Abstract
This paper claims that the word order restriction observed in Old Japanese (OJ) does not indicate that Japanese was once a wh-movement language, counter to Watanabe's (2002) Minimalist analysis. The apparent wh-movement effect is epiphenomenal of the interplay of several constraints sensitive to the profiles of case morphemes. More specifically, the word order restriction reflects a kind of mismatch where the nominal case framed is imposed upon a clause. This curious OJ word order restriction and its loss in Middle Japanese (MJ) simply reflect the gradual and dynamic development of the morphological case system, rather than a change of wh-parameter. The change in the profile of case particles is partly lexical, but it has systematic consequences on the surface syntactic structure. With the mixed category analysis, adopted from Malouf (2000), and Boersma-type OT allowing optionality, the proposed analysis gives a comprehensive picture of the diachronic facts of the language.