Speakers have a variety of referring devices for introducing new characters into the universe of discourse, such as proper names, definite and indefinite descriptions, etc. In this paper I shall discuss the prerequisites for the speakers’ preference of bare proper names over proper names accompanied by some sort of categorizer, extension, or a modifier, as in the following examples: Ja včera vstretil Serežu vs. Ja včera vstretil vašego Serežu vs. Ja včera vstretil vašego prijatelja Serežu.
It is generally believed that among the conditions for the use of bare proper names are the interlocutors’ familiarity with the referent, the ability of both interlocutors to identify the referent with a proper name, and the referential knowledge being retrievable by the addressee. While all of these conditions are necessary for the use of bare proper names and, in fact, some of them are stronger than others, neither of them constitutes a necessary and sufficient condition. It is impossible to explain the difference between the examples above by means of these conditions. Rather, I would like to argue that it is the interaction of the strongest condition (that of retrievability) with other pragmatic factors, such as the type of knowledge and the type of relationships between the interlocutors and the referent that make the choice of a bare proper name likely and/or desirable. The purpose of this paper is to formulate (soft) rules for the likelihood of the use of a bare proper name and to define discourse situations where more than a bare proper name is needed or where a proper name is inappropriate to refer to a person, despite the fact that both interlocutors know the referent and his/her name. This is obtained by a calculus of discourse situations that takes into account all the relevant factors. The procedure would be, then, to scan discourse situations which differ only in one factor, while keeping all the others equal.
The three types of knowledge/familiarity postulated as relevant for the choice of a referring device are empirical knowledge, encyclopedic knowledge, and knowledge by hearsay.
The social and personal status of the interlocutors is defined in relative terms as being “socially higher/lower,” “socially equal and close,” “socially equal and remote,” etc.
The degree of familiarity of the interlocutors with the referent is also defined in relative terms. Two basic situations are relevant: (a) both interlocutors are equally familiar with the referent; (b) one of the interlocutors is more familiar with the referent than the other.
The results of the experiment are viewed within the theoretical framework of the Transactional Discourse Model (Yokoyama) and evaluated against the background of the “classical” and “causal” theories of reference (Russel/Carnap/Searle and Kripke/Donnellan, respectively).