本論文根據對日文例子的分析判斷以及對語意(Kuroda 2005)、資訊結構(Vallduví 1992)、理想理論(Prince and Smolensky 1993)和資訊限制(Choi 1999)的概念,發現日文的WA有五種功能(Completive Focus, Contrastive Tail, Link, Contrastive Link, Contrastive Focus)以及GA有三種功能(Completive Focus, Contrastive Focus, Tail),並提出運用在理想理論的限制等級是:LINK > PROM > NEW,只要是有「link」功能的WA就一定要在句子的最前面。透過上下文分析、WA-GA組合、語意、資訊結構、限制等級以及SOV的字序,決定了攪拌現象的可能性,並且也解釋日文局部攪拌現象的字序。這些發現也許能延伸到雙賓動詞並且也將對於嘗試要學習日文WA和GA有很大的幫助。
Based on judgments of Japanese data and the notions of Kuroda’s (2005) Semantics, Vallduví’s (1992) Information Structure, Optimality Theory (OT) (Prince and Smolensky 1993), and Choi’s (1999) Information Constraints, this thesis discovers that WA has five functions (Completive Focus, Contrastive Tail, Link, Contrastive Link, Contrastive Focus) and GA has three functions (Completive Focus, Contrastive Focus, Tail), and I show the ranking of constraints is: LINK > PROM > NEW applied in a typical OT chart. When WA is a link, the WA-phrase must be sentence-initial. Through context, WA-GA combination, semantics, information structure, ranking, and default SOV order, they determine scrambling possibilities and explain word orders in Japanese local scrambling. The findings may be able to be extended to double object construction and will also be useful for people trying to learn the particles WA and GA in Japanese.