|
内容記述 |
Three levels of adequacies for linguistic analysis have been discussed in the framework of generative grammar so far ; that is, observational adequacy, descriptive adequacy, and explanatory adequacy. In addition to these adequacies, a higher level of linguistic analysis was pointed out by Chomsky (2004) : how perfect is the language designed? This is appealed to the language universality hypothesized under generative grammar. At the same time, the differences among languages have been analyzed as well, which is well known as parameters. If we assume that there are parameters in a language design, then, it is natural that the realization of parameters in a derivation be coherent ; it shouldbe an ideal model for a language design that any parameters are realized at the same point in the course of derivations such as PF or Spell Out. However, this is not the case. Some are realized in PF, like head parameter, and other are differentiated by the existence of an EPP or an OCC feature in the derivation. I will claim in this paper that such a model is not desirable for a language design in Minimalist Program, considering the process of language acquisition by children. Furthermore, I will point out that the head parameter, which is assumed to be applicable in PF component, has an effect on not only the word order but also the nature of movement in the language, in spite of the claim that PF rules are not core in a derivation. I conclude, following Baker (2001), that such a situation shows that there must not be any connection between head parameters and the culture in which a language exists, and that parameters should be reduced to the natures of the vocabulary, which are to be acquired in the process of acquiring language by children. |
KJ00005595179.pdf (819.6 kB)
