This paper explores the theoretical consequences of surface-to-surface correspondence constraints in Optimality Theory, in comparison to a derivational approach employing sequential derivational mechanisms like cyclicity and ordered default rules. The empirical object of investigation are the facts of the g-N(angma) alternation and variation in Tokyo Japanese. The central argument of the paper is that the OT analysis is better equipped to handle the facts of the alternation, which are truly intricate and interesting, since an (at first glance) simple process of allophony is complicated by subpatterns of optionality vs. obligatoriness, governed by factors such as the presence vs. absence of related words and the influence of Rendaku voicing.
No takes yet. Share an insight, caveat, or question.
Itô et al. (1996) studied this question.