Page 196 - Encyclopedia of Philosophy of Language

Truth and Meaning
however, that the point is not to insist on a distinction, but to correct an erroneous conception of (semantic) rules. This is suggested in other passages, where it is stressed that the determination of the use of a word by rules is not complete; new situations can always arise which are not covered by the rules (so-called open-texture).
2.5 Pragmatic Rules
The notion of pragmatic rules is one of rules governing linguistic activity in respects other than those of syntax and semantics, in respects peculiar to communication. This is an area where it is difficult to distinguish fea- tures characteristic of linguistic practice as such from other factors, such as (everyday)humanpsychology, cultural or social norms, and contexts; indeed the desirability of employing that distinction is also ques- tionable. Nonetheless, elaborate pragmatic analyses in terms of linguistic rules have been carried out, above all concerning individual speech acts and fea- tures of conversational interaction.
J. L. Austin (1976) introduced the notion of illo- cutionary acts, speech acts such as asserting, prom- ising, commanding, congratulating. A speech act belongs to one of these categories by virtue of intrinsic properties, as opposed to properties depending on further reactions of the hearer, by virtue of which the act can be characterized as an act of scaring, amusing, or persuading. In contradistinction to acts of these kinds, called perlocutionary, illocutionary acts were held by Austin to involve conventions, but he did not develop this idea. It was later developed by Searle (1969), who extracted a number of rules for various kinds of illocutionary act. As regards promising, Sear- le's main example, there are five rules, understood as governing the use of expressions such as I promise and other linguistic devices indicating the illocutionary type of promising (and in virtue of this Searle charac- terizes these rules as semantic). The first of these requires that such an expression be uttered only in the context of predicating a future action of the speaker (the propositional content rule), the second that it be uttered only if the hearer prefers performance of that action to nonperformance and the speaker also believes this of the hearer, the third that it be uttered only if that action would not obviously be performed anyway (preparatory rules), the fourth that it be uttered only if the speaker intends to perform the action (the sincerity rule). These four rules are held to regulate the practice of promising. The fifth, on the other hand, the so-called essential rule, is seen by Searle as constitutive of that practice, as a rule which makes promising possible. This is the rule which (pro- vided that the requirements of the first three rules are met) holds that such an utterance counts as under- taking an obligation to perform the action in question. These rule statements no doubt capture standard fea- tures of promisory utterances, even though Searle's
conception of the rules themselves, especially the fifth, has been subject to discussion.
A different kind of pragmatic rule is the one which Grice (1989) has labeled conversational maxims. It includes rules such as: Make your contribution as informative as required (for the current purposes of the exchange)! Do not say that for which you lack adequate evidence! Be relevant! By means of reference to such rules Grice explains varieties of so-called con- versational implicature, as in the phenomenon of deliberately conveying or implying something else than one is literally saying.
3. Is Linguistic Practice Ride-governed?
The idea that linguistic practice is essentially rule- governed, that the meaning of linguistic expressions is determined by rules, is intimately connected with a conception of linguistic capacity as a kind of knowl- edge. On this conception a speaker stands in a cog- nitive relation to his own mother tongue; his ability to use it is a way of knowing the meaning of its expressions. Given the further idea that the meaning of a linguistic expression is arbitrary, the speaker's knowledge must be a knowledge of rules. These two tenets, about determination of meaning by rules and about knowledge of one's language, are almost invariably discussed together.
3.1 Language as Conventional
The most basic conception of meaning as determined by rules is that of meaning as conventional. One argu- ment against this (Davidson 1984) is that we can give an account of what a speaker means by his words without requiring that the speaker knows the meaning of his words, and hence without requiring that he knows conventions. The basic point of another well- known argument (Quine 1976) is that, since pro- ponents of the view must ultimately appeal to unstated conventions, the claim that speakers go by con- ventions runs the risk of becoming empty (for more on this, see Convention).
Sellars, too (1974), stresses that only antecedently stated rules can be said to be obeyed. His conclusion, however, is that for this reason rules which determine meaning must be so-called ought-to-be rules (compare Sect. 2.4). In this way Sellars hopes to avoid a regress: knowledge of language requires knowledge of rules, which is knowledge of rule formulations, which in turn requires knowledge of language, and so on. However, since just conforming to ought-to-be linguistic rules, in Sellars's view, falls short of constituting under- standing, it is not clear that the regress is really avoided.
3.2 Wittgenstein on Normative Linguistic Practice
In Wittgenstein's view (1958) there are other ways of stating rules than that of providing full-fledged verbal
174