Token-reflexivity is a notion introduced by H. Reichenbach to treat indexical expressions in his Fregean descriptive semantic system. Reichenbach claims that the semantical functions, such as reference and truth conditional import, depend on the properties of the uttered token itself. On the other hand, D. Kaplan constructed his own Theory of Direct Reference for indexical expressions, as revising Reichenbach's problematic ideas including token-reflexives, using some technical maneuver. Namely, Kaplan abstructed the notion of "token", which is originally extralinguistic entity, to the notion of "occurence in a context", and brought it as a type into his own system. There seems to be a difference between Reichenbach's Theory of Token-Reflexives and Kaplan's Theory of Direct Reference: the former requires semantical conventions specific to the indexical expressions, and the latter inquires a comprehensive convention common to indexical and non-indexical expressions. Although Kaplan's type-oriented theory has been regarded as solving some serious problems implicated by Reichenbach's token-oriented theory, it is confronted with another problem with so-called "true demonstratives", which requires non-linguistic demonstration to fix the referent. Of course the token-oriented theories have also some difficulties as seen in the Reichenbach's or D.H. Mellor's. It seems to me that this problem shows a serious and ingrained dilemma between type-oriented semantical theory and token-oriented theory; between a pure semantical theory and a hybrid theory containing extra-linguistic entities in it.

投稿論文