Fusion and Other Segmental Processes in Malay: The Crucial Role of Prosody
Committee: Adamantios Gafos (Chair), Harry van der Hulst, John Singler, Mark Baltin, Ray Dougherty
The variegated data relating to nasal substitution (or fusion) in Malay have long escaped unified analysis. Past analyses have sought to explain the occurrence of fusion, assuming without exception that the conditioning environment is morphological. I argue that the answer lies unquestionably in consideration of prosody. Crucially, I claim that the demand for feature recoverability interacts with the prosodic structure of the language. I assume as primitives of phonological representations the notions of "headedness" and "dependency" espoused in works on dependency and government based models of phonology. The head-dependent asymmetry I focus upon relates to the level of the word. Thus I claim that fusion is the combined result of licensing only the features of a potential coda segment, and an opposing pressure for dependent foot syllables to be as light as possible, i.e., open. By contrast, fusion is blocked in head foot syllables, which I claim are unmarked when closed (Weight-to-Stress Principle, Prince and Smolensky 1993). I claim further that the segments of head foot syllables must express all featural information of their corresponding input segments. Head feet are thus "privileged domains" (Beckman 1998, Alderete 1995). As a consequence of my analysis, all data relevant to fusion that have hitherto been cited as exceptions are, instead, predictable.
While seeking independent support in the language for this analysis, I find evidence that, far from demanding special treatment, fusion is more appropriately seen as just one instance of a more general pattern, where many segmental changes occur in response to head-dependent asymmetries at the level of the prosodic word. I combine this substantive theory of representations with a theory of constraint interaction to show that a ranking of head foot constraints over general phonotactic and prosodic constraints accounts for the distribution of all these processes, not just fusion.