New Perspectives on Endangered Languages. Bridging gaps ...

<strong>New</strong> <strong>Perspectives</strong> on Endangered Languages.
Bridging gaps between Sociolinguistics, Documentation and Language
Revitalization
José Antonio Flores Farfán
Fernando Ramallo

Contents
José Antonio Flores Farfán and Fernando Ramallo
Exploring links between documentation, sociolinguistics and language
revitalization: An introduction
Alexandra Yurievna Aikhenvald
The social life of a language: will Manambu survive?
Nancy C. Dorian
The Private and the Public in Documentation and Revitalization
Bruna Franchetto
Bridging Linguistic Research and Linguistic Documentation: The Kuikuro
Experience (Brazil)
Lenore A. Grenoble
Language vitality and revitalization in the Artic
Pieter Muysken
The demise and attempted revival of Uchumataqu (Uru). Values and actors
Sabine Reiter
Linguistic Vitality in the Awetí indigenous community: a case study from the
Upper Xingu multilingual area
Fernando Ramallo and José Antonio Flores Farfán
Epilogue
Subject index
Author index
2

Exploring links between documentation, sociolinguistics and language
revitalization: An introduction
José Antonio Flores Farfán
Fernando Ramallo
Since its inception, the documentation program has posed a series of open
quests. Relating documentation, sociolinguistics and language revitalization,
establishing an agenda of research priorities, constitutes an important issue
which invites developing a productive debate to which this book hopes to
contribute. Pursuing to link the documentary, sociolinguistic and
revitalization programs a series of tasks are at stake. In the light of the
potential cross-fertilization of these interdisciplinary fields, prefiguring and
exploring the need to connect documentation and revitalization is a crucial
one. The chapters presented in this book are outstanding contributions to this
purpose. As major representatives of the documentary linguistics field, all
participants aim at the advancement of finding connections between the
three agendas. In the quest of an exegesis of such trinity, let us mention and
briefly discuss some of these open questions, touching upon what to our
mind are critical issues.
The considerable rapid growth of the linguistics documentation field has
witnessed the emergence of a number of documentary linguistics trends with
different emphases. The identification of types of documentation carried out
3

in the last decades lead to the construction of two opposite paradigms, even
as conflictive ways of developing documentary research which
notwithstanding nurture each other. Bearing in mind the aim for which the
present volume was called upon, linking three very rich and complex
agendas, let us divide documentary practices in a continuum; ranging from
what we would provisionally call active documentation or documentation
oriented to the community versus received documentation or documentation
with a major concern for scientific descriptions of endangered languages. An
important tenet in and by itself, the latter no doubt has allowed establishing
important repositories of digital archives of several representatives of the
endangerment continuum worldwide. This trend constitutes a growing field
which involves the creation of open digital archives (see for example the
DoBeS program: http://www.mpi.nl/DOBES/; see also Franchetto or Reiter
in this volume). Even when this is an important task, it is yet not enough,
especially in the light of communities’ perspectives, in particular those
oriented to the retention and survival of their ancestral tongues.
As we all know, the documentary practice can become very demanding and
time consuming in the endeavor of collecting and registering data. Reaching
the point where at times there is little space to fulfill other objectives beyond
language documentation per se. A wider conception should overcome the
risk of jeopardizing one of the most important utopias of the documentary
program, linking the needs and interests of speakers themselves to the
4

documentary practice, which might be left only at the level of a desideratum.
Even when schematic, this distinction allows outlining a series of issues to
be taken care of in terms of theoretical, methodological as well as empirical
grounds to develop best practices for language documentation and
revitalization. Leaving speakers interests out of the linguistic documentation
processes often clashes with actors’ perspectives, in which the use of the
language belongs to a much more complex linguistic ecology.
The documentary program inspired by the interfaces between documentation
and sociolinguistic theory and method plus the revitalization imperative
requires finding connections between several forms of constructing the
documentation process. What is documented, who documents who, when
and how the documentation takes place, in which conditions, and so on? The
documentation itself is only one part of such ecology, in turn directly linked
to social and political issues. Outstandingly, power differentials affect and in
the last analysis determine documentation as expressed in forms and types of
corpora we construct in direct interaction with speakers. In this respect, types
of data are constructed and negotiated according to specific power dynamics
and hierarchies, a point rarely explicitly acknowledged---not to say
conceived as part of the documentation endeavor. Let us devote some
attention to this issue providing some illustrations.
Issues of power in documentary linguistics
5

Let us briefly consider the politics of power in field linguistics, on which
documentary linguistics is based. Especially in the field of endangered
languages, which has witnessed a growing interest in the last few decades,
posing a series of interesting quests regarding among others the issue at
hand; namely, the materialization of power in and by discourse in the
making of documentary research. The interest that the field of endangered
languages in its relationship to power and discourse represents is evidenced
in a number of facts even entailing the reconfiguration of priorities of the
linguistics agenda itself, in turn pinpointing to the political economy of
power in academia. The reconfiguration of priorities in the linguistic agenda
has witnessed the emergence of two “new” fields; namely, language
documentation and revitalization, which much more work devoted to the
first than to the second, producing a gap between these two fields that we
should start to seriously overcome if we are to effectively contribute to the
future of endangered communities and their languages.
Due to this state of affairs, we will devote some space to analyze the prism
of power and discourse and their dilemmas and caveats in these emergent
fields. By definition, in contrast to documentary linguistics, descriptive
(field) linguistics swiftly distances itself from speakers (cf. Himmelmann
2006) in the extreme case establishing an exploitation relationship with the
subjects which are considered objects, exerting epistemic violence,
conceiving actors as means to an end; that is to say depositories of data to
6

produce phonologies, grammars, dictionaries and advancing in the
interpretation of the typology of the world languages, demonstrating what is
and what is not possible in terms of its diverse structures.
As is well known data in descriptive linguistics is edited ---a power and
ideological gesture. For example, contact phenomena can be considered
marginal, secondary or at best parenthetic, at times totally discarded.
Moreover, interrogatory elicitations even produce induced forms according
to the linguists’ “descriptive” theoretical dreams, not to speak of biases. The
context in which linguistic data is “elicited” is omitted thus deprived from its
interactive nature, discourse texture and construction; including of course the
power relationship from which it stems. Another good illustration of a power
issue is the selection of “best informants”, which represent the linguistic
ideal of a speaker —an “uncontaminated”, “pure” (ideally monolingual) but
ironically a bilingual one. As an overall result of a descriptive or extractive
linguistics practice, highly artificial varieties are ironically produced, as we
will illustrate in more detail below, eloquently pointing to the linguist’s -
“informant” political economy.
Sociolinguistics has demonstrated that such ideal speaker does not exist,
pinpointing to a wide open variety of “real” speakers, an investigative
priority of the documentation, sociolinguistics and revitalization agendas.
Giving rise to a series of concepts and terminology which correspond to
7

different types of individuals, ranging from owners of the language, re-
callers, semi or even pseudo, last and of course death speakers of the
language, among several others (for examples of a wide range of such very
different speakers see Dorian or Muysken, this volume).
On the other hand, documentary linguistics arises (although not for the most
part explicitly) precisely in response to these and other descriptive biases; it
pretends to be exhaustive with respect to the construction of data, another
defining and characteristic feature of the documentary program. This is
expressed in the production of metadata in the form of thoroughly
annotations of the context(s) in which materials are gathered, with
commentaries on different aspects of the situation, speakers, etc., —
especially with respect to the ways in which the linguistic materials are
collected. Moreover, as part of the documentary linguistics agenda, it
pursues to be multipurpose, including coping with the needs and
expectations of the communities (cf. Gippert et al. 2006). This last
desideratum, the main focus of language revitalization, is not always the first
and foremost purpose of documentary linguistics. In the field of language
revitalization, which is most of all a political project, the main interface with
academia implies balancing power between researchers and the researched,
pursuing to establish a set of priorities directed to reversing language shift
(cf. Fishman 1991). In this sense, the field of language revitalization implies
a new type of relationship between academia and speakers of endangered
8

languages, giving rise to new (power) epistemologies in the making of
research (cf. Franchetto, this volume; Flores Farfán 2005).
Voices from the field
In Flores Farfán research in indigenous communities, he has come across a
number of instances of different hierarchical discourse practices as those
manifested in the use of research instruments. Applying a questionnaire and
other research instruments invites an interesting revision of the power issues
at stake in the interaction linguist - “informant”, which in turn are
linguistically indexicalized. In passing consider that historically actually
most dictionaries are collections of pieces of power interactions, which in
the case of Mesoamerican languages such as Nahuatl or Maya have
produced a well documented contact history giving rise to a number of
neologisms and circumlocutions to express pre contact, inexistent cultural
items, such as horses, sheep or other animals, along with religious
terminology. Some of these words have become part of the repertoire of the
communities up to this day and indeed express key new cultural acquisitions:
e.g. tomin “money”, some faded away representing ephemeral uses of
language, a telling process of the imposition of a hegemonic tongue versus
the ongoing disappearance of an increasingly compartmentalized language.
For instance, in Friar Molina’s Mexicano dictionary of 1571, a missionary
variety arose, induced by evangelization needs, which together with religious
neologisms include lexical items pinpointing to a very careful speech on the
9

side of Molina’s collaborators, suggesting not only a high attention to
discourse, but the condescending attitude towards the Friar, as part of a
power relationship materialized in discourse. For example, noamoch “my
book” (lit. my codex) in extemporaneous speech would normally be reduced
to namoch; body parts would obligatorily be possessed, yet in the dictionary
we encounter absolutive items such as maytl, “hand” or tzontli, “hair”, etc.,
pretty much induced forms (cf. Flores Farfán 2009).
Similar issues still in deed emerge in the practice of contemporary field
linguistics. Within two attitudinal trends that emerge in the making of field
linguistics—a cooperative versus a non-cooperative one. Recall that one
leads to a situation of an extreme condescending attitude, materialized in the
emergence of inexistent, at times extremely creative, ideolectal linguistic
expressions, pretty much strange to everyday usage, such as the ones just
quoted above. A very careful speech and phenomena of hypercorrection and
purism are also present as part of the above mentioned attitude. Thus
speakers in contact with linguists produce “new” and at times even
unintelligible forms or varieties of their languages! Highly indexing the
power differential between the researcher and the so called “informant” —
also a pretty telling term if thought as indexing a person conceived as a
depositary of data meant to be extracted.
10

Among other indexicalizing functions that loans may fulfill, nativization
might also represent another power (dulcification) strategy on the side of
hyper-cooperative speakers. Converging with a tendency to avoid
borrowings as much as possible, in turn shows an extreme purist attitude,
indirectly pinpointing to the endangered status of the language (for
interesting examples see for instance Hill & Hill 1986; Flores Farfán 2009).
This is the case of several words that are elicited via questionnaires, such as
kinship terminology in which speakers choose to produce forms as “the
brother of my mother” instead of simply uttering the borrowing for “uncle”,
in use in actual practice in the Nahuatl tongue. For instance, in Balsas
Nahuatl instead of using the Spanish loan tio, “uncle”, which corresponds to
the everyday syncretic use, in its possessed form, no-tio, “my uncle”, in
elicitation speakers utter descriptive forms or circumlocutions such as iikniw
de nonaan “the brother of my mother” (cf. Flores Farfán 2001). This has
been occurring already for centuries. Even when utilizing spontaneous
triggering discourse instruments inspired by experimental sociolinguistics,
such as oral or visual stimuli, amates “bark wood painted paper” stimuli,
applied to study generational differences in Mexicano speech, speakers still
pay attention to the presentation of the self as part of a power-discourse
ritual, manifested for instance in the use of numerals (which in everyday
speech are mainly in Spanish) and Classical Nahuatl (Mexicano)
innovations, such as no longer used postpositions, e.g. ipan, “on top”,
instead of the modern common preposition among adult and young
11

generations pan “on top”; compare: ipan matlaktli waan see amatsiintli “in
the eleventh amate”, instead of pan amate once).
Paradoxically, linguistic ideologies such as purism legitimize specific
practices such as abandoning the use of an indigenous language in settings
like the primary (home) or secondary spaces of socialization (schools);
perpetuating destructive ideas about the nature of the language in question.
Such ideas include judgments about the generally negative possibilities of
the indigenous tongue, reaching the point where some speakers consider that
their language does not allow a written grammar or constitutes only a
“dialect”, is incomplete, defective and the like. Degrees of purism in the
endangered language constitutes also an ideological mechanism of
legitimating power, showing a confrontation of interests of certain speakers
who intend to present themselves as “real”, “authentic”, “good” actually
even the best speakers of a language (cf Tsunoda 2005), as the above
examples suggest.
Field linguistics is an excellent example of a power situation in which such
ideologies about languages abound, are deployed and even perpetuated in
prescriptive forms such as whole grammars. Thus induction of rare or
inexistent forms, the selection of “informants”, discarding some speakers
and data over another, cataloguing and classifying, editing the materials, etc.,
12

have as an ultimate expression purist varieties of languages which eloquently
indexes of the power relationship between the linguist and his “informants”.
Power differentials directly affect and thus are a key part of the
documentation process. These facts invite sociolinguistics and
documentation practitioners to reconsider the role and place of power in the
materialization and expression of discourse differences and how this exercise
could be turned in favor of speakers of the languages themselves. This
awareness invites to articulate a different approach, outstandingly with
respect to the quest for more and better forms of revitalization, which after
all pursues to balance power relationships in the making of (documentary)
research.
Filling the gap between documentation and revitalization thus invites a more
speaker-centered approach to the documentation discipline not only
incorporating speakers in the documentation process but opening the
possibility of allowing a more equal participation on the side of the speaker
in the whole process. This poses a series of open questions. These include
identifying and analyzing speakers’ expertise knowledge as manifested in
heterogeneity of players and actors, in linguistic structures as expressed in
typologies of speakers arranged according to a series of criteria such as
power and variability or the vitality or degrees of endangerment of the
languages, etc. Serious consideration of these amongst other many other
13

variables would allow developing new forms of documentation closer to the
growing concern of revitalization. For instance, carrying out active or even
activist documentation by speakers themselves alongside with developing
language revitalization methods, analytical abilities and the production of
educational materials in their own language would probably have a positive
cumulative effect in the retention and maintenance of endangered languages,
opening interesting possibilities for their future, counteracting the destructive
forces that today still overwhelmingly favor language shift worldwide
(Florey 2008). This conception, which implies taking the next step beyond
fieldwork, is what Grinevald (2003) refers to as “fieldwork by speakers of
the language community” compared to other forms of threatened language
documentation: “on a language”, “for the language community” and “with
speakers of the language community”.
The present volume
The purpose of this volume is to contribute to the debate regarding the
perspective of documenting languages “with revitalization in mind” (Amery
2009). Each of the six works included in this volume compose a prominent
contribution to many of the challenges we have just set forth. We hope these
efforts contribute to a significant advancement of the discussion about the
importance of working collaboratively on the three agendas in regressive
sociolinguistic contexts.
14

In Chapter 1, Alexandra Y. Aikhenvald discusses the possibilities of survival
of Manambu, a language spoken by a few thousand people on the island of
<strong>New</strong> Guinea. In a line similar to the previous chapter, the author places the
possibilities of this language’s survival in the value which the community
has for it in a context of great linguistic diversity and multilingualism, with
recent changes in socioeconomic conditions and with numerous endangered
languages. As in other communities in <strong>New</strong> Guinea, Tok Pisin has achieved
such a level of community presence that not only is it now the majority
language in communicative interactions - albeit with much code-switching -
, especially among youth who consider it to be “easier”, but it is also now the
language of authority. Yet knowledge of Manambu adds value and its study
and use in strongly ritualized exchanges elevates the symbolic power and the
status of the people inside the community. This has been complemented by
the strengthening of a conservative attitude towards language, painted in
purist overtones, whose yearning is to protect the language from any foreign
contamination, while understanding that this is the best remedy for its
survival. It is a case of diglossic reversion, linked to the reproduction of that
which is sacred, and communication with the divine, something that in these
communities has great value in the sense of maintaining a legendary and
historic oral culture, demonstrating the importance of that which is sacred in
the preservation of marginalized languages.
15

In Chapter 2, Nancy Dorian addresses a new and rarely documented aspect
of the practice of documentation and revitalization. When a private variety
expands its use to include formal domains, new tensions arise from the
fieldwork in the community. Faced with the confidentiality problems
associated with working in a private environment that bases its entire
methodological strategy on the creation of a framework of trust among
actors, including the external researcher’s use of the language in
communities where the formal use of minority or indigenous languages is
less common, the researcher that opts for working in public contexts may
find himself or herself with less cooperation on the part of the community in
the measure of that his or her work privileges communicative practices and
language maintenance strategies very far, formally and functionally, from
the traditional ones. Such is the case of the construed variety in academia or
schools, which encounters resistance from the every-day communicative
practice of the community, according to Dorian.
In chapter 3, Bruna Franchetto embarks on a deep reflection upon the
relationship that exists in all fieldwork between the researcher and the
indigenous population of the subject communities. From her 30 years of
experience among the Kuikuro (Brazil), the author sheds light on the
complex processes involved in an indigenous community accepting a work
of documentation without ever losing its dominant position as a witness of
immaterial knowledge with differential values, frequently with conflict, for
16

the different agents (Dorbin, 2005). The chapter invites us to reflect upon the
true reach of documentation, such as scientific work which must provide
rewards for all rather than some. In other terms, this would fall under
folklorization of the documentation as an end in itself. The success of the
documentation will depend not only on the scientific and technological
advances, but also in good measure on the researcher’s capacity to integrate
his or her expectations into a system of foreign and distant cultural and
communicative references. In this sense the language speakers themselves
develop documentation about the researcher, the hunter becomes the hunted
and in practice they demand of him or her to play by their rules in order to
complete their own documentation. Even more so, the chapter shows that
different types of documentation are effectively possible, as mentioned
earlier. The work of Franchetto with the Kuikuro of Brazil is a paradigmatic
example of how quality documentation can go hand in hand with
revitalization processes, or even still with the strengthening of a threatened
language, as is the case with the empowerment of the language-speakers
involved in the processes. An example of this is the training of Kuikuro
videographers.
In Chapter 4, Lenore Grenoble presents the case-study of two communities
in which Evenki is spoken. Evenki is a threatened Tunguska language with a
significant dialectical variety spoken in Siberia. This type of situation is a
vivid example of the need to approach jointly the practices that the
17

documentation and revitalization ought to exercise in order to attempt to
favorably orient the future of these communities of language-speakers and
their languages, while maintaining awareness of the context of linguistic
contact and the sociolinguistic dynamics generated within it. From this
perspective, the more oriented towards revitalization the documentation is
the more effective and relevant it will be. The production of materials for the
education system, for example, becomes a necessary strategy for the
construction of an effective pedagogy for the preservation of Evenki. This
sentiment is echoed by the Evenki speakers in their communication with
linguists. Grenoble’s chapter is a sublime example of how good
documentation is essential to unravel revitalization processes and how the
coordination of the agendas of local language-speakers with those of the
linguists is possible and results in a “win-win” situation.
In Chapter 5, Pieter Muysken analyzes the causes of the Uchumataqu (Uru)
language’s disappearance and the recent attempts to revitalize it through an
actor-centered approach originating from the need to bring together the
interests of different participants and focusing on the connections among
sociolinguistics, documentation and language revitalization. In a context of
great eco-linguistic fragility, this language has been replaced by Spanish and
Aymara, currently the dominant language in the Bolivian Altiplano. The
author bases his argument first with the consideration that the decline of
Uchumataqu witnessed many external factors (migratory movements,
18

changes in economic models, population loss, exogamy and ecological
transformation) as well as local values and attitudes of the community and
local actors implicated in their maintenance and intergenerational
transmission. This chapter is of much interest in that it demonstrates that the
reversion of displacement is possible, as is the revival of a language and
culture considered extinct. This is not a case in which the emergence of a
national state, as in the case of Hebrew, or the defense of a language by a
powerful bourgeoisie, as in the case of Catalan or Basque favors the revival
of a language; rather, revitalization is emerging in the context of the
vindications of a marginalized oral culture such as that of Uru, which was
considered assimilated by the Aymara. Uru now has a favorable breeding
ground in the current context of political vindication of the indigenous
groups in Bolivia.
Last but not least, in Chapter 6, Sabine Reiter describes the crossfertilization
between sociolinguistics and documentation that she has developed during
five years of work in the Awetí community (Brazil). In the multilingual
context of Parque Indigena do Xingu (The Indigenous Park of the Xingu),
indigenous communities that speak different languages coexist and share a
common cultural system that favors cooperative relationships and mixed
marriages among communities, resulting in a complex multilingual situation.
The Awetí community has recently divided itself into two villages with
partially differentiated linguistic contact situations, but still subject to the
19

strict linguistic policies of Upper Xiguan society which decree that
notwithstanding the parents’ language, all children must at least learn the
language group of their village’s ethnic group. Even though this suggests
some level of assurance for the survival of the language, unique from other
mega-diverse linguistic situations, apparently the penetration of other
languages (Kamaiurá and Portuguese) are having a negative impact on the
vitality of Awetí. In an effort to contribute to its revitalization, the author
highlights the interest sparked by the documentation project, including the
involvement of Ametí youth.
From the beginning, all of these are first steps that we must continue to
strengthen to achieve the possibility of a labor of “revitalization”, recovery
and even promotion in the long term of threatened languages. This book in
its entirety, with its contribution of new and exciting results, lays the
foundation for this.
Conclusions
Understanding sociolinguistics as a theoretical and methodological
framework which attempts to promote change and/or social development in
human communities presents important epistemological challenges. A
sociolinguistics of development (cf. Djite 2008), in which the revitalization
of linguistic communities is the priority, opens new perspectives for the
emerging field of linguistic documentation, in which the societal aspects of
20

esearch have frequently been marginal. The need to focus on the
documentation of linguistic communities which seeks to contribute to the
revitalization of these communities requires changes in both theoretical and
methodological perspectives. Instead of creating mere museum pieces for the
future (as has been the major trend up to now), it is necessary to revitalize or
reactivate the actual use of endangered languages in the present time in order
to place language(s) at the center of the community’s development.
In the state of the art in the new field of documentary linguistics
sophisticated archives and repositories have been produced to handle and
preserve large amounts of data. These have played a prominent role in
producing robust corpora, yet not so much in the revitalization of
endangered languages. Archiving has been largely limited to the concerns of
a small group of individuals, practitioners in documentation, specifically
linguists with an interest in language description and typology. At times
these activities are grounded in best practice guidelines, paired with the use
of software, and coding schemes for interlinear speech data analysis and
transcription. Yet documentation pursues links with wider interests and
audiences, in particular speakers, as has been well established in the field (cf.
Gippert et al 2006). For instance, even when speaking of sustainable data,
almost no attention has been paid to the discussion of methodology both for
archiving plus presenting teaching and revitalizing materials. There are few
efforts of revitalization in modular and interactive multimedia for example.
21

Indeed the field of documentation lacks anything that is not sequenced
annotations or linguistic descriptive documents, with the interpretation that
links knowledge of the language with knowledge of the code. If we are to
really contribute to stop the fatal decline of many languages of the world,
disciplines such as sociolinguistics should play a much more definite role in
defining the links between both agendas, which also has lot to offer to the
area of sociolinguistics. This is what the present volume also aims at.
Acknowledgements
The editors wish to thank all of the people who made this volume possible.
First, the six authors for their generous contributions. We are very thankful
to all of the peer-reviewers who have contributed their critical comments in
order to improve the volume’s chapters: Gregory Anderson, David Bradley,
Wolf Dietrich, Maurizio Gnerre, Jane H. Hill, Nikolaus P. Himmelmann,
Gary Holton, Patricia Kwachka, Marcus Maia, Marilyn Martin-Jones, Denny
Moore and Pedro Plaza. Our most sincere appreciation goes to Kees Vaes
and to Gunther Senft for giving us the opportunity to inaugurate the new
collection Culture and Language Use: Studies in Anthropological
Linguistics by John Benjamins in which this volume appears. Last but not
least we would like to acknowledge the support that Lucía González
Gallardo provided in preparing the manuscript for publication.
References
22

Amery, R. 2009. Phoenix or relic? Documentation of languages with
revitalization in mind. Language Documentation & Conservation,
3(2): 138-148.
Djité, P. G. 2008. The Sociolinguistics of Development in Africa. Clevedon:
Multilingual Matters.
Dobrin, L. M. 2005. When our values conflict with theirs: linguists and
community empowerment in Melanesia. In Language Documentation
and Description 3, Peter K. Austin (ed.), 42-52. London: The Hans
Rausing Endangered Languages Project.
Fishman, J. 1991. Reversing Language Shift: Theoretical and Empirical
Foundations of Assistance to Threatened Languages. Clevedon:
Multilingual Matters.
Flores Farfán, J. A. 2005. Intervention in indigenous education. Culturally-
sensitive materials for bilingual Nahuatl speakers. In Mexican
indigenous languages at the dawn of the 21st century. Margarita
Hidalgo (ed.), 301-323. Berlin: Mouton de Gruyter.
Flores Farfán, J.A. 2009. Variación, ideologías y purismo lingüístico. El
caso del mexicano o náhuatl. México DF: CIESAS.
Florey, M. 2008. Language activism and the “new linguistics”: expanding
opportunities for documenting endangered languages in Indonesia. In
23
in Language Documentation and Description 5, Peter K. Austin

(ed.), 120-135. London: The Hans Rausing Endangered Languages
Project.
Gippert, J., Himmelmann, N. P. & Mosel, U. (eds.) 2006. Essentials of
Language Documentation. Berlin-<strong>New</strong> York: Mouton de Gruyter.
Grinevald, C. (2003). Speakers and documentation of endangered languages.
In Language Documentation and Description 1, Peter K. Austin
(ed.), 52-72. London: The Hans Rausing Endangered Languages
Project.
Hill, J.H. & Hill, K.C. 1986. Speaking Mexicano: Dynamics of Syncretic
Language in Central Mexico. Tucson, AZ: University of Arizona
Press.
Himmelmann, N.P. 2006. Language documentation: What is it and what is it
good for. In Essentials of Language Documentation, J. Gippert, N.
Himmelmann & U. Mosel (eds.). 1-30. Berlin-<strong>New</strong> York: Mouton de
Gruyter.
Tsunoda, T. 2005. Language endangerment and language revitalization. An
introduction, Berlin: Mouton de Gruyter.
24

Abstract
The social life of a language: will Manambu survive?
Alexandra Y. Aikhenvald
The island of <strong>New</strong> Guinea is probably the most linguistically diverse and
complex area in the world. The Sepik river basin displays cultural as well as
linguistic diversity and fragmentation, perhaps more so than other areas of
<strong>New</strong> Guinea. Many of the Sepik languages show signs of endangerment.
Manambu, from the Ndu language family, is spoken by about 2500 people.
Many Manambu children acquire Tok Pisin, the local lingua franca, as their
first language, using it in their day-to-day communication. The paper shows
that the value placed on the language by its speakers — and a number of
cultural and economic trends in modern-day Papua <strong>New</strong> Guinea — are
favourable to slowing down the process of impending language shift, and
improving the perspectives for language survival.
Keywords: Papua <strong>New</strong> Guinea multilingualism, Manambu survival, Tok
Pisin, language ideologies, purism and retention.
1. Linguistic diversity and language endangerment in the Sepik area of
<strong>New</strong> Guinea
The <strong>New</strong> Guinea region is the most linguistically diverse and
complex area in the world, with over 1,000 languages spoken in an area of
25

about 900,000 square kilometres. Three to four hundred languages spoken
there belong to the Austronesian family. Other, non-Austronesian, languages
are often referred to as 'Papuan' (see Foley 1986: 1-3; Aikhenvald and
Stebbins 2007). The term 'Papuan' is a rough denomination which covers
over sixty genetically unrelated language families and a fair number of
isolates in the area.
Within the <strong>New</strong> Guinea area, the Sepik River Basin (which includes
East Sepik and Sandaun Provinces) is, linguistically, the most complex spot
within <strong>New</strong> Guinea. 1 About 200 languages are spoken in this area, a density
unparalleled anywhere else in the world. The Sepik river basin displays
cultural as well as linguistic diversity and fragmentation, perhaps more so
than any other area of <strong>New</strong> Guinea. Reasons for this include geographic
diversity, inaccessible swampy or mountainous terrains, patterns of language
contact and language attitudes (see Foley 1986, 1988: 167-8; Aikhenvald
2004; Aikhenvald and Stebbins 2007). The average size of language
communities is significantly lower than in the <strong>New</strong> Guinea Highlands (cf.
<strong>New</strong>ton 1967).
1 The Sepik River is the largest river system in Papua <strong>New</strong> Guinea (with a catchment of
77,700 square kilometres). The river runs 1,126 kilometres from its sources in the
mountains to the sea. The extreme genetic diversity among the non-Austronesian languages
in <strong>New</strong> Guinea, with numerous families interspersed with isolates, remains a puzzle for
comparative linguists. The languages of <strong>New</strong> Guinea have suffered — perhaps, more than
any other area in the world — in the hands of 'lumpers', with their attempts to put languages
together into 'stocks', 'macro-stocks' and 'phyla' based on just a few look-alikes. Laycock
and Z'graggen (1975) postulated their 'Sepik-Ramu phylum' based on a number of
typological similarities, and two lexical similarities (terms for 'child' and 'pig'), for which
cognate sets were not provided.
26

Quite a few languages are spoken in fairly isolated villages on the
banks of remote rivulets, or perched on top of mountains. In some known
cases, such settlement patterns are due to relatively recent wars and forced
migrations. Gala (or Ngala) is spoken by about 150 people in just one
village, Swakap (or Swagup), located at a junction of a black-water river
running between the Sepik and a point a few miles up the April River. The
Gala are the remainder of a group decimated by the combined forces of the
Manambu and the Kwoma in the nineteenth century (the stories of the Gala
wars are well remembered by many indigenous peoples of the area, and have
been the subject of anthropological and linguistic discussions: see Bragge
1990; Harrison 1993; Aikhenvald 2009b). The hill-dwelling Yessan-Mayo
are said to have taken refuge there as a result of intratribal disputes (see
Bowden 1997: xviii-xxv).
Sadly, many of the languages of the Sepik area are endangered.
Numerous languages with a small number of speakers have effectively
undergone language shift. Children tend to acquire Tok Pisin, am English-
based Creole and a national lingua franca rather than the vernacular as the
first language, and full competence in the vernacular is only found among
adults. A classic case of such shift is Taiap (isolate spoken in Gapun village,
spoken by about 100 people) documented by Kulick (1987, 1992a, b) and by
Kulick and Stroud (1990).
Language socialisation in Taiap involves the conceptualisation of
Tok Pisin as a symbol of modernity and sought-after prosperity, while the
27

vernacular is associated with 'backwardness'. A somewhat similar example is
Yimas (Lower Sepik: Foley 1991: 4-6) spoken by about 250 people in two
villages. 2
Languages spoken in largish communities also tend to succumb to
the pressure of the national language. The number of people who identify
themselves as Western Iatmul (also known as Nyaura) reaches at least
30,000 people. Yet more than half of them are more comfortable speaking
Tok Pisin or Papua <strong>New</strong> Guinea English than their native language (Pius
Bonjui, p.c.; Aikhenvald 2009b). And many children do not speak Iatmul
any longer.
Iatmul, and its close relative Manambu, belong to the Ndu language
family, one of the few well established Papuan families in the Sepik Basin.
In terms of number of speakers, the Ndu family is the largest in the Sepik
area. It consists of at least six languages spoken by over 100,000 people
along the course of the middle Sepik River and to the north of it (see
Laycock 1965; Aikhenvald 2004 and 2008 on the composition of the
family).
We now turn to a situation in which a small language can be
maintained despite pressure from the national language, Tok Pisin, and the
2 Foley reports that everyone under forty is at least bilingual in Tok Pisin. Foley (1991: 5)
concludes that 'it is perhaps too early to say that the Yimas language is dying, but it must be
admitted that the prognosis is not good.' Murik (1,200, Lower Sepik) appears not to be
learnt by children any more and neither is Abu' Arapesh (with over 5,000 people: Nekitel
1985). Cemaun Arapesh (Lise Dobrin, p.c.) has less that 100 fluent speakers, while Tok
Pisin is employed by everyone. Makopin (Northern Arapesh, Torricelli family: Nidue 1990:
65-6) is currently spoken in a situation of a relatively stable triglossia; however, with the
increasing number of young people learning Tok Pisin, its survival within the next fifty
years is problematic.
28

impending threat of language shift. The case in point is the Manambu
language, and the prospectives for its survival.
2. Manambu today
Manambu 3 is currently spoken by about 2500 people in three major
villages (Avatip, Malu, and Yambon) in the Ambunti Province, and by
smallish expatriate communities in major cities of Papua <strong>New</strong> Guinea. At
present, Manambu has no monolingual speakers. 4 Most children acquire Tok
Pisin, the local lingua franca, as their first language, using it in day-to-day
communication. Papua <strong>New</strong> Guinea English is used in school, and by urban
Manambu in most settings. In the rural settings, both Tok Pisin and
Manambu are used at home, and also in rituals, still performed but in a
reduced form (compared to what was documented earlier, e.g. by Harrison
1990). Tok Pisin is dominant in village meetings, parent-teacher meetings at
school, and in church (where Manambu is also used, but to a limited extent).
That is, Tok Pisin and Manambu are in a diglossic situation.
Even ritual activities involve Tok Pisin. For instance, meetings
accompanying the funerary ritual KEkEtEp involve a large amount of
3 The materials on Manambu are based on my fieldwork conducted throughout 12 years (see
Aikhenvald 2008). My Manambu corpus comprises over 1000 pages of texts, word lists,
recordings of dialogues and conversations, songs and also participant-observation. I am
grateful to my Manambu family in Avatip, and in Malu, for sharing their language with me.
A full description of Manambu phonology is in Chapter 1 of Aikhenvald (2008). This paper
follows the principles of phonological transcription outlined there.
4 I was told that there were two children in Avatip who did not know Tok Pisin; however,
no-one seemed to know how to locate them. A few elderly ladies avoid speaking Tok Pisin
claiming that their knowledge is not good enough. However, they can communicate in that
language when they need to (for instance, when their grandchildren from cities come to
visit).
29

interaction in Tok Pisin. This has to do with their major topic which
concerns the distribution of material wealth between the parties and power
relationships in general. The powerful 'big' men (numa-dE du, big-masc.sg
man 5 ) frequently code-switch with Tok Pisin, even during the name debate
ceremony (see below). We will see that Tok Pisin and, even more so,
English are the languages of authority. This, and also the fact that men are
exposed to the outside world more than women, explains why men of all
generations code-switch more than women. 6
The increase in the number of outsiders coming to live in the
villages, mostly as a result of mixed marriages, makes proficiency in Tok
Pisin even more important. A number of teachers at the Avatip primary
School come from other areas, and do not know any Manambu. And the only
nurse in Avatip was from the Enga province; she could communicate in
Manambu, but only just.
Tok Pisin-Manambu-English multilingualism is fairly recent, just as
in many other places in <strong>New</strong> Guinea (see Aikhenvald and Stebbins 2007).
Proficient speakers of Tok Pisin were few and far between in the late 1950s.
However, this does not mean that the Manambu used to be monolingual. Up
until recently, the Manambu used to know a fair amount of the neighbouring
Iatmul (from the same family) — borrowed words, incantations and spells
5 Abbreviations used here are: COM - comitative; fem - feminine; FUT - future; IMPV -
imperative; LK - linker; masc - masculine; NEG - negative; NEG.SUB - negative
subordinate; pl - plural; PROH - prohibitive; SEQ - sequential; sg – singular.
6 The fact that men generally tend to code-switch more than women was frequently
commented on by my consultants of both sexes.
30

used to be the basis for ceremonial styles, but now are on their way out. The
Manambu used to speak and understand the languages of their closest
trading neighbours, the Kwoma/Washkuk. Only a few old people still have
this knowledge (see Aikhenvald 2009a, on language contact patterns in the
area).
At first sight, the prospects for language survival appear rather
dismal: Tok Pisin seems to be intruding into every sphere of life, and is
being increasingly used by children and by their care-takers. We will now
look at the patterns of use of the two languages.
3. Tok Pisin and Manambu
When children and teenagers speak among themselves, they use
hardly any Manambu. A substantial amount of parent-child communication
in the village context involves Tok Pisin, or even English, especially with
young children. Instead of saying to a toddler Mæya! 'Come' a mother would
say Yu kam! in Tok Pisin. Alternatively, a mother may use the topical noun
phrase in Tok Pisin, and the verb in Manambu. In (1), 'your name' appears in
Tok Pisin. The Manambu verb is omitted in the second occurrence of this
sentence (here and throughout the paper Tok Pisin forms are underlined
within examples; English forms are in bold face both in Manambu and in the
English translations):
(1) nem bilong yu aw nem bilong yu yet
31

name of you IMPV+speak name of you self
‘Say your name, your own name!’
Not infrequently, a command is in Tok Pisin, too, as in (2). Only the
personal cross-referencing -a is Manambu:
(2) sekan sekan anti-a
shake.hand shake.hand auntie-3fem.sg
'Shake hand, shake hand, (she) is an auntie'
When one day I expressed worry about little Joana (about 3 years
old) speaking only Tok Pisin, her mother reassured me:
(3) yi-kE-na wali-kudi wukEn
speak-FUT-3fem.sgNON.PAST white-language together
tEp-a kudi wukEn
village-LK language together
'She will speak, Tok Pisin together with the village language
(that is, Manambu)'
This attitude reflects the reality. 7 At the very early age, children are
given instructions in both languages simultaneously. That much of care-
7 Adult speakers of Manambu lament the language loss by some children in the urban
32

takers' interaction with toddlers and children in general involves commands
and instructions comes as no surprise. Command words and imperatives are
among the first forms acquired by children in the Kaluli-speaking area of
<strong>New</strong> Guinea (Schieffelin 1985), and by the Ku Waru in the Eastern
Highlands Province (Rumsey 2003). The reason is simple: their care-takers,
including mothers and older children, use directives and imperatives more
often than any other forms to address them. This is consistent with patterns
all over the world (see Berman 1985, and a summary in Chapter 9 of
Aikhenvald 2010).
A typical command for a child to listen to what the care-taker says is:
(4) [ñEn awuk] [harim tok ada]
you.fem IMPV+listen listen speech be'.IMPV
'You listen! Listen to speech!'
Repeating the same word, phrase, clause or sentence twice — once in
Manambu and once in Tok Pisin — can be conceived of as a way of teaching
children two languages simultaneously. This is reminiscent of the
phenomenon of 'lexical pairing'. In a number of languages, a native term
would be accompanied by a borrowed one, as a mechanism for 'gradual
mediation' of grammatical change in progress. This has been described by
environment (this is unlike adult speakers of Australian Aboriginal languages who often do
not recognise that the languages are endangered; see, for instance, Schmidt 1990). The
statement in (3) is not about language loss; it is about late language acquisition.
33

described by Hajek (2006: 170), for Tetun Dili, the major lingua franca of
East Timor in contact with Portuguese.
In the case of Manambu mothers, repeating the command twice in
different languages is a further way of getting the point across: if the child
has not understood it in one language (or pretends not to), they will have to
understand it in the other one.
Repetition typically involves the whole clause, as in (3).
Alternatively, a command in Manambu appears rephrased with a Tok Pisin
clause (5).
(5) yata-u mama i karim yu
carry.in.one's.arms-1sgIMPV mama PREDICATE.MARKER carry you
‘Shall I carry you? (Manambu) Shall Mummy carry you?’(Tok Pisin)
The Manambu mothers I talked to referred to Tok Pisin as being
'easier' to acquire than Manambu. Indeed, many perceive Tok Pisin as a
'short cut'. In many cases, Tok Pisin allows to use one simple word where in
Manambu one would require making a choice out of many.
Consider commands like 'come!', 'go!' or even saying 'good-bye'. In
Tok Pisin, one can say yu go! (you go!) as a farewell formula. In contrast,
traditional Manambu speech etiquette requires adding an appropriate address
form if one wishes to be nice and polite.
34

Choosing an appropriate term of address is not simple matter.
Roughly, the world is classified into totemic categories each of which is
owned by a subclan. Membership within a subclan is inherited from one's
father. So, if my father belongs to the Wulwi-ñawi subclan traditionally
associated with sun and moon (which 'belong' to this clan), I can be greeted
as bap 'moon'. It is also good style to greet someone using names of objects
and totems belonging to one's mother's clan. So, if my mother belongs to the
Gla:gw clan which 'owns' everything dark, belonging to the bush and the
earth, and also bow and arrow, I can also be greeted as amnEbi 'bow and
arrow'). That is, an appropriate loving way of farewelling me would be:
(6) maya bap
go.2sgIMPV moon(totem.of.addressee's.father's.clan)
amnEbi
bow.arrow(totem.of.addressee's.father's.clan)
35
'Go, moon (that is, the one who belongs to the clan Wulwi-
ñawi), bow and arrow (that is, the one whose mother belongs
to the clan owning bow and arrow'
Tok Pisin provide a conveniently 'short' way of avoiding this
complexity — this is considered appropriate for small children (who, as yet,
are not 'sensible', kwam tE-na-di (crazy be-ACTION.FOCUS-
3plNON.PAST)).

Manambu abounds in polysemous terms. Tok Pisin helps
DISAMBIGUATE POLYSEMY. The verb wukE- (see (4)) can mean 'hear', 'listen',
'smell', 'obey' and 'worry'. By using a Tok Pisin term wori 'worry' to replace
wukE- a speaker avoids a potential misunderstanding.
The same applies to yata- 'carry in one's arms (a smallish object)' in
(5). Manambu has a number of verbs distinguishing different types of
carrying (on one's head, on one's shoulder, in one's arms (as something big),
or under the armpit': Aikhenvald 2008: 356). The Tok Pisin verb karim
covers any type of carrying. It is believed to be 'easy' for a child to 'hear', and
to use. Manambu has one term kE- covering 'eat, drink, consume (e.g.
smoke)' (see Aikhenvald 2009b, for a detailed discussion). Tok Pisin
distinguishes kaikai 'eat' and dringim 'drink' — and some speakers,
especially mothers of pre-school children, use these words when urging
children to eat or to drink, to ensure that the distinction is made.
Similarly to other indigenous languages of the area, Manambu has a
complex kinship system. In my experience, children do not acquire it fully
until they are well into their teens. And it is very convenient for a mother to
refer to someone as anti 'auntie' (as in (2)) as a cover-term rather than going
into detailed relationships.
Tok Pisin can be used as a way of disambiguating polysemous items
in communication from adult to adult. The verb vE- 'see, look' in Manambu
is also used in the meaning 'read'. To avoid misunderstanding, adults use the
Tok Pisin ridim instead of the polysemous term. Traditional Manambu had
36

the same word meaning 'truth' and 'lie' (sua:l). The two meanings can be
disambiguated with different constructions: for instance, a collocation sua:l
taka- (truth/lie put-) means 'tell a lie', and the noun phrase mEya sua:l (real
truth/lie) means 'real truth' (more details are given in Aikhenvald 2008).
Nowadays, many use the two different terms from Tok Pisin, giaman 'lie'
(originally from English gammon) and tru 'true', or tru stori.
The lure of a quick-and-easy 'shortcut' is strong for most people —
with the exception of traditionally-minded elderly ladies who feel insecure in
their Tok Pisin. It appears to be easier to say stori 'story' than to choose an
exact Manambu term — gabu-ma:j 'traditional story', or wa-saki-ma:j 'true
story'. A short Tok Pisin word hat 'hat' is easier than Manambu aba-wapi
(head+LK-clothing). A lengthy ap tE-na-d-E mi (bone/strength 'stand'/have-
ACTION.FOCUS-3masc.sgNON.PAST-LK wood) 'hard wood' is often
replaced with short strongpela (Tok Pisin) mi (Manambu), with the same
meaning. The same applies to numbers higher than ten which have a very
complex structure in Manambu (Aikhenvald 2008: 239-41, 2004). As a
result, Tok Pisin and English numbers are preferred: speakers themselves say
that this is 'an easy' and 'a lazy' option.
In summary: Tok Pisin (but not English) plays a special role in
language socialization: it is considered 'easier' than Manambu and is
employed as a 'baby talk'.
37

This takes us to further functions of code-switching with Tok Pisin,
and with English. 8
A further major function of Tok Pisin is that of a language of
authority, also known as 'DIRECTIVE' function. Consider the following
interaction. The mother was getting more and more annoyed with a five-
year-old girl trying to pull bits off the string bag the mother was knitting.
She starts telling her off in Manambu (7), and then switches to a mixture of
English and Tok Pisin (8): this makes the threat real and immediate. the girl
runs away. English forms are in bold face.
(7) a kur-tukwa da-n ada
then do-PROH go.down-SEQ sit:IMPV
‘Don't do (it), sit down’
(8) naughty yu stupid idiot
naughty you stupid idiot
bai mi pait-im yu nogut tru
FUT I hit-TRANSITIVISER you bad/strong really
'Naughty, stupid idiot, I will hit you really strongly'
8 The status of frequently used Tok Pisin and English forms as loans or as code-switches is
not easy to ascertain (as is often the case with such distinction: see discussion in Aikhenvald
and Dixon 2006: 333-4 and references there). For the purposes of this paper, we treat all
Tok Pisin and English inserts as code-switches. The problem of potential influence of
Manambu on the Tok Pisin of the area (in the spirit of Ramallo's 2007 discussion of contactinduced
change in varieties of Spanish spoken in the areas of Spain whose inhabitants were
not originally Spanish-speaking) is a matter for future study.
38

This kind of language socialization is strongly reminiscent of the
situation in Gapun village analysed by Kulick (1987). In both cases, code-
switching and language mixing vary depending on speech genre, and have a
special illocutionary force. That Tok Pisin is indeed the language of
authority is evident. Not just parents, but also school teachers and members
of the village council use Tok Pisin in most circumstances when they need to
show who is in charge.
English is also in the process of becoming an even stronger 'power-
talk', especially for those for whom Tok Pisin is the major language anyway.
A group of young boys gathered around Joel's house were obviously up to
no good. My adoptive sister went up to them to enquire about their
intentions, and addressed them as 'Hello boys, what are you doing here?' The
boys muttered something unintelligible and disappeared. Using English was
her way of asserting her authority over the wayward youngsters — and she
herself made no bones about it, explaining to me that, had she used
Manambu or Tok Pisin, they would not have 'listened'. This is very similar to
the ways in which the majority languages — be it Spanish, or Portuguese, or
English — are used throughout the world by minority speakers to claim
power over others (see discussion, and references, in Aikhenvald 2002: 187-
211, Hill and Hill 1986: 364-86 and also Gumperz 1982). The motivation for
this is fairly straightforward: speakers of the majority language are the ones
in the position of authority, and so those who speak minority languages
39

employ these same languages to acquire an air of authority; this may happen
consciously or not.
Another function of Tok Pisin is FILLING A PERCEIVED GAP in the
language. Manambu does not have a dedicated contrast marker, so tasol 'but'
is used. A similar explanation would hold for the introduction of sapos 'if'
(Manambu has a fairly elaborate system of conditionals, but not a word for
'if'), and for the disjunction o. 9
Similar examples abound for lexical items. Filling a lexical gap
accounts for the use of Tok Pisin forms mas 'must', tabu 'forbidden', and the
verb laik 'like'. The concepts of 'liking' and 'wanting' can be expressed by
using the desiderative mood, a speech report construction (e.g. 'I say you
go!, meaning 'I want you to go'), a complex predicate with the form mEyakw
kwa- or a body part construction with the noun mawul 'the inside'. But each
of these constructions is rather complex, and fraught with additional
overtones. Most speakers find it easier to say wun laiki-na-wun 'I like' or
wun ma: laik (I NEG like) 'I don't like' rather than wun ma: wa-na-wun (I
NEG say-ACTION.FOCUS-1sgNON.PAST) 'I am saying no'. Note that the
latter form is in fact ambiguous between 'I don't want', 'I don't like' and 'I
refuse' ('I say "no" to'). Here, the introduction of Tok Pisin results in a
resolution of potential ambiguity. And this takes us to the next point.
9 This appears to agree with a cross-linguistic tendency to borrow clause-linking devices
(see Dixon 2009). Curnow (2001) surveys suggestions for 'hierarchies of borrowing' and
finds none of them convincing (also see the summary in Aikhenvald 2006).
40

In addition to this, code-switching may be used to stress a particular
point (see Hill and Hill 1986: 377). Tok Pisin forms often sound more
expressive than their Manambu equivalents. If someone really approves of
something, they are likely to exclaim Nambawan! (from English number
one), rather than Vyakat-a-mEy (good-LK-very) 'very good'; or tru-a! (true-
3fem.sgPRED) 'it is true!' rather than mEy-a sual-a 'it is really a true story!'.
During the Saki (name debate) ceremony, such exclamations were
accompanied with an often unanimous sound of appreciation (sounding like
a very loud AAAã-ˆˆˆnÃ). And a negative Nogat! (from English no and got)
reflects a much stronger rejection than a simple Manambu ma: or ka. This
difference in illocutionary force between Tok Pisin and Manambu can be
linked together with the role of Tok Pisin as 'power-talk', for the newly
introduced power of the foreign, 'white man's' society.
4 The power of knowledge, and the value of words
Speakers often conceive of their languages in terms of lexicon,
putting primary value on the knowledge of words. As Thurston (1987: 93)
put it, 'since people generally construe languages as being collections of
words, it is primarily by lexical form that linguistic groups identify linguistic
contrasts among themselves'. A feature of the Manambu is their focus on
ownership of ancestral names and totems, and on traditional personal names.
The Manambu culture is — similarly to many other Sepik cultures —
centred on exchange and value assigned to outside material and non-material
41

goods. In many Sepik societies, incantations and even names and individual
words were traded and bought (see Harrison 1990: 20-3, for a general
perspective). These 'acquisitions' used to surface in various ritual speech
styles, many of them effectively lost in modern days. One of the major
rituals remains a name debate, saki. Each clan and subclan owns a limited
number of names belonging to its ancestors. If a subclan decides that a name
considered their own was inappropriately bestowed onto a member of
another subclan, the name ownership is debated — as it would be in court
(see Harrison 1990, for a full-length study of name debates in their
traditional form; further discussion of modern shorter versions of names
debates is in Aikhenvald 2008: 11-15). Speakers of Manambu value the
knowledge of and the proficiency in totemic terms and personal names each
of which links its bearer to an important ancestor or a mythical personality.
To acquire respect within Manambu society, a man has to be a distinguished
orator. This involves the capacity of defending a name belonging to one's
subclan — that is, knowing the genealogies, and the ancestors going many
generations back.
Children, especially boys, are taught to recite names of their
ancestors in Manambu— even before they learn how to speak the language.
M., a four-year old son of my classificatory sister, spoke mostly Tok Pisin.
His father, K., a man in his early thirties, considered himself 'in training' as
an incipient orator: he could recite names for fourteen generations of his
subclan. Playing with M., he would teach him to repeat names, explaining in
42

Tok Pisin who the actual characters were. As a result, M. learnt at least some
names before he could speak Manambu.
The 'value of words' appears to be at the heart of a marked tendency
to use the meanings of Manambu words — and not Tok Pisin or English
imports — to refer to newly introduced items, and activities. The
maintenance of these appears to be ensured by language purists concerned
with the survival of Manambu, and the general attitude to 'rubbish talk' —
Manambu peppered with Tok Pisin. In public places outside the Manambu
speaking villages, many of these words serve the function of 'secret talk' —
everyone would understand mani (Tok Pisin for 'money'), but hardly anyone
would make any sense of kabak (lit. 'stone') 'a lot of money' or sa:n (lit. shell
valuable) 'money'.
For instance, rifle is called jarkañ (a term for bamboo shoot
originally used as a storing tube), and jarkañ lE-kE væy (bamboo.tube
3fem.sg-POSS+3fem.sg spear, lit.'bamboo tube's little spear') is the word for
'bullet'. A term for long piece of bamboo kañgu is used to refer to a
policeman (by reference to policemen carrying a long bamboo-like rifle on
their shoulder). The word jElEg 'ten shell valuables strung together' is used
for a ten-kina note. KabasEk 'seed' now means 'rice', tEk-E-mi
(seed/little.fruit-LK-tree) 'seed, little fruit of a tree' is a term for 'medicine';
wali-na:gw (white.man-sago) refers to 'biscuit', and wali-gus (white.man-
paddle) means 'outboard motor'. Few people say kar for 'car' — most prefer
val 'canoe', and hardly anyone says umbrella — aba-ñEg (head+LK-leaf) is
43

used instead. The term for a plane is mi val (high canoe); though many do
switch to balus, a Tok Pisin term. Telefon is in competition with mi ma:j
(high talk). The term jEb 'design, drawing' is also used for 'letter, literacy'; its
major competitor is English letter. And the term taba-ñE (hand+LK-sun) is
used in the meaning of 'watch' more than the English word watch.
Manambu purists go even further in suggesting lexically 'pure'
innovations covering educational, financial and religious terminology, e.g.
suku-mawul (carve-insides) 'patience, that is, the mindset of a carver', kalipa-
dE du (teach-3masc.sg man) 'male teacher', sa:n warapwi-dana tamiy
(money change-3pl area) 'stock exchange', du-awa kwa-mar-na ta:kw (man-
COM lie-NEG.SUB-ACTION.FOCUS+3fem.sg.NON.PAST woman, lit.
'woman who does not lie with men') 'virgin', God ma:j krayin kalpa-di (God
speech carry+SEQ teach-3plNON.PAST, lit. 'teachers carrying God's
speech') 'disciples' and even NEma-dE Du (big-masc.sg man) or Mayir (lit.
spirit) 'The Christian God'.
It is too early to evaluate the long-term effects of language
engineering by this group of purists. Conservative attitudes toward
loanwords are known to have hampered efforts to maintain endangered
languages; 'unrealistically severe older-speaker purism can discourage
younger speakers' (Dorian 1994; similar points were raised by Hill and Hill
1986: 140-1). And Hamp (1989) suggested that if a minority language
survives next to a larger dominant language, it has to allow for a certain
amount of borrowing of morphemes. On the other hand, a certain amount of
44

purism may stop otherwise unlimited borrowing and code-switching with
Tok Pisin. More importantly, it helps Manambu speakers to preserve the
words of their own.
There is a certain amount of language consciousness and resistance
to borrowings among some young villagers. One of the urban Manambu,
when she visited Malu village, asked a ten-year-old boy to tie her canoe for
her, using a Tok Pisin verb pas 'be closed, stuck':
(9) val a-pas
Canoe IMPV-tie
'Tie the canoe!'
The boy corrected her, saying:
(10) a-pas ma: atawtak aw
IMPV-tie NEG IMPV+tie+put IMPV+say
''Not "apas", say "tie"
Since this urban Manambu is known to be a purist and an authority
on Manambu culture, the boy could have just been trying to get his own
back, demonstrating that he knows enough to correct the 'authority'. But
opportunities like this no doubt enhance the language awareness of speakers.
Children who are proficient in Manambu feel valued by their seniors. It was
45

a matter of particular pride for eleven-year old T., to be able to correct her
classificatory mothers and grandmothers whenever they let slip a Tok Pisin
word. No matter how much Tok Pisin she uses herself, she is a language
revivalist in the making.
There are further indications that speakers value Manambu and are
not prepared to just let it go. An important group of these are the urban
Manambu returning to the villages where they were born.
5. Back to grassroots: the impact of the urban Manambu
A number of cultural and economic trends in modern-day Papua <strong>New</strong>
Guinea could also be favourable to slowing down the process of impending
language shift, and improving the perspective for language survival.
Manambu is spoken by 200 or so people in urban communities,
including Port Moresby, Lae, Madang, Wewak and also Kokopo. Within the
urban communities, Manambu is employed as a sort of 'secret language' and
an in-group means of communication emblematic in itself. Many of the
urban Manambu have made their way into what they call the 'white man's
world': they include the former Chief of Staff of the Army, a number of
lieutenant-colonels and colonels, public servants of high rank, and a number
of High Commissioners and Embassy staff, who have worked in many
countries, from <strong>New</strong> Zealand to the Philippines. These are village-born
people who had received some of their 'training' in totemic terms and
Manambu lore from their fathers and mothers before they left the villages to
46

study (at age 7-10). And now they are retiring. Many are going back to the
villages which have more to offer than the dangerous and insecure life in
gang-ridden towns. Life in the village is cheaper and a pension goes further
than in a town.
In addition, the sheer prestige of Manambu speakers — such as Paul
Badaibæg Kat, a village councillor and the owner of a number of shops; the
late Eric Yuamalen, a renowned school teacher; John Sepaywus, a former
village councillor, and others — in a larger Sepik context helps maintain the
high prestige of the language, and of the traditional knowledge. Avatip is
lucky in having Leo Yabwi Luma, a highly knowledgeable and respected
man and an elaborate connoisseur of the language, and of the tradition, as the
headmaster of its primary school.
Within a broader context, the fact that many members of the urban
elite in the cities are Manambu provides further appreciation of those who
know the language, and the lore. A growing number of the Manambu
acquire higher degrees overseas and return to their country to work in
universities and various companies. These people create role-models for the
new generation — those who want to be 'like uncle Leo', or 'like uncle Joel',
would try and learn at least some Manambu to maintain their status in the
village, and outside it.
As the retiring urban elite return to the village, they make a point of
'learning' how to be big men and orators. This has already brought another
influx of interest and culture revival into the village. The activity of an
47

overseas anthropologist and an overseas linguist each working with highly
respected Manambu speakers boosts the prestige of the language as a means
of day-to-day communication; this has already initiated a community-based
project of Manambu language and culture documentation and teaching (see
Hornberger and King 2000: 185, on the importance of such activities in the
context of Quechua language maintenance). The 'revival' of the Manambu
school program is now under way, under the leadership of Leo Yabwi Luma.
Despite signs of the expansion of Tok Pisin and English, Manambu is not as
endangered as one might think.
To conclude: the role models of returning urban Manambu, and the
power and enduring value of cultural knowledge conceived as the
knowledge of important words and names, is what will ensure that the
language lives on. Tok Pisin and Papua <strong>New</strong> Guinea English are there to
stay, as languages of outside power and authority conceptualized as
attributes of the 'white man's' world. In addition, Tok Pisin is a useful tool
for 'facilitating' learning language by children who cannot yet access cultural
knowledge. But neither of these intruders will shatter the position of
Manambu, as the language of spiritual and symbolic power, which, in this
Sepik culture, is the most valued. Manambu remains an emblematic
language for the people it 'belongs' to, and the 'owners' are determined to
pass it on.
48

References
Aikhenvald, Alexandra Y. 2002. Language Contact in Amazonia. Oxford:
Oxford University Press.
Aikhenvald, Alexandra Y. 2004. “Language endangerment in the Sepik area
of Papua <strong>New</strong> Guinea”. In Lectures on Endangered Languages: 5 -
From Tokyo and Kyoto Conferences 2002, O. Sakiyama and F. Endo
(eds), 97-142. Suita, Osaka: The project 'Endangered languages of
the Pacific Rim' (ELPR).
Aikhenvald, Alexandra Y. 2006. “Grammars in contact: a cross-linguistic
perspective”. In Grammars in Contact: A Cross-linguistic Typology,
Aikhenvald and Dixon (eds), 1-66. Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. 2008. The Manambu language of East Sepik,
Papua <strong>New</strong> Guinea. Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. 2009a. “Language contact along the Sepik River,
Papua <strong>New</strong> Guinea”. Anthropological Linguistics 50: 1-66 (2008,
published 2009).
Aikhenvald, Alexandra Y. 2009b. “"Eating", "drinking" and "smoking": a
generic verb and its semantics in Manambu”. In The Linguistics of
eating and drinking, John <strong>New</strong>man (ed), 91-108. Amsterdam: John
Benjamins.
Aikhenvald, Alexandra Y. 2010. Imperatives and Commands. Oxford:
Oxford University Press.
49

Aikhenvald, Alexandra Y. and R. M. W. Dixon. 2006. Grammars in
Contact: A Cross-linguistic Typology. Oxford: Oxford University
Press.
Aikhenvald, Alexandra Y. and Tonya N. Stebbins. 2007. “Languages of
Papua <strong>New</strong> Guinea”. In Vanishing Languages of the Pacific, Michael
Krauss and Osahito Miyaoka (eds), 239-66. Oxford: Oxford
University Press.
Berman, Ruth.1985. “The acquisition of Hebrew”. In The cross-linguistic
study of language acquisition. Volume 1. The Data, Dan I. Slobin (ed),
255-372. Hillsdale: NJ: Lawrence Erlbaum Associates.
Bowden, Ross. 1997. A Dictionary of Kwoma, a Papuan Language of North-
east <strong>New</strong> Guinea. Canberra: Pacific Linguistics.
Bragge, L. 1990. “The Japandai migrations”. In Sepik Heritage. Tradition
and Change in Papua <strong>New</strong> Guinea, N. Lutkehaus, C. Kaufmann, W.
E. Mitchell, D. <strong>New</strong>ton, L. Osmundsen & M. Schuster (eds), 36-49.
Durham: Carolina Academic Press.
Curnow, Timothy J. 2001. “What language features can be "borrowed'?”. In
Areal diffusion and genetic inheritance: problems in comparative
linguistics, Alexandra Y. Aikhenvald and R. M. W. Dixon (eds),
412-36. Oxford: Oxford University Press.
Dixon, R. M. W. 2009. “The semantics of clause linking in typological
50
perspective”. In The semantics of clause linking. A cross-linguistic

typology, R. M. W. Dixon and Alexandra Y. Aikhenvald (eds), 1-55.
Oxford: Oxford University Press.
Dorian, Nancy. 1994. “Purism vs. compromise in language revitalisation and
language revival”. Language in Society 23: 479-94.
Foley, W. A. 1986. The Papuan Languages of <strong>New</strong> Guinea. Cambridge:
Cambridge University Press.
Foley, W. A. 1988. “Language birth: the processes of pidginisation and
creolisation”. In Language: The Socio-cultural Survey. Vol. IV of
Linguistics: The Cambridge Survey, F. J. <strong>New</strong>meyer (ed), 162-84.
Cambridge: Cambridge University Press.
Foley, W. A. 1991. The Yimas Language of <strong>New</strong> Guinea. California:
Stanford University Press.
Gumperz, J. J. 1982. Discourse strategies. Cambridge: Cambridge
University Press.
Hajek, John. 2006. “Language contact and convergence in East Timor: the
case of Tetun Dili”. In Grammars in Contact: A Cross-linguistic
Typology, A. Y. Aikhenvald and R. M. Dixon (eds), 163-78. Oxford:
Oxford University Press.
Hamp, Eric. 1989. “On signs of health and death”. In Investigating
Obsolescence. Studies in Language Contraction and Death, N.
Dorian (ed), 197-210. Cambridge: Cambridge University Press.
Harrison, Simon J. 1990. Stealing People's Names. History and Politics in a
Sepik River Cosmology. Cambridge University Press: Cambridge.
51

Harrison, Simon J. 1993. The mask of war: Violence, Ritual and the Self in
Melanesia. Manchester: Manchester University Press.
Hill, Jane and K. C. Hill. 1986. Speaking Mexicano. Dynamics of Syncretic
Language in Central Mexico. Tucson: The University of Arizona
Press.
Hornberger, N. H. and K. A. King. 2000. “Reversing Quechua language shift
in South America”. In Can Threatened Languages Be Saved?
Reversing Language Shift, Revisited: A 21st Century Perspective, J.
Fishman (ed), 166-94. Clevedon: Multilingual Matters.
Kulick, Don. 1987. “Language shift and language socialisation in Gapun. A
report on fieldwork in progress”. Language and Linguistics in
Melanesia 17: 125-50.
Kulick, Don. 1992a. Language Shift and Cultural Reproduction:
Socialisation, Self, and Syncretism in a Papua <strong>New</strong> Guinean Village.
Cambridge: Cambridge University Press.
Kulick, Don. 1992b. “Language shift as cultural reproduction”. In Culture
change, language change. Case studies from Melanesia, T. E. Dutton
(ed), 7-26. Canberra: Pacific Linguistics.
Kulick, Don. and C. Stroud. 1990. “Code-switching in Gapun: Social and
linguistic aspects of language used in a language shifting
community”. In Melanesian pidgin and Tok Pisin, J. W. M. Verhaar
(ed), 205-34. Amsterdam: John Benjamins.
52

Laycock, D. C. 1965. The Ndu Language Family (Sepik District, <strong>New</strong>
Guinea). Canberra: Linguistic Circle of Canberra Publications.
Laycock, D. C. and J. Z'graggen. 1975. “The Sepik-Ramu phylum”. In <strong>New</strong>
Guinea Area Languages and Language Study, Vol. 1. Papuan
Languages and the <strong>New</strong> Guinea Linguistic Scene, S. A. Wurm (ed),
731-64. Canberra: Pacific Linguistics.
Nekitel, Otto. 1985. Sociolinguistic Aspects of Abu': A Papuan Language of
the Sepik area, Papua <strong>New</strong> Guinea. PhD thesis, The Australian
National University.
Nidue, J. A. 1990. “Language use in a <strong>New</strong> Guinea Village: a triglossic
profile of Makopin I”. Language and Linguistics in Melanesia 21:
47-69.
<strong>New</strong>ton, Douglas. 1967. “Oral tradition and art history in the Sepik district”.
In Essays on the verbal and visual arts. Proceedings of the 1966
Spring Meeting of the American Ethnological Society, J. Helm (ed),
200-15. Seattle: American Ethnological Society.
Ramallo, Fernando. 2007. “Sociolinguistics of Spanish in Galicia”.
International Journal of the Sociology of Language 184: 21-36.
Rumsey, Alan. 2003. “Language, desire, and the ontogenesis of
intersubjectivity”. Language and Communication 23: 169-87.
Schieffelin, Bambi. 1985. “The acquisition of Kaluli”. In The cross-
linguistic study of language acquisition. Volume 1. The Data, Dan I.
Slobin (ed), 525-92. Hillsdale: NJ: Lawrence Erlbaum Associates.
53

Schmidt, Annette. 1990. The loss of Australia's Aboriginal language
heritage. Canberra: Australian Aboriginal Studies Press.
Thurston, W. R. 1987. Processes of change in the languages of northwestern
Britain. Canberra: Pacific Linguistics.
54

Abstract
The Private and the Public in Documentation and Revitalization
Nancy C. Dorian
Relatively “private” and relatively “public” aspects of language use,
acquisition, and transmission come into question both in linguistic fieldwork
and in language revitalization efforts. Sequestered dyadic fieldwork tends to
encourage confidentiality, team-based or whole-community documentation
may prevent self-censorship slippage, but neither approach resolves acute
informed-consent dilemmas. The former approach may encourage the
researcher’s local-language acquisition, the latter require the researchers’ use
of a dominant-language lingua franca. School-based revitalization transposes
intimate-setting minority languages into public settings where solidary
dialect usages are replaced by codifications and coinages, perpetuating the
isolation of residual speaker communities. Availability of school
transmission risks further neglect of home transmission. Long-established
closed-community use of a minority language may prevent recruitment of
second-language learners. In the more extreme environments of language
endangerment and loss, researcher involvement in attempted solutions for
revitalization and transmission obstacles has contentiously but arguably
importantly followed involvement in description and documentation.
55

Keywords: Varieties of Welsh in the public and private spheres; fieldwork
methodology and practice; second language learners in the context of
language loss.
1. Introduction
Among those who are strongly supportive of efforts to revitalize
receding languages, as well as to document them, it is well recognized that
the barriers to success are enormous. Fishman’s 1991 volume Reversing
Language Shift discusses many of those barriers, and voluminous
contributions to the literature on this topic have expanded the discussion. In
what follows here, I focus initially on potential problems in the linguistic
fieldwork setting and the language revitalization setting, arising from
different positions with regard to relatively private versus relatively public
language use, acquisition, and transmission. Efforts to lessen such problems
and the prospects for lessening them further are discussed thereafter, with
particular focus on the increasing professional involvement in these efforts.
2. The fieldwork context
2.1 The problems of “private” vs. “public” fieldwork and informed consent
Of late there has been a strong emphasis on teamwork in language
documentation, with researchers cooperating to record language use in a
fuller range of cultural contexts and also to cover spheres outside the strictly
linguistic, such as ethnobotany and traditional song (Wittenburg 2003;
56

Himmelmann 2008). Full documentation is now also assumed to include
visual as well as auditory recording (Csató and Nathan 2003; Wittenburg
2003), so that more of the discourse context will be recoverable and
analyzable for future researchers (and revitalizers), as well as for those
currently making the record. Archiving, too, is typically fuller and more
responsible, both as to completeness and as to accessibility (Thieberger and
Musgrave 2007).
One of several advantages in approaching fieldwork in this way is the
relative publicness of the activity. The arrival of a team of documenters
carrying recording machinery of various kinds is a conspicuous occurrence
in most small-language settings, and if community cultural activities are
being filmed, many local participants will be involved. Some funding
agencies routinely ask for some sort of written consent from the community
in which the documentation is undertaken, so that the project will also be as
broadly consensual as possible (Grinevald 2006). Serious questions still arise
about the degree to which individuals or communities can provide informed
consent for projects that are essentially uninterpretable in terms of the local
cultures (Dobrin 2008), but the enterprise is at any rate openly and often
publicly undertaken. Researchers are also increasingly prepared to share
recordings, films, or some printed outcome of this kind of work with the host
community (Terrill 2002); in some cases this is an objective of the project
from the outset.
57

Prior to the late 1990s, multiple-researcher documentation projects
were probably still few in number by comparison with more traditional
fieldwork projects in which a lone researcher went “into the field” to make a
record of a language, usually with more descriptive than documentary
objectives. Videotaping was also relatively uncommon, undertaken more at
the initiative of the individual researcher than in response to professional
expectations. Some climates and some field settings favored working out of
doors, so that the researcher’s activities were open to a great deal of public
scrutiny. In other climates and settings, however, the researcher disappeared
within some enclosed structure and conducted the research work in
considerable privacy, usually on a more or less dyadic basis. In open-air
settings much that passed between sources and researcher was audible to
others and self-censorship on the part of the local speaker(s) was probably
automatic. Self-censorship was less likely in the privacy of a home or a
sequestered workroom, and in addition the non-judgmental ear of someone
not connected to the community by blood or marriage could produce a
freedom of expression neither party originally anticipated, especially if the
work continued over a longer period of time.
The very intimacy of such sequestered fieldwork encourages a trust
that over time reduces inhibitions. This was certainly the case in some of
Milroy’s work in Belfast, where she describes the work she did with one
family in the following terms: “Much of the conversation had focused on
the disastrous and pathetic effect upon the family of the civil unrest in
58

Belfast and its function was plainly cathartic; many recordings resembled
therapy sessions more closely than sociolinguistic field tapes” (Milroy 1987:
90).
My own experience in fieldwork done prior to the 1990s was that
long-term, sequestered fieldwork produced personal and at times very
uninhibited content some of which was quite unsuitable for general-access
archiving; it would have been equally unsuitable as the basis for printed
materials that might promote revitalization, such as story books and
autobiographical sketches. The more spontaneous and lively a speaker’s
recordings were, the more unsuitable they were likely to be for such
purposes. In effect they represented a sharing of confidences on the part of
the speaker. One of my best sources produced relatively little that I could
feel free to publish as text or to archive, even with time restrictions on
archive access. Her stories were full of life, but they frequently told, with
gusto and in rich detail, of alleged misbehavior on the part of fellow-
villagers or other identifiable figures. They were superb resources for
descriptive and sociolinguistic purposes, and as a researcher I derived much
benefit from them; but it was only long after her death, with the buffer of
many intervening years, that I ventured to ask for (and did receive)
permission from her remaining family to archive a few carefully selected
portions of some of her stories. I took the same steps in the case of
problematic material recorded by two other (likewise deceased) sources. But
some material remains in my judgment too potentially offensive to archive at
59

all, even with long-term access restrictions. Memories are very long in small
villages, and families tend to remain in place for generations. Where a local
reputation is at issue, sensibilities are understandably acute.
2.2 Overcoming problems related to privacy in fieldwork
The confidentiality problems occasioned by fieldwork sessions that
take place in private locations between a single fieldworker and a single
primary source (or perhaps with some other members of the household
present as well) are somewhat less likely to arise in 21 st -century conditions
for one reason already noted, the increasing adoption of a team-based
fieldwork that lessens the frequency of sequestered, dyadic interaction
between researcher and source. In addition, the level of discussion about
ethical issues of confidentiality and informed consent has risen steadily in
workshops, conferences, and the scholarly literature, so that researchers go
into the field with greater awareness of the need to protect sources’ privacy.
A growing focus on archiving, prompted by recognition of the large number
of languages likely to pass out of regular use and of the limited time
available to record them, requires researchers to consider just what ought to
be permanently on record and to consult with communities on the matter.
Not infrequently, the community’s own interest in having their language on
record in enough detail to support revitalization efforts is the moving force
behind the documentation in the first place.
60

Even so, the issues of privacy and confidentiality remain difficult
ones. Milroy offers an example in the form of a recurrent problem she faced
in her Belfast fieldwork: a recording session might be underway, by
permission of the participants, when people who were not present at the time
permission to record was discussed arrived unexpectedly and joined in the
conversation; they might not even be aware that recording was going on
when they joined in. Yet stopping the session to re-negotiate permission to
record in each such case would have been fatal to any degree of naturalness
(Milroy 1987: 89).
Because of the elusiveness of a truly informed consent from people
to whom the methods, purposes, and products of research are bound to
remain to some extent obscure, the researcher’s best efforts may not be
enough to ensure adequate consent, as one of my own fieldwork experiences
can illustrate. In 1976 I embarked on an oral history project in one of the
three fishing communities where I had been doing fieldwork intermittently
since 1963. I conceived of this as a way of making a record of what was by
then a way of life that lingered only in the memories of the elderly and also
as a way of making some return to the community for their unstinting
generosity in sharing their language with me over the years. I approached a
husband and wife in their seventies with whom I had already done linguistic
work and asked them whether they would be willing to act as the central
sources for an oral history; they were ideal for the role because each of them
had had experience of work connected both with the local line fishing and
61

with the national herring fishery. I gave them time to discuss the project
between themselves, and when they agreed to participate I wrote an informal
agreement into one of my field notebooks, stating that they would allow me
to use the material they recorded while I would be careful to protect
identities; this we all signed. We proceeded with this work during the
summers of 1976 and 1978, and when I had a manuscript ready I posted it to
them, so as to give them an opportunity to have anything they objected to
removed. This was to be their story, after all, and I wanted it to reflect their
lives in a way that felt both accurate and acceptable to them. I was also
aware that as an outsider to the community I might not, even after what was
by then 16 years, be able to identify exactly what would or would not be
objectionable to local sensibilities. Sadly, the husband died while the
manuscript was in the mail; but at his widow’s request I removed two items
of somewhat personal information from the text. In spite of all my
precautions, however, when the book came out the widow was distressed by
two other items she had overlooked at first reading. She now wanted these
removed as well, which of course was not possible at that point, and it took
the kindly intervention of two of her children to persuade her that she was
not, as she feared, overly publicly exposed.
The flaw in my earnest efforts to make the finished oral history
completely inoffensive was that the publication process itself was unfamiliar
to my sources: it was not obvious to the surviving spouse, as I had supposed
it would be, that the manuscript phase of the book was the only stage at
62

which anything she objected to could be taken out. This was a woman
literate in English whose lifestyle in a familiar-seeming, first-world
environment was not at that point exotic or unusual in any obvious way, yet
her genuinely informed consent turned out to be much more difficult to
achieve than I had imagined. It is easy to see how much harder it would be
to achieve well-informed consent in a cultural context that differed more
radically from the researcher’s own. Clearly a good deal of responsibility
falls to the researcher when it comes to the protection of privacy.
3. The language revitalization context
3.1 Minority languages in private and public spheres
Where the vitality of a language is high and its dominance
unquestioned, use of the language will normally be broad-based. It is likely
to serve equally, for example, for intimate family life and for more public
gatherings. If the language is written, it will generally be used both for
private notes and letters and for more formal purposes as well. Resort to
other languages is typically voluntary under these circumstances.
Purely statistically speaking this profile is uncommon, since by far
the greatest number of all languages coexists with at least one other more
dominant language. The more dominant language routinely shows greater
vitality in the sense of enjoying either a larger or a socioeconomically better-
placed population base, more official support, and much wider public use as
the result of adoption by the national or regional government and its
63

educational, administrative, and judicial systems. Some small languages in
competition with a more dominant language lose ground in a pattern that
results in their being reserved for sacred purposes, in invocations, prayers,
and the like. More often, however, the pattern of retreat is to the local
neighborhood or to just the home and the kin circle, perhaps with some
special use of a more formal register persisting in oratory or religious ritual.
A small minority language still well established in private spheres
but not much used in public spheres is relatively easily documented,
assuming access to the speech community is granted. If people still use the
language in the streets or in courtyards and homes, and if they are willing to
be recorded on tape or film, then discourse in many rich forms can be
documented and preserved, for the community’s own use and for the
scientific record. But these same languages are less well placed where
revitalization is concerned, in particular if the method adopted for promoting
the language is schooling, as is increasingly common.
On the one hand speakers of a long disfavored language may be glad
of any sociopolitical developments that give their language enough
legitimacy to claim a place in the educational system. But on the other hand
a considerable problem with fit is likely to arise in the early stages of school
introduction. If the language has been used mainly in home and
neighborhood settings, transferring it from intimate settings into a formal
and public setting like the school can produce an inhibiting sense of
inappropriateness, for teachers and students alike. If the schools have a long
64

history of excluding local minority languages, children from minority-
language homes may be particularly uncomfortable about encountering their
home language in the classroom and being asked to use it in that
environment. Indigenous community members recruited as language
teachers might be expected to have no difficulty in using their languages in
the school setting, yet school use of the local indigenous language may be
problematic for them as well. In Tlaxcala, Mexico, eight individuals who
were candidates for positions as indigenous language teachers in a
revitalization initiative were observed during parts of a teacher-training
program mounted between August and December of 1999. Although the
candidates were all speakers of Mexicano (otherwise known as Nahuatl, an
indigenous Uto-Aztecan language), the observer reported that such use as
they made of Mexicano during their course participation was symbolic,
except for the one occasion when they were being tested on their indigenous-
language skills (Messing 2003: 82). Despite the avowed revitalization
purpose of the training course, the classroom was “a formal context, …
without sufficient intimacy and solidarity between speakers to warrant more
use of Mexicano” (Messing 2003: 82).
It should be noted, however, that the reverse of this development is
also known to happen: minority-language speakers who become teachers of
their indigenous language may adapt well to school use of the language
while failing to use it at home with their own children, promoting public-
sphere use but neglecting private-sphere use (Hinton 2009). Parent-child
65

transmission, usually the swiftest and most complete route to mastery of the
target language, is then replaced by purely school-based transmission, which
Fishman in his benchmark study of reversing language shift convincingly
depicts as ineffective (Fishman 1991: 368-70). Schooling is very often the
chosen locus of revitalization efforts, all the same, precisely because it
moves revitalization from the less accessible realm of personal motivation
into the public and potentially more maneuverable realm of educational
policy-making.
Once the classroom becomes the setting for revitalization efforts, the
requirements of formal teaching present certain foreseeable problems for
speakers of private-sphere forms of the minority language. In the interests of
promoting literacy and broader use, coinage and codification make their
appearance, and their adoption changes the school-promoted version of the
language into one that differs from the variety spoken locally. Coinage of
new terms is needed so that speakers can deal with topics that are either
seldom discussed in the minority language or are discussed by drawing on
many loanwords from a more widely used language. Codification is needed
in order to replace limited-currency local dialect forms with forms that will
be recognized in all regions where the minority language is spoken. The new
written-language forms are necessarily unfamiliar to children who normally
speak or hear only the local private-sphere form of the language, and if the
classroom teachers are not strictly local the pronunciations favored in school
may be unfamiliar as well. More importantly, they will be unfamiliar to the
66

children’s parents and grandparents. If local children are taught to produce
the variety promoted by the schools, the result may prove counterproductive,
at least initially, for the community. King describes, for example, the
generational divide that emerged in two Ecuadorian Quichua-speaking
communities when elderly local speakers were confronted with the school
variety their grandchildren were being taught: grandmothers did not want to
converse in Quichua with grandchildren whose speech was full of unfamiliar
lexicon (King 2001: 95). In a southeastern Welsh locality, Jones documents
support for school-promoted Standard Welsh successful enough to have
prevented even passive recognition of the original local dialect. She found
that when local children who had acquired school-taught Welsh responded to
a matched guise test, they did not so much as recognize certain long-standing
features of their own locality’s Welsh and instead identified them as features
of northern or western dialects (Jones 1998: 117). If the formal school
version of the indigenous language becomes normative in this fashion, the
local speech variety may remain essentially where it was: a private-sphere
language confined to use in the home and in certain other intimate and
solidary settings, perhaps especially among the elderly, as in the case
discussed by King. Under these circumstances the local variety is likely to
continue to lose speakers by attrition and transmission failure, the
phenomenon that usually prompted the revitalization effort in the first place.
Bridging the gap between public and private spheres of minority-
language use may depend, in long-term revitalization efforts, on the degree
67

to which proficient school-taught speakers prove willing to carry their
acquired language beyond the classroom, into some level of more general
social use, and also on the degree to which native speakers prove willing to
accept such second-language speakers into their conversational networks.
Neither development is a reliable outcome of revitalization undertakings, but
hope of such an outcome motivates many language-support movements, and
the demonstrated attainability of more modest goals (younger people with at
least some active knowledge of the heritage language and some familiarity
with its lexicon and structure, as in the case of Tolowa individuals who have
passed through the Tolowa language program in northern California; see
Collins 1998a: 264) fuels continuing community support for school-based
programs.
3.2 Native speaker status as a private-group right
Most of us have acquired some additional language or languages
through schooling without anyone ever challenging our right to learn those
languages. Native speakers may well object to our accents or the way we
handle the grammar of their languages, and some may try to avoid speaking
with foreigners who speak their languages particularly badly, but it would
not occur to them to them to warn us off trying to use their languages at all.
Languages such as French, Spanish, Russian, English, and German are
learned by legions of schoolchildren and university students in Europe and
68

the Americas, for example, as increasingly are Japanese, Chinese, and
Arabic, and the right to acquire them is taken for granted.
No such freedom of acquisition is assumable in the case of small
minority languages. Some have been spoken for generations in essentially
closed communities within which only birthright members have access to the
minority language. In some cultural contexts the language is deeply
associated with the territory in which it is spoken, and access to both land
and language is restricted, at least ideally, to members of the indigenous-
language group. This is famously the case in Australia, for example, where,
as Amery puts it, “languages are owned, in the same way that art designs are
owned by particular groups or clans”, and “senior individuals are recognized
as the owners or custodians of the language”, so that permission must
generally be obtained from them to teach the language in a formal course
(Amery 2000: 44). Even without an ownership concept as fully developed
and asserted as in the Australian context, it may still happen that the
minority language comes to be so closely associated with a particular
population that it becomes unusual, and possibly unwelcome, for others to
acquire it. At that stage would-be learners can be seen as trying to adopt an
identity that belongs by rights only to the native-born.
This issue arises somewhat counterintuitively in connection with
adult second-language acquisition of Scottish Gaelic. The number of Gaelic
speakers in Scotland has been in decline for more than a hundred years and
the language has only in the last two or three decades enjoyed any
69

governmental support to speak of. With funding for Gaelic initiatives
dependant on the continued existence of a population of Gaelic speakers and
users, one might suppose that learners would be welcomed by the native-
speaking population, but this is far from universally the case. One problem
arises from the fact that the identity “Gael” has been heavily romanticized in
a backward-looking way (Chapman 1978), so that second-language learners
may seem incongruously distant from that identity if they have no Highland
ancestry or only very long-ago and partial Highland ancestry. A second
problem arises at a more practical but not unimportant level: Gaelic learners
bring limited second-language skills to interactions with a bilingual native-
speaker population that is fully competent in English, the primary language
of most learners. Only the most patient and sympathetic of native speakers
are willing to converse with learners who speak Gaelic poorly when fluent
English is available to both parties. A third problem arises from the fact that
most learners are either English monolinguals or speakers of two major
(non-minority) languages and are accustomed to using the languages they
speak for all purposes. In keeping with that model of linguistic behavior,
they introduce Gaelic into contexts where local community members do not
normally use the language; beyond this, activist learners wish to promote the
use of Gaelic outside the traditional Gaelic-speaking parts of Scotland. For
some native speakers these are unnatural roles for Gaelic, and because they
violate local norms for Gaelic use, they seem artificial and off-putting. (All
of these problems are lucidly discussed by MacCaluim in a study of the
70

potential value of Gaelic learners to reversing language shift in Scotland;
MacCaluim 2007.)
Note that in the Scottish context, and no doubt in others, two
“private-language” obstacles can be seen to coincide: not only may native
speakers consider the language a near-exclusive birthright privilege of their
own group, but they may also have grown so accustomed to its exclusively
private-sphere use that learners’ attempts to expand its functions are viewed
as illegitimate.
3.3. Overcoming problems related to acquisition and use outside the home
If schools are the only setting in which use of the local language is
promoted among children, problems are hard to avoid. Aside from the
difficulties noted above – the absence in school of an intimacy and solidarity
otherwise associated with the language, elimination of strictly local dialect
features in the process of codification, and the introduction of unfamiliar,
newly coined lexicon – there are often some young people in the larger
minority-language ethnic community who respond poorly to the school
environment in general. Alienation arising from a long history of majority-
language hostility to the minority group is one common source of such a
response, all the more so if the minority language was firmly kept out of the
schools before a recent policy reversal.
The effectiveness of classroom instruction is often in question, too,
as in the school-based efforts to promote Ecuadorian Quichua described by
71

King (2001), where neither materials nor teaching strategies were well
enough developed to move the instructional program forward. Even where
better instruction is available, however, minority-language advocates have
often pointed out that school promotion creates problems (so much so that
Flores Farfán recommends avoiding dependence on schooling entirely in the
Nahua communities he has worked with in Mexico; Flores Farfán 2001:
191). In general, revitalization seems to proceed more effectively if any
school instruction that may be available is supplemented by culturally
appropriate activities outside school. More effective revitalization of the
Keres language, in <strong>New</strong> Mexico’s Pueblo de Cochiti, has relied, for
example, on the embedding of language learning in traditional community
practices such as visiting and community clean-up projects, so that younger
community members are brought naturally into contact with older members
who are skilled habitual speakers of Keres (Pecos and Blum-Martinez 2001).
In the Solomon Islands, according to Wurm (1999), Äyiwo, a non-
Austronesian language with an elaborate noun-class system, a complex
noun-phrase concordance, and other morphologically challenging features,
had begun to show simplifications and losses in the version of the language
spoken by young people. In response islanders undertook a gradual
revitalization process in the course of which an alphabet, a dictionary, and a
text collection were produced, and Äyiwo literacy and features of the
traditional forms of the language were introduced into some schools. But by
deliberate policy young people were also encouraged to take part in
72

traditional crafts such as carving and canoe-building, so that such linguistic
features as the mode-of-action prefixes that attached to verbs could be
demonstrated and acquired in a natural context of tool-using (Wurm
1999:171). In northern California, where since the 1920s Tolowa children
have learned English first and Tolowa only if their circumstances were
unusually favorable to its acquisition, Collins describes the school-based
language program mentioned above, underway since the 1970s, as enjoying
considerable support from the indigenous community and some success in
producing a widely shared though non-fluent knowledge of the ancestral
language (Collins 1998a: 264); but he also notes the vital importance of
traditional fishing and dancing activities in supporting a Tolowa identity and
a cultural ideal that in turn support the school language program (Collins
1998b: 178-93).
As for the inclination of some native speakers to consider use of the
ancestral language an exclusive in-group privilege, some will no doubt find
themselves unable or unwilling to depart from this position. Others, faced
with the choice between accepting second-language speakers whose
rendition of the language is “inauthentic” in various ways (including
phonological and grammatical deviations from traditional norms) or
accepting loss of the language altogether, will make their peace with the
deviations (Ó Baoill 1987: 102). How many make this second choice has
considerable potential significance for the continued oral survival of some
form of the language. In Scotland, for example, a small but growing number
73

of children from non-Gaelic-speaking homes are emerging from immersion
schooling as fluent speakers of Gaelic. Demand for Gaelic immersion
schooling has continued to rise, and if a shortage of teachers and funding can
be overcome the numbers will certainly rise further. Yet at the same time,
because of continuing transmission failure in the traditional Gaelic heartland,
the number of monolingual English-speaking children from Gaelic-speaking
homes is rising as well. Monolingual young people from traditionally
Gaelic-speaking areas are particularly inclined to resent learners, seeing
them as laying claim to a Gaelic identity that rightfully belongs to
themselves, even if they have not acquired the language (MacCaluim 2007:
96). At the moment the tide still seems to be flowing against home
transmission in the rural heartland, even while Gaelic-medium education
strengthens, especially in the cities. This is likely to exacerbate the tensions,
and the longer-term outcome is not clear.
4. The researcher’s role
The practices of researchers bear directly on the privacy-related
problems that arise in fieldwork, whereas any convictions that academic
researchers may hold about the significance of schooling and of second-
language learners for the survival of a language are unlikely to have any
bearing on how willing native speakers are to accept a school-based
language program and to welcome second-language speakers of their
language.
74

In a more general sense, however, the attitudes of researchers may
nonetheless have an impact on native speakers’ attitudes and behaviors, at
least in some cases. In 1991, at what proved to be the dawn of an era of
rising concern about endangered languages, Dixon pointed to the potential
value of documentation for the self-image of peoples whose languages were
at risk (Dixon 1991: 254). Alongside such concrete steps as implementing
bilingual education programs and developing a stock of written literature, he
mentioned “helping people to value and cherish their traditional language”
as a possibly useful support measure (Dixon 1991: 253). He was exhorting
his fellow-linguists in this case, but he credited missionary work in East
Sepik Province in Papua <strong>New</strong> Guinea with a strong revival of Urat (Dixon
1991: 246) and he stated that “any attention that is paid to a local language,
whether by linguist or missionary, is likely to enhance the speaker’s image
of that language, and of themselves, and can only have a beneficial effect”
(Dixon 1991: 247). Though it is not as hard to think of exceptions to this
generalization as one could wish, a good many researchers do subscribe to
the notion that a fieldworker who takes the trouble to learn the local
language sends a message about the worth of that language (Dorian 2001:
149; Dobrin 2008: 318), and that linguists who produce written materials at
the request of a community whose language was previously unwritten, or
very rarely written, enhance the standing of that language (Terrill 2002).
Since negative messages about the value of the receding language are
prominent among the factors that bring transmission to an end, linguists’
75

affirming actions can help to revalorize a small local language “as an
important and viable language which still has an important role to play even
in the changing urbanizing world” (Terrill, speaking of Lavukaleve, the
language of the Solomon Islands community for which she provided a
storybook and a dictionary, the latter at the community’s request, 2002:
210).
From this point of view the more traditional dyadic form of
fieldwork, at least if it is long-term and encourages the researcher to acquire
the local language, may have an advantage over team-based documentation
projects, which are both unavoidably intrusive (Thieberger and Musgrave
2007) and also expensive to mount, and for the latter reason are perhaps less
likely to be sustained over a long stretch of time. Above all, a resident
fieldworker who immediately sets out to learn the local language avoids
inadvertently modeling the advantages of language shift, whereas a team of
affluent and technologically well-equipped outsiders does the opposite if on
entering the community they are heard to use a wider-currency language
regularly among themselves.
In his 1991 article, Dixon maintained that “the work of
documentation and that of language maintenance naturally go hand in hand”
(Dixon 1991: 254). So they should, perhaps, but we have plentiful
testimony from members of indigenous communities that from their point of
view this has by no means always been the case. A 2008 conference on
“Native American Languages in Crisis: Exploring the Interface between
76

Academia, Technology and Smaller Native Language Communities” at the
University of Pennsylvania pointed to the very different agendas of
academic linguists and indigenous peoples as a long-standing problem and
brought together members of both groups to discuss that problem.
Participants hoped to turn a common interest in small-language survival into
a more mutually beneficial partnership by identifying and developing best
practices for strengthening indigenous American languages at serious risk
and by giving indigenous people more control over research and its results.
As with the problem of informed consent, this issue is now very much on the
table and will presumably receive increasing attention as we go forward. A
good example of movement forward appears in Florey’s recent account of a
two-session sequence of workshops in Indonesia, designed to create or
expand the capacity of Indonesian researchers to document Indonesian
languages (Florey 2008).
5. Looking both behind and ahead at revitalization issues
The most fundamental issues in connection with documentation and
revitalization relate to will and mobilization; they must be addressed initially
within the communities in question (Dauenhauer and Dauenhauer 1998).
Whether outside researchers may have anything to offer is among the
questions for internal discussion. Quite apart from any community doubts
about the advisability of seeking outside expertise, however, there are voices
within fieldworker ranks arguing that researchers have no business trying to
77

intervene in contexts of language shift, possess no great expertise in the
relatively applied forms of linguistics that intervention calls for, and are in
any case badly needed for the jobs of description and analysis for which they
are specially trained (Ladefoged 1992; <strong>New</strong>man 2003; Matras 2005).
Certainly communities do not always welcome researchers into their
midst. Callaghan (ms: 9) tells of being denied entry to the home of the last
speaker of Marin Miwok and Grinevald (2001: 290) of being expelled from
a community in Bolivia; the hostility in both cases arose from the potential
sources’ prior negative experiences with non-Indians. Wilson describes
anthropological fieldwork with Tsimihety (speakers of a regional dialect of
Malagasy) in Madagascar which was effectively resisted, the Tsimihety will
to “freedom from outside intrusion” extending to keeping the researcher
poorly informed, though without active hostility (Wilson 1992: 162).
Dwindling speaker communities do not necessarily see their speech form as
candidates for revitalization, furthermore. Matras reports that in the
community in which he grew up, where only the elderly spoke Yiddish, non-
transmission of Yiddish was not regarded as tragic by the people involved
(Matras 2005: 227). He also describes the Domari-speaking community he
worked in (in Jerusalem, among the Dom, or “gypsies”, whose ancestral
community language is Indo-Iranian) as one in which “ongoing language
death is accepted by the speakers and the community” (Matras 2005: 244).
Many receding languages are spoken by small and poorly placed
populations, are unwritten, and receive no official support. If the speaker
78

community has had a traditional economic base of some particular sort, it
has typically come under increasing pressure during at least the last half-
century from demographic, social, and economic changes associated with
land takeovers, resource depletion, urbanization, nation-state development,
and economic globalization. If the community has endured a degree of
stigma as the result of lesser technological development, low-income
subsistence modes, illiteracy, or perceived minority-group distinctiveness,
that stigma is likely to have remained strong over the same period in
proportion to any continuing distance from regional or national
socioeconomic norms. It is not surprising that some such communities have
little inclination to regret the loss of their ancestral language, especially if
they appear to have a reasonable chance of eventual assimilation. If their
distinctiveness was largely related to place of residence, occupation, and
income level, for example, and not to differences in physical appearance,
ancestral language loss may well shrink in importance for them when set
beside the social and economic rewards of assimilation. Even if the prospect
for improvement amounts only to some small degree of economic
advancement, an ancestral language may be abandoned with little or no
regret.
Matras does not perceive any conscious attempt among the Dom to
integrate into the surrounding Arab society; but he notes that with their
recent transition from a cohesive ethnic group with a long history as
nomadic metalworkers to a settled “clan” of poor urban wage-laborers, there
79

was no basis for cleaving to traditions (including their language) that marked
a stigmatized common group origin (Matras 2005: 242-43). Acceptance of
language loss under conditions of this sort has sometimes been characterized
as “language suicide”, since it appears that no resistance is mounted to the
loss (Dennison 1977: 16). In view of the lengthy period of stigmatization and
discrimination that precedes the apparent acquiescence, however, the
“suicide” terminology seems to blame the wrong party, the victims rather
than the perpetrators. Such is my view, at any rate, after several decades of
work with a speaker population whose members, in the wake of generations
of stigma associated first with their subsistence mode and subsequently also
with their language, likewise expressed no desire for the revitalization of
their local speech form (Dorian 1987).
Matras takes strong exception to “salvation linguists”, as he calls
them, who see themselves as stepping in to “save” an endangered language.
He points out that communities do not take a single attitude or speak with a
single voice, and that not every community wants its language rescued by a
linguist (Matras 2005: 227). Yet it is not uncommon today for speakers of a
receding language to take the initiative themselves, asking to have a linguist
study and record their language (Nagy 2000; Grinevald 2005) or hiring a
linguist to help in creating or refining a writing system, putting together a
dictionary, or preparing teaching materials, as do some indigenous
Australian and North American groups (Wilkins 1992; Debenport 2009).
Hinton describes vividly the yearning for recovery of ancestral languages
80

that brought leaders of the Native California Network together with linguists
in 1992 to find effective ways to preserve or restore their languages (Hinton
1994: 221-22). The success of the Indonesian workshops mounted by Florey
and her colleagues likewise indicates that local speakers wish both for
outside expertise and for greater expertise of their own in furthering
documentation and revitalization.
Dixon, in his 1991 paper, pointed out that communities often remain
unaware of the risk to their language until it has grown too late to change the
situation (Dixon 1991: 231). This seems an odd observation initially, but it is
not so counterintuitive as it seems. Fluent speakers who have reached middle
age still have available a generation older than themselves who regularly
speak their language. They themselves make ample use of the language, and
they may simply fail to register the degree to which young people and
children are using some more widely spoken language instead. As Kulick
demonstrated for the Papua <strong>New</strong> Guinean village where he worked, they
may also not register the degree to which they themselves are failing to use
the local language with their children (Kulick 1992). It is only as they
become the older generation themselves that some local-language speakers
look around and realize that there are no speakers coming along behind
them: they are the last remaining speakers, and unless heroic measures are
taken their language will disappear with them. At this point the attention of
specialist outsiders may rather suddenly be seen as useful where it was not
before.
81

Grinevald (2006) has pointed out that interventionist agendas
developed among linguists in very particular contexts, namely those in
which patterns of language loss had already reached extreme proportions:
North and South America and Australia. While in Europe some regional
languages were acknowledged to be used less than others – hence the
terminology EBLUL, European Bureau of Lesser Used Languages, for the
body that represents them -- and in South Asia and Africa some languages
were recognized as spoken only by certain tribal groups in a context of
widespread multilingualism, in the Americas and Australia the context was
massive indigenous language loss, already far advanced and in prospect even
more severe. In America and Australia, where dwindling numbers of
indigenous-language speakers confronted this prospect, indigenous
communities were voicing acute concerns about their languages and
beginning to agitate for revitalization; the Americanist and Australianist
linguists with whom these matters were raised were among the first to
express a sense of professional obligation to intervene on behalf of small and
receding languages (Grinevald 2006: 340-41). It was no accident that the
linguists and activists who contributed to the 1992 issue of the journal
Language that famously raised the issue of language endangerment within
the U.S. professional community all worked with American or Australian
languages (Hale, Krauss, Watahomigie, Yamamoto, Masayesva Jeanne, and
England 1992).
82

Linguists have long had their own purely professional reasons for
wishing to document and describe receding languages (more typically to
describe than to document, although Himmelmann 2008 argues powerfully
for the greater value of documentation). Matras articulates the chief such
reason at the end of the article cited above when he speaks of “the urgent
task of securing a diverse linguistic sample corpus for the sake of future
generations of students of language” (Matras 2005: 248). But Woodbury
represents a growing voice in the profession when he writes that “it is
becoming less and less viable for linguists to think of the stakeholders in
language documentation to be constituted only of a vaguely-conceived
scientific posterity” (Woodbury 2003: 39), and the same may be said where
language description and analysis are concerned.
The unhappy fact is that whatever linguistic professionals do, it will
be inadequate. The complexity and richness of language and its cultural
context are such that they escape all our efforts to capture them. Our
descriptions will prove to be more incomplete, our analyses more imperfect,
and our documentations more limited than we imagine. Twenty years from
now a new generation of linguistic professionals will wonder how we could
have failed to raise the questions that interest them most or document the
kinds of linguistic behavior that have come to preoccupy them. Whether we
scatter ourselves across the globe in what promises to be a last-minute
attempt to record more receding languages, or embed ourselves deeply in
particular small-language contexts in an attempt to record and understand
83

them more fully, much will elude us. But that is no reason for not trying.
The forces arrayed against the survival of small languages are formidable,
and our efforts to provide support for them are likely to prove inadequate.
But that is also no reason for not trying.
References
Amery, Rob. 2000. Warrabarna Kaurna! Reclaiming an Australian
language. Lisse: Swets and Zeitlinger.
Austin, Peter K. and Grenoble, Lenore A. 2007. Current trends in language
documentation. In Language documentation and description 4, P. K.
Austin (ed.), 12-25. London: The Hans Rausing Endangered
Languages Project.
Callaghan, Catherine A. Unpublished manuscript. Why work with the last
living speaker of a language?
Chapman, Malcolm. 1978. The Gaelic vision in Scottish culture. London:
Croom Helm.
Collins, James. 1998a. “Their ideology and ours”. In Language ideologies:
Practice and theory, B. B. Schieffelin, K. A. Woolard, and P. V.
Kroskrity (eds), 256-270. <strong>New</strong> York: Oxford University Press.
Collins, James. 1998b. Understanding Tolowa histories: Western
hegemonies and Native American responses. <strong>New</strong> York: Routledge.
Csató, Eva A. and Nathan, David. 2003. “Multimedia and documentation of
84
endangered languages”. In Language Documentation and description

1, P. K. Austin (ed.), 73-84. London: Hans Rausing Endangered
Languages Project.
Dauenhauer, Nora Marks, and Dauenhauer, Richard. 1998. “Technical,
emotional, and ideological issues in reversing language shift:
Examples from Southeast Alaska”. In Endangered languages:
Language loss and community response, L. A. Grenoble and L. J.
Whaley (eds), 57-98. Cambridge: Cambridge University Press.
Debenport, Erin. 2009. “Listen so you can live life the way it’s supposed to
be lived”: Paradoxes of text, secrecy and language at a <strong>New</strong> Mexico
pueblo. PhD dissertation, University of Chicago.
Dennison, Norman. 1977. “Language death or language suicide?”.
International Journal of the Sociology of Language 12: 13-22.
Dixon, R. M.W. 1991. “The endangered languages of Australia, Indonesia,
and Oceania”. In Endangered languages, R. H. Robins and E. M.
Uhlenbeck (eds), 229-55. Oxford: Berg.
Dobrin, Lise. 2008. “From linguistic elicitation to eliciting the linguist:
Lessons in community empowerment from Melanesia”. Language
84(2): 300-24.
Dorian, Nancy. 1987. “The value of language-maintenance efforts which are
unlikely to succeed”. International Journal of the Sociology of
Language 68: 57-67.
85

Dorian, Nancy. 2001. “Surprises in Sutherland: Linguistic variability amidst
social uniformity”. In Linguistic fieldwork, P.<strong>New</strong>man and M. Ratliff
(eds), 133-51. Cambridge: Cambridge University Press.
Fishman, Joshua A. 1991. Reversing language shift. Clevedon: Multilingual
Matters.
Flores Farfán, José Antonio. 2001. “Culture and language revitalization,
maintenance, and development in Mexico: The Nahua Alto Balsas
communities”. International Journal of the Sociology of Language
152:185-197.
Florey, Margaret. 2008. “Language activism and the ‘new linguistics’:
expanding opportunities for documenting endangered languages in
Indonesia”. In Language documentation and description 5, P. K.
Austin (ed.), 120-135. London: Hans Rausing Endangered
Languages Project.
Grinevald, Colette. 2001. “Encounters at the brink: Linguistic fieldwork
among speakers of endangered languages”. In Lectures on
endangered languages: 2, O. Sakiyama and F. Endo (eds), 285-313.
Kyoto: Endangered Languages of the Pacific Rim.
Grinevald, Colette. 2005. “Why the Tiger Language and not Rama Cay
Creole? Language revitalization made harder”. In Language
documentation and description 3, P. K. Austin (ed.), 196-24.
London: Hans Rausing Endangered Languages Project.
86

Grinevald, Colette. 2006. “Worrying about ethics and wondering about
‘informed consent’: fieldwork from an Americanist perspective”. In
Lesser-known languages of South Asia Anju Saxena and Lars Borin
(eds), 339-370. Berlin: Mouton de Gruyter.
Hale, Ken, Krauss, Michael, Watahomigie, Lucille J., Yamamoto, Akira Y.,
Craig, Colette, Masayesva Jeanne, LaVerne, and England, Nora C.
1992. Endangered languages. Language 68(1): 1-42.
Himmelmann, Nikolaus. 2008. “Reproduction and preservation of linguistic
knowledge: Linguistics’ response to language endangerment”.
Annual Review of Anthropology 37: 337-350.
Hinton, Leanne. 1994. Flutes of Fire: Essays on California Indian
languages. Berkeley: Heyday Books.
Hinton, Leanne. 2009. “Language revitalization at home”. Paper delivered
at the First International Conference on Language Documentation
and Conservation, Manoa, Hawaii, March 14, 2009. Accessible at
http:/scholarspace.manoa.hawaii.edu/handle/10125/5961.
Jones, Mari C. 1998. Language obsolescence and revitalization: Linguistic
change in two sociolinguistically contrasting Welsh communities.
Oxford: Oxford University Press.
King, Kendall A. 2001. Language revitalization processes and prospects:
Quichua in the Ecuadorian Andes. Clevedon: Multilingual Matters.
87

Kulick, Don. 1992. Language shift and cultural reproduction: Socialization,
self, and syncretism in a Papua <strong>New</strong> Guinean village. Cambridge:
Cambridge University Press.
Ladefoged, Peter. 1992. Discussion note. Another view of endangered
languages. Language, 68: 809-811.
MacCaluim, Alasdair. 2007. Reversing language shift: The social identity
and role of adult learners of Scottish Gaelic. Belfast: Cló Ollscoil na
Banríona.
Matras, Yaron. 2005. “Language contact, language endangerment, and the
role of the ‘salvation linguist’”. In Language documentation and
description 3, P. K. Austin (ed.), 225-251. London: Hans Rausing
Endangered Languages Project.
Messing, Jacqueline H. E. 2003. Ideological multiplicity in discourse:
Language shift and Bilingual schooling in Tlaxcala, Mexico. PhD
dissertation, University of Arizona.
Milroy, Lesley. 1987. Observing and analyzing natural language. Oxford:
Basil Blackwell.
Nagy, Naomi. 2000. “What I didn’t know about working in an endangered
language community: Some fieldwork issues”. International Journal
of the Sociology of Language 44: 143-60.
<strong>New</strong>man, Paul. 2003. “The endangered languages issue as a hopeless cause”.
88
In Language death and language maintenance: Theoretical,

practical and descriptive approaches, M. Janse and S. Tol (eds), 1-
13. Amsterdam: John Benjamins.
Ó Baoill, Dónall P. 1987. “Phonological borrowing in Irish and problems of
orthographical representation”. In Third International Conference on
Minority Languages: Celtic Papers, G. Mac Eoin, A. Ahlqvist, and
D. Ó hAodha (eds), 89-103. Clevedon: Multilingual Matters.
Pecos, Regis, and Blum-Martinez, Rebecca. 2001. “The key to cultural
survival: Language planning and revitalization in the Pueblo de
Cochiti”. In The green book of language revitalization in practice, L.
Hinton and K. Hale (eds), 75-82. San Diego: Academic Press.
Terrill, Angela. 2002. “Why write books for people who don’t read? A
perspective on documentation of an endangered language from
Solomon Islands”. International Journal of the Sociology of
Language 155/56: 205-219.
Thieberger, Nick and Musgrave, Simon. 2007. “Documentary linguistics and
ethical issues”. In Language Documentation and description 4, P. K.
Austin (ed.), 26-36. London: The Hans Rausing Endangered
Languages Project.
Wilkins, David. 1992. “Linguistic research under Aboriginal control: A
personal account of fieldwork in Central Australia”. Australian
Journal of Linguistics 12(1): 171-200.
Wilson, Peter J. 1992. Freedom by a hair’s breadth: Tsimihety in
Madagascar. Ann Arbor: University of Michigan Press.
89

Wittenburg, Peter. 2003. “The DOBES model of language documentation”.
In Language documentation and description 1, P. K. Austin (ed),
122-139. London: Hans Rausing Endangered Languages Project.
Woodbury, Anthony. 2003. “Defining documentary linguistics”. In
Language documentation and description 1, P. K. Austin (ed.), 35-
51. London: Hans Rausing Endangered Languages Project.
Wurm, Stephen A. 1999. “Language revivalism and revitalization in Pacific
and Asian areas”. International Journal of the Sociology of Language
137: 163-72.
90

Bridging linguistic research and linguistic documentation: the Kuikuro
Abstract
experience (Brazil)
Bruna Franchetto
This article aims at discussing the relation between linguistic research and
documentation projects based on a long term field experience among the
Kuikuro, a Carib speaking people living in Southern Amazonia (Brazil).
Kuikuro is an endangered language, spoken by about 600 individuals, who
suffer the encroachment of Brazilian national society, whose dominant
language is Portuguese. Whereas the linguistic documentation project started
in 2000, with the additional proposal of recording ethnographic contexts of
language use, the linguistic research began much earlier. The Kuikuro
expectations and representations on research and documentation are
described as well as their cultural presuppositions and consequences. Three
moments characterize, along almost thirty years, the development of the
relations between the researcher(s) and the native people: first, a suspicious
reception; then, the domestication of the outsider; finally, the documentation
project was assumed by the Kuikuro themselves, who never gave up their
political and cultural agentivity.
91

Keywords: Amazonian languages, Carib, Upper Xingu, Linguistic
Documentation, Endangered Languages, DOBES Program.
1. Introduction 10
Aurore Monod Becquelin, Emmanuel de Vienne and Raquel Guirardello-
Damian, members of one of the first teams of the DOBES Program
(Documentation of Endangered Languages), presented a very interesting
paper at the DOBES International Conference, held in Frankfurt in
September 2004, just published in the book Lessons from Documented
Endangered Languages (Harrison, Rood and Dwyer 2008). In this article,
entitled ‘Working together: The interface between researchers and the native
people – The Trumai case’, the authors propose to analyze and comprehend
their experience and the limits of the concept of documentation among the
Trumai. The Trumai and Kuikuro 11 people live in the same region of
southern Amazonia (Brazil), known as the Upper Xingu, forming part of the
10 A first version of this article was presented at the DOBES workshop held in June 2006 at
the Max Planck Institute of Psycholinguistics (Nijmegen). I acknowledge the DOBES
Program, whose support made possible the documentation of the Upper Xingu Carib
language (Kuikuro dialect) from 2001 to 2005, as well as the CNPq (Brazilian Counsel for
Scientific and Technological Development). I acknowledge also the Kuikuro community of
the Ipatse village and the other members of the project: Mara Santos, Carlos Fausto,
Michael Heckenberger, Afukaká Kuikuro, Mutuá and Jamalui Mehinaku, and the Kuikuro
video-makers. The author works among the Kuikuro, doing linguistic and ethnographic
research, since 1976.
11 The Kuikuro are all native speakers of one of the two variants of the Upper Xingu Carib
language, one of the two southern branches of the Carib family (Meira and Franchetto
2005); according to the last census (2006), the Kuikuro population numbers about 600
people, living in four villages. Trumai is an isolated language spoken, today, at different
degrees of fluency, by no more than 50 individuals in a population of 120.
92

same native sociocultural, multiethnic and plurilingual system, albeit to
different degrees and with distinct configurations of linguistic endangerment.
The aim of this paper is to try to respond to some of the same
questions posed by Becquelin, de Vienne and Guirardello-Damian at the
start of their article, explored through the complementary lens of myself and
the Kuikuro: How can language and culture, conceived as human
interactions, be reduced to objects, facts and texts? For whom are they
preserved? How does the native community experience what they perceive
as the loss of their tradition? What is their understanding of our purposes and
aims in documenting their language? How may our work change the native
way of life?
Firstly, I shall talk about my experience among the Kuikuro prior to
the arrival of the DOBES project and after its implementation, in order to
show that the Kuikuro never were, in fact, objects or passive recipients:
rather, they have always been ‘agents’ (possessing their own agency) in their
relations with the researcher. Secondly, I shall discuss how the DOBES
project, which started in 2001, has almost inevitably ended up being
appropriated by the Kuikuro themselves, adapting the project to their own
desires and strategies, especially over the last two years (2004 and 2005).
This does not mean all of the impasses and problems pointed out by our
‘Trumai’ colleagues have been solved. I refer in particular to two dilemmas:
(i) a hiatus exists between our conception of ‘culture’ and ‘knowledge’ and
that of the Kuikuro and Trumai; documentation, our own invention, seems to
93

entrench a ‘folklorization of the culture’ already set in motion by a long
history of relations between Whites and Indians in the Upper Xingu; (ii) yet
“if we do not record this material, the linguistic and cultural knowledge of a
unique system may disappear without trace.” (Bequelin et al, 2008: 65).
Our work with the Kuikuro is a practical realization of the ideas
proposed by Bequelin, de Vienne and Guirardello-Damian at the end of their
article as a potential solution to this dilemma. This approach involves: (i) not
reinforcing the dichotomy between tradition and modernity; (ii) recording
not only the traditional practices, but also the changes taking place today;
(iii) enabling the Kuikuro to make use of new recording technologies and
equipment to record both their own world and the non-indigenous world. It
is still early for any in-depth analysis of the results of this ‘adventure,’ but
the work in progress is extremely exciting and it seems to be proving ‘worth
the effort.’
This experience shows that the success of a documentation project
largely depends on the capacity to integrate anthropological sensibility and
ethnographic knowledge with the real, multiple and often contradictory
expectations of the ‘object’ population, at the same time involving and
supporting the local users and producers of the documentation, most of the
time young people caught in a precarious equilibrium between the ‘old’ and
the ‘new’. These young people from the small communities speaking
minority, besieged languages are the living and reactive laboratories whose
94

experiments may determine the eventual preservation of a language and the
culture transmitted by it, or their demise.
2. Before DOBES
In 1981 I was in Ipatse village. During the ‘festival’ called kwambü,
held in November that year, a woman composed a song allowing me to take
part in the ritual – in fact, compelling me to participate. Kwambü is the name
of an itseke (Spirit-Beast), a mask and a joyful and noisy ‘festival,’ where
composers and signers publicly create and transmit individual
‘messages/songs’ called tolo (bird, pet). These songs are commentaries,
gossip and accusations – or responses to accusations and admonitions
(Franchetto 1997, 2001). Like other participants in the ‘festival,’ I had to
wear a costume: in my case, transform into a caricature of myself. I dressed
up as a ‘researcher,’ sporting a large safari hat, a photographic camera and
tape recorder draped around my neck, a blanket wrapped around my waist,
and a large rucksack on my shoulders. I went singing from house to house
for hours on end amid peals of laughter from my hosts. The song imposed on
me denounced my servitude to the schemes of the Whites and, at the same
time, announced the start of my ‘emancipation’ (Franchetto 1986). This was
the song: 12
12 The Kuikuro data are transcribed using the current phonemic orthography established by
the Kuikuro teachers and by ourselves, the linguists. The correspondences between written
symbols and the sound they represent (when not obvious) are as follows: ü (high central
unrounded vowel), j (palatal voiced consonant), g (uvular flap), ng (velar nasal), nh (palatal
nasal), nkg (pre-nasalized voiced velar nasal). The abbreviations for glosses in the
interlinear morpheme-to-morpheme translation are: 1 first person, 12 first inclusive dual, 2
95

e-enge-lü-ko-i u-e-nhügü
2-eat-PNCT-PL-COP 1-come-PNCT
‘to eat/frighten/kill you I came’
kuk-imbuki-ne-tagü-ko
12-spy-TR-CONT-PL
‘(she) is spying on all of us’
akatsige
really
Buguna heke ige-i
Bruna ERG DPROX-COP
‘it is Bruna’
engelükoi uenhügü
‘to eat/frighten/kill you I came’
A brief analysis of the content/message of the song serves to unveil
the perception which the Kuikuro at that time had of me and our
relationship. It also helps reveal what has and has not changed since then.
second person, CONT continuative aspect, COP copula, DPROX deíctic.proximity, ERG
ergative, PL plural, PNCT punctual aspect, REL relator, TR transitivizer.
96

In the song, I, Buguna, presented myself as someone who enge,
eats/frightens/kills. This is what the itseke are like: I was an itseke, one of the
most dangerous kind, the kuk-enge-ni (12-eat-AGNR), ‘the eaters.’
Mutuá, a Kuikuro teacher, defined the term itseke for a dictionary
being compiled for the Kuikuro Archive in the following way:
itseke ekisei kukengeni, kugehüngüha ekisei, inhalüha ingilüi; itseke
kukilüha ngiko heke kukengeni heke; itseke ekisei kukotombani
kukügünu hata.
Itseke is the one who eats us, s/he isn’t a person (kuge), s/he can’t be
seen. Itseke is the one that harms us with invisible arrows when we’re
sick.
Itseke, says Mutuá, is not kuge and is defined by its predatory
impulse and invisibility.
Itseke designates a specific and distinct ontological category, and
indexes a mode of existence different from the ordinary. In this sense, the
term is similar to our notion of ‘spirit,’ though it does not designate the
immaterial part of a person constituted by the body/soul duality. The
Kuikuro usually translate itseke as ‘bicho’, ´beast/animal,’ a translation with
some interesting implications. The itseke populating the Xinguano world are
mostly animals in their condition as persons, endowed with intentionality,
reflexivity and a communicative capacity. They are omnipresent in mythic
97

time, but it is neither easy nor good to see itseke in ‘ordinary life’ (with the
exception of shamans), since we only see them when we are sick or on the
verge of dying.
The itseke are also kuge, insofar as they are persons who possess their
own subjective perspectives, and manifest themselves to Xinguano shamans
in human form. They are set apart by being particularly powerful, elusive
and aggressive.
In sum, the ontological category itseke indicates an extraordinary
context. In historical narratives or everyday events, however, the category is
used as an explanatory shortcut for anything strange or incomprehensible.
The Xinguano mode of dealing with human or non-human aggressive
alterity is to try to tame it through the ties of ritual exchange. Historically,
this was how the pluri-ethnic and multilingual Upper Xingu system took
shape. The social device that made the Upper Xingu cultural amalgam
possible was ‘relational interweaving;’ in other words, the long-term
development of cordial relations through visits, presents and marriages,
produces a network of identities of higher density than the network of
differences, especially linguistic differences (Fausto, Franchetto and
Heckenberger 2008). Among other things, the Xinguano rites are designed to
tame the non-human other, itseke. The entities represented in this kind of
ritual – through songs, choreography, masks, flutes – are itseke. These rituals
promote the conversion of a dual relation of predation into a collective act of
reciprocity. This was the aim of my participation in the kwambü of 1981 and
98

of my song: to tame the kagaiha/itseke, as the Upper Xinguanos have always
done, taming Indians and Whites and luring them into their network of
endless reciprocity.
In first getting to know the Kuikuro, I had to overcome my surprise
(and frustration) on discovering that they were not among the survivors of a
romantic ‘golden age of primitive communism’.
The terms oto, ‘owner’, and ihipügü, ‘payment’ denote core concepts,
institutions, values and practices in the Upper Xingu, even today. Gertrude
Dole (1958: 126) had already observed back in the 1950:
The Kuikuro are constantly preoccupied with the exchange of goods
and services and with equalizing accounts. This theme permeates the
whole culture. One of the expression most frequently heard is “Tüma
ihipügüi?”, which means “What’s the price” and which may be used
in various contexts, being equivalent to “What did you receive”, or
“What did you have to pay”... The desire to acquire material goods
appears to antedate the first recorded contact with Whites, but native
acquisitiveness has also been fostered by contact with civilization.
A man pays his parents-in-law for his wife’s vagina (bride wealth);
other payments are made to learn formal verbal genres, to acquire a name, to
win over a lover, to marry and to obtain secret information. An ihipügü is
needed to obtain services from a wide variety of specialists, including:
99

100
killers, those who ritually wash the dead person’s kin at the end of the
mourning period, people who paint those in reclusion when they make their
public reappearance during the large-scale ‘festivals,’ people who help in
swidden work or house building, and so on. ‘Paying’ for hospitality in the
village is expected and indeed forms the pre-condition for being made
welcome. Almost everything has a ‘price’ (ihipügü) and this aspect regulates
every kind of exchange, a fact evident in the ulukí, a ritualized ‘market’ held
frequently throughout the year.
Practically nothing exists that could be described as ‘collective’
property, lacking an ‘owner,’ oto, not even the school. A relation of identity
exists between a thing and its owner. Having an ‘owner’ makes something
exist socially, for someone in relation to someone else.
‘Presents’ are an institution of the relations with Whites, inaugurated
by Karl Von den Steinen, the German ethnographer considered the first
White to arrive in the region “in peace, without killing anymore,” at the end
of the 19 th century. The older Kuikuro say: “We saw White things and our
hearts were struck, women saw them and their hearts were struck, from the
very first time.” ‘Struck the heart’ of the victims of the itseke, beings who, as
we have seen, can cause sickness and death; the fish ingested by a
menstruating woman – a breach of an alimentary rule – “strikes her heart.”
Ever since my first arrival in the Kuikuro village, custom dictated I take the
‘presents’ to the middle of the village in front of the ‘men’s house’ for them

101
to be displayed, evaluated and distributed while I waited in an attitude of
‘shame/respect.’
The Kuikuro say that Whites don’t like to exchange, make payments
or be inundated with demands, but do like to give and receive ‘presents.’
This, though, was a rhetorical mask for the imperatives of reciprocity in
response to the criticisms made by Whites in response to the exaggerated
demands of the Upper Xinguanos and a generic condemnation of the
introduction of money by employees of the Brazilian State charged with
‘protecting and assisting’ the Indians.
Excluded from the prestations and counter-prestations that cement
ties of kinship and alliance, the researcher is, however, from the outset
involved in the native system of exchanges and payments. And he or she
represents an important source of resources. The individual ‘payment’ of
‘informants’ and specialists took place almost invariably in secret, while the
disputes over the resources provided by the researcher occasionally led to
highly tense situations. Privileges mean some individuals are potential
victims of jealousy, ‘gossip’ and accusations. Various levels of mediation
were needed for me to be able to formulate my requests to the oto, ‘owners,’
of different genres of verbal arts and negotiate the demanded ihipügü. After
all, they would say, in the White world even food has to be paid for and
money is an omnipresent and omnipotent measure. All this seemed to me to
raise the ‘price’ above internal levels and forced me into the tortuous use of
tactics of persuasion.

102
In the 1980s, money began to make its presence felt in the Upper
Xingu, with no prospect of a return to the old times and irrevocably altering
the ihipügü system of equations of value between exchanged objects; it
began to function as a kind of table of ‘tribal rates,’ regulating the access of
researchers, journalists and photographers, who were all lumped into the
same category.
On arriving in the village for the first time, it appeared that
everything had already been decided. I quickly realized that I had no
freedom of choice, neither over possible ‘informants,’ nor over my own
identity vis-à-vis the Indians. A set of relations and meanings would
condition my steps, leaving me narrow room for manoeuvre. Hence, for a
long time, I was considered “an anthropologist who wants to hear stories;”
the figure of the linguist was a novelty that only gradually took shape. Over
the years I recorded dozens and dozens of akinhá, narratives, since
‘everything has a story,’ and narration is a fairly open-ended genre.
However, it didn’t take long for the akinhá oto (owners of the art of
narration) to discover that they could ask for payment for their knowledge
from Whites. After all, don’t Whites fill books with stories and histories?
And books are merchandise that are bought and sold.
Although in the past the ‘outsider’ was potentially suspect, since the
Kuikuro generally supposed that anyone who abandons their place of origin
for another has been accused of sorcery, this position had already became an
institutionalized resource by the time I first arrived in the village. The

103
Indians also know we live in their villages in apparent temporary poverty.
All of us Whites return ‘to our place’ where, it is presumed, there is a
fantastic quantity and diversity of people and things.
Present giving, requesting, paying and exchanging –life in society–
formed the modalities through which the Kuikuro absorbed me into relations
of reciprocity and dependency. The famous Xinguano art of involvement is a
political skill involving a mix of diplomacy and manipulation – an ethos
which spins webs, casts nets and snares the domestication of the other. It is a
power game, but one which is decentralized, diffuse and networked.
Let’s return to the kwambü song. As well as being an itseke, I, in the
song, was also a ‘spy.’ The meaning of the root imbukine is to “send a
woman to sleep with the enemy.” I was a woman sent by the Whites to spy
on the Kuikuro and bring home everything I had learnt, written and
recorded, to deliver the ‘stolen’ treasure to my similars, the enemies.
Twenty five years ago, the technology at my disposal were
notebooks, pens, a tape recorder and a photographic camera (a cheap kind).
The owners’ (oto) proved to be reluctant to pass on their knowledge
to someone ‘from outside.’ Sheets of paper and tapes were in my possession
and would circulate in the outside world beyond their control. The Kuikuro
were not yet familiar with tape recorders, but by the end of the 1980s there
were already five recorders in the village. They entered the circuit of
exchanges and payments – consumed until they became useless. Stripped
down, their parts served as body decorations, utensils, toys and rubbish.

104
The tape decks and batteries introduced by researchers were among
the first requests made by the Indians. While working at least, the tape
recorder was, and continues to be, used to record songs and musical
performances in the intra and intertribal ‘festivals,’ but never narratives or
formal speech. The tapes circulated through the houses and villages,
borrowed, exchanged and given. The performances of singers and ritual
specialists were publicly evaluated. Tapes were recorded, blanked an re-
recorded until worn out; they were never kept as mnemonic records or
documents. Although crystallizing a performance, making it indefinitely
repeatable, the forms of recording developed by the Kuikuro still allowed
variations which approximated orality and separated it from writing, the
indelible reification of a performance.
These recording and writing technologies meant that I, an exceptional
student, came to pose a threat. Transcription allowed me to reproduce a
performance in a very short amount of time, dispensing with the process of
memorization that successive ihipügü. On the other hand, this mechanical
performance made me an extremely quick-learning performer, but totally
incompetent in terms of mastering the verbal arts. Serious conflicts surfaced
with the recording of certain songs defined as expensive or prohibited. My
proposals of ‘payment’ seduced a few singers willing to risk flouting rules
and hierarchies. Tensions and rumours erupted into open accusations, a
theme of male reproofs in public speeches and the cause of temporary
outbursts of factional infighting. Women, finally, sought me out

105
clandestinely at night, asking me to ensure the tapes were stored as carefully
as possible and not shown to anyone. I had to promise in solemn, complicit
whispers.
Recording informal, everyday conversations was strictly prohibited,
since it could publicly expose ‘gossip,’ or private politics in other words.
The representation of speech as the voice, part of the body, an extension of
the person, was an impediment to using the tape recorder freely. Older
people said I wanted to ‘steal’ their voice, separating it from them. They
claimed they were close to death and that were their dead voice – a living
fragment – heard by their children, the latter would be filled with otonu, a
‘longing’ or ‘nostalgia’ which comprises the suffering of the living and an
evocation of the akunga (soul) already on its journey to the village of the
dead – a return exorcized by mourning rites. Under these conditions,
recording was never easy and only took place after insistent requests,
explanations, the intervention of mediators and the final approval.
Recordings, all of them, had a ‘real’ price to be paid with goods I had on me
or which I could buy in town.
Reaction to the tape recorder was very similar to that caused by the
camera. Roland Barthes (1984) wrote that photography is the contemplation
of death in a spectrum/ghost, a micro-experience of death.
I continued to return to the Kuikuro, year after year, and I learned
time can change everything. I became an eternal apprentice. From being an
androgynous, infantile object of distrust – the Whites ‘steal’ land, lives,

106
narratives, words, names – I became a familiar person. The song which the
Kuikuro made me sing in the Kwampü festival, amid general laughter, in
November 1981, marked a passage or catharsis, my transformation from an
itseke into almost-kuge, an almost-person. The Kuikuro describe this process
as a ‘becoming kuge,’ ukugetilü, a verb which designates the process of
adopting Xinguano way of being, as well the process through which a
previously wild animal accepts its condition as a pet and becomes tame.
Making other laugh establishes reconciliation, the return home, as episodes
from various mythic narratives relate.
Now I was no longer like other Whites, who “arrive, go away and
never return, not even sending news,” I acquired another ambiguity, this
time positive: I spoke Kuikuro sufficiently well to be welcome as kin, I
became old (an older woman no longer has blood, nor shame), and I turned
into an itankgo (woman-chief); the Kuikuro filled me with pride and bound
me even further to obligations of reciprocity (a woman-chief can never say
no).
3. A.D.: After DOBES
The DOBES documentation project arrived in 2001 and was received
as tisakisü ongitegoho (‘made for keep our words’). The worst was over, we
could say; it was finally possible to test out ‘working with’ the Indians,
especially since there was no other alternative for executing a Project among
the Kuikuro. From the 1980s to 2001, White things invaded the village:

107
clothing, utensils, school, TV satellite dishes and television sets, CD and
DVD players, books, motorbikes, cars, Portuguese. The challenges
(familiarizing the Indians with new technologies and new objects, gradually
intensifying the acts of recording and writing) could be reconfigured, but the
underlying doubts remained and other challenges arose. ‘Working with’
meant trying to forge a real partnership, effect an exchange in the realm of
the intangible and unknown. I say ‘try’ since at no point were we certain that
we could free ourselves of an equal relationship, latent distrust, and silent
calculations of interest on both sides. We have tried, though, to be sincere,
open, humble, solidary and above all honest, while the Kuikuro ‘repay’ us
with honesty, friendship and above all a continuous effort to maintain and
renew mutual trust.
The letter which the young Kuikuro, our permanent consultants,
wrote, printed and posted on the wall of the Project’s office in Rio de Janeiro
says it all: BIG BROTHER KUIKURO, with the signatures of each of them
added on each ‘work trip’. Fine-tuned irony which reveals the continuity and
distance of old representations at one and the same time. Spending days or
months in Rio de Janeiro in an apartment working on transcriptions and
translations, revising the lexicon and grammar, at an intense pace is like (as a
metaphor) being imprisoned in a reality show, experiencing suffering and
reclusion with the prospect of some kind of gain, palpable in terms of
money, less palpable in terms of the goal of ‘safeguarding our entire
language’ for the future. Furthermore, a reality show exists thanks to new

108
recording and communication technologies, but above all because of
countless spies.
Time can change everything. In 2001, two Kuikuro enrolled on an
undergraduate course: their degree ceremony was in June 2006. One of them
discovered linguistics, working with us. He fell in love with the
morphological surgery of a highly agglutinative language, the intellectual
pleasure of research; now he is a master student in anthropological
linguistics. As he says: ‘thinking in and on the language, the documentation
project rescued me from darkness’.
This diffuse talent for linguistics can in part be explained by living in a
multilingual system, which traverses day-to-day life and produces a
sophisticated metalinguisitic reflexivity and discourse among the Upper
Xinguanos (Franchetto 2003). The Project has taken enormous advantage of
this and, at the same time, propelled this characteristic to develop in a
direction we had never imagined. For people living in a society with many
languages, including, today, Portuguese, the lexicon is an object provoking
intense interest and translation is a fascinating process. We turned
translation, in its various levels and with all its potential pitfalls, into a work
in partnership which, although slow, had enriched everyone and enabled the
input of linguistic and ethnographic data that we would probably have been
unable to obtain otherwise.
What I said about the tape recorder, in the past, applies today to the
video camera: the same distrust when the camera is in the hands of Whites.

109
With an aggravating factor: video is voice and photo in movement, a
frightening reproduction of the seen and the heard. On the other hand, the
younger Kuikuro were increasingly focused on the outside, the world of the
cities and their ‘marvels,’ turning their backs on village life, its rhythms and
its ‘poverty.’ It was in this context that in 2003 the Project took an
unexpected turn.
Our use of the video camera signified an unequivocal act of
violence; since the start of the Project, we put the video camera in the
hands of Indians who learnt to use it by themselves after reading through
the manuals with us. Some youths expressed a strong desire to learn how
to ‘make videos.’ As one of the Kuikuro video-makers told us: “since I
was little I followed the Whites from television that visited us here, I
couldn't take my eyes off them and what they were doing and afterwards I
wondered how films appear on the television and tapes; I said to myself: I
want to learn how to do this.” We contacted a Brazilian NGO – Vídeo nas
Aldeias – which has been training indigenous filmmakers for over twenty
years. A series of video workshops were planned in Ipatse village. At the
same time, we helped set up AIKAX, the Kuikuro Association, with the
purpose of raising funds from governmental and non-governmental
organizations for projects designed, realized and managed by the Kuikuro
themselves.
4. Beyond DOBES

The linguistic documentation project was the stimulus providing
the Kuikuro with the possibility and opportunity to conceive and initiate
their own documentation project. In 2003, the Kuikuro chiefs announced
their concern over the direction taken by the younger generations; they
decided to run a parallel multimedia project documenting the ‘culture’
(ügühütu): that is, all the rituals, including associated narratives, music and
songs, as well as statements and interviews obtained from the ‘owners’ of
each ritual and specialists such as musicians, singers and shamans. They
also decided that each specialist collaborating with the ‘cultural’ Project
would receive adequate payment.
110
A dialogue recorded in one video produced by the young Kuikuro
filmmakers, and here translated from Kuikuro to English, summarizes
wishes and expectations of the main chief and of the oldest Kuikuro. Inside
the kwakutu (the ‘men’s house’ in the middle of the village), J., one of the
most important ritual specialist, is formally giving to the chief T. a box
containing all the recordings produced by the documentation of the kagutu
flute songs:
T. Have you finished the recordings?
J. They finished late afternoon yesterday
T. Come here grandfather! All you singers, come here!
Listen, boys! I worry about you. That’s why I thought: let’s
safeguard our songs! Because when we’re no longer around,

or songs come to an end. When grandfather dies, you die, you
die too. So, they’re going to ask: hey, how are we going to
sing? That’s why you’re recording for our grandchildren, for
our grand-grandchildren and for their children too, so we’ll
keep our songs. That’s was my idea. I want to thank you like
the White do. Give me a hug! Boys, these are our things.
J. Here is the little I learnt, that’s what I recorded. I hope it
turned out well. Do you think our brother will learn? Who
knows? Do you think our tradition will last? The smell of the
Whites is very strong. Our brothers don’t want to know about
anything else. Our children are already becoming White.
111
More than fifteen different intra and inter-tribal rituals still exist
among the Kuikuro; they characterize a supra-local cultural community, as
they are known and performed by all the Upper Xingu groups. They are
structured around a set of songs, one or more mythic narratives and a precise
choreographic routine. The ritual mediates sociopolitically between humans
and cosmopolitically between humans and non-humans, transforming a
private sickness into a public act of collective reciprocity with the spirits.
Ritual is the locus of what Xinguanos nowadays call ‘our culture,’ an
imported notion that translates native terms such as tisügühütu, ‘our way of
being,’ and tisakisü, ‘our speech’. Ceremonial life is where Xinguano culture
is objectified and it forms one of the key mechanisms in producing the

112
identity of the whole, while simultaneously displaying the political
autonomy of the local groups.
The older generation continues to be suspicious and past problems
have not yet been completely overcome. I should like to give just a few
examples.
Every ritual marks a public status to some extent, preferentially the
status of those deemed to be ‘chiefs’ or ‘nobles’. This condition of ‘chief’
(anetü) must be inherited and at the same time ritually confirmed. ‘Chief-
ness’ is at once an objective genealogical fact and a political fact subject to
strategic manipulations. The primary form of gaining such recognition
involves becoming the ‘chief of the guests’ (hagito anetügü in Kuikuro)
during intertribal rituals. To be ‘chief of the guests’ presupposes
knowledge of ‘chief’s speech’ (anetü itaginhu), the patrimony of a true
chief (Franchetto 1993, 2000).
There exists a long tradition of formal apprenticeship in the Upper
Xingu. The transmission of shamanic knowledge, songs and instrumental
music, as well as the ceremonial chiefly speech, demands the establishment
of a relation between master and apprentice, mediated by payments in luxury
goods and lasting for a number of years. Tape recorders – and today videos –
have shortened and subverted the time span and modalities of this relation.
Today one can pay to record a set of songs or speeches from the ‘chief’s
speech’ and learn them alone at home. However, many song masters
continue to teach (and allow the recording) of parts of the set only each time.

113
The researcher becomes the guardian of precious, secret and disputed
archives and needs to know how to maintain trust and, at the same time,
negotiate honestly. This is far from easy tasks.
On the other hand, won over by the new technologies (videos and
computers), the young have begun to dedicate themselves intensely to the
documentation work, defining autonomously the research themes that led
them to enter into contact with the older community members and their
knowledge. Addicted to videos and computers, though, they move ever
further away from the increasingly rare contexts where ‘traditional’
knowledge is really transmitted. Paradoxically, the Project may be a means
of salvation and destruction at one and the same time.
We are in the middle of a fascinating process, whose final result, for
the Kuikuro, we still cannot foresee. But we can note a few facts. The
cultural documentation project has obtained its own sources of funding,
managed by the Kuikuro. The old figure of the ‘informant,’ if it ever existed,
is no more and the ‘consultants’ have already become ‘authors.’ In
coauthorship with ourselves and Video nas Aldeias 13 , the youngsters from
the Kuikuro video team have already produced two videos that have received
a great deal of comment, as well as prizes at film exhibitions in Brazil and
abroad: Ngune Elü (The day when the moon menstruated) and Imbe Gikegü
13 Established in 1987, the ONG Vídeo nas Aldeias (VNA, Video in the Villages) is a
pioneer project in the field of indigenous audiovisual production in Brazil. Since its
beginnings, the project’s goal has been to support indigenous peoples’ struggles in order to
strengthen identities and territorial and cultural heritages, through audiovisual resources and
a shared production with the indigenous peoples Vídeo nas Aldeias works with.

114
(The Smell of Pequi). The young Kuikuro video-makers, as well as being
excellent cinematographers, are completing their training in editing. They
tirelessly edited the material for an exhibition opened to the public in Rio in
August 2006 with the name ‘Tisakisü: tradition and new technologies of
memory.’ As well as providing a space for publicizing the successes of the
documentation projects, the exhibition gives a suggestion of their impasses
and contradictions. They themselves had the idea of filming and editing two
video-clips of meta-documentation, entitled ‘Kakisü ügühütu - Documenting
the language’ and ‘Kahehijü ügühütu - The custom of the camera’. 14 Using
these clips, the aim is to explain their work, in the village and the city, to the
world and especially to those who stay behind in the villages and still have
difficulties in understanding and accepting. All these videos are now in a
DVD with subtitles in Portuguese, English, Italian, Spanish and French.
old, says:
In the video Kahehijü ügühütu, the filmamaker Takumã, 20 years
When I was a child, about 5 years old, the Whites arrived
here, photographer and filmmakers, which was when I saw
their things, big cameras, like the Globo TV crew who came
here a while ago. I used to spy on them, walking behind them,
14 K-aki-sü ügühütu (12-word.language-REL way.of.being) can be translated literally as ‘the
way of being of our language’; k-ahehi-jü ügühütu (12-trace-PNCT way.of.being) means
‘the way of being of our filming’. The root ahehi means a specific modality of
representation which includes to draw maps, plans, graphs or charts, as well to write, to
make photos and to make film.

thinking: what machines are these? I was still a child, I didn’t
know. Then the chief had an idea of implementing a
documentation project in our village, so we don’t lose our
custom.
115
The young filmmakers have now their own projects (documenting
the life of Indians in the city, new forms of shamanism) and are being sought
out by other indigenous associations and by well-known ‘White’
documentarists. The problem perhaps, will be to keep them linked to their
community’s project, whose completion requires years of involvement and
application.
The DOBES Project, in concluding, is merging into the Project (or
projects) of the Kuikuro. The Kuikuro always were and continue to be the
agents of everything that concerns them, with the exception of the illnesses,
deaths and rubbish (material and immaterial) which the Whites have brought
and continue to bring.
In commenting on the phases, successes and impasses of three
decades experience studying, documenting and living a language spoken by
400 Indians between the savannah and forest of southern Amazonia, a
number of questions deserve mentioning by way of conclusion.
For the Kuikuro, speaking is an integral part of their ügühütu or ‘way
of being,’ a term which can also be used to translate our notions of
‘language’ and ‘culture.’ The essence of the latter is an amalgam of

116
sedimented properties and conditions that are always subject to personal and
historical dynamics and rooted in festivals, dances, music and songs. It is no
coincidence that the Kuikuro accepted a proposal to document their
‘language’ by transforming it into ‘documenting songs,’ a project presented
as a ‘documentation of the culture.’ Nor is it accidental that the Kuikuro, like
many indigenous peoples still largely unaffected by missionization and
schooling, both recent experiences for them, have resisted the attempts to
teach reading and writing in their own language through an ‘intercultural and
bilingual’ school in their villages.
Reifying a language as an object of conceptual and political
operations derives from the ‘way of being’ of the Whites, the literate and
colonizing west. One example of this is Christian missionization, especially
those missions that apparently subscribe to the motto ‘save languages, kill
cultures’ (and save languages through writing and the Scriptures). I was
never able to make the Kuikuro think of their language as something with its
own ontology and existence. The only exceptions are the few young men
who have undergone lengthy schooling as far as higher education
(undergraduate and postgraduate studies) and who have worked intensively
with linguists and anthropologists as part of the process of becoming
researchers themselves, dedicated to understanding and learning how to reify
a language and a ‘way of being’ (their own). We met on the frontier between
kugetilü (become a Xinguano person) and kagaihatilü (become White),
conditions arising from transformations – of the body and its coverings, the

117
word and forms of saying, memory and its supports, taste, the ‘way of being’
– that are not necessarily definitive and irreversible.
Employing all their predatory diplomacy, the Kuikuro measured and
modulated my penetration into their ügühütu. They freely applauded my
effort to learn and speak Kuikuro, to dance and sing, eat with them and work
like a real woman. Undoubtedly this attitude, along with the non-
objectification of the ‘language’ – which for us seems like a form of
obfuscation derived from a lack of awareness – is associated with a strategy
of self-preservation with positive consequences for what we consider the
‘health’ of a language. However, the language documentation project arrived
at a moment when this resistance was at a low ebb: Portuguese was
becoming increasingly heard and used with the overwhelming influx of
industrial goods and visitors, symptoms of the abandonment of ügühütu by
many young people. We offered new technologies and new knowledge that
we thought would be able to dovetail the curiosity of these youths with the
anxiety of chiefs and elders to preserve their language and culture. The new
mnemonic technology introduced by our equipment froze words, phrases,
names, narratives, explanations, speeches, and individual and family
recollections. This mechanical memory is now freezing the more than 1,500
songs which the ritual masters still remember. As an experiment, this may or
may not prove successful, since it is impossible to foresee or project into
either the near or distant future. Everything depends on the young Kuikuro

118
generations and what happens to the integrity of their territory, rivers and
lakes, swiddens and their own physical health.
I am certain that the project of documenting a language, when
combined with a well-honed anthropological sensibility, can help strengthen
a threatened minority language, not so much through its transformation into
written texts, but primarily through the perpetuation of the knowledge that
the language supports. And the best way of revitalizing a language is by
documenting it – in the broad, pragmatic, political and participatory sense
assumed here – before we are left recovering merely fragments of a lacunar
memory in order to try to resuscitate it.
REFERENCES
Barthes, R. 1984. “A fotografia por Roland Barthes”. Folha Ilustrada. São
Paulo: Folha Ilustrada (1º de setembro de 1984).
Dole, G. 1958 “Ownership and Exchange among the Kuikuro Indians of
Mato Grosso. São Paulo”. Revista do Museu Paulista, N.S. X: 124-
133.
Fausto, C., Franchetto, B. and Heckenberger M. J. 2008. “Language, ritual
and historical reconstruction: towards a linguistic, ethnographical and
archaeological account of Upper Xingu Society”. In Lessons from
Documented Endangered Languages, D. K. Harrison, D. S. Rood.
and A. Dwyer (eds), 129-158. Amsterdam: John Benjamins
Publishing Company (Typological Studies in Language 78).

119
Franchetto, B. 1986. Falar Kuikúro. Estudo etnolingüístico de um grupo
caribe do Alto Xingu. PhD Thesis, Rio de Janeiro: Programa de Pós
Graduação em Antropologia Social, Museu Nacional, UFRJ.
Franchetto, B. 1993. “A celebração da história nos discursos cerimoniais
kuikúro (Alto Xingu)”. In Amazônia Etnologia e História Indígena,
E. Viveiros de Castro and M. Carneiro da Cunha (eds), 95-116. São
Paulo: NHII/USP, FAPESP.
Franchetto, B. 1997. “Tolo Kuikúro: Diga cantando o que não pode ser dito
falando”. Invenção do Brasil, Revista do Museu Aberto do
Descobrimento 57-64.
Franchetto, B. 2000. “Rencontres rituelles dans le Haut Xingu: la parole du
chef”. In Les Rituels du Dialogue. Promenades Ethnolinguistiques en
Terres Amérindiennes, A. Becquelin Monod and Ph. Erikson (eds),
481--510. Nanterre: Societé d´Ethnologie.
Franchetto, B. 2003. “L’autre du même: parallélisme et grammaire dans l’art
verbal des récits Kuikuro (caribe du Haut Xingu, Brésil)”. Amerindia
28 (numéro Langues Caribes): 213--248. Paris: AEA.
Franchetto, B. 2001. “Ele é dos outros. Gêneros de fala cantada entre os
Kuikuro do Alto Xingu”. In Ao Encontro da Palavra Cantada:
Poesia, Música e Voz, C. N. Mattos, E. Travassos and F. T. Medeiros
(eds), 40-52. Rio de Janeiro: 7 Letras/CNPq.

120
Meira, S. and Franchetto, B. 2005. “The Southern Cariban languages and the
Cariban family”. International Journal of American Linguistics 71
(2): 127-190. Chicago: Chicago University Press.
Monod Becquelin A., De Vienne E. and Guirardello-Damian, R. 2008.
“Working together: Interface between researchers and the native
people - The Trumai case”. In Lessons from Documented
Endangered Languages, D. K. Harrison, D. S. Rood and A. Dwyer
(eds). 43-66. Amsterdam: John Benjamins Publishing Company
(Typological Studies in Language 78).

Abstract
Language vitality and revitalization in the Artic
Lenore A. Grenoble
121
A prime case study for exploring the potentially conflicting agendas of
language documentation and revitalization is Evenki, a Tungusic language
spoken by approximately 5000 people living in small villages scattered
throughout much of Siberia. Historically the Evenki people were nomadic
herders and hunters, moving across vast regions in Siberia. As a result, there
is significant dialect variation in what remains of Evenki today. This
variation, coupled with attrition, poses significant issues for language
documenters and language activists committed to revitalization. The present
paper examines the use of Evenki in two distinct regions to investigate how
practices in documentation and revitalization can mutually inform one
another, with a focus on the role of sociolinguistics in determining possible
outcomes.
Keywords: Language endangerment in Siberia, Evenki, language activism,
interface between language documentation, education and revitalization.
1. Introduction
Ever-growing concern over the loss of a large number of the world’s
languages has resulted in a concerted response by the linguistic community

122
to document these languages while still possible. This in turn has brought
about a marked increase in international funding opportunities for
documentation as well as the establishment of a number of language archives
devoted to the long-term preservation of endangered language data.
Response from the linguistic community to the call for work on endangered
languages began as a lively debate and has resulted in the emergence of
language documentation as a robust new field (Gippert et al. 2006;
Himmelmann 1998, 2006; Woodbury 2003). One issue that quickly emerged
was the relative paucity of adequately trained linguists to carry out the work,
and so a number of focused training programs have emerged to fill in that
gap. These include programs with broad reach, targeted for a wide spectrum
of linguists, such as the Endangered Language Academic Programme
(ELAP) in the Department of Linguistics, School of Oriental and African
Studies (SOAS), London, which offers a one-year Master’s of Arts degree in
language documentation; the 3L (Lyon, Leiden, London) International
Summer School on Language Documentation and Description; and the
biannual Institute on Field Linguistics and Language Documentation
(InField). Some programs, such as the annual American Indian Language
Development Institute, are aimed specifically at community members and
offer such courses as “Documenting your language through film” as well as
“Language immersion methods for language revitalization.” There are also
special targeted programs such as the Saami Language Documentation and

123
Revitalization school (in Bodø, Norway, 1-11 March 2010), 15 a program
which focuses on documentary linguistics, but also includes practical
training for language maintenance and revitalization, with special attention
on how to use language archives and corpora. This is just one of a number of
such focused training programs.
With the solid support of current funding agencies and the dedicated
commitment of linguists, and training programs to ensure future linguists all
in place, the field seems well-positioned to carry on for the indefinite future.
With the experience of the past now informing our future responses, it is
time to question two of the core assumptions underlying language
documentation: (1) that documentation is the best use of linguists’ time and
resources in addressing language endangerment; and (2) that the goals of
documentation are even achievable. These challenges arise from two
intertwined and fundamental issues. First, language documentation is at best
a second priority for many endangered language communities, who are most
often more interested in language revitalization and in developing language
education programs and materials. The second issue stems from the goals of
documentation itself, which aims “to provide a comprehensive record of the
linguistic practices characteristic of a given speech community”
(Himmelmann 1998: 166). This goal rests on the implicit assumptions that
15 Information about ELAP is found at http://www.hrelp.org/courses/; the 3L International
Summer School at http://www.hrelp.org/events/3L/3L.html; Infield at
http://darkwing.uoregon.edu/~spike/Site/InField_2010.html; AILDI at
http://www.u.arizona.edu/~aildi/; the Winter School of Saami Language Documentation and
Revitalization is at http://www.linguistik.uni-kiel.de/sldr/.

124
there exists a linguistic variety which can be clearly defined, that there are
speakers who can be identified as speaking it, and that this variety can
somehow be “captured” in use in different domains. In reality linguistic
communities are not homogenous entities; there is great variation within
individual communities and from community to community.
Answers to such questions are even less clear where the language is
undergoing attrition with accompanying structural changes. This may at first
appear to be a relatively trivial issue, for surely the need is to document
whatever is spoken while still possible. Linguists understand that “pure” and
“authentic” varieties of language simply do not exist. But closer examination
of the issue leads to the core questions of who the documentation is for, what
purposes a documentation is meant to serve, and exactly what should be
documented. If it is for the linguistic community, then historical linguists,
general linguists and typologists all need to know as much as possible about
the language before language shift took place; they are generally interested
in documenting fluent and, if at all possible, monolingual, speakers. If it is
for the speaker community, then the issue of authenticity is hard to ignore,
however it may be defined. Speakers often identify prestige varieties and,
along with them, may place higher social value on the speakers who speak
them. Where only a handful of speakers remain, communities seeking to
revitalize their language, or at least to obtain more information about it,
often do so as part of larger agendas of establishing identity, cultural and
linguistic and, frequently, as part of political and power agendas as well.

125
Clearly, language documentation is of critical importance; it is not
only an appropriate and good use of linguists’ time, but good documentation
is also central to revitalization. Yet the goals of communities and linguists
can be at odds with one another, or at least not share a common focus, more
thought needs to be given to the needs of the speaker communities and the
linguistic purposes of a documentation corpus, and more work needs to be
done to align the goals of both groups. More research is pressingly needed in
the area of contact-induced change and language attrition, an area of
research which can do much to provide needed information to both sets of
communities.
In the present paper I present two contrastive case studies of the
ramifications of these issues for one and the same language in Siberia,
Evenki (a Tungusic language), spoken in two separate, non-contiguous
geographic regions. Before turning to an examination of these different
situations, I first present the position of Evenki with relation to other Arctic
indigenous languages (§2), and then within the context of Siberia (§3), and
then turn to a more detailed discussion of the documentation and
revitalization of Evenki (§4).
2. Indigenous languages in the Arctic
Evenki is just one of a number of endangered Siberian languages
which are collectively part of a larger group of endangered Arctic languages.
Siberian indigenous people are in dialogue with an international cohort of

126
people engaged in the study, documentation and revitalization of languages
throughout the Arctic. There are approximately 40 or so indigenous
languages spoken in the Arctic today, with the exact count depending both
on how one determines what the Arctic is, and how one determines what a
language is; 16 it is common practice within Russia to refer to the languages
of “Siberia, the North and the Far East,” while Arctic residents outside of
Russia refer simply to the Arctic. Evenki is one of the languages spoken in
the Arctic, although the territory inhabited by its speakers extends south of
the Arctic as well. There is a long-standing and deep interest among
indigenous peoples in the Arctic and the far North to revitalizing their
languages, many of which are in serious trouble (see Krauss 1997).
Indigenous languages throughout the Arctic are threatened by a combination
of factors: the spread of languages of wider communication, both the
national and extra-national, including global, languages (e.g. Norwegian and
English both put pressure on different Saami varieties in Norway, Russian,
Sakha (Yakut) and English in parts of the Russian Federation, and so on).
Demands for language rights and the resources to develop, revitalize
and maintain indigenous languages are fundamental goals in and of
themselves. They are also often deep-seated parts of larger political agendas
of indigenous groups world-wide. Nowhere is this clearer than in Arctic
16 Depending on whether one uses strict physical-geographic criteria, or political and
administrative criteria as determined by the countries in the Arctic, the region referred to as
the “Arctic” can be strictly interpreted as the area above the Arctic Circle, or less strictly
interpreted as the region defined as the Arctic by the AMAP (Arctic Monitoring Assessment
Programme, a subgroup of the Arctic Council; see http://www.amap.no/). A full discussion
of the classification issues of the Arctic languages is beyond the scope of the present paper.

127
Communities. The Salekhard Declaration of 26 October 2006 at the Fifth
Ministerial Meeting of the Arctic Council stated that the Ministers of the
eight Arctic States should “encourage Member States and other parties to
support the cultural diversity of the Arctic and especially uphold and
revitalize the indigenous languages, support the Arctic Indigenous
Languages Symposium and welcome further projects in this important field”
(see Salekhard Declaration) and mandated a symposium to study the
problem and make recommendations. This resulted in the Arctic Indigenous
Languages Symposium, held in Tromsø, Norway in October 2008, which in
turn produced a set of recommendations. 17 These provide clear instructions
to a range of groups— the Arctic Council, UNESCO, indigenous groups,
and national and local governments—for ways to support the revitalization
of indigenous languages in the Arctic. They specifically advocate that the
Arctic Council “lead and facilitate inter-regional, international, and
intergovernmental activities to strengthen the use of indigenous languages
across the Arctic” (Section 5b). Their recommendations propose using
existing legislation to promote and support language revitalization, including
the European Charter for Regional or Minority Languages (1992); the
Convention on the Protection and Promotion of the Diversity of Cultural
Expressions (2005) and the Convention for the Safeguarding of the
Intangible Cultural Heritage (2003).
17 The final recommendations of the Symposium are at
http://www.arcticlanguages.com/presentations/20081021_arcticlanguages_recommendation
s.pdf.

128
This discussion is relevant here for two reasons. First, it places the
Siberian indigenous languages within the broader context of on-going
international discussions and support for language revitalization. The
movement for language rights in the Arctic has helped strengthen the
position of the Russian Association for Indigenous Peoples of the North
(RAIPON), the advocacy group for small-numbered indigenous peoples,
both within the Russian Federation and internationally. RAIPON has been an
active participant in all of these meetings; its First Vice President, Pavel
Sulyandziga, is also a member of the United Nations Permanent Forum on
Indigenous Issues and is himself Udege, one of the Tungusic groups of
Siberia. International organizations such as the UN and the Arctic Council
have been an influential source for important information and advocacy for
Siberian indigenous peoples. Second, shifting research paradigms in the
Circumpolar North place linguists in a relatively new position. Within the
context of research in the Arctic, it is no longer possible for external
scholars—scientists or social scientists—to work independently from and
without concern for the interests of local residents. This is point is made
clear in a variety of forms. Recommendations of the United Nations
Permanent Forum on Indigenous Issues to actively promote and engage
indigenous peoples and their values in research can be found in their
recommendation database on their website. The Tromsø Declaration, on the
occasion of Sixth Ministerial Meeting of the Arctic Council in Tromsø, April
2009, both acknowledges the contribution of indigenous peoples and

129
recognizes the urgent need for developing and revitalizing their languages
(Tromsø 2009: 3, 5). The Arctic Social Science Program of the National
Science Foundation in the US The Arctic Social Sciences Program
“especially encourages projects that are circumpolar and/or comparative;
involve collaborations between researchers and those living in the Arctic”
(National Science Foundation 2009).
3. Evenki and the Siberian linguistic landscape
Evenki is a Tungusic language spoken by approximately 5335
speakers out of an ethnic population of 35,527 or less (data from the 2002
All-Russian census; www.raipon.org). Speakers presently live in small
villages scattered throughout much of Siberia. Evenki is one of eight
Tungusic languages spoken in Siberia today; the remaining Tungusic
languages are spoken to the south, in China. The Tungusic languages are
SOV, left-branching, agglutinative languages characterized by ATR vowel
harmony. All of the Siberian Tungusic languages are seriously endangered.
Even, a language closely related to Evenki, is arguably the most robust, as
just less than 25% of its population speaks the language, but the total number
of speakers is only 4743. 15% of the Evenki population speak the language.
Other Siberian Tungusic languages, such as Negidal, Oroch or Orok, are
moribund (see Appendix 2 for more data.)
Historically, Evenki were nomadic reindeer herders and hunters and
roamed over much of Siberia, where they had contact with speakers from

130
multiple different linguistic groups. Their territory was larger than that of
any other Siberian people, as is true to this day. They currently live in a wide
region of Siberia, extending from the Ob and Yenisey river basins in the
west, to the far eastern coast in the East—to the banks of the Sea of Okhotsk,
encompassing the Sakhalin peninsula. Their territory stretches from the
many tributaries of the Lower Tunguska river and the Vila, Amga and Aldan
river basins in the Northwest to the Angara and Podkammenaya Tunguska
river basins in the South. Evenki currently live in Siberia in concentrated
areas in the Amur basin, in southern Sakha and, to a lesser extent, in the
Evenki district in the Lake Baikal region and Krasnoyarsk krai. As nomadic
herders and hunters, they were in contact with speakers of a number of other
Siberian languages, in particular with Buriat (Mongolic) in the Baikal and
more western regions of Siberia, and Sakha (Turkic) in the region of Sakha
(Yakutia), and with many smaller indigenous groups, especially other
Tungusic speakers, Even, Negidal, Oroch and Orok, depending on their
territory, and with non-Tungusic languages, such as Chukchi in the Arctic.
In addition, they have had centuries of sustained contact with Russian
speakers. Russian contact with the western-most groups of Evenki was
established by the end of the 16th century; it had extended far into Siberia, to
even the most eastern Evenki regions, by the early 17th century (Forsyth
1992).
The post World War II era and beyond has been marked by ever-
increasing language shift away from the indigenous languages to Russian,

131
resulting in profound changes in the phonology and morphosyntax of many
Siberian indigenous languages, and by accelerated attrition in many groups
to the point of serious language endangerment. In addition to Russian
contact, the Siberian language context is characterized by intermittent but
long-term contact of a number of language groups, because historically
many indigenous groups were nomadic hunter-herders. Many of these
people were characteristically multilingual, and some are to this day.
Depending on the general region, contact could be between speakers of a
variety of languages, primarily Tungusic, Turkic and Mongolic, as well as
Uralic and Eskimo-Aleut, along with a handful of language isolates. In the
Soviet period, contact with Russian intensified due to Soviet language
policy. One result of this early and continuous Russian presence in Siberia is
sustained language contact between Russian speakers and speakers of the
various Siberian languages over the course of centuries. This contact has had
a profound impact upon indigenous and non-indigenous languages alike
(see, for example, Blankenhorn 2003; Harrison and Anderson 2008;
Nevskaja 2000). While a few changes are areal (see Anderson 2003 for a
discussion of areal phonetics), in general change is unidirectional, with
Russian affecting the contact language, although there are a few
counterexamples (e.g., Krasovicky & Sappok 2000). Contact-induced
change has been more multi-directional in the case of contact between
different Siberian languages (see Pakendorf 2007, for example).

132
During the Soviet period, it became accepted practice to divide
indigenous groups into those which were “small” or “small-numbered,” i.e.,
those with a population of less than 50,000, and those which were not. This
practice is continued to this day. By this measure, there are two large
indigenous languages in the Siberian, Buriat (a Mongolic language) and
Sakha (Yakut, a Turkic language), and several in the Siberian South (e.g.
Tuvan, a Turkic language). By this measure, there were 26 officially
recognized small-numbered indigenous groups during much of the Soviet
period. In the post-Soviet era, this number increased first to 40 in 2000 by
decision № 255 of the Government of the Russian Federation, with a 41st
language officially added in 2005 (see Appendix 1). Some ten or so are
spoken in Southern Siberia. Of the remaining languages, approximately 14
are spoken within the official territory of the Arctic (Einarsson et al. 2004),
and the territory of some of these (in particular Evenki) extends throughout
Siberia. They represent a large number of different language families: Altaic
(Tungusic and Turkic); Eskimo- (or Inuit-) Aleut (with Yupik and Aleut);
Uralic (Finno-Ugric and Saami); Chukotko-Kamchatkan; and two isolates
(Ket and Yukaghir, although Ket has recently been identified as Athabascan
(Vajda nd.).
Language shift and attrition in the Siberian Far North are occurring at
a rapid rate; all of the small-numbered indigenous languages are endangered
to some degree, and many are moribund. Of the 30 or so small-numbered
languages of the Russian Arctic and Far North, only Nenets (Uralic) is

133
spoken by more than half of its ethnic population. The fluency rates of
others, such as Yupik (11.3% of a population of 1750) or Nivkh (5.3% of
5162) are shockingly low in terms of percentage and total numbers of
speakers (Appendix 2). A number of groups have responded to the
continuing and rapid language shift, both within their local communities and
at a more pan-Siberian level. A set of recommendations similar to those
advocated by other Arctic groups were put forth at the recent VI Congress of
Indigenous Peoples of the North, Siberia and the Far East, held in Moscow
in April 2009 (where the Tromsø recommendations were distributed to all
participants in Russian translation). At this time, the President of RAIPON
Sergey Kharuchi (2009) spoke directly of a funding shortage for the
publication of textbooks and other pedagogical and educational materials, as
well as the lack of funds for “belletristic literature in indigenous people’s
languages.” Added to this is a general shortage of teachers who can teach
mathematics, science and foreign languages. One impact is that less than
70% of children over 15 living in traditional areas have access to primary
education and access to only partial, incomplete secondary education.
Kharuchi’s remarks speak to the awareness of language shift and the need to
offset it through education as a key concern of Siberian indigenous peoples.
3.1 Two Evenki regions
The speaker population of Evenki was greatly reduced over the
course of the Soviet era. Although census data at best give approximate

134
estimates of the actual number of speakers, there has clear been a significant
decline in terms of both raw numbers as well as percentage of the ethnic
population which speaks the language. At the time of contact, nearly all
Evenki were fluent in their language. According to the 1959 Soviet census,
56% of an ethnic population of 24,710 Evenki used Evenki as their first and
primary language. As we have seen (§3) at present only 15% or less speak
Evenki; many linguists in Siberia consider this figure high. To this day,
those Evenki who maintain a traditional lifestyle continue to herd and hunt,
and there is a strong correlation between language retention and a traditional
lifestyle. A contrastive study of vitality and revitalization in two different
regions, Evenkiya, where language shift is more pronounced, and the
Republic of Sakha (Yakutia), where language retention is more robust, show
how different regional and local ecologies come into play and suggest active
roles for external linguists in documentation and revitalization.
3.2 The example of Evenkiya
Evenkiya is the Evenki municipal region, theoretically the heart of
Evenki at an official, organizational level. The town of Tura served as the
administrative center for the Evenki Autonomous District (okrug) from its
founding in December 1930 until its closing on 31 December 2006. It
continues to function as the administrative center of the Evenki municipal
region, or Evenkiya, which is now (politically and administratively) part of

135
Krasnoyarsk Territory (kraj) 18 , functioning as subunit of it. In effect, this
change in status constitutes a diminishment of Evenki regional autonomy.
While the region had official status as an Autonomous District, it has some
level of independence. This was admittedly small, but even that has been
essentially eliminated. Now the administration in Tura reports to the
administrative center of Krasnoyarsk Territory. The overall population of
this larger region is predominantly Russian and is itself dominated by the
city of Krasnoyarsk, the third largest city in Siberia with a population of
948,507 (estimated for 2009; 909,341 as of the 2002 census). According to
the 2002 census, the total population of the Krasnoyarsk Territory was
3,023,525. Of this, just over 88% were Russians, while only 0.29% is
Evenki. The other indigenous groups in Krasnoyarsk Territory account for
an even smaller percentage of the population than do Evenki. Nenets is
arguably comparable at 0.21%, but Ket is only 0.05% and Sakha (Yakuts)
0.08%.
Evenkiya is a subdivision within Krasnoyarsk Territory; it has a
population density of only about 0.03 people/km 2 as of 2008 19 and
encompasses a territory of 767.60 square kilometers in the Northern region
of Russia, and is thus comparable in size to Turkey or Chile. Approximately
10% of the land is tundra and 75% taiga, with the majority of the area
18 Both okrug and kraj are official terms within the Soviet Union and the Russian Federation
which imply different levels of autonomy and authority—a kraj is a larger and more
powerful administrative region than an okrug.
19 See the official Evenkiya website for more information; facts here are adapted from it
(http://www.evenkiya.ru/new/index.html); (28 August 2008).

136
covered with permafrost. In addition to Russians who inhabit the region,
there are assorted groups of indigenous peoples living there, including not
only Evenki but also Kets, Nenets and Yakuts. The indigenous population
constitutes 59% of the working population of the region. Evenkiya is thus
characterized by very low population density with a number of indigenous
languages spoken in an area under the dominance of the national language
which is also the regional language and lingua franca of Krasnoyarsk
Territory, Russian. At the regional level for Evenkiya, basic demographics
are a critical factor. The changes in the administrative structure which make
Evenkiya part of Krasnoyarsk Territory mean that Evenki have become a
very small minority within an area overwhelmingly dominated by ethnic
Russians. In addition, the problems of a small town like Tura 20 (population
5616 of which approximately 1000 are Evenki), or perhaps more poignantly
of the two Evenki villages with the highest ratio of Evenki, Surunda (total
population 478, of which 475 are Evenki) and Ekonda (total population 257,
of which 243 are Evenki), easily appear insignificant to an administration
struggling with the problems of a much larger population. Just a quick
calculation shows that the actual numbers of school-age Evenki children in
any of these villages is very small and so they are not likely to be an
educational priority unless other, more compelling reasons are found by
external authorities to make them one. Moreover, these villages are remote
20 http://www.evenkya.ru/rus/?id=obsh&sid=admterdel&ssid=5; (21 August 2008). I am
grateful to Olga Kazakevich for the Evenki population estimate for Tura.

137
and relatively inaccessible, and thus all the more easily determined to be
inconsequential or even disregarded.
Perhaps surprisingly, the existence of Tura as an administrative
center which is home to Evenki from a variety of regions (and speakers of
differing regional varieties) appears to have inhibited Evenki vitality. On a
field trip to Tura in the summer of 2008 Nadezhda Bulatova and I discovered
that speakers there, regardless of fluency levels, consistently reported that
they did not speak Evenki with others because the dialect differences
interfered in comprehension. Strictly speaking this was not true, since they
understood Bulatova quite well. This points to a potential role for
sociolinguistics in explaining dialect differences and promoting dialect
awareness, along with increasing speakers’ awareness of ongoing language
shift due to lack of use.
3.3 The case of Evenki in Sakha (Yakutia)
The case of Evenki living in Sakha (Yakutia) is significantly
different from that of Evenki in Evenkiya to warrant an analysis of the
different variables. These center around political autonomy, language
attitudes, language density and basic demographics. Sakha (Yakutia) is a
Republic within the Russian Federation, a status which grants it the greatest
level of autonomy possible in the Federation. It is the largest subnational
governing body in the world at 3,100,000 km 2 ; Krasnoyarsk Territory is the
third largest. The population of Sakha, however, is significantly smaller, at

138
949,280 as of the 2002 census. Ethnic Russians make up only 41.15% of the
population of the Republic; the largest group is the Sakha, or Yakuts people,
at 45.54%. Sakha is a Turkic language and so distinct from both Russian
(Indo-European) and Evenki (Tungusic). There are 18,232 Evenki in Sakha,
or 1.92% of the population (2002 census). In contrast to the case of
Evenkiya, several factors come to the fore. First of all, ethnic Russians are
not the majority, but rather second to what—at the national level—is a clear
ethnic minority, the Sakha. Second, indigenous groups constitute a
significant percentage of the entire population which is, in order of ranking:
Sakha, Russian, Evenki, Even (1.23%), as opposed to Krasnoyarsk Territory,
which is much more ethnically homogenous. Not only do Russians constitute
the overwhelming majority at 88.38%, but the second largest group is
Ukrainian, at 2.37%. Given these demographics, it is not surprising to find
that Evenki in Sakha live in a region which has been and continues to be
multilingual and multiethnic.
3.4 The standard language and dialect variation
At the time of Russian contact, Evenki people were called Tungus by
Russian explorers, and that group usually included at least the Negidal;
Negidal itself was identified as a distinct language only in the 1950’s. The
earliest published documentation of the language comes from the mid-19th
century with descriptive grammars such as Kastren (1856) or Adam (1873).
Yet at the time of the Bolshevik revolution, the Tungusic languages were

139
still largely not described and not well understood by Russian linguists.
Little work had been done to sort out and describe the differing but closely
related Tungusic varieties. This fell to linguists sent to Siberia as part of the
Bolshevik nativization campaign. One of its goals was to increase literacy
rates throughout the newly formed Soviet Union rapidly. It reached the
North somewhat later than elsewhere in the Soviet Union, but by the 1920’s
linguists and ethnographers were hard at work in recording, describing, and
writing pedagogical materials for many of the languages of the North.
A standard, prescriptive norm for Evenki was created in the 1930’s
by Soviet linguists, using the Nepa dialect as a basis. In 1952 the dialect
basis for the standard variety was shifted to the Poligus dialect, a change
which had little substantive difference in the standard language itself. Both
of these are from the Southern group of dialects which are characterized,
among other things, by the etymologically correct distribution of [s] and [h].
The differences between the standard language and the varieties spoken in
the eastern part of Siberia, in particular in Sakha (Yakutia) are significant
enough that language leaders in these parts have been working to create
pedagogical materials in the eastern varieties. The other significant issue is
the vast number of dialect differences of Evenki. They are traditionally
divided into three basic groups based on the isoglosses of inherited *s: the
Southern group, which maintains inherited [s] both word-initially and word-
internally (sulakiː ‘fox’; asi ‘woman’); the Northern group, (hulakiː ‘fox’;
ahi ‘woman’); and the Eastern group, where [s] is maintained word-initially

140
but has changed to [h] internally (sulakiː ‘fox’; ahi ‘woman’). There are a
number of lexical and morphological differences between different varieties
as well. For example, the Southern dialects use a deontic morpheme -matʃin,
which is not found in the Eastern dialects, where use of the morpheme -ŋat is
historically the norm, although it has largely been replaced by a borrowed
Russian form naːda ‘should’ (Grenoble 2000). The number of cases varies
across the dialects as well, with 11 in the standard language and as many as
13 in some dialects. Nominal plural morphology exhibits variation spread
across a geographic continuum (Grenoble & Whaley 2003a. For discussion,
see Bulatova 1987; Gorcevskaja 1954; Konstantinova 1964; Vasilevich
1948; for a brief overview in English, see Bulatova and Grenoble 1998;
Nedjalkov 1997).
The decision to base the standard language on one of the southern
dialects was probably reasonable at the time. The territories of speakers of
both the Nepa and (later) the Poligus dialects were relatively centrally
located. Both dialects were robustly spoken. Unfortunately the Poligus
dialect is now extinct and language shift is advanced in the regions where the
southern dialects are spoken; instead, it is now the eastern dialects where
more Evenki are fluent in their language.
The standard language, however, has never become established as a
cross-dialect norm. The reasons are relatively clear. Beyond the linguistic
challenges of learning an artificial language, there is little use for a written
form of Evenki. The spoken language is used to this day by people who

141
maintain a traditional lifestyle and engage in hunting, fishing and herding
activities; it is the language of the herders and of those living in villages. It
was supplanted by Russian (and in some areas Sakha or Buriat) as a written
language in all domains and as a language of wider communication in the
Soviet era, a process well underway by the mid-1960’s.
Dialect differences make standard Evenki (as well as other varieties)
very challenging for some potential learners. For example, standard Evenki
has phonemic distinction in vowel length, and the distinction between long
and short vowels is very salient for some speakers. For others, though,
anecdotal evidence suggests that the distinction has been lost. Whether this
loss is part of a larger scenario of language attrition or reflects dialect
differences is impossible to say at this time. A focused investigation of
vowel length is needed. But for speakers who do not hear long vowels,
reinstating them to meet the norms of the standard language amounts to
basic memorization. This is just one example, but it illustrates how what
would appear to be small issues can evolve into major ones. Research on
language attitudes is desperately needed, but my own experience is that
semi-speakers and non-speakers are insecure and self-conscious about their
lack of language skills. The dialect differences at first appear to be minor
but, taken together, and coupled with already existing issues in language
acquisition—the very same issues which led to language shift in the first
place—can be a serious impediment to the acquisition of literacy. As a
somewhat different example, in 1999 Nadezhda Bulatova and I published, in

142
desktop fashion, a small booklet of folklore which she had collected in
Sakhalin, entitled simply Nimngakar ‘Folktales’. We later distributed this to
speaker communities in the Amur Basin and Sakha (Yakutia). This was a
collection of several traditional folktales, narrated by fluent native speakers
in the Sakhalin region and recorded by a fluent native speaker (from the
Amur region). Although many were very appreciative of the effort, some
speakers pulled me aside and reported that this was not their language, “I do
not speak this way” or “This is not my language.”
It is difficult to know how to interpret such reactions. Is it the case
that the dialect differences are so great that one standard cannot serve all
different groups? Or is it rather the case that the lack of genuine literacy in
Evenki impedes acceptance of a single standard variety? In considering these
issues, it is important to bear in mind that Evenki speakers are bilingual in
Russian and receive education in Russian. They are, perhaps to varying
degrees, literate in Standard Russian. There is significant variation between
the Russian orthographic system and the phonetics of Russian and there is
certainly dialect variation within Russian. Most Evenki speakers are exposed
to Moscow Russian through the media and to more local, Siberian dialects,
through contact with local speakers, as well as through contact with other
Evenki, many of whom use Russian as a preferred language of
communication. Yet the perceptions of the standard—as an artificial and
foreign language in the case of Evenki, and as the correct and proper form in
the case of Russian—are radically different. Is this because the dialect

143
differences in Evenki are indeed more salient than in Russian, because the
prestige of Russian is so much greater, or because speakers are accustomed
to accepting the variety learned in the schools to such an extent that they
ignore differences which are at odds with the local variety? More research is
needed to answer these questions, but they are relevant to those endangered
language situations, where communities aim to create a standard variety as
part of language revitalization.
4. The educational system
There are several key impediments to successful language
revitalization embedded in the current educational system. One is a lack of
control at a local level, as the number of class hours devoted to the study of
the local (or, to use the term of Russian bureaucracies, “national”) language
is determined by regional and national governments, not by local authorities.
The Russian Federation is notable for strong central control of language and
education policies and allocation of resources. Article 68 of the Constitution
of the Russian Federation guarantees makes Russian the official language of
the Federation and, at the same time, recognizes the rights of individual
Republics to develop their own regional languages.
Article 69 singles out the rights of “small-numbered” peoples, i.e.,
peoples whose population is 50,000 or less, which is an official classification
in Russia. Recognizing their special status, it guarantees them rights
“according to universally recognized principles and norms of international

144
law,” without specifying what these are. Thus although the Constitution
guarantees the right to use one’s native language as a basic right of all
citizens, it makes no provisions for reconciling this with the fact that
Russian is the single official national language. The nationwide Unified
State Exam (EGÈ [ЕГЭ], edinyj gosudarstvennyj èkzamen; see
http://www1.ege.edu.ru/) reinforces its national status, effective on a
national level since the beginning of 2009. This standardized nationwide test
requires that all children take exams in Russian and in mathematics in order
to receive a school diploma. Additional standardized tests are required for
entry into any program of higher education. These include the possibility of
language tests in English, French, German or Spanish; there are no tests in
the indigenous languages of the country.
4.1 Evenki vitality and education
The current state of Evenki instruction is mixed. In those regions
where Evenki live in relatively large groups (in the Republics of Buryatia
and Khabarovsk, in the Krasnoyarsk Territory and in the Amur and Chitin
regions), Evenki is currently taught as a required subject at the elementary
level and as required or optional in middle schools in the so-called
“national” (ethnically-based) schools. Evenki is also taught in the pre-
schools in these regions, but there is insufficient evidence as to how
widespread actual Evenki instruction is. Evenki is also taught in four
universities in the Russian Federation. It has been taught at the Gertsen

145
University 21 in St. Petersburg since 1953; for the academic year 2002-03, a
total of 54 students studied Evenki at its Institute of Peoples of the North. In
the Yakutsk State University in the name of M. K. Ammosov since 1989; in
the Buriat State University (with on average 6-7 students/year) and in the
Khabarovsk pedagogical university since 1986.
Thus Evenki is in a very different position than those endangered
languages which are completely undocumented and lacking description.
Superficially it enjoys a position in the formal education system from pre-
school to graduate study. But despite these apparent advantages, there are a
number of significant challenges to any concerted revitalization effort.
Broadly, these are:
1. The existing educational system and its traditions;
2. Insufficient resources, financial and human;
3. A small and disperse target population;
4. Great dialect differentiation.
These issues are interrelated. Not only do the national laws regarding
standardized tests determine the curriculum, but a strong centralized federal
government also determines the allocation of resources. The disperse
population makes it difficult to find sufficient human resources to staff
classes, for example, because more teachers are needed to teach relatively
small, disparate classes. One of the impediments to Evenki education is a
general lack of suitable pedagogical materials. There are two fundamental
21 The full official name is currently the Russian State Pedagogical University in the name
of A. M. Gertsen.

146
problems: a lack of appropriate materials and of a clear understanding of
what would constitute such materials. Textbooks continue to be published in
“standard” Evenki, although it is now far removed in structure from the
dialects more robustly spoken. Many are written on the assumption that they
will be used by speakers of the language, whereas in fact children are
frequently semi-speakers or second-language learners who need to access
Evenki through their first language, which is generally Russian. In Sakha
there is currently work being done on creating textbooks based on eastern
dialects. Existing materials may be well suited to one dialect group but fail
to address dialect differences of other groups. One obvious need is a collated
compendium of dialect differences written in a form accessible to both
teachers and linguists (see 3.3). The recent publication of Myreeva’s (2004)
dictionary goes a long way to filling this need, although it has the
shortcoming of being only unidirectional (Evenki-Russian).
In addition to the issue of the linguistic content of textbooks is the
issue of their overall availability. The government does support the
publishing efforts of the Evenki Pedagogical Center in Tura but the books
are printed in relatively small print runs and there are major distribution
issues. Textbooks are not available for use outside of the school; rather they
are used for the duration of a class period and then returned to the teacher, so
children do not even have the opportunity to review them in the evening.
A general lack of financial resources only adds to this problem. The
small population size means that relatively little federal money is allocated

147
to Evenki education. It would be unrealistic to think that better language
documentation and description alone would help surmount the very more
basic problem of a lack of financial resources to promote Evenki language
education. This issue is exacerbated by very low numbers of potential
students. Because Evenki tend to live in villages with a total population of
200-400, the actual number of school-age children in any given village is
small. In some places, like Tura, Evenki is taught primarily in the boarding
school (which is itself a “national” school), which is home to not only the
children of nomadic herders and hunters but also orphans and children from
impoverished families who cannot support their children. To center hopes
for true revitalization in schools populated by children suffering from a host
of social problems is not realistic.
Overall enrollment data is outdated but if we assume that the
numbers have at best held steady and not decreased over the last decade,
then class size is still overwhelmingly low, even in those regions with the
densest Evenki populations. The enrollment for 1997-1998 is summarized in
Table 1:
Table 1: Evenki enrollments, 1997-1998 (data from Burykin and Parfenova
2003).
Region Elementary Middle
pupils schools pupils schools

Amur 198 5 124 5
Chitin 16 3 32 3
Khabarovsk 53 1 22 1
Evenki AO 539 17
Krasnoyarsk 29 1
Sakha 409 8 509 8
Buriatia 89 5
148
For example, in elementary schools in the Republic of Sakha
(Yakutia) there were only 409 pupils in total studying Evenki in this time
period, and in the Amur Oblast, only 198 pupils in a total of 5 schools; in the
Evenki AO 539 pupils in 17 schools for the same period.
This brings us to the issue of human resources. First is the obvious
fact that the small, disperse population basis makes it difficult to argue for
increased resources for education due to state governments needs for cost-
efficiency. The small number of pupils come from a variety of dialects and
exhibit a great range in their overall fluency and pedagogical needs. Second
is the issue faced by many language communities undergoing language shift:
the lack of qualified teachers and speakers. These issues are compounded by
the dialect variation across this small number of speakers (section 3.3)
further increases the need for more, different teachers and pedagogical
materials.
4.2 Educational innovation in Sakha (Yakutia)

149
One result of the demographics of Sakha (§3.3) is local policies
which are more favorable to indigenous languages, including the majority
language of the Republic, Sakha (a Turkic language). Moreover, language
attitudes are more favorable toward multilingualism, and there is a general
sense of pride in public descriptions of Sakha’s multilingual, multiethnic
state. Despite the high prestige of Sakha and Russian, and what would
appear to be an added burden of knowing two languages in addition to one’s
native tongue (or three, if one adds English as an important language
globally), conditions for language revitalization and maintenance are in
general more favorable here than elsewhere.
This is borne out by the general retention rates of Evenki, which are
higher than in Evenkiya. Sakha is also home to a number of innovative and
progressive education programs for indigenous language schooling, in
particular to the nomadic schools. The concept of nomadic schools
(kochevye shkoly) dates to the 1920’s and 1930’s but was largely abandoned
during the Soviet period. It has been gaining new support in recent years.
The Sakha project targets children who live among the herds with their
parents, specifically Even, Evenki, Yukaghir and Chukchi groups and is
supported by the Sakha government, according to the Law for Nomadic
Schools in the Republic of Sakha [O kochevyx shkolax Respubliki Sakha
(Jakutija)], signed 22 July 2008 (Vladimirov 2008).
The Sakha model proposes seven different models of nomadic
schools. These embrace a number of innovative teaching methods, including

150
parents as teachers and use of the internet as a means of connecting teachers
to students. These include (1) models which combine traditional schooling in
the villages and home schooling by parents; (2) nomadic network schools
which move from one herd to another, so that pupils spend some time in the
“traditional” schools when the nomadic school is located with their herd and
are home-schooled otherwise; or (3) a model which involves having a
teacher accompany the herding brigades so as to school the children as they
live in their normal environment. (Sarviro 2007 and Semenova 2008 provide
details on these models and others.) Emphasis in all of these models is
placed not only on the indigenous language, but also on learning the values
and traditions of the ethnic group. The first nomadic school in Taimyr
opened in October 2008, equipping the single teacher in the school with a
field telephone and two laptop computers. The government is required, by
law, to update and publish textbooks and other pedagogical materials,
including teacher training manuals, not only on language but also on
teaching traditional culture.
Such innovation is possible due to a combination of factors—
regional legislation, positive language attitudes, high levels of
multilingualism—all of which support Evenki language revitalization. The
schools face difficulties in finding qualified teachers who are willing to work
in them, and an ongoing lack of appropriate pedagogical materials. The
schools were conceptualized to maximize the use of technology—field
telephones and the internet enable parents engaged in home schooling to stay

151
in touch with the teachers, or teachers in the field to stay in touch with their
home base.
4.3 Linguists and revitalization
The Russian Federation poses a set of significant political and
educational issues which are probably insurmountable for most linguists,
who are in no position to influence them. It is important to at least be aware
of the existence of such factors, as they can be serious hindrances to any
programs, regardless of how solidly they may be formulated (Grenoble &
Whaley 2006: 25-35). But there are issues within the existing educational
system where linguists could be of greater help. First is the issue of the how
Evenki is taught in the schools. Historically, children entered the classroom
fluent in their local variety of Evenki and, theoretically, would have needed
instruction in the standard and in literacy. Today, few children enter the
school system fully fluent in Evenki and most use Russian as a primary
language. Most textbooks and pedagogical materials are still being created as
if the target audience were fluent speakers, and are not suitable for second-
language learners. In addition, the standard language, as it exists today, is a
form of the language spoken by no one, used only in textbooks. In this
respect it differs from standard varieties of some major languages, which
may differ more or less from colloquial varieties, depending on speaker,
domain, and other situational factors, but are recognized varieties of the
language which educated speakers use, or at least strive to use. In the

152
classroom, the children encounter a variety which is very different from
what they might hear at home or in the village, and it is a variety that is not
reinforced anywhere. In some cases, the teachers may have imperfectly
learned the standard, and exhibit transference between their own variety and
the standard, often when they are not fluent in their own variety. (For
example, speakers between [s] and [h], including hypercorrections, replacing
an etymologically expected [h] with [s].)
Second, there has been an overall failure to create domains of usage
for any written form of Evenki: since all speakers not only speak Russian but
receive formal schooling in it, Russian is the language of written
communication (Grenoble & Whaley 2003b). Of course this situation is
compounded by the use of Russian as a national language and in all forms of
media, and furthered by the relatively new Unified State Examination. This
issue dates back to the creation of the standard Evenki language; it was
simply an artificially created form which could be used alongside Russian.
The weekly newspaper Evenki Life (Èvenkijskaja zhizn’) is published in
Russian, although it does feature a page in Evenki. This, however, is
generally a translation of news from Russian into Evenki, i.e., it generally
contains information that would be otherwise accessible without any
knowledge of Evenki.
There are several fundamental linguistic issues which need to be
resolved in order to meet community desires for language education; at
present, the research of many external linguists does little to address them.

153
From the standpoint of the community, there are a number of issues
centering around variation: What form of the language should be
documented when there are competing varieties? Which should serve as the
basis of pedagogical materials? How do we determine which speakers
should be documented? Some of the thorniest questions for documenters and
for community activists alike center around the issue of just which speakers
are to be considered authoritative, whose speech should be documented. If
the community opts to revitalize, then what variety or varieties should serve
as the base for revitalization? How does one handle linguistic variation in
such cases? (In the Evenki villages I have worked in, there is general
consensus about who is a “good” speaker and who is not. (In this vein, see
also Evans 2001, which addresses issues of speakers and ownership of
language.) The more challenging task has been getting access to those who
are perceived as “bad” speakers, who are of equal interest to me, but
community members often do not want me to record such speakers.) All of
these questions stem from the central issue of language variation. From the
standpoint of the linguist, variation is central to understanding language
change and sociolinguistic variation. No documentation can be complete
without a study of variation. Even when these are not insurmountable
problems, there are no easy answers to these questions, in part because each
individual case depends upon too many locally determined variables. It is,
however, possible to identify some broad solutions. The first of these is
documenting variation in its present state. If the speaker population is very

154
small, this goal may well coincide with the goal of documenting all
remaining speakers. With larger populations, such complete documentation
may not be feasible, given temporal and financial constraints. Yet
differences in key features can be documented, as can generational variation.
Speakers tend to be aware of dialectal diversity, even if they are couched in
different terms. Alongside the act of documenting variation, linguists can
help educate speakers in dialect awareness, to understand that variation is the
natural result of language change and is found in vital languages which are
robustly spoken.
To return to the Evenki case, dialect differences extend beyond
phonological and lexical differences to morphology. Dialect differences,
briefly outlined in Section 3.4, can be relatively trivial or relatively salient,
depending on which varieties are compared. These include phonology, the
lexicon, and differences in both nominal and verbal morphology. Such
variation is not readily described in a single place and is not accessible to
non-scholars. Most dialect information was collected in the middle of the
last century, when the language was still robustly spoken, and published in
scholarly journals or monographs, often with small print runs, often in
Siberian publishing houses. They may be difficult for scholars to access and
virtually impossible for community members to find. The descriptions
themselves, often based on materials of more than half a century ago, are
possibly out-of-date and certainly need to be verified or at least compared to
more modern varieties. They do not include sociolinguistic variation as this

155
type of data was not collected in the Soviet era, when even the notion of
social stratification was ideologically at odds with the Soviet doctrine. They
are out-of-print and thus relatively inaccessible to communities. Moreover,
they are written for linguists, not for community members, and will not be
readily understood by non-specialists. Thus the existing dialect descriptions
are inadequate for community purposes, although they do provide a strong
basis for the beginning of new studies of and approaches to variation.
Russian has already had a measurable influence on Evenki. The
degree of influence depends on geographic region and on the type of
speakers concerned. It ranges from impact on the phonetic system (which
may explain the loss of phonemic vowel length for some speakers) to the
lexicon to morphosyntax. Changes in verbal government, such as
replacement of the historically expected accusative or ablative with the
dative case as the complement of speech verbs are almost certainly due to
the impact of Russian. Other changes, such as morphological and stylistic
shrinkage, are likely more the result of attrition than interference per se. But
the bottom line is that the varieties of Evenki spoken today have changed
considerably since the language was codified in the 1930’s. Documentation
of this change is important, as is broad discussion of their implications. On
the one hand they may seem alarming to language purists, on the other,
languages such as Russian and English show significant influence from other
languages, and any discussion of the impact of contact on Evenki should
place such changes in the broader context of contact-induced change.

156
Overall, the differences between dialects extend beyond phonetics
and include both the lexicon and morphology; the phonetic differences alone
are at times sufficient enough to cause difficulties in comprehension
(Lavrillier 2005: 439). For Evenki this suggests that a single standardized
variety may not provide the best solution for language revitalization. One
could argue that this is a strategy which has already failed in the past. The
documentation of variation, in a form which is both accessible and usable for
community members, coupled with in-depth information about its
significance, would be a valuable contribution to Evenki revitalization.
5. Documentation meets revitalization: Conclusion
Work in both areas, documentation and revitalization, can be
greatly informed by work in language variation and work in second
language acquisition, using concrete examples from such work on
Heritage Russian on the one hand, and the lack of such research on
Evenki, on the other. Present documentation practices intersect little, if at
all, with the interests of language activists in Siberia. This is not to say
that there have been no collaborations between external academics and
community members. One striking example is the traveling Evenki
school established by Alexandra Lavrillier, who won a prestigious Rolex
award to support the project in 2006.
One important point that I have hinted at above is the important
role that an external linguist can play by providing access to and links

with the world beyond the local community or even beyond the regional
and national governments in which those communities are embedded.
This includes facilitating the flow of information to and from other
indigenous communities to learn from their experiences. It would be too
simplistic—and patronizing—to claim that the external linguist should
play the role of the educator. Rather, the appropriate role is one of
facilitator and collaborator. What the linguist brings to the table is the
knowledge of the field of linguistics, its traditions and theoretical
assumptions, as well as access to other communities and their practices
and experiences.
The outcomes of a well-planned documentation project extend
beyond the actual documentation itself. The potential for training and
capacity-building is tremendous and so the work should be oriented from
the outset with this important goal in mind, as a driving force in the
structure of the project itself. The documentation can be planned to be
not just accessible, but beneficial, to the language community. This may
well mean reorienting the project at a core conceptual level to document
linguistically interesting features of a language, but to include language
in interaction which is important for a community which wishes to speak
and use the language on a daily basis. In both the communities discussed
here, documentation is taking place on an ongoing basis by community
members themselves. It is not full-fledged (professional) linguistic
documentation with audio (or video) recordings which are archived as
157

permanent records, in the definition of Himmelmann (2006), for
example. It is, however, community documentation, and of equal or
greater value to the communities themselves. For example, the
Pedagogical Center in Tura has advanced the publication of a number of
books which document Evenki traditional knowledge, including books
on Evenki flora (Pikunova & Pikunova 2004), animal life (Pikunova &
Pikunova 2008), both published in Evenki, and a description of
traditional crafts and beliefs (Saf’jannikova 2006), in Russian with
Evenki lexicon.
Finally, one untapped source of truly helpful research for
language communities is work on heritage language learners, including
studies of second language acquisition and attrition. This is a rapidly
growing area of research in much of North America and Europe where
immigrant populations have put both new demands on teaching resources
and have opened up new opportunities for doing such studies. Very little
work of this kind is being done in endangered language situations,
presumably due to a strong belief among linguists that the pressing work
involves documenting only fluent speakers while still possible. Again,
this is driven by a need for external researchers to have access to the
intact linguistic system, to the extent that this is possible. Such
documentation is equally important for communities who, in the process
of revitalization, need to determine which forms should be used. Just
158

which system a community decides to promote may well depend on the
information linguists provide them.
Yet projects which study and document the attrition process are
at least equally valuable to both parties. They provide information which
can be essential to constructing effective pedagogy. In the summer of
2008 Nadezhda Bulatova (Institute of Linguistic Research in St.
Petersburg) and I conducted a pilot study of vitality and use of Evenki in
Tura, the capital of Evenkija. We found a number of forms which would
not be acceptable to fluent speakers. It is difficult to determine whether a
given token reflects the impact of Russian, or influence from a dialect
other than the speaker’s native or first dialect; or is symptomatic of
language attrition, imperfect learning of or influence from the standard
variety; an instance of language change, either independent, language-
internal or contact-induced (but not a reflection of shift or attrition). The
pilot study was conducted with 12 adult speakers, all of whom came
from different regions and spoke what they considered to be different
dialects. They varied in the overall fluency and communicative
competency. This study was too small to reach definitive conclusions
although we were able to identify possible trends in language shift. It
became clear that a larger, more thorough study of language variation
and shift would be useful not just to linguists interested in this topic, but
to the language planners and pedagogues who are creating the
curriculum for Evenki instruction. They also saw the need to understand
159

the mechanics of language shift and heritage language acquisition so as
to improve their successes in the classroom.
This argues for a rethinking of current research paradigms.
Clearly, not all linguists will be interested in exploring such models and
it is not reasonable to expect them to. But at a time when language
communities are demanding to be more deeply engaged in co-
constructing research that involves them, linguists may find themselves
forced to new approaches. I have illustrated some of the issues and
possible solutions here with the case study of Evenki in Siberia because
these are communities with which I am personally engaged. The
solutions for one individual community may not apply to others. This is
certainly true across language groups and across national borders, but is
also true within the same language group. For Evenki, what is effective
in Sakha (Yakutia) may well not be applicable in Evenkija, where the
population demographics, overall language vitality, and contact issues
differ significantly.
APPENDIX 1
“Small-numbered” indigenous Siberian languages.
160
The following 26 ethnic groups were officially recognized during much of
the Soviet period as “small-numbered indigenous peoples,” along with the
ISO 639-3 codes (where available) for their languages:

1. Aleut [ale];
2. Chukchi [ckt];
3. Chuvan [xcv];
4. Dolgan [dlg];
5. Enets (Forest Enets [enf] and Tundra Enets [ent] are grouped together);
161
6. Eskimo (the standard Russian name for the language generally referred to
by linguists as Yupik, ISO 639-2 ypk);
7. Evenki [evn];
8. Even [eve];
9. Itelmen [itl];
10. Ket [ket];
11. Khanty [kca];
12. Koryak [kpy];
13. Mansi [mns];
14. Nanai [gld];
15. Negidal [neg];
16. Nenets [yrk];
17. Nganasan [nio];
18. Nivkh [niv];
19. Oroch [oac];
20. Orok [oaa];
21. Saami (all Saami varieties spoken in the Russian Federation);
22. Selkup [sel];

23. Tofalar, no ISO code, Turkic/Altaic;
24. Udege [ude];
25. Ulch [ulc];
26. Yukaghirs (Northern Yukaghir [ykg] and Southern Yukaghir [yux])
162
The following the ethnic groups were added to the above, by the decision of
the Government of the Russian Federation № 255 of 24 March 2000.
27. Alutor, formerly grouped with Koryak;
28. Chelkans, Turkic/Altaic;
29. Chulym [clw],
30. Kamchadal, Turkic/Altaic not recognized 1927-2000;
31. Kerek [krk], formerly grouped with Koryak;
32. Kumandin, Turkic/Altaic;
33. Shors [cjs];
34. Soyots, Turkic/Altaic;
35. Taz, newly recognized;
36. Telengit, Turkic/Altaic;
37. Teleuts, Turkic/Altaic;
38. Tubalars, Turkic/Altaic;
39. Tuvinian-Todzhin [tyv];
40. Veps [vep]

163
In 2005 the Izhma-Komi attained official status as the 41 st small-numbered
indigenous people:
41. Izhma-Komi
Source: Arctic Network for the Support of Indigenous Peoples of the
Russian Arctic (ANISPRA),
http://www.npolar.no/ansipra/english/items/Off_Rec_Indigenous.html
(7 May 2008)
APPENDIX 2
Indigenous languages of the Russian North and Siberia (2002 All Russian
census, www.raipon.org).
Language
Eskimo-Aleut
population Speakers % tot pop women spkrs % women
1 Inuit 1509 243 16.1% 787 128 16.3%
2 Aleut 540 7 1.3% 278 1 0.4%
Chukotko-Kamchatkan
3 Chukchi 15,767 4356 27.6% 8324 2360 28.4%
4 Kerek 8 0 0 6 0 0
5 Koryak
6 Alyutor 12 1 8.3 4 0 0
7 Itel’men 3180 86 2.7% 1715 55 3.2%
8 Kamchadal 2293 2287 99.7% 1179 1175 99.7%

Tungusic
164
9 Even 19,071 4743 24.9% 11,012 2549 23.1%
10 Evenki 35,527 5335 15.0% 18,522 2890 15.6%
11 Negidal 567 9 1.6% 318 4 1.3%
12 Ul’ch 2913 51 1.8% 1547 31 2.0%
13 Nanai 12,160 1388 11.4% 6524 737 11.3%
14 Oroch 686 14 2.0% 364 10 2.7%
15 Orok 346 3 0.9% 178 2 1.1%
16 Udihe 1657 71 4.3% 866 35 4.0%
Nivkh
17 Nivkh 5162 275 5.3% 2783 171 6.1%
Yukagir
18 Yukagir 1509 243 16.1% 787 128 16.3%
19 Chuvan 1087 40 3.7% 604 18 3.0%
Ketic
20 Ket 1494 333 22.3% 782 158 20.2%
Uralic
21 Nganasan 834 368 44.1% 472 217 46.0%
22 Enets 237 56 23.6% 115 29 25.2%
23 Nenets 41,302 27,206 65.9% 22,035 14,390 65.3%
24 Sel’kup 4249 1077 25.3% 2221 570 25.7%
25 Khanty 28,678 10,427 36.4% 15,645 5739 36.7%
26 Mansi 11,432 1312 11.5% 6265 782 12.5%
27 Saami 1991 303 15.2% 1109 173 15.6%

165
28 Veps 8240 1880 22.8% 4901 1193 24.3%
Turkic
29 Dolgan 7261 4338 59.7% 3999 2406 60.2%
30 Kumandin 3114 670 21.5% 1701 392 23.0%
31 Telengit 2399 2300 95.9% 1248 1189 95.3%
32 Teleut* 2650 1382 52.2% 1440 773 53.7%
33 Tofalar 837 32 3.8% 416 15 3.6%
34 Tubular 1565 385 24.6% 779 184 23.6%
35 Tuvin-
Todzhin
4442 4401 99.1% 2288 2265 99.0%
36 Chelkan 855 448 52.4% 479 251 52.4%
37 Chulym 656 77 11.7% 338 44 13.0%
38 Shor 13,975 4058 29.0% 7520 2264 30.1%
Soyot** 2769 2613 94.4% 1377 1305 94.8%
Sino-Tibetan
† Taz 276 0 0 145 0 0
*Teleut language = southern dialect of Altay
**apparently currently speak a dialect of Buryat, although they are
ethnically Turkic
REFERENCES
Adam, L. 1873. Grammaíre de la langue toungouse. Paris: Maisonneuve
et Cie.

American Community Survey. 2008. 2008 American Community 1-Year
Estimates. US Census Bureau American Fact Finder.
http://factfinder.census.gov/servlet/DTTable?_bm=y&-
context=dt&-ds_name=ACS_2008_1YR_G00_&-
CONTEXT=dt&-mt_name=ACS_2008_1YR_G2000_B02001&-
tree_id=306&-redoLog=false&-all_geo_types=N&-
currentselections=ACS_2006_EST_G2000_B02001&-
geo_id=01000US&-search_results=01000US&-format=&-
_lang=en (14 December 2009)
Anderson, Gregory D. S. 2003. Towards a phonological typology of
native Siberia. In Current trends in Caucasian, East European
and Inner Asian Linguistics, Dee Ann Holisky & Kevin Tuite
(eds), 1-22. Amsterdam: John Benjamins.
Blankenhorn, Renate. 2003. Pragmatische Spezifika der Kommunkation
von Russlanddeutschen in Siberien. Entlehnung von
Diskursmakern und Modifikatoren sowie Code-switching.
Frankfurt am Main: Peter Lang.
Bulatova, Nadezhda Ja. 1987. Govory èvenkov amurskoj oblasti.
Leningrad: Nauka.
Bulatova, Nadezhda Ja. and Grenoble, Lenore A. 1998. Evenki
[Languages of the World Materials 141]. Munich: Lincom.
Burykin, Aleksej A. and Parfenova, Olga S. 2003. Èvenkijskij jazyk. V.
Ju. Mixal’chenko (ed.), Pis’mennye jazyki mira: Jazyki Rossijskoj
166

Federacii. Sociolingvisticheskaja ènciklopedija. [The Written
Languages of the World: Languages of the Russian Federation.
The Sociolinguistic Encyclopedia] Volume 2, 640—666.
Moscow: Academiia.
Einarsson, Níels, Larsen, Joan Nymand, Nilsson, Annika, and Young,
Oran R. 2004. Arctic Human Development Report. Akureyri,
Iceland: Stefansson Arctic Institute.
Evans, Nicholas. 2001. “The last speaker is dead—long live the last
speaker!”. In Linguistic Field Work, P. <strong>New</strong>man and M. Ratliff
(eds), 250–281 Cambridge: Cambridge University Press.
Forsyth, James. 1992. A History of the Peoples of Siberia: Russia’s
North Asian Colony, 1581-1990. Cambridge: Cambridge
University Press.
Gippert, Jost, Himmelmann, Nikolaus P., and Mosel, Ulrike (eds). 2006.
Essentials of Language Documentation. Berlin: Mouton de
Gruyter.
Gorcevskaja, V. A. 1954. Slovarnye osobennosti podkamennotugusskix
govorov evenkijskogo jazyka. Uchenye zapiski Leningradskogo
pedagogicheskogo instituta imeni A. I. Gercena, volume 101.
Leningrad: Nauka.
Grenoble, Lenore A. 2000. “Morphosyntactic change: The impact of
Russian on Evenki”. In Languages in Contact, Dicky Gilbers,
167

John Nerbonne & Jos Schaeken (eds), 105-120. Amsterdam:
Rodopi.
Grenoble, Lenore A. and Whaley, Lindsay J. 2003a. “The case for
dialect continua in Tungusic: Plural morphology”. In Current
Trends in Caucasian, East European, and Inner Asian Linguistics
Papers: In Honor of Howard Aronson, Dee A. Holisky & Kevin
Tuite (eds), 97-122. Amsterdam: John Benjamins.
Grenoble, Lenore A. and Whaley, Lindsay J. 2003b. “Evaluating the
impact of literacy: The case of Evenki”. In When Languages
Collide: <strong>Perspectives</strong> on Language Conflict, Language
Competition, and Language Coexistence, Johanna Destefano,
Neil Jacobs, Brian Joseph & Ilse Lehiste (eds), 109-121.
Columbus, OH: Ohio State University Press.
Grenoble, Lenore A. and Whaley, Lindsay J. 2006. Saving Languages.
An Introduction to Language Revitalization. Cambridge:
Cambridge University Press.
De Graaf, Tjeerd. 2005. Kollekcii narodov Severa v fonogrammarxive
Pushkinskogo doma. [‘Collections of the Peoples of the North in
the Phonogram Archive of the Pushkin House.’] Saint Petersburg:
St. Petersburg State University.
Harrison, K. David and Anderson, Greg S. 2008. “Tofa language change
and terminal generation speakers”. In Lessons from Documented
168

Endangered Languages, K. David Harrison, David Rood and
Arienne Dwyer (eds), 243-270. Amsterdam: John Benjamins.
Himmelmann, Nikolaus. 1998. “Documentary and descriptive
linguistics”. Linguistics 36:162-195.
Himmelmann, Nikolaus. 2006. “Language documentation: What is it and
what is it good for?”. In Essentials of Language Documentation,
J. Gippert, N. Himmelmann and U. Mosel, 1-30. Berlin: Mouton
de Gruyter.
Kastren, M. A. 1856. Grundzuge einer tungusischen Sprache nebst
Kurzem Wörterverzeichnis. St. Petersburg.
Kharuchi, Sergey N. 2009. Report of the President of the Russian
Association of Indigenous Peoples of the North, Siberia and the
Far East to the VI Congress of Indigenous Peoples of the North,
Siberia and the Far East of Russian Federation. [written form
circulated at the Congress] Moscow, April 2009.
Konstantinova, O. A. 1964. Èvenkijskij jazyk. [The Evenki Language]
Moscow-Leningrad: Nauka.
169
Krasovicky, Alexander and Sappok, Christian. 2000. The Russian dialect
system in contact with Tungus languages in Siberia and the Far East.
In Languages in Contact, Dickey Gilbers, John Nerbonne & Jos
Schaeken (eds), 199-207. Amsterdam: Rodopi.
Krauss, Michael. 1997. “The indigenous languages of the north: A report on
their present state”. In Northern Minority Languages: Problems of

170
Survival, Hiroshi Shoji and Juha Janhunen (eds), 1-34. (Senri
Ethnological Studies 44). Osaka: National Museum of Ethnology.
Lavrillier, Alexandra. 2005. “Dialectes et norme écrite en évenk
contemporain (langue toungouse)”. In Les langues ouraliennes
aujourd’hui: Approche linguistique et cognitive, M. M. J. Fernandez-
Veste (ed), 433-446. Paris: Librairie Honoré Champion.
Myreeva, Anna N. 2004. Èvenkijsko-russkij slovar’. [Evenki-Russian
Dictionary] Novosibirsk: Nauka.
National Science Foundation. 2009. Arctic Social Science Program.
Synopsis.
http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=13425&org=A
RC&sel_org=ARC&from=fund (11 December 2009)
Nevskaja, Irina A. 2000. “Shor-Russian contact features”. In Languages
in Contact, Dicky Gilbers, John Nerbonne and Jos Schaeken
(eds), 238-298. Amsterdam: Rodopi.
Pakendorf, Brigitte. 2007. Contact in the Prehistory of the Sakha
(Yakuts). Linguistic and Genetic <strong>Perspectives</strong>. Utrecht: LOT.
Pikunova, Z. N. & Pikunova, I. R. 2004. Ènciklpedija prirody. Buga
dzjarin ènciklopedija. Flora. [Encyclopedia of Nature. Flora] St.
Petersburg: Prosveshchenie.
Pikunova, Z. N. & Pikunova, I. R. 2008. Ènciklpedija prirody. Buga
dzjarin ènciklopedija. Fauna. [Encyclopedia of Nature. Fauna]
Novosibirsk: Nauka.

Saf’jannikova, T. M. 2006. Raduga krasok sonkana. [Rainbow of colors
of the reindeer birthing season] Krasnoyarsk: Sibirskie promysly.
Sarviro, Julia. 2007. Sur la piste des rennes: les écoles nomades de
Sibérie.
http://www.unesco.org/education/nomadicschoolsEDPortal_fr.pd
f (4 July 2008)
Semenova, Svetlana. 2008. Nomadic schools of the Republic of Sakha
(Yakutia). [Kochevye shkoly RS (Ja)]. Presentation at the Arctic
Indigenous Languages Symposium, 20-21 October, 2008.
Tromsø, Norway.
Tromsø Declaration, on the Occasion of Sixth Ministerial Meeting of the
Arctic Council. Tromsø, Norway, 29 April 2009. http://arctic-
council.org/filearchive/Tromsoe%20Declaration-1..pdf (15
November 2009)
United Nations Declaration on the Rights of Indigenous Peoples.
http://www.un.org/esa/socdev/unpfii/documents/DRIPS_en.pdf
United Nations Permanent Forum on Indigenous Issues.
Recommendations database.
http://www.un.org/esa/socdev/unpfii/en/recommendations.htm
Vajda, Edward. forthcoming. A Siberian link with the Na-Dene.
Archaeological Papers of the University of Alaska, Volume 6,
<strong>New</strong> Series, 75-156.
171

Vasilevich, G. M. 1948. Ocherki dialektov evenkijskogo (tungusskogo)
jazyka. Leningrad: Uchpedgiz.
Vladimirov, Leonid. 2008. Razrabatyvaetsja zakon o kochevyx shkolax.
[The law for nomadic schools is being developed] Official web-
server for the organs of the state government of the Republic of
Sakha (Yakutia), Novosti: JaSIA.
http://sakha.gov.ru/main.asp?n=6363 (3 August 2009)
Woodbury, Anthony C. 2003. “Defining documentary linguistics”. In
Language Documentation and Description, Volume 1, P. K.
Austin (ed), 35-51. London: Hans Rausing Endangered
Languages Project.
172

The demise and attempted revival of Uchumataqu (Uru). Values and
Abstract
actors
Pieter Muysken
In this paper I analyze in detail the processes that have led to the virtual
disappearance of Uchumataqu, an Uru language of the Bolivian altiplano.
Four possible causes of the decay of the language are explored: (a) urban
migration; (b) socio-economic restructuring and ethnic reorientation; (c)
population decrease; and (d) demographic growth and increasing exogamy
with neighbouring Aymara groups; (e) a fragile ecology. Using data from
ethnographic research, oral history, analysis of earlier published and
unpublished sources, and parish registers, I conclude that the interaction of
several of these factors combined was probably responsible for
Uchumataqu's demise, and in the final part of the paper I contrast the fate of
Uchumataqu with that of several other seriously endangered languages in
Bolivia, and review current attempts to revive the language and their
sociolingusitic embedding.
Keywords: Revival of endangered languages in Bolivia, Uru, Aymara,
actor-centered perspective, language ecology.
1. Introduction

175
The seemingly inevitable loss and the potential revival of small
languages, all over the world, pose a number of questions, questions
primarily concerned with values. Central is of course the question: What is a
language worth (and here ‘worth’ can be defined in many ways) to whom?
This question is not answerable in the abstract, since a number of parties are
involved. Hence we need an actor-centered approach. What is a language
worth to whom? First of all, there is the speaker and the speech community.
When identifying the last speakers of an endangered language, the question
of dating the death of a language is particularly complex. What is your
language worth to you as an individual, and as a member of the speech
community? The fate of the language is not entirely individual
responsibility. A second party is the larger political entity: What is a
language worth to the country in which it is spoken? Do the dominant
powers in a country favor or abhor linguistic diversity within their borders,
or does it leave them cold? Language displacement takes place as a function
of power relationships between speakers of the dominant language and
speakers of the subordinated one. What is a language worth to Mankind? Is
language diversity, as manifested in the continued existence of numerous
small languages, important to the world as a whole, and why? A question
important to researchers, what is a language worth to Science? Is it
important to record as many as possible of the currently rapidly dying
languages, and why? Does the continued use of a sample of sixty or so
diverse larger languages not suffice for the scientific need to study diversity?

176
Do we need a speech community, or is a video or audio record of a dying
language sufficient documentation, or is documentation primarily a means of
revitalizing a language rather than an end in itself?
These are difficult questions, and they have been phrased here in
much too simple terms. The main point is, though, that ultimately we cannot
avoid confronting the question of value, with respect to languages. Clearly,
values are closely linked to ideologies, as will become clear in the next
section. In this paper I will discuss the process of the demise and attempted
revival of Uchumataqu in Bolivia, a by now extinct language, that possibly
has been always quite small, part of the equally small Uru language family.
Uchumataqu has now been replaced by Aymara. It illustrates many of the
problems we are confronted with.
In section 2 a brief overview of some of the discourses on language
diversity is given, outlining the principal positions. Section 3 contains a few
remarks on the language situation in Bolivia, and a few vignettes of specific
languages in danger of extinction. In section 4 I present the Uru language
group, and in 5 the earlier studies on the language, while the main features
and genetic status of the language are discussed in section 6. Section 7
describes the speech community, as it has developed over time, and section 8
discusses the various possible causes for the decline of the language. In
section 9, I try to place the current reactions to the decline of Uchumataqu in
an actor-centered perspective, focusing on the the connections between
sociolinguistics, language documentation, and revitalization.

2. Discourses on language diversity
177
Discourses on language diversity have filled a noisy bazar, and I
cannot do justice to all of them here. A good starting point in looking at the
discourses on diversity is the work of the early students of bilingualism.
Haugen’s (1972) The ecology of language introduced the ecological
metaphor for language into multilingual contexts, a metaphor taken over in
much current work. Fishman focused on the different functions of language
and language endangerment as functional restriction. His (1991) Reversing
language shift is very practically oriented, and outlines the various steps to
be taken in language revival efforts.
Albó’s article in Spanish “El futuro de las lenguas oprimidas en los
Andes” [The future of the oppressed languages of the Andes] (1977)
introduces the discussion of language endangerment in Bolivia in terms of
the notion of linguistic oppression, and poses that there are only two options
for an oppressed language: loss and revitalization. Hornberger (1988: 237),
on the basis of a study carried out in Puno (Peru) points to the need to embed
projects of language revitalization in an overall process of social and
political change. She writes “…what is needed for effective maintenance
planning and effective use of schools as agents for language maintenance is:
autonomy of the speech community in deciding about use of languages in
their schools and a societal context in which primary incentives exist for the
use of one, two, or multiple languages in that and every other domain”.

178
From a more distant theoretical perspective, the sociologist De
Swaan (2004) raises the issue of linguistic sentimentalism and hints at the
bad faith of the identity linguists, who condemn small groups to speaking a
minor language, but profit from being part of an international speech
community themselves. De Swaan supports his position with two arguments.
First of all, language does not equal identity. Groups can maintain separate
identities without speaking a separate language. Second, de Swaan
underlines the autonomy of speech communities to decide for themselves on
what they speak. They have no need for an external actor in this, if a clear
cut distinction between internal and external actors makes sense at all.
Similar arguments have been raised by Peter Ladefoged (1992) and Jan
Blommaert (2001) in this respect, based on their experiences in the African
context.
A very different point of view is taken by Skuttnab-Kangas and
Phillipson (1995), who stress Linguistic Human Rights: to speak your own
community language is an inalienable human right. This recommendation
has not been taken over as such by UNESCO, which however has issued a
Universal declaration on cultural diversity and action plan (2001).
Recently, in public discourse a link has been established between
bio-diversity and linguo-diversity, sometimes under the label of ecological
linguistics. Organizations such as Terralingua stress this link, and thus
language rights activists have been linked to ecological activists. Journalist
Lorenzo Carrasco speaks of the ‘Máfia verde’ [green mafia] when referring

179
to this general approach, often in relation to the defense of the Amazonian
rainforest. The parallel language/species is not without problems, however.
There can be bilingualism in communities, but not always multi-species co-
existence. Some species may be put into a reserve, but speakers cannot be
treated in the same way, tend to be mobile and engage in complex
interactions with outsiders.
Advocates of the English-only movement in the US have also
stressed the economic costs of maintaining separate languages. On one site,
the website of U.S. English, Inc., it says: ‘According to the Canadian
Government, the annual cost of providing government services in both
official languages is three cents per Canadian per day. While many sources
discredit the theory that official multilingualism costs only pennies per
person daily, even this amount spent on only two languages would
dramatically impact the American economy.’ However, in these economic
calculations, the costs of excluding groups from the public discourse have
not been taken into account.
A strong critique of the language revitalization movement is
presented in Heller and Duchêne (2007) who, basing themselves on case
studies mostly in the European context, stress that assumptions underlying
the defense of minority languages are frequently concealed, and that often
there are interested parties that remain hidden. They also critique the
commodification of minority languages, where language documentation is
portrayed as creating a speaker-proof package of linguistic materials.

180
Two recent papers (May 2005; Dobrin et al. 2007) aim to reconcile
some of these widely different positions, aiming at a balanced perspective.
All these discourses, spanning a wide range of the political and academic
spectrum, and drawing on experiences from all continents, show a large
range of perspectives on the issue of language endangerment. It is too early
to draw valid generalizations, but possibly the concerns about
endangerments are much stronger in the Americas and Australia than in the
Old World, and possibly stronger among linguists than among social
scientists or politicians. Similarly, speakers in different parts of the world
differ widely in their views on this issue. I think it is fair to say that taking
into account the local political context is crucial in an evaluation of language
revival activities.
3. Language in Bolivia
Bolivia has ca. 9 million inhabitants (the total in the 2001 Census
was around 8.261.000). The majority of these indicate that they speak
Spanish (6821 thousand), and substantial minorities speak Quechua (2281
thousand) and Aymara (1525 thousand). The third original Amerindian
language is Guaraní, with around 62 thousand speakers, and the thirty odd
other languages taken together have less than 49 thousand speakers.
Of the ca. 35 languages spoken in Bolivia in recent times
Uchumataqu is not unique in having almost disappeared. A number of

181
languages were classified by Crevels (2002) as seriously endangered or
nearly extinct. These are presented in Table 1.
Table 1: Languages listed as moribund or possibly extinct in Crevels 2002.
Language Language family Number of speakers Number of ethnic group
Reyesano TACANAN

182
Causes of decay of the Leko include the land invasions in the early
20th century due to the quinine and rubber boom, which lead to ethnic
restructuring and a shift to Spanish. During the Revolution of 1952 there was
a strong tendency towards ethnic redefinition of indigenous peoples as
campesinos ‘peasants’. Currently the Leko communities are under threat due
to the migration of Aymara and Quechua “colonos” from the highlands,
which has led to fragmentation of the original Leko communities.
Itonama, according to the census of 2001, had 389 speakers (2791 in
the self-declared ethnic group). Crevels discovers only a handful of very old
actual speakers, however, and refers to these as rememberers (“recordantes”)
rather than speakers (“hablantes”). Population went from ca. 6000 in 1700 to
300 in 1967.
Causes of decay are similar to those for Leko, but go further back. In
the 16th century cases of epidemics and slavery are reported. In the 17th
century the Itonama suffered from the Jesuit policy of “reduction” of
indigenous groups into settlements. The 18th century saw an increase in
commercial exploitation, culminating in the rubber boom ca. 1900. In the
1930s the Chaco war against Paraguay was waged, in which many Itonama
died, and the Revolution of 1952 led to ethnic redefinition of the Itonama as
campesinos. These stories illustrate the fate of many of the lowland groups. I
will now turn to the highlands, where a slightly different situation holds.
4. Uru

183
The Bolivian altiplano [high plain] is now predominantly Aymara-
and Spanish-speaking, but in earlier times several other small language
families were present. One is Puquina, a language which has been related to
the Arawakan language family on the basis of the shape of its pro-clitic
personal pronouns (Torero 1987). Puquina has been extinct for perhaps
almost a century. Some of its vocabulary has been preserved in the mixed
language Callahuaya (cf. e.g. Muysken 1996).
Uru. The other small language family is the one that concerns us
here, Uru. In the colonial period its speakers have occupied part of the
borders of Lake Titicaca, Río Desaguadero, Lake Poopo, Río Lacajahuira,
Lake Coipasa, and Río Lauca, in short the main altiplano water system, in
the departments of La Paz and Oruro. Now three groups are left that are most
frequently considered ethnically Uru: the inhabitants of Iruitu [official
hispanicized name: Irohito], on the banks of the Desaguadero River near
Lake Titicaca, the descendants of the speakers of Murato or Chholo on Lake
Poopo, and the Chipaya near Lake Coipasa (e.g. Olson 1967; Porterie
Gutierrez 1990; Cerrón Palomino 2001, 2005). In addition, three other
groups feel sufficient affinity with the Uru that they joined the recently
formed Nación Originaria Uru (2001) [Indigenous Uru Nation]: the
community of San Juan de Coripata (Carangas Province, Department of
Oruro, Bolivia), the community of Isluga (northern Chile), and the
inhabitants of the ‘floating islands' on Lake Titicaca near Puno (Peru). All in

184
all, the ‘Uru nation’ would at most 2,000 peopleThere are reports of a total
of only 1 190 Urus for the Uru-chipaya, Murato, Iruito groups.
There are also linguistic data from Ch’imu, a village near Puno, Peru,
recorded by Lehmann (1928a), reflecting a now extinct variety clearly
related to the other Uru languages. Apaza Apaza (2000) presents data from
the Aymara dialect of Uyuni and Coipasa suggesting lexical traces of an
antecedent Uru variety in the region. The data for Murato, formerly spoken
near Lake Poopo, are scant and possibly not all reliable. Recently, new
recordings have become available for Murato (Schumacher et al. 2009).
Table 2 presents numerals for Murato, Chipaya, and Uchumataqu or Uru; the
results suggest a clear split between Uru and Chipaya, on the one hand, and
Murato on the other. The glossary provided in Miranda and Moricio (1992),
which is claimed to contain words from the Murato linguistic heritage, has
an overlap of six out of 58 items with the vocabulary I have collected in
Iruitu: ara ‘totota root’, kakunas (Mur)/kaku (Uch) ‘type of net’, liwis
(Mur)/liwi (Uch) ‘boleadora, used for hunting’, qhana (Mur)/qaana (Uch)
‘fishing implement’, sipita ‘trap used for flamingoes’, untu ‘llama fat’. The
correspondences are clear but they could be borrowings reflecting a common
life style. Thus it is not clear whether the speakers of Murato may have had a
different ethnic origin from the Uru peoples, and simply shared their aquatic
life style. In addition, /liwi-/ is also used in Quechua; /sipita/ and /untu/ are
also common items in Aymara, and the Quechua of Northern Potosí.

Table 2: Numerals in the three Uru languages
Murato Iruitu Chipaya (Métraux)
1 Maitin Chhi Tshiy
2 Paitín Piski pizk
3 Iritín Chip chhep
4 Yunkati Paxpiku paqpik
5 Tajso Taxnuku -
6 Takarko taxchuku -
7 Kollkhana Tuuku -
8 Chipana Quunku -
9 Chaleco Sanqu -
10 k’ati Qhalo -
185
(Sources: Chipaya: Métraux 1935b/36; Murato: Guerra Guitiérrez 1984; Iruitu: fieldwork
data, Distrito Nacionalidad Indígena Urus de Irohito 2005)
External links. Attempts have been made to identify the Uru
languages with Puquina (Créqui-Montfort and Rivet 1925; 1926; 1927), and
to link it to Mayan (Olson 1964, 1965). Both of these attempts proved to be
unsuccessful. The proposed cognates with Mayan languages are not
convincing. Puquina and Uru share some typological features (notably the
possibility of attaching subject pronouns to the verb, cf. Adelaar on Puquina
in Adelaar and Muysken, 2004), as may be expected given their prolonged
coexistence on the Altiplano. Nonetheless, shared basic vocabulary is almost
nil and shared grammatical morphemes are absent, making a genetic

186
relationship highly unlikely. Fabre (1995) has pointed out the existence of
shared vocabulary between Uru and lowland Bolivian languages such as
Tacanan. These correspondences are convincing, but do not unequivocally
point to genetic relationships. Earlier highland-lowland contacts are equally
possible. The possible affiliation of the Uru languages with other language
families in the region will need to be reconsidered again when the Bolivian
lowland languages have been better described, a project currently under way
in various research centers around the world, including La Trobe
(Melbourne), U. of Texas (Austin), University of Lyon, and Radboud
University Nijmegen (Netherlands).
Only one Uru group has managed to retain its language, the Chipaya.
The Murato only recall a few words of their original language, Chholo,
which probably ceased being spoken before 1950. The Uru of Iruitu likewise
have lost their language, Uchumataqu, although not completely. Both the
Murato and the Uru of Iruitu are now Aymara-speaking, while the Chipaya
tend to be bilingual in Chipaya and Aymara. Members of all three groups
also know Spanish as a second or third language, to varying degrees.
Why and when did Uchumataqu cease to be spoken in Iruitu? Before
going on, it is useful to clarify nomenclature, which is somewhat confusing.
Consider Table 3, where the main terms used are explained.
Table 3: Names for ethnic groups and languages used

People Language
Generic Uru Uru
Iruitu Uru of Iruitu Uchumataqu
Santa Ana de Chipaya Chipaya Chipaya/Chipay tago
Lake Poopo Murato Chholo
Aymara Aymara
187
Historically, the situation is even more complicated, because there
was no exact correlation over time between ‘Uru’ ethnicity and a nomadic
aquatic life style, on the one hand, and ‘Aymara’ ethnicity and a sedentary
agricultural life style, on the other (see section 4 below).
What makes the situation of Uchumataqu special is that it has been
documented fairly continuously from 1893 until 2001/2002, the moment at
which I visited the community three times for a number of weeks and
worked with the last reasonably fluent speaker. Not only can we trace the
numerous changes in the language during the last century of its existence, we
can also place all consultants listed in the family trees of the five families
that constituted the community in the 20th century. Most researchers who
studied the language mention by name who they worked with. In addition,
we can link the sociolinguistic commentary in the different sources to the
remembered past of the older community members. Finally, there is
information, albeit incomplete, in the parish records about marriages and
births, allowing us to reconstruct some of the family trees and explore

188
patterns of ethnic exogamy. This paper illustrates the possibilities of
combining ethno-historical data with current fieldwork data. In addition, it
contributes to the growing body of knowledge about settings of language
attrition and death.
To conclude this introduction, I should mention that, while
Uchumataqu is not the vernacular community language of Iruitu any more, it
is not completely gone. Many people, ´recordantes´, know a number of
words and fixed expressions, and there is a small circle of Uchumataqu
amateurs, who try to recollect as much as possible of the language. In the
1992 census 87 people claimed to be speakers of Uchumataqu (Albó 1995).
On the basis of the work done in 2001 two simple cartillas, booklets with
text for reading, have been prepared for the school, and an hour a week of
Uchumataqu instruction is planned (cf. Muysken 2002). Now the last known
speaker, Julia Vila, who died in 2004, lies buried in a nameless grave in the
small Iruitu cemetery, marked with a small wooden cross and three stalks of
totora reed.
5. Earlier studies
A survey of the extensive research on the Uru of Iruitu and on
Uchumataqu is given in Appendix 1 Table 4.
It is clear that a dozen or so researchers have left research reports
about the community and the language, and others may well have been there
but without leaving tangible results. These researchers had a number of

189
different nationalities and language backgrounds; very few were Bolivian
nationals. What is striking is how few of the research results have been made
available to the Bolivians, let alone members of the community. The
important early work by Uhle (1893) and Lehmann (1928b) was never
published and has remained in German manuscript form in archives in
Berlin. The very rich data base of the French anthropologist Jehan Vellard
(1949; 1950; 1951; 1954; 1967) was mostly published in an inaccessible
French series. Even now it is hard to find, and represents the language in a
phonetic notation which is difficult to interpret. Currently, these materials
have been digitalized, compared and analyzed in Hannss (2008). Only
recently are there sources available to the community (notably Colette
Grinevald’s 1995 field notes and my own field data; Distrito Nacionalidad
Indígena Urus de Irohito 2005), and very little has actually been published.
A second remarkable finding is that Iruitu must have been quite a hot
spot for scientific visitors at one time. Between 1928 and 1933 no less than
four researchers spent time in the then tiny community. There was a special
(self-selected?) subgroup of “informants” who could spoon out the same
information to different researchers, often in the form of set phrases. The
resulting reservoir of canonized cultural and linguistic knowledge may have
helped the community through hard times later. The results of fieldwork are
appropriated by the community and become part of the linguistic tradition.
A third aspect that should not remain unnoticed are the activities of
Lorenzo Inda (1947-..), great grandson of Manuel Inta (+1865 - +1948), the

190
main informant of Jehan Vellard. Lorenzo Inda was not brought up speaking
Uchumataqu, but learned quite a bit from Vellard's other main informant,
Carlos Quispe (+1900 - 1962), and from his grand uncle, Luis Inta (+1890 -
+1960). In any case, Lorenzo Inda became a powerful community and local
politician. At a course at THOA (Taller de Historia Oral Andina [Workshop
on Andean Oral History]) he learned the rudiments of linguistic analysis. In
1982 he did his own fieldwork with Eduardo Inta (1891- ca. 1985), not a
close relative (see below), on the language. The result is an unpublished
Uchumataqu-Aymara word and phrase list. In 1985 a facsimile edition of his
handwritten illustrated history of Iruitu was published in La Paz.
6. Uchumataqu
Together with Chipaya or Chipay tago, as said before, Uchumataqu
is the sole well-documented member of the Uru language family. The
languages are quite close, and currently efforts towards unification in
orthography are under way; unification in vocabulary and grammar is less
realistic in the short term. Uchuma may derive from Juchusuma, reportedly
the native name of the Río Desaguadero, and taqu means `language'. Hence,
`language of the (Desaguadero) River people'. An alternative explanation
(Olson 1964) is that Uchuma derives from the personal pronoun uchumi
`we', and hence `our language'. However plausible this may sound,
Juchusuma and its variants occur as an ethnonym for the Uru of Iruitu in the

191
colonial sources, and there is a settlement near Lake Poopo called
Juchusuma. This makes a derivation from uchumi less likely.
To describe the features of the language is somewhat risky, since
these have undergone changes over the years. Nonetheless, overall
characteristics are that it tends to be verb-final with relatively limited
derivational morphology, case markers or postpositions for the oblique
arguments, little if any participant-marking on the verb, and a semi-
obligatory indicative particle placed somewhere in declarative sentences
(often on the verb). The following sentence from Vellard (1949: 151)
constitutes a representative example (PA = past; IND = indicative):
wakpa-chul tsini waki chork k'ota-kis huxk-u-chay
all-together egg Guaqui big lake-to leave-PA-IND
‘All eggs have gone to the big lake of Guaqui’
Notice that modifiers precede the noun, and that the place name
Guaqui is simply preposed to the noun it specifies. There is a past tense
marker -u-, preceding the indicative enclitic -chay. Recent work by Hannss
(2008) confirms the hypothesis that Uchumataqu and Chipaya were
originally quite closely related; one of the grammatical features mentioned
by Cerrón Palomino (2001; 2005) for Chipaya, systematic gender
distinctions, is not documented in any of the Uuchumataqu sources,
however.

7. Irohito
192
The community of Irohito [´iru´witu] is located on the banks of the
Desaguadero river, about 25 kms. from Lake Titicaca, which empties into
this river (besides there being a considerable amount of evaporation). In
good times, the river is very wide at this point, although mostly rather
shallow. It contains (or used to until recently) a rich ecosystem with totoras
(reeds), water plants, fish, and water birds. The Uru are completely
dependent on the river. They hunt and fish (selling their surplus catch
elsewhere), and feed their cattle with reeds and plants from the river. In
earlier times, the totora played an even more important role, serving as the
basic material for their reed boats, and for the roofs of their houses. Now
boats are made of wood, and roofs of various materials including roof tiles
and eternit. Nonetheless, dependence on the river is still the hallmark of Uru
life.
As far as we know, the community has never been very large,
ranging from 7 in the earliest source to as many as 80 in the early 20th
century. In Appendix 2 the population estimates provided by the different
researchers are given. P Presently, the community is larger than ever,
with over 200 members This contradicts what Vellard predicted half a
century ago (1951: 51) “le groupe a été frappé à mort et ne se reformera
plus” […the group has been struck dead and will not reshape itself any
more.]. Regarding Vellard, it should be said that he was using racial criteria

193
rather than ethnic ones; so all Uru descending from mixed marriages would
not count for him. Still, he would be very surprised to see Iruitu now. For
Altiplano standards, a thriving community with a school, a small (if
generally locked up) museum, two community centers, and a soccer field.
Indeed, a model for neighboring communities, and Uru in its ethnic self-
definition even if there is progressive aymarization.
In the colonial period the Uru occupied a relatively more important
position in the area than today. In 1657, for example, 6 out of 15 yanaconas
serving with the cacique of Jesús de Machaca come from Ancohaque
(Wachtel 1990: 431). The second family name of the cacique, Gabriel
Fernández Guarachi, would later become one of the characteristic Uru
family names. However, it is also clear from Wachtel’s detailed account that
the relation between the Uru and the Aymaras was far from being a simple
dichotomy. In the Pre-Columbian period both groups coexisted, together
with the Puquina, in the Altiplano, possibly with ecological specializations.
The Uru were concentrated along the water system of lakes Titicaca, Poopo,
and Coipasa and the rivers connecting them. In the colonial period, many
Uru were aymarized, and in the early colonial period the Uru of Jesús de
Machaca were split into those assimilated to the more sedentary and
terrestrial Aymara life style, the Uru of Ancoaqui, and those retaining their
more nomadic and aquatic existence, the Uru of Iruitu. Wachtel (1990: 425),
referring to the period 1570-1575, writes:

194
Les Urus d’Ancohaque, par exemple, au fur et à mesure que s’efface
l’identité pacaj, se perçoivent, et sont perçus, simplement comme des
Indiens autochtones de Jesús de Machaca, des originarios, au même
titre (ou presque) que les originarios aymaras. Presque, dans la
mesure où ils ne paient qu’un tribut de 2 pesos par tête et par an,
tandis que les membres des autres ayllus doivent 5 pesos. Mais cette
différence s’avère elle-même relative dès lors que le tribut
d’Ancohaque fait contraste, à son tour, avec celui des Iru-Itus, qui ne
paient que 6 réaux.»
[The Uru of Ancoaqui, for instance, the further the pacaj identity
fades away, perceive themselves, and are perceived, as the native
Indians of Jesús de Machaca, as originarios, in the same way (or
almost) as the Aymara originarios. Almost, inasmuch as they pay a
tribute of only 2 pesos per capita per year, while the members of the
other ayllus owe 5 pesos. But that difference is itself relative since
the tribute of Ancoaqui contrasts, in turn, with that of the Iruitus, who
pay only 6 reales.]
According to Wachtel (1990: 374) the process of aymarization of the
Uru population was in large part accomplished around 1680, and slowed
down in the 18 th century. During the process, the Uru communities in contact
with the colonial authorities were replenished with new groups coming from
the lake. At the same time, the aquatic Uru populations were increasingly

195
marginalized and set apart from the sedentary Uru and the Aymara.
Consequently there were rebellions from the riverine Uru’s, also referred to
as the Ochosumas and the Iruitu’s, against the sedentary Uru and the Aymara
of Ancoaqui, San Andrés, and Guaqui.
At present, the original Uru of Ancoaqui are considered Aymara,
even if they have a special relationship with the Uru of Iruitu, as evidenced
in the frequency of marriages between the two communities (see below).
Even though there is an ideology of distinctness, the differences between the
Uru and the Aymara should not be exaggerated. Other people along the river
also feed their cattle with reeds and water plants gathered in boats, even if
most Aymara living more inland have no experience with water. There are
absolutely no discernable physical differences between the two groups.
8. The decline of Uchumataqu
Brenzinger and Dimmendaal (1992: 3) divide the approaches to
language decline into two main categories:
1. Developments in the external environment: political, historical and
economic factors;
2. Reactions of the speech community: language use, attitudes, speaker
strategies.

196
It would not be appropriate to think of these as competing clusters,
but rather as different levels of analysis. Taking this as our starting point,
five external explanations, alternative or complementary to each other, will
be given for the demise of Uchumataqu in Iruitu:
a. Urban migration: as more and more Uru went to live in the cities, old
traditions declined and the language withered away.
b. Economic and socio-cultural restructuring: the Uru started resembling
the neighboring Aymara of Ancoaqui more and more in their living
habits.
c. Population decrease: the number of members of the community
diminished, and hence the transmission of the language was interrupted.
d. Exogamy: because Uru men were more and more prone to marry Aymara
women, children were not raised in the language any more. Aymara
women are known for their practice of raising their children in Aymara.
e. Ecology. The delicate ecological basis of a water-based economy may
have proven fatal for traditional Uru language and culture in the long
run.
Combining data from different sources, I will look at these
explanations one by one. However, first the sources of information need to
be discussed a bit further.

8.1 Data sources
197
To explore thefive ‘external’ explanations, we need to study all the
available data on the demographic and social history of the community
during the 20 th century, where possible in the light of accounts of consultants
from the community. An important source of information is constituted by
the parish records available for the early period in the parish center in Qurpa
(transferred there from Jesús de Machaqa). They cover much of the
following periods:
1. Baptisms: 1911-1935, 1939-
2. Weddings: 1872-1902, 1912-1925, 1932-
This information can be matched with genealogical information
gathered in Iruitu itself, with the information in the cemetery of Iruitu, and
with the data provided in the various descriptions. My exploration of the
parish records did not produce fully reliable results, for a number of reasons:
(a) The records contain gaps where registry books were lost or stolen.
(b) In addition, the priests were not always careful or did not
understand their parishioners very well. Thus the last name Inta appears as
Inta, Inda, Ita, Inca, and Ingra. The same person appears with the first name
Felisa, Feliciana, and Felicidad in various places.
(c) Since Iruitu was situated near the margin of the Jesús de
Machaqa/Qurpa parish, sometimes people may have had their children

198
baptized e.g. in the neighboring parish center of Guaqui. I have not studied
the Guaqui records, which presumably contain only a very small number of
Iruitu entries.
(d) The Jesús de Machaqa/Qurpa records themselves only have about
one out of every 200 births and weddings as pertaining to Iruitu (reflecting
the small proportion of Uru viz à viz the total Aymara population). Since the
handwriting of the priests over time has not gained in clarity, I may have
missed a few births and weddings (hopefully not many).
(e) There is an identification and demarcation problem. Since there
are only four family names in Iruitu, I have used the last name criterium to
single out Uru births and weddings. If the father or mother had one of the
family names in Table 4, I counted them as Uru.
Table 4: Spelling variants of Uru family names in the parish records
Standard name Variants
Inta Inta, Inda, Inca, Ita, Ingra
Vila Vila, Guila, Villa
Quispe Quispe
Salinas Salinas, Salina
Huaranca Baranca, Guaranca, Huarachi
In the parish records, the large majority of the places of birth for
these family names are given as in Table 5.

199
Table 5: Places of birth indicated for Uru children and places of birth
indicated for Uru wedding partners
Place name Number of births
Ancuaqui, Ancoaqui, Hankoaqui, Jancoaqui, Ancohaqui 23
Ancoaq i urus 1
Ancoaqui abajo 2
Ancoaqui (Iruito) 2
Eruguello, Iru Guito, Iruitu, Iruuito, Iroguito, Iruguito, Irohuito 21
Urus 1
-- 1
Total 51
To be sure, Iruitu is a part of, or appended to, Ancoaqui (

200
Additional data sources include the small and decayed cemetery in
Iruitu, community records, and interviews with inhabitants of Iruitu about
their ancestors. However, the results from these are sometimes confusing,
since remembered names may be distinct from the name recorded in the
baptismal register. On the basis of these sources we can now turn to the
possible causes of the decay of Uchumataqu in more detail.
8.2 Urban migration
In many parts of South America migration of community members,
often cyclical, to the large cities has lead to progressive Hispanicization and
consequently, to loss of the original language. Young men go to work in the
city, sometimes taking their families with them, and return to their home
village occasionally, bringing Spanish back with them. Indeed, many Uru
now live in the El Alto neighborhood of the capital La Paz and some also in
other cities such as Cochabamba and Santa Cruz. In occasional soccer
championships organized in the region, the urban Uru from (Iruitu???)
participate with their own team, the Uru Fris (< Eng. free). However, this
migration appears to be relatively recent, while Uchumataqu stopped being
spoken quite some time ago, given that there is almost no one who
remembers the language enough to speak it actively. Furthermore,
Uchumataqu did not loose ground to Spanish, as would be expected if
language loss were due to urban migration (as elsewhere in Latin America),
but to Aymara, which is now the dominant language in Iruitu. We will see

201
below that in the 1940s many Uru’s resettled elsewhere, but not in the city of
La Paz.
8.3 Economic restructuring and changes in residence patterns and life style
Complementary to these processes there is economic and
consequently socio-cultural and ethnic restructuring. Many small peoples
have simply been integrated into larger neighboring groups, losing their
original mode of subsistence, residence patterns, and life style.
Like many, even most, Andean communities, Iruitu is divided into
two parts, a northern part called alay wariyu (

202
inhabitants. If we accept the idea that the Inta split up in the course of the
twentieth century, these clans could correspond to the present-day four Iruitu
families. Métraux also mentions two moieties: ti achai (anan-saya), xanacha
(mana-saya). However, these terms mean roughly ‘this one’ and ‘the one
over there’, and the two moieties may well correspond to the main division
given in (2). According to Lorenzo Inda, the four families originally lived
together in a tight compound, while later they dispersed to their present
much more dispersed locations.
The original settlement was a small compound of five or six large
round houses, one for each family or clan, divided into two parts with a
sacred house in the center. This house is told to have contained a hole in the
ground, through which the Uru deity, Uphaña, would speak. The compound
was located in the center of the village, near the chapel and cemetery. At the
time there was only fishing and hunting, no agriculture. This must have been
the situation encountered by Polo and Uhle around 1900.
The idea of a single compound is compatible with Uhle’s observation
in 1893 that there were only 27 members of the group, as well as with the
reduced population figures given for the earlier period. Possibly the
compound existed as late as 1900 or 1910. However, we know from
Palavecino’s (1949) detailed description of the situation in 1933 that there
were separate compounds at that point, square yards surrounded by a low
wall with several buildings, one per extended family. This is what we find
today, particularly in the southern part of Iruitu: clusters of square

203
compounds, either adjacent or separated by an alley. The difference between
what Palavecino describes and the present situation is that the round main
dwelling of each compound has been replaced by a square dwelling.
Palavecino’s observation is also compatible with Posnansky’s (1934) list of
16 households in 1931.
Thus, if the memory of a single compound is correct, the
transformation to a community of multiple compounds must have taken
place well before 1930, presumably concommittant with the demographic
growth of the early 20 th century. In addition to demographic aspects, this
restructuring also had an effect on the economy, in which breeding of
livestock and planting tubercles and some barley became a secondary source
of food.
However, the restructuring did not imply that the community became
more like the surrounding Aymara communities. In these, the compounds
are very wide apart, each isolated from the others, while in Iruitu, houses
remain very much clustered per family group. Thus the restructuring did not
imply an aymarization in itself. In addition to a difference in residence
patterns, the Uru distinguish themselves from their Aymara neighbors in
their boat-building and boating skills, their reliance on fishing and hunting of
water fowl, and their use of reeds, totora, as cattle fodder. Unfortunately, the
occupations mentioned for the Uru in the parish records are not specific
enough about their way of living (though one is listed as fisherman); cf.
Table 6.

204
Table 6: Uru professions listed in the early 20 th century parish
records. ‘Workers’here are those employed on agricultural estates.
8.4 Population decrease
Year Profession
1915 Labrador ‘worker’
1915 Tejedora ‘weaver’ (fem.)
1916 Labrador ‘worker’
1916 Tejedora ‘weaver’ (fem.)
1916 Pescadores ‘fishermen’
1917 Labradores ‘workers’
1918 Labradores ‘workers’
At first sight, population decrease cannot have been responsible for
the decline of Uchumataqu. At present, Iruitu has more inhabitants (and all
stemming from families that originated there) than ever before in its history.
However, the present day situation does not reflect the demographic
development throughout the 20 th century. A large part of the explanation for
the rebirth of Iruitu from 1948 onwards is that the cause for the population
decline between 1930 and 1949 - drought - has been absent (see section 5.6
below). Many families left for the communities in the region of Guaqui and
Taraco. Others sought refuge in nearby Aymara communities and married
Aymara women. Vellard mentions that there may have been an even worse
drought around 1870. This may also explain the small size of the community

205
in 1896, and the low numbers given in Table 4 of 21 in the year 1877. In the
oral history of the Uru droughts play a prominent role.
However, Aymara linguistic influence in the community does not
date from the droughts in the 1940s. Vellard notes (1949) that at the time of
his first visit in 1938 some men spoke fluent Aymara (which is not
surprising given that the nearest Aymara community, Ancoaqui, is within
easy walking distance). Other men spoke it haltingly and most women did
not speak it at all. Nonetheless, we will see below that Aymara was already
making itself felt in the internal character of the Uchumataqu language in the
1940s, even with speakers not fluent in Aymara like Manuel Inta (Vellard
1949).
8.5 Exogamy
The main reason why Uchumataqu was under Aymara influence even
in the first half of the 20 th century is exogamy. The available data on births
and marriages are summarized in Tables 7 and 8.
Table 7: Ethnicity of the parents of the registered Uru births, 1913-1934
Period Mother Father
1913-22 8 Uru, 2 Aymara 8 Uru, 1 Aymara
1923-34 16 Uru, 20 Aymara 34 Uru

206
Table 8: Ethnicity of the participants of the registered Uru weddings, 1873-
1894 and 1915-1923
Period Wife Parents wife Husband Parents
1873-
1894
1915-
1923
husband
Witnesses Godparents/
6 U 4 UU, 1 UA 5 U, 1 A 4 UU, 1 U 10 U 2 AA
4 U, 3
A
3 UU, 1 U, 3
UA
(U = Uru; A = Aymara; UA = Uru father, Aymara mother, etc.)
7 U 3 UU, 4 U 5 UU, 2
UA
Presenciales
7 AA
These data show that in the early part of the 20th century marriages
with non-Uru partners were relatively rare, while after 1923 their number
increases rapidly.
According to present-day oral sources, marriage with Aymara
women was traditionally prohibited, supposedly because Aymara women
inevitably became ill due to the aquatic activities. At this point in time, the
community subsisted exclusively from hunting and fishing. At one time,
possibly in the 1910s, a member of the Vila family marries an Aymara
woman from across the river. This was around the time that the compound
broke up, and the families resettled in the wider area. Cattle holding became
important, and the cows were being fed totora reeds, a practice which
continues to the present. As the community thrived, more and more young
Uru men married outsiders. Though there have been a number of marriages

207
with Aymara women from neighboring communities, there are no entire
Aymara families living in the community at present.
At the moment of the drought, a substantial minority of the people in
Iruitu were Aymara-speaking. Thus, when the community was reduced to a
handful of people around 1940, at least some of these may have been non-
Aymara speaking. It is even possible that more people managed to survive
the drought in Iruitu with ties to neighboring Aymara communities than
people without such ties, under the assumption that the Aymara economies
were less tied to the water resources of the lake.
8.6 A fragile ecology
The long droughts between 1940 and 1948 have led to the complete
decay of Irohito. In the mid forties the Río Desaguadero was reduced from a
shallow river at least a kilometer wide to a small trickle. People were
deprived of their livelihood, which hinged on fishing, aquatic birds, and
totora reeds for the cattle, and thus faced with the choice of starving or
leaving. Currently, the river has lost much of its fauna, due to chemical
pollution and possibly overfishing.
8.7 The speech community
After discussing the five external factors that may have contributed
to the decline of Uchumataqu, we may now turn to the speech community
itself and its language attitudes. This paper could have been called ‘the bad

208
fortune of a language’ because it shows the vulnerability of language
transmission in a small community. Though the Uru as a group managed to
survive the drought of the 1940s and at present are stronger than ever, they
did not manage to maintain their language, a loss which is now dearly felt in
the community. However, it is not so clear this was felt as a loss at the time.
Nancy Dorian (1998: 3) writes that ‘languages are seldom admired to
death but are frequently despised to death.’ The question is whether the
differences in prestige of Aymara and Uchumataqu contributed to the shift
from Aymara to Uchumataqu in the community. This question is difficult to
answer directly, since we cannot really know what language attitudes
prevailed in the thirties and forties of the previouscentury. (Sheer numbers,
nonetheless, may hold the answer: surrounded by the Aymara people and
language, it would appear a difficult endeavor to cling to Uchumataqu).
We may surmise that the bilingualism there was have been mostly
asymmetrical: speakers of Uchumataqu have been bilingual in Aymara as
well, but speakers of Aymara probably did not know Uchumataqu. The
exception may have been the adjacent community of Ancoaqui, both because
of its frequent contacts with the Uru and because it had an Uru origin itself.
Furthermore, it is clear that Aymara is held in high esteem by its speakers.
However, the ‘people of the lake’, as Vellard (1954) called them, also had a
keen sense of identity, even if they were considered savages by mainstream
Bolivian society. Thus a simple account in terms of differences in general
prestige may not work here; above I have argued that it was at the micro-

209
level of child raising patterns, coupled with ecological crisis, that the
transmission potential for Uchumataqu diminished fatally.
A possibility that could be raised on the basis of the discussion in
Woolard (1989) is that Aymara was used inside of the community more by
those wanting to distinguish themselves sociolinguistically from other, less
socially prominent community members. However, it seems to me that the
more prominent members of the community later took on the role of
guardians of Uchumataqu, as is evidenced from the names of the consultants
of the investigators who visited the community previously, and the way they
are described. This is certainly the situation at present: the dozen or so adult
males most interested in rescuing the language could be viewed like the
miniature version of the Iruitu Rotary Club. This is also in line with a
suggestion made by Woolard (1989: 364) about purism as a possible channel
for intra-community linguistic self-profiling. In addition, we can view this
type of activity in terms of the assertion of internal stratification and power.
9. An actor-centered approach to language death and revival
Although a detailed comparison of the different settings of language
death for all the languages involved still has to be made, a few general
differences between the Uchumataqu setting and many languages in the
lowlands come to the fore.
First of all, the shift was to Spanish, in most of the cases of lowland
languages dying out. Crevels (2002) documents the numerous invasions by

210
Spanish-speaking colonists from the cities and from the highlands, and
progressive deterioration of the original habitats of the lowland Indian
communities. Traditional Indian communities become peopled with
mestizo’s and whites, and Spanish then becomes the dominant language.
Ethnic boundaries become vague and ethnic identities are replaced by local
identities. In contrast, Uchumataqu has lost ground to Aymara.
Only in the case of the Jorá and Pacaguara, the ethnic groups
themselves have also become very small, in the first case due to ethnocide in
the 1950s (Crevels 2002: 26). In most cases, communities remain large, and
become gradually assimilated into mainstream lowland regional culture. The
Uru remain ethnically distinct, even though this is not manifest in their
language.
What Uchumataqu has in common with many lowland Bolivian
languages is the keen interest in its survival by at least some of the speakers.
In the contemporary political setting, in which for the first time in history
Indian political movements play an important role, there is considerable
emphasis nationally on the preservation of the different cultures in Bolivia.
This national interest does not always get transferred to the local level.
However, the continued existence of a distinct native language can play a
role in land claims on the basis of traditional communities. This has
furthered the efforts towards linguistic preservation, after many years of
pressure towards assimilation and neglect. This point to the need for an

211
actor-centred approach, returning to the issues raised at the beginning of this
paper.
In actor-centered approaches, the different participants involved in
the sociolinguistic setting are all explicitly highlighted, in terms of their
immediate interests and more long-term objectives. Actors in the story of
Uchumataqu are many. In Table 9 a first impression is given of the actors
involved, starting with the ‘linguist’, myself, recognizing of course the
difficulty of adopting an autobiographical perspective.
In our documentation project in Bolivia we have to policy to let PhD
students work on languages with a fair speaker number, so that a reasonable
product (minimally a grammar in this particular case, given the requirements
of the funding agency) can be expected. For postdoctoral researchers and
myself as PI we choose languages with higher risk in terms of possible
speaker numbers. So I went to Irohito with an open mind, although of course
with the hope to be able to still record some materials and thus potentially
contribute to a revitalization effort.
In the case of Uchumataqu, the role of the actual principal consultant,
tía Julia, was not a very active one, but the local group of enthusiasts was all
the more so, linking up to the community leadership.
Table 9: Impressionistic sketch of the main actors involved, in the
documentation and potential survival of Uchumataqu

Interests and objectives
Linguist Record materials for academic institution or granting agency
“Informant” , “consultant”,
or “collaborator”
212
Interested but particularly so because inspired by local group of
Uchumataqu enthusiasts
Community leadership Competition for prestige and power in the community
Community Attempts to diversify the economic base of the community
Local teacher Little interest, underpaid, Aymara background
Provincial politicians Orientation on Aymara constituency
Ministry of education Due to low speaker numbers only moderately interested in the
community and language, in spite of official bilingual education
policy
Surrounding communities Aymara communities on whom the Uru were traditionally dependent
Indian political groups /
tribal associations
Growing current importance of Inter-Uru association, but not in period
of fieldwork
NGO’s / foreign embassies Not very active in the area
National academics Interested in documenting and revitalizing but without resources
Eventually it will be interactions of the community leadership with
larger organizations in the altiplano political arena which will determine
whether the currently existing resources for the revitalization of the language
will be brought into the educational programs in a more serious way. When I
visited Iruitu in August 2009, many members of the community had moved
elsewhere for the time being, to La Paz-El Alto, to other parts of lake
Titicaca, or to the lowlands, to find a means of subsistence. Only if the
community remains economically viable and socially intact is there a chance
of revival of Uchumataqu.

APPENDIX 1
213
Overview of the research carried out over the years on Iruitu and
Uchumataqu
Year visit Researcher language and place
publication
nature and size of data
given
1893 Max Uhle German manuscript (Berlin) word list, grammatical
sketch
1901 José Toribio Polo Spanish (Lima) word list, ethnographic notes
1928 Walter Lehmann German manuscript (Berlin) word lists, comparative notes
1931 Arturo Posnansky Spanish (La Paz, Buenos
Aires)
A few words, ethnographic
notes
1931 Alfred Métraux French (Paris) word and phrase list,
ethnographic notes
1933 Enrique Palavecino Spanish (Buenos Aires) Ethnographic notes
ca. 1940 Weston La Barre English (U.S.) Ethnographic notes
1938-51 Jehan Vellard French (Lima, Paris) texts, word lists, grammar
notes, ethnographic notes
+ 1970 Ronald Olson English (Chicago) word lists, comparative notes
Resident Lorenzo Inda Spanish (La Paz), Aymara
manuscript
word list, phrases,
ethnohistory
1978 Nathan Wachtel French (Paris) ethno-history
1995 Colette Grinevald Spanish typoscript (La Paz) word list, orthography
2001-2 Pieter Muysken Spanish typoscript (La Paz) word list, phrases, grammar
notes, orthography

APPENDIX 2
214
Demographic data for Iruitu and Ancoaqui (colonial period). The Aymara
population of Jesus de Machaca is given for the colonial period (Wachtel
1990 : 390) and the 19 th century (Choque 1988)
Date and source Inhabitants of Iru Itu Urus Ancoaqui Aymaran pop. Machaqa
1575 Wachtel 157 650
1685 Wachtel 7 32 429
1721 Wachtel 17 28 240
1724 Wachtel 19 29 237
1767 Wachtel 25 67 347
1786 Wachtel 35 68 359
1792 Wachtel 45 90 375
1797 Choque 55 404
1803 Choque 57 426
1807 Choque 46 355
1817 Choque 52 412
1838 Choque 36 475
1877 Choque 21 534
1893 Uhle 16 men /14 women /9
children
1931 Posnansky 16 families

1933 Palavecino 14 or 15 families
1949 Vellard a dozen people
1951 Vellard 22 adults (16 Uru, 6
mixed or Aymara)
1978 Wachtel 120 members
community, 25 family
heads
2001 Muysken 221 members
2001 Census
References
community
Census 2001: 124 Uru
speakers (87 ethnic
group)
215
Adelaar, W.F.H., in collaboration with P.C. Muysken. 2004. The languages
of the Andes. Cambridge: Cambridge University Press.
Albó, X. 1977. El futuro de las lenguas oprimidas en los Andes. Documento
de trabajo 33. Lima: Universidad Nacional Mayor de San Marcos,
Centro de Investigación de Lingüística Aplicada.
Albó, X. 1995. Bolivia plurilingüe. Guia para planificadores y educadores.
Cuadernos de investigación 44. 2 volumes, Maps, Appendices. La
Paz: Unicef - Cipca.

216
Apaza Apaza, I. 2000. Estudio dialectal del aymara. Caracterización
lingüística de la región intersalar de Uyuni y Coipasa. La Paz:
Instituto de Estudios Bolivianos.
Blommaert, J. 2001. “The Asmara Declaration as a sociolinguistic problem:
Reflections on scholarship and linguistic rights”. Journal of
Sociolinguistics 5: 131-155.
Brenzinger, M. and G. J. Dimmendaal.1992. “Social contexts of language
death”. In Language death. Factual and theoretical explorations with
special reference to East Africa, M. Brenzinger (ed), 3-5. Berlin and
<strong>New</strong> York: Mouton de Gruyter.
Cerrón Palomino, R. 2001. “Bosquejo estructural del chipaya”. Revista
Lengua 12: 87-109.
Cerrón Palomino, R. 2005. El chipaya o la lengua de los hombres del agua.
Lima: Editorial PUCP.
Choque, R. 1988. Historia de Machaca. La Paz, Bolivia: Cipca.
Crevels, M. 2002. “Why speakers shift and languages die: an account of
language death in Amazonian Bolivia”. In Current Studies on South
American Languages M. Crevels, S. van de Kerke, S. Meira, and H.
van de Voort (eds), 9-31. ILLA 3, Leiden: CNWS.
Créqui-Montford, G. de and P. Rivet. 1925. “La langue Uru ou Pukina”.
Journal de la Société des Américanistes de Paris 17: 211-244.
Créqui-Montford, G. de and P. Rivet. 1926. “La langue Uru ou Pukina”.
Journal de la Société des Américanistes de Paris 18: 111-139.

217
Créqui-Montford, G. de and P. Rivet. 1927. “La langue Uru ou Pukina”.
Journal de la Société des Américanistes de Paris 19: 57-116.
Distrito Nacionalidad Indígena Urus de Irohito. 2005. El idioma
Uchumataqu. Manuscript, Irohito, Bolivia.
Dobrin, L., P. K. Austin and D. Nathan. 2007. “Dying to be counted: the
commodification of endangered languages in documentary
linguistics”. In Language Documentation and Description 6, P. K.
Austin, O. Bond and D. Nathan (eds.), 37-52. London: Hans Rausing
Endangered Languages Project.
Dorian, N. C. 1998. “Western language ideologies and small language
prospects”. In Endangered languages. Current issues and future
prospects, L. A. Grenoble and L. J. Whaley (eds), 3-21. Cambridge:
Cambridge University Press.
Heller, M. and A. Duchêne. 2007. Discourses of Endangerment: Ideologies
and Interests in the Defense of Languages. London: Continuum.
Fishman, J.A. 1991. Reversing Language Shift. Theoretical and Empirical
Foundations of Assistance to Threatened Languages. Clevedon:
Multilingual Matters.
Fabre, A. 1995. “Lexical similarities between Uru-Chipaya and Pano-
Takanan languages: Genetic relationship or areal diffusion?”. Opción
11 (18): 45-73.
Grinevald, C. et al. 1995. Curso de uchumataqu. Manucript, Irohito/La Paz.
Guerra Gutiérrez, A. 1984. Oruro. Realidad socio-cultural. Oruro: Cordeor.

218
Hanss, K. 2008. Uchumataqu. The lost language of the Urus of Bolivia. A
grammatical description of the language as documented between
1894 and 1952. Doctoral dissertation, Radboud University Nijmegen.
Leiden: ILLA 7, CNWS Publications.
Haugen, E. 1972. The Ecology of Language. Pasadena, CA: Stanford
University Press.
Hornberger, N.H. 1988. Bilingual education and language maintenance: a
southern Peruvian case. Dordrecht: Foris Publications.
Inda, L. 1985. Historia de los urus. Comunidad Irohito de Yanapata. La Paz:
HISBOL.
Kerke, S, van de. 2000. “Case marking in Leko”. In Indigenous languages of
Lowland South America, H. van der Voort and S. van der Kerke
(eds), 25-39. Leiden: Universiteit Leiden, CNWS.
LaBarre, W. 1941. “The Uru of the Río Desaguadero”. American
Anthropologist 43: 493-522.
Ladefoged, P. 1992. “Another view of endangered languages”. Language 68:
809-11.
Lehmann, W. 1928a. Vocabulario de la lengua uro de Ch’imu. Manuscript.
Berlin: Ibero-Amerikanisches Institut.
Lehmann, W. 1928b. Vocabulario de la lengua uro de Irohito. Manuscript.
Berlin: Ibero-Amerikanisches Institut.
May, S. 2005. “Language rights: Moving the debate forward”. Journal of
Sociolinguistics 9: 313-347.

219
Métraux, A. 1935a. “Contribution à l'etnographie et à la linguistique des
indiens Uru d'Ancoaqui (Bolivie)”. Journal de la Société des
Américanistes de Paris 27: 75-110.
Métraux, A. 1935b. “Les Indiens Uru-Čipaya de Carangas”. Journal de la
Société des Américanistes de Paris 27: 111-127, 325-415.
Métraux, A. 1936. “Les Indiens Uru-Čipaya de Carangas”. Journal de la
Société des Américanistes de Paris 28: 337-394.
Miranda Mamani, L.; Moricio Choque, D.; and Alvarez de Moricio, S. 1992.
Memorias de un olvido. Testimonios de vida Uru-Muratos.
Recopilado por Rossana Barragán (ASUR). La Paz: ASUR/HISBOL.
Molina B., R. and X. Albó C. 2006. Gama étnica y lingüística de la
población boliviana. La Paz: Sistema de las Naciones Unidas de
Bolivia.
Muysken, P. 1996. “Callahuaya”. In Contact languages, S.G. Thomason
(ed), 427-448. Amsterdam: Benjamins.
Muysken, P. 2001. “El Uchumataqu (Uru) de Irohito. Observaciones
preliminares”. Revista Lengua 12: 75-86.
Muysken, P. 2002. “Uchumataqu: Research in progress on the Bolivian
Altiplano”. International Journal on Multicultural Societies 4: 235-
237.
Nación Originaria Uru. 2001. “Kiriwill Qamanak_tan nij_
cheqanchištanpacha tiy wajtha Qalltiniki Uru_ / Estatutos orgánicos

220
y reglamentos de la Nación Originario Uru. Año 0-2000”. Oruro:
Centro diocesano de pastoral social (CEDIPAS).
Olson, R. D. 1964. “Mayan affinities with Chipaya of Bolivia I”.
International Journal of American Linguistics 30: 313-324.
Olson, R. D. 1965. “Mayan affinities with Chipaya of Bolivia II”.
International Journal of American Linguistics 31: 29-38.
Olson, R. D. 1967. “The syllable in Chipaya”. International Journal of
American Linguistics 33: 300-304.
Palavecino, E. 1949. “Los indios Uru de Iruito”. Runa 2: 59-88.
Polo, J. Toribio. 1901. “Los indios Urus del Perú y Bolivia”. Boletín de la
Sociedad Geográfica de Lima 10: 445-482.
Porterie-Gutierrez, L. 1990. “Documentos para el estudio de la lengua
chipaya. Presentación de R. Howard-Malverde”. Amerindia 15: 157-
191.
Posnansky, A. 1915. La lengua Chipaya. La Paz.
Posnansky, A. 1934. “Los urus o uchumi”. Actas del 25o Congreso
Internacional de Americanistas 1932, La Plata. I: 235-300. Buenos
Aires.
Schumacher, A., N. Böcker and F. Condori Mollo. 2009. “El chholo, lengua
de los uru-muratos”. In Lenguas de Bolivia, M. Crevels and P.
Muysken (eds), 00--00. La Paz: Ediciones Plural.
Skuttnab-Kangas, T. and Phillipson, R. 1995. Linguistic human rights.
Overcoming linguistic discrimination. Berlin: Mouton de Gruyter.

221
Swaan, A. de. 2004. “Endangered languages, sociolinguistics, and linguistic
sentimentalism”. European Review 12: 567-580.
Torero, A. 1987. “Lenguas y pueblos altiplánicos en torno al siglo XVI”.
Revista Andina 5: 329-406.
Uhle, M. 1893. Grundzüge einer Uro-Grammatik. Manuscrit. Berlin: Ibero-
Amerikanisches Institut.
Vellard, J. 1949. “Contribution à l'étude des Indiens Uru ou Kotsuñs”.
Travaux de l'Institut Français d'Études Andines, tome I (Paris-Lima):
145-209.
Vellard, J. 1950. “Contribution à l'étude des Indiens Uru ou Kotsuñs”.
Travaux de l'Institut Français d'Études Andines, tome II (Paris-
Lima): 51-89.
Vellard, J. 1951. “Contribution à l'étude des Indiens Uru ou Kotsuñs”.
Travaux de l'Institut Français d'Études Andines, tome III (Paris-
Lima): 3-39.
Vellard, J. 1954. Dieux et parias des Andes. Les ourous, ceux qui ne veulent
pas êtres des hommes. Paris: Éditions Émile-Paul.
Vellard, J. 1967. “Contribución al estudio de la lengua uru”. Centro de
Estudios Lingüísticos, Universidad de Buenos Aires.
Wachtel, N. 1990. Le retour des ancêtres. Les indiens Urus de Bolivie XXe-
XVIe siècle. Essai d'histoire régressive. Paris: Gallimard.
Woolard, K.A. 1989. “Language convergence and language death as social
processes”. In Investigating obsolescence: studies in language

222
contraction and death, N. Dorian (ed), 355-367. Cambridge:
Cambridge University Press.

Linguistic vitality in the Awetí indigenous community: a case from the
Abstract
Upper Xingu multilingual area
Sabine Reiter
The sociolinguistic study, carried out over a period of five years, delineates
the development of the Awetí language, spoken by a small Tupian group in
the Upper Xingu multilingual area in Central Brazil, during a critical
moment when the community split into two.
The empirical analysis on individual language competences and language
use in the Awetí community has revealed a high level of bilingualism with
Kamaiurá, a related language of the Tupi-Guaranian branch. Portuguese,
used as a lingua franca and dominant language of bilingual instruction, plays
a limited role within the community.
The overall pattern of distribution of languages for each Awetí village in
combination with developments from outside may allow some conclusions
about the viability of the Awetí language in the near future.
Keywords: Linguistic vitality, bilingualism, sociolinguitics and
documentation
1. Introduction

224
The Awetí language is currently spoken by a small indigenous
community living in two villages in the national reserve of the Parque
Indígena do Xingu, State of Mato Grosso, Central Brazil. The Awetí people
live in a multilingual area where – together with other ethnic groups – they
have merged into the Upper Xingu cultural system. The multi-ethnicity of
the area is a result of the fact that for more than three centuries indigenous
populations from various regions have been settling there, some of them
after they were expelled from their natural habitats by European colonialism,
others, more recently, after being relocated in the Parque Indígena do Xingú.
This reserve, with an extension of 2.9 Mio hectares, was the first to be
installed in 1961 by the Federal Brazilian Government. At present, this
national reserve is inhabited by representatives of the four major indigenous
language families of Brazil, Arawak (Mehinaku, Waura, Yawalapiti), Carib
(Kuikuru, Kalapalo, Nahukwa, Matipu, Ikpeng) Tupí (Kamaiurá, Awetí,
Kayabi, Juruna) and Gê (Suyá, Txucarramãe/ Kayapó), as well as a language
isolate (Trumai). Of these, all Arawak, the Kuikuru, Kalapalo, Nahukwa,
Matipu, Kamaiurá, Awetí and the Trumai belong to the Upper Xingu cultural
system at the headwaters of the Xingu river in the southern part of the
reserve. This interethnic system is characterized by ceremonial cooperation,
intermarriages and economic interdependencies. Due to a specific language
policy most of the groups have been able to preserve their native languages
which are the primary badge of a distinct ethnic identity within a common
culture. The other ethnic groups who live in the northern areas of the park,

225
also known as “Lower Xingu”, have kept apart from this cultural system (Cf.
Dole 2001: 63). Of the thirteen groups originally taking part in the Upper
Xingu society as registered at first contact in the late 19th century, three are
already extinct, basically due to epidemics or intertribal hostilities. Currently
there are roughly 2.900 individuals in the Upper Xingu, each ethnic group
having between 103 (Matipu) and 509 (Kuikuro) members, living in 18
villages (one to three per group). 22
The following study delineates the development of the Awetí
language in two Awetí villages for the years since 2002 when part of the
population, basically members of one large family of mixed Awetí-
Kamaiurá origin, separated from their community in the main village,
hereafter referred to as “village A”, and built up a new “village B”. The
inhabitants of both places were additionally joined by family members
previously living in a Kamaiurá village and dominant in that language. This
had consequences not only for the linguistic composition of individual
households but for the villages in general. In 2007, with the arrival of the last
two households in village B, the population structure of both places seemed
to have consolidated and an appropriate moment was given for a first
prognosis on the fate of the Awetí language after the division of the
community. The fact that the language is spoken by a small number of
individuals who are concentrated in two villages offered the unique
22 For demographic information see www.socioambiental.org, last confirmed 11/01/2009.
The substantial numbers of different ethnic group members living at indigenous health posts
as well as smaller settlements of one or two houses are not counted as “villages”.

226
opportunity to portray its development in a specific microcosm where any
kind of change from outside or within the group has a perceptible
repercussion on the speech community. 23 By thus meeting the requirements
recently stated by Himmelmann (2009) who postulates the “need for detailed
and fine-grained descriptions of endangerment scenarios in order to arrive at
a useful and comprehensive typology of such scenarios” the study, carried
out over a period of five years, is meant to make a contribution. Although
the overall results will permit to make some predictions about the
continuation of the Awetí language in the near future, such statements
should be regarded as preliminary. Only a longitudinal study of at least
another decade will be able to give more definite answers with regard to the
vitality of this language, given that at present the effects of many outside
factors are only at an incipient stage. The primary aim of this study is rather
to show how both individual in-group as much as general outside
developments may have a major impact on the linguistic behavior of a
group.
2. Data used for the study
The study is based on different kind of material which could be
gathered during participation in the Awetí Language Documentation
23 The few Awetí speakers who live outside these villages could be neglected for the study
since they mostly consist of exogamous spouses (wives) who have little opportunity for a
monolingual communication in Awetí

227
Project. 24 General characteristics of and recent developments in the Upper
Xinguan society as stated by other researchers will be confirmed,
supplemented or contrasted by personal observations during four field-trips
to village A, between July 2002 and August 2005, as well as additional
insights obtained in informal conversation and unstructured interviews with
community members. On the linguistic level, several phenomena have been
discerned from the data obtained at documenting the language and by
personal observation of group interaction. In the empirical analysis, which
can only be summarized here in its most significant results, the influence of
other indigenous languages and Portuguese on the language use in the Awetí
community was investigated. It is based on census data collected between
1998 and 2007 and on a sociolinguistic survey, eliciting linguistic
competences and patterns of language use of 30 representative community
members. 25 Structured interviews by means of a questionnaire were carried
out in May 2003 and June 2004 with 16 male and 14 female inhabitants of
village A. 26 All subjects, born between 1931 and 1991, were chosen among
the adolescent and adult population due to the observation that adolescents
mostly reflect cultural changes, while adults pass on their native languages
24 This DOBES project, under supervision of Prof. Dr. Hans-Heinrich Lieb, was carried out
between 2000 and 2006 by Dr. Sebastian Drude and the author in collaboration with the
Awetí community. See www.mpi.nl/DOBES/projects for detailed information on project
and corpus.
25 A full-length account of the empirical analysis can be found in the Awetí-DOBES-archive
26 Due to the ongoing separation process at that time it was not possible to get permission
from the authorities of village A to visit the new village B or even work with visiting
members of this faction. However, five of the interviewed subjects later moved to village B,
and the missing information for this village could finally be obtained in late 2007 by help of
an unstructured interview with Waranaku Awetí, the school-teacher who had just moved
there and is a well-informed and competent representative of this part of the group.

228
to their descendents. Information on the linguistic behaviour of young
children, the largest group within the Awetí population (cf. figures in the
appendix), could indirectly be obtained from their parent’s answers to the
questionnaire. One of the male subjects was ethnically considered a Trumai,
and one of the female subjects was a Kamaiurá. All other subjects were
Awetí. A combination of census and questionnaire data made it possible to
draw a clear picture of the language use not only of the subject him-/herself
but of whole families and households. All information given by different
members of the same nuclear family or household was highly consistent.
3. Some ethnographic information on the Awetí
When the Awetí were first contacted by the German explorer Karl
von den Steinen in 1887, they were all living in one village which in
Steinen’s opinion, judging from his observation of frequent visits from other
ethnic groups, had a kind of mediator function in the Upper Xingu cultural
system (cf. Steinen 1968: 110, 154). Steinen did not give any number for the
Awetí population he encountered, but in the following decades all Xinguan
groups were severely reduced by several epidemics (cf. Heckenberger 2001:
94. For a detailed account of the population sizes registered by different
authors for some of the years between 1924 and 1996 see Stoeckler Coelho
de Souza 2001: 360). In 1924, a first demographic estimate was given which
numbered the overall Awetí population to about 80 individuals (cf.
Vasconcelos 1945: 70). In 1954, an official census registered only 23 Awetí

229
who had survived an epidemic of measles and were at that time living
together with members of other ethnic groups at the local health post now
referred to as Posto Leonardo (cf. Silva 1972: 357-8, 376).
Today there are more than 160 members of the Awetí ethnic group,
whereby a person whose father or mother is Awetí and who grew up in an
Awetí village and thus speaks the language is usually considered a group
member. 27 The majority of these live in one of two Awetí villages, and few
individuals either at the Posto Leonardo or – if married to members of other
ethnic groups – in their respective villages. The Awetí main village A was
founded in the mid 1950s by three major political factions. In November
2007 it had 92 inhabitants living in eight houses. 28 It is situated at the right
banks of the river Tuatuarí and about six km from the river Curisevo. The
second village B, also at the Tuatuarí, is about 20 km further to the north and
around five km from the Posto Leonardo. It was founded in 2002 by one of
the three political factions which split off from the rest of the population. In
November 2007, 67 individuals were registered in five houses. 29
The villages, like those of the other Xinguan groups, consist of
spacious oval-shaped houses arranged in a circle around a centre which
27 This largely confirms what was stated by Gregor (1977: 307) for the Mehinaku, an
Arawakan group and nearest neighbours of the Awetí. According to his observations, there
were no expressed rules for membership to the ethnic group which rather depends on a
combination of attributes such as language competence, kinship ties, public opinion and
place of residence. Basso (1973: 43/44) observed for the Kalapalo (Carib) that a person is
considered a member of a village group if he/ she were born in that village
28 See figure 1 in the appendix for details. Male and female individuals were graphically
distinguished. The older generations up to the birth year of 1960 were summarized in
columns each covering a decade because their overall number is comparatively low. From
1961 onwards community members were summarized in age-groups covering five years
each, also to account for a proportionally high birthrate in the last few decades
29 See figure 2 in the appendix and previous note.

230
represents the public space where people of all household groups meet and
interact. The houses, private areas which are not entered by “outsiders” to
the household without a specific reason, are often inhabited by ten
individuals or more. An Awetí household normally consists of a single
nuclear family, i.e. a husband, one or – currently in three cases – two wives,
and not seldom seven or more children per woman. In addition, the
husband’s or the wife’s old parents, spouses of the children as well as
grandchildren can form part of the household. As soon as a young couple has
a certain number of children and the required social and economic
conditions, the man usually builds a house for his nuclear family.
Both Awetí men and women nowadays tend to marry at a fairly early
age, women predominantly between 15 and 18, men usually only a few years
older. As is the case with all Upper Xinguan groups, intermarriages are not
infrequent although spouses from the same ethnic group are clearly
considered a preferred choice. Pressure to marry within the group seems to
have increased in recent years. As a result, there are currently several young
Awetí women in village A, well above the age of getting married, who still
live with their parents for lack of an eligible husband in their own
community. Advantageous for the inner unity in village B is a high
possibility for cross-cousin marriages in the younger generations (i.e. for
individuals born in 1980 or later). This kind of marriage arrangement
between the children of siblings of the opposite sex is the most preferred in
the Upper Xingu society. Since the foundation of village B two such

231
marriages, with two young men marrying each other’s sisters, have already
taken place.
By far the most of the non-Awetí husbands and wives are Kamaiurá,
the only other Tupian group belonging to the shared cultural system. 30 In
2007, this applies to five women in village A and three men and two women
in village B. Considerably less frequent are marriages with Mehinaku,
Yawalapiti, Trumai and Kuikuro. In 2007, an Awetí woman from village B
got married to an Ikpeng. This is very unusual since members of this Carib
group from the northern part of the reserve do not normally participate in the
intermarriage system of the Upper Xinguan people. As the most common
pattern it can be observed that intermarriages are patrilocal, i.e. husbands
from other groups live in their in-laws’ houses for some time before –
together with their wives and children – returning to their own villages (cf.
Gregor 2001: 190). In the Awetí community, too, permanent members from
other ethnic groups are typically women.
Traditionally, all the groups participating in the Upper Xinguan
society – except for the Trumai and the Yawalapiti 31 – follow a strict
language policy which contributes to keeping the languages apart up to the
individual level. According to this policy, the language spoken in the village
and acquired by all children as a first language is that of the ethnic group the
30 Franchetto (2001: 141) observed for the Kuikuro the same tendency of preferably
marrying within one’s own group and – in second place – with other Carib peoples.
31 These two groups have mostly given up their own languages and adopted one or several
of the others. In the specific case of the Trumai, there is a strong tendency for language shift
towards Portuguese (see Franchetto 2001: 139/140, Guirardello 1999: xvi)

232
village population belongs to. Other-ethnic spouses continue to speak their
own language and acquire a passive knowledge of the community’s
language. In personal interethnic communication among couples, or among
other people within and outside the village each person uses his/her own
language or, more recently, Portuguese. 32 Children of mixed couples –
especially when it is their mother who has a different linguistic background
– grow up bilingually, speaking the language of the other ethnic group with
the respective parent inside the house. In contact with members of the group
their other-ethnic parent or close relative belongs to, such bi- or multilingual
individuals may adopt the role of interpreters. The Upper Xinguan language
policy, as could be confirmed by the sociolinguistic survey (cf. section 4.3),
is still valid for the Awetí.
In comparison to other Xinguan groups, the Awetí were living in
relative isolation during the second half of the 20th century. One reason can
be seen in the peripheral geographic position of their main village A which
can only be reached on narrow, frequently overgrown paths of many
kilometres through the forest. Another reason is their low number and
consequently not very powerful position in the multiethnic system (cf. Seki
2000: 34). This position may be further weakened by the absence of inner
unity, expressed in the recent division of the village 33 and an ongoing
32 In non-personal interethnic communication during intertribal ceremonies interaction takes
place predominantly via a system of non-verbal codes shared by all participants and an
extremely formalized use of the language (cf. Basso 1973a: 8).
33 Heckenberger (2001: 102) states that divisions of villages of a certain size into two are not
unusual in the Upper Xingu and have occurred with frequency after the demographic
increase of the last few decades. The latest example is that of the Mehinaku who

233
disagreement between the two remaining political factions in village A who
both have a legitimate claim for leadership of the group. In 2007, tensions
within the group had reached a critical stage with the attempt of an
impeachment process against the current chief who, as a consequence, left
village A in 2008. There is not yet a legitimate successor due to the refusal
of leading authorities from the other Xinguan groups to accept the young
Awetí candidate from the opposite faction.
Although contact with the outside world has increased in the last
decades, most of the inhabitants of the Awetí villages continue in the
traditional Xinguan way. They live on slash-and-burn subsistence
agriculture, basically cultivating bitter manioc, and fishing. Industrial food
plays an inferior role since most Awetí lack the financial resources to buy it.
However, other industrial products, mostly for fishing, hunting and cooking
purposes, have been introduced to the community and the demand is
constantly increasing. In more recent times more sophisticated technical
equipment (television sets, radios, DVD-players, shortwave radio sets,
electricity generators) have been added to this list. As their means of
transport the Awetí nowadays use aluminium boats with two two-stroke
outboard motors, two Toyota off-road vehicles 34 and a considerable number
experienced a split in early 2004. According to Heckenberger’s observations, however, such
divisions usually happen in villages of 200 inhabitants and more, which is neither for the
Awetí nor for the Mehinaku the case. In both cases conflicts with regard to group leadership
and distribution of resources from outside among the factions have been primary reasons for
the separation.
34 One Toyota currently circulates between the towns of Canarana, Gaúcha do Norte and the
bank of the Curisevo river where the Awetí leave their boats when coming to town. The
other moves between the Awetí harbour and the village. There are roads between indigenous

234
of bicycles. In order to refill their supplies, to fuel their engines and to
maintain their technical equipment, the community is in constant need of
monetary resources. Very little of this is covered by material and financial
support from the Federal Government (FUNAI). For additional funds, the
Awetí, organized in an association, have been participating in a couple of
projects, stimulated and financed by NGOs. Few members of the community
have a regular income paid by the Brazilian government: two primary-school
teachers and health assistants, one in each village, and an environmental
agent who is responsible for waste disposal and for the solar-driven pump of
the well in village A. 35 These positions are very prestigious and therefore
given to male group-members of a higher rank in the internal hierarchy. The
major source of income, however, continues to be the selling of traditional
cultural artefacts (hammocks, baskets, stools etc.) in town.
The Awetí, in general, do not need to work outside their native
villages, even though a few of the men have occasionally done so, working
as guides to the area or as hired farm workers. Neither does anyone pursue a
career outside the village, presumably due to the fact that a comprehensible
near-standard variety of Portuguese is almost exclusively spoken by group
members under the age of thirty and that literacy is only slowly spreading
among young adults since the late 1990s, when the bilingual village primary-
villages in the Xingu Park but none yet permanently connects the Awetí village A with any
other village.
35 Most of these earn a “salário mínimo”, a minimal salary which at the moment is at around
400 R$ i.e. about 150 € a month. The two school-teachers earn more (Drude 2008, personal
communication).

235
school was started by two young Awetí men who, in the same period, had
begun to participate in a teacher’s training course organized by the non-
governmental organization Instituto Socioambiental (ISA). 36 Apart from the
village chief who as an adolescent had been taught to speak and write
Portuguese by military officials at the Posto Leonardo, these two were the
first members of the community to receive a school education. Both finished
their training in 2003, and since 2005 one of them has been continuing his
studies at the State-University of Mato Grosso, where he was joined by his
colleague in 2008.
4. The Awetí language
The Awetí language, according to recent analyses (cf. Drude 2006),
is the only known representative of a branch of its own within the Tupí
stock. It is genetically most closely related to the languages of the Tupí-
Guaraní branch and to Mawé, spoken in the states of Amazonas and Pará
and equally representing a branch by itself.
4.1 Lexical influences from other languages
There are various phenomena in the Awetí language which are
arguably contact-induced. A feature which has not been reported for many
other Tupian languages, although not unknown in Amazonian languages, is
36 Since 1994, ISA has been organizing training courses for members of the Xinguan
indigenous communities in which these are trained to become primary-school teachers (see
www.socioambiental.org). The current two teachers, Awajatu and Waranaku Awetí, started
to participate in these courses in 1997.

236
the existence of genderlects (cf. Drude 2002). A male and a female variety of
the Awetí language are distinguishable by few features such as different
forms for personal and deictic pronouns, demonstratives, discourse particles
and a very small number of lexical items. One may hypothesize that the two
varieties are residues of two closely related Tupian languages. This idea is
supported by the Awetí’s oral history which informs that their ancestors used
to live in two villages; one inhabited by a people called Enumania, the other
by the “Awytyza ’ytoto”, the “real” Awetí. These were closely related by
matrimonial ties. According to a historical narrative captured in the DOBES
corpus, the inhabitants of the Awetí village one day were all killed by the
Tonoly. 37 Therefore the Awetí of today consider themselves descendents of
the surviving Enumania village which was possibly inhabited by a number of
ethnically mixed couples, due to patrilocal intermarriage possibly consisting
of Enumania men married to Awetí women. 38
Apart from these minor differences between male and female speech
which may be the result of language contact, 39 the Awetí lexicon contains
some words from other Upper Xinguan languages. These, on the one hand,
often denominate elements of a shared cultural environment (e.g. kalahang
37 There is evidence in the narrative that these, too, were of Tupian origin.
38 According to genealogical information presented by Stoeckler Coelho de Souza (2001),
most Awetí of today are descendents of the Enumania. The same author (2001: 371), based
on information from a Kuikuro and an Awetí oral narrative, estimates that the Awetí have
joined with the Enumania before 1750, before establishing themselves in the Upper Xingu
region.
39 Bastos (1989, cited in Franchetto 2001: 121/2), whose reconstruction work on the
Xinguan peoples is based on the Kamaiurá oral history, draws up an even more complex
contact situation. He states that two more Tupian groups, the Ka’atyp and Wyrawat,
together with the Enumania have merged into the Awetí. Seki (2000: 34), on the other hand,
claims that the Ka’atyp were one of the five sub-groups of the Kamaiurá

237
‘pineapple’ and majãku ‘certain type of basket’ both of Arawakan origin (cf.
Stoeckler Coelho de Souza 2001: 392, footnote 6), or Keri, ‘moon’ in
Arawakan, which in Awetí has turned into the proper name of the
protagonist Moon in a narrative). On the other hand, it frequently occurs in
less careful speech that bilingual speakers of Awetí and Kamaiurá use a
Kamaiurá word instead of its equivalent in Awetí. Examples from the
DOBES corpus, pointed out by Awetí consultants, are:
(1) (kal_kamukuaka)
Pyw w=e-tigaw-ut tororororo,
IDEO-take 3coref=REL-stool-PAST IDEO-go.down.throat,
n=ewizako mo'at tororororo, apykap tororo tororo tororo tororo.
3=after person IDEO, stool IDEO-swallow
‘He took (what used to be) his own stool, it went down its throat, after that a
person, it went down its throat, a stool, [the beast] swallowed it [all].’
(2) (awu_fishing)
Tiwk wakupa a'yt a-mokym tiwk.
IDEO-catch corvine PART 1sg-catch IDEO
‘I caught a corvina.’

238
In (1) both the Awetí word tigap as well as the equivalent Kamaiurá
word apykap for ‘stool’ are used in one sentence. In (2) the Kamaiurá word
wakupa is used instead of Awetí tywamunuryt to refer to a species of fish.
The first process has been described by Franchetto (2001: 147) as a
restricted linguistic interference on the word-level, amplifying or modifying
the lexicon, which is observable in most Xinguan languages. The second
process is an instance of individual code-mixing which – given the high
percentage of Awetí-Kamaiurá bilinguals – has a potential of influencing the
language use of the group as a whole.
While few nouns and even less items of other word classes from
related or unrelated Upper Xinguan languages have found their way into
Awetí, it could be observed that a number of ideophones and interjections
seem to be very similar or even identical across Upper Xinguan languages. 40
For sixteen Awetí (Aw) ideophones correspondences could be found in
Kamaiurá (Kam), Waurá (Wa) and/or Kalapalo (Kal). Examples are:
(3)
‘blow’
powder’
(a) fuu fuu (Aw, Kal)
(b) pii pii pii (Aw) ‘sow corn’, pyy pyy pyy (Wa) ‘throw
40 Examples are from Basso (1995), Guirardello (1999: 226/7), Richards (1977, 1988) and
Seki (2000: 104ff.). The orthography of Waurá and Trumai was adapted to Awetí and
Kamaiurá conventions

ground’
give’
‘walk’
‘arrive’
239
(c) pok (Aw) ‘put to ground’, mbuk (Kam) ‘fall to
paky (Wa) ‘put’, bok, puk, pokï (Kal) ‘put,
(d) tyk tyk tyk (Aw, Kam), tytytytyty (Wa), ti ti ti ti (Kal)
(e) tyky (Aw, Kam), tiki (Kal)
Nine of the interjections used in Awetí occur in the scarce data
available on other Upper Xinguan languages, whereby all of the interjections
listed for Kamaiurá also occur in Awetí. With Trumai Awetí shares atsy, an
expression of disgust, and ko, indicating ignorance of the speaker. In
Kalapalo as well as in Awetí teh is used for the admiration of beauty.
Comparative analyses between Xinguan languages based on large corpora of
linguistic data – currently only at an incipient stage – will probably yield
further correspondences resulting from language contact. 41
A more recent lexical feature reflecting ongoing cultural changes
within the Awetí community can be observed in the language use of
adolescents and young adults. These tend to address their siblings by their
proper names instead of a kinship term as used to be common practice and as
41 One initiative consists in the cooperation of the three former DOBES projects Awetí,
Kuikuro and Trumai, also including data from the languages Waurá, Mehinaku and Bakairi.
The analyses resulting from a symposium held in Rio de Janeiro in 2008 are to be published
online at www.museunacional.ufrj.br.

240
is still done by younger children and older people. In general, the use of
“apelidos”, kind of nicknames, often Portuguese-based, is widespread among
community members. This is in order to avoid the pronunciation of proper
names traditionally accompanied by a complex system of restrictive rules. 42
With regard to the use of more specialised vocabulary in Awetí (names of
fish, insects, plants, geographical places, etc.) it was noticed that this is not
necessarily part of common knowledge but rather restricted to the lexicon of
older and more traditional (male) speakers.
Up to the present the Awetí language shows a very low number of
loanwords from Portuguese which are integrated into the Awetí linguistic
system to varying degrees. They are primarily used to fill in the lexical gaps
in newly introduced contexts: white men’s clothing, counting, time-related
concepts (days of the week, months, etc.), certain food items (milk, coffee,
rice), and technical terms. The examples in (4) show different stages of
integration into the Awetí language. They are all taken from the DOBES
corpus which means that they also occur in relatively careful speech where
speakers were aware of the recording situation.
(4)
(a) (a)-lata-’jyt (< port. lata ‘tin’ + aw. diminutive)
(b) kahawa (< port. garrafa ‘bottle’; aw. kujãkomyt)
(c) meio dia (port. ‘midday’; aw. apyterype )
42 Another strategy to avoid a relationship term is the usage of a Portuguese kinship term
instead. This was already observed for the Kalapalo by Basso (1973a: 12ff.) in the 1960/70s.

241
Example (a), referring to a tin (or plastic) cup, is in widespread use
among the Awetí. In addition to the Awetí diminutive suffix -‘jyt the word is
optionally pronounced with a vowel at the onset of the first syllable. This
corresponds to the Awetí lexicon which does not have any noun starting with
a lateral. The second example consists of a Portuguese term transferred into
the Awetí phonological system. For this process there are many examples,
whereby these words most often refer to items from the industrial society
which were introduced into Awetí daily culture some time ago. The third
phenomenon shows items which are pronounced in a Portuguese-like
fashion. Apart from more modern technical terms this applies most often to
time concepts and numbering. There is a counting system in Awetí but this,
in comparison to Portuguese, is neither elaborate nor economical. Portuguese
quatro (‘four’) thus is preferred to mokõj mokõjput, according to a young
Awetí consultant, who claimed that only the first three numbers in Awetí are
still used by everyone. Money, on the other hand, is entirely counted in
Portuguese. <strong>New</strong> words have been coined for few items of the modern
world, such as kwaza’jyt (‘the little bright / intelligent one’) for a recording
device. Among the older generation, however, this phenomenon seems to
have been more frequent. An example is given by the preferred term to refer
to a torch: While alatena, derived from Portuguese laterna ‘lantern’, is
common among younger speakers, older people still use a kind of loan
translation, taty a’ang (‘image of the moon’).

242
A difference with regard to the pronunciation of loanwords could
also be observed between older and younger people, especially women.
While older women were using words phonologically adapted to their own
language, these words were pronounced with a Portuguese phonology by
younger women with equally little command of that language.
The Awetí have preserved a positive attitude towards their language
which is their primary means to express their ethnic identity in a multiethnic
environment. Full proficiency in Awetí is highly valued and – according to
its speakers – only reached at an advanced age if at all. It can be measured
by the negative reactions with regard to code-mixing observed on many
occasions during the recording and analysis of recorded speech in the
DOBES Awetí Language Documentation Project. All Awetí speakers seem
to be well aware of Kamaiurá loanwords. Whenever a speaker used a word
identified as Kamaiurá, consultants tended to find excuses for this such as
“he has been living among the Kamaiurá for too long”. Speakers with a
strong tendency towards Portuguese loanwords or expressions were
considered non-proficient and ridiculed.
A reflection of the Awetí’s concern to keep their language alive in all
its functional domains, was that the major part of the adult population, in
particular those with traditional knowledge, were eager to pass it on and
actively participated in building up a vast corpus of linguistic and cultural
data during the DOBES project. In turn, the data, made available to the
community by recordings on tape and CD and especially as material for

243
instruction, may have a positive effect on the attitude of younger community
members who could generally be observed to be more indifferent towards
indigenous traditions. For transcription purposes during the documentation
project several young speakers were trained to become fluent writers of
Awetí. The impact of other products of the documentation project on the
speakers’ linguistic competences and attitude towards their language may
only be perceivable in the future. A revised orthography, adapted to recent
research results and elaborated in close cooperation with the two Awetí
teachers, could help to enhance the status of Awetí as a language with a
written standard (see Drude, S., Awete, W. and Awetí, A. ined.). A concise
grammar and an Awetí-Portuguese dictionary, currently in preparation, as
well as various other analyses may further contribute to heighten its prestige,
especially in comparison to the national language. 43 The archive itself may
probably only be of interest to the speech-community in the remote future
but the documentation project, nonetheless, has already had a positive effect
on the group’s position in the multiethnic system. No specific attitude could
be observed with regard to Kamaiurá, the other dominant indigenous
language. It is spoken by those who are entitled to use it and respected by the
others, like any other language belonging to the Upper Xinguan system. The
fact that the Kamaiurá community has three times as many members as that
of the Awetí and that their language has been exhaustively studied does not
43 An “emblematic function” of written material as a means to “bolster the prestige of the
indigenous language in the community” was perceived by Terrill (2002: 216) during her
research among the Lavukal people on the Solomon Islands.

244
seem to make it more prestigious in the perception of the Awetí or influence
their linguistic orientation. 44
4.2 Individual linguistic competences
The empirical study explores to what extent other indigenous
languages, brought in by interethnic marriages, and Portuguese as a contact
language with the national society are known and used in the Awetí
community.
A combination of census and questionnaire data confirmed that most
Awetí living in either of the two villages are of mixed ethnic origin. Among
the 30 community members participating in the sociolinguistic survey only
one old woman declared to have had four Awetí-speaking grandparents.
Seventeen subjects were of mixed descent in their parents’, the rest in their
grandparents’ generation. In order to illustrate the ethnic composition of
each household in village A and B in 2007, the ethnic and linguistic origin of
the parents and grandparents of the nuclear couples of the respective houses
are listed in table 1 and 2.
44 A grammar of Kamaiurá was published by Seki (2000), and Moore (2007: 34/35) rated its
linguistic description with “3”, meaning (almost) complete, whereas the study of the Awetí
language was rated with “1” on a scale from 0 to 3, relating to the publication of several
articles and a master thesis. The imminent publication of the DOBES corpus as well as other
analyses by members of the project team may change this rating within the next few years.

Table 1: Linguistic origin of the nuclear couple of each household in village
A*
245
Household ♀ ♀♀ ♀♂ ♀♀♀ ♀♀♂ ♀♂♀ ♀♂♂
1 Kam. Meh. Kam. Yaw. Meh. Meh. Kam.
3 Aw. Aw. Kui. Yaw. Aw. Kui. Kui.
4 Sep.
7 Aw. Aw. Aw.(Tru.) Yaw. Aw. Aw.(Tru.) Aw.
8 Aw. Aw. Kam. Aw. Aw. Aw. Kam.
9 Aw. Kam. Aw. Kam. Kam. Aw. Aw.
10 Aw. Yaw. Aw. Yaw. Yaw. Kam. Aw.
12 Aw. Aw. Aw.(Tru.) Yaw. Aw. Kam. Aw.(Tru.)
Household ♂ ♂♀ ♂♂ ♂♀♀ ♂♀♂ ♂♂♀ ♂♂♀
1 Aw. Aw. Aw. Aw. Yaw. Yaw. Aw.
3 Aw. Kam. Aw. Kam. Kam. Aw. Aw.
4 Aw. Kam. Aw. Kam. Kam. Kam. Aw.
7 Aw. Meh. Aw. Meh./Wa. Wau. Kam./Aw. Aw.
8 Aw. Kam. Aw. Kam. Kam. Aw. Aw.
9 Aw. Aw. Kam. Aw. Aw. Aw. Kam.
10 Aw.(Tru.) Kam. Aw.(Tru.) Kam. Kam. Aw.(Tru.) Aw.
12 Aw. Aw. Aw. Kam. Aw. Aw. Kam.
Table 2: Linguistic origin of the nuclear couple of each household in village
B
Household ♀ ♀♀ ♀♂ ♀♀♀ ♀♀♂ ♀♂♀ ♀♂♂
1 Kam./Aw. Aw. Kam. Yaw. Aw. Kam. Kam.
2 Kam. Kam. Kam. Kam. Kam. ? ?
3 Aw. Yaw. Aw. Yaw. Aw. Aw. Aw.
4 Aw./Kam. Aw. Kam. Yaw. Aw. Kam. Kam.
5 Aw. Yaw. Aw. Yaw. Aw. Kui. Aw.
6 Aw. Aw. Aw./Kam. Aw. Aw. Kam. Aw.
7 Aw./Kam. Aw. Kam. Yaw. Aw. Kam. Kam.
Household ♂ ♂♀ ♂♂ ♂♀♀ ♂♀♂ ♂♂♀ ♂♂♀
1 Kam. Kam. Kam. ? ? Kam. Kam.
2 Aw./Kam. Aw. Kam. Yaw. Aw. Kam. Kam.
3 sep.
4 Aw. Meh. Aw. Meh./Wa. Wau. Kam./Aw. Aw.
5 Kam. Kam. Kam. Kam. Kam. Kam. ?
6 Aw./Kam. Aw. Kam. Yaw. Aw. Kam. Kam.
7 Aw. Aw. Aw.(Tru.) Yaw. Aw. Aw.(Tru.) Aw.

246
* The symbols and abbreviations are as follows: ♀♀ refers to “wife’s mother”, ♀♂ “wife’s
father”, ♀♂♀ “wife’s father’s mother” etc. The languages spoken by these individuals at the
same time determine their membership in the respective ethnic group: Awetí (Aw.),
Kamaiurá (Kam.), Kuikuro (Kui.), Mehinaku (Meh.), Waurá (Wa.), Yawalapiti (Yaw.).
Those, for whom two languages are listed, are of mixed ethnic origin and have spent part of
their lives in both places. They are dominant in the first of the two languages. Some
individuals are ethnically considered Trumai but their dominant language was claimed to be
Awetí: Aw. (Tru.). The “gaps” in the numbering of households in village A indicate the
households which in 2003/4 had not yet moved to village B. The male head of household 4
in village A and the female head of household 3 in village B used to form a couple in village
A before the separation of the community. The last two households of village B, formerly
household 5 and 11 in village A, in 2007 were temporarily living in house 5.
The tables demonstrate that in every nuclear family, with the
exception of one in village A and a split household in village B, either
husband or wife has a non-Awetí parent, whereby the overwhelming
majority of these are Kamaiurá. 45
The sociolinguistic questionnaire further explored whether a
multiethnic family background had an influence on individual linguistic
competences and language use. Table 3 summarizes the individual
competences in Xinguan languages as stated by the subjects.
Table 3: The subjects’ competences of indigenous languages in 2003/4
First language** Second language No competence
Active Passive
45 In village A this applied to half of the men and women, in village B to eight out of 14,
equally distributed over the houses. The individuals with the same ethnic composition in
village B (♀1, ♀4, ♀7, ♂2, ♂6) are daughters and sons of the founders of the village who
themselves are listed as the nuclear couple of house 5.

Awetí 25 (+1) 2 2 -
Kamaiurá 5 (+2) 13 5 5
Mehinaku (+1) 1 3 25
Waurá - 1 3 26
Yawalapiti - 2 4 24
Kuikuro - - 1 29
247
** “First language” here refers to the first language a child uses in communication.
Numbers in parentheses indicate that the subject had more than one first language, usually
that of the mother and that of the surrounding village, but at the time of the interview was
only using the other one.
It was differentiated between “active” and “passive” competences,
the former meaning that the language was used in communication, the latter
that it was understood but not spoken. A general observation is that
individuals rate themselves as competent speakers of another indigenous
language, when they feel entitled to use it due to appropriate kinship
relations and when their competences are well-developed indeed (see e.g.
Gregor 1977: 30; Franchetto 2001: 141-2). The answers to the questionnaire
suggest that individuals living in the Awetí community did not develop an
active competence in a non-native language, even if they had been exposed
to it since an early age, if neither their parents nor grandparents belonged to
that ethnic group.
Kamaiurá clearly stood out as a second language. It was actively
spoken by two thirds of all subjects, including seven who acquired it as a
first (or an additional first) language. Only five subjects, all of them females,
declared to have no knowledge of that language at all. The linguistic

248
influence from the other Xinguan groups was not significant due to the small
number or absence of native speakers living among the Awetí.
With regard to the subjects’ knowledge of Portuguese a combination
of two variables seemed to be relevant: sex and age. In the older generations,
men more than women had access to a reduced contact variety of this
language which they acquired in an uncontrolled way. Since the mid 1990s,
younger people of both sexes have been learning a near-standard variety of
Portuguese at school. In 2003/4, literacy was developed to differing degrees
among these subjects. Seven male subjects rated themselves as having an
active and seven as having a passive competence. Among the female
subjects only four claimed to be able to speak the language and another three
to fully or partially understand it. Portuguese was considered an important
means of communication, especially with non-indigenous people, by
younger community members of both sexes. Most men were highly
motivated to know the language well, whereas women and old people
showed considerably less interest. Some of the younger men and women,
mostly born in the 1970s, without access to school-education expressed their
regret for not having been able to learn the language systematically.
4.3 Patterns and domains of language use
Two levels of results – an individual and a “communal”, i.e.
regarding the community as a whole – need to be distinguished for the part

249
of the questionnaire which deals with language use within the family, as well
as in and outside the villages.
On an individual level the results of the study in many respects
reflected the traditional language policy of the Upper Xinguan area (cf.
Basso 1973a: 6-7; Gregor 1977: 30; Franchetto 2001: 141-142). All
interviewed subjects, independently of whether they grew up in an Awetí or
Kamaiurá village, declared to have acquired the language of the village they
grew up in as a first language, and in 2003/4 this was also true for all
children in village A. 46 The traditional rule, that adults from other ethnic
communities who come to live in the Awetí village are not “entitled” to
speak the community’s language, in 2003/4 still seemed to be intact, too.
This applied to one subject of the questionnaire and three further women
who were mentioned indirectly as mothers of subjects. They had merely
acquired a passive knowledge of the village’s language. Among interethnic
couples, usually, each continues with his/her own language while
understanding that of the other. In communication with Awetí monolinguals,
other-ethnic members of the community claimed to be assisted by
interpreters of mixed-ethnic origin. Only those other-ethnic spouses who had
one Awetí parent were allowed to speak this language. 47 A possible
innovation in this respect was reported in 2007 for a young Awetí-Ikpeng
46 The children of one household, however, were growing up with two first languages, one
being the language of the village, the other Kamaiurá spoken in the house.
47 This was explicitly confirmed in 2007 by one of the consultants who explained that a
woman in his village was “pure Kamaiurá” and therefore only entitled to speak that
language, whereas a young man, who had previously been living among the Kamaiurá was
of mixed Kamaiurá-Awetí origin and now also speaking Awetí in village B to which he had
migrated (Waranaku Awetí, personal communication).

250
couple in village B, who were using Portuguese as a lingua franca among
themselves. Such tendencies, however, must be further observed in order to
find out whether this is a temporary feature, valid until the spouses have
developed passive competences in each other’s languages, or whether it will
eventually replace the traditional language policy for interethnic marriages.
The bilingualism of a subject with parents or grandparents from a
different ethnic group was either actively or passively developed, depending
on the quality and amount of input from this language the person
experienced throughout his/her earlier childhood as well as a specific family
history or even individual preferences. In the case of subjects who never left
the Awetí village they grew up in and whose non-Awetí parent or
grandparent living in their household was the only member from that group
or one of a few within the community this resulted in passive bilingualism
where the other ethnic parent spoke his/her language, and the child replied in
that of the surrounding village. This applied to all languages except
Kamaiurá. Due to the presence of “pure” Kamaiurá and mixed-ethnic
individuals in the community there was more opportunity to speak this
language. Several subjects who had migrated from a Kamaiurá- to the
Awetí-village A in early childhood reported to code-switch when speaking
with Kamaiurá or bilingual community members. Siblings among these
subjects either continued to use Kamaiurá with each other, speaking Awetí
with younger siblings born in the village, or completely shifted to Awetí. In
all but one case, however, Awetí had become the dominant language of

251
communication whereby migration at an early age seemed to have been a
decisive factor. 48 Another relevant variable is the sex of the speakers.
Daughters who spend more time in the house in the company of their other-
ethnic mothers or grandmothers tended to have a more developed
competence in this language than sons.
On a “communal” level the active or passive bilingualism of most
members of the Awetí community in combination with an increasing number
of Kamaiurá migrants to both places had changed the proportions of
language use in individual households. The degree of individual bilingualism
and language-dominance in each household and the resulting overall pattern
of distribution for each village may allow some conclusions about the future
vitality of the Awetí language. In village A, in 2003/4 only in one of the
houses Kamaiurá was the dominant household-language. This situation had
become even more pronounced after the arrival of nine relatives from a
Kamaiurá village of whom only one adult was an active, and possibly three
adults were passive bilinguals with Awetí. In 2007, all three mothers of the
eleven children in this house did not speak Awetí, and another young
Kamaiurá woman was expecting her first child. In only one of the remaining
seven houses Kamaiurá was actively spoken by one Kamaiurá household-
member and – in interaction with her – by some of her children and
grandchildren. In the other houses Awetí was the only active language,
48 This applied to four subjects from three different households who had left their Kamaiurá
environment between the ages of three and six. An older brother of one of them, who
arrived in the Awetí village at the age of 13, even ten years later had not yet developed an
active competence in this language, even though he had an Awetí father.

252
although a considerable part of the adolescent and adult population was
capable of understanding Kamaiurá.
Table 4: Dominant language per household in village A in November 2007
Village A Household members Languages used Dominant language
Household 1 12 adults***, 11 children Kam., Aw. Kamaiurá
Household 3 4 adults, 5 children Aw., Kam. Awetí
Household 4 3 adults, 3 children Awetí Awetí
Household 7 7 adults, 5 children Awetí Awetí
Household 8 9 adults, 7 children Awetí Awetí
Household 9 5 adults, 4 children Awetí Awetí
Household 10 4 adults, 3 children Awetí Awetí
Household 12 3 adults, 7 children Awetí Awetí
*** Individuals born in 1992 and earlier were registered as “adults”.
In 2008, the Kamaiurá-dominant household, holding 25% of the
village population, separated from the other seven houses to build a new
village in the vicinity. Only the old Awetí-speaking father of the head of
household and current Awetí chief remained in village A. He had in fact
been the only person in the household with whom all active bilinguals had
been speaking Awetí. In this future village the public language under current
conditions will undoubtedly be Kamaiurá. One may further hypothesize that
the younger children of this household who used to speak Awetí outside the
house will stop to do so when living in a village where there are no other
Awetí-speaking children.
In village B the dominant language varied from household to
household, depending on whether its members had spent most of their
childhood and adult lives in the Awetí or in the Kamaiurá community. In

253
2007, of the five houses only two were Kamaiurá-dominant whereby the
number of household-members in both cases was proportionally high –
together amounting to 46% of the village population. Virtually all of these
were active or passive bilinguals. 49 By contrast, there were two houses where
Awetí is the only language spoken. All-but-one of the inhabitants neither
spoke, nor did they understand Kamaiurá very well. In the fifth house the
only Kamaiurá household-member had a passive competence in Awetí, his
wife and two of his adult children were bilinguals with an Awetí bias, and a
young Ikpeng son-in-law was communicating with the inhabitants of the
village in Portuguese. In table 5, the two recently-arrived households 6 and
7, at that time living in house 5, are listed separately because they used to
live in their own houses in village A. All children growing up in either of
these two households in 2007 were Awetí monolinguals. For 2008, both
households were planning to rebuild their own houses, together with three
young families from the Kamaiurá-dominant households. In these three new
houses it remains to be seen which will become the dominant household-
language, since at least in two of them one part of the nuclear couple has
spent a major part of his/her childhood in the Awetí village A. A factor
favoring the continuation of Awetí as the primary means of communication
is that in this village in 2007 all Kamaiurá-speakers had an active or passive
competence in Awetí, whereas a minimum of nine adults and 17 children
were monolingual Awetí speakers.
49 In household 2 a young mother of a girl was the only monolingual Awetí speaker.

Table 5: Dominant language per household in village B in November 2007
254
Village B Household members Languages used Dominant language
Household 1 7 adults, 11 children Kam., Aw. Kamaiurá
Household 2 7 adults, 6 children Kam., Aw. Kamaiurá
Household 3 5 adults, 3 children Aw., Yaw. Awetí
Household 4 4 adults, 6 children Awetí Awetí
Household 5 5 adults Aw., Kam., Port. Awetí
Household 6 2 adults, 6 children Awetí Awetí
Household 7 2 adults, 3 children Awetí Awetí
Most public communication among community members in both
villages takes place in Awetí. For village B it was reported in February 2008
that five adults of the entire population were exclusively or predominantly
using Kamaiurá in the public sphere (Drude personal communication).
Except as a contact language, used in town or with visitors to the
village, Portuguese has restricted functions in the Awetí linguistic
community. It must be emphasized, however, that these contact situations
have become much more frequent in the last few years, Awetí men travelling
to the nearest-by Brazilian towns (Gaúcha do Norte, Canarana) on a regular
basis to sell their artefacts or attend meetings with other Xinguans, and
representatives of the outside society (e.g. nurses, researchers, members of
NGOs, etc.) coming to the villages and sometimes staying for several weeks.
In neither of the two villages Portuguese plays a role in the domains
characterizing the use of indigenous languages, except at school which has
already had an impact on the language use and attitude of the younger
generation. Many of these, but men more frequently than women, are

255
making use of it in interaction not only with people from outside the reserve
but also with young Xinguans from other communities.
In spite of the fact that bilingual education is emphasized, the
language of instruction is mostly Portuguese. For practical reasons (i.e.
availability of teaching material and broader context of use), literacy in
Portuguese is also given more importance to than in Awetí. The two Awetí
school teachers have a crucial role in the transmission of the national
language within the community. Their advanced learner variety can be
characterized as grammatically and phonologically approaching the standard
language, although lexically somewhat reduced and influenced by regional
and indigenous terms from the Upper Xingu contact variety of Portuguese. 50
In November 2007, one of the Awetí teachers was living in each of the
villages. According to the teacher Awajatu Awetí, the school in village A
was operating with two age groups, each taught during three to four hours on
five days per week. The languages of instruction were Portuguese and
Awetí, whereby only one language at a time was used, divided into blocks of
two weeks each. 51 The second teacher, Waranaku Awetí, had taken up his
work in the school of village B in September 2007. In November that year he
reported that the two teachers of the village were teaching one age group
each, whereby the second language of instruction of his new colleague was
50 Since the acquisition of Portuguese as a second language by these two individuals was
directed from the very beginning, their variety does not show many of the “pidginizing”
elements listed by Emmerich and Paiva (2000: 455).
51 In 2008, a young man without any teacher-training started to regularly teach the younger
children in village A, using Awetí as the only medium of instruction. The older pupils
continued to be taught on a less regular basis by the bilingual teacher (Waranaku Awetí,
email communication 27.02.09).

256
not Awetí but Kamaiurá. The Awetí-dominant teacher had started to teach
the younger children. He informed that every person in the school –
including the teachers – was using the indigenous language he/she felt most
comfortable with, i.e. either Awetí or Kamaiurá. This was possible due to a
high proportion of bilinguals in the village population. 52 According to the
observations of the Awetí teacher, the children in village B were good at
reading and speaking Portuguese. This may be due to the fact that before his
arrival these pupils had been exclusively alphabetized in the national
language. 53 Additionally, the adolescents of village B, due to the vicinity of
the Posto Leonardo, have the opportunity to practice their Portuguese with
members of the national society working there and with peers from other
ethnic groups. The Awetí school-teacher introduced the same scheme to the
school of village B which was in use in village A, i.e. alternating between
Portuguese and Awetí as languages of instruction at intervals of two weeks.
In 2007, both Awetí teachers started to use data from the DOBES corpus in a
print format. 54 There are still logistical problems, such as a continuous
supply with print-outs or the financing of book-publications, to be overcome,
52 In February 2009 only the Kamaiurá-dominant teacher and some of his siblings among
the pupils were using Kamaiurá. The other pupils were responding in Awetí (Waranaku
Awetí, email communication 27.02.09).
53 Waranaku Awetí (2007, personal communication) stated that there were no school-books
written in Kamaiurá in the village school and that the two teachers did not dare to use any in
order “not to get into trouble with the Kamaiurá of other places”. Since there is a strong
tendency observable among the village inhabitants to identify themselves with the Awetí
rather than the Kamaiurá, he estimated that the other teacher would stop to use Kamaiurá as
a medium of instruction in the near future. However, in August 2008, this second teacher of
village B was participating in a DoBeS-workshop for indigenous teachers at the University
of Campinas as a representative of the Kamaiurá speech-community (personal observation)
54 Their first choices were a lengthy narrative about the group’s Awetí-Enumania origins to
be used in history lessons and a descriptive text on how to fabricate a hammock.

257
and some data is still being revised for orthographical consistency, but the
material in Awetí at disposition for instruction could be increased and may –
on the long run – counterbalance the Portuguese bias in literacy.
As already observed by Emmerich (1992: 75), radio telephony is
another domain where Portuguese is frequently used as a lingua franca with
interlocutors from outside the reserve and with people from other Xinguan
communities with whom one does not have any other language in common.
In each of the Awetí villages there is a shortwave radio set which is
continually in operation so that this way especially the female inhabitants of
the respective house have a regular input from other languages.
Further Portuguese input is given through media consume. With
regard to the exposure to mass media, a strong increase could be observed.
In 2002, one TV-set in the chief’s house of village A was irregularly put into
operation for a couple of hours at night and attended by most villagers, and
two radios were listened to in other houses. In 2005 there were already three
TV-sets and DVDs, another generator as well as three additional radios. In
November 2007, TVs and DVDs were registered for two of the five houses
of village B, while two of the other houses had a radio. All TV-sets and other
electronic devices of the village could be operated by one powerful
generator. Each evening, the inhabitants of all houses were watching
national telenovelas and the news, occasionally soccer-matches and DVDs
of traditional Upper Xinguan festivals or action movies. Little can be said as
to a continuous supply of fuel resources necessary to operate the electronic

258
devices, but at least the younger and mobile population of village B had an
additional possibility to watch television at the nearby Posto Leonardo. The
questionnaire data revealed different tendencies in the way mass media was
made use of when comparing uneducated and illiterate with more educated
community members. While those with little or no formal instruction tended
to consume mass media indiscriminately, more educated people, including
illiterate male members of the group with a higher level of traditional
knowledge, used these devices as a source of information. This suggests that
input from the media alone does not increase the level of competence in
Portuguese. Additional instruction and – more importantly – a positive
attitude are needed to effectively make use of this input source.
Another development indirectly caused by massive access to mass
media is the loss of the Awetí oral literary tradition. Currently there are two
competent Awetí story-tellers, one at a very advanced age, the other born in
1952, and both living in village A. Among adolescents and young adults
there was not much interest in continuing with this tradition. However,
during transcription work many of them came along to listen to the
recordings on the computer, and recently a son of one of the story-tellers has
apparently taken up to acquiring this traditional knowledge. An additional
effect of the DOBES project is that the narratives collected in the corpus
circulate as tape-recordings and CDs in the community and may in this more

259
modern fashion continue to be passed on to the younger generations even
though this does not replace the elaborate art of story-telling. 55
A further innovation, added to the Xinguan infrastructure in 2006, is
the installation of several computers in the reserve providing access to the
internet. One of these can be used at the Posto Leonardo and renders
communication with the outside world possible without having to pay the
enormous costs of a telephone call via shortwave radio. In how far this new
medium of communication is used by members of the Awetí community, not
much can be stated at the present moment except that some young male
adults of both villages have their personal email addresses and that an
irregular Portuguese email contact with the two Awetí teachers could be
established.
5. Conclusion and future prospects
The overall low number of Awetí speakers and the recent split of the
speech community are two factors which by themselves may have a negative
impact on the vitality of the language. This is further aggravated by the
presence of two other dominant languages: Kamaiurá and Portuguese. Most
Awetí are of mixed Awetí-Kamaiurá origin and have at least a passive
competence of this language. In the years after the separation a more
substantial migration of Kamaiurá speakers to both Awetí villages has
influenced the proportional use of the two languages. A point in favour of
55 See Franchetto’s (this volume) description of how recordings of traditional knowledge
have changed the relation between masters and apprentices in the Kuikuro community.

260
the vitality of the Awetí language under these conditions is that the Upper
Xinguan language policy is still intact so that the two languages – although
used in the same functional domains – are clearly differentiated with regard
to the person speaking or spoken to. However, the traditional division of a
Xinguan village into a public (i.e. village plaza) and a private sphere (i.e.
household unit) promotes areas of differing language use and individual
bilingualism. By a detailed analysis of language competences and use in
both villages it could be shown that in village A in 2007 only one household
was Kamaiurá-dominant. In village B all Kamaiurá-speakers, living in two
households, were bilingual in comparison to a noticeable number of Awetí
monolinguals, many of them children, in the other houses. The move of one
of the Awetí school-teachers to village B in 2007 marked an important step
towards further fortifying the Awetí language in that place so that in 2009
Awetí had become the preferred language of most pupils.
Another strong indicator for the vitality of Awetí is its high
transmission rate. After the departure of the Kamaiurá-dominant household
in 2008, there are no children growing up in village A who do not have
Awetí as their first and dominant language, and in village B this also applies
to the majority of children.
Although the Awetí village A, on the surface, is characterized by a
more traditional way of life of its inhabitants, this can be regarded as a
consequence of the fact that little access to financial resources prevents the
community from participating in more “innovative” structures. Another

261
factor in the same direction is its relative geographic and political isolation.
For village B, given its direct vicinity to the Brazilian health post, cultural
pressure from the national society may be more evident.
Both places experience a strong input from Portuguese but the
language mostly occurs in other domains than the indigenous languages. Not
only that it is the primary language of bilingual education and used as a
contact language for trading purposes. Its overall presence is further
strengthened by radio telephony and the growing importance of mass media
in both villages whereby the latter, as a side-effect, promotes the loss of the
Awetí oral literary tradition. Currently, there are not yet many fluent
speakers of Portuguese in both communities, and literacy in Portuguese is
low, but the effects of bilingual education, introduced more than a decade
ago, can already be observed in the language attitude and use of the younger
generation.
There are few contact phenomena, mostly on the lexical level, in the
Awetí language itself, and instances of code-mixing are still socially marked
as unacceptable. Although the variety spoken by adolescents and younger
adults, in comparison to older speakers, appears to be slightly reduced in its
lexicon, a positive attitude towards the Awetí language, marking the ethnic
identity of the inhabitants of both villages, can still be observed in all
generations.
The documentation project in various aspects has made a
contribution towards creating means for the retention of the Awetí language

262
and culture in a constantly changing environment. It should by no means be
regarded as a “revitalization program” but as a way to enhance the prestige
of Awetí within the community and an opportunity to document it at a stage
where it is still vital, much in line with what Crystal (2000: 112f.) postulates
as “preventive linguistics”. During their participation in the project, several
young Awetí transcription assistants became more interested in their oral
history and systematically learned to write the language. Material in Awetí
for instruction on important cultural matters could be produced from corpus
data, and in general many aspects of traditional life were captured and stored
in an archive which will be accessible to future generations and can be
continued by the Awetí themselves. An example of how this may be done is
given by Franchetto’s (this volume) account of the Kuikuro speech
community’s consultants assuming the position of authors of their own
documentation during and following their participation in the DOBES
program. In 2009, among the Awetí, too, several young men from village B
have started documentation on video. This was made possible with prize-
money from the Ministry of Culture received for a DOBES documentary,
and access to monetary resources must in fact be seen as the key condition
for successfully carrying out any activity in this direction.
In summary, a detailed analysis of language competences, use and
attitude within the group has shown that the continuation of the Awetí
language is not as threatened in the near future as the overall factors for
language endangerment may indicate. Thus, the recent developments in the

263
Awetí linguistic community neatly illustrate Himmelmann’s (2009)
observation that the “potential which any single factor might have for
inducing language shift is only realized in a specific and complex
constellation of factors”. Although the Awetí belong to the Upper Xingu
cultural area where all participating ethnic groups are exposed to comparable
outside factors with more or less the same intensity, the study has shown that
– apart from some overall tendencies valid for all of them – there is also
variation which goes back to individual developments most often from
within the group. Since the study accompanied a decisive moment in the
group’s history – the split of one village into two, the development outlined
here is highly significant. For the Yawalapiti and the Trumai the dispersion
of their communities led to language loss and shift towards other indigenous
languages and Portuguese. Their separation, then, was a direct consequence
of the population reduction through epidemics, whereas also after the recent
division the internal social structure of the Awetí community has remained
relatively intact. However, what has been outlined here with regard to the
vitality of the Awetí language in the two villages can still not be
characterized as a stable situation and definitely does not apply to the same
extent to the Awetí culture. In this respect, an increasing acculturation
through mass media can be expected as much as in other Xinguan groups
and predominantly in the younger generations with a basic school-education.
Their interest to learn Portuguese is a clear indicator for an interest in the
Brazilian national society as a whole. This attitude expressed by young male

264
Awetí community members may on the long run turn out to be the incipient
stage of a possible process of language shift towards Portuguese as in the
Upper Xingu has already occurred with the Trumai and throughout Brazil
with a major part of the indigenous population. Such a future development
for the Awetí may be counteracted by joint efforts from within the
community to maintain the indigenous language and traditions alive for
which the wide acceptance and usage of the possibilities of the DOBES
archive and especially the creation of regular income sources can be
important prerequisites.
APPENDIX

References
265
Basso, Ellen. 1973. The Kalapalo indians of central Brazil. <strong>New</strong> York: Holt,
Rinehart & Winston.
Basso, Ellen. 1973a. “The use of Portuguese relationship terms in Kalapalo
(Xingu Carib) encounters: changes in a central Brazilian
communications network”. Language in Society 2 (1): 1-21.
Basso, Ellen. 1995. The last Cannibals. A South American Oral History.
Austin: University of Texas Press.
Crystal, David. 2000. Language Death. Cambridge: CUP.
Dole, Gertrude E. 2001. “Retrospectiva da história comparativa das culturas
do Alto Xingu: Um esboço das origens culturais alto-xinguanas”. In
Os Povos do Alto Xingu. História e Cultura, B. Franchetto and M.
Heckenberger (eds), 63-76. Rio de Janeiro: Editora da UFRJ.
Drude, Sebastian. 2002. “Fala masculina e feminina em Awetí”. In Línguas
Indígenas Brasileiras. Fonologia, Gramática e História [Atas do I

266
Encontro Internacional do Grupo de Trabalho sobre Línguas
Indígenas da ANPOLL], A. Cabral, Suelly Câmara and A. D.
Rodrigues (eds), 177-190. Belém: Editora UFPA.
Drude, Sebastian. 2006. “On the position of the Awetí language in the Tupí
family”. In Guaraní y Mawetí-Tupí-Guaraní. Estudios históricos y
descriptivos sobre una familia lingüística de América del Sur
[Regionalwissenschaft Lateinamerika 11], W. Dietrich and H.
Symeonidis (eds), 11-45. Berlin: LIT Verlag.
Drude, Sebastian, Awete, Waranaku and Awetí, Awajatu (in press). “A
ortografia da língua Awetí”. Revista de Estudos e Pesquisas. Brasília:
FUNAI: CGEP/CGDTI.
Emmerich. Charlotte. 1992. “O português de contato no Parque Indígena do
Xingu, Mato Grosso, Brasil Central”. Estudos 13: 57-90.
Emmerich, Charlotte and Paiva, Maria da Conceição de. 2000. “O português
indígena do Alto Xingu, Brasil Central: mudanças e perspectivas
vinte anos depois”. In O Português Brasileiro: Pesquisas e Projetos,
S. Große and K. Zimmermann (eds), 451-464. Frankfurt a.M.: TFM.
Franchetto, Bruna and Heckenberger, Michael J. (eds). 2001. Os Povos do
Alto Xingu. História e Cultura. Rio de Janeiro: Editora da UFRJ.
Franchetto, Bruna. 2001. “Linguas e história no Alto Xingu”. In Os Povos
do Alto Xingu. História e Cultura, B. Franchetto and M.
Heckenberger (eds), 111-156. Rio de Janeiro: Editora da UFRJ.

267
Gregor, Thomas. 1977. Mehinako: The Drama of Daily Life in a Brazilian
Indian Village. Chicago: University of Chicago Press.
Guirardello, Raquel. 1999. A Reference Grammar of Trumai. PhD
dissertation, Rice University.
Heckenberger, Michael J. 2001. “Epidemias, índios bravos e brancos:
contato cultural e etnogênese do Alto Xingu”. In Os Povos do Alto
Xingu. História e Cultura, B. Franchetto and M. Heckenberger, 77-
110. Rio de Janeiro: Editora da UFRJ.
Himmelmann, Nikolaus P. 2009. “Language endangerment scenarios: A case
study from northern central Sulawesi”. In Endangered Languages of
Austronesia, M. Florey (ed), 00-00. Oxford: Oxford University Press.
Moore, Denny. 2007. “Endangered languages of lowland tropical South
America”. In Language diversity endangered, M. Brenzinger (ed),
29-58. Berlin: de Gruyter.
Richards, Joan. 1977. Orações em Waurá. Summer Institute of Linguistics.
Série Lingüística Nr. 7: 141-184.
Richards, Joan. 1988. A estrutura verbal Waurá. Summer Institute of
Linguistics. Série Lingüística Nr. 9: 197-218.
Seki, Lucy. 2000. Gramática do Kamaiurá. Língua Tupi-Guarani do Alto
Xingu. Campinas: Editora da Unicamp.
Silva, Pedro Agostinho da. 1972. “Informe sobre a situação territorial e
demográfica no Alto Xingu”. In La situación del indígena en
América del Sur, G. Grunberg (ed), 355-379. Uruguay: Tierra Nueva.

268
Steinen, Karl von den. 1968/ 1894. Unter den Naturvölkern Zentral-
Brasiliens. Reiseschilderung und Ergebnisse der Zweiten Schingú-
Expedition 1887-1888. <strong>New</strong> York / London: Johnson Reprint
Corporation.
Stoeckler Coelho de Souza, Marcela. 2001. “Virando gente: notas a uma
história Awetí”. In Os Povos do Alto Xingu. História e Cultura, B.
Franchetto and M. Heckenberger (eds), 358-400. Rio de Janeiro:
Editora da UFRJ.
Terrill, Angela. 2002. “Why make books for people who don’t read? A
perspective on documentation of an endangered language from
Solomon Islands”. International Journal of the Sociology of
Language 155/156: 205-219.
Vasconcelos, Vicente de P.T. da Fonseca. 1945. Expedição ao Rio Ronuro.
Rio de Janeiro: CNPI / Imprensa Nacional.

Epilogue
Fernando Ramallo & José Antonio Flores Farfán
269
The principal reasoning for the justification of this book is that if we are to
overcome the gap that exists between documentary practice and the
revitalization of languages, it is necessary to contemplate the role of a
critical sociolinguistics that positions threatened language-speakers at the
forefront of the action and debate. Considering the current sociolinguistic
regression taking place in many parts of the world, the sociolinguist’s
responsibility in the reproduction and preservation of linguistic knowledge
(Himmelmann 2008) is paramount. Of course, the dialogue between
documentarians and sociolinguists should produce methodological clues that
reverberate into increasingly sustainable approaches to address the
challenges related to work in the field of linguistic diversity. In order for this
to happen it is also necessary for sociolinguistics to delve deeper into the
importance of the presence and function of language in the sociopolitical
landscape and in the construction and reproduction of power relations, social
exclusion and inequality that also operate in the academic world. Naturally,
this has much to do with the understanding of the term sociolinguistics, what
it contributes to the advancement of knowledge and what talking about
sociolinguistic information presumes to accomplish, among other things

270
(Hambye and Siroux 2009). This entails positioning oneself in a
sociolinguistics that is context-sensitive, that goes deep into the difficulties
of the field and that, in addition to trying to overcome the tensions that are
generated by its very own internal debate, not only favors cross-fertilization
between different linguistic disciplines, but most of allcontributes to the
emancipation of linguistic communities.
As it is a relatively new field, documentary linguistics has only recently
begun to benefit from the theories, methods and advances of sociolinguistics.
This is largely due to the forms in which work in the area of linguistic
documentation has been developed. Although significant changes are being
produced in what we understand as the appropriate direction, much in the
sense enunciated by the present volume, much of the work done in the past
two decades has been designed based on proposals that are far from what we
would call revitalization sociolinguistics and this continues to be the
prevailing trend.
It is true that documentary linguistics during the last several decades has
contributed a good deal to the renovations of its theoretical focuses and its
traditional methodological strategies. Since the beginning of the 1990s
knowledge of linguistic extinction processes and of the world panorama of
threatened languages has fostered a critical conscience about the future
viability of linguistic diversity, going as far to provoke a shift in linguistic

271
research priorities. Even though this is by no means a new problem, the last
two decades have witnessed the configuration of standardized research and
action protocols with the local communities of speakers of languages in
regressive sociolinguistic situations, not to speak of the emergence of best
practices protocols.
In the face of continuing decrease in linguistic diversity on the planet,
anthropologists, linguists, documentarians, sociolinguists, educators, etc.
have for decades worked tirelessly on language documentation projects,
linguistic community projects, linguistic repertoire projects and registering
other communicative practices. Such endeavors, based on the logic of the
protection of immaterial heritage, have served to create vast linguistic
archives albeit their primary purpose has not always been and rarely is
revitalization; rather it has been linguistic research itself. (cf. Austin &
Grenoble 2007).
In any case, the missing element has frequently been greater professional
cooperation among people with similar and even dissimilar interests, such as
academic interests when compared with those of the language-speaking
communities. The consequence of such absence of cooperation is that rarely
have personal agendas converged in the search of solutions which transcend
the personal gains experienced by the researchers or groups of investigators,
outstandingly linguists. For this reason it is essential that we continue

272
pushing for a quality standard which demands practices destined to promote
a model of collaborative research or engaged linguistics that reports benefits
not only for those performing the research but for all of those being
researched. Moreover, appropriation of the research itself by the subjects of
investigation themselves is an outstanding goal for the purposes of
revitalization. In fact, this research-subject logic must give way to a new
model of cooperative work that guarantees the creation of those vast
documentary archives but also aims to strengthen the linguistic revitalization
processes which seek to give continuity to linguistic groups, getting them to
dialogue, something which only until recently is beginning to take shape.
This is why the need for further supportive efforts on the part of
documentarians and sociolinguists, naturally with the cooperation of the very
speakers of the languages, is such an urgent matter, as the authors of this
book have emphasized in one way or another.
A more sociolinguistic approach to the issue of documentation is understood
not only as documenting linguistic and cultural practices. Pursuing to change
specific linguistic ideologies with respect to endangered languages, and
becoming relevant for the interest and perspectives of speakers themselves,
are just a couple of a series of goals which in turn raise a number of
questions that require further discussion. Some such topics that are invited
by the guests editors include: How can we not limit linguistic archives and
repositories to linguists? How can we start thinking in even a more general

273
public including non speakers? Documentary linguistics requires going
beyond “pure” linguistic description, for which sociolinguistics provides
theoretical, methodological and empirical ground. Bringing together the
active practice of the two agendas in one supporting an integrated
epistemology leads almost naturally to the issue of revitalization. Even when
these three subfields in linguistics have much in common and could benefit
from each other, as suggested little communication within them has been so
far established. Because of this, it is important to promote a model of
documentary linguistics in which interdisciplinarity is not seen as a handicap
but rather as an asset in the development of the linguistic communities and in
which different strands of linguistic knowledge can contribute to a single
goal, namely, the revitalization of endangered languages. This must be done
with the active participation of community members, a sine qua non
condition to ensure the survival of linguistic diversity in the near future
References
Austin, P.K. & L.A. Grenoble (2007). “Current trends in language
documentation” in P.K. Austin (ed.). Language Documentation and
Description, vol. 4, 12-25.
Hambye, Ph. & J.L. Sirioux (ed.) (2009). “Analysing language as a social
practice: reflections on some implications for the analysis of
language”. Sociolinguistic Studies, 3(2), 131-147.

274
Himmelmann, N.P. (2008). “Reproduction and Preservation of Linguistic
Knowledge: Linguistics’s Response to Language Endangerment”.
Annual Review of Anthropology, 37, 337-350.