Congressus XI Internationalis Fenno-Ugristarum IV [antikvár]

Niina Aasmae - Karl Pajusalu - Tatiana Zirnask Tartu VARIABILITY OF STRESS ASSIGNMENT AND VOWEL DURATIONS IN ERZYA AND MOKSHA 1. Introduction The present paper provides a comparison of empirical data pertaining to the assignment of stress, the temporal characteristics of vowels, and patterns of vowel distribution in several varieties of the Erzya and Moksha languages. With regard to the set of data, some of the general features of the prosody of Erzya and Moksha are outlined. Major ideas concerning the prosody of Erzya and Moksha derive from the works of H. Paasonen (1897, 1903, 1910). Illustrating variation of stress in several sub-dialects of Erzya and Moksha H. Paasonen suggested the idea of Proto-Mordvin stress, which might have been similar to that in contemporary Moksha. Namely, dominantly initial stress could shift to a subsequent syllable having a highly sonorous vowel (a, a). Prosody variation in Erzya has also been noted by A. A. Shakhmatov (1910) in a treatise on two sub-dialects. Both linguists expressed the point of view that vowel inventory in Proto-Mordvin can have included reduced vowels conditioned by the influence of dynamic stress. As far as the origin of movable stress in Erzya is concerned it has been assumed to have developed later. In the treatment of questions related to the phonology of Erzya and Moksha, these assumptions, including the authors' proposition that folk verse in Mordvin is syllabic, have been relied upon until now. The ideas concerning language prosody in Erzya and Moksha have been largely extended in the works of Paavo Ravila (1929, 1973) and Erkki Itkonen (1946, 1970). Few attempts have been made to re-consider the views formulated a century ago. In a relatively recent publication treating the question whether Kalevala metre is akin to the metre of Mordvin folk verse (Korhonen 1994), the author finds a contradiction between the proposition that Proto-Mordvin was prosodically closer to modern Moksha than to Erzya, and the definition of Mordvin verse as syllabic. M. Korhonen writes that Erzya with its "grammatically totally irrelevant" stress "is prosodically one of the most typical languages favouring a syllabic metric system", while such features of Moksha as "a shifting dynamic accent with reduced vowels in the unstressed syllables by no means point to a purely syllable-timed language". According to M. Korhonen (1994: 78-79), detailed further research is required to find a solution to the problem. The theme of prosody in the research of Erzya and Moksha over long time has been peripheral. For lack of empirical data, no definitive answers to the questions raised in the early studies have been found. One of these is what factors might condition stress alternation. In the case of Erzya, there is a supposition that stress assignment is conditioned by the rhythm of an utterance; the trochaic foot has been regarded as the underlying rhythmic unit in the language (Ryabov 1932; Lewy 1937; Aasmae 1982; Lehiste et al. 2003). The mobility of stress in Moksha has been explained as a purely phonetic phenomenon conditioned by the relative sonority of vowels in a word (e.g.: Devayev and Tsygankin 1970: 19) or/and as a means for marking a morphological relation (e.g.: Ravila 1929: 118-119). De-verbal nouns and infinitives, e.g. erama - ii': ; : V . I 7 ¦