On the empirical basis and formal derivation of penultimate stress systems
Bert Vaux
November 2021

The Classical Sanskrit stress system is of theoretical and empirical interest due to its apparent four-syllable stress window, yet there has only been one attempt to analyse it formally (Keydana 2016), which fails to generate the facts it is designed to account for. Existing analyses of other systems with a right edge stress window that can extend to a fourth syllable have special conditions (e.g. only LLLL) that enable modeling with binary feet. This doesn’t work for Sanskrit. The existence of the Sanskrit sort of system is also problematic for Kager’s stress window typology. I propose a grid-based analysis a la Halle and Idsardi which doesn’t encounter these problems. It also doesn’t encounter the problems Kager, McCarthy, etc. are trying to avoid, especially the midpoint pathology, due to the different character of RBP vs OT, where e.g. RBP has no competition between L and R alignment. I suggest moreover that more careful inspection reveals that the empirical and conceptual bases of so-called penultimate systems--both live (Mongolian, Palestinian Arabic) and dead (Plautine Latin, Classical Arabic, Classical Sanskrit)--are suspect and need to be worked out more carefully before larger typological or theoretical conclusions can be definitively reached.
Format: [ pdf ]
Reference: lingbuzz/006824
(please use that when you cite this article)
Published in:
keywords: phonology, perception, sanskrit, stress, arabic, phonology
Downloaded:113 times


[ edit this article | back to article list ]