In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • Postvelar harmony by Kimary Shahin
  • Andrew Nevins
Postvelar harmony. By Kimary Shahin. (Current issues in linguistic theory 225.) Amsterdam: John Benjamins, 2003. Pp. 344. ISBN 158812217. $150 (Hb).

Kimary Shahin takes the reader through a detailed investigation, demonstrating that pharyngealization harmony and uvularization harmony are distinct processes. The acoustic and phonological studies focus on two languages: the Abu Shusha dialect of Palestinian Arabic (PA), and St’àt’imcets Salish (St’). S’s leading proposal is that pharyngealization harmony involves spread of the feature [Retracted Tongue Root]. Uvularization harmony, by contrast, involves spread of both [RTR] and the feature [Dorsal].

In PA, pharyngealization of vowels has three sources. One is adjacency to a laryngeal, pharyngeal, or uvular consonant, for example, /ʔ, h, ʕ, ħ, ʁ, χ/, which S demonstrates has acoustic effects on neighboring vowels: a raised F1 and a lowered F2. The affected vowels are the nonlow vowels. The second source of [RTR] alternations in PA is due to closed-syllable pharyngealization (81ff.), which S shows to cause F1 raising and F2 lowering, perhaps due to articulatory undershoot. Finally, §2.4.3 documents nonlocal harmony. A vowel in an open syllable becomes [RTR] even when it is not in the local environment to do so, for example, [lɪ.bʊ.ʔə] ‘lioness’, [ʁɪ.nɪ.mə] ‘goat’, due to a pharyngeal/uvular elsewhere in the word. Arguably, this is due to vowel harmony from the vowel tautosyllabic with the [RTR] consonant. A vowel in an open syllable may also become [RTR] if the neighboring vowel became [RTR] through closed-syllable laxing, that is, [kʊ.tʊb] ‘books’, [tɪ.bɪn] ‘straw’. An interesting additional phenomenon from the perspective of the phonology is that long vowels and stem-final vowels (even when not strictly word-final) are opaque to pharyngealization harmony.

Uvularization harmony (often called ‘emphasis spread’ in the literature) affects consonants and causes backing and F2-lowering in low vowels (i.e. / æ, æː/ become [ɑ, ɐː]), and is triggered by emphatic consonants. There are underlying emphatics, which can be observed by the minimal pairs [tiːn] ‘figs’ vs. [t̙iːn̙] ‘mud’. S proposes that emphatic consonants bear a secondary place specification of [Dorsal] and [RTR] in addition to their primary place specification [End Page 230] (e.g. [Coronal] in the consonants of [t̙iːn̙]). One of the most fascinating facts to emerge from S’s study of PA is that the high vowels are completely transparent to uvular harmony across them (§2.5.5), and do not become phonetically backed, for example, [m̙ʊ.hʊ.’r̙-ɐːt̙], *[m̙ʊ>.hʊ>.’r̙-ɐːt̙] ‘fillies’, [b̙-ʕ̙ɜ>.t̙i-’nɐ̙-ʃ], *[b̙-ʕ̙ɜ>.t̙ɨ-’nɐ̙ː-ʃ] ‘he doesn’t give it to us’ (158). S shows that even when comparing the [tiːn]/[t̙iːt̙] pair, ‘those vowels [in tokens of emphatic environments] did not reach and maintain a lowered F2 target’ (159). The phonological and phonetic data lead S to conclude that NO-GAP, a putatively inviolable constraint on Gen that has recently become fashionable within the OT literature, must in fact be violable (161).

Ch. 3 is devoted to St’, which also has distinct pharyngealization and uvularization harmony. St’ has the three underlying vowels /i, u, æ/ and epenthetic mid-central vowels. Pharyngealization harmony causes [RTR] spread to vowels preceding (but not following) a guttural, for example, [tɪʁ’w-in’] ‘to untie something’. Unlike PA, however, St’ lacks the more spectacular cases of vowel harmony transmitting pharyngealization nonlocally to the left. In addition, St’ lacks closed-syllable pharyngealization. Uvularization harmony in St’ causes backing of low vowels and epenthetic vowels preceding an emphatic, for example, [pətʃkəɬ] ‘leaf’ vs. [ɬɅt̙ʃ] ‘to cave in’. Like PA, high vowels in St’ are unaffected by uvularization harmony, and S provides acoustic support here as well. In addition, a guttural in St’ does not cause the backing that is characteristic of uvularization, lending support to S’s claim that uvularization is distinct from pharyngealization in St’ as well as PA.

This is a commendable study, and the comparison between typologically unrelated St’ and PA make for important conclusions about the universal and language-particular nature of principles of postvelar harmony. It may be a difficult read for those who do not already have some familiarity with...

pdf

Share