From correspondence to realism

When once adopted on those grounds, especially by minds that, like that of the early Russell, happen to be sympathetic to empiricism, while at the same time repelled by any form of Idealism, the Correspondence Theory can very easily be reworked to provide the basis of a form of Realism – call it Referential Realism – definable by its opposition to two opposing views: Conventionalism and Relativism. The Referential Realist shares Locke's conviction that what we say has a bearing on reality only if its content is “conformable” to “some real being.” The Referential Realist need not, of course, commit himself to any view concerning the identity of the ultimate content-bearing elements of a language. It is indifferent to him, for instance, whether he sets up his position in terms of names and predicate-expressions, or in terms of sentences, or in terms of theoretically articulated collections of sentences. His claim is, simply, that only if some semantic contents, whatever linguistic entities they may attach to, in some way correspond to, or mirror, actually existing elements of Reality, will it be possible, in the language concerned, to construct propositions having a bearing on reality. Conventionalism, from the Referential Realist's point of view, is the claim that all the entities picked out by the content-bearing expressions of a natural language might be linguistic constructs: entities wholly constituted by linguistic convention. Referential Realism is not, of course, incompatible with the claim that some may be.

Realism Restored

“To say that self-sufficient thought always refers to a thought enmeshed in language,” says Merleau-Ponty, “is not to say that thought is alienated or that language cuts thought off from truth and certainty.” If we have done nothing else in the foregoing pages, we have offered a rich series of amplifications of that teasing remark. Analytic philosophers, though, are apt to take a dim view of talk of thought being enmeshed in language. The atmosphere of many departmental coffee-rooms when that sort of thing comes up is nicely caught by J. A. Fodor in a passage which we have found occasion to quote already (Chapter 11 §ⅶ n. 20).

Philosophers since Russell have, broadly speaking, taken a commitment to Referential Realism, in one or other of its many forms, to be essential to the preservation of the sort of “respectable Realism” Fodor here invokes.

Paradox and understanding

In “A Puzzle About Belief,” Saul Kripke argues that linguistic moves to all appearances normal in reporting the beliefs of others can be shown to generate paradoxical results. The paradox lies in the impossibility of giving a straight answer to this question: Does Pierre, or does he not, believe that London is pretty? We can neither say that Pierre believes London to be pretty nor that he believes it not to be pretty; worse still, we cannot say that he believes neither that it is nor that it is not pretty; and certainly we cannot say that he believes both that it is and that it is not pretty. As straightforward as this outline of the options appears, it is mistaken. There is indeed a straight answer to the question: viz., that Pierre believes that London is not pretty and that Londres est jolie.

In this chapter, we show that the supposed paradox is one in appearance only, and that the appearance rests on a Kripke's covert vacillation between two conceptions of linguistic understanding, a weak, or “minimal” one, and a “strong” one. The weak conception allows Kripke to set up the example that allegedly generates his paradox; but only the strong allows the generation of a philosophically significant paradox. However, a conception of linguistic understanding strong enough to generate the paradox turns out to be strong enough to block it.

From Mill to Kripke

So far we have presented the argument as an attack on the Orthodox View defined in Chapter 5 ii. The Orthodox View, which at any rate until quite recently deserved the name, was there defined as a compendium of three doctrines: (1) the doctrine that the meaning, or Bedeutung, of a proper name is the individual it picks out; (2) the doctrine that a speaker can attach no assertoric content to an indicative sentence employing a proper name N unless he or she knows which individual N does, in fact, pick out; (3) the doctrine that what places a speaker in a position to meet the condition spelled out by (2) is the possession of an identifying description of the individual in question. That collection of doctrines would, no doubt, fairly adequately capture the content of philosophical orthodoxy between 1930 and 1970. It does not, however, at least in its entirety, capture the views of a majority of philosophers at present. Doctrines (1) and (2) are still widely held. What has changed is that there is now widespread scepticism concerning doctrine (3).

In order to simplify the exposition of our argument in its early stages, our account has so far ignored the inherent tension between doctrine (1) and doctrine (3). Doctrine (1) contends that the meaning of a proper name “N” is its bearer, N. But if that is so, how can knowing the meaning of a proper name amount to knowing some predicate P which N appens to satisfy, as doctrine (3) contends?

One or two systems?

There is much that is common between the view so far presented and ‘relevance theory’. The emphasis upon the psychological dimension of utterance interpretation, and the rejection of Minimalism, are among the important features that the two frameworks share. But there are also some differences.

One difference that has been recently the focus of attention on the part of relevance theorists is this. According to me, the primary pragmatic processes involved in comprehension are not ‘inferential’. Only when the unreflective, normal process of interpretation yields weird results does a genuine inference process take place whereby we use evidence concerning the speaker's beliefs and intentions to work out what he means. There is no doubt that our ability to do so is an important part of our conversational competence, but the question at issue is: how essential is that inferential ability? Can linguistic communication proceed without it, at some basic level, or is it from the very start constituted by it? Following Millikan and Burge, I reply that communication is not constitutively inferential. As Burge writes,

The basic triad

Anyone who has reflected on the sentence meaning/speaker's meaning distinction knows that a simple distinction is in fact insufficient. Two equally important distinctions must be made. First, there is the distinction between the linguistic meaning of a sentence-type, and what is said (the proposition expressed) by an utterance of the sentence. For example, the English sentence ‘I am French’ has a certain meaning which, qua meaning of a sentence-type, is not affected by changes in the context of utterance. This context-independent meaning contrasts with the context-dependent propositions which the sentence expresses with respect to particular contexts. Thus ‘I am French’, said by me, expresses the proposition that I am French; if you utter the sentence, it expresses a different proposition, even though its linguistic meaning remains the same across contexts of use.

Second, there is a no less important distinction between what is actually said and what is merely ‘conveyed’ by the utterance. My utterance of ‘I am French’ expresses the proposition that I am French, but there are contexts in which it conveys much more. Suppose that, having been asked whether I can cook, I reply: ‘I am French.’ Clearly my utterance (in this context) provides an affirmative answer to the question. The meaning of the utterance in such a case includes more than what is literally said; it also includes what the utterance ‘implicates’.

Mandatory vs optional

What characterizes contextual ingredients of the optional variety is the fact that their contextual provision is not mandatory – it is not required in virtue of a linguistic convention governing the use of a particular construction (or class of constructions). In context, it may be that that ingredient is ‘required’; but then it is required in virtue of features of the context, not in virtue of linguistic properties of the expression-type. A contextual ingredient is mandatory in the relevant sense, and is provided through saturation, only if in every context such an ingredient has to be provided (precisely because the need for saturation is not a contextual matter, but a context-independent property of the expression-type). This, then, is the criterion we must use for deciding whether a contextual ingredient results from an optional pragmatic process or from saturation: can we imagine a context in which the same words are used normally, and a truth-evaluable statement is made, yet no such ingredient is provided?

To illustrate the contrast between the two types of pragmatic process, let us consider the phenomenon of ‘null instantiation’ (to use Fillmore's terminology), where the direct object of a transitive verb is not syntactically realized, or at least not overtly. There are two sorts of case, which must be sharply distinguished. In indefinite null instantiation (INI), the argument role corresponding to the direct object is existentially quantified instead of being assigned a particular value.

Around the middle of the twentieth century, there were two opposing camps within the analytic philosophy of language. The first camp – ideal language philosophy, as it was then called – was that of the pioneers, Frege, Russell, Carnap, Tarski, and so on. They were, first and foremost, logicians studying formal languages and, through them, ‘language’ in general. They were not originally concerned with natural language, which they thought defective in various ways; yet, in the 1960s, some of their disciples established the relevance of their methods to the detailed study of natural language. Their efforts gave rise to contemporary formal semantics, a very active discipline whose stunning developments in the last quarter of the twentieth century changed the face of linguistics.

The other camp was that of so-called ordinary language philosophers, who thought important features of natural language were not revealed but hidden by the logical approach initiated by Frege and Russell. They advocated a more descriptive approach and emphasized the pragmatic nature of natural language as opposed to, say, the language of Principia Mathematica. Their own work gave rise to contemporary pragmatics, a discipline which, like formal semantics, developed successfully within linguistics in the past forty years.

Central in the ideal language tradition had been the equation of, or at least the close connection between, the meaning of a (declarative) sentence and its truth-conditions. This truth-conditional approach to meaning is perpetuated, to a large extent, in contemporary formal semantics.

In this book we have been concerned with the contribution context makes to truth-conditional content. How essential is that contribution, and how much controlled by linguistic conventions? These are two of the main questions we have dealt with, in an attempt to revive the debate between Literalism and Contextualism. In this chapter we will consider a fundamental dimension under which context contributes to truth-conditional content: the so-called ‘circumstance of evaluation’.

Truth-evaluation (or semantic evaluation more generally) requires not merely a content to evaluate, but also a ‘circumstance’ against which to evaluate that content. As Austin once put it, ‘it takes two to make a truth’. The circumstance of evaluation is not an aspect of the content to be evaluated, but an entity with respect to which that content is evaluated. Still, according to the theory of situations to be introduced in this chapter, the circumstance of evaluation is an aspect of content in a broader sense of ‘content’. And that aspect of content is irreducibly contextual.

Modality

The notion of circumstance of evaluation is familiar from modal logic. In modal logic, propositions are evaluated relative to ‘possible worlds’. The possible worlds are necessary to truth-evaluation, but they are not themselves represented in the propositions that we evaluate. Thus ‘I am French’ (said by me) is true, with respect to a world w, iff I am French in w; but the sentence ‘I am French’ only talks about me and the property of being French.

It may be thought that the pragmatic approach to what is said advocated in the first four chapters of this book blurs the commonsensical distinction between what is (literally) said and what is non-literally conveyed. What is said, in the pragmatic sense, already incorporates the derived, non-literal values resulting from primary pragmatic processes of the optional variety. Is there still room for a contrast between literal and non-literal speech? Wilson and Sperber think this distinction belongs to folk-linguistics and must be repudiated. In this chapter, however, I will attempt to show that the ordinary, folk-theoretical notion of non-literal use can be rescued. It corresponds to a range of phenomena which constitute a natural class, from a phenomenological point of view.

Non-literal uses as non-minimal departures from literal meaning

Let us start with a sense of the phrase ‘literal meaning’ which is reasonably clear and raises no particular problem. In that sense, the literal meaning of a linguistic expression is its conventional meaning: the meaning it has in virtue of the conventions which are constitutive of the language. Thus understood literal meaning is a property of the expression-type; for it is the expression-type which the conventions of the language endow with a particular meaning. Literal meaning, in that sense, I will dub ‘t-literal meaning’ (with ‘t’ standing for ‘type’), in order to distinguish the sense just introduced from other possible senses of the phrase ‘literal meaning’.

The modulation of sense

In context the meaning of words is adjusted or ‘modulated’ so as to fit what is being talked about. Sense modulation is essential to speech, because we use a (more or less) fixed stock of lexemes to talk about an indefinite variety of things, situations and experiences. Through the interaction between the context-independent meanings of our words and the particulars of the situation talked about, contextualized, modulated senses emerge, appropriate to the situation at hand. The meaning of a word can thus be made contextually more specific, or it may, on the contrary, be loosened and suitably extended, as in metaphor. It may also undergo ‘semantic transfer’, etc.

According to many authors among those who have studied the phenomenon, modulation is the process whereby the meaning of a given word is affected by the meanings of other words in the same sentence. Thus the meaning of the adjective ‘light’ is affected by the meaning of the noun it modifies: a light lunch is not light in quite the same sense in which a piece of luggage is said to be light. According to Jonathan Cohen, this is one of the big differences between natural language and formal languages: ‘artificial languages satisfy an insulationalist account whereas natural languages require an interactionist one’.

Five positions

I take it that there currently are five basic positions concerning the role of context in the determination of truth-conditions. I have already attempted to refute two of these positions. Before proceeding, it will be helpful to identify all five, and to take a bird's eye view of the theoretical landscape.

The five positions I am about to list can be ordered on a scale, one end of which is occupied by the extreme position I call Literalism, and the other end by another extreme position, which I call Contextualism. Of the intermediate positions I will describe, some fall on the literalist side and others on the contextualist side, depending on where they are situated on the scale (see table 6.1, p. 86).

The debate between Literalism and Contextualism was at the forefront of attention in the philosophy of language of the middle of the twentieth century. It is widely believed to have been settled (in favour of Literalism in some version or other) but I think that is a mistake. The alleged refutations of Contextualism that have been offered actually refute nothing; while Literalism strikes me as by and large indefensible. So I think the history of twentieth-century philosophy of language ought to be rewritten. I will not do so in this book, however.

Enrichment, loosening and transfer

Secondary pragmatic processes are ‘post-propositional’. They cannot take place unless some proposition p is considered as having been expressed, for they proceed by inferentially deriving some further proposition q (the implicature) from the fact that p has been expressed. In contrast, primary pragmatic processes are ‘pre-propositional’: they do not presuppose the prior identification of some proposition serving as input to the process. Another difference is the fact that secondary pragmatic processes are conscious in the sense that normal interpreters are aware both of what is said and of what is implied and are capable of working out the inferential connection between them. Primary pragmatic processes are not conscious in that sense. Normal interpreters need not be aware of the context-independent meanings of the expressions used, nor of the processes through which those meanings are enriched or otherwise adjusted to fit the situation of use. Unless they are linguists or would-be linguists, they are aware only of the output of the primary processes involved in contextual adjustment.

Saturation is a primary pragmatic process. If the uttered sentence is ‘She is smaller than John's sister’, then in order to work out what is said I must (at least) determine to whom the speaker refers by the pronoun ‘she’ and what the relevant relation is between John and the mentioned sister. Were saturation a secondary pragmatic process, I would have to proceed in reverse order, that is, to identify what is said in order to determine those things.

At the beginning of chapter 6, I said that the debate between Literalism and Contextualism, which stood at the forefront of attention in the middle of the twentieth century, is widely, and wrongly, believed to have been settled in favour of (an attenuated version of) Literalism. My aim, in this book, has been to revive that debate, and to show that Contextualism is still a live option. By ‘Contextualism’ I mean the view according to which it is speech acts, not sentences, which have a determinate content and are truth-evaluable: sentences themselves express a determinate content only in the context of a speech act. In this concluding chapter I will summarize the discussion and deal with a few residual issues.

Alleged arguments against contextualism

First, I should say something of the philosophical arguments which, in the sixties, led to the demise of Contextualism. Two arguments, in particular, have been so generally taken to constitute a complete and final refutation of Contextualism, that I cannot close this book without saying what is wrong with them. One argument, due to Peter Geach, makes use of what he calls the ‘Frege point’. The other, due to Paul Grice, invokes a principle which he calls ‘Modified Occam's Razor’.

The Frege Point is the view that ‘a thought may have just the same content whether you assent to its truth or not; a proposition may occur in discourse now asserted, now unasserted, and yet be recognizably the same proposition’.