.<br />
317
318
CHAPTER 8:<br />
VESSEL FORM VARIABILITY AND VESSEL<br />
Introduction:<br />
FUNCTION<br />
<strong>The</strong> moderate degree <strong>of</strong> vessel brokenness <strong>of</strong> the Roviana assemblages (Chapter 5), in<br />
which many sherds are identifiable to one or more vessel parts, permits an analysis <strong>of</strong><br />
assemblage vessel form characteristics, which was carried out using a metric vessel contour<br />
approach (introduced in Chapter 1). <strong>The</strong> classification in this <strong>chapter</strong> is materialist in that<br />
it is a measurement device created by the analyst from observed variability, for the<br />
purposes <strong>of</strong> controlling to some extent for functional variation in the seriation analysis in<br />
Chapter 12.<br />
What are the sources <strong>of</strong> form variability? Are there constraints imposed by the<br />
material and method <strong>of</strong> construction? What aspects <strong>of</strong> observed form variability are related<br />
to vessel function? Is temporal variation in form constrained or directed by function, or<br />
might it fluctuate in response to some natural or social environmental condition? Which<br />
aspects <strong>of</strong> form variability might be seen as stylistic, or free to drift in a stochastic manner<br />
(Dunnell 1978, 2001)? This last question is key for the chronological aims <strong>of</strong> the analysis,<br />
while the other questions are the reverse side <strong>of</strong> the same coin. By seeking general-purpose<br />
run-<strong>of</strong> the mill vessel forms for use in seriation, some gross functional variation is removed<br />
from the mix, reducing the possibility <strong>of</strong> incorrectly seriating functional variation.<br />
Some characteristics <strong>of</strong> Roviana vessel forms are inferred from negative evidence.<br />
For example, the lack <strong>of</strong> any flat base sherds will be discussed and interpreted in this<br />
319
<strong>chapter</strong>. <strong>The</strong> main positive evidence for vessel forms is upper vessel contour variation,<br />
especially the measurements <strong>of</strong> rim angle, rim depth, rim Vcurve, neck angle, neck Vcurve,<br />
shoulder depth/Vcurve and carination angle/Vcurve (see Chapter 4 for definitions). Body<br />
sherds provided an additional line <strong>of</strong> evidence which was largely experimental, but which<br />
showed up some interesting differences between site assemblages. <strong>The</strong> precision with<br />
which these body form data are replicable has not been assessed, and there are difficulties<br />
in associating body forms with rim/neck/shoulder forms in assemblages comprising a<br />
mixture <strong>of</strong> production styles, so primary emphasis is laid on upper vessel contour for<br />
classification <strong>of</strong> vessel forms.<br />
In accordance with the theoretical discussion in Chapter 1, functional classes are<br />
kept broad, and multi-functional or general purpose forms (e.g. cooking/serving/liquid<br />
transport) are expected to dominate assemblages, on the basis <strong>of</strong> use-life theory.<br />
Subordinate (long-use-life) functional classes, for example large storage jars or ritual<br />
vessels, were also expected to be present in lower frequencies in the assemblages.<br />
Substantial form variation was expected within such broad functional classes.<br />
Sample size comes to the fore when dealing with long-use life or low-frequency<br />
functional classes, and there is no attempt here to characterize these in any detail. Analysis<br />
focuses on classification <strong>of</strong> and variability within the largest general purpose functional<br />
class. Comparison <strong>of</strong> site sample characteristics is also undertaken, using a measure <strong>of</strong><br />
sherd restriction factor, the ratio between total sherd area for a site, and rim EVE for<br />
restricted vessels. Covariation between thickness, vessel size and vessel form is examined,<br />
as it relates to the functional structure <strong>of</strong> assemblages and the characterization <strong>of</strong> vessel<br />
function. Sherd strength was not tested, but some qualitative remarks on this subject<br />
relating to assemblage functional structure are included.<br />
Methods for Inferring Vessel Form from Sherds:<br />
Rim forms alone were not regarded as a reliable guide to vessel function (this is a<br />
320
departure from the analyses <strong>of</strong> others in the general region, e.g. Summerhayes and Wickler,<br />
who in some instances regard rim forms as proxies for vessel forms (see Chapter 2). If a<br />
fundamental distinction is made between jars and bowls, it is essential that only those lip<br />
sherds which are sufficiently large to discriminate between these categories <strong>of</strong> vessel<br />
contour are taken as evidence. For very broken assemblages, neck sherds by contrast can<br />
be used to demonstrate the presence <strong>of</strong> restricted jars, as this contour is diagnostic <strong>of</strong><br />
restriction if the sherd is large enough to orient using the neck horizon. Similarly,<br />
carination sherds and sherds from the junction between flat bases and vessel sides can be<br />
used to infer carinated and flat-based vessel forms respectively. Neck/shoulder sherds can<br />
be used also to infer generally carinated form, although the form <strong>of</strong> the carination itself<br />
cannot be specified. One would expect to find carination sherds in the assemblage too in<br />
such a case. Small lip sherds cannot be inferred to represent bowls, unless independent,<br />
convincing evidence is presented to demonstrate a correlation between the particular<br />
morphology <strong>of</strong> lips and overall vessel morphology.<br />
Figure 79: Initial classification <strong>of</strong> vessel forms.<br />
321
Function and Temper:<br />
Much <strong>of</strong> the Roviana pottery is relatively strong, compared with some samples <strong>of</strong><br />
terrestrial origin (For example Reef-Santa-Cruz materials or modern Mailu pottery in<br />
storage at <strong>Auckland</strong> <strong>University</strong>). This is reflected in sherd size in some sites. A lot <strong>of</strong> the<br />
tall thick everted rims from Honiavasa for example are usually very strong, highly<br />
tempered, difficult to snap with pliers, and are quite elastic (can be dropped on the<br />
linoleum-covered concrete <strong>of</strong> the laboratory and will bounce). This seems to be evidence<br />
for either unusually strong manufacture for some production units, or preservational bias<br />
towards strong fabric (see Chapter 7). If the former were the case this would tend to<br />
support an understanding <strong>of</strong> the assemblages as predominantly cooking and/or general<br />
purpose vessels although this may relate to drying technology during pre-firing stages <strong>of</strong><br />
manufacture (drying in a torrid-zone tropical environment may require rapid fire-drying or<br />
drying in direct sunlight, necessitating increased temper concentration to reduce shrinkage-<br />
cracking (see Chapter 1). Alternately, terrestrial deposition might lead to a substantial<br />
weakening <strong>of</strong> fabric through chemical weathering, an effect not present in the sea, or<br />
perhaps immersion strengthens some sherds. Many sherds from the submerged Mulifanua<br />
site in Samoa were,<br />
“...dense and largely unaffected by age or wear. Under the petrographic<br />
microscope it was apparent that zeolites had filled in some pores. This is<br />
a phenomenon caused by sea water immersion....(Petchey 1995:57).”<br />
Some white mineral veins in some <strong>of</strong> the Roviana pottery may be a similar phenomenon.<br />
Form: Initial Nominal Classification:<br />
<strong>The</strong> following classifications <strong>of</strong> vessel form are combinations <strong>of</strong> the major nominal form<br />
attributes as described in Chapter 4. <strong>The</strong>se are preliminary classifications that are<br />
recombined into functional classes following a metric analysis <strong>of</strong> form variability within<br />
322
classes. Some are mutually exclusive, others (e.g. Form 1 and Form 6a) are<br />
complementary, in that similar vessel forms are being inferred from independent lines <strong>of</strong><br />
evidence. Some <strong>of</strong> the initial descriptive categories turn out, on further analysis <strong>of</strong><br />
variability later in the <strong>chapter</strong>, to be quite arbitrary divisions <strong>of</strong> a continuum <strong>of</strong> variation,<br />
and are lumped in a functional classification as a consequence.<br />
Form 1: Conical-to-Hyperboloid Shoulder Form and Carination:<br />
Hyperboloid form is similar to conical form but concave in vertical pr<strong>of</strong>ile (Rice 1987:219)<br />
with a corner point (Rice 1987: 218) or carination. Classification criteria in the table <strong>of</strong><br />
form data (see Chapter 4 and attached CD-ROM table “Form.db”) were “carination type”<br />
with a value “H”and “shoulder type” with a value “L”. <strong>The</strong>se forms were mainly found in<br />
the Honiavasa site (see also Figure 7, Figure 8, Figure 9 ). <strong>The</strong>y are interpreted as<br />
representing carinated restricted jars with a conical/hyperboloid shoulder, everted rims and<br />
round bases (round bases by negative evidence:- absence <strong>of</strong> flat bases in assemblages)<br />
(Figure 79). One sherd from Hoghoi had this shoulder form, and was probably broken at<br />
the carination (HG.99.061). <strong>The</strong>re was also one hard carination sherd in the Punala sample,<br />
which has not been analysed in detail. Most, though, were from Honiavasa.<br />
Form 2: Unrestricted /Weakly Restricted Rimmed Jars or Deep Bowls:<br />
This class was subdivided as follows:<br />
Form 2a:<br />
Form 2a (Figure 79, Figure 80, Figure 81) is an unrestricted jar or bowl form with a<br />
slightly excurvate rim and a thickening <strong>of</strong> the upper body giving the external appearance<br />
<strong>of</strong> a carination, without any obvious interior join, probably originating through a two-piece<br />
construction method where the excurvate rim is joined to a rounded body/base.<br />
(Classification criteria are “rim type” has value “K”, “shoulder type” does not have value<br />
“L”). This is a small group, defined mainly on their construction method and usually<br />
323
weakly restricted or unrestricted form, not easily differentiated from compound rimmed<br />
vessels. Because there were only a few <strong>of</strong> these vessels, and because decoration was<br />
confined to fingernail impression or excision <strong>of</strong> the outer rim (an attribute omitted from<br />
seriations), these sherds had no effect on the seriations.<br />
Form 2b:<br />
Form 2b (Figure 79, Figure 80) is represented by a single large quartz-calcite tempered<br />
sherd, and is thought to indicate a shallow unrestricted ellipsoidal base topped by a corner<br />
point (hard carination) and a narrow conical restriction. At the top <strong>of</strong> the conical restriction<br />
there is another corner point in the pr<strong>of</strong>ile, with a tall excurvate (hyperboloid) attached rim.<br />
Apart from the distinct corner points and the shallow base/body or lower rim, this sherd<br />
bears some resemblance to Form 2a, but was given a separate class as the exotic quartz-<br />
calcite temper in its fabric suggests a different manufacturing centre to the Form 2a sherds<br />
(suggesting the differences between this sherd and the rest <strong>of</strong> Form 2a may turn out to be<br />
more than just “noise” despite occurrence in the sample as a single sherd). Classification<br />
criteria are “rim type” with a value <strong>of</strong> “K” and shoulder type “L”.<br />
This unrestricted vessel form is sufficiently deep, rounded and well-tempered to<br />
have had similar functions to more restricted vessel forms, but equally, due to a shallow<br />
base and unrestricted access to the contents, could have fulfilled a variety <strong>of</strong> serving, non-<br />
fluid cooking, drying, washing and other functions, and would have been a fairly stable<br />
form not liable to complete capsize due to the shallowness <strong>of</strong> the base. Because it was<br />
undecorated, it was automatically excluded from all seriation analyses.<br />
Form 2c:<br />
<strong>The</strong>se were everted-rim to vertical-rim unrestricted jars or weakly restricted jars defined<br />
reasonably securely from rim/neck attributes (especially neck angle), and forming the least-<br />
restricted end <strong>of</strong> a continuum <strong>of</strong> rim/neck/shoulder variation (Figure 79, Figure 80),<br />
which included everted rim restricted jars to straight or incurvate restricted/unrestricted<br />
jars. Classification criteria were “rim type” with a value <strong>of</strong> “X”, “S” or “V” and “neck<br />
type” with a value <strong>of</strong> “T”.<br />
324
Figure 80: Examples <strong>of</strong> unrestricted Form 2 variants: Form 2a (top); Form 2b (upper<br />
middle); Form 2c (lower middle); and Form 2d (bottom).<br />
325
Figure 81: Additional Form 2a Sherds (top 6); the two decorated gambrelled vessels<br />
from Nusa Roviana are a short-rim variant <strong>of</strong> Form 2a; the lower sherd is transitional<br />
between Form 2a and Form 2d, being unrestricted.<br />
326
<strong>The</strong>se sherds do not display any morphological evidence <strong>of</strong> multi piece construction, and<br />
are inferred to be modeled from a single lump <strong>of</strong> clay by paddle and anvil techniques (or<br />
perhaps paddle and finger in some cases). Anvil impressions are common on the vessel<br />
interior below the neck, but are usually absent from the interior <strong>of</strong> the rim, presumably as<br />
a result <strong>of</strong> wiping <strong>of</strong> the rim or forming with a large anvil/tool <strong>of</strong> some sort.<br />
Form 2d:<br />
This class comprised unrestricted bowls and/or jars (Figure 79, Figure 80, Figure 81).<br />
<strong>The</strong>re were few examples, classified together by absence <strong>of</strong> any suggestion <strong>of</strong> a neck,<br />
together with excurvate rim inflecting into a globular body. <strong>The</strong>re are only two examples<br />
within this category, from two sites (Hg.10.103 and HV.5.205). Classificatory criteria in<br />
“Form.db” are: “rim type” has a value <strong>of</strong> “X” and “neck type” has a value <strong>of</strong> “N”.<br />
Form 2e:<br />
This class comprised a single sherd, suggestive <strong>of</strong> an unrestricted jar or deep bowl, with<br />
a flat thickened lip and punctate/fingernail impressed decoration (sherdA32, see images<br />
appended on CDROM). Classification criteria were “lip type” value “L” and “shoulder<br />
type” value “T”.<br />
Form 3: Inverted Rims:<br />
Form 3 vessels (Figure 79, Figure 82) were thought to be small, inverted (direct) rimmed<br />
vessels (classification as direct is tentative: there is a possibility that these are the heavily<br />
flared everted rims <strong>of</strong> small-orifice restricted vessels such as the water jar illustrated in<br />
Ross (Ross 1996:78, 2b water jar). Classification criteria were:<br />
• the lip was definitely represented on the sherd (to avoid confusion with rounded<br />
shoulders <strong>of</strong> everted-rimmed necked pots);<br />
• rim depth was sufficient to suggest absence <strong>of</strong> a neck;<br />
327
• rim form was inverted “rim type” had a value “T” in “Form.db”;<br />
• no carination was recorded (“carination type” had a value <strong>of</strong> “9"), to differentiate<br />
from one rim <strong>of</strong> this form which had a carination near the lip.<br />
Judging by the horizontal curvature <strong>of</strong> the rim, some <strong>of</strong> these sherds were from tiny<br />
vessels, about the size <strong>of</strong> an egg-cup. Vessel B.6/B.7 from Zangana was inferred to have<br />
had loop handles (Figure 83), the only example <strong>of</strong> this handle form, suggestive <strong>of</strong><br />
suspension during use, and conceivably having a specialized cooking/heating/drying<br />
function at some height above a fire which would have made support on three stones<br />
impractical. Given the uncertainties surrounding the overall vessel forms and sizes and<br />
functional implications <strong>of</strong> Form 3 these were omitted from seriations.<br />
Form 4: Shallow Carinated Bowls:<br />
<strong>The</strong>re is one example <strong>of</strong> this form, in an exotic quartz-calcite fabric ( Figure 79,Figure<br />
83).<br />
Form 5: Reworked Body Sherd:<br />
<strong>The</strong> single instance <strong>of</strong> Form 5 (Figure 79, Figure 83) was a large shallow spherical sherd<br />
with a circular lip <strong>of</strong> unusually high lip angle, with small irregularities <strong>of</strong> the lip suggesting<br />
possible reworking <strong>of</strong> a spheroidal body sherd. <strong>The</strong> sherd was found at Paniavile, and may<br />
be a potting stand or mold <strong>of</strong> some sort, or alternatively fits a description <strong>of</strong> frying pan<br />
sherds attested ethnographically in Oceania (Ross 1996:70). As this sherd was undecorated<br />
it had no effect on seriations.<br />
Form 6: Everted-rim Restricted Vessels:<br />
<strong>The</strong>se can be inferred with confidence from the numerous neck sherds indicating this form.<br />
In many cases, but not all, neck sherds can be approximately oriented by latitudinal<br />
decoration and corner point horizon to ascertain whether the vessel is clearly restricted<br />
328
Figure 82: Form 3 inverted restricted vessels.<br />
329
Figure 83: Form 3 or Form 4 (top left) and another example <strong>of</strong> a short-rim variant <strong>of</strong><br />
Form 2a (top right); the sole example <strong>of</strong> Form 3 with loop handle(s) (2 nd to top); the<br />
only confirmed Form 4 carinated bowl, in exotic temper (2 nd to bottom) and a large<br />
base sherd or frying pan, Form 5 (bottom).<br />
330
(Form 6) or unrestricted/weakly restricted (Form 2c). An analysis <strong>of</strong> metric variability<br />
within this large group <strong>of</strong> sherds was undertaken to investigate functional variation.<br />
Metric Variability Within Form 6:<br />
If orifice diameter is related to function, and fluid storage or fluid transport vessels are<br />
expected to have had small orifices, and cooking vessels larger orifices, this might produce<br />
a bimodal or multimodal distribution <strong>of</strong> orifice size (for which neck “Hcurve” measurement<br />
is used as a proxy here). Neck Hcurve for all such vessels taken together, using size<br />
intervals <strong>of</strong> 10mm, looked approximately normally distributed, but with a hint <strong>of</strong> multimodality<br />
in the tails <strong>of</strong> the distribution (Figure 84).<br />
Neck Hcurve seems largely constrained between 95mm and 155mm, although this<br />
could be a sample size or measurement effect. This makes some sense from the point <strong>of</strong><br />
view <strong>of</strong> function, since Hcurve <strong>of</strong> 95 mm is about the smallest that will admit a hand<br />
(remembering that Hcurve is measured external to the neck so neck thickness must be<br />
subtracted from Hcurve to yield orifice radius). General purpose vessels that require<br />
abrasive cleaning need to be large enough to admit a hand (this is the difference between<br />
pots or jars on the one hand and flasks or bottles on the other).<br />
To test whether this aspect <strong>of</strong> the distribution was perhaps being obscured by<br />
measurement error on small sherds, either through curve irregularity or taphonomic<br />
rounding <strong>of</strong> sherd edges, the same data were run with cases excluded where sherd area was<br />
less than 25cm2 . In this bar chart (Figure 85) sample size was much reduced, which<br />
Count<br />
30<br />
20<br />
10<br />
0<br />
100.00 200.00 300.00<br />
EXTERNAL NECK RADIUS<br />
331<br />
Bars show counts<br />
Figure 84: External neck radius, all sites, size<br />
intervals 10mm.
can be expected to introduce a greater level <strong>of</strong> jaggedness to the pattern. <strong>The</strong> pattern was<br />
clearer rather than obscured, suggesting that either measurement error was a factor on<br />
small sherds, or that the smaller vessel size class was represented principally by small<br />
sherds. Although at least one vessel style within Form 6 is represented predominantly by<br />
small sherds, as will be discussed below, this was not thought to be the explanation for the<br />
discontinuity, as neck Hcurve was generally larger than 96mm radius for that group.<br />
<strong>The</strong> relatively small number <strong>of</strong> necks measuring outside these apparent constraints<br />
were <strong>of</strong> interest. Were those necks <strong>of</strong> less than 95mm Hcurve perhaps functional variants;<br />
or evidence <strong>of</strong> multi-part construction; evidence that some potters (perhaps children) had<br />
very small hands (there is a photograph <strong>of</strong> a mother and daughter making pottery in Specht<br />
& Fields 1984 which illustrates this possibility):17); or were these merely irregularities in<br />
curvature, unrepresentative <strong>of</strong> average vessel neck radius? Were necks measuring larger<br />
than 150mm radius (the upper tail <strong>of</strong> the frequency distributions) perhaps other functional<br />
variants; unrepresentative irregularities in curvature; or a large variant specific to a<br />
particular period or production unit?<br />
Count<br />
20<br />
15<br />
10<br />
5<br />
0<br />
100.00 200.00 300.00<br />
EXTERNAL NECK RADIUS<br />
332<br />
Bars show counts<br />
Figure 85: All sites, sherds >25cm 2 , neck Hcurve intervals<br />
10mm.
Form 6a:<br />
<strong>The</strong> 95mm lower limit <strong>of</strong> external radius <strong>of</strong> the bulk <strong>of</strong> necks approaches the minimum<br />
needed for the hand <strong>of</strong> the potter to access the vessel. Only one excurvate rim sherd<br />
provides secure evidence for an external neck horizontal radius below 95mm (Figure 86),<br />
with a internal orifice radius measured as 47mm. This sherd EVE measured 25%, making<br />
irregularity <strong>of</strong> manufacture an unlikely explanation for the anomalous neck size. This sherd<br />
was from the Honiavasa site, where two-piece or multi-piece construction is evidenced by<br />
a number <strong>of</strong> slab-built carination sherds (Form 1). No neck exteriors are measurable for<br />
the Honiavasa carinated sherds, but measurement from drawn pr<strong>of</strong>iles for which the<br />
location <strong>of</strong> the CVA had been established from measurement elsewhere on the sherd<br />
suggest the following approximate data: 80mm, 90mm, 100mm and a single example <strong>of</strong><br />
less than 70mm. Together with the neck measurement <strong>of</strong> HV.04.133, these data provide<br />
two examples <strong>of</strong> orifices too small to get a large hand through, indicating that at least some<br />
vessels <strong>of</strong> this form were unlikely to have been used in any function requiring direct<br />
abrasive cleaning. <strong>The</strong>se two necks were interpreted as evidence for an everted-rim<br />
carinated vessel class with a probable use <strong>of</strong> fluid transport/storage, poorly represented in<br />
the Honiavasa assemblage, and not securely evidenced in samples from other sites<br />
(remembering there were single sherds with hard carinations at Hoghoi, Paniavile and<br />
Punala).<br />
Whether all Honiavasa vessels represented by carination sherds were so used is<br />
unknown. <strong>The</strong> effect <strong>of</strong> this vessel form on seriations is discussed in Chapter 12. Form 6a<br />
and Form 1 are regarded as possibly functionally equivalent, although the larger orifice for<br />
some Form 1 vessels could confer utility as multi purpose vessels including cooking<br />
functions.<br />
Form 6b:<br />
Of the 32 sherds which measured above 160mm external neck radius, some<br />
333
Figure 86: Form 6 “Neck Hcurve” variation; the two top sherds are too small to get a<br />
hand into (Form 6a), while the two lower sherds have head-sized or larger orifices<br />
(Form 6b).<br />
334
seemed inexplicable by curve irregularity alone but larger sherds in this category confirm<br />
the presence <strong>of</strong> a larger, less frequently discarded or preserved Form 6 vessels. Due to low<br />
numbers these do not affect the outcome <strong>of</strong> seriations in Chapter 12. Counts <strong>of</strong> vessel<br />
forms are given in Table 26.<br />
Form 6c:<br />
<strong>The</strong> general situation regarding the neck radius <strong>of</strong> everted rim restricted vessels, taking all<br />
sites together, is that the vast majority <strong>of</strong> reliable Hcurve observations fall between 95mm<br />
and 155mm radius. This range represents substantial variation in the physical size <strong>of</strong> vessel<br />
orifice, from adult hand-sized to adult head-sized, and might encompass virtually any<br />
function one might think <strong>of</strong> for a vessel <strong>of</strong> this general form, from wet/dry storage or<br />
fermentation (possibly with a tied cover) to boiling/simmering to feed a variety <strong>of</strong> group<br />
sizes, serving, drying/roasting (e.g. Canarium processing), ritual consumption or <strong>of</strong>fering,<br />
soaking, washing, transport <strong>of</strong> wet or dry materials or food processing (e.g. collection <strong>of</strong><br />
grated foodstuffs, settlement <strong>of</strong> starch grains, expression <strong>of</strong> coconut cream, processing <strong>of</strong><br />
cream into oil). <strong>The</strong> list is limited more by imagination than by vessel form. <strong>The</strong> necks tell<br />
us that these are pots rather than flasks, bowls or platters, and that these occur in a variety<br />
<strong>of</strong> neck sizes, within quite broad but strict size limits predominantly.<br />
<strong>The</strong>re is substantial form variability and decorative variability within the numerically<br />
large Form 6c (Table 26), and undoubtedly (on the evidence <strong>of</strong> rim form and evidence<br />
from body and base sherds presented below) functional variation also. <strong>The</strong>re seems little<br />
basis for separating Form 2c from Form 6c for the purposes <strong>of</strong> seriation, as assessment <strong>of</strong><br />
the degree <strong>of</strong> restriction relied on measurement <strong>of</strong> sherd orientation, which was inaccurate.<br />
Moreover the amount <strong>of</strong> restriction seemed to form a continuum, and therefore<br />
classification into Form 2c and Form 6c is arbitrary. Forms 2c and 6c combined seemed<br />
to be the best candidate <strong>of</strong> all the vessel forms on which to perform form-controlled<br />
seriations, due to large sample size, presence in all sites and decorative variability.<br />
335
Table 26: Counts <strong>of</strong> vessel form classes.<br />
Form Paniavile Hoghoi Miho Honiavasa Gharanga Nusa<br />
Roviana<br />
Not<br />
Classified<br />
336<br />
Kopo Zangana Total<br />
517 708 273 264 199 91 17 677 2746<br />
Form 1 1 1 9 1 12<br />
Form 2a 4 1 1 2 2 10<br />
Form 2b 1 1<br />
Form 2c 5 6 3 1 2 3 2 22<br />
Form 2d 1 1 2<br />
Form 2e 1 1<br />
Form 2f 1 1 2<br />
Form 3 2 1 1 6 10<br />
Form 4 1 1<br />
Form 5 1 1<br />
Form 6a 1 1<br />
Form 6b 1 1<br />
Form 6c 116 142 103 165 73 21 5 170 801<br />
Everted Rim Form Variation:<br />
Seventy-seven examples <strong>of</strong> everted rims from restricted vessels were complete in pr<strong>of</strong>ile<br />
from lip to neck, for which rim angle and rim depth were measurable. Neck angle and neck<br />
Vcurve were measurable for 60 <strong>of</strong> these sherds, and rim Vcurve for 54. <strong>The</strong>se data<br />
permitted a rough analysis <strong>of</strong> rim and neck form variability within this class (Forms 6 and<br />
2c).<br />
Rim eversion angle and rim depth for Honiavasa and Hoghoi sites illustrate the<br />
extremes <strong>of</strong> assemblage difference in rim form for the everted-rim restricted vessel class<br />
(Figure 87). Multiple models can be constructed to explain this difference: there may be<br />
two or more contemporaneous functional classes <strong>of</strong> vessel represented by groups in these<br />
data, or there may have been a gradual shift in production style over time, with overlap in<br />
occupation span accounting for Hoghoi rims being made in Honiavasa forms. <strong>The</strong>re
Figure 87: Relationship between rim angle and rim height for two sites, all<br />
measurements shown.<br />
may have been a sudden shift in trade-exchange patterning during the occupation span <strong>of</strong><br />
Hoghoi, after occupation ceased at Honiavasa, that brought a new style <strong>of</strong> pottery to the<br />
area or perhaps there was a shift in economy, encouraging the adoption <strong>of</strong> a new rim form;<br />
or alternatively a change in pottery diversity over time, where Hoghoi and Honiavasa<br />
occupations are non-contemporaneous but Honiavasa had two functional classes <strong>of</strong> pottery<br />
while Hoghoi had three.<br />
Measurement error (particularly <strong>of</strong> rim angle, given the difficulty <strong>of</strong> rim orientation<br />
measurement) is unlikely to explain the separation between these two sites, but might tend<br />
to blur the overall relationship between rim depth and rim angle in all assemblages. Using<br />
only those cases where EVE was greater than 9%, for which rim angle measurement error<br />
can be expected to be smaller, the plot <strong>of</strong> rim depth against rim angle for all sites was more<br />
expressive <strong>of</strong> a negative correlation (Figure 88).<br />
337
Figure 88: Rim angle and rim depth, all sites, filtered so that EVE is greater than<br />
9%, to reduce the effects <strong>of</strong> measurement error.<br />
This clarification <strong>of</strong> the relationship between rim depth and rim angle suggests that<br />
measurement error was creating some <strong>of</strong> the groupings in the data in Figure 87 (the<br />
greater the EVE, the more confidence we can have that the sherd is correctly oriented, on<br />
which the rim angle measurement depends). Re-examining the relationship between<br />
Honiavasa and Hoghoi rim form using EVE>9% as a filter retains a separation <strong>of</strong><br />
Honiavasa and Hoghoi rim forms, but with more linearity. It seems likely that there are<br />
technological or functional constraints on rim form, which might explain the linearity in<br />
these data. Tall rims tend to be less everted. This may be due to technical difficulty in<br />
forming a heavily-everted tall rim from a single lump <strong>of</strong> clay using paddle and anvil<br />
technique, or alternately, may reflect a continuum <strong>of</strong> intended-use variation. <strong>The</strong>se factors<br />
may combine in a temporal phyletic series, placing Honiavasa and Hoghoi at opposing ends<br />
<strong>of</strong> a temporal continuum, or may not.<br />
Outliers above this negative correlation tendency are probably slab-constructed<br />
rims, formed from an arc <strong>of</strong> clay cut from a flat slab, and the most extreme outlier <strong>of</strong> this<br />
338
sort is from the Honiavasa site. <strong>The</strong>re is other evidence for slab construction in that site,<br />
where Form 1 shoulders are most common, and where many tall everted rims are snapped<br />
from the vessel in the neck region, suggesting a weak join at the neck.<br />
<strong>The</strong> results <strong>of</strong> this analysis suggest that rim form variation <strong>of</strong> everted rim restricted<br />
jars may be constrained both by manufacturing processes and/or by functional variation,<br />
and that the effect <strong>of</strong> this variation on seriations should be assessed or controlled for.<br />
Inferring Vessel Form from Body Sherds:<br />
Curve fitting using brass template curves allowed accurate estimation <strong>of</strong> the evenness <strong>of</strong><br />
curves, and the radii <strong>of</strong> even curves. Errors <strong>of</strong> parallax that arise when using curvature<br />
charts were eliminated.<br />
<strong>The</strong> observed arrangement <strong>of</strong> interior anvil marks in latitudinal circles on larger<br />
sherds which could be oriented with confidence allowed approximate orientation <strong>of</strong> many<br />
less complete body-only sherds to be inferred. Using this method it was possible to<br />
measure the convexity <strong>of</strong> the exterior vertical pr<strong>of</strong>ile (Vcurve), and to take a second<br />
measurement normal to this curve (Hcurve). This was not necessarily a measure <strong>of</strong><br />
latitudinal curvature, and in most cases is better likened to a “great circle” curve in<br />
navigation, a curve marking the intersection <strong>of</strong> the earth’s surface and a plane passing<br />
through the earth’s centre. This curve was measured by choosing the tightest curve that<br />
would sit conformably on the sherd surface at right angles to the Vcurve measurement line.<br />
Curves which were uneven were noted but not entered as Vcurve or Hcurve.<br />
Body sherds for which anvil marks were absent or not patterned, and for which<br />
orientation could thus not be inferred, were measured in the dimension <strong>of</strong> their maximum<br />
and minimum curvatures. Where these two measurements differed they were noted in the<br />
“Notes” field <strong>of</strong> the “Form.db” table. Where they were equal (indicative <strong>of</strong> a convex<br />
spheroidal sherd surface) they were entered as Vcurve and Hcurve measurements. <strong>The</strong>re<br />
is thus a higher probability <strong>of</strong> measurement for spheroidal sherds than for cone/can forms,<br />
which is significant for the plots presented below. While all body sherds were measured<br />
in this manner, a minimum size filter <strong>of</strong> 25cm 2 was applied to the data analyzed, to reduce<br />
the possibility <strong>of</strong> sampling error due to an unknown level <strong>of</strong> curve evenness. <strong>The</strong>se<br />
339
methods are similar in overall concept to the two curvature method (Hagstrum &<br />
Hildebrand 1990). Results for all sites combined are presented in Figure 89 (Honiavasa<br />
is shown using a symbol different to the other site to show up a difference in the patterning<br />
<strong>of</strong> observations for that site.<br />
Figure 89:Body sherd curvature measurements, showing spheroidal sherds<br />
(diagonal alignment) and other canister/conical forms (e.g. s-shaped pattern<br />
<strong>of</strong> Honiavasa sherd measurements).<br />
<strong>The</strong> truncation <strong>of</strong> Vcurve data in Figure 89 at 300mm radius represents the upper limit <strong>of</strong><br />
the curve template set, and the difference between a radius <strong>of</strong> 300mm and a straight line<br />
is probably within the measurement error <strong>of</strong> the technique, given the sherd sizes under<br />
study.<br />
<strong>The</strong> other obvious patterning is concentration <strong>of</strong> the data along the line <strong>of</strong><br />
sphericity, where one curve fits the sherd in both dimensions <strong>of</strong> measurement. Most <strong>of</strong><br />
these measurements probably represent globular (spheroidal) bodied vessels <strong>of</strong> up to<br />
200mm body radius. Spheroidal sherds smaller than 100mm radius may simply be the<br />
340
smallest examples <strong>of</strong> spheroidal bodies, or may be the tightly curved bases <strong>of</strong> vessel bodies<br />
that are semi-conical in shape, approximating an upside-down parabola in external base<br />
pr<strong>of</strong>ile (see Figure 35). Sherds at the upper end <strong>of</strong> the line <strong>of</strong> sphericity are fairly gently-<br />
curving, near-flat, suggesting flattish base sherds from squat globular vessels or carinated<br />
vessels with a flattish rounded base and built-up sides.<br />
Sherds falling substantially below or above the spheroidal group are either the<br />
flatter side portions <strong>of</strong> conical lower bodies, or tightly curving (in vertical pr<strong>of</strong>ile) but<br />
horizontally moderate s<strong>of</strong>t shoulders or s<strong>of</strong>t carinations.<br />
By site, samples show some variation when plotted in this manner (Figure 90).<br />
Most sites show a cluster <strong>of</strong> spheroidal sherds above 100mm radius, extending to<br />
approximately 150-200mm. <strong>The</strong> spread in this cluster for Paniavile is probably<br />
measurement bias in the first analyzed sample, created by choosing the 150mm curve<br />
before others, and tending to accept this reading. This tendency was actively resisted by<br />
more careful curve fitting in other site assemblages. Hoghoi, Miho and Honiavasa have<br />
what are potentially pointy bases or small (less than 100mm radius) globular bodies. Miho<br />
and Zangana had outliers <strong>of</strong> low curvature.<br />
<strong>The</strong> Honiavasa sample showed a tighter clustering <strong>of</strong> spheroidal sherds than other<br />
samples, between 100 and 150mm radius, suggesting smaller-bodied globular vessels were<br />
the norm here. This sample also had a number <strong>of</strong> tight Vcurve measurements <strong>of</strong> around<br />
50mm associated with Hcurves <strong>of</strong> about 120mm suggesting a s<strong>of</strong>t carination body form,<br />
which did not occur in other sites among the body-only sherd samples (there was a s<strong>of</strong>t<br />
carination in the Hoghoi site sample).<br />
341
Figure 90: Comparison <strong>of</strong> body sherd curvature measurements by site.<br />
342
Gharanga clearly has the lowest incidence <strong>of</strong> sphericity, which is potentially significant for<br />
vessel function. A comment by the landowner that these sherds were probably indicative<br />
<strong>of</strong> pots broken during water collection from the Gharanga stream provides an additional<br />
hint that perhaps there could be some functional information in this plot. If water jars<br />
tended to be wall-sided in body shape, or biconical (unrestricted spheroid base, restricted<br />
cone or restricted hyperboloid shoulder) in shape, and the Gharanga site was a specialized<br />
water collecting site, one would expect this patterning in the data. <strong>The</strong>re are strong<br />
decorative similarities between Hoghoi and Gharanga, and yet marked dissimilarity in this<br />
plot, which may be functional difference. It should be noted that vertical-sided jars would<br />
produce restricted tall ovaloid or restricted tall elipsoid body sides with low Vcurve (Rice<br />
1987:219), but there should also be a minority <strong>of</strong> spherical sherd forms in an assemblage<br />
resulting from the breakage <strong>of</strong> bases <strong>of</strong> these vessel forms, which is the pattern seen in the<br />
Gharanga data although some <strong>of</strong> the spherical sherds that have made it through the filter<br />
are rather large at up to 200mmx200mm, indicating globular rather than pointed bases.<br />
Body Sherd Thickness: Form 7 Robust Low Curvature Body Sherds:<br />
While many sherds yielded large-radius curvature measurements, some unusually thick<br />
sherds within this category are <strong>of</strong> interest, as their large curves and robust construction<br />
suggest substantial vessels (Figure 91). Of the thicker sherds, those which yielded even<br />
curvature measurements are shown in Figure 92.<br />
343
Figure 91: Robust base sherds and cannister-shaped large body sherd (the latter having<br />
exotic quartz-calcite temper).<br />
344
<strong>The</strong>se sherds seem likely to represent large and heavy vessels. Three are cylinder-shaped<br />
in body, while the other is so flat as to be difficult to measure, with one thick tightly-<br />
Figure 92: Curvature <strong>of</strong> robust body sherds (thicker than 14mm).<br />
curved spheroidal sherd with a conical to spherical transition, likely to be the slightly<br />
pointed spherical base <strong>of</strong> a large vessel. Shott’s theory <strong>of</strong> vessel size and use-life predicts<br />
that such vessels will be under-represented in assemblages due to infrequent use/breakage<br />
(Shott 1996). In light <strong>of</strong> this well supported theory <strong>of</strong> assemblage formation, unusual<br />
sherds <strong>of</strong> these dimensions should be viewed as rare glimpses <strong>of</strong> vessel forms likely to<br />
have been a prominent feature <strong>of</strong> past economy, and <strong>of</strong> low archaeological visibility due<br />
to low breakage rates combined with low vessel sample sizes in the case <strong>of</strong> the Roviana<br />
sites.<br />
One can speculate on the functions <strong>of</strong> such vessels: storage <strong>of</strong> fluids, oils, starches<br />
or dried or otherwise preserved or fermented foods, cooking or serving for large groups<br />
on unusual occasions, or perhaps even provisioning/watering <strong>of</strong> canoes (one <strong>of</strong> these<br />
sherds was <strong>of</strong> exotic quartz-calcite temper). It seems clear that due to the paucity <strong>of</strong><br />
345
evidence for such vessels and the likelihood <strong>of</strong> functional divergence from the run-<strong>of</strong> the<br />
mill or frequently broken vessel form, that these sherds should not be used in any temporal<br />
characterization <strong>of</strong> assemblages. Absence <strong>of</strong> decoration automatically excluded them from<br />
the seriation analysis in Chapter 12.<br />
Vessel Function from Evidence for Use-Alteration:<br />
<strong>The</strong> varying degree <strong>of</strong> surface ablation <strong>of</strong> sherds from Roviana intertidal sites, evidenced<br />
by pedestalled temper grains or rounded sherd edges, suggests that surface sooting, if it<br />
was present, has generally been removed by postdepositional abrasion by water-born sand.<br />
Heat spalling was not recorded for the Roviana materials due to the difficulty <strong>of</strong><br />
distinguishing heat spalling from salt crystallisation spalling.<br />
One Form 6 vessel from Hoghoi had an encrustation <strong>of</strong> soot on the exterior (see<br />
Chapter 12), thought to have been preserved as a result <strong>of</strong> burial in s<strong>of</strong>t sediments for<br />
much <strong>of</strong> the past three millennia. One sherd from this vessel was recovered by shallow<br />
excavation, the other was recovered from the mud surface. Both had sooting. No other<br />
evidence <strong>of</strong> surface sooting was seen, which is not to say that it does not exist (soot may<br />
well be retained in surface crevices or in decorative punctations, or possibly grown over<br />
with coral). It is not assumed that sooting relates to the pot’s primary or intended use, as<br />
secondary or impromptu use over a fire without cleaning cannot be ruled out.<br />
One other sherd displayed near-surface carbon, which was dated by AMS. A plain<br />
body sherd from Honiavasa site was submitted for radiocarbon dating <strong>of</strong> a seed inclusion.<br />
<strong>The</strong> date turned out to be about 9000bp, and the laboratory also dated a thin subsurface<br />
carboniferous rind or zone near the surface <strong>of</strong> the vessel. This also yielded an age <strong>of</strong><br />
around 9000bp, suggesting that the clay incorporated organic materials <strong>of</strong> this age which<br />
had vaporised and carbonized near the surface <strong>of</strong> the vessel during firing, and that this was<br />
not use alteration.<br />
346
Twenty large body sherds were examined as a pilot study to test the incidence <strong>of</strong><br />
macro-cracking (visible to the naked eye), micro-cracking (visible under 10x<br />
magnification), pitting and spalling. Exterior and interior surfaces <strong>of</strong> sherds were compared<br />
(eleven from Miho, five from Hoghoi and four from Honiavasa). Five sherds showed<br />
interior macrocracking without exterior macrocracking. Old macrocracking <strong>of</strong> the exterior<br />
occurred in seven cases and in all but two cases this was matched by interior<br />
macrocracking. Most instances <strong>of</strong> microcracking appeared to be post-recovery as interior<br />
faces <strong>of</strong> the micro-cracks were unweathered and without marine growth, thought to be due<br />
to shrinkage on drying. Microcracking <strong>of</strong> the exterior was noted on eight sherds, matched<br />
on the interior in five cases. <strong>The</strong>re were a further three cases <strong>of</strong> interior microcracking<br />
unmatched on the exterior. Three cases <strong>of</strong> exterior spalling were noted, all from the sample<br />
<strong>of</strong> eleven Miho sherds. Three cases <strong>of</strong> interior spalling were recorded, all unmatched by<br />
exterior spalling. Two <strong>of</strong> these were from Miho and one was from Honiavasa. Exterior<br />
pitting was recorded on thirteen sherds, matched on the interior in eleven cases. Three<br />
further cases <strong>of</strong> interior pitting were recorded which were unmatched on the exterior. This<br />
pitting could have been post-depositional solution pitting <strong>of</strong> organic inclusions or<br />
something similarly s<strong>of</strong>t.<br />
<strong>The</strong>se pilot data seemed to hold little promise that any strong pattern <strong>of</strong> differential<br />
alteration <strong>of</strong> interior and exterior body sherds would be seen, and no further use-alteration<br />
data was collected. It would have been worthwhile, had the time been available, to look<br />
microscopically at indentations on body/shoulder sherds to try to find retained traces <strong>of</strong><br />
carbon. <strong>The</strong> virtual absence <strong>of</strong> evidence for use alteration should not be read as evidence<br />
for absence in the past, due to the slight polishing effect on sherd surfaces <strong>of</strong> water-borne<br />
calcite detritus in the swash zone.<br />
Secondary Smoothing <strong>of</strong> Vessel Interiors, and Interior Neck Form:<br />
347
It was noted that some interior pr<strong>of</strong>iles <strong>of</strong> restricted, everted rim vessels had a smooth<br />
upper body interior and a hard angle at the point <strong>of</strong> interior restriction (Figure 36 and<br />
Figure 93). It was assumed that this pattern was produced by secondary smoothing <strong>of</strong> the<br />
interior using a larger anvil than that used for primary forming, in contrast to those vessels<br />
having an uneven interior composed <strong>of</strong> multiple small anvil/finger impressions. Some<br />
vessels had thin walls with an even interior, but without the hard line around the interior<br />
neck. <strong>The</strong> possibilities are either that these variants represented temporal changes in<br />
potting skill/technique or that these are style or functional variation within/between<br />
assemblages. <strong>The</strong>se neck/shoulder interior variants are identified in table “Form.db” and<br />
in table “Flat.db” appended on CD-Rom.<br />
Four hundred and thirty-one sherds were recorded as having a restricted neck with<br />
both the neck and shoulder represented. Of these, 308 were <strong>of</strong> area greater than 15cm 2 ,<br />
chosen as a reasonable lower size limit, below which classification <strong>of</strong> shoulder interior<br />
evenness or neck interior form was considered insecure. Interior evenness data by site are<br />
given in Table 27. Hoghoi and Miho have the greatest abundance <strong>of</strong> even interiors, while<br />
Paniavile yielded only a small sample and Gharanga had a low abundance <strong>of</strong> even<br />
interiors.<br />
Table 27: <strong>The</strong> attribute “even/not even” by site.<br />
Site Count E Count not<br />
E<br />
348<br />
Total % E % not E<br />
Paniavile 8 14 22 36 64<br />
Hoghoi 17 19 36 47 53<br />
Miho 22 26 48 46 54<br />
Honiavasa 30 62 92 33 67<br />
Gharanga 7 28 35 20 80<br />
Zangana 23 39 62 37 63<br />
<strong>The</strong> relative abundance <strong>of</strong> hard neck interior pr<strong>of</strong>iles to not hard grouped Gharanga,
Figure 93: Hard interior neck pr<strong>of</strong>ile and even interior body/shoulder pr<strong>of</strong>ile (top); a<br />
s<strong>of</strong>ter neck interior pr<strong>of</strong>ile (middle) and uneven interior pr<strong>of</strong>ile (bottom).<br />
349
Hoghoi and Paniavile together (Paniavile had the smallest sample size, and is therefore the<br />
least reliable data), while Miho and Zangana yielded similar relative abundances. <strong>The</strong><br />
Honiavasa site was the odd one out, with intermediate relative abundance (Table 28).<br />
<strong>The</strong> relative abundance <strong>of</strong> even interiors with hard pr<strong>of</strong>iles to even interiors<br />
without hard pr<strong>of</strong>ile (see Figure 36 for definition <strong>of</strong> these attributes) grouped sites Miho,<br />
Honiavasa and Zangana together, as distinct from the Hoghoi samples. Hoghoi, and<br />
Gharanga / Kopo, if sample size is ignored, tended to have thin-walled vessels with even<br />
interiors and s<strong>of</strong>t interior pr<strong>of</strong>ile at the neck (Table 29). This tendency is highly correlated<br />
Table 28: Relative abundance <strong>of</strong> “hard neck” interior pr<strong>of</strong>iles to “not hard”.<br />
Site hard not hard total % hard % not hard<br />
Paniavile 3 19 22 14 86<br />
Hoghoi 5 31 36 14 86<br />
Miho 15 33 48 31 69<br />
Honiavasa 18 74 92 20 80<br />
Gharanga 4 31 35 11 89<br />
Nusa Rov. 5 4 9<br />
Kopo 0 4 4<br />
Zangana 18 44 62 29 71<br />
Table 29: Relative abundance <strong>of</strong> even interiors with hard neck interior pr<strong>of</strong>iles to even<br />
interiors without hard pr<strong>of</strong>ile.<br />
site even and hard even not hard total % e and s % e not s<br />
1(Paniavile)) 2 6 8<br />
2 (Hoghoi) 1 16 17 6 94<br />
3 (Miho) 9 13 22 41 59<br />
4 (Honiavasa) 14 16 30 47 53<br />
5 (Gharanga) 2 5 7<br />
6 (Nusa Rov.) 3 0 3 100<br />
7 (Kopo) 0 2 2 100<br />
8 (Zangana) 11 12 23 48 52<br />
350
with rim form (this Chapter) and decoration (Chapter 9).<br />
<strong>The</strong>se three relative abundances serve to separate Hoghoi from Miho/Zangana, and<br />
Honiavasa from all <strong>of</strong> these. Miho/Zangana have the highest relative abundance <strong>of</strong> hard<br />
neck interiors, while those site assemblages which have lower numbers <strong>of</strong> hard necks can<br />
be separated into two groups, one <strong>of</strong> which has even interiors highly correlated with hard<br />
necks, the other <strong>of</strong> which has even interiors with a s<strong>of</strong>ter neck, indicative <strong>of</strong> a different<br />
production style or technique. <strong>The</strong> low abundance <strong>of</strong> even interiors at Gharanga is <strong>of</strong><br />
interest, as decoratively Gharanga, Hoghoi and the northern end <strong>of</strong> Zangana bear many<br />
similarities (explored further in Chapters 11 and 12), and this is reflected also in Table 27;<br />
but in terms <strong>of</strong> their body sherd form these assemblages show some differences (this<br />
<strong>chapter</strong>) and this is reflected in their neck angles and rim forms also (this <strong>chapter</strong>). <strong>The</strong> low<br />
evenness count for Gharanga compared to Hoghoi may have more to do with a difference<br />
in site function than a temporal difference indicated by other aspects <strong>of</strong> vessel form,<br />
whereas the differences in evenness abundance and neck interior hard pr<strong>of</strong>ile abundance<br />
between other sites seem to correspond with stylistic differences, and are thought therefore<br />
to be more safely viewed as a temporal variant within the Roviana Lagoon region.<br />
Shoulder Form Classes <strong>of</strong> Form 6 Vessels:<br />
Two shoulder variants were recorded for restricted everted-rimmed jars: hard shoulders<br />
(see Figure 34 for a definition <strong>of</strong> these and for examples see Figure 94) and the much<br />
more common s<strong>of</strong>t shoulders (e.g. Z.11.572, Figure 94). <strong>The</strong>se variants were reminiscent<br />
<strong>of</strong> two forms recorded for the Kalinga <strong>of</strong> the Phillippines (Longacre 1991). In that case<br />
they were contemporaneous geographic variation <strong>of</strong> no functional significance within a<br />
local region. <strong>The</strong>re seems no evidence at present to suggest that these are correlates <strong>of</strong><br />
functional differences, and the Kalinga case <strong>of</strong>fers uniformitarian support for this.<br />
351
Figure 94: Hard shoulder variants <strong>of</strong> Form 6 (top and middle) as distinct from the more<br />
common s<strong>of</strong>t shoulder form (bottom).<br />
352
<strong>The</strong>se differences most likely reflect either slight differences in manufacturing technique<br />
or simply different production styles, either contemporaneous or temporal. Due to the<br />
possibility that this represents contemporaneous variation, this attribute was not used in<br />
seriations.<br />
Sherd Restriction Factor:<br />
Sherd restriction factor is a measure <strong>of</strong> the amount <strong>of</strong> surface area <strong>of</strong> sherds in an<br />
assemblage in relation to the amount <strong>of</strong> orifice represented (Smith 1985:269). Calculation<br />
<strong>of</strong> sherd restriction factor was made using EVE as a unit <strong>of</strong> rim circumference rather than<br />
cm as used by Smith. This might create some orifice-size bias in the results, but as neck<br />
Vcurve was not highly structured by site, this should not be too much <strong>of</strong> a problem. It<br />
would have been preferable to measure circumferences <strong>of</strong> lips, necks and carinations in cm<br />
as well as EVE.<br />
Table 30: Sherd restriction factor comparison <strong>of</strong> site samples.<br />
Site total sherd surface<br />
area (cm 2 )<br />
353<br />
total rim circ. (%<br />
EVE)<br />
Paniavile 9438 375 2517<br />
Hoghoi 9421 448 2103<br />
Miho 8438 274 3080<br />
Honiavasa 12487 814 1534<br />
Gharanga 4976 215 2314<br />
Zangana 11694 493 2372<br />
Sherd restriction<br />
factor (cm 2 per<br />
100%)<br />
Miho had the highest sum <strong>of</strong> sherd surface area per everted-rim vessel equivalent (Table<br />
30). Honiavasa had by far the lowest, while the other sites formed a group (in those cases<br />
where sample size permitted comparisons). <strong>The</strong> first question in relation to these data is<br />
whether collection intensity has biased the ratio <strong>of</strong> rim sherds to body sherds, and here<br />
Honiavasa has the lowest collection intensity, with Hoghoi and Zangana the highest. Miho,<br />
in spite <strong>of</strong> having an intermediate intensity <strong>of</strong> collection, not much
Table 31: Preservational bias <strong>of</strong> vessel parts as an explanation for differences in sherd<br />
restriction factor (data has the form average thickness (count) standard deviation).<br />
Site Lip<br />
Thickness<br />
(mm)<br />
Paniavile 8.21(66)<br />
2.83<br />
Hoghoi 6.32(84)<br />
1.78<br />
Miho 6.45(65)<br />
1.60<br />
Honiavasa 8.81(102)<br />
1.96<br />
Gharanga 7.11(34)<br />
1.45<br />
Zangana 7.76(80)<br />
2.37<br />
Rim<br />
Thickness<br />
(mm)<br />
7.49(141)<br />
2.12<br />
5.96(135)<br />
1.51<br />
7.21(111)<br />
1.5<br />
8.33(171)<br />
1.79<br />
7.02(49)<br />
1.56<br />
7.25(126)<br />
1.79<br />
Neck<br />
Thickness<br />
(mm)<br />
9.49(128)<br />
2.12<br />
7.72(152)<br />
2.12<br />
8.84(104)<br />
2.24<br />
10.28(173)<br />
2.34<br />
8.33(74)<br />
1.68<br />
8.93(169)<br />
2.45<br />
354<br />
Shoulder<br />
Thickness<br />
(mm)<br />
8.91(75)<br />
2.49<br />
7.14(137)<br />
2.03<br />
8.37(97)<br />
2.42<br />
9.50(146)<br />
2.48<br />
8.04(68)<br />
1.65<br />
7.84(135)<br />
2.18<br />
Body<br />
Thickness<br />
(mm)<br />
8.21(278)<br />
2.26<br />
6.67(355)<br />
1.93<br />
7.84(112)<br />
2.13<br />
8.69(139)<br />
2.06<br />
7.24(80)<br />
1.39<br />
7.03(247)<br />
1.87<br />
L:N<br />
Ratio<br />
L:B<br />
Ratio<br />
0.87 1.00<br />
0.81 0.95<br />
0.73 0.82<br />
0.86 1.01<br />
0.85 0.98<br />
0.87 1.10<br />
greater than Honiavasa, had the highest restriction factor, where for Honiavasa the figure<br />
was almost half. <strong>The</strong> intensively collected Hoghoi and Zangana sites had near average<br />
restriction factor. <strong>The</strong>re therefore seems to be almost no correlation between collection<br />
intensity and sherd restriction factor.<br />
When the ratio between average lip thickness and average body thickness was<br />
compared for site assemblages (Table 31), Miho, the site with the highest sherd restriction<br />
factor, had the lowest lip to body thickness ratio, and lip to neck thickness ratio, predicting<br />
significant postdepositional bias against lip sherds in this assemblage. Honiavasa has the<br />
thickest and most rugged lip sherds <strong>of</strong> all sites, slightly thicker on average than body<br />
sherds, <strong>of</strong>fering clear support for a taphonomic rather than vessel form / functional<br />
interpretation <strong>of</strong> restriction factor.<br />
Tall, weakly everted rim sherds may be particularly fragile and may be under-<br />
represented in assemblage EVE measurements due to failure to survive. <strong>The</strong>re is some<br />
support for such an assessment <strong>of</strong> the Miho sample and the southern area <strong>of</strong> the Zangana<br />
sample, which included some particularly tall but incomplete everted rims (Figure 95).
Chapter Summary:<br />
<strong>The</strong> primary evidence used in the identification and interptetation <strong>of</strong> vessel forms was<br />
classification <strong>of</strong> vertical contour <strong>of</strong> the upper vessel. Form variability was analysed as a<br />
means <strong>of</strong> instituting some control for functional variation in trying to identify stylistic<br />
variation, or variation subject to drift over time. As vessel form is inferred from sherds<br />
rather than observation <strong>of</strong> whole pots, epistemological issues arise in the identification <strong>of</strong><br />
form, and throughout the analysis the aim was to depend as little as possible on the form<br />
<strong>of</strong> lip/rim/sherds for information on body form.<br />
Thus analysis <strong>of</strong> vessel parts was separated into various sections, with more<br />
emphasis than is usual on body sherds, restriction factor, shoulder forms and neck/interior<br />
characteristics. Also, negative evidence was used: the absence <strong>of</strong> flat bases, stands, etc was<br />
noted. <strong>The</strong> paucity <strong>of</strong> bases identified in the Roviana data (only a few sherds were<br />
identified as bases, and these were rounded) strongly suggests that rounded, uniform<br />
thickness bases were overwhelmingly the norm. Six broad functional vessel form classes<br />
were identified through examination <strong>of</strong> upper body contour variation, while a seventh was<br />
defined on the basis <strong>of</strong> body sherd thickness and form.<br />
Forms 1-6 were defined using combinations <strong>of</strong> nominal descriptive variables. All<br />
assemblages were dominated by Form 6 (everted rim restricted vessels), identified securely<br />
by neck sherds. This vertical contour information was supplemented with metric analysis<br />
<strong>of</strong> neck and rim form, body sherd curvature and sherd restriction factor. Analysis <strong>of</strong> metric<br />
variables recording lip, rim, neck, shoulder and body curvatures, angles and thickness in<br />
some cases, resulted in re-evaluation <strong>of</strong> the nominal combinations <strong>of</strong> Forms 1-6.<br />
355
Figure 95: Tall, fragile everted excurvate rims.<br />
356
Form 6 was subdivided into 6a, 6b and 6c using neck horizontal curvature variability.<br />
Analysis <strong>of</strong> neck and orifice sizes indicated that the bulk <strong>of</strong> vessels <strong>of</strong> these forms had<br />
orifices <strong>of</strong> hand-size up to head-sized, suggesting substantial size variation. Form 6c was<br />
merged with Form 2c due to a subjective distinction between these classes, which also<br />
need not be salient to functional variation.<br />
Forms 2c and 6c had general-purpose form characteristics consistent with cooking,<br />
liquid transport, serving, storage, and processing functions. An analysis <strong>of</strong> rim variation<br />
within Form 2c/6c identified a negative correlation between rim height and eversion angle,<br />
which tied into ideas about form strength that came up in the analysis <strong>of</strong> restriction factor,<br />
and may also indicate some functional differences regarding suitability for liquid transport,<br />
stacking, and access for stirring or ladling. Short, heavily everted rims are at the<br />
stirring/ladling end <strong>of</strong> this continuum, while tall weakly everted rims tend more towards<br />
liquid transport, with the trade<strong>of</strong>f being a tendency to have a fragile thin lip (under-<br />
represented in the sample), except where slab construction is used, resulting in a more<br />
robust tall rim. Forms 2c/6c are the subject <strong>of</strong> a form/decoration analysis in the next<br />
<strong>chapter</strong>.<br />
Analysis <strong>of</strong> body sherd variability found some differences between sites. <strong>The</strong><br />
Honiavasa site was characterized as having mainly spherical bodies less than 150mm in<br />
radius, with relatively few conical or cylindrical sherds. Gharanga was at the opposite<br />
extreme, dominated by cylindrical sherd forms. Sample sizes were small for these analyses<br />
when smaller sherds were filtered out, but do illustrate the uses to which body sherds can<br />
be put in comparing samples <strong>of</strong> relatively plain potsherds.<br />
Surface ablation <strong>of</strong> sherds, while moderate in may cases, was sufficient to remove<br />
sooting in most cases, making use-alteration study difficult. Sooting was noted on one<br />
multi-purpose everted-rim Form 6c vessel (see Chapter 12). A pilot study <strong>of</strong><br />
microcracking, macrocracking and spalling found that most instances <strong>of</strong> these had<br />
probably occurred post-recovery. Some potential for soot preservation in crevices was<br />
357
noted but was not followed up due to time constraints.<br />
Form 7 comprised a small number <strong>of</strong> robust body sherds. One <strong>of</strong> these seemed<br />
likely to be the tightly rounded base <strong>of</strong> a sub-conical vessel body, while the others were<br />
wall-sided cylinder forms <strong>of</strong> relatively large horizontal radius. <strong>The</strong>se seemed likely to be<br />
rare preserved examples <strong>of</strong> high-use life large robust vessels, possibly storage vessels. One<br />
<strong>of</strong> three sherds was <strong>of</strong> exotic quartz-calcite hybrid temper.<br />
Sherd restriction factor was calculated for site samples, but was found to be subject<br />
to taphonomic bias. <strong>The</strong> part thickness data which were used to justify this conclusion<br />
draw attention to differences in lip and rim strength between sites, which created<br />
differences in the number <strong>of</strong> vessel equivalents represented by lip EVE, relative to total<br />
sherd surface-area in site assemblages. This will be shown to be significant for the analysis<br />
<strong>of</strong> decorative variation in the next <strong>chapter</strong>.<br />
A concern with sampling errors and measurement errors was maintained<br />
throughout the <strong>chapter</strong>. Various size filters were applied during analyses to systematically<br />
exclude or reduce the impact <strong>of</strong> measurement errors or poor representations <strong>of</strong> vessel<br />
characteristics.<br />
Form 1 and Forms 2c/6c are analysed decoratively in the next <strong>chapter</strong>, where the<br />
relationships between vessel form variability and decoration are investigated further. Here,<br />
the question for seriation is whether decorative variation is strongly correlated with form<br />
variation, or whether these vary independently. In the latter situation, we approach more<br />
closely a set <strong>of</strong> attributes for analysis that are not strongly structured by vessel function,<br />
but rather by style, and thus more likely to show drift with time.<br />
358
CHAPTER 9:<br />
CERAMIC DECORATIVE CLASSIFICATION AND<br />
Introduction:<br />
DECORATION/FORM VARIABILITY:<br />
Having made relatively detailed attribute descriptions <strong>of</strong> sherds, the analyst is faced with<br />
the task <strong>of</strong> refining which <strong>of</strong> the recorded variation is salient to the research questions. One<br />
crude but widely used option is to use all descriptive variability in seriation. In the worst<br />
scenario these data are used in a similarity matrix clustering approach, which will yield<br />
assemblage groupings, the significance <strong>of</strong> which may be hard to assess. <strong>The</strong> approach<br />
taken here is an exploratory analysis <strong>of</strong> covariance between variables in the data, with<br />
regard to theories <strong>of</strong> ceramic variation.<br />
Sherd decoration was recorded primarily as nominal or category data (with mark<br />
dimensions and spacing measured but mostly unused). A primary task in this <strong>chapter</strong> is to<br />
move from descriptive decorative information about sherds to information about variability<br />
<strong>of</strong> decoration at the scale <strong>of</strong> pottery vessels, as the units <strong>of</strong> behaviour across which<br />
decoration is patterned during manufacture. Some patterns are difficult to reconstruct from<br />
sherds (larger incised patterns for example). <strong>The</strong> first section deals with classification <strong>of</strong><br />
these using zone markers, which are identifiable across a range <strong>of</strong> levels <strong>of</strong> brokenness due<br />
to location at rugged corner points <strong>of</strong> the vessel pr<strong>of</strong>ile (lip, neck or carination).<br />
Simple bands <strong>of</strong> decoration, and the permutations <strong>of</strong> these across the vessel parts,<br />
provide a window into decorative structure that is, like the linear zone-markers <strong>of</strong> some<br />
<strong>of</strong> the larger patterns, insensitive to brokenness levels. <strong>The</strong> flat data structure in “Flat.db”<br />
(appended on CD) permits an evaluation <strong>of</strong> part representation and assemblage<br />
information content: the “diagnosticness” <strong>of</strong> sherds. Sherds that represent more <strong>of</strong> the<br />
vessel pr<strong>of</strong>ile tell us more about the structure <strong>of</strong> decoration across the vessel. Sample sizes<br />
359
for various part combinations are given and discussed.<br />
Covariation <strong>of</strong> decoration with vessel part is investigated using these samples, and<br />
this exploratory data analysis is used to refine ideas about the vessel production styles<br />
introduced mainly as illustrations in Chapter 3. A detailed analysis <strong>of</strong> lip<br />
impression/notching metric variability is given, allowing assessment <strong>of</strong> the temporal<br />
saliency <strong>of</strong> these decorative attributes in general, and <strong>of</strong> some widely used classificatory<br />
schemes.<br />
In a section integrating form variability with decorative variability, ideas are<br />
developed about which vessel decorative styles might be strongly constrained by or<br />
correlated with artefact form/function. Decorative types that crosscut form types can be<br />
better argued to be temporal styles (in the Dunnellian sense) rather than functional types,<br />
as there is some evidence for functional variation within the decorative class. Decorative<br />
attributes strongly correlated with particular forms may be poor attributes for seriation as<br />
there is a possibility that the analyst is seriating contemporaneous decorative variability<br />
related to function.<br />
Classification <strong>of</strong> Linear Motifs:<br />
As discussed in Chapter 1, a structural formulaic analysis is suited to simple arrangements<br />
<strong>of</strong> elements, but larger, more varied, or less structured patterns encourage the use <strong>of</strong><br />
broader classification criteria. Sample size is also an issue in making decisions about<br />
classification. <strong>The</strong> more vessels there are in a sample the more the classification can split,<br />
without running into sample size problems for counts <strong>of</strong> attributes. Conversely, where a<br />
sample is smaller, classificatory units may need to be lumped. Faced with the need to boost<br />
sample sizes <strong>of</strong> complex and highly varied linear motifs located on the rim, shoulder or<br />
body, and avoid the sparse data that would result from a splitting approach, these were<br />
categorized using two criteria<br />
• whether or not the decorative zone was delimited by a linear marker (as opposed<br />
to for example, a band <strong>of</strong> wavy deformation at the lip or a band <strong>of</strong> opposed pinch<br />
360
fingernail impression at the neck, or simply absence <strong>of</strong> a zone marker band)<br />
• whether the primary pattern was composed <strong>of</strong> single or double/triple lines)<br />
<strong>The</strong>se criteria divided the motifs into three groups, those with linear zone markers<br />
comprising double or triple lines delimiting the primary pattern (Class 1), those without<br />
linear zone markers, and with the primary pattern laid out in single lines, <strong>of</strong>ten but not<br />
necessarily with filled motifs (Class2), and unclassified motifs (Class 3).<br />
Class 1 comprised the following pattern codes: CH9, C10, C11, C12, C13, CL5,<br />
GM3, GM4, RL1. (See Table 15)<br />
Class 2 comprised the following pattern codes: CHV, CH2, CH3, CH4, CH5, CH6,<br />
CH7, CH8, C15, C17, C18, C19, C20, C23, C24, C25, CT3, CTW, GM5, GM7, PBR,<br />
RBR. (See Table 15)<br />
<strong>The</strong>se criteria were developed to express perceived differences between<br />
Honiavasa/Nusa Roviana incised/stamped motifs and all other sites.<br />
This classification-by-zone-marker owes a lot to Mead (Mead et al. 1975),<br />
although it is a lumping <strong>of</strong> Mead’s zone marker categories, and is similar to Wickler’s<br />
classification <strong>of</strong> incised motifs for the Buka Lapita-phase reef sites (Wickler 2001: 112-<br />
113), although in that case vessel corner points were viewed as zone markers, so the<br />
distinction here is slightly finer. <strong>The</strong> classification used here differs substantially from motif<br />
inventory approaches (e.g. Anson 1983, Bedford 2000), being a lumping <strong>of</strong> heterogenous<br />
motifs according to broad classificatory criteria. This has the benefit <strong>of</strong> increasing sample<br />
size for Class1 and Class 2 decoration, where a motif-based splitting approach would leave<br />
the data with almost as many classes as observations, i.e. sparse.<br />
Part Representation in the Potsherd Sample:<br />
<strong>The</strong> diagnostic value <strong>of</strong> sherds for the development <strong>of</strong> a classification utilizing decorative<br />
structure across the vessel depends on how complete a representation <strong>of</strong> vessel form and<br />
361
decoration the sherds provide. All sherds are not equal in the information they convey<br />
about decorative structure across the vessel, so sherd count alone is not an accurate<br />
representation <strong>of</strong> the sample size bearing on the analytical problem. Sample sizes for the<br />
various permutations <strong>of</strong> vessel part representation are given in Table 32, Table 33 and<br />
Table 33. <strong>The</strong>se information are given to convey the overall level <strong>of</strong> fragmentation <strong>of</strong> the<br />
sherd assemblage, particularly as it pertains to decoration and form analysis, and<br />
complement information given elsewhere on sherd area and weight, and lip EVEs.<br />
Table 32: Part representation, all sherds including necked sherds but excluding<br />
carinated and/or inverted-rim sherds.<br />
lowermost part<br />
lip rim neck shoulder body<br />
lip 22 266 91 60 14<br />
topmost rim 118 68 167 37<br />
part neck 223 210 62<br />
shoulder 233 75<br />
body 1053<br />
Table 33: Part representation for inverted-rim sherds (no neck corner point expected<br />
from morphology <strong>of</strong> sherd).<br />
lowest part<br />
rim shoulder/carination body<br />
topmost part lip 9 2 3<br />
362
Nine-hundred and fifty eight sherds were not identified by vessel part. <strong>The</strong> bulk <strong>of</strong> these<br />
would be body sherds or shoulder sherds with a convex exterior surface. For smaller sherd<br />
sizes shoulders and bodies are impossible to distinguish from each other. By site, there<br />
were differences in part representation that are thought to relate to the different breakage<br />
patterns <strong>of</strong> various pottery styles. This is illustrated by the data in Table 35.<br />
<strong>The</strong> Honiavasa site has the highest individual part counts, but the lowest relative<br />
Table 34: Carinated sherds (excluding carinated sherds with inverted rims).<br />
rim 3<br />
363<br />
lowest part<br />
carination body<br />
neck 5 2<br />
Topmost part shoulder 10 7<br />
Site Lip<br />
observation<br />
count<br />
carination 3 2<br />
Table 35: Form strength variation for different pottery styles and the effect on part<br />
representation.<br />
Rim<br />
observation<br />
count<br />
Neck<br />
observation<br />
count<br />
Shoulder<br />
observation<br />
count<br />
Paniavile 80 146 125 82 2<br />
Hoghoi 85 138 151 143 23<br />
Miho 66 115 107 98 9<br />
Honiavasa 103 171 170 148 4<br />
Gharanga 34 52 74 70 9<br />
Nusa Roviana 9 31 25 20 2<br />
Kopo 3 7 8 8 1<br />
Zangana 80 134 172 144 10<br />
total all sites 460 794 832 713 60<br />
LRNS sherd<br />
count
proportion <strong>of</strong> Lip-Rim-Neck-Shoulder sherds (LRNS). As sherd size is on average larger<br />
for this site than others, this low LRNS count is not an artefact <strong>of</strong> relative harshness <strong>of</strong><br />
taphonomic regime, but is instead interpreted as a relative weakness <strong>of</strong> the construction<br />
<strong>of</strong> necks in the site which caused many sherds to break at the neck rather than elsewhere.<br />
Similarly, the high LRNS count at Hoghoi is thought to relate to high form strength <strong>of</strong><br />
short, heavily everted rims common in this site. <strong>The</strong> low LRNS count at Miho and<br />
Paniavile have something to do with rim fragility also.<br />
Covariation Between Vessel Part and Decoration:<br />
Lips and Rims:<br />
Unbounded incision <strong>of</strong> the rim was matched by deformation into a wave form more <strong>of</strong>ten<br />
than expected from sample sizes, but this could have been sampling error as sample size<br />
for this decorative permutation was low, due to very few unbounded incision occurrences<br />
having the lip and rim in one piece (14 cases in total) (Table 36). Punctation <strong>of</strong> the rim<br />
was matched by plain lips more <strong>of</strong>ten than expected from sample sizes (15 examples where<br />
nine would have been expected from sample sizes). Plain rims showed no particular<br />
patterning with lip decoration beyond what would be expected from sample sizes.<br />
Lips and Necks:<br />
Lips and necks, being non-adjacent vessel parts, suffer from smaller sample size due to<br />
breakage (155 in total:Table 37). Plain necks do not seem to depart significantly from the<br />
incidences expected from sample sizes. This suggests lip decoration is not highly correlated<br />
with neck decoration for sherds that survive in this configuration. This does not mean that<br />
such structure was absent in the past, but suggests if it was present, it occurred<br />
364
Table 36: Lip decoration and rim decoration.<br />
Lip Decoration<br />
Class 2 punctation<br />
band<br />
365<br />
Rim Decoration<br />
unclassifie<br />
d<br />
plain Total<br />
Count<br />
unclassified 1 (7%) 3 (11%) 5 (11%) 27 (8%) 36 (8%)<br />
plain 5 (36%) 15 (54%) 16 (36%) 94 (27%) 130<br />
(30%)<br />
discontinuous deformation 2 (14%) 3 (7%) 38 (11%) 43 (10%)<br />
impression, “wav” pattern 2 (7%) 8 (2%) 10 (2%)<br />
deformation, “wav” pattern 5 (36%) 1 (4%) 5 (11%) 69 (20%) 80 (18%)<br />
fingernail pinching, single band 1 (7%) 3 (7%) 24 (7%) 28 (6%)<br />
impression, band on top edge 2 (7%) 4 (9%) 28 (9%) 34 (8%)<br />
impression, band on inner edge 3 (7%) 20 (6%) 23 (5%)<br />
impression, bands on both edges 4 (14%) 1 (2%) 8 (2%) 13 (3%)<br />
impression, bands on outer edges 1 (4%) 3 (7%) 31 (9%) 35 (8%)<br />
applied nubbins, band 1 (2%) 1 (0%)<br />
Count 14 (100%) 28<br />
(100%)<br />
on forms for which the lip and neck were easily separated.<br />
44 (100% ) 347<br />
(100%)<br />
433<br />
(100%)<br />
<strong>The</strong> low number <strong>of</strong> classified linear motifs occurring on the neck suggests this<br />
occurrence is an inaccuracy on the part <strong>of</strong> an individual potter, and that the mental<br />
template to which the potter or potters were working did not include decorating the neck<br />
in this manner (more on this below).<br />
Sherds from Honiavasa site are commonly broken at the neck. It was suspected<br />
that many <strong>of</strong> these Honiavasa necks were plain as usually there was no trace <strong>of</strong> decoration<br />
near the break. Only two lip/neck sherds represent the numerous opposed-pinch<br />
fingernail-impressed necks, suggesting that these necks are attached to a fragile rim-lip<br />
form. This is an important point for taphonomic inference and supports the idea that many
<strong>of</strong> the Honiavasa vessels were slab-constructed (see Chapter 7). <strong>The</strong> common tall rim<br />
form at Miho may have been an attempt to replicate slab-constructed tall everted rims as<br />
seen at Honiavasa, but using the one-piece method.<br />
Table 37: Lip decoration and neck decoration.<br />
Neck Decoration<br />
Lip Decoration<br />
Class<br />
2<br />
Unclassified 1<br />
(100%)<br />
punctation<br />
band unclassifie<br />
366<br />
plain band<br />
pinchin<br />
nubbins<br />
band<br />
Total<br />
2 (15%) 3 (75%) 12 (9%) 18<br />
(11%)<br />
Plain 4 (31%) 1 (25%) 45 2<br />
(34%) (100%)<br />
Deformation,<br />
discontinuous.<br />
Impressed bands,<br />
both edges, “wav”<br />
pattern<br />
Deformation<br />
“wav” pattern<br />
Pinching, single<br />
band<br />
Impression, band,<br />
top face<br />
Impression, band,<br />
inner edge<br />
Impression, band,<br />
both edges<br />
Impression, band,<br />
outer edge<br />
Column total 1<br />
(100%)<br />
1 (8%) 15<br />
(11%)<br />
2 (15%) 15<br />
(11%)<br />
1<br />
(100%)<br />
53<br />
(34%)<br />
11 (8%) 11<br />
(7%)<br />
2 (2%) 2 (1%)<br />
16<br />
(10%)<br />
10 (7%) 10<br />
(6%)<br />
17<br />
(11%)<br />
3 (23%) 6 (4%) 9 (6%)<br />
1 (8%) 8 (6%) 9 (6%)<br />
13 (100%) 4 (100%) 134 2<br />
(100%) (100%)<br />
10 (7%) 10<br />
(6%)<br />
1<br />
(100%)<br />
155<br />
(100%)
Lips and Shoulders:<br />
Here sample size is even more <strong>of</strong> a problem (71 observations in total), and a severe bias<br />
towards rugged LRNS forms is likely. No significant departures from the expected pattern<br />
based on attribute sample sizes are evident (Table 38).<br />
Table 38: Cross-tabulation <strong>of</strong> lip decoration with shoulder decoration.<br />
Lip Decoration<br />
Class<br />
2<br />
pinching,<br />
multi-band<br />
Shoulder Decoration<br />
band<br />
punctation<br />
367<br />
unclassified plain pinch<br />
band<br />
unclassified 2 1 1 1 1 6<br />
plain 1 6 2 1 19 29<br />
deformation,<br />
discontinuous<br />
deformation,<br />
“wav” pattern<br />
2 5 7<br />
6 6<br />
pinch band 3 3<br />
impressions,<br />
band, top face<br />
impressions,<br />
band, inner edge<br />
impressions,<br />
band, both edges<br />
impressions,<br />
band outer edge<br />
1 1 5 7<br />
3 1 4<br />
1 1 2 4<br />
1 4 5<br />
TOTAL 1 14 5 4 46 1 71<br />
TOTAL
Rims and Necks:<br />
Here sample size is improved to 389 due to part propinquinity (Table 39). Class 1<br />
bounded linear motifs on the rim are rare, and occur in association with single examples<br />
<strong>of</strong> unclassified neck modification, plain neck and opposed pinch fingernail-impressed band<br />
respectively. (See also sherd HV.2.297 in Figure 9 which has Class 1 decoration <strong>of</strong> the rim<br />
and a band <strong>of</strong> single fingernail impressions at the neck.) Class 2 unbounded linear motifs<br />
occur predominantly in association with fingernail-pinched necks, with a minority <strong>of</strong><br />
examples linked to plain necks also. Punctate rims are rare, and sample size is insufficient<br />
to see any patterned association with necks. Plain rims are matched more <strong>of</strong>ten than<br />
expected from sample sizes by punctate necks or by plain necks, and more rarely than<br />
expected by necks with a band <strong>of</strong> fingernail pinching.<br />
Table 39: Cross-tabulation <strong>of</strong> rim decoration and neck decoration.<br />
Neck Decoration<br />
Rim Decoration<br />
Class 2 pinching band unclas- plain pinchnub- TOTA<br />
multipunctasifiedingbins L<br />
bandtion single<br />
band<br />
Class 1 linear 1 1 1<br />
3<br />
(5%) (0%) (2%)<br />
(1%)<br />
Class 2 linear 1<br />
2 9 19<br />
31<br />
(20%)<br />
(11%) (3%) (46%)<br />
(8%)<br />
pinching,<br />
1<br />
1<br />
multi-band<br />
(100%)<br />
(1%)<br />
punctation,<br />
20 1 1 22<br />
band<br />
(7%) (2%) (33%) (6%)<br />
unclassified 1<br />
1 5 12 9<br />
28<br />
(20%)<br />
(6%) (26%) (4%) (22%)<br />
(7%)<br />
plain 3<br />
16 11 261 11 2 304<br />
(60%)<br />
(94%) (58%) (86%) (27%) (67%) (78%)<br />
TOTAL 5 (100%) 1 17 19 303 41 3 389<br />
(100%) (100%) (100%) (100%) (100%) (100%) (100%)<br />
368
Rims and Shoulders:<br />
Sherds representing both rims and shoulders numbered 243, a better sample size than lips<br />
and necks, indicating the more rugged nature <strong>of</strong> these sherds (Table 40). <strong>The</strong>re were fewer<br />
plain rims than expected that had Class 2 shoulders, fewer plain rims with shoulders having<br />
multiple bands <strong>of</strong> pinching, and fewer plain rims than expected with shoulders having a<br />
single band <strong>of</strong> pinching. Class 2 shoulders tended to have unclassified decoration, usually<br />
consisting unclassified or incomplete unbounded linear motifs (Figure 12). <strong>The</strong>re were<br />
more punctate rims with multiple bands <strong>of</strong> pinching on the shoulder than expected from<br />
sample sizes. <strong>The</strong>re were more plain rims with plain shoulders than expected from the total<br />
column percentage plain.<br />
Table 40: Cross-tabulation <strong>of</strong> rim decoration and shoulder decoration.<br />
Shoulder Decoration<br />
Rim Decoration<br />
Class 1<br />
linear<br />
Class 2<br />
linear<br />
Class 1 linear 1<br />
(10%)<br />
Class 2 linear<br />
pinching, multiband<br />
punctation,<br />
band<br />
2<br />
(20%)<br />
unclassified 4<br />
(40%)<br />
plain 1 3<br />
(100%) (30%)<br />
TOTAL 1 10<br />
(100%) (100%)<br />
Necks and Shoulders:<br />
punctati<br />
on<br />
multiband<br />
369<br />
punctat<br />
ion<br />
band<br />
unclass. Plain pinching<br />
band<br />
1 1<br />
(7%) (1%)<br />
3 16<br />
(21%) (8%)<br />
1<br />
(6%)<br />
7<br />
1 10<br />
(41%)<br />
(7%) (5%)<br />
1<br />
3 7<br />
(6%)<br />
(21%) (4%)<br />
8 5 6 156<br />
(47%) (100%) (43%) (82%)<br />
17 5 14 190<br />
(100%) (100%) (100%) (100%)<br />
3<br />
(50%)<br />
3<br />
(50%)<br />
6<br />
(100%)<br />
Total<br />
3<br />
(1%)<br />
24<br />
(10%)<br />
1<br />
(0%)<br />
18<br />
(7%)<br />
15<br />
(6%)<br />
182<br />
(74%)<br />
243<br />
(100%)<br />
Here part propinquinity benefitted sample sizes, and 449 observations pertain to this
elationship (Table 41). Of the 15 punctate necks, nearly half had multiple bands <strong>of</strong><br />
fingernail pinch on the shoulder, the rest were plain. Of the 348 plain necks, 87% were<br />
matched by plain shoulders, but there were a number <strong>of</strong> examples <strong>of</strong> multiple bands <strong>of</strong><br />
fingernail impression (14) and single bands <strong>of</strong> punctation (8) also. Plain necks occurred less<br />
<strong>of</strong>ten than expected for Class 1 and especially Class 2 linear motifs on the shoulder. A<br />
single band <strong>of</strong> pinching at the neck occurred more <strong>of</strong>ten than expected with Class 2<br />
shoulders. <strong>The</strong>se were matched in ten cases by Class 2 unbounded incision. Lateral bands<br />
<strong>of</strong> applied nubbins did not form a large sample in this analysis, but in four <strong>of</strong> seven cases<br />
were matched by Class 1 bounded incision on the shoulder, while only three cases were<br />
matched by plain shoulders.<br />
Table 41: Cross-tabulation <strong>of</strong> neck decoration and shoulder decoration.<br />
Shoulder Decoration<br />
Class Class 2 pinching punc- plain pinchin nubbin TOTAL<br />
1 linear multitationunclas- g s<br />
linear band band sified band band<br />
Class 2<br />
5 2<br />
7<br />
linear<br />
(1%) (33%) (2%)<br />
Pinching,<br />
1 1<br />
2<br />
multi-band<br />
(6%) (5%)<br />
(0%)<br />
Punctation,<br />
7<br />
8<br />
15<br />
single band<br />
(32%)<br />
(2%)<br />
(3%)<br />
Unclassified 1 3<br />
6 10 1<br />
21<br />
(12%) (18%)<br />
(24%) (3%) (17%) (5%)<br />
Plain 2 3 14 8 13 304 3 1 348<br />
(25%) (18%) (64%) (100%) (52%) (84%) (50%) (100%) (78%)<br />
Pinching, 1 10<br />
6 32<br />
49<br />
single band (12%) (59%)<br />
(24%) (9%)<br />
(11%)<br />
Band <strong>of</strong> 4<br />
3<br />
7<br />
nubbins (50%)<br />
(1%)<br />
(2%)<br />
TOTAL 8 17 22 8 25 362 6 1 449<br />
(100%) (100%) (100%) (100% (100%) (100%) (100%) (100%) (100%)<br />
Neck Decoration<br />
370
Summary <strong>of</strong> Decorative Structure Across the Vessel:<br />
Lips were more likely to be decorated than other parts (30% <strong>of</strong> lips were plain, while 78%<br />
<strong>of</strong> rims were plain, 78% <strong>of</strong> necks were plain, and 81% <strong>of</strong> shoulders were plain. Although<br />
the types <strong>of</strong> decoration present on lips was quite strongly structured by site (Table 42),<br />
there was little information to be had from the sherds themselves as to which lip decorative<br />
attributes went with which other decorative attributes, due to sherd breakage.<br />
Lip/rim sherds showed a weak relationship between Class 2 unbounded incision<br />
<strong>of</strong> the rim and wave deformation <strong>of</strong> the lip, although this could have been a sample-size<br />
error and the vast majority <strong>of</strong> wave-deformed lips which had a rim attached had plain<br />
rims. <strong>The</strong>re were no examples <strong>of</strong> unbounded incised rims with impressed decoration, but<br />
there were sample size problems with unbounded incised rims. <strong>The</strong> low number <strong>of</strong><br />
complete lip/rim sherds for this rim decorative category suggests a fragile lip/rim form.<br />
Punctate rims, by contrast, were more likely to have plain lips or impressed lips, with only<br />
one example <strong>of</strong> wave deformation. While there was one example <strong>of</strong> a punctate neck with<br />
wave-deformation <strong>of</strong> the lip, like punctate rims, punctate necks had either plain lips or<br />
impressed lips in general. In spite <strong>of</strong> small sample size, punctate shoulders showed a<br />
similar lip patterning to punctate rims and necks, and multiple bands <strong>of</strong> fingernail<br />
impression on the shoulder followed this tendency towards either plain or impressed lips<br />
also. <strong>The</strong>re was only one sherd relating unbounded incision <strong>of</strong> the shoulder to lip<br />
decoration (plain), a further indication <strong>of</strong> the fragile lip-rim form associated with<br />
unbounded incision.<br />
A band <strong>of</strong> opposed pinch fingernail impression on the lip did not occur with any but<br />
plain rims and necks, with the exception <strong>of</strong> a single sherd which had dentate-stamped<br />
bounded linear motif on the vessel body. This may relate simply to sample size, but the<br />
concentration <strong>of</strong> sherds with this lip decoration in a single site on particular vessel forms<br />
will be discussed further in Chapter 12. (This lip decoration was almost exclusively found<br />
in the Lapita-phase Honiavasa site, which is consistent with the occurrence <strong>of</strong> this<br />
decorative motif on a single dentate sherd from the Nusa Roviana site.)<br />
371
Unbounded incision <strong>of</strong> the rim was most commonly associated with a single band<br />
<strong>of</strong> opposed pinch finger-nail impression at the neck, and less commonly with a plain neck,<br />
while by contrast punctation <strong>of</strong> the rim was overwhelmingly associated with a plain neck<br />
(20 cases), with one instance <strong>of</strong> a single band <strong>of</strong> fingernail opposed pinching and one<br />
instance <strong>of</strong> a band <strong>of</strong> applied nubbins at the neck.<br />
Sherds with both rim and shoulder were not common, due to breakage, but<br />
suggested that where rims with unbounded incision were matched in the majority <strong>of</strong> cases<br />
by plain shoulders or occasionally unbounded incision on the shoulder, rims with<br />
punctation were less likely to be plain-shouldered, and where the shoulder was decorated<br />
for a punctate rim, all examples (7) were multiple bands <strong>of</strong> opposed pinch fingernail<br />
impression.<br />
Necks with a single band <strong>of</strong> opposed-pinch fingernail impression were most<br />
commonly plain shouldered, but where the shoulder was decorated in such cases, the<br />
decoration was almost always unbounded incision (10 examples). Only in three cases were<br />
plain necks associated with unbounded incision on the shoulder, and in a further three<br />
cases plain necks were associated with a single band <strong>of</strong> opposed pinch fingernail<br />
impression on the shoulder. <strong>The</strong>se latter three cases could be regarded as decorative<br />
imprecision, where the intention was to apply a band <strong>of</strong> fingernail impression to the neck,<br />
but the execution was less precise than the academic definition <strong>of</strong> body part in the present<br />
instance. Two cases where the neck had unbounded incision and the shoulder had a single<br />
band <strong>of</strong> opposed pinching were regarded as further examples <strong>of</strong> the pattern having slipped<br />
down a little.<br />
Where plain necks were associated with decorated shoulders, the decoration was<br />
most likely to be a band <strong>of</strong> punctation or multiple bands <strong>of</strong> fingernail impression. Punctate<br />
necks were sometimes matched by undecorated shoulders, and sometimes by multiple<br />
bands <strong>of</strong> fingernail impression.<br />
While sample size was small, the majority <strong>of</strong> examples <strong>of</strong> a band <strong>of</strong> applied<br />
nubbins on the neck were associated with bounded linear motifs on the shoulder, with the<br />
remainder associated with plain shoulders.<br />
372
Interpretation <strong>of</strong> Decorative Co-variation by Vessel Part:<br />
While the data presented could never prove that there are emic types in the assemblages,<br />
it is consistent with the idea that the etic decorative structure <strong>of</strong> a motif, its analytical<br />
meaning, varies according to its position on the vessel. In this view, a single band <strong>of</strong><br />
fingernail impression on the lip has a different attribute value to a band <strong>of</strong> fingernail<br />
impression at the vessel neck. Fingernail impression as multiple bands at neck or shoulder<br />
have the same value, a value which differs from either a single band <strong>of</strong> pinching at the lip,<br />
or a single band <strong>of</strong> pinching at the neck. Unbounded incision has a similar value whether<br />
it is on rim or shoulder. A single band <strong>of</strong> punctation has the same decorative value whether<br />
it occurs on rim, neck or shoulder, a case <strong>of</strong> the analyst being more specific about location<br />
on the vessel than the potter <strong>of</strong> the past. <strong>The</strong>se data supported the classification <strong>of</strong> sherd<br />
assemblages in terms <strong>of</strong> analytical decorative vessel classes, formulated in terms <strong>of</strong> the<br />
structure in the cross-tabulations given above. <strong>The</strong>se were useful for a number <strong>of</strong> purposes:<br />
to provide a univariate decorative description for spatial analysis and for easy comparison<br />
with form and fabric information below.<br />
Miho Decorative Class:<br />
Fingernail pinching <strong>of</strong> the neck, for example, commonly occurred with a plain rim or<br />
shoulder, but when decorated, almost always occurred with unbounded incision on those<br />
parts. While it was not possible with this information to classify undecorated sherds, it<br />
seemed reasonable to regard sherds with unbounded incision (neck or shoulder), and<br />
sherds with a single band <strong>of</strong> fingernail pincging at the neck, as belonging to a single group<br />
(Miho-class decoration after the Miho site where this seemed prevalent). It is possible, and<br />
likely, that the attributes which make up this grouping have independent rates <strong>of</strong> change,<br />
and a seriation using correspondence analysis in Chapter 12 uses the short list <strong>of</strong> attribute<br />
counts given in Table 42 for this reason, rather than the more normative decorative<br />
classes.<br />
<strong>The</strong> assignment <strong>of</strong> lip sherds to such a grouping is problematic due to poor<br />
373
preservation <strong>of</strong> lip-rims for this group (see Chapter 11 for a spatial approach to this<br />
problem). Wave-deformation <strong>of</strong> lips seemed to be associated with unbounded incision, but<br />
the predominance <strong>of</strong> plain rims with this lip form suggested caution, as did spatial<br />
information (Chapter 11). It is conceivable that a class <strong>of</strong> plain vessels with wave-<br />
deformation <strong>of</strong> the lip could hide within a classification that lumped all wave-deformed lips<br />
with unbounded incised rims and single-band pinched necks.<br />
Gharanga/Kopo Decorative Class:<br />
Another set <strong>of</strong> attributes that seem to group to form a style are punctation <strong>of</strong> the rim, neck<br />
or shoulder and multiple bands <strong>of</strong> fingernail pinching <strong>of</strong> the shoulder/upper body.<br />
Punctation <strong>of</strong> the rim, neck or shoulder may occur with plain adjacent parts, but where<br />
other decoration is present below this band <strong>of</strong> punctation, it was always multiple bands <strong>of</strong><br />
opposed pinch fingernail impression. This decorative association was termed<br />
Gharanga/Kopo style decoration in Chapter 3 after the sites where it is common. Rare<br />
instances <strong>of</strong> punctation <strong>of</strong> the lip occurred with a variety <strong>of</strong> decoration, and were omitted<br />
from this analysis as not highly correlated with any other decorative attribute.<br />
Attribute Frequencies:<br />
Attributes as defined in relation to vessel part above and their frequencies are given in<br />
Table 42. <strong>The</strong> data for Zangana site has been provided as a single sample, but also is split<br />
into two areas, Zangana South and Zangana North, as a result <strong>of</strong> clear large-scale spatial<br />
structure <strong>of</strong> decoration at Zangana (explained in detail in Chapter 11). <strong>The</strong> two Class 2<br />
linear motifs on the rim at Zangana North were sherds A21 and A22, found just outside<br />
the arbitrary boundary <strong>of</strong> Zangana North/Zangana South used in the preparation <strong>of</strong> Table<br />
42 (the boundary used in preparation <strong>of</strong> the attribute frequencies being in the vicinity <strong>of</strong><br />
the modern wharf at the end <strong>of</strong> the Munda-Zangana road).<br />
374
Table 42: Counts <strong>of</strong> occurrences <strong>of</strong> decorative attributes.<br />
Attributes<br />
Paniavile<br />
Hoghoi<br />
Miho<br />
375<br />
Honiavasa<br />
Sites<br />
nubbins, band at neck 2 8 1 11<br />
pinching, band at lip 5 22 1 1 28<br />
pinching, band at neck 20 4 21 1 1 2 20 1 19 69<br />
pinching, multi-band 3 10 2 11 2 1 1 28<br />
lip impression, band inner edge 2 5 1 22 3 1 1 34<br />
lip impression, band outer edge 11 7 3 4 4 2 4 2 2 35<br />
lip impression, band top face 1 10 10 6 1 6 2 4 34<br />
lip impression, band both edges 1 4 3 6 2 2 16<br />
lip impression, staggered both<br />
edges to form “wav” pattern<br />
1 5 2 2 8<br />
lip deformation, “wav” pattern 25 11 22 2 4 3 20 10 10 87<br />
lip deformation, discontinuous 8 8 12 2 1 1 1 6 2 4 39<br />
punctation, single band<br />
anywhere<br />
4 27 4 19 1 1 12 11 1 68<br />
Class 1 linear motif 16 1 17<br />
Class 2 linear motif rim 14 1 22 14 2 12 51<br />
Class 2 linear motif shoulder 10 2 3 1 8 8 24<br />
stamped decoration (dentate or<br />
wavy)<br />
6 1 7<br />
Total 104 92 100 93 56 11 2 98 34 64 555<br />
Variability <strong>of</strong> Impressed Lips:<br />
Lip notching and crenulate lips are <strong>of</strong>ten cited as a temporal markers, and in this section<br />
a sample <strong>of</strong> sherds with lip impression are analysed to examine the discreteness <strong>of</strong> these<br />
categories and to ask whether this type <strong>of</strong> simple decoration should be regarded as a<br />
homologous trait or as a simple analogous trait. In the Roviana sample, bands <strong>of</strong><br />
Gharanga<br />
Nusa Roviana<br />
Kopo<br />
Zangana combined<br />
Zangana North<br />
Zangana South<br />
Total
impression on the inner or outer and top edges <strong>of</strong> lips co-occurred with single bands <strong>of</strong><br />
punctation and multiple bands <strong>of</strong> opposed pinching lower on the vessel, and failed to occur<br />
in the single band pinched neck group. In the Honiavasa assemblage such lips were<br />
common in the absence <strong>of</strong> any punctation or multiple-band pinching, suggesting that lip<br />
impression may be a very generalized technique that may well be reinvented over time or<br />
occur across production units, and should be avoided for the purposes <strong>of</strong> seriation. <strong>The</strong><br />
presence <strong>of</strong> lip impression on late-prehistoric terrestrial plainware from Site 25, dated to<br />
1400AD provided support for this view <strong>of</strong> lip “notching” as a simple decorative analogue.<br />
Review <strong>of</strong> Lip Notching as a Temporal Marker:<br />
Specht developed a theory that Watom Lapita and Buka-Phase Lapita were characterized<br />
by a broad, deep, widely spaced notching <strong>of</strong> the lip (which looks like a v-shaped notch in<br />
Specht’s photograph), a decorative style that did not occur later, although “simple<br />
notching” was found throughout the Buka sequence (Specht 1969:218). In a similar vein,<br />
for post-Lapita pottery from New Ireland, Golson differentiated three classes <strong>of</strong> notching<br />
or crenulate decoration (v-shaped incision, parallel-sided incision/indentation, and wide<br />
stick or thumb indentations (scallops) (Golson 1991, 1992). Similarly, White and Downie<br />
differentiated several lip treatments <strong>of</strong> this sort (White & Downie 1980). Like Specht,<br />
Wickler noted rare examples <strong>of</strong> deep u-shaped impressed lip notches in site DAF (Wickler<br />
2001:120), as well as a variety <strong>of</strong> less bold notching or impression, classified in a similar<br />
way to the Roviana data above, according to position on the rim (see also Kirch et al.<br />
1991, Reeve 1989). Given that various people have classified pottery using the dimensions<br />
and form <strong>of</strong> notching or impression in various ways, it seemed that the sample <strong>of</strong> Roviana<br />
notched/impressed rims could usefully be analysed using the metric data that had been<br />
376
ecorded, to pose the question whether there were discrete classes in the Roviana data, or<br />
whether these classes are better regarded as arbitrary measurement devices, in which case<br />
they require careful metric definition. Metric data recorded for bands <strong>of</strong> lip impressions is<br />
analysed below using bivariate plots. <strong>The</strong> relationship between form and decoration is also<br />
examined, as one form attribute seems to be correlated with the location <strong>of</strong> impressions,<br />
suggesting that decorative location was an analogous decorative similarity arising from<br />
form similarity in a single attribute, rather than any general similarity in form.<br />
Lip Notching /Impression/Crenation in the Roviana sample:<br />
<strong>The</strong>re is no particular pattern to the metric data concerning banded parallel impressions on<br />
the outer lip (see Figure 96). Neither, significantly, is there any strong pattern discernable<br />
in the distribution <strong>of</strong> mark section. Not only does there seem to be little correlation<br />
between the size <strong>of</strong> impressions and the spacing <strong>of</strong> their arrangement around the rim, but<br />
there is no obvious correlation between these measurements and the shape <strong>of</strong> the marks<br />
in section, although there may be a slight tendency for narrower impressions to be more<br />
<strong>of</strong>ten v-shaped. Under Specht’s proposed Buka-Watom model, we would expect to see<br />
a positive correlation between mark width and mark spacing. This does not appear to be<br />
the case for the Roviana data, which raises the broader question for Lapita and related<br />
pottery styles whether the distinction drawn by both Specht and by Wickler is purely an<br />
arbitrary one, where occasional outliers with coarse decoration are being classified as a<br />
separate notching type, in which case early notching is simply more varied than later.<br />
Metric Data on the size and spacing <strong>of</strong> single-band impression on the inner<br />
(interior) edges <strong>of</strong> lips displays moderate positive correlation. Larger marks tend to be<br />
377
more widely spaced, while smaller, more closely spaced marks are almost always v-shaped<br />
in section (Figure 97). This latter pattern is likely to be a matter <strong>of</strong> tool choice: if the<br />
potter wished to execute a fine, closely spaced pattern <strong>of</strong> notching, then the edge <strong>of</strong> a<br />
fingernail or other fine object was used. This finer decoration was dominated by the<br />
Honiavasa site, with occasional examples from other sites too. This prevalence <strong>of</strong> the<br />
Honiavasa site in these data raises the question whether this attribute is a useful temporal<br />
marker. An argument will be put forward in the section below that it is not, that location<br />
<strong>of</strong> notching on the inner lip is related to lip orientation angle rather than directly relating<br />
to time.<br />
A similar scatterplot analysis <strong>of</strong> the size, shape and spacing <strong>of</strong> single-band<br />
impression on the top/end face <strong>of</strong> lips is reported (Figure 98). Again a moderate positive<br />
correlation can be seen between the variables, with a suggestion <strong>of</strong> a cluster <strong>of</strong> points from<br />
various sites <strong>of</strong> small, closely spaced v-section marks. This scatterplot is the closest any<br />
<strong>of</strong> the data come to showing a real distinction between fine edge-impressions and crenulate<br />
lips, but again, the overall pattern is one <strong>of</strong> a moderate positive correlation between mark<br />
width and mark spacing, with a variety <strong>of</strong> mark sections among the coarser expressions <strong>of</strong><br />
the pattern. <strong>The</strong>se is no strong patterning by site, which would be expected if the pattern<br />
was a reliable temporal marker. <strong>The</strong> virtual absence <strong>of</strong> this pattern from Gharanga and the<br />
absence <strong>of</strong> Nusa Roviana from the plot are most likely sample size effects.<br />
For sherds having double-band impression, on both edges <strong>of</strong> the lip (Figure 99)<br />
metric variables seemed weakly correlated, and mark section was either v or u, with no<br />
particular clustering in the plot. <strong>The</strong>re was no significant grouping by site.<br />
378
Figure 96: Lip impression, single band on outer edge <strong>of</strong> lip, labeled<br />
by mark section: u=u-shaped, v=v-shaped, o=oblique v, w=w-shaped,<br />
s=flat-bottomed groove.<br />
Figure 97: Lip impression, single band on inner edge <strong>of</strong> lip, labeled by<br />
mark section: u=u-shaped section, etc..<br />
379
Figure 98: Lip impression, single band on top face <strong>of</strong> lip, labeled by<br />
mark section: u=u-shaped, etc..<br />
Figure 99: Lip impression, bands on both edges <strong>of</strong> the lip. Labeled<br />
by mark section: u=u-shaped, etc..<br />
380
Conclusions, Lip Notching/Impression/Crenation:<br />
<strong>The</strong>re is a hint in the Paniavile data in Figure 96 <strong>of</strong> the pattern <strong>of</strong> coarse impression on the<br />
outer lip noted by Specht for Buka and Watom, and by Wickler for Buka, but the<br />
Honiavasa site, which is so much more evocative <strong>of</strong> Lapita proper, for various reasons<br />
discussed in Chapter 3 and Chapter 12, shows a pattern dominated by small impressions<br />
along the inner lip, with a weak correlation between mark width and mark spacing.<br />
Crenate lips, as a broad impression <strong>of</strong> the top <strong>of</strong> the lip, are not well supported as<br />
being a distinct class <strong>of</strong> decoration. Taking broad impressions as a group, without<br />
worrying too much about where on the lip they fall, there seems no pattern by site to the<br />
shape <strong>of</strong> the impressions. Either v-shaped or u-shaped can be broad and widely spaced, and<br />
crenation as a lip decoration grouping seems to have no strong structure in relation to site<br />
in the Roviana data set. <strong>The</strong>re is a continuum <strong>of</strong> variation between small impressions<br />
closely spaced and broad impressions widely spaced for impressions located on the inner<br />
lip and on the top <strong>of</strong> the lip, suggesting that a classificatory distinction between “crenate”<br />
and fine, closely spaced impression would be arbitrary rather than natural.<br />
Location <strong>of</strong> Lip Impression and Lip Form:<br />
<strong>The</strong> Roviana data is more strongly site-structured by decorative location than by size or<br />
spacing <strong>of</strong> marks. Impressions on the inner edge <strong>of</strong> the lip are common in the Honiavasa<br />
sample, and occur in some numbers in the Hoghoi sample (Table 42). Is this homologous<br />
similarity, which might place the Hoghoi punctate-neck/pinched shoulder pottery style (on<br />
which the attribute occurs at Hoghoi) closer in time to the Honiavasa site than to the Miho<br />
decorative style, or is this simple decorative technique merely an analogous similarity that<br />
would incorrectly order a seriation?<br />
381
Lip orientation angle is the sum <strong>of</strong> rim angle and lip angle, as shown in Figure 100.<br />
While lip angle was easily measured for most sherds with a well-preserved lip, rim angle<br />
was difficult to measure for all but the largest sherds, and the method used was not very<br />
precise (precision varied with EVE, but rim angle was seldom measured with a precision<br />
better than ±10 degrees. <strong>The</strong> difficulty <strong>of</strong> measuring rim angle must be borne in mind when<br />
interpreting lip orientation data.<br />
A comparison in terms <strong>of</strong> lip orientation between band <strong>of</strong> impression on the top<br />
face <strong>of</strong> the lip, band <strong>of</strong> impression on the inner edge <strong>of</strong> the lip and band <strong>of</strong> impressions on<br />
the outer edge <strong>of</strong> the lip suggests that lip orientation is influencing location <strong>of</strong> impressions<br />
(Figure 101). Although sample sizes are small and rim orientation was measured with poor<br />
precision, these data favour an explanation <strong>of</strong> the decorative similarity between Hoghoi and<br />
Honiavasa samples (bands <strong>of</strong> impression on the inner edge <strong>of</strong> the lip) as analogous rather<br />
than homologous. This interpretation is supported also by the occurrence <strong>of</strong> “bpi” pattern<br />
on Gharanga-style sherds (discussed further below).<br />
Figure 100: Calculation <strong>of</strong> lip orientation angle.<br />
382
Decorative Attributes and Vessel Form:<br />
Vessel form variability suggested subdivision <strong>of</strong> some decorative styles. Correlation<br />
between vessel form and decoration was obvious in some cases (e.g. Form 1 carinated jars<br />
and Class1 bounded linear motifs), but more difficult to approach quantitatively in others,<br />
for example in the large group <strong>of</strong> everted-rim jars. This latter group was the focus <strong>of</strong><br />
investigation <strong>of</strong> the relationship between decorative variability and form variability. <strong>The</strong><br />
classes <strong>of</strong> sherd relevant to this vessel class were everted rims, neck sherds, hard shoulders<br />
and s<strong>of</strong>t shoulders. No instances <strong>of</strong> decoration on the body <strong>of</strong> such vessels was recorded.<br />
Everted Rims:<br />
Figure 101: Lip orientation and location <strong>of</strong> impressions.<br />
In Chapter 8 a negative correlation between rim angle and rim height was suggested. Rim<br />
depth is re-plotted below for everted rims <strong>of</strong> five decorative attribute combinations<br />
(defined in detail below). Only two <strong>of</strong> the lip-rim-neck sherds with measurements<br />
pertaining to this relationship had Class 2 linear motif decoration and/or a band <strong>of</strong> pinching<br />
at the neck, a further indication <strong>of</strong> the fragility <strong>of</strong> lips associated with the Miho decorative<br />
class (Figure 102).<br />
383
Figure 102: Rim depth by decorative class.<br />
Sherds with Gharanga/Kopo-class decoration, comprising a band <strong>of</strong> punctation in the<br />
general region <strong>of</strong> the neck, and/or multiple bands <strong>of</strong> fingernail pinching on the shoulder<br />
region, can be subdivided using differences in rim form, in that pinching <strong>of</strong> the shoulder is<br />
restricted to a short-rim group, while punctation cross-cuts rim depth variability. If the<br />
short-rim variants, with or without fingernail impression, are classed as a production style<br />
on the basis <strong>of</strong> this attribute, many <strong>of</strong> the undecorated sherds would be incorporated into<br />
the class, as would the three short-rim outliers in the wave-deformed lip class in Figure<br />
102. This short-rim form, which commonly has the decorative attributes <strong>of</strong><br />
rim/neck/shoulder punctation and/or a matrix <strong>of</strong> pinching on the shoulder, was termed<br />
“Gharanga style” in Chapter 3, after the site where it was first noted.<br />
“Kopo style” is defined as those vessels with “Rimdepth” greater than 20mm, but<br />
with a band <strong>of</strong> punctation in the neck region (here style is not used in the Dunellian sense,<br />
but rather denotes a classification, as form is involved in its definition, which may have<br />
384
a functional aspect). Many vessels <strong>of</strong> this style have weakly everted rims and only slightly<br />
restricted neck, and may well be a contemporaneous functional variant <strong>of</strong> the Gharanga<br />
style. <strong>The</strong> taller, more vertical rim <strong>of</strong> the Kopo style may have been more suited to liquid<br />
transport, although access to the contents for stirring or ladling would have been less, and<br />
the form would have been less suited to use <strong>of</strong> a torque or tongs for lifting, and less suited<br />
to the use <strong>of</strong> a tied cover.<br />
Wave-deformation <strong>of</strong> the lip (fourth from left in Figure 102) only overlaps in rim<br />
depth with the Kopo style by three outliers. <strong>The</strong>se sherds (from Gharanga site) may be<br />
interpreted as homologous or phyletic links between the Gharanga/Kopo styles and wave<br />
deformation <strong>of</strong> the lip (the latter attribute weakly associated with the Miho decorative<br />
type). This is because Miho-style and Kopo-style vessels are functionally equivalent as far<br />
as can be established from form, and are likely to be temporal stylistic variants <strong>of</strong> each<br />
other. It follows that we might expect one to evolve into the other over time, and we might<br />
thus expect crossover styles forming an evolutionary continuum. <strong>The</strong> rarity <strong>of</strong> such links<br />
could result froms time-gaps in current sampling <strong>of</strong> variability.<br />
<strong>The</strong> sample <strong>of</strong> sherds with wave deformation on which the neck is still represented<br />
are typically tall, with one example <strong>of</strong> 100mmm “Rimdepth”. What cannot be seen from<br />
Figure 102 are the lip-rim sherds that are incomplete, missing the neck, and thus not<br />
measurable for rim depth.<br />
Bearing in mind the relationship between rim angle and rim depth already<br />
established in Chapter 8, the maximum measurable depths <strong>of</strong> incomplete decorated rims<br />
was <strong>of</strong> interest given the small sample <strong>of</strong> complete Miho decorative type rims, those<br />
measurable for rim angle and depth. (<strong>The</strong>re are several indications <strong>of</strong> a fragile rim form<br />
for this style, and any information pertaining to the form <strong>of</strong> what can be assumed to be<br />
many missing rims is <strong>of</strong> interest). Incomplete rims <strong>of</strong> Miho decorative type (including any<br />
385
Class 2 decorated rims that may have had wave deformation <strong>of</strong> the lip) yielded rim depths<br />
>48mm, >60mm, >65mm, >79mm, >69mm, >81mm, >76mm, >45mm, and >62mm,<br />
adding further support to the notion <strong>of</strong> a tall fragile everted rim form associated with this<br />
decorative type.<br />
Incomplete plain rims with wave deformation <strong>of</strong> the lip, which seem to have some<br />
relationship to Miho decorative type in view <strong>of</strong> five occurrences <strong>of</strong> this lip decoration on<br />
Class 2 incised lip/rim sherds, yielded the following minimum rim depth measurements: one<br />
sherd had measurable rim angle <strong>of</strong> 20 degrees, with a rim depth >65mm, and a further six<br />
tall incomplete rims yielded minimum depths <strong>of</strong> >26mm, >40mm, >47mm, >60mm,<br />
>79mm and >65mm, consistent with the data from complete rims, and consistent with the<br />
forms <strong>of</strong> Miho style decorated rims.<br />
For lips decorated with a pinched band, found mainly at Honiavasa, rim depth<br />
ranged from a minimum complete depth <strong>of</strong> 25mm (n=6) to a maximum <strong>of</strong> >65mm (n=11).<br />
<strong>The</strong>re was therefore no overlap in rim form with the Gharanga type, but overlap with the<br />
Kopo and Miho decorative classes and associated plain-rimmed wave-deformed lips.<br />
Forms <strong>of</strong> Plain Everted Rims:<br />
Sherds comprising at least the lip, rim and neck for which rim angle and rim depth were<br />
measured, and which were devoid <strong>of</strong> decoration, were plotted (Figure 103) and can be<br />
seen to be a mixture <strong>of</strong> the production styles or types discussed above. <strong>The</strong> category <strong>of</strong><br />
plain everted rim cross-cut the distinction between Gharanga and other forms, suggesting<br />
it was not a fine-grained category <strong>of</strong> stylistic or functional information, and should not be<br />
used in seriations when the aim is high temporal resolution.<br />
386
Figure 103: Rim depth and rim angle <strong>of</strong> undecorated liprim-neck<br />
sherds.<br />
Rim Form and Discontinuous Deformation <strong>of</strong> the Lip:<br />
Discontinuous deformation <strong>of</strong> the lip was present on only ten everted rim sherds that were<br />
complete from lip to neck, suggesting either a weakness at the rim/neck correlating with<br />
this decorative attribute or an unrestricted vessel form. Rim depth for this sample ranged<br />
from 21mm to 73mm (n=10), and rim angle for all measured sherds <strong>of</strong> this lip decoration<br />
ranged from 15 degrees to 50 degrees (n=10). <strong>The</strong> small samples <strong>of</strong> observations suggest<br />
that given an expanded sample there may be some overlap with the Gharanga form class,<br />
but lack <strong>of</strong> association <strong>of</strong> this lip decoration with any classified decoration on other vessel<br />
parts other than two examples <strong>of</strong> Class 2 unbounded linear motifs suggest that any such<br />
overlap was probably minor. <strong>The</strong> tendency for these sherds to be thick, chunky sherds<br />
suggests slab-construction in many cases.<br />
Everted Rim Form and Bands <strong>of</strong> Impression on the Lip:<br />
<strong>The</strong> four locational classes <strong>of</strong> lip impressed bands (bpi, bpo, bpt, bpb) were found in all<br />
sites, and analysis <strong>of</strong> decorative covariation between parts above found that these variants<br />
<strong>of</strong> lip impression were found in association with both plain sherds and with<br />
387
punctate/multi-band opposed pinch decoration on rim, neck and shoulder. <strong>The</strong> presence<br />
<strong>of</strong> these decoration classes in quantity in the Honiavasa site assemblage, from which<br />
Gharanga/Kopo decoration was absent, suggested that rather than these otherwise plain<br />
sherds being linked in some way to the Gharanga/Kopo decorative type, these were<br />
probably a set <strong>of</strong> lip attributes with a broad functional-temporal spread, possibly <strong>of</strong> little<br />
use for seriation. Sherds with this lip decoration were plotted labeled by lip decoration<br />
class (Figure 104).<br />
Although the pattern is not as marked for rim angle as it was for lip orientation<br />
angle (Figure 101), bands <strong>of</strong> impression on more everted rims tend to be on the interior<br />
edge <strong>of</strong> the lip, while on less everted rims impression tends to be on the outer edges.<br />
Impression on the top face is found across the range <strong>of</strong> rim angles, while impression on<br />
both edges ranges between 25 degrees and 55 degrees rim angle. On the rim depth scale<br />
the data fall into two groups, the sub-20mm group, which is argued above to be the<br />
Gharanga style <strong>of</strong> decoration, and the tall-rim group, in which a variety <strong>of</strong> decorative styles<br />
can be found. Lip impression, because it cross-cuts such a broad range <strong>of</strong> other decorative<br />
and form variability, is therefore regarded as likely to be analogous similarity, until proven<br />
otherwise, and is tracked carefully in the seriations in Chapter 12.<br />
Figure 104: Location <strong>of</strong> bands <strong>of</strong> lip impression in relation<br />
to rim form variability.<br />
388
Figure 105: Neck thickness comparison <strong>of</strong> Gharanga/Kopo<br />
and Miho styles/types.<br />
Neck Form and Neck Decoration:<br />
<strong>The</strong> rim-neck-shoulder area <strong>of</strong> vessels is generally robust due to form strength and part<br />
propinquinity. <strong>The</strong> Miho decorative class, most commonly identified from a single band<br />
<strong>of</strong> fingernail pinching in the neck region, was <strong>of</strong>ten seen to have a thickened neck, as in<br />
Figure 33. Gharanga-Kopo styles, on the other hand (defined as a decorative class by a<br />
band <strong>of</strong> punctation in the general region <strong>of</strong> the neck and multiple bands, or a matrix <strong>of</strong><br />
fingernail pinching in the region <strong>of</strong> the shoulder), tended to maintain a fairly constant<br />
thickness from lip to body. Metric data supporting this observation are given in Figure<br />
105.<br />
Neck thickness for Miho decorative type was on average about 2mm thicker than<br />
for Gharanga/Kopo type. Median neck Vcurve for Gharanga/Kopo decoration style was<br />
substantially smaller than for Miho decoration style (Figure 106). This information<br />
suggests that while Miho-style rims are likely to be under represented due to taphonomic<br />
389
ias, Miho-style necks are rugged as a result <strong>of</strong> form and thickness, and likely to be over-<br />
represented relative to other vessel styles/parts. This is borne out by sample sizes <strong>of</strong> the<br />
different form attributes. While only two Miho-style rims were complete from lip to neck,<br />
the count <strong>of</strong> Miho-style necks in Figure 106 is 48.<br />
Sherds which were plain on the rim, neck and shoulder, and thus not part <strong>of</strong> the<br />
Miho style by definition (with the possible exception <strong>of</strong> some wave-deformed lips; but the<br />
relationship between wave-deformation <strong>of</strong> the lip and Miho-style rims and necks is not<br />
clear) were <strong>of</strong>ten <strong>of</strong> much larger neck Vcurve than sherds <strong>of</strong> Miho style. This difference<br />
shows up in the median Vcurve <strong>of</strong> such sherds in Figure 106. Plain necks <strong>of</strong> large Vcurve<br />
were especially common in the Honiavasa site.<br />
Figure 106: Neck Vcurve for Gharanga/Kopo decoration,<br />
Miho Decoration, and plain rim-plain neck-plain shoulder<br />
sherds.<br />
390
Chapter Summary:<br />
Linear motifs were analytically classed as either those with linear zone markers or those<br />
without (with an unclassified dustbin category also). This classification is robust across a<br />
range <strong>of</strong> levels <strong>of</strong> brokenness, as the classificatory criteria are located at the edges <strong>of</strong><br />
design zones, these being the corner points <strong>of</strong> vessel pr<strong>of</strong>ile, locations which preserve well<br />
due to robust form. Patterns comprising formulaic expressions <strong>of</strong> repeated elements,<br />
together with the two linear motif classes, formed the set <strong>of</strong> decorative patterns that were<br />
used in exploratory analysis <strong>of</strong> decoration structure across vessels and in relation to vessel<br />
form variability.<br />
Class 1 motifs were largely confined to the Honiavasa assemblage, as was a lip<br />
zone marker <strong>of</strong> a band <strong>of</strong> opposed fingernail pinches. Similarly, a band <strong>of</strong> applied nubbins<br />
at the neck was virtually confined to the Honiavasa assemblage, as were sharply carinated<br />
vessel forms (Form 1), which were almost exclusively associated with Class 1 linear motifs.<br />
Deformation <strong>of</strong> the lip into a lateral wave occurred in two instances in the<br />
Honiavasa assemblage (both <strong>of</strong> these were coarse finger-impression variants <strong>of</strong> the wave<br />
form rather than lateral shear-deformation seen in most other sites). Lateral wave-<br />
deformation was more prevalent in Nusa Roviana, Paniavile and Zangana sites, with a<br />
small number <strong>of</strong> occurrences at Hoghoi and Gharanga. Wave deformation <strong>of</strong> the lip<br />
generally does not appear with punctation, with rare exceptions. While punctation was<br />
dominant at Zangana North, and Class 2 linear motifs dominated at Zangana South, wave<br />
deformation did not show this spatial patterning, occurring in several instances in both<br />
areas, suggesting possibly that wave deformation <strong>of</strong> the lip, while related to Class 2 linear<br />
motifs, may have had an extended production span relative to Class 2 linear motifs.<br />
Unlike the Honiavasa site, where the few observed instances <strong>of</strong> neck decoration are<br />
predominantly a band <strong>of</strong> applied circular nubbins (and occasionally a band <strong>of</strong> single<br />
fingernail impressions repeated), neck decoration <strong>of</strong> the other collections taken as a whole<br />
is commonly a single band <strong>of</strong> opposed-pinch fingernail impression. Where this co-occurs<br />
391
on sherds with classified decoration, the decoration is almost always unbounded linear<br />
incised (Class 2 linear). <strong>The</strong> combination <strong>of</strong> wave-deformed lip, linear incised Class 2 rim<br />
and shoulder (perhaps with some idiosyncratic applied decoration), and a band <strong>of</strong> opposed<br />
pinch decoration at the neck, is the classic expression <strong>of</strong> what I have termed the Miho style<br />
in Chapter 3 and elsewhere (Felgate 2002), but the attribute analysis in this <strong>chapter</strong><br />
suggests that a less normative approach as taken here reveals that attributes comprising<br />
this style vary independently to some extent, and may, if this variation is temporal, each<br />
have their own trajectories and rates <strong>of</strong> change.<br />
Within the Miho style there is some pattern by site regarding whether Class 2<br />
motifs occur on the rim or shoulder, with Class 2 motifs on the shoulder being almost as<br />
common as on the rim at Paniavile and Zangana South, while they occur mainly on the rim<br />
only at Miho (where Class 2 motifs occur on the shoulder it is usually motif C18 that is<br />
present). This distributional information, where the spatially distant Paniavile and Zangana<br />
South sites share a decorative similarity, while the relatively close Miho and Zangana<br />
South sites differ, suggests there may be temporal variation within the Miho style, visible<br />
in the details <strong>of</strong> attribute structure across the vessel. Accordingly, Class 2 decoration <strong>of</strong><br />
the rim and Class 2 decoration <strong>of</strong> the shoulder are treated separately in the seriations in<br />
Chapter 12.<br />
Miho-style attributes usually (but not always) occurred on excurvate rimmed<br />
restricted pots (Form 6c), and a case was made for these having a rugged neck form but<br />
a fragile rim/lip, leading to a taphonomic bias in part representation, where few complete<br />
lip/rim/neck sherds have survived.<br />
Punctation occurs predominantly as a single band in the general region <strong>of</strong> the neck<br />
(sometimes on the adjacent rim, as though the vessel was decorated while upside-down,<br />
and other times on the shoulder, clearly decorated while upright). Although a couple <strong>of</strong><br />
sherds <strong>of</strong> flared everted form had dual bands <strong>of</strong> punctation (these may have been<br />
unrestricted bowls) the single punctate band motif generally occurred on medium-height<br />
vertical-to short heavily everted rims, in the latter case <strong>of</strong>ten associated with multiple bands<br />
<strong>of</strong> pinching on<br />
392
the shoulder, forming a matrix <strong>of</strong> pinches, or a band <strong>of</strong> vertically oriented raised bars in<br />
effect, rather like some Honiavasa applied decoration. <strong>The</strong> tall punctation-only vessels<br />
were named Kopo-style, while the short-rim heavily-excurvate variants which sometimes<br />
had the matrix <strong>of</strong> pinching were called Gharanga style. Gharanga/Kopo showed a tendency<br />
for thinner necks than Miho style.<br />
<strong>The</strong>re are various applied decorations in all site assemblages, most <strong>of</strong> which seemed<br />
idiosyncratic and therefore difficult to classify. This lack <strong>of</strong> clear patterning to the applied<br />
decoration meant that in the main (with the exception <strong>of</strong> a band <strong>of</strong> circular nubbins at the<br />
neck) it was omitted from consideration in seriation analysis in Chapter 12.<br />
Homologous vs. Analogous Similarity:<br />
Analysis <strong>of</strong> lip impression/notching variability suggested there was little basis for<br />
regarding similarities between Honiavasa and Hoghoi in this regard as anything other than<br />
analogous similarity in the absence <strong>of</strong> evidence to the contrary, although the effect <strong>of</strong> these<br />
variables on seriations is considered in Chapter 12. By contrast, the Honiavasa pottery and<br />
the Miho/Gharanga-Kopo styles all share what seems likely to be a homologous trait in<br />
their neck decoration. In the Honiavasa site there were a number <strong>of</strong> instances <strong>of</strong> necks<br />
being decorated with a single band <strong>of</strong> circular nubbins. <strong>The</strong> Miho style is characterized<br />
as having a band <strong>of</strong> opposed pinching at the neck, which <strong>of</strong>ten has the appearance <strong>of</strong> a<br />
band <strong>of</strong> raised lumps <strong>of</strong> clay. In the Gharanga and Kopo styles, these are replaced by a row<br />
<strong>of</strong> punctations (almost always circular) made using a tool (access for the finger-techniques<br />
<strong>of</strong> applied nubbin and opposed pinch is restricted by the Gharanga rim form, and this<br />
could explain a change in tool used). Taken together, these three neck decorative attributes<br />
account for almost all bands <strong>of</strong> decoration at the neck or in close proximity to the neck<br />
393
(148 instances in total), and <strong>of</strong>ten occur on otherwise plain sherds, suggesting they are a<br />
primary component <strong>of</strong> a Roviana intertidal-site pottery design system, and that the three<br />
decorative techniques are three expressions <strong>of</strong> the same pattern, a row <strong>of</strong> circular elements.<br />
<strong>The</strong> three expressions <strong>of</strong> this pattern are thus likely to be homologous similarity between<br />
styles.<br />
This design feature may be phyletically related to the eyes in the Lapita faces as<br />
suggested by Spriggs (Spriggs 1990). Such an interpretation could be extended to suggest<br />
that these are either the eyes <strong>of</strong> the pot personified, or the pot as representation <strong>of</strong> person<br />
or persons, ancestral or mythical. <strong>The</strong> persistence <strong>of</strong> this design structure across the<br />
variability <strong>of</strong> the various styles suggests a temporal persistence, or, in Dunnell’s terms, a<br />
function, since the pattern is not drifting, only the technique <strong>of</strong> execution, at least at the<br />
temporal scale <strong>of</strong> the sample. <strong>The</strong> rate <strong>of</strong> change <strong>of</strong> this feature <strong>of</strong> the design system must<br />
be constrained, since it is so conservative compared to other design attributes.<br />
Eyes might have a protective ide<strong>of</strong>unction; after all, the contents, which are in most<br />
cases likely to be food, provide both sustenance and threat <strong>of</strong> poisoning, the latter either<br />
through inadvertent contamination or through malice (the distinction between these may<br />
be peculiarly modern; a result <strong>of</strong> modern microbiological knowledge). I note that previous<br />
explanations for Lapita “faces” as reviewed in the opening <strong>chapter</strong>s have tended to be<br />
soci<strong>of</strong>unctional rather than ide<strong>of</strong>unctional in that there has been a strong emphasis on ritual<br />
relating to the maintenance <strong>of</strong> Austronesian house societies and/or prestige exchange. Here<br />
the suggestion is put forward that for the Roviana pottery at least, a more prosaic efficacy<br />
may be being attempted, simply protecting the contents through ritualistic manufacture<br />
including the application <strong>of</strong> decorative eyes. Where the house-ancestor and prestige<br />
exchange theories suggest male manufacturers, an ide<strong>of</strong>unctional explanation <strong>of</strong> the sort<br />
suggested here is more consistent with utilitarian production by women.<br />
In this <strong>chapter</strong> a spatial division <strong>of</strong> style between Zangana North and Zangana<br />
South has been introduced, which is examined in more detail in Chapter 11, prior to the<br />
ceramic seriations in Chapter 12.<br />
394
Introduction:<br />
CHAPTER 10:<br />
LITHICS<br />
<strong>The</strong> main focus <strong>of</strong> this research has been ceramics, but the intertidal/shallow water lag<br />
deposits <strong>of</strong> resistant materials have the potential to provide good samples <strong>of</strong> lithic artefacts<br />
and manuports. <strong>The</strong> analysis <strong>of</strong> lithics had several practical applications in meeting the<br />
overall methodological objectives <strong>of</strong> this research. Firstly, the form <strong>of</strong> lithic artefacts has<br />
the potential to slot the Roviana materials into an existing Culture-Historical framework,<br />
providing traditional and useful independent evidence for chronological corroboration,<br />
however coarse.<br />
Secondly, this study has a central aim <strong>of</strong> providing a methodological scoping<br />
exercise for these intertidal sites: how might we proceed in seeking to write prehistory<br />
from the materials at hand? <strong>The</strong> intertidal artefact/manuport scatters have a conspicuous<br />
lithic component, due to the rugged nature <strong>of</strong> lithics. We have so little preserved to work<br />
with that it is impossible to ignore the lithic component <strong>of</strong> assemblages in seeking to<br />
develop applied archaeological method. <strong>The</strong> preserved lithic material can be used in<br />
sourcing studies, in behavioural studies, and in taphonomic studies, and this <strong>chapter</strong><br />
provides pilot studies along each <strong>of</strong> these lines <strong>of</strong> enquiry. Substantive results are not earth<br />
shattering, but the attempt is made in order to think about what sort <strong>of</strong> information might<br />
usefully be acquired from these sites and how it might be used.<br />
Although Reeve reports an obsidian point from Paniavile, and Wickler found<br />
copious quantities <strong>of</strong> obsidian on the Lapita reef sites <strong>of</strong> Buka, no obsidian, and only two<br />
small flakes <strong>of</strong> chert were found in the present study. This may be an observer bias, or a<br />
different setting which alters the visibility <strong>of</strong> these materials, but the expectation that these<br />
395
would be present was not met during extensive searching with the aim <strong>of</strong> locating obsidian.<br />
Shell adzes have been numerous in some early ceramic sites (more than 200 from<br />
AN-6 on Anuta) (Kirch & Rosendahl 1973:67). While there are many small Tridacna<br />
adzes littering the surface <strong>of</strong> some late-prehistoric sites <strong>of</strong> the Roviana area, none have<br />
been found so far from the intertidal sites, other than the three Tridacna adzes reportedly<br />
recovered from Paniavile by villagers, illustrated by Reeve (Reeve 1989). <strong>The</strong> lack <strong>of</strong> these<br />
in all recent collections may be either a preservation, observer, or visibility bias or a<br />
behavioural difference between the occupants <strong>of</strong> the collection sites in the past (we cannot<br />
distinguish between these possibilities on current evidence). In order to compare like with<br />
like, in terms <strong>of</strong> raw material properties and manufacturing constraints, no comparisons are<br />
made between the stone adzes recovered from the Roviana intertidal sites and the shell<br />
adzes recovered from early ceramic sites elsewhere.<br />
A section about water-rounded and fractured lithic manuports attempts a<br />
petrographic classification based on a macroscopic type series. Petrographic results 1 allow<br />
reclassification into broad source groupings. <strong>The</strong>se groupings will be used in Chapter 11<br />
to investigate whether manuport source classes are spatially structured at Hoghoi.<br />
Additionally, manuport size frequency distribution is used to make some behavioural<br />
inferences about procurement strategies and uses <strong>of</strong> these. <strong>The</strong> results are valuable in that<br />
they raise a series <strong>of</strong> questions on these topics which show the value in focusing this sort<br />
<strong>of</strong> attention on these materials. Also, many <strong>of</strong> the nondescript fragments that looked as<br />
though they held little archaeological value individually turned out to be more exciting<br />
when analyzed petrographically, and suggest that intensive analysis <strong>of</strong> all lithic materials<br />
can pay major dividends for this type <strong>of</strong> site. A wealth <strong>of</strong> information content is there,<br />
1 Petrographic descriptions were transcribed by the author from verbal descriptions<br />
generously provided by Dr Robin Parker, <strong>of</strong> the Geology Department, <strong>University</strong> <strong>of</strong><br />
<strong>Auckland</strong>.<br />
396
awaiting detailed analysis.<br />
Formation processes are a major focus <strong>of</strong> the ceramic analysis. <strong>The</strong> lithics, like the<br />
ceramics, can provide information on formation processes through brokenness,<br />
completeness and remaining use life. For lithic artefacts, brokenness and completeness can<br />
be used in the same way as ceramics, to estimate a breakage population, except that<br />
artefacts like adzes have bilateral symmetry rather than circular symmetry about a central<br />
axis. <strong>The</strong> difference between a breakage population and the extant deposit may inform on<br />
formation processes if a sufficient sample is available to allow an estimate <strong>of</strong> breakage<br />
population within useful limits. This was not the case with the sample <strong>of</strong> adzes obtained,<br />
but an expanded sample would potentially allow this sort <strong>of</strong> approach. Remaining use life<br />
can be assessed for those adzes that are recovered whole, by dimensional comparison with<br />
unused preforms. Where the degree <strong>of</strong> resharpening is low, a substantial use life remains,<br />
and if this is the dominant pattern, site abandonment or destruction by a sudden natural<br />
disaster or attack is a possibility. Where adzes are all re-sharpened to stubby dimensions<br />
with little remaining use life other than recycling to smaller tools, gradual discard through<br />
unexceptional processes is more likely. Some rudimentary comments can be made on the<br />
Roviana materials in this regard, as one unused preform from a local landowner/collector<br />
was recorded.<br />
Sourcing studies can only be taken so far within the limited resources <strong>of</strong> a thesis,<br />
but petrographic descriptions given in this <strong>chapter</strong> illuminate a diversity <strong>of</strong> geological raw<br />
materials. <strong>The</strong> geological origins from which these were sourced and the uses to which<br />
these were put are limited by little more than the imagination at present, and it is hoped<br />
that these data will demonstrate the desirability <strong>of</strong> making these materials the object <strong>of</strong> a<br />
sustained program <strong>of</strong> research and cultural heritage resource management.<br />
397
Review:<br />
Formation theory suggests that unlike utilitarian pottery, stone axes may have long use-<br />
lives and low discard rates (Shott & Sillitoe 2001:276), leading to an expectation that<br />
sample size will be smaller for stone axes than for pots. It is assumed here that the same<br />
holds for stone adzes. Turner’s discussion <strong>of</strong> adze use-life (Turner 2000: 231-296) does<br />
not cover the relative representation <strong>of</strong> adzes and other materials in archaeological<br />
samples, but rather focuses on the various use-life states <strong>of</strong> recovered adzes and<br />
experimental adzes. <strong>The</strong> conclusion that, “Both experimental and archaeological data<br />
support the probability that adzes had long use lives and underwent considerable<br />
morphological changes as a result <strong>of</strong> intensive curation (Turner 2000:296).” supports that<br />
assumption.<br />
In line with this expectation, few stone adzes have been recovered from Lapita-age<br />
sites in Near Oceania. Four are reported from site SAC at Watom (Green & Anson 2000b:<br />
59), and “...a few flakes... (Golson 1991:255) are reported from Lasigi. <strong>The</strong>re is no<br />
mention <strong>of</strong> stone adzes in any <strong>of</strong> the Arawe Lapita site reports (Gosden 1989, 1991a, b,<br />
Gosden et al. 1989, Gosden & Webb 1994). Few are reported from excavations in the<br />
remote Oceanic islands <strong>of</strong> the southeast Solomons or Vanuatu, with the exception <strong>of</strong> the<br />
Reef/Santa Cruz sample (much <strong>of</strong> which may have been surface collected, and excavated<br />
area was large). On Anuta, for example, no stone adzes were found either from<br />
excavations at AN-6 or from surface collections, while more than two hundred shell adzes<br />
were recovered (Kirch & Rosendahl 1973:66). Thirteen stone adzes or flakes from adzes<br />
were reported as analysed petrographically from the Lapita sites <strong>of</strong> the Reef/Santa Cruz<br />
Islands (Green 1976a:259), although it is not known whether these are all from separate<br />
adzes; the number <strong>of</strong> adzes represented by that sample may be less than thirteen. <strong>The</strong>re are<br />
also a number <strong>of</strong> other as-yet unreported stone flakes (most with signs <strong>of</strong> grinding) in the<br />
398
Reef-Santa-Cruz collections, some <strong>of</strong> which are clearly fragments <strong>of</strong> adzes (Sheppard<br />
2003: pers. comm.). Apparently no stone adzes were recovered during the extensive<br />
program <strong>of</strong> survey and excavation <strong>of</strong> Lapita and post-Lapita sites on Mussau, as there<br />
seems to be no reporting <strong>of</strong> any such items in the most recent volume on the subject,<br />
although no conspicuous mention is made <strong>of</strong> this absence in the report (Kirch 2001).<br />
In Vanuatu, the best Lapita samples at present are from surface sites on Malo.<br />
Hedrick found two stone adzes here (one plano-lateral and one plano-convex) (Hedrick<br />
n.d). Ten shell adzes were recovered from the Atanoasao site on Malo, seven from surface<br />
collection and three from excavation, but no stone adzes were found (Galipaud 1998).<br />
Recent intensive testpitting on a number <strong>of</strong> early Vanuatu sites (Bedford 2000: 205-206)<br />
recovered one Tridacna adze from Bedford’s early period (3000-2800bp), from the<br />
Arapus site, while a number <strong>of</strong> shell adzes have been recovered from later contexts,<br />
together with a single rectangular-sectioned stone adze from Bedford’s 2800-2500bp<br />
period.<br />
In contrast to this dearth <strong>of</strong> evidence from excavated Lapita sites on land in island<br />
Melanesia, there has been a relative flood <strong>of</strong> stone adzes and flakes from reef sites in Near<br />
Oceania, which may say more about the nature <strong>of</strong> the sampling process than any behavioral<br />
differences (large areas are being sampled by the surface collections, as done also by Green<br />
in the extensive surface collections and area-excavations in the Reef-Santa Cruz sites).<br />
Three complete and four fragmentary stone adzes were recovered from Paniavile by local<br />
collectors (Reeve 1989:55), which were either quadrangular or Plano-convex in cross<br />
section.<br />
Green and Davidson’s Samoan Adze typology (Green & Davidson 1969) as<br />
recently modified by Wickler for a relatively large Island Melanesian sample (Wickler<br />
2001:181) provides the classificatory units for Wickler’s study <strong>of</strong> Buka stone adzes. Green<br />
and Davidson based their Samoan typology on previous schemes <strong>of</strong> Buck’s and Poulsen’s,<br />
399
ut with more emphasis on cross section attributes, rather than the emphasis on attributes<br />
<strong>of</strong> the butt, lugs, or curvature <strong>of</strong> the back as in Duff’s scheme (Green & Davidson 1969).<br />
Green and Davidson’s type IVa and type V were characteristic <strong>of</strong> early Samoan ceramic<br />
sites, and were rare in surface collections. <strong>The</strong>se are similar in cross section, being flat on<br />
one surface and convex-to-trapezoidal on the other, but type IVa is sharpened with the<br />
edge at the flat side, producing a flat cut, while type V is sharpened with a horseho<strong>of</strong> -<br />
shaped edge along the convex side, which produces a gouge-shaped cut.<br />
Green and Davidson’s types VI, VII and VIII are all triangular-sectioned,<br />
sharpened either way to produce either a narrow straight edge or a narrow gouge, and<br />
some <strong>of</strong> these are large and heavy, but many are small and fragmentary. Type VI was<br />
thought to be present throughout the Samoan sequence. Type VII was a narrow deeper<br />
version <strong>of</strong> type V, with a prominent narrow gouge-shaped cutting edge, as though for<br />
producing grooves or rebates, and was found only in early contexts. Type VIII is rare in<br />
the Samoan sample, and although only known from surface contexts, was inferred to be<br />
an early form from its rarity. This was a heterogeneous type, most subtypes <strong>of</strong> which had<br />
a straight cutting edge and a triangular -sectioned back, and were a broad triangle in cross-<br />
section (broader than type VI).<br />
Wickler looked at 18 adzes and 19 triangular-sectioned flaked stone tools from the<br />
Buka reef sites. Wickler departed from Green and Davidson’s typology in that fragments<br />
<strong>of</strong> triangular-sectioned flaked stone tools were classified separate to the adzes (Wickler<br />
2001:175, 177), when these may well have been broken fragments <strong>of</strong> small type VI adzes.<br />
Wickler describes ten <strong>of</strong> these as flakes <strong>of</strong> dark green lithified mudstone, and nine as<br />
andesite, both materials used in the manufacture <strong>of</strong> the adzes in that sample. Five <strong>of</strong> these<br />
had ground or polished margins. Both factors suggest that at least some <strong>of</strong> these fragments<br />
could be classified as small Green and Davidson type VI adze fragments. Whether the term<br />
adze is appropriate to these small narrow forms or whether they were used as chisels<br />
400
is unimportant in the current context, where we are dealing with small fragments only, and<br />
are limited to discussion largely <strong>of</strong> differences in cross-section.<br />
Roviana Lithic Artefacts:<br />
Lithic artefacts from the intertidal included seven relatively complete adzes (Figure 107,<br />
Figure 108, Figure 109, Figure 110), a number <strong>of</strong> small fragments <strong>of</strong> what were probably<br />
adzes (not illustrated), two chert flakes, a worked sandstone slab, notched as though for<br />
an anchor and a number <strong>of</strong> small cobbles with depressions on the edges and ends, semi-<br />
cubic in form, that are interpreted as Canarium hammers (Figure 111). <strong>The</strong>se differ from<br />
an ethnographically known hafted nut hammer found widely on land in the region, and are<br />
identified by the use-indents on the six faces <strong>of</strong> the cube. <strong>The</strong>y are typically composed <strong>of</strong><br />
a relatively s<strong>of</strong>t rock that is inferred to have indented easily with use, compared to the<br />
hafted variety, which is not to say these were too s<strong>of</strong>t to be functional.<br />
Petrographic descriptions <strong>of</strong> lithic artefacts are given in Table 43. <strong>The</strong>se data<br />
suggest a diversity <strong>of</strong> lithic resources were exploited for the manufacture <strong>of</strong> artefacts,<br />
including at least two sources <strong>of</strong> adze rock, one represented by A.12, A.31 and A.6, the<br />
other, more diverse group represented by the remainder <strong>of</strong> the adze fragments. <strong>The</strong> second<br />
petrographic grouping includes substantial differences in the surface appearance <strong>of</strong> the<br />
stone, with one adze and one fragment from Hoghoi being glossy black in appearance,<br />
while other fragments from Zangana and Miho within this broad petrographic grouping<br />
were more weathered in appearance, in a variety <strong>of</strong> hues. <strong>The</strong>se surface differences suggest<br />
that if further analysis was carried out to characterize diagenesis within this grouping it<br />
could be further subdivided on a systematic basis.<br />
401
Figure 107: Planilateral sectioned adze from Hoghoi initial surface collection (top),<br />
plano-convex-sectioned type V adze from Zangana (middle), and planilateral adze<br />
fragment from Zangana (bottom).<br />
402
Figure 108: Images <strong>of</strong> the adzes shown in the preceding illustration.<br />
403
Figure 109: Green “type VI” or “type VIII” triangular-section adze fragment from<br />
Miho (top); butt-end <strong>of</strong> a plano-lateral-sectioned adze from Zangana South (middle)<br />
and a fragment <strong>of</strong> a trapezoidal-sectioned Green “type IV” adze from Zangana.<br />
404
Figure 110: Images <strong>of</strong> adzes illustrated on previous page.<br />
405
Figure 111: Canarium hammerstone from H5 ceramic findspot (top) (a similar artefact<br />
was found at Hoghoi); waisted sandstone slab from Zangana (middle); and chert flakes<br />
from Hoghoi (bottom).<br />
406
Table 43: Petrographic descriptions <strong>of</strong> lithic artefacts.<br />
Artefact ID Sit<br />
e<br />
B1.T50 10-<br />
15<br />
Unit Description<br />
8 13 In hand specimen a pale green-grey fine-grained adze fragment. In thin section a fine-grained sedimentary or volcanic<br />
rock. It is impossible to be more determinate without further analysis, as recrystallization <strong>of</strong> sedimentary clasts and<br />
primary volcanic crystallization could have the same petrographic result.<br />
HG.17.1 2 17 In hand specimen a glossy black stone, being a small fragment <strong>of</strong> an adze with polishing visible on two faces. In thin<br />
section this shows a fine-grained texture with sedimentary banding, with a possibility also <strong>of</strong> recrystallization.<br />
HG.99 2 99 A complete adze in a glossy black stone, similar in appearance to HG.17.1. This adze is reminiscent <strong>of</strong> Green and<br />
Davidson’s type V, occurring only at the early end <strong>of</strong> the Samoan sequence [Green, 1969 #375:25] In thin section,<br />
sedimentary or volcanic: comprising fine material in a very fine-grained matrix, with some evidence <strong>of</strong> feldspar<br />
recrystallization, with clay-rich silty minerals perhaps present as well.<br />
B3.T10 5-10 8 89 Mid-grey fine-grained appearance in hand specimen, weathered to a green-grey on the surface. In thin section coarser<br />
grained with micaceous clay precursors, dominated by plagioclase, with some Pyroxenes.<br />
A.12 8 48 In hand specimen a dark grey rock with a dull but generally unweatherd surface, being the sharpened end <strong>of</strong> a thin flat<br />
adze. In thin section a sedimentary rock, coarser grained. diagenesis indications include feldspar recrystallization and<br />
clay precursors. <strong>The</strong>re is no obvious calcite, and the matrix is dominated by plagioclase feldspars. Noted for<br />
recrystallized opaque pyrite.<br />
A.31 8 94 In hand specimen a complete but heavily resharpened plano-convex “horseho<strong>of</strong>”(Green and Davidson type 5) or<br />
gouge-edged adze in a glossy brown-grey stone (mid-grey when sawn). In thin section a coarser-grained sedimentary<br />
rock, similar to A.12., with feldspar dominant, diagenetic recrystallization <strong>of</strong> a variety <strong>of</strong> minerals, including opaque<br />
pyrite.<br />
A.6 8 23 In hand specimen the midsection <strong>of</strong> an adze, probably plano-convex, with surface weathered to a dark brown-grey,<br />
and mid-grey when sawn. In thin section the same as A.31 and A.12.<br />
B1.T50 0-<br />
5m<br />
8 12 A small grey rock fragment without obvious polished or ground surfaces. Mid-to-dark-grey when sawn. In thinsection<br />
it is unclear whether this is sedimentary or volcanic in origin, due to an advanced stage <strong>of</strong> recrystallization,<br />
making it difficult to see original textures. Dominated by plagioclase feldspar in various stages <strong>of</strong> recrystallization.<br />
Fine material also looks feldspathic, but includes a pyroxene/Ferromagnesian component. Pale-green chlorite is<br />
possibly present. Texture possibly indicates a plutonic origin, but this is tentative, and would require further analysis<br />
and thought to characterize this with more confidence.<br />
Miho 3 1 butt-end <strong>of</strong> a triangular-sectioned or fractured adze? Lightly ground or water rolled? Surfaces are dark-brown-grey,<br />
with sawn surfaces being mid grey to dark grey. This rock is a lot fresher than “B1.T50 0-5", with only slight<br />
recrystallization. Fine-grained plagioclase crystals are set in an even finer-grained matrix <strong>of</strong> recrystallized feldspathic<br />
character. This might be an immature, poorly-sorted sedimentary rock, or a crystal-rich igneous rock with a low<br />
carrying fluid content. XRF analysis would clarify whether chemistry indicates volcanic or sedimentary origin and<br />
diagenesis. X-ray diffraction would show up any quartz, clay minerals or calcite.<br />
P199 1 1 In hand specimen a thin small flake <strong>of</strong> fine-grained stone weathered to a golden-brown color, with light-grey sawn<br />
color. Possibly reworking debitage? In thin section this has a very fine-grained volcanic matrix carrying scattered<br />
plagioclase and pyroxene phenocrysts.<br />
HV.04.141 4 4 A flat sandstone slab 10x7x1cm, weathered to a golden-brown color on the surface and light grey when sawn. In thin<br />
section this is well supplied with pyroxenes, plagioclase and quartz, these being immature, angular, poorly sorted,<br />
with a silty clay cement. No Calcite. This is a packed stone with larger grains forming a self-supporting framework.<br />
Clinopyroxene/augite are present(a green pyroxene is probably augite). Plagioclase can show simple and multiple<br />
twinning as well as zoning. Amphibole and hornblende are present as well as a fine-grained foraminiferous silt clast.<br />
B(Zangana) 8 B In hand specimen this was a fin-shaped sherd <strong>of</strong> sandstone weathered to a golden brown, about 12cm maximum<br />
dimension, with a light grey color when sawn. In thin section a predominantly calcite cement with occasional<br />
unfragmented foraminiferal shell, suggesting calcite is a direct perecipitant, or a biogenetic calcite fine fraction, or<br />
recrystallization. <strong>The</strong> matrix is dominated by silt-sized glass fragments and some crystal products( feldspars including<br />
plagioclase, pyroxenes, and opaques). This was interpreted a s a cemented volcanic ash dominated by glassy<br />
fragments, with crystals co-genetic with the glass, which is bubble-fractured by expanding gasses. Microprobing the<br />
feldspars and glass or grinding/acid-washing/XRF would identify the type <strong>of</strong> volcanism involved.<br />
Z76 8 76 A coarse grey sandstone slab with pedestalled black mineral grains, 16x12x2cm and subrectangular in form, with two<br />
notches forming a waist, interpreted as a sandstone grindstone modified for secondary use as an anchor stone. Sawn<br />
colour matches the surface colour. In thin section fill <strong>of</strong> immature, angular, unsorted pyroxene, plagioclase and augite<br />
up to 1mm in length in a calcite cement. Fine-grained silty volcanogenic clasts are also present, which sometimes<br />
display fragments <strong>of</strong> coarser pyroxene and plagioclase. No confirmed quartz.<br />
It seems clear there are diverse resources being used over the period <strong>of</strong> deposition <strong>of</strong> these<br />
materials. Additional samples, more concerted analysis, and discovery and<br />
407
characterization <strong>of</strong> source quarries are needed to explicate patterns <strong>of</strong> lithic raw material<br />
transport. Adzes/fragments from Zangana <strong>of</strong> recrystallized fine-grained rock showed a<br />
variety <strong>of</strong> surface colours, including a pale greenish example. This suggests that raw<br />
material procurement may have been influenced by surface color, as there is some<br />
congruence here with the green adze rocks noted by Green for the Reef-Santa Cruz Lapita<br />
sites, the green plano-convex example from Samoa (Fijian origin?), the adze in green<br />
alteration dacitic welded tuff from Tonga, and green meta-sedimentary adzes from Naigani<br />
(Green 1979, 1994).<br />
A number <strong>of</strong> adzes and axes from Paniavile held in the private collection <strong>of</strong> Mr Sae<br />
Oka were photographed during the course <strong>of</strong> the study (Figure 112). Some small waisted<br />
axes and one unfinished adze preform were reportedly collected from Gharanga, by the late<br />
Mr Phillip Lanni (Figure 113, Figure 114). <strong>The</strong> preform is much longer than the complete<br />
adzes, while cross-section size is similar to the other adzes. If the preform is any indication,<br />
the adzes collected are less than half their original length, suggesting a lengthy use life<br />
involving a massive grinding resharpening labour or possibly more rapid<br />
reflaking/resharpening to repair edge damage.<br />
Sandstone artefacts/manuports similarly suggest a diversity <strong>of</strong> sources, as yet<br />
poorly represented in the collections. <strong>The</strong>re is clearly vast scope for artefact studies and<br />
lithic sourcing research for this type <strong>of</strong> intertidal ceramic/lithic scatter. HV.04.141 (Table<br />
43) is unlikely to be from the New Georgia group due to the abundance <strong>of</strong> quartz in the<br />
sand-sized component, as sediments rich in quartz should not occur given the geological<br />
characteristics <strong>of</strong> the area, although some forearc sediments have a plutonic component<br />
which might be relevant here (the nearest such source would be southern Rendova or<br />
Tetepare). <strong>The</strong> cemented volcanic ash (sample B-Zangana South, Table 43) might<br />
originate within the New Georgia group, as rocks originating as volcanic ashes are present<br />
on Ranongga, and possibly Southern Rendova or Tetepare.<br />
408
Figure 112: Artefacts photographed from Oka collection and reported to be from the<br />
Paniavile site: shell and stone adzes (top); stone adzes (middle) and waisted<br />
tools/weapons and a pineapple club (bottom).<br />
409
Figure 113: Waisted axes photographed from the Lanni collection, courtesy <strong>of</strong> the late<br />
Mr. Phillip Lanni, found in the vicinity <strong>of</strong> Gharanga Stream.<br />
410
Figure 114: Un-ground adze preform photographed from Lanni collection<br />
courtesy <strong>of</strong> the late Mr. Phillip Lanni, reportedly found at the Gharanga<br />
site.<br />
Sample Z76, from Zangana (the waisted sandstone slab in Figure 111), could conceivably<br />
be from Ranongga/Rendova/Tetepare, as the forearc region has had a complex history <strong>of</strong><br />
volcanism, uplift <strong>of</strong> the forearc region followed by subsidence, associated with erosional<br />
features, and subsequent rapid and accelerating uplift in response to subduction <strong>of</strong> the<br />
Coleman seamount (Mann et al. 1998). <strong>The</strong>se processes have created a variety <strong>of</strong><br />
siltstones, mudstones and sandstones <strong>of</strong> the Tetepare Formation.<br />
Chert Flakes/Fragments:<br />
Two chert flakes were found at Hoghoi (Figure 111). One (HG.20.22) was thin sectioned,<br />
411
and identified as microquartz 2 , but did not have any micr<strong>of</strong>oraminifera that might narrow<br />
the range <strong>of</strong> options for a source region. It is likely to be exotic to the New Georgia group<br />
as cherts are not expected there. <strong>The</strong> nearest known sources <strong>of</strong> chert are in Vella Lavella,<br />
where it occurs in minor amounts (Sheppard, pers. comm.), Santa Ysabel, Nggela or<br />
Malaita (Sheppard 1993:124).<br />
Analysis <strong>of</strong> Hoghoi Water-rounded and Fractured Volcanic Manuports:<br />
A transect-collected assemblage <strong>of</strong> 345 lithic manuports from the Hoghoi site was<br />
described in the field by matching to a type series <strong>of</strong> 23 stones or stone fragments. <strong>The</strong><br />
characteristics <strong>of</strong> degree <strong>of</strong> fragmentation (water rounded, fragmented with rounding, or<br />
fragment with no water-rounded cortex visible), type, and mass (to the nearest 50g) were<br />
recorded for each stone. <strong>The</strong>se data are given in table “Hoghoi_lithics.db” appended on<br />
CD. <strong>The</strong> type series was retained and petrographic thin-sections were prepared for each<br />
type. A combination <strong>of</strong> physical properties in hand-specimen and optical petrographic<br />
properties were used to reclassify the type series as shown in Table 44.<br />
Porphyritic olivine basalt is the most widespread rock type in the New Georgia<br />
group and occurs on all <strong>of</strong> the main islands except Simbo (Dunkley 1986:13). Classes 1,<br />
2, 10, 12, 16 and 18 fall into this category. <strong>The</strong> nearest source is Rendova, lying about<br />
12km distant across the Blanche channel, and this is a major source <strong>of</strong> ovenstones for the<br />
Roviana lagoon area in modern times. A slightly longer but more sheltered trip 15km east<br />
<strong>of</strong> Hoghoi can locate similar stones in mainland riverbeds draining into the Lagoon.<br />
Class 15 is a feldspar-phyric basalt or andesite, another important rock type that<br />
covers large areas <strong>of</strong> New Georgia, Vangunu, and southern Rendova. <strong>The</strong> nearest source<br />
<strong>of</strong> this rock type to the Hoghoi site is the Hura River (Dunkley 1986:22), about five km<br />
distant across the Lagoon. Vesicular and scoracious lava flows are present in the Picritic<br />
lavas eroding into the Viru Harbour area and Marovo Lagoon (Dunkley 1986:16), and<br />
2 Cherts were examined petrographically by Dr Peter Sheppard <strong>of</strong> the<br />
Anthropology Department, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
412
Classes 4 and 5 might be from such a source, which would involve transport from Viru<br />
Harbour at a minimum, a distance <strong>of</strong> about 50km. Kolombangara, at a similar distance<br />
from Hoghoi, is also noted by Dunkerly as having scoria cones. Closer sources are not<br />
precluded on current evidence. Class 6 has similar mineral composition to Class 5, and may<br />
be from a similar source.<br />
Class 7 is an olivine basalt with olivines serpentinised rather than altered to<br />
iddingsite or magnetite, suggesting a magnesium-rich magma. Sepentinised olivines are<br />
noted in the gabbros <strong>of</strong> the Choe intrusive complex (Dunkley 1986:31), but whether this<br />
means volcanic rocks in that area are more likely to show serpentine rather than iddingsite<br />
as the alteration product <strong>of</strong> olivine is unknown, but this suggests a specific question for<br />
source sampling in the future. If it does turn out that serpentine is characteristic <strong>of</strong> Viru<br />
harbour rocks, then the Class 4 and 5 scoriacious manuports, Class 6, and Class 7 could<br />
all be a pointer to procurement <strong>of</strong> a minority <strong>of</strong> manuports at a greater distance than is the<br />
norm today. Scoracious rocks in particular might have a non-cooking function, as their<br />
heat capacity is low. That function could relate to their relatively low specific gravity, and<br />
surface roughness which might confer desirable properties as net weights or possibly<br />
abrasives.<br />
Looking only at those manuports that were water-rounded and unfractured in form,<br />
a comparison <strong>of</strong> size (average weight) by class is given in Table 45. Some classes appear<br />
mainly as small cobbles or pebbles <strong>of</strong> less than 120g average weight (for example the<br />
large Class 18, and also classes 3, 5, 7, 10 and 16), while other classes average more than<br />
240g (Classes 1 and 2). This is most likely a slight difference in the survival rate <strong>of</strong> larger<br />
stones if used for cooking (i.e. Class 18 is more likely to shatter than Class 1) but the<br />
larger number <strong>of</strong> complete water rounded stones in Class 18 is a bit <strong>of</strong> a mystery, unless<br />
these were preferred for some other function, such as fishing or net weights. Perhaps this<br />
is just sampling error, as the number <strong>of</strong> occurrences <strong>of</strong> complete water-rounded stones is<br />
low for each class.<br />
413
Table 44: Petrographic classification <strong>of</strong> manuports collected at Hoghoi.<br />
Sampl<br />
e #<br />
count class Description<br />
HG1 67 1 Medium grey color, phenocrysts visible to the naked eye. In thin section a groundmass <strong>of</strong> Plagioclase laths and fine opaques<br />
with Pyroxene phenocrysts and a brown-rimmed alteration mineral (probably Iddingsite).<br />
HG2 36 2 In hand specimen a pinkish rock with pale green phenocrysts. In thin section the groundmass was predominantly Plagioclase<br />
laths with subordinate Olivines altered to Iddingsite. Phenocrysts were large Pyroxenes. No Feldspar phenocrysts.<br />
HG3 21 3 Fine grained spherulitic texture, a brown glassy matrix carrying scattered large Feldspar phenocrysts as well as Feldspar<br />
elongate quench crystals, giving stong indications <strong>of</strong> rapid cooling forming glass followed by recrystallization from a series<br />
<strong>of</strong> points. A large crystal <strong>of</strong> Calcite is present, pseudomorphing pre existing Olivine or Pyroxene.<br />
HG4 21 4 Pinkish in hand specimen with small vesicles. In thin section a groundmass <strong>of</strong> medium-textured Plagioclase and Pyroxene<br />
crystals.<br />
HG5 26 5 In hand specimen pinkish, fine-grained, s<strong>of</strong>t to saw. A dense, fine-grained texture with some vesicularity. Reddish<br />
groundmass, with mainly Plagioclase phenocrysts and some Pyroxene.<br />
HG6 1 6 In hand specimen pale grey to pale green and fine-grained in hand specimen. In thin section this had a very fine-grained<br />
Plagioclase groundmass with Plagioclase phenocrysts, similar to HG5 but without the vesicular texture.<br />
HG7 10 7 In hand specimen grey, medium grained, even textured. In thin section a plagioclase groundmass and abundant pyroxene<br />
phenocrysts were noted, and also an alteration product <strong>of</strong> Olivine, Serpentine, indicating a relatively magnesium-rich<br />
magma.<br />
HG8 2 1 A coarse-grained grey rock in hand specimen, and in thin section a coarse volcanic Plagioclase groundmass with abundant<br />
Pyroxene phenocrysts and smaller brown alteration mineral, possibly Iddingsite.<br />
HG9 13 9 Grey and medium-textured in hand specimen. In thin section a very fine groundmass <strong>of</strong> unidentified mineral grains with<br />
large phenocrysts <strong>of</strong> Plagioclase and smaller Pyroxene phenocrysts.<br />
HG10 18 10 In hand specimen a dark-grey rock with greenish phenocrysts. In thin section a fine grained volcanic matrix, some small<br />
crystals <strong>of</strong> an alteration mineral, possibly Iddingsite, with phenocrysts <strong>of</strong> Pyroxene and Plagioclase.<br />
HG11 15 5 In hand specimen a s<strong>of</strong>t pinkish fine-grained rock, and in thin section a dense, fine-grained texture carrying large phenocrysts<br />
<strong>of</strong> Plagioclase and Pyroxenes, with a possible Olivine alteration product in the groundmass.<br />
HG12 12 12 In thin section a coarse-grained phenocrystic volcanic carrying Clinopyroxene, Augite, occasional Orthopyroxene<br />
(Hypersthene), red Iddingsite. (This alteration product <strong>of</strong> Olivines occurs in Iron-rich magmas and could result from use <strong>of</strong><br />
heat as oven-stones).<br />
HG13 6 5 In hand specimen s<strong>of</strong>t, pinkish and fine grained. In thin section a very fine-grained volcanic groundmass with mostly<br />
Plagioclase phenocrysts.<br />
HG14 4 14 A fine-grained light-grey water-rounded pebble. In thin section there is dense accumulation <strong>of</strong> Plagioclase and Pyroxenes in<br />
layers, could be plutonic. No Olivine. Fine-grained matrix suggests volcanic.<br />
HG15 10 15 Huge Feldspar/Pyroxene phenocrysts (some 1cm) in a matrix <strong>of</strong> fine Plagioclase. Some brown-rimmed smaller phenocrysts<br />
are possibly an altered pre-existing Olivine.<br />
HG16 14 16 In hand specimen a dark grey medium-grained rock similar to HG7. In thin section a matrix <strong>of</strong> very fine Plagioclase contains<br />
subordinate medium-sized Plagioclase phenocrysts and dominant large Pyroxene phenocrysts.<br />
HG17 7 17 In hand specimen a very-fine-grained dark grey angular fragment, in thin section a very fine grained volcanic matrix carrying<br />
scattered phenocrysts <strong>of</strong> Plagioclase and a dark brown/red alteration mineral, with occasional Pyroxene phenocrysts (see also<br />
GW8 in ?), interpreted as a fragment <strong>of</strong> adze rock.<br />
HG18 57 18 Dark grey in hand specimen with greenish phenocrysts, similar to HG 7 and HG 16. In thin section a medium-textured<br />
Plagioclase groundmass with Pyroxene phenocrysts and a lot <strong>of</strong> medium-sized opaques. Some brownish minerals that are<br />
possibly an alteration mineral <strong>of</strong> Olivine or Calcite.<br />
HG19 6 19 Vesicular in hand specimen with pink-to grey color zonation (heat alteration rom use?). In thin section a very fine<br />
groundmass (mineral not identified) with rounded red-to-brown phenocrysts filling some vesicles.<br />
HG20 4 2 In hand specimen dark grey and coarse-grained, and in thin section a coarse-grained texture, predominantly plagioclase<br />
groundmass with large Pyroxene phenocrysts. Large plagioclase is absent. Iddingsite phenocrysts present<br />
HG<br />
“incise<br />
d”<br />
pebble<br />
HG<br />
40-45<br />
pound<br />
er<br />
1 3 As below, except texture is slightly coarser and includes pyroxene phenocrysts.<br />
1 3 In thin section a fine grained sperulitic texture, a glassy matrix carrying scattered large Feldspar phenocrysts as well as<br />
Feldspar elongate quench crystals, giving strong indications <strong>of</strong> rapid cooling forming glass followed by recrystallization from<br />
a series <strong>of</strong> points.<br />
414
Table 45: Size comparisons between petrographic classes <strong>of</strong> lithic manuports.<br />
Class Count <strong>of</strong> class Sum <strong>of</strong> mass (g) Mean(St.D.) mass (g)<br />
1 15. 3600. 240(252)<br />
2 10. 2600. 260(353)<br />
3 17 1300 76(26)<br />
4 2. 300. 150(71)<br />
5 8. 825. 103(123)<br />
7 4. 400. 100(71)<br />
10 7. 550. 79(27)<br />
14 3. 575. 192(267)<br />
16 9. 1075. 119(146)<br />
18 31. 3650. 118(188)<br />
Taking water-rounded (complete) cobbles <strong>of</strong> all classes together (Figure 115), there is a<br />
suggestion in a histogram <strong>of</strong> sizes that stone collecting strategies operated between the<br />
limits <strong>of</strong> roughly 50g ( the scale used was not very accurate) and 1150g, although the<br />
lower limit may reflect an archaeological collection threshold, and the upper limit is<br />
imprecise due to small sample size <strong>of</strong> larger complete stones (heavier stones than this<br />
were present among the fractured cobbles, with the largest stone being 1900g, with mean<br />
mass for all stones collected being 200g, and Standard deviation <strong>of</strong> 212g). Most <strong>of</strong> the<br />
larger stones within these limits would have been fractured by heat, but the smaller stones,<br />
from 50g to 150g, seem to have been either not subjected to heating or have survived<br />
better. <strong>The</strong>se smaller stones are an inconvenient size for domestic cooking from the<br />
perspective <strong>of</strong> the modern local practice <strong>of</strong> handling stones individually with bamboo<br />
tongs, and may have had another function, such as net weights or fishing weights, for large<br />
nets such as turtle or dugong nets, which required heavier weights due to their added<br />
buoyancy.<br />
415
Figure 115: Water-rounded lithic manuports, cortex<br />
complete.<br />
Figure 116: Size distribution <strong>of</strong> fractured lithic manuports,<br />
either with some cortex or without.<br />
416
Ethnographic cooking practice need not reflect those <strong>of</strong> 3000 years ago though.<br />
If the weight frequency distribution <strong>of</strong> complete stones (Figure 115) is contrasted<br />
with the weights <strong>of</strong> fractured stones (Figure 116) it seems that the complete recovered<br />
examples underestimate the upper weight limit <strong>of</strong> procurement, as some fractured stones<br />
approach 2kg in weight.<br />
<strong>The</strong> spatial distribution at Hoghoi <strong>of</strong> the petrographic classes defined in this section<br />
is discussed in Chapter 11.<br />
Chapter Summary:<br />
Lithic manuports systematically surface collected at Hoghoi included a diversity <strong>of</strong> volcanic<br />
rocks, with coarse porphyritic textures and cortex suggestive <strong>of</strong> origins as transported<br />
water-rounded cobbles. Sizes in excess <strong>of</strong> 2kg were targeted in some cases, but there was<br />
no evidence for any rocks larger than this. <strong>The</strong>se coarse textures, and the fractured forms<br />
<strong>of</strong> most rocks suggests a cooking function. Some fine-grained sperulitic-textured<br />
fragments with quench-textures suggest rocks poorly suited to oven-stone use (liable to<br />
explode?), and some <strong>of</strong> these may be fragments <strong>of</strong> flaked artefacts. At least one fine-<br />
grained recrystallysed sedimentary or volcanic fragment in the manuport collection is likely<br />
to be a fragment <strong>of</strong> an adze, as petrographic characteristics matched those <strong>of</strong> some <strong>of</strong> the<br />
adzes collected.<br />
A number <strong>of</strong> smaller porphyritic water-rounded pebbles were collected, which do<br />
not fit with local ethnographically observed oven-stone size selection. <strong>The</strong>se may be<br />
evidence <strong>of</strong> other non-cooking functions, such as net weights or fishing sinkers, or possibly<br />
slingstones even, or may indicate a shift in cooking practices over time. <strong>The</strong> presence <strong>of</strong><br />
scoracious basalts is inconsistent with a cooking function also, and these may originate<br />
from Viru or Kolombangara, although more local sources may exist, as yet<br />
417
unreported. While most olivine basalts are most likely from Rendova, serpentinised olivines<br />
may indicate different sources, but this has yet to be tested.<br />
A diversity <strong>of</strong> adze materials indicated the potential for sourcing studies to<br />
illuminate raw material procurement patterns for this site type. At least one plano-convex<br />
horseho<strong>of</strong>-edged adze form was recovered from Zangana, regarded by Green as indicating<br />
a homologous link with Lapita at this site, as these forms are characteristic <strong>of</strong> the Lapita<br />
horizon and derivatives (Green, pers. comm.). Comparison <strong>of</strong> the complete adzes with an<br />
unused preform suggests the complete used examples have been heavily resharpened to<br />
less than half their original length prior to discard, suggesting little remaining use life.<br />
Although the sample <strong>of</strong> such items is small, it does suggest accumulation from gradual<br />
refuse discard rather than discard as a result <strong>of</strong> sudden or catastrophic abandonment <strong>of</strong><br />
sites or extinction <strong>of</strong> inhabitants. <strong>The</strong> presence <strong>of</strong> a number <strong>of</strong> small fragments <strong>of</strong> adzes,<br />
identifiable only through raw material characteristics, suggests either a taphonomic regime<br />
capable <strong>of</strong> smashing these rugged artefacts, or reworking <strong>of</strong> adzes into smaller artifacts,<br />
damning any notion that the observed quantities <strong>of</strong> even rugged lithic artefacts can be read<br />
as directly measuring the intensity/duration <strong>of</strong> discard.<br />
Other lithic items such as sandstone fragments and artefacts are not always easy to<br />
find among coral gravels, but the diversity <strong>of</strong> sources indicated by petrographic analysis<br />
suggests these can be a major component <strong>of</strong> sourcing studies, and are worth putting more<br />
effort into finding and analysing.<br />
This <strong>chapter</strong> demonstrates that archaeological values <strong>of</strong> this type <strong>of</strong> site lie as<br />
much in the tiny nondescript lithic fragments as in the complete artefacts. One is <strong>of</strong> little<br />
value without the other. <strong>The</strong> complete artefacts alone would mislead if read as the<br />
quantities and diversity <strong>of</strong> adzes or whatever discarded in the past. Similarly, the<br />
fragments alone would be difficult to recognize as artefact fragments without petrographic<br />
analysis <strong>of</strong> both these and the complete or identifiable adzes. Intensive investigation <strong>of</strong><br />
lithic variability can unlock the cultural and postdepositional formation processes <strong>of</strong> these<br />
418
sites. It is clear from this lithic analysis, as it was from the various ceramic <strong>chapter</strong>s, that<br />
the information content <strong>of</strong> these sites is fragile as a result <strong>of</strong> this interdependence: take the<br />
fancy bits away and it gets harder, if not impossible to understand the fragments, and the<br />
fancy pieces alone do not give an accurate picture <strong>of</strong> site formation.<br />
<strong>The</strong> raw materials and forms <strong>of</strong> the Roviana flaked and polished stone adzes<br />
recovered from the intertidal sites are very similar to the materials recovered from the<br />
Buka Lapita sites on the reef flat. This similarity is so marked as to suggest that detailed<br />
petrographic and morphological comparisons would be justified, although it was not<br />
possible to do this within the time-constraints <strong>of</strong> the present study.<br />
Comparisons between the Buka sample and the Roviana sample <strong>of</strong> adzes suggest<br />
these are expressions <strong>of</strong> a very similar range <strong>of</strong> adze forms and selection <strong>of</strong> raw materials,<br />
most economically explained as homologous similarity. This supports the notion <strong>of</strong> a<br />
Lapita/post-Lapita stone adze repertoire, if not an adze kit, which is not to say that there<br />
was no change over time or space, but which acknowledges that given the limited current<br />
sample, particularly for near-Oceania, a coarse stone adze horizon can be seen, which<br />
probably occupies a broader temporal span than elaborate dentate-stamped pottery, and<br />
serves to link the “post-Lapita” Roviana intertidal sites to the Lapita ceramic series by<br />
homologous similarity <strong>of</strong> the lithic adze component <strong>of</strong> sites.<br />
<strong>The</strong> near-Oceania Lapita-horizon stone adze sample is now dominated by the items<br />
recovered from the Buka/Nissan reef sites and the Roviana intertidal sites. This suggests<br />
that further sampling <strong>of</strong> this settlement type holds the prospect <strong>of</strong> major dividends for this<br />
rare class <strong>of</strong> artefact, and that a large sample attained by this means can be expected to<br />
enable detailed sourcing and variability studies to proceed in the future, to an extent never<br />
before possible in Near-Oceania due to sample size constraints.<br />
419
420
Introduction:<br />
CHAPTER 11:<br />
INTRASITE SPATIAL STRUCTURE OF<br />
CERAMIC AND LITHIC VARIABILITY:<br />
Intrasite spatial structure is here defined as the patterned distribution <strong>of</strong> potsherds,<br />
artefacts, and/or manuports across sites (Wandsnider 1996). Intrasite spatial structure in<br />
the Roviana intertidal scatters is assumed to result from processes such as patterned<br />
discard (which may or may not have included contemporaneous differentiated activity<br />
areas), wave sorting <strong>of</strong> archaeological materials, selective trapping and preservation <strong>of</strong><br />
archaeological materials by site micro-topography and/or postdepositional pattern-<br />
alteration by humans.<br />
Site occupation span can be expected to have varied from site to site, and sites in<br />
favourable locations may have been occupied more than once. Even within a continuous<br />
occupation, it seems likely that the extent and locus <strong>of</strong> settlement may have changed over<br />
time, the latter either randomly, or possibly through unknown socially-determined<br />
mechanisms such as the avoidance <strong>of</strong> places formerly used by the deceased. Such processes<br />
should be expected to create, over the generations, chrono/stylistic spatial structure within<br />
sites. Ultimately the process <strong>of</strong> distinguishing whether stylistic spatial structure represents<br />
chronological change in style or contemporaneous diversity is one that occurs in dialogue<br />
with one’s ideas on ceramic classification and the nature <strong>of</strong> the stylistic series (Baxter<br />
1994:23). For a poorly-known sequence like the Roviana one, with no well-established<br />
fine-grained stylistic chronology, any conclusions must be regarded as preliminary working<br />
hypotheses.<br />
421
Review <strong>of</strong> Methods <strong>of</strong> Spatial Analysis:<br />
While intersite distributional mapping <strong>of</strong> ceramic styles or types within a geographic<br />
approach was common in European archaeology (e.g. Hodder & Orton 1976)and has been<br />
applied in the Pacific (Irwin 1985), ceramic intrasite spatial studies have been relatively<br />
rare in archaeology (Fontana 1998) compared to intersite or distributional studies, many<br />
examples being student dissertations ( e.g. Henderson 1998, Henderson 1992, Robertson<br />
2001). <strong>The</strong>re are notable exceptions to this tendency, some local (Best 1984:594-617,<br />
Cowgill et al. 1984, Fontana 1998, Sheppard & Green 1991).<br />
During the 1980s, debate in intrasite spatial analysis centred around spatial<br />
clustering and the methods by which archaeologists could identify spatially clustered<br />
patterns (Baxter 1994:23-24). LuAnn Wandsnider noted that emphasis has shifted in recent<br />
studies from a functional view <strong>of</strong> site structure in which activity patterning ought to be<br />
easily read and toolkits easily identified spatially (e.g Blankholm 1991), to a broader and<br />
more complex formational view <strong>of</strong> spatial structure, where the interaction <strong>of</strong> both cultural<br />
and natural formation processes are acknowledged as responsible for the structure <strong>of</strong><br />
archaeological deposits, and the identification <strong>of</strong> specific formational processes has become<br />
the goal (Wandsnider 1996). Congruent with this shift has been a shift in method from<br />
partitive typological approaches seeking to characterized artifact clusters within sites as<br />
functionally discrete to a more inclusive set <strong>of</strong> methods that examine distributions <strong>of</strong> a suite<br />
<strong>of</strong> formationally sensitive attributes <strong>of</strong> artefacts across sites. Although Wandsnider does<br />
not specifically make this point, it appears from the historical data presented that a shift in<br />
technique <strong>of</strong> analysis accompanied this trend, from computer identification <strong>of</strong> spatial<br />
clusters, to a recent state <strong>of</strong> practice in which it is more common to rely on visual<br />
inspection <strong>of</strong> density maps <strong>of</strong> formationally sensitive indicators (Wandsnider 1996:Table<br />
1).<br />
An analysis <strong>of</strong> intrasite stylistic spatial structure can be used to make inferences<br />
about changes in the extent/location <strong>of</strong> settlements over time. Most open sites are<br />
palimpsests <strong>of</strong> formation events and processes: in relation to intrasite patterning generally,<br />
422
Kintigh describes the problem <strong>of</strong> overlap or mixing <strong>of</strong> deposits as “perhaps the outstanding<br />
issue before (spatial archaeology)” (Kintigh 1990), and I suggest that the prospects for<br />
advance in this respect are best for the special case <strong>of</strong> ceramic archaeology, where<br />
relatively rich stylistic information can be expected to be a sensitive chronological<br />
indicator.<br />
Selection <strong>of</strong> Sites for Intrasite Spatial Analysis:<br />
Entire-sites were collected as single units during initial intertidal surface collections in early<br />
1996. By 1997, when intertidal survey was concentrated in the Kaliquongu region <strong>of</strong><br />
Roviana Lagoon, the aim was to try to document intrasite spatial structure for sites. Hence,<br />
the sites Zangana, Gharanga (2 nd collection), Honiavasa and Hoghoi (2 nd collection) were<br />
surface-collected in multiple transects. <strong>The</strong> exception was the Miho site, which was <strong>of</strong><br />
more limited extent, where significant along-shore spatial structure seemed unlikely.<br />
Intertidal ceramic/lithic scatters at H5, H6 and H7 were not surface collected due to time<br />
constraints and the small samples <strong>of</strong> sherds present. Largely aceramic intertidal lithic sites<br />
at Mbolave Island, H4, H7 and at Elelo/Mbanga were left undisturbed for the present.<br />
Objectives:<br />
Analysis <strong>of</strong> vessel completeness (Chapter 5) showed that only a small percentage <strong>of</strong> the<br />
deposited pottery was recovered, suggesting a harsh taphonomic regime for most sherds,<br />
resulting in the complete destruction <strong>of</strong> most vessels, and lucky survival in good condition<br />
for a minority <strong>of</strong> sherds from a minority <strong>of</strong> vessels. It seemed unlikely therefore that<br />
spatial distribution <strong>of</strong> sherds corresponding directly with activity areas (Schiffer 1995b)<br />
would have been preserved, although the possibility was not ruled out. Of more direct<br />
interest though, is the question <strong>of</strong> whether there were any broad-scale differences within<br />
423
sites in the distribution <strong>of</strong> fabric or style. Such differences might be expected to survive<br />
the taphonomic regime, and would be significant to chronological analysis <strong>of</strong> style, as sites<br />
could potentially be split into smaller seriation units on the basis <strong>of</strong> such distributional<br />
patterning, potentially lessening the problem <strong>of</strong> chronological mixing <strong>of</strong> styles within site<br />
assemblages.<br />
If wave-sorting <strong>of</strong> sherds has been a primary factor in creating spatial structure,<br />
small sherds should be found along the strandline, the upper intertidal, the swash zone,<br />
should be relatively free <strong>of</strong> sherds, and larger sherds should be found at the deeper margins<br />
<strong>of</strong> sites, in shallow water at low tide. <strong>The</strong> Zangana sherd size data was analysed in these<br />
terms. Furthermore, if wave-sorting <strong>of</strong> sites has been a primary factor in creating spatial<br />
structure, along-shore movement <strong>of</strong> smaller sherds would create a spread distribution <strong>of</strong><br />
small sherds in relation to large sherds, which latter should more nearly reflect the location<br />
and extent <strong>of</strong> primary discard. <strong>The</strong> along-shore linear nature <strong>of</strong> the Hoghoi ceramic scatter<br />
was considered an ideal vehicle for an analysis <strong>of</strong> this sort, where the narrow across-shore<br />
extent <strong>of</strong> the site precluded much in the way <strong>of</strong> analysis <strong>of</strong> across-shore sorting. For the<br />
Hoghoi site, lithic manuports were used in addition to ceramics to address these questions.<br />
<strong>The</strong> Honiavasa and Gharanga sites were collected in unequal-area transects. In the<br />
case <strong>of</strong> the Honiavasa site this was due to collection during a rising tide in deeper water<br />
and windy weather, where collection proceeded along five pace-measured transects<br />
between an anchored canoe and the shore, arranged radially in a fan formation around<br />
Honiavasa Point. <strong>The</strong> accuracy <strong>of</strong> measurement <strong>of</strong> area <strong>of</strong> the resulting transects is<br />
unknown, so relative measures <strong>of</strong> assemblage composition by transect are used wherever<br />
possible.<br />
One aim was to seek, through spatial structure, to untangle the sources <strong>of</strong><br />
decorative variability, and another was to take a distributional approach to sherd size,<br />
treating artefact fragments as sedimentary particles. While there may be some lag deposits<br />
<strong>of</strong> larger sherds that preserve behavioural or activity patterning, these also may be highly<br />
424
correlated with site microtopography and minimally correlated with specific activity areas.<br />
Method:<br />
A statistical approach was eschewed in favour <strong>of</strong> visual inspection <strong>of</strong> graduated symbol<br />
maps representing sherd counts or other quantification <strong>of</strong> cell compositional attributes,<br />
similar to the display used for ground visibility data in the Hvar Archaeological Survey in<br />
Yugoslavia (Gaffney et al. 1991). A difference from the Hvar displays that should be noted<br />
is that in the present study the dot diameter is related to the square root <strong>of</strong> the raw transect<br />
data, whereas in the Hvar survey a limited number <strong>of</strong> dot sizes was used to display data<br />
that had been converted to ranges or categories.<br />
An advantage <strong>of</strong> this graduated dot approach over isopleth displays is that the<br />
displayed data (symbol sizes) are derived fairly directly from the raw data (sherd counts<br />
or sums <strong>of</strong> sherd areas by cell), making a visual assessment <strong>of</strong> sample size effects by<br />
transect possible. An advantage <strong>of</strong> this type <strong>of</strong> display over choropleth displays was that<br />
the elongated transect shapes <strong>of</strong> the field collection strategy would have been distracting<br />
to the eye had these displays been used, whereas size-graduated circular symbols created<br />
a directionally-unbiased visual effect. Were frequency data or ratio data displayed instead,<br />
significant sample size information would be less accessible.<br />
Sherds were coded by numbered collection unit for intra-site spatial analysis in<br />
table “Flat.db” as follows:<br />
425
Paniavile (site 1): 1=cave assemblage collected prior to ‘89?<br />
2=Jan 1996 colln.<br />
Gharanga: (site 5) 1=Jan 96<br />
3=August 96 colln.<br />
2=G west <strong>of</strong> stream 98<br />
3=G East <strong>of</strong> stream 98<br />
Hoghoi (Site 2) 1-21=5m wide transects<br />
99=initial surface collection by Kenneth Roga August 97<br />
130=hg13t(subsurface test)<br />
150=hg15a(additional)<br />
151=hg15s(subsurface test)<br />
152=hg15t (another test)<br />
Honiavasa (Site 4) 1-5=radial transects 1-5<br />
Kopo (Site 7) 1=single collection unit<br />
Miho (Site 3) 1=single collection unit<br />
Nusa Roviana (Site 6) 1=single collection unit<br />
Zangana (Site 8) 1-105= transect units as per Figure 117(collected August97)<br />
110=initial surface collection (June97) <strong>of</strong> Zangana<br />
111=last collection (August 1997) <strong>of</strong> Zangana<br />
Digital site maps were drawn for Sites 2, 4 and 8 (Hoghoi, Honiavasa and Zangana) from<br />
field notes and tape/compass maps using AUTOCAD drafting s<strong>of</strong>tware. <strong>The</strong>se maps were<br />
linked to ceramic thematic data tables using ArcInfo and MapInfo programs. MapInfo was<br />
used to create site graduated dot thematic maps.<br />
426
Figure 117: Intertidal collection units at Zangana: numbers are values in “unit” column<br />
in table “Flat.db” appended on CD. Units without numbers are those which yielded no<br />
sherds.<br />
427
Zangana:<br />
Natural Formation Processes at Zangana:<br />
Across shore size sorting. <strong>The</strong> question: can I demonstrate that appreciable numbers <strong>of</strong><br />
sherds have moved, and if some sherds may not have moved much, can this be shown?<br />
It would be unsatisfactory to analyse the Zangana materials purely in term <strong>of</strong> distance from<br />
shore in relation to sherd size, for two reasons: firstly, the shoreline has undergone<br />
substantial modification during WWII as the site <strong>of</strong> a major American landing on New<br />
Georgia, and more recently during refurbishment <strong>of</strong> the road and wharf circa 1988;<br />
secondly, the topography <strong>of</strong> the intertidal zone varies significantly between the North and<br />
South <strong>of</strong> the site.<br />
Zangana Sherd Count<br />
All temper classes<br />
80<br />
40<br />
8<br />
New Georgia Mainland Roviana Lagoon<br />
428<br />
35m<br />
�<br />
Figure 118: Spatial distribution <strong>of</strong> the sherd sample at<br />
Zangana.
<strong>The</strong> dense cluster <strong>of</strong> ceramic sherds at the north end <strong>of</strong> the site (Figure 118) occur<br />
where the intertidal is characterized by relative microtopographic roughness, and overall<br />
narrowness <strong>of</strong> the intertidal and shallow subtidal (around 15m from the high water mark<br />
to the lower limit <strong>of</strong> the intertidal). This microtopographic roughness is thought to have<br />
acted as a sherd trap and shield, and supplies the primary explanation for the preservation<br />
<strong>of</strong> sherds in this area. By contrast, the intertidal zone south <strong>of</strong> the wharf is generally about<br />
30m wide during the low tides, with sherds tending to be found in the recesses <strong>of</strong> a more<br />
gentle microterrain.<br />
North <strong>of</strong> the Zangana wharf there are two clusters <strong>of</strong> transects with reasonable<br />
sample sizes, interspersed by an area <strong>of</strong> s<strong>of</strong>t sediment most likely deposited as a result <strong>of</strong><br />
the sheltered conditions from the SE tradewind and southerly sea breezes created by the<br />
earthen wharf. South <strong>of</strong> the wharf there is a band <strong>of</strong> transects along the shoreline which<br />
have marginal samples <strong>of</strong> around eight sherds each, and a zone or band towards the lower<br />
limit <strong>of</strong> the intertidal zone which includes some slightly larger samples, comprising three<br />
groups <strong>of</strong> transects.<br />
Zangana Average Sherd Area<br />
(cm2)<br />
110<br />
55<br />
11<br />
New Georgia Mainland Roviana Lagoon<br />
429<br />
35m<br />
�<br />
Figure 119: Across-shore size sorting at Zangana.
<strong>The</strong> larger sherds are along the seaward margins <strong>of</strong> the intertidal zone in most cases<br />
(Figure 119), suggesting selective removal <strong>of</strong> smaller sherds and probably also <strong>of</strong> sherds<br />
not fortuitously sheltered/trapped by microtopography. This pattern suggests the seaward<br />
margin <strong>of</strong> Zangana South is a lag deposit, while the landward margin is a secondary<br />
accumulation <strong>of</strong> wave-transported sherds. <strong>The</strong> large sherd counts from transects at the<br />
north end <strong>of</strong> the site comprise mainly small sherds, with a couple <strong>of</strong> larger averages<br />
occurring only in transects with small samples. <strong>The</strong>re is a suggestion though <strong>of</strong> occasional<br />
larger sherds around the seaward margins <strong>of</strong> this area also.<br />
Sherds immediately to the north <strong>of</strong> the wharf are all small, and it is here that WWII<br />
and subsequent effects are probably greatest, with trampling and recent sherd breakage<br />
likely to be extreme due to modern foot traffic; also, accumulation <strong>of</strong> s<strong>of</strong>t sediments<br />
following rebuilding <strong>of</strong> the wharf in the 1990s has reduced archaeological visibility here.<br />
<strong>The</strong> more extensive tidal flat to the south <strong>of</strong> the wharf displays a more coherent<br />
pattern <strong>of</strong> size sorting across the shoreline, with all <strong>of</strong> the strandline sherd samples having<br />
average sherd sizes <strong>of</strong> less than 8cm 2 . All this supports the intuitive impression formed<br />
during field collection that the strandline comprised almost exclusively small, water<br />
rounded sherds, while the seaward limits <strong>of</strong> the intertidal zone yielded larger sherds with<br />
little evidence <strong>of</strong> rolling damage, and only minor sand abrasion in many cases, marine<br />
growth probably being a factor in prevention <strong>of</strong> the latter.<br />
A further question: was there a smaller locus or multiple loci <strong>of</strong> settlement in the<br />
past that has been spread into the present 3000m 2 distribution by wave-transport <strong>of</strong><br />
sherds? It seems clear from the above that wave processes have created a pattern <strong>of</strong> small<br />
sherds along the strandline which are relatively mobile and larger sherds further seaward<br />
that are less so. Is there a spread <strong>of</strong> small sherds and a lag deposit <strong>of</strong> larger sherds? Or<br />
perhaps multiple spaced lag deposits <strong>of</strong> larger sherds interspersed with small sherds<br />
representing spread from multiple occupation loci? This latter possibility seems borne out<br />
to some extent by the distribution in Figure 119, particularly to the south <strong>of</strong> the wharf. A<br />
band <strong>of</strong> larger sherds is distributed sub-parallel to the shoreline, possibly following a pre-<br />
bulldozer shoreline around what would have been the less prominent prehistoric Zangana<br />
point. Whether clusters in this distribution are a random sample size effect or a<br />
taphonomic effect relating to microtopographic variation are questions for future research<br />
430
on these sites. A third explanation has not been ruled out, that these clusters <strong>of</strong> larger<br />
sherds forming a line along the shore are a settlement pattern, a line <strong>of</strong> discard zones<br />
surrounding house locations.<br />
If sherd size and quantity are combined by summing sherd areas by transect, giving<br />
weight to those transects having the greatest total surface area <strong>of</strong> pottery, a picture that<br />
could be interpreted in this way emerges (Figure 120). South <strong>of</strong> the wharf, there are a line<br />
<strong>of</strong> four holes in the distribution <strong>of</strong> potsherds which might result from a line <strong>of</strong> structures<br />
surrounded by throw zones for broken pottery. A test <strong>of</strong> this idea would be to examine<br />
whether these gaps could be random, and whether they correlate with microtopographic<br />
variation, but these tests will not be attempted here.<br />
Zangana Sum <strong>of</strong> Sherd Areas<br />
(cm2)<br />
620<br />
310<br />
62<br />
New Georgia Mainland Roviana Lagoon<br />
431<br />
35m<br />
�<br />
Figure 120: Possible linear settlement patterning in the<br />
distribution <strong>of</strong> large sherds at Zangana south.
If this distribution does represent a settlement pattern, this would imply either long-<br />
duration stability in settlement layout, or short duration <strong>of</strong> occupation, as, were one <strong>of</strong><br />
these conditions not the case, such patterning would be obscured by overprinting.<br />
Spatial Structure in Ceramic Style at Zangana:<br />
<strong>The</strong> above discussion established that post-depositional taphonomic effects have probably<br />
obscured most small-scale settlement patterning, with the possible exception <strong>of</strong> the<br />
Figure 121: Initial point-provenanced collection <strong>of</strong><br />
decorated sherds at Zangana (subsequent collection<br />
transects shown for spatial reference).<br />
southern-seaward part <strong>of</strong> the site. Initial surface collection <strong>of</strong> a small number <strong>of</strong> sherds at<br />
Zangana targeted only decorated sherds. Point locations <strong>of</strong> sherds were recorded.(Figure<br />
121)<br />
432
<strong>The</strong>se data seemed to suggest Miho styles (Class 2 decoration and pinched band at the<br />
neck were found mainly to the south <strong>of</strong> the wharf, and Gharanga/Kopo styles (punctate<br />
band) were found to the north <strong>of</strong> the wharf. A single sherd with a Honiavasa style Class<br />
1 bounded linear incised motif was found to the north <strong>of</strong> the wharf. <strong>The</strong> subsequent, more<br />
comprehensive, 1/5 sample transect surface collection enabled stylistic spatial structure to<br />
be established with more confidence. <strong>The</strong> distribution <strong>of</strong> ceramic styles at Zangana was<br />
compared to the overall quantity distribution. Wavy-lipped sherds and Gharanga/Kopo<br />
multi-band fingernail-pinched/punctate styles were distributed as expected from the<br />
distribution <strong>of</strong> the overall sherd sample (Figure 122, Figure 123).<br />
Deformation <strong>of</strong> the lip into a wave form<br />
(Sherd count)<br />
10<br />
5<br />
1<br />
New Georgia Mainland Roviana Lagoon<br />
Class 1 linear motif incised/fingernail-pinched sherds showed a different distribution to<br />
433<br />
35m<br />
�<br />
Figure 122: Lip deformation into a wave present in both<br />
Zangana-North and Zangana-South.
Band <strong>of</strong> punctation/multiple bands pinching<br />
(sherd count)<br />
10<br />
5<br />
1<br />
New Georgia Mainland Roviana Lagoon<br />
434<br />
35m<br />
�<br />
Figure 123: Bands <strong>of</strong> punctation were restricted to Zangana-North.<br />
the one expected from the overall sherd distribution, with sherds <strong>of</strong> this style being found<br />
mainly to the south <strong>of</strong> the wharf (Figure 124).<br />
<strong>The</strong> pattern expected from the overall distribution <strong>of</strong> all sherds would be for the<br />
bulk <strong>of</strong> Miho-style sherds to be found at the north end <strong>of</strong> the site, with only a few sherds<br />
found south <strong>of</strong> the wharf. Instead, 24 Miho sherds were found either near the wharf or on<br />
the reef flat to the south, with only four found at the north end if the site, where the<br />
majority <strong>of</strong> sherds were found. <strong>The</strong> chances <strong>of</strong> this happening randomly have not been<br />
calculated, but would be very slim.
Figure 124: Unbounded linear incision or necks banded with<br />
pinching at Zangana.<br />
Two alternate explanations for this patterning were considered, firstly that the Miho<br />
Class1/pinch style, the Miho “wav” lip deformation style, and the Gharanga/Kopo<br />
punctate/multi-band-pinch styles form a time-series, and are spatially segregated in this<br />
instance by a shift or expansion/contraction <strong>of</strong> the locus <strong>of</strong> settlement over time (possibly<br />
in the reverse order to that listed); secondly, that this patterning represents<br />
contemporaneous, spatially segregated stylistic diversity, interpretable as diversity <strong>of</strong><br />
manufacturing traditions or diversity <strong>of</strong> activity patterning. This second explanation seems<br />
unlikely, as the scale <strong>of</strong> the spatial strucure is too large to be activity patterning, and even<br />
if one end <strong>of</strong> a village were producing different pottery discard is unlikely to be so neatly<br />
structured.<br />
435
Zangana Temper Classes<br />
(sherd Count)<br />
70<br />
35<br />
7<br />
Quartz-calcite hybrid<br />
Placered volcanic<br />
Unplacered feldspathic<br />
New Georgia Mainland Roviana Lagoon<br />
436<br />
35m<br />
�<br />
Figure 125: Distribution <strong>of</strong> temper classes at Zangana.<br />
Distribution <strong>of</strong> Zangana Pottery Tempers:<br />
Temper class sherd counts did not show strong patterning across the site (Figure 125).<br />
Sherds tempered with unplacered volcanic/feldspathic/hornblendic stream sands are more<br />
common in the northern part <strong>of</strong> the site (see an enlarged detail <strong>of</strong> Figure 125 in Figure<br />
126). Quartz-calcite tempered sherds do not seem to have any particular departure from<br />
the pattern expected from overall sherd quantities, suggesting that this anomalous temper<br />
group is not highly spatially correlated with any <strong>of</strong> the decorative attributes. This suggests<br />
that, if the Zangana stylistic spatial structure represents a time series <strong>of</strong> styles, the quartz-<br />
calcite temper transcends this series, occurring throughout.
Zangana Temper Classes<br />
(sherd Count)<br />
100<br />
50<br />
10<br />
Unplacered feldspathic<br />
Placered volcanic<br />
Quartz-calcite hybrid<br />
Figure 126: Detail <strong>of</strong> distribution <strong>of</strong> temper classes at Zangana-<br />
North.<br />
Zangana Summary:<br />
In summary, at Zangana the pottery distribution shows evidence <strong>of</strong> wave size-sorting,<br />
where lag deposits <strong>of</strong> larger sherds are mainly found at the seaward margin <strong>of</strong> Zangana<br />
South. <strong>The</strong>re may be some settlement patterning in this part <strong>of</strong> the site, but the possibility<br />
<strong>of</strong> correlation with mictotopographic variation has yet to be examined, and random<br />
distribution might look something like the pattern observed. Stylistically, there is spatial<br />
structure at Zangana, with the Miho style having a different distribution to other styles.<br />
This may indicate a shifting locus <strong>of</strong> settlement across a ceramic time-series, or alternately<br />
a contemporaneous expression <strong>of</strong> manufacturing diversity across a settlement. None <strong>of</strong> the<br />
major temper types exhibit any clear spatial structure, although there is a (statistically<br />
untested) suggestion <strong>of</strong> a negative correlation between Miho style ceramics and unplacered<br />
volcanic temper.<br />
437
Hoghoi:<br />
Questions <strong>of</strong> taphonomy and other site formation questions at Hoghoi were addressed by<br />
mapping data from the second surface collection, which comprised a linear along-shore<br />
arrangement <strong>of</strong> 22 contiguous intertidal transects, each 5x12m, oriented with the long axis<br />
at right angles to the shoreline (the first surface collection the previous year had been a<br />
grab sample <strong>of</strong> 105 items, mostly decorated pottery).<br />
Natural Formation Processes at Hoghoi: Along-shore Size Sorting:<br />
<strong>The</strong> question: was there a smaller site in the past that has been spread into the present<br />
110m distribution by sherd/manuport transport? This was investigated by looking at the<br />
average size <strong>of</strong> sherds. Is there a spread <strong>of</strong> small sherds and a lag deposit <strong>of</strong> larger sherds?<br />
Or perhaps multiple spaced lag deposits <strong>of</strong> larger sherds interspersed with small sherds<br />
representing spread from multiple occupation loci? <strong>The</strong> overall sherd count per transect<br />
identifies those transects where sherd sample size is sufficient to address this question<br />
(Figure 127).<br />
Transect sample sizes<br />
(sherd count)<br />
150<br />
75<br />
15<br />
0 0 12 24 12 5 3 4 64 149 95 67 39 23 66 9 28 23 46 31 10 0<br />
0-5m 15-20m 30-35m 45-50m 60-65m 75-80m 90-95m 105-110<br />
Figure 127: Distribution <strong>of</strong> the sherd sample at Hoghoi.<br />
Transects 0 through 10m, 25 through 40m, 75-80m and 100 through 110m were excluded<br />
as having too small a sample <strong>of</strong> sherds to be reliable indicators <strong>of</strong> along-shore size-sorting<br />
processes.<br />
438<br />
�<br />
<strong>The</strong> fairly large sherd count at 70-75m is immediately adjacent to a transect with
a very low sherd count thought during field collection to have been the subject <strong>of</strong> coral<br />
boulder clearance for a copra-loading canoe-landing (the 70-75m transect appears to have<br />
been the recipient <strong>of</strong> the coral cleared from the 75-80m transect. It can be seen in figure<br />
12 that the average sherd area for these 66 sherds is noticeable smaller than for<br />
neighbouring transects, suggesting many <strong>of</strong> these sherds may have resulted from trampling<br />
and loss <strong>of</strong> wave-shelter in the “cleared” transect immediately to the west.<br />
Average sherd area<br />
(cm2, showing sample sizes)<br />
0-5m<br />
30<br />
15<br />
3<br />
0 12 24 12 5 3 4 64 149 95 67 39 23 66 9 28 23 46 31 10 0<br />
15-20m 30-35m 45-50m 60-65m 75-80m 90-95m 105-110<br />
Figure 128: Lack <strong>of</strong> sherd size variation at Hoghoi, except at 35-40m (n=4).<br />
439<br />
�<br />
<strong>The</strong>re is little evidence in Figure 128 that the site comprises concentrated lag-deposits <strong>of</strong><br />
large sherds flanked by transects containing only smaller sherds. <strong>The</strong> 45-50m sample <strong>of</strong><br />
149 sherds, which is the most dense scatter <strong>of</strong> sherds, is not composed <strong>of</strong> unusually large<br />
sherds, but the 50-60m transects have very slightly larger than average sherd area and<br />
reasonably high density <strong>of</strong> deposits.<br />
Lithics at Hoghoi:<br />
Data on lithic manuports, many <strong>of</strong> which appear to be ovenstones, was mapped to further<br />
investigate the evidence for along-shore size sorting. Lithic manuports tended to be larger<br />
on average between the 25m mark and 60m, in the region where sherd density is highest<br />
(Figure 129).
Average weight <strong>of</strong> Hoghoi Manuports (g)<br />
Sample count shown<br />
200<br />
100<br />
20<br />
3 3 3 11 0 8 20 9 44 53 11 29 21 12 14 8 30 12 19 23 7 4<br />
0-5m 20-25m 40-45m 60-65m 80-85m 100-105m<br />
Figure 129: Larger average manuport mass from 25m to 60m at Hoghoi.<br />
Of the smaller lithic manuports, many were unfractured water rounded cobbles, prompting<br />
speculation that these might have some non-cooking function such as net weights or sling-<br />
stones. Distribution by form and size classes is mapped in Figure 130.<br />
Hoghoi Manuport Forms/Sizes<br />
(Quantified by Count)<br />
40<br />
20<br />
4<br />
Rounded stones
showing little evidence <strong>of</strong> the strong clustering at 45-50m in evidence in the ceramic<br />
distribution, or the distribution <strong>of</strong> larger stones.<br />
It seemed reasonable to conclude that these smaller rounded stones were more<br />
transportable than ceramic sherds, and thus had been spread by wave action, obliterating<br />
any activity patterning. A lithic manuport type series, established through petrographic and<br />
macroscopic classification, identified several geological source groups within the<br />
ovenstone assemblage (Chapter 10). When manuports were mapped according to source<br />
group (Figure 131), Classes 3, 5 and 18 dominate at the eastern end <strong>of</strong> the site.<br />
Hoghoi Manuport Classes<br />
Quantified By Count<br />
30<br />
15<br />
Class 1<br />
Class 2<br />
Class 3<br />
Class 5<br />
Class 18<br />
3<br />
0-5m 20-25m 40-45m 60-65m 80-85m 100-105m<br />
Figure 131: Size-sorting <strong>of</strong> manuport petrographic classes at Hoghoi, suggestive<br />
<strong>of</strong> different size-procurement patterns by source.<br />
In addition to the size sorting effect there is also a possible difference in procurement<br />
pattern by source, with smaller stones being selected from some sources than others.<br />
Differences in the size <strong>of</strong> stones available or selected from different sources is likely to<br />
have produced the patterning in Figure 131. <strong>The</strong> only large Class 18 stone (1100g) was<br />
found in unit 10 (45-50m), <strong>of</strong>fering further support for a size-sorting wave-transport<br />
explanation for spatial structure at Hoghoi, although re-use <strong>of</strong> materials with remaining use<br />
life from an earlier, more easterly occupation cannot be ruled out.<br />
441
Distribution <strong>of</strong> Hoghoi Ceramic Tempers:<br />
Ceramic tempers <strong>of</strong>fer further evidence relating to the question <strong>of</strong> size/density sorting by<br />
waves as opposed to along-shore behavioural spatial structure (Figure 132). Quart-calcite<br />
and unplacered tempers are slightly more common from 65-80m than elsewhere, but since<br />
these tempers may have lower specific gravity than placered volcanic tempers with a high<br />
ferro-magnesian component, density sorting by along-shore wave processes could explain<br />
this patterning (but analysis showed no difference in specific gravity by temper).<br />
Distribution <strong>of</strong> pottery tempers<br />
(Quantified by sherd count)<br />
130<br />
65<br />
13<br />
Quartz-calcite hybrid temper<br />
Unplacered lithic-feldspathic<br />
Placered volcanic<br />
0-5m 20-25m 40-45m 60-65m 80-85m 100-105m<br />
Figure 132: Distribution <strong>of</strong> potsherd temper classes at Hoghoi.<br />
This analysis <strong>of</strong> the distribution <strong>of</strong> temper types suffers from sample size problems,<br />
and highlights the need to make the best possible use <strong>of</strong> the samples that have survived the<br />
last two-to-three millennia. <strong>The</strong> hundred or so larger sherds uplifted during the initial<br />
collection, together with the adzes collected at the same time, would have alleviated this<br />
problem to some extent.<br />
Distribution <strong>of</strong> Ceramic Styles at Hoghoi:<br />
<strong>The</strong> collection <strong>of</strong> a hundred decorated sherds or lithic artefacts from this site on first<br />
discovery has had a greater adverse effect on the stylistic component <strong>of</strong> this spatial analysis<br />
than on the manuport or temper analyses. This highlights the need for careful spatial<br />
recording on these fragile sites. <strong>The</strong>re is so little spatial stylistic information from the<br />
second collection at Hoghoi that the information was not mapped, and some general<br />
comments only are made.<br />
<strong>The</strong>re seems to be proportionately more deformation <strong>of</strong> the lip into a wave form<br />
442
and Class 2 linear motif on the rim or shoulder between 80m and 95m than between 40<br />
and 65m, especially at 80-85m,although numbers are so small that this could easily be<br />
sampling error or quantification error, but punctate band/multiple bands <strong>of</strong> pinching are<br />
found right along the site, albeit in small numbers ( peaking at six instances <strong>of</strong> punctate<br />
band at 45-50m).<br />
Hoghoi Summary:<br />
Sherd quantities suggest a strongly clustered distribution <strong>of</strong> potsherds in the vicinity <strong>of</strong> the<br />
45-50m transect, with a variable density across other transects. <strong>The</strong> largest mean sherd<br />
sizes, by a very small margin, were around the 45-50m transect, suggesting along shore<br />
size-sorting is operative to a slight extent, and that there is a slight lag effect in the vicinity<br />
<strong>of</strong> these larger (on average) sherds. <strong>The</strong> spatial distribution <strong>of</strong> lithic manuports also<br />
suggested along-shore size sorting (somewhat surprisingly in view <strong>of</strong> the higher specific<br />
gravity <strong>of</strong> the lithic manuports, but perhaps the flat shape <strong>of</strong> sherds makes them less subject<br />
to transport by rolling). Compositional differences and possible functional differences<br />
between manuport samples from the main 45-50m sherd cluster and the western sherd<br />
distribution are most likely a result <strong>of</strong> size-sorting by waves, added onto a difference by<br />
source in procurement strategies (small stones were taken from the Class 18 source for<br />
some reason).<br />
Initial collection <strong>of</strong> a hundred or so larger sherds from the site prior to more<br />
detailed transect collection made a spatial analysis <strong>of</strong> stylistic structure impossible, as these<br />
were the bulk <strong>of</strong> the sherds diagnostic <strong>of</strong> the styles defined in Chapter 9. This highlights<br />
the need to record the locations <strong>of</strong> items carefully for this fragile site type, which is easily<br />
rendered uninformative by casual surface collection. This finding is important for future<br />
resource management and research.<br />
Honiavasa:<br />
443
Sherds were collected in five pace-measured transects at Honiavasa, so the site maps on<br />
which the spatial analysis which follows are based are only roughly to scale, and transect<br />
areas may in practice have been unequal. Consequently, sherd counts cannot be translated<br />
into sherd densities, but relative frequencies <strong>of</strong> sherd attributes and taphonomic measures<br />
such as average sherd sizes can be used with confidence. Overall sherd sample sizes are<br />
reasonable for all five transects when looking at the distribution <strong>of</strong> sherd size and<br />
composition (Figure 133).<br />
Sherd Count<br />
110<br />
55<br />
11<br />
Honiavasa Passage<br />
� 35m<br />
444<br />
113<br />
98<br />
45<br />
76<br />
110<br />
Honiavasa Point<br />
Figure 133: Distribution <strong>of</strong> the sherd sample at Honiavasa.<br />
Taphonomy:<br />
Honiavasa has the largest mean sherd size <strong>of</strong> all sites in the study. <strong>The</strong> site had not<br />
previously been discovered by archaeologists. Honiavasa is located opposite the modern<br />
village <strong>of</strong> Sasavele, across the reef passage, and is not subject to intensive foot traffic or<br />
intensive children’s play activities. <strong>The</strong> site as collected was largely subtidal in water up<br />
to a metre deep at low water on 20 September 1997, and may extend into deeper water.<br />
This means that for most <strong>of</strong> the year the site is well submerged and sherds are seldom<br />
subject to high-energy wave transport, despite the reef-passage location. Sherd size<br />
increases slightly with increasing depth <strong>of</strong> water from east to west (Figure 134). <strong>The</strong><br />
centre transect is that most exposed to wave -induced sherd damage, as it is here the reef<br />
flat proper begins, and here that refracted swell curling around from the passage to the east<br />
breaks at low tide. This is reflected in sherd size, although low collection intensity in this<br />
transect due to a rising tide may have been a factor.
Average sherd area (cm2)<br />
33<br />
16.5<br />
3.3<br />
Honiavasa Passage<br />
� 35m<br />
445<br />
Honiavasa Point<br />
Figure 134: Effects <strong>of</strong> wave refraction (and collection intensity?) on sherd<br />
size at Honiavasa: the western margin is exposed to waves from Honiavasa<br />
channel, which expend their energy in a swash zone at about the centre <strong>of</strong><br />
the site at low tide.<br />
Distribution <strong>of</strong> Temper Types:<br />
Transect 2 (2 nd from right, n = 76) shows an increased frequency <strong>of</strong> placered volcanic<br />
/unplacered pyroxenic tempers (Figure 135), otherwise the relative proportions <strong>of</strong> the<br />
three major temper classes are reasonably consistent across the site. (Placered volcanic<br />
with volcanic rock fragments is a subgroup <strong>of</strong> the placered volcanic group, and difficult<br />
to distinguish reliably without petrographic analysis. Variations in the relative frequencies<br />
in which this temper was identified may have more to do with variations in surface etching<br />
resulting from cleaning than with behaviour in the past, hence shown in the same shade as<br />
placered volcanic temper.)<br />
Temper Types<br />
110<br />
55<br />
11<br />
Placered volcanic temper<br />
Pv with V.rock fragments<br />
Unplacered lithic/feldspathic<br />
Quartz-calcite<br />
Roviana Lagoon<br />
�<br />
Honiavasa Point<br />
0 30m<br />
Figure 135: Distribution <strong>of</strong> pottery tempers at Honiavasa.
Figure 136: Co-joining sherds from deeper western margin <strong>of</strong> Honiavasa..<br />
<strong>The</strong>re is a pattern <strong>of</strong> slightly increased incidence <strong>of</strong> quartz-calcite tempers at the eastern<br />
and western margins <strong>of</strong> the site. Analysis <strong>of</strong> vessel completeness did show, however, that<br />
the western margins <strong>of</strong> the site yielded larger sherds, and one particularly complete vessel<br />
including the only two co-joining sherds recovered during the entire study (Figure 136),<br />
so perhaps taphonomic variation should not be hastily discounted in explaining variation<br />
in the distribution <strong>of</strong> Quartz-calcite tempered sherds. <strong>The</strong> same could hold for unplacered<br />
lithic/feldspathic sherds, which might be slightly weaker, hence smaller and more easily<br />
transported than placered volcanic sherds.<br />
Distribution <strong>of</strong> Decorative Attributes at Honiavasa:<br />
Due to the relative plain-ness <strong>of</strong> the Honiavasa pottery, analysis <strong>of</strong> the spatial distribution<br />
<strong>of</strong> styles at Honiavasa is faced with severe sample size problems (Figure 137). <strong>The</strong><br />
relatively high frequency <strong>of</strong> notched lips <strong>of</strong> two types, and relatively low frequency <strong>of</strong><br />
carinations/Class 1 linear motifs at the western end <strong>of</strong> the site may be a sampling error,<br />
446
Distribution <strong>of</strong> decorative attributes<br />
(Quantified by sherd count)<br />
30<br />
15 3<br />
Band <strong>of</strong> impression on the top face <strong>of</strong> the lip, or on the outer edge<br />
Carinated vessels or Class 1 linear motifs<br />
Band <strong>of</strong> fingernail pinching on the outer edge <strong>of</strong> the lip<br />
Roviana Lagoon<br />
�<br />
Honiavasa Point<br />
0 30m<br />
Figure 137: Distribution <strong>of</strong> pottery decorative attributes at<br />
Honiavasa.<br />
especially in view <strong>of</strong> higher vessel completeness in transect 5.<br />
Honiavasa Summary:<br />
While sherd size distribution suggests some taphonomic variation across the site consistent<br />
with field observation <strong>of</strong> site morphology and wave exposure, the data on Honiavasa<br />
temper distribution suggests that mean sherd size as a taphonomic indicator may mask<br />
additional significant taphonomic variation. Recovery <strong>of</strong> two large cojoining sherds from<br />
transect five at the deeper western margin <strong>of</strong> the site alongside Honiavasa passage provides<br />
a clue that taphonomic degradation has been less severe in this part <strong>of</strong> the site. If<br />
preservation is better at transect 5 this may indicate that the relatively large sample <strong>of</strong><br />
sherds from that deeper location actually represents a relatively small number <strong>of</strong> vessels<br />
in a more complete state, a biased sample from which several diagnostic pieces may derive<br />
from a single vessel.<br />
Chapter Summary and Conclusions.<br />
447
In drawing together the information from analyses <strong>of</strong> these three sites, the first issue is<br />
whether an inter-site comparison <strong>of</strong> any <strong>of</strong> the themes <strong>of</strong> these analyses, size distribution,<br />
temper distribution, or stylistic distribution, can be informative. A question in this regard<br />
mentioned above in relation to Zangana stylistic distribution was whether a comparison <strong>of</strong><br />
the distribution <strong>of</strong> styles between sites would show up regularities in such patterning, or<br />
whether these might be considered as random variation from site to site. A further question<br />
for intersite comparisons that arose in relation to a structured distribution <strong>of</strong> tempers at<br />
Hoghoi was whether the anomalous quartz-calcite temper group separated out spatially<br />
at all in other sites. <strong>The</strong> Honiavasa site analysis raised questions about the methods used<br />
to assess intrasite taphonomic variation, the significance <strong>of</strong> which can now be discussed<br />
in relation to other sites.<br />
Class 2 Linear Motifs and Neck Pinching:<br />
<strong>The</strong>re was a clear pattern <strong>of</strong> spatial segregation <strong>of</strong> Class 2 linear motifs or neck pinching<br />
(Miho style) from punctate bands/multi-band pinching (Gharanga/Kopo style) at Zangana,<br />
which justifies Zangana South being treated separately to Zangana North in seriation<br />
(Chapter 12). At Hoghoi, no firm conclusion <strong>of</strong> this nature could be drawn, and these<br />
styles do not occur at Honiavasa. Of the two other sites where the Miho style occurs there<br />
is no repetition <strong>of</strong> the pattern observed at Zangana, so the hypothesis that this spatial<br />
structure represents a contemporaneous socio-ceramic spatial structure is rejected. This<br />
leaves the competing alternative hypothesis, that the Miho style and the Gharanga/Kopo<br />
styles form a time-series, and are spatially segregated in this instance by a shift or<br />
expansion/contraction in the locus <strong>of</strong> settlement over time. This conclusion is significant<br />
for seriations in Chapter 12.<br />
Spatial Separation <strong>of</strong> Quartz-Calcite Temper:<br />
At Hoghoi, the distribution <strong>of</strong> quartz-calcite temper showed some structure, but this did<br />
not seem to be the case at Zangana. At Honiavasa there was a hint <strong>of</strong> similar structure, and<br />
448
in neither case Honiavasa or Hoghoi) can taphonomic sorting or simply small-sample error<br />
be ruled out as an explanation.<br />
Implications for Resource Management and Future Research:<br />
Spatial structure in collection sites has been shown to be key to interpreting these sites.<br />
Wave sorting processes and swash-zone taphonomic effects can be seen in all three sites<br />
analysed in this <strong>chapter</strong>. Also, the difficulty in analysing stylistic spatial structure at Hoghoi<br />
was increased by removal <strong>of</strong> an initial grab sample. Similar sampling strategies at Paniavile<br />
by a series <strong>of</strong> archaeological teams and by locals meant that no spatial analysis could be<br />
attempted there.<br />
It is clear that future research and resource management <strong>of</strong> such surface sites can<br />
usefully apply detailed spatial recording, to the level <strong>of</strong> point-provenancing <strong>of</strong> artefacts and<br />
recording <strong>of</strong> micro-topography, otherwise formation processes <strong>of</strong> these scatters <strong>of</strong> pottery<br />
are unlikely to be understood. It seems unlikely to me that some <strong>of</strong> the fundamental<br />
difficulties in seriating surface collections, such as unraveling temporal mixing from<br />
different periods, and distinguishing activity patterning from taphonomic spatial sorting or<br />
microtopographic trapping, can be resolved satisfactorarily unless such recording is done<br />
carefully.<br />
449
450
Introduction:<br />
CHAPTER 12:<br />
CHRONOLOGY<br />
In this <strong>chapter</strong> information from thermoluminescence, radiocarbon determinations, and<br />
ceramic seriation is presented, followed by a discussion which attempts to draw this<br />
diverse information into a provisional chronology. Implications for future research and<br />
resource management are also discussed.<br />
Seriation measures time as a continuum <strong>of</strong> gradual change, in that rates <strong>of</strong> change<br />
cannot be measured by seriation. In seriation chronological types or attributes are treated<br />
as arbitrary units <strong>of</strong> measurement <strong>of</strong> the process <strong>of</strong> change rather than as essential classes<br />
in the data, allowing change over time between one type and another (O'Brien & Lyman<br />
2000:52-56). <strong>The</strong> seriator is not interested in establishing fixed boundaries <strong>of</strong> a type, for<br />
example, whether something is Lapita or not, but rather in discovering which aspects <strong>of</strong><br />
artefact variability can be used to construct a coherent evolutionary series with which to<br />
tell time. By contrast, in using absolute dates one is usually (but not always) concerned to<br />
know which chunk <strong>of</strong> time is being dated (which event). Paradoxically, although<br />
radiocarbon age is a continuous variable, applications <strong>of</strong> the technique seek to date<br />
occupations, layers, horizons, and thus seek a discontinuous measure <strong>of</strong> time (for an<br />
extended discussion <strong>of</strong> this see O'Brien & Lyman 2000). “Is this Lapita? How old is type<br />
X? When was this stratigraphic unit deposited? When does Lapita end? Absolute dating<br />
and seriation thus work best when combined in a dialogue in which the flow <strong>of</strong> time within<br />
and between events can be calibrated by the age <strong>of</strong> the events and inform on the properties<br />
<strong>of</strong> the events, such as relative order.<br />
451
In this <strong>chapter</strong> the radiocarbon dates and thermoluminescence dates used are direct<br />
dates associated with the manufacture or use <strong>of</strong> particular pots. Sites are temporal<br />
mixtures, but the direct dates give the age <strong>of</strong> pots, not sites, and thus partially escape the<br />
discontinuous view <strong>of</strong> time <strong>of</strong> dating by stratigraphic association.<br />
Patchy sampling <strong>of</strong> a ceramic series might look like natural groupings in the<br />
ceramic data, which could be misinterpreted as culture-historical phases. Distinguishing<br />
between these alternatives, patchy sampling and discrete phases <strong>of</strong> occupation, requires<br />
that other lines <strong>of</strong> evidence besides ceramic variability be taken into account to assess<br />
whether there is historical continuity and heritable continuity in the sample.<br />
If, for example, at some period the action moved <strong>of</strong>fstage, out <strong>of</strong> the sampling<br />
frame, and then descendants <strong>of</strong> the same people returned later, while there might be a<br />
discontinuity in the sampled ceramic evolutionary record, other factors such as type <strong>of</strong><br />
settlement or use <strong>of</strong> raw materials and adze maufacturing techniques might be more<br />
conservative, and provide evidence for heritable continuity.<br />
<strong>The</strong> Roviana ceramic data falls quite easily into styles or types (as discussed in<br />
Chapter 9). This tendency <strong>of</strong> decorative variability to come in types is regarded below as<br />
more likely to be telling us something about the patchy view <strong>of</strong> time presented by the<br />
available sample, than indicating any sort <strong>of</strong> cultural replacement. <strong>The</strong> argument is made<br />
below that there is evidence for heritable continuity across the entire sample, from<br />
settlement location type and from adzes, as well as from occasional sherds that are<br />
intermediate between the styles or types that most decorated sherds can be easily assigned<br />
to. <strong>The</strong> tendency for sherds to group neatly most likely results from historical<br />
discontinuities in the sample rather than any lack <strong>of</strong> heritable continuity. This does not<br />
automatically mean that there were hiatuses in occupation, just that the sample available<br />
to the analyst is historically incomplete, in that some periods are not well represented.<br />
452
Possibly settlements from these periods have been destroyed by waves due to exposed<br />
location, are hidden under s<strong>of</strong>t sediments, or simply have yet to be found, or alternatively<br />
people may have indeed have been largely absent from the sampling frame at some times.<br />
We cannot discriminate between these possibilities on current evidence.<br />
Radiocarbon data from a series <strong>of</strong> AMS dates is presented and discussed in the<br />
following section, followed in turn by a section on TL dating, and a section on seriation.<br />
TL dates do not give any absolute chronology, but provide useful data on the relative ages<br />
<strong>of</strong> units used in the seriation, despite problems with anomalous fading <strong>of</strong> the TL signal<br />
from the feldspar grains used in the analysis. A series <strong>of</strong> CA seriations are performed,<br />
examining the effect <strong>of</strong> controlling for vessel form in the selection <strong>of</strong> attributes used.<br />
AMS Radiocarbon Dates on Potsherds:<br />
As pottery was collected from the lower margins <strong>of</strong> the swash zone predominantly, spatial<br />
associations between organic materials and the pottery were unlikely to be temporal<br />
associations. Four AMS radiocarbon dates were obtained from sherds (Table 46). <strong>The</strong>se<br />
involve three different modes <strong>of</strong> association between pot and dated carbon,<br />
• the formation <strong>of</strong> the clay deposit from which the pot was made,<br />
• mixing <strong>of</strong> the clay with environmental impurities during manufacture <strong>of</strong> the vessel<br />
• accumulation <strong>of</strong> smoke-derived carbon on the surface <strong>of</strong> the vessel during use.<br />
Also, there is evidence from two <strong>of</strong> the dates to suggest that distinguishing between all <strong>of</strong><br />
these modes <strong>of</strong> association is possible, suggesting direct dating <strong>of</strong> pottery styles is now a<br />
practical reality in Oceanic archaeology.<br />
453
Table 46: Radiocarbon determinations from pottery (calibrated using OxCal 3.5,<br />
Stuiver et al. 1998 atmospheric data).<br />
Sample ID Fraction<br />
Dated<br />
Lab<br />
Number<br />
HV.4.143 “seed” NZA-<br />
12345<br />
HV.4.143 blackened<br />
edge<br />
P.68 charcoal<br />
inclusion<br />
HG.15s.1 fine soot on<br />
sherd<br />
surface<br />
NZA-<br />
12346<br />
Years BP Cal. Age<br />
2sigma<br />
8250+/-370 8300BC-<br />
6400BC<br />
9742+/-80 9350BC-<br />
8800BC<br />
AA33504 2130+/-90 390BC-<br />
30AD<br />
NZA-1253 2619=+/-<br />
45<br />
454<br />
900BC-<br />
550BC<br />
Site Unit<br />
Honiavasa 4<br />
Honiavasa 4<br />
Paniavile Jan 1996<br />
Hoghoi 15s<br />
<strong>The</strong> dates from sherd HV.4.143 (Figure 138) most likely date the geomorphological<br />
formation <strong>of</strong> the clay deposit from which the pottery was eventually made. <strong>The</strong> round<br />
“seed” (identity as a seed is unconfirmed) dated to within 500 years in age <strong>of</strong> a dark<br />
carbon-rich zone near the surface <strong>of</strong> the sherd, visible in the break. <strong>The</strong> carboniferous zone<br />
can be expected to contain small amounts carbon expelled from the clay on firing,<br />
potentially a mix <strong>of</strong> old clay-derived carbon and new carbon mixed in at manufacture,<br />
while the round “seed” could have been included either at the time <strong>of</strong> manufacture <strong>of</strong> the<br />
pot or at the time <strong>of</strong> geological formation <strong>of</strong> the clay. As both ages are substantially older<br />
than the expected age <strong>of</strong> the pottery from style and provenance, we can be reasonably<br />
confident that these do not date the manufacture <strong>of</strong> the pot.<br />
<strong>The</strong> charcoal inclusion in sherd P.68 from Paniavile must predate firing <strong>of</strong> the<br />
vessel and is <strong>of</strong> an age broadly consistent with our expectations <strong>of</strong> the approximate<br />
antiquity <strong>of</strong> the intertidal pottery series (circa-or-post-Lapita) (Figure 139).
Figure 138: A method <strong>of</strong> identifying sub-fossil organic inclusions?<br />
(Image supplied by Rafter Radiocarbon Lab)<br />
Radiocarbon determination<br />
2600BP<br />
2400BP<br />
2200BP<br />
2000BP<br />
1800BP<br />
1600BP<br />
Atmospheric data from Stuiver et al. (1998); OxCal v3.5 Bronk Ramsey (2000); cub r:4 sd:12 prob usp[chron]<br />
AA33504 : 2130±90BP<br />
68.2% probability<br />
360BC (15.3%) 290BC<br />
240BC (52.9%) 40BC<br />
95.4% probability<br />
390BC (95.4%) 30AD<br />
1000CalBC 500CalBC CalBC/CalAD<br />
Calibrated date<br />
500CalAD<br />
Figure 139: Calibration <strong>of</strong> radiocarbon determination from a charcoal inclusion in<br />
a sherd from Paniavile.<br />
455
Sherd HG.15s.1 was dated using smoke-derived carbon which had formed a fine soot on<br />
the surface <strong>of</strong> the sherd, preserved as a result <strong>of</strong> shallow burial in s<strong>of</strong>t mud (Figure 140,<br />
Figure 141, Figure 142). <strong>The</strong> sherd had been treated by the archaeologist with methyl<br />
methacrylate (B72). Pretreatment at the Rafter lab included soaking in acetone, scraping<br />
<strong>of</strong>f <strong>of</strong> a soot sample, followed by a series <strong>of</strong> organic solvent washes (two cycles <strong>of</strong> hexane,<br />
two cycles <strong>of</strong> isopropanol and four cycles <strong>of</strong> acetone, prior to the usual acid/alkali/acid<br />
sequence) (Prior 2000). This allows a reasonable level <strong>of</strong> confidence that the methyl<br />
methacrylate was removed and the 14C age is reliable. <strong>The</strong> extent to which the soot dates<br />
the use <strong>of</strong> the pot is dependent on the average age <strong>of</strong> the fuel used in the fires, so there is<br />
some potential for an old wood effect.<br />
Radiocarbon determination<br />
2900BP<br />
2800BP<br />
2700BP<br />
2600BP<br />
2500BP<br />
2400BP<br />
2300BP<br />
2200BP<br />
Atmospheric data from Stuiver et al. (1998); OxCal v3.5 Bronk Ramsey (2000); cub r:4 sd:12 prob usp[chron]<br />
NZA12353 : 2619±45BP<br />
68.2% probability<br />
830BC (65.2%) 780BC<br />
775BC ( 3.0%) 765BC<br />
95.4% probability<br />
900BC (86.7%) 750BC<br />
690BC ( 3.5%) 660BC<br />
640BC ( 3.6%) 590BC<br />
580BC ( 1.6%) 550BC<br />
1000CalBC 800CalBC 600CalBC 400CalBC<br />
Calibrated date<br />
Figure 140: Calibration <strong>of</strong> a radiocarbon determination from smoke-derived<br />
carbon on a sherd from Hoghoi.<br />
<strong>The</strong> dates from Paniavile and Hoghoi do not overlap at 95% confidence interval, and<br />
suggest that the total duration <strong>of</strong> the intertidal/stilt house settlement pattern was at least<br />
160 years and possibly much longer. <strong>The</strong>se dates suggest there is likely to be significant<br />
456
temporal variability within the total intertidal-zone pottery sample, which is good news for<br />
seriation, as the possibility that the decorative variation which forms the data <strong>of</strong> the<br />
seriation is contemporaneous non-temporal variability <strong>of</strong> some sort is reduced.<br />
<strong>The</strong> older <strong>of</strong> these two dates is from Hoghoi, while the younger is from Paniavile,<br />
but as each <strong>of</strong> these sites contains a mixture <strong>of</strong> styles, and as neither <strong>of</strong> the dated sherds<br />
is clearly identifiable to particular styles, these dates don’t tell us much about the relative<br />
ages <strong>of</strong> styles. Although the Hoghoi site has a lot <strong>of</strong> Gharanga-Kopo style sherds, sherds<br />
<strong>of</strong> the dated sample are from an area <strong>of</strong> the site (transects 12 and 15) where attributes<br />
identified with the Miho style occur also, and the form <strong>of</strong> the vessel, with tallish thin<br />
Figure 141: AMS sample taken from blackened sherd at far left, found on<br />
the surface <strong>of</strong> unit 12 at Hoghoi. <strong>The</strong> sherd to the right, found in a<br />
subsurface test <strong>of</strong> unit 15 at Hoghoi, may be from the same vessel, and also<br />
has a sooted surface.<br />
excurvate rim, is more like the Miho style than Kopo style (Figure 141, Figure 142). <strong>The</strong><br />
only decoration, if it can be called that, is a deep impressed groove around the neck. <strong>The</strong><br />
vessel is unusually small in horizontal dimensions, which may have been a factor<br />
contributing to its preservation through increased form strength.<br />
457
Figure 142: Vessel with surface sooting from Hoghoi dated by AMS<br />
radiocarbon.<br />
Another sherd in the illustrations thought to be from this vessel was located by test<br />
excavation within 20m <strong>of</strong> the found location, on the surface in square 12. <strong>The</strong>se were<br />
attributed to the same vessel by virtue <strong>of</strong> similar form, the grooving at the neck, and similar<br />
sooting on the body surface, not seen on any other sherds.<br />
<strong>The</strong> dated sample from Paniavile was recovered from a plain sherd comprising a<br />
s<strong>of</strong>t shoulder and globular body, which cannot be identified to any stylistic grouping or<br />
period.<br />
<strong>The</strong>rmoluminescence (TL) Dates from Quartz-Calcite Sherds:<br />
Applicability <strong>of</strong> TL:<br />
<strong>The</strong>rmoluminescence dates the time a pot was last heated to a sufficient temperature to re-<br />
set the luminescence clock (about 450/C) (Feathers 1997). In most cases this will date<br />
either the manufacture <strong>of</strong> the pot or its last use. This assumption is particularly secure for<br />
materials discarded into the sea. This makes TL particularly useful for surface<br />
archaeological distributions in the intertidal zone and in shallow water.<br />
458
Uncertainties arising from factors affecting the precision <strong>of</strong> radiocarbon dates such as old<br />
carbon and contamination from geologic carbon do not apply to thermoluminescence.<br />
Calibration <strong>of</strong> radiocarbon dates sometimes involves a reduction in precision, when the<br />
calibration curve is flat, as in the period 760BC-250BC (Figure 138 and Figure 140). This<br />
flatness makes charcoal-dating the transformations <strong>of</strong> the Lapita cultural complex<br />
following its initial spread across Oceania extremely difficult.<br />
<strong>The</strong> disadvantage <strong>of</strong> thermoluminescence dating lies in “its dependence on<br />
numerous complex variables that can be difficult to estimate in any given situation”<br />
(Feathers 1997:4). TL dating can be used as an ordinal measure <strong>of</strong> age, an interval measure<br />
or as a ratio scale measurement, with a reduction in precision as one moves up the scale<br />
between these. Long-duration events such as an occupation can be dated with more<br />
precision by dating more <strong>of</strong> the objects involved (Feathers 1997:7).<br />
Luminescence Physics:<br />
<strong>The</strong> following is a synthesis from a recent TL review (Feathers 1997:11-54) with<br />
comments relevant to the Roviana case added.<br />
A material that has been exposed to natural radiation will emit a faint light when<br />
heated, proportional to the amount <strong>of</strong> energy absorbed, and proportional to time elapsed<br />
since such accumulation began. This relationship breaks down when enough time has<br />
elapsed for the material dated (the dosimeter) to become saturated with radiation, to the<br />
point where it can trap no more energy. <strong>The</strong> most common natural radioactive isotopes are<br />
40 K, 238 U and 232 Th, which have long half-lives and effectively constant global levels,<br />
allowing estimation <strong>of</strong> dose rates impinging on archaeological samples. <strong>The</strong>re are other<br />
minor contributors to natural radiation flux. Terrestrial natural radiation can be divided into<br />
three kinds (Alpha, Beta, and Gamma rays) <strong>of</strong> which Beta and Gamma rays are effective<br />
in inducing luminescence, by virtue <strong>of</strong> their penetrating power (Alpha rays do affect the<br />
surfaces <strong>of</strong> grains and small grains).<br />
Successful dating requires a source <strong>of</strong> radiation in the environment <strong>of</strong> the item to<br />
be dated, or even within the item itself, and a natural dosimeter within the item (which<br />
459
may also be a source <strong>of</strong> radiation, as in the cases <strong>of</strong> potassium feldspar, calcite or zircon).<br />
Quartz and sodic feldspars are good dosimeters, but lack an internal radiation source.<br />
None <strong>of</strong> the usual dosimeters will become saturated in less than about 100Ka (thousand<br />
years) so saturation is not a serious problem for Lapita archaeology.<br />
Fading is said to occur when the accumulated latent signal is not stable through<br />
time, but leaks away at ambient temperatures, making the date too young, either by<br />
mechanisms for which the probability <strong>of</strong> fading is known (thermal fading) or by poorly-<br />
understood and difficult-to-predict mechanisms (anomalous fading). <strong>The</strong>rmal fading can<br />
be more <strong>of</strong> a problem in hot climates, and for older samples, but in the case <strong>of</strong> the Roviana<br />
pottery thermal fading is less <strong>of</strong> a factor as the pottery has been underwater for most <strong>of</strong> its<br />
relatively short depositional history and has not been subject to high ambient temperatures.<br />
Anomalous fading can render a sample undatable, and is thought to have affected the<br />
Roviana samples, being a major obstacle to accurate calendrical dating in this case.<br />
Anomalous fading is more common with feldspars, and seldom affects quartz, unless the<br />
latter has feldspar impurities.<br />
To derive the age <strong>of</strong> the sample two measurements are required, equivalent dose<br />
and annual dose. <strong>The</strong> age is the equivalent dose (the total amount <strong>of</strong> radiation absorbed<br />
since last zeroed) divided by the annual dose. Equivalent dose is measured either by the<br />
additive dose technique, whereby the material is irradiated in increments and the growth<br />
curve <strong>of</strong> the emitted luminescence is compared to that <strong>of</strong> a standard, or by the regeneration<br />
technique, whereby the luminescence <strong>of</strong> the sample is measured, then zeroed by heat or<br />
light, then rebuilt by incremental dosing, producing a regeneration curve. <strong>The</strong> shapes <strong>of</strong><br />
these response curves (the sensitivity <strong>of</strong> the sample) are used in both techniques to<br />
estimate the age <strong>of</strong> the sample. Sensitivity to dose can change during measurement.<br />
Linearity in the growth curve cannot be assumed. <strong>The</strong> causes <strong>of</strong> sensitivity changes during<br />
measurement are not all well understood. After a zeroing event, the luminescence signal<br />
can build at a different rate, depending on the nature <strong>of</strong> the zeroing event.<br />
Measuring emission is more difficult for young samples like pottery, as the<br />
amount <strong>of</strong> light emitted is less. Selection <strong>of</strong> an appropriate spectral filter may be required<br />
460
to maximise the signal from such samples. Detection equipment tends to be designed for<br />
older samples than typically analysed by archaeologists, and may lack sufficient resolution<br />
for the task in some labs.<br />
Measuring dose rate requires correction for moisture content, correction for<br />
uncertainty regarding the external dose rate, and the possibility <strong>of</strong> changes in the dose rate<br />
through time. Water absorbs some <strong>of</strong> the radiation, and correction for this factor is easiest<br />
where the environment is known to have been either saturated or fully dry during the time<br />
since zeroing (most likely fully saturated in the Roviana case). Alpha and beta radiation,<br />
with little penetrating ability, are relatively easy to estimate from the properties <strong>of</strong> the<br />
sample itself, but gamma radiation, with a range <strong>of</strong> up to 30cm in soil, comes mostly from<br />
outside the sample, requiring some idea <strong>of</strong> the homogeneity <strong>of</strong> the environment in this<br />
respect. <strong>The</strong> Roviana sites with a scatter <strong>of</strong> volcanic ovenstones and volcanic-tempered<br />
pottery in a predominantly calcareous environment, are potentially inhomogeneous in this<br />
respect.<br />
“Particularly troublesome is the presence within the 30cm sphere <strong>of</strong> large<br />
clasts with a radioactivity different from that <strong>of</strong> the surrounding sediment”<br />
(Feathers 1997:31).<br />
Placement <strong>of</strong> zeroed dosimeters in the sample location, or use <strong>of</strong> portable gamma-<br />
ray scintillometers can calculate the modern dose to which the sample is exposed, but<br />
where there is a history <strong>of</strong> turbation or movement <strong>of</strong> the sample, calculation <strong>of</strong><br />
environmental gamma dose will not be as clear-cut, but may be less variable between<br />
samples from a site due to an averaging effect over time. Perhaps the issues for the<br />
Roviana materials is how long have these been in lag deposits as opposed to a stable<br />
matrix, and whether the dose rate was lower or more variable between sherds when these<br />
were deeply immersed, compared to the present dose rate. A detailed sea-level history<br />
combined with placement <strong>of</strong> dosimeters in experimental deposits could partially answer<br />
these questions.<br />
Weathering can affect internal and external dose rate through postdepositional<br />
movement <strong>of</strong> radionuclides. For example water can remove U but leave Th intact. Clay<br />
461
minerals in ceramics are end products <strong>of</strong> weathering, and in the case <strong>of</strong> the Roviana<br />
ceramics, have clearly been exposed to salt water. Low fired ceramics (less than 1000/C,<br />
as in the Roviana case as calcite reef detritus is unsintered), are particularly prone to these<br />
effects, but if such a disequilibrium is detected for radionuclides with short half-lives a<br />
correction can sometimes be applied, especially where the leaching effect is more or less<br />
constant. If radionuclides <strong>of</strong> longer half-life are involved this is more difficult, and<br />
uncertainty about dose rate can be the result. Detection <strong>of</strong> such disequilibrium is largely<br />
a matter <strong>of</strong> cost: a variety <strong>of</strong> measurement techniques can be applied, the cheaper <strong>of</strong> which<br />
measure alpha radiation only, and estimate other radiation from elemental abundance<br />
combined with natural isotopic abundance, or by Neutron Activation Analysis. High-<br />
resolution gamma spectroscopy is expensive, requires long counting times and large<br />
samples, but measures the gamma activities <strong>of</strong> several radionuclides, and this supplies a<br />
means to assess equilibrium.<br />
Luminescence signal can be measured from individual grains (inclusion dating),<br />
usually from either quartz or feldspar, or, where possible, zircon, which has abundant<br />
radioactive impurities that negate the need for estimation <strong>of</strong> external dose. Use <strong>of</strong><br />
individual grains facilitates choice <strong>of</strong> dating procedures and eliminates influence from clay<br />
minerals. Fine grain analysis, making no attempt to separate particular minerals, is required<br />
where large grains are unavailable, and despite some problems, can be useful where the<br />
environmental dose is complicated because the influence <strong>of</strong> gamma radiation is less with<br />
the inclusion <strong>of</strong> alpha-irradiated fine grains.<br />
Low signal output is the commonest cause <strong>of</strong> low precision when dating young<br />
ceramics. <strong>The</strong> best results come from samples in which there are abundant quartz grains<br />
<strong>of</strong> optimum size. Rocks which have been sufficiently heated <strong>of</strong>ten produce reliable dates,<br />
using either TL or OSL. <strong>The</strong> main reason lithic materials are seldom used for TL dating<br />
is that there is seldom reason to believe they have been sufficiently heated, although there<br />
are also greater sample preparation problems with lithics. I see no reason why ovenstones<br />
should not have been heated as highly if not higher than open-fired calcite-tempered<br />
pottery.<br />
462
Roviana TL Data:<br />
Eight pottery samples from four archaeological sites were submitted to the <strong>University</strong> <strong>of</strong><br />
Washington Anthropology Department TL laboratory. All were from the quartz-calcite<br />
temper group, which held the best prospect for isolating grains suitable for TL dating<br />
(Feathers 2002).<br />
Environmental Dose:<br />
Nine sediment samples and one ovenstone sample were measured to aid in the estimation<br />
<strong>of</strong> environmental radiation. Radioactivity <strong>of</strong> sherds was higher than that <strong>of</strong> sediments<br />
samples. Volcanic manuports from Zangana yielded higher radioactivity measurements<br />
than the calcite reefal detritus samples. Environmental dose rate was calculated by<br />
averaging the dose from the sediment samples for each site, with the manuport<br />
radioactivity omitted.<br />
Equivalent Dose and Fading:<br />
Two <strong>of</strong> the smaller samples, one from Gharanga and one from Miho, which also lacked<br />
sufficient quartz gains for inclusion dating, were analysed by the fine-grain method, and<br />
yielded equivalent dose measurements <strong>of</strong> low precision, due to weak signal-to-noise ratio.<br />
<strong>The</strong> other sherds were initially analyzed using the inclusion method on extracted quartz,<br />
but unreasonably old ages with high error terms were obtained, leading Feathers to<br />
conclude there was a spurious luminescence component, and this approach was aborted.<br />
Inclusions obtained by density separation <strong>of</strong> Potassium feldspars and Aluminium<br />
Silicates produced a sufficiently strong signal for dating, but four <strong>of</strong> theses samples showed<br />
evidence for anomalous fading (which can make the age too young). A correction for<br />
fading was applied to three <strong>of</strong> these samples, which resulted in low precision. Results are<br />
given in Table 47.<br />
Feathers rejected the fading correction for uw476 as it seemed too old based on<br />
the information supplied to him, which was that the occupation being dated ran from<br />
463
about 2500-1900BP, however, the corrected age, as initially supplied by him by email, is<br />
entirely consistent with the expected age <strong>of</strong> the Honiavasa Lapita site, and is therefore<br />
included in Table 47. Looking at the ages uncorrected for fading, the two dates from<br />
Honiavasa were the oldest (uw475 and uw476), while two <strong>of</strong> the dates from Paniavile<br />
(UW 472 and uw474) and the date from Gharanga (uw471) formed a young group. <strong>The</strong><br />
other date from Paniavile could be grouped with the Miho date and the Zangana south date<br />
( uw473, uw477, uw478). While Feathers does not think the dates can be used as relative<br />
ages as they are statistically indistinguishable when corrected for fading, the argument will<br />
be made below that the uncorrected ordination <strong>of</strong> ages, ignoring their standard deviations,<br />
agrees with the seriation favoured below, and can be reconciled with the AMS C14<br />
evidence also. This correlation with the seriation additionally supports the inference drawn<br />
from the decorative analysis <strong>of</strong> the quartz-calcite hybrid temper, which was that this temper<br />
group was imported over a substantial period.<br />
Table 47: <strong>The</strong>rmoluminescence dating results. Precision shown is at confidence limits<br />
<strong>of</strong> one standard deviation.<br />
Lab# Site Sherd Grain size Date (ka) Date after<br />
fading<br />
correction<br />
(ka)<br />
uw471 Gharanga GH.183 fine 0.984±0.166 1.608±1.043<br />
uw472 Paniavile C.49 coarse 1.095±0.364 2.079±1.202<br />
uw473 Paniavile R379a4 coarse 1.801±0.296<br />
uw474 Paniavile P.165 coarse 0.790±0.363 0.950±0.495<br />
uw475 Honiavasa HV.5.120 coarse 1.903±0.290<br />
uw476 Honiavasa HV.1.48 coarse 1.903±0.290 2.339±0.750<br />
uw477 Miho MH.61 fine 1.798±0.278<br />
uw478 Zangana B.1 coarse 1.817±0.338<br />
464
UW474 suggests (at two standard deviations) that firing or use <strong>of</strong> quartz-calcite pottery<br />
continued until after 1940 BP (BP=before 2001 for this TL data) or in calendrical years,<br />
61AD. In combination with the radiocarbon evidence from NZA-12353 (Figure 140)<br />
UW474 suggest a minimum total temporal span for the intertidal-zone sites from 550BC<br />
to 61AD, or 611 years.<br />
Seriation:<br />
<strong>The</strong> technique chosen for seriation was correspondence analysis (CA), using the<br />
WINBASP program developed by Bonn <strong>University</strong>. Correspondence analysis will only<br />
produce a temporal seriation if the types/attributes in terms <strong>of</strong> which the units are being<br />
described are chronologically sensitive. “<strong>The</strong>re must be reasons to believe this in the first<br />
Table 48: Definitions <strong>of</strong> attribute codes used in seriation tables and plots.<br />
N_A_B_NB Neck, applied decoration, being a band, <strong>of</strong> nubbins<br />
L_F_BOPC Lip, fingernail impression, being a band, <strong>of</strong> opposed pinching<br />
N_F_B_PC Neck, fingernail impression, being a band, <strong>of</strong> opposed pinching<br />
F_MB__PC Fingernail impression, multiple bands, <strong>of</strong> opposed pinching<br />
L_I___BI Lip, impression, band <strong>of</strong>, inner edge <strong>of</strong> lip<br />
L_I___BO Lip, impression, being a band <strong>of</strong>, outer edge <strong>of</strong> lip<br />
L_I___BT Lip, impression, being a band <strong>of</strong>, top <strong>of</strong> lip<br />
L_I___BB Lip, impression, being a band <strong>of</strong>, both edges<br />
L_I__WAV Lip, impression, into a wave pattern (staggered inner and outer edges)<br />
L_D__WAV Lip, deformation, into a horizontal wave<br />
L_D__OCI Lip, deformation, discontinuous into individual spout-like impressions<br />
G_P____B anywhere on vessel (G=general), punctation, single horizontal band<br />
CLASS1LM Class 1 linear motif, anywhere on vessel<br />
R_CLS2LM Rim, class 2 linear motif<br />
S_CLS2LM Shoulder, class 2 linear motif<br />
STAMPING Anywhere on vessel, stamped decoration (including dentate)<br />
465
place (Shennan 1997:342).” <strong>The</strong> decorative attribute incidences (sherd counts) tabulated<br />
in Table 42 were used in the initial seriation reported below. To control for form variation,<br />
a number <strong>of</strong> these attributes were excluded from a second seriation, producing a different<br />
result to that <strong>of</strong> the total attribute set.<br />
In the seriations, attributes from Table 42 are coded as in Table 48. Schematic<br />
representation <strong>of</strong> attributes is given in Figure 143.<br />
Seriation 1: All Attributes, All Vessel Forms:<br />
Sample sizes by site for the various attributes are given in Table 42. Component 1 and<br />
Component 2 represent more than 84% <strong>of</strong> the total inertia, which reassures that the counts<br />
<strong>of</strong> the various attributes reduce well into two dimensions <strong>of</strong> variability (Table 49). <strong>The</strong> Qlt<br />
column <strong>of</strong> Table 50 indicates that some attributes are not well characterized by these two<br />
components (impressions on the outer lip and impressions on the top <strong>of</strong> the lip, have the<br />
lowest quality scores, while wave-impression <strong>of</strong> the lip, deformation <strong>of</strong> the lip by<br />
impression, and Class 2 linear motif on the shoulder score fairly low, suggesting these<br />
latter attributes are only moderately well characterized by the two summary components).<br />
<strong>The</strong> Mass column indicates that lip deformation into a wave, a band <strong>of</strong> pinching at the<br />
neck, and banded punctation have the greatest weight among the types by virtue <strong>of</strong> their<br />
greater incidence, while stamping, lip impression into a wave and a band <strong>of</strong> applied nubbins<br />
at the neck have the least sample-size-related weight <strong>of</strong> the types/attributes. <strong>The</strong>re are no<br />
outliers among the attribute contributions to inertia (Inr). Pinched band at the lip, Class 1<br />
linear motifs, band <strong>of</strong> punctation, and band <strong>of</strong> impressions on the inner lip have the greatest<br />
attribute contributions to inertia.<br />
<strong>The</strong> first Ctr column (Table 50) shows that applied band <strong>of</strong> nubbins at the neck<br />
(N_A_B_NB) contributes 8.0% <strong>of</strong> the inertia in Component 1 (80/1000), band <strong>of</strong> pinching<br />
at the lip (L_F_BOPC) contributes 22.5% <strong>of</strong> the Component 1 inertia (225/1000), etc.<br />
Thus we see that Component 1 is a product principally <strong>of</strong> three attribute incidences, these<br />
being: Class 1 linear motif incidence (CLASS1LM), band <strong>of</strong> pinching <strong>of</strong> the lip<br />
(L_F_BOPC), and a band <strong>of</strong> impressions on the inner lip (L_I___BI).<br />
466
Figure 143: Attributes used in seriations, groupings explained in <strong>chapter</strong> conclusions.<br />
467
Component 2 inertia is a result principally <strong>of</strong> the distribution through sites <strong>of</strong> bands <strong>of</strong><br />
punctation (G_P____B). Multiple bands <strong>of</strong> pinching, a band <strong>of</strong> pinching at the neck, and<br />
Class 2 linear motif on the shoulder are also important contributors to Component 2<br />
inertia. A plot <strong>of</strong> type (attribute) correspondence values for the two components is given<br />
in Figure 144.<br />
Table 49: Relative contribution <strong>of</strong> first and second components <strong>of</strong> CA using all<br />
attributes and forms.<br />
Component Iterations Norm Eigenvalue % Inertia Cumulative%<br />
1 28 0.051 0.583806 51.3 51.3<br />
2 8 0.082 0.376932 33.1 84.4<br />
Table 50: CA diagnostics by attribute, all attributes included.<br />
Attribute<br />
Qlt Mass Inr Comp1 Cor Ctr Comp2 Cor Ctr<br />
N_A_B_NB 971 20 42 -1534 962 80 -148 9 1<br />
L_F_BOPC 969 50 127 -1616 909 225 415 60 23<br />
N_F_B_PC 921 124 76 508 372 55 616 549 125<br />
F_MB__PC 854 50 63 298 62 8 -1062 791 151<br />
L_I___BI 982 61 94 -1307 977 179 -92 5 1<br />
L_I___BO 63 63 15 74 20 1 -106 43 2<br />
L_I___BT 0 61 20 0 0 0 -9 0 0<br />
L_I___BB 893 29 32 -195 30 2 -1043 863 83<br />
L_I__WAV 421 14 44 327 30 3 -1172 391 52<br />
L_D__WAV 669 156 42 409 542 45 198 127 16<br />
L_D__OCI 369 70 23 339 304 14 157 65 5<br />
G_P____B 951 122 136 299 70 19 -1056 881 362<br />
CLASS1LM 983 31 120 -2071 959 225 327 24 9<br />
R_CLS2LM 895 92 76 593 371 55 704 524 121<br />
S_CLS2LM 563 43 39 427 177 13 632 387 46<br />
STAMPING 840 13 49 -1899 815 78 328 24 4<br />
468
Figure 144: Correspondence plot (attributes) using all attributes and forms.<br />
<strong>The</strong> same sort <strong>of</strong> diagnostic statistics are given for the sites (Table 51). <strong>The</strong> two summary<br />
components give a good representation <strong>of</strong> the variability at Honiavasa (999/1000).<br />
Similarly, but to a lesser extent, the variability at all other sites except the three with the<br />
least mass (small sample sizes) is well represented. <strong>The</strong> Honiavasa site is an outlier in terms<br />
<strong>of</strong> its contribution to total inertia, contributing more than 80% <strong>of</strong> Component 1 inertia.<br />
Gharanga makes the greatest contribution to Component 2 inertia, with Hoghoi, Miho and<br />
Zangana South making significant contributions to Component 2 also. Honiavasa’s<br />
contribution to Component 2 inertia is negligible (1.8%).<br />
469
Table 51: CA diagnostics by site, all attributes included (*=inertia outlier).<br />
SITES Qlt Mass Inr Comp1 Cor Ctr Comp2 Cor Ctr<br />
PANIAVIL 703 187 70 348 284 39 422 418 88<br />
HOGHOI__ 817 165 87 174 50 9 -676 767 201<br />
MIHO____ 794 180 98 507 416 79 484 378 112<br />
HONIAVAS 999 167 418* -1674 985 803 201 14 18<br />
GHARANGA 866 101 164 158 13 4 -1255 852 421<br />
NUSAROVI 39 20 24 -112 9 0 203 30 2<br />
KOPO____ 154 4 15 417 38 1 -733 116 5<br />
ZANNORTH 487 61 41 327 140 11 -516 348 43<br />
ZANSOUTH 761 115 84 520 327 53 599 434 109<br />
A plot <strong>of</strong> these two components is given in Figure 145. <strong>The</strong> correspondence scores in the<br />
Cor columns provide the coordinates for these plots. Honiavasa is an outlier along the<br />
Component 1 axis, the most similar site along Component 1 being Nusa Roviana, while the<br />
other sites separate into two groups at opposing ends <strong>of</strong> Component 2. Nusa Roviana has<br />
one example <strong>of</strong> a Class 1 linear motif (sherd NR34), but is closer to Honiavasa more by<br />
default, by absence <strong>of</strong> characteristics, leading to a location on the plot close to the average,<br />
or 0,0 position. <strong>The</strong> single occurrence <strong>of</strong> stamping at Nusa Roviana will also have had a<br />
slight effect on the plot. It must be remembered that Nusa Roviana was not well<br />
characterized by these two components. Stamping and Class 1 linear motif occurred on the<br />
same sherd, so the similarity between Nusa Roviana and Honiavasa hinges on a single<br />
sherd, and thus should not be made too much <strong>of</strong>.<br />
470
Figure 145: Correspondence plot (sites) using all attributes and forms.<br />
As discussed in the form and decoration <strong>chapter</strong>s (Chapters 8 and 9) there are some<br />
decorative attributes that are highly correlated with aspects <strong>of</strong> vessel form variation, and<br />
which may be decoration specific to vessels <strong>of</strong> a particular function. In order to minimise<br />
the risk <strong>of</strong> incorrectly seriating contemporaneous functional variation, these attributes were<br />
excluded from analysis below. Excluded were: Class 1 linear motifs; bands <strong>of</strong> applied<br />
nubbins at the neck; and stamping; all highly correlated with carinated vessel forms.<br />
Multiple bands <strong>of</strong> fingernail pinching were also excluded due to occurrence only on short-<br />
rim heavily excurvate Gharanga-type vessels.<br />
Also, some lip impression similarities between Gharanga form/decoration type and<br />
some Honiavasa tall excurvate rims are likely to be analogous rather than homologous<br />
similarity (see Chapter 9), and banded lip impressions are thus excluded from subsequent<br />
471
seriations, with the exception <strong>of</strong> lip impressions on both edges <strong>of</strong> the lip forming a wave<br />
pattern; this decorative attribute occurred mainly in the Gharanga site on tall weakly<br />
excurvate rims within the “general purpose” Form 6c/2c category, and indications from<br />
this pattern <strong>of</strong> occurrence are a relatively restricted temporal occurrence.<br />
Seriation 2: Form-correlated Attributes Removed:<br />
Cumulative inertia <strong>of</strong> the first two components is 88%, suggesting that the patterning in<br />
the data summarizes well into these components (Table 52). <strong>The</strong> three components<br />
together produce an average quality <strong>of</strong> characterization for the attributes <strong>of</strong> 81%<br />
(812/1000) (Table 52), with three attributes only moderately characterized (wave-<br />
deformation <strong>of</strong> the lip, discontinuous finger-deformation <strong>of</strong> the lip and Class 2 linear motif<br />
on the shoulder). Of these, Class 2 linear motif on the shoulder has a low mass figure,<br />
related to small sample size. Component 1 contributions to inertia are almost wholly<br />
dominated by fingernail pinch band at the lip, while the major contributor to Component<br />
2 inertia is banded punctation (G_P___B).<br />
Class 2 linear motif on the rim and a band <strong>of</strong> pinching at the neck make significant<br />
contributions to Component 2 inertia also, while the reasonably numerous wave<br />
deformation <strong>of</strong> the lip and discontinuous finger-deformation <strong>of</strong> the lip classes make very<br />
little contribution to the inertia <strong>of</strong> either <strong>of</strong> the first two components. This is particularly<br />
interesting, as these ubiquitous decorative techniques may indicate either a long production<br />
span, or possibly a temporally central production span within the Roviana intertidal milieu,<br />
where occupation spans <strong>of</strong> sites tend to incorporate discard <strong>of</strong> these attributes by virtue<br />
<strong>of</strong> their temporal centrality. Component three may hold the key in this regard, because<br />
these attributes, and particularly the related technique <strong>of</strong> lip impression into a wave, have<br />
a lot <strong>of</strong> inertia there. <strong>The</strong> correspondence scores <strong>of</strong> these attributes for Components 2 and<br />
3 are plotted in Figure 146.<br />
<strong>The</strong> position <strong>of</strong> pinching <strong>of</strong> the lip on this plot can be disregarded as it is an outlier<br />
472
in Component 1, which is not shown. Lip impression into a wave pattern is also an outlier,<br />
in the low-inertia Component 3, but small sample size must be born in mind here. Although<br />
most examples occur in a single site, absence from others might be a sample size effect.<br />
<strong>The</strong> tight grouping <strong>of</strong> neck pinching, Class 2 linear motif on the rim, and Class 2 linear<br />
motif on the shoulder confirms the “Miho” decorative style as a relatively coherent entity,<br />
as suggested in Chapter 9. <strong>The</strong> separation <strong>of</strong> lip deformation into a wave and discontinuous<br />
finger deformation <strong>of</strong> the lip from the “Miho style” grouping results largely from the spatial<br />
separation <strong>of</strong> these attributes within the greater Zangana site, although wave-deformation<br />
<strong>of</strong> the lip occurs in almost all sites to some extent, which is why these attributes lie close<br />
to the intersection <strong>of</strong> the axes. Overall it is reassuring that the attributes order in the same<br />
way in Component 2 as they did in Figure 144 despite omission <strong>of</strong> some potential problem<br />
attributes.<br />
Table 52: Eigenvalues and contributions to intertia <strong>of</strong> CA components, for data<br />
excluding form-correlated attributes.<br />
Component Iterations Norm Eigenvalue % Inertia Cumulative<br />
%<br />
1 28 0.051 0.611850 53 53.0<br />
2 9 0.004 0.407938 35.3 88.3<br />
3 10 0.041 0.080177 6.9 95.2<br />
Table 53: CA diagnostic table for attributes, form-correlated attributes omitted (*=intertia<br />
outlier).<br />
Name Qlt Mass Inr Comp1 Cor Ctr Comp2 Cor Ctr Comp3 Cor Ctr<br />
L_F_BOPC 1000 75 479* 2676 969 876 479 31 42 -21 0 0<br />
N_F_B_PC 959 184 56 -111 35 4 -547 850 135 -161 74 60<br />
L_I__WAV 972 21 81 -591 80 12 1255 359 83 -1530 533 624<br />
L_D__WAV 638 233 26 -165 213 10 -133 138 10 192 288 107<br />
L_D__OCI 485 104 18 -90 39 1 -62 19 1 295 426 113<br />
G_P____B 996 182 240 -551 199 90 1099 792 538 89 5 18<br />
R_CLS2LM 862 136 64 -158 46 6 -663 809 147 -62 7 7<br />
S_CLS2LM 586 64 35 36 2 0 -525 441 43 -298 142 71<br />
473
Turning to the table <strong>of</strong> scores for sites (Table 54) it is clear that all sites except Kopo and<br />
Nusa Roviana are well characterized by the three components <strong>of</strong> the analysis. <strong>The</strong> mass<br />
column indicates that these latter sites had small sample sizes, and were given little weight<br />
in the calculation <strong>of</strong> inertia values. Honiavasa site is still an outlier in its contribution to<br />
total inertia, by virtue <strong>of</strong> a contribution <strong>of</strong> almost 87% <strong>of</strong> Component 1 inertia, which, as<br />
can be seen in Figure 146, is mostly to do with the incidence <strong>of</strong> a band <strong>of</strong> pinching at the<br />
lip. Gharanga site dominates the contributions to Component 2 inertia, but Hoghoi, Miho<br />
and Zangana South are important also, the first two by virtue <strong>of</strong> the high incidence <strong>of</strong><br />
punctation, the last because <strong>of</strong> neck pinching and Class 2 linear motif on the rim. A plot<br />
<strong>of</strong> the first two component correspondence scores is given in Figure 147.<br />
Figure 146: Correspondence plot (attributes) excluding form-correlated attributes<br />
474
Table 54: CA diagnostics by site, form-correlated attributes omitted (*=inertia outlier).<br />
SITE Qlt Mas<br />
s<br />
Inr Comp1 Cor Ctr Comp<br />
2<br />
475<br />
Cor Ctr Comp<br />
3<br />
Cor Ctr<br />
PANIAVIL 840 230 40 34 6 0 -410 834 95 14 1 1<br />
HOGHOI__ 948 144 104 -439 231 45 727 634 187 264 84 126<br />
MIHO____ 822 225 71 -192 101 14 -502 691 139 104 30 30<br />
HONIAVAS 1000 75 476* 2661 965 86<br />
7<br />
508 35 47 6 0 0<br />
GHARANGA 989 80 180 -609 143 49 1358 710 362 -595 136 354<br />
NUSAROVI 325 19 9 -248 114 2 -102 19 0 322 192 24<br />
KOPO____ 464 5 13 -410 61 1 811 237 9 678 166 31<br />
ZANNORTH 872 70 44 -409 230 19 528 382 47 435 260 164<br />
ZANSOUTH 957 152 63 -107 24 3 -551 635 113 -377 297 270<br />
In comparison to Figure 145, Figure 147 shows a different ordering <strong>of</strong> sites in<br />
Component 1.<br />
Figure 147: Correspondence plot (sites) excluding form-correlated<br />
attributes.
While this might seem relatively minor in view <strong>of</strong> the stability <strong>of</strong> Component 2 ordering<br />
between these two plots, Component 1 is crucial to the overall relative dating <strong>of</strong> the three<br />
broad styles that make up the series (the Honiavasa sample, Miho style and Gharanga<br />
style) as it is Component 1 which shows which <strong>of</strong> the site assemblages is more similar to<br />
the Honiavasa Lapita site. Where previously Nusa Roviana, Gharanga and Hoghoi were<br />
closest to Honiavasa in terms <strong>of</strong> Component 1, now Paniavile is closest, by virtue <strong>of</strong> the<br />
presence <strong>of</strong> five examples <strong>of</strong> opposed-pinch band on the outer lip there. Zangana South<br />
is next closest due to a single occurrence <strong>of</strong> the same attribute there. <strong>The</strong>se sample sizes<br />
show how tenuous the homologous links between Honiavasa and the other sites become,<br />
when the (potentially analogous) lip impression similarity is excluded. <strong>The</strong>re are echoes in<br />
this situation <strong>of</strong> the dilemma faced by Specht in deciding whether the transition from Buka<br />
phase to Sohano phase on Bougainville was gradual, but poorly sampled, or whether<br />
cultural replacement had occurred. It seems clear in the present case that the Honiavasa<br />
site is a site largely alone, with the exception <strong>of</strong> occasional sherds with similar attributes<br />
occurring elsewhere, and that the current sample is insufficient to address this question<br />
from ceramic stylistic evidence alone.<br />
With the omission <strong>of</strong> lip impressions Gharanga is now the most distant site in<br />
Component 1, and all the sites in which a single band <strong>of</strong> punctation is dominant are now<br />
at the opposite end <strong>of</strong> Component 1 to the Honiavasa site. Ordering in Component 1 <strong>of</strong><br />
these other sites roughly mirrors ordering in Component 2, with the group nearest to<br />
Honiavasa in Component 1 (Paniavile, Miho, Zangana South and Nusa Roviana) forming<br />
a separate group in Component 2 also.<br />
Seriation 3: Honiavasa Excluded and Form-correlated Attributes Excluded:<br />
If Honiavasa, the outlier, is excluded from the data, some <strong>of</strong> the minor components might<br />
be expected to assume greater weight in the CA, which is not the case (Table 57, Table<br />
55, Table 56 and Figure 148).<br />
476
Table 57: CA eigenvalues and component contributions to inertia omitting Honiavasa<br />
data and form-correlated attributes<br />
Component Iterations Norm Eigenvalue % Inertia Cumulative<br />
1 8 0.024 0.426248 72 72.0<br />
2 21 0.048 0.081960 13.9 85.9<br />
3 15 0.048 0.046658 7.9 93.8<br />
Table 55: CA diagnostics for attributes, omitting Honiavasa data and form-correlated<br />
attributes.<br />
ATTRIBUTE Qlt Mas<br />
s<br />
Table 56: CA diagnostics by site, omitting Honiavasa data and form-correlated<br />
attributes.<br />
SITE Qlt Mass Inr Comp1 Cor Ctr Comp2 Cor Ctr Comp3 Cor Ctr<br />
PANIAVIL 978 249 122 441 673 114 27 2 2 296 302 465<br />
HOGHOI__ 947 156 188 -779 852 222 -249 87 118 75 8 19<br />
MIHO____ 986 243 117 411 594 96 -142 71 59 -302 321 475<br />
GHARANG<br />
A<br />
990 87 342 -1402 844 400 575 142 350 -105 5 20<br />
NUSAROVI 233 20 16 32 2 0 -325 230 26 16 1 0<br />
KOPO____ 472 6 25 -855 280 10 -680 177 33 -193 14 5<br />
ZANNORT<br />
H<br />
Inr Comp1 Cor Ctr Comp2 Cor Ctr Comp3 Cor Ctr<br />
L_F_BOPC 908 17 58 691 243 19 284 41 17 1106 623 455<br />
N_F_B_PC 944 197 95 494 855 112 145 74 50 -67 16 19<br />
L_I__WAV 981 23 156 -1300 424 92 1457 533 599 -311 24 48<br />
L_D__WAV 561 246 36 92 97 5 -184 390 101 80 74 34<br />
L_D__OCI 567 107 36 34 6 0 -313 487 128 -123 75 34<br />
G_P____B 996 197 444* -1151 991 610 -76 4 14 39 1 7<br />
R_CLS2LM 936 147 107 578 780 116 30 2 2 -256 153 208<br />
S_CLS2LM 877 66 69 539 473 45 332 179 89 371 224 196<br />
855 75 78 -590 566 61 -416 281 158 73 9 9<br />
ZANSOUTH 932 165 113 500 616 97 355 311 254 -45 5 7<br />
477
Only a single dimension <strong>of</strong> high inertia remains in the data, which is dominated by the<br />
inertia <strong>of</strong> bands <strong>of</strong> punctation, and it produces groupings much like Component 2 in the<br />
previous plots, with Gharanga at one extreme (tailed by Kopo, with a tiny sample, and<br />
Hoghoi, then Zangana North). Zangana South is at the opposite extreme, with Paniavile<br />
and Miho nearby, and Nusa Roviana tending in the same direction, although sample size<br />
for that site is negligible. <strong>The</strong> second component has only 14% inertia, and mainly involves<br />
the count <strong>of</strong> impression <strong>of</strong> the lip into a wave pattern, which occurs five times at Gharanga<br />
and twice at Zangana South.<br />
Discussion <strong>of</strong> Correspondence Analyses: the Alternatives:<br />
<strong>The</strong> search for temporal variability centered on the exclusion <strong>of</strong> form-correlated decorative<br />
attributes. Exclusion <strong>of</strong> CLASS1LM and STAMPING made no difference to the<br />
Figure 148: Correspondence plot (sites), Honiavasa sample and form-correllated<br />
attributes omitted from data-set.<br />
478
ordination <strong>of</strong> sites or attributes, with the exception <strong>of</strong> Nusa Roviana, a small site sample<br />
with a single sherd on which two <strong>of</strong> the excluded attributes occurred (Class 1 linear motifs<br />
and stamping). <strong>The</strong> excluded sherd tended to make Nusa Roviana more similar to<br />
Honiavasa, the latter site being an outlier in all analyses.<br />
<br />
<strong>The</strong> effect on the seriation <strong>of</strong> excluding bands <strong>of</strong> lip impression is worthy <strong>of</strong> note.<br />
If these analysis are interpreted in temporal terms, then two temporal ordinations <strong>of</strong><br />
groupings in the data (groupings are summarized in Table 58) can be constructed, either<br />
A-B-C or A-C-B (or a mirror image <strong>of</strong> either <strong>of</strong> these, as seriation does not specify a<br />
direction for time). <strong>The</strong>se two alternatives, when combined with the notion that the<br />
Honiavasa (A) phase is most Lapita-like, and therefore earliest (for which there is some<br />
supporting evidence from TL), yield two alternative developmental modes. If Honiavasa<br />
Table 58: Summary <strong>of</strong> variability<br />
Phase<br />
Symbol<br />
A<br />
B<br />
C<br />
Multipurpose vessel<br />
decoration<br />
<br />
one-piece construction PINCHED<br />
BAND LIPS, IMPRESSED LIPS<br />
(especially the inner lip), some<br />
coarse wave deformation <strong>of</strong> the lip<br />
One-piece construction PINCHED<br />
BAND NECKS, CLASS 2<br />
LINEAR MOTIF RIMS, CLASS 1<br />
LINEAR MOTIF SHOULDERS,<br />
tall excurvate rim form with thin<br />
fragile lips, thick robust necks.<br />
Tendency towards wavedeformation<br />
<strong>of</strong> the lip, when<br />
preserved.<br />
One-piece construction<br />
PUNCTATE BAND in the vicinity<br />
if the neck, IMPRESSED LIPS,<br />
some lip impression into a wave<br />
pattern<br />
479<br />
Temporally associated<br />
decorative attributes and<br />
vessel forms/functions<br />
Robust Slab-Constructed carinated<br />
vessels with class 1 linear motifs,<br />
stamped decoration, and band <strong>of</strong><br />
applied nubbins at the neck, some<br />
compound or collar rims, some<br />
evidence for shallow bowls and robust<br />
storage vessels with low breakage<br />
rates.<br />
Various small bowls or flared rims,<br />
low discard rate <strong>of</strong> large robust vessels<br />
(storage vessels?)<br />
Gharanga short-rim variant, thinwalled<br />
in many cases, with multiple<br />
bands <strong>of</strong> fingernail pinching on the<br />
shoulder and bands <strong>of</strong> lip impression<br />
(especially the inner lip), occasional<br />
discard <strong>of</strong> punctate bowls.
is early, Miho style (B) is intermediate and Gharanga style(C) is late then the period over<br />
which discard occurred is characterized by a change from slab-constructed forms to one-<br />
piece forms, initially using a tall rim similar to that used in slab construction, but eventually<br />
replacing this for some purposes with a more rugged short, strongly outcurved rim form<br />
that survives well in the swash zone in spite <strong>of</strong> very thin construction in some cases.<br />
Alternatively, if Gharanga/Kopo style is earlier than Miho style, then the transition to thin,<br />
short-rim pots from Honiavasa heavy excurvate-rimmed pots is more rapid, the loss <strong>of</strong> the<br />
heavily carinated Class 2 decorated pots or jars is also relatively rapid, and by the terminal<br />
stage, Miho-style linear motifs (unbounded this time) reappear, along with tall rims arising<br />
from Kopo-style tall rimmed punctate-decorated vessels.<br />
Conclusions: Integrating 14 C, TL and Seriation:<br />
14 C data suggested we can be reasonably confident that the Roviana intertidal sites have<br />
a total occupation span <strong>of</strong> at least 160 years, unless there is an old-wood problem with the<br />
smoke-derived date from Hoghoi. It was not possible to associate either <strong>of</strong> these dates<br />
with a particular style <strong>of</strong> ceramic production, although the thin-necked pot from Hoghoi<br />
with a tall everted rim seems intermediate in some attributes between the Gharanga style<br />
and either Honiavasa plain pottery or Miho-style pottery.<br />
<strong>The</strong> addition <strong>of</strong> eight TL dates expands the minimum occupation span <strong>of</strong> the<br />
intertidal sites to 611 years. Also, if the TL ages from quartz-calcite hybrid tempered<br />
sherds uncorrected for fading are taken as an ordinal-scale indication <strong>of</strong> age ( Dickinson’s<br />
conclusion was that these tempers all originate from a single coastline, which might<br />
provide some confidence that, although they fade anomalously, they may all fade in the<br />
same anomalous way), then the Honiavasa site is likely to be oldest, while some <strong>of</strong> the<br />
pottery from Miho, Paniavile and Zangana South is mostly intermediate in age between<br />
Honiavasa and Gharanga. Other pottery from Gharanga and Paniavile is younger still.<br />
More TL dates are necessary to test this argument, and could pr<strong>of</strong>itably use some <strong>of</strong> the<br />
decorated Q-C tempered sherds held in the collections at present, especially sherd MH033,<br />
480
decorated with a band <strong>of</strong> punctation at the neck.<br />
styles, either:<br />
or:<br />
Seriation analyses can be summarized as producing two different sequences <strong>of</strong><br />
HoniavasaMihoGharanga/Kopo<br />
HoniavasaGharanga/KopoMiho<br />
(the seriation was more fine-grained than this, being attribute-based, but the major<br />
chronological issue can be phrased in terms <strong>of</strong> styles). TL results suggest that the<br />
Honiavasa site is older than the others, and that the Miho style predates Gharanga/Kopo,<br />
although more evidence is needed to test this. This conclusion rests on an assumption that<br />
the feldspar grains in quartz-calcite hybrid tempers used in TL dating all fade in a similar<br />
manner, allowing a relative chronology.<br />
<strong>The</strong> seriation analysis tended to show groups <strong>of</strong> attributes clustering into styles,<br />
which suggested the homologous links between Honiavasa site and the other sites were<br />
few (but present), which is most economically explained as resulting from historical<br />
discontinuity in the sample. Also, the phyletic links that could show the relationship<br />
between Miho and Gharanga/Kopo styles in detail were largely absent (but not completely<br />
absent). This again is interpreted as a historical gap in the sample, rather than indicating<br />
any essential, immutable reality to the Miho and Gharanga types. In the absence <strong>of</strong><br />
evidence to the contrary, any conclusion that these data clusters represent cultural<br />
replacement <strong>of</strong> any sort is unwarranted. As discussed in the next <strong>chapter</strong>, there is non-<br />
ceramic evidence for heritable continuity across the sample. If the corpus <strong>of</strong> sites were<br />
much larger but these discontinuities persisted, then explanation for the discrete styles<br />
might be sought in historical events. <strong>The</strong> current state <strong>of</strong> sampling <strong>of</strong> the New Georgia<br />
group does not rule out gradual continuous changes in some form and decoration<br />
attributes over time for the larger region. Whether such changes all happened at Roviana<br />
Lagoon, but outside the current site sample, or whether the action moved <strong>of</strong>fstage, beyond<br />
the sampling frame, at some times, is unknown from the present evidence.<br />
481
482
Introduction:<br />
CHAPTER 13:<br />
SUMMARY AND CONCLUSIONS<br />
<strong>The</strong> key question laid out in Chapter One was whether there was a gap in the past<br />
distribution <strong>of</strong> Lapita, consistent with an avoidance or leap-frogging colonization model,<br />
or whether the near-oceanic Solomon Islands, as represented by New Georgia, were<br />
settled early in the Lapita ceramic series. As the early Roviana ceramics are all in the sea<br />
there were special questions concerning formation processes <strong>of</strong> this particular record, and<br />
also a need for adaptation <strong>of</strong> archaeological method in some respects to fit this<br />
archaeological landscape.<br />
<strong>The</strong> broad question concerning Lapita distribution across the region <strong>of</strong> Near<br />
Oceania was broken down into a number <strong>of</strong> more specific topics: how can the Roviana<br />
intertidal record be interpreted in behavioural terms (and to what extent is patterning due<br />
to natural rather than cultural formation processes)? How can a high-resolution chronology<br />
be constructed to allow fine chronological control and enable the historicist approaches<br />
that are so <strong>of</strong>ten attempted using inappropriately coarse radiocarbon chronologies? How<br />
much <strong>of</strong> a sample is needed, and what sort <strong>of</strong> sample? how can sample size be quantified?<br />
What potential is there for bias given particular sets <strong>of</strong> formation processes, and what are<br />
the likely biases present in the samples?<br />
In tackling these questions, conclusions <strong>of</strong> significance to resource management<br />
and future research were drawn regarding the fragility <strong>of</strong> these swash-zone sites, the<br />
poverty <strong>of</strong> information remaining, and extreme sensitivity in terms <strong>of</strong> information content<br />
to further removal <strong>of</strong> material. In spite <strong>of</strong> the apparent poverty <strong>of</strong> the record, the high<br />
archaeological visibility <strong>of</strong> lag deposits <strong>of</strong> pottery and lithics means there is unusually<br />
483
good potential for fine-grained chrono-stylistic studies.<br />
Summary:<br />
A review <strong>of</strong> approaches to ceramic classification and seriation method in Chapter<br />
1 identified a need for materialist approaches to artifact classification capable <strong>of</strong> capturing<br />
stylistic drift, compatible with a descent-with-modification mode <strong>of</strong> ceramic change, in<br />
order to develop fine-grained stylistic chronology. <strong>The</strong> review identified a need to control<br />
for functional variability, as a means <strong>of</strong> isolating temporal variability.<br />
A review <strong>of</strong> theories <strong>of</strong> form-function correlation found that functional classes were<br />
best kept broad, and that use-life theory predicts that general purpose vessels that included<br />
a cooking function should dominate vessel samples. Seriation should focus on forms that<br />
fell into this broad use-category, and on decoration that cross-cut form-function classes.<br />
<strong>The</strong> review identified a disjuncture between descriptive/classificatory units for<br />
Lapita and post-Lapita (Mead’s system or Anson’s system principally for Lapita, and the<br />
Frost-Irwin attribute combination approach for later pottery). In order to span the Lapita-<br />
post-Lapita period these differences needed to be resolved. <strong>The</strong> relative simplicity <strong>of</strong><br />
decoration on utilitarian pottery, which dominates post-Lapita assemblages at least, meant<br />
that descriptive and classificatory schemes needed to be sensitive to slight differences in<br />
decoration. Also, in relation to units <strong>of</strong> quantification, there is a need for measures <strong>of</strong><br />
attribute frequency and sample sizes that are not biased by differences in assemblage<br />
brokenness, or differential preservation <strong>of</strong> parts <strong>of</strong> the vessel. Accordingly, it was<br />
determined that units <strong>of</strong> description and classification should pay close attention to<br />
location on the vessel, and should focus on those parts <strong>of</strong> the vessel that preserve well and<br />
are easily identified even as small fragments, allowing the structure <strong>of</strong> decoration across<br />
the vessel as represented by sherd samples to be captured.<br />
A review <strong>of</strong> temporal constructs for Lapita (Chapter 2) found that temporal<br />
484
esolution <strong>of</strong> the available chronologies was alarmingly similar to the various definitions<br />
<strong>of</strong> Lapita, and could be broadly characterized as culture-historical in resolution, with<br />
Lapita continuing to have the status <strong>of</strong> a secure temporal horizon, but with potential in<br />
current higher-resolution constructions <strong>of</strong> Lapita temporal variability for mis-assignment<br />
<strong>of</strong> other dimensions <strong>of</strong> variability to time. This is partly a result a lack <strong>of</strong> clear sample<br />
evaluation, an important step, particularly when constructing occurrence seriations and<br />
interpreting motif-sharing. <strong>The</strong>se factors undermine the security <strong>of</strong> current constructions<br />
<strong>of</strong> the “Lapita Ceramic Series”.<br />
While temporal constructs are seen as secure by several investigators, for example,<br />
Anson's “Early Far Western/Western” distinction (Anson 1983, 1986, 1987, 1990, 2000),<br />
Summerhayes’ Early/Middle/Late universal Lapita series (Summerhayes 2000a, 2001,<br />
2002), and Sand’s Southern Lapita series for New Caledonia (Sand 1997a, 1999, 2000,<br />
2001), there is a tendency for site samples to be slotted in to these schemes without much<br />
control for functional variation especially, and geographic variation to a lesser extent.<br />
While I would not suggest these constructions are necessarily wrong, these are seen here<br />
as important works in progress rather than a secure foundation for comparative dating <strong>of</strong><br />
the Roviana materials.<br />
While Best’s sequence at Lakeba was judged to be a more secure stratigraphic<br />
demonstration <strong>of</strong> temporal changes from Lapita to other styles, uncertainties remain, and<br />
the Eastern chronology need not apply elsewhere. Constructing a phyletic seriation to<br />
argue that it does, as Spriggs does in “<strong>The</strong> Changing Face <strong>of</strong> Lapita” (Spriggs 1990) and<br />
as Best does in “A View from the East” (Best 2002), is putting the cart before the horse.<br />
Secure regional sequences need to be compared and cross-matched to assess whether there<br />
are universal Lapita-wide parallel changes before doing this. It cannot be assumed that<br />
there are such changes, to justify picking a phyletic series from across vast swaths <strong>of</strong> the<br />
Pacific, because this assumes the answer to a primary question to be asked <strong>of</strong> the material.<br />
In this final <strong>chapter</strong> some <strong>of</strong> the difficulties encountered in trying to reconcile the Roviana<br />
485
early ceramics with other temporal constructs will be elaborated.<br />
An increased emphasis on oceanography (wave processes) and the consequences<br />
<strong>of</strong> these for archaeological preservation and probability <strong>of</strong> detection was advocated in<br />
Chapter 3. Previous survey method and results in near Oceania were reviewed, and the<br />
results <strong>of</strong> two surveys <strong>of</strong> the Roviana Lagoon region were presented in terms <strong>of</strong> a sample-<br />
surveying approach. For Lapita, the informant prospection survey (Roviana survey)<br />
obtained similar results to other informant-prospection surveys in the Near-Oceanic<br />
Solomon Islands, i.e. no hits (other than a single sherd), in spite <strong>of</strong> informants volunteering<br />
the locations <strong>of</strong> a number <strong>of</strong> post-Lapita or Lapita-derived intertidal sites. More intensive<br />
intertidal coverage <strong>of</strong> the complex coastline <strong>of</strong> the more restricted Kaliquongu area located<br />
a single Lapita site.<br />
Although a sample <strong>of</strong> one suggests attribution <strong>of</strong> different results to different<br />
survey methods should be made with caution, I think we have a better idea, as a result <strong>of</strong><br />
the Kaliquongu survey, how to locate sites <strong>of</strong> the Lapita period in the New Georgia area.<br />
Despite the small site-sample size for Lapita, Lapita recorded site density (as opposed to<br />
past site density) was well within the range recorded elsewhere in Near Oceania, <strong>of</strong>fering<br />
some evidence in support <strong>of</strong> a model <strong>of</strong> rapid and comprehensive spread <strong>of</strong> Lapita rather<br />
than leap-frog colonization or avoidance modes. Given that the Kaliquongu survey covered<br />
only a small fraction <strong>of</strong> the New Georgia intertidal zone and only one major reef passage<br />
into the chain <strong>of</strong> landlocked lagoons encircling New Georgia, it is highly likely that<br />
temporal-stylistic diversity will increase with further survey <strong>of</strong> similar locations around<br />
New Georgia, which, if this turns out on further analysis to be the case, would in turn<br />
strengthen support for a model <strong>of</strong> early, rapid spread. While speculative, this <strong>of</strong>fers a<br />
testable hypothesis for future work.<br />
What is clear from the Kaliquongu results is that we would need a lot more<br />
evidence than currently in hand to rule out early Lapita settlement <strong>of</strong> the New Georgia<br />
group. Furthermore, given the results <strong>of</strong> the review <strong>of</strong> the Lapita ceramic series, it is<br />
486
difficult to say on current evidence just how old the Honiavasa materials are, although they<br />
suggest a sufficient period <strong>of</strong> divergence from an early-Lapita ancestor to lack pedestalled<br />
bowls and flat-based dishes (although absence from the current sample could be a sampling<br />
error given the small sample <strong>of</strong> slab-built and carinated Lapita forms currently recorded<br />
from Roviana Lagoon).<br />
Chapter 4 gave details <strong>of</strong> the ceramic database, and coding <strong>of</strong> units <strong>of</strong> ceramic<br />
description in the core files on the appended data CD. Core sherd attributes <strong>of</strong> size, weight<br />
and fabric were recorded in the “Master.db” table, while thickness, form and decoration<br />
were coded in separated tables organizing multiple records by vessel part for each sherd.<br />
This is a departure from the “diagnostic sherds’ approach where analysis focussed only on<br />
a small subset <strong>of</strong> sherds which met (usually unstated) criteria. <strong>The</strong> approach allows close<br />
control for part representation when querying the data, and allows convenient access to<br />
a large set <strong>of</strong> attributes through a form layout <strong>of</strong> master record and client tables. Salient<br />
attributes from this relational database were reconstructed into a single flat table for spatial<br />
analysis and seriation.<br />
Fabric classes included several placering variants <strong>of</strong> volcanic sands, most <strong>of</strong> which<br />
are thought to be local, and an exotic quartz-calcite hybrid temper, thought on current<br />
evidence to be from a continental location to the west, but other less exotic sources cannot<br />
be ruled out on current evidence. <strong>The</strong>re were few calcite-only tempered sherds. <strong>The</strong><br />
quartz-calcite tempers were much more quartzose than any quartz-calcite hybrid tempers<br />
from the Bismarck Archipelago, with ratio <strong>of</strong> coarse K-feldspar-to-Plagioclase suggestive<br />
<strong>of</strong> granitic origin. This raises the prospect <strong>of</strong> low frequency <strong>of</strong> trans-Solomon Sea or trans-<br />
Coral-Sea pottery transfer, which would accord well with some <strong>of</strong> the materials from<br />
Reef-Santa-Cruz Lapita sites, and with Irwin’s voyaging models, but is anomalous in that<br />
no other sites in Near Oceania are known to have yielded similar pottery temper sand<br />
evidence (Felgate & Dickinson 2001). <strong>The</strong> explanation preferred by Felgate and<br />
Dickinson is thus unlikely to find much favour with the wider Pacific archaeological<br />
487
audience unless such a source <strong>of</strong> pottery is physically located, which is beyond the scope<br />
<strong>of</strong> the current research, so the implications <strong>of</strong> this temper remain an open question.<br />
Form <strong>of</strong> vessel parts was recorded with the aim <strong>of</strong> providing sufficient information<br />
to enable a reasonably accurate depiction <strong>of</strong> vessel shape variability. Nominal<br />
morphological classes such as everted rim, restricted neck, carination, etc were thus<br />
supplemented with a series <strong>of</strong> orientation, curvature and distance measurements. <strong>The</strong>se<br />
together with thickness measurements at a series <strong>of</strong> defined points enable more traditional<br />
nominal classes <strong>of</strong> vessel form to be deconstructed for the purpose <strong>of</strong> an exploratory<br />
analysis <strong>of</strong> variability, in tune with the materialist, evolutionist approach to identification<br />
<strong>of</strong> stylistic and functional variation.<br />
Similarly, decoration was coded by vessel part, employing formulaic descriptions<br />
<strong>of</strong> the patterned layout <strong>of</strong> elements for simple banded decorations, and reference to<br />
drawings for more complex or larger, or fragmentary designs. <strong>The</strong>re was a conscious effort<br />
in coming up with this scheme to separate decorative technique from decorative element,<br />
and from the layout <strong>of</strong> elements. Metric variability <strong>of</strong> formulaic banded designs was<br />
recorded, some <strong>of</strong> which information turned out to be salient to some <strong>of</strong> the basic research<br />
questions when analysed in Chapter 8.<br />
Chapter 5 provided a detailed investigation <strong>of</strong> vessel brokenness and completeness,<br />
with several practical outcomes important to the conclusions <strong>of</strong> the thesis. While many <strong>of</strong><br />
the samples comprised quite large sherds, vessel completeness as evidenced by<br />
construction <strong>of</strong> lip-sherd vessel families was extremely low, with mostly singleton lip<br />
sherds present. A total breakage population was estimated by constructing a virtual<br />
assemblage <strong>of</strong> whole rims broken in the manner <strong>of</strong> the sample, and iteratively sampling it<br />
at various levels to see what range <strong>of</strong> sampling fractions yielded completeness data similar<br />
to that <strong>of</strong> the real sample (Felgate & Bickler n.d). A statistical approach using the same<br />
completeness data and a “number <strong>of</strong> species” algorithm obtained similar results to the<br />
simulation, but the simulation results were preferred as more precise due to the use<br />
488
<strong>of</strong> brokenness information.<br />
Around 99% <strong>of</strong> the sherdage <strong>of</strong> the original breakage population has not made it<br />
into the sample, and the breakage population vessel count would have been at least five<br />
times larger than the vessel count represented in the lip sample. As little <strong>of</strong> the missing<br />
pottery was buried on site, a taphonomic regime resulting in substantial sherd attrition was<br />
inferred, leading to the conclusion that there was high potential for taphonomic bias in the<br />
recovered sample. Also, accumulations-based inferences that use quantities <strong>of</strong> recovered<br />
pottery to estimate intensity/duration <strong>of</strong> occupation are not applicable to data <strong>of</strong> this<br />
quality. This finding led to a question being raised in regard to Wickler’s inference <strong>of</strong> low<br />
intensity <strong>of</strong> occupation for similar sites in the Buka region. Also, comparisons using<br />
relative abundances <strong>of</strong> styles should be made with caution, particularly when comparing<br />
with well-preserved sites such as the Arawe Islands stilt villages and the Mussau stilt<br />
villages, where preservation appears to be extremely good and taphonomic bias is less<br />
likely. Similarly, comparison with terrestrial sites should be made with extreme caution,<br />
as while biases are likely in heavily gardened sites like those <strong>of</strong> the Reef/Santa-Cruz area,<br />
we do not yet know whether the biases are the same, or whether they are likely to create<br />
differences in stylistic composition <strong>of</strong> assemblages.<br />
How fragile is the site type? Chapter 5 informed on the state <strong>of</strong> preservation <strong>of</strong><br />
samples in a particular setting, the Roviana lagoon, where wave exposure is uniformly low<br />
and sea level history is similar all sites in the study. Even in this sheltered setting, the sites<br />
are in a very poor state <strong>of</strong> preservation, although visibility for those that remain exposed<br />
on the surface is high. This finding is <strong>of</strong> fundamental importance to the central question<br />
<strong>of</strong> this thesis, how to interpret the Lapita gap in Near-Oceania Solomon<br />
Islands/Bougainville. If the sites are poorly preserved in this sheltered setting, along most<br />
other coastlines exposure to a height level in relation to wave processes where<br />
archaeological visibility is high would result in total destruction <strong>of</strong> the ceramic component<br />
in most cases. Along such coastlines, where sea levels have fallen since Lapita times we<br />
489
should be looking for vestiges <strong>of</strong> sites such as stormwash accumulations rather than in-situ<br />
deposits, or deeply submerged sites if there is a history <strong>of</strong> relative rise in sea levels.<br />
<strong>The</strong> low state <strong>of</strong> completeness <strong>of</strong> the Roviana intertidal pottery provides some<br />
assurance that sherd counts are mostly independent observations. Sherd counts thus<br />
provide a useful measure <strong>of</strong> observation sample size, and the relative frequencies <strong>of</strong> pottery<br />
attributes in site samples. It should be noted that the same cannot be said <strong>of</strong> well preserved<br />
sites, where many sherds in the sample can derive from a single vessel.<br />
Chapter 6 extended the analysis <strong>of</strong> site preservation, adding the wave exposure<br />
component to a model <strong>of</strong> preservation. While information on prevailing winds conflicted,<br />
and environmental information on the extent and depths <strong>of</strong> reefs, sheltering sediments and<br />
seagrass beds was not <strong>of</strong> sufficient quality to generate precise comparative results, it is<br />
clear that none <strong>of</strong> the sites where pottery was found in quantity have more than 6 km <strong>of</strong><br />
wave exposure, and most have mitigating factors that prevent any substantial waves from<br />
affecting the sites at low tide. Thus the results <strong>of</strong> Chapter 5 can be extended to other<br />
regions as follows: where fetch is greater than 6km, and sea-levels have undergone a<br />
similar fall as in the Roviana setting, preserved intertidal ceramic sites are not expected to<br />
have survived except as lithic scatters or storm-redeposited material. This is just a<br />
hypothesis at present given some <strong>of</strong> the measurement uncertainties, but provides a rule-<strong>of</strong>-<br />
thumb starting point from which future survey <strong>of</strong> this sort can be targeted towards<br />
relatively sheltered locations.<br />
Analysis <strong>of</strong> formation process evidence in Chapter 7 considered a series <strong>of</strong> models<br />
<strong>of</strong> cultural formation processes, followed by a consideration <strong>of</strong> the evidence pertaining to<br />
natural or taphonomic processes as they affected ceramics. <strong>The</strong> absence <strong>of</strong> adjacent<br />
evidence on land, and the absence <strong>of</strong> erosion features other than solution scarring <strong>of</strong> Plio-<br />
Pleistocene upraised reefs are evidence against erosion <strong>of</strong> terrestrial sites into the sea. <strong>The</strong><br />
fresh state <strong>of</strong> preservation <strong>of</strong> Acropora corals on the upraised shore platform to landward<br />
<strong>of</strong> most sites suggests a recent high stand, the upper limit <strong>of</strong> coral growth, at which time<br />
490
the site locations would have been s<strong>of</strong>t backreef lagoonal sediments unaffected by swash<br />
processes. Paleoshoreline data (Mann et al. 1998) strongly support this interpretation, and<br />
suggest a high stand <strong>of</strong> at least 1.5m above present at 3000 BP, probably more.<br />
Subsequent sea-level fall led to to emergence <strong>of</strong> the backreef ceramic sites into the swash<br />
zone, such as it is. Coral growth accretions on some sherds also support this conclusion,<br />
as no modern coral growth occurs at these levels. Additional evidence against terrestrial<br />
settlement is complete absence <strong>of</strong> acid-leaching <strong>of</strong> carbonate grains (disregarding surface<br />
etching caused by cleaning sherds in acetic acid).<br />
Taking all these factors together, including the recent information from geology<br />
which was unavailable at the outset <strong>of</strong> the study, coral artificial islets seem an unlikely<br />
formation process for these sites, as the water would have been at least chest-deep at low<br />
tide, and probably deeper, requiring phenomenal labour to construct walling for a<br />
settlement. Stilt dwellings over shallow water, with piles set in s<strong>of</strong>t sediments now<br />
removed by swash processes as a result <strong>of</strong> uplift, are the most likely scenario.<br />
<strong>The</strong> artefact inventory <strong>of</strong> all intertidal sites was similar, dominated by pottery and<br />
lithic manuports, with rare adzes, flakes <strong>of</strong> meta-basalt or metamorphosed mudstone, very<br />
rare chert flakes, and a variety <strong>of</strong> abrader manuports. Coral abraders were not seen, but<br />
the corraline gravels (dominated by small Acropora fragments)forming the matrix <strong>of</strong> most<br />
sites would have made these difficult to find. <strong>The</strong>re was no evidence for site abandonment<br />
in the artefact inventory, or none that had survived postdepositional processes. All<br />
complete adzes were heavily resharpened in comparison to one unused preform in a private<br />
collection, suggesting sites were the focus <strong>of</strong> extended settlement resulting in the discard<br />
<strong>of</strong> ceramics, ovenstones/net weights and worn out adzes.<br />
Taphonomic analysis and spatial analysis (the latter at Zangana) <strong>of</strong>fered support<br />
for a transportation model <strong>of</strong> sherd attrition, while presence <strong>of</strong> ferromagnesian mineral<br />
grains in a sediment sample from the coralline Hoghoi shoreline suggested dissolution<br />
was also a factor in sherd attrition at some stage in the history <strong>of</strong> the sites. Taphonomic<br />
491
analysis <strong>of</strong> sherd samples most clearly supported the collector and collection intensity<br />
models, suggesting the potsherd scatters are highly sensitive to collection events. This<br />
finding is important for archaeological resource management and also places an additional<br />
burden on future investigators to thoroughly document the provenance <strong>of</strong> each item<br />
collected, and attend to desalination procedures. We cannot pick up a sample expecting<br />
to be able to go back later and get more <strong>of</strong> the same. This also suggests proactive<br />
education on the significance and fragility <strong>of</strong> the resource is desirable at the local level.<br />
Were artifact collectors to be successful in illegally selling items collected from such sites,<br />
we could expect the accessible surface-site information resource to rapidly degrade to an<br />
irretrievable level.<br />
Taphonomic analysis in Chapter 8 also provided information on the types <strong>of</strong> bias<br />
likely to be present in the ceramic samples. Strength variation between temper classes as<br />
evidenced by sherd size was low, but sherd thickness was clearly related to sherd size,<br />
suggesting thinner sherds are weaker and break more easily, and that thinner pottery is<br />
likely, or thinner parts <strong>of</strong> pots are likely to be underrespresented in the lag deposits as a<br />
result. Data was presented showing that some forms were stronger than others: Miho style<br />
pottery tended to be thicker at the neck and thin and fragile at the lip, suggesting this style<br />
is principally represented by neck sherds. Gharanga-style vessels with short, heavily-<br />
excurvate-to-rolled rims preserve well due to form strength, and lips <strong>of</strong> this form are thus<br />
more commonly attached to rims, leading to a larger count <strong>of</strong> Lip/Rim/Neck/Shoulder<br />
sherds.<br />
<strong>The</strong> analysis <strong>of</strong> pottery form variability and function in Chapter 8 focused on<br />
vertical contour <strong>of</strong> the upper vessel, and also used negative evidence, the absence <strong>of</strong> flat<br />
based sherds and stands (although there was one possible stand from Zangana). Six broad<br />
functional vessel form classes were identified on the basis <strong>of</strong> upper body contour<br />
variation, and a seventh was defined using body sherd thickness and form. <strong>The</strong>se classes<br />
were not all mutually exclusive, as some <strong>of</strong> the larger and more robust upper vessel sherds<br />
492
perhaps should have been classed together with the Form 7 sherds. <strong>The</strong> aims <strong>of</strong> the<br />
functional classification were the identification <strong>of</strong> a general purpose class that can be<br />
expected to have high use rates, exposure to thermal shock and short use life/ high discard<br />
rate, for use in form-controlled seriation analysis. A sub-class <strong>of</strong> Form 2 and a sub-class<br />
<strong>of</strong> Form 6 were judged to best meet these criteria, and were used for seriation analysis in<br />
Chapter 12.<br />
In Chapter 9 decorative variability was explored, and the Zangana site was split<br />
into two areas on the basis <strong>of</strong> decorative differences. Class 1 linear motifs<br />
(bounded/defined with double-line zone-markers, either dentate or incised) were largely<br />
restricted to the Honiavasa site, where they always occurred in association with carinated<br />
Form 1 vessels, as did a band <strong>of</strong> circular nubbins at the neck. It was these sherds that had<br />
led to the assignment <strong>of</strong> the site to the Lapita period from first discovery. Also present in<br />
some quantity were everted-rim necked vessels with either lip notching/impression <strong>of</strong><br />
various sorts, or a band <strong>of</strong> pinching on the outer edge <strong>of</strong> the lip.<br />
<strong>The</strong>re were only two examples <strong>of</strong> deformation <strong>of</strong> the lip into a wave form at<br />
Honiavasa, both <strong>of</strong> these coarsely executed, whereas this decorative technique was<br />
ubiquitous at Nusa Roviana, Paniavile and Zangana. Deformation <strong>of</strong> the lip into a<br />
discontinuous series <strong>of</strong> small spout-like decorative finger impressions was most common<br />
in Miho site, and occurred in all other sites for which sample size was adequate. <strong>The</strong>re<br />
were only two occurrences <strong>of</strong> this attribute at Honiavasa, compared with 22 occurrences<br />
at Miho. Class 2 linear motifs on the shoulder or rim were common at Miho, Paniavile and<br />
Zangana South, and absent from Honiavasa. A single band <strong>of</strong> fingernail pinching at the<br />
neck was common in the same sites as Class 2 linear motifs, and <strong>of</strong>ten occurred on the<br />
same sherds. This complex <strong>of</strong> decoration was termed the Miho decorative class or “Miho<br />
style” after the site where it was first recognised.<br />
<strong>The</strong> decorative attribute <strong>of</strong> a single punctate band at the neck was common in two<br />
sites, and was associated either with tall weakly everted rims above a slight neck<br />
493
estriction (Kopo style) or with short heavily everted rims (Gharanga style), the latter <strong>of</strong>ten<br />
associated also with multiple bands <strong>of</strong> fingernail pinching on the shoulder. Gharanga/Kopo<br />
style sherds showed a tendency towards thinner necks than Miho style, supporting the<br />
conclusions in Chapter 7 regarding the nature <strong>of</strong> stylistic-taphonomic bias.<br />
A heterogeneous sample <strong>of</strong> applied decoration proved difficult to classify and was<br />
omitted from the attributes used in seriations, although there is some discussion <strong>of</strong> possible<br />
temporal variability given below.<br />
Analysis <strong>of</strong> lip impression/crenation/notching in relation to lip form found a<br />
correlation between lip orientation angle and the location <strong>of</strong> decoration which cross-cut<br />
drastic differences in vessel form and decoration, suggesting that lip impression attributes<br />
could cause faulty ordering <strong>of</strong> a seriation. Location <strong>of</strong> lip impression bands seemed to be<br />
correlated with lip orientation, but as this measurement was dependent on rim orientation<br />
measurements, which had poor precision, this requires confirmation from an enlarged<br />
sample <strong>of</strong> sherds <strong>of</strong> 10% EVE or greater (below which rim orientation is difficult to<br />
measure).<br />
Neck decoration for all styles (Honiavasa, Miho and Gharanga-Kopo styles) was<br />
dominated by a band <strong>of</strong> circular elements encircling the neck, while technique <strong>of</strong><br />
execution varied between these styles. Within an evolutionary definition <strong>of</strong> style and<br />
function, following Dunnell, this suggests, if these are temporal types, that the pattern is<br />
constrained and remaining constant, while the technique <strong>of</strong> execution varies, either<br />
through drift or selection. <strong>The</strong> dominance <strong>of</strong> this pattern at the neck <strong>of</strong> the pot and the<br />
temporal constancy suggests the pattern is fundamental to the decorative ideas <strong>of</strong> the<br />
potters, and is functional in some way. A parallel was seen with Spriggs’ “changing face<br />
<strong>of</strong> Lapita” in that these circular elements could be seen as the eyes in the Lapita face, and<br />
a protective ide<strong>of</strong>unction was suggested, where the eyes were an example <strong>of</strong> the<br />
“technology <strong>of</strong> enchantment or the enchantment <strong>of</strong> technology” (Gell 1992), perhaps<br />
conferring protection from illness or poisoning to the contents <strong>of</strong> the pot and to eaters <strong>of</strong><br />
494
the contents.<br />
Analysis <strong>of</strong> lithics in Chapter 10 focussed on petrographic classification <strong>of</strong> lithic<br />
manuports, analysis <strong>of</strong> size distribution <strong>of</strong> these, morphological and petrographic<br />
description <strong>of</strong> adzes, and petrographic description <strong>of</strong> sandstone abrader fragments. <strong>The</strong><br />
diversity <strong>of</strong> sources represented suggests enormous potential <strong>of</strong> such sites for lithic<br />
sourcing studies, as a means <strong>of</strong> reconstructing patterns <strong>of</strong> raw material transport in the<br />
past. This conclusion should add archaeological value to these intertidal/underwater<br />
surface scatters, and encourage more intensive recording and sampling in the course <strong>of</strong><br />
future research and resource management.<br />
Some <strong>of</strong> the unfractured water-rounded lithic manuports seemed too small to have<br />
been procured as ovenstones, although small stones are sometimes collected for this<br />
purpose in Oceania today (Nojima 2002: Pers. Comm.). Another functional possibility is<br />
as net weights for large nets such as used ethnographically for turtle or dugong, both <strong>of</strong><br />
which are common today in the Lagoon, dugong especially in extensive seagrass meadows<br />
adjacent to Hoghoi.<br />
While most manuports probably were sourced from Rendova, some may have<br />
originated from further afield (Viru Harbour or Kolombangara, both about 50km distant),<br />
although this suggestion requires testing through examination <strong>of</strong> likely source streams.<br />
Adzes are from a diversity <strong>of</strong> sources, showing a preference for metamorphosed<br />
siltstones or metamorphosed fine-grained volcanic rocks. While some <strong>of</strong> these adze rocks<br />
are conceivably from Tertiary sedimentary formations <strong>of</strong> the forearc chain <strong>of</strong><br />
Ranonnga/Southern Rendova/Tetepare, 15km or more distant from Roviana, the<br />
recrystallization evident petrographically may indicate an older origin further afield,<br />
perhaps among the complex geology <strong>of</strong> Choiseul, or possibly some <strong>of</strong> these result from<br />
more local contact metamorphism associated with plutonic intrusions.<br />
Adze forms are consistent with the Buka/Nissan Lapita reef-site evidence, and<br />
suggest heritable continuity with Lapita, supporting Reeve’s contention that the Paniavile<br />
495
material was Lapita-derived. This evidence for heritable continuity was also useful for<br />
seriation, as a demonstration <strong>of</strong> heritable continuity is important for evolution-based<br />
techniques such as seriation. <strong>The</strong> Near-Oceania stone adze sample from the Lapita period<br />
is now dominated by the materials obtained from reef/intertidal sites, suggesting surface<br />
archaeology holds the key to acquiring a large sample with which to study variability <strong>of</strong><br />
this rare artefact class in detail.<br />
Petrography <strong>of</strong> abrasives indicated procurement from a variety <strong>of</strong> sources, and the<br />
diversity <strong>of</strong> sources (no repeats) suggests the parent population <strong>of</strong> sources utilized is large.<br />
Our sample <strong>of</strong> these sorts <strong>of</strong> materials is far from saturated. Detailed petrographic and<br />
morphological comparisons with materials recovered by Wickler from Buka would be a<br />
rewarding exercise.<br />
Analysis <strong>of</strong> spatial structure (Chapter 11) in the ceramic data and the lithic Hoghoi<br />
data identified evidence for size and density sorting, supporting the sherd transport natural<br />
formation process model <strong>of</strong> Chapter 7. <strong>The</strong> analysis also looked for structured distribution<br />
<strong>of</strong> tempers and styles. <strong>The</strong>re was a clear separation <strong>of</strong> Miho decorative class from<br />
Gharanga/Kopo decorative class at Zangana, supporting division into Zangana North and<br />
Zangana South for seriation. By contrast, wave-deformation <strong>of</strong> the lip occurred ten times<br />
in each <strong>of</strong> the two halves, suggesting this attribute had a different temporal trajectory to<br />
the Miho decorative class despite association on some sherds with unbounded incised rim<br />
decoration.<br />
<strong>The</strong>re was no strong evidence for a spatial separation <strong>of</strong> the exotic quartz-calcite<br />
temper class from other tempers, which suggests, in conjunction with TL dating results,<br />
stylistic analysis, and occurrence in all sites, that this class <strong>of</strong> temper had an extended<br />
period <strong>of</strong> production and importation, despite the low sherd count.<br />
In Chapter 12 three lines <strong>of</strong> chronological evidence were integrated: these being<br />
radiocarbon dates; thermoluminescence; and ceramic seriation. C 14 data were interpreted<br />
as allowing reasonable confidence that the Roviana intertidal sites have an occupation<br />
496
span <strong>of</strong> at least 160 years, unless there is an old-wood problem with the determination on<br />
smoke-derived carbon from the surface <strong>of</strong> a sherd in the Hoghoi site. Direct dating was<br />
successful in two out <strong>of</strong> three attempts, with one case <strong>of</strong> dating <strong>of</strong> a sub-fossil organic<br />
inclusion clearly documented. <strong>The</strong> prospects for AMS direct dating <strong>of</strong> pottery are good,<br />
and throw the spotlight onto developing techniques for locating additional samples within<br />
sherds, both from the current collection and from future samples. Demonstrating that it is<br />
now possible to derive independent chronological data from the physical properties <strong>of</strong><br />
sherds and lithics in Near Oceania in this way, using TL and Radiocarbon, places added<br />
archaeological value on “disturbed” archaeological contexts such as intertidal lag scatters.<br />
<strong>The</strong> addition <strong>of</strong> eight TL dates expands the minimum occupation span <strong>of</strong> the<br />
intertidal pattern to 611years at 2 s.d.. If the TL dates from the Quartz-calcite temper are<br />
read as relative ages, disregarding corrections for anomalous fading, these suggest<br />
Honiavasa is the oldest site, while Miho-style pottery from Miho, Zangana and Paniavile<br />
is mostly slightly younger, with the youngest dates coming from Paniavile and Gharanga.<br />
More TL data are needed to test the notion that the uncorrected ages provide a relative<br />
chronology due to origin from a common source, and TL/carbon pairs would be an ideal<br />
scenario. If a systematic model <strong>of</strong> Luminescence fading can be developed for the<br />
feldspathic component <strong>of</strong> the quartz-calcite temper, then TL dating could become a key<br />
tool for defining Lapita chronology in the Western Solomon Islands.<br />
Seriation using CA produced different orderings depending on which attributes<br />
were included, and the results <strong>of</strong> seriation from which lip impressions were excluded were<br />
preferred, but this cannot be independently confirmed on current data. <strong>The</strong> tendency for<br />
ceramic variability to coalesce into three styles (Honiavasa, Miho and Gharanga/Kopo was<br />
partly an result <strong>of</strong> choice and sample sizes <strong>of</strong> attributes, but is likely also to be a<br />
consequence <strong>of</strong> a historically incomplete sample. Honiavasa site has very little in common<br />
ceramically with the other sites, although enough to suggest heritable continuity. Taken<br />
together with continuity in settlement types, the preferred explanation for the three styles<br />
497
is that we are missing the transition from Honiavasa utilitarian ceramics to Miho and<br />
Gharanga/Kopo utilitarian ceramics, unless some <strong>of</strong> the ubiquitous attributes span this gap,<br />
for example lip deformation on plain rims. Additional sampling <strong>of</strong> the regional record is<br />
required, including areas beyond the current sampling frame, as concluded on other<br />
evidence in Chapter 3 also. Should these stylistic discontinuities persist with additional<br />
sampling, this conclusion would have to be reevaluated.<br />
When the three different chronologies are synthesised, a Honiavasa>Miho-<br />
style>Gharanga/Kopo style sequence seems most likely. Wave-deformation <strong>of</strong> the lip<br />
seems likely to begin during the production span <strong>of</strong> the Miho style and continue till tall-rim<br />
Kopo-style production commences. Gharanga-style short-rim vessels may be<br />
contemporaneous functional variants <strong>of</strong> the Gharanga/Kopo period, or may be a<br />
development out <strong>of</strong> Kopo style, with multi-band pinching <strong>of</strong> the shoulder, and thinner<br />
construction being innovations following the reduction <strong>of</strong> rim height.<br />
Under Balfet’s model as applied by Irwin (Balfet 1965, Irwin 1985), this sequence<br />
may document replacement <strong>of</strong> domestic idiosyncratic low-skill production with specialist<br />
standardised technically-accomplished production. <strong>The</strong> chunky Form 1 carinated jars in the<br />
Honiavasa site with individualistic decoration, and heavy Form 6 pots, have been replaced,<br />
if this series is correct, by progressively thinner and simpler, more standardised pots <strong>of</strong> a<br />
more general-purpose nature, with better thermal properties as a result <strong>of</strong> reduction in<br />
thickness, also potentially allowing more rapid drying prior to firing.<br />
External Comparisons:<br />
Having critiqued the use <strong>of</strong> presence-absence data for comparing subsamples <strong>of</strong> past<br />
behaviour in Chapter 1 and 2, I am unwilling to make such comparisons using the<br />
Honiavasa data as I believe these will simply mislead. <strong>The</strong> Honiavasa sample contains a<br />
498
small corpus <strong>of</strong> carinated vessels with Lapita motifs, but sample size is insufficient for<br />
relative-frequency based motif comparisons also. Comparisons are thus restricted to some<br />
general statements. <strong>The</strong> Honiavasa carinated vessels with Class 1 motifs have clear<br />
homologous similarity to the Lapita ceramic horizon. <strong>The</strong>re are no flat bases or cylinder<br />
stands in the sample, but this does not mean these were absent in the past. <strong>The</strong> complex<br />
curvilinear motifs characteristic <strong>of</strong> early Lapita according to Best are not present in the<br />
sample. Ishimura’s recent analysis, by contrast, has these techniques <strong>of</strong> decoration evolving<br />
over a more lengthy period across the temporal dimension <strong>of</strong> the Lapita horizon.<br />
A number <strong>of</strong> Anson’s motifs, or similar motifs are present in the small Honiavasa<br />
carinated vessel sample. Mead motifs 16, something similar to motif 11 or 16.3, something<br />
similar to motif 15, and a geometric motif similar to 17.1 or 30 are present among the small<br />
sample <strong>of</strong> carinated Form 1 vessels. Mead’s 3-dimensional design elements N1.1, VB2,<br />
TB3.1, and TB3.3 occur in the Honiavasa sample. Design elements 1.2, 5 and 6 occur at<br />
Honiavasa. <strong>The</strong>se characteristics also serve to link Honiavasa to the Lapita ceramic<br />
horizon.<br />
Carinated vessel forms at Honiavasa are characteristic <strong>of</strong> Summerhayes’ early<br />
Lapita in the Bismarck Archipelago (Summerhayes Form V), but the rare occurrences <strong>of</strong><br />
dentate-stamping argue for a later date under the Summerhayes scheme. While this site<br />
was initially described as late-Lapita (Felgate 2002) I would now prefer to have a detailed<br />
regional sequence as a basis for cross matching, and am less confident in the temporal<br />
constructs <strong>of</strong> others than at the outset <strong>of</strong> analysis.<br />
I initially thought that the Roviana intertidal ceramics must date to around 500BC<br />
or younger, based on the information from Watom and Buka, and the first C14 date from<br />
Paniavile was consistent with this assignment. Feathers, however, was forced to reject his<br />
measurements from Honiavasa as too old, when he compared the results obtained to the<br />
expected age. Since then, a date likely to be around 800BC from Hoghoi has suggested<br />
that the Honiavasa Lapita subtantially predates 500BC, also that the TL measurements<br />
499
were giving a reasonable ordination chronology, and that Lapita is effectively over by<br />
about 800BC at Roviana Lagoon.<br />
As a result <strong>of</strong> this information, combined with the results <strong>of</strong> a critical review <strong>of</strong> the<br />
underpinnings <strong>of</strong> current constructs <strong>of</strong> Lapita variability, I am now inclined to put the<br />
absolute age <strong>of</strong> the Honiavasa carinated pottery at no younger than 850BC, and possibly<br />
older. Miho style pottery may be being produced by about 800BC, and my inclination,<br />
although as discussed in Chapter 12 this is unconfirmed by solid evidence, is to have<br />
Gharanga/Kopo style produced sometime between 800BC and 0.AD.<br />
Several occurrences <strong>of</strong> compound rims at Honiavasa provide similarity with rim<br />
pr<strong>of</strong>iles from Buka (especially Wickler 2001: Figure 4.2k from site DAF). <strong>The</strong>se rims are<br />
listed by Sand as characteristic <strong>of</strong> the Southern Lapita Province (Sand 2000), and the Buka<br />
and Roviana materials provide a suggestion that they are a more widespread trait. Poulsen<br />
calls these “flange” rims for Tonga, and these are Mead’s 3-dimensional design element<br />
2.8, defined from a Sigatoka sample, so they are clearly widely distributed, although the<br />
best examples are from the large whole vessels recently found in New Caledonia. Given<br />
the widespread distribution <strong>of</strong> these I hesitate to attribute them to a late date, and regard<br />
them as potentially part <strong>of</strong> the repertoire <strong>of</strong> rim/vessel forms <strong>of</strong> the initial spread into<br />
Remote Oceania, and thus having an origin <strong>of</strong> similar antiquity to the curvilinear/face<br />
motifs such as Mead M33. Alternatively, the widespread occurrence <strong>of</strong> these rim forms<br />
may arise through reticulate processes in RemoteOceania and the Near-Oceanic Solomon<br />
Islands/Buka. I am tempted to call these an indication <strong>of</strong> a “Central Lapita Province”,<br />
rather than Western Lapita, the latter something <strong>of</strong> a misnomer when Western Lapita is<br />
found as far east as Tonga (eg. Burley & Dickinson 2001).<br />
It is possible that the Roviana early ceramic sequence, in spite <strong>of</strong> the low number<br />
<strong>of</strong> sites, few dates and patchy sample, is amplifying the temporal resolution obtainable<br />
from excavated samples from elsewhere in Near Oceania for the Lapita/post-Lapita<br />
transition, as predicted by the extended review in the opening <strong>chapter</strong>s <strong>of</strong> the nature and<br />
500
potential <strong>of</strong> seriation <strong>of</strong> surface sites as an approach to the archaeology <strong>of</strong> a region. <strong>The</strong><br />
Garua pottery from FSZ looks like a mixture <strong>of</strong> attributes that are spatially separated in the<br />
Roviana surface sites. Wavy stamping, as illustrated by Summerhayes for the Garua sample<br />
(Summerhayes 2000a:146, Figure 9.2, sherds 1620, 1101, 1927) occurs on several<br />
Honiavasa sherds, in similarly abstract arrangements, with hints <strong>of</strong> use as a latitudinal zone<br />
marker in some cases. Also, Summerhayes' Sherd 1275, for instance, is a typical Roviana<br />
wave-deformed lip, occurring in high frequency at Miho, Paniavile, Zangana and Nusa<br />
Roviana, but only rarely and in an odd, coarse, heavy variant at the Honiavasa Lapita site.<br />
<strong>The</strong> separation <strong>of</strong> Honiavasa and Paniavile using linear motifs provides another clue<br />
that in the Roviana surface-collected data we may be seeing an amplification <strong>of</strong> temporal<br />
resolution in the relatively unrestricted landscape with plenty <strong>of</strong> places to move a stilt<br />
settlement to. Miho style is dominated by variants <strong>of</strong> Mead motif 18, while motif 24 also<br />
occurs several times. On the shoulder <strong>of</strong> Miho-style vessels it is not uncommon to see a<br />
simplified form <strong>of</strong> motif 16, being just the diagonal lines forming a zigzag, expressed using<br />
the single-line (Mead GZ3) rather than twinned lines. <strong>The</strong> motif listed as Mead M16.1<br />
(Green 1979) (different to that illustrated by Mead) is present at Paniavile, where Miho<br />
style predominates, and is associated with pinching at the neck, a feature diagnostic <strong>of</strong> the<br />
Miho style.<br />
This idea <strong>of</strong> temporal amplification obtains additional support from the<br />
occurrences <strong>of</strong> Gharanga/Kopo style punctation elsewhere. <strong>The</strong> characteristic punctation<br />
or stick impression diagnostic <strong>of</strong> Gharanga/Kopo style, where the still-plastic clay is<br />
deformed inwards slightly as a bump on the interior surface <strong>of</strong> the neck region occurs on<br />
more than 60 sherds in the sample, and is completely absent from the large Honiavasa<br />
vessel sample. This decoration is illustrated by Anson from Ambitle (Anson 1983: Figure<br />
X), and occurs on two sherds at site DES on Nissan (Wickler 2001:120), and at the EKQ<br />
site on Mussau (Kirch et al. 1991: Figure 4c). Compared to these samples, the Roviana<br />
501
sample, particularly from Gharanga and Hoghoi, represents the parent lode, unless this is<br />
simply analogous similarity, but if the latter is the case, why is it so rare elsewhere?<br />
<strong>The</strong> single quartz-calcite tempered example <strong>of</strong> this class <strong>of</strong> decoration may indicate<br />
an extended zone <strong>of</strong> production, design emulation, or exchange <strong>of</strong> potters as well as pots,<br />
as suggested by Summerhayes also from other data. Most <strong>of</strong> these vessels have placered<br />
volcanic temper, and it is not certain that there is much benefit in petrographic comparisons<br />
with the other examples from Buka and Papua New Guinea, although it never hurts to<br />
look.<br />
Gharanga-style multiple bands <strong>of</strong> fingernail impression on the shoulder occur in low<br />
frequency in many samples elsewhere, and has generally been investigated at the level <strong>of</strong><br />
decorative technique rather than at the level <strong>of</strong> structured pattern across the vessel (e.g.<br />
Anson 1983:Figure XI) although Bedford is much more specific in this respect (see for<br />
example Bedford 2000: early and late Ifo ware, Figures 7.2-7.4). Until chronologies are<br />
more clearly defined it will be difficult to distinguish between homologies and analogies<br />
in respect <strong>of</strong> these simple decorative techniques, and Bedford’s reassessment <strong>of</strong> the incised<br />
and applied relief “tradition” is apposite here too, providing a caution against jumping to<br />
hastily to conclusions and making unwarranted connections.<br />
If there is any connection, and/or if decorative technique and pattern is separating<br />
out by site at Roviana, where it is mixed in with other pottery to the northwest or Vanuatu,<br />
either the Roviana surface scatters are amplifying chrono-stylistic structure in the data and<br />
thus temporal resolution, where this is more temporally mixed in excavated samples to the<br />
northwest, or the trajectories <strong>of</strong> ceramic change are so different between Roviana and the<br />
Buka/New Britain area that Roviana pottery may have changed beyond recognition as<br />
Lapita by 800BC, while Lapita-style is still supposedly thriving in the Bismark<br />
Archipelago.<br />
Whether Gharanga/Kopo pottery is thus a valuable index fossil (O'Brien & Lyman<br />
2000), and similar sites further north are yet to be discovered, or whether this is just<br />
502
analogous similarity (unlikely in view <strong>of</strong> the rarity <strong>of</strong> occurrence further north) cannot be<br />
resolved until temporal variability is better understood at the local level in each <strong>of</strong> these<br />
areas. <strong>The</strong> preferred explanation is that we have yet to fully sample and define temporal<br />
variability fully for the Lapita period and what came after, in New Britain, at Buka, and at<br />
Roviana Lagoon, and that the current samples from throughout the region, including<br />
Roviana and Buka and the Bismarck Archipelago, are insufficiently historically complete<br />
and site-structured to allow the sort <strong>of</strong> high-resolution chronology that can be used to<br />
discriminate between homologous and analogous similarity <strong>of</strong> simpler designs between<br />
regions.<br />
Late prehistoric adzes from Gharanga and Paniavile tie into Specht and Wickler’s<br />
terrestrial chronology, and suggest intertidal settlement may have continued until circa<br />
1500 AD, but this is speculative as the adzes are reported to have come from the intertidal,<br />
but are not securely provenanced. Even if these were found in the intertidal scatters, this<br />
does not mean they are the same age as the bulk <strong>of</strong> pottery in those sites, as surface<br />
collections, to a greater extent than some buried sites can easily incorporate materials from<br />
various periods.<br />
<strong>The</strong> Lapita Gap as an Area <strong>of</strong> Low Probability <strong>of</strong> Detection :<br />
<strong>The</strong> Roviana study shows that Lapita is most likely continuously distributed across near<br />
Oceania, because the gap in the recorded distribution can now be seen to reflect<br />
preservation and visibility rather than past behaviour. Now that we know the poor state<br />
<strong>of</strong> preservation <strong>of</strong> the emergent Roviana stilt villages, and the likely depth <strong>of</strong> water in<br />
which they were constructed, the absence <strong>of</strong> evidence from Choiseul, Ysabel and Ngela<br />
is explicable by the geological evidence for subsidence in those areas (Bruns et al. 1986).<br />
If the main chain adjacent to the New Georgia group is subsiding while Roviana is<br />
uplifting, sites at the sorts <strong>of</strong> locations evident in the Roviana data will be either grown<br />
503
over with coral or buried in s<strong>of</strong>t reefal detrital sediments, and thus be both well preserved<br />
and invisible, like the Mulifunua Ferry Berth site in Western Samoa, discovered by<br />
fortuitous dredging. In areas with more substantial uplift than Roviana, such as the forearc<br />
chain <strong>of</strong> Rendova, Tetepare and Ranonnga, preservation is likely to be worse than in the<br />
Roviana Lagoon, in some areas through wave exposure, but also because uplift has<br />
brought the pottery through the swash zone where it cannot survive wave exposure. This<br />
can be expected to have substantially reduced probability <strong>of</strong> detection. Evidence suggesting<br />
a Lapita gap in the Near-Oceanic Solomon Islands as avoidance or leap-frog colonization<br />
is thus substantially eliminated.<br />
This thesis can be tested by looking in places where high probability <strong>of</strong> past<br />
settlement, archaeological visibility, preservation and survey methods combine to promote<br />
a high probability <strong>of</strong> site detection. <strong>The</strong> scale <strong>of</strong> survey required is suggested in Chapter<br />
3 to be <strong>of</strong> the order <strong>of</strong> 900km 2 <strong>of</strong> lagoon, or 40 reef passages, or about 150km <strong>of</strong> coastline<br />
with high detectability, to detect a sample <strong>of</strong> ten or more sites with good samples <strong>of</strong><br />
Lapita-era and a larger number <strong>of</strong> derivatives <strong>of</strong> Lapita. With this sort <strong>of</strong> information in<br />
hand, the behavioural questions that require us to be able to tell time with high temporal<br />
resolution (Spriggs 2001) including consideration <strong>of</strong> spatial factors such as site density and<br />
mobility models (Anderson 2002) can be matched to appropriate data.<br />
Intertidal-Zone and Shallow-Water Archaeology:<br />
This thesis has been as much about figuring out how to proceed in researching an intertidal<br />
and shallow water archaeological distribution as about the distribution <strong>of</strong> Lapita pottery.<br />
<strong>The</strong> key <strong>chapter</strong>s in this regard are those on survey (Chapter 3) and formation processes<br />
(Chapters 5, 6, 7 and 11). <strong>The</strong>se could all have been vastly improved had the<br />
environmental variables used (wave-exposure; sediment characteristics and vegetation<br />
cover; relative sea level and tidal fluctuations; and collection site micro-topography) and<br />
504
the locations <strong>of</strong> artefacts and manuports been recorded more systematically and in more<br />
detail. This would have necessitated either a lot more survey time or much better<br />
equipment (probably both). In retrospect, if accurate electronic survey equipment was<br />
being used to capture these environmental details, point-provenancing <strong>of</strong> all sherds and<br />
lithics would have been a simple matter too.<br />
<strong>The</strong> wet tropical environment <strong>of</strong> the Roviana study is hard on such gear, which<br />
must be carefully maintained to function well, may fail unexpectedly, and adds to the<br />
expense and logistics <strong>of</strong> fieldwork Many <strong>of</strong> the suggestions for practice might seem trivial<br />
to those working in urban environments or in the rural hinterland <strong>of</strong> industrialised societies,<br />
where power supplies are usually available either from the main grid or from vehicle<br />
batteries, and where electronic equipment is available either through <strong>University</strong><br />
departments or Resource Management Agencies or hire companies, and is <strong>of</strong>ten used as<br />
a matter <strong>of</strong> course. Some <strong>of</strong> these suggestions might be very difficult to implement,<br />
however, by resource management agencies in non-industrialised nations or regions, like<br />
Solomon Islands, particularly where funding is meagre. <strong>The</strong> core information, though, that<br />
cannot be reconstructed after the event <strong>of</strong> picking up a sample, is where the piece came<br />
from, which is brought home to me by analysis <strong>of</strong> data which reflects my own poor efforts<br />
in this regard.<br />
A low-tech solutions to such recording can include detailed triangulation mapping<br />
<strong>of</strong> sample locations using measuring tapes from a permanently marked baseline, and is thus<br />
possible with simple equipment, although time consuming. Desalination <strong>of</strong> porous materials<br />
can be done in rainwater, as long as some method <strong>of</strong> labelling artefacts to retain their<br />
provenance information can be devised. Even cheaper, if this cannot be done, is to<br />
encourage protection <strong>of</strong> the site at the local level, where possible, and to leave the items<br />
where they are unless there is an immediate threat to their survival. Non-destructive in-situ<br />
recording <strong>of</strong> form and decoration for research purposes is always an option if there is<br />
sufficient skilled labour available. In some circumstances marine growth on sherds will<br />
505
obscure the details <strong>of</strong> decoration, in which case laboratory cleaning might be required.<br />
Broad fabric classifications could also be done in the field in many cases using a 10x hand<br />
lens, if a small corner <strong>of</strong> each sherd is snapped, with a type-series retained for petrographic<br />
analysis. <strong>The</strong>se suggestions for practice may not all turn out to be entirely practical, but<br />
are the range <strong>of</strong> possibilities I would experiment with in any future fieldwork <strong>of</strong> this sort.<br />
<strong>The</strong> focus on formation processes, which involved treating sherds and lithics as<br />
sedimentary particles, required descriptive systematics adapted to that purpose. Sherd size,<br />
in particular, was recorded in more detail than is the norm in Pacific archaeology, as was<br />
lithic manuport size and form. Point provenancing <strong>of</strong> artefacts, and recording <strong>of</strong> orientation<br />
could be useful developments <strong>of</strong> this approach.<br />
While the search for undisturbed buried villages continues in Lapita archaeology,<br />
there has been little explicit consideration <strong>of</strong> the problem <strong>of</strong> what to do with these when<br />
they are found (exceptions to this trend are noted in the introduction to Chapter 11).<br />
Relatively well-preserved stilt-house remains have been found in the Arawes and Mussau,<br />
and terrestrial buried landscapes have been reported from Manus and Vanuatu, but none<br />
<strong>of</strong> these have been extensively excavated, because to do so would take an enormous<br />
amount <strong>of</strong> time, money and skilled labour. Settlement pattern studies in the Reef-Santa<br />
Cruz islands and Fiji using both the surface and excavation record have been rare<br />
exceptions rather than the rule, enabled in part by shallow burial. Investigation <strong>of</strong> deeply-<br />
buried places has tended to become a search for layercake chronology, largely due to the<br />
difficulties <strong>of</strong> excavating anything other than a small fraction <strong>of</strong> the areas in a<br />
reconnaissance mode. <strong>The</strong>re is nothing intrinsically wrong with this, provided stringent<br />
attention is paid to identification <strong>of</strong> formation processes, and age-depth relationships are<br />
carefully constructed rather than assumed to be readable as time. <strong>The</strong>se relatively small<br />
holes will seldom be easily demonstrated to contain representative samples <strong>of</strong> variability<br />
for any <strong>of</strong> the layers, due to limited spatial sampling and the difficulty <strong>of</strong> quantifying<br />
506
samples, an area in which this thesis has sought to make some progress.<br />
Well preserved ceramic sites yield a lot <strong>of</strong> sherdage, which can foster complacence<br />
that the sample obtained from a small excavation is representative <strong>of</strong> the settlement as a<br />
whole, both across space and through time. <strong>The</strong> analysis in Chapter 5 provides an example<br />
showing how distantly related physical quantity <strong>of</strong> ceramics and sample size can be. It is<br />
this aspect <strong>of</strong> a tendency to search for layer upon layer <strong>of</strong> buried villages that is the focus<br />
<strong>of</strong> much <strong>of</strong> the critique in Chapter 2, and the argument is put forward in various places in<br />
this thesis that the temporal resolution <strong>of</strong> a seriation chronology constructed from a large<br />
number <strong>of</strong> large samples will almost inevitably surpass that obtained from reconnaissance<br />
excavation squares.<br />
In making external comparisons using the Roviana data above, the suggestion is<br />
put forward that some <strong>of</strong> the difficulty in matching the Roviana series to that from other<br />
areas may be to do with differences in temporal resolution, where horizontal structure is<br />
separating out styles that are mixed in other buried contexts. An argument for large area<br />
excavations and surface collection has been championed by Green for many years, and in<br />
the lithics <strong>chapter</strong> (Chapter 10) it is clear that the stone adze sample from Lapita-era sites<br />
in Near Oceania is dominated by materials from surface collections (this probably holds<br />
true for remote Oceania also, if large-area excavations are counted as equivalent in the<br />
scope <strong>of</strong> sampling to surface collections).<br />
I put it to research archaeologists interested in Lapita that this is probably true also<br />
<strong>of</strong> pottery, it is just that we have been lax about quantifying sample size, as few have been<br />
willing to battle with the problem <strong>of</strong> establishing vessel brokenness and completeness, and<br />
to develop methods <strong>of</strong> assessing sample size in various sampling contexts using these<br />
properties <strong>of</strong> samples. For these reasons we should value the surface record, including lag<br />
deposits in the sea, especially in Near Oceania (where settlement in such locations in the<br />
past is now well documented), as a source <strong>of</strong> large, easily acquired samples <strong>of</strong> ceramics<br />
and lithics, that can potentially form the backbone <strong>of</strong> high-resolution chronologies on<br />
507
which so much behavioural inference depends.<br />
Landscapes which <strong>of</strong>fer surface exposure <strong>of</strong> samples <strong>of</strong> Lapita-age materials<br />
sufficiently large to allow exploration <strong>of</strong> ceramic variability, <strong>of</strong> which the New Georgia<br />
region is one, should not be too hastily bypassed in the noble quest for Wobst’s mythical<br />
single point where all the region’s culture-historical complexes can be found superposed,<br />
separated by sterile deposits.<br />
<strong>The</strong> End<br />
508
REFERENCES<br />
Allen, J., 1984. Pots and Poor Princes: A Multidimensional Approach to the Role <strong>of</strong><br />
Pottery Trading in Coastal Papua. In S. E. Van Der Leeuw and A. C. Pritchard<br />
(eds) <strong>The</strong> Many Dimensions <strong>of</strong> Pottery: Ceramics in Archaeology and<br />
Anthropology. Amsterdam: Universiteit van Amsterdam,<br />
Allen, J., Gosden C., 1991. Report <strong>of</strong> the Lapita Homeland Project. Canberra:<br />
Department <strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, ANU.<br />
Allen, J., Gosden C., White J. P., 1989. Human Pleistocene Adaptations in the Tropical<br />
Island Pacific: Recent Evidence from New Ireland, a Greater Australian Outlier.<br />
Antiquity 63: 548-61<br />
Allen, J., Specht J., Ambrose W. R., Yen D. E., 1984. Lapita Homeland Project: Report<br />
<strong>of</strong> the 1984 Field Season Department <strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific<br />
Studies, ANU, Canberra<br />
Ambrose, W., 1991. Lapita or Not Lapita: <strong>The</strong> Case <strong>of</strong> the Manus Pots. In J. Allen and C.<br />
Gosden (eds) Report <strong>of</strong> the Lapita Homeland Project. Canberra, Australia:<br />
Department <strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian<br />
National <strong>University</strong>, pp.103-12.<br />
Ambrose, W., Gosden C., 1991. Investigations on Boduna Island. In J. Allen and C.<br />
Gosden (eds) Report <strong>of</strong> the Lapita Homeland Project. Canberra, Australia:<br />
Department <strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian<br />
National <strong>University</strong>, pp.182-8.<br />
Anderson, A., 2002. Mobility Models <strong>of</strong> Lapita Migration. In G. R. Clark, A. J. Anderson<br />
and T. Vunidilo (eds) <strong>The</strong> Archaeology <strong>of</strong> the Lapita Dispersal in Oceania:<br />
Papers from the 4th Lapita Conference, June 2000, Canberra, Australia.<br />
Canberra: Pandanus Press, Research School <strong>of</strong> Pacific and Asian studies, ANU,<br />
pp.15-23.<br />
509
Anson, D., 1983. Lapita Pottery <strong>of</strong> the Bismarck Archipelago and Its Affinities. PhD<br />
thesis. <strong>University</strong> <strong>of</strong> Sydney.<br />
---, 1986. Lapita Pottery <strong>of</strong> the Bismarck Archipelago and Its Affinities. Archaeology in<br />
Oceania 21: 157-63<br />
---, 1987. Reply to Kirch Et Al. Archaeology in Oceania 22: 127-8<br />
---, 1990. Aspiring to Paradise. In M. Spriggs (ed) Lapita Design, Form and<br />
Composition: Proceedings <strong>of</strong> the Lapita Design Workshop, Canberra, Australia,<br />
December 1988. Canberra Australia: Department <strong>of</strong> Prehistory, Research School<br />
<strong>of</strong> Pacific Studies, Australian National <strong>University</strong>, pp.53-8.<br />
---, 2000. Reber-Rakival Dentate-Stamped Motifs: Documentation and Comparative<br />
Implications. New Zealand Journal <strong>of</strong> Archaeology 20(1998): 119-35<br />
Anthes, R. A., 1982. Tropical Cyclones, <strong>The</strong>ir Evolution, Structure and Effects.<br />
Meteorological Monographs No. 41. American Meteorological Society.<br />
Arnold, D. E., 1985. Ceramic <strong>The</strong>ory and Cultural Process. Cambridge <strong>University</strong> Press.<br />
Arqueologicas, P. N. D. P., 1970. Brazilian Archaeology in 1968: And Interim Report on<br />
the National Program <strong>of</strong> Archaeological Research. American Antiquity 35 (1): 1-<br />
23<br />
Aswani, S., 1997. Customary Marine Tenure and Artisanal Fishing in the Roviana and<br />
Vonavona Lagoons, Solomon Islands: <strong>The</strong> Evolutionary Ecology <strong>of</strong> Marine<br />
Resource Utilization. PhD thesis. Anthropology, <strong>University</strong> <strong>of</strong> Hawaii.<br />
Balfet, H., 1965. Ethnographic Observations in North Africa and Archaeological<br />
Interpretation:<strong>The</strong> Pottery <strong>of</strong> the Maghreb. In F. R. Matson (ed) Ceramics and<br />
Man. New York: Viking,<br />
Bath, M., Deguara J., 2003. Southern Hemisphere Tropical Cyclones: Historical Tracks<br />
and Data. Electronic maps prepared using Australian Bureau <strong>of</strong> Meteorology<br />
data. http://australiasevereweather.com/cyclones/history.htm<br />
Baxter, M. J., 1994. Exploratory Multivariate Analysis in Archaeology. Edinburgh<br />
510
<strong>University</strong> Press.<br />
Bedford, S., 1999. Lapita and Post-Lapita Ceramic Sequences from Erromango, Southern<br />
Vanuatu. In J. C. Galipaud and I. Lilley (eds) <strong>The</strong> Pacific, 5000-2000 BP:<br />
Colonizations and Transformations. Paris: IRD Editions, pp.127-37.<br />
---, 2000. Pieces <strong>of</strong> the Vanuatu Puzzle: Archaeology <strong>of</strong> the North, South and Centre.<br />
PhD thesis. Department <strong>of</strong> Archaeology and Natural History, Research School <strong>of</strong><br />
Pacific and Asian Studies, Australian National <strong>University</strong>.<br />
---, 2002. Personal Communication: to M. Felgate. <strong>Auckland</strong><br />
Bellwood, P., 1978. Man's Conquest <strong>of</strong> the Pacific. Sydney: Collins.<br />
Best, S., 1984. Lakeba: <strong>The</strong> Prehistory <strong>of</strong> a Fijian Island. PhD thesis. Anthropology<br />
Department, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
---, 2002. Lapita: A View from the East. New Zealand Archaeological Association<br />
Monograph 24. <strong>Auckland</strong>.<br />
Bickler, S. H., 1998. Eating Stone and Dying: Archaeological Survey on Woodlark<br />
Island, Milne Bay Province, Papua New Guinea. PhD thesis. Department <strong>of</strong><br />
Anthropology, <strong>University</strong> <strong>of</strong> Virginia.<br />
Binford, L. R., 1962. Archaeology as Anthropology. American Antiquity 44: 455-70<br />
---, 1964. A Consideration <strong>of</strong> Archaeological Research Design. American Antiquity 29:<br />
425-41<br />
Blankholm, H. P., 1991. Intrasite Spatial Analysis in <strong>The</strong>ory and Practice. Aarhus:<br />
Aarhus <strong>University</strong> Press.<br />
Braun, D. P., 1983. Pots as Tools. In J. A. Moore and A. S. Keene (eds) Archaeological<br />
Hammers and <strong>The</strong>ories. New York, London: Academic Press, pp.107-36.<br />
Brenchley, P. J., Harper D. T., 1998. Paleoecology: Ecosystems, Environments and<br />
Evolution. Chapman and Hall.<br />
Bruns, T. R., Cooper A. K., Mann D., Vedder J. G., 1986. Seismic Stratigraphy and<br />
Structure <strong>of</strong> Sedimentary Basins in the Solomon Islands Region. In J. G. Vedder<br />
511
(ed) Geology and Offshore Resources <strong>of</strong> Pacific Island Arcs - Central and<br />
Western Solomon Islands. AAPG Circum-Pacific Council for Energy and Mineral<br />
Resources, 4, pp.177-214.<br />
Bunge, J., Fitzpatrick M., 1993. Estimating the Number <strong>of</strong> Species: A Review. Journal <strong>of</strong><br />
the American Statistical Association 88 (421): 364-73<br />
Burley, D. V., 1994. Settlement Pattern and Tongan Prehistory. Journal <strong>of</strong> the Polynesian<br />
Society 103 (4): 379-411<br />
Burley, D. V., Dickinson W. R., 2001. Origin and Significance <strong>of</strong> a Founding Settlement<br />
in Polynesia. Proceedings <strong>of</strong> the National Academy <strong>of</strong> Sciences 98 (20): 11829-31<br />
Chao, A., 1984. Nonparametric Estimation <strong>of</strong> the Number <strong>of</strong> Classes in a Population.<br />
Scandinavian Journal <strong>of</strong> Statistics 11: 265-70<br />
Chao, A., Lee S.-M., 1992. Estimating the Number <strong>of</strong> Classes Via Sample Coverage.<br />
Journal <strong>of</strong> the American Statistical Association 87 (417): 210-7<br />
Clark, G. R., 1999. Post-Lapita Fiji: Cultural Transformation in the Mid-Sequence. PhD<br />
thesis. Department <strong>of</strong> Archaeology and Natural History, Research School <strong>of</strong><br />
Pacific and Asian Studies, Australian National <strong>University</strong>.<br />
Clarke, D. L., 1968. Analytical Archaeology. London: Methuen.<br />
Colwell, R. K., 1997. Estimates: Statistical Estimation <strong>of</strong> Species Richness and Shared<br />
Species from Samples. Connecticut: http://viceroy.eeb.uconn.edu/estimates.<br />
Cowgill, G. L., 1972. Models, Methods and Techniques for Seriation. In D. L. Clarke (ed)<br />
Models in Archaeology. London: Methuen, pp.381-424.<br />
---, 1975. A Selection <strong>of</strong> Samplers: Comments on Archaeostatistics. In J. W. Mueller (ed)<br />
Sampling in Archaeology. Tucson: <strong>The</strong> <strong>University</strong> <strong>of</strong> Arizons Press, pp.258-74.<br />
Cowgill, G. L., Altshul J. H., Sload R. S., 1984. Spatial Analysis <strong>of</strong> Teotihuacan: A<br />
Mesoamerican Metropolis. In H. Hietala (ed) Intrasite Spatial Analysis in<br />
Archaeology. Cambridge <strong>University</strong> Press, pp.154-95.<br />
Davenport, W., 1972. Preliminary Excavations on Santa Ana Island, Eastern Solomon<br />
512
Islands. Archaeology and Physical Anthropology in Oceania 7 (3): 165-83<br />
David, B., Wilson M., 1999. Re-Reading the Landscape: Place and Identity in NE<br />
Australia During the Late Holocene. Cambridge Archaeological Journal 9 (2):<br />
163-88<br />
David, N., 1972. On the Life Span <strong>of</strong> Pottery, Type Frequencies, and Archaeological<br />
Inference. American Antiquity 37 (1): 141-2<br />
Davis, W. M., 1928. <strong>The</strong> Coral Reef Problem. American Geographic Society Special<br />
Publication 9.<br />
Day, T. J., 1980. A Study <strong>of</strong> the Transport <strong>of</strong> Graded Sediments. Wallingford, England :<br />
Hydraulics Research Station.<br />
de Barros, P., 1982. <strong>The</strong> Effects <strong>of</strong> Variable Site Occupation Span on the Results <strong>of</strong><br />
Frequency Seriation. American Antiquity 47 (2): 291-315<br />
De Boer, W. R., 1983. <strong>The</strong> Archaeological Assemblage as Preserved Death Assemblage.<br />
In J. A. Moore and A. S. Keene (eds) Archaeological Hammers and <strong>The</strong>ories.<br />
New York: Academic Press,<br />
Dickinson, W. R., 2000a. Petrography <strong>of</strong> Sand Tempers in Prehistoric Sherds from<br />
Roviana Lagoon, New Georgia, Solomon Islands. Tucson: Unpublished Report,<br />
<strong>University</strong> <strong>of</strong> Arizona, Geology Department, pp.10.<br />
---, 2000b. Petrography <strong>of</strong> Sand Tempers in Prehistoric Watom Sherds and Comparison<br />
with Other Temper Suites <strong>of</strong> the Bismarck Archipelago. New Zealand Journal <strong>of</strong><br />
Archaeology 20 (1998): 161-82<br />
Dickinson, W. R., Green R. C., 1998. Geoarchaeological Context <strong>of</strong> Holocene Subsidence<br />
at the Ferry Berth Lapita Site, Mulifanua, Upolu, Samoa. Geoarchaeology 13 (3):<br />
239-67<br />
Donovan, L. J., 1973. A Study <strong>of</strong> the Decorative System <strong>of</strong> the Lapita Potters in Reefs and<br />
Santa Cruz Islands. MA thesis. Department <strong>of</strong> Anthropology, <strong>University</strong> <strong>of</strong><br />
<strong>Auckland</strong>.<br />
513
Duff, A. I., 1996. Ceramic Micro-Seriation: Types or Attributes? American Antiquity 61<br />
(1): 89-101<br />
Dunkley, P. N., 1986. <strong>The</strong> Geology <strong>of</strong> the New Georgia Group, Western Solomon Islands<br />
British Geological Survey Overseas Directorate, Keyworth, Nottinghamshire<br />
Dunnell, R. C., 1970. Seriation Method and Its Evaluation. American Antiquity 35 (3):<br />
305-19<br />
---, 1971. Sabl<strong>of</strong>f and Smith's "the Importance <strong>of</strong> Both Analytic and Taxonomic<br />
Classification in the Type-Variety System". American Antiquity 36 (1): 115-8<br />
---, 1978. Style and Function: A Fundamental Dichotomy. American Antiquity 43: 192-<br />
202<br />
---, 1986. Methodological Issues in Americanist Artifact Classification. In M. B. Schiffer<br />
(ed) Advances in Archaeological Method and <strong>The</strong>ory. Academic Press, pp.149-<br />
207.<br />
---, 2001. Foreword. In T. D. Hurt and G. F. M. Rakita (eds) Style and Function:<br />
Conceptual Issues in Evolutionary Archaeology. Westport, Connecticutt: Bergin<br />
and Harvey, pp.xiii-xxiv.<br />
Egl<strong>of</strong>f, B. J., 1971. Collingwood Bay and the Trobriand Islands in Recent Prehistory.<br />
PhD thesis. Australian National <strong>University</strong>.<br />
---, 1973. A Method for Counting Ceramic Rim Sherds. American Antiquity 38 (3): 351-3<br />
---, 1979. Recent Prehistory in Southeast Papua. Terra Australis. 4, Canberra:<br />
Department <strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, ANU.<br />
Ericson, J. E., Read D. W., Burke C., 1972. Research Design: <strong>The</strong> Relationships between<br />
the Primary Functions and the Physical Properties <strong>of</strong> Ceramic Vessels and <strong>The</strong>ir<br />
Implications for Distributions on an Archaeological Site. Anthropology UCLA 3<br />
(2): 84-95<br />
Feathers, J. K., 1997. <strong>The</strong> Application <strong>of</strong> Luminescence Dating in American Archaeology.<br />
Journal <strong>of</strong> Archaeological Method and <strong>The</strong>ory 4 (1): 1-67<br />
514
---, 2002. Luminescence Dating <strong>of</strong> Pottery Sherds from the Solomon Islands. Unpublished<br />
Analysis Report, Seattle: <strong>University</strong> <strong>of</strong> Washington.<br />
Felgate, M. W., 2002. A Roviana Ceramic Sequence and the Prehistory <strong>of</strong> Near Oceania:<br />
Work in Progress. In G. R. Clark, A. J. Anderson and T. Vunidilo (eds) <strong>The</strong><br />
Archaeology <strong>of</strong> the Lapita Dispersal in Oceania: Papers from the 4th Lapita<br />
Conference, June 2000, Canberra, Australia. Pandanus Press, Research School<br />
<strong>of</strong> Pacific and Asian studies, ANU, pp.39-60.<br />
Felgate, M. W., Bickler S. H., n.d. Pots as Species: Estimating a Parent Population <strong>of</strong><br />
Vessels from a Sherd Sample. Unpublished Manuscript in the Posession <strong>of</strong> the<br />
Authors.<br />
Felgate, M. W., Dickinson W. R., 2001. Late-Lapita and Post-Lapita Pottery Transfers:<br />
Evidence from Intertidal-Zone Sites <strong>of</strong> Roviana Lagoon, Western Province,<br />
Solomon Islands. In M. Jones and P. J. Sheppard (eds) Proceedings <strong>of</strong> the 2001<br />
Austalasian Archaeolmentry Conference. <strong>Auckland</strong>: Research Papers in<br />
Anthropology and Linguistics, pp.105-22.<br />
Ferrari, B., Adams J., 1990. Biogenetic Modifications <strong>of</strong> Marine Sediments and <strong>The</strong>ir<br />
Influence on Archaeological Material. International Journal <strong>of</strong> Nautical<br />
Archaeology and Underwater Exploration 19: 139-51<br />
Flemming, N. C., 1983. Survival <strong>of</strong> Submerged Lithic and Bronze Age Artifact Sites: A<br />
Review. In P. M. Masters and N. C. Flemming (eds) Quaternary Coastlines and<br />
Marine Archaeology. Academic Press, pp.135-73.<br />
Fontana, V., 1998. Procedures to Analyse Intra-Site Pottery Distribution, Applied to the<br />
Neolithic Site <strong>of</strong> Fimon, Molino Casarotto (Italy), Housesite No.3. Journal <strong>of</strong><br />
Archaeological Science 25 (11): 1067-72<br />
Frost, E. L., 1974. Archaeological Excavations <strong>of</strong> Fortified Sites on Tavenui, Fiji. Asian<br />
and Pacific Archaeology Series. 6, Honolulu: Social Science Research Institute,<br />
<strong>University</strong> <strong>of</strong> Hawaii.<br />
515
Gaffney, V. L., Bintcliff J., Slapsak B., 1991. Site Formation Processes and the Hvar<br />
Survey Project, Yugoslavia. In A. J. Sch<strong>of</strong>ield (ed) Interpreting Atrefact Scatters:<br />
Contributions to Ploughzone Archaeology. Oxford: Oxbow Books,<br />
Galipaud, J.-C., 1996. New Caledonia: Some Recent Archaeological Perspectives. In J.<br />
M. Davidson, G. Irwin, B. F. Leach, A. Pawley and D. Brown (eds) Oceanic<br />
Culture History: Essays in Honour <strong>of</strong> Roger Green. New Zealand Journal <strong>of</strong><br />
Archaeology Special Publication, pp.297-305.<br />
---, 1998. <strong>The</strong> Lapita Site <strong>of</strong> Atanoasao: Malo, Vanuatu ORSTOM, Port Vila<br />
Gardin, J. C., 1967. Methods for the Descriptive Analysis <strong>of</strong> Archaeological Material.<br />
American Antiquity 32: 13-30<br />
Gell, A., 1992. <strong>The</strong> Technology <strong>of</strong> Enchantment and the Enchantment <strong>of</strong> Technology. In<br />
J. Coote and A. Shelton (eds) Anthropology, Art and Aesthetes. Oxford:<br />
Clarendon Press, pp.40-63.<br />
Gifford, E. W., Shutler R. J., 1956. Archaeological Excavations in New Caledonia.<br />
Anthropological Records. 18(1), Berkeley and Los Angeles: <strong>University</strong> <strong>of</strong><br />
California Press.<br />
Goldmann, K., 1971. Some Archaeological Criteria for Chronological Seriation. In F. R.<br />
Hodson, D. G. Kendall and P. Tautu (eds) Mathematics in the Archaeological and<br />
Information Sciences. Edinburgh: Edinburgh <strong>University</strong> Press,<br />
Golson, J., 1971. Lapita Ware and Its Transformations. In R. C. Green and M. Kelly (eds)<br />
Studies in Oceanic Culture History. Honolulu: Department <strong>of</strong> Anthropology,<br />
Bernice P. Bishop Museum, pp.67-76.<br />
---, 1991. Two Sites at Lasigi, New Ireland. In J. Allen and C. Gosden (eds) Report <strong>of</strong> the<br />
Lapita Homeland Project. Canberra, Australia: Department <strong>of</strong> Prehistory,<br />
Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian National <strong>University</strong>, pp.244-59.<br />
---, 1992. <strong>The</strong> Pottery from Lasigi, New Ireland. Presented at Poterie Lapita et<br />
516
Peuplement: Actes Du Colloque Lapita, Janvier 1992, Noumea, New Caledonia<br />
Gorecki, P., 1985. Test Excavation at Lemau, New Ireland Province and Report <strong>of</strong> an<br />
Excursion to Langovai, New Hanover, New Ireland Province. In J. Allen (ed)<br />
Lapita Homeland Project: Report <strong>of</strong> the 1985 Field Season. Melbourne: La Trobe<br />
<strong>University</strong>, pp.20-2.<br />
---, 1992. A Lapita Smoke Screen ? In J. C. Galipaud (ed) Poterie Lapita Et Peuplement:<br />
Actes Du Colloque Lapita, Janvier 1992. Noumea, New Caledonia:<br />
O.R.S.T.R.O.M., pp.27-47.<br />
Gosden, C., 1985. An Archaeological Survey <strong>of</strong> the Arawe Islands, West New Britain.<br />
Lapita Homeland Project: Report <strong>of</strong> the 1985 Field Season. Melbourne: La Trobe<br />
<strong>University</strong>, pp.54-5.<br />
---, 1989. Prehistoric Social Landscapes <strong>of</strong> the Arawe Islands, West New Britain Province,<br />
Papua New Guinea. Archaeology in Oceania 24 (2): 45-58<br />
---, 1991a. Learning About Lapita in the Bismarck Archipelago. In J. Allen and C. Gosden<br />
(eds) Report <strong>of</strong> the Lapita Homeland Project. Canberra, Australia: Department <strong>of</strong><br />
Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian National <strong>University</strong>,<br />
pp.260-8.<br />
---, 1991b. Towards an Understanding <strong>of</strong> the Regional Archaeological Record from the<br />
Arawe Islands, West New Britain, Papua New Guinea. In J. Allen and C. Gosden<br />
(eds) Report <strong>of</strong> the Lapita Homeland Project. Canberra, Australia: Department <strong>of</strong><br />
Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian National <strong>University</strong>,<br />
pp.205-16.<br />
Gosden, C., Allen J., Ambrose W. R., Anson D., Golson J., et al, 1989. Lapita Sites <strong>of</strong> the<br />
Bismarck Archipelago. Antiquity 63: 561-86<br />
Gosden, C., Head L., 1994. Landscape - a Usefully Ambiguous Concept. Archaeology in<br />
Oceania 29 (3): 113-6<br />
Gosden, C., Webb J., 1994. <strong>The</strong> Creation <strong>of</strong> a Papua-New Guinean Landscape:<br />
517
Archaeological and Geomorphological Evidence. Journal <strong>of</strong> Field Archaeology<br />
21: 29-51<br />
Gould, S. J., Vbra E., 1982. Exaptation- a Missing Term in the Science <strong>of</strong> Form.<br />
Paleobiology 8: 4-15<br />
Graves, M. W., 1998. <strong>The</strong> History <strong>of</strong> Method and <strong>The</strong>ory in the Study <strong>of</strong> Prehistoric<br />
Puebloan Pottery Style in the American Southwest. Journal <strong>of</strong> Archaeological<br />
Method and <strong>The</strong>ory 5 (4): 309-43<br />
Graves, M. W., Cachola-Abad C. K., 1996. Seriation as a Method <strong>of</strong> Chronologically<br />
Ordering Architectural Design Traits: An Example from Hawai'i. Archaeology in<br />
Oceania 31 (1): 19-32<br />
Green, R. C., 1974a. Excavation <strong>of</strong> the Prehistoric Occupations <strong>of</strong> Su-Sa-3. In R. C.<br />
Green and J. Davidson (eds) Archaeology in Western Samoa: Volume II.<br />
<strong>Auckland</strong>: <strong>Auckland</strong> Institute and Museum, pp.108-62.<br />
---, 1974b. Review <strong>of</strong> Portable Artefacts from Western Samoa. In R. C. Green and J.<br />
Davidson (eds) Archaeology in Western Samoa: Volume II. pp.245-74.<br />
---, 1976a. Lapita Sites in the Santa Cruz Group. In R. C. Green and M. M. Cresswell<br />
(eds) Southeast Solomons Culture History. Wellington, New Zealand: <strong>The</strong> Royal<br />
Society <strong>of</strong> New Zealand, Bulletin 11, pp.245-65.<br />
---, 1976b. A Late Prehistoric Sequence from Su' Ena Village, Uki. In R. C. Green and M.<br />
M. Cresswell (eds) Southeast Solomons Culture History. Wellington, New<br />
Zealand: <strong>The</strong> Royal Society <strong>of</strong> New Zealand, Bulletin 11, pp.181-91.<br />
---, 1976c. A Late-Prehistoric Settlement in Star Harbour. In R. C. Green and M. M.<br />
Cresswell (eds) Southeast Solomons Culture History. Wellington, New Zealand:<br />
<strong>The</strong> Royal Society <strong>of</strong> New Zealand, Bulletin 11, pp.133-47.<br />
---, 1978. New Sites with Lapita Pottery and <strong>The</strong>ir Implications for an Understanding <strong>of</strong><br />
the Settlement <strong>of</strong> the Western Pacific. Working Papers in Anthropology,<br />
Archaeology, Linguistics and Maori Studies. 51, <strong>Auckland</strong>, New Zealand:<br />
518
Department <strong>of</strong> Anthropology, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
---, 1979. Lapita. In J. D. Jennings (ed) <strong>The</strong> Prehistory <strong>of</strong> Polynesia. Cambridge,<br />
Massachusetts: Harvard <strong>University</strong> Press, pp.27-59.<br />
---, 1990. Lapita Design Analysis: <strong>The</strong> Mead System and Its Use; a Potted History.<br />
Presented at Lapita Design, Form and Composition: Proceedings <strong>of</strong> the Lapita<br />
Design Workshop, Canberra, Australia, December 1988, Canberra Australia<br />
---, 1991a. <strong>The</strong> Lapita Cultural Complex: Current Evidence and Proposed Models.<br />
Bulletin <strong>of</strong> the Indo-Pacific Prehistory Association 11: 295-305<br />
---, 1991b. Near and Remote Oceania -- Disestablishing "Melanesia" in Culture History.<br />
In A. Pawley (ed) Man and a Half: Essays in Pacific Anthropology and<br />
Ethnobiology in Honour <strong>of</strong> Ralph Bulmer. <strong>Auckland</strong>: <strong>The</strong> Polynesian Society,<br />
pp.491-502.<br />
---, 1991c. A Reappraisal <strong>of</strong> the Dating for Some Lapita Sites in the Reef/Santa Cruz<br />
Group <strong>of</strong> the Southeast Solomons. Journal <strong>of</strong> the Polynesian Society 100: 197-<br />
207<br />
---, 1992. Definitions <strong>of</strong> the Lapita Cultural Complex and Its Non-Ceramic Component.<br />
in Poterie Lapita et Peuplement: Actes Du Colloque Lapita, Janvier 1992, pp. 7-<br />
21. Noumea, New Caledonia: O.R.S.T.R.O.M.<br />
---, 1994. Prehistoric Transfers <strong>of</strong> Portable Items During the Lapita Horizon in Remote<br />
Oceania: A Review. Ippa 15th Congress: Chiang Mai -Jan. 5-12.<br />
Green, R. C., Anson D., 1991. <strong>The</strong> Reber-Rakival Site on Watom. Implications <strong>of</strong> the<br />
1985 Excavations at the SAC and SDI Localities. In J. Allen and C. Gosden (eds)<br />
Report <strong>of</strong> the Lapita Homeland Project. Canberra, Australia: Department <strong>of</strong><br />
Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian National <strong>University</strong>,<br />
pp.170-81.<br />
---, 2000a. Archaeological Investigations on Watom Island: Early Work, Outcomes <strong>of</strong><br />
Recent Investigations, and Future Prospects. New Zealand Journal <strong>of</strong> Archaeology<br />
519
20 (1998): 183-97<br />
---, 2000b. Excavations at Kainapirina (SAC), Watom Island, Papua New Guinea. New<br />
Zealand Journal <strong>of</strong> Archaeology 20 (1998): 29-94<br />
Green, R. C., Davidson J. M., 1969. Description and Classification <strong>of</strong> Samoan Adzes. In<br />
R. C. Green and J. M. Davidson (eds) Archaeology in Western Samoa. <strong>Auckland</strong>:<br />
<strong>Auckland</strong> Institute and Museum, pp.21-32.<br />
Hagstrum, M. B., Hildebrand J. A., 1990. <strong>The</strong> Two-Curvature Method for Reconstructing<br />
Ceramic Morphology. American Antiquity 55 (2): 388-403<br />
Hamilton, D., Woodward R., 1984. A Sunken 17th Century City: Port Royal, Jamaica.<br />
Archaeology 37: 38-45<br />
Hedrick, J. D., n.d. Archaeological Investigation <strong>of</strong> Malo Prehistory:Lapita Settlement<br />
Strategy in the Northern New Hebrides. Unfinished draft PhD thesis. <strong>University</strong><br />
<strong>of</strong> Pennsylvania.<br />
Henderson, A. G., 1998. Middle Fort Ancient Villages and Organizational Complexity<br />
in Central Kentucky. PhD thesis. university <strong>of</strong> kentucky.<br />
Henderson, H. M., 1992. <strong>The</strong> Strickler Site: <strong>The</strong> Analysis <strong>of</strong> a Nineteenth to Twentieth<br />
Century Urban Farmstead in Markham County, Ontario (Canada). MA thesis.<br />
<strong>University</strong> <strong>of</strong> Toronto.<br />
Hendren, G. H., 1976. Recent Settlement Pattern Changes on Ulawa, Southeast Solomon<br />
Islands. In R. C. Green and M. M. Cresswell (eds) Southeast Solomons Culture<br />
History. Wellington, New Zealand: <strong>The</strong> Royal Society <strong>of</strong> New Zealand, Bulletin<br />
11, pp.149-59.<br />
Henrickson, E. F., 1990. Investigating Ancient Ceramic Form and Use. In W. D. Kingery<br />
(ed) <strong>The</strong> Changing Roles <strong>of</strong> Ceramics in Society: 26,000 B.P. To the Present.<br />
Westerville, Ohio: <strong>The</strong> American Ceramic Society, Inc., pp.83-117.<br />
Henrickson, E. F., Mcdonald M. A., 1985. Ceramic Form and Function: An Ethnographic<br />
Search and Archaeological Application. American Anthropologist 85: 630-43<br />
520
Hodder, I., Orton C., 1976. Spatial Analysis in Archaeology. New Studies in<br />
Archaeology. Cambridge <strong>University</strong> Press.<br />
Hydrographic-Office, G. B., 1971. Pacific Islands Pilot Volume I: <strong>The</strong> Western Groups<br />
Comprising the Territory <strong>of</strong> Papua Including the Louisiade Archipelago, the<br />
North-East and North Coasts <strong>of</strong> New Guinea, the Solomon Islands, the Bismarck<br />
Archipelago, and the Caroline and Marianas Islands. Hydrographer <strong>of</strong> the Navy.<br />
Irwin, G., 1972. An Archaeological Survey in the Shortland Islands, B.S.I.P. MA thesis.<br />
Anthropology Department, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
---, 1985. <strong>The</strong> Emergence <strong>of</strong> Mailu: As a Central Place in Coastal Papuan Prehistory.<br />
Terra Australis. 10, Canberra: Department <strong>of</strong> Prehistory, Research School <strong>of</strong><br />
Pacific Studies, ANU.<br />
---, 1992. <strong>The</strong> Prehistoric Exploration and Colonization <strong>of</strong> the Pacific. Cambridge<br />
<strong>University</strong> Press.<br />
Ishimura, T., 2002. In the Wake <strong>of</strong> Lapita: Transformation <strong>of</strong> Lapita Designs and Gradual<br />
Dispersal <strong>of</strong> the Lapita Peoples. People and Culture in Oceania 18: 77-97<br />
Jensen, K., Jensen J., Clegg C., 1999. Inferring Intensity <strong>of</strong> Site Use from the Breakdown<br />
Rate and Discard Patterns <strong>of</strong> Fire-Cracked Rock at Playa View Dune. Utah<br />
Archaeology 12: 51-64<br />
Kintigh, K. W., 1984. Measuring Archaeological Diversity by Comparison with Simulated<br />
Assemblages. American Antiquity 49 (1): 44-54<br />
---, 1990. Intrasite Spatial Analysis: A Commentary on Major Methods. In A. Voorrips<br />
(ed) Mathematics and Information Science in Archaeology: A Flexible<br />
Framework. Bonn: Holos, pp.165-200.<br />
Kirch, P. V., 1985. Archaeological Fieldwork on Eloaua and Emananus Islands, St.<br />
Matthias Group, New Ireland Province. In J. Allen (ed) Lapita Homeland Project:<br />
Report <strong>of</strong> the 1985 Field Season. Melbourne: La Trobe <strong>University</strong>, pp.14-9.<br />
---, 1987a. Is <strong>The</strong>re an Early Far Western Lapita Province? Sample Size Effects and New<br />
521
Evidence from Eloaua Island. Archaeology in Oceania 22: 123-6<br />
---, 1987b. Lapita and Oceanic Cultural Origins: Excavations in the Mussau Islands,<br />
Bismarck Archipelago, 1985. Journal <strong>of</strong> Field Archaeology 14: 163-80<br />
---, 1988a. Niuatopupapu: <strong>The</strong> Prehistory <strong>of</strong> a Polynesian Chiefdom. Monograph 5,<br />
Seattle: Burke Museum, Seattle.<br />
---, 1988b. Problems and Issues in Lapita Archaeology. In P. V. Kirch and T. L. Hunt<br />
(eds) Archaeology <strong>of</strong> the Lapita Cultural Complex: A Critical Review. Seattle:<br />
Thomas Burke Memorial Washington State Museum Research Report No. 5,<br />
pp.167-5.<br />
---, 1988c. <strong>The</strong> Talepakemalai Lapita Site and Oceanic Prehistory. National Geographic<br />
Research 4 (3): 38-342<br />
---, 1997. <strong>The</strong> Lapita Peoples: Ancestors <strong>of</strong> the Oceanic World. Blackwell.<br />
---, 2001. Lapita and Its Transformations in near Oceania: Archaeological Investigations<br />
in the Mussau Islands, Papua New Guinea, 1985-88. Berkeley: <strong>University</strong> <strong>of</strong><br />
California.<br />
Kirch, P. V., Green R. C., 2001. Hawaiki, Ancestral Polynesia: An Essay in Historical<br />
Anthropology. Cambridge <strong>University</strong> Press.<br />
Kirch, P. V., Hunt T. L., 1988. <strong>The</strong> Spatial and Temporal Boundaries <strong>of</strong> Lapita. In P. V.<br />
Kirch and T. L. Hunt (eds) Archaeology <strong>of</strong> the Lapita Cultural Complex: A<br />
Critical Review. Seattle: Thomas Burke Memorial Washington State Museum<br />
Research Report No. 5, pp.9-31.<br />
Kirch, P. V., Hunt T. L., Weisler M., Butler V., Allen M. S., 1991. Mussau Islands<br />
Prehistory: Results <strong>of</strong> the 1985-1986 Excavations. In J. Allen and C. Gosden (eds)<br />
Report <strong>of</strong> the Lapita Homeland Project. Canberra, Australia: Department <strong>of</strong><br />
Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian National <strong>University</strong>,<br />
pp.144-63.<br />
Kirch, P. V., Rosendahl P., 1973. Archaeological Investigation <strong>of</strong> Anuta. In D. E. Yen and<br />
522
J. Gordon (eds) Anuta: A Polynesian Outlier in the Solomon Islands. Honolulu:<br />
Bishop Museum,<br />
Krantz, G. S., 1968. A New Method for Counting Mammal Bones. American Journal <strong>of</strong><br />
Archaeology 72: 286-8<br />
Lampert, R. J., 1966. Archaeological Reconnaissance in Papua and New Guinea.<br />
Canberra: Unpublished Mimeograph, Department <strong>of</strong> Prehistory, Research School<br />
<strong>of</strong> Pacific Studies, ANU.<br />
Le Blanc, S. A., 1975. Micro-Seriation: A Method for Chronologic Differentiation.<br />
American Antiquity 40 (1): 22-38<br />
Lep<strong>of</strong>sky, D., 1988. <strong>The</strong> Environmental Context <strong>of</strong> Lapita Settlement Locations. In P. V.<br />
Kirch and T. L. Hunt (eds) Archaeology <strong>of</strong> the Lapita Cultural Complex: A<br />
Critical Review. Seattle: Burke Museum, Seattle, pp.33-48.<br />
Lilley, I., 1985. Archaeological Fieldwork in Duke <strong>of</strong> York Islands, East New Britain<br />
Province. In J. Allen (ed) Lapita Homeland Project: Report <strong>of</strong> the 1985 Field<br />
Season. Melbourne: La Trobe <strong>University</strong>, pp.41-2.<br />
---, 1986. Prehistoric Exchange in the Vitiaz Strait, Papua New Guinea. PhD thesis.<br />
Australian National <strong>University</strong>.<br />
---, 1991. Lapita Sites in the Duke <strong>of</strong> York Islands. In J. Allen and C. Gosden (eds) Report<br />
<strong>of</strong> the Lapita Homeland Project. Canberra, Australia: Department <strong>of</strong> Prehistory,<br />
Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian National <strong>University</strong>, pp.164-9.<br />
Linton, R. E., 1925. Archaeology <strong>of</strong> the Marquesas Islands. Bernice P Bishop Museum<br />
Bulletin 23. Honolulu: Bayard Dominick Expedition Bulletin Number 10.<br />
---, 1944. North American Cooking Pots. American Antiquity IX (4): 369-80<br />
Longacre, W. A., 1991. Sources <strong>of</strong> Ceramic Variability among the Kalinga <strong>of</strong> Northern<br />
Luzon. In W. A. Longacre (ed) Ceramic Ethnoarchaeology. Tucson: <strong>University</strong><br />
<strong>of</strong> Arizona Press, pp.95-111.<br />
Lyman, R. L., Wolverton S., O'Brien M. J., 1998. Seriation, Superposition, and<br />
523
Interdigitation: A History <strong>of</strong> Americanist Graphic Depictions <strong>of</strong> Culture Change.<br />
American Antiquity 63 (2): 239-61<br />
Macmillan, D. H., 1966. Tides. New York: American Elsevier Publishing Company Inc.<br />
Mann, P., Taylor F. W., Lagoe M. B., Quarles A., Burr G., 1998. Accelerating Late<br />
Quaternary Uplift <strong>of</strong> the New Georgia Island Group (Solomon Island Arc) in<br />
Response to Subduction <strong>of</strong> the Recently Active Woodlark Spreading Centre and<br />
Coleman Seamount. Tectonophysics 295: 259-306<br />
Marquardt, W. H., 1978. Advances in Archaeological Seriation. Advances in<br />
Archaeological Method and <strong>The</strong>ory. Journal <strong>of</strong> Archaeological Method and<br />
<strong>The</strong>ory, pp.257-314.<br />
Mccoy, P. C., Cleghorn P. L., 1988. Archaeological Investigations on Santa Cruz<br />
(Nendo), Southeast Solomon Islands: Summary Report. Archaeology in Oceania<br />
23: 104-15<br />
McNutt, C. H., 1973. On the Methodological Validity <strong>of</strong> Frequency Seriation. American<br />
Antiquity 38 (1): 45-60<br />
Mead, S. M., 1975. <strong>The</strong> Decorative System <strong>of</strong> the Lapita Potters <strong>of</strong> Sigatoka, Fiji. In S.<br />
M. Mead, H. Birks, L. Birks and E. Shaw (eds) <strong>The</strong> Lapita Pottey Style <strong>of</strong> Fiji and<br />
Its Associations. Wellington: <strong>The</strong> Polynesian Society, pp.19-43.<br />
Mead, S. M., Birks L., Birks H., Shaw E., 1975. <strong>The</strong> Lapita Pottery Style <strong>of</strong> Fiji and Its<br />
Associations. Wellington: <strong>The</strong> Polynesian Society.<br />
Miller, D., 1985. Artefacts as Categories: A Study <strong>of</strong> Ceramic Variability in Central<br />
India. New Studies in Archaeology. London: Cambridge <strong>University</strong> Press.<br />
Miller, D., Roe D., 1982. <strong>The</strong> Solomon Islands National Sites Survey: <strong>The</strong> First Phase.<br />
Bulletin <strong>of</strong> the Indo-Pacific Prehistory Association 3: 47-51<br />
Mills, B. J., 1989. Integrating Functional Analyses <strong>of</strong> Vessels and Sherds through Models<br />
<strong>of</strong> Ceramic Assemblage Formation. World Archaeology 21 (1): 133-47<br />
Nojima, Y., 2002. Personal Communication: to M. Felgate.<br />
524
O'Brien, M. J., Lyman R. L., 2000. Seriation, Stratigraphy and Index Fossils: <strong>The</strong><br />
Backbone <strong>of</strong> Archaeological Dating. New York: Kluwer Academic/Plenum.<br />
Open <strong>University</strong> Oceanography Team, 1989. Waves, Tides and Shallow-Water Processes.<br />
Oxford: Pergamon Press.<br />
Oreihaka, E., 1997. Freshwater and Marine Aquatic Resources in Solomon Islands.<br />
Orton, C., 1982. Computer Simulation Experiments to Assess the Performance <strong>of</strong><br />
Measures <strong>of</strong> Quantity <strong>of</strong> Pottery. World Archaeology 14 (1): 1-20<br />
---, 1993. How Many Pots Make Five?- an Historical Review <strong>of</strong> Pottery Quantification.<br />
Archaeometry 35 (2): 169-84<br />
---, 2000. Sampling in Archaeology. Cambridge Manuals in Archaeology. Cambridge<br />
<strong>University</strong> Press.<br />
Orton, C., Tyers P., Vince A., 1993. Pottery in Archaeology. Cambridge Manuals in<br />
Archaeology. Cambridge <strong>University</strong> Press.<br />
Orton, C., Tyers P. A., 1991. Counting Broken Objects: <strong>The</strong> Statistics <strong>of</strong> Ceramic<br />
Assemblages. Proceedings <strong>of</strong> the British Academy 77: 163-84<br />
Parker, V. N. M., 1981. Vessel Forms <strong>of</strong> the Reef Islands Se-Rf-2 Site and <strong>The</strong>ir<br />
Relationships to Vessel Forms in Other Western Lapita Sites <strong>of</strong> the Reef/Santa<br />
Cruz and Island Melanesian Area. MA thesis. Anthropology Department,<br />
<strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
Petchey, F. J., 1995. <strong>The</strong> Archaeology <strong>of</strong> Kudon: Archaeological Analysis <strong>of</strong> Lapita<br />
Ceramics from Mulifanua, Samoa, and Sigatoka, Fiji. MA thesis. Anthropology<br />
Department, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
Phelan, M., 1997. Scratching the Surface: <strong>The</strong> Lapita Pottery <strong>of</strong> Makegur, Papua New<br />
Guinea. MA thesis. School <strong>of</strong> archaeology, La Trobe <strong>University</strong>.<br />
Plog, S., 1985. Estimating Vessel Orifice Diameters: Methods and Measurement Error. In<br />
B. A. Nelson (ed) Decoding Prehistoric Ceramics. Carbondale and Edwardsville:<br />
Southern Illinois <strong>University</strong> Press, pp.243-53.<br />
525
Poulsen, J. I., Polach H. A., 1972. Outlier Archaeology: Bellona. A Preliminary Report on<br />
Field Work and Radiocarbon Dates. Archaeology and Physical Anthropology in<br />
Oceania 7: 184-214<br />
Prior, C., 2000. Personal Communication: Rafter Radiocarbon Laboratory, Institute <strong>of</strong><br />
Geological and Nuclear Sciences, Lower Hutt, New Zealand, to M. Felgate.<br />
Rathje, R., 2000. Lapita Development as the First Holiday Inns. In A. Anderson and T.<br />
Murray (eds) Australian Archaeologist: Collected Papers in Honour <strong>of</strong> Jim Allen.<br />
Canberra: Coombs Academic Publishing. pp 62-68.<br />
Reeve, R., 1989. Recent Work on the Prehistory <strong>of</strong> the Western Solomons, Melanesia.<br />
Bulletin <strong>of</strong> the Indo-Pacific Prehistory Association 9: 44-67<br />
Rice, P., 1982. Pottery Production, Pottery Classification, and the Role <strong>of</strong><br />
Physicochemical Analyses. In J. S. Olin and A. D. Franklin (eds) Archaeological<br />
Ceramics. Washington DC: Smithsonian Institute Press, pp.47-56.<br />
---, 1987. Pottery Analysis: A Sourcebook. <strong>University</strong> <strong>of</strong> Chicago Press.<br />
Robertson, I. G., 2001. Mapping the Social Landscape <strong>of</strong> an Early Urban Centre. PhD<br />
thesis. Arizona State <strong>University</strong>.<br />
Roe, D., 1992. Investigations into the Prehistory <strong>of</strong> the Central Solomons: Some Old and<br />
Some New Data from Northwest Guadalcanal. In J.-C. Galipaud (ed) Poterie<br />
Lapita Et Peuplement: Actes Du Colloque Lapita, Janvier 1992. Noumea, New<br />
Caledonia: O.R.S.T.R.O.M., pp.91-101.<br />
---, 1993. Prehistory without Pots: Prehistoric Settlement and Economy <strong>of</strong> Northwest<br />
Guadalcanal, Solomon Islands. PhD thesis. Australian National <strong>University</strong>.<br />
Rogers, G., 1973. Report on Archaeological Survey in Niuatoputapu Island, Tonga.<br />
Working Papers in Anthropology, Archaeology, Linguistics and Maori Studies No.<br />
28. Department <strong>of</strong> Anthropology, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
Ross, M. D., 1996. Pottery Terms in Proto Oceanic. In J. M. Davidson, G. Irwin, B. F.<br />
Leach, A. Pawley and D. Brown (eds) Oceanic Culture History: Essays in Honour<br />
526
<strong>of</strong> Roger Green. New Zealand Journal <strong>of</strong> Archaeology Special Publication, pp.67-<br />
82.<br />
Rouse, I. B., 1960. <strong>The</strong> Classification <strong>of</strong> Artifacts in Archaeology. American Antiquity 25:<br />
313-23<br />
Rowland, M. J., 1989. Population Increase:Intensification or a Result <strong>of</strong> Preservation?<br />
Explaining Site Distribution Patterns on the Coast <strong>of</strong> Queensland. Australian<br />
Aboriginal Studies 2: 32-42<br />
Ruig, J., 1999. Collectors as Taphonomic Agents in the Archaeological Record. In M.<br />
Mountain and D. Bowdery (eds) Taphonomy: <strong>The</strong> Analysis <strong>of</strong> Processes from<br />
Phytoliths to Megafauna. Canberra: ANH publications: Australian National<br />
<strong>University</strong>, pp.109-15.<br />
Rye, O. S., 1981. Pottery Technology: Principles and Reconstruction. Manual on<br />
Archaeology 4. Washington: Taraxacum.<br />
Sand, C., 1997a. <strong>The</strong> Chronology <strong>of</strong> Lapita Ware in New Caledonia. Antiquity 71: 539-47<br />
---, 1997b. Variété De L'habitat Ancien En Nouvelle-Calédonie: Étude De Cas Sur Des<br />
Vestiges Archéologiques Du Centre-Nord De La Grande Terre. Journal, Société<br />
des Océanistes - Paris 104: 39-66<br />
---, 1998. Archaeological Report on Localities WKO013a and WKO013b at the Site <strong>of</strong><br />
Lapita(Kone, New Caledonia). Journal <strong>of</strong> the Polynesian Society 107: 7-33<br />
---, 1999. Lapita and Non-Lapita Ware During New Caledonia's First Millenium <strong>of</strong><br />
Austronesian Settlement. In J.-C. Galipaud and I. Lilley (eds) <strong>The</strong> Pacific from<br />
5000 to 2000 BP: Colonisation and Transformations. Patis: Editions de IRD,<br />
pp.139-59.<br />
---, 2000. <strong>The</strong> Specificities <strong>of</strong> the "Southern Lapita Province", the New Caledonian Case.<br />
Archaeology in Oceania 35: 20-33<br />
---, 2001. Evolutions in the Lapita Cultural Complex: A View from the Southern Lapita<br />
Province. Archaeology in Oceania 36: 65-76<br />
527
Schiffer, M. B., 1987. Formation Processes <strong>of</strong> the Archaelogical Record. Albuquerque:<br />
<strong>University</strong> <strong>of</strong> New Mexico Press.<br />
---, 1995a. Archaeological Context and Systemic Context. Behavioural Archaeology:<br />
First Principles. Salt Lake City: <strong>University</strong> Of Utah Press, pp.25-34.<br />
---, 1995b. Archaeology as Behavioural Science. Behavioural Archaeology: First<br />
Principles. Salt Lake City: <strong>University</strong> Of Utah Press, pp.46-54.<br />
---, 1995c. Toward the Identification <strong>of</strong> Formation Processes. In M. B. Schiffer (ed)<br />
Behavioural Archaeology: First Principles. Salt Lake City: <strong>University</strong> Of Utah<br />
Press, pp.171-95.<br />
Schiffer, M. B., Sullivan A. P., Klinger T. C., 1978. <strong>The</strong> Design <strong>of</strong> Archaeological<br />
Surveys. World Archaeology 10 (1): 1-28<br />
Seber, G. A. F., 1986. A Review <strong>of</strong> Estimating Animal Abundance. Biometrics 42 (2):<br />
267-92<br />
Sergeantson, S. W., Gao X., 1995. Homo Sapiens Is an Evolving Species: Origins <strong>of</strong> the<br />
Austronesians. In P. Bellwood, J. J. Fox and D. Tryon (eds) <strong>The</strong> Austronesians:<br />
Historical and Comparative Perspectives. Canberra: Department <strong>of</strong> Anthropology,<br />
Research School <strong>of</strong> Pacific and Asian Studies, Australian National <strong>University</strong>,<br />
Shackley, M. L., 1974. Stream Abrasion <strong>of</strong> Flint Implements. Nature 248: 501-2<br />
---, 1978. <strong>The</strong> Behaviour <strong>of</strong> Artefacts as Sedimentary Particles in a Fluvatile Environment.<br />
Archaeometry 20: 55-61<br />
Sharp, N. D., 1988. Style and Substance: A Reconsideration <strong>of</strong> the Lapita Decorative<br />
System. In P. V. Kirch and T. L. Hunt (eds) Archaeology <strong>of</strong> the Lapita Cultural<br />
Complex: A Critical Review. Seattle: Burke Museum, pp.61-81.<br />
---, 1991. Lapita as Text: <strong>The</strong> Meaning <strong>of</strong> Pottery in Melanesian Prehistory. Bulletin <strong>of</strong> the<br />
Indo-Pacific Prehistory Association 11: 323-32<br />
Shennan, S., 1997. Quantifying Archaeology. Edinburgh: Edinburgh <strong>University</strong> Press.<br />
Shepard, A. O., 1963. Ceramics for the Archaeologist. Washington DC: Carnegie<br />
528
Institution <strong>of</strong> Washington.<br />
Sheppard, P. J., 1993. Lapita Lithics: Trade/Exchang and Technology. A View from Thr<br />
Reefs/Santa Cruz. Archaeology in Oceania 28: 121-37<br />
---, 1996. Annual Report <strong>of</strong> the New Georgia Archaeological Survey 1996, Roviana<br />
Lagoon Survey Year 1 Centre for Archaeological Research, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong><br />
---, 2003. Personal Communication: to M. Felgate.<br />
Sheppard, P. J., Felgate M., Roga K., Keopo J., Walter R., 1999. A Ceramic Sequence<br />
from Roviana Lagoon (New Georgia, Solomon Islands). In J.-C. Galipaud and I.<br />
Lilley (eds) <strong>The</strong> Pacific from 5000 to 2000 BP: Colonization and<br />
Transformations. Paris: Editions de IRD, pp.313-22.<br />
Sheppard, P. J., Green R. C., 1991. Spatial Analysis <strong>of</strong> the Nenumbo (Se-Rf-2) Lapita<br />
Site, Solomon Islands. Archaeology in Oceania 26: 89-101<br />
Shott, M. J., 1989a. Diversity, Organization, and Behaviour in the Material Record.<br />
Current Anthropology 30 (3): 283-315<br />
---, 1989b. On Tool-Class Use Lives and the Formation <strong>of</strong> Archaeological Assemblages.<br />
American Antiquity 54 (1): 9-30<br />
---, 1996. Mortal Pots: On Use Life and Vessel Size in the Formation <strong>of</strong> Archaeological<br />
Assemblages. American Antiquity 61 (3): 463-82<br />
---, 1998. Status and Role <strong>of</strong> Formation <strong>The</strong>ory in Contemporary Archaeological Practice.<br />
Journal <strong>of</strong> Archaeological Research 6 (4): 299-329<br />
Shott, M. J., Sillitoe P., 2001. <strong>The</strong> Mortality <strong>of</strong> Things: Correlates <strong>of</strong> Use Life in Wola<br />
Material Culture Using Age-at-Census Data. Journal <strong>of</strong> Archaeological Method<br />
and <strong>The</strong>ory 8 (3): 269-302<br />
Shutler, R., Shutler M. E., 1964. Potsherds from Bougainville Island. Asian Perspectives<br />
8: 181-3<br />
Sinopoli, C. M., 1991. Approaches to Archaeological Ceramics. New York: Plenum<br />
Press.<br />
529
Siorat, J. P., 1990. A Technological Analysis <strong>of</strong> Lapita Pottery Decoration. in Lapita<br />
Design, Form and Composition: Proceedings <strong>of</strong> the Lapita Design Workshop,<br />
Canberra, Australia, December 1988, pp. 59-82. Canberra Australia: Department<br />
<strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, Australian National <strong>University</strong><br />
Skibo, J., Schiffer M. B., 1987. <strong>The</strong> Effects <strong>of</strong> Water on Processes <strong>of</strong> Ceramic Abrasion.<br />
Journal <strong>of</strong> Archaeological Science 14: 83-96<br />
Skibo, J. N., 1992. Pottery Function: A Use Alteration Perspective. Interdisciplinary<br />
Contributions to Archaeology. New York: Plenum Press.<br />
Smith, M. F., Jr., 1983. <strong>The</strong> Study <strong>of</strong> Ceramic Function from Artefact Size and Shape.<br />
PhD thesis. <strong>University</strong> <strong>of</strong> Oregon.<br />
---, 1985. Toward an Economic Interpretation <strong>of</strong> Ceramics: Relating Vessel Size and<br />
Shape to Use. In B. A. Nelson (ed) Decoding Prehistoric Ceramics. Carbondale<br />
and Edwardsville: Southern Illinois <strong>University</strong> Press, pp.254-309.<br />
Snow, B. E., Shutler R., 1985. <strong>The</strong> Archaeology <strong>of</strong> Fuga Moro Island. Cebu City: San<br />
Carlos Publications.<br />
Sokal, R. R., Sneath P. H. A., 1963. Principles <strong>of</strong> Numerical Taxonomy. San Francisco:<br />
W. H. Freeman and Co.<br />
Solow, 1994. On the Bayesian Estimation <strong>of</strong> the Number <strong>of</strong> Species in a Community.<br />
Ecology 75 (7): 2139-42<br />
Spaulding, A. C., 1953. Statistical Techniques for the Discovery <strong>of</strong> Artefact Types.<br />
American Antiquity 18: 305-14<br />
Specht, J., 1968. Preliminary Report <strong>of</strong> Excavations on Watom Island. Journal <strong>of</strong> the<br />
Polynesian Society 77: 117-34<br />
---, 1969. Prehistoric and Modern Pottery Industries <strong>of</strong> Buka Island, T.P.N.G. PhD thesis.<br />
Research School <strong>of</strong> Archaeology and Natural History, Australian National<br />
<strong>University</strong>.<br />
---, 1974. Lapita Pottery at Talasea, West New Britain. Antiquity 48: 302-6<br />
530
---, 1977. Book Review - the Lapita Pottery Style in Fiji and Its Associations. Mankind 11<br />
(1): 75-6<br />
---, 1981. Obsidian Sources at Talasea, West New Britain, Papua New Guinea. Journal<br />
<strong>of</strong> the Polynesian Society 90 (3): 337-56<br />
---, 1985. Report on Archaeological Research in the Arawe and Kandrian Areas <strong>of</strong> West<br />
New Britain Province. In J. Allen (eds) Lapita Homeland Project: Report <strong>of</strong> the<br />
1985 Field Season. Melbourne: La Trobe <strong>University</strong>, pp.48-53.<br />
---, 1991. Kreslo: A Lapita Pottery Site in Southeast New Britain, Papua New Guinea. In<br />
J. Allen and C. Gosden (eds) Report <strong>of</strong> the Lapita Homeland Project. Canberra,<br />
Australia: Department <strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong><br />
Australian National <strong>University</strong>, pp.189-204.<br />
---, 2002. Personal Communication: to M. Felgate. Noumea<br />
Specht, J., Fields J., 1984. Frank Hurley in Papua: Photographs <strong>of</strong> the 1920-1923<br />
Expeditions. Bathhurst, NSW, Australia: Robert Brown and Associates in<br />
association with the Australian Museum Trust.<br />
Specht, J., Fullagar R., Torrence R., 1991. What Was the Significance <strong>of</strong> Lapita Pottery<br />
at Talasea? Bulletin <strong>of</strong> the Indo-Pacific Prehistory Association 11: 281-94<br />
Spennemann, D. H. R., 1987. <strong>The</strong> Impact <strong>of</strong> Cyclonic Surge on Archaeological Sites in<br />
Tonga. Bulletin <strong>of</strong> the Indo-Pacific Prehistory Association 7: 75-87<br />
Spriggs, M., 1985. Archaeological Fieldwork on Nissan Island, North Solomons Province.<br />
In J. Allen (ed) Lapita Homeland Project: Report <strong>of</strong> the 1985 Field Season.<br />
Melbourne: La Trobe <strong>University</strong>, pp.58-63.<br />
---, 1990. <strong>The</strong> Changing Face <strong>of</strong> Lapita: Transformation <strong>of</strong> a Design. In M. Spriggs (ed)<br />
Lapita Design, Form and Composition: Proceedings <strong>of</strong> the Lapita Design<br />
Workshop, Canberra, Australia, December 1988. Canberra Australia: Department<br />
<strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, Australian National <strong>University</strong>,<br />
pp.83-122.<br />
531
---, 1991. Nissan, the Island in the Middle. Summary Report on Excavations at the North<br />
End <strong>of</strong> the Solomons and the South End <strong>of</strong> the Bismarcks. In J. Allen and C.<br />
Gosden (eds) Report <strong>of</strong> the Lapita Homeland Project. Canberra, Australia:<br />
Department <strong>of</strong> Prehistory, Research School <strong>of</strong> Pacific Studies, <strong>The</strong> Australian<br />
National <strong>University</strong>, pp.222-43.<br />
---, 1993. Island Melanesia: <strong>The</strong> Last 10 000 Years. In M. Spriggs, D. Yen, W. Ambrose,<br />
R. Jones, A. Thorne and A. Andrews (eds) A Community <strong>of</strong> Culture: <strong>The</strong> People<br />
and Prehistory <strong>of</strong> the Pacific. Canberra: Department <strong>of</strong> Prehistory, Research<br />
School <strong>of</strong> Pacific Studies, ANU, pp.187-205.<br />
---, 1997. <strong>The</strong> Island Melanesians. <strong>The</strong> Peoples <strong>of</strong> Southeast Asia and the Pacific.<br />
London: Blackwell.<br />
---, 2000. <strong>The</strong> Solomon Islands as Bridge and Barrier in the Settlement <strong>of</strong> the Pacific. In<br />
A. Anderson and T. Murray (eds) Australian Archaeologist: Collected Papers in<br />
Honour <strong>of</strong> Jim Allen. Canberra: Coombs Academic Publishing, ANU,<br />
---, 2001. Who Cares What Time It Is? <strong>The</strong> Importance <strong>of</strong> Chronology in Pacific<br />
Archaeology. In A. Anderson, I. Lilley and S. O'Connor (eds) Histories <strong>of</strong> Old<br />
Ages: Essays in Honour <strong>of</strong> Rhys Jones. Canberra: Pandanus Books, Research<br />
School <strong>of</strong> Pacific and Asian studies, ANU,<br />
Stewart, D. J., 1999. Formation Processes Affecting Submerged Archaeological Sites: An<br />
Overview. Geoarchaeology 14 (6): 565-87<br />
Stoddart, D. R., 1969a. Geomorphology <strong>of</strong> the Marovo Elevated Barrier Reef, New<br />
Georgia. Philosophical Transactions <strong>of</strong> the Royal Society <strong>of</strong> London, Series B,<br />
Biological Sciences 255 (800): 383-402<br />
---, 1969b. Geomorphology <strong>of</strong> the Solomon Islands Coral Reefs. Philosophical<br />
Transactions <strong>of</strong> the Royal Society <strong>of</strong> London, Series B, Biological Sciences 255<br />
(800): 355-82<br />
Summerhayes, G., 1996. Interaction in Pacific Prehistory. PhD thesis. School <strong>of</strong><br />
532
Archaeology, Faculty <strong>of</strong> Humanities, La Trobe <strong>University</strong>.<br />
---, 2000a. Lapita Interaction. Terra Australis 15. Canberra: Department <strong>of</strong> Archaeology<br />
and Natural History and Centre for Archaeological Research, Australian National<br />
<strong>University</strong>.<br />
---, 2000b. Recent Archaeological Investigations in the Bismarck Archipelago, Anir - New<br />
Ireland Province, Papua New Guinea. Bulletin <strong>of</strong> the Indo-Pacific Prehistory<br />
Association 19: 167-74<br />
---, 2000c. What's in a Pot? In A. Anderson and T. Murray (eds) Australian<br />
Archaeologist: Collected Papers in Honour <strong>of</strong> Jim Allen. Canberra: Coombs<br />
Academic Publishing, ANU, pp.291-307.<br />
---, 2001. Lapita in the Far West: Recent Developments. Archaeology in Oceania 36: 53-<br />
63<br />
---, 2002. Defining the Chronology <strong>of</strong> Lapita in the Bismarck Archipelago. In G. R. Clark,<br />
A. J. Anderson and T. Vunidilo (eds) <strong>The</strong> Archaeology <strong>of</strong> the Lapita Dispersal in<br />
Oceania: Papers from the 4th Lapita Conference, June 2000, Canberra,<br />
Australia. Pandanus Press, Research School <strong>of</strong> Pacific and Asian studies, ANU,<br />
pp.25-38.<br />
Swadling, P. L., 1976. <strong>The</strong> Occupation Sequence and Settlement Pattern on Santa Ana.<br />
In R. C. Green and M. M. Cresswell (eds) Southeast Solomons Culture History.<br />
Wellington, New Zealand: <strong>The</strong> Royal Society <strong>of</strong> New Zealand, Bulletin 11,<br />
pp.123-32.<br />
Tani, M., Longacre W. A., 1999. On Methods <strong>of</strong> Estimating Ceramic Uselife: A Revision<br />
<strong>of</strong> the Uselife Estimates <strong>of</strong> Cooking Vessels among the Kalinga, Phillipines.<br />
American Antiquity 64 (2): 299-308<br />
Tarlton, F. J., 1996. Effects <strong>of</strong> Southwest Pacific Tropical Cyclones on Archaeological<br />
Sites. MA thesis. Anthropology Department, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
Terrell, J. E., 1976. Perspectives on the Prehistory <strong>of</strong> Bougainville Island, Papua New<br />
533
Guinea: A Study in the Human Biogeography <strong>of</strong> the Southwestern Pacific. PhD<br />
thesis. Anthropology, Harvard <strong>University</strong>.<br />
Thurman, H. V., 1981. Introductory Oceanograpgy. Sydney: Charles E. Merrill Publishing<br />
Company.<br />
Torrence, R., Stevenson C., 2000. Beyond the Beach: Changing Lapita Landscapes on<br />
Garua Island, Papua New Guinea. In A. Anderson and T. Murray (eds) Australian<br />
Archaeologist: Collected Papers in Honour <strong>of</strong> Jim Allen. Canberra: Coombs<br />
Academic Publishing, ANU, pp.324-45.<br />
Torrence, R., White J. P., 2002. Tattooed Faces from Boduna Island, Papua New Guinea.<br />
In G. R. Clark, A. J. Anderson and T. Vunidilo (eds) <strong>The</strong> Archaeology <strong>of</strong> the<br />
Lapita Dispersal in Oceania: Papers from the 4th Lapita Conference, June 2000,<br />
Canberra, Australia. Pandanus Press, Research School <strong>of</strong> Pacific and Asian<br />
studies, ANU, pp.135-40.<br />
Trigger, B., 1989. A History <strong>of</strong> Archaeological Thought. Cambridge <strong>University</strong> Press.<br />
Turner, M., 2000. <strong>The</strong> Function, Design and Distribution <strong>of</strong> New Zealand Adzes. PhD<br />
thesis. Anthropology, <strong>University</strong> <strong>of</strong> <strong>Auckland</strong>.<br />
Vanderwal, R. L., 1973. <strong>The</strong> Torres Strait: Protohistory and Beyond. Occasional Papers.<br />
Anthropology Museum, <strong>University</strong> <strong>of</strong> Queensland, pp.157-94.<br />
Varien, M. D., Mills B. J., 1997. Accumulations Research: Problems and Prospects for<br />
Estimating Site Occupation Span. Journal <strong>of</strong> Archaeological Method and <strong>The</strong>ory<br />
4 (4): 141-91<br />
Varien, M. D., Potter J. M., 1997. Unpacking the Discard Equation: Simulating the<br />
Accumulation <strong>of</strong> Artefacts in the Archaeological Record. American Antiquity 62<br />
(2): 194-213<br />
Wahome, E. W., 1999. Ceramics and Prehistoric Exchange in the Admiralty Islands,<br />
Papua New Guinea. PhD thesis. Australian National <strong>University</strong>.<br />
Wandsnider, L., 1996. Describing and Comparing Archaeological Spatial Structures.<br />
534
Journal <strong>of</strong> Archaeological Method and <strong>The</strong>ory 3 (4): 319-84<br />
Ward, G. K., 1976. <strong>The</strong> Archaeology <strong>of</strong> Settlements Associated with the Chert Industry<br />
<strong>of</strong> Ulawa. In R. C. Green and M. M. Cresswell (eds) Southeast Solomons Culture<br />
History. Wellington, New Zealand: <strong>The</strong> Royal Society <strong>of</strong> New Zealand, Bulletin<br />
11, pp.161-80.<br />
Waterhouse, J. H. L., 1926. A Roviana Phrasebook. Sydney: Epworth Press.<br />
Waters, M. R., 1992. Principles <strong>of</strong> Geoarchaeology: A North American Perspective.<br />
<strong>University</strong> <strong>of</strong> Arizona Press.<br />
Weisler, M. I., 2001. Lapita Rockshelters <strong>of</strong> Elouaua and Mussau Islands (Sites EHM,<br />
EHN, EKO, EKP, and EKQ). In P. V. Kirch (ed) Lapita and Its Transformations<br />
in Near Oceania. Berkeley: Archaeological Research Facility, <strong>University</strong> <strong>of</strong><br />
California, pp.146-74.<br />
White, J. P., Coroneos C., Neall V., Boyd W., Torrence R., 2002. FEA Site, Boduna<br />
Island: Further Investigations. In S. Bedford, C. Sand and D. V. Burley (eds) Fifty<br />
Years in the Field. Essays in Honour and Celebration <strong>of</strong> Richard Shutler Jr's<br />
Archaeological Career. <strong>Auckland</strong>: New Zealand Archaeological Association<br />
Monograph 25, pp.101-8.<br />
White, J. P., Downie J. E., 1980. Excavations at Lesu, New Ireland. Asian Perspectives<br />
23 (2): 193-220<br />
White, J. P., Gorecki P., O'Brien R., Anderson P., 1985. Excavations at Bal<strong>of</strong> and<br />
Explorations Elsewhere in the New Ireland Province, 17 August - 23 September<br />
1985. In J. Allen (eds) Lapita Homeland Project: Report <strong>of</strong> the 1985 Field<br />
Season. Melbourne: La Trobe <strong>University</strong>, pp.9-.<br />
White, J. P., Harris M.-N., 1997. Changing Sources: Early Lapita Period Obsidian in the<br />
Bismarck Archipelago. Archaeology in Oceania 32: 97-107<br />
Wickler, S., 1995. Twenty-Nine Thousand Years on Buka: Long-Term Cultural Change<br />
in the Northern Solomon Islands. PhD thesis. Anthropology Department,<br />
535
<strong>University</strong> <strong>of</strong> Hawaii.<br />
---, 2001. <strong>The</strong> Prehistory <strong>of</strong> Buka: A Stepping Stone Island in the North Solomons. Terra<br />
Australis 16. Canberra: Department <strong>of</strong> Archaeology and Natural History and<br />
Centre for Archaeological Research, Australian National <strong>University</strong>.<br />
Wickler, S., Spriggs M., 1988. Pleistocene Human Occupation <strong>of</strong> the Solomon Islands,<br />
Melanesia. Antiquity 62: 703-6<br />
Willey, G. R., Phillips P., 1958. Method and <strong>The</strong>ory in American Archaeology. Chicago:<br />
<strong>University</strong> <strong>of</strong> Chicago Press.<br />
Wobst, H. M., 1983. We Cant See the Forest for the Trees: Sampling and the Shapes <strong>of</strong><br />
Archaeological Distributions. In J. A. Moore and A. S. Keene (eds)<br />
Archaeological Hammers and <strong>The</strong>ories. Academic Press, pp.37-85.<br />
Yen, D. E., 1976. Inland Settlement on Santa Cruz Island (Nendo). In R. C. Green and M.<br />
M. Cresswell (eds) Southeast Solomons Culture History. Wellington, New<br />
Zealand: <strong>The</strong> Royal Society <strong>of</strong> New Zealand, Bulletin 11, pp.203-24.<br />
536