ANAPHORIC PRONOUNS: UNDER LINGUISTIC CONTROL OR SIGNALLING PARTICULAR DISCOURSE REPRESENTATIONS?
Francis Cornish · Journal of Semantics · 1986
The article is a contribution to the debate between Tasmowski & Verluyten (1982, 1985) and Bosch (1983, 1984, 1987) as to how the form as well as the interpretation of anaphoric pronouns is determined. TV rightly criticize B's tests as to whether a particular third-person pronoun is functioning semantico-syntactically or referential-anaphorically; however, their examples and arguments do not warrant the conclusion that there is no substantive distinction to be drawn between the two types of pronoun use. Many of TV's examples in this connection merit further analysis, which leads to very different conclusions from the ones they arrive at. There is not a single dichotomy between two types of pronoun use, but a dine, the crucial factor differentiating each position on the cline being the degree to which the pronoun's discourse referent or its intension is presupposed by the speaker. In section 2, I argue that the traditional notion ‘antecedent’, as espoused by TV, should be abandoned, and that it is in terms of the discourse model representation by means of which each discourse referent is encoded in the discourse model that anaphoric pronouns refer. Finally, in section 3, the role of the ‘agreement’ of anaphoric pronouns in gender and number is examined, and the conclusion is drawn that this is not a necessary condition for pronominal anaphora. Referential-anaphoric pronouns are relatively independent indexical expressions, and their gender and number features may be manipulated by the speaker to achieve a variety of types of reference to a particular discourse referent. Suggestions as to fruitful areas for future research in the field of pronominal anaphora are derived from the foregoing discussion.