In this paper, it has been shown that theoretical necessity requires the newly defined operation Move in Chomsky (1998, 1999) and the Multiple Spell-Out framework proposed by Chomsky (1998) and Uriagereka (1999), etc. With such theoretical backgrounds...
In this paper, it has been shown that theoretical necessity requires the newly defined operation Move in Chomsky (1998, 1999) and the Multiple Spell-Out framework proposed by Chomsky (1998) and Uriagereka (1999), etc. With such theoretical backgrounds, I examined Ochi's (1999) PF merger analysis, given that the distributed morphology (cf. Bobaljik 1994, 1995, Lasnik 1994) should be adopted. In order to account for the non-cyclic property of adjunct within the Multiple Spell-Out model, it is assumed that adjunct adverb like quickly is inserted into the derivation acyclically. However, in order not to derive the example like Who did John kissed?, it was proposed that for a simple interrogative sentence, only C can have a verbal affix and PF merger with its tense morpheme is realized via do-support.
In addition, I also explored how to interpret the several outcomes of multiple Spell-Out. As an alternative for Ausin's (1999) matching principle, the Multiple Spell-Out Condition was proposed. This condition makes every cyclic maximal node of all functional categories a candidate for Spell-Out, while only the last outcome of Spell-Out get the legitimate interpretation among all the outcomes. However, vP is optionally spelled-out since we assumed that it is spelled-out only when its outer SPEC is filled with the moved element like an object after feature agreement.