Page 464 - Encyclopedia of Philosophy of Language

Pragmatics and Speech Act Theory
relations between utterance function and form. It can- not be accepted to the extent that the felicity con- ditions can only be heterogeneous and involve also the extralinguisticworld, since they concern the place in which language and world meet—the com- municative situation.
In a first, extremely sketchy approach, it is possible to distinguish between constitutive and regulative rules (Searle 1969; for a discussion of the notion of 'rule,' see Black 1962). There are presuppositions act- ing as constitutive rules concerning the what: the per- formance of a given act. There are presuppositions acting as regulative rules concerning the how:style, degree of indirectness, politeness; Keenan (1971) was especially thinking of these.
Further, there are pragmatic presuppositions con- cerning: (a) communication in general, linked to the utterance act (e.g., preparatory conditions in the sense of Fillmore, absence of obstacles to communication, sharing a code); and (b) pragmatic presuppositions specific to the type of speech act. In any case 'we cannot hope to understand the way someone is using the word presupposition unless we are able to discover what is thought to result when a presupposition fails' (Garner 1971:27).
In order to achieve this purpose, it would be useful to devise a typology of different kinds of pre- suppositional failures which is based on Austin's (1962) theory of infelicities (for semantic treatments see Vanderveken 1991). In order to overcome the dis- tinction between constative utterances (true or false) and performative utterances (happy or unhappy), Austin identifies different ways in which both can fail, to conclude that assertions too, as acts, are subject to the infelicity of performatives. Conditions can be divided into two groups, according to the different kind of consequence brought about when something goes wrong with one of them. The first group is formed by those conditions which, if disregarded, make the act void (which is the case with misfires: the act is purported but void). The second group is made up of those conditions whose breach makes the act an abuse (the act is professed but hollow). This criterion dis- tinguishes A and B conditions on the one hand and F conditions on the other (the Greek letter is used by Austin to stress the difference in nature of the latter
conditions). A conditions concern the existence of a procedure (Al) and the right to invokeit (A2), whereas B conditions regard execution, which must be correct (Bl) and complete (B2).
If A conditions (misinvocations) are not satisfied, the act is void. A failure in the act of reference belongs to this type. According to Austin, in the case of failure of an existential presupposition (e.g., 'G's children are bald, but G has no children,'), the utterance is void: the referential failure leads to an infelicity analogous to a failure of A conditions for performatives. Speci- fying Austin's characterization, one can establish with
Donnellan (1966) a difference between referential and attributive uses of definite descriptions. One can say 'The man in the purple turtleneck shirt is bald' (the example is Stalnaker's 1970) to qualify a specific person, Daniels, as bald. This is an example of ref- erential use of the definite description which in such cases is just a means to identify the person about which one wants to speak, e.g., picking him out from a group. The same utterance can also be used when one, through ignoring who actually wears the purple turtleneck shirt, wants to qualify him as bald. This is an example of attributive use of the definite descrip- tion. As Stalnaker (1970:285) underlines, the conse- quences of a failure of referential presuppositions on the success of the speech act are marginal in the sense that the speaker 'may still have successfully referred to someone,' even if there is no entity that satisfies the description (if there is no man wearing the purple turtleneck shirt). The consequences of a failure of attributive presuppositions are, on the contrary, dra- matic because 'nothing true or false has been said at all.' Or, one can suggest, with Lyons (1977:602), a link between existential presupposition and theme: 'It is only when a refering expression is thematic that
failure of the existential presupposition results in what Strawson [...] would call a truth-value gap.'
Also, a nonfulfillment of Austin's B conditions (misexecutions) makes the act void: Austin has in mind the ambiguity of reference in particular, the use of vague formulas and uncertain references (p. 36) (Bl, flaws) or the lack of uptake on the part of the hearer (B2, hitches), especially as regards acts that require a specific uptake. ExtendingAustin's idea, one could consider also some conversational phenomena, such as turn taking and topic change or, in textual linguistics, the different conditions of coherence and cohesion as examples affecting the correctness and completeness of the execution of a procedure.
A nonfulfillment of F conditions, on the contrary, leads to an abuse: the act is performed, but is unhappy. This is the case with different kinds of insincerity, for example when not believing what one is saying (Fl, insincerities), in the possible world opened up by the utterance, or in breaching a commitment later on (F2, 'breaches of commitment'), by not keeping a promise or, as far as assertions are concerned, by being inco- herent in the world following the issue of the utter- ance.
5. Toward a Pragmatic Definition of Pragmatic Presupposition
The typology of Austin's infelicities is a philosophical approach to the problem of linguistic action, seen as a kind of social action. The types of infelicity are a step forward with respect to the recurring, undifferentiated notions, such as that of 'appropriateness,' in the defi- nition of pragmatic presupposition. They help dis-
442