Tatiana Kachkovskaia (Saint Petersburg State University) will present her work on pre-boundary lengthening in Russian in the Research Colloquium Phonology at the Institute of English Linguistics. The talk will be held online on 21 May 2021 at 9.45 a.m. CEST (UTC+2). If you like to join, please contact Fabian Schubö.
Language-specific differences in the domain of final lengthening: evidence from Russian
The lengthening of final segments at ends of major prosodic units is considered a language-independent phenomenon. However, a number of languages have revealed specific traits concerning the question of which part of the final word is affected. In many publications, the domain of final lengthening is reported to be the last rhyme preceding the boundary. At the same time, there are languages where segmental duration is reserved for other linguistic purposes—e.g., languages with contrastive vowel length. One could hardly imagine a lengthening of the word-final short vowel without any compensatory effects on the preceding vowels, as this may lead to a change in meaning. Multiple proof of such compensation has been published for Japanese, Finnish and Estonian.
Our data for Russian were obtained using a 20-hour corpus of read speech recorded from 4 speakers. The domain and degree of final lengthening were explored with regard to the factors of segment type, boundary depth, type of melodic movement within the nucleus, and presence or absence of silent pause at the juncture.
Russian has no contrastive vowel length, but duration is the primary correlate of lexical stress (the secondary correlate being vowel quality, due to significant reduction of unstressed vowels). This may be the reason why our data show a strong lengthening effect not only on the last rhyme, but also on the last stressed vowel. E.g., in the word /mama/ (“mother”, with stress on the first syllable), the domain of final lengthening includes both of the vowels, even when the word does not carry nuclear stress. This mechanism enables to maintain the durational contrast between the stressed and unstressed vowels, which is critical as lexical stress in Russian has a distinctive function. Thus, what we observe here is another example of how a linguistic system may impose restrictions on the language-independent phenomenon of final lengthening. At the same time, this evidence for Russian bears resemblance to the case of Dutch, where the domain of final lengthening is the final rhyme with the exception of words ending in a schwa.