Palatalization and coarticulation in Russian.

Abstract

Previous investigations have suggested that contrastive palatalization in Russian blocks vocalic coarticulation by constraining the movement of the tongue during consonant production. This hypothesis assumes that palatalization is realized identically for all consonants and presupposes the existence of a depalatalization gesture for nonpalatalized consonants. Detailed descriptions of Russian suggest that palatalization is not realized identically for all consonants and that some or all nonpalatalized consonants may lack active depalatalization gestures. To the extent that palatalization is not uniform, its effects on coarticulation may vary; where there is no depalatalization gesture vocalic coarticulation may exist. These possibilities were investigated with an acoustic study of Russian. Eight St. Petersburg (Leningrad) natives produced multiple repetitions of (kedvCV) stimuli (C = /b/, /d/, /b$\sp{\rm j}$/, /d$\sp{\rm j}$/, V = /i/, /a/, /u/) and (kVCeCe) stimuli (V = /i /, /a/, /u/, C = /b/, /d/ (combined factorially)). The first stimulus type tests for anticipatory coarticulation, the second for carryover coarticulation. F1 and F2 were measured at the onset, center, and offset of each vowel. Anticipatory and carryover vocalic coarticulation in F1 was noted in both reduced vowels for all speakers. F2 effects at the end of the pre-stress vowel were consistent with anticipatory rounding for nearly all speakers, and with nonrounding effects for some. Carryover vocalic coarticulation in F2 pervaded the post-stress vowel for most speakers; some showed it in the final vowel as well. The predicted differences in coarticulation across different consonants were observed in the speech of some speakers. Consonantal coarticulation was also found. The palatalization of C3 was evident in F1 of the initial vowel for all speakers and in F2 for some. The place of articulation of C2 affected F2 of the final vowel for some speakers. The temporal extent of carryover vocalic coarticulation in Russian was similar to that of an American English control group, but the effect was only half as large. At the same time, the influence on the vowels of the consonant across which coarticulation occurred was larger in Russian than in English. These results suggest that coarticulation is more extensive in Russian than had been supposed but that it is indeed more constrained than in English. The fact that Russian has greater consonant-vowel effects than English at those points where its vowel-vowel effects are smaller supports the claim that consonantal constraints are responsible for Russian's reduced vocalic coarticulation.