Special meeting of the Phonology Group today: Manami Hirayama

There will be a special meeting of the phonology group next today (September 6th) at 10 am in SS560A. Manami Hirayama (Ritsumeikan University) will be presenting her research on high vowel devoicing in Japanese; all are welcome to attend. The title and abstract follow below.

Testing the visibility of morphology in postlexical phonology: Evidence from vowel devoicing in Japanese

It has been argued in the literature of high vowel devoicing in standard Tokyo Japanese (HVD) that HVD makes reference to morphological boundaries. For example, McCawley (1968) observes that accent shifting in verb and adjective alternations triggered by HVD has the stem as the domain. Vance (1992), looking at nominal compounds with two potential undergoers of HVD in two successive syllables, argues that the presence of the boundary prohibits HVD: those adjacent to the boundary fail to undergo the process. Yoshida (2004) argues that what he calls “compound boundaries” blocks the occurrence of HVD more frequently than “morpheme boundaries” do.

The visibility of word-internal structure in HVD is a challenge to phonological theories. For example, in the theory of Lexical Phonology (e.g., Kiparsky 1982, Mohanan 1986), processes in the postlexical domain do not refer to word-internal structure (only phonological information is available in that domain), while HVD, being non-structure-preserving and non-categorical, occurs postlexically. The reference to the word-internal information is also not expected in the prosody-morphology/syntax interface hypothesis, where phonological/phonetic rules refer only to prosodic structure, not directly to morphological/syntactic structure.

In this talk, I explore the effects of morphological boundaries in HVD through two studies I have conducted. One is preliminary results from a production experiment with ten speakers. The other is a dictionary study on the lexical accent variation related to HVD (Hirayama & Giriko 2012). I argue that it is not the morphological boundary but likely the phonological boundary that HVD refers to. This supports the view of Lexical Phonology and the prosody-morphology/syntax interface hypothesis discussed above.