A Comparative Study of Scramblings

语法：通过与G.Grewendorf、J.Sabel(日耳曼语)、Z.Boskovic(斯拉夫语)、S.Miyagawa(阿尔泰语)和J.Jayaseelan(德拉维亚语)的广泛联合研究，明确了日语乱码的突出特性。在这些属性中，最基本的是运动没有在语义表示中反映出来。在此基础上，对日语混杂现象进行了具体的分析。这一分析发展了极简主义理论，通过删除句法特征来循环解释连锁链，并成功地解释了日语乱码的主要特征，包括词根重建性质、着陆点的A/A混合性质以及踪迹严格遵守适当的约束条件。最后，有人提议解释为什么日语而不是其他语言会有这种乱码。该假说认为，这是因为日语在短语结构构建中使用最纯粹的“合并”形式。这也解释了日语中的其他独特现象，如轻动词结构和泛化论元省略。研究结果：对2~4岁的被试进行了一些实验。结果表明，在非常早期(大约2岁)就已经习得了打乱，这与文献中报道的发现相矛盾。特别是，研究结果清楚地表明，扰乱的习得比被动的早得多，这支持了扰乱不是由特征驱动的假说。这一结论与句法研究中的观点是一致的，支持了日语由于在短语结构构建中使用合并而具有独特性质的观点。

Syntax : The outstanding properties of Japanese scrambling were made clear through extensive joint research with G. Grewendorf, J.Sabel (Germanic languages), Z.Boskovic (Slavic languages), S.Miyagawa (Altaic languages), and J. Jayaseelan (Dravidian languages). The most fundamental among those properties is that the movement is not reflected in semantic representation. A concrete analysis of Japanese scrambling was proposed on the basis of this result. The analysis develops the minimalist theory so that chains are interpreted cyclically by deletion of syntactic features, and it successfully explains the major properties of Japanese scrambling, including the radical reconstruction property, the mixed A/A' nature of the landing site, and the strict obedience of the trace to the proper binding condition. Finally, a proposal was made to explain why Japanese, and not other languages, has this type of, scrambling. The hypothesis is that this is due to the fact that Japanese employs 'Merge' in its purest form in phrase structure construction. This accounts for other unique phenomena in Japanese as well, for example, the light verb construction and extensive argument ellipsis.Acquisition : Some experiments were conducted on the acquisition of scrambling with 2〜4 year old subjects. The results indicate that scrambling is already acquired at a very early stage (about 2 years of age), in contradiction with the findings reported in the literature. In particular, the results show clearly that scrambling is acquired much earlier than passive, and this supports the hypothesis that scrambling is not feature-driven. This conclusion is in perfect harmony with and hence, supports the proposal in the syntactic research that Japanese has scrambling with unique properties because of the way it employs'Merge' in phrase-structure building.