Speech acts and conversational interaction By Michael L. Geis (review)

REVIEWS Speech acts and conversational interaction. By Michael L. Geis. Cambridge, New York & Melbourne: Cambridge University Press, 1995. Pp. xiv, 248. Reviewed by Peter Tiersma, Loyola Law School, Los Angeles In this work Geis proposes a new approach to speech acts which he calls dynamic speech act theory (DSAT). Although based to some extent on traditional speech act theory as developed by J. L. Austin and John Searle, DSAT seeks to incorporate the insights of conversation analysis and artificial intelligence research, as well as recent work on politeness and register. Ultimately, the aim is to produce a 'conversation machine' that can account for the conversational competence of speakers. G begins Ch. 1 with a brief overview of traditional speech act theory. One of the shortcomings of speech act theory, according to G, is that its practitioners have generally focused on constructed examples such as the following: (I)I request that you give me hot chocolate with whipped cream. Of course, there has been a great deal of discussion of more indirect ways of saying the same thing, such as: (2)Can I have a hot chocolate with whipped cream? Yet requests do not occur in isolation, as examples (1) and (2) suggest but rather as part of multiturn sequences such as: (3)Customer: Do you have hot chocolate? Clerk: mmhmm Customer: Can I have hot chocolate with whipped cream? Clerk: Sure ((leaves to get)). Although one might be able to analyze the conversation in (3) as consisting of several distinct speech acts, each associated with one tum, G demonstrates that requests and the like frequently require more than one turn to be accomplished. For example, a person who wants to achieve the goal of requesting a ride from someone might begin a telephone conversation by informing the intended ride-giver that his car is stalled, tell her in a later tum that he needs to be at a particular destination, and subsequently ask her whether she is busy or plans to be near his required destination anytime soon. Because all of these actions are part of his strategy of requesting a ride, it seems arbitrary to identify any single utterance as constituting the speech act of requesting. G's conclusion is that traditional speech theory does not deal all that well with naturally occurring conversations. For reasons such as these, G aims to avoid the association of speech acts with particular linguistic constructions, positing instead that actions like requesting or making offers are best viewed not as speech acts but rather as communicative actions (12). Yet I am not quite ready to toss the notion of speech acts entirely, and I expect many others would react similarly. Even though it is true that many prototypical speech acts seem to consist of single sentences dreamed up by philosophers in ivory towers, and thus may present an idealized vision of how speakers accomplish acts like requesting or promising, the work of speech act theorists based on such isolated sentences has proved enormously insightful, even if their data are somewhat artificial. Perhaps more fundamentally, not all language consists of colloquial conversations. While G makes a good case for the notion that speech acts have limited relevance in the analysis of conversation, in less conversational varieties of language (consider legal language or formal written language), speech acts are still a very useful analytical tool and can more sensibly be linked to specific linguistic structures than is true in most conversations. In Ch. 2, 'Meaning and force' (33-53), G elaborates on his point that what is important about an utterance is not what type of action it performs, or what type of illocutionary force it has, but instead 'what it contributes to the work of the interaction in which it occurs' (33). He posits 137 138LANGUAGE, VOLUME 74, NUMBER 1 (1998) a separate level of meaning in conversation: s-meaning, the significance of an utterance event, similar to what is sometimes called 'pragmatic meaning'. An utterance can have either transactional significance, or interactional significance, or both. G then explains why utterance generation , as well as politeness and register variation, requires this level of meaning. Ch. 3 is entitled 'The structure of communicative interactions' (54...