This paper is research on change of the meaning of the verb which it originally has by the meaning of a complement. The sentence showing transitivity will turn into an objective type reflexive usage sentence, if a body noun comes to an objective complement. If a body noun comes to a dative complement, it will become a dative type reflexive usage sentence, and if a body noun comes to a benefactive complement, it will become a benefactive type reflexive usage sentence. The subdivision of these three types was carried out further. The semantic element which becomes the standard of a classification is [±visibility], [±relativity] and [±resultativity]. Each type can be classified into the further subsort type according to the combination of these meaning element.

The purpose of this paper is to investigate the distributional and functional differences among ‘caki, casin, and caki casin’ in the Korean sentences. In general, Korean reflexive ‘caki’ is known to be capable of both short-distance and long-distance binding, while ‘casin and caki casin’ seem to have a higher tendency to short-distance binding (Yang 1986, Hong 1986, Kim 2003 and others). On the other hand, it has been researched that ‘caki’ can be used to represent an antecedent such as a ‘statue’ depicting someone, but it can not be replaced by ‘casin, and caki casin’ (Lim 1987, Jackendoff 1992, Lidz 2001) In this paper, I discuss these issues in comparison with the characteristics of se-anaphor and self-anaphor in European languages. After these discussions, I will try to directly examine the examples cited in the research. Experiments were conducted with about 12 students who major in Korean language and 8 students do not, to examine whether the interpretations of the sentences in this study are correct.

This study investigates Korean speakers' interpretations of the Korean polymorphemic reflexive caki-casin with respect to interactions between its traditionally assumed syntactic requirements and pragmatic conditions for its interpretational preferences. Despite the fact that caki-casin is poly-morphemic, the study on the interpretations of caki-casin in this paper shows that the behavior of caki-casin differs from that of the English poly-morphemic himself in that it is sensitive to the pragmatic information for its interpretations. With manipulation of pragmatic information, Korean speakers showed a tendency to choose a non-local antecedent or even a non-c-commanding one despite the presence of syntactically legitimate local or c-commanding potential antecedent.

The purpose of this paper is to show the internal structures of Korean reflexives such as 'caki, casin and caki casin.' According to Boskovich (2012), Korean is an NP language in which the determiner is positioned in the specifier of NP. In this paper, following his suggestion, it is claimed that 'caki' appear in the position of the specifier of the NP and 'caki casin' is placed in the specifier and head of the NP.: it follows that 'casin' is derived in the head position of the NP. With regards to some problems DP analyses of the Korean reflexives pose, it is pointed out that since Korean is not a DP language, the determiner cannot be the head of the relevant phrase and that since Korean is a head final language, even if Korean would be a DP language, the head D should be in the rightmost position. Lastly, this paper shows the complex structures combined the structures of 'caki. casin, caki casin' with the structural cases such as '-ul/lul' under the structure of the NP.