Descriptive names : a contribution to the semantics of referring expressions
A theory of descriptive names is developed and defended against several objections. Descriptive names pose an interesting challenge to any theory of reference, since they possess both features of proper names and definite descriptions, i.e. of expressions which are often considered to be radically different. These features are referentiality and descriptive sense. The thesis takes as its point of departure Gareth Evans's theory of descriptive names, improves upon it and discusses several other authors and related theories along the way. Chapter I provides an brief introduction to the topic and an abstract of the main lines of argument. Chapter 2 argues that descriptive names possess both referential status and descriptive sense, and that these qualities constitute the two most basic elements of the notion of descriptive reference (which is contrasted with Russellian reference). It is demonstrated that not all names introduced by description are descriptive names, a claim which is given additional substance by a comparison between Evans's and Kripke's accounts of such names. Chapters 3 and 4 deal with two major challenges to the possibility of descriptive names. Chapter 3 explores the possibility of a truth-conditional theory of meaning for descriptive names, but it is shown that if we follow Evans's suggestion that the semantic value of a descriptive name is to be construed according to model theory - namely, as an entity distinct from the referent (a set) - such a theory will result in treating descriptive names as predicates, and thus eliminate then qua referring expressions. Similar accounts given by other authors are also examined and found to be problematic. I conclude by rejecting the model-theoretic notion of semantic value. Chapter 4 addresses a second challenge, posed by the fact that if a descriptive name has a descriptive sense, then given a Russellian analysis of definite descriptions, descriptive names must be quantifier phrases, and thus, again, non-referring expressions. It is argued that if this is true, then the use of negative free logic is unnecessary. Using the idea of rigidified descriptions, it is shown that Evans's arguments, based on modality and simplicity considerations, fail to save both the referentiality and descriptive sense of descriptive names while semantically dissociating them from descriptions. I show that descriptive names can be treated as shorthand for rigidified descriptions and thus semantically on a par with the latter, which, as I demonstrate, is still consistent with Evans's own (convincing) solution of the puzzle of the contingent a priori. Nevertheless, this still does not guarantee the referentiality of descriptive names. Chapter 5 presents in detail the argument that we can only save the referentiality and descriptive sense of descriptive names if we treat definite descriptions as referring expressions. Several negative arguments undermining the most influential defences of the Russellianism are given and three positive accounts of referring descriptions, Wettstein's, Sainbury's and Strawson's, are critically discussed, finally settling, with some proviso, for Strawson's. Finally, the principles of a 'Fregean' free logic for Strawsonian semantics are sketched, and I suggest ways in which a truth theory could be expressed by means of these principles. Chapter 6 summarises the achievements, sketches possible research concerning descriptive names and concludes that the analysis of descriptive names is useful in at least three ways: it provides us with means to, first, solve problems that arise from the introduction of artificial expressions such as descriptive names (e.g. the problem of the contingent a priori), second, to better understand our natural language and its relation to formal theories of meaning, and, last but not least, to give a strong rationale for a referential treatment of definite descriptions. Chapter 7 includes the bibliography and Chapter 8 a list of axioms and formulas.