Page 446 - Encyclopedia of Philosophy of Language

Pragmatics and Speech Act Theory
turn, implies making a distinction between the 'gram- matical' and the 'user' point of view on the basis of how language and context relate, whether they do this with, or without, grammar's helping hand. The important notion of context, however, and the role that it plays in the expression of grammatical and pragmatic relations, is not addressed.
Still, the difference between a 'grammatical' and a 'user-oriented' point of view is precisely in the context: on the first view, one considers linguistic elements in isolation, as syntactic structures or pans of a gram- matical paradigm, such as case, tense, etc.; whereas on the second, the all-important question is how these linguistic elements are used in a concrete setting,that is, in a context.
A definition of pragmatics that limits 'context,' and references to context, to what is (or can be) gram- matically expressed has, of course, a big advantage: it excludes a number of irrelevant factors from the scope of investigation. One could formally define pragmatics as 'the study of grammatically encoded aspects of contexts,' establishing a function 'that assigns to utter- ances the propositions that express their full meaning in context' (where proposition is to be understood as the logical equivalent of a sentence), or, alternatively, 'a function from utterances to contexts, namely the contexts brought about by each utterance' (Levinson 1983:31). Among the irrelevant, excluded factors, one could mention (to take a classic example) the presence of food in the mouth while speaking; this may be part of some context, yet it is not a linguistic factor, and maybe not even a pragmatic one.
As Levinson has it, 'the main strength of this defi- nition of pragmatics is that it restricts the field to purely linguistic matters' (1983:11). So far, so good: but restricting the field of pragmatics to 'purely linguistic matters' does not seem to be too interesting a definition from apragmatic point of view. Certainly, not all 'extralinguistic' factors can always and every- where be safely excluded from a pragmatic evaluation. A truly pragmatic consideration has to deal with the context as a user's context, and cannot limit itself to the study of grammatically encoded aspects of contexts, as the 'grammaticalization requirement' seems to imply.
The next section examines the grammaticalization problem in more detail.
2.3.4 Why Grammar Fails: Implicatures and Pre- suppositions
In order to establish why a 'strictly grammatical' definition of pragmatics must fail, two cases will be looked into that are often used to illustrate the problems of the 'grammaticalization' of supposedly pragmatic relationships. The cases in question are con- versational implicature and presupposition; together they constitute two of the most important disputed
areas in the borderland of logic, semantics, and prag- matics.
By conversational implicature is meant the principle according to which an utterance, in a concrete con- versational setting, is always understood in accord- ance with what one can expect in such a setting. Thus, in a particular situation involving a question, an utter- ance that on the face of it does not make 'sense' can very well be an adequate answer. If speaker A asks speaker B
What time is it?
it makes perfectly good sense to answer
The busjust went by,
given a particular constellation of contextual factors, including the fact that there is only one bus a day, and that it passes B's house at 7:45 each morning; furthermore, that A is aware of this, and that A takes B's answer in the 'cooperative spirit' in which it was given, viz., as a relevant answer to a previous question. Notice, however, that there are no strictly 'gram- maticalized' items in this interchange that could be identified as carriers of such information about the context. Hence, under the interpretation of prag- matics-as-strictly-grammatical, such relevant infor- mation about the users and their contexts is not taken into consideration (cf. Levinson 1983:98).
The other exampledeals withpresuppositions. It is by no means always the case that pragmatic meaning is linguistically (that is, grammatically) encoded; usually, it is said to be 'presupposed' (Lakoff's cat and assorted other animals, referred to in Sect. 1.4, are cases in point).
Compare also the following pair of utterances: Gorbachev called Yeltsin a real Marxist, and then Yeltsin
insulted him and
Yeltsin called Gorbachev a real Marxist, and then Gor- bachev insulted him.
The first sentence seems normal, given what is known from the early 1990s about the two statesmen and feelings about eastern European leaders. In the second sentence, however, Yeltsin is trying (rather unsubtly, perhaps) to shower praise on his former president— so Gorbachev could retort with an insult. But even though the second sentence is odd, there are no gram- matical means of establishing or recognizing that odd- ity: the presuppositions are all hidden.
Things get even worse in the following pair of sen- tences:
Bush called Gorbachev a real Marxist, and then Gor- bachev insulted him
and
424