ADAMS AND THE VALIDITY OF CONDITIONAL INFERENCES

From the fact that indicative conditionals have no truth conditions, conventional wisdom will have us inferring that they have no logic. On a standard definition, which I relied on in 8.2, a sentence S follows from a class P of premises iff it is not logically possible for the members of P to be true and S to be false. But this way of thinking surely concedes too much to one modern conception of the function of logical theory. For on the traditional conception logic is the study of reasonable argument; and indicative conditionals show up in arguments both good and bad, so that something needs to be said about how we tell which is which. A theorist of argument who said nothing about indicative conditionals, just because they have no truth-values, would be avoiding a central and familiar class of cases. Indeed, we might well respond by denying that someone whose theory cannot tell us that modus ponens is an acceptable form of argument and affirming the consequent is not, understands what an argument is at all.

OVERVIEW

Because our theory is computational it will treat beliefs as representations; and it says that some mental processes are computations with these and other representations. Every representation which is a belief has also a property which reflects the degree of that belief; and every desire is a representation with a property that reflects its strength; i.e. in the technical sense introduced in Chapter 3, the subjective ‘desirability’ of what is thereby desired. Together these determine utilities, which determine action. What is required by a computational theory of decision is that our minds are so constituted that, given these materials, they compute the expected utilities that underlie our actions. This computation is a mental process, a process which is carried out in us by our central nervous systems, and which, like all processes, takes time. But it is a process which can also go wrong, a process in which computational errors can occur. If the computations are completed, then we come to be in a state of preferring some options to others and this causes us to do one of those most preferred basic actions; and if the computation is without error as well as complete, then that action will be a basic action with maximum expected utility.

OVERVIEW

The central claim of this book is that the meaning of an asserted sentence is, in a certain sense, the content of the belief that it expresses. So I must begin with a story about beliefs and their contents. In fact, as I have said, my story is functionalism, the story encapsulated in the remark of Armstrong's which provides my epigraph. I need to give the outlines of functionalism's picture of the mental because the general account of belief and of the contents of beliefs is needed if I am to make good my central claim. Since my interest is mainly in assertion, and in beliefs as what assertions express, I concentrate mainly on giving an account of the mental state of believing; but because, as I shall argue, that account can only be given in the context of a theory that also takes account of desires, I am obliged also to say something about desire.

The theory I hold is functionalist: for it claims that we can state sufficient a priori truths about the causal roles of mental states in interaction with each other and with events outside the mind, to be able to individuate those mental states by their causal roles.

AFTER BEHAVIOURISM

In the sixth of the Meditations on First Philosophy, which he published in 1641, Descartes expresses the core of the dominant philosophy of mind of the last three centuries:

Introduction

The single most obvious way in which the relationship between language and context is reflected in the structures of languages themselves, is through the phenomenon of deixis. The term is borrowed from the Greek word for pointing or indicating, and has as prototypical or focal exemplars the use of demonstratives, first and second person pronouns, tense, specific time and place adverbs like now and here, and a variety of other grammatical features tied directly to the circumstances of utterance.

Essentially deixis concerns the ways in which languages encode or grammaticalize features of the context of utterance or speech event, and thus also concerns ways in which the interpretation of utterances depends on the analysis of that context of utterance. Thus the pronoun this does not name or refer to any particular entity on all occasions of use; rather it is a variable or place-holder for some particular entity given by the context (e.g. by a gesture). The facts of deixis should act as a constant reminder to theoretical linguists of the simple but immensely important fact that natural languages are primarily designed, so to speak, for use in face-to-face interaction, and thus there are limits to the extent to which they can be analysed without taking this into account (Lyons, 1977a: 589ff).

The importance of deictic information for the interpretation of utterances is perhaps best illustrated by what happens when such information is lacking (Fillmore, 1975: 38–9).

Introduction

In this Chapter we shall be centrally concerned with the organization of conversation. Definitions will emerge below, but for the present conversation may be taken to be that familiar predominant kind of talk in which two or more participants freely alternate in speaking, which generally occurs outside specific institutional settings like religious services, law courts, classrooms and the like.

It is not hard to see why one should look to conversation for insight into pragmatic phenomena, for conversation is clearly the prototypical kind of language usage, the form in which we are all first exposed to language – the matrix for language acquisition. Various aspects of pragmatic organization can be shown to be centrally organized around usage in conversation, including the aspects of deixis explored in Chapter 2 where it was shown that unmarked usages of grammatical encodings of temporal, spatial, social and discourse parameters are organized around an assumption of co-present conversational participants. Presupposition may also be seen as in some basic ways organized around a conversational setting: the phenomena involve constraints on the way in which information has to be presented if it is to be introduced to particular participants with specific shared assumptions and knowledge about the world. The issues touch closely on the distinction between given and new (see e.g. Clark & Haviland, 1977), and concern constraints on the formulation of information (that is, the choice of just one out of the indefinitely many possible descriptions of some entity – see Schegloff, 1972b), both of which are important issues in conversational organization.

Introduction

In these conclusions we shall try to tie together some of the loose strands of thought that have run through this book, by considering the relation between pragmatics and other disciplines. One discipline will stand noticeably absent: philosophy, the ‘prodigal provider’, cannot easily re-absorb the empirical studies that it has spawned (but cf. Atlas, 1979). The general tenor of this book has been the description of how, from original, mostly philosophical concepts, a series of empirical modes of investigation have developed, which jointly form the climate of the Anglo-American tradition in pragmatics. As the quotation indicates, Austin foresaw, and indeed hoped for, just this development of a field that he, perhaps more than any other single individual, did most to promote.

Introduction

In the previous Chapter we discussed conversational implicature as a special kind of pragmatic inference. Such inferences cannot be thought of as semantic (i.e. as pertaining to the meanings of words, phrases and sentences) because they are based squarely on certain contextual assumptions concerning the co-operativeness of participants in a conversation, rather than being built into the linguistic structure of the sentences that give rise to them. We turn in this Chapter to another kind of pragmatic inference, namely presupposition, that does seem at least to be based more closely on the actual linguistic structure of sentences; we shall conclude, however, that such inferences cannot be thought of as semantic in the narrow sense, because they are too sensitive to contextual factors in ways that this Chapter will be centrally concerned with.

The reader should be warned of two things at the outset. The first is that there is more literature on presupposition than on almost any other topic in pragmatics (excepting perhaps speech acts), and while much of this is of a technical and complex kind, a great deal is also obsolete and sterile. The volume of work is in part accounted for by a long tradition of philosophical interest which, because it is much referred to in the linguistic literature, will be briefly reviewed in 4.1.

Introduction

Of all the issues in the general theory of language usage, speech act theory has probably aroused the widest interest. Psychologists, for example, have suggested that the acquisition of the concepts underlying speech acts may be a prerequisite for the acquisition of language in general (see e.g. Bruner, 1975; Bates, 1976), literary critics have looked to speech act theory for an illumination of textual subtleties or for an understanding of the nature of literary genres (see e.g. Ohmann, 1971; Levin, 1976), anthropologists have hoped to find in the theory some account of the nature of magical spells and ritual in general (see e.g. Tambiah, 1968), philosophers have seen potential applications to, amongst other things, the status of ethical statements (see e.g. Searle, 1969: Chapter 8), while linguists have seen the notions of speech act theory as variously applicable to problems in syntax (see e.g. Sadock, 1974), semantics (see e.g. Fillmore, 1971a), second language learning (see e.g. Jakobovitz & Gordon, 1974), and elsewhere. Meanwhile in linguistic pragmatics, speech acts remain, along with presupposition and implicature in particular, one of the central phenomena that any general pragmatic theory must account for.

Given this widespread interest, there is an enormous literature on the subject, and in this Chapter we cannot review all the work within linguistics, let alone the large and technical literature within philosophy, from which (like all the other concepts we have so far reviewed) the basic theories come.