Reference, Existence and Truth in Discourse.

Abstract

It is a long established and still respectable claim in the linguistic discipline that sentences containing reference to non-existent objects have no truth-value. This thesis is an attempt to provide a richer and more accurate account of the interesting relations which connect the existence or non-existence of objects and the expressions by which speakers attempt to refer to them with the range of truth-values assigned to sentences containing such expressions. After an introductory chapter whch defines the main terms used in the thesis and discusses preliminary issues, the second chapter is taken up with a critical review of the history of presupposition in linguistics. The important early theoretical contributions are surveyed and the relevant later theories discussed, particularly those analyses which acknowledge the complexities of the relationshp between failed presuppositions and truth-values . These are evaluated with respect to their empirical and theoretical adequacy and the chapter concludes with a summary of the outstanding problems. Chapter three contains a discussion of determiners focusing on the existential aspects of their meaning. An analysis of existential force as a scalar phenomenon is proposed and examples of each of the types of determiner distinguished by the proposed existential scale are examined. h the fourth chapter, questions surroundmg the nature of the existence of objects are discussed. It is proposed that different types of existence are viewed in terms of different existential locations; and that these existential locations can be modelled as possible worlds. Some preliminary questions concerning the nature of possible worlds are addressed and arguments from the linguistic and philosophcal literature in favour of viewing possible worlds as existent entities or as abstract constructs are reviewed. The chapter concludes by defending the modal realistic stance whch maintains that all possible worlds actually exist. Chapter five contains a detailed discussion of the members of the set of possible worlds and describes a rich structure whch can be imposed on the set in the form of accessibility relations of several types: counterpart relations which define inter-world proximity; temporal links whch can be used to identifL distinct temporal stages of worlds as chronological counterparts; and familiarity relations which connect individuals to sub-sets of worlds via epistemic llnks. The next chapter presents a theory of how referring expressions are interpreted by means of locating their referents in possible worlds. The processes involved in accessing worlds of all types are discussed and then a set of rules is laid out which governs the choice of referential location on the basis of the relative accessibility of competing potential reference worlds, where accessibility is determined by the structure of llnks imposed on the set. The application of the rules to a range of increasingly complex sentences is discussed. The final chapter of the thesis is concerned with the potential problems and wider implications associated with adopting the location theory of reference presented, First, it is shown that two related types of sentences, those containing deictic pronouns and names, which were problem cases for previous theories, can be accounted for. We then discuss the implications, concerning the failure of Leibniz's Law, which arise from treating all referring expressions as picking out their extensions. Finally there is a discussion of the implications of adopting the framework set out in the current thesis for the interpretation of a wider range of predicate types and further areas of research are suggested.