Vowel and Consonant Sonority and Coda Weight: A Cross-Linguistic Study Matthew Gordon, Carmen Jany, Carlos Nash, and Nobutaka Takara University of California, Santa Barbara 1. Introduction In many stress systems, there are certain “heavy” syllable types that preferentially attract stress from other syllable types. For example, in Yana (Sapir and Swadesh 1960), stress falls on the rightmost heavy syllable in a word (1a), where heavy syllables include those containing a long vowel (CVV) or a coda consonant (CVC). In words lacking any heavy syllables, stress falls on the initial syllable (1b). (1) a. sibúmk’ai ‘sandstone’, suk’óːniyaː ‘name of Indian tribe’, tsiniyáː ‘no’ b. p’údiwi ‘women’ Languages vary in terms of which syllables are considered heavy. For example, unlike Yana, which treats both CVV and CVC as heavy, Khalkha Mongolian only treats CVV as heavy (Bosson 1964, Poppe 1970, Walker 1995). Primary stress in Mongolian falls on the rightmost non-final heavy syllable (2a), or on the final syllable if the only long vowel is word-final (2b). In words with only light syllables, primary stress is on the first syllable (2c). Heavy syllables not receiving primary stress carry secondary stress (2d). (2) a. áːrùːl ‘dry cheese curds’, ùːrtáegàːr ‘angrily’, úitgartàe ‘sad’ b. galúː ‘goose’ c. xáda ‘mountain’, únʃisan ‘having read’ d. dolóːdugàːr ‘seventh’, ulàːnbáːtaràːs ‘Ulaanbaatar (ablat.)’ The dichotomy between languages that treat CVC as heavy, like Yana, and those that treat CVC as light, like Mongolian, accounts for the greatest source of cross-linguistic variation in weight criteria. In his survey of syllable weight in 408 languages, Gordon (2006) identifies 136 with weight-sensitive stress systems. Of these 136 languages, he finds 42 that treat CVC as heavy and 35 that treat CVC as light. The next most frequent type of weight distinction, which is based on the distinction between schwa and other vowels, is only found in 12 languages. Recent years have witnessed a significant increase in the exploration of the phonetic correlates of syllable weight (e.g. Maddieson 1993, Hubbard 1994, 1995, Broselow et al. 1997, Goedemans 1998, Ham 2001, Gordon 2002a, 2005) with much of this research focused on identifying the phonetic basis for the language specific parameterization of CVC weight. Studies have demonstrated a link between phonological weight and a number of phonetic properties, both acoustic and perceptual. For example, Broselow et al. (1997) show that differences in the weight of CVC syllables between languages correspond to differences in duration: languages with light CVC display closed syllable vowel shortening whereas those with heavy CVC do not. They claim that this duration difference reflects a difference in moraic structure, such that the vowel and the coda consonant share a mora in languages * The authors gratefully acknowledge comments from the audience at WCCFL 26 and the assistance of our language consultants without whom this work would have been impossible. The research presented here was funded by National Science Foundation Grant BCS-0343081. © 2008 Matthew Gordon, Carmen Jany, Carlos Nash, and Nobutaka Takara. Proceedings of the 26th West Coast Conference on Formal Linguistics, ed. Charles B. Chang and Hannah J. Haynie, 208-216. Somerville, MA: Cascadilla Proceedings Project.