J.L. Austin on Performative Utterances [Alberto Torrigiotti]

The verification theory of meaning developed by A.J. Ayer and others in the Vienna Circle could be accused of committing the descriptive fallacy because it takes as its point of departure the assumption that descriptions with the prospect of being verified as true or false are the only kinds of utterances that deserve to be studied. By stipulating that the meaning of a statement just amounts to its truth value under the conditions of verification, it presupposes the meaninglessness of utterances for which no verification of truth value can be given, thus pretheoretically excluding a significant chunk of utterances. The major focus within the theory then is to study the conditions under which a statement can be verified as being true or false.

J.L. Austin calls attention to the fact that many of our utterances don't set out to be true or false because they don't describe something correctly or incorrectly, but instead perform speech-acts; he believes that these performative utterances are an important part of language use so he doesn't want to call them meaningless like the logical positivists. He avoids committing the descriptive fallacy by denying that these performative utterances are descriptions, thus shifting directions from talk of truth conditions to talk of speech-acts. Austin's theory adds several things to our understanding of language, not the least of which is its ability to account for motivations, other than attempting to say something true, that underlie the use of utterances.

My question for Austin is "What do you make of statements which take the form of speech-acts, e.g., 'I punish you', but which don't (can't?) perform the act they purport to on the occasion of being uttered?" When I say 'I promise you', this utterance alone suffices to perform the illocutionary act of making a promise and the perlocutionary act of assuring one that I will uphold my end of the bargain, but this is not the case if we substitute any number of verbs in the place of 'promise'. Does Austin say anything on this matter? I'm aware that Searle has said a bit in lectures about this topic.