Quotational Constructions

The utterance of any expression x ostends or makes manifest the customary referent of x, x itself, and related matter. If x appears in quotation marks then the presumed intention behind the utterance is to pick out something other than the customary referent (either instead of it or in addition to it). The consequences of these ideas, taken from my 1998 work, are here drawn out in application to a variety of quotations: metalinguistic citation, reported speech, scare-quoting, echo-quoting, loan words, and titles. My earlier work on quotation makes several claims – call them theses, if you will. (T1) Every speech act creates multiple ostensions. For instance, in saying hippopotamus one ostends, or makes manifest, a sequence of sounds or letters, a lexeme, a concept, a referent, and more. (T2) Ostension makes it possible for a speaker not only to draw attention to such ostensibles, but to recognizably intend to do so. (T3) Consequently any expression may be either used or mentioned. (T4) Quotation marks are used to signal mentioning, and thus serve to disambiguate. (T5) Quotations may be used for many reasons, e.g. reporting speech, scare-quoting, metalinguistic citation. (T6) Quotations, i.e. quotation marks plus quoted matter, refer to quoted matter; quotation marks by themselves do not. (T7) At least some versions of the Identity Theory are mistaken. (T8) The Description Theory is mistaken, quotations are not descriptions. (T9) The Demonstrative Theory is mistaken, quotation marks are not demonstratives. (T10) The Name Theory is mistaken, quotations are not names. These claims variously call for elaboration, fresh justification, and emendation. (T10) The Name Theory no longer has any proponents so far as I can tell. Philosophy does make progress! (T9) New arguments against the Demonstrative Theory are so considerable as to fill their own paper: Saka (2004b). (T8) Description Theories ultimately invoke proper names, demonstratives, or displays and hence should respectively be classified as versions of the Name, Demonstrative, and Identity Theories. Treating them as a distinct and unified group, as I and others did earlier, is infelicitous. (T7) 188 QUOTATIONAL CONSTRUCTIONS Though I continue to find the Identity Theory’s label troubling, I have acquired deeper appreciation for its substance from Washington (1992) and Reimer (1996). My own work goes beyond theirs, possibly in ways they would disapprove of, but it belongs to the same family. It also has affinities with Geurts & Maier (this volume). (T6) To be precise, speakers refer; expressions do not. Language consists of speech acts and their underlying competence, not lifeless soundwaves/inkblots and not abstract objects. Although it is sometimes convenient to speak as if expressions refer, such talk is loose and potentially misleading. The speech-act/competence point of view captures that which denotational/truth-conditional semantics misses – that interpretations are constructed dynamically and imaginatively. (T5) Quotations may be used for assorted purposes. Yet these purposes, by and large, are special cases of one phenomenon. (In contrast, for instance, capitalization serves to mark proper names and sentence beginnings, which are unrelated: we do not want a unifying theory of capitalization.) The bulk of this paper is dedicated to showing how different kinds of quotation can be explained on the basis of the same underlying principles, namely T4 – the central thesis of my original paper – and T3. T2, a peripheral point, is justified by T1 and further defended in Saka (2004b). Section 1 situates T1 in my larger Deweyan (naturalistic, irrealist, and open-ended) outlook. Section 2 defends T2 and T3. The remainder of the paper is devoted to T4-T6, especially T5. To the extent that the present work is plausible, it supports T7-T10. 1. Ostension and Construction: Principle (P) Every time you vocalize you vibrate the air, and in so doing you make manifest to your audience certain phonetic phenomena. The phonetic phenomena, in turn, activate phonological representations in the audience, which in turn trigger recognition of lexical entries including conceptual content and projections of syntactic structures (multiple sets thereof in the case of lexicosyntactic ambiguity), eventuating in referential models. What’s more, the route from sound to referent is robust. Thus, I can plead “don’t think of ferocious furry-footed caterpillars”, and you will think of ferocious furry-footed caterpillars. All I have done is present some ink, but the ink, or more directly speaking your perception of it, willy-nilly caused you to think of something other than the ink. In short, speech acts directly ostend sounds and thereby deferringly ostend, according to principles of association, lexemes and mental models (for more on ostension, see Saka 1998: 125).