사이시옷 첨가의 준거와 사이시옷의 기저 요소 유무에 대한 고찰

Abstract

I studied two subject in this thesis. The first is a problem about a criteria judging the saisiot insertion in a compound, the second is about wether inserted saisiot is underlying elements or not. So far, many people have judged saisiot insertion by two phenomena. The ons is a glottalization of initial consonant of the following noun in a compound. The other is [n] insertion between two constituents in a compound. But that isn't a conditioned environment of saisiot insertion. Because glottalizaion and [n] insertion is just one of the resultant phenomenon of saisiot insertion. Also it is difficult to prove that a glottalization of initial consonant of the following noun and [n] insertion in a compound is caused just in saisiot insertion. That is to say, other reason is able to be related in that phenomena. To be present, Many people agree that the saisiot insertion is decided according to a semantic relation of front ant rear constituents of compound. If that is true, I think, it is right that we judge wether saisiot is inserted or not by a semantic relation of front ant rear constituents of compound. Additionally, I raised a problem whether the inserted saisiot is underlying element or is an element inserted during derivation process. In relation to this problem, I suggested that we need to classify compounds two sorts. the one is reconstructed after saisiot insertion, the other is non-reconstructed. The latter is a case that the saisiot is inserted during derivation process. At present, I don't show a criteria sorting two cases clearly. But I just suggested that the saisiot insertion is diachronic phenomenon at least when the C1 is voiced consonant /n, m, ŋ, l/ in the environment of (C)VC1#C2V(C).