|Title:||Anti-Homophony Blocking and its Productivity in Transparadigmatic Relations|
|Comment:||2006 Boston University doctoral dissertation|
|Abstract:||This dissertation addresses 'anti-homophony blocking' in transparadigmatic relations, where an application of a particular phonological process is blocked in order to avoid homophony creation by neutralization of distinct inputs between morphologically unrelated words.
Past research was concerned with anti-homophony blocking but only within the inflectional paradigm. The possibility that this principle is also applied to transparadigmatic relations has not been pursued. In recent literature, anti-homophony constraints in paradigmatic relations have been proposed (Crosswhite 1999, 2001, among others) within the framework of Optimality Theory (Prince and Smolensky 1993). However, no attempt has been documented that proves that anti-homophony blocking is in fact a productive process. I examine these two key issues: first that anti-homophony blocking applies to transparadigmatic relations; second that it is productive, using a case of anti-homophony blocking in Japanese.
The main data comes from "contracted forms" (Kikuzawa 1935, Toki 1975) in derived environments in Japanese, created by syncope along with lenition or deletion of the adjacent consonant. Within the framework of Optimality Theory, I will demonstrate that the contraction process and anti-homophony blocking in transparadigmatic relations are accounted for by particular constraints and ranking specific to the contraction grammar. I propose an anti-homophony constraint called CONTRAST, which is integrated into the contraction grammar. Analyses are given as to why homophony is created in inflectional morphology, as it could be counterevidence to my claim of anti-homophony blocking. I will argue that the anti-homophony principle must be phonology-internal which is embedded in the phonological grammar.
I conducted an experiment to test the extent to which anti-homophony blocking is part of the phonological grammar of Japanese, which provides some evidence in support of the claim that contraction and anti-homophony blocking are productive processes. Using a Japanese corpus, I found that there is no positive influence of word frequency and word familiarity on the occurrence and blocking of contractions.
This dissertation concludes that anti-homophony blocking is not limited to an inflectional paradigm but also occurs in transparadigmatic relations, and it is part of the productive phonological grammar.