The contrastive hierarchy in phonology 2009 Dresher.pdf - CUNY ...

More documents

Recommendations

Info

obstruent, not /k/: /h/ is not sonorant, not labial, not dorsal, and not dental (7.58). Therefore, [s, z, S] /h/. (7.58) NZ Ma#ori contrastive specifications (Herd 2005) non-sonorant [sonorant] qp wo labial non-labial /m, w, n, r, N/ ty ei /p/ /f/ non-dorsal [dorsal] ei g [dental] non-dental /k/ g g /t/ /h/ The different contrastive roles played by /h/ in these languages suggests that they have different ‘pattern alignments’, in Sapir’s terms, despite their very similar phonetic realizations. The differing status of /h/, as well as the presence of /t/ in NZ Ma#ori but not in Hawaiian, also account for the very different contrastive status of /k/ in each language: general default consonant in Hawaiian, and dorsal obstruent in NZ Ma#ori. 7.7.3. Summary As mentioned, loan phonology is a diverse phenomenon, and it is unlikely that a single approach can account for all patterns of loanword adaptations. But it suffices for our purposes to show that there exists a class of cases in which loan phonology is sensitive to the contrastive structure of a language; in particular, to the contrastive feature hierarchy. The Polynesian examples discussed above provide a compelling case of this type. 326
7.8. The acquisition of distinctive features and contrasts Following the pioneering work of Jakobson (1941) and Jakobson and Halle (1956) discussed in §4.3, the notion of a contrastive hierarchy has been fruitfully applied in acquisition studies, where it is a natural way of describing developing phonological inventories (Pye, Ingram and List 1987, Ingram 1988, 1989, Levelt 1989, Dinnsen et al. 1990, Dinnsen 1992, 1996; see Dresher 1998a for a review). For example, Fikkert (1994) describes the development of segment types in onset position in Dutch as in (7.59). (7.59) Development of Dutch onset consonants (Fikkert 1994) consonant /P/ Stage 1 u m obstruent /P/ sonorant /N/ Stage 2 ueim ueim plosive fricative nasal liquid/glide Stage 3 g g g g /P/ /F/ /N/ /L/J/ In Stage 1 there are no contrasts. The value of the consonant defaults to the least marked onset, namely an obstruent plosive, designated here as /P/. The first contrast (Stage 2) is between obstruent and sonorant. The former remains the unmarked option (u); the sonorant defaults to nasal, /N/. At this point children differ. Some expand the obstruent branch first (Stage 3a), bringing in marked fricatives, /F/, in contrast with plosives. Others (Stage 3b) expand the sonorant branch, introducing marked sonorants, which may be either liquids, /L/, or glides, /J/. Continuing in this way we will eventually have a tree that gives all and only the contrasting features in the language. 327
Page 1:
The Contrastive Hierarchy in Phonol
Page 4 and 5:
course with Keren Rice at the Cente
Page 7 and 8:
Table of Contents Acknowledgements
Page 9:
6.7. Underspecified inputs.........
Page 12 and 13:
(Chen 1972). In final open syllable
Page 14 and 15:
For example, given a language in wh
Page 16 and 17:
difficulty Czech speakers have in r
Page 18 and 19:
four-vowel system /i, e, a, u/. Eve
Page 20 and 21:
sonority (say, /a/) and one of less
Page 22 and 23:
Halle (1959) against the structural
Page 24 and 25:
former approach, pairwise compariso
Page 26 and 27:
Chapter 7 presents a series of case
Page 28 and 29:
these features. And yet, none of th
Page 30 and 31:
2.2. Contrastive specification by p
Page 32 and 33:
to /p/, that is, a voiceless nasal
Page 34 and 35:
collaborators. Rather than make pai
Page 36 and 37:
. If the set is found to consist of
Page 38 and 39:
This kind of reasoning, which I wil
Page 40 and 41:
edundancy, and this conflation of t
Page 42 and 43:
in (2.13). /p/ is characterized as
Page 44 and 45:
voiced, and the same holds for [nas
Page 46 and 47:
The problem of too many features do
Page 48 and 49:
(2.21) Three-vowel system, features
Page 50 and 51:
[+back]. The addition of the featur
Page 52 and 53:
. Specifications according to the p
Page 54 and 55:
pairwise method. One might reason t
Page 56 and 57:
the program generated up to 52 colu
Page 58 and 59:
properly contrastive specifications
Page 60 and 61:
discussion here, because they inter
Page 62 and 63:
There are different kinds of binary
Page 64 and 65:
(2.30) Contrastive specifications:
Page 66 and 67:
algorithm that assigns feature valu
Page 68 and 69:
feature ordering, for there is empi
Page 70 and 71:
contrastive in a given language; th
Page 72 and 73:
(3.1) Different phonetics, similar
Page 74 and 75:
the wrong place in the pattern of D
Page 76 and 77:
c. Consonants obstruent sonorant vo
Page 78 and 79:
proposing certain contrastive relat
Page 80 and 81:
The phonologically relevant marks o
Page 82 and 83:
1. First, divide the inventory by t
Page 84 and 85:
[T]he opposition d-n (as in French)
Page 86 and 87:
that these features are contrastive
Page 88 and 89:
a contrastive hierarchy, but not in
Page 90 and 91:
only the fact that lips and tip of
Page 92 and 93:
ts, s, z/ and /k, x, ƒ/, respectiv
Page 94 and 95:
3.3.3.2. Vowel systems Polabian In
Page 96 and 97:
(3.14) Triangular system with schwa
Page 98 and 99:
three vertical sets, as in (3.18).
Page 100 and 101:
Though the above analysis is the ty
Page 102 and 103:
In these examples there appears to
Page 104 and 105:
classify them according to their fu
Page 106 and 107:
Trubetzkoy’s criteria, variation
Page 108 and 109:
o o^ O O^ a) e) o) O) ´ Vocality +
Page 110 and 111:
Subtree (3.28b) depicts the contras
Page 112 and 113:
Twaddell (1935)…’ 16 Second, th
Page 114 and 115:
After considering the merits of eac
Page 116 and 117:
3.6.2. Hockett’s approach in prac
Page 118 and 119:
decisions here are not arbitrary (t
Page 120 and 121:
obstruent and sonorant coming secon
Page 122 and 123:
chapter). I have argued that issues
Page 124 and 125:
(3.39) Diagnostics used in identify
Page 126 and 127:
(3.41) Contrast and phonological ac
Page 128 and 129:
and (2) contrastive features are as
Page 130 and 131:
emphasis on dichotomies as a fundam
Page 132 and 133:
implication’ that are ‘universa
Page 134 and 135:
in the low vowels (0.11121). These
Page 136 and 137:
they observe that a set of eight ph
Page 138 and 139:
illustrates that the concept of the
Page 140 and 141:
esult in all phonemes receiving a v
Page 142 and 143:
n m n@ r l l’ i u e o a Vocality
Page 144 and 145:
approach. Perhaps because of this J
Page 146 and 147:
(4.10) Contrastive hierarchy: Stand
Page 148 and 149:
acoustics. Halle sets out a number
Page 150 and 151:
allophonic realizations of phonemes
Page 152 and 153:
tS S Z k kJ g x j r rJ l lJ vocalic
Page 154 and 155:
the sonorant l). The rule that chan
Page 156 and 157:
and redundant features: only the fo
Page 158 and 159:
Phonological theory, however, did n
Page 160 and 161:
. Phonological rules i. [ ] [-g] i
Page 162 and 163:
Stanley (1967: 416) considers the c
Page 164 and 165:
edundant features. This argument wo
Page 166 and 167:
equire a different tree in which [s
Page 168 and 169:
158
Page 170 and 171:
c. Consonants: velopalatal vs. labi
Page 172 and 173:
arguments of Bloomfield 1933). Inst
Page 174 and 175:
ules to which a purely formal evalu
Page 176 and 177:
nevertheless interact with the phon
Page 178 and 179:
the rule replacing [m F] as the com
Page 180 and 181:
(5.4) Markedness matrices for vowel
Page 182 and 183:
emains /a, i, u/. The vowels /i, u/
Page 184 and 185:
d. XI [u high] [-high] / Kean prop
Page 186 and 187:
capture co-occurrence restrictions
Page 188 and 189:
distinguish between contrastive and
Page 190 and 191:
manipulates only +/- values, not m/
Page 192 and 193:
Linking crucially depends on the fe
Page 194 and 195:
of underspecification theory and le
Page 196 and 197:
marked on sonorants in the lexicon
Page 198 and 199:
Kiparsky’s analysis thus mirrors
Page 200 and 201:
constrain this process by a hierarc
Page 202 and 203:
RU does not fully adopt the Contras
Page 204 and 205:
minimality. This leaves us with the
Page 206 and 207:
Similar results can be shown to be
Page 208 and 209:
for example, the underlying vowel s
Page 210 and 211:
5.3.3. Contrastive Specification St
Page 212 and 213:
(5.31) Pasiego Spanish vowels: Poss
Page 214 and 215:
all logical redundancies, nor does
Page 216 and 217:
Hume 1995, Halle 1995), Government
Page 218 and 219:
.Vocoids laryngeal [spread] [constr
Page 220 and 221:
5.4.2. Terminal features and the co
Page 222 and 223:
A fourth difference is that the var
Page 224 and 225:
Intuitively, the two segments diffe
Page 226 and 227:
epresentations beyond the needs of
Page 228 and 229:
218
Page 230 and 231:
6.2. Optimality Theory OT (Prince a
Page 232 and 233:
convert the coda into an onset, and
Page 234 and 235:
eing equal, this implies that a son
Page 236 and 237:
6.3.2. Contrast as an emergent Kirc
Page 238 and 239:
6.4. The contrastive hierarchy in O
Page 240 and 241:
6.4.2. Inputs What should be the in
Page 242 and 243:
(6.12) Contrastive hierarchy for Cl
Page 244 and 245:
(6.14) Classical Manchu constraint
Page 246 and 247:
epresentation because it, too, viol
Page 248 and 249:
We can mimic the effect of the Pair
Page 250 and 251:
limiting specifications to those th
Page 252 and 253:
and the rest of the features are re
Page 254 and 255:
to adopt privative features, where
Page 256 and 257:
6.7. Underspecified inputs The noti
Page 258 and 259:
(6.29) Implicational constraints fo
Page 260 and 261:
In (6.31), these constraints are ad
Page 262 and 263:
252
Page 264 and 265:
(6.17) Full specification of all fe
Page 266 and 267:
(6.21) Underspecification by means
Page 268 and 269:
(6.24) Illicit fully specified inpu
Page 270 and 271:
(6.27) Underspecified input Input:
Page 272 and 273:
262
Page 274 and 275:
value: the fact that all the alloph
Page 276 and 277:
7.2.1. Contrast and markedness MCS
Page 278 and 279:
(7.2) Contrastive features for dial
Page 280 and 281:
analysis is ‘abstract’ with res
Page 282 and 283:
universal; we will see in §7.4 tha
Page 284 and 285:
feature [high], the contrastively [
Page 286 and 287: phonological analysis. This should
Page 288 and 289: 7.3.2. Implicit feature hierarchies
Page 290 and 291: elations among features. This is no
Page 292 and 293: dialects with similar-looking inven
Page 294 and 295: this burden exists independently of
Page 296 and 297: (7.12) Classical Manchu vowel syste
Page 298 and 299: 7.4.1.1. ATR harmony All vowels in
Page 300 and 301: (7.17) Stems with only /i/: Suffixe
Page 302 and 303: no surprise, therefore, to see this
Page 304 and 305: [coronal] is contributed by /i/, bu
Page 306 and 307: 296 c. Xibe: [low] > [coronal] > [l
Page 308 and 309: (7.24) Turkish vowel system coronal
Page 310 and 311: (7.26) ATR harmony: /Q/ alternates
Page 312 and 313: ordered before one based on place (
Page 314 and 315: hierarchy we arrived at earlier; in
Page 316 and 317: (7.37) Role of IO-IDENT[back] (adap
Page 318 and 319: contrastive features to the stem-fi
Page 320 and 321: (7.40) Pasiego metaphony does not a
Page 322 and 323: O/, but its scope is limited to the
Page 324 and 325: one step further, arguing that the
Page 326 and 327: spreading [high] to a mid vowel uns
Page 328 and 329: (7.47) Bumo Izon labial and alveola
Page 330 and 331: and, in the framework of generative
Page 332 and 333: French contrastive feature hierarch
Page 334 and 335: (7.54) NZ Ma#ori adaptation of Engl
Page 338 and 339: While the contrastive hierarchy has
Page 340 and 341: fundamental perceptual-cognitive bu
Page 342 and 343: their perceptual system to the cont
Page 344 and 345: figure in the phonology. I conclude
Page 346 and 347: the postlexical or phonetic compone
Page 348 and 349: However, when /r&/ devoices it does
Page 350 and 351: 340
Page 352 and 353: 342
Page 354 and 355: theory is indirectly sensitive to t
Page 356 and 357: Constraint (8.1a) provokes labial h
Page 358 and 359: 8.3. Dispersion-theoretic approache
Page 360 and 361: evaluation of phonetic distance app
Page 362 and 363: specified [+back]. The vowels /i/ a
Page 364 and 365: change, traditionally known as the
Page 366 and 367: primarily signalled by the second v
Page 368 and 369: fronting. To implement this idea fo
Page 370 and 371: allophone of /i/, as shown in (8.13
Page 372 and 373: In the analysis sketched above what
Page 374 and 375: emerge as one of the five contrasti
Page 376 and 377: give the correct results for any ar
Page 378 and 379: vowel [ˆ] as having 100% of the co
Page 380 and 381: as noncontrastive for [back], the a
Page 382 and 383: on). Of the possible potential clas
Page 384 and 385: symmetrical. The choice of a featur
Page 386 and 387:
(Broe 1993, Frisch 1996), I have gi
Page 388 and 389:
(8.23) Features for Bumo Izon stops
Page 390 and 391:
(8.26) Similarity measures for Hawa
Page 392 and 393:
essentially the same as the one ass
Page 394 and 395:
Clements argues that the robustness
Page 396 and 397:
feature specifications, a major wea
Page 398 and 399:
observed by Stevens, Keyser & Kawas
Page 400 and 401:
there are vowels that are [+high, -
Page 402 and 403:
. Features that make each other log
Page 404 and 405:
[round] > [back] (8.38b). This may
Page 406 and 407:
and contradicts the intuition, shar
Page 408 and 409:
Thus, the lack of voiced sonorants
Page 410 and 411:
uniting phonologists who do not oth
Page 412 and 413:
short back (or central) low vowel i
Page 414 and 415:
In a hierarchical conception of con
Page 416 and 417:
406
Page 418 and 419:
These two concepts have not always
Page 420 and 421:
410
Page 422 and 423:
Archangeli, Diana and Douglas Pulle
Page 424 and 425:
Bolognesi, Roberto (1998). The phon
Page 426 and 427:
Cherry, E. Colin, Morris Halle and
Page 428 and 429:
Linguistics (Special Issue on Contr
Page 430 and 431:
Association, 82-93. Département de
Page 432 and 433:
Fikkert, Paula (1994). On the Acqui
Page 434 and 435:
Hall, Beatrice L. and R. M. R. Hall
Page 436 and 437:
Heffernan, Kevin (2007). The role o
Page 438 and 439:
Itô, Junko and Armin Mester (1986)
Page 440 and 441:
Kaisse, Ellen M. and Patricia A. Sh
Page 442 and 443:
Kuhl, Patricia, K. A. Williams, Fra
Page 444 and 445:
Martinet, André (1960). Éléments
Page 446 and 447:
Paradis, Carole (1988). On constrai
Page 448 and 449:
Radis&ic@, Milica (2007). Velar /g/
Page 450 and 451:
Rose, Sharon (1993). Coronality and
Page 452 and 453:
Steriade, Donca (1995). Underspecif
Page 454 and 455:
Walker, Rachel (2001). Round licens
show all

The contrastive hierarchy in phonology 2009 Dresher.pdf - CUNY ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?