抄録
Throughout the 20th Century, linguists have always been looking for the best way to properly describe and explain interfaces among branches of linguistics. This has been inevitable partially due to the double- (or n-tuple-) faced nature of linguistics. Among various trials and errors, the paths drawn by generative linguistics are worth special attention. Its "modular" view of language has made possible a tremendous amount of new findings which were not possible before. However, this "modularity" issue must now come under scrutiny, especially within the components of grammar.
