The chief concern of this dissertation is to investigate a fundamental, yet unsolved problem within the phonology of Proto-Indo-European (PIE): the process of syllabification. I show that by analyzing the much more easily reconstructable word-edge clusters we may predict which types of consonant clusters can occur word-medially, provided that we assume a special status for certain consonants at word’s edge. Having thus analyzed the entire PIE phonological system, I believe I have developed the first working hypothesis of Indo-European syllabification, which we may now use to pre- dict which types of syllable-driven rules of consonant deletion and vowel epenthesis occurred within PIE. My dissertation argues that there existed at least five phonologi- cal processes of this type. The second half of the dissertation focuses on the problem of Sievers’ Law, through which I argue for the tendency in PIE to keep morphemes syllabically distinct, in accordance with a high-ranking constraint ALIGN. I conclude by proposing that the assumption of morphological relevance in the syllabic derivation provides us with a mechanism to reconcile the well-established principle of ONSET MAXIMIZATION with the reconstructable parsing of VCCV sequences as VC.CV.
Byrd, Andrew M., "Reconstructing Indo-European Syllabification" (2010). Linguistics Faculty Publications. 55.