A STUDY ON PROJECTION FROM LEXICAL CONCEPTUAL STRUCTURE

本文研究了动词的词汇语义及其向句法结构和词结构的投射。本研究的主要发现是词汇概念结构向非动词复合词的投射，在对非动词复合词的研究中，我们发现，对屈折变化提出的规则/联想记忆二分法也可以推广到某些复合词;也就是说，通过使用这种区分，我们已经能够将日语VC分类为参数VC和调整VC，并解释了它们不同的生产力以及迄今为止在这些复合词的研究中未解释或未注意到的各种对比行为。特别是在附加/结果VC的词汇语义方面，研究发现，尽管这类动词的解释似乎是随机的，但我们可以用词汇概念结构（LCS）来系统地分析它们。由于人们普遍认为附加语的位置是在主成分论元结构的层面上确定的，因此，附加语的语义必须在主成分论元结构中作为谓语的名物化和谓语所选择的附加语的修饰来捕捉。通过这样做，我们可以解释他们的歧义以及他们的差异，从英语的VC参考不同的谓词和选择的谓词在LCS这些动词的基础上。

In this study we have studied the lexical semantics of verbs and their projection to both syntanctic structures and word structures. The nobable findings of this study is in the investigation of the projection from Lexical Conceptual Structure to deverbal compounds.In the study on deverbal compounds we have found that the rule/associative memory dichotomy that has been proposed for inflection can be extended to some case of compounding as well ; that is, by using this distinction, we have been able to classify Japanese VC into argument VC and adjuct VC,and account for their different productivity as well as various contrasting behaviors which hitherto have been left unexplained or unnoticed in the existing studies on these compounds.In particular, in regard to the lexical semantics of adjunct/result VCs, it has been found that although this type of VCs appear to have random interpretations, we can analyze them systematically using Lexical Conceptual Sturcture (LCS). Since it is a widely held view that adjunct positions are soecified at the level of LCS ant not argument structure, the semantics of adiunct VCs must be captured at LCS as nominalization of a predicate and modification by the adjunct selected by the predicate. By doing so, we could account for their ambigities as well as their differences from the English VCs by lreferring to different predicates and selected adjuncts kin LCS these VSs are based on.