type:text

This paper defends the traditional view that words can refer (and not only language users do so), and that the reference of words are essentially linked to the understanding of sentences in which they occur. It examines certain issues involving the concept of reference reflected in three views that have been aired in recent years. 1) The revival of the so-called Russellian thoughts by Gareth Evans and John MacDowell, namely that a person's thinking that he has a thought about a particular object does not guarantee that he has one. They have claimed that certain causal relations with the object has to obtain between the language user and the object, in addition to the correct use of a singular proposition to think about it. 2) The belief that, singular propositions containing proper names, as well as propositions with names of natural kinds, play a special role in linking language to the world, and that this is so because of a causal story involving the naming event and the subsequent handing down of the correct use of the name. 3) The negative view expressed more recently by Putnam and Davidson in different ways that since we have no guarantee that words can succeed in referring to a determinate object, one should do without reference; and the linked (but, actually not related, negative view), that reference presupposes a confrontation of a object identified independently of any conceptual system with our language and thought, a view implied by Putnam and Rorth. If reference implies such an impossible correlation, we would naturally have to cast aside reference. The paper tries to show that reference does not, and cannot imply such a confrontation of language with objects presumed to be identified independently of any conceptual system, and that causal links have little to do with reference. We claim that one can fix the identity of the object designated by a name or any singular term more and more precisely, by using general propositions. This does not entail that these general propositions give the sense of the name. The possibility that the same natural kind word could refer to a different stuff in a different possible world does not show that it is an additional causal link that enables us to refer to the actual natural kind, or that the inadequacy of the causal theory should make us give up reference. If we hold on to our belief in access to reality and to truths, then we cannot give up reference as Davidson suggests we do, in his 'Reality without Reference' of 1977.

1. 序 2. 一般命題と単称命題 3. 単称命題とレアリズム 4. 対応と指示 5. 因果関係と指示

文学部創設百周年記念論文集I Treatise