Performative Linguistics: Speaking and Translating as Doing Things with Words

Citation preview

Performative Linguistics

‘‘This is an exciting book, that raises important questions about the relationship between linguistics and translation. Robinson’s challenging approach is to be welcomed, and the clarity of his arguments adds to the impact of his ideas. You might not agree with everything he says, but you will never be bored for a moment.’’ Susan Bassnett, University of Warwick, UK ‘‘This book lays out a much-needed interdisciplinary map for coordinated advances in translation studies, intercultural studies, and a rejuvenated discipline that should still be called ‘linguistics’.’’ Anthony Pym, Rovira University,Virgili, Spain

J. L. Austin famously distinguished between ‘‘constative’’ utterances that convey information and ‘‘performative’’ utterances that perform actions. In this groundbreaking new book, Douglas Robinson argues that Austin’s distinction can be used to understand linguistic methodologies. Robinson uses Austin’s model to introduce a new distinction between ‘‘constative’’ and ‘‘performative’’ linguistics. Constative linguistics, Robinson suggests, includes methodologies aimed at ‘‘freezing’’ language as an abstract sign system cut off from the use of language in actual speech situations. Performative linguistics, on the other hand, covers methodologies aimed at exploring how language gets used or ‘‘performed’’ in those speech situations. Robinson then tests his hypothesis on a series of complex speech acts, including translation, deception, and allusion, and shows that a performative approach to the study of language can explain ALL linguistic complexities, from the ordinary to the extraordinary. Drawing on a range of language scholars and theorists including Austin, Grice, Peirce, Bakhtin, Burke, and Derrida, Performative Linguistics consolidates the many disparate action-approaches to language into a single coherent new paradigm for the study of language as speech act, as performance ^ as doing things with words. Douglas Robinson is Professor of English at the University of Mississippi, USA. His publications include Who Translates? (2001), Becoming a Translator (Routledge, 1997), Western Translation Theory from Herodotus to Nietzsche (1997), Translation andTaboo (1996), andTheTranslator’sTurn (1991).

Performative Linguistics Speaking and translating as doing things with words

Douglas Robinson

First published 2003 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, NewYork, NY 10001 Routledge is an imprint of theTaylor & Francis Group This edition published in the Taylor & Francis e-Library, 2004. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”

ß 2003 Douglas Robinson All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book has been requested ISBN 0-203-22285-7 Master e-book ISBN

ISBN 0-203-27719-8 (Adobe eReader Format) ISBN 0-415-30036-3 (Print Edition)

Imagine that you enter a parlor.You come late.When you arrive, others have long preceded you, and they are engaged in a heated discussion, a discussion too heated for them to pause and tell you exactly what it is about. In fact, the discussion had already begun long before any of them got there, so that no one present is qualified to retrace for you all the steps that had gone before.You listen for a while, until you decide that you have caught the tenor of the argument; then you put in your oar. Someone answers; you answer him; another comes to your defense; another aligns himself against you, to either the embarrassment or gratification of your opponent, depending upon the quality of your ally’s assistance. However, the discussion is interminable. The hour grows late, you must depart. And you do depart, with the discussion still vigorously in progress. Kenneth Burke,The Philosophy of Literary Form

Contents

xi

Acknowledgments PART I

Performatives 1 Introduction: linguistics and translation studies Constative vs. performative linguistics: a first formulation The problem of translation in the study of language 6 The problem of linguistics in the study of translation 11 Constative and performative linguists of translation 16 A map of the argument 18 2 Constative and performative linguistics

3

3

23

The discovery of the performative 23 The constative recuperation of the performative 24 The performative as crime and scandal 26 Language games 28 A brief history of the constative and the performative 29 Integrational linguistics 34 3 Translatorial performatives The translator’s personlessness Double reception 47 Levels of literature 52

41

42

PART II

Iterations (the performative back-story) 4 Iterability Parasitism 57 Iterability 60

57

viii

Contents

Iterability imaged 64 Translator, iterator 66 A map of Part II 68 70

5 Somatic markers Reading the somatic theory Somatic markers 72 Propositions 77

70

6 The translator’s habitus

82

Norm theory 82 Habitus theory 91 7 Double-voicing

95

Austin’s binary impasse 95 Dialogism 99 Double-voicing 106 Pandemonium 110 Double-voiced translation 113 Language change 115 PART III

Implicatures (performative uptake) 8 Conversational implicature

123

The cooperative principle and its maxims 123 Constative survivals 126 The maxims as cultural habits 128 9 Translation as ideological implicature

132

Constative applications of Grice to translation 133 Venuti’s critique of Grice 138 A Gricean reading of Venuti 142 10 Il-, per-, and metalocutionary implicature Illocutionary and perlocutionary implicature 144 Metalocutionary implicature 148 Translation as self-discovery 153

144

Contents 11 Intendants and interpretants

ix 160

The interpretant triad 161 Interpreters and utterers 164 Performative dialogic 168 12 Conventional implicature and language change

171

Of moons and square-dancing 171 Immediatization and language change 178 13 Conversational invocature

182

I remember 182 J’m’en souviens 189 14 Metalocutionary implicature and cross-cultural misunderstandings

193

Grice’s maxims and cultural relativity 193 Conversations with foreigners 197 The failure of Grice’s original model 199 Cognitive dissonance 202 Metalocutionary implicature 205 Maxim-formation and reimmediatization 211 Metalocutionary iterability 214 Conclusion

217

The constative crisis 217 The next thing? 221 Notes Bibliography Index

225 246 255

Acknowledgments

This book has been some time in gestation, beginning in the early fall of 1987, when I gave a trial lecture at the University of Tampere, Finland, on the difference between constative and performative linguistics. I have also incorporated two early published articles of mine (from 1985 and 1986) into my argument here, in radically transformed ways ^ the material on illocutionary, perlocutionary, and metalocutionary implicature in Chapter 10 and the material on Peirce in Chapter 11 ^ and my thinking on the subject was salutarily shaped by interchanges with the editors of Kodikas/Code/Ars Semeiotica and The Journal of Pragmatics, respectively, as well as with Herman Parret in the late 1980s in connection with my article on Grice in the latter. In 1992 I wrote an early draft of the book without the translation tie-ins; that book bogged down a few chapters after Grice on metalocutionary implicature, and the draft sat on various floppy disks for seven years, waiting for new inspiration, only sporadically discussed with interested souls like Bob Ashley, who ended up reading the new draft as well and, as usual, giving me invaluable feedback. That earlier version also took significant impetus from a graduate seminar I taught on language theory at the University of Mississippi around that same time, for which thanks are due especially to Dave Powell, Debra Rae Cohen, and Dianne Bunch, and from lengthy discussions with other grad students, especially Bill Kaul. The new version of the book likewise arose out of a teaching situation, this time a doctoral seminar in the Department of Translation Studies at the University of Vic in Spain, where I taught in the winter of 1999^2000 on a Fulbright: I assigned Grice’s ‘‘Logic and Conversation’’ and Venuti’s attack on Grice from The Scandals of Translation and then, while preparing for the class, wrote a good deal of what here appears as Part III (most of Chapters 8^10). I would especially like to thank Richard Samson, a colleague who sat in on the course, and my students Cristina Mallol i Macau, Angela Colomer i Font, and Anna Argemi-Catlla¤ de Carson. Thanks also to Martha Tennent, Dean of the Translation Faculty in Vic, and the other colleagues with whom I discussed various translation-related issues while there, especially Ron Puppo, Sheila Waldeck, Victor Obiols, Eva Espasa, Xesca Bartrina,

xii

Acknowledgments

and Maria Gonzalez. And of course thanks are due to the Fulbright Commission for making my five months in Spain possible. One of the great things professionally about being based in Barcelona for five months was my close proximity to Anthony Pym, just a one-hour train ride away in Tarragona. He and I met several times during my stay there and discussed matters performative and implicative at length, both in person and over e-mail and the phone. As my discussion in Chapter 3 should make clear, Pym is the only translation scholar to have tackled the issue of performative utterances in anything like what I would call a performative spirit; all the other translation scholars who have addressed Austinian or Gricean issues have been constative linguists concerned to anchor all of their comments in equivalence. This of course makes speech acts and implicatures things that only the original author is by definition capable of perpetrating: the translator’s only possible interest in Austin or Grice, for these others, lies in recognizing source-text performative and implicative speech acts effectively enough to reproduce them accurately in the translation. After the book was complete I taught a graduate seminar in linguistics based around it at the University of Mississippi, and my students in that class helped me rethink many of my ideas as well, and rewrite a few passages based on our discussions: thanks to Ingrid Haynes, Max Heidel, Carson Chittom, Natal’ya Litvinova, and Lola Kaufova. The last chapter of the book, on cross-cultural miscommunication, grew out of my preparations for a cross-cultural discourse pragmatics class that I taught to 20 seniors at Ole Miss the next spring. Thanks also to Don Dyer, who helped me with the Bulgarian in Chapter 1, and Manon Bergeron, who helped me with the Que¤be¤cois French in Chapter 13. Oxford, Mississippi, 2002

Part I

Performatives We were to consider, you will remember, some cases and senses (only some, Heaven help us!) in which to say something is to do something; or in which by saying or in saying something we are doing something. This topic is one development ^ there are many others ^ in the recent movement towards questioning an age-old assumption in philosophy ^ the assumption that to say something, at least in all cases worth considering, i.e. all cases considered, is always and simply to state something. This assumption is no doubt unconscious, no doubt is precipitate, but it is wholly natural in philosophy apparently. We must learn to run before we can walk. If we never made mistakes how should we correct them? J. L. Austin, HowTo DoThings With Words (1962: 12)

1

Introduction: linguistics and translation studies

Constative vs. performative linguistics: a first formulation Linguistics is the study of language: even etymologically this is an obvious fact. In the twentieth century, however, the term came to signify a single fairly narrow approach to language and to exclude everything else of interest that might theoretically be included within it. Ferdinand de Saussure and Noam Chomsky were linguists; Ludwig Wittgenstein and Kenneth Burke were not. Roman Jakobson was a linguist; Mikhail Bakhtin was not. Emile Benveniste was a linguist; Jacques Derrida was not. The first glimmering of the idea that generated this book came one day when I realized that the students of language whom I found least interesting were called ‘‘linguists,’’and the students of language whom I found most interesting were called something else ^ philosophers of language, critical theorists, literary scholars. I had long pondered the conundrum in my own professional life that language fascinated me but linguistics repelled me: why? If I loved to learn languages and speak and write languages and translate from one language to another and think about language, surely I should love to study linguistics as well? But I didn’t. I read Wittgenstein and Burke and Bakhtin and Derrida on language and was enthralled; I studied linguistics as an undergraduate and postgraduate student, and tried to read in it later as well, and kept throwing it down in disgust. I studied language professionally, published on language, but shuddered at the thought that I might ever be considered a linguist (and certainly never was, by any self-proclaimed linguists among my readers). How could this be? How could the term ‘‘linguistics’’ have become so narrowly specialized, so jealously circumscribed, that avid students of language like myself would shun it, and it them? The two philosophers of language whose theories form the intellectual core of this book, J. L. Austin and H. Paul Grice, have been assimilated to the ‘‘linguistic’’mainstream, but only tentatively and problematically, and rather peripherally. Austin and Grice, it is clear from the remarks of linguists on their work, despite the massive impact that that work has had on linguistic theory, are not ‘‘true’’ linguists. Is it possible, I began to wonder, that ‘‘linguistics’’ might be defined more broadly, more inclusively, so as to cover the full range of scholarship on

4

Performatives

language? Might there be a sense of the term ‘‘linguistics’’ that could usefully and accurately describe the sorts of theoretical and practical work on language done by Mikhail Bakhtin andJacques Derrida? This book was born out of my wranglings with this problem. What if, I wondered, Saussure was wrong and language is not la langue, not a more or less stable sign system, not structure, but something altogether different? What if the whole Saussurean system was wrong: what if the operative methodological choice was not between studying structure and speech (la parole), as Saussure insisted, but between structure and act? These ruminations spawned the distinction that lies at the core of this book back in the late 1980s, at the University of Tampere, Finland. I was reading J. L. Austin’s HowTo DoThings With Words, and puzzling over the moment that puzzles everyone in that book, when halfway through his book Austin discards his exciting and productive performative^constative distinction because all utterances perform actions and convey information. What if, it struck me then, we took the terms to apply not to utterances but to approaches to utterances, linguistic methodologies? Constative linguists would be those interested in stable (‘‘constatic’’) patterns, structures, rules, with la langue, language in the null context, language as a set of structural properties and the logical interrelations among those properties, existing objectively outside of all human cognition and social use and describable using an objectivist methodology based on formal logic; performative linguists would be those interested in actual language use in real-world contexts, in the relationships between actual speakers and writers and actual interpreters, specifically in how humans perform verbal actions and respond to the verbal actions performed by others. Since my own sympathies were overwhelmingly with the performativists, with those who saw language as act, against the constativists who saw it as structure, I was first inclined to dismiss constative linguistics as bad theory, an approach to the study of language based on a philosophically discredited paradigm ^ which is to say, on error and illusion. It was only gradually, through numerous discussions with intelligent and articulate constative linguists, that my thinking on the subject began to move in a more pluralistic direction: to recognize that we need both constative and performative linguistics. In English, as in German, French, Spanish, and other European languages, we always put the definite article in front of the noun and all adjectives: ‘‘The girl was walking through the woods.’’ Always.We would never say *‘‘Girl the was walking through woods the.’’ In Bulgarian that is precisely how they do it: there is a definite particle -to/-ta attached to the end of the noun or the first adjective: (momicheto vu˘ rveshe prez gorata: momiche ‘‘girl’’ þ to, gora ‘‘forest’’ þta). Always. In Finnish, Russian, and other languages the definiteness of a noun is not signaled syntactically at all, and speakers of those languages learning English are hard pressed to understand the necessity of deciding whether something is definite or indefinite. These syntactic structures are not ‘‘acts’’performed by speakers of the various languages; they are more or less stable patterns inherited and used by those speakers. It is

Linguistics and translation studies

5

reasonable to assume that a constative syntactician will have more useful things to say about this stable pattern than performative linguists attuned to situationally contingent acts performed by speakers. As I will show in Part II, a performative linguistics can trace the ‘‘iterative’’acts by which these syntactic structures came to have what relative stability they have, and thus has greater explanatory power in a historical purview than constative linguistics, which tends to assume stability as a matter of course. And cognitive linguists have shown that a conception of syntactic structure as the product of cognitive acts of structuring or categorizing can help linguists solve many knotty analytical problems left by an uncritically ‘‘static’’ or constative conception of structure. However, it should also be clear that constative linguists have developed methodological tools for the analysis of stability that continue to be useful in a wide variety of noncontext-specific cases. We might say: constative linguists tend to be more attentive to those aspects of language that do not change contextually, and that therefore come to seem like stable ‘‘objective’’ structures akin to the foundation of a building; while performative linguists tend to be more attentive to those aspects of language that depend on individual speakers’ and listeners’ perceptions of contextual features and desire to manipulate those features in personally and socially significant ways, and that therefore come to seem like social acts akin to smiling ironically or patting someone on the back. Since constative linguistics has for almost a century been the dominant and indeed almost the only form of language study accepted as ‘‘linguistics,’’ I do not consider the methodological elaboration of constative linguistics to be a particularly pressing task. Anyone who knows anything at all about linguistics as it is taught in most universities today knows basically what I mean by constative linguistics. The almost total exclusion of what I am calling performative linguistics from the discipline of ‘‘linguistics’’makes its elaboration considerably more urgent. Hence, obviously, my titling of the book Performative Linguistics and not, say, Performative and Constative Linguistics. This book is designed to expand the discipline of linguistics to include performative approaches, and focuses on that expansion. I should note, however, that ‘‘constative linguistics’’ takes two rhetorical forms in this book: it is both a modest local quest for stable structures, which will always be an important part of the linguistic project, and a more arrogant universalizing quest for a single dogmatic formalism that will dominate all linguistic analysis. When I praise constative linguistics in what follows, I am imagining it positively in terms of the former modest quest; when I attack it, I am attempting to dislodge the latter arrogant one. To the extent that constative linguists recognize that the search for stable structures is only one important analytical task undertaken by linguists, they will be willing to make room for, and work alongside of, performative linguists. I have no quarrel with these scholars. To the extent that some constative linguists have continued to follow Saussure in assuming that the study of the stable abstract structures of la langue

6

Performatives

is the only acceptable analytical task undertaken by linguists, they have been contemptuous and dismissive of the language scholars that I am calling performative linguists, have closed the disciplinary gates against these scholars, ruled their work beyond the disciplinary pale. I offer some fairly strong critiques of these latter thinkers, especially in Part I. But those critiques are aimed not at constative linguistics per se, rather at the universalizing tendencies behind certain hegemonic forms of constative linguistics that would elevate the study of stable structures into the study of all language. The constative notion of language as stable objective structure is a limited philosophical construct that can be quite useful in analyzing those aspects of language that do not vary greatly from context to context; constative linguists who embrace that construct as a useful fiction have a great deal to offer their more contextually, socially, actionally oriented performative colleagues. By the same token, the performative notion of language as contextually contingent act is also a limited philosophical construct that can be quite useful in analyzing those aspects of language that vary strikingly from context to context; constative linguists who dogmatically deny the usefulness of this latter fiction may, in what follows, be handled somewhat roughly. Let me also say here at the outset of my argument that I am not claiming a high degree of originality in what follows. Most of the radical pioneering work I present here has been done by others: Austin and Grice, Derrida and Bakhtin, many others whose names do not appear in these pages. What I am here calling ‘‘performative linguistics’’ is far from a‘‘new’’approach to language. It has been around and in some circles highly influential for at least half a century. Apart from my somatic theory of language, which as far as I know is original with me, my only original contribution in this book is the consolidation of existing ‘‘peripheral’’ or ‘‘nonmainstream’’ or ‘‘extra-linguistic’’ theories of language under the rubric ‘‘performative linguistics.’’ I came up with the idea of using constatives and performatives as descriptors of linguistic methodologies, and of expanding the field of linguistics to include both. Most of the rest of what you will read here is the work of far greater minds than my own.

The problem of translation in the study of language Constative linguistics, as I say, is the study of stable linguistic forms, the structures that we inherit and use without conscious awareness or expressive purpose. As long as linguistics is conceived as ‘‘basically’’about those structures ^ say, phonemes, morphemes, and syntactic structures ^ it will seem natural to restrict linguistic methodologies to what I am calling the constative. What I propose to do in this book is to start at the other end of the linguistic spectrum, with a speech act that is traditionally considered so complicated, so problematic, so rife with irresolvable methodological difficulties as to be virtually beyond the pale of linguistic study: the act of translation. It is, I will be arguing, at this extreme that the need for a broader paradigm for linguistics studies becomes most clearly evident. If translating is regarded as a language

Linguistics and translation studies

7

act, a use of language ^ and what else can it be considered? ^ then linguistics, the study of language, should be able to explain it. If we agree to include the act of translation within the purview of linguistics, then we are going to need a performative branch of linguistics. Let’s see how that works. The subtitle of this book,‘‘Speaking and Translating as Doing Things With Words,’’ is a somewhat tendentious one, because speaking and translating have been perceived by many scholars in both the linguistics and the translation studies camps as sufficiently different that any casual juxtaposition of the two is suspect. In the introduction to his volume in the St. Jerome ‘‘Translation Theories Explained’’ series, for example, Translation and Language, Peter Fawcett notes that the relationship between linguistics and translation studies has long been a‘‘troubled’’one: Since linguistics is the study of language and has produced such powerful and productive theories about how language works, and since translation is a language activity, it would seem only common sense to think that the first had something to say about the second. Indeed in 1965 the British scholar John Catford opened his book A Linguistic Theory of Translation with the words: ‘‘Clearly, then, any theory of translation must draw upon a theory of language ^ a general linguistic theory.’’ In exactly the same year, however, the famous American theoretical linguist Noam Chomsky was rather more skeptical about the implications of his own theory for translation, saying that his theory ‘‘does not, for example, imply that there must be some reasonable procedure for translating between languages’’ (1965: 30). Although no expert in translation, Chomsky nonetheless divined that there was something about the activity that put it beyond reason. Perhaps he had read what the academic Ivor Richards (1953: 250) said about translation: ‘‘We have here indeed what may very probably be the most complex type of event yet produced in the evolution of the cosmos.’’ (1997:1) Well, yes: translation is the most complex type of event ever produced, if you look at it through a rigid enough lens. If your model for studying language is derived from the study of phonemes, say, translation is going to look pretty frighteningly complex. All we need to do to get past Richards’ extreme view, however, is to develop a model that is more in tune with the complex dynamics of social action. The specific problem for the constative study of translation is that, in order to maintain a sense of language as more or less stable structure, constative linguists ^ here excepting Firthian (1957) context-of-situationists, Hallidayan (1978) systemic-functionalists, sociolinguists like Labov (1966, 1972a,b, 1994) and Gumperz (1971, 1982), and anthropological linguists like Sapir (1949, 1955), Whorf (1941/1956), Hymes (1962), Geertz (1966), Hanks (1999), and Silverstein (2001) ^ need to stay out of the volatile world of the communicative situation. The stable structures of constative or ‘‘formalist’’

8

Performatives

linguistics are the methodological byproducts of a conceptual lens called the ‘‘null context’’ ^ an imaginary context that has been analytically cleansed of all contextual variation, which is to say cleansed of everything that might complicate it in real-world ways. When constative formalists do venture out into speech situations, they typically reduce them to a tidy depersonalized (mechanized) null-context formula borrowed from early information theory: Sender ! Message ! Receptor The Sender, in this model, possesses a quantum of information, a message, which he or she wants to send to the Receptor. As we will see in Part II, especially in connection with the dialogical theories of Mikhail Bakhtin in Chapter 7, this ‘‘telementational’’model (to use Roy Harris’s term, for which see pp. 34^9) is based on a mind-as-machine paradigm that is utterly inadequate for describing the intersubjective complexities of human communication. But never mind that for now: that is a problem for performative linguists, and so is not part of what bothers constative linguists about translation. The problem arises for constative linguists when they attempt to insert translation into this informationtheory model. Where exactly do you slot the translator? Nowhere, really. The translator is a Receptor who becomes a new Sender; but making the translator a Sender implies that she (secondary and all-too-human in this depersonalizing model, therefore in hegemonic patriarchal purview implicitly female) somehow has her own Message to send, which can’t be, because then the Message sent would no longer be a translation but something else. This means that the translator must somehow come to be seen as a nonsending Sender who intervenes in the communicative situation without actually intervening in the communicative situation ^ a level of philosophical complexity with which constative linguists do not traditionally feel comfortable, so that this particular move usually has to be methodologically mystified. The translator should either be a Sender or a nonsender, not both. In order to minimize the problems this model raises, constative linguists have historically resorted to two basic stratagems: 1. Ignore translation altogether. By far the most popular solution. 2. Deal with translation but ignore the translator. Compare the source text and the target text in quest of equivalences and nonequivalences. Treat translation as a fairly mechanistic process (not one performed by a human being with her own experiences or thoughts or feelings, in a specific social and historical context shaped by a multitude of conflicting forces) that somehow manages to create target texts that convey the original message with varying degrees of conformity to the source text. This generates ideally dehumanized models for the ‘‘translation’’ process that look something like this: Sender ! Message ðSLÞ ! Receptor ðSLÞ & Message ðTLÞ ! Receptor ðTLÞ

Linguistics and translation studies

9

The translator’s rightful place in this model, obviously, is just below Sender, who even in the most stripped-down version of this model should be Sender (SL), followed by the translator as Sender (TL). The elision of the translator from the model makes two interrelated constative pretenses possible: that the communicative situation in translation is still about the original Sender sending a Message to a Receptor, and that translation is a purely instrumental or mechanistic process that does not involve the intervention of a human communicative (and potentially distortive) mediator. That things are considerably more complex than this even in a monolingual communicative situation should be clear. In this book I plan to run ‘‘speaking’’ and ‘‘translating’’ in parallel not in order to make a pitch for a more complex model of language use based on the complexity of translation, but rather to show how complex, in a performative purview, in a linguistics of the act, both monolingual and interlingual speaking and writing are. In Parts II and III, for example, we will see that the basic translation situation ^ a new Sender sending to a new Receptor communicative material originally sent by a different Sender to a different Receptor, and invariably changing that material a bit in resending it ^ is in fact quite common (indeed probably unavoidable) in monolingual speech situations as well. We will be examining many cases of this complexity in the course of the book. For now, look at just one, offered by Livingston (1997:19): Once I was talking with a friend on the phone. I told her about the letter I had just gotten from a Korean student looking for books unavailable in Korea: he had introduced himself as ‘‘a nonswimmer in English,’’ and closed with ‘‘thanks for hearing my awkward voice.’’ She told me she had just been swimming with some friends in a lake where schools of little fish would nibble at their body hairs, darting out of the way of any hand that moved into their midst. I told her how I had, many years before, visited a conservatory with a friend, and how we had watched a child squatting next to a fishpond, pointing to the bright carp swimming in the murky water and (because he wasn’t quite able to say the word ‘‘fish’’) delightedly exclaiming over and over, ‘‘FEEE . . . FEEE . . .’’ (a nebulous word for a nebulous shape somewhere between absence and presence, between ‘‘fort’’ and ‘‘da’’); and how years later I told my lover this story, and how years later still she remembered that she had been there with me and that we had seen the child together. There was a lot of intermittent background static in the phone connection. After a while it seemed to get louder and its rhythms became discernible as those of a phone conversation, although it remained only almost (but not quite) intelligible. We wondered if they could hear us. We tried to hail them with shouts, but their garbled conversation just continued. At one point it seemed to us that they may have paused to listen or wait for our disturbance to pass, but it may have been just a lapse in their conversation: an angel passed. It was very disconcerting to address

10

Performatives them: we had begun talking to each other, assuming, as usual, that we were two endpoints of the phone line between us. Then the line changed dimensions; in trying to talk to it, we changed with it. It estranged and intrigued us, rendered our own conversation ambiguous, as if the phone line had broken up into our body hairs nibbled by fish, or rendered us flesh and fish in turn, or the water in which alien species swim. We liked it, but we had to stop talking.

Livingston’s verbal memory of the boy and the fish is taken over by his lover, who remembers having been there herself, retells the story as if it had happened to her: is this not a quite ordinary monolingual parallel to the ‘‘cosmic complexity’’ of translation? Talking on the phone, Livingston and his friend become uncomfortably aware of another conversation that is both ‘‘on’’ and ‘‘not on’’ the same line; they try to enter it, but seem to fail; they wonder whether they have in fact entered it but just don’t notice it. Is this not a monolingual parallel to the source-language Sender^Receptor dialogue moving toward translation, seeking through the translator’s mediation to enter into a new dialogue in the target language? Is this all not fairly common in ordinary language use? Of course, bear in mind that when I talk about ‘‘ordinary language use,’’ I’m talking about the performance of speech acts and thus about the subject matter proper of performative linguistics. I don’t mean what constative linguists call ‘‘ordinary language.’’ Constative linguists occasionally attempt to move into the realm of language use (la parole or ‘‘performance’’), which they think of as a secondary expression or execution of ‘‘true language’’ (i.e. stable abstract form); because their models won’t let them deal with much complexity in the alien world of use, they have to keep it simple.‘‘Ordinary language’’ in a constative purview is any language use that lends itself to formalist reduction according to their models. The Livingston example is too complex for their models, so constative linguists must treat it as ‘‘special’’ or ‘‘extraordinary’’ language and exclude it from their considerations, as if it were not ‘‘really’’ language at all, or as if it were some secondary or parasitic form of language that can safely be ignored. What I am arguing here is that this methodological bias has led us to ignore far too much ‘‘ordinary’’ language. Far too many of the ways we use language are too complex for constative linguistics. Performative linguistics thrives on examples like the one Livingston offers, or similarly complex examples taken from joke-telling, story-telling, play-acting, pretending, and so on, including translating. Performative linguistics offers another methodological bias, of course, which makes it difficult to explain, say, why English-speakers would rather not put an adverb between a transitive verb and its direct object: for that we need constative linguistics. But if we want to study language in all its complexity, we need both biases. We need an expanded paradigm for linguistic study.

Linguistics and translation studies 11

The problem of linguistics in the study of translation Not only have linguists typically not been particularly interested in translation; with a few exceptions, notably Eugene A. Nida, translation scholars have traditionally (and since the 1960s increasingly) been chary of linguistics. As Fawcett notes (1997: 2), the French translation scholar Maurice Pergnier (1993 : 9) underscores the continuing impulse in the field to liberate translation studies from linguistics, and Marianne Lederer (1994 : 87) writes: ‘‘I hope in this way to bring out the reasons why translation must be dealt with on a level other than the linguistic’’ (translation Fawcett’s). Mona Baker (2000 : 21) suggests some reasons why this situation might have arisen: During the 1950s, 1960s, and 1970s, when Translation Studies was still struggling to establish a place for itself in the academy, linguistics had already come to be recognized as a fully fledged ‘‘scientific’’ discipline, and had developed an impressive range of research methods and tools of analysis. Given that language is the raw material of translation and that Translation Studies needed a role model to follow in order to establish itself in the academy, linguistics naturally became the main source of theoretical and pedagogical insights. Initially, partly because linguistics itself was less sophisticated in its approach to language at the time, linguistic studies of translation were often extremely simplistic in their approach to meaning and highly prescriptive in their pronouncements on translation. There were far fewer descriptive studies than there are now. The thrust of much linguistically oriented work was that clear guidelines had to be developed in order to ensure that translators had a set of ready-made, reliable solutions for linguistic difficulties. These difficulties were perceived as essentially formal in nature: lack of equivalence at word level, culturespecific items, difficult syntax, non-matching of grammatical categories such as gender, and so on. . . . Initially, then, much of what went on under the banner of linguistics in Translation Studies proceeded along the following lines: here is a source text in language A; how can it best be translated into language B given what we know about the characteristics of languages A and B? Baker is herself a (constative) linguist; her reasonably fair historical overview is subtly shaded in favor of traditional linguistic approaches, portraying the fatal flaws translation scholars have seen in constative linguistic methods as largely a thing of the past, based on unfortunate methodological oversimplifications early on that have now been overcome. One could easily rewrite this passage from an antilinguistic point of view, especially the one that has come to be called the ‘‘cultural-studies’’ approach (and that has come since the early 1990s to dominate the field), to highlight the survival in linguistics of those methodological failings that have caused dominant trends in the field to shift

12

Performatives

decisively away from linguistics: 1. ‘‘Linguistics had already come to be recognized as a fully fledged ‘scientific’ discipline, and had developed an impressive range of research methods and tools of analysis.’’ It is precisely this aspiration of becoming truly and fully ‘‘scientific’’ that has since the early 1990s most fatally consigned constative linguists’ ‘‘impressive range of research methods and tools of analysis’’ to the peripheries in translation studies. The key pitfall for would-be scientists of language is that, apart from acoustic phonology, language has no objective existence susceptible to the methods of inquiry accepted by quantitative science. The ideal ‘‘structures’’ that constative linguists study are in fact intersubjective constructs maintained and regulated by social groups and operating in and through individual speaker subjectivities. The methodological pretense that these exist in some stable objective form and state, and can therefore be studied as if they were physical objects, condemns the ‘‘science’’ of language to a most unscientific disciplinary backwater. 2. ‘‘Partly because linguistics itself was less sophisticated in its approach to language at the time, linguistic studies of translation were often extremely simplistic in their approach to meaning and highly prescriptive in their pronouncements on translation.’’ Constative linguistics still seems remarkably unsophisticated and simplistic in its approach to language when compared with the work of language philosophers like Ludwig Wittgenstein, Martin Heidegger, Mikhail Bakhtin, or Jacques Derrida. Indeed the work of language philosophers that constative linguists respect and cite and adapt for their own work, like John Searle, seems remarkably unsophisticated and simplistic as well. And the prescriptivism of early linguistic studies of translation has mainly just gone underground, into ostensibly ‘‘descriptive’’ studies based on the assumption that translation is and must always be based on linguistic equivalence, and anything that is called translation but isn’t according to prevailing linguistic models demonstrably equivalent simply isn’t translation. This is still prescriptive; just more subtly so.1 3. ‘‘The thrust of much linguistically oriented work was that clear guidelines had to be developed in order to ensure that translators had a set of ready-made, reliable solutions for linguistic difficulties. These difficulties were perceived as essentially formal in nature: lack of equivalence at word level, culture-specific items, difficult syntax, nonmatching of grammatical categories such as gender, and so on.’’ From within a constative linguistic viewpoint, certainly, it seems as if great progress has been made from these early naive approaches to translation. Constative linguistic translation studies have become more descriptive, less focused on the best way to translate, less exclusively directed to the solution of practical problems that arise in translating. Earlier formalisms have been expanded and complicated through a new attention to ideological markers and other ‘‘sociolinguistic’’ factors; as Baker (2000: 22) notes, linguistically oriented scholars of translation have increasingly been incorporating cultural-studies concerns

Linguistics and translation studies 13 into their models: During the 1980s and 1990s, a new generation of translation scholars with training in linguistics began to carry out more sophisticated descriptive studies. These focused more on exploring what actually happens in translation than on exploring what should happen or what can potentially happen. The starting-point for much of this work was the idea that meaning is diffuse, in the sense that it is not located in the word or grammatical category but is signalled by a variety of means which cross the traditional boundaries of word, phrase, clause, sentence, and even text. It is also unstable, in that it is only realized in text, and a ‘‘dictionary’’ meaning may even be negated within a specific textual environment. Moreover, meaning is now understood to be culturally constructed and all language use is seen as mediated (culturally, ideologically, cognitively). Language, both generally and in translation, has come to be seen as intimately linked to the social and cultural context in which it is produced. Baker does not mention the cultural-studies origins of these ideas; they are simply part of the maturation process of linguistics. She does not mention the field’s shifting disciplinary hegemonies, the fact that linguistics was the dominant explanatory framework in translation studies until the early 1990s, when it began to be eclipsed and peripheralized by cultural studies, and linguistically oriented translation scholars increasingly felt the need to draw on the methodological assumptions of their disciplinary opponents in order to retain a significant voice in disciplinary debates; as Baker describes it here, the change just sort of happened. Nor does she mention the fact that, outside of cognitive linguistics (see e.g. Lee 1992), linguistic studies of ideology remain even today highly formalistic, focusing for example on formal markers of ideology: what has happened, as I suggested above, is that the methodological scope of formalism has been expanded from the word, sentence, and text levels to the social and political, while nevertheless retaining the scientistic bias in favor of highly idealized and decontextualized language, language conceived in terms of larger and more sociopolitical co-texts but methodologically still nullified contexts. Eugene Nida is widely considered to be the founder of the field of translation studies, the first translation scholar to systematize what had been for a very long time a very ad hoc affair. And Nida’s systematizations proceeded specifically along linguistic lines, first in the 1940s and 50s, then, with signal impetus from theTG grammar of Noam Chomsky (1957,1965), particularly in the 1960s, in two big books, Toward a Science of Translating (1964) and, with Charles R. Taber, TheTheory and Practice of Translation (1969). Though critics of Nida (see Gentzler 1993: 44^8, Meschonnic 1973) argue persuasively that he oversimplified Chomsky and somewhat rashly ignored Chomsky’s own strictures against the project of deriving a general theory of translation from his work, clearly Chomsky’s notion of surface and deep structure provided him with a productive

14

Performatives

linguistic framework on which to hang a dualistic model of the translation process: analyze the source-text surface structure down into its deep structure, make the transfer to the target language at the level of the deep structure, then restructure the message in terms of the target-language surface structure. When J. C. Catford’s (1965) Firthian approach to translation came out, it too seemed to many to be the clearest exposition of the problems of translation that anyone would ever need.2 Voices began to be raised against the monolithic linguistic orientation of translation studies in the early 1970s, by people like Itamar Even-Zohar and Gideon Toury in Israel and, around the same time, James Holmes and Andre¤ Lefevere and Theo Hermans and others in the Low Countries. By the late 1970s (especially beginning with a watershed conference held in 1976 in Leuven, Belgium) these scholars had formed a school of sorts, a‘‘camp,’’an approach, variously called polysystem theory or just plain ‘‘translation studies’’; eventually it would shape-shift into the term James Holmes invented for it, descriptive translation studies (DTS). Susan Bassnett’s 1980 book Translation Studies was an early popularization of this approach. Toury too published a 1980 book that would be enormously influential, In Search of a Theory of Translation (and updated that book under the title DescriptiveTranslation Studies and beyond in 1995; for introductions, see Gentzler 1993: Chapters 4^5 and Hermans 1999). More antilinguistic approaches were born in the 1980s, most notably the skopos/functionalist/action-oriented approaches of Hans Vermeer, Justa Holz-Ma«ntta«ri, Christiane Nord, and others, especially in the signal year 1984, which saw the publication of Katharina Reiss and Hans Vermeer’s Grundlegung einer allgemeine Translationstheorie (‘‘Foundations of a General Theory of Translation’’) and Justa Holz-Ma«ntta«ri’s Translatorisches Handeln (‘‘Translatorial Action’’). These scholars ignored linguistic issues entirely, dismissed the question of equivalence impatiently, focused instead on the translator as a professional intercultural communicator, hired by a client to produce a certain sort of text whose parameters are determined by the client, in the service of the target audience. (For a useful introduction, see Nord 1997.) Think-aloud protocol (TAP) research on translation also begins in this period: in an attempt to get at the ‘‘black box’’ inside the translator’s head, TAP researchers got translators working alone and in groups to verbalize their thoughts as they translated, taped the verbalizations, and analyzed them in search of cognitive strategies. Like the skopos people, TAP scholars like Krings (1986,1988), Lo«rscher (1991), Kussmaul (1991,1995), and Tirkkonen-Condit (1991) also ignored questions of linguistic equivalence, drawing to some degree on psycholinguistics but far more heavily on cognitive psychology. Their concern was almost exclusively with what the translator does ^ or, to use Krings’early book title, what happens in the translator’s head. In the late 1980s and early 1990s two more antilinguistic schools of translation studies sprang up, the postcolonialists and the feminists. Postcolonial translation studies was born out of cultural anthropology: as anthropologists were increasingly influenced by cultural studies paradigms they began to realize that

Linguistics and translation studies 15 they were basically translators and their translation assumptions were basically colonial. They went abroad to live with a ‘‘primitive’’ people, learned their language,‘‘mastered’’ their culture, then returned to ‘‘civilization’’ to translate what they had learned into their European language. Increased awareness of the implicit ideological hierarchies in this field-work ethos began to transform cultural anthropology by the mid-1980s. Translation scholars building on this foundation have stressed the political and ideological roles translation has played in the spread of empire: the ways in which translators as colonial agents have effectively silenced the colonized they are supposedly speaking for (Asad 1987, Cheyfitz 1991, Niranjana 1992), the complexity of the interaction between colonizer and colonized in the translation situation (Rafael 1993), and the significance for translation of power differentials between source culture and target culture in the postcolonial context ( Jacquemond 1992,Venuti 1995,1998). Feminist translation scholars have been engaged in the interlinked tasks of recovering the lost or neglected history of women as translators and translation theorists (Krontiris 1992, Robinson 1995, Simon 1996), articulating the patriarchal ideologies undergirding mainstream Western translation theory (Chamberlain 1988/1992), and formulating a coherent and effective feminist practice of translation (Maier 1980, 1984, 1989, Lotbinie're-Harwood 1991, Godard 1989, Levine 1992, D|¤ az-Diocaretz 1985). All of these scholars either pointedly ignore the formalistic methodologies of linguistic approaches or launch principled attacks on them as complicit in colonial and patriarchal ideologies (see especially Venuti 1998, and, for discussion, Chapter 9). (For introductions, see Robinson1997b on postcolonialism and von Flotow1997 on feminism.) But still the linguists stood strong. In fact, the main translation theorists being taught to students worldwide as late as 1995, based on my own discussions with teachers and students in various countries that I visited that year, were Nida, Catford, and Peter Newmark ^ another linguist. By the time of the current writing, the winter of 2001^02, this has changed in most places, and students in theory courses around the world have read Toury and Lefevere and Hermans,Vermeer and Nord,Venuti and Pym. Anthony Pym is, in fact, one of the very few translation scholars to combine exciting inquiries into translation as culture, translation as profession, and translation as economy with specifically (though transformatively) linguistic inquiries into translation as language. An Australian scholar trained in sociology at the Ecole des Hautes Etudes en Sciences Sociales, Pym is currently professor of English linguistics in the department of English and German at the Universitat de Rovira i Virgili inTarragona, Spain; and despite his reputation as basically a cultural-studies theorist of translation, he continues to integrate linguistic approaches into his wide-ranging theoretical repertoire. Nevertheless, he too notes the inherent clash between linguistic approaches as they are traditionally conceived and the specific discursive needs of translation (Pym1993: 47): Although most linguists approach translation as a comparison of two texts, few translators see their work that way and few translations are

16

Performatives actually received that way. To be sure, any translator must interpret the linguistic structures of the source text, but that is not what they are paid for. Their real linguistic work is the production of a specifically translational text for a new reader and a new situation. Similarly, the reception of a translation certainly involves the interpretation of linguistic structures, but it also concerns the reception of a text that is somehow specifically translational. Linguists who merely compare texts rarely appreciate these two aspects of properly translational phenomena. They can see the input and the output, but not the projection of an active link between the two. Linguistics is thus mostly a study of textual results but not of translational phenomena.

Discourse analysis, Pym notes, is the one branch of linguistics that supposedly addresses itself to issues of production and reception that might be useful in a translation-studies purview; but unfortunately the few linguists who have attempted to apply discourse analysis to the study of translation have hobbled themselves methodologically by tying all discursive studies of translation to equivalence: But when we look at the supposedly translational research carried out in this field ^ most prominently in Hatim and Mason’s Discourse and the Translator ^, we find that the discourses concerned are exclusively those of source and target texts that are supposed to conform to some kind of necessary matching. They are not those of the translator, nor of translational reception. This traditional focus has effectively been liquidated by mostly non-linguistic descriptive approaches that are now unable or unwilling to define translation in terms of any necessary correspondence between two texts. (See also Pym 1991.) I will be taking a close look at Pym’s use of linguistics in the study of translation, specifically Austin’s distinction between performatives and constatives, in Chapter 3.

Constative and performative linguists of translation If, then, we think of constative linguists of translation as those who see translation in terms of ‘‘saying,’’ restating in the target language more or less precisely what the source author said in the source language, and performative linguists of translation as those who see translating as ‘‘doing,’’ doing something to the target reader, then the contemporary scene comes to look rather different. Then, obviously, the politically engaged cultural theorists of translation ^ the postcolonialists and the feminists ^ become performative linguists: translating as colonizing, or as fighting the lingering effects of colonialism (Cheyfitz, Rafael, Niranjana); translating as resisting global capitalism (Venuti); translating as fighting patriarchy, as liberating women (and men) from

Linguistics and translation studies 17 patriarchal gender roles (Maier, Godard, Levine, Lotbinie're-Harwood, D|¤ az-Diocaretz). But then, this model can also help us see a repressed performative linguistics in Eugene Nida as well, whose ideal translator is at least implicitly engaged in the activity of converting people to Christianity. The theological orthodoxy in translation theory from Augustine to Nida, of course, would deny this, insisting that they are not working to convert readers so much as simply providing those readers with unfettered access to the Bible, which they can do with as they like. But Nida’s most famous theory of translation, that of ‘‘dynamic equivalence,’’ is explicitly based on reader response, on the attempt to have an impact on a target reader that is functionally equivalent to the impact the original text had on the source reader. Attempts to have impacts on readers are, it seems to me, very much in the realm of performative linguistics (when, of course, they are admitted). Also to be included in the camp of performative linguistics, it seems clear to me, are the other influential groups that I’ve mentioned above: DTS, skopos theory, and TAP theory. They are all concerned with the translator as a doer, an actor on variously conceived cultural (DTS), professional (skopos), and cognitive (TAP) stages. For theorists in these schools it doesn’t matter what the translated text looks like; or, well, it does, but not to the theorist, only to the receiving culture (DTS), the client (skopos), or the translator himself or herself (TAP). What matters theoretically is what the translator does, and what complex forces influence that doing. Now perhaps this seems like no great gain: taking all the approaches of the so-called ‘‘cultural turn,’’all the action-oriented theories that have overwhelmed and overthrown the hegemony of linguistics, and lumping them together under a new name ^ even if that name is rather tendentiously ‘‘linguistics.’’ What, you may ask, has changed here? Before we had linguistics vs. cultural studies, now we have constative linguistics vs. performative linguistics, but all that has changed are the names. But the changes are in fact more substantial than that. For one thing, TAP theory is usually not included in cultural studies: it draws on psycholinguistics, and therefore, for the cultural-studies people, is part of the despised enemy. Skopos theory is relatively new (15 or 20 years old), and antilinguistic, therefore more or less tolerated by cultural studies, but it has never really been part of the cultural turn either. In his Translation Studies Reader, in fact, Lawrence Venuti (2000: 217) disparages Holz-Ma«ntta«ri’s approach as a reduction of ‘‘translation to an assembly-line process of text production, a Fordism that values mere efficiency.’’ Thinking about approaches to translation in terms of saying versus doing, too, restating versus acting, helps us to make the connection between translation-as-conversion in the missionary subtext of Nida’s (ostensibly constative) work and the postcolonial studies of translation as colonization, especially perhaps Rafael’s on the use of translation to convert the Tagalogs to Christianity. It also helps us to make the connection without, say, condemning Nida for imagining the translator as a part of the missionary

18

Performatives

project: Nida covertly and Rafael overtly are both portraying the translator as a doer, as someone acting on the target audience, not simply holding a mirror or a window up to the source text. More important, drawing on speech-act theory to discuss theories of language and translation can help us flesh out the performative project of these recent radical theories more fully. What would it mean, to stick with the Austinian vocabulary that I’ll be smuggling into Gricean implicature in Chapter 10, to analyze translations in terms not only of locutions but of illocutionary force and perlocutionary effect as well? What is the illocutionary force of a Bible translation, a literary translation, an advertising translation? What kinds of perlocutionary effects do translations have? What speech acts are various specific translations intended to perform, and what speech acts do they come to perform throughout the history of their reception? What happens to our conceptions of translation when we imagine it not as stable equivalence that is never quite stable and never quite equivalent, but (my focus in Part II) as what Jacques Derrida calls ‘‘iteration,’’ a repetition of the same that always alters the ‘‘same,’’ translation as reperformed language? And, anticipating the Gricean theory that I will be exploring in Part III, what sorts of things do translators implicate? Translators are normatively expected to be invisible handmaidens of the original text; if they want to act on the reader, if they want to do something through their translations, most likely they are going to have to rely heavily on what Grice calls conversational implicature, saying things without spelling them out. How exactly do translators do that? I’m going to be arguing in Chapter 9, for example, that Gricean implicature offers a powerful model for explaining what Lawrence Venuti’s ideal translator does: flouts target-cultural maxims of fluency cooperatively, signaling surreptitiously that his or her intention is not simply to confuse or irritate the reader but to have some sort of salutary ethical or political impact.

A map of the argument The book is divided into three parts. Part I centers around the William James lectures delivered at Harvard in 1955 by the Oxford language philosopher J. L. Austin, published posthumously as How To Do Things With Words (1962). Specifically, as I say, I take his distinction between constative and performative utterances as an organizing principle for thinking about linguistic methodologies. My main task here is to distinguish between constative linguistics and performative linguistics, by way of charting out a signal methodological expansion of the linguistic discipline. In Chapter 2 I will be ringing some changes on J. L. Austin’s original distinction, looking at the attempts made by Emile Benveniste and Jerrold Katz to recuperate the constative^performative distinction for a constative linguistics, Shoshana Felman’s reading of Austin as a prankster who loved scandal, the work Ludwig Wittgenstein was doing on language games at Cambridge at the same time Austin was developing the performative at Oxford, a brief history of the performative and the constative

Linguistics and translation studies 19 (and the victory of the constative), and a quick survey of what happens to a critique of formalist linguistics when one omits the performative (Roy Harris’ integrational linguistics). Chapter 3 is the book’s first major analytical discussion. In it I engage Anthony Pym in a discussion of the key issue, perhaps the definitive issue for the entire book, of whether a translation might ever be thought of as a performative utterance. In conservative translation theory, as I suggested above in connection with the Sender ! Message ! Receptor model, a translation is by definition a vehicle for someone else’s utterance, not an utterance in its own right; there is no way for it to take on the characteristics of a performative. The conservative assumptions behind this approach are, I would argue, specifically constative, and undermining them along performative lines is one of my main tasks in this book. I will be recurring to Pym’s example and analytical problem twice again, briefly, later in the book (in Chapters 4 and 13). Part II offers one fairly complex line of thinking in response to the salutary problem raised by Austin of parasitic speech acts. Austin says that language doesn’t just convey information; it also performs actions, speech acts. A minister stands before the altar and says to the bride and groom,‘‘I now pronounce you man and wife,’’ and by that very saying, they are made man and wife.3 The utterance performs the action. But then Austin raises the question: what happens when an actor on stage dresses up as a minister and pronounces those same words before a stage altar to two other actors dressed up as bride and groom? Clearly, in that context the utterance does not perform the action. The performative utterance ‘‘I now pronounce you man and wife’’ does not marry the two actors. Why not? He couldn’t answer that question, so he called those speech acts ‘‘nonserious’’ or ‘‘parasitic’’ and set them aside. In a 1971 article on Austin and related topics entitled ‘‘Signature Event Context,’’ Jacques Derrida argued that (to put it simply) an utterance must be performable, and thus distortable, in order to be spoken at all. This performability he called ‘‘iterability,’’ the capacity to be reused, which also invariably involves the capacity to be misused, misperformed, changed or twisted in some new way. I outline Derrida’s theory in Chapter 4, then, by way of offering an explanation of how iterability might be channeled through actual human bodies, take a new look at the somatic theory of language that I developed in The Translator’s Turn (Robinson 1991). The somatic theory, which I present here in terms of the neurological studies of Antonio Damasio in Chapter 5, and Daniel Simeoni on the translator’s habitus in Chapter 6, says that language is both stored and ideologically and idiosyncratically inflected through what Damasio calls ‘‘somatic markers,’’ body responses that the autonomic nervous system learns from our experience to send us in order to guide our rational thought processes. Simeoni’s work on the translator’s habitus seems to me an enormously valuable contribution to somatic theory, because it so complexly sociologizes the somatic: it was all too easy for readers of TheTranslator’s Turn to reduce my argument to the mindless contention that translators should translate any way they damn well feel. Simeoni will help, I hope, to forestall such reductions of my argument

20

Performatives

here. In Chapter 7 I jump back in time to 1929, to Mikhail Bakhtin’s book on Dostoevsky’s poetics, where he offers his theory of ‘‘double-voicing’’: the notion that every word ever spoken is saturated with all the dialogues it has ever been used in, or rather, to put that more realistically, that each language user brings to every use of a word the memory of all the past dialogues s/he ever heard it in. This seems to me a wonderful analytical formulation of the abstract philosophical conceptJacques Derrida calls iterability ^ even though, anachronistically as I present it, the analytical formulation precedes the concept it fleshes out by just over four decades (but then Bakhtin’s work did not become well-known in theWest until a few years after Derrida wrote on Austin). I subtitle Part II ‘‘The performative back-story’’ because I am specifically interested there in theories of how we become capable of performing speech acts, what brings us to the performative brink: the theories I explore there flesh out the ‘‘back-story’’ of language learning, of a speech community’s regulatory training in performative competence. Derrida says that language becomes usable, speakable, writeable, performable, through its iterability, its capacity for being repeated in new contexts, which tends to transform it. My somatic theory says that our bodies channel these repetitions through‘‘somatic markers,’’ which help us ‘‘remember’’ (though unconsciously) what we have learned from experience in ways that facilitate rapid and complex action-oriented decision-making. And Bakhtin’s theory of double-voicing allows us to trace the history of polyvocal repetition analytically, to show how a voice is transformed in and through a series of iterations. Part III, then, brings us to the Oxford and Berkeley language philosopher H. Paul Grice, specifically to the lecture in the 1967 William James series at Harvard, published in 1975 as ‘‘Logic and Conversation,’’ that first proposed the concept of conversational implicature. Part III in general follows Grice in taking on the problem of how we can understand speakers (the part is subtitled ‘‘Performative uptake’’) who do not say what they mean, who seem to be saying things that they can’t possibly, or perhaps just shouldn’t, be saying. How does this work? The actor on the stage; the speaker in metaphors; the speaker or writer who deliberately avoids saying what s/he wants to say: how does this person make himself or herself understood? This is essentially a different way of framing Austin’s problem of parasitic speech acts: Austin dealt with whether such speech acts exist, whether they should really be considered speech acts at all (Part I); Derrida, Robinson, and Bakhtin dealt with the role such ‘‘parasitic’’ speech acts play in language history, language change, linguistic evolution (Part II); Grice deals with how they function to make understanding possible (Part III). Specifically, Grice suggests that the utterer and interpreter in such a dialogue manipulate shared assumptions or ‘‘maxims’’ about what it would entail to be speaking cooperatively, and so guide intention and/or interpretation so as to make understanding possible. In this theoretical framework, in fact, the ‘‘problem’’ of the actor playing a minister and marrying two people on stage that Austin stumbled over becomes so unproblematic as hardly to be worth Grice’s attention: there are conventions governing the

Linguistics and translation studies

21

performing of speech acts in a play, and thus also, and more importantly, governing the interpretation of such speech acts by an audience. Nobody takes a stage marriage to be a real one, or the stage ‘‘I now pronounce you man and wife’’ to be a ‘‘serious’’ performative utterance: it’s just a line in a play. But in the theory of implicature, this is a fairly ordinary state of being, something that happens all the time off-stage as well as on: people are always saying things that others don’t take at face value, because there are fairly widely accepted conventions or at least assumptions about how such things are to be interpreted. Derrida was right (but Grice said it first): this ‘‘deviation’’ from the ‘‘usual’’ order of things is not deviant, because the ‘‘usual’’ order isn’t usual at all, it’s just an analytical fiction ^ what I would identify specifically as a constative analytical fiction, the assumption that the rule (and the straightforwardness it entails) is primary and ordinary and any deviation from it is ‘‘parasitic.’’ Grice, like Derrida after him, says that such supposedly ‘‘parasitic’’ utterances, or conversational implicatures, are perfectly ordinary, the stuff of everyday conversation. Part III offers a series of analytical expansions of this idea. Chapter 8 presents the theory itself in a light that will help me in later chapters to shift Grice’s explanatory ground slightly but significantly for a performative approach. Chapter 9 shows how Lawrence Venuti’s attack on Gricean implicature as a theory of translation not only misses the point in Grice but fails to see how Grice’s theory can explain what Venuti has been trying to say better than Venuti himself. Chapter 10 slots Grice into Austin’s locutionary/illocutionary/ perlocutionary framework, exploring illocutionary and perlocutionary implicatures; then goes beyond illocutionary and perlocutionary implicature to explore the possibility of a metalocutionary implicature, an implicative intention or interpretation that turns the speaker’s and/or interpreter’s attention back onto the larger ideological and psychosocial implications of what is going on in the current language-use situation. Chapter 11 slots Grice into an expanded version of Charles Sanders Peirce’s interpretant triad, in order to facilitate the differentiation of conversational from conventional implicature in Chapter 12 and maxim-flouting implicature from ‘‘invocative’’ implicature in Chapter 13. Chapter 14 ties up the remaining loose ends of my Gricean ruminations by taking a close look at the problem of cross-cultural misunderstanding. How do we explain, first, and then learn to deal with, the fact that native speakers of other languages bring such divergent assumptions (‘‘maxims’’) to bear on conversational behavior? The book is big, too big for coziness, too big for a quick read; but it is also nowhere near big enough. The task I have set myself is far too big for one book, or even possibly for one person. And indeed as I say I do not think of myself as having invented this task ^ expanding linguistics to include performative approaches ^ but rather to have picked it up where others have left off, to have taken the ongoing dialogue on alternative linguistic methodologies a few steps further. Still, I do attempt to push and prod all these divergent language theories into a single overarching and coherent shape, which I’m the first to be

22

Performatives

theorizing as performative linguistics; and that is probably too ambitious a goal. With the possible exception of Paul Grice, the magnitude of my design makes it impossible for me to deal with anyone’s argument in adequate detail. I have to argue at such a high level of generality that many salient elements of each theorist’s contribution drop away. This is even true of Austin’s contribution; even more true of Derrida’s, Bakhtin’s, and Peirce’s; and the many painstaking attempts dozens of other innovative thinkers have made to redirect the linguistic current, the Firthians, the Hallidayans, the Chomskyites, the sociolinguists, the anthropological linguists, the critical discourse analysts, the cognitive linguists, the communication scholars ^ all these hardly even register as a blip on my screen. This certainly does not mean that I believe Austin, Grice, Derrida, Robinson, Bakhtin, and Peirce to be the only ‘‘performative’’ linguists in the world, or even that performative linguistics is the only viable alternative to ‘‘mainstream’’ (constative or segregational) linguistics. This book is only one consolidation of alternative linguistic methodologies; there are many other ways of setting up ‘‘the next thing’’ after the downfall of constative hegemony, and no doubt better ways of setting up the performative ‘‘next thing’’ as well. I only hope that you find this way instructive enough that you too will enter the fray, enter the lists, take a stand.

2

Constative and performative linguistics

The discovery of the performative In the William James lectures that J. L. Austin delivered at Harvard University in 1955, which were later (posthumously) published as How To Do Things With Words (1962), he distinguished first between two kinds of utterances: ‘‘constatives,’’ for conveying information, and ‘‘performatives,’’ for performing actions. Students of language had long thought of language as primarily a communicative medium, a channel by which human beings (and perhaps other animals as well) tell each other things,‘‘state’’ things, pass on information that they consider important. But, Austin notes, think of such utterances as ‘‘I now pronounce you man and wife,’’ ‘‘I bet you five pounds,’’ ‘‘I christen this ship the Mr Stalin,’’ ‘‘I promise I’ll be there on time’’: these utterances don’t state things, they do things. They perform the actions that they seem superficially to be merely describing. The minister marrying a couple does not perform the action of actually joining them in wedlock by some nonverbal ritual and then merely reflect or record that ritual with words; it is the words themselves that perform the action. The action is verbal. This is, clearly, an exciting discovery. It seems so obvious to us, once Austin has pointed it out, that we shake our heads at our own blindness, our failure to notice it before. We cast our eyes back over the history of language philosophy up to Austin, read in it carefully, marvel that no one noticed it earlier, no one said what seems so overwhelmingly obvious now: scholars writing about language before Austin, and even a good number of language scholars after Austin, refer repeatedly to language as a communicative medium, a channel for saying things, full stop. But then, halfway through his book, Austin backs off from his own wonderful discovery. In lectures seven and eight he decides that the distinction doesn’t really work: performatives convey information too, he says; constatives perform actions. Yes, in saying ‘‘I now pronounce you man and wife’’ the minister is actually marrying the couple; but the minister is also conveying to the gathered congregation that the couple is now married. And saying ‘‘My plane arrives at 8:05’’ conveys information, but it also does something, for example gets your interlocutor to the airport to pick you up on time. It is impossible, Austin finally

24

Performatives

concludes, to think of an utterance that does not perform some action; and obviously, if language philosophers had thought of language exclusively in terms of the conveying of information for all these centuries, these ubiquitous performatives must be ‘‘constating’’ things as well. And so he spends his remaining lectures developing a new model for the study of language: assuming that every utterance does something, performs an action ^ is, in his key term, a‘‘speech act’’ ^ he proposes to analyze utterances on three levels, the locutionary (what is said), the illocutionary (what is being done in the saying of it), and the perlocutionary (the effect the speaker has on the listener by or through the saying of it). Thus ‘‘My plane arrives at 8:05’’ is a simple locutionary statement of fact with (for example) the illocutionary force of a plea (‘‘Please come get me!’’) and the perlocutionary effect of getting the listener to the airport on time (and, more mediately, getting that person to arrange his or her day around that trip to the airport, to spend time planning the drive, parking, the pick-up, the return home, possible dinner and/or sleeping arrangements, etc.). Now clearly, here, Austin has not really abandoned the performative; what he has done, in fact, is given it pride of place, made it his ‘‘god-term,’’ as Kenneth Burke would say, defined all language use in terms of it. If every utterance is a speech act, then every utterance is, according to his first distinction, a performative. The first distinction lives on in the wings of the second: the constative becomes the ‘‘locution,’’ the mere superficial external form of an utterance that seems to traditional language philosophers to be intended to convey information; but beneath that surface every utterance performs actions, has both an illocutionary force and a perlocutionary effect.

The constative recuperation of the performative It is, of course, a pity to lose the performative/constative distinction. There does seem to be a significant difference between an utterance like ‘‘My plane arrives at 8:05,’’ which, no matter what kind of extralinguistic effect it has on the person on the other end of the telephone line, does clearly convey information, and one like ‘‘I now pronounce you man and wife,’’ which might be thought of as conveying information after a fashion but really basically performs an action. They really are different types of utterances. And it is helpful to have categories for such differences, names for them. It helps us organize our world mentally, to divide it up into clearly demarcated compartments. And as long as we take that desire to make sense of things by demarcating them ^ drawing boundaries between them, clarifying them categorically by ignoring the overlaps between them ^ as a fundamentally and thus universally human desire, then Austin’s willingness to scrap the distinction will seem to us inherently perverse. And we will do whatever we can, especially if we are linguists or analytical language philosophers, to recuperate it. And in fact this is precisely what has happened. In the two decades following Austin’s lecture series at Harvard, a number of highly influential and respected

Constative and performative linguistics

25

linguists and language philosophers argued persuasively and successfully for the reinstatement of the performative/constative distinction to describe these idealized types of utterances. Here, for example, are Emile Benveniste from Problems in General Linguistics in 1966 and Jerrold Katz from Propositional Structure and Illocutionary Force in 1977: We see no reason for abandoning the distinction between the performative and the constative. We believe it justified and necessary, provided that one maintain it within the strict conditions of use that sanction it, without letting the consideration of the ‘‘result obtained’’ intervene since this is the source of confusion. (Benveniste 1966: 238) Austin mistakenly thought that cases like (5.97)^(5.102) were counterexamples simply because he restricted himself to talking about utterances and actions, to performance. . . . The competence-performance distinction provides us with the benefits of idealization. Like the use of idealization in the theory of gases and mechanics, we do not have to state laws in terms of real objects and events ^ in our case, utterances and the speech acts they perform. Instead, we can state them in terms of their idealized objects and events such as sentence types, their meaning in the language, that is, their meaning in the null context, and context types. . . . Thus, classifying sentences into performatives and constatives can be carried out on the basis of their senses in the language. (Katz 1977:184^5) Note, however, how Benveniste and Katz go about reinstating the distinction: by reducing ‘‘noise,’’ reducing ‘‘confusion,’’ eliminating from the equation all of the real-world variables that originally convinced Austin that you really could not make the distinction stick. Fundamentally, Benveniste and Katz are arguing that Austin properly should have ignored those variables himself. That is, after all, what analytical thinkers do. They idealize, they remove themselves from the variability of the real world to the relative clarity of philosophical ideality, where things make more sense, where simple dualistic distinctions work. The real world is a place of confusion, of conceptual impurity; in order to philosophize, which is to say, to analyze, to think analytically, the thinker must look through impurities to the true underlying pure and ideal essences they conceal, must as Katz says idealize the real, and deal with the clarities of the ideal, not the confusions of the real. The impurities that Austin succumbed to, what we might call the invasion of the ideal by the real, for Benveniste are ‘‘results’’ and for Katz ‘‘performance’’; for both writers, clearly, these must be banished from the idealizing equation before it will work (again). Results and performance are both, in this conception, unidealized events as secondary effects: a result is what follows from or leads out of an antecedent or an

26

Performatives

originating event, and performance in Chomsky’s (1965) terms is the ‘‘performing’’ or executing or realizing of a perfectly stable ^ which is to say ideal, and therefore logically or chronologically (actually ideologically) ‘‘prior’’ ^ competence. Thus for Benveniste constatives and performatives are only philosophically confusing when they are allowed to result in something, when they are placed in the context of their real-world circumstances and viewed in terms of their consequences; for Katz they are confusing when they are ‘‘performed,’’ when (to exfoliate Chomsky’s dramatic metaphor) the ideal script is actually played on stage, where it becomes subject to all the disturbing variety of human error and deviation. Katz is more insistent about idealization than Benveniste: ‘‘the benefits of idealization,’’ he says, and ‘‘the use of idealization in the theory of gases and mechanics,’’ and ‘‘idealized objects and events.’’ In case we are not familiar with the mechanism of idealization in linguistics, he also provides a short list of the relevant operations, two in particular, typification and decontextualization: the abstraction from actual utterances of ‘‘sentence types’’and from actual contexts of ‘‘context types’’ (and the idealized god-term for that series, the ‘‘null context’’), and the thematization of these ideal types as ‘‘meaning’’ and ‘‘the language.’’ This does, obviously, have the effect of recuperating the performative for the philosophy of language, and for constative linguistics. But at what a cost! What kind of ‘‘performative’’ is it, after all, that has been purified of ‘‘performance’’? Austin speaks specifically of the performing of speech acts; Katz would now disallow this performing. Do we still have speech acts, then, or not? Is an unperformed act still an act? Is speech that is not spoken, even inwardly, still speech? What would a ‘‘speech act’’ be like if removed from the world of speech and acts?

The performative as crime and scandal It may not be too much to say that from a constative point of view, J. L. Austin’s crime in HowTo DoThings With Words lay not merely in dismissing the performative, but in announcing it without proper safeguards, without sufficient idealization. First he talks about performance, about acts, and names one kind of speech act after performance; then he allows performance and acts to displace the term itself, the idealization. This is a philosophical misdemeanor of a rather serious sort. But Austin seems born to a life of crime. For example, note how, after announcing the performative, he quickly finds himself swamped in the unpredictable variety and illogicality of human behavior ^ that very mess that constative linguistics was an attempt to tidy up. What happens, he asks, if I walk breezily up to an unattended ship on the stocks, smash the bottle on the stem, and say ‘‘I name this ship the Mr Stalin’’ (1962: 23)? Suppose even that it was supposed to be given that name, but I wasn’t the person supposed to christen it ^ was it named the Mr Stalin, or wasn’t it? If the minister gives a baby the wrong name ^ misremembers or mispronounces it, say ^ was

Constative and performative linguistics

27

the baby named that name or not? Can a minister baptize a penguin, or marry two monkeys? Austin at Harvard three years into Eisenhower’s first term is like a particularly zany stand-up comic, the Oxford don as cosmic buffoon, imagining a hundred ludicrous situations that conjure up the ludicrous variety of real life, infinitely more bizarre than the craziest novel or movie. In The Literary Speech Act: Don Juan with J. L. Austin, or Seduction in Two Languages, Shoshana Felman explores at some length this ‘‘festival’’ mood in Austin, this feeling that, as she says,‘‘Austin is constantly having a good time’’ (1983: 102). This is not, Felman argues, merely a personality trait, merely an unfortunate (perhaps criminal) pathology in an otherwise serious thinker; it is intrinsic to Austin’s philosophical method and essential to its success, which is consistently thematized in terms of failure: The act of failing thus leads, paradoxically, to an excess of utterance: manifest through its pleasure, independent of the ‘‘felicity’’ of its search for knowledge, the Austinian ‘‘force of utterance’’ is constantly in excess over the meaning of the theoretical statement. It is precisely this excess of energy that is continually discharged through humor. ‘‘[Some] French authors,’’ writes Freud, ‘‘describe laughter as a ‘de¤tente’. . . . We should say that laughter arises if a quota of psychical energy . . . has become unusable, so that it can find free discharge’’ [ Jokes, p. 147]. Humor indeed is preeminently not a ‘‘saying’’ but a ‘‘doing’’: a ‘‘making (someone) laugh.’’ If Austin is continually taking and giving the pleasure of jokes, it is because, paradoxically, the supreme performance of the body’s failing itself is that of making jokes [ faire de l’esprit]. To be sure, the ‘‘joke’’ participates in the ‘‘doing’’ of seduction, stemming in its turn, like the banquet Don Juan offers, from the gratuitousness of the gift of pleasure. As Freud writes,‘‘the psychical process in the hearer . . . can scarcely be more aptly described than by stressing the fact that he has bought the pleasure of the joke with very small expenditure on his own part. He might be said to have been presented with it ’’ ( Jokes, p. 148). However, Austin’s invitation to laughter is not just (yet another) invitation to pleasure but, more specifically, an invitation to the pleasure of scandal: ‘‘we shall most interestingly,’’ as Austin put it, ‘‘have committed the act of bigamy.’’ The laughter provoked by a joke turns the reader into an accomplice: an accomplice precisely in scandal. (113) Austin does love scandal, in the most lurid and meretricious sense of that word; he loves the spectacle of the minister baptizing the monkey and marrying the penguins, of the chance passerby inadvertently christening the ship. He loves the farce of the banana peel, the gleeful oops! celebrated utterly without dignity (after his death, mostly) on TV shows like Candid Camera and Sports Bloopers and America’s Funniest HomeVideos. But the scandal he delights in is also more than sheer farce. Etymologically a scandal is a stumbling block, and Austin is constantly throwing stumbling

28

Performatives

blocks (or banana peels) in the path of classical philosophy. As Felman writes, ‘‘In fact, the Austinian performance of humorous slipping aims above all at shaking the institution of prejudice itself, the institution of beliefs or received ideas: ‘To feel the firm ground of prejudice slipping away is exhilarating . . .’ [HT, p. 61]’’ (121).4 Exhilarating, maybe, for someone like Austin who loves pratfalls; less exhilarating for the constative linguist who cherishes his dignity and the institutionalized received ideas that help him maintain it.

Language games The performative as pratfall: will this work? If I say merely ‘‘The performative, for example practical jokes, magic tricks, scandalous witticisms, clever blasphemies, double entendres, slippery evasions and inversions, euphemisms, subversive remarks, half-deliberate slips of the tongue,’’ do you see what I’m getting at? Is that enough? Can you go on from there? Have I said enough about Austin for now? Or try this: ‘‘The performative, for example language games, each of them a loose collection of unruly creative acts, play and playfulness, the putting on of plays, the staging of social relations.’’ I’m thinking of Ludwig Wittgenstein (1953/ 1984: no. 66), of course: Consider for example the proceedings that we call ‘‘games’’. I mean board-games, card-games, ball-games, Olympic games, and so on. What is common to them all? ^ Don’t say: ‘‘There must be something common or they would not be called ‘games’’’ ^ but look and see whether there is something common to all. ^ For if you look at them you will not see something that is common to all, but similarities, relationships, and a whole series of them at that. To repeat: don’t think, but look! ^ Look for example at board-games, with their multifarious relationships. Now pass to card-games; here you find many correspondences with the first group, but many common features drop out, and others appear. When we pass next to ball games, much that is common is retained, but much is lost. ^ Are they all ‘‘amusing’’? Compare chess with noughts and crosses. Or is there always winning and losing, or competition between players? Think of patience [solitaire]. In ball games there is winning and losing; but when a child throws his ball at the wall and catches it again, this feature has disappeared. Look at the parts played by skill and luck; and at the differences between skill in chess and skill in tennis. Think now of games like ring-a-ring-a-roses; here is the element of amusement, but how many other characteristic features have disappeared! And we can go through the many, many other groups of games in the same way; can see how similarities crop up and disappear. Not surprisingly, the playful open-endedness of Wittgenstein’s approach to language has made his constative followers as uneasy as Austin’s. Constative

Constative and performative linguistics

29

linguists want to regulate language, stabilize it, objectify it. Just as Austin’s followers have needed to codify speech acts, therefore, so too have Wittgenstein’s followers needed to codify language games, to write books with titles like The Vocabulary of Politics,The Logic of Moral Discourse,The Language of Literary Criticism, The Logic of the Social Sciences, The Logic of Religious Language, The Language of Education ^ rulebooks, basically, for various language games, imagined as distinct logics or grammars.5 But Wittgenstein too is exploding toward the performative, arguing that ‘‘the meaning of a word is its use in the language’’ (no. 43) and taking ‘‘the language’’ not in the constative sense invoked by Jerrold Katz ^ the ‘‘null context’’ ^ but performatively, as whatever people say and write, as a loose and unbounded collection of verbal acts or performances. These, for instance: Giving orders, and obeying them ^ Describing the appearance of an object, or giving its measurements ^ Constructing an object from a description (a drawing) ^ Reporting an event ^ Speculating about an event ^ Forming and testing a hypothesis ^ Presenting the results of an experiment in tables and diagrams ^ Making up a story; and reading it ^ Play-acting ^ Singing catches ^ Guessing riddles ^ Making a joke; telling it ^ Solving a problem in practical arithmetic ^ Translating from one language into another ^ Asking, thanking, cursing, greeting, praying. (Wittgenstein 1953/1984: no. 23) Is that the complete list? Of course not. There is no complete list. But if there’s no complete list, how can we ever hope to determine the shared essence of them all? Forget shared essences. Study the multiplicity. ‘‘Don’t think, but look.’’ (There is no complete list of performative linguists, either, but whatever partial list one were to draw up,Wittgenstein would be on it.)

A brief history of the constative and the performative The idea that language is fundamentally communicative or informative ^ message-bearing, constative ^ was first formulated by Plato, and through Plato became integral to Christian theology, and thus to that thousand-year period from, roughly, Augustine to Port Royal that was foundational for modern Western linguistics. Conceptually in Greek the word as logos was a static picture of reality, a stable container of information; and medieval theologians idealized the Logos-as-God’s-Word, under the aegis of Jesus and in the dual form of the Scriptures and their own pronouncements, precisely as the only reliable vehicle

30

Performatives

of truth. Without the linguistics of information, of communication, of reliably borne messages, there could be no theology; without platonic Christian theology as God’s Logos, as the secure stable container of God’s message to humans, there could be no constative linguistics. From the very start mainstream Western linguistics is normatively constative, wrapped up with ‘‘communication,’’ with the hegemonic transmission of information from point A to point B. But this ‘‘from the very start’’ is misleadingly simple. Constative linguistics didn’t just start. It was gradually mortared together to stem an ancient performative tide, to clear a space for calm certainty, for stable truth, for predictability within an already existing state of what was perceived as linguistic flux, chaos, confusion. After all, the performative, the incantation, the spell ^ the ritual words that shape reality ^ might be thought of as the primary tool of animistic religion. I speak, and it comes to pass. I discover your secret name and pronounce it in the spell which I cast upon you, and you are suddenly taken sick, even die (or you were sick, and are healed). I invoke a god or a spirit, and it appears ^ not in response to my summons, but in it, or through it. The tribe sings and dances victory over an enemy, and the enemy is defeated; the performative magic generates the courage and the strength the tribe needs to emerge victorious. Words have power over events natural and supernatural.Words do things. This primitive ‘‘performative linguistics’’ is patently an early attempt to control the utterer’s environment: words are invested with the power to effect change because they are human society’s most powerful technology of the day. The constative assault on the performative begins with the uneasy sense (let me put it philosophically) that performative utterances are insufficiently transcendental to effect absolute (or even adequate) control over one’s environment ^ that, because they are immanent in history, because they take their power from the bodily speaking of individual human beings in specific ritual situations, they have no absolute power. All you have to do to undo a spell, for example, is to get a rival shaman or witch to utter a counterspell. Magic has no access to a universal spiritual locus of truth, and thus has no power to arbitrate between conflicting magics. Magic, to put it in derogatory constative terms, is just a bunch of scared bodies going around casting spells on each other. The same problem plagues the modern performative as well. Austin devotes a good deal of space in Lectures II throughVII of HowTo DoThingsWithWords to the question of what makes a performative utterance ‘‘happy’’ ^ what defining conditions must be present in order for an utterance to perform the action that we think it ought to be performing. There must be a conventional procedure governing the verbal action, the person performing the procedure must be the right person to perform it, the procedure must be performed correctly and completely, if people performing the procedure are expected to do it with the proper thoughts or feelings, they must have those thoughts or feelings, and must live in accordance with them after the procedure is completed (Austin 1962: 14^15). One of the implications of these conditions that Austin himself only

Constative and performative linguistics 6

31

hints at, but that seems absolutely essential to me, is that it is each individual community that determines what will count as a performative, and when one has been successfully performed. If the minister about to marry a couple is drunk, for example, one congregation may simply suffer through the wedding ceremony in silence while another may raise a hue and cry and declare the minister unfit to proceed. The latter congregation is effectively deciding that the performative ‘‘I now pronounce you man and wife’’ uttered by this drunken minister will not marry the couple; a drunken minister is no minister at all and must be replaced by someone who better fits the community’s concept of a ‘‘proper’’ minister. This is, obviously, a judgment call: nowhere in the Bible or the religious or civil laws does it state that a drunken minister is unfit to marry a couple. But the authority is ultimately vested in the performers of ritual performatives by the community, by each community, in each performative situation, and it can be withdrawn by that community as it sees fit. This also means that differences of opinion may well arise among the community over whether a performative was successfully performed or not. A student of mine at the University of Mississippi told our class the story of her wedding the previous summer in a rural Mississippi black Baptist church. Her husband came from the north and was used to ministers being seminary-trained and speaking standard English in calm, measured tones; the wild sing-song preaching of the self-taught lay preacher in her church felt so alien to him, so unpastoral, that he frequently joked afterward that they weren’t married at all. Her Mississippi family loved their preacher, and yelled ‘‘Amen brother!’’ all through the ceremony.7 This ‘‘disintegration’’ of a single unified universal point of view is endemic in Austin, who would no doubt claim (and I would agree) that it is endemic in reality as well. Where the performative reigns ^ and a performative linguist would want to insist that it always reigns in the social world ^ conflicting communities frequently vie for the right to determine when a performative is to be judged as successful, when as a failure. If only one out of the fifty U.S. states legalizes marrying gays and lesbians, is a gay or lesbian couple married in that state to be considered married in other states as well? Is the minister or civil servant that married them to be considered the ‘‘right’’ person to utter the performative ‘‘I now pronounce you husband and husband, wife and wife’’? The officiant was obviously considered the right person when it happened, in the appropriate community; but since other communities do not recognize gay and lesbian marriages, it is difficult to universalize ^ and ultimately impossible to make a universalization stick. Inter-faith marriages are often contested in just this way: the bride’s community may insist that only a wedding performed by their priest is a real wedding, really marries the couple, while the groom’s community insists just as firmly and dogmatically that only one performed by their priest is valid. The matter is absolutely certain from within each community; considerably less certain from the point of view of the bride and groom themselves, who may have to try to live in both communities at once. Some countries allow polygamy; if a man with three legal wives moves to

32

Performatives

a country that does not allow it, should all three women be legally considered his wives? Should they all be eligible for social benefits as his spouses? International treaties have been worked out to determine the official policy in such cases, but those treaties do not necessarily extend to the actual communities in which these people live. A couple (or members of a polygamous marriage) may be shunned by a community that, for whatever reason, do not consider them legally married and refuse to socialize with people living in what they consider to be sin. The same is true of the other performatives as well. Watch a group of boys arguing over whether one of them really made a bet, and whether he now owes the other guy the $5 he supposedly bet him.‘‘I didn’t bet, we never shook on it!’’ ‘‘It doesn’t matter whether you shake on it, all you have to do is say it!’’ ‘‘Yeah, but I had my fingers crossed behind my back!’’ The constative scholarly community that has sought to make Austin their own, beginning in the 1960s with Benveniste and Searle, has worked hard to undo this ‘‘competing powers’’ aspect of the performative, which it inherits from millennia of such conflicting communities vying over who has the most powerful spell or incantation. The answer from a constative point of view, obviously, is to universalize the performative: ignore the real communities that will go on fighting about these things no matter what the theorists say, raise the discussion to the ‘‘proper’’ rational (i.e. abstract, transcendental, ideal) level, legislate for theory what can never be perfectly imposed on social reality, and declare the problem solved. But the performative power of magic has proved extremely resistant to constative repression, and almost impossible to dislodge; even today we skeptical, scientific moderns find ourselves at least behaviorally believing, acting as if we believed, that verbalizing a streak of good luck can change that luck. We think, for example ^ we catch ourselves thinking, despite our own ‘‘true’’ lack of belief in any of this ^ that the act of saying ‘‘I can’t believe nothing’s gone wrong yet’’ can actually make something go wrong. The performative jinx. This performative magic is inimical to monotheism, which wields power through the centralized and bureaucratized control of truth. Monotheistic religions have always worked hard, therefore, to eradicate it, both by demonizing it (magicians are invoking the devil) and by ridiculing it (superstition is unscientific). To be sure, even Christianity retains powerful traces of performative magic. The Hebrew Bible begins with a story about the performative utterance ‘‘Let there be light’’ (Genesis 1:3), and in the New Testament the Word Made Flesh walks around Galilee healing the sick with the performative words (disguised as a constative) ‘‘Your faith has made you well’’ (e.g. Matthew 9:22; Mark 10:52; Luke 8:48, 18:42). The Pharisees’ reaction to this magic is strongly indicative of the direction monotheism is (and has been) heading: Jesus is thematized as a minion of the Evil One, channeling satanic power through his magic (Matthew 12:22^29; Mark 3:22^27). Christianity does in the end validate Jesus’s performative healing, and even hands it down to the twelve disciples; but the power of words to effect change is severely circumscribed, restricted to the priesthood, and by the time of the letters of platonizing Paul, two or three

Constative and performative linguistics

33

decades later, there is already a strong move away from the power of words, toward the truth of words. Increasingly in Christianity the word has an impact on human existence only insofar as people recognize its truth ^ never directly, never by creating or dissolving whole human institutions through the enunciation of a dread word. The ancient performative power of words to effect real change is retained in the sacramental rituals of baptism, absolution, and marriage, of course; to say ego te absolvo is to absolve, to cleanse the sinner of sin. But this power is closely guarded, restricted to an elite few (who for a great many centuries performed the ritual in a language most church-goers did not understand) and in any case severely qualified by the anxious in nomine Patri, et Filii, et Spiritus Sancti: if a human being’s words do have the power to effect real change, it is only as the agent of God’s performative Logos.8 Most important, for ordinary sinners tapping into this power is strictly taboo:9 ‘‘Again you have heard,’’ Jesus tells his disciples,‘‘that it was said to the men of old,‘You shall not swear falsely, but shall perform to the Lord what you have sworn.’ But I say to you, Do not swear at all, either to heaven, for it is the throne of God, or by the earth, for it is his footstool, or by Jerusalem, for it is the city of the great King. And do not swear by your head, for you cannot make one hair white or black. Let what you say be simply ‘Yes’or ‘No’; anything more than this comes from evil’’ (Matthew 5:33^37). The performative power to ‘‘make one hair white or black’’ by merely uttering the words let it change color is God’s, and God’s alone; if a human being (especially one unordained by God to wield his power) attempts to perform actions with words, by promising or vowing in the name of God or the earth or the city, the power is not from God but from Satan. Anything more than yes and no, anything more than a simple response to truth-questions, the interrogative form of the constative, is evil. And, by extension, it is evil also to challenge the ancient assumption that language is for conveying information ^ to suggest that words might be used for ‘‘anything more’’ than stating things. This goes too far, you will say, and of course it does: even if some medieval theologian had overtly denounced performative linguistics, we moderns ^ secular thinkers of the last three or four centuries ^ do not commonly think of acceptable and unacceptable modes of thought in the eschatological terms of ‘‘good’’ and ‘‘evil.’’ But the taboo remains with us, wielding power from our ideological unconscious. We feel it in our reluctance to admit the full force of Austin’s insight. We feel it in our relief when John Searle ‘‘saves’’ us from the potential anarchy of Austin’s little book in Speech Acts. Authors of introductory texts on linguistics in the decades since Austin wrote feel it in their deep-seated need to ‘‘forget’’ the performative in answering the question,‘‘What is language for?’’Austin himself, in the passage I quoted in the epigraph to Part I, describes this need to forget the performative as a kind of childhood disease, a stage philosophy had to pass through on its way to bigger and better things. This is a little of Austin’s incredible anarchism, his boyish, pranksterish iconoclasm (of which more follows): the theological substrate of constative linguistics, the naturalization of religious sanctions, ‘‘unconscious’’ but ‘‘precipitate,’’ is a trivial mistake, a booboo we were lucky to

34

Performatives

make because, having made it, we can now correct it. Oops! Oh well, no use crying over spilled milk! Austin is aware, of course, of how deep this constative assumption lies in our ideological knowing, how ‘‘natural’’ and therefore intuitively right it feels, and thus how counterintuitive his performative corrective must be.‘‘CAN SAYING MAKE IT SO?’’ he asks in a section heading to his first lecture, and continues: ‘‘Are we then to say things like this: ‘To marry is to say a few words,’ or ‘Betting is simply saying something’? Such a doctrine sounds odd or even flippant at first, but with sufficient safeguards it may become not odd at all’’ (7). But of course Austin certainly does not provide ‘‘sufficient safeguards’’ from the point of view of such constative followers of his as Emile Benveniste, Jerrold Katz, and John Searle; his love is for tearing down safeguards, not building them. It is useful, perhaps, to recall that Austin was not the first to articulate a performative linguistics ^ that we are faced, here at the beginning of the book, looking backward toward the beginning of our civilization, not with an articulate hegemonic linguistics and an inchoate sublimated magic, but with two distinct linguistic traditions of similar age (though admittedly dissimilar histories of social legitimation). It may bolster our confidence to see the ‘‘saying things/doing things’’ opposition as a modern avatar of the ancient debate between logic and rhetoric, description and persuasion; between Plato and the Sophists (and the ancient Greek poets), between the Rabbinical tradition and the Kabbalists, between Augustine and the Gnostics. In A Map of Misreading, Harold Bloom (1975: 42) describes a similar clash between the Greek word for word, logos, with its associations of static visual structure, and the Hebrew word for word, davhar, with its associations of dynamic human action. In Faust Goethe fought the dominant constative linguistics of the West by retranslating the opening line of John’s Gospel from ‘‘In the beginning was the word [das Wort]’’ to ‘‘In the beginning was the deed [dieTat]’’ ^ patently an attempt to recover the peripheralized rhetorical tradition and to ground language not in inert transcendental structure but in creative human actions. And the German romantic tradition since Goethe has insisted on this revisionism, from Hegel through Marx, Nietzsche, and Heidegger to Gadamer (who inTruth and Method [1960/1989: Section 3.2.A] rereads the history of medieval Christianity so as to find a dynamic, creative, active Logos in Augustine and Thomas Aquinas). But this historical reassurance requires archeological research, the painstaking excavation of systematically buried precursors. Constative linguists have not wanted these counterhegemonic traditions handed down. Constative linguistics, like all other hegemonic Western sciences, is founded on the repression of the magical, the mystical, the rhetorical, and the romantic. To find those things in the history of Western thought, we have to dig.

Integrational linguistics The methodological power of the performative to transform linguistics can also be illustrated negatively, by measuring the relatively weaker power of

Constative and performative linguistics

35

an alternative to Saussurean linguistics that eschews it. Several such alternatives could be examined, including the ‘‘empiricist’’ (contra ‘‘idealist’’) one offered by Geoffrey Sampson (1986,1997); let me focus here, however, on only one, what I take to be the most influential alternative paradigm being offered today to Saussurean (or what Harris calls ‘‘segregational’’) linguistics, the ‘‘integrational linguistics’’of Roy Harris. Harris has been an indefatigable critic of what I am calling constative linguistics since the late 1970s, with his books On the Possibility of Linguistic Change (1977) and Communication and Language (1978), and then more pointedly in the early 1980s, with his attacks on Chomskyan TG-grammar, The Language Makers (1980) and The Language Myth (1981). His opposition to Chomsky was at once spirited and scholarly, and therefore for some of us highly persuasive, but isolated enough, and perhaps ‘‘spirited’’ enough (in an academic context this ‘‘spirit’’ may sound to some like hysteria) to be fairly easily dismissible by the linguists he was attacking and their many followers as a crank. Linguistics was still very strong in the late 1970s and early 1980s; voices raised against it did sound unserious, even irresponsible. By the late 1980s, with books like The Language Machine (1987), and Reading Saussure (1987), Harris’s critique had expanded to encompass the entire Western tradition behind orthodox linguistics. By the late 1980s he had also begun to attract a number of sympathetic and even enthusiastic readers (including me); some of these became his collaborators in the books of the 1990s, such as Taylor J. Talbot, with whom he wrote Landmarks in Linguistic Thought: The Western Tradition from Socrates to Saussure (1989), Hayley Davis, who coedited a volume with Taylor J. Talbot called Redefining Linguistics (1990) based around Harris’s ideas, and George Wolf, who coedited an essay collection with Harris called Integrational Linguistics: A First Reader (1998) and, the previous year, edited a collection for John Benjamins entitled Linguistics Inside Out: Roy Harris and His Critics (1997).10 In 2000 this group founded the International Association for the Integrational Study of Language and Communication (point your browser at http://www.gold.ac.uk/academic/eng/iaislc/), which has been active in promoting Harris’s oppositional approach in conferences and publications. Since the late 1980s, as I say, Harris’s term for what I am calling constative linguistics has been ‘‘segregational’’ linguistics; by obvious contrast his own alternative to that approach he calls ‘‘integrational’’ linguistics. His negative critique of the former began early, in a somewhat inchoate way in the 1970s, and became significantly more systematic and pointed in the 1980s. Harris only began working out his ‘‘integrational’’ alternative to segregational linguistics in the books and articles of the 1990s, and indeed some of the books of the late 1990s have the term in their titles: the ‘‘first reader’’ that he coedited with George Wolf, mentioned earlier, Signs, Language and Communication: Integrational and Segregational Approaches (1996), and Introduction to Integrational Linguistics (1998). Harris’s term ‘‘segregational’’ comes from his perception that the key idealizing move made by the mainstream linguistic tradition

36

Performatives

that he attacks is to segregate what they are pleased to call ‘‘language’’ from actual language use, from what I am calling the performance of speech acts: in Saussurean terms, la langue, or the pure abstract fixed code that supposedly ‘‘underlies’’ language use, is theoretically and methodologically segregated from la parole, or speech, language as actually used by real human beings. This also entailed, Harris shows, the segregation of ‘‘linguistic science’’ from all other sciences, such as psychology (what goes through the mind of the user of language while using it), sociology (how language communities regulate linguistic behavior), rhetoric and literary studies (how language is used for persuasive and expressive purposes), physiology (how language sounds are produced and received by the human body), and acoustics (how the sounds of language are transmitted across sound waves). As Harris (1990: 21^2) writes: What in fact that definition amounted to was a decision to restrict the concept ‘‘language’’ in a particularly narrow way. In the first place, it restricted language to speech; and then it restricted speech to the production of determinate strings of phonemes, segmentable into determinate substrings, each identifiable as the manifestation of a determinate linguistic sign. Each linguistic sign was assumed to have a determinate form, a determinate meaning, and a determinate capacity for linear combination with other linguistic signs. In brief, it was a linguistics which could handle the phenomena of speech only in so far as a speaker’s vocalization was reducible to a set of determinable phonological forms with determinate meanings and a determinate combinatorial pattern. Any aspects of speech not reducible to this schema were simply ignored. The basic contention of the present paper is that the fundamental error of contemporary linguistics is still the fundamental error of Saussure’s original thesis. It involves the process of abstraction by which certain phenomena are segregated from the continuum of human communication, and these segregated phenomena are then, rather capriciously, set up for academic purposes as constituting the linguistic part of communication. The defining error by which Harris thematizes mainstream linguistics, then, is this idealizing separation or segregation, this insistence on abstracting ideal forms out of the apparently chaotic welter of actual language use and ignoring everything that lies outside the idealization; if that is the main problem, obviously, the solution will be to integrate what has been segregated. Harris’s critique of segregational linguistics is wideranging; it amounts to a systematic unpacking of the implications of this segregationalism for the twentieth-century history of linguistics (taken from Harris 1990): 1. The principle of arbitrariness. Signs only have an arbitrary meaning in the abstract, not in actual communicative contexts. Insisting on arbitrariness

Constative and performative linguistics

37

only serves to separate what linguists call ‘‘language’’ from actual language use. 2. The principle of linearity. Not all language use is linear (the writing and reading of poetry mostly isn’t, for example, and it has been shown that people will typically read a page of prose, too, more like a painting, the eye skipping all over the page, than linearly). The only language use that is obviously linear is speech, and even its linearity is problematic, not only because the ear too jumps around, forward and backward on the supposedly unidirectional linear track of speech; phenomena like simultaneous interpreting and our ability to converse with one person while simultaneously listening in on another conversation near us also militate against the reduction of all speech to linearity. 3. The focus on speech to the exclusion of writing. In literate societies it is impossible to separate speech and writing psychosocially; each influences the other in powerful and deep-seated ways. This critique of Saussurean linguistics was begun byJacques Derrida in Of Grammatology, in fact, in the late 1960s (1967/1974). Speech can only be segregated from writing in the abstract, in the hypothetical realm of la langue (which linguists beginning with Saussure of course hypostatize as the true language that underlies actual use). 4. The insistence on studying fixed codes. Fixed codes do not, of course, exist outside of the idealizing linguist’s imagination; they are also hypothetical calculi designed to make analysis easier. But the effect of studying those hypothetical fixed codes to the exclusion of actual language regularities, with all their overlapping complexities and contradictions, is once again to segregate the linguist from the real world of linguistic communication. 5. The theory of telementation. This theory, according to which speakers embed thoughts in auditory form in order to transport them to the minds of their listeners, was developed out of a cross between Saussure’s notion of a ‘‘speech circuit’’ and the early information-theory models of communication that I mentioned in the introduction. Because these models were based on machine communication, they definitively lack human agency (subjectivity, intentionality, will): according to the theory of telementation, thoughts simply get encoded in linguistically significant sounds, in which form they then get carried across the sound waves to the listener’s ear, and through the ear into the brain, where they get decoded as meaning. It could be argued, in fact, that the methodological exclusion of human agency is strategic, because humans are notoriously susceptible to error and distortion. If we can remove humans, speakers and listeners, from the equation as far as possible, if we imagine them not so much as wanting to say this or that (or wanting to understand what they hear in certain ways) as simply functioning more or less automatically as the carriers or channels of the fixed linguistic codes, then it becomes easier to imagine the hypothetical perfection on which this model is built. 6. The focus on synchrony to the exclusion of diachrony. This is an almost inescapable consequence of the insistence on fixed codes: if your code must be fixed,

38

Performatives

clearly, it cannot change. The best way to avoid that whole problem, therefore, is to exclude diachronic perspectives on language and focus on single stop-frame slices of linguistic life. This obviates the necessity of trying to imagine how language might ever move from one stop-frame to the next. Or, in the series of critiques that Harris (1990) offers of this notion: a. The synchronic study of fixed codes makes it impossible to explain how a person could ever learn unfamiliar words by asking someone else. If the asker and the asked share the same fixed code, they must already know the word; if they don’t both know it, they don’t share the same fixed code, and therefore have no possible way of rectifying their differences. Of course asking the meaning of a word is extremely common in real life, and requires no great ingenuity to explain in psychosocial terms. This theoretical stupidity among the segregationalists arises solely out of the need to segregate the fixed code from the corrupting influence of actual language users, especially as that influence functions over time. b. The synchronic study of fixed codes also makes it impossible to explain where the fixed code came from. This for segregational linguists is a nonissue, of course: a diachronic question that Saussure said they didn’t need to worry about. Where do languages come from? How do people (little children, foreigners) get inducted into them? How is language regulated by the community that speaks it? These are all ‘‘segregated’’ off as ‘‘nonlinguistic’’ questions. c. In this perspective, clearly, linguistic change too becomes a theoretical impossibility. No one can ever introduce a new word, because by definition new words fall outside of the existing fixed code and therefore cannot be understood or learned by its current users. But fortunately the principle that language is to be studied synchronically rather than diachronically enables segregational linguists to ignore this theoretical conundrum too. d. ‘‘A fourth objection to a linguistics based on the fixed-code theory is that even if A and B were using the same fixed code they would never be able to be sure of this. For if B wishes to verify that the ideas A wished to convey are indeed those which B interprets A’s utterances as conveying, B must either elicit further utterances from A or else assess A’s reactions to further utterances by B, or both’’ (Harris 1990: 34). This objection is shaky, however, because it is based on the assumption that the use of these fixed codes by real people in actual language-use situations is somehow relevant to linguistic study, when of course according to Saussure and his followers it is not. One could go on with this; the integrationalist critique of segregational linguistics is comprehensive, and convincing. My point in introducing this critique here early on in my argument, however, is to note that to date ‘‘integrational linguistics’’ has consisted almost entirely in this comprehensive and convincing critique. Invariably, when Harris and his colleagues get around to writing about the ‘‘integrationalist program,’’ their remarks become vague and evasive and quickly turn back to negatives: ‘‘A demythologized linguistics,’’

Constative and performative linguistics

39

Harris (1990: 45^6) writes, ‘‘(that is to say, a linguistics liberated from the telementational fallacy) would sponsor a type of programme which will be briefly outlined in what follows. Once a telementational model is rejected, one can reject along with it the orthodox constraints on what a language is conceived to be.’’ And he goes on to list those rejected constraints at such great length that he never really gets around to offering a positive program. This is a recurring phenomenon in the work of Harris and the other integrational linguists he has gathered around himself: again and again, throughout the books and articles of the 1990s, they promise the reader an integrationalist program and then find themselves enumerating one more time all the theoretical elements that they won’t need from segregational models.11 Integrational linguistics has mustered a powerful critique of segregational linguistics; it has not yet offered a detailed methodological alternative to it. And I would suggest that this has something to do with the particular thematization that Harris developed of the key Saussurean error: segregation. Harris and his colleagues repeatedly point out crippling segregations in the works of their mainstream opponents and then say only that an integrational linguistics would integrate these things. Yes, well, right, one wants to say, that’s obvious, that follows logically from the evils of segregation . . . but how? What should be integrated, and what steps should be taken to bring that integration about? What theoretical and analytical models should be developed to make integration effective? Integration is only the necessary first step toward a new linguistic paradigm, and, big and drastic and difficult as that step is, given the magnitude of the theoretical hegemony these people are seeking to dislodge, just mentioning its necessity does not help us imagine what lies beyond integration. The problem is similar to that faced by feminist and postcolonial theorists: having shown at great length that patriarchal gender identities and colonial identities are harmful and destructive, how do we go about imagining something beyond them, the next step past a negative critique? In this light the power of the performative should be obvious. Behind the performative as a speech act lies performance, drama, the entire rich world of the theater, acting, staging, pretending, histrionics. That world offers copious methodological metaphors for the study of language. To perform is to act in a social context ^ precisely what a radical post-Saussurean linguist will want to imagine speakers doing when they speak, responders doing when they respond. And we already have a complexly fleshed-out sense of what all performing might entail, from our experience of the theater, from performance studies, from pop-psychological concepts of ‘‘performance anxiety,’’ and so on. ‘‘Performance’’ is conceptually and imagistically a far richer field than ‘‘integration.’’ As the book proceeds, I will be borrowing other metaphors for this paradigm shift as well, from other theorists who have attacked mainstream linguistics: Jacques Derrida’s (re)iterations, Antonio Damasio’s somatic

40

Performatives

markers, Mikhail Bakhtin’s dialogue, Paul Grice’s conversational implicatures. Each of these metaphorical fields will be presented, however, under the aegis of performance.‘‘Performative linguistics’’ is about real human beings performing language as an actor performs his or her role ^ doing things with words in a complex social context.

3

Translatorial performatives

Before we move on to more complex figurations of performance, let us pause to put Austin’s own conception of the performative to a strenuous test: the act of translation. Do translators, in reproducing other people’s discourse, ever themselves perform speech acts? Can translation ever be thought of as performative? A good many translation scholars have likened translating to performing in another sense, the translator as performer, the translator standing in the same relation to the original text as the musical performer does to the score (see Wechsler 1998 for an overview) ^ but can the translator ever be thought of as uttering a performative utterance in Austin’s sense? A few constative linguists have attempted to apply speech act theory to translation, mostly by taking speech act theory as a useful linguistic description of what goes on in language ^ in a source text, say ^ and expecting the translator to reproduce those speech acts accurately in the target language. In Discourse and theTranslator Hatim and Mason state categorically that ‘‘the translator will seek to relay the illocutionary force of each speech act in turn’’ (1990: 61), adding that translators should really be able to identify speech acts by type, so as to be able to translate them most effectively into the target language. Fawcett (1997: 128^9), noting that Hatim and Mason’s position is probably extreme ^ ‘‘a fairly literal translation will in very many cases produce the desired effect,’’ without the translator’s needing to know what the speech acts are called, or even that they are speech acts ^ suggests that constative linguists seem to agree that speech acts will not differ radically from language to language. This may in fact be why speech act theory has not proven as productive for a constative linguistics of translation as Gricean implicature, the topic of Part III: Obviously, when a translator has to translate the more obvious ritualistic performatives (of the ‘‘I now pronounce you man and wife’’ variety), then a literal translation will not do, and the appropriate form of words needs to be used. It is quite surprising, therefore, not to find these formulas in bilingual dictionaries. Other speech acts of the non-ritualistic kind, such as promising, betting and so on, should also pose no problem, since, as Palmer says (1981: 166),‘‘speech acts are probably independent of the actual languages’’ (although he adds an immediate, second qualifier in the form

42

Performatives of ‘‘at least to some degree’’). Searle also accepts that ‘‘translations of the sentences in question will often, though by no means always, produce sentences with the same indirect illocutionary act potential of the English examples’’ (1975: 68), just as Ho«nig and Kussmaul say of a sentence to be translated into German ‘‘However it is translated, the illocutionary force of the sentence would not change’’ (1982: 80). (Fawcett 1997:128^9)

Fawcett goes on to summarize other constative work done on speech acts in translation, notably the work of Snell-Hornby (1988) and Leuven-Zwart (1990) on cultural differences in the use and uptake of speech acts between the source and the target language; but since all of this research is constative, aimed at helping translators restate the source text accurately in the target language, I won’t go into it further here.

The translator’s personlessness The one performative linguist of translation to undertake a study of the specifically translatorial performative, as I mentioned in Chapter 1, is Anthony Pym. The argument as Pym develops it begins in his 1992 book Translation and TextTransfer, where he posits that the translator, in the act of translating, cannot say ‘‘I am translating’’ ^ cannot, in other words, make the translation into a first-person utterance. Translators, Pym argues, are by definition conveyors of other people’s words, not (while translating, in the actual translation-asutterance) the utterers of their own words. If this means that the translators-asutterers have no first person, it also means that they have no second person either: if translators cannot speak as themselves, neither can they address the readers of the translation (again, in the translation ^ Pym makes a distinction between the translator-as-translator and the translator-as-author-of-the-translator’s-preface).‘‘One could say,’’ he writes,‘‘that the utterance ‘I am translating’ is no truer than is its extended form ‘You are reading the translation I am now doing’. The translational second person must be as anonymous as the translator’’ (1992: 57). He also notes that this premise of the translator’s linguistic personlessness is congruent with a constative approach to translation (which he doesn’t call by that name, of course), and may in fact have predisposed linguistic scholars of translation to constative assumptions: ‘‘This analysis should further explain why translational discourse is happier referring to words rather than to subjectivity, to neutral objects rather than to positioned first or second persons. If the translator’s position is to be without linguistic manifestation, so must the first and second persons of properly translational discourse’’ (ibid.). This is roughly where he picked up the argument in a paper he gave at the Current Trends in Translation Theory conference in Szombathely, Hungary, in late 1992, the same year as Translation and Text Transfer appeared, and published the following year with the title ‘‘Performatives as a Key to Modes of

Translatorial performatives

43

Translational Discourse’’ (Pym 1993). Quite clearly, he notes there, Austin’s performatives require first-person address: I pronounce, I bet, I promise, and so on. Theoretically, then, if translators are denied access to the first person, they must be denied access to performatives as well. The translator, according to his theory from Translation and Text Transfer, cannot ‘‘do things with words’’: cannot perform speech acts. And his initial ruminations seem to confirm this. He poses a hypothetical simultaneous interpretation from English into French in the opening ceremonies of an academic conference: 2a.

I declare the meeting open.

Half a second later this is ‘‘simultaneously’’ translated as: 2b.

Je de¤clare ouverte la re¤union.

Here we find that although both utterances are well-formed performatives, only the chairperson’s statement (2a) can properly perform. The interpreter’s version (2b) will necessarily have a constative function with respect to the utterance that actually opened the meeting. In fact, its discursive value could easily be rewritten as third-person reported speech: 2c.

Le Pre¤sident vient de de¤clarer ouverte la re¤union.

Thus, a properly translational relationship between two performative forms seems to imply that only one of those forms, the non-translational one, can actually perform. The second utterance, which arrives just a half-second too late, has its function blocked by the presence of an anterior first person, visible to all receivers of the translation. That is, the communication scene is already occupied by a first person with full capacity to perform. Chairpeople can open conferences; interpreters cannot. Or more generally, translational discourse seems by definition to exclude the possibility of a fully performative discursive function. And obversely, a translation that has a fully performative discursive function might then no longer be properly translational. (1993: 50^1) If this is in fact true, if translators qua translators never utter performatives, if the speech acts they perform are merely the constative ‘‘statements’’ or restatements of someone else’s speech acts, then, one might also conclude resignedly, translation theory should by rights be constative as well. No performatives in translation, no performative studies of translation. But in fact, Pym shows, this latter does not follow. Even if his initial suspicion is correct and translations never serve as their translators’ performative utterances, translators still do something, still perform speech acts ^ just constative ones. ‘‘The interpreter’s version (2b),’’ he writes, ‘‘will necessarily have a constative function with respect to the utterance that actually opened the meeting.’’ The

44

Performatives

interpreter in this example is not performing the action of opening the meeting, but is performing a speech act nevertheless, which Pym wants to call a constative, because the translator can be imagined as saying not ‘‘I declare this meeting open’’ but ‘‘the president declares this meeting open.’’ The translator (con)states something; performs a constative speech act. Pym’s methodological focus, in other words, is what translators do ^ not, as for constative theorists, the dead letter of the translated text. His focus is on translatorial action; and if that action proves not to be a performative, if translators in and through their translations never actually open meetings, they still are doing something ^ even constatives are speech acts! ^ and Pym submits that it should be the concern of the translation scholar to figure out what that something is. I would add: that should be the concern of the performative translation scholar. My reservation about this first tentative stage in Pym’s analysis is that I’m not convinced that the translator is (con)stating anything. If s/he is, if to say into a microphone in the interpreting booth at the back of a conference ‘‘Je de¤clare ouverte la re¤union’’ is to state that the president has just declared the meeting open, it is not an overt constative. If anything, it is an implicit constative ^ something we don’t yet have the conceptual tools to begin to comprehend (more of this in Chapter 10). In straight Austinian terms, at least in its ‘‘ideal form’’ (staying for the moment with Pym’s adherence to the constative Austinian/ Benvenistean tradition), it is not a constative at all but what Austin would call a ‘‘parasitic’’ performative: it looks like a performative, it has the ideal form of a performative, but it doesn’t perform the action that it seems to be performing. It doesn’t open the meeting. It is like a performative uttered by an actor on stage: the actor playing a minister who says to the actors playing a bride and a groom,‘‘I now pronounce you man and wife,’’ does not thereby marry the couple. And in precisely the same way, the interpreter rendering ‘‘I declare the meeting open’’ into French does not thereby open the meeting.12 Pym runs into this problem partly, it seems to me, because he is working with a highly simplified Austinian conceptual framework: no parasitic speech acts; and neither the Derridean nor the Gricean solution to the problem of the parasitic, which we will be returning to in Parts II and III, respectively. But I think it is also more than that. Pym is also to some degree caught between conceptual frameworks, here ^ between the action-orientation of performative linguistics and the structure-orientation of constative linguistics. He wants to analyze translations in terms of performative and constative speech acts ^ he wants to study the ways in which translators do something in translating ^ but in order to do so, in order to treat performative and constative utterances as recognizably different in structure, he has to rely heavily on the constative recuperations of Austin’s terminology offered by Benveniste and others, who speak of the ideal form of performatives and constatives. For Austin, as we’ve seen, performatives were first doing something with words and constatives were saying something with words; once he realized that performatives say things too, and constatives perform actions, he collapsed the dualism into the new framework of locutions, illocutions, and perlocutions, the three parts of

Translatorial performatives

45

a ‘‘speech act.’’Austin never theorized constatives as speech acts; but the constative Austinian tradition, beginning with Benveniste, does, and Pym follows the latter here. He wants to study translations as speech acts, as actions performed in real-world contexts (to be, in my term, a performative linguist); but he wants to study those speech acts in terms of their idealized structures. He wants, in other words, to hover back in the early lectures of HowTo DoThings With Words: he wants to determine whether a given apparent performative ‘‘uttered’’ by the translator is a true performative or not, based on whether it has the proper performative structure and whether it meets certain formalized conditions for ‘‘felicity’’: (A. 1) There must exist an accepted conventional procedure having a certain conventional effect, that procedure to include the uttering of certain words by certain persons in certain circumstances, and further, (A. 2) the particular persons and circumstances in a given case must be appropriate for the invocation of the particular procedure invoked. (B. 1) The procedure must be executed by all participants both correctly and (B. 2) completely. (G.1) Where, as often, the procedure is designed for use by persons having certain thoughts or feelings, or for the inauguration of certain consequential conduct on the part of any participant, then a person participating in and so invoking the procedure must in fact have those thoughts or feelings, and the participants must intend so to conduct themselves, and further, (G. 2) must actually so conduct themselves subsequently. (Austin 1962:15) Drawing implicitly on this Austinian formulation of conditions for the ‘‘felicity’’ of a performative, Pym seems to be arguing that the interpreter saying into a microphone in the interpreting booth at the back of a conference ‘‘Je de¤clare ouverte la re¤union’’ is not the appropriate person to declare the meeting open (A. 2), and does not have the proper thoughts or feelings of someone charged with opening the meeting (G. 1), so cannot be thought of as performing the action of opening the meeting. By this token, ‘‘Je de¤clare ouverte la re¤union’’ would be an infelicitous performative; but because, like the actor uttering a performative on stage, the interpreter did not intend for his or her utterance to open the meeting but was doing something else, in Austin’s terms it is in fact not a failed or infelicitous performative but a parasitic performative. That this is an inadequate solution to the question of what the interpreter was doing in saying those words in French should be perfectly clear; as we will see Derrida insisting next chapter, calling a thing ‘‘parasitic’’ and excluding it from one’s inquiry is basically just a way of saying ‘‘I don’t know what’s going on in it.’’ If the

46

Performatives

interpreter’s words were not intended to open the meeting, what were they intended to do? Pym’s solution, which is one step in the right direction but still inadequate, is to suggest that we think of the interpreter’s utterance as a constative: ‘‘The president declares the meeting open.’’ This has the virtue that in the real world interpreters do sometimes resort to third-person address, to make it clear that they are not uttering the speaker’s words in their own name; but obviously it is still far from a solution. Having offered the tentative formalistic suggestion, then, that translators may in fact never offer their translations as performative utterances, Pym goes on to consider cases in which this suggestion seems to fall apart ^ cases where a translation does seem to have a performative function. This for me is the most interesting part of Pym’s paper, the part where he begins to expand the realm of translation in performative linguistics by examining the translator as an ethical agent, and I want to devote the rest of this chapter to a close look at it. Pym takes the example from Annie Brisset’s (1991) remarks on Michel Garneau’s Que¤be¤cois French translation of Shakespeare’s Macbeth: In Shakespeare’s Macbeth, Macduff is reacting to the news that his estate has been destroyed and his wife and children massacred: 3a.

I cannot but remember such things were, That were most precious to me. Did heaven look on, And would not take their part!

The utterance ‘‘I cannot but remember’’ should be seen as functionally equivalent to the performative ‘‘I remember,’’ since the dramatic speech is in fact the act of remembrance. But if we can accept this as a performative discursive function, the same should surely be said of its Que¤be¤cois translation by Michel Garneau: 3b.

C’que j’ava’s d’plus pre¤cieux dans l’monde, chu t’oblige¤ d’commencer A m’en souv’nir. Comment c’est que l’bon dieu peut laisser fe¤re Des affe’res pareilles? Sans prendre la part des faibles?

As Brisset points out, the phrase ‘‘A m’en souv’nir’’ in 3b is peculiarly performative not just because of the dramatic situation ^ which could still involve a Que¤be¤cois actor constating the grief of a Scottish Macduff ^ but because the translation involves a constant semantic shift to associate the situation of Scotland with that of Que¤bec (seen in the rendering of the neutral ‘‘their part’’ as ‘‘la part des faibles’’). This target-text performative function also depends on the translation being performed within the specific social context of Que¤bec: ‘‘The wording of this resolution echoes the declaration Je me souviens (‘I remember’) which is such a prominent feature of Que¤be¤cois social discourse (it is on every vehicle’s license plate)’’

Translatorial performatives

47

(Brisset 1991:126). The translational phrase ‘‘a' m’en souv’nir’’ calls up a powerful contextual phrase that enables a functionally performative ‘‘I remember’’ to be performed not just by the actor but also by a specific audience receiving this translation. (1993: 52) The situation Brisset describes here is, of course, Pym realizes, a complex one. Do we want to say that the translator Michel Garneau uttered the translation as a performative, or merely that this particular target audience received the translation as one? This is on the one hand a rather constative quibble: in a performative linguistics of translation specific translations or translatorial utterances don’t ‘‘possess’’ the quality of performativity but are assigned that quality by people in specific speech act situations. Thus if Michel Garneau or the Que¤be¤cois theater audience took the French Macbeth to be a performative utterance, then for them, at that moment, it was. There is no ‘‘merely’’ about an audience’s reception. It is one active construction of the speech act situation. It is not just a reception of the drama; it is itself a drama. Without an audience to receive a speech act as this or that, a performative or a constative, a remembrance or a local nationalistic slogan, there quite simply is no speech act. Reception dramatically creates the drama as what the receivers take it to be.

Double reception On the other hand, asking the question about reception and its impact on the nature of the speech act raises important methodological issues for the performative linguist of translation. As Pym writes: One could argue that since the content remembered through the targettext performative (Que¤bec) is quite distinct from that of the source-text performative (Macduff ’s family), the translated text should lose translational status for as long as the receiver has forgotten about the specific context of Shakespeare’s play. On the other hand, there is no a priori reason to discount the possibility of a double reception process in which a Scottish family and Que¤bec could be simultaneously present through the discursive functioning of a translational performative. But there is an important condition here. Such a double reception is only possible in the specific social context of Que¤bec, number plates and all, where circumstances allow the second person to participate in the performative (that is, to recognise it as a performative and to accord the speaker authority to perform it). If the translation were received, say, in France, absence of this particularly participative receptive context would imply a merely observational second person and thus effective annulment of the translational performative. (1993: 53)

48

Performatives

Pym here sees two interpretive options; but implicit in his shortlist of options are at least two more (2 and 3), the argumentative steps that he leaped over en route from (1) to (4): 1. Translation, as he originally argued in Translation and Text Transfer, will not allow the translator-as-translator a second-person address to the audience. If Garneau does address his Que¤be¤cois audience, or if they take his translation to be addressing them, and in so doing forget that this is a play about Scotland, then it ceases to be a translation. (This is close to an argument from equivalence: if Garneau ‘‘adds’’ local connotations to the text, he is rewriting Shakespeare, not translating. But Pym has very carefully defined translation in terms not of textual correspondence, a constative approach, but rather of I^you address and speech acts, people talking to people, a performative one.) 2. Translation must be reimagined as allowing second-person address. If Garneau in and through his translation is taken to be addressing his Que¤be¤cois audience with the I’s and you’s of performatives, then the theory is wrong and must be revised. His French Macbeth is a performative translation. 3. Translations only figure as speech acts in specific situations, the address of a person (the translator) to another person or persons (the audience) in a certain place and time. It is not that Garneau’s Macbeth is or is not a performative translation; it is that it was taken that way in Que¤bec at a particular historical moment (the moment when Que¤be¤cois license plates echoed his translation). It would not be a performative translation in France, or perhaps even in Que¤bec twenty or fifty or a hundred years from now. 4. It is possible to integrate the strict theoretical definition of translation in (1) with the loose revised definitions in (2) and (3) by imagining a dual audience, and thus a dual uptake: any audience watching the play in translation, including perhaps one in Que¤bec that did not hear or think of the connection to the license-plate slogan, that understands Macduff ’s outburst to be a performative outburst of grieving remembrance of his family interprets the play in terms of Macduff ’s explicit I-address (and perhaps Shakespeare’s implicit I-address as well), hence as a Shakespearean utterance without performative intervention from the translator; the Que¤be¤cois audience receives it as Garneau’s performative outburst of grieving remembrance of French Canada’s sufferings at the hands of the English. So Garneau’s translation both is and is not taken to be a performative. Step (4) is, of course, where Pym ends up; it seems to be the terminus of his argument at this point.13 But there are problems here, perhaps because Pym does not spell out exactly what the ‘‘double reception’’ would entail: he writes only, somewhat cryptically, of a ‘‘double reception process in which a Scottish family and Que¤bec could be simultaneously present through the discursive functioning of a translational performative.’’ Present to whom? What kind of

Translatorial performatives

49

‘‘double reception process’’? What kind of ‘‘throughness’’? In (4), above, I interpreted that line to mean that the remembered Scottish family (murdered by Macbeth) and the remembered history of Que¤bec (oppressed by the English) might be simultaneously present to two different audiences, one of which reads Macduff ’s remembrance ‘‘straight,’’ without translatorial performatives linking it to Que¤bec, the other ‘‘through the discursive functioning of a translational performative’’ linking it specifically to Que¤bec. But Pym’s idea might be read differently as well, to mean the simultaneous presence of the murdered Scottish family and the history of Que¤bec, say on two different levels, in a single audience ^ even, for the sake of clarity, a single member of the Que¤be¤cois audience. I’m thinking of theories of irony, here, explanations of how irony works.14 There are three different models: (a) posit two different audiences, one that gets the irony, another that does not; (b) posit a single understanding receptor that posits the same two different audiences, one getting it, the other not; (c) posit that the person who gets the irony and the person who does not get it are the same person, inhabit the same body. Applied to the Macbeth example, this would yield: a. A Que¤be¤cois audience, sitting in the theater in Montreal watching a performance of Garneau’s translation, hearing echoes of Que¤be¤cois nationalism in it, and another audience, either in the theater in Montreal but not Que¤be¤cois or otherwise not sensitive to Que¤be¤cois echoes in the translation, or elsewhere, say in France. For the former, the Garneau Macbeth is performative and not a translation; for the latter, it is a translation and not performative. b. A single Quebecker, sitting in the theater in Montreal watching a performance of Garneau’s translation, hearing echoes of Que¤be¤cois nationalism in it, and imagining another audience that does not get it, does not hear those echoes. As in theories of irony, this single person imagining both sides of the uptake takes an added intellectual pleasure from imagining a hypothetical failure to understand what they understand. For this person, the Garneau Macbeth is performative and not a translation, but imagines the hypothetical nonunderstanding viewer taking it to be a translation and not performative. c. A single Quebecker, sitting in the theater in Montreal watching a performance of Garneau’s Macbeth, and simultaneously seeing it as both Shakespeare and Garneau, understanding that it is at once about Scotland and about Que¤bec, and therefore that it is both a nonperformative translation and a performative original work. The problem with both (a) and (b), obviously, is that they force us to make a dualistic choice between alternatives that look equally attractive, which makes the binary distinction (nonperformative translation vs. performative nontranslation) feel very arbitrary and contrived, like something one insists on out of a discomfort with complexity, something whose main value is logical clarity

50

Performatives

rather than explanatory power. The Quebecker in (b) hears a brief evanescent translatorial performative in the play, a single performative moment in the final act that seems to be addressed by the Que¤be¤cois translator to the Que¤be¤cois audience: this is going to push him or her to deny the name of ‘‘translation’’ to the entire French play? It doesn’t seem likely, or particularly fruitful. The relative virtues of (c) should be clear: .

. .

.

like (b), it o¡ers a model of audience response that is internally split and contradictory but lodged in the head of a single viewer, which grounds the interpretation in a single viewing and obviates the necessity of positing widely di¡erent responses in di¡erent times and places; this is also appropriate to a performative perspective in which each drama, each dialogue between utterer and interpreter is uniquely complex and important, and shouldn’t be abstracted and generalized to play a role in a large-scale constative schema involving idealized dualisms (about Scotland vs. about Que¤bec, Que¤be¤cois French vs. Parisian French; nonperformative translation vs. performative nontranslation); also as in (b), that split single response is plausibly human (people do respond to things in contradictory ways); unlike (b), it doesn’t require the Que¤be¤cois theatergoer to imagine a nuance-challenged viewer in the theater that night, or another audience in some other place and time that doesn’t get it; some people do imagine such things, of course, especially intellectuals who take pride in understanding more than ordinary people, so that the problem with (b) is not so much a lack of realism as the conceptual requirement it imposes of imagining a hierarchized dual audience, one right, the other wrong; above all, it provides a plausible and balanced model for the performative linguist who responds complexly in context, and wants to be able to understand complexity without polarizing or hierarchizing: the Macbethas-translated-Shakespeare and the Macbeth-as-original-performative readings simultaneously present alternative interpretations between which the viewer hermeneutically vacillates, generating the possibility that both interpretations are equally valid and thus that the play is right now, in this speci¢c speech act situation, in this dialogue between the producers of the play here tonight (Shakespeare, Garneau, the director, the actors) and its consumers in tonight’s audience, both a translation and a performative original work ^ which is to say, a translatorial performative.

This, it seems to me, solves one of Pym’s problems: the clash between, on the one hand, his insistence inTranslation andTextTransfer that translations be defined in terms of the impossibility of first-person address, along with the corollary of that definition here in this article, that this will allow the translator no direct performative utterances to the audience; and, on the other, Annie Brisset’s persuasive claim that Garneau is (taken to be) saying and doing something local, performing a speech act for the Que¤be¤cois audience. Pym likes the rigor of

Translatorial performatives

51

his definition, and would like to hold onto it; but the Brisset example, if he takes it seriously, as indeed he is inclined to do, seems to undo it. How can he hold onto his definition and still admit the force of Brisset’s argument? He tries to have his cake and eat it by positing a dual audience, but that just divides and conquers him, and he ends by waffling: ‘‘For my present concerns, however, it is of some interest to accept the possibility (although not the necessity) of at least partly translational performatives. This possibility may be written into our working hypothesis in the following way: When a source text has a performative function, the target text can be both at least partly performative in discursive function and properly translational if reception conditions allow the second person of the translational discourse to be participative’’ (53). That last is the giveaway: when is the second person of a real discourse not participative? He needs that proviso because he is imagining the definitional problems that will arise if one side of his dual audience is in Que¤bec taking the play to be (at least partly) about Que¤bec and the other side is in France or somewhere else taking the play to be about Scotland. In that case, which would it be? The root problem through all of this, I suggest, is the persistence in Pym’s emergent performative approach of constative assumptions about stable essences: ‘‘the target text can be both at least partly performative in discursive function and properly translational,’’ he writes, concerned to define the being or essence of the target text through the perceptions of his two audiences. Looking to the power of audience response to solve the problem shows Pym’s performative instincts; but he still wants audience response to determine the nature of the text, rather than simply recognizing that people see texts in different ways. Hence the Garneau translation becomes for him ‘‘properly translational’’ and yet still ‘‘at least partly performative in discursive function.’’ If some member of the audience sees the play as basically about fishing, do we then have to say that the play is ‘‘at least partly about fishing’’? Of course not. Once we jettison the objectivist^constativist insistence on stable definitions of textual meaning or essence, the analytical key becomes not what the text is or is not but how the the text is dialogically construed in specific social situations. The move in (c) to a conflicted response in a single participative second-person Quebecker in the audience addressed by the Garneau translation makes this methodological shift clear: that single audience member trying to decide whether the play in Garneau’s French is a translation or a performative original work can still be seen as trying to apply to that decision the criterion of first-person address, asking himself or herself whether Garneau is addressing him or her as ‘‘I’’ to ‘‘you’’ (making the play a performative original work) or whether the only first- and second-person addresses in the play are assigned to characters. Thus Pym gets to keep his translational criterion of I/you-address in a more explicitly performative framework; all he has lost is the ability (and, hopefully, the desire as well) to determine whether the Garneau Macbeth ‘‘really is’’ or ‘‘really isn’t’’ or ‘‘is at least partly’’ a translation. The determination is made situationally by dialogical participants, from their own local and topical points of view; it cannot be made, nor need it be made,

52

Performatives

by the translation scholar himself in any once-and-for-all constative objectivist^essentialist sense.

Levels of literature But this doesn’t solve all the problems that this example raises. Most importantly, it doesn’t solve the level shifts implicit in Pym’s movement from Macduff ’s performative address to Ross and Malcolm, to Shakespeare’s (possibly performative) address to an English audience, to Garneau’s (again possibly performative) address to a Que¤be¤cois or French theater audience. What is going on there? What status do performatives have when uttered by (a) a character to a character, (b) an author to his or her local audience, and (c) a translator to an audience in the target culture? Pym gives a good deal of thought to the problematic relations between (b) and (c) but skirts entirely around the problems arising between (a) and (b/c). He also neglects to mention that Shakespeare himself has no first-person address in the play: it is not only translators that are denied access to the first person! (Some novelists and playwrights have attempted to create a first-person address for themselves in their texts, creating a character called, for example, The Author, who speaks supposedly in the author’s ‘‘own true’’ voice; but this character is inevitably just another character, not the actual author speaking in his or her own voice.) If translation is to be defined by the impossibility of first-person address, then perhaps Shakespeare’s original English play must be seen as a translation too? And if the impossibility of first-person address makes it impossible for novelists and playwrights to ‘‘do things with words,’’ utter performatives through their works, then do writers never attempt to convey some implicit (say, symbolic or allegorical) message in their work? Literary theorists have sometimes tried to claim that writers never say anything in their works, never make claims about the world indirectly to their readers or audiences over the heads of their characters (Sir Philip Sidney’s ‘‘Now, for the poet, he nothing affirms, and therefore never lieth’’ [Adams 1971/1992: 155] is the most famous); but that is decidedly a minority opinion among theorists, and if it were true (and its consequences could somehow be enforced) it would put an awful lot of literary critics out of business. What is tripping Pym up here, I want to suggest, is precisely the same problem that tripped Austin up, which I’ll discuss in greater detail in Part II: neither of them knows quite what to do with speech acts in the complexly layered discourse of literature. Austin said it clearly: an actor on stage uttering a performative is not performing an action (or at least not the action that one seems to be performing). The actor playing Macduff is not ‘‘remembering’’ Macduff ’s family in saying ‘‘I cannot but remember’’ or ‘‘m’en souv’nir.’’ Nor for that matter was Shakespeare in writing those words in English. The shift in audience uptake involved in moving from the character-to-character level to the source-author-to-source-audience level to the translator-to-target-audience level is far too complicated, and as yet still too mysterious, for us to cut corners and

Translatorial performatives

53

simply assume a kind of rough correspondence across levels. As I mentioned above, because he couldn’t figure out how any of this level-shifting worked, Austin called performatives uttered in a play or a poem ‘‘parasitic’’ or ‘‘nonserious’’ performatives, which, clearly, doesn’t solve the problem, just postpones it. All right, so they are parasitic, so what: what are they? How do they work? And above all, how exactly does it happen that a translator can be seen by Annie Brisset as addressing the Que¤be¤cois audience directly, performing the act of encouraging them too to remember? And what kind of address is it, when a Que¤be¤cois actor on the stage, playing Macduff in French, says ‘‘m’en souv’nir’’? Is it a first-person address to a second-person audience? If so, who ‘‘owns’’ the first-person pronoun? The character, the actor, the author, the translator, or some strange combination of all four? Should we imagine the translator as using the body of the actor and the name of the character and the foreign words of the author as the vehicles of his first-person address? Or should we imagine him walking out on stage and saying ‘‘m’en souv’nir’’ directly to the audience? That seems far-fetched, unless of course we imagine that the actor hired to play Macduff is also hired to do the new Que¤be¤cois translation, and as a kind of avant-garde Que¤be¤cois nationalist move is asked by the director to signal to the theater audience, through body language or signboards or however else such things might conceivably be accomplished, that he is speaking as the translator as well as Macduff. What we need in order to be able to deal adequately with these mindnumbing complexities, the history of speech act theory suggests, is H. Paul Grice’s theory of implicature ^ or specifically, as I am going to be arguing in Part III, a rather more complex theory of implicative uptake based on Grice. We are almost ready for that, in fact, but not yet. First, in Part II, we will need to consider a series of theoretical perspectives that either are, directly, or might fruitfully be taken to be, responses to Austin on parasitic speech acts: Derrida on iterability (after which I will return briefly to the Garneau example, in Chapter 4); Damasio and Robinson on somatics; Simeoni on the translator’s habitus; Bakhtin on double-voicing. Specifically, as my subtitle to Part II suggests, these perspectives will flesh out the ‘‘back-story’’ of the performative: not a speech-act history of linguistics, such as I offered in Chapter 2, but rather a theoretical history of individual speech acts from the standpoint of the subjectivity, both personal and collective, of the language user. How do language users come to feel capable of performing speech acts? How do they come to have any concept of what a speech act is or could be, what doing this or that would signify and what signifying this or that would do? How is linguistic repetition structured in subjective history so as to make both stability and change possible, the stability and change that are both necessary for communication, enough stability or regularity that understanding is possible, enough change or innovation that speech acts have power in novel situations? What stabilizes the subjectivity of speech acts over time, and how are novel elements introduced into it?

54

Performatives

Once we are done with all that, we will be ready for Gricean implicature, in Part III, where we will be discussing among other things the ways in which translators (like all writers and speakers) can do things with words implicitly. In Chapter 13, we will return for the last time to the Garneau example, placing it in the specific context of allusion or ‘‘invocation,’’ invoking shared knowledge that is never explicitly stated.

Part II

Iterations (the performative back-story) If I speak of great stability, it is in order to emphasize that this semantic level is neither originary, nor ahistorical, nor simple, nor self-identical in any of its elements, nor even entirely semantic or significant. Such stabilization is relative, even if it is sometimes so great as to seem immutable and permanent. It is the momentary result of a whole history of relations of force (intra- and extrasemantic, intra- and extradiscursive, intra- and extraliterary or -philosophical, intra- and extraacademic, etc.). In order for this history to have taken place, in its turbulence and in its stases, in order for relations of force, of tensions, or of wars to have taken place, in order for hegemonies to have imposed themselves during a determinate period, there must have been a certain play in all these structures, hence a certain instability or non-self-identity, nontransparency. Jacques Derrida, ‘‘Afterword’’ (1988:145)

4

Iterability

Parasitism J. L. Austin writes, in his second Harvard lecture on the ‘‘Conditions for Happy Performatives’’: Secondly, as utterances our performatives are also heir to certain other kinds of ill which infect all utterances. And these likewise, though again they might be brought into a more general account, we are deliberately at present excluding. I mean, for example, the following: a performative utterance will, for example, be in a peculiar way hollow or void if said by an actor on the stage, or if introduced in a poem, or spoken in a soliloquy. This applies in a similar manner to any and every utterance ^ a sea-change in special circumstances. Language in such circumstances is in special ways ^ intelligibly ^ used not seriously, but in many ways parasitic upon its normal use ^ ways which fall under the doctrine of the etiolations of language. All this we are excluding from consideration. Our performative utterances, felicitous or not, are to be understood as issued in ordinary circumstances. (1962: 21^2) ‘‘Walt Whitman,’’ he adds later,‘‘does not seriously incite the eagle of liberty to soar’’ (104). As we’ve just seen in Chapter 3, this exclusion has serious implications for a performative linguistics of translation. Not only does it become very difficult, as Anthony Pym’s discussion of the Macbeth example that he took from Annie Brisset shows, to determine how to read uptake in a case where there is a clear difference between a character’s reading of another character’s performative and an audience’s reading of a translator’s performative; any performative (taken to be) uttered by a translator would or perhaps should seem to be just as ‘‘parasitic’’ or ‘‘nonserious’’ as one uttered by an actor on stage. But this is much more than a problem for translation theory. Austin’s exclusion of some performatives from his purview (because they are performances that don’t fit his system!) threatens the very performative nature of his analysis.

58

Iterations (the performative back-story)

Insofar as I want to derive performative linguistics from Austin’s book, this exclusion poses grave problems for my case. If Austin does truly believe that a performative utterance is ‘‘hollow or void’’ when performed by an actor on stage, then Jerrold Katz is well within his rights in excluding performance from the performative. The constative revision of Austin was performed by Austin himself. Then John Searle too is right to exclude from his constative study of the promise (which he chooses to call an illocutionary act, dropping the term ‘‘performative’’ from his constative performance altogether) all ‘‘marginal, fringe, and partially defective promises’’:15 In order to give an analysis of the illocutionary act of promising I shall ask what conditions are necessary and sufficient for the act of promising to have been successfully and non-defectively performed in the utterance of a given sentence. . . . If we get such a set of [necessary and sufficient] conditions [for the successful promise] we can extract from them a set of rules for the use of the illocutionary force indicating device. The method here is analogous to discovering the rules of chess by asking oneself what are the necessary and sufficient conditions under which one can be said to have correctly moved a knight or castled or checkmated a player, etc. We are in the position of someone who has learned to play chess without ever having the rules formulated and who wants such a formulation. We learned how to play the game of illocutionary acts, but in general it was done without an explicit formulation of the rules, and the first step in getting such a formulation is to set out the conditions for the performance of a particular illocutionary act. . . . [Note here the analogy with chess, a game with universally stable rules that are maintained by official rule committees and enforced in chess matches everywhere in the world in exactly the same way. If Searle is to justify his constative analysis in a performative speech-act tradition pioneered by Austin, he has to imagine the language game of promising not on the analogy of, say, backyard pick-up games of football or baseball when you don’t have the right equipment or enough space or enough players, which are much closer to how language is actually used, but rather on that of chess, which is not only highly formal and thoroughly rule-governed but policed ^ chess does not simply ‘‘have’’ these rules, the rules are developed and maintained and enforced by committees and judges.16 He also has to assume that ordinary speakers want some set of absolutely stable universal rules for promising and other illocutionary acts.] So described, my enterprise must seem to have a somewhat archaic and period flavor. One of the most important insights of recent work in the philosophy of language is that most nontechnical concepts in ordinary language lack absolutely strict rules. [Note the hedging here: ‘‘most nontechnical concepts in ordinary language lack absolutely strict rules.’’ Does this imply that some nontechnical concepts in ordinary language do have absolutely strict rules? or that some of them only have rather strict rules? The Wittgensteinian position that he is alluding to (see Chapter 2, pp. 28^9) is that concepts in ordinary language never have absolutely strict rules ^ but Searle wants to leave a loophole

Iterability

59

here, some road back through to the kind of hegemonic monolithic idealization that Wittgenstein was challenging.] The concepts of game, or chair, or promise do not have absolutely knockdown necessary and sufficient conditions, such that unless they are satisfied something cannot be a game or a chair or a promise, and given that they are satisfied in a given case that case must be, cannot but be, a game or a chair or a promise. [This in general is the threat to constative stability posed by the nascently performative ‘‘ordinary-language’’ philosophy pioneered in the early 1950s by Wittgenstein at Cambridge and Austin at Oxford as well as others in other places. This is the dangerous trend that Searle must in some principled way thwart or stifle if he is to recuperate Austinian speech-act theory for a constative linguistics. So he continues:] But this insight into the looseness of our concepts, and its attendant jargon of ‘‘family resemblance’’ should not lead us into a rejection of the very enterprise of philosophical analysis; rather the conclusion to be drawn is that certain forms of analysis, especially into necessary and sufficient conditions, are likely to involve (in varying degrees) idealization of the concept analyzed. In the present case, our analysis will be directed at the center of the concept of promising. I am ignoring marginal, fringe, and partially defective promises. This approach has the consequence that counterexamples can be produced of ordinary uses of the word ‘‘promise’’ which do not fit the analysis. Some of these counterexamples I shall discuss. Their existence does not ‘‘refute’’ the analysis, rather they require an explanation of why and how they depart from the paradigm cases of promise making. (1969: 54^5) Yes, ‘‘certain forms of analysis’’ are indeed ‘‘likely to involve idealization of the concept analyzed.’’ This is if anything a fairly substantial understatement, perhaps a form of modesty, since in the West all philosophical analysis always involves idealization. The complexity of any given real-world phenomenon is such that analysis is impossible without some form of conceptual reduction, which is to say, idealization. But there is a world of difference between the idealizing analysis of an Austin, who idealizes ‘‘infelicities’’ along with ‘‘happy performatives,’’ ‘‘marginal, fringe, and partially defective promises’’ along with ‘‘paradigm cases of promise making,’’and the idealizing analysis of John Searle, who simply ignores things that don’t fit his narrowly constative approach. Austin wants an ideal model of human speech acts in general; Searle wants an ideal model of ‘‘the center of the concept of promising.’’ In fact, it would not be too much to argue that Austin wants to model the failure of his own model, so that he can test whether it might work in the real world; indeed when it fails to meet his various real-world tests, he throws it away. Searle wants to protect his model against failure by excluding any real-world variables that might endanger it. Hence his decision to rule out real acts of promising, ‘‘counterexamples’’ that do not ‘‘fit the analysis’’: the act of idealizing the ‘‘centers’’of an act consigns actual speech acts, actual linguistic performances, to the margins of philosophical discourse, the fringes, the periphery, where they cannot intrude upon

60

Iterations (the performative back-story)

what Dell Hymes (1972) has called this ‘‘garden of Eden’’ perspective. Where Katz thematized that perspective in Chomskyan terms, as ‘‘competence,’’ Searle returns to Saussure: ‘‘It may seem,’’ he writes, clearly concerned to protect himself against his constative colleagues’ accusations that he is selling out to the enemy, studying ‘‘speech’’ instead of ‘‘language,’’ the way people actually talk rather than the idealized system of language ^ ‘‘It may seem that my approach is simply, in Saussurian terms, a study of ‘parole’ rather than ‘langue’. I am arguing, however, that an adequate study of speech acts is a study of langue’’ (1969:17). As I say, however, what disturbs me about all this as a performative linguist ^ as someone who wants to claim Austin as predecessor and to denounce Searle as a usurper ^ is that Austin himself manifestly authorizes what I would like to see as Searle’s ‘‘deviation’’ from the true performative form of speech-act philosophy. If Austin too excludes the performance of speech acts on stage, it is only a short step, hardly a step at all, from there to the exclusion of speech acts performed offstage; for who is to say that I am not ‘‘acting,’’ ‘‘performing’’ an inner script in everyday life, whenever, as Erving Goffman (1959/1980) says, I want to dramatize or present myself in certain ways (which is almost always)? Noam Chomsky says this very thing, that everyday speech acts (not his word, of course) are performances of an ideal competence; that is Katz’s rationale for excluding real speech acts from the study of the ‘‘speech act.’’ In this light, John Searle is no usurper at all, any more than Emile Benveniste or Jerrold Katz is. They are ‘‘serious,’’ which is to say legitimate, heirs to Austin’s constative project ^ not parasites; not marginal or fringe cases; not usurpers.

Iterability This might be cause for despair among the performative camp, were it not for the fact that Austin’s constative project is clearly at odds with his (already demonstrated) performative project. Like my hypothetical Quebecker at the end of Chapter 3, Austin seems to be of two minds, here, and it is at least open to question whether he intended to institute a new form of constative linguistics but unfortunately allowed certain inappropriate speech acts to disrupt his discourse (the defendants’ case), or whether he intended to institute a new form of performative linguistics but unfortunately found himself unable to break free of certain hegemonic assumptions inherited from constative linguists (the plaintiff ’s case). In support of my argument that Austin is divided specifically in my (our) favor, that his unfortunate lapse in the matter of ‘‘parasitical’’ speech acts in fact only further corroborates the performative reading of his will, I callJacques Derrida to the stand:17 I would therefore pose the following question: is this general possibility [that performatives may suffer a ‘‘sea-change’’ and become unserious, parasitic] necessarily one of a failure or trap into which language may fall

Iterability

61

or lose itself as in an abyss situated outside of or in front of itself ? What is the status of this parasitism? In other words, does the quality of risk admitted by Austin surround language like a kind of ditch or external place of perdition which speech [la locution] could never hope to leave, but which it can escape by remaining ‘‘at home,’’ by and in itself, in the shelter of its essence or telos? Or, on the contrary, is this risk rather its internal and positive condition of possibility? Is that outside its inside, the very force and law of its emergence? In this last case, what would be meant by an ‘‘ordinary’’ language defined by the exclusion of the very law of language?. . . For, ultimately, isn’t it true that what Austin excludes as anomaly, exception,‘‘non-serious,’’citation (on stage, in a poem, or a soliloquy) is the determined modification of a general citationality ^ or rather, a general iterability ^ without which there would not even be a ‘‘successful’’ performative? So that ^ a paradoxical but unavoidable conclusion ^ a successful performative is necessarily an ‘‘impure’’ performative, to adopt the word advanced later on by Austin when he acknowledges that there is no ‘‘pure’’ performative. (1988:17) Derrida’s purpose in this 1971 deconstruction of Austin entitled ‘‘Signature Event Context’’ was, as he has made amply clear since, to take Austin’s side ^ what I have been calling Austin’s performative side ^ precisely by exposing and analyzing this critical waffling on the matter of performance. Iterability, the concept Derrida develops to get at the complexities of the old and the new in language, is a kind of generalized performability, a trace of past performances and an inclination toward future performances that carries within it the ever-present possibility of ‘‘misuse,’’ misperforming. It is, Derrida argues, precisely because iterability makes misuse (the transformative performance of another person’s utterance) possible that it enables both Searle to idealize Austin on speech acts and Derrida (and me) to deidealize Searle on Austin. Iterability is the reusability of speech acts (or, as Derrida insists, of any signs or marks), and thus both the enabling condition and the limitation of all definition: in order to define something, let’s say a performative speech act, one must reperform it in a new context, the context specifically of definition. This new context both is and is not what constative linguists call the ‘‘null context.’’ It is in the sense that it has been isolated from the context of its earlier use, so that, for example, the performative ‘‘I now pronounce you husband and wife’’ is first uttered in a wedding, then in a book on speech acts. The latter utterance is made in the ‘‘null context’’ in a fairly simplistic mathematical sense: the wedding is one context, and when that ‘‘one’’ is subtracted from the equation, the remainder is zero, or the null context. What the metaconcept18 of iterability does, however, is to infect both (1) the ‘‘original’’ context (the ‘‘one’’) with the possibility of being repeated in an uncontrollable number and variety of contexts (including not only books on speech acts, but lay people playing

62

Iterations (the performative back-story)

around with the wedding ceremony, ministers playfully or seriously marrying monkeys, and so on), and (;) the ‘‘reiterated’’ context (the ‘‘null’’) with the repetition, and thus with a trace of the supposedly subtracted or isolated ‘‘original’’ context. Because the possibility of reuse (and thus what may in some new use need to be defined as a misuse) is present in every ‘‘original’’ or ‘‘ordinary’’ context, a wedding in which the minister says ‘‘I now pronounce you husband and wife’’ is never simply ‘‘one’’ context which, subtracted in a philosophical discussion, would yield the ‘‘null context.’’ It is always a fraction between one and zero, a fractal between binary poles, a fractured context or contextuality. Because the fact of reuse (or misuse) is present in every subtracted or ‘‘abstracted’’ context, it is impossible to subtract contexts down to zero, down to a zero degree of abstract ideality, which eliminates the possibility of ever attaining a null context. Or, as Derrida says in ‘‘Limited Inc abc . . .,’’ his 1977 response toJohn Searle’s attack on his original Austinian deconstruction: Secondly: the iterability of the mark does not leave any of the philosophical oppositions which govern the idealizing abstraction intact (for instance, serious/non-serious, literal/metaphorical or sarcastic, ordinary/parasitical, strict/non-strict, etc.). Iterability blurs a priori the dividing-line that passes between these opposed terms, ‘‘corrupting’’ it if you like, contaminating it parasitically, qua limit. What is re-markable about the mark includes the margin within the mark. The line delineating the margin can therefore never be determined rigorously, it is never pure or simple . . . [But] why (for whom) should this possibility appear as a menace, as a purely ‘‘negative’’ risk, as an ‘‘infelicity’’? Once it is iterable, to be sure, a mark marked with a supposedly ‘‘positive’’ value (‘‘serious,’’ ‘‘literal,’’ etc.) can be mimed, cited, transformed into an ‘‘exercise’’ or into ‘‘literature,’’ even into a ‘‘lie’’ ^ that is, it can be made to carry its other, its ‘‘negative’’double. But iterability is also, by the same token, the condition of the values said to be ‘‘positive.’’ The simple fact is that this condition of possibility is structurally divided or ‘‘differing-deferring’’ [diffe¤rante]. (1988: 70)19 What is interesting and useful in ‘‘Limited Inc abc. . .’’ for my project of naming performative linguistics is first of all the concept of iterability, of course, of which more in a moment; but also the playful work Derrida does with Searle as Austin’s putative heir (1988: 93^4), the source of my courtroom conceit a few paragraphs ago. Derrida is, as I say, concerned to deconstruct Austin in order to recuperate Austin ( from his putative heirs, and from the exclusion of the parasitic that seems to authorize those heirs) for what I am calling performative linguistics. Searle, however, takes him to be attacking Austin and rushes to the master’s defense ^ to the defense specifically of the master as constative linguist, of the Austin who would have theorized the parasitic (constatively) had he lived. Derrida recognizes (in my terms) that Austin is also a constative linguist, and that Searle is partly justified in

Iterability

63

dramatizing himself as Austin’s rightful heir: ‘‘Shall we say that such a lack of interest in the effects of context [in Searle’s work on speech acts] marks a corruption or degeneration of the Austinian heritage, or an alteration of Austin’s intentions, of what he meant to say? No, because what Austin said and did was sufficiently ambiguous, in its iterability, to authorize such an exclusion as well’’ (1988: 78). But then, invoking iterability on a larger scale, he turns the tables on Searle: ‘‘If one were fond of this word [‘corruption’], and of the evaluation it carries with it, one would have to say that corruptibility, too, is part of the heritage and of its legitimation’’ (78). In my terms, Austin’s text in its inner divisions can authorize or legitimize either a constative or a performative heritage; but because he did not (and could not, because no one can) perfectly control the direction his intellectual heritage took after his death, because he found himself so uneasily divided on the issue of the performability (or what Derrida calls the iterability) of speech acts, and because the corruptibility of his heritage is an example of the general iterability of all signs, those who stress performance and performability ^ viz., people like Derrida and me ^ follow his intellectual lead most faithfully. Having said all this, however, let me now begin to dissociate myself just slightly from my witness. The concept of iterability is excitingly productive for a performative linguistics; it opens up a whole new logic of performance that is ‘‘in’’Austin, even if only in his failure to explore the horns of the dilemma across which he found himself stretched. It is in fact so productive that I’m increasingly inclined to think that not just Part II but this entire book might be thematized as a series of reiterations of the theory of iterability ^ though the reiteration is so radical in places that by the end of the book it may no longer be recognizable (which is, in any case, one of the problematics charted by iterability). Iterability is repeatability. Iterability is the repeat-performability of all speech. It is the quality of our speech acts that comes from our having learned them through other people, from other speech acts ^ by learning to emulate what we have seen other people do. It is, as I want to show in Chapter 7, very similar to what Mikhail Bakhtin calls ‘‘polyphony’’ or ‘‘double-voicing’’ or ‘‘internal dialogism,’’ the quality every word we utter has of being saturated with every other dialogue it has ever been used in (certainly, for us as speakers or hearers, every other dialogue we have ever heard it used in, including ones where we have used it ourselves). It is the mutability of language in repetition, difference in the repetition of the same. It is the fact that whenever we repeat something, we change it; whenever we restate something, we reperform it. Iterability is at the heart of the sort of problematic (as well as less obviously problematic) language games one might extrapolate out of Wittgenstein’s lists that I cited in Chapter 2: say, concealing information, telling secrets, doubting the veracity of what someone else is saying, lying, cheating, protecting someone from an unpleasant truth, chiding, teasing, threatening, and so on. Also talking in your sleep, raving in delirium: this too is iterability, because where do the oneiric or delirious speech acts come from except previous speech acts?20 It is at the heart of all imitation: when we mimic a friend, when we retell a story in what we

64

Iterations (the performative back-story)

want to claim is someone else’s voice, when we write a short story or a play, when we act, when we translate. It is, obviously, as I’ll show in a later section of this chapter, an explosively productive idea for the study of translation, because it makes not just translations but all texts and all utterances iterations, and will enable us to trace the similar iterabilities of source and target texts. But iterability is also, at least as Derrida explores it, restrictively and I think counterproductively abstract and structural ^ (anti)logical, and almost, as I hinted in my discussion of it above, (anti)mathematical. Derrida is very little concerned to image the iterable; the most he gives us is lists like ‘‘can be mimed, cited, transformed into an ‘exercise’ or into ‘literature,’ even into a ‘lie’’’ (1988: 70), ‘‘as though literature, theater, deceit, infidelity, hypocrisy, infelicity, parasitism, and the simulation of real life were not part of real life!’’ (90),‘‘the possibility of being repeated, and thus potentially [e¤ventuellement] of being mimed, feigned, cited, played, simulated, parasited, etc.’’ (91^2), and so on. It is possible that this is just, in Kuhn’s (1970) terms, a matter of revolutionary and normal science: Derrida does the revolutionary conceptual groundwork and leaves the working out of examples to his followers. I think in fact that there is more to it than that. Even in Derrida’s own terms, even in terms of the iterable, the effect of listing such stagings or imagings of the iterable is to foreshorten its extension ^ and simultaneously, of course, to extend its foreshortening. The effect of listing is at least dual, but the duality folds back on itself: it at once transforms the items in his lists (miming, citation, exercise, literature, lying, theater, deceit, infidelity, hypocrisy, infelicity, simulation, feigning) into interchangeable instantiations of the same and explodes its differential applicability to infinity, so that anything could (and perhaps should) become an instantiation of iterability. Either way (and both ways) iterability is emptied out into (in)difference ^ a differentiation that is simultaneously an undifferentiation. This doesn’t bother Derrida,21 but it bothers me, largely because I believe performance or performativity or performability can most powerfully transform linguistics not at a high level of philosophical abstraction ^ even an abstraction that complicates the constative as radically as does Derrida’s ^ but at the level of felt experiential details and their social determinants, at the level of idio- and ideosomatized moments in actual lived dramas, performed by the real people present, actors and audience alike. If everything is iteration, if iterability covers the entire field of experience, then maybe we need some smaller concepts as well, some specific images of iterability to work with, some analytical models for working out specific iterations.

Iterability imaged What effects, for example, do institutional practices have on the iterability of print literature as opposed to the theater (and of a play as opposed to a pantomime or an opera, Broadway as opposed to street theater or your children’s school Christmas pageants, the theater of the ‘‘theater’’ as opposed to

Iterability

65

the theater of the law or education or war or whatever, and so on), or of citation as opposed to exercise? Think of the various institutionalized scenes of writing and reading, producing and consuming, marketing and selling, agenting and negotiating, hiring and paying, entertaining and legitimating, convincing and teaching, that shape the iterability of a novel, a play, a scholarly quotation, a pedagogical exercise. What effect does ideological software like ‘‘morality’’ or ‘‘conscience’’ have on the iterability of lying, deceit, infidelity, hypocrisy, simulation, and feigning? What difference for iterability does it make if ideology is understood as operating intellectually, as a political sign-system that governs rational decisions, or as operating somatically, as the felt political unconscious (see Chapters 5^6) that inclines us both to the making of an ideologically correct decision and to the naturalization or rationalization of that decision as ‘‘voluntary’’or ‘‘intentional’’? Will any given iteration ^ let us say, hypocrisy ^ be governed by the ‘‘same’’ law of iterability in people of different classes, genders, races, sexual preferences, educational backgrounds, birth orders, menstrual cycles and other biological rhythms? How is the iterability of feigning, for instance, altered by electrical stimulation of the amygdala or the nucleus accumbens, a hippocampectomy, heavy dosages of antipsychotic drugs, Alzheimer’s disease, anomic aphasia, schizophrenia, hypnosis, or cocaine addiction? What slow ripples does idiosyncratic experience send through the viscosity of iterability? Derrida might reply to all this, simply, that iterability accounts for and contains all variations because it is itself a structure of variability. My questions would then be no more than the working out of Derrida’s global insight at the local level ^ a necessary project, to be sure, but one that does not undermine the globality of the insight. As Derrida himself would insist, the ‘‘structure’’ or the ‘‘law’’ of iterability does always undermine (while also making possible) the globality of any insight; and if, so undermined, in the sense of being globally reperformed and transformed, iterability cannot (and still does) have a‘‘structure’’ or a ‘‘law,’’ so be it. But does this help? As an explanation of change ^ what I’m calling here in Part II the ‘‘performative back-story,’’ how we got to the brink of performance ^ the metaconcept of iterability is structurally equivalent to positing a psychic chaos generator, or attributing all mutability to Satan. In effect it’s a way of saying ‘‘I don’t know why it happens, or how.’’ Maybe my problem is that I’m not enough of a philosopher to rest easy, or even uneasy, in the kind of intellectual Moebius strip that Derrida favors; I want iterability to take me out into the world, out into the texture of performance (or performability), the felt realities, and I am willing to sacrifice quite a lot of philosophical rigor in order to let that happen. To anticipate, then ^ to ‘‘corrupt’’ or ‘‘contaminate’’ iterability with its own iterability by way of mapping a tentative path for the book to follow (and deviate from): we are always acting, in both senses of the word, whether we have memorized our lines from a specific script for a specific play or, more generally, from ‘‘life,’’ from previous speech encounters; and our acting always relies

66

Iterations (the performative back-story)

on that ‘‘script,’’ a socially regulated pattern for our behavior, but in dynamic, situationally contingent ways. We roughly follow the script, but because the script never quite specifies every detail in every scene, we also constantly ad lib ^ and in some sense, because the script is multiple and our memories are bad, we ad lib the script itself. Ad libbing the script itself is iterability, but imaging it forth ^ imagining it, for example, in overtly theatrical terms, as here, or legal ones, as in Part I ^ makes a difference in and for the way we understand the philosophical concept, and thus (because the concept ‘‘exists’’ only in its iterability) in and for the concept ‘‘itself.’’ It matters immensely how we ad lib the script of iterability: ad libbed philosophically, it becomes philosophy; ad libbed theatrically, performatively, it becomes a theatrical performance; ad libbed juridically, it becomes ‘‘the law,’’or courtroom drama; ad libbed naturalistically, as if spontaneously, it becomes a performance in the ongoingness of our everyday lives. There are people who rail against the ‘‘corruption’’of everyday speech by television and movies; who despair at the fact that most people in a postmodern age construct their sentences, and indeed their personalities, around imitations of their favorite sitcoms. Derrida’s concept of iterability simply says that if we aren’t imitating sitcoms we’re imitating something else, some other performance of language, because all language use is the reperforming of past performances. And we never get it ‘‘quite right.’’ We always modify what we perform. We always introduce novel elements into our performances. Performances are always transformations, never pure reproductions. This is a potential cause for despair among constativists, who value purity and stability above all else; hence the need we saw in John Searle to idealize speech acts into the realm of purity and ignore everything impure. For performativists, this ‘‘impurity’’ is bread and butter: it is, in fact, the complexity of the real world.

Translator, iterator The title of this section obviously plays on the infamous Italian aphorism, traduttore, traditore: translator, traitor; translator, traducer. Since treason is a form of iteration, but only one form of iteration,‘‘translator, iterator’’ invokes Derrida in order to take a larger view of translation, one in which treason or the betrayal of the source text is only one thing that translators do. More, since the bitter Italian accusation that translators invariably betray the source text is predicated on the assumption that any change brought about in the transmission of a text is a bad thing, and Derrida’s concept of iterability is based on the assumption that, bad or good, change in the transmission of language is inevitable, the somewhat pompous Latinate aphorism in my title also signals a move to a more tolerant attitude toward translating ^ one that refuses to condemn the work translators do. Consider once more the example that Anthony Pym takes from Annie Brisset. The translator, Michel Garneau, reiterates Macbeth for Quebec: obviously! That is what translators do: they reiterate a text written in one language in another. He reiterates the Shakespeare play specifically in

Iterability

67

Que¤be¤cois French, but this does not in and of itself signal any particular kind of translatorial address to the audience, right? ^ because translators do by definition translate into the target language, the language spoken by their target audience. He localizes, yes, but then in the broadest sense all translation is localization: making a nonlocal text available to a local audience. Suppose, to begin in a fairly conservative place, that Garneau intends no performative address to the Que¤be¤cois audience; suppose that the echo in his words ‘‘a' m’en souv’nir’’of the nationalistic motto on Que¤be¤cois license plates,‘‘je me souviens,’’ never even occurs to him. As far as he is concerned, he is not rewriting the play as a Que¤bec nationalist proclamation; he is simply translating it into Que¤be¤cois French. Suppose further that no one in the audience hears the echo either. The first person to ‘‘hear’’ it is Annie Brisset. What then shall we say is going on? The reason I put it so baldly is that I’m guessing that some such suspicion has occurred to you already, and you may be inclined to reject the example on the grounds that the whole thing is somewhat unreal, even perhaps unrealistic, based not on what actually happened in that theater, let alone what is actually going on in Garneau’s text, or Garneau’s head, but rather on some Que¤be¤cois scholar’s need to find something to say about a translation, some reason to write and publish an article on it. If the Que¤be¤cois audience actually took Garneau’s Macbeth to be a nationalist performative utterance, if Garneau himself intended for it to be taken that way, then maybe we can talk about whether it is still a translation, and correspondingly whether it is possible for a translation to act as a performative utterance. If not, if this is just some scholar’s attempt to be clever, well, then, the whole example is a waste of our time. This is, clearly, a thoroughly constative approach to the issue. The performativity of the translation, if we are going to posit it constatively, must be in the translation. It must be a stable objective feature or property of the translation. It must have been put into the translation by the translator, or at the very least recognized in the translation by its audience. From a constative point of view this latter is somewhat iffy, since audiences can be wrong about such things, but if the whole audience recognized it, if thousands and thousands of Quebeckers took the play to be a Que¤be¤cois nationalist performative, then (we can grudgingly admit) it probably is one. If only one scholar, who needs to come up with ingenious interpretations like this in order to get published, to make a name for herself, etc., ‘‘sees’’ it, it probably isn’t there. In a performative perspective, especially one looking at texts and utterances through Derrida’s concept of iterability, these rigid walls come tumbling down. The ‘‘text’’ as Shakespeare originally wrote it is an iteration of other words, other speech acts. Every reading of it, every performance of it is another such iteration. Every mention of it is an iteration too. And iterability means that one is simultaneously repeating something that went before and adapting that thing to the current speech situation, localizing it. It is impossible to read Macbeth in English without iterating it in all sorts of ways, all of them repetitions of the original iteration, all of them intensely local, intensely situational.22

68

Iterations (the performative back-story)

A francophone Quebecker reading Macbeth in the original English might very well come upon the line ‘‘I cannot help but remember’’ and be unable to avoid remembering the slogan on the local license plates, ‘‘je me souviens.’’ For that reader, in that moment, Shakespeare’s original play becomes a Que¤be¤cois nationalist performative: Shakespeare is talking to me! To us Quebeckers! If we imagine Michel Garneau deliberately making his translation into such a performative, this could very well have been the germ of his idea to do so, this sort of iterative moment in his reading of the original. It might even have been the germ of Annie Brisset’s reading: even before she saw Garneau’s translation she could have iterated Shakespeare’s play lectorially in English, and then, reading the French translation, come across ‘‘a' m’en souv’nir’’ with a thrill of recognition. In this perspective, clearly, it doesn’t matter who the ‘‘original’’ utterer of a perceived or felt or (re)iterated performative is, character or author or translator; the receptor who is suddenly galvanized with the conviction that this line is a performative addressed to me will construct in his or her own imagination an interpretive context that will somehow account for any cognitive dissonance. Macduff is talking to me? No problem: it’s actually Shakespeare talking to me through Macduff, projecting himself into the consciousness of a fictitious Scot, who, as the representative of a country long dominated by the English, might very well be imagined as having something to say to Quebeckers. And so on. In the performative perspective of iterability, intentions flow through speech acts, channeling the past into the present through the localizing acts of iterators. Each such act changes the iterable intentions, but in and through those changes also preserves (something of ) them.

A map of Part II Iterability here in Part II will be only the first of two rather different answers I’ll be considering to Austin’s question about parasitic speech acts. Paul Grice’s theory of conversational implicatures in Part III will be the second, an answer grounded not in what I’m here calling the ‘‘performative back-story’’ of repetition and change, but in cooperation and noncooperation, adherence to and transgression of norms, as the basis for ‘‘performative uptake.’’ Even in Part II, though, I am going to be exfoliating Derrida’s theory quite radically. Most important, I’m going to be emphasizing the temporal dynamic that is everywhere implicit (and often explicit) in Derrida’s discussion, the sense in which iterability is most importantly a historical process. In fact the quotation from Derrida’s ‘‘A fterword’’ to Limited Inc that I’ve set as the epigraph to Part II sums up quite neatly, though still to my mind rather vaguely, the temporal or historical concerns I want to explore. Gerald Graff asks him about a passage in his first book, Of Grammatology, to the effect that there are stabilizing forces in writing and reading, that meaning is never entirely cut adrift, that there are limits and even ‘‘guardrails’’ to semantic undecidability. Derrida replies that language and social experience in general

Iterability

69

are stabilized by numerous social forces, but that that very stability implies and indeed requires the possibility of destabilization. I agree wholeheartedly. What Derrida there calls the ‘‘whole history of relations of force’’ is in fact precisely what I want to consider at some length here in Part II. What I want to insert into that history is a somewhat more detailed depiction of those ‘‘forces.’’ Specifically, I propose to flesh it out with two things I think it is lacking: a social somatic (an embodied channel of transmission, so that we know how it works in human beings, whom Kenneth Burke called ‘‘bodies that learn language’’) and a pragmatic analytic (a set of tools for tracing its operation in specific speech acts, so that we know what it’s good for). I will be rehearsing the somatic theory of language I developed in The Translator’s Turn (Robinson 1991), with support this time from the Portuguese neuroscientist Antonio Damasio (Chapter 5), and the French sociologists Pierre Bourdieu and Daniel Simeoni (Chapter 6). The analytic I will be developing out of the Russian language theorist Mikhail Bakhtin (Chapter 7). My idea throughout will be to explore the ways in which repetition both maintains stability and destabilizes things: how somatic markers make this repetitive retention and rearticulation of ‘‘the same’’ possible in both individual bodies and in whole societies (Robinson, Simeoni); how the dialogical repetition (retransmission and retonalization) of ‘‘voices’’ creates both the stability of linguistic structure and language change (Bakhtin). My goal in Part II is to offer a complex performative explanation of the linguistic ‘‘facts’’ that feed constative linguistics ^ the relative stability of linguistic and cognitive structure ^ while also allowing for the dynamism of change.

5

Somatic markers

Iterability as performability. Iterability as quotability. Iterability as corruptibility. Language is constantly both emerging out of and cycling back into and through a performative social realm in which its ability to ‘‘signify,’’ to serve as a channel of communication, depends on its being performed, and more specifically, according to Derrida, on its capacity for being performed and reperformed, for being quoted, reused ^ which is to say, at least from some perspectives, its potential for being misperformed, misquoted. Nor are these ‘‘negative’’ perspectives necessarily constative: a person who feels her words have been misquoted need not have recourse to constative theories of perfect linguistic stability in order to argue that some writer or commentator got her wrong. Arguing over who’s right and who’s wrong is part and parcel of the performative project too. Even insisting that you are absolutely right is a performative act. It is only when groups of people agree to act as if this sort of insistence could actually make it so that we begin to move into the constative realm: when certain kinds of rhetorically controlled insistence begin to convince enough people that Truth or Fact can somehow be transcendentally stabilized by the right kind of procedural safeguards. As I noted at the end of Chapter 4, I want to do two things to this theory of iterability here in the rest of Part II: I want to suggest a somatic channel of transmission for it, here and in Chapter 6, so that it can be recognized as a realistic description of how humans actually function in the social realm (rather than, say, as mere perverse poststructuralist cleverness); and in Chapter 7 I want to draw on Mikhail Bakhtin’s theory of double-voicing in order to build a model of iterative origins, a kind of dialogical history of (re)iteration(s) based on the repetition of words.

Reading the somatic theory The somatic theory of language was one of the central themes of my first book on translation, TheTranslator’s Turn (1991): that our understanding of language, our memories of language, our use and reuse of language, our language-related choices and decisions are all ‘‘somatically marked.’’ That we have a feeling for words and phrases, registers and styles, either when someone else is speaking or

Somatic markers

71

writing or when we are doing so ourselves, either when we are working in a single language or when we are engineering a transfer from one to another; and that all of our decisions about language, including what word or phrase would be best or what would be most ‘‘equivalent,’’are channeled through these feelings. This seems to me a theory with a great deal of explanatory power.The somatic theory of language can, for example, plausibly account for both the repetition and the change in Derrida’s theory of iterability: both the fact that we use language in ways we have heard it (felt it) used before and the fact that, in reusing it, we invariably change it. It is also the most widely misunderstood set of claims I have ever made. So closely have my various approaches to translation been associated with this somatic theory, in fact, and so simplistically has that theory been understood, that I have grown rather tired of meeting people who say things like ‘‘oh yeah, you’re the guy who says people should translate any way they feel,’’or of reading passing references toTheTranslator’sTurn in other books on translation that reduce that entire argument to the simple-minded assertion that gut feelings invariably rule (or ideally should rule) the translator’s work. Clearly, I didn’t work hard enough in that book to forestall that particular reduction. I kept insisting, all through the book’s opening chapter, that the somatics of language use don’t take the place of reason; they are the necessary ground of reason. It’s not that we have a choice between thinking and feeling and really should allow ourselves to feel more, to base more of our language-related decisions on feelings; it’s that there is no thinking without feeling. Despite my admonitions, an alarming number of readers seem to have taken that book to be saying that translators don’t have to think; they can just feel. Two notable exceptions to this swollen tide of reductionism have been Alan Melby, inThe Possibility of Language (1995), whom I want to discuss briefly here, and Daniel Simeoni, in ‘‘The Pivotal Status of theTranslator’s Habitus’’ (1998), to whom I will be returning in Chapter 6. Melby has for several decades been one of the leading figures in machine-translation research; his book is an eloquent philosophical and ethical reflection on the necessary constraints on his research, the necessary failure of the original dream of machine translation. Like BarHillel in the 1960s, Melby finally came to the conclusion that the fully-automatic high-quality translation of unrestricted texts (FAHQT of UT) would never work: machine translation is only, and will only ever be, feasible if we give up the requirements that it be fully automatic (human-assisted machine translation, HAMT, is currently feasible, and computer-aided translation or CAT is currently a lucrative business), high-quality (scientists needing only the gists of a large body of texts, say dissertation abstracts from around the world, may prefer to rely on extremely rough machine translations), or performed on unrestricted texts (the machine translation of artificially constrained texts in very narrowly defined linguistic domains, like weather reports, has been quite successful). Trying to figure out why this was the case, why machines were unable to process language like human beings, Melby happened uponTheTranslator’s Turn, with its argument that humans process language somatically ^ not exclusively

72

Iterations (the performative back-story)

somatically, again, since various rational and analytical procedures are at play as well, but always also somatically ^ and realized that this is indeed a key difference between human and machine translators: It seems that the agentive approach to language is consistent with the somatic approach of Robinson (1991). There is no transcendent, disembodied meaning in human translation. In machine translation, we create an artificial world within which we can get away with pretending that there is transcendent meaning. For Robinson, as for Lakoff (1987), the human body is the starting point for meaning. The senses are not irrelevant to understanding. George Steiner, the respected translation theorist, once commented that a blind person should be able to produce translations indistinguishable from those of a seeing person. Practicing translators who teach translation have found otherwise. (1995:177) Let me note parenthetically that Lakoff, whose work in cognitive linguistics is thoroughly performative, does not deal with somatics per se, with the somatic marking of behavioral choices by the autonomic nervous system; he is interested in grounding cognitive models in subjectivity, in the way actual human beings experience things, which for him, rightly, always entails embodiment:‘‘Thought is embodied, that is, the structures used to put together our conceptual systems grow out of bodily experience and make sense in terms of it; moreover, the core of our conceptual systems is directly grounded in perception, body movement, and experience of a physical and social character’’ (1987: xiv). I should also note that Melby’s argument is far more complex and wideranging than my selective quoting would suggest; I don’t want to give the impression that his complex understanding of human language is simply the working-out of my somatic theory. But that theory does form one solid plank in his contention that humans and machines process language in radically different ways, and thus that machine translation is only ever going to work in areas where it is practical to ignore those differences, especially narrow linguistic domains like weather reports.

Somatic markers A couple of years after Melby’s book came out, I had a long and fruitful conversation with him at an American Translators Association conference, and in the course of that conversation he directed me to a book that had come out while he was finishing his, and so didn’t appear in his bibliography or his argument: Antonio Damasio’s Descartes’ Error: Emotion, Reason, and the Human Brain (1994). Damasio is M. W. Van Allen Professor of Neurology at the University of Iowa College of Medicine and one of the world’s leading brain scientists; based on decades of research on patients whose emotional centers in the prefrontal cortex had been severely damaged or entirely destroyed, he came

Somatic markers

73

to the conclusion, presented in his book, that emotion is absolutely essential for reason. More specifically, he argued that reason ^ in the specific sense in which ‘‘the purpose of reasoning is deciding and . . . the essence of deciding is selecting a response option, that is, choosing a nonverbal action, a word, a sentence, or some combination thereof, among the many possible at the moment, in connection with a given situation’’ (Damasio 1994: 165) ^ is only made possible by ‘‘somatic markers’’ (his term). His research-based conclusions so strikingly confirm my own from three years earlier, based not on research but on introspection, that I want to present them in some detail here.23 Damasio’s research begins with the puzzling behavior of patients with a specific kind of damage to the prefrontal cortex: they can reason, they pass all reasoning and decision-making tests with flying colors, but they can’t hold down a job. Before the accident or the stroke or the brain surgery that changed their lives, they were highly competent professionals making hundreds of responsible decisions a day; after they returned to work following the accident or stroke or surgery, their behavior was irreparably changed, their competence destroyed. They would spend too much time on peripheral details; they had no sense of proportion, or of priorities. Their bosses would ask them to put aside what they were doing and handle an urgent matter, and they would agree to do so, but would not adjust their priorities; two hours later their boss would return for the rush job and they would not even have started in on it yet. They could not hierarchize. Yet there seemed to be nothing wrong with them. So ‘‘normal’’ did they seem otherwise, in fact, that it was almost impossible for them to obtain disability benefits; their bosses and the doctors they went to all concluded that they were just lazy, that they were shamming. Gradually, Damasio and his team pinpointed the problem: these people with prefrontal brain damage were no longer able to feel normally. They were pleasant and intelligent and seemed perfectly normal, if perhaps a bit ‘‘contained’’ emotionally; but that more or less ‘‘normal’’ front concealed a deeper change in their brain function that, in fact, involved an inability to use emotion to ‘‘manage’’ or channel rational processes. Sometimes, in fact, this ‘‘emotional cutout’’ was beneficial: Not too long ago, one of our patients with ventromedial prefrontal damage was visiting the laboratory on a cold winter day. Freezing rain had fallen, the roads were icy, and the driving had been hazardous. I had been concerned with the situation and I asked the patient, who had been driving himself, about the ride, about how difficult it had been. His answer was prompt and dispassionate: It had been fine, no different from the usual, except that it had called for some attention to the proper procedures for driving on ice. The patient then went on to outline some of the procedures and to describe how he had seen cars and trucks skidding off the roadway because they were not following these proper, rational procedures. He even had a particular case in point, that of a woman driving ahead of him who had entered a patch of ice, skidded, and rather

74

Iterations (the performative back-story) than gently pulling away from the tailspin, had panicked, hit the brakes, and gone zooming into a ditch. One instant later, apparently unperturbed by this hair-raising scene, my patient crossed the ice patch and drove calmly and surely ahead. He told me all this with the same tranquillity with which he obviously had experienced the incident. (193)

This patient’s lack of emotional response probably saved him from the same ditch, possibly from serious bodily harm. As Damasio remarks, ‘‘Most of us would have had to use a deliberate overriding decision to stop us from hitting the brakes, out of panic or out of sheer feeling for the unfortunate driver in front of us’’ (193). But this patient felt none of that. He was the perfectly rational agent imagined by the classic rationalist tradition in philosophy, and in fact often posited by driver’s manuals as well. No emotion intruded on what he rationally knew about driving on slippery roads. His unperturbed rationality may even have saved his life. But then Damasio tells another story about this same patient on the next day: I was discussing with [him] when his next visit to the laboratory should take place. I suggested two alternative dates, both in the coming month and just a few days apart from each other. The patient pulled out his appointment book and began consulting the calendar. The behavior that ensued, which was witnessed by several investigators, was remarkable. For the better part of a half-hour, the patient enumerated reasons for and against each of the two dates: previous engagements, proximity to other engagements, possible meteorological conditions, virtually anything that one could reasonably think about concerning a simple date. Just as calmly as he had driven over the ice, and recounted that episode, he was now walking us through a tiresome cost-benefit analysis, an endless outlining and fruitless comparison of options and possible consequences. It took enormous discipline to listen to all of this without pounding on the table and telling him to stop, but we finally did tell him, quietly, that he should come on the second of the alternative dates. His response was equally calm and prompt. He simply said,‘‘That’s fine.’’ Back the appointment book went into his pocket, and then he was off. (193^4) ‘‘This behavior,’’ Damasio comments, ‘‘is a good example of the limits of pure reason. It is also a good example of the calamitous consequence of not having automated mechanisms of decision making’’ (194). And this is, in fact, the heart of his theory: that our emotions help us sort through the millions of minute details that ‘‘pure reason’’ would have to work with; that our emotions help us hierarchize and prioritize those details by ‘‘marking’’ or ‘‘biasing’’ them somatically. This means that we do not need to deal with complexities that our

Somatic markers

75

brains are not equipped to handle all at once. To illustrate this, Damasio pictures a situation in which you are the owner of a business and must decide whether to meet with a potential client who will bring your company much-needed income but who is also your best friend’s worst enemy, and notes that the unraveling complexity of that decision is just too much for pure reason to handle. ‘‘At best,’’ he writes, without somatic markers your decision will take an inordinately long time, far more than acceptable if you are to get anything else done that day. At worst, you may not even end up with a decision at all because you will be lost in the byways of your calculation. Why? Because it will not be easy to hold in memory the many ledgers of losses and gains that you need to consult for your comparisons. The representations of intermediate steps, which you have put on hold and now need to inspect in order to translate them in whatever symbolic form required to proceed with your logical inferences, are simply going to vanish from your memory slate. You will lose track. Attention and working memory have a limited capacity. In the end, if purely rational calculation is how your mind normally operates, you might choose incorrectly and live to regret the error, or simply give up trying, in frustration. (172) Damasio argues that ‘‘the cool strategy advocated by Kant, among others, has far more to do with the way patients with prefrontal damage go about deciding than with how normals usually operate’’ (172). How do normals operate? Here is Damasio’s general statement of his ‘‘somatic-marker hypothesis,’’ which basically spells out the same somatic theory I offered in TheTranslator’sTurn: Consider again the scenarios I outlined. The key components unfold in our minds instantly, sketchily, and virtually simultaneously, too fast for the details to be clearly defined. But now, imagine that before you apply any kind of cost/benefit analysis to the premises, and before you reason toward the solution of the problem, something quite important happens: When the bad outcome connected with a given response option comes into mind, however fleetingly, you experience an unpleasant gut feeling. Because the feeling is about the body, I gave the phenomenon the technical term somatic state (‘‘soma’’ is Greek for body); and because it ‘‘marks’’ an image, I called it a marker. Note again that I use somatic in the most general sense (that which pertains to the body) and I include both visceral and nonvisceral sensation when I refer to somatic markers. What does the somatic marker achieve? It forces attention on the negative outcome to which a given action may lead, and functions as an automated alarm signal which says: Beware of danger ahead if you choose the option which leads to this outcome. The signal may lead you to reject, immediately, the negative course of action and thus make you choose

76

Iterations (the performative back-story) among other alternatives. The automated signal protects you against future losses, without further ado, and then allows you to choose from among fewer alternatives. There is still room for using a cost/benefit analysis and proper deductive competence, but only after the automated step drastically reduces the number of options. Somatic markers may not be sufficient for normal human decision-making since a subsequent process of reasoning and final selection will still take place in many though not all instances. Somatic markers probably increase the accuracy and efficiency of the decision process. Their absence reduces them. This distinction is important and can easily be missed. The hypothesis does not concern the reasoning steps which follow the action of the somatic marker. In short, somatic markers are a special instance of feelings generated from secondary emotions. Those emotions and feelings have been connected, by learning, to predicted future outcomes of certain scenarios. When a somatic marker is juxtaposed to a particular future outcome the combination functions as an alarm bell. When a positive somatic marker is juxtaposed instead, it becomes a beacon of incentive. (173^4)

These somatic markers are not innate but learned, Damasio says. This is not ‘‘mystical biologism,’’ Lawrence Venuti’s (1992: 323) derogation of the somatic theory in The Translator’s Turn, but a neurological explanation of a social^ psychological process that may, in fact, be the primary channel of the collective (‘‘ideological’’) regulation of individuals by societies. It rests on a ‘‘hard-wired’’ neural system, yes: ‘‘The neural basis for the internal preference system,’’ Damasio writes, ‘‘consists of mostly innate regulatory dispositions, posed to ensure survival of the organism’’ (179), typically through‘‘the ultimate reduction of unpleasant body states and the attaining of homeostatic ones.. . . The internal preference system is inherently biased to avoid pain, seek potential pleasure, and is probably pretuned for achieving these goals in social situations’’ (179). But the specific somatic markers that individuals develop are not innate. They are conditioned into us through social experience, experience that is both unique to the individual and thus idiosyncratic (what I called in The Translator’s Turn ‘‘idiosomatic’’) and highly collectivized and regulatory (what I called ‘‘ideosomatic’’). Members of groups (families, genders, social classes, national cultures, etc.) respond in similar ways to similar situations because those groups tend to ‘‘train’’ the bodies of their members to mark stimuli in similar ways ^ the source of ‘‘ideosomatics’’ ^ but this training is never monolithic. Each individual belongs to many different groups, and the overlaps between them will be different for different individuals; and every individual has many social experiences that no one else has, indeed that may from a collective point of view be entirely ‘‘random,’’ so that the somatic markers associated with those experiences will be strikingly different from those of any other members of the various groups to which a given individual belongs. This latter, again, would be what I called ‘‘idiosomatics.’’

Somatic markers

77

Every use of language is both collectively controlled, or ideosomatic, and thus a form of repetition, and individually divergent or creative, idiosomatic, and thus a vehicle of change: the somatics of language, then, as the embodied channel of iterability.

Propositions I want to run through the somatics of language twice, in this chapter and the next. Here in the remainder of this chapter, let me summarize my contentions about the somatics of language use from TheTranslator’s Turn systematically, in eight numbered paragraphs. In Chapter 6, I will move on to a full-fledged somatic sociology, through the work of Daniel Simeoni on the translator’s habitus. 1. Our sense of how to use language in real-world contexts (both speech and writing) is marked somatically. We feel that this word or construction or usage is better than that, more appropriate than that, more accurate than that, more equivalent to its counterpart in an original text than that. 2. These somatic markings not only invariably accompany logical or systematic analyses of language (except in patients with prefrontal damage and sociopaths); they also experientially precede those analyses, and may sometimes even preempt or preclude them. This frequently makes our sense of rightness or wrongness about our own language use emotionally strong but intellectually weak. When a foreigner or a child asks us whether some vaguely strange or unidiomatic string of words is the ‘‘right’’ or ‘‘best’’ way to say something, for example, we may have a strong sense that there is something wrong with it but have no idea what, or even whether it should be considered ‘‘incorrect’’ or just ‘‘awkward;’’ or we may have a strong sense that it is wrong but not be able to invent a rule that it is breaking, or even to figure out what makes it wrong, or to come up with a ‘‘better’’ or ‘‘more correct’’ way of saying it. This is even true of many highly intellectual and analytical language users who have not been specifically trained in constative linguistics. Many literature professors, for example, trained in the critical and articulate reading of difficult literary texts, hate to teach students how to write because they find it so difficult to pinpoint just what is wrong with a student paper and to suggest how the student might go about making it better.24 3. Our language use is somatically marked for ‘‘correctness’’and ‘‘incorrectness,’’ ‘‘appropriateness’’and ‘‘inappropriateness,’’ ‘‘effectiveness’’and ‘‘ineffectiveness,’’ ‘‘high’’ and ‘‘low’’ and various ‘‘middle’’ styles,‘‘eloquence’’ and ‘‘crudeness,’’ and so on through a long process beginning at birth and continuing throughout our lives; this is the process by which language use is collectively regulated, so that members of the same language community ‘‘mean’’ roughly the same things by the same words and phrases, and make more or less reliable guesses as to what other people are trying to say.25 This regulation controls (brings into rough mutual conformity) what I will be calling in Chapter 11 the ‘‘immediate intendant’’and the ‘‘immediate interpretant’’of an utterance. It is also the source

78

Iterations (the performative back-story)

of the ‘‘maxims’’ that we will see Paul Grice theorizing in Chapter 9 ^ even if, as I will be showing in later chapters (10 and 11), utterers and interpreters ‘‘invoke’’ those maxims primarily unconsciously, somatically, and only construct conscious analytical articulations of them secondarily, if pushed hard enough (say, by a Gricean linguist teaching your class). 4. This collective regulation of language use requires that we be trained (in childhood and throughout our lives, in mostly unconscious ways) to mark usages generally considered ‘‘correct’’ or ‘‘acceptable’’ with positive somatic responses, and those generally considered ‘‘incorrect’’ or ‘‘unacceptable’’ with negative ones. The markers are typically so mild as to be almost imperceptible to the person experiencing them; they most commonly occur completely below the level of conscious awareness (they can be picked up by instruments like a polygraph machine that detects tiny changes in the conductance of electrical current on the skin). The person experiencing somatic markers may, in fact, feel positive markers primarily as an absence of negative markers, as when we scan through a well-written piece of student writing nodding ‘‘yes, yes, this is all fine, no problems here.’’ Slightly stronger positive somatic markers may include a general sense of well-being, comfort, friendliness, openness, receptiveness, and so on. Even stronger positive markers may move us to laughter, or love, or tears (of happiness), or outrage (on behalf of a victim with whom we are encouraged to identify), or any number of other powerful emotions. The impact on listeners or readers of a brilliant poem or persuasive speech or sermon or insightful psychotherapy session may attest to this sort of positive somatic marking: the positive somatic markers attached to some forms of speech or writing can make people fall in love with the utterer, or change their lives in some radical way. Negative somatic markers include various forms of ‘‘correctness anxiety,’’ the unpleasant feeling we get when someone around us uses language ‘‘incorrectly’’ (foreigner talk, lower-class talk, student writing, mistranslation, etc.) and we feel an urge to correct it or distance ourselves from it. Anxiety markers include such unpleasant autonomic responses as constrictions of the throat or chest, sweaty or clammy palms, a churning in the stomach, indigestion, a taste of bile in the mouth, etc. This does not mean, of course, that whenever we hear language we consider incorrect we go into seizures, or burst into tears (a reductio ad absurdum that I have frequently heard from critics of the somatic theory: somatic markers typically aren’t dramatic, so they don’t exist), or otherwise even become aware of our unpleasant somatic responses. Most language purists who feel constantly obliged to correct other people’s ‘‘bad grammar’’ and other usage divagations are typically able to sustain the illusion that their need to correct usage is purely intellectual, mental ^ they feel, and feel compelled by, negative somatic markers, but have learned so successfully to ignore somatic signals that the behavior that to other people seems so ‘‘kneejerk’’ (physically conditioned) seems to them rigorously logical.26 5. This somatic marking of language use does not mean either that we never reason logically about language; we do.27 Indeed the problems we often have reasoning logically can themselves be traced back to the operation of

Somatic markers

79

somatic markers. Somatic markers are learned, experience-based: they are part of ‘‘what we know,’’ what we have learned in the course of living our lives. This is not mysticism or biologism. Because they are learned, however ^ because they are conditioned by a variety of real-world experiences ^ they often conflict. For example, grammarians trained in the ideosomatics of logic (ideologically conditioned by the Western intellectual tradition since Plato and Aristotle, and especially since Augustine and Aquinas, to believe that logic is the underlying form of all reality) may feel inclined to oversystematize the grammar of a language; but to the extent that they speak that language effectively (especially if they are native speakers of that language), they will also have been trained in the highly complex situational ideosomatics of how the language actually works, so that ‘‘the language’’ may haul them back from the transgrammatical brink of overgeneralization through the operation of somatic markers. It will feel wrong to overgeneralize. They will be reading along in their own draft and start getting a bad feeling, a sense that something’s wrong here; this somatic marker will keep pushing them until they reason through the problem and realize that they are overgeneralizing. This also means that in reasoning logically about grammatical structure they will feel a constant somatic tension between the conflicting pulls of rules and exceptions, ideal structures and contextual slippages, regulatory and creative impulses, standard and regional dialects, sociolects and idiolects, and so on. The debate between prescriptive and descriptive grammar could in fact be traced back to conflicting somatic markers: prescriptivism is driven by positive somatic markers supporting a certain sociolect (associated with a given social class and/or race and/or education level and/or historical period ^ typically, I would hazard a guess, the period of the prescriptivist’s late adolescence or early adulthood) and ringed around with negative somatic markers (correctness anxiety) undermining other sociolects; descriptivism is driven by positive somatic markers supporting tolerance for linguistic diversity as experienced in a wide range of social groups and negative somatic markers directed against prescriptivists (supporting images of them as crabby narrow-minded people who wield language rules like weapons). Again, this does not mean that prescriptive and descriptive grammar can or should be reduced to somatic markers: both obviously involve a great deal of systematic logical reasoning as well. But they are never purely intellectual. 6. The somatic marking of language is never completely collective. Our sense of the language(s) we use is somatically marked by experience; that experience is largely but never perfectly shaped and regulated by collective forces. Our specific individual experiences with specific individual cats and dogs and trees and rocks, let alone the vast array of human, social, cultural phenomena, will charge our sense of the words and phrases for those things with a slightly but significantly divergent idiosomatic current. Hence the possibility (and frequency) of misunderstandings, and of arguments over the meanings or extensions or connotations of words, or the interpretation of poems. Hence also the fact of linguistic change: the volatility of the somatic

80

Iterations (the performative back-story)

marking of language use ensures that language as it is used in a given community is always in flux, always surging and churning, always generating and selecting from among new expressive potentials. (For two different performative schematizations of language change, see Chapters 7 (Bakhtin), pp. 115^19, and 12 (Grice/Peirce). 7. The somatic marking of language increases the speed and economy of actual language use. We could not run through all the (logical) grammatical rules we have ever memorized when we speak; there would not be time. Learners of foreign languages, taught the ‘‘rules’’ in complex abstract ways, typically find it extremely difficult to start speaking it, or to understand what native speakers are saying. You can’t mentally review all the complex rules for the usage and formation of the Spanish subjunctive, for example, and still choose between *‘‘espero que viene’’and ‘‘espero que venga’’ (I hope s/he comes) fast enough to utter the sentence. You have to have a single rule of thumb in order to say anything; and eventually you have to develop something like a native feel for when the subjunctive is used. Even when we write, with plenty of time to think and check and revise, it would be far too great a strain on our attention and working memory to check every usage against some conscious internal (let alone external) grammar or lexis or pragmatic rulebook. Actual language use, as Daniel Dennett (1991:145) writes, is ‘‘ballistic’’: we are constantly uttering and interpreting language out at the leading edge of our reasoning ability, so rapidly that we have only the vaguest idea of what we are planning to say or write, and, while someone else is speaking, or when we are reading someone else’s words, of what they are trying to say to us. This speed is only possible due to the processing shortcuts produced for us by somatic markers. 8. A somatics of language helps us situate language use firmly in the realm of performance. As I wrote inTheTranslator’sTurn: Part of learning a language well is watching what native speakers’ bodies do when they speak it: how they move their mouths, how they gesture and shift their weight, how they stumble over words, where and how they pause, how they use stress for emphasis ^ in general, how they stage their speech. . . . You have to sensitize yourself to body signals, to project yourself imaginatively into the other person’s body. In a sense, the foreign-language learner has to be at once a playwright, a director, and an actor, in the foreign language: he or she must generate out of fleeting impressions living, breathing images of native speakers inside his or her own body, create them as vehicles for identification ^ and then become them, grow into them, body them forth. (1991:16^17) And, of course, we do this not only while learning foreign languages, but while learning all language, including our first or ‘‘native’’ language(s). Small children learning to speak are learning how to stage their thoughts in speech, in their bodies, both outwardly (using tone and stress and pitch, facial expressions

Somatic markers

81

and gestures and body position) and inwardly (feeling something of what other people feel when they use certain words and phrases in certain ways, marking those words and phrases somatically). Throughout our lives we continue to learn to stage and restage language, to perform it in and with and through our bodies, registering small changes in other people’s somatic performances, judging from those changes when and how the meaning and usage of words and phrases are changing, trying out those changes in our own speech and writing, inflecting those performances with our own personalities, our own experiences, our own needs and expectations. (This is what Jacques Derrida means by iterability and, as we’ll see in Chapter 7, what Mikhail Bakhtin means by double-voicing.) A familiar word is tonalized differently: it was given a ‘‘straight’’ or serious tonalization before, and now is increasingly tonalized humorously, ironically, for purposes of ridicule. One person does this, ironizes the word, gives it a camp performance, and you think of it as an expression of this person’s general idiosomatic take on things, maybe even of his or her current mood. Someone else does it, someone who cannot have heard it from the first person; then another someone, and soon you note a trend. At some point in this change sequence you start experimenting with the new tonalization too. You feel it yourself, first; then use it that way in conversation with friends, externalizing your idiosomatic response, helping to push it toward the ideosomatic. You help the language community reperform the word, revise its ideosomatic charge. These somatic observations about language seem counterintuitive to constative linguists, because they have worked so hard to purge language use of use, of embodied performance; and so influential have these constative linguists been in our civilization, say since St. Augustine, that many people who would not consider themselves (constative) linguists respond to the somatic theory with profound skepticism.They don’t feel any somatic responses to words; therefore this whole theory must be rubbish. The people who know most about the somatics of language are theater people, actors and directors, and the performance theorists who base their theorizing on the theater: they know that all language is performed with the body, with tone and stress and pitch and facial expression and gesture and movement and position on the outside, with feeling on the inside ^ and as Damasio says (1994: 143), e-motion is simply the moving-outward of feelings, the ex-pression of what is inwardly felt. Bad actors say words without feeling, or without feelings that audiences find (i.e., somatize as) ‘‘persuasive’’ or appropriate for the words they are saying in the context they are saying them in. And bad actors are not to be found only on the stage. ‘‘Real life’’ is full of them too.

6

The translator’s habitus

Language, then, is performed in and through and by the body. We make decisions about what word or phrase to use with the help of somatic markers. We remember what words and phrases mean, not just their denotations but their connotations and collocations and implications, with the help of somatic markers. Communication is made possible through the collective regulation of language that somatic marking achieves: members of a language community continue to use words and phrases in roughly the same way because language users are trained to mark them somatically in roughly the same way. This is ideosomatics: ideology as somatically stored and channeled iterative repetition. Language changes because speakers and listeners only learn to mark those words and phrases in roughly the same ways. There are always individual divergences. This is idiosomatics: individual creativity as somatically experienced and expressed iterative difference. But how? Chapter 5 was about the mechanism of somatic marking in general. Now we need to look at the actual transmission of somatic markers from person to person: to begin to shift the somatic theory from the individual to the group, specifically through a look at Daniel Simeoni’s application of Pierre Bourdieu’s habitus theory to translation.

Norm theory Simeoni (1998) begins his discussion of the translator’s habitus with a critical review of Gideon Toury’s descriptive ‘‘norm-theoretical’’ approach to translation, specifically in Descriptive Translation Studies and beyond (Toury 1995), according to which the process of becoming a professional translator involves the internalization of social and professional norms: In my reading of the tentative model described in DTS & beyond, translator interaction with the ‘‘environment’’ (p. 248) takes the shape of a series of incremental stages, starting with initiation and ending up with the recognition by peers and others of his/her full-fledged competence (pp. 249ff.). Thus do learners become trained, step by step, in the skills of the profession. We are led to visualize a slow process of inculcation,

The translator’s habitus

83

emphasizing the translator’s gradual relief from the shackles of external pressures as s/he internalizes normative behaviour ever more deeply in his or her practice. This interplay of constraints on the activity of translating is valuable, for it gives us a sense of the kind of limitations to be expected from a purely mentalistic approach to translating. Those constraints are of course social (‘‘environmental circumstances’’ in Toury’s words, p. 246). Only by becoming internalized do they give an impression of being part of the mental apparatus of the translator. (Simeoni 1998: 5) ‘‘Toury is adamant,’’ Simeoni notes on the next page (6), ‘‘that the horizon of the successful translator heralds near-complete submission to the norms effective in the subsector(s) of society in which s/he is professionally active. . . . Norms have the upper hand. Translators adhere to them more often than not.’’ In Simeoni’s reading, then, Toury’s norm theory is modeled on something like Nietzsche’s internalization of mastery, or what Althusser (1971) calls ‘‘subjectification’’: there are forces in society, collective, regulatory forces, which Toury calls ‘‘norms,’’ that constrain translators’ professional decisions externally at first, but that are gradually internalized until each translator internally replicates the hegemonic or normative regimen prescribed by society (or some relevant segment of society) for his or her profession. This internalization model is quite obviously close enough to the ideosomatic model I presented in Chapter 5 to warrant a more detailed comparison. In the ideosomatics of language, the language learners are corrected by external authorities until they begin to internalize those corrections somatically, to ‘‘somatize’’ them; once the corrections or linguistic norms are somatized, they operate, through what Damasio calls somatic markers, inside the learners like the individualized representatives of collective authority, guiding them increasingly reliably to usages widely accepted as correct in the language community to which they belong. Those norms are at first external regulatory forces constraining the individual speaker’s speech behavior, so that we might imagine the situation as one agent forcing another agent to submit to its authority; but gradually, as they are internalized, the norms become internal somatic regulatory agents ‘‘wielding’’ the speaker from within, to the end of generating collectively acceptable speech acts. It should be clear by now how this model can be extended to explain how norms for rational behavior are inculcated in large populations, or how people learn to interpret the speech acts of others through a similar ideosomatization process (see Chapter 12), or how, as in Toury, novice translators become professionals through ‘‘near-complete submission to the norms effective in the subsector(s) of society in which [they are] professionally active.’’ In other words, the theory of ideosomatics contains a theory of social norms, and of the inculcation of those norms in individuals. More specifically, TheTranslator’sTurn contained a norm theory of translation and of the translator’s

84

Iterations (the performative back-story)

training (see esp. Robinson 1991: 109^17). Since norm theory became in the 1990s one of the most ‘‘serious’’ and highly respected branches of translation studies, to say that TheTranslator’s Turn contains a norm theory is essentially to claim for it a critical disciplinary significance. As Simeoni shows, however, norm theory as propounded by Gideon Toury and his many followers in the DTS camp lacks four key elements, three of which are supplied by somatic theory (and incorporated from somatic theory into habitus theory by Simeoni), the fourth of which is supplied only by habitus theory ^ which is why I want to take the time to unfold the latter here. Habitus theory offers a signal expansion of somatic theory. 1. The first missing element in norm theory is an account of how norms are perceived, grasped, and learned by the individual. Simeoni calls Toury’s norm theory ‘‘behaviorist,’’ but I see little evidence inToury’s writing that he has even this well-developed a social psychology. Certainly he had nothing like a Skinnerian model of afferent conditioning in Descriptive Translation Studies and beyond (Toury 1995), the book Simeoni is explicitly critiquing. By his more recent contribution to the Aston conference on norms (Toury 1998), he had done some reading in social anthropology, specifically an article entitled ‘‘Social Creativity’’ by one J. Davis (1994), who offers him something like a social-psychology, but one that is strikingly lacking in social or psychological specificity: As J. Davis ^ an anthropologist who tried to systematize the notion of ‘‘social creativity’’ and render it serviceable in explaining the making and maintenance of social groups ^ recently put it: People use their given sociability to create agreements about actions. So, our worlds achieve the appearance of stability and regularity because we agree that certain actions are acceptable in appropriate circumstances, and others are not. (Davis 1994: 97, italics added) Rather than given, (tacit or explicit) ‘‘agreements about actions’’ are always negotiated, with or without the intervention of language. Such negotiations, which require some time, result in the establishment of social conventions, according to which members of the group will behave when they find themselves under particular circumstances. They often do so in the form of behavioural routines.‘‘What we create is ^ within agreed limits ^ a predictable event, from which certain choices have been excluded’’. ‘‘So when we are [socially] creative we attempt to create order and predictability and to eliminate choice, or at any rate to confine choice within certain prescribed limits’’ (Davis,1994: 97). (Toury 1995:14^15) Not only can social negotiations be carried out with or without the intervention of language (paragraph 3.1), but also the norms themselves

The translator’s habitus

85

which would govern behaviour need not be formulated at all: They may well remain implicit. (16) In an established group, norms are basically acquired by the individual ^ a newcomer to the group on whatever grounds ^ in the process of his/her socialisation. . . . Very often, these norms ^ or even the basic agreements and conventions ^ go on being negotiated throughout one’s entire life in the group, for instance, when members struggle to establish their own positions within the group (or vis-a-vis its other members). (17) And so on. Toury’s discussion obviously relies heavily on the negotiation metaphor of rationalist social theory (for discussion, see Pym 1998: 112); the idea is, at least figuratively, that members of a group sit around and discuss how they are going to regulate behavior in the group, negotiate until they reach some sort of collectively acceptable agreement. That this is only a metaphor, that groups don’t usually engage in such negotiation behavior, Toury too knows full well: the negotiations, he says, need not be carried out in language. Nor should we imagine the negotiations proceeding by means of hand signs or some such. The words ‘‘negotiation’’ and ‘‘agreement’’ are figures for something else; but what that something else is, Toury doesn’t tell us, presumably because Davis doesn’t either, which probably means that Toury has no idea (maybe in fact Davis is as much in the dark himself ). Something happens in groups; norms, conventions, routines, and so on are somehow collectively formulated and passed on to every member of the group. But how? What is the tenor of this metaphor? How are norms developed, transmitted, internalized? The somatic theory was my attempt to answer this question: through somatic ‘‘contagion’’ we internalize group norms in the form of ‘‘somatic markers,’’ autonomic responses that ‘‘bias’’ our decision-making processes toward ‘‘good’’ choices (marked as ‘‘desirable’’ or somatically pleasurable by social hegemony) and away from ‘‘bad’’ choices (marked as ‘‘undesirable’’ or somatically unpleasant). Because of the felt strength of these somatic markers, when they are activated we not only feel considerable somatic pressure to conform, to make whatever decision it is we’re making in accordance with group norms; we also feel compelled to put pressure on others to conform as well. This explains where norms come from, how they are transmitted throughout a social group, and how they act on each member of the group to enforce normative behavior. Simeoni recognizes the explanatory power of the somatic model as well; while nowhere insisting that habitus theory is intrinsically somatic, he too, like Alan Melby, has read TheTranslator’s Turn and been persuaded that somatic channeling does usefully account for our experience of social norms, or

86

Iterations (the performative back-story)

‘‘field structures’’: Field structures in general, whether loose or rigid, complex or simple, gain in comprehension as soon as they are perceived as felt structures. Structure as a nexus of social norms is embodied and customarily somatized in daily routines. Try opposing a structure (external or internal); it will resist. This is what makes the social world as experienced a vector of common violence either suffered, accepted, channeled or under certain circumstances, obviously enjoyed. In individual situations, an easing of the tension felt while translating may be construed as evidence for a greater availability of internalized options on the scale of one’s native and acquired skills and a concomitant decrease in felt pressure. An elevation of the tension on the other hand might be testimony to stronger pressure to comply with external ^ i.e. not (yet?) internalized norms ^ a situation accompanied by a feeling of relative unnaturalness. This alternation of tension and relative ease is a common experience in translating, as is the feeling of gradual easing as the task progresses ^ until it turns into a new routine. . . . Too much routine of course is likely to reactivate the pain, as aptly noted by Douglas Robinson (1991). An area of research in which psychologically oriented experimentation might contribute a better understanding of the translator’s task relates to the bodily inscription of many such routines. How is it that we can experience pleasure in translating texts commonly described as difficult? What does the repetitiveness of the task contribute to the feelings in question? (Simeoni 1998: 27^8) 2. Toury’s norm theory also lacks the iterative thru-put element of somatic theory, which Simeoni insists on for habitus theory as well: the notion that people who internalize norms thus become carriers of norms, passing them on to other members of the group, and invariably in slightly transformed or ‘‘personalized’’ ways. As Derrida at least implicitly suggests in his theory of iterability, and as we will see further in connection with Bakhtin’s theory of double-voicing in Chapter 7, human beings are not machines, capable of transmitting language or other social material unchanged. They always contribute something individual to whatever passes through them. To somatize a social structure is to be transformed by it; but it is also to transform that structure in return, at least for those to whom one then passes it on. As Simeoni puts this,‘‘the habitus is both . . . ‘structured’, i.e. the nexus of dispositions at any given time is given a structure in the course of individual social lives; the habitus is acquired and shaped, neither innate nor a haphazard construction . . . [and] a ‘structuring’ mechanism, i.e. the bundle of dispositions thus acquired contribute[s] directly to the elaboration of norms and conventions’’ (21^2). Simeoni also notes that Toury’s Tel Aviv colleague Itamar Even-Zohar has recently incorporated something like this structured/structuring tension into his discussion of the transmission of ‘‘repertoires,’’ urging scholars to examine repertoires in terms of the principles both of ‘‘organizedness’’ and of

The translator’s habitus

87

‘‘organizingness,’’ and calling for ‘‘special attention to the activity of the makers of repertoire who are at the same time agents of transfer’’ (Even-Zohar 1997: 381). Toury too, it should be noted, having read Simeoni’s article before writing his Aston paper, built a head-nod toward this same thru-put idea into it, but tellingly, with an authoritarian twist: ‘‘Very often, these norms ^ or even the basic agreements and conventions ^ go on being negotiated throughout one’s entire life in the group, for instance, when members struggle to establish their own positions within the group (or vis-a-vis its other members). Moreover, certain individuals may be more instrumental than others in effecting changes in the norms, depending on the status and position they have acquired in the group’’ (Toury 1998: 17). The implicit assumption here is that you have to have a position of power in a group to help shape group norms: presumably because he relies on the metaphor of ‘‘negotiating’’ to explain how norms are formulated, Toury thinks that in order to have any kind of impact at all on group norms you have to be important enough in the group to sit at the negotiating table. One never gets the sense, reading Toury’s work, that he sees translators or members of any social group as agents whose actions always impact the group and its norms; for him translators are basically the passive receivers and obeyers of group norms, never their creators, carriers, or revisors. (Unless, of course, they are elevated into a higher ‘‘status or position’’ in the group and are thus given some say in the (re)formulation of norms as well. Just what kind of ‘‘status or position’’ in the translation profession would give a translator that kind of say, Toury doesn’t tell us; and in fact I think it would be difficult to come up with such a ‘‘status or position,’’ since such things are not ‘‘negotiated’’anywhere; translation norms arise out of the structured/structuring interactions of the marketplace, not out of the work of committees or other decision-making bodies to which translators might be appointed as the representatives of ‘‘practitioners’’or some such.28) 3. Related to this structured/structuring effect is the lack in Toury’s normtheoretical model of a sense of the individual’s complex social freedom. True, in his Aston lecture Toury pointedly allows the translator ‘‘freedom of choice,’’ but specifically only in the sense that those governed by rules always have the option ^ the perverse option, inToury’s view ^ of breaking the rules: To put it bluntly, it is always the translator herself or himself, as an autonomous individual, who decides how to behave, be that decision fully conscious or not. Whatever the degree of awareness, it is s/he who will also have to bear the consequences. Remember the notion of sanctions (paragraph 3.7)? At the same time, it is clear that, even though there is always the possibility that one would be willing to take the risks which unconventional, non-normative decisions entail, under normal conditions, a translator would tend to avoid negative sanctions on ‘improper’ behaviour as much as obtain the rewards which go with a ‘proper’ one. Needless to say, it would make an interesting

88

Iterations (the performative back-story) project to study the [negative and positive] sanctions that may be associated with translational behaviour and their (possible and actual) effects on instances of performance within defined sociocultural settings (see the way Daniel Simeoni has recently related the concept of ‘norm’ and Bourdieu’s notion of ‘habitus’ in the specific context of translation, Simeoni, 1998). (20; square and round bracketsToury’s)

The Tourian translators have internalized the norms and know the consequences for transgressing them; if they are willing to accept those consequences, they may decide to transgress. But this is like saying that every citizen in a police state is free to break the laws and go to jail, or that every believer in a wrathful god is free to defy that god’s laws and go to hell: a very grudging kind of freedom, at best. Simeoni begins to take us past this implicit totalitarianism by noting that norms never perfectly govern any individual’s behavior. For one thing, norms conflict: the same freelancer may be asked in the course of a single month to do a back-translation, sticking as closely as possible to the original syntax to show the client whether the original translation was properly done; to localize a piece of software, adjusting measurements from English to metric, date formats from month-date to date-month, and so on; to give a client the gist of a letter over the phone; and to edit the work of another translator. What are the ‘‘norms’’ governing this translator’s professional behavior? Clearly, they are different in each of these jobs. Simeoni asks: Could the elusive faculty of translating today primarily be one of adjusting to different types of norms, making the most of them under widely varying circumstances (the image of Dryden serving different masters, and advising translators to steer a middle course, would then be truly emblematic)?. . . In a different order of concerns, could the increasing variety of tasks they are being asked to perform (different clients and contracts, integrating diverse computer skills, working increasingly in their second or even third languages, sometimes stretching their expertise to the fuzzier domain of ‘‘information and consulting services’’) have alerted translators to the relativity of the demands placed on them, thereby causing some degree of cognitive dissonance in their historically imposed submissiveness, making them perhaps also more receptive to Translation Studies? Could it be, circumstances permitting, that the mythical belief in pure, untainted service will eventually prove more and more difficult to sustain? (13^14) This variability in the norms that (externally or internally) constrain translators’ behavior leaves them considerably more freedom of movement than Toury recognizes. Translators in Simeoni’s conception become far more than

The translator’s habitus

89

the passive receivers and obeyers of norms; they become the active interpreters and appliers of norms. This perception is thoroughly borne out by my own experience as a translator as well. For example, clients and agency project managers always seem to neglect to tell me when a job I have been sent is a back-translation. This seems counterproductive to me, since the purpose of a back-translation is typically very different from that of a front-translation: an initial translation may be intended to be fluent or easily accessible in the target language, in order to address a target audience as effectively as possible, while the translation of that translation back into its original source language is intended to demonstrate to the client that the original translator did his or her job properly, or to identify problem areas in the translation. Fluency or easy accessibility is not a high priority; strict accuracy is. If I assume that a back-translation job is in fact a front-translation job, I may feel tempted to ‘‘fix things up,’’ smooth out awkwardnesses, straighten out misleading references, etc. This might be beneficial if the result of my work were intended for publication, but will be counterproductive in a back-translation, where the possible problems in my source text (which is some other translator’s target text) are the whole point. So why don’t clients and agencies tell translators these things? Who knows. The upshot of their not telling me, though, is that I have to guess, and call the client or agency to confirm my guess, and translate accordingly. I have to interpret the nature of the job so as to apply the appropriate translational norm. This is far more freedom of action thanToury’s model of simply breaking the rules and suffering the consequences would allow. The translator’s ‘‘norm-governed’’ freedom is even clearer in the translation of badly written texts. Almost every conservative translation theorist agrees that the translator should improve a badly written text, and not simply transfer to the target language the original’s shoddiness as accurately as possible. There is less agreement as to how, in doing this, the translator is to proceed. The idealized advice is to make all improvements in consultation with the author, but in my experience, again, this is a frustrating process for the ‘‘norm-governed’’ translator because the authors of badly written texts (when they can be tracked down, of course) are typically no help at all in disambiguating their texts. Their editing skills were too weak to fix up the text before it went out to you, and they most often remain too weak to help you fix it up after you’ve torn your hair out over it. So you come up hard against the limits of the norms that say that the translator must serve, submit to, be subservient to the source author: in order to serve source authors who are bad writers, you have to exceed them, go beyond what sketchy and shaky authority they can muster. Every time you interact with such an author, your sense of your subservience both to the source author and to the norm that enforces subservience to the source author is eroded, until finally you begin to reformulate the norm in more radical ways:‘‘Translate subserviently, unless the author’s an idiot, in which case serve the poor target reader by imposing more order on the text than the author could ever hope to imagine.’’ Hence the job I once did for a Finnish professor of physical education who sent

90

Iterations (the performative back-story)

me a paper he was going to be reading at an international conference: the Finnish text was hopelessly garbled, lacked all argumentative coherence, so I fixed it up for him, moved sentences around in paragraphs and paragraphs around in the paper. This was radical enough, at least in a normgoverned purview, though it is certainly a common enough practice in the real world. But the professor was in a huge hurry and drove out to my house to pick it up, and when I handed it to him I told him exactly what I had done, breaking the tacit norm that says ‘‘if you improve a badly written text, make the changes tacitly.’’ Did I suffer sanctions or other dire consequences for my transgressive behavior? No I did not.The professor thanked me profusely, indeed rather abjectly, confessing his own hopelessly inadequate writing skills and his gratitude to anyone who made one of his pieces work properly. Indeed my idiosomatic theory of translation in The Translator’s Turn is only slightly more radical than Simeoni’s insistence on the translator’s active and constructive freedom in internalizing, interpreting, and applying conflicting norms. This of course was the part of the somatic theory that got some readers’ danders up, because they thought I was saying ‘‘translate any way you damn well feel’’; but all the theory of idiosomatics said was that ideosomatic programming is never monolithic, collective regulation therefore never perfect. Social experience is at ‘‘best’’ (or, depending on the extent of your love of law and order, at ‘‘worst’’) only partially regulated by collective forces like norms and conventions. The theory of idiosomatics grows out of the perception that we all develop our sense of group norms in different combinations of structured/ structuring interactions with other group members, so that the specific norms we construct out of those interactions will always be more or less ‘‘idiosyncratic.’’ Translation is defined socially and professionally, Simeoni agrees, in terms of subservience; as a result, translators tend to internalize an ethos of subservience, complete with specific norms for achieving that subservience in given professional circumstances (job interactions). This means that ‘‘idiosomatic’’ translation will most likely be just as subservient to norms as conservative theorists would want ^ just different norms, or different combinations of norms, or norms differently interpreted. Translators sent lots of badly written texts to translate, for example, will internalize very different norms from those sent only masterpieces of eloquence, concision, and wit. My theory of the idiosomatics of translation is not about creative anarchy, the freedom to translate any way the translator feels ^ that is a red herring growing out of lazy binary excludedmiddle thinking, according to which, if I don’t enforce absolute fidelity to the text, I must be clamoring for some sort of carte blanche. The only real difference between the translator’s act conceived somatically and the translator’s act conceived norm-theoretically is that in the former I recognize that the translator plays an active role in identifying and internalizing subservience norms, and is exposed to a specific array of such norms somewhat at random, as the marketplace coughs up jobs for him or her to do, so that his or her norm-governed professional behavior will differ idiosomatically from that of other translators.

The translator’s habitus

91

4. Finally, what is lacking in both Toury’s norm theory and my somatic theory, and amply provided in Simeoni’s habitus theory, is a complex vision of the social world and its internalization by the individual. My somatic theory is based on a more complex vision of collective regulation and individual action than norm theory, but I am not a sociologist, and tend to reduce social interaction to the relative simplicity of one-on-one encounters. In this sociologically challenged state of mine, I find habitus theory one of the most productive and exciting contributions to performative linguistics ^ performative sociology, perhaps ^ to come down the pike in a long long time.

Habitus theory In his ‘‘brief conceptual genealogy’’ of the term habitus, Simeoni explains that it is a classical Latin translation of Aristotle’s hexis,‘‘a quality of being or ‘disposition’ characterized by stability and permanence, as opposed to diathesis (changing disposition) and pathos (simple accident)’’ (Simeoni 1998: 15). Hexis for Aristotle is closely related to ethos, or ‘‘character’’; both are shaped by the choices one makes, and in turn shape further choices.‘‘Virtue,’’ in fact, he calls hexis proairetike, which Castoriadis glosses as ‘‘habitus dependent on and creator of choices.’’ In the Rhetoric he distinguishes between hexis as ‘‘those dispositions only which determine the character of a man’s life’’ and stylistic variation, insisting that hexis ‘‘is ‘reproduced’ in his or her style of speaking’’ (all three passages quoted by Simeoni 1998:15). Both Aristotle’s Greek term and the Latin term translated from it had lost conceptual currency by the twentieth century, when the Latin version was resuscitated by the German sociologist Norbert Elias to mean ‘‘the kind of generalized ‘embodied social learning’ whereby ‘the fortunes of a nation over the centuries become sedimented [in] its individual members’ (1996:10). According to Elias, nations are carriers of a specific cultural habitus which is transmitted to natives as a matter of routine yet imperious socialization’’ (Simeoni 1998: 16). The term was again picked up by the art historian Erwin Panofsky in his 1948 Wimmer lectures, published in 1951 as Gothic Architecture and Scholasticism, in order to trace a strong cultural connection between the two terms of his title: gothic architecture in a 100-mile radius around Paris between the early twelfth and the late thirteenth century and the scholasticism of the Sorbonne, he argued, crossfertilized each other, fed each other ideas or assumptions or ‘‘dispositions’’ that go beyond mere‘‘parallelisms’’or ‘‘influences.’’ Simeoni takes his conceptualization of the habitus most proximately, however, from Pierre Bourdieu, who translated Gothic Architecture and Scholasticism into French in 1967 and wrote a postface commentary for his translation. There he calls the habitus ‘‘a ‘stenograph’ for any ‘system of dispositions’ specific to (and active in) not only a nation-state but the ‘fields’ within it. When he first developed his own full version of the concept (1972: Part 2),’’ Simeoni adds, ‘‘he was able to operationalize the notion of individual differentiation within a single culture without jeopardizing

92

Iterations (the performative back-story)

the centrality of organized systemic properties’’ (Simeoni 1998: 16). Simeoni quotesJohn B. Thompson’s definition of the concept as theorized by Bourdieu: The habitus is a set of dispositions which incline agents to act and react in certain ways. The dispositions generate practices, perceptions and attitudes which are ‘regular’ without being consciously coordinated or governed by any ‘rule’. . . . Dispositions are acquired through a gradual process of inculcation in which early childhood experiences are particularly important. Through a myriad of mundane processes of training and learning, such as those involved in the inculcation of table manners. . . . the individual acquires a set of dispositions which literally mold the body and become second nature. The dispositions produced thereby are also structured in the sense that they unavoidably reflect the social conditions within which they were acquired. (quoted in Simeoni 1998:16^17) The motto for Bourdieu’s concept of habitus as it applies to the individual’s choice of profession, Simeoni says, might be ‘‘Don’t even think of entering a field if your habitus does not match the requirements’’ (17): don’t even dream of becoming an opera singer if you are not already a very good singer, for example; and, by extension, don’t dream of becoming a translator unless you have some considerable skill in and love for foreign languages. And even these ‘‘requirements’’are only the tip of the habitus iceberg: the opera singer would be expected not only to be an excellent singer but to enjoy performing before audiences (shy, retiring people need not apply), to travel (homebodies need not apply), and so on; the translator would be expected to enjoy devoting long hours to sitting at a computer fiddling with words, so that a person who hated sitting staring at texts for long hours or worrying about minor semantic nuance differences should be discouraged from moving into the translator’s profession. But the habitus is also much more than the set of dispositions that incline us toward (and develop out of ) our choice of profession: But the extension of the concept is such that it applies not only to the state-national fields of cultural production or, for that matter, economic production, but to the social environment at large and to the circumstances of everyday consumption, lifestyles or preferences. In such an environment, influences are so entangled on account of the fuzziness of field borders and the relationships within, that the larger social constituency can hardly be called a ‘‘field’’ or, at least, it should be viewed as a much more complex and messy system, where most participants are not even aware of the rules, or worse, misread them as being similar to those in force in the specialized fields in which, in the best of cases, they happen to be active professionally. The overall effect is one of generalized misunderstanding or of partial understanding. Illustrations are easy to find: Although nothing would seem to stop us from buying newspapers, magazines or postcards

The translator’s habitus

93

at random, it is a fact that we stay away from some of these items almost naturally, as it were, going instead for ones better suited to our tastes. We do not really ‘‘understand’’ how such items [as the ones we do not prefer] can find buyers.. . . Our decisions in social life make sense for us alone, convincing us that our choices are the only valid ones, all others being either futile, distasteful or plain wrong. (ibid.:17^18) An interesting place to look for examples of this ‘‘generalized misunderstanding’’ (which might serve as a very rough paraphrase for ‘‘iterability’’) is in the kind of theorizing about translation one finds among translators with significantly different professional habitus: . .

.

.

the translator of technical texts tends ‘‘naturally’’ to assume that translation is all about accuracy and has nothing to do with £uency or style; the translator of advertising copy tends to assume that translation is about £uency and reader impact and to dismiss strict semantic accuracy as a petty academic concern; the court interpreter tends to thematize accuracy in terms of strict neutrality, and ‘‘naturally’’assumes that a biased translation is no translation at all; the elite academic translator of literary classics or di⁄cult scholarly texts tends to assume that translation is about giving the reader a feel for the foreignness of the original, and cannot fathom how other translation scholars can blithely talk about ‘‘meeting the needs of the market’’or ‘‘giving target readers what they want.’’

Each different professional habitus conditions the person living it to generalize from it to all translation, to consider the ‘‘rules’’ or ‘‘regularities’’ of his or her own experience somehow universal, intrinsic to ‘‘translation in general,’’and to regard those with different views as somehow perverse or twisted. To the translator or translation scholar working within it each habitus feels like a whole world, like the whole world, because our experience changes us somatically. We somatize what we learn so that we can remember it more efficiently (rapidly and unconsciously) later. And as the ‘‘norms’’ that we mark somatically vary with experience, so too do the translators whose habitus emerge out of those experiences, out of those divergent somatizations of norms, tips, clues, hunches, and so on. Somatic markers are learned, but they feel like instincts: hence our unthinking assumption that our intuitions about the world we live in are not only ‘‘accurate’’ but somehow universally human, as hard-wired into our brains as the fight-or-flight reflex. It rarely occurs to us just how constructed our ‘‘intuitive’’ sense of the world actually is. We will be moving on to Mikhail Bakhtin’s dialogical theory of language in the next chapter. Before we do, let me iterate the somatics of iterability one more time.

94

Iterations (the performative back-story)

Our experience shapes us so powerfully, gives such change-resistant form to our habitus or ‘‘dispositions,’’ because we store what we experience somatically. At the broadest level, this is obviously and trivially true: we have bodies; our brains are part of our bodies; everything we learn with our brains we learn in a bodily sense, somatically. On a slightly less trivial level, Antonio Damasio shows persuasively that we learn not just with our brains (our ‘‘central nervous system’’ or CNS) but with our peripheral nervous systems (PNS) as well, the nerves running through our entire body and connecting up with our brains. Most important, we learn somatically in the sense that whatever our experience teaches us is important we mark somatically, so that we don’t have to remember it consciously, and so that we don’t have to hold all the minute details relevant to a decision-making process in our minds at once. This is the somatic channel through which we internalize the norms and dispositions of our social group(s): in order to belong, to get along, to get by, to succeed, to be liked and loved, to be respected and admired, we adapt our thoughts and feelings and behaviors to those of the various groups to which we belong. We learn what is acceptable and what is unacceptable, and seek to conform ourselves as effectively as possible to the former, and to unlearn or else carefully hide (suppress or repress) the latter. We internalize and obey social and professional standards, assumptions, attitudes, values, beliefs, practices. In Derrida’s terms, we repeat these things in our own assumptions and attitudes and values and practices and so on. We iterate them. We somatize them in order to more effectively iterate them. We are able to iterate them because we have somatized them, made them our own, embodied them in powerful and indeed often indelible ways. In an important sense we are our iterations. But as Derrida reminds us, iterations involve change too, not just repetition of the same. Whatever we somatize, we modify. Whatever we repeat, we alter. This is true not only because, as Daniel Simeoni reminds us, we belong to many different social groups and somatize the dispositions characteristic of each into a complex ‘‘mosaic habitus’’ (Simeoni’s term29) that is uniquely our own, but also because each human being has a separate nervous system concerned with its own survival, and will somatize experiences in his or her own way. We are mammals with a strong need to belong, to share feelings, to be surrounded by people like us; but we are also separate creatures that cannot help but experience the world at least partly idiosyncratically. As Simeoni puts it, we are both structured and structuring: we are shaped somatically by our encounters with other people, and we shape them in return. The specific people we happen to come in contact with shape us, and we shape them, in ways that are significantly different from those that would obtain had we come into contact with some other specific people. Iterability is not, then, mere French poststructuralist sleight-of-mind. It is a highly realistic performative description of the social behavior of human beings. At least, that’s how I’m performing it.

7

Double-voicing

Austin’s binary impasse J. L. Austin was an analytical philosopher; that meant, among many other things, that he was trained in binary thinking, the primary analytical mode in the Western intellectual tradition. And as we saw in Part I, constative linguists like Emile Benveniste (1966) and Jerrold Katz (1979) took Austin’s binary distinction between constative and performative utterances to be his main contribution to linguistic theory. In his recuperation of Austin’s ‘‘central’’ ideas for constative linguistics, John Searle (1969) focused on the illocutionary act, which he was at great pains to distinguish as clearly as possible, in traditional binary fashion, from the locutionary act of uttering certain phonemes that carried a certain meaning. Indeed in Searle’s hands this locutionary versus illocutionary distinction became a reformulation of the constative versus performative distinction, rendering Searle’s binarization of Austin’s argument structurally congruent with Benveniste’s and Katz’s. Austin’s own argumentation, however, ran in the opposite direction: toward the systematic erosion of binary distinctions. It fact it appears, reading Austin, as if he only set up binaries in the first place so as to erode them ^ posed binary distinctions heuristically, in order to bring certain heretofore ignored linguistic phenomena into bold relief, with the primary idea of then systematically ‘‘counterexampling’’ them, throwing them into the pot of ordinary language use and stirring furiously until the original binary distinction was all mixed into the same chunky performative stew. This strategy is especially clear in Lectures VII, on the distinction between the constative and the performative, in the course of which he discovers that ultimately there is no real difference between them as acts (stating or informing is a speech act just like marrying or betting or promising) and ends up junking the whole thing and starting over, and IX, on the distinction between the illocutionary act and the perlocutionary act. It is this latter chapter that I want to read briefly here, by way of setting up a rather spectacular performative solution to a crucial Austinian problem, offered by the dialogical theory of Mikhail Bakhtin. Austin’s problem in Lecture IX, to begin with, is that he wants to see perlocutions not as the effects of actions (though he does also call them that), but as actions

96

Iterations (the performative back-story)

in their own right: illocutionary acts are actions with a certain conventional force (promising, betting, naming, marrying, informing, etc.), perlocutionary acts actions that bring about some significant change in another person’s thoughts or behavior. And at first glance there is a good deal to say for such a distinction. Austin (102) gives the example: ‘‘He said to me,‘You can’t do that’’’ (locutionary act),‘‘He protested against my doing it’’ (illocutionary act), and ‘‘He pulled me up, checked me,’’ ‘‘He stopped me, he brought me to my senses,’’ ‘‘He annoyed me’’ (perlocutionary act). There is, surely, an obvious and significant difference between merely protesting an action and actually stopping it? But as Austin delves into this distinction, it gradually comes apart at the seams. Indeed he predicts on p. 110 that this distinction is in fact the ‘‘likeliest to give trouble’’: ‘‘For clearly any, or almost any, perlocutionary act is liable to be brought off, in sufficiently special circumstances, by the issuing, with or without calculation, of any utterance whatsoever, and in particular by a straightforward constative utterance (if there is such an animal).’’ If perlocutionary acts are performed in or by or through illocutionary acts, which are in turn often but not always performed in or by or through locutionary acts, and if there is no sure way to guarantee in advance that a certain chain reaction will ensue from the performance of a certain locutionary act, how can we ever seriously hope to distinguish these things? I can pull you up, check you, stop you, bring you to your senses in any number of ways: I can tap you on the shoulder and give you a certain significant look; I can block your path; I can explain to you at length just what a foolish thing this is you’re attempting; I can call you a silly ass; I can encourage you to do it for the wrong reasons; I can play on your vanity, or your fear; I can get your spouse and children and parents and siblings and all your friends to pressure you; and so on. Any of these (il)locutionary acts might work as a perlocutionary act, but there is no guarantee that the desired perlocutionary act ever will be executed. It does seem as if perlocutions are the consequences or effects of illocutionary acts, and not acts in their own right. But this is too simple too. A few pages later (116) Austin offers an example: ‘‘I cannot be said to have warned an audience unless it hears what I say and takes what I say in a certain sense. An effect must be achieved on the audience if the illocutionary act is to be carried out.’’ If this is true, an illocutionary act must entail or engender a perlocutionary act or fail. But the warning example perfectly reflects the mind-numbing complexity Austin is probing here: in fact we can say, and frequently do, things like ‘‘I warned him, but he just wouldn’t listen!’’ This suggests that the illocutionary act of warning might in fact be somehow systematically distinguished from the perlocutionary act of making the warning stick, warning him out of what he was planning; but Austin is also right when he says that warning somebody requires that the person take the warning ‘‘in a certain sense.’’ What sense? Does he have to agree not to do what he planned? Clearly not; that would be ‘‘stopping’’ the person, not warning. But he does have to take the warning seriously, right? If I utter what sounds like a warning and the person laughs, because it’s clear to him that I am just joking, playing around with the exterior form of warning, it seems reasonable to

Double-voicing

97

claim that I haven’t actually warned him at all. If I’m not joking, if his assumption that I am is based on a bad reading of what I’m trying to do, then I will have to keep revising my warning until it actually sinks in, secures, as Austin says (117), the desired uptake. Only then can I be said to have warned him. The problem with this for Austin as for us his readers is that it seems to erode the distinction not only between an illocutionary act and a perlocutionary act, but between me and you as well, between the speaker and the listener as separate beings. If what I do can only be classed as a certain kind of action if you react in a certain way to it, then the action is not mine alone; it is ours. It is ‘‘shared,’’ in some strange collective way that we do not yet have the conceptual apparatus to understand. The action of warning is not individual; it is systemic. A socius performs it, a socius of two, and both play integral and intertwined roles in it, warning and being warned, which are not two separate actions but rather two mutually dependent aspects of the same action. If my warning doesn’t warn you, if you refuse to be warned by my warning, I didn’t warn you at all! The same is true of intimidating, insinuating, suggesting, cheating, annoying, and many other apparently individual illocutionary acts that cannot (except by constative violence) be separated from the perlocutionary acts that they require for their completion. If I try to intimidate you, I don’t succeed unless you actually are intimidated. If I insinuate or suggest something, hint at something in a roundabout manner, you have to get it or my insinuation or suggestion fails. If I cheat you and you notice it and refuse to be cheated, or if you know that I cheat everyone and have deliberately entered into the transaction with the surreptitious intention of giving me charity without hurting my pride, I haven’t cheated you at all. The perlocutionary act of ‘‘annoying’’ someone is the product of a group effort that may in fact be very complex.We can imagine, for example, three-way situations in which ‘‘annoying’’ is an interpretation imposed dramatically on the interaction of an apparent annoyer and an apparent annoyee: a mother may say to her son who is talking to a stranger in a park,‘‘Stop annoying that nice man,’’ and the man may smile back,‘‘Oh, don’t worry, ma’am, he’s not annoying me.’’ Is the boy performing the perlocutionary act of annoying the man, or not? Most likely, here, the boy’s behavior is annoying the mother more than the man (perhaps what is annoying her is the possibility that the man is only pretending not to be annoyed); but the mother is not likely for all that to discard her belief that the boy is in fact annoying the man, despite his claims not to be annoyed. ‘‘Annoying’’ here is clearly a collective action performed by all three, a minidrama whose roles in fact vary with the point of view of each person engaged in playing it out. From the mother’s point of view, we might imagine, the child is the annoyer, the man is the annoyance victim, and she herself is the annoyance interpreter and apologist, the mediator who attempts on her child’s behalf to soothe any annoyance the man might feel. From the child’s point of view his interaction with the man may simply be an interesting social encounter and his mother an annoying parent who won’t let him do what he wants; or maybe there is something about the man that annoys him, and he wants to

98

Iterations (the performative back-story)

annoy the man back. (He may also be completely unaware of this latter drama and go right on thinking of it in terms of a normal harmless social encounter that his mother perversely insists on interpreting as annoying.) From the man’s point of view, too, several different things may be going on at once: he enjoys the boy but is annoyed at the mother’s insistence that he is feeling annoyed; or he does find the boy’s behavior annoying, on one level, but on another level (say one where he remembers being that age himself and yelled at by his mother for annoying strangers) is annoyed with himself for feeling that way and with the mother for reminding him of his own mother. In all this, then, is it a reasonable project to isolate the boy’s illocutionary and perlocutionary acts from the interconnected dramatic situation, and determine with analytical clarity and certainty whether the boy really is performing the perlocutionary act of annoying the man? Only, I would suggest, in a methodologically nervous constative world. But Austin never quite notices this transpersonal interconnectedness that hovers about his illocutionary/perlocutionary distinction; he goes on cheerfully trying to sort things out in terms of separate bodies interacting, apparently failing to notice that it never quite works. He tries to distinguish between the various forms of ‘‘effect’’ required for an illocutionary act to be successful, ‘‘securing uptake, taking effect, and inviting a response,’’ (118) and ‘‘the producing of effects which is characteristic of the perlocutionary act’’ (118), but this distinction too quickly collapses: The perlocutionary act may be either the achievement of a perlocutionary object (convince, persuade) or the production of a perlocutionary sequel. Thus the act of warning may achieve its perlocutionary object of alerting and also have the perlocutionary sequel of alarming, and an argument against a view may fail to achieve its object but have the perlocutionary sequel of convincing our opponent of the truth (‘‘I only succeeded in convincing him’’). What is the perlocutionary object of one illocution may be the sequel of another. For example, warning may produce the sequel of deterring and saying ‘‘Don’t’’, whose object is to deter, may produce the sequel of alerting or even alarming. Some perlocutionary acts are always the producing of a sequel, namely those where there is no illocutionary formula: thus I may surprise you or upset you or humiliate you by a locution, though there is no illocutionary formula ‘‘I surprise you by . . .’’, ‘‘I upset you by . . .’’, ‘‘I humiliate you by . . .’’ The many confusions Austin sows in Lecture IX in his exuberant and rather feckless attempts to sort these things out stem, then, I’m suggesting, from his ideologically ‘‘normal’’ insistence on seeing the two or more actors in these minidramas as separate individuals, rather than as the interdependent components of a social dialogue. If we shift our ideological (and thus methodological) focus from the individual as a separate monad to the socius, then most of Austin’s problems disappear: the actions performed in social

Double-voicing

99

contexts are always interactive, always performed not by individuals but by the entire socius involved in them.

Dialogism This is in fact the radically dialogical direction in which Mikhail Bakhtin points us ^ in particular in two early works, his book on Dostoevsky from 1929 , ‘‘Problems of Dostoevsky’s Creative Art,’’ revised by Bakhtin himself in 1963 as , translated by Caryl Emerson in 1984 as Problems of Dostoevsky’s Poetics), and a long essay on language and the novel from 1934^5 (‘‘ ,’’ translated by Caryl Emerson and Michael Holquist in 1981 as ‘‘Discourse in the Novel’’).30 In his introduction to the English translation of Bakhtin’s Speech Genres essay collection, Michael Holquist writes: ‘‘The Problem of the Text’’ is typical of most works from Bakhtin’s last years in that it is not so much an essay as a series of entries from the notebooks in which Bakhtin jotted down his thoughts. Keeping such notebooks was a habit he had developed in his youth and one he maintained throughout his career. This lifelong dialogue with himself accounts for many of the features that characterize Bakhtin’s style (or, more accurately, one of Bakhtin’s styles): the allusive structure of his remarks and the repetitiveness that so often bothers readers trained to value more economical and forensic presentation. Anyone expecting a finished, consecutively prosecuted argument in these pieces that have been torn out of the notebooks is bound to be frustrated. But the suspension of such expectations reveals a style that has its own rewards: not the pleasure we derive from an author who compels us to believe his logic is ineluctable, but the excitement that comes from seeing a mind at work while it is at work. (Bakhtin 1986: xvii) I do, in fact, find reading Bakhtin exciting. I have been reading him for almost twenty years now, and every time I go back to him I rediscover the somewhat daunting extent to which my most ‘‘original’’ thoughts on language come from him. His prose, often repetitive, sometimes bombastic, is nevertheless so rich with transformative insights that my first impulse in using his ideas in a piece of my writing is always to quote him at length ^ especially, I suppose, from ‘‘Discourse in the Novel,’’ which I did, in fact, in TheTranslator’sTurn. But I am going to resist the temptation to do the same here. Bakhtin and Burke were the central theoretical geniuses of that earlier book of mine; Austin and Grice are the main figures here, and all the other theorists I draw ideas from, especially Derrida, Simeoni and Bourdieu, and Bakhtin (and even that earlier Robinson) here in Part II, and Charles Sanders Peirce in Part III,

100

Iterations (the performative back-story)

should I feel coach my argument more from the wings, letting the two English speech-act philosophers occupy center stage. And so I propose to reduce Bakhtin’s complex theoretical musings on language to a more or less syllogistic series of propositions about the dialogism of language, and, when I quote, quote less directly from Bakhtin and more from two excellent overviews of his thought, Michael Holquist’s Dialogism: Bakhtin and his World (1990) and Gary Saul Morson and Caryl Emerson’s Mikhail Bakhtin: Creation of a Prosaics (1990). This I hope will greatly facilitate argumentative economy ^ not a virtue widely associated with either Bakhtin’s writing or mine, but I think this time a necessary one. I will also give page references to the relevant passages in English translations of Bakhtin’s works in parentheses; and, of course, when Bakhtin’s particular formulation of a point is both absolutely crucial to my argument here and unusually concise, I will quote it as well. The distinction I am drawing in this book between constative and performative linguistics is roughly congruent with Bakhtin’s basic distinction between theoretism and dialogism. Theoretism for Bakhtin is based on the Platonic notion that transcendental forms (especially structures, laws, and rules) are more ‘‘real’’ than actual events; theoretists, he insists, seek to read through human actions and social events to the pure and stable forms that supposedly undergird them. From the purified perspective of these forms, acts and events are seen as the mere clumsy or distorted instantiations of general abstract laws or rules, and thus as relatively irrelevant to the scholar of language (the constative linguist), of the psyche (the constative psychoanalyst), or of society (the constative sociologist). Outlining the organizing principles of dialogism, very briefly, will be the burden of this chapter. But even more briefly, by way of getting started, let me quote Morson and Emerson: That which makes real dialogue a live process transcends received models, none of which allow for unfinalizability. The problem with theoretism is that it begins analysis from the wrong place. According to Bakhtin, ‘‘we cannot break out into the world of events from within the theoretical world. One must start with the act itself, and not with its theoretical transcription’’ (KFP, p. 91). This kind of ‘‘transcription’’ reduces time to a mere parameter and loses the real historicity of an event. In theoretism, Bakhtin observes, ‘‘the temporality of the actual historicity of existence is only a moment of abstract cognized history’’ (KFP, p. 89). (1990: 50) This methodological imperative of ‘‘starting with the act itself ’’ is clearly very close to what I have been calling performative linguistics. As I hope I have made clear by now, this methodological shift from the constative to the performative, from the abstract law to the individual act, entails a radical realignment of scholarly directedness: it means basing all understanding on precisely that which is most strongly quarantined in dominant

Double-voicing 101 (constative/scientistic/theoretistic/objectivist) scholarly paradigms, namely what people do and say, the ordinariness of everyday life. Bakhtin was one of the twentieth century’s most eloquent prophets of that ordinariness. And now let us proceed to the propositions. I make no pretense of comprehensivity; I have selected from the vast wealth of Bakhtin’s dialogical thought only those principles that best suit my argumentative needs here. The principles I have chosen are central to Bakhtin’s dialogism, but they are not the only central ones, and anyone who knows Bakhtin well will no doubt feel frustrated upon arriving at the last proposition and finding three or four favorite Bakhtinian ideas unmentioned. 1. All being is dialogical. The part of this that interests me is specifically social being, humans interacting with each other, in particular speaking and writing to each other; but Bakhtin insists that dialogue is at the heart of all being, including that of inanimate objects. This is his way of solving the subject^ object problem, which tends to founder on the assumption that anything that is not a subject is by definition an object. Attempts to modify the philosophy of subjects and objects so that subjects deal with other subjects, or so that subjects come to share something called ‘‘intersubjectivity,’’don’t really escape this pitfall; they really only paste smiling ‘‘subject’’ faces on objects and agree to proceed as if the objects so designated really did have feelings and thoughts and so on. Bakhtin does not want to start with subjects as separate monads and then attempt to connect them; he wants to start with dialogue and show how all being is interrelated by it. Or as Michael Holquist (1990: 20^1) writes: Dialogism argues that all meaning is relative in the sense that it comes about only as a result of the relation between two bodies occupying simultaneous but different space, where bodies may be thought of as ranging from the immediacy of our physical bodies, to political bodies and to bodies of ideas in general (ideologies [but as we’ll see Bakhtin doesn’t really have a somatics for dealing with such things as ‘‘bodies of ideas’’]). In Bakhtin’s thought experiments, as in Einstein’s, the position of the observer is fundamental. If motion is to have meaning, not only must there be two different bodies in a relation with each other, but there must as well be someone to grasp the nature of such a relation: the non-centeredness of the bodies themselves requires the center constituted by an observer. But unlike the passive stick figures who are positioned at a point equidistant between two railway trains in the cartoons often used to illustrate Einsteinian motion, Bakhtin’s observer is also, simultaneously, an active participant in the relation of simultaneity. Conceiving being dialogically means that reality is always experienced, not just perceived, and further that it is experienced from a particular position. Bakhtin conceives that position in kinetic terms as a situation, an event, the event of being a self. For Bakhtin, Morson and Emerson (1990: 50) add,‘‘it is inaccurate to speak of entering into dialogue, as if the components that do so could exist in any

102

Iterations (the performative back-story)

other way. To be sure, particular dialogues may break off (they never truly end), but dialogue itself is always going on. In his linguistic work, Bakhtin endeavors to describe language so that dialogue is not a subsequent act of combination but is itself the starting point.’’ 2. The self is constituted socially, and thus dialogically. We are not born ‘‘separate’’ or ‘‘egotistical’’ creatures who, once we reach a certain age, begin to reach out to others, as Jean Piaget claimed; we are born undifferentiated from other people and our environment, and we gradually develop a self in dialogue with others, by internalizing parts of them that we gain access to through ‘‘fellowfeeling’’ (somatics ^ this part I’ve added to Bakhtin), and by sending our own intentions and desires and anxieties and fears to them. As a result, they are always in us and we are always in them.‘‘Our own’’ voices are made up of the voices of others, which we personalize, tonalize out of different organismic needs, but which still retain a powerful charge of dialogical engagedness.‘‘Our own’’ ideas, values, beliefs, social norms, are all dialogical constructs, primarily shared things that we pass back and forth across the boundaries marked by our skin. As Morson and Emerson (1990: 50^1) write, for Bakhtin neither individuals nor social entities ever constitute a monad. They are much looser, ‘‘messier,’’ and more open than that. The most interesting and most unfinalizable aspects of any interaction arise from the relative disorder of the participants. Morever, we usually think of monads as wholes with clear boundaries. Bakhtin warns us, however, that neither individuals nor any other social entities are locked within their boundaries. They are extraterritorial, partially ‘‘located outside’’ themselves. Thus, Bakhtin refers to the ‘‘nonselfsufficiency’’ of the self (TRDB, p. 287).‘‘To be means to be for another, and through the other for oneself. A person has no sovereign internal territory, he is wholly and always on the boundary; looking inside himself, he looks into the eyes of another or with the eyes of another’’ (ibid.). Indeed, the very metaphor of ‘‘territory’’ and ‘‘boundary’’ is faulty, according to Bakhtin. If one insists on using it, one necessarily winds up with paradoxical formulations. Bakhtin insists that cultural entities are, in effect, all boundary: One must not, however, imagine the realm of culture as some sort of spatial whole, having boundaries but also having internal territory. The realm of culture has no internal territory: it is entirely distributed along the boundaries, boundaries pass everywhere, through its every aspect. . . . Every cultural act lives essentially on the boundaries: in this is its seriousness and significance; abstracted from boundaries it loses its soil, it becomes empty, arrogant, it degenerates and dies. 3. Language as a medium of social exchange is constituted dialogically. Language does not exist in the abstract formal state of pure stability imagined by

Double-voicing 103 constative linguists ^ what Saussure called la langue. Language does not even exist in the form Saussure called la parole,‘‘speech’’ as the excluded other of constative linguistics. Saussurean parole is the instantiation of the stable structures of la langue, or what Noam Chomsky derogated as the mere human ‘‘performance’’ of the beautifully abstract and formal realm of language ‘‘competence.’’ Language is dialogue. It exists only in and through dialogue. Without dialogical encounters there is no language; there are only sounds and marks, abstract signs lacking all interactive human performativity, directedness, intentionality. As Morson and Emerson (1990: 142) write, for Bakhtin ‘‘there are no ‘neutral words,’ no ‘no one’s words.’ ’’ Language is always someone’s words, words uttered by someone in dialogue with someone else, intended by the speaker to have a certain kind of impact on that someone else in that particular dialogue. Bakhtin sometimes puts this differently: ‘‘language’’ as defined by (constative or theoretistic) linguists is static and formal, and the dialogism of discourse is ‘‘more’’ than that, so discursive dialogism is ‘‘more’’ than language and what is needed is not linguistics but metalinguistics (see e.g. Bakhtin 1984: 182^3). I would say that that particular conception of language is a constative fantasy, not language; what is needed is a linguistics that is attentive to the performative realities of language rather than trying to pretend they don’t exist. (Bakhtin has not had much of an impact on linguistics; he has, however, had a significant impact on communication studies, for example Linell 2001.) 4. Every utterance is ‘‘internally dialogized ’’ in the sense of being oriented both back to the utterance that gave rise to it and forward to the response that it will provoke. In ‘‘Discourse in the Novel,’’ Bakhtin calls this an orientation toward the listener’s response, a projected expectation that prestructures the utterance around what the listener may say in reply: ‘‘The word in living conversation is directly, blatantly, oriented toward a future answer-word: it provokes an answer, anticipates it and structures itself in the answer’s direction. Forming itself in an atmosphere of the already spoken, the word is at the same time determined by that which has not yet been said but which is needed and in fact anticipated by the answering word. Such is the situation in any living dialogue’’ (Bakhtin 1981: 280). In this model the speaker’s interpretation of the previous utterance (the ‘‘already spoken’’), the speaker’s new intention, and the listener’s expected interpretation (the ‘‘answer-word’’) coexist within the utterance, like a miniature image of the dialogue in progress. I will be doing more with this image in connection with Peirce’s interpretant triad in Chapter 11, and in a series of dialogical applications of that revised model in Chapters 12^14; for now let me note simply that it is ultimately only confusing to say that the dialogue is introjected into the utterance, as if the utterance were a container that could hold such things. The dialogical image is actually internal to the speaker’s sense of the utterance. Bakhtin means this, I think; he himself repeatedly insists that words don’t do things, people do. And this person-oriented view is at least implicit in his description: ‘‘The word in living conversation is directly, blatantly, oriented toward a future

104

Iterations (the performative back-story)

answer-word’’: oriented by someone, presumably. Not just generally oriented. Oriented as part of a passive construction implying a human agent, not as an essentializing adjective.‘‘It provokes an answer’’: it becomes the occasion of an answer, someone feels provoked by it to answer. But then he says ‘‘anticipates it and structures itself in the answer’s direction,’’and ‘‘Forming itself . . . ’’: here the verbs anticipates, structures, and forms all presume agent status for the word, the word as something capable of anticipating things, structuring itself, and forming itself. By the middle of his career, two decades later, Bakhtin has come to see language in terms of its ‘‘addressivity’’ (obraschennost 0 ), which I take to be a more strenuously dialogical way of saying its ‘‘iterability.’’ ‘‘Thus,’’ Bakhtin writes in the early 1950s in ‘‘The Problem of Speech Genres,’’ ‘‘addressivity, the quality of turning to someone, is a constitutive feature of the utterance; without it the utterance does not and cannot exist. The various typical forms this addressivity assumes are constitutive, definitive features of various speech genres’’ (1986: 99). In the next paragraph he distinguishes between ‘‘utterances’’ (which he is interested in) and ‘‘language,’’ which he continues to abandon to the (constative) linguists: As distinct from utterances (and speech genres), the signifying units of a language ^ the word and the sentence ^ lack this quality of being directed or addressed to someone: these units belong to nobody and are addressed to nobody. Moreover, they in themselves are devoid of any kind of relation to the other’s utterance, the other’s word. If an individual word or sentence is directed at someone, addressed to someone, then we have a completed utterance that consists of one word or one sentence, and addressivity is inherent not in the unit of language, but in the utterance. (Bakhtin 1986: 99) What he means to say, presumably, is not that ‘‘the signifying units of a language’’ are real things that exist in parallel with utterances (and speech genres) and somehow lack addressivity, but rather that ‘‘the signifying units of a language’’as conceived by (constative) linguists lack addressivity, and in fact do not exist. They are impoverished conceptual constructs that Bakhtin invokes mainly by contrast with his richer (performative) (meta)linguistic approach. 5. Every element of language comes to us dialogically, through actual dialogical exchanges, and remains saturated with the voices of its past users. This idea is the meat of what I want from Bakhtin, his contribution to the ‘‘performative back-story’’ that I am tracing here in Part II, and I want to spend a bit more time with it, both here in the numbered series of propositions and after. This time, after having succumbed to temptation and quoted Bakhtin at some length in (4), I want to go back to quoting his commentators, specifically again Morson and Emerson (1990:139): Words ‘‘remember’’ earlier contexts, and so achieve a ‘‘stylistic aura,’’ often misconceived as the word’s ‘‘connotations’’ around a semantic center.

Double-voicing 105 This aura is, in fact, the effect of manifold voices that do not reduce to unity or yield a center. In using a word, speakers may intone the word so as to question the values present in its aura and the presuppositions of its earlier usage. In other words, the word may be ‘‘reaccented’’ (as sentimental was in the nineteenth century). As they accumulate and come to be shared, reaccentuations add to and alter the already-spoken-about quality of the word. This process is an essential factor in shaping a word’s evolution. What this Bakhtinian approach cannot explain, however, is just how words might ‘‘remember’’ anything. This is partly the same problem I drew your attention to above, in (4), in connection with the word or the utterance that was somehow capable of anticipating an answer, structuring itself, and forming itself: words are, after all, disturbances of the sound waves, black squiggles on paper or the computer screen; they have no memory. Or, as constative linguists would say, they are abstract mental constructs represented by certain phonetic sounds and/or certain graphological marks: also, clearly, things without memory. But in (4) it was at least possible to imagine the utterer doing all the things Bakhtin said the word or the utterance was doing; here it isn’t. Language users might be imagined as remembering the specific contexts a given word was used in, the contexts in which they heard it used, but Bakhtin seems to mean something more than this, something taking place at a larger or more collective level of generality. What we need to explain this ‘‘word memory,’’ which I agree is crucial for a performative understanding of language change and language stability, is the somatic theory: the bodies of the human users of a language mark words somatically, through a potentially infinite series of evaluative tonalizations, and those somatic markers are stored in the memory of those users. (Bakhtin’s notion of dialogical ‘‘tonalization’’ is in fact his strongest nod toward a somatic theory of dialogue. After all, we tonalize words with our bodies ^ our vocal chords, our lungs, our gestures, etc. We feel tonalization in other people’s dialogical utterances with our bodies. Tonalization is a thoroughly somatic activity.) To the extent that those somatic markers are collectively regulated (‘‘ideosomatic’’), the memory of past values and presuppositions connected with a specific word may come to be thought of as ‘‘remembered’’ by the word itself, contained within it, or else laid out around its perimeter like the fields around a walled city, the standard image usually implicit in the idea of a word’s connotations or ‘‘peripheral semantic fields.’’ Or, in the more Bakhtinian vocabulary Morson and Emerson invoke, these collectively regulated ideosomatic markers can be construed as forming a word’s ‘‘stylistic aura,’’ an image that likewise imagines ‘‘style’’ as a periphery (this time one of light rather than of fields) fanning out from a central object (in religious art, the saint’s head). But as Bakhtin tirelessly reminds us (without however referring specifically to somatics), the somatic markers that a speech community programs its members to attach to any given word are never perfectly regulated, therefore

106

Iterations (the performative back-story)

never perfectly collective. They are channeled in and by and through specific speech acts, specific communicative situations, specific dialogues, and what is channeled through them is not only collective regulation (ideosomatics) but individual tonalizations as well (idiosomatics). And what is stored for each word and each group of words in each language user’s memory is a cluster of somatic markers inherited from past usages, past experiences of the word’s use, experiences that are by definition saturated with both the regulatory tonalizations of the community and the idiosyncratic tonalizations of individual users. (Indeed it could be argued that the regulatory tonalizations of the community are in practice just a convergent accretion of the idiosyncratic tonalizations of individual users: when the specific tonalizations that are somatized are all more or less the same, they tend to have the force of collective regulation; when they diverge, they seem more idiosyncratic.) This suggests something like the following ‘‘edited’’version of Morson and Emerson’s remarks on pp.104^5: In using a word, speakers may intone the word so as to question the values present in the somatic markers they have stored for it, and the presuppositions they have somatized from its earlier usage. In other words, they may ‘‘reaccent’’ the word (as speakers of English did to the word sentimental in the nineteenth century). As large numbers of speakers gradually accumulate and pass on such reaccentuations, as these reaccentuations more or less converge in the somatic memories of whole populations of language users, the already-spoken-about quality of the word grows and shifts in their memories. This process is an essential factor in shaping a word’s evolution.

Double-voicing 6. This internal dialogism of language cannot be eradicated, but it can be rhetorically suppressed or concealed, and typically is when speakers want to give the impression that the language they are using to communicate with is the transparent vehicle of truth or fact (scholarly discourse), or the vehicle of their individual ideas alone (the traditional notion that ‘‘style is the man’’). This is the first hierarchical level of the model that Bakhtin developed in his Dostoevsky book; I will be taking the model as Part II’s signal analytical framework for the working-out of specific cases of iterability. In the Dostoevsky book Bakhtin called this level ‘‘direct, unmediated discourse directed exclusively toward its referential object, as an expression of the speaker’s ultimate semantic authority’’ (1984: 199). Of course it is not really ‘‘direct’’ or ‘‘unmediated,’’ nor is it ‘‘directed exclusively toward its referential object’’; the internal dialogism of language makes such a thing impossible. All utterances are mediated by dialogue, both the past dialogues in which they have accreted what meaning they have and the current dialogue that is giving significant shape to the speech acts being performed in it; this means that even utterances that seem to be directed exclusively toward an object are attempting to suppress or conceal their directedness toward a listener,

Double-voicing 107 toward an answer-word. It is a bit confusing, again, to label utterances on this level ‘‘direct referentially oriented discourse,’’ as if they really were direct and referentially oriented; but Bakhtin makes clear in the text that what he means is that the speaker thinks of his or her utterance in this way; that the speaker does not attempt to manipulate the internal dialogism of every utterance: Direct referentially oriented discourse recognizes only itself and its object, to which it strives to be maximally adequate. If in the process it imitates someone or learns from someone, this does not change things in the slightest: that is merely the scaffolding, which is not incorporated into the architectural whole even though it is indispensable and taken into account by the builder. The act of imitating someone else’s discourse and the presence of various influences from other people’s words, while recognizably clear to the historian of literature and to any competent reader, do not enter into the project that discourse has set itself. If they do enter in, that is, if the discourse itself does contain a deliberate reference to someone else’s words, then again we would have before us discourse of the third type, and not the first. (Bakhtin 1984:187) Bakhtin calls this level and the next ‘‘single-voiced discourse’’ (1984: 189): discourse where the speaker does not attempt to exploit the multiplicity of voices inherent in all dialogical utterances in any way but attempts to convey only a single voice, his or her own in this first type, that of another person in the second. 7. The internal dialogism of utterances is also typically suppressed or concealed when speakers want togive the impression that they are conveyinganother person ’s words without distortion or change ^ in quoting, for example, or even paraphrasing, or translating. Under this category we would find the traditional ideals for journalism, academic scholarship, and translation. Bakhtin uses this level, which he calls ‘‘objectified discourse,’’ to describe represented speech in novels, speech made to sound purely like the voice of a single character, or even like that of a characterized narrator. Obviously, the voice of a character will be saturated with the author’s voice as well, and when the author’s voice is filtered through that of a characterized narrator, it is common for the character’s voice to contain elements of both the author’s and the narrator’s voices; but in this type Bakhtin wants to trace the maintenance and creation of the impression that no other voices get mixed into the character’s speech, that the character speaks like a real person, a person conceived monadically (traditionally) as a separate individual, not one constituted (as Bakhtin would insist we all are) through dialogue. Objectified discourse, in other words, like direct referentially oriented discourse, is an idealized illusion. This is the illusion that translators are traditionally expected to create as well: that their voices do not mix with those of their source authors (though of course they do), that in their translation the source author speaks with a single voice, his or her own. Bakhtin (1984: 188) provides

108

Iterations (the performative back-story)

other examples: ‘‘the scholarly article ^ where various authors’ utterances on a given question are cited, some for refutation and others for confirmation and supplementation,’’ and drama, which involves ‘‘not a clash of two ultimate semantic authorities, but rather an objectified (plotted) clash of two represented positions, subordinated wholly to the higher, ultimate authority of the author.’’ 8. Most common in ordinary speech, however, and most interesting for Bakhtin as a performative linguist, is the situation where speakers do not attempt to suppress or conceal the internal dialogism of language, but rather seek to manipulate it to their own ends. Here, Bakhtin says, the controlled impression of ‘‘monologue’’ or monologism ^ the illusion of a single voice ^ is broken or weakened. Now the speaker wants his or her words to resound overtly with (at least) two voices: hence his term for this category, ‘‘double-voiced discourse.’’ This category Bakhtin splits twice, once into ‘‘passive’’ and ‘‘active’’ double-voiced discourse, the former again into ‘‘unidirectional’’ and ‘‘varidirectional’’ passive double-voiced discourse. Passive double-voiced discourse arises when the represented voice has no power to inflect the voice of the author or speaker, is passive with respect to the author or speaker who is ‘‘using’’ it for his or her own purposes; in active doublevoiced discourse there is an overt conflict between the two voices, and the ‘‘other’’ voice (the represented one) threatens to seize control from the author or speaker. a. Unidirectional passive double-voiced discourse Bakhtin associates with stylization, which Morson and Emerson (1990:150^1) describe as follows: The stylizer adopts the discourse of an earlier speaker or writer whose way of speaking or writing is regarded as essentially correct and in accord with the task to be accomplished. A critic revives a style inherited from a previous generation he admires, or a writer adopts a style marked as belonging to a particular earlier school because he has concluded that he agrees with that school. One might ask why this discourse need be considered double-voiced and dialogic in the second sense, if the two speakers want to do essentially the same thing. The answer is that, as we have seen, agreement, no less than disagreement, is a dialogic relation. There is a radical and qualitative difference between, on the one hand, one speaker being consistent with himself, and on the other, two speakers who happen to agree, each from his own perspective. As we will see below, stylization as ‘‘unidirectional passive double-voiced discourse’’ is a very useful concept for explaining the particular coming together of two voices in translation, especially when the translator essentially likes and supports the source author’s voice and wishes to transmit it as unchanged as possible ^ and, of course, when the translator is willing to present the translation as a translation, as something produced, and not, as is sometimes expected or demanded, as the original work somehow magically transported into the target language.

Double-voicing 109 Bakhtin (1984: 190^1) also detects unidirectional passive double-voicing in various kinds of narration (narrator’s narration, author’s narration, Ich-Erza«hlung, etc.): Here, too [as in stylization], someone else’s verbal manner is utilized by the author as a point of view, as a position indispensable to him for carrying on the story. But the shadow of objectification falling on the narrator’s discourse is much denser than in stylization, and the conventionality much weaker. Of course, the degree of both may vary greatly. But the narrator’s discourse can never become purely objectified, even when he himself is one of the characters and takes upon himself only part of the narration. His importance to the author, after all, lies not only in his individual and typical manner of thinking, experiencing, and speaking, but above all in his manner of seeing and portraying: in this lies his direct function as a narrator replacing the author. Therefore the author’s attitude, as in stylization, penetrates inside the narrator’s discourse, rendering it to a greater or lesser degree conventional. The author does not display the narrator’s discourse to us (as he does the objectified discourse of a hero), but utilizes it from within for his own purposes, forcing us to be acutely aware of the distance between him and this alien discourse. b. In stylization and narration the double-voicing is ‘‘unidirectional’’ in the sense that the author/speaker and the other voice they are representing tend in the same direction: they agree, they’re pulling on the same oar, they want to say basically the same things, just in slightly different voices. Bakhtin’s primary example of varidirectional passive double-voicing is parody, where the author or speaker uses the other voice in order to make fun of it, or of something that the other voice holds dear. In speech this is often done through tonalization: the other’s words are repeated verbatim but in a sarcastic or ironic tone that makes it very clear to listeners that the speaker finds the other’s utterance absurd, ridiculous, stupid, ignorant, dangerous, whatever. Another important realm of varidirectional passive double-voicing is the unreliable narrator in fiction, whose narrative ‘‘speech’’ is characterized (thus ‘‘objectified’’on the second level) and ‘‘stylized’’ (thus passive on the third level), but also clearly distinguished from the voice of the author, who may despise the values represented by the narrator, may be ridiculing them, or at the very least making it clear that the narrator’s voice is not to be trusted as a source of narrative truth or fact. Bakhtin follows the Russian formalists in calling this type of narration ‘‘skaz’’ (and Bakhtin’s English translators leave the Russian word untranslated); English examples might include Ring Lardner’s boob narrators (for discussion see Robinson 1992), or Nabokov’s Kinbote in Pale Fire. These are ‘‘varidirectional’’ because the speaker’s voice and the parodied voice (or the author’s voice and the unreliable narrator’s voice) tend in different directions, and ‘‘passive’’ because the parodied voice has no power to disrupt that of the speaker who is parodying it.

110

Iterations (the performative back-story)

c. When the represented voice gains that power and begins to disrupt the speaker, Bakhtin calls the double-voicing that results ‘‘active’’: ‘‘in stylization, in the narrated story and in parody the other person’s discourse is a completely passive tool in the hands of the author wielding it. He takes, so to speak, someone else’s meek and defenseless discourse and installs his own interpretation of it, forcing it to serve his own new purposes. In hidden polemic and in dialogue, on the contrary, the other’s words actively influence the author’s speech, forcing it to alter itself accordingly under their influence and initiative’’ (1984: 197). This category has always seemed a bit confusing to me: as long as double-voicing is understood hierarchically, as something a speaker or author does to other voices, it seems unrealistic to imagine one of those other voices rising up in rebellion against the dominant one. In Bakhtin’s discussion of the passive types the author/speaker is a ‘‘he’’ and the other discourse ‘‘he’’ is manipulating is an ‘‘it’’; in his description of this active type here both voices become ‘‘its,’’ but Bakhtin still retains the hierarchical distinction between ‘‘the author’’ and ‘‘the other,’’ making it seem as if what is happening now is that an ‘‘it,’’ the other’s voice, is forcing the ‘‘he,’’ the speaker or author, to talk in a different way.

Pandemonium Based on Daniel Dennett’s neurophilosophical theories in Consciousness Explained (1991), in fact, it would make more sense to run Bakhtin’s hierarchy the other way: to flip it on its head and see active double-voicing as the primary or default form, the ‘‘natural’’ form of speech. In this view it would be the most common thing in the world for a multitude of voices to rise up within the speaker, all of them his or her own (of course, because they rise up from within), and all of them other voices as well, voices from somewhere (of course, because as Bakhtin insists we get all of our words and phrases from other people), and they mix and mingle in his or her words, vying for ascendancy. As Dennett (1991: 304) writes: We have scant access to the processes by which words ‘‘occur to us’’ to say, even in the cases where we speak deliberately, rehearsing our speech acts silently before uttering them. Candidates for something to say just spring up from we know not where. Either we find ourselves already saying them, or we find ourselves checking them out, sometimes discarding them, other times editing them slightly and then saying them, but even these occasional intermediate steps give us no further hints about how we do them. We just find ourselves accepting or discarding this word and that. If we have reasons for our judgments, they are seldom contemplated before the act, but only retrospectively obvious. (‘‘I was going to use the word jejune but stopped myself, since it would have sounded so pretentious.’’) So we really have no privileged insight into the processes that occur in us to get us from thought to speech. They might be produced by a pandemonium, for all we know.

Double-voicing 111 And Dennett comes to use this metaphor of the ‘‘pandemonium,’’ the place of all the demons, to image forth the way utterances arise within us, as if ferried out by competing word-demons (daimones in the pagan Greek sense, not devils): We can suppose that all of this happens in swift generations of ‘‘wasteful’’ parallel processing, with hordes of anonymous demons and their hopeful constructions never seeing the light of day ^ either as options that are consciously considered and rejected, or as ultimately executed speech acts for outsiders to hear. If given enough time, more than one of these may be silently tried out in a conscious rehearsal, but such a formal audition is a relatively rare event, reserved for occasions where the stakes are high and misspeaking carries heavy penalties. In the normal case, the speaker gets no preview; he and his audience learn what the speaker’s utterance is at the same time. (1991: 238) ‘‘In the Pandemonium model,’’ Dennett writes a few pages later (241), ‘‘control is usurped rather than delegated, in a process that is largely undesigned and opportunistic; there are multiple sources for the design ‘decisions’ that yield the final utterance, and no strict division is possible between the marching orders of content flowing from within and the volunteered suggestions for implementation posed by the word-demons.’’31 There is, in other words, no such thing as a single ‘‘thinker’’ in the brain that ‘‘thinks thoughts,’’ intends meanings, and then starts planning how to express them in the most effective and accurate way possible. Thus also, there is no single ‘‘speaker’’ who marshals all the ‘‘other’’ voices at ‘‘his’’ disposal and structures them tidily for public consumption; or perhaps, to put this in more Bakhtinian terms, there are individual human ‘‘speakers,’’ but they too are internally dialogized, double-voiced or rather multiply voiced, and only give the impression of speaking as single beings because their words come out of the mouth or fingertips attached to a single body. This revised Bakhtinian model for speech would suggest that his more extreme descriptions of active double-voicing are in fact the basic state of human speech production, and the ‘‘lower’’ rungs on his hierarchy, the varidirectional and unidirectional passive forms and the objectified and direct unmediated forms, are the result of concerted rational self-control and image-management. A paragraph like this from the Dostoevsky book, for example, describing active double-voicing, sounds to me like human dialogical business-as-usual, with only a few slight modifications: Internally polemical discourse ^ the word with a sideward glance at someone else’s hostile word ^ is extremely widespread in practical everyday speech as well as literary speech, and has enormous style-shaping significance. Here belong, in everyday speech, all words that ‘‘make digs at others’’ and all ‘‘barbed’’ words. But here also belongs all self-deprecating

112

Iterations (the performative back-story) overblown speech that repudiates itself in advance, speech with a thousand reservations, concessions, loopholes and the like. Such speech literally cringes in the presence or the anticipation of someone else’s word, reply, objection. The individual manner in which a person structures his own speech is determined to a significant degree by his peculiar awareness of another’s words, and by his means for reacting to them. (Bakhtin 1984:196)

All I would want to edit in this passage is that last sentence: the terminus of Bakhtin’s own argument would seem to be that the individual manner in which a person structures his or her own speech is determined to a significant degree not by his or her awareness of another’s words, but simply by the influence of other people’s words over his or her thinking, aware or not. Bakhtin’s sentence seems itself to be heading in this direction: ‘‘the individual manner in which a person structures his own speech’’ sounds supremely rationalist, as if the speaker were in total control, structured his speech in an individual manner; but this proud syntactic subject is converted by the passive construction ‘‘is determined’’ into the implicit direct object of an action performed not by the person but by something inside the person, both ‘‘awareness’’ and ‘‘means.’’And ‘‘means’’ is even less conducive to a rationalist understanding of this process than ‘‘awareness’’ is: the speaker happens to have ‘‘means for reacting’’ to other voices in his or her head, and those means to a significant degree determine the individual manner in which s/he structures his or her own speech. What makes the manner individual is not rational control, in other words, but simply the fact that the speech production of this particular speaker is shaped by a specific collection of internalized voices and ‘‘means for reacting’’ to them. To be sure, we do exert rational control over our speech; Dennett draws our attention to this when he writes of ‘‘silently trying out’’ alternative utterances ‘‘in a conscious rehearsal’’ before actually speaking; but, he reminds us,‘‘such a formal audition is a relatively rare event, reserved for occasions where the stakes are high and misspeaking carries heavy penalties.’’ This is in fact a very Bakhtinian perspective on rational control: messes for him are experientially primary, and order and control are ideals to which we but imperfectly strive. Hence the importance for him of starting with the act, with the performance, with the event, with dialogue, rather than with perfectly ordered speech conceived transcendentally as‘‘pure structure,’’as the constative linguists do. And so in constructing a more ‘‘performative’’ revision of his model of double-voicing, I do it very much in the spirit of Bakhtin’s own approach, assuming that he would have presented his ideas in this way himself had he worked through them to the end: First type: overt multiple voicing . the speaker’s internal dialogue or pandemonium of voices is frankly made public, re£ected openly in speech or writing . Bakhtin’s active type: the sideward glance at someone else’s hostile word, the word with a loophole, etc.

Double-voicing 113 Second type: overt multiple voicing with a hierarchy imposed . the speaker gives the impression of being in control of the other voices that striate his or her speech . Bakhtin’s passive type, both varidirectional (parody and the use of an unreliable narrator) and unidirectional (stylization and the use of a reliable narrator) Third type: covert multiple voicing with a hierarchy imposed . the speaker gives the impression of speaking with only a single voice, either his or her own or someone else’s . Bakhtin’s objecti¢ed and direct unmediated types This revised model begins with the performative subjectivity of the speaker and adds rational constraints ^ more or less the way things work in the real world, as Bakhtin himself would doubtless insist. The revised model thus avoids the unfortunate impression that Bakhtin’s formulation gave that it is possible to speak in a single voice, that direct unmediated speech is possible, that a speaker or writer can represent another person’s speech without retonalizing it ^ and thus that double-voicing is something superadded to speech, something extra, something perhaps a little strange. Multiple voicing in my revision of the model is fundamental to human speech; it is always with us, always a constitutive impulse for our speaking and writing. We learn through long hard practice to cover it up, to pretend to be speaking as single voices, our own, mainly, or, when we quote another person, purely as that other person; but even at best we never fully succeed, never fully purge our speech of the cacophony of other voices that make up our unspoken thoughts. And we don’t even always try: we do so, as Dennett suggests, primarily when a great deal rides on what we say, when ‘‘misspeaking carries heavy penalties.’’

Double-voiced translation This revised Bakhtinian model has such obvious applications to the study of translation that I will forebear going into them in detail; if speakers and writers always channel lots of voices into what comes out of their mouths or pens, it will not be surprising to us if translators do the same. Translators are, after all, as human as the rest of us; they wade their way through the pandemonium of speech-production just like any speaker or any writer. Clearly, however, some (perhaps even most) translators are obliged by the marketplace to conceal the welter of voices they channel through their translations. The translator of a toxic waste environmental impact statement, for example, is expected to produce a text that sounds more or less like a similar statement in the target language. This means, in my revised Bakhtinian terms, the carefully controlled reduction of a potential vocal pandemonium to what I called just above the ‘‘third type,’’ covert multiple voicing with a hierarchy imposed, where the speaker gives the impression of speaking with only a single voice, either his or her own or someone else’s. The translator works hard to

114

Iterations (the performative back-story)

give the impression of ‘‘objectifying’’ another person’s discourse, passing on the words of another without change, so that the target-language reader can read the translation under the impression that it is direct and unmediated. But of course that is just an impression, one that it requires a great deal of skill to create ^ just as it requires a great deal of skill to reduce the verbal pandemonium in your head so as to‘‘speak clearly,’’or ‘‘talk like a book.’’ There are also, however, translations that are more overtly multiply voiced. Some of these might be described in terms of the revised second type, either stylized (‘‘unidirectional’’) or parodic (‘‘varidirectional’’); others in terms of the revised first type, with hidden polemics, sideward glances, and so on. Recognizing that stylization necessarily adds at least one voice (the stylizer’s) to a discourse makes it possible to accept with equanimity the vocal shift that all translation inevitably involves; for even at its least obtrusive, translation is obviously a form of stylization. And whenever the stylizing translator has a distinctive recognizable style in the target language ^ Pope’s Homer, say, or the translations of Ezra Pound ^ it will be useful to discuss the addition of the translator’s voice to the mix in terms of Bakhtin’s theory of double-voiced stylization. Wherever the translator has a distinctive voice and deviates from the original in some striking and obvious way ^ disagrees with it, wants to create some sort of ironic distance from it ^ we move into the second-type realm of parody. Many of the poetic translations that I discussed in the ‘‘tropics’’chapter of TheTranslator’s Turn were ‘‘parodic’’ in this way, especially perhaps my propagandistic version of Guille¤n (1991: 155), which parodically exposed the poet’s sexism. But a radical version like the Zukovskys’ Catullus too, it seems to me, should be considered ‘‘parodic’’ in this way ^ not because the Zukovskys should be thought of as ‘‘making fun’’ of Catullus, but because the translation is a radically divergent revoicing of the original, and charges the text with powerfully ‘‘varidirectional’’ energies.The revised first type appears rarely; most translators are not willing to argue with the original this radically. I did a lot of it inTheTranslator’sTurn, using overt mistranslation to problematize and complicate some of the theoretical texts I was using, including, in fact, a passage or two from Bakhtin’s ‘‘Discourse in the Novel’’ (1991:103^4), which I more or less overtly pushed in the direction of my own somatic theory by revoicing Caryl Emerson and Michael Holquist’s English translation. The most obvious case of overt active multiply-voiced translation in that book, though, was a rendition of a passage from Derrida’s ‘‘Des tours de Babel,’’ where I was at once illustrating and problematizing Derrida’s own ‘‘perverse’’approach to translation: Le roi a bien un corps (et ce n’est pas ici le texte original mais ce qui constitue la teneur du texte traduit) mais ce corps est seulement promis, annonce¤ et dissimule¤ par la traduction. L’habit sied mais ne serre pas assez strictement la personne royale. Ce n’est pas une faiblesse, la meilleure traduction ressemble a' ce manteau royale. Elle reste se¤pare¤e du corps auquel cependant elle se conjoint, l’e¤spousant sans l’e¤spouser. (Derrida 1985: 237)

Double-voicing 115 The king does have a body, of course, a real one, no body-as-original-text (the metaphor that Benjamin plays with) but that which turns the original toward the translation, that which grounds the translation somatically. This body is not the cape or clothing that Benjamin and Derrida both make so much of, body as separable from soul, clothing as separable from body, words as separable from meaning, signifiers as separable from signifieds, that whole dualistic tradition going back to Augustine and Paul and Aristotle and Plato.Those clothes are only the dissembling promise or announcement of a translation that will never be, the habitual (ideosomatically programmed) promise of perfect new clothes that will never materialize and wouldn’t fit even if they did. In perfectionist theories this perfect imperfection is no weakness; it’s the resolute denial of weakness, the determined assertion of ideal strength, potency, the king’s erection under the wide folds of his cape, ideal fortitude: the perfectionist translation resembles the king’s or emperor’s new cape, which is visible only to idealists who believe strongly enough in its reality to make it seem real, but none of this is the real body I’m talking about, I’m talking about the body as somatic response. (Robinson 1991: 237) Constativists will ask, of course: why should this be considered a translation at all? From a performativist point of view, the answer is simple: in this passage I am reperforming a French text in English. Derrida’s voice surges through my performance, but so do various voices of my own. Using Bakhtin’s schema will help us to track the voices that make up any text, including the series of crosslingual reperformances I’ve been setting up here: the apparently ‘‘objectified’’ toxic waste environmental impact statement, the ‘‘stylized’’ translation of Homer by Alexander Pope, the ‘‘parodic’’ translation of Catullus by Celia and Louis Zukovsky, and my active polemic reperformance of Derrida. Bakhtin can help us see that there is a clear progression here, as the careful rhetorical control of voice slips, as more and more of the vocal pandemonium in the translator’s head gets infused into the translation. The constative insistence on rigid distinctions between translating (which reproduces the original unchanged) and adapting or parodying or rewriting does not help us track this progression. A Bakhtinian performative schema of double-voicing not only helps us view translation in terms of what the translator does; it is analytically, too, far more capacious than similar constative models. It does more analytical work.

Language change The revised Bakhtinian schema is, then, a more integrally performative model, a less vestigially constative model than Bakhtin’s original formulation. It is also one better suited to the analytical exploration of iterability, the temporal dynamic of language stability and change: this revised Bakhtinian model will enable us to trace the performative creation of language history through repeated transformative performances of ‘‘old’’ voices in ‘‘new’’

116

Iterations (the performative back-story)

dialogical contexts. How does language change? How does it retain stability in the midst of change? Where do the regularities of language come from? The somatic theory explains the physiological and emotional (experiential) bases of linguistic stability, regularity, and change: socially accepted utterances are somatically marked positive, and thus almost unconsciously retained, repeated (double-voiced, incorporated into their own speech) by individual speakers with as little change as they can manage (ideosomatics); but somatic markers are only a very rough and approximate channel of collective control, and individual speakers are always inflecting somatized speech regularities with their own personalities as well (idiosomatics). What Bakhtin’s dialogical model of double-voicing can contribute to Derrida’s iterability and my somatics is a temporal analytic, a method for tracing specific retentions and innovations in language history. (I will be exploring another such temporal analytic, based on Paul Grice’s concept of conventional implicature, in Chapter 12.) Consider the possibility, which Bakhtin might be willing (simplifying his theories slightly) to assert as a fact, that double-voicing is the prime motor of all language evolution. This would mean that everything in language, all words, all phrases, all syntactic structures, all pronunciations, all regional and class dialects (including standard ones), all sociolects and idiolects, all connotations and collocations, all pragmatic implicatures and uptakes, everything linguistic comes out of double-voicing: speakers of a language experience language dialogically, are (as selves, as thinkers, as speakers) themselves dialogized by the social use of language that surrounds them, are thus filled with the dialogized voices of other people, and, when they speak, pour those voices into their own utterances. Those utterances are constructed almost unconsciously, most often, and frequently take their utterers somewhat by surprise, but since the unconscious strategies that all speakers have developed for constructing meaningful utterances are first and foremost dialogized, they will invariably come out oriented toward the words of others: past speakers who have used these particular utterances or bits and pieces of them, the person or persons who most recently spoke in the current dialogue, and the ‘‘answer-word’’ that they will most likely provoke in that same dialogue. The more consciously and carefully and analytically the speaker constructs his or her utterances, the more they will come to resemble the idealized ‘‘sentences’’ or (better) ‘‘structures’’ studied by constative linguists, but, humans being what they are, they will never quite live up to those high transcendental ideals, and will always have to be consigned in the end, perhaps a bit wistfully, to the rubbish bin of ‘‘parole’’ or ‘‘performance.’’ But the mere fact of approaching that ideal does not mean that these utterances should be studied in terms of it; from a performative point of view they should be studied in terms of what they are doing, not of what they are failing to do. And through this dialogical process surge all the constitutive elements of language: phonemes, phonemic and allophonic variation, suprasegmental phonetic units, the use of phonetics for both meaning and tonalization (and phonetics is, after all, the most obviously corporeal of linguistic elements, the

Double-voicing 117 mode of linguistic expression that is most directly channeled through the body); morphemes, morphemic and allomorphic variation, our sense of what a word is and where its boundaries lie and what can be done with it in a sentence, how it can be molded to different discursive ends; syntagmemes, the building blocks of word order, predications, the sequencing and relationalizing of words and phrases for purposes of understandability and emphasis and other forms of expressivity; sememes, the meanings and connotations of words and phrases as we have heard other people use them; various text-linguistic structuring devices for building internal coherence and cohesion in an utterance, including various forms of backward and forward reference; styles, registers, and subregisters, jargons and argots, genres, various forms of social orientation to authority, reliability, group membership, and so on. To put that more schematically (see Table 7.1), language is constantly being (re)born or (re)created ^ iterated ^ out of double-voicing on several different levels: On the four levels described in Table 7.1, I am suggesting, all language is constantly coming into being. All structural, communicative, expressive, and social stability is constantly being maintained (enforced and reinforced) through the respective channels of double-voicing, for the specific purposes of achieving social belonging, differentiation, and understanding. The various channels of double-voicing involve repetition of past voices, the voices of other members of the language community, and that tends to conserve existing structures; but as Jacques Derrida (1988: 145) insists in the passage I set as the epigraph of Part II, it is precisely the relative stability of such systems that makes them destabilizable, and precisely the constant need to guard against destabilization that makes them relatively stable. Bakhtin (1981: 270) agrees: Unitary language constitutes the theoretical expression of the historical processes of linguistic unification and centralization, an expression of the centripetal forces of language. A unitary language is not something given [dan] but is always in essence posited [zadan] ^ and at every moment of its linguistic life it is opposed to the realities of heteroglossia.33 But at the same time it makes its real presence felt as a force for overcoming this heteroglossia, imposing specific limits to it, guaranteeing a certain maximum of mutual understanding and crystalizing into a real, although still relative, unity ^ the unity of the reigning conversational (everyday) and literary language,‘‘correct language.’’ The ‘‘project’’ (Ru zadanie) or ‘‘task’’ (Ru zadacha)34 of unifying or stabilizing a language is a double-voiced one carried out by all of its speakers, through their own conservative and conformist impulses, as well as (but only secondarily) by various language academies and other linguistic authorities; and because of the nature of double-voicing, it is always only more or less successful. Stable structures are destabilized in the very act of their most careful and even

2. telling stories, telling jokes, dramatizing events, mimicking others, bragging, addressing a superaddressee,32 arguing, agreeing, listening supportively, consoling, etc.

3. incorporating into your remarks ‘‘hooks’’ from the last thing somebody else said, building anticipations of others’ responses (objections, criticisms, arguments, admiration, boredom) into your utterances 4. building utterances out of words, phrases, structures heard from other people; making the forms of those structures conform to those used by others

2. expressive, individualistic: persuasiveness, impressiveness, uniqueness, originality, cleverness, vivacity, attractiveness, friendliness, emotional sensitivity, emotional strength, emotional interdependence, humorousness, etc.

3. communicative: dialogical orientations toward the utterances of others in the current conversation

4. making yourself understood by using commonly accepted forms in a commonly accepted way

1. making yourself the same as others by conforming your voice to those of socially admired or respected speakers: parents, teachers, preachers, politicians, scholarly authorities, news commentators, favorite characters from television or ¢lms 2. making yourself di¡erent from others by transformatively incorporating into your speech the voices of widely accepted models of attractive uniqueness: raconteurs, humorists, members of this or that elite (upper class, intelligentsia, etc.), representative men and women, etc. 3. making yourself understood by manipulating conversational strategies for e¡ective uptake

1. quoting, alluding, paraphrasing, summarizing, modeling, imitating, stylizing

1. social, conformist: authority, reliability, integrity, authenticity, knowledgeability, competence, intelligence, altruism, sympathy, group solidarity, group conformity, sense of humor, etc.

4. structural: structures of sentences (syntax), meanings of words (semantics), pronunciations of sounds (phonetics), etc.

Aim or rationale

Double-voiced channels

Values internalized

Table 7.1

Double-voicing 119 worshipful repetition, and must be restabilized, which again leaves them open to destabilization. By the same token, however, the project or task of transforming a language is also a double-voiced one carried out by all of its speakers, through their more restless or innovative impulses. Not only are the members of a language community constitutionally incapable of transferring stability without change; they are constitutionally incapable of wanting to transfer stability without change for very long. Everyone wants stability, and stability is more or less what they get; but nobody wants boredom, and the best defense against boredom is a slight new twist or spin on the same old thing. The result is constant innovation within relative stability. Tonalizations change rapidly and radically within a more or less stable field defining the impact or uptake of tones. Pronunciations vary significantly within the same family, and even from hour to hour within the speech of the same speaker; most of these variations dissipate without taking hold, but inevitably some do have a more lasting or at least widespread impact on the language (‘‘nucular’’ for ‘‘nuclear,’’ ‘‘supposably’’ for ‘‘supposedly’’). The meanings of words are constantly subject to critical semantic shifts, as homophonic drift confuses native speakers (‘‘We’re presently [at present] experiencing a downturn,’’ ‘‘I bought this specially [or especially] for you’’), as they bypass overfine distinctions (‘‘We’ll need to share it between [or among] the three of us’’) or drop words they feel are extraneous (‘‘It’s fine as far as John [is concerned]’’), or as they get tired of meaning the same old things with the same old words and use them figuratively or otherwise inventively (‘‘You know damn well,’’ the hero of the Kevin Smith movie Mall Rats yells at his friend and friend’s ex-girlfriend in the climactic game show, ‘‘that you’re retarded for each other,’’meaning madly in love with each other). The constative linguistics textbooks tell us that syntactic structures change most glacially, but any attentive speaker of a language can name a dozen areas where syntactic shifts are in progress: ‘‘I wish I knew what he was saying about you and I, you and me, you and myself,’’ ‘‘I wish he was home, I wish he were home,’’ ‘‘You should have laid it on the table, lain it on the table, put the damn thing on the table.’’ The most tolerant language-lovers among us, the ones who pride ourselves on keeping up with language change and dismissing prescriptivists with superior contempt, could probably name a handful of new usages that grate on our ears, and that we just know we’ll never get used to, as when my teenaged daughters say to their friends ‘‘So I’m like ‘What’d he say?’ and she’s like ‘I don’t remember.’ ’’35 This isn’t the last model I will be offering for linguistic evolution; as I say, I have another in Chapter 12. But it is I think a powerful and useful one, one that does explain all linguistic phenomena in a thoroughly performative way.

Part III

Implicatures (performative uptake) Transactions between a person and other persons or things closely connected with him are liable to be very different as regards their concomitants and results from the same sort of transactions involving only remotely connected persons or things; the concomitants and results, for instance, of my finding a hole in my roof are likely to be very different from the concomitants and results of my finding a hole in someone else’s roof. Information, like money, is often given without the giver’s knowing to just what use the recipient will want to put it. If someone to whom a transaction is mentioned gives it further consideration, he is likely to find himself wanting the answers to further questions that the speaker may not be able to identify in advance; if the appropriate specification will be likely to enable the hearer to answer a considerable variety of such questions for himself, then there is a presumption that the speaker should include it in his remark; if not, then there is no such presumption. H. Paul Grice,‘‘Logic and Conversation’’ (1988: 38)

8

Conversational implicature

Part II was a series of theoretical approaches to the unsolved problem that Austin left us: perceiving that the novelist or the poet or the actor on the stage is doing something, but something other than actually promising, betting, and so on, and not knowing what to do with that observation analytically, he called such speech acts ‘‘parasitical’’ and set them to one side. The section’s organizing concept of iterability was born out of Jacques Derrida’s signature deconstructive move of taking the remainder or excluded ‘‘supplement’’of someone’s argument and making it into the main set: not only is the parasitic not to be excluded from the ‘‘serious’’ study of speech acts, it becomes the key to understanding all speech acts. Still, as I have noted several times, iterability as Derrida develops it is a general philosophical category, not a methodological tool for analyzing specific ‘‘misuses’’or ‘‘misperformings.’’36 Saying that all language is based on the possibility of such misuse is an important step toward a performative reframing of linguistics, as it showcases the performances that make it possible for us to know what we know, see what we see, say what we say: Part II was, you will recall, specifically about the ‘‘back-story’’of performance, comparable to the exposition in a novel where the narrator goes back and gives the reader a summary of the main events leading up to the beginning of the tale as the narrator is telling it. But iterative/somatic/polyphonic theory does not help us solve the problem of uptake, the difficult matter of how interpreters can understand what Austin wanted to call ‘‘nonserious’’or ‘‘parasitic’’ speech acts.

The cooperative principle and its maxims It is precisely this kind of ‘‘misperformative’’ analytical tool that the English language philosopher H. Paul Grice (Austin’s colleague at Oxford up until Austin’s death in 1961) attempted to provide in the series of lectures he delivered at Harvard in 1967 ^ the same series, in fact, the William James lectures, at which Austin had presented his theory of speech acts twelve years earlier.37 Grice poses the question: how is it possible for us to imply things, to convey intended meanings that we do not make explicit? How do our interlocutors manage to understand us correctly? This has been a recurring problem for

124

Implicatures (performative uptake)

constative linguists and language philosophers, since their transcendental (systemic, rule-governed) model requires that communication only be imagined as possible if it obeys the rules, and various evasive speech acts fail to do so ^ and yet succeed as speech acts. Grice argues that implied speech acts do in fact break the rules, but in a controlled fashion. We all bring certain assumptions and expectations to speech situations, Grice says, and it is possible for us to manipulate those assumptions and expectations so as to make ourselves understood indirectly. By breaking the rules on one rather superficial level, we signal to our listeners that we are in fact obeying them on a deeper level ^ and thus get understood. Grice identifies three of these hierarchical levels, in fact, and postulates their stability and universality in order to explain individual speakers’ability to manipulate them ‘‘transgressively’’ and still be understood. In this insistence on conceptual constancy Grice remains very much in the constative camp, and has, in fact, thanks to the work of Levinson (1983), Sperber and Wilson (1986), and others, been picked up by the constative mainstream as the father of ‘‘linguistic pragmatics’’ ^ the science of the relation between signs and their interpreters, which constative linguists investigate by idealizing interpreter and context types. But in his constancy-threatening concern with transgression, with breaking the rules, he is also patently pushing at the soft sides of his constative enchilada. Here is the paragraph from Grice’s ‘‘Logic and Conversation’’ outlining the three levels, with rationalism (rational talk exchanges) at the bottom (and first in the paragraph), general cooperativeness (directed toward a common purpose) in the middle, and specific cooperative rules or maxims at the top: Our talk exchanges do not normally consist of a succession of disconnected remarks, and would not be rational if they did. They are characteristically, to some degree at least, cooperative efforts; and each participant recognizes in them, to some extent, a common purpose or set of purposes, or at least a mutually accepted direction. This purpose or direction may be fixed from the start (e.g. by an initial proposal of a question for discussion), or it may be so indefinite as to leave very considerable latitude to the participants (as in a casual conversation). But at each stage, SOME possible conversational moves would be excluded as conversationally unsuitable.We might then formulate a rough general principle which participants will be expected (ceteris paribus) to observe, namely: Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged. One might label this the COOPERATIVE PRINCIPLE. (1989: 26) He then goes on to itemize the ‘‘maxims’’ of the cooperative principle (CP), based on Kant’s four categories of quantity, quality, relation, and manner. In the category of quantity he identifies two maxims: ‘‘Make your contribution

Conversational implicature 125 as informative as is required (for the current purposes of the exchange)’’ and ‘‘Do not make your contribution more informative than is required’’ (26). In the category of quality he identifies a supermaxim,‘‘Try to make your contribution one that is true,’’ and two maxims, ‘‘Do not say what you believe to be false’’ and ‘‘Do not say that for which you lack adequate evidence’’ (27). In relation he sees only one maxim, ‘‘Be relevant’’ (27). In manner, finally, under the supermaxim ‘‘Be perspicuous,’’ he lists four maxims: ‘‘Avoid obscurity of expression,’’ ‘‘Avoid ambiguity,’’ ‘‘Be brief (avoid unnecessary prolixity),’’and ‘‘Be orderly’’ (27). Grice’s central claim is that we make ourselves understood by deliberately flouting the maxims of the CP ^ breaking the rules of rational, cooperative discourse in such a way as to ensure that our listeners will be able to figure out just what we’re trying to say.38 Thus, for example: ‘‘A is writing a testimonial,’’ Grice writes, ‘‘about a pupil who is a candidate for a philosophy job, and his letter reads as follows: ‘Dear Sir, Mr X’s command of English is excellent, and his attendance at tutorials has been regular.Yours, etc.’’’ (33). Clearly, A provides too little information, and can thus be taken as deliberately flouting the first maxim of quantity,‘‘Make your contribution as informative as is required.’’ He’s breaking a cooperative rule, being apparently uncooperative; but the very fact that he bothered to write the recommendation in the first place indicates that he is in actuality cooperating. Therefore, X’s prospective employer will surmise, A must be trying to tell me something indirectly. He’s implying or insinuating something; his explicit statement has an implicit meaning that deviates from the explicit one, and the unstated implication is what I’m supposed to believe. The missing information regards X’s proficiency at philosophy; by omitting that, A must be implying that X is no good at philosophy. Grice calls this ‘‘conversational implicature,’’ a situationally creative channel of implicit communication that he opposes to the dead metaphors and other habitualized phrases of ‘‘conventional implicature.’’ (Grice does not really theorize conventional implicature; he just mentions it in passing. I will be theorizing it in Chapter 12.) Metaphorical language was, you will recall, one of the slippery slopes of speech-act theory as theorized by Austin and Searle: ‘‘Walt Whitman does not seriously incite the eagle of liberty to soar.’’ Dead metaphors (or generally tropes) do not trouble Austin and Searle much precisely because, in Grice’s terms, their implicature (what the utterer means by them, how they should be taken) has been conventionalized; the interpretive pathways in our neural networks are so well worn by frequent repetition that we don’t even notice the trope when it hits us, don’t flinch as it passes by. Dead metaphors have all the phenomenological impact of what Austin and Searle valorized as serious, literal utterances. Metaphorical language only became a problem ^ and got declared ‘‘non-serious’’ or ‘‘parasitic’’ ^ when it was situationally creative, innovative, invented for the purpose of the conversation. But where Austin found himself unable to explain how an utterance could be at once metaphorical and ‘‘serious,’’ Grice has a plausible theory: the clash between the CP and one of its maxims, between our assumption that the speaker is speaking rationally and cooperatively and our perception that the

126

Implicatures (performative uptake)

speaker is flouting one of the CP’s maxims, compels us to make an interpretive leap. If I say to you, in one of Grice’s faintly goofy examples, ‘‘You are the cream in my coffee,’’ I am flagrantly flouting the first maxim of quality,‘‘Do not say what you believe to be false.’’ ‘‘I am not cream,’’ a literalist might sniff,‘‘and I am most certainly not in your coffee.’’ But if you assume that I am being cooperative, you will have to move past the obstacle this interpretation raises to understanding. Calling you cream literally would be disruptive, and on the assumption that I’m not trying to be disruptive you try out a metaphorical interpretation: I like cream in my coffee; I’m implying that I like you. Grice also notes that this metaphorical interpretation can in some cases then be reinterpreted ironically: if I’m furious with you and have just spent fifteen minutes chewing you out, and you say meekly,‘‘Does that mean you don’t like me?,’’ I might well reply disgustedly, ‘‘No, you’re the cream in my coffee.’’ Here we have two levels or stages of conversational implicature. If I am being cooperative, I can’t mean it literally (‘‘You are a dairy product’’); but then the CP requires some degree of consistency, and in the overheated conversation we’re having, a ‘‘straight’’ (or what Austin and Searle might reluctantly call ‘‘serious’’) metaphorical interpretation like ‘‘I like you’’ would be inconsistent with my anger. Therefore, I must mean my utterance ironically: ‘‘I don’t like you,’’or perhaps only,‘‘Your behavior makes it hard for me to think of you right now in terms of liking.’’ This obviously solves Austin’s problem, and in a significantly different way than Derrida. Derrida’s solution was to articulate the structure of utterances ^ to reify utterances as inwardly divided, and overlapping across the dividing line, between past and future, old and new, anticipation and repetition. Grice shifts his calculus to the interpretive activity of the listener; and while he is at pains to constatize (stabilize, formalize, idealize) the hierarchical assumptions underlying that activity, he is nevertheless willing to allow the listener considerable latitude in the making of interpretive leaps. Grice moves decisively into the arena of performative linguistics when he assumes that implicature carries with it no guarantee of understanding; when he assumes that in order to be understood the speaker must perform his or her implied meaning convincingly, and that some performances are inevitably, in an imperfect human world, more successful than others.

Constative survivals Still, Grice should not be thought of as simply turning over the constative traces with performative abandon. He remains very much a constative rationalist/idealist/formalist thinker in his assumptions that rationality and cooperation provide the necessary foundation for conversation, and that cooperative maxims are stable universals.39 And it is here, with these assumptions, that I want to begin my critique of Grice. First, is it really true that for a conversation to‘‘work’’ ^ to serve whatever purpose its participants want it to serve in any given context ^ it must be rational?

Conversational implicature 127 In fact it isn’t.We all, at various times and in a large variety of different contexts, find ourselves interpreting irrational discourse, sudden irruptions of something or other in our friends’ conversational contributions, bursts of petulance, semiconcealed jabs, lies, pretenses, non sequiturs, and so on. The better we know the people who speak irrationally, the better able we are to interpret what they’re doing conversationally. Irrationality is not the sort of absolute block to conversational success that Grice makes it out to be. What about cooperation, then? Must we assume that a participant in a conversation is cooperating with us (or with the implicit or explicit purpose or direction of the conversation) before his or her contribution will make sense? It’s ten o’clock in the evening, and your three-year-old is practically asleep on her feet, yawning, bursting into tears at the slightest provocation; try to get her into bed, however, and she explodes with the rage of three-year-old exhaustion: ‘‘I hate you, I hate you, I hate you!’’ This is uncooperative discourse; but is it understandable? Of course it is. She’s sleepy. Or this: a colleague in a faculty meeting is fighting a proposed curricular change tooth and nail; half the department is offering him plausible rationales for the change, each speaker trying to come up with an effective or operative reason, one that will break through the brick wall of his resistance and make him reconsider (the administration will like it, it’s more consistent with our current curriculum or department image, the students are clamoring for it, etc.), but nothing works. The reasons he offers for opposing it are rubbish and everyone knows it; they are transparent subterfuges. Everyone knows, though no one says it aloud ^ indeed no one has actually even discussed it with him ^ that the real reason for his opposition, and his refusal to cooperate with the current purposes of the exchange, is that he is afraid the proposed change will channel students away from a course he has taught for twenty years. He is not ‘‘flouting maxims;’’ he is not, to shift to the Austinian vocabulary that I echo in my subtitle to Part III, manipulating uptake. He is raising a conversational fortress around himself and doing his pigheaded best to obstruct accurate interpretation of his conversational intentions; but that accuracy is obtained anyway (though not inevitably). The fact is, participants in a conversation will make an interpretive leap whenever easy access to understanding is blocked ^ whether the blockage is intentional or not,‘‘guided’’ or not. We are socialized to crave understanding of others’ utterances, and to keep actively constructing images of understanding until we are (at least provisionally, pragmatically) satisfied that we have understood well enough to go on. We don’t need to ground these utterances in the transcendental, as constative linguists assume; people are a lot better at making sense of things without stable reference to an unchanging ground than constative linguists would have us believe. We do, of course, quite often seek stable grounds for our interpretations; the fact that we can’t always find them, and that we can never find perfectly stable grounds, doesn’t rid us of the constant (perhaps even obsessive) quest for certainty. And Grice is sufficiently attracted to what Lecercle (1990) calls the

128

Implicatures (performative uptake)

linguistic ‘‘remainder,’’ to a more open-ended and open-minded tolerance for the unsystematizable in speech, to realize that this quest is more of a nervous longing than a foundational condition for understanding: he hints, after all, that the longing is ‘‘normal’’ or normative, which is to say socially controlled, when he says that ‘‘our talk exchanges do not normally consist of a succession of disconnected remarks,’’ or that ‘‘they are characteristically, to some degree at least, cooperative efforts.’’ In Western society, since at least the rise to power of the middle classes in the Reformation and Renaissance, the expectation of conversational rationality and cooperation is ‘‘normal,’’ indeed normative, controlled by the liberal ideology that has dominated Western society during those centuries and continues to dominate it. In Madness and Civilization (1961/1982) Michel Foucault traces the gradual imposition of rationalist ideological norms on conversation through the seventeenth and eighteenth centuries, and the gradual social transformation of those who failed to satisfy those norms in their conversation into the ‘‘insane.’’ Since that time the socio-ideological pressure exerted on speakers to speak rationally and cooperatively has given us (relatively) new ground to stand on in our quest for certainty, for understanding; but that ground is not transcendental. It is not universal. It is socially controlled and thus historically contingent.40

The maxims as cultural habits The same can be said for Grice’s maxims, so richly redolent of the white masculine professional middle-class culture of the past hundred-odd years in England and the U.S. (Grice was an English-born philosopher whose life spanned the twentieth century: he taught at St. John’s College, Oxford, from 1938 to 1967, the year of his William James lectures and the year he moved to Berkeley, where he served as professor of philosophy until his death in 1988.) As we will see in greater detail in Chapter 9, in connection with various constative applications of Grice to translation studies, his maxims have been widely criticized as not universal but very culture-bound. They are the emotionally invested (ideosomatic) ideals of a race and class dedicated to prosperity and propriety, a class and professional group dedicated to ethical and intellectual reliability, and a professional group and gender dedicated to protection against emotion and self-knowledge. Grice himself notes, in fact, that the maxims are learned: ‘‘It is just a wellrecognized empirical fact that people do behave in these ways; they learned to do so in childhood and have not lost the habit of doing so; and, indeed, it would involve a good deal of effort to make a radical departure from the habit. It is much easier, for example, to tell the truth than to invent lies’’ (29). The maxims are cultural habits, not universal principles governing human behavior. But somehow Grice manages to forget that cultures differ, that people learn different behavioral norms in childhood. Thus when he notes that ‘‘it would involve a good deal of effort to make a radical departure from the habit,’’ he explains that fact not through the relative tenacity of ideosomatic marking (Chapter 5)

Conversational implicature 129 but through a tacit appeal to universalism: it just is easier to tell the truth than to invent a lie. It is easier for all humans in all cultures, he’s implying, because the maxim of quality is a universal, an expression of an ideal symbolic logic that governs all conversation everywhere. The cooperative principle and its maxims are not, clearly, Grice’s fantasies; they exist, and they have a powerful shaping influence on our lives. They are, I would suggest, the ideological code or episteme, or one section of the episteme, that rules the men (and nowadays a few women) who rule contemporary Western society. Provide the information required and no more: don’t talk too much; don’t run your mouth; don’t indulge in empty conversational ‘‘stroking’’ or other phatic nonsense; stick to the information, the facts; don’t wander off on speculative or hypothetical tangents about emotions or the afterlife or what Joe’s wife might or might not be doing behind his back. Don’t write big sprawling florid chatty books like this one, and hope to have them published by an academic press. Tell the truth: don’t invent things; don’t tell stories that might have happened, or that should have happened, or that explain what really happened better than the truth would have done; don’t pretend that a story someone else told you actually happened to you; don’t shade the truth to protect your own or someone else’s feelings. Be relevant: don’t deviate from the course of conversation dictated by others; don’t let your imagination run along lines that others consider irrelevant; don’t analyze a speaker’s ideas or behavior in terms other than those he himself has proposed (if for example he insists that the conversational exclusion of excess verbiage and made-up stories as irrelevant is ‘‘natural’’ and ‘‘human’’ and thus ‘‘universal,’’ it will be irrelevant and impertinent of you to launch a feminist or postcolonial critique of that idea). Be perspicuous: reduce the complexity of everything you think and feel to a simplified syllogistic clarity; make sure that you have completed the process of syllogistic reduction before you open your mouth, so that you don’t find yourself groping and stumbling your way to an expression of what you really want to say; don’t say silly or playful or iconoclastic things that tend to muddy the conversational waters. The overarching rule is: obey these rules (at least at some level, even if superficially you appear to be flouting them) or become hopelessly ‘‘other,’’ marginal, fringe, partially defective ^ un(der)employed, poor, homeless, exploited, oppressed. Grice says that disruptiveness ^ read ‘‘deviation from these ideological norms’’ ^ makes conversation impossible; I say that disruption only makes conservative or conformist conversation impossible.41 Having stumbled upon the ideological contingency of these maxims, though, what shall we do with them? Supplement them with parallel maxims for other groups? Shall we be pluralist about the CP and its maxims? We could imagine maxims for women, for people of African descent, for Hispanics, for three-year-olds, even, say, for fourteenth-century Augustinian monks; or we could be more specific still, envisioning maxims for, say, middle-aged conservative antifeminist housewives, young professional progressive but antifeminist women, liberal white straight feminists, Hispanic working-class

130

Implicatures (performative uptake)

radical lesbian Catholic feminists, black working-class Marxist academically trained feminist theorists not currently employed by a university, and so on. Certainly, we could imagine a separate set of maxims for each language community in the world. Would this do it? We could imagine each set of maxims as a formalized Wittgensteinian language game, each with its own discrete and therefore formalizable rules. Wouldn’t this be the democratic thing to do? We could, in other words, push Grice’s incipient awareness of the linguistic remainder ‘‘back’’ in a more ideologically normal formalizing direction, back toward ideal methodological stability, back toward stable rule-governed interpretive systems ^ back toward what seems universal because it is ideological, and because the function of ideology is to naturalize the local as the universal, the recent as the timeless, the social as the biological. This is, in fact, more or less what Grice’s followers among linguistic pragmaticians have attempted to do ^ although, because constative linguistics normatively represses the specifically ideological functioning that underlies the drive to idealize, it is usually less rather than more. They have tinkered with his maxims, beefed them up, added substance and ‘‘coverage’’ in an attempt to make them as universal as Grice claims they already are. Only then, they believe, only when they have nailed down the universal (ideal, transcendental) maxims underlying all speech, will they be able to make reliable truth claims about conversational implicature. Well, they won’t. Language doesn’t ‘‘possess’’ a stable ideal essence that can be discovered and formalized by linguists. Language is social behavior, and to the extent that it is controlled at all, it is controlled by local hegemonic social forces that are profoundly conservative but never monolithic or universal or perfectly successful. What I will be suggesting about maxims throughout Part III is that they are activated dynamically by the utterer in the act of intending, and again by the interpreter in the act of interpreting: that they exist, but unconsciously, ideosomatically, and are only brought to analytical consciousness (and thus tentatively formulated as maxims) when a specific interpretive need arises. The maxims are fundamentally social assumptions governing language use; the reason Grice had such trouble formulating them in any kind of universal (‘‘constative’’) way was that social assumptions vary from society to society, and indeed from social group to social group. In a performative perspective this implies several things. First, that the maxims are born out of dynamic speech exchanges. Second, that they only ‘‘exist’’ ‘‘in the language’’ through each social group’s gradual conventionalization (what in chapters 11^14 I will be calling ‘‘immediatization’’) of those dynamic exchanges, so that wandering out of one social group and into another (let alone out of one language and into another) will always involve a certain amount of uncertainty as to the maxims that govern speech. Third, that each group member’s sense of how the group has conventionalized or immediatized those dynamic exchanges will be based on

Conversational implicature 131 actual experience, and may differ in small but potentially significant ways from that of other members of the group. And fourth, following out of points one through three, that it is methodologically preferable to work out the maxims that each language user is constructing in a given speech situation rather than to posit a stable set of maxims governing the entire exchange (and indeed, in Grice, all exchanges). This is already a fairly radical critique of Grice; still more radical critiques are to come. I have dedicated Part III to an extended discussion of Gricean implicature, and will be developing my sense of its analytical usefulness slowly, over the course of the next six chapters: expanding upon Grice’s specific theories along Austinian lines and applying those expanded forms of his theories to translation (Chapters 9^10), and adding the interpretant triad of Charles Sanders Peirce to the mix (Chapters 11^14) in order to set up a more detailed analytical framework for the study of what exactly is going on in implicature. As I undo this Gricean principle, drastically expand that, turn this Gricean theory on its head, totally replace that, it may come to seem as if there is nothing left of Grice but the name ^ that ‘‘Gricean conversational implicature’’ becomes in the end but a hook on which to hang my own theories. And in fact my revisions of Grice here in Part III are far more thoroughgoing than were those of Austin, Derrida, or Bakhtin. They are also lengthier: I have dedicated a single chapter each to those previous theorists, fully seven to Grice. When we recall that each of those earlier theorists developed their ideas at book length, however, Grice in a single lecture,42 the surprising thing is not that I revise him so extensively but that his short lecture/article is so endlessly productive. Grice did not elaborate on the most interesting analytical possibilities he raised; he left most of them to those who came after. The fascinating theoretical avenues that he pointed down in passing but did not stop to follow are surprisingly many. I take the core of Grice to be the observation that we are able to convey implicit messages successfully by deliberately flouting common conversational assumptions or expectations. This Gricean core I do expand slightly, from implicit messages to implicit speech acts (Chapter 10); but it survives my revisions more or less intact. Everything else in Grice’s lecture is subjected to thoroughgoing critique. Before I launch into those more radical critiques, however, I want in Chapter 9 to look at some fairly timid applications of Grice to the study of translation.

9

Translation as ideological implicature

Grice himself in fact provides a translation example of conversational implicature: Take the complex example of the British General who captured the province of Sind and sent back the message Peccavi. The ambiguity involved (‘‘I have Sind’’/‘‘I have sinned’’) is phonemic, not morphemic; and the expression actually used is unambiguous, but since it is in a language foreign to speaker and hearer, translation is called for, and the ambiguity resides in the standard translation into native English. (1988: 36) This for Grice is an example of the flouting of the maxim of manner, ‘‘be perspicuous’’: it is nonperspicuous to convey the message ‘‘I have Sind’’ by means of a Latin word that must be translated into English as ‘‘I have sinned’’ and then understood homophonically to mean that the city of Sind is now in English hands. Grice is not particularly interested in the role played by translation in this implicature; his discussion revolves around whether the general is not also saying ‘‘I have sinned,’’ which Grice calls ‘‘the straightforward interpretant.’’ ‘‘If,’’ he writes, ‘‘the author of Peccavi could naturally be supposed to think that he had committed some kind of transgression, for example, had disobeyed his orders in capturing Sind, and if reference to such a transgression would be relevant to the presumed interests of the audience, then he would have been conveying both interpretants: otherwise he would be conveying only the nonstraightforward one’’ (36). The fact that the general sent one or both interpretants in Latin, and that working out the implicature involves both translation back from Latin into English and an understanding of the clever English pun built into the Latin word, seems peripheral to Grice. A full Gricean analysis of this example, it seems to me, would have to begin with the question of what the general was implicating in sending a message in Latin to English military headquarters: in other words, what the general was doing in or by translating. For this we would need something like the expansion of Grice offered by Charles Altieri (1981: 88), which he

Translation as ideological implicature 133 calls ‘‘expressive implicature’’: Expressive implicature applies clearly in utterances where aspects of tenor or mode are foregrounded. I take paradigm cases of expressive implicature to be those in which one is invited to see a speaker as intentionally focussing pragmatic, nonconventional implicatures of an utterance in order to call attention to qualities of the agent-speaker revealed in the act of saying something in a particular way. Expressive implicature, then, is a way of developing Austin’s initial account of performatives in a fashion that may ultimately show the utility of Grice for questions of style in literature and in ordinary experience. On Gricean foundations we can speak of linguistic choices as ways of exploiting deviations from norms which entail behavioral as well as semantic and syntactic frameworks of expectations.43 The choice of a punning translated message instead of an English one is an act of expressive implicature: in choosing to write Peccavi rather than ‘‘I have Sind’’ the general is conveying messages other than the two possible interpretants Grice analyzes (‘‘I have transgressed’’and ‘‘I have captured the city’’). Just what those other messages are, just what paths the general’s expressive implicature follows, would require a lengthy and largely speculative analysis; but they might include such things as ‘‘we English generals in this war are educated gentlemen who know Latin’’ and ‘‘look how witty I am.’’ The analysis of the general’s expressive implicature would thus need to be historically and ideologically attuned to the expressive weight carried by knowledge of Latin (and especially the witty and urbane use of Latin for cross-linguistic puns) in a specific cultural situation in mid-nineteenth century England: its use to signal ‘‘superiority’’ markers in terms of class (gentlemen), gender (male), and, perhaps, colonial relations (English over Indian) as well. The use of a dead language in a war-time message might also be justified, of course, in terms of encryption: the general might be imagined as assuming that Latin would be known to his English staff but not to Indians intercepting his message. Translation would in this case be a performative act intended to conceal the intended message from unintended receptors.

Constative applications of Grice to translation But as I say, Grice considered none of this; and, useful as his theory of conversational implicature clearly is to the study of translation, neither have any of the linguistically minded translation scholars who have sought to apply Grice to the translation process. This I think is for good constative reasons: as long as translators are thought of constatively as by definition incapable of performing a speech act, it is not going to occur to scholars to ask whether they are capable of implicating anything conversationally in or through a translation either. This in turn means that Grice is going to have to be put to considerably less interesting work as a mere describer of the pragmatics of the

134

Implicatures (performative uptake)

source text, which the translator will be expected to transfer accurately to the target language. This methodologically normative ‘‘failure’’ on the part of the constativists to study translation as a complex social action performed by the translator once again makes the act of translation an instructive initial test case for a performative reading of Grice’s theory. The constative linguists of translation who have applied Grice to the study of translation include Basil Hatim and Ian Mason in Discourse and the Translator (1990) and The Translator as Communicator (1997), Basil Hatim again in Communication Across Cultures (1997), Mona Baker in In OtherWords (1992), Albrecht Neubert and Gregory M. Shreve in Translation as Text (1992), Peter Fawcett in Translation and Language: Linguistic Theories Explained (1997), and Kirsten Malmkjaer in ‘‘Cooperation and Literary Translation’’ (1998). For all seven of these linguistically oriented theorists, as I say, the pathway to applying Grice in translation studies lies through the presumption of equivalence: whatever the source author or speaker does verbally, they all believe, the translator must seek to follow it in as effective and representative a manner as possible. This means that conversational implicature for all seven of them is something that is to be found only in the source text, never in the translation ^ or rather, to put that more clearly, the only conversational implicature that should ever be found in the translation is the one modeled more or less faithfully on that perpetrated by the source author. The translator is never to be thought of as an independent pragmatic agent implicating something conversationally, by flouting maxims; s/he remains a kind of amateur pragmatician who studies the source text for implicature and seeks to replicate it as effectively as possible in the target language: no actor on the actual implicature stage himself, no flouter of maxims herself, merely a photographer or a sound engineer charged with getting other people’s implicature recorded right. Or else, especially for Neubert and Shreve and Fawcett, the maxims become guides to equivalent transfer: ‘‘Translators are also subject to the maxim of quantity,’’ Neubert and Shreve (1992: 76) write, ‘‘when they create their texts for L2 readers.’’ [L2 ¼ ‘‘second language.’’] And Fawcett (1997: 132): ‘‘The maxim of quality, of telling things as they are, ought to be at the very heart of translation in the concept of fidelity, and one of the hopes of a linguistic approach to translation was to produce precisely that. From a linguistic point of view, the only departures from the maxim of quality in translation should relate to those situations in which there is no linguistic expression available in the target language.’’ And in fact, as this last quote from Fawcett suggests, our seven constative linguists of translation do also account for shifts from the source text to the target text, largely by drawing attention to the culture-bound nature of the maxims (Baker) or to the differences in what people in different cultures will consider ‘‘relevant’’ or ‘‘irrelevant,’’ ‘‘too much’’ or ‘‘not enough’’ (Hatim and Mason, Neubert and Shreve). Hatim and Mason (1990: 96), for example, give the example of a translator for a Spanish in-flight magazine omitting part of the text that seems likely to be irrelevant to the English-speaking reader; they call

Translation as ideological implicature 135 this an application of Grice’s maxim of relevance, suggesting that the omitted parts do not seem likely to be relevant to English-speaking readers.‘‘Even in full translations,’’ they argue,‘‘translators can and do take responsibility for omitting information which is deemed to be of insufficient relevance to TT readers’’ (96). [TT ¼ ‘‘target text.’’] Judging from their skittishness about this ‘‘responsibility’’ that translators can and do take, this seems to be a potentially controversial point for them: ‘‘To what extent translators have licence to take such decisions,’’ they conclude their discussion of this particular example,‘‘is a legitimate subject for debate; when does ‘improved relevance’ become unacceptable intrusion or dereliction of duty? We submit, however, that sensitivity to the issue of relevance in text processing is a necessary part of the translator’s skills’’ (96). While they refuse to take a stand on this touchy subject, their rhetoric suggests that what they are surreptitiously attempting to do here is to expand the notion of equivalence to cover cases where an interpretive application of a Gricean maxim leads to a significant alteration in the text but can nevertheless, precisely because of this continuing adherence to the relevance maxim, still be considered equivalent. And in fact this is the most radical example they give ^ the closest their ideal translator ever comes to actually flouting a maxim. Elsewhere they are repeatedly concerned with recognizing maxims in the source text so as to be in a position to warn translators against flouting or violating them in the target text. On the very next page (1990: 97) after the Spanish in-flight magazine case, for example, they discuss the use of synonyms and other equivalent expressions in a text in French, referring to a certain M. Tricot not by his name but by his post, le conseiller d’Etat. ‘‘No violation of the maxim of relation is involved here since, for ST users, the expressive device is entirely conventional.’’ [ST ¼ ‘‘source text.’’] If the translators are not careful, however, they may translate these individual items (Mr Tricot. . . The Councillor of State) without concern for Gricean maxims: this ‘‘would appear to violate the maxim of relation, since neither co-text nor context provides evidence for the assumption that Mr Tricot is a member of the Council of State.’’ The French reader knows that MrTricot is a Councillor of State; the English reader, presumably, does not. In order to adhere to the maxim of relation, then, and to produce in these Gricean terms an equivalent translation, the translator must render differently, either providing the job identification in the co-text or sticking to ‘‘Mr Tricot’’ throughout. Mona Baker takes a somewhat different approach: following critiques of Grice’s maxims that see in them reflections of the culture of educated English professional males (or some such) rather than universal descriptions of human speech behavior (Clyne 1981, Thomson 1982, Loveday 1982), she argues that translators attempting to render a text effectively into a different culture must be prepared to shift into a whole new set of maxims. This is a more radical position than that adopted by Hatim and Mason or Neubert and Shreve, who tend to assume that the maxims are universal (cross-culturally stable) but their application or interpretation in different cultures will differ; Baker at least

136

Implicatures (performative uptake)

potentially cuts the translator adrift from the more or less firm moorings that the maxims might be imagined to provide, to reimagine them for his or her target culture. Thus for example Clyne (1981) suggests that, unlike English, German discourse is non-linear and favours digressions. In some extreme cases, such as Fritz Schutze’s Sprache soziologisch gesehen, there are ‘not only digressions [Exkurse], but also digressions from digressions. Even within the conclusion, there are digressions’ (ibid.: 63). Not only does the maxim of Relevance need to be redefined in view of these comments, but the non-linear organization of German discourse also seems to require a reassessment of another maxim: ‘Be brief ’. Clyne (ibid.: 63) explains how ‘every time the author returns to the main line of argument, he has to recapitulate up to the point before the last digression, resulting in much repetition’. One wonders how an organizational feature such as this relates to the maxims of Relevance and Manner. Can this apparent violation of the maxims render a German text partially incoherent if it is not adjusted in translation? An English translation of a German book, Norbert Dittmar’s Soziolinguistik, was apparently felt to be chaotic and lacking in focus and cohesiveness, although the original was considered a landmark in its field by Germans (Clyne,1981). (Baker 1992: 236) The interesting question here, of course, is: ‘‘Can this apparent violation of the maxims render a German text partially incoherent if it is not adjusted in translation?’’ The question is complicated by a telling vagueness in the manner of its asking: 1. Is it a violation of the maxims? Baker says apparent, suggesting that some people (especially native speakers of English) might be tempted to call it one, but it probably isn’t (at least not for German speakers). 2. If it is a violation of the maxims for English speakers, how can that ‘‘render a German text partially incoherent’’? The answer to that, of course, is that we mean at least two things by a ‘‘German text,’’ (a) one written in German, and (b) one written in some other language (say, English) but identified as originally German. Thus if the digressive style Clyne is describing is widely accepted in German academic circles, and thus may be assumed to be violating no maxims in Germany, then the violation of Grice’s maxims is only apparent in (a) but may become very real in (b). 3. What kind of ‘‘adjustment in translation’’ are we talking about, and what stable foundation can we imagine it resting on? For Grice, conversational implicature rests on the foundation of the maxims, which Baker agrees vary from culture to culture, rendering that particular foundation untrustworthy; what else is there? Grice does of course have a back-up: the maxims rest on the foundation of the cooperative principle or CP: ‘‘make your conversational contribution such as is required by the accepted purpose or direction of the talk exchange in which you are engaged’’ (Grice 1989: 26), and this is in fact the

Translation as ideological implicature 137 foundation to which Baker retreats: We have seen that the suggestion that Grice’s maxims are universal is difficult to justify. A more plausible suggestion would be that all discourse, in any language, is essentially co-operative and that the phenomenon of implicature (rather than the specific maxims suggested by Grice) is universal. In other words, the interpretation of a maxim or the maxims themselves may differ from one linguistic community to another, but the process of conveying intended meaning by means of exploiting whatever maxims are in operation in that community will be the same. This position is much more tenable, particularly since it seems to be a feature of language use in general that is based partly on adhering to constraints and partly on manipulating constraints to produce special effects. (Baker 1992: 238) But note that Baker is still not tendering the possibility that the translator might be the one ‘‘manipulating constraints to produce special effects.’’ The source author creates special effects by flouting maxims considered binding in his or her culture, or perhaps by manipulating other constraints; the translator seeks not to create his or her own special effects but to reproduce the author’s special effects for an audience in a different culture. This entails a greatly expanded conception of equivalence, one perhaps closer to that offered by the descriptive translation studies (DTS) people, like GideonToury (see Chapter 6, pp. 82^91), for whom translation often involves significant cultural shifts mandated by the target culture. Indeed equivalence is so broad in this conception, including such things as the replacement of potentially offensive jokes by less offensive jokes, in the Arabic translation of an English book called Arab Political Humor (Baker 1992: 234), that many constative translation scholars would doubtless not want to dignify it with the term. But Baker still wants to portray the translator as a reproducer of implicative effects, a secondary agent whose job description does not include initiating conversational implicature. The most philosophically sophisticated of the recent constative attempts to apply Grice to translation is that offered by the Danish UK-based translation scholar Kirsten Malmkjaer, who invokes Davidson (1973) on Grice to argue that Grice’s theory both cannot account for translation and must account for translation in order to work at all. The problem as Malmkjaer sees it is similar to the one Austin faced with ‘‘parasitic’’ speech acts: just as to Austin a speech act does not seem ‘‘serious’’ if performed by someone other than the ‘‘original’’ or ‘‘ordinary’’ speaker (if performed on stage, say, or in a poem), Gricean implicature does not seem to work if performed by a mediator: According to the theory, for a proper case of telling to have taken place, the writer must intend the reader to form a belief in virtue of his or her recognition of the writer’s intention that s/he form this belief. In translation, this picture is skewed by the intervention of the translator.

138

Implicatures (performative uptake) The reader of a translation typically finds it meaningful; s/he feels that s/he is being told something. But s/he cannot be being told this because s/he recognizes any intentions of the writer: s/he is recognizing the tellingintentions of the translator. But the translator is not the author and there is no place in the theory for a mediator of messages. Perhaps such a place could be created. However, I think that such a project would run into difficulties because of the reliance it would have to make on intentions. The translator would be likely to be cast in the role, first, of a recogniser of the writer’s intentions to produce responses in a readership, and next as a purveyor of these, via a new medium, to a different readership than any the writer is likely to have had in mind. Such a theory would share important characteristics with the theory of dynamic translation (Nida,1964) and would, in my view, run into serious difficulties concerning the notion of equivalent effect ^ effect here being renamed response. (1998: 35^6)

It is fairly easy, by now, having come to Grice through the theories of iterability, somatics, and especially double-voicing in Part II, and Anthony Pym’s reflections on the performativity of Michel Garneau’s translation at the end of Part I, to obviate Malmkjaer’s objection: she is attempting to stabilize intention along constative lines, as something statically present in the mind of the author just once. If it is mediated to a new audience, it must somehow be kept intact, unchanged; and that is by definition impossible, since now the ‘‘speaker’’ is different, the translator rather than the original author. In a performative purview all this seems perfectly irrelevant. Translation is to Gricean implicature as parasitic speech acts are to the ‘‘serious’’ speech acts they are based on, further performances of the ‘‘same’’ that inevitably change what they perform because all speech acts change what they perform. All speakers in the performative linguistics of Derrida and Bakhtin, even original authors, are ‘‘mediators of messages’’ who reperform what they have heard others say.

Venuti’s critique of Grice Lawrence Venuti, one of the best-known translation scholars in the world today, originally made his name in the translation studies field by putting a Marxist spin on the German romantic theory of translation offered in 1813 by Friedrich Schleiermacher. In that lecture Schleiermacher argued that ‘‘assimilative’’ translation that ‘‘brings the author to the reader’’ simply panders to the reader, gives the reader more of what the reader already knows, and so has no beneficial transformative effect on the target culture; the only really valuable approach to translation ‘‘brings the reader to the author,’’ which is to say leaves a feeling of the foreign in the translated text. Beginning in his 1986 article ‘‘The Translator’s Invisibility’’ and continuing in the essay collection he edited in 1992, RethinkingTranslation, and his 1995 book of the same title as his first essay, The Translator’s Invisibility, Venuti has insistently given this ethical

Translation as ideological implicature 139 argument a Marxist twist, arguing that global capitalism spearheaded by the United States has systematically stripped foreign texts of their foreignness, reduced them to the same blandness that is produced locally, and that to resist this assimilative pressure the dissident (literary) translator must foreignize his or her translations. In his 1998 book The Scandals of Translation, however, Venuti takes a significantly different tack, launching a strong new two-pronged critique of the study of language in translation: one positive, coming out of French poststructuralism, Gilles Deleuze and Felix Guattari on minoritarian and majoritarian writing and Jean-Jacques Lecercle on the unsystematizable ‘‘remainder’’ that constantly thwarts any analytic of language; the other negative, directed generally at formalistic and scientistic approaches to translation, especially linguistic approaches based on Gricean theories of implicature. On the surface, neither critique is particularly surprising: Venuti has always drawn heavily on poststructuralist analyses, even though his Marxist ideology and Schleiermacherian model of translation have tended to render his approach more rationalist and binary than playfully deconstructive; and as a prominent representative of the cultural-studies approach to translation Venuti has always been unsympathetic to linguistic approaches. A closer look, however, suggests that Grice is a rather strange target for Venuti’s attack. After all, Grice offers what is essentially a transgressive approach to language, based on the notion that people are able to mean things that they don’t say by breaking the rules. Based on Venuti’s theoretical work to date, this should by rights recommend Grice to him. How much closer can an analytical language philosopher get to the poststructuralist notion that Lecercle calls the ‘‘remainder’’? As I will show, also,Venuti reads Grice so simplistically as to miss entirely the exciting and productive ways in which Grice’s theory dovetails with his own, or with theoretical ideals that he holds dear. Venuti begins his opening chapter of Scandals with a personal manifesto, situating himself in the contemporary cultural landscape and setting up his philosophy of language, based on the French poststructuralist theories of Deleuze and Guattari in their Kafka book and Jean-Jacques Lecercle in TheViolence of Language: Following Deleuze and Guattari (1987), I rather see language as a collective force, an assemblage of forms that constitute a semiotic regime. Circulating among diverse cultural constituencies and social institutions, these forms are positioned hierarchically, with the standard dialect in dominance but subject to constant variation from regional or group dialects, jargons, cliches, and slogans, stylistic innovations, nonce words, and the sheer accumulation of previous uses. Any language use is thus a site of power relationships because a language, at any historical moment, is a specific conjuncture of a major form holding sway over minor variables. Lecercle (1990) calls them the ‘‘remainder.’’ The linguistic

140

Implicatures (performative uptake) variations released by the remainder do not merely exceed any communicative act, but frustrate any effort to formulate systematic rules. The remainder subverts the major form by revealing it to be socially and historically situated, by staging ‘‘the return within language of the contradictions and struggles that make up the social’’ and by containing as well ‘‘the anticipation of future uses’’ (Lecercle 1990:182). (Venuti 1998: 9^10)

It is with this theory of language in mind, then, that Venuti tackles Grice and the Gricean translation theorists I reviewed just above, in particular Hatim and Mason (1990), Baker (1992), and Neubert and Shreve (1992). In the work of these theorists, Venuti claims, ‘‘Translation is . . . theorized on the model of Gricean conversation, in which the translator communicates the foreign text by cooperating with the domestic reader according to four ‘maxims’: ‘quantity’ of information,‘quality’ or truthfulness,‘relevance’ or consistency of context, and ‘manner’ or clarity’’ (21). I agree with Venuti on the constative applications of Grice to translation; interestingly, however, Venuti seems to lump Grice in with them, as if they were all saying basically the same thing. He seems, in fact, to have read Grice through those applications, to have read Hatim and Mason, Baker, and Neubert and Shreve on Grice first, and then assumed that these constative applications had more or less exhausted Grice’s theoretical usefulness. For Venuti the core of Grice’s argument is the maxims, maxim-based cooperation, which he sees as restricting and repressing variation and transgression; his recurring refrain all through his section on Grice and Gricean translation theories, is that the Lecerclean remainder in the act of translation violates the maxims: The remainder likewise threatens other Gricean maxims. It violates the maxim of truth, or the ‘‘virtual reality’’ created in the translation (Neubert and Shreve 1992: 79), because the variables it contains can introduce a competing truth or break the realist illusion. (22) Yet even with technical translation [where, he’s just told us, translators are constrained ‘‘by their contracts or by the conditions of their employment in agencies to honor those maxims’’] certain situations and projects may arise to warrant a violation of a Gricean maxim. (24) Diplomatic and legal interpreting can easily involve violations of the Gricean maxim of truthfulness. (24) In these cases, the translator not only violates some conversational maxim, but raises the question of whether the translation practice can

Translation as ideological implicature 141 accurately be called mere communication as opposed to mystification, naturalization, or palliation. (25) The one time he does pause to note Grice’s main point, the flouting of the maxims in order to mean what one does not overtly say, he makes it sound as if for Grice this flouting and the implicature that results are failures of some sort, counterexamples to his argument in favor of cooperation, and thus limitations on his claims: ‘‘Grice interestingly admits,’’ Venuti writes ^ that word ‘‘admits’’ doing a lot of rhetorical work in his rather tendentious summary ^ ‘‘that language is much more than cooperative communication when he proceeds to argue that the maxims are routinely ‘flouted’ in conversation, ‘exploited’ by interlocutors to open up a substratum of ‘implicature,’ such as irony’’ (21). Yes, Grice does say this: it is, after all, the primary thrust of his argument. It is not something he ‘‘admits,’’ as if a little shamefully; it is the main thing he came to say.44 Austin can’t figure out what to do with figurative and literary language, so he calls it ‘‘parasitical’’; Grice answers that question by suggesting that ‘‘parasitical’’ speech acts form part of a much larger category of transgressive speech acts, speech acts that take their force by flouting some set of assumptions, which Grice then proceeds, but only secondarily, to formulate: the CP and its maxims. First, the problem: people saying things they don’t literally mean, meaning things they don’t overtly say. Surely a significant part of the ‘‘remainder.’’ Then, the solution: it works through the flouting of conventional expectations. Then, finally, a tentative sketch of those expectations, the CP and its maxims. Venuti gets it exactly backwards: first, for Grice as he reads him, stultifying conventions of cooperation; then, as if apologetically, the ‘‘admission’’ that people sometimes violate convention and thus fail to cooperate. Based on this upside-down reading, Venuti generalizes from Grice to linguistics and concludes: ‘‘Linguistics-oriented approaches, then, would seem to block the ethical and political agenda I envisaged for minoritizing translation’’ (22). It’s tempting to respond to this that if your ethical and political agenda is narrow enough, and dogmatic enough, almost everything comes to look as if it would block it; but as my preliminary critique of Grice’s constativism in Chapter 8 suggests, Venuti is also right in suspecting that there is something wrong with Grice’s approach. For one thing, as we’ll see again below, the very act of submitting some part of the ‘‘remainder’’ to logical analysis does in effect redefine that analyzed part as no longer part of the remainder ^ by definition, the ‘‘remainder’’ is that which forever eludes systematic analysis, that which is left over after all the systematizers have done their work. And while Grice begins with an interest in the remainder, and feels driven to his analytical formulations by that interest in a problem that has not yet been adequately explained, he also quite clearly remains strongly attached to the analytical system he has devised to explain that problem, his maxims and the rationalist model of cooperation that they rest

142

Implicatures (performative uptake)

on; and his followers in the field of linguistic pragmatics (which those followers created out of Grice) are even more so.

A Gricean reading of Venuti Still, it seems clear that Venuti’s attack on Grice is significantly misdirected; it should be aimed at a few linguists who apply Grice rather conservatively to translation studies, not at Grice. All we need to do to obviate Venuti’s complaint about ‘‘linguistics,’’or ‘‘Grice,’’ in fact, is to extend the possibility of implicating things ^ meaning things one doesn’t say outright ^ to the translator as well. Then the translator becomes precisely whatVenuti wants him or her to become, an ethical agent taking a stand, saying something to the reader by flouting the maxims (or the cooperative principle, or what have you ^ more on this in a moment). The maxim of quantity as Grice formulates it, for example, is ‘‘Make your contribution as informative as required. Do not make your contribution more informative than required’’ (1989: 26). Suppose the translator flouts this maxim, withholds ‘‘required’’ information, or says more than is required, while still obviously ‘‘cooperating’’ with the target reader, treating that reader as a serious and intelligent person deserving of a serious and intelligent translation (not, in other words, flouting the maxim of quantity simply in order to confuse the reader or play some other cultural mind-game). Then we are in the territory of propagandistic translation, clearly. A good example of this sort of cooperative maxim-flouting in the translation field might in fact be the feminist translation of sexist texts, especially where the translator is determined to signal to the reader that the text is sexist, by highlighting more or less peripheral sexist traces in the text: saying more to the reader than the author would have liked, in order to signal to the reader that a certain kind of ideological vigilance is in order when reading the text (for a detailed example, see Robinson1991:148^58). The pragmaticians Dan Sperber and Deirdre Wilson famously reduce all of Grice’s maxims to relevance, on the grounds that ‘‘To communicate is to claim an individual’s attention; hence, to communicate is to imply that the information communicated is relevant’’ (1986: vii), and Hatim and Mason note that this is true because ‘‘in communication we pay attention only to information which seems relevant to us (greatest cognitive effect for minimum processing effort)’’ (1990: 95).45 But of course relevance may vary widely: not just authorial relevance, source-lectorial relevance, and target-lectorial relevance, as conservative relevance models would have it, but translatorial relevance as well. The source author may feel that his sexist remarks are irrelevant, or only marginally relevant, part of the (properly) ignored periphery of his text; and he presumably would want his readers to read him accordingly. For the feminist translator, however, this ‘‘periphery’’ takes on new relevance, and as an ethical agent she has to decide what to do about it: leave it as is, refuse to translate the whole text, air-brush it out, or highlight it (for discussions of these see Lotbiniere-Harwood 1991). These decisions are ideological ones related to the desired uptake of relevance by the target reader, and, through that

Translation as ideological implicature 143 uptake, the proactive selection and subjectification of that reader (the feminist translator, in other words, may be presumed to be looking for feminist readers, either targeting readers who are already feminists or trying to help convert readers into feminists). Indeed Venuti’s own political agenda in translation sounds to me almost exactly like the ‘‘cooperative maxim-flouting’’ model Grice imagines: like Grice’s hypothetical interlocutors, the Venutian translator seeks to flout culturally dominant maxims for translation (‘‘translate fluently, so that the translation seems to have been written originally in the target language’’), but seeks specifically to do so cooperatively, so that the reader understands that this translation approach is not simply a sign of translator incompetence or contempt for the reader’s needs but a culturally and politically meaningful speech act, or translation act.46 Venuti’s ideal dissident translator flouts fluency maxims cooperatively in order to implicate something to the reader, to signal without spelling it out that fluency is complicit with capitalism and should be resisted. It is only if one assumes that the foreignizing or minoritizing translator’s readers recognize such cooperation on the translator’s part, in fact, that the translating styles Venuti favors can be construed in context as the expression of an ethical stance. Otherwise, they collapse into the rubbish heap of ‘‘bad translation’’ from whichVenuti has worked so hard to rescue them. Surely this solves all Venuti’s problems with Grice? Let the translator implicate things too, let the translator flout fluency maxims cooperatively so as to signal his or her participation in a larger ethical project that the reader should take seriously, and Venuti’s critiques of ‘‘Grice’’or ‘‘linguistic pragmatics’’ simply dry up and blow away ^ which is to say that his critiques weren’t really of Grice at all, who in retrospect now seems to be an excellent model for Venuti’s theory, but of equivalence theory surviving in the name of Grice. If all a theory of translation has to do to satisfyVenuti is to allow the translator to play an active role in the resisting of hegemonic ideologies, that can be done easily within the confines of Gricean pragmatics. Which is not to say that I’m here to defend Grice against Venuti; only, perhaps, that I am harder to satisfy. Grice’s theory of conversational implicature is riddled with problems, and I suggest that a series of closer looks may well point us a way past at least some of them. The series here in the rest of Part III: in Chapter 10, implicature slotted into Austin’s theory of illocutionary force and perlocutionary effect, and ‘‘metalocutionary’’ implicature offered as a channel of ideological self-discovery; and in Chapters 11^14, complication through Charles Sanders Peirce’s interpretant triad, first a discussion and modification of Peirce’s model (Chapter 11), then the modified Peircean model applied to the problem of conventional implicature through the concept of ‘‘immediatization’’ (Chapter 12), then conversational implicature and immediatization expanded to include a theory of implicative ‘‘invocature’’ (Chapter 13), and finally the notions of metalocutionary implicature, conversational invocature, and immediatization used to explore the cultural relativism of the maxims (Chapter 14).

10 Il-, per-, and metalocutionary implicature

Grice’s theory of conversational implicature is, as I hope to show throughout Part III, almost endlessly productive; but it is also, and Ithink this may have a lot to do with Venuti’s bad reading of him, so sketchy that it has remained wide open to constative iterations. His cooperative maxims, as we saw in Chapter 8, are patently culture-bound in a casual, not-quite-thought-out manner, so that, as we saw in Chapter 9, in a hegemonically constative intellectual tradition they seem to invite reformulations in more ‘‘universal’’ terms. As we will see in Chapters 11 and 12, his analytic of dialogue is pretty bare-bones, and his distinction between conversational and conventional implicature has hardly enough bone in it even to be called skeletal. In Chapter 13 I will argue that the type of conversational implicature that he analyzes, one based on the flouting of maxims, is only one such type that we can detect in actual language-use situations. And in Chapter 14 I will show that his analytical model collapses entirely when it is applied to cross-cultural misunderstandings; there I will be drawing together the various threads of Part III, especially metalocutionary implicature from the end of this chapter and the idea of immediatization from Chapters12 and 13, to offer a comprehensive new model of implicative analysis.

Illocutionary and perlocutionary implicature To start with, then, let us note that there is something suspiciously constative about Grice’s theory, in Austin’s original narrow sense of the word: suspiciously informational or message-bearing. In Grice there is an almost exclusive concern with the transmission of messages, with understanding what the other person is not quite saying openly, or getting the other person to understand what you’re saying. Surely a performative account of this theory should push it a bit more sternly toward saying as doing? As Austin reminds us halfway through his book, of course, all saying, even that which conveys information, is doing. All speech is speech acts. Still, there must be more to the speech act(s) of implicature than constating things implicitly. All of Grice’s examples revolve around concealed information: the banker who hasn’t been to prison yet, the philosophy professor whose student is assiduous but not very good at philosophy, the metaphorical and ironic interpretations of ‘‘You are the cream in my coffee,’’ the general who

Il-, per-, and metalocutionary implicature 145 captures Sind and wires back ‘‘Peccavi.’’ Surely there might also be performative implicature ^ a wager or a promise, say, that isn’t spelled out? Or, to invoke Austin’s later and fuller framework, couldn’t we distinguish between ‘‘locutionary’’ implicature (the one kind Grice discusses), ‘‘illocutionary’’ implicature, and ‘‘perlocutionary’’ implicature? So imagine:47 a 17-year-old American middle-class girl sits down with her father in the living room for a heart-to-heart talk, and in the course of the conversation says: 1.

Dad, I’ve been thinking, I’d like to move out of the house.

Her father replies: 2.

Go right ahead, drive another nail into your mother’s coffin.

Intuitively we sense the presence of some implicature or another here, and in the kind of constative laboratory conditions that Grice favors, where speaker-intention and hearer-interpretation are both perfectly and simultaneously accessible to the objective observer,48 it would probably be possible to determine a ‘‘definitive’’ implicature for it. In the dialogical situation itself, especially when viewed performatively, things are more difficult. If I am the daughter, for example ^ since in real dialogues I can only be one person at a time, and the performative linguist is more interested in real dialogues (or at least in the idealized limitations of real dialogues) than ideal ones ^ (2) may well leave me puzzled. Is my father giving me permission to move out or not? This would be illocutionary implicature: performing a classic speech act without spelling it out. My interpretation will depend largely on my sense of my father, of his character, of his present mood; and even the best-informed interpretation will never be transparently reliable. My gut reaction may be (especially if I’m a whiz in conversational logic) that the logical proposition to which his utterance might be reduced (if you leave, your mother will die) reverses the import of the ‘‘Go right ahead’’ ^ thus, he’s denying me permission to leave. But that’s bound to be unsatisfactory, for several reasons. One is that, even at the simplest interpretive level, telling me to go for the wrong reason doesn’t deny me permission, it seeks to work on me so as to make me change my mind. This would be perlocutionary implicature: trying to affect my response without spelling it out. Another problem is the complexity of human moods, motivations, selfprojections. If my father likes to think of himself as a humorist, for example, I may take the apparently bitter irony of his remark as I first understood it to be a joke: the irony may be more conventional than bitter, a phrase picked up from television, which he is employing lightly, mock-melodramatically. But then if he is joking, does that mean that he will let me go, or that, say, he just wants to set a jovial tone for our conversation? These would be alternative illocutionary implicatures, but even if I could be relatively certain of my interpretation at this level, I might still remain in the dark about his perlocutionary intentions ^ about

146

Implicatures (performative uptake)

what effect he meant his joking to have on me. Does he want to jolly me into changing the subject? Or does he find the whole conversation so unpleasant and indeed my presence in the room so distracting (or whatever) that he wants to irritate me into leaving him alone (he knows how much I hate his levity when I want to talk about something important)? Or, perhaps, does he want to teach me to make my own decisions without waiting for permission from him, and so deliberately confuses me so as not to tie my hands with his authority? The complexity of this expanded model of implicature should be clear. When we add to the Austinian framework of locution/illocution/perlocution and the Gricean framework of literal/implied speech acts49 the additional feature that several levels of pretense can be built into implied speech acts, we get a schema that looks something like this: I. II. III.

literal (a) locution1 (b) illocution1 (c) perlocution1 implied (a) locution1,2 (b) illocution1,2 (c) perlocution1,2 pretended (a) locution2,3 (b) illocution2,3 (c) perlocution2,3

The multiple subscripts in (II) and (III) indicate the alternative possibilities of either operating on all three levels of implicature (I/II/III) or of implying the same thing you say or do literally by pretending to say or do something else (I/I/II). Thus in (I) above, I may as father be (taken by my daughter to be): I.

literally (c) confusing her (perlocution1) by (b) giving her permission to move out (illocution1) on the grounds that (a) her leaving would kill her mother (locution1) but II. implicitly (c) getting her to drop the whole subject (perlocution2) by (b) denying her permission to move out (illocution2) on the grounds that (a) her mother and I would miss her too much (locution2), while at the same time III. pretending to (c) keep her happy (perlocution3) by (b) joking about the whole situation (illocution3) Or I may be (taken by my daughter to be): I.

literally (c) confusing her (perlocution1) by (b) giving her permission to move out (illocution1) on the grounds that (a) her leaving would kill her mother (locution1) but II. implicitly (c) trying to make her see that she is an adult and capable of making her own decisions (perlocution2) and (a) telling her that what her mother and I feel is less important than what she feels (locution2), and therefore (b) giving her permission to move out (illocution1), but achieving all this by III. pretending (c) to irritate her (perlocution3) by (b) denying her permission to move out (illocution2) on the grounds that (a) her mother and I would miss her too much (locution3).

Il-, per-, and metalocutionary implicature 147 And if that isn’t complex enough, note that there are no natural limits to the levels of pretense (III, IV, V, . . . , n) that can be employed: pretending to joke about something, for example, might be said to involve pretending to take it seriously on one level and joking about it on another, or pretending to take it seriously on one level, pretending to joke about it on another, and really taking it seriously on a third. It may be objected that no one could possibly convey such devious intentions in a single utterance; but I would deny that it is possible to guarantee the ability to convey any intention in a single utterance. Interpretation is too complexly contingent, too finely dependent on an uncountable number of factors, for interpretive reliability to be guaranteeable. The objection that no one (or no normal person) would be devious enough to try is another kind of claim ^ one steeped in the axiomatic white educated male middle-class ideology that we saw Grice formulating in Chapter 8. We’re all devious enough, sometimes, in certain situations, to concoct intentions of these levels of complexity; and we all are quite willing to interpret our friends’ intentions with equal complexity, given circumstances that arouse our suspicions (or generally our imaginations) and our need to unravel our interpretive responses fully ^ as when making or understanding a point in a marital quarrel or academic debate, or pulling off or exposing an elaborate swindle, requires getting the phrasing precisely right. Whether such interpretations are ever accurate, whether such intentions are ever wholly or perfectly successful ^ these are questions of another order altogether. Complex as this formulation of all the possible implicit illocutions and perlocutions is, of course, it remains formalistic, another attempt to systematize the remainder; once again the doors of constative linguistics open wide to welcome us. Not only that: my schematization of illocutionary and perlocutionary implicature remains firmly in the realm of what game theoreticians would call tactical complexity: complexity useful to someone playing by the rules. I implicated this rule-obeying surreptitiously by making the ‘‘literal’’or explicit locutionary, illocutionary, and perlocutionary acts in my two scenarios identical, as if the literal meaning and effect of utterances were somehow ‘‘given’’ or ‘‘transparent,’’ indeed as if they were somehow ‘‘literal,’’ intrinsic properties of the ‘‘letter.’’ That they aren’t ^ that they are ideologically constituted in and by the discourse of a specific interpretive community, that they are‘‘rules’’programmed into the somatic responses of a given group or culture ^ should by now be obvious. The father’s instruction to move out in order to kill the mother is perlocutionarily confusing (induces in the daughter intellectual bafflement marked by somatic signs of anxiety or unease) only within a given set of historically contingent assumptions about family life ^ which nevertheless feel real, feel normal, feel intuitively right, due to the somatic power of ideological programming. An American middle-class teenager would not be inclined to take such an instruction at face value (would feel anxiety signals at the very possibility of taking it at face value) partly because, unlike some prescientific societies, the American middle class doesn’t believe in the power of emotional

148

Implicatures (performative uptake)

distress to cause death (would in fact rather not think about the somatics of emotion at all), partly also because the American middle class is not in the habit of instructing children to kill their mothers. Tactically speaking, of course, this sort of information is useless. The tactician doesn’t need to know why the game is played in a certain way or what alternatives there are to playing that way (how it’s played elsewhere) ^ only that that is how it’s played here, in his or her own interpretive community. This is how it feels right to play it. The tactician remains within a given inscription, a given set of social conventions. Strategically, however ^ to invoke the other key game-theoretical term,‘‘strategy’’ as opposed to ‘‘tactics’’ ^ it becomes critical to gain knowledge about the structure of the game you’ve been placed in: who placed you there, who wrote the rules and who enforces them, and what alternatives there might be. Strategy is a way of stepping outside a game, of placing it in a new perspective, of learning to understand the contingent but real-feeling power of its rules, and of gaining power over that controlling inner feeling. Effective strategy requires, that is, a self-reflexive sense (‘‘theorizing’’) of the authority you’ve internalized, an ability to work past the somatic ‘‘naturalness’’of your beliefs and behavior (your ‘‘practice’’) to an understanding of how you came to harbor all that. It requires a willingness to destabilize the world as you’ve been taught to feel right understanding it in order to gain a greater freedom of movement in the dual realms of action and thought. This is the subject I will be exploring in the last two sections of this chapter, under the heading ‘‘metalocutionary implicature.’’

Metalocutionary implicature In Chapter 8 I argued that we don’t need an analytical foundation in rationalism and cooperation for intention or interpretation to be possible: we intend, we interpret, and merely posit those foundational things along the way, if it seems helpful. If we intend something and our interlocutor doesn’t get it, we try again. If we interpret incorrectly and our interlocutor responds with impatience or shock or disgust to our construction of what s/he was trying to say, we try again. That’s all. This performative approach to the analysis of dialogue, which I will be schematizing along Peircean lines in Chapter 10, undermines the basic analytical premises on which Grice based his theorizing, most important among them the theoretistic assumption that analytical categories are primary and usage secondary, that the ‘‘rules’’ somehow exist somewhere (in Plato’s heaven, in some abstract place in our minds) and then somehow get obeyed or broken in actual language use. This assumption for performative linguists is not only a boring and outdated and philosophically discredited (in a word, thoroughly constative) way of proceeding; it is also quite unproductively distant from the way these things actually work in the real world. My assumption throughout this section on Grice, as throughout the entire book, is that actions are experientially primary, rules secondary: that by acting in the social sphere we do assimilate and internalize (somatize) collective regularities that we may

Il-, per-, and metalocutionary implicature 149 later formulate as ‘‘rules,’’ but the regularities are felt before they are formulated, and they are not rules until the ex post facto moment that someone treats them as such. In the constative intellectual tradition this attitude seems preanalytical, of course, or even antianalytical, based on the somewhat scandalous notion that people act first and only then (sometimes, maybe) analyze, which is no doubt why analytical philosophers and constative linguists and other objectivist thinkers have eschewed it; but as I hope my analytical models have shown thus far and will continue to show here in Part III, it is possible to analyze things usefully and realistically without recourse to an idealized anachronology of the act. And as in my discussions of iteration in Part II, in the two main spins I propose to put on Grice’s theory of implicature, the first here in Chapter 10, the second drawing on Peirce in Chapters 11^14, I want to put this performative sequence to work: first actions, then rules; first the activity of theorizing, then the theory. Let me begin the first spin with the Venutian critique of Grice that I discussed in Chapter 9: Venuti charged Grice with complicity in the status quo, the repression of difference and dissidence, the rigorous attempt to make the world conform to the cooperative principle and its maxims. Since Grice is throughout concerned with the many different ways people flout those maxims successfully, I suggested that Venuti had hold of the wrong end of Grice: that if we simply imagine the translator as an implicator, as someone who flouts standard assumptions about translation in order to convey an implicit message to his or her audience through the act of translating, Gricean implicature offers a useful schematization of Venuti’s own theory. In fact there are other significant parallels between Grice’s approach and Venuti’s as well. Both thinkers, for example, are rationalists who believe that action should be founded on principle: Grice on rational cooperation,Venuti on specific dissident ideologies. The Venutian translator should be encouraged to resist dominant ideological inscriptions ^ but not any old way. Resistance for Venuti should be systematic, organized by a coherent counterideology, and specifically a counterideology on the left, not the right. His old dualism, domesticating versus foreignizing, specifically meant complicity with capitalism (domestication) versus Marxist resistance to capitalism (foreignizing). His new approach uses the Deleuzean concept of minoritarianism to champion feminism against patriarchal sexism, queer theory against heterosexism, postcolonialism against First World hegemony, and so on. And I agree with most of this. I too prefer these particular resistant ideologies to orthodoxy, or to anything radically reactionary. What bothers me in Venuti is what I would identify as his constative strain: his insistence on the priority of principle. If you’re going to fight capitalism, you had better obtain a thoroughgoing training in Marxism. If you’re going to fight patriarchy, you had better first learn everything there is to know about feminism. And so on. The resistant ideology has to form a counterfoundation for the struggle ahead. I don’t agree. This seems to me a top^down, authoritarian survival of the Old Left: first, rigorous ideological analysis and thorough ideological training;

150

Implicatures (performative uptake)

only then, principled counterhegemonic action. My performative approach is perhaps more reminiscent of the New Left that sprang up in the sixties: act first, figure out what you’re doing, and how and why, as you go. And while you’re doing it, have fun. Venuti is specifically interested in the radicalization of the translator, the translator’s transformation from a neutral apolitical instrument into an active, politically aware, ethically responsible agent who performs speech acts rather than merely reproducing them as accurately as possible. And clearly, the first step in the radicalization process is consciousness-raising, becoming aware of what is being done to you. And, just as clearly, consciousness-raising requires ideological analysis. The problem is, ideological analysis when done by someone else and presented whole and complete to the unconvinced reader does not perform the desired conversion to the new way of thinking; it constates it. It presents the ‘‘true’’ ideology as a factual representation of the way things are, and expects the would-be convert to transform his or her perception of the world (and, secondarily perhaps, all future actions based on that perception) so as to conform to the newly apprehended reality. Experientially speaking this is too much at once; it is overwhelming in its systematicity, its seamless web of analytical truths. It leaves no room for tentative exploration, for groping performatively toward an experiential reformulation. Ideological analysis is useful, even essential, for the reader who is convinced of a radical view but confused as to its extension; it is useless, even counterproductive, for the reader who remains attached to a hegemonic view. Hegemony is an inward panopticon; it blocks doubt by repressing alternative views. In that mindset, a full-fledged alternative view can only appear as an oddity, a freak or sport ^ an elaborate hoax, perhaps, concocted by Martians or the mentally ill.There is no ingress into it. What is needed, then, is an experiential ingress, a performative utterance that performs not sudden conversion but emergent awareness: a sense simultaneously of security (this is nothing world-shattering, rest easy!) and of unease (something’s wrong here, try to work it out!) that nudges the reader gradually, cautiously, out of the hegemonic mindset that feels so natural, so real, so comprehensive, so complete, and into a confrontation with the leading edge of the repressed. I’d like to suggest that one ingress of this sort might be cobbled together out of Grice’s theory of implicature and Austin’s theory of locutions ^ that there is (or that I can bring performatively into existence by naming) a form of implicature, which I want to call ‘‘metalocutionary,’’ that performs the action of enhancing self-reflexivity, of nudging a speaker toward awareness of his or her own implied meanings. For example, in my father^daughter scenario from the first section of this chapter, the seventeen-year-old daughter has been raised to believe in an all-embracing dualism between ‘‘living at home’’ as childish unfreedom and ‘‘moving out’’ as adult freedom. Does she realize this in saying (1) ‘‘Dad, I’ve been thinking, I’d like to move out of the house’’? Probably not. Her assumptions are structured by that dualism, but she is probably not aware of being so

Il-, per-, and metalocutionary implicature 151 structured. If she is aware of the dualism at all, it is as a simple ‘‘reality’’: live at home and your parents tell you what to do; move out and do what you want to do. That’s the way the world works. Period. In the game-theoretical terms I invoked there in that first section, of course, this is tactics, playing the game as effectively as possible within the rules. This is as far, clearly, as hegemony wants us to go. As I also noted in that earlier section, game theoreticians contrast tactics with strategy, which involves rethinking the entire game from a position above or outside it. What, then, might constitute a new, transformative strategy? What could drive the daughter to a radicalizing awareness of, let us say, her submission not only to her father’s control (permission/denial of permission) but also to the law that defines seventeenyear-olds as minors subject to the authority of parents or guardians? What could goad her into ‘‘thinking’’more transformatively about moving out ^ about the ways in which her attempt to resist her father by moving in with a boyfriend, say, is a patriarchally inscribed desire to replace one male supporter with another? Her father might in fact radicalize her, or plant the dialogical seed of radical renaming in her, by responding with a koan that is less doubly binding than (2) ‘‘Go right ahead, drive another nail into your mother’s coffin’’: 3. You know, freedom doesn’t start at that front door. What does this mean? Like the coffin remark, this new paternal statement about freedom is apparently uncooperative (thus at least potentially implicative) in Grice’s sense, since it patently evades the issue of the daughter staying or going by shifting laterally to a discussion of freedom and its boundaries. The daughter could once again interpret the utterance in terms of illocutionary and perlocutionary implicature, asking herself whether the remark about freedom constituted permission or the denial of permission, and whether her father is just trying to stall or irritate or confuse her; but she could also respond by interrogating the connections between this response and her own utterance that provoked it. What have I been thinking about moving out? Have I in fact been thinking that I will then be free? Will I? Does freedom start at the front door or doesn’t it? If it does, why is he saying the opposite? Is he just trying to thwart my attempt to move out, or is he pushing me toward a rethinking of my assumptions about moving out? If it doesn’t, then moving out doesn’t necessarily or automatically put me in a state of freedom; but if that’s the case, where does freedom start? Is he saying that there is no such thing as freedom? Or that I am freer here than I would be if I moved out? Or that I have to struggle toward freedom by other means than simply changing my address? And so on. What I see going on in this series of interpretive responses is a questioning of the interpreter’s own original intentions ^ which, since they were at least partly (politically) unconscious, were also ‘‘implicit’’and thus need to be worked out.What did I mean? What was I trying to say? I’m calling this performative reflexivity metalocutionary implicature, the strategic step past the tactics of locutionary, illocutionary, and perlocutionary

152

Implicatures (performative uptake)

implicature. Where those three work within existing game-rules, by whatever ‘‘maxims’’ each individual interpretive community has programmed its members with, metalocutionary implicature would be an explorative (interpretive or intentional) act that precipitates a metaunderstanding of the game you’re already engaged in playing according to rules you don’t even know you know, the act of coming to view the playing of the game from without, temporarily or partially lifting yourself outside of your own tactical situatedness and thus working toward strategic self-discovery. Here Grice’s theory of, or rather theorizing about, conversational implicature may again prove helpful. Grice has, as I mentioned above, been taken to task for insisting that his maxims are universal, when they are so obviously the ideological norms of his group, white middle-class professional males of twentieth-century England and the United States; but in this metalocutionary perspective it is more fruitful to look not at his universality claims but at the process by which he articulated the maxims. I said before that Grice didn’t make them up; they exist, as the episteme governing white professional male behavior in the West, at least, and probably more specifically, in England and the U.S. But this is simplistic. Where do they exist? How? In what form? How did Grice find out about them? They ‘‘exist,’’ in fact, to the extent that they exist at all, as programmed ideosomatic responses: it feels right for a professional Anglo-American middle-class male to say no more than is necessary, to keep all speech acts strictly relevant and perspicuous, etc., because those ideologically normative assumptions have been iterated ideosomatically, inscribed by hegemony on his body, on the bodies in fact of all members of the group. But those ideosomatic responses are ordinarily part of the (somatized) political unconscious; they don’t exist as maxims until they are spelled out, which would suggest that Grice’s maxims are in an important sense ‘‘made-up,’’ are Grice’s own critical fictions, mental constructs by which he seeks to explain how it feels to convey a meaning that he doesn’t make explicit. Note that this is already a metalocutionary leap: in formulating his theory of the cooperative principle, its maxims, and how they can most effectively be flouted for implicative purposes, Grice was essentially reaching toward a metalocutionary insight into his own ideosomatic inscription, interpreting the metalocutionary implications of his own conversational habits. He was able to ‘‘create’’ (or formulate) the rules only by breaking the rules of ideological repression. In game-theoretical terms, he rose above the tactics of the conversational games he played in day-to-day life to a strategic perception of those games from above. His mistake, then, lay in presenting (and, no doubt, in perceiving) his insight as an objective representation of the game, a formulation of the unwritten rules by which it was generally played, rather than as his personal attainment of strategic mastery of the game. His mistake, in other words, lay in attempting to formalize his metalocutionary strategy into tactical rules rather than offering as interpretive key the strategy of bringing tactics into consciousness.

Il-, per-, and metalocutionary implicature 153 I will in fact be pushing this metalocutionary strategy farther in the remaining four chapters of the book, most overtly and extensively in Chapter 14, after a lengthy Peircean detour that will establish the key concept of ‘‘immediatization.’’ I will be pushing it specifically into the recognition that metalocutionary implicature is in effect the channel through which maxims are constituted as maxims in every individual conversation. This apparently cart-before-the-horse principle is based on the perception with which I began Part III, and reiterated in the beginning of this chapter: that dialogical understanding doesn’t rest securely upon the ideal foundation of rationality or cooperation or anything else, or any other formal principle or concept, either; it is, rather, a political bone of contention, something to be fought over, vied for, experimented upon, groped for, rebuilt, reiterated, over and over again, in actual dialogues. This political iteration of Derrida’s concept of iterability suggests that all conversation involves not merely tactical maneuvers within shared game rules, but also strategic maneuvers designed to reshape the game rules, to play the game by new rules. These new rules may be developed, advanced, and imposed (partially, tentatively, more or less successfully) on other speakers by a single speaker in rebellion against her or his collective conditioning, or by a group of speakers in search of a new (or constantly changing) group identity, a new (fresh, unfamiliar, ‘‘spontaneous’’) foundation for group solidarity.

Translation as self-discovery I suggested last chapter that we might, following Grice, imagine translators as ethical agents actually implicating things in their translations, flouting maxims and being strategically uncooperative, in the service of a translational ethics larger and more politically aware and responsible than merely that of (constative) representational accuracy. Now, following up on my discussion of metalocutionary implicature, I want to lodge a few concluding observations on a performative theory and practice of translation, one based not only on doing something with words but also on learning something about what one is doing. This will be my last extended discussion of translation in the book; in the four chapters that remain I will only make passing references to translating and focus more fully on speaking. The delayed-action nature of translating obviously mires it in mind-numbing paradoxes. The German original says ‘‘In German we say,’’ but you’re translating it into English, and it seems ridiculous in English to talk about what ‘‘we’’ say in German. If you’re translating a sacred text or a literary classic, the Bible or Homer or Shakespeare, you’re striving to create in the present a version of a book that has had enormous influence on hundreds or thousands of years of cultural history before you, indeed on you as well. How do you convey the antiquity and the historical significance of ‘‘your’’ text that you are just now in the process of creating?

154

Implicatures (performative uptake)

Every translator and translation scholar who has ever lived has recognized the many insoluble problems posed by translation; most have come to accept them as part of the job, and thus not to worry too much about solving them ^ undoing the paradoxes of time and place ^ but simply to work around them as best they can. This is tactics: playing the game by the rules. What I am suggesting is that translators and translation scholars be pushed toward a strategic metalocutionary rethinking of those problems that would give them useful ‘‘big picture’’ perspectives on what they are doing, and why. What is this process that we’re participating in, that keeps throwing all these paradoxes at us? Why is it that we’re expected to make Jesus sound like the boy next door, or Homer like Ezra Pound? Why are we expected to erase all traces of our own voices, our own interpretive contribution to the translation, our own situatedness in the time and place of the target reader? Rather than radicalizing students tout court, as Venuti seems to be proposing, I’m arguing that we should teach them to pay close metalocutionary attention to the problems or paradoxes of translation, and specifically teach them some strategies for working through the paradoxes to a new and transformative understanding of their role in the process as a whole, linguistic, professional, cultural, and ideological ^ an understanding that will arise out of their own current translational experiences and be applied in turn to their own future translational experiences, in an ongoing series of self-transformation that might be called ‘‘lifelong radicalization.’’ In fact the problems or paradoxes that students might be taught to recognize as triggers for self-transformative theorizing can be divided into the four categories listed just above: linguistic, professional, cultural, and ideological. 1. Linguistic. Many of the linguistic markers in the source text, because they refer specifically to the source author addressing source readers, will pose nearly insuperable problems for the translator. Any linguistic marker that specifically identifies the language of the text as the source language (‘‘As we say in German. . .’’), for example, will have this effect. More complexly, and more interestingly, deictics, including personal pronouns (I/my, you/your, we/our), demonstrative pronouns (this, that), adverbs of time and place (now, soon, here, nearby, recently, forthcoming) should awaken all the translator’s suspicions, and direct his or her attention to what exactly is going on at this spot (for a general discussion of deictics in translation, see Fawcett 1997: 94^6). Hatim and Mason (1990: 91) give a good example of the problems caused by personal pronouns, which they take from Harris (1981): The court interpreter, on the other hand, is likely to suffer from ‘‘role conflict’’. Harris (1981) notes how even seating position (nearer the defence or nearer the prosecution) can affect clients’ trust of the interpreter’s neutrality. He also relates that, in the particular case he observed, the interpreter preferred third-person reported speech to first-person speaking in order to mark her own neutrality, thus: ‘‘Le pre¤sident vous demande. . .’’ (‘‘the presiding judge is asking you. . .’’) and ‘‘Die Zeugin

Il-, per-, and metalocutionary implicature 155 antwortet. . .’’ (‘‘the witness’s answer is that. . .’’). The need for this distancing device became apparent when one of the witnesses turned on the interpreter and asked: ‘‘Why are you asking me these pointless questions?’’ (Harris 1981: 198), thus obviously identifying the interpreter as an agent of the court. The usual solution to these problems, as advocated by most translator handbooks and trainers, is to find some fairly unobtrusive compromise that will minimize the problem without actually solving it: switching to the third person, for example. Writing in English ‘‘As we say in German’’ and letting the reader deal with any cognitive dissonance that arises; letting him or her attribute it to the ‘‘willing suspension of disbelief ’’ that the reading of all translations involves. And I would agree that in actual translating tasks, this sort of compromise solution is probably best. What I’m arguing for, in other words, is not necessarily a new way of translating ^ foreignizing, say, or minoritizing, though the radicalization process may well lead to some sort of significant transformation of translation methods ^ but rather the metalocutionary use of these linguistic complexities or paradoxes to help the translator nudge himself or herself out of uncritical assumptions about what it means to be this particular person translating this particular text in this particular time and place, for this client, to this purpose, etc. Maybe I am acting as an agent of the court? Maybe in my very acquiescence to the court interpreter’s ethics of neutrality, I am complicit in the legal system that has accused you? Maybe it’s impossible for any human being to be neutral, and I should see my ‘‘role conflict’’as a clear signal that I am either for you or against you? Or do I actually take the court’s side when I interpret the judge or the prosecuting attorney and the defendant’s side when I interpret him or her? Is it possible to take sides alternatingly? In shifting to third-person address ^ translating every ‘‘I’’ as ‘‘he’’ or ‘‘she’’ ^ am I in fact creating a kind of third ‘‘side’’ for myself, a place outside the right-orwrong, guilty-or-innocent agon of the court? Obviously, court interpreters will not be able to pour these metalocutionary ruminations into their actual words in the courtroom, for fear of losing their jobs; but if the ethical considerations that such ruminations awaken in their imagination go far enough, and become disturbing enough, they may well bring about a significant change in their professional self-concept, and perhaps eventually in their job profile as well. An interpreter who begins to feel profoundly uncomfortable in the ‘‘role conflict’’ that court interpreting invariably entails may decide to change jobs, may get into community interpreting, for example, where overt advocacy is allowed; or if enough court interpreters begin to feel disquiet of this sort, an interpreters’ union may be able to bring about changes in the professional ethics governing the behavior of interpreters in court, as Sabine Fenton (1997) notes is happening in New Zealand, where interpreters increasingly are sworn in as expert witnesses rather than being treated as neutral conduits of information between speakers of two different languages.

156

Implicatures (performative uptake)

2. Professional. A whole host of professional markers or signals should alert the practicing translator that something ideologically interesting and perhaps significant and disturbing is afoot. This is the area where what Venuti calls the translator’s ‘‘invisibility’’ comes into play: work with pay but little or no recognition; the author’s name on the front cover, the translator’s name buried in the acknowledgments or not mentioned at all; translations reviewed as if they were original works, with no mention of the translator’s name, no discussion of the translation as a rendering of an original in another language, and so on. This relative invisibility may lead to growing irritation, which may lead to inquiries: what exactly is the situation here? In the nonliterary field, once the translator discovers or realizes that the source author is almost always invisible as well, that discovery may be enough to soothe the irritation: I’m not being singled out for invisibility. On the other hand, in nonliterary translation the translator’s work is regularly ‘‘touched up’’ (often, at least from the translator’s point of view, destroyed) by invisible editors and published or released in forms that the translator (if the translator ever sees it, in fact) does not like and does not want to acknowledge as her or his own. What should be done about this? Anything? Is there anything that can be done? Through professional translator associations, perhaps? If not, how do I feel about that? 3. Cultural. This is a problem area that translation theorists have always been painfully aware of, and offered various remedies to: cultures differ. It is the area that Hatim and Mason and Baker address in their applications of Grice’s conversational implicature to translation, Hatim and Mason by noting that the maxims may have to be interpreted differently in different cultures, Baker by insisting more radically that different cultures may just flat-out have different cooperative maxims (Chapter 9). Hatim and Mason’s applications of Grice in The Translator as Communicator (1997) move in a slightly different direction, coming out of Brown and Levinson (1987) on politeness: all humans, Brown and Levinson insist, possess a social fiction called ‘‘face,’’ defined as either negative face (the freedom not to be forced to do other people’s bidding) or positive face (feeling good about yourself and recognizing that others respect you as well). Polite social behavior is designed to protect the face of others (and yourself); but there are many social behaviors that threaten face, called ‘‘face-threatening acts,’’ or FTAs. Brown and Levinson list a hierarchical series of strategies for minimizing face-threat, from avoiding the FTA altogether at the top through various mitigating moves (do it off the record, do it on the record but with redressive action) to bald unredressive on-the-record FTAs. In their application of this model to translation studies, Hatim and Mason once again assume that the translator is going to want to strive for some sort of equivalent effect in the target language, which means adjusting for different politeness codes in the target culture; but applying my trans-Gricean theory of metalocutionary implicature to this problem, we can imagine a translator asking the pointed questions of himself or herself as ethical agent: how polite do I feel the source author or speaker is being to me, how polite do I want to be to my readers or listeners, etc. A translator who decides that the apparently polite surface of

Il-, per-, and metalocutionary implicature 157 a source text actually masks a deeper insulting condescension, or even malice or evil, may in grappling with that metalocutionary interpretation ultimately choose to highlight the deeper level in his or her translation, in order to pass a metalocutionary implicature on (from his or her own imagination, not the author’s!) to target readers. 4. Ideological. Well, these categories overlap significantly, of course, since in an ideological perspective all of the problems we have been considering are at root ideological, and translators are always mired in problems of transfer across cultural difference, and most problems arise in linguistic form, even professional ones (over the phone or in e-mail or snail-mail or fax correspondence, for example). But we might use the categories to draw attention here to the most obviously ‘‘political’’ issues, especially those on a left-leaning or oppositional or ‘‘radical’’ agenda, such as economic exploitation, sexism, classism, racism, heterosexism, and so on. What does the translator do when asked to translate a neonazi text? Translate it without a fuss? Refuse to take the job? Translate it with subtle metalocutionary implicatures strewn through it, with the hope of instigating a transformative process in the (presumably nazi) reader? What if you work in-house at an advertising/translation agency, one of whose clients is a photocopier manufacturer that advertises its photocopiers in terms that you begin to see as defamatory to women? ‘‘It’ll do everything your secretary does now,’’ the ad copy you are asked to translate says,‘‘except ask for a raise.’’ ‘‘You could hire a beautiful woman to do the work this photocopier does, but what would your wife say?’’And so on. Nothing flamingly misogynist, but the sexism of the ads grows on you, and begins to bother you. What do you do? Do you ask your manager to give the photocopier ads to someone else in the office? Do you recommend that the agency suggest to the company that they try out a different approach for your particular target culture(s)? Do you translate metalocutionarily, highlighting sexist cues so as to awaken the target reader’s suspicions and possibly enhance his or her subsequent radicalization? Quite clearly I do not believe that there is any one right solution to any of these situations. I have no panacea for the ethical and ideological problems facing translators once they become aware of their own ethical agency. I don’t recommend a certain translating style or method that will supposedly combat capitalist or sexist or some other hegemonic ideology most effectively. What I recommend instead is a highly local and individuated radicalization process, arising inductively and experimentally out of very specific translation jobs, reacting metalocutionarily to the paradoxes that beset all translation, and taking whatever resistant or dissident or even self-protective form the translator feels comfortable undertaking. This process should not be imagined as solipsistic, of course; it occurs in dialogical interaction not only between the translator and the various other agents involved (author or speaker, client, reader or listener) but with other translators as well, both during training (teachers, other students) and in the working world (colleagues, fellow members of translator organizations). But neither should it be imagined as ideally collective, imposed on the mass of current or future translators by some

158

Implicatures (performative uptake)

radicalizing force from above. It really only works as the empowering dialogue each individual participates in with other members of the social group, at a pace and in a direction that each individual finds subjectively (emotionally and ideologically) tolerable. The expansions of Gricean implicature in this chapter, then, move past the implicit saying of things to a series of implicit doings: acting on others in implicit ways through illocutionary and perlocutionary implicature, acting on oneself, one’s self-awareness, through metalocutionary implicature. Grice opened up an exciting new channel of speech-act analysis by postulating that we convey messages that we do not make explicit by flouting the cooperative assumptions that speakers of a language bring to any conversation; the simple act of interweaving Grice’s conversational implicature with Austin’s tripartite locution/illocution/perlocution distinction makes it clear that we can do far more than convey implicit messages in this way. We can also do implicit things with words. And one of the implicit things we can do is to learn, to become increasingly aware of the conversational situation in which we find ourselves, how others are constructing it, how we ourselves are constructing it, how those constructs limit us and how those limitations might be partially or temporarily overcome. This latter process, metalocutionary implicature, is the source not only of Grice’s theory of conversational implicature but of all theorizing: the smooth (or even not-so-smooth) flow of life and language balls up, entangles, turns back in on itself, and finally comes grinding to a halt while metalocutionary reflexivity takes over, and generates a new metaunderstanding of the current situation and its implications for wider contexts as well. And as we saw in the father^daughter example, above, it is possible for speakers to (attempt to) initiate such a metalocutionary entanglement in their listeners: the father, wanting his daughter to think more complexly about ‘‘freedom,’’ says (3) ‘‘You know, freedom doesn’t start at that front door.’’ As Bob Ashley writes in private correspondence, it is possible for the speaker to initiate what he calls a‘‘stall,’’or ‘‘vibratory stasis’’: A stream-lined, frictionless response of immediately ‘‘knowing’’ what the speaker had intended, in this instance, is exactly contrary to what the speaker intends. Hence, in a rather odd way, the deliberate intention to find a way to make the hearer ‘‘puzzled’’ initiates a stall, forcing the hearer to ‘‘dwell’’ upon the possible content, yes, but importantly, also the speaker’s modality or attitude toward what she is saying. Much of your hypothetical questioning is aimed at the hearer attempting to figure out the speaker’s own commitment to what she is saying. Only a stall in the seriality of the comprehension can allow the space for the paradigm of puzzle-piece choices to be surveyed. Of course the problem we always have in this sort of situation is that we can never know what exactly will make the other ‘‘stall.’’50 Grice himself tended, like

Il-, per-, and metalocutionary implicature 159 a good constativist, to analyze both sides of a conversation at once, as if human beings could ever know both even in the past or present, let alone the future. What we need next, it seems to me, is an expansion of Grice’s model that will account for this uncertainty, this inability each of us has ever to project an intention that is guaranteed to be interpreted correctly, or to come up with an interpretation that is guaranteed to reproduce the speaker’s intention accurately. In chapter 11 I build such a model out of the ‘‘interpretant triad’’ of Charles Sanders Peirce; in Chapter 12, apply it to problems of conversational and conventional implicature; in Chapter 13 use it to discuss another form of implicature based on ‘‘remembering’’ and invoking past speech acts. In Chapter 14, finally, I return to the possibility that metalocutionary implicature might offer a pathway of interpretive discovery for the working out of conversational implicatures and maxims across cultural boundaries.

11 Intendants and interpretants

My ruminations on illocutionary and perlocutionary implicature might be read as an answer to the question of what J. L. Austin might have thought about Grice’s theory of implicature: let us imagine implicature not just in terms of unstated information but also of implied speech acts. My ruminations on metalocutionary implicature came out of the same set of concerns, obviously, but they were more deeply motivated by an imagined answer to the related (and perhaps somewhat belated) question of what Jacques Derrida might say about Grice’s theory. Taking as my model Derrida’s deconstruction of Austin, considered in Chapter 4 (pp. 60^3), where Derrida’s strategy was to turn the excluded parasitical ‘‘remainder’’ of Austin’s argument into the engine that pulled the entire train, I surmised that the focus of his attention in Grice would be on the failure of the cooperative principle and its maxims to provide a firm foundation for rule-breaking behavior. People mean things they don’t say and still manage to communicate by flouting maxims, defying conventional expectations, breaking rules ^ which turn out not to have any stable existence that one can point to. How does that work? A Derridean performative linguist might want to flip Grice on his head by suggesting that breaking the rules creates the rules ^ that a ‘‘rule’’ (to put that neat aphorism more philosophically) is only a virtual structure that is incipiently carried and ultimately realized (invoked or even ‘‘performed’’) by the transgressive iterability of every utterance. This was the methodological shift that led to my theorization of a ‘‘metalocutionary’’ implicature. Put in terms of rulebreaking-as-rule-creating, this may sound like typical deconstructive cleverness; translated laterally out of the deconstructive idiom, however, (re)iterated as performance, it also carries a good deal of tentative or preliminary intuitive weight. For example, a friend says something odd, something that seems intuitively, because it defies easy understanding, to be breaking the rules; in order to figure out what she is saying, you construct a transgressive scenario: ‘‘She can’t mean that; she doesn’t really believe that; she’s lying to me, I can feel it; she’s lying, and she wants me to know she’s lying, so I’ll figure out what she’s really trying to say.’’ Having come up with that, I’ll project it back to the time prior to her speaking, so that the rule seemed to precede the transgression.51

Intendants and interpretants 161 Or, schematically, from the interpreter’s point of view: First comes (1) subjective puzzlement, a ‘‘stall,’’ a state of not-knowing or not-understanding, which seems to demand (2) a subjective explanation, often of an active, strenuous sort that pushes past normative bounds and then, retroactively, generates (3) a projective objectifying explanation that reverses the actual chronology, idealizes the sequence utterance^interpretation^transgression^rule as rule^transgression^utterance^interpretation, which yields (the impression of) (4) objective understanding, a sense that we now ‘‘know’’ what the speaker intended to say. But how does this work from the speaker’s (intender’s) point of view? First comes (1) subjective transgression, a sense of drawing on uncodified linguistic resources, or what Lecercle (1990) calls the ‘‘constitutive remainder’’of linguistic rules ^ a sense of speaking grammatically and with recognizable lexical items but from outside the boundaries of conventional speech. The uncertainty this wandering or errance provokes will motivate the speaker to attempt to manipulate or guide the listener’s interpretation (the four steps schematized above) so as to facilitate understanding ^ but how? Here my enumeration breaks down; I don’t know how to go on from (1), and Lecercle, whose bookTheViolence of Language I picked up in search of an answer, doesn’t know either. Lecercle is wonderful on the transformative power of the remainder, the ways in which transgressive linguistic play continually constitutes and reconstitutes language; but his interest lies more in the breakdown of communication than in the transgressive transformation of communication. What do we do to make ourselves understood if we can’t count on stable structures of rationality and cooperation to guarantee reliable interpretation? How do we transgress the ideologically controlled boundaries of rationality and cooperation and still channel the semantic and somatic impact of our words?

The interpretant triad If you can’t break through a barrier, stop, backtrack, and go around. I want specifically to try and flesh out a tentative answer to the question from two paragraphs ago (‘‘but how?’’) by making an excursus into Charles Sanders Peirce’s theory of the interpretant triad early in the last century, from a letter to LadyWelby on March 14,1909.52 This seems a useful place to go from an iterabilized interpretive act, because Peirce’s conception of the sign’s ‘‘own peculiar Interpretability before it gets any Interpreter’’ so clearly anticipates Derrida’s concept of the sign’s iterability: My Immediate Interpretant is implied in the fact that each Sign must have its own peculiar Interpretability before it gets any Interpreter. My Dynamical Interpretant is that which is experienced in each act of Interpretation and is different in each from that of any other; and the Final Interpretant is the one Interpretative result to which every Interpreter is destined to come if the Sign is sufficiently considered. The Immediate

162

Implicatures (performative uptake) Interpretant is an abstraction, consisting in a Possibility. The Dynamic Interpretant is a single actual event. The Final Interpretant is that toward which the actual tends. (1958: 414)

Implicit in this triad are Peirce’s three Universes: the immediate interpretant is a First, an idea as abstract potentiality; the dynamic interpretant is a Second, a brute reality or object as event; and the final interpretant is an Aufhebung or Third, a consummate incorporation of idea and event into a ‘‘destination,’’ a destined end as truth. Despite all the obvious problems with these three Universes, I confess to liking them a good deal, and want to suggest some things we can do with them once we’ve tinkered with them a little, ironed some of their wrinkles out. Well, perhaps ‘‘wrinkles’’ is a bit euphemistic. I have a major problem with Peirce’s conception of Thirdness. It sounds almost religious to me, like a sacred teleology. In fact, Peirce takes it from Hegel, and Hegel in the mid- to late nineteenth century did amount to a kind of secular religion, at least for intellectuals. Peirce believed, as he wrote to Lady Welby on May 20, 1911, that ‘‘there is real progress. Every science proves it, and we must believe that the world is governed by a living spirit’’ (427). But it is one thing to believe such things, quite another to found a semiotics on it. The almost invariable collapse of Peirce’s triadic thought into reductive dualisms among his non-Hegelian followers (which is virtually everybody) suggests that his ideology and his method may be irretrievably at odds. On the other hand, Peirce is useful for my performative project because he does insist so strongly, and I think rightly, that some kind of Third is needed to combat dualistic reductions. Without Peirce to remind me, I might very well relax into a vague generalized poetic acceptance of Secondness: hey, stop worrying about definitional rigor, clear categories are a constative tic, let it flow. Without Peirce, and equally rigorous performative thinkers like Anthony Pym, it would be all too easy to fall into a low-grade binary distinction between constative and performative linguistics according to which constative linguistics deals with Firsts (abstract structures in the null context) and performative linguistics deals with Seconds (speech acts in real contexts), and that’s that. Peirce hates such tidy dualisms, and is constantly pushing at them until they yield aThird. But what kind of Third will work? If we reject Peirce’s quasi-religious Hegelianism, what shall we put in its place? One standby answer to this need is the constative one: cling to an idealized Third at all costs in order to guarantee interpretive stability and thus validity (which is to say, ideality). The key to this idealization is the restructuring of Secondness so as to eliminate the system-busting variability implied by Peirce’s insistence on brute dynamic otherness (John Searle defining idealized ‘‘central’’ or ‘‘paradigm’’ cases of promising as ‘‘ordinary,’’ for example, and excluding real-world counterexamples from his analytical purview). In this sort of revision

Intendants and interpretants 163 the dynamic interpretant would become not each act of interpretation, but the idea of such an act, the idealized map or plan of such an act in the null context (or in various context types), which is implicit in and controlled by the immediate interpretant. Following Katz on performance, we might call this the dynamic interpretant ideally deprived of its dunamis and reduced (or ‘‘raised’’) to an image of stasis. Secondness would thus become a cipher in a rigid formal system,Thirdness a sham.53 Another answer, a kind of kneejerk performative one such as I imagined briefly three paragraphs ago, might be to embrace the dynamic difference of each act of interpretation so wholeheartedly that we give up all hope of aThird, or, better, that we see the interchangeability of Firstness and Secondness ^ or in Austinian terms, of locution and illocution, or in Derridean terms, of iterability-as-sameness and iterability-as-difference ^ as the only Third we’re ever going to get. This is, in fact, probably where my preliminary Derridean revision of Grice’s conversational implicature points us: to the conclusion (as I say) that breaking the rules creates the rules, or, in Peircean terms, that each dynamic approach to an immediate interpretant lays its own private hermeneutical foundation.54 But fleshing Derrida’s abstract iterability out with Gricean utterers and interpreters will take us past that, I think, to a dialogized iterability that I was beginning to develop a few paragraphs back when I got stuck. Dialogue, as Peirce himself hints more than once in his work, might provide a post-Hegelian Third that would get us past the impasses of both Peirce’s Hegelianism and Derrida’s anti-Hegelianism. Here, for example, is Peirce from ‘‘What Pragmatism Is’’: Two things here are all-important to assure oneself of and to remember. The first is that a person is not absolutely an individual. His thoughts are what he is ‘‘saying to himself,’’ that is, is saying to that other self that is just coming to life in the flow of time. When one reasons, it is that critical self that one is trying to persuade; and all thought whatsoever is a sign, and is mostly of the nature of language. The second thing to remember is that the man’s circle of society (however widely or narrowly this phrase may be understood) is a sort of loosely compacted person, in some respects of higher rank than the person of an individual organism. It is these two things alone that render it possible for you ^ but only in the abstract, and in a Pickwickian sense ^ to distinguish between absolute truth and what you do not doubt. (1958:191) Peirce’s ‘‘first thing’’ reminds us of the internal dialogism of Bakhtin that we considered in Chapter 7: ‘‘Meditation,’’ Peirce wrote in the ‘‘Neglected Argument’’ of 1908, ‘‘takes the form of a dialogue’’ (375), which suggests that Peirce would have us study not only the logic of signification but also its dialogic, the ‘‘talking-across’’ or interpersonal context in which signs are generated and

164

Implicatures (performative uptake)

modified, used and understood. The ‘‘second thing’’ points implicitly in the same direction: ‘‘Unless truth be recognized as public,’’ Peirce writes to Lady Welby on December 23, 1908 ^ ‘‘as that of which any person would come to be convinced if he carried his inquiry, his sincere search for immovable belief, far enough ^ then there will be nothing to prevent each one of us from adopting an utterly futile belief of his own which all the rest will disbelieve’’ (398). But, of course, a mere conception of truth as public isn’t going to prevent that either. What usually prevents the adoption of futile beliefs ^ and only usually, none of this being infallible ^ is no Hegelian ideal but precisely our Pickwickian situatedness in communal dialogue, in dialogized sociality. Not only must we get along with people, which means putting our ideas into interactive practice; as Bakhtin reminds us, we initially got our ideas from other people too, developed them dialogically in somewhat idiosyncratic (idiosomatic) ways along regularized (ideosomatized) guidelines adopted from other people.

Interpreters and utterers And this requires that we study signs not in the abstract, nor merely in their relation to interpreters, but as the occasions for actual dialogues between interpreters and utterers. But for some reason Peirce doesn’t feel comfortable with this explicit formulation ^ any more than Derrida does, for that matter, six decades later. Earlier in the March 14, 1909, letter to Lady Welby he compares his interpretant triad to her own sense^meaning^significance triad, and notes that The greatest discrepancy appears to lie in my Dynamical Interpretant as compared with your ‘‘Meaning.’’ If I understand the latter, it consists in the effect upon the mind of the Interpreter that the utterer (whether vocally or by writing) of the sign intends to produce. My Dynamical Interpretant consists in the direct effect actually produced by a Sign upon an Interpreter of it. They agree in being effects of the Sign upon an individual mind, I think or upon a number of actual individual minds by independent action on each. (1958: 413) Quite clearly, Peirce’s conception lacks the interpersonal aspect of Lady Welby’s: Peirce posits a sign in a vacuum, or at most a sign divorced from its communal context and interpreted in privacy by single individuals. But aren’t signs used? Aren’t they intended to be interpreted in certain ways by people, motivated by the desire to bring about a certain kind of responsive act? Later in the same letter Peirce tries to clarify his position: My Interpretant with its three kinds is supposed by me to be something essentially attaching to anything that acts as a Sign. Now natural Signs and symptoms have no utterer; and consequently have no Meaning, if

Intendants and interpretants 165 Meaning be defined as the intention of the utterer. I do not allow myself to speak of the ‘‘purposes of the Almighty,’’ since whatever He might desire is done. Intention seems to me, though I may be mistaken, an interval of time between the desire and the laying of the train by which the desire is to be brought about. But it seems to me that Desire can only belong to a finite creature. (1958: 414) Here Peirce’s problematic relation to Christianity (the collectivized ideological utterer that uttered him, generated him as sign, to which he as interpreter later struggles to respond as independently as possible) begins to trip him up. Why not speak of the ‘‘purposes of the Almighty’’? ‘‘Since whatever He might desire is done’’ only begs the question: that’s a specific theological construct of a possible god that binds neither that god as utterer nor any given human being or society as interpreter. Take the once-popular prophet of doom Hal Lindsey for instance, in The Late Great Planet Earth: writing in the late seventies, Lindsey took the recent history of conflict in the Middle East to be a God-given sign that the Biblical end of the world was near. The return of Israel to the Holy Land in 1948, the bitter hostility of the Arab states and Russia to Israel, the inflamed issue of oil ^ all this ‘‘proved’’ to Lindsey and his followers that God intended to destroy the world very soon (in 1988, in fact, one Biblical generation after the return of Israel55 ). Whether there really was a god such as Lindsey posited, and whether that god actually ‘‘uttered’’ these specific intentions in the Book of Revelation and other biblical apocalypses, had nothing to do with it, really. The dialogue was specifically set up by Lindsey and other popular apocalyptists, as interpreters of the signs of the times. Things don’t just sort of ‘‘act’’ as signs, as Peirce the constative thinker wants to insist. Interpretation doesn’t just ‘‘attach’’ itself to ‘‘anything that acts as a sign.’’ People interpret, and when they interpret they posit an utterer, who intends something by the sign. The sign ‘‘becomes’’ a sign only in a dialogical context. Whether there really is an utterer who intended the sign to be interpreted in just that way is a secondary (ontological rather than pragmatic) question. And in the interpretation of human signs ^ words, say ^ the role played by an utterer in shaping the interpretation should be plain. In fact, I don’t see how we can avoid speaking of intention; but here again Peirce mysteriously evades the issue. Intention as ‘‘an interval of time between the desire and the laying of the train by which the desire is to be brought about’’ is a mechanistic structural abstraction (in fact, an iterability) that in pragmatic terms will take us nowhere. The ‘‘interval of time’’ depersonalizes desire into a formal lack, the silence between two beats of the metronome. Pragmatically speaking it would have been more productive to define intention precisely as the ‘‘train’’ that is laid in order to bring the desire about; but even this assumes that desire is brought about in an automatic, mechanistic fashion, simply by marking out the path of its fulfillment. If, looking at it dialogically, we can assume that the utterer desires the proper understanding of his or her intention (and a responsive act

166

Implicatures (performative uptake)

based on that understanding), then it might make sense to see the ‘‘train’’ by which the utterer tries to bring that desire about as the intentional equivalent of the interpretive ‘‘train’’ by which the interpreter tries to understand the utterer’s intention. So far from mechanically carrying out some abstractly conceived desire, in other words, this ‘‘train’’ might well be understood as the sign itself, imagined dialogically as both the interpretive apparatus for reaching an understanding of the utterer’s intention and the intentional apparatus for shaping the interpreter’s interpretation. This would mean expanding Peirce’s interpretant triad ^ in fact, doubling it, mirroring it with an intendant triad. If, for example, the immediate interpretant in Peirce’s scheme is an abstract potential of interpretability, an ability to be interpreted, it is obviously linked to, but not necessarily identical with, something that I want to call the immediate intendant, an abstract ability to be used for intentional purposes. I’m not particularly happy with the ‘‘abstract’’ part of that: as the somatic theory presented in Part II would suggest, the ‘‘conventions’’ of language or of any other semiosis are not abstract at all, not mentalist ideals floating about in some realm of forms or recorded in dictionaries or language users’ heads, but felt, programmed into our somatic responses.56 For the most part, in fact, our somatic responses to words are controlled so effectively by our ideological programming that the immediate elements, the interpretant and intendant, will be identical; but they need not be, as in cross-cultural misunderstandings (my topic in Chapter 14), or ^ an extreme case ^ when a chimpanzee smiles and we humans take the smile for an expression of pleasure. If we knew more about divine intendants it might be possible to say whether God’s immediate intendant and Hal Lindsey’s immediate interpretant were identical; but since we don’t, we have only the interpretant triad to go on there.57 Even where the immediate elements are close enough that we might as well consider them identical, however, a shift to the dynamic element reminds us that the ideosomatically programmed ‘‘immediate’’ interpretability/intendability of a sign is never omnipotent (and in fact, because it is mediated by programmed somatic response, it is never immediate either). People respond complexly (‘‘dynamically’’) to specific situations. Their ideosomatic response is always complicated and deepened by idiosomatic response, so that, say, the ‘‘meaning’’ of the word ‘‘cat’’ for me will be shaped by muscle and emotional memories not of cats in general, but of actual cats my family had when I was small, and other cats I have known or experienced during my life, and by my feelings about them. If I grew up hating cats the word ‘‘cat’’ will have a negative charge for me; I’ll hear hissing and spitting in it, I’ll feel the stiff back and the claws. If I grew up loving cats my response will be very different ^ depending, of course, on other extenuating circumstances, like how much time I spent with the cats of my childhood. If I was an only child and spent hours every day with a favorite soft cuddly cat, obviously I’m going to have a different response to the word ‘‘cat’’ than if I had lots of siblings and lots of other animals, and liked the cats well enough but didn’t really pay that much attention to them.

Intendants and interpretants 167 All this will contribute to my dynamic (response to the) use of the word ‘‘cat’’ in conversation ^ my dynamic intendant if I’m using it, my dynamic interpretant if I’m responding to someone else’s use. But not just that: also my mood, my reading of the situation, my emotional needs and sense of my dialogical partner’s emotional needs, all the details that make up an actual situation or Second. A woman says of another woman,‘‘I can’t stand that cat,’’and I respond dynamically to the hostility in her voice and actions, out of my own sense of the woman she’s referring to, my uneasiness with this (for me alien) scratching and hissing sense of the word ‘‘cat,’’ my ideological disapproval of traditional feminine competition, and so on. It’s all relevant. And it all has an effect both ways, for both intention and interpretation. Since all but the most ritualized dialogues involve varying degrees of idiosyncratic creativity on both the interpretant and the intendant sides, both participants in a dialogue will have to be continually constructing new images of the entire ‘‘train,’’ interpretive images when they are interpreting, intentional images when they are intending (we will be looking more closely at this problem in ritualized dialogues in Chapter 13). To be understood, the utterer will have to take into account not only the somatically programmed or ‘‘conventionalized’’ meanings of signs, the immediate element, but whatever he or she knows of the interpreter’s dynamic response to certain words and phrases, certain attitudes, certain scenarios; and to understand and respond properly, the interpreter will have to take into account not only those same conventionalized responses (assuming they are the same) but whatever he or she knows of the utterer’s dynamic responses. These dialogical images of the other person’s participation in the conversation I want to call ‘‘final’’ ^ the final interpretant and the final intendant. They aren’t final in Peirce’s sense of being ‘‘destined’’ to be reached by every intender or interpreter; they’re only final in the pragmatic sense of forming the basis for action. The final interpretant is the interpreter’s understanding of how to take the sign’s interpretive potential (its conventional or common meaning as he or she feels it) in this particular context in order to apprehend how the utterer is using it. And the final intendant is the utterer’s attempt to impose his or her own dynamic shape on the sign’s intentional potential in order to guide the interpreter to an appropriate interpretive response. Here it is shown in a diagram (see Figure 11.1).

DIrp UTTERER FInd

IInd/IIrp DInd

(FInd ¼ final intendant DInd ¼ dynamic intendant IInd ¼ immediate intendant

Figure 11.1.

FIrp INTERPRETER

FIrp ¼ final interpretant DIrp ¼ dynamic interpretant IIrp ¼ immediate interpretant)

168

Implicatures (performative uptake)

Note that, in a reconstructed chronological sequence, the immediate element here is, in fact (as Peirce would say), First: it’s programmed into us before we know what’s happening, before we are conscious of learning language, and we bring it like a somatic reserve to every dialogue we enter. The dynamic element is Second: it is our specific response to specific situations, our feel about certain signs not in general but in specific real dialogues. Putting the final element Third rounds the schema off nicely, gives it an idealized touch: we synthesize the immediate and dynamic elements and generate an overall interpretive or intentional image of each utterance.

Performative dialogic In actual dialogues, however ^ as I began to suggest in the beginning of this chapter, in my discussion of how breaking the rules creates the rules ^ things usually go the other way. We understand what someone is saying, we say what we want to say, i.e. we start with a final interpretant or final intendant that contains and in some sense constitutes the dynamic and immediate elements as components of our interpretive or intentional act. Peirce himself leans toward the idealized or ontologized (constative) former approach, what in the beginning of this chapter I called the ‘‘anachronologized’’ approach, wanting to see Thirdness as a synthetic destination toward which antithetical Firstness and Secondness strive. But he too says (in the ‘‘Neglected Argument’’) that ‘‘the third Universe comprises everything whose being consists in active power to establish connections between different objects, especially between objects in different Universes’’ (359, emphasis mine), which suggests that final interpretants and intendants don’t ‘‘represent’’ meaning but generate it, constitute it as a connectedness. Meaning doesn’t ‘‘exist’’ in signs; people use signs to create meaning dynamically, and signs ‘‘take on’’or are ‘‘invested with’’ meaning only in and by that process. In a performative view, the process would go like this. The immediate intendant is a regulatory tool which the utterer actualizes dynamically (using the dynamic intendant as regulatory act), in accordance with his or her final image of intentional connectedness: actualizes it as a norm-fiction, which the utterer tries to convince the interpreter is normative, as an aid to apprehending the turn or deviation apparently enacted in the dynamic intendant. Similarly, the immediate interpretant is an explanatory tool which the interpreter actualizes dynamically (using the dynamic interpretant as explanatory act), in accordance with his or her final image of interpretive connectedness: actualizes it as a calculus-fiction, an explanatory device that doesn’t bind the dynamic interpretant but comes to stand in some kind of significant relation to it. Needless to say, the process doesn’t always work. The final elements may diverge: the participants in a dialogue may misunderstand each other. It happens all the time. Speech is too fast, too ‘‘ballistic,’’ as Daniel Dennett says (see Chapter 5, above), for utterers and interpreters to run through the

Intendants and interpretants 169 cumbersome analytical apparatus I’m schematizing here with any kind of close systematic accuracy. We wing it, both when we intend and when we interpret, and winging it leads to frequent misunderstandings. But misunderstanding is never ideally ‘‘final’’: the close dialogical interaction of final interpretant and final intendant always make it possible (though never inevitable) to move past misunderstanding to a pragmatically ‘‘true’’ or working understanding. The utterer may fail to regulate the interpreter’s dynamic act of interpretation: may lay an inadequate ‘‘train’’ for the conveyance of his or her intentions, may provide insufficient guidance to interpretation, and, as we saw in Chapter 10 on illocutionary, perlocutionary, and metalocutionary implicature and will see again in more detail in Chapter 13 on conversational invocature, may even be only partly conscious of his or her own intentions. This would obviously make the regulation of interpretation rather haphazard. But if the interpreter indicates by his or her response that the final interpretant diverges from the final intendant, it is still possible to build another ‘‘train,’’ to restructure the final intendant in a new utterance. Even misinterpretation can teach the utterer to direct his or her regulatory dialogical image more effectively toward this particular interpreter, or to become (metalocutionarily) more conscious of it him- or herself. Or the interpreter may fail to explain the utterer’s dynamic act of intention: may interpret inattentively, may misconstrue the utterer’s construction of a regulatory norm. But again it is possible to test interpretations by acting on them (verbally or otherwise) and thus provoking new attempts at clarification. Misinterpretation too can teach the interpreter to structure his or her explanatory dialogical images more accurately, to work toward a better image of the utterer’s intentions, and for that matter of his or her own interpretive behavior. Intention and interpretation are thus creative acts ^ but acts that must submit for success to the potentially distortive force of a dialogical other. As we have seen repeatedly in this book, of course, there are also deliberately distortive dialogical acts, uncooperative acts by which an utterer misleads an interpreter or an interpreter misreads an utterer. Misinterpretation can teach both participants in a dialogue to structure their dialogical images more appropriately, but doesn’t do so automatically. The whole range of lying, cheating, swindling, pretending, evading, manipulating, and the like on the one hand, and deliberate deafness or misunderstanding on the other, reminds us that just idealizing understanding doesn’t make it come true in the real world. Still, the difficulty we have lying effectively to people we know well suggests that long dialogical experience with a given utterer allows an interpreter to construct relatively reliable final interpretants even despite misdirective (un- or anticooperative) final intendants. Every time we engage in conversation with a person we work to understand not only his or her intendants but an interpretative sense of how that person constructs intendants, and perhaps also how that person’s typical processes of intendant-construction resemble or differ from those of other people. This means that effective lying will require the construction of an effective final intendant whose purpose is not understanding

170

Implicatures (performative uptake)

but misunderstanding: one that directs interpretation away from the ‘‘truth’’ (the uncomfortable intendant that the utterer wants to conceal) toward a dummy final intendant that is both harmless (will not cause embarrassment or anger or other unpleasant emotions) and plausible (will fit the interpreter’s dynamic sense of what is going on in the context). This also suggests that intendants and interpretants may work on several levels: in the case of lying, for example, we would expect to find a concealed intendant, a dummy intendant that is made public, and an overarching intendant that seeks effectively to present the latter in place of the former. We will be looking at such complexities in greater detail in Chapters 12 through 14, and returning, in fact, in that concluding chapter, to modify the expanded Peircean diagram radically to account for cross-cultural misunderstandings. Similarly, deliberate misunderstanding by definition entails an attempt to conceal true understanding, and this attempt is relatively transparent to people we know well ^ unless we are able to construct such an effective final interpretant that we convince the utterer that we really didn’t understand. Even so, after a number of such responses the utterer may realize our inability to understand is no interpretive failure (which would reflect a failure on his or her part to structure an effective final intendant) but an opposing final intendant with a message like: I don’t want to hear what you’re saying. Grice is probably right: the best guarantee that a dialogue will progress toward mutual understanding would probably be a certain cooperative spirit, a certain goodwill. But of course, outside of the kind of ideal model constative linguists favor, we can never be absolutely sure that a dialogue is progressing in such a spirit, still less that an expected future dialogue will progress in such a spirit, nor even that a perfectly cooperative dialogue (if we could ever point to one) would produce understanding. All we know is that sometimes people do cooperate, and that sometimes cooperative dialogues lead to mutual understanding.

12 Conventional implicature and language change

Now let’s return to Grice. Specifically, in this chapter I want to start slotting my modified Peircean framework into Grice’s theory of conversational and conventional implicature. The doubled triadic model with immediate, dynamic, and final intendants and interpretants will at this point enable us to distinguish much more clearly than Grice manages between conversational and conventional implicature, and the schema I offer here will also suggest further useful complications of the notion of implicature in Chapters 8 and 10, as well as of the Bakhtinian model of language change that I offered back in Chapter 7.

Of moons and square-dancing Let us imagine a scene.58 A thirteen-year-old boy is spending the summer with his rather conventional grandmother. She keeps coming up with wildly implausible activities for him, like carrying groceries to shut-ins, singing songs at the rest home where a friend of hers lives, or working on her jigsaw puzzle with her.What he mostly wants to be doing is practicing his skateboard jumps down at the high school, watching MTV, playing video games, and chatting with his friends back home on ICQ (it turns out his grandma bought a computer with a 28.8 modem about five years ago and hasn’t even taken it out of the box; it’s a dinosaur, but it’s better than nothing, so he sets it up for her, downloads ICQ and a few other good things for her, and spends most of his time at home on-line). He is a good boy, and doesn’t want to hurt his grandmother’s feelings, but he is not about to waste his summer doing stupid boring things, either, so he keeps having to invent excuses. Finally, though, he can’t take it any longer. She has found an ad in the paper for square-dancing lessons down at the parks department, and keeps at him and at him to sign up for them. He feeds her one excuse after another, but nothing avails. After three or four days of this he grows exasperated, and says to her, keeping his voice light and friendly: ‘‘You know, Grandma, what you should do with that square-dancing?’’ ‘‘What?’’ ‘‘You should fold it four ways. . .’’

172

Implicatures (performative uptake) ‘‘Yes?’’ ‘‘And stick it,’’ big friendly smile,‘‘where the moon don’t shine.’’ ‘‘Doesn’t shine, dear.’’ ‘‘Right.’’ ‘‘What does that mean, dear?’’ He just rolls his eyes. ‘‘So what about the square-dancing, dear? Can I sign you up?’’

Now, clearly, the grandmother has not read this phrase ‘‘fold it four ways and stick it where the moon don’t shine’’ to mean no, absolutely not, and stop pestering me about it! She’s never heard the phrase, obviously doesn’t know that the third-person don’t goes with it. This ignorance, presumably, would make the boy’s concealed message conversational implicature ^ something that she should be able to work out. If, of course, she has heard it many times, and understands what her grandson is driving at without having to work it out (say she only pretends to be a harmless sweet old granny, and, in fact, is worldly and tough as nails), it would be conventional implicature. How can we distinguish analytically between the two? Are we forced to conclude that, because the difference lies in the subjectivity of each individual interpreter, there is no ‘‘real’’ difference at all, that conventional implicature doesn’t really exist? That argument too has been made. Or are we to conclude that the difference exists, but purely subjectively, purely in the mind of the interpreter? Ultimately, yes, probably; but that is not a big problem in a performative linguistics. There is, in fact, a good deal more that a performative linguist can say about that ‘‘purely subjective’’difference, and I want to say it, here and in the next two chapters. To begin with, then, the grandmother has to work out the implicature. Her grandson has carefully avoided giving her any body signals that would make it clear just how violently he is rejecting her latest suggestion; she is going to have to use Gricean pragmatics to figure out what he’s saying. What does she do, analytically?59 Using the dual Peircean model, we can tentatively imagine the grandmother starting (perhaps almost unconsciously) with an immediate interpretant entailing linguistic meaning ^ word-meaning, sentence-meaning, dictionary or denotative meaning ^ and then trying to make that jibe with a dynamic interpretant, in order to generate a final interpretant or meaning. An interpretant is, of course, an interpretive image of an intendant, so that we could also express that last sentence by saying that the grandmother begins with an image of the immediate intendant and tries to make that jibe with her image of the dynamic intendant in order to generate a detailed image of the boy’s final intendant (not just no, but stop pestering me with stupid suggestions). But since I do not propose to explore here in this chapter the possibility of misunderstanding, the situation in which intendants and interpretants at all levels most strikingly and significantly diverge ^ I want to save that problem for Chapter 14 ^ let me cut corners a bit analytically by not distinguishing between intendants and interpretants, referring instead to immediate, dynamic, and final elements.

Conventional implicature and language change 173 This does not mean that intendants and interpretants necessarily or even typically coincide; only that their (quite common, indeed at the broadest level inevitable) discrepancies are not relevant to my purposes here. Let’s start with the dynamic element. It would include at least two subelements: context, the specific situation in which the utterance is uttered (in this case the grandmother’s proposal and the boy’s submerged exasperated response to it), and reference, the act of referring or pointing verbally to specific real-world actions and objects (in this case presumably the square-dancing, the moon and its shine, and the actions of folding and sticking). The dynamic element is considerably more complex than this, but again, let’s keep things as simple as possible. Even at the level of analytical simplicity I’m proposing, this model proliferates complexity at an alarming rate. In Gricean terms, the dynamic context is where the question of cooperation would be determined, and the dynamic reference is the channel of information. According to Grice’s model, in order to have conversational implicature we must have a conflict between these two: an assumption on the interpreter’s part that the speaker is cooperating, which clashes with the perception that the information or reference is (apparently) uncooperative. In this case, we can imagine the grandmother assuming that her grandson, while failing to cooperate fully with her in the square-dancing idea (not saying yes), is cooperating verbally or conversationally in the sense of speaking frankly, not attempting to conceal his true feelings or intentions from her. Assuming that the boy wants to convey his true feelings, then, if the grandmother finds the actual utterance puzzling, she will perceive a clash between dynamic context and dynamic reference and feel compelled to work out implicature. In other words, even in this stripped-down model (I’m going to be adding complications as I go), the interpretive conflict that prompts the need to work out implicature is located within the dynamic interpretant, which is to say that in the case of conversational implicature the dynamic interpretant is in itself conflictual. By the same token, the speaker wanting to implicate something without spelling it out will need to build a conflictual dynamic intendant, projecting contextual cooperation and apparently uncooperative reference. The dynamic intendant is potentially very complex in this case. One could argue that the boy both does and does not want his grandmother to work out the conversational implicature: does, so she’ll let him be, but does not, so she won’t get her feelings hurt. But let’s set that complication aside. Let’s just say, again, for the sake of argumentative economy, that for a need to work out implicature to arise, for either the speaker (intended implicature) or the listener (interpreted implicature), the dynamic element must be conflictual, involving projected/assumed cooperation in the context and apparent noncooperation in the reference. In order for the dynamic reference to signal apparent noncooperation, that appearance must be grounded in the immediate element, in (again, starting simply) the linguistic meaning of the utterance, what Austin called the locution,

174

Implicatures (performative uptake)

or what constative linguists would call ‘‘the meaning of the utterance in the null context.’’ We can now unpack that immediate element with ease: Immediate element: linguistic meaning of ‘‘You should fold that squaredancing four ways and stick it where the moon don’t shine!’’ Word-meaning: You ¼ the specific contextual addressee(s) should ¼ ought to; are encouraged to fold ¼ crease and bend over on top of itself that ¼ the particular thing named next square-dancing ¼ a certain type of dancing; the grandmother’s plan to send her grandson to square-dancing classes; the ad in the paper for square-dancing classes four ¼ 4 ways ¼ manners or fashions; times and ¼ do this next thing in addition to the former thing stick ¼ put or insert roughly it ¼ the same thing referred to above as ‘‘that square-dancing’’ where ¼ in a place specified in the next four words the ¼ there’s only one of the thing named next moon ¼ the earth’s natural satellite, conceived here as a body reflecting sunlight don’t ¼ negates the action mentioned next; it doesn’t happen (marked syntactically for lower-class address, possibly, in a middle-class speaker’s speech, for humorous purposes) shine ¼ give off light Sentence-meaning: fold the square-dancing four times and put it in a place protected from moonlight. If this is the nonimplicative meaning that the grandmother generates out of the coincidence of the immediate element and the dynamic reference, it doesn’t work, clearly. It doesn’t make sense. It doesn’t seem cooperative. What does moonlight have to do with the square-dancing classes? What difference does it make whether the moonlight shines on the classes or not? What possible relevance could the four folds have? And how do you fold square-dancing anyway? Could he be referring to the newspaper in which I saw the ad? In order to get from this analytical impasse to a final interpretant, the grandmother has to factor in two contextual elements: her grandson’s apparent lack of enthusiasm for her plan, and the absolute absence of signals that he is lying or being evasive or otherwise refusing to cooperate conversationally. If, then, the boy is cooperating conversationally (trying to tell his grandmother what he thinks) and refusing to cooperate behaviorally (trying to say no!)

Conventional implicature and language change 175 then (the grandmother might be imagined as speculating) the place protected from moonlight must be a place that the speaker associates with distance and darkness, a lack of pressing visibility, out-of-sightness and thus out-of-mindness. ‘‘Hm,’’ we can imagine this grandmother musing about her grandson’s final intendant.‘‘If the moon doesn’t shine on it, presumably the sun doesn’t either. And possibly electric lights as well. Which would mean that neither he nor I could see it. It would be hidden away somewhere. Could he be telling me, nicely, to set the square-dancing idea aside? To put it in some out-of-the-way place, like the basement, or some junk drawer, where he won’t ever have to see it or hear about it again? Could he be implying something like ‘Forget about it, Grandma’?’’ Clearly, here, the grandmother has reached an adequate final interpretant: no. She has missed what Charles Altieri would call her grandson’s expressive implicature entirely, of course ^ the fact that the implicated place where the moon don’t shine is her grandmotherly rectum, that her sweet grandson is expressively implicating a fairly uncomfortable and degrading procedure (‘‘shove it up your ass’’) ^ but this may be quite satisfactory for her in numerous ways. She might be shocked, horrified, scandalized, disgusted, deeply troubled to find herself forced to think about such a procedure so graphically. She might prefer not to receive such an overt invitation from her beloved grandson’s lips. All she really wants to know is whether he’ll go to square-dancing or not, and having reached the final interpretant no, that’s enough. She doesn’t need to understand the full range of his expressive final intendant. However, note that if she is determined to build a comprehensive expressive final interpretant, the materials for the requisite analytical process are potentially present here ^ namely, in the lower-class syntax of ‘‘where the moon don’t shine.’’ If, as the grandmother’s correction suggests, the boy is a middle-class child who otherwise does not use third-person don’t, his use of it here signals some sort of expressively significant ‘‘slumming,’’ some rhetorical descent into substandard English and the values presumably reflected by it. From the grandmother’s point of view, those values would be associated with crudeness, a lack of (middle-class) decency and propriety: this would signal, though perhaps still only vaguely, the expressive implicature behind ‘‘stick it where the moon don’t shine.’’ It may well be that she could not get from ‘‘stick it where the moon don’t shine’’ to ‘‘shove it up your ass’’ on your own; that the analytical process that would bring this grandmother to a full expressive final interpretant would involve asking a friend what the phrase means. This would entail a collectivized model of implicature that to my knowledge Grice and his followers nowhere deal with: implicative ‘‘uptake’’ becomes possible only through dialogical consultation, only through a group effort. I won’t explore it further here either; but want to mention it as a significant avenue for further research. Why shouldn’t the working-out of conversational implicature involve collective efforts? Grice’s assumption that an individual

176

Implicatures (performative uptake)

will be able to work out the implicated meaning on his or her own, without group consultation, is just another reflection of the individualistic culture(s) in which he lived. So what do we have so far? Immediate element ‘‘fold up the square-dancing and put it in a place protected from moonlight’’ Dynamic element Dynamic context emotional/behavioral context: failing to cooperate verbal/conversational context: appearing to cooperate Dynamic reference ‘‘fold up the square-dancing and put it in a place protected from moonlight’’ Final element expressive implicature: ‘‘shove it up your ass’’ (possibly missed by this interpreter) metaphorical implicature: ‘‘put it in a place where I won’t have to see it or hear about it’’ actional implicature: ‘‘I’m not going square-dancing’’ What makes this series of embedded implicatures interesting for my purposes here is that at some indefinite or at least shifting point along a sorites series between what is said and what is meant, the implicatures become conventional: if the grandmother has heard the phrase ‘‘stick it where the moon don’t shine’’ before, that is conventional implicature; if she hasn’t heard that one, but has heard ‘‘shove it up your ass,’’ and either reaches that expressive final intendant on her own or has someone spell it out for her, then the expressive implicature is conversational and the metaphorical implicature is conventional. If she has heard both before, they are both conventional implicatures. If she has heard neither, then both are conversational and the third-level implicature is conventional. Identifying this shift, then, will help us not only distinguish between the two but trace the origins of implicative and other conventions.60 What is still missing in this formulation, obviously, are the maxims. Where do they fit in? This is where the revised Peircean model gets interesting. Clearly, for the model to work at all, the maxims would have to be part of the immediate element, part of the ‘‘a priori’’ baggage passed on to all speakers by the culture; but how do they get into that baggage for individual speakers? How do they become part of the default operating system for intenders intending and interpreters interpreting? I suggested above that the maxims are ‘‘activated’’ or ‘‘actualized’’ dynamically in the act of intending or interpreting: that the specific (dynamic) need for a maxim as an explanatory backdrop for implicature gives each maxim what conscious or active analytical form it has. In other words, the maxims are only

Conventional implicature and language change 177 ‘‘immediate’’ as an unconscious or ideosomatic potential that is activated by the intender or interpreter needing to build or explain a dynamic conflict. They do not exist in consciously articulated form prior to their dynamic activation. They are not agreed-upon rules which all speakers have consciously memorized (let alone ‘‘negotiated’’!) and then consciously ‘‘flout.’’ In the act of intending or interpreting, flouting comes first, and only secondarily ‘‘pulls’’ the ‘‘maxim’’ up out of immediate potentiality into actual dynamic play (though as we will see at the end of this chapter the relationship between the immediate and the dynamic is considerably more complicated than this). Thus the grandmother (who is also becoming, in my reconstruction, something of an amateur performative linguist) might be imagined as being driven to construct a series of maxims to explain the clashes she perceives between reference and context ^ a series that it might be useful to present hierarchically, as if she typically started at the top and worked downward: Maxims governing the behavior of grandchildren asked to engage in some activity deemed interesting or appropriate or educational for them by grandparents, as might be presumed to be lurking somewhere in the ideosomatic/immediate mind of the grandmother in our scenario, and activated by her in order to work out the speaker’s implicature: Maxim 1: cooperate behaviorally (do whatever the important adults in your life ask you to do) Maxim 2: cooperate emotionally (be friendly, speak in a friendly tone) Maxim 3: cooperate verbally (tell your parents and grandparents (etc.) whatever you know without trying to hide anything) Maxim 4: don’t use lower-class English Maxim 5: don’t refer to taboo body parts Maxim 6: don’t encourage people to perform actions that might cause unnecessary physical discomfort We might, then, rewrite the complete schema with the maxims included under the immediate element and then noted parenthetically at the relevant insertion point for each, the place where the grandmother would need to postulate or‘‘activate’’each in order to make sense of what she is hearing andthinking: Immediate element Linguistic meaning ‘‘fold up the square-dancing and put it in a place protected from moonlight’’ Maxims (M) Dynamic element Dynamic context emotional/behavioral context: showing an unwillingness to cooperate (violating M1)

178

Implicatures (performative uptake) verbal/conversational context: acting friendly (obeying M2), appearing to cooperate (obeying M3), but using lower-class language (flouting M4) Dynamic reference ‘‘fold up the square-dancing and put it in a place protected from moonlight’’ Final element expressive implicature: ‘‘shove it up your ass’’ (apparently obeying M5, but in the dynamic context of flouting M4 ultimately flouting M5 too on a deeper level) metaphorical implicature: ‘‘put it in a place where I won’t have to see it or hear about it’’ (flouting M6) actional implicature: ‘‘I’m not going square-dancing’’

Or, narratively: the grandmother notes straight off that her grandson is not particularly inclined to cooperate with her plan to send him to square-dancing classes (M1). However, he is polite about it (M2) and does not seem to be lying or otherwise concealing his true intentions (M3), and for the grandmother this is, as Grice predicts, an important platform from which to launch her final interpretant. The grandson’s obvious violation of cooperative Maxim 1, combined with his apparent compliance with Maxims 2 and 3, together tend to make the grandmother conclude that the dynamic reference ‘‘fold up the square-dancing and put it in a place protected from moonlight’’ must be referring to a dark or hidden place, or even, given his flouting of Maxim 4, a crudely negative or unpleasant place, like the rectum, but euphemistically, in accordance with Maxim 5 (based on suspicions aroused by his flouting of Maxim 4, the grandmother might imagine the boy as superficially obeying Maxim 5 by not referring explicitly to her rectum, but flouting it at a deeper and more implicative level, referring implicitly to her rectum in ‘‘where the moon don’t shine’’). The resulting expressive implicature ‘‘shove it up your ass,’’ taken literally, then flouts Maxim 6, leading to a metaphorical implicature ‘‘put it in a place where I won’t have to see it or hear about it,’’ leading finally (via the assumption that the boy must have a good reason to flout all these maxims) to a third-level implicature ‘‘I’m not going square-dancing.’’

Immediatization and language change The sprinkling of maxim-activations through these various ‘‘insertion points’’ in the dynamic and final interpretants suggests, as I adumbrated in Chapter 11, that in the speech habits of individual language users maxims originate dynamically, as explanatory fictions in specific conversational contexts, and only gradually, through frequent repetition of conversations like this one, get ‘‘immediatized’’ as a priori maxims or rules of thumb governing how the utterer or interpreter (in this case the grandmother) should allow people

Conventional implicature and language change 179 to speak to him or her. This immediatization process will most likely not be entirely or even mostly idiosyncratic; in the grandmother’s case, for example, it will be guided by her experience as a grandchild herself, fifty years ago, and her experience as a parent watching her mother interact with her child; and her mother’s and grandmother’s experience will have been guided by other grandmothers, and so on: what we might call, following Simeoni in Chapter 6, socialization to the grandmother’s habitus. And the grandmother is certainly not likely to feel the ‘‘immediacy’’ of the maxims as purely idiosyncratic: like other deep-seated elements of her grandmotherly habitus, they will feel like givens, even universals, not just rules governing grandmotherly interactions with grandchildren, or adult interactions with children in general, but commonsensical rules of thumb for all human beings. As Daniel Simeoni made clear, one’s habitus tends to come to seem universal, so that the deviations of others from that habitus appear irrational, even perverse. In the model schematized above, then, both the linguistic meaning and the maxims are gradually immediatized dynamic elements: linguistic meaning through tens of thousands of dialogical encounters, as Bakhtin’s double-voicing theory (Chapter 7) and my somatic theory (Chapter 5) suggested, and the maxims through the gradual somatization of social/professional habitus (Chapter 6). What the shifting scale between conversational and conventional implicature suggests, furthermore, is that conventional implicature is born out of an immediatization of the dynamic elements of context and reference as well. How, for example, does the grandson in my scenario learn to use the phrase ‘‘stick it where the moon don’t shine’’? We can imagine him first hearing the phrase ‘‘shove it up your ass,’’ and at first being tempted to take it literally; having worked out the (conversational) implicature that time, however, every time he heard it afterwards he could skip the analytical processes detailed above and go straight to ‘‘I won’t do it’’ or some other contextually appropriate negative final interpretant. The conflicted dynamic interpretant assumed cooperation þ apparent noncooperation will have been immediatized, so that it goes without saying ^ or rather, without analyzing ^ and as a result the implicative final interpretant goes without saying as well, becomes conventionalized. Later, perhaps, he hears the phrase ‘‘stick it where the moon don’t shine,’’ and once again has to work out the (conversational) implicature, that it is a cute humorous euphemism for ‘‘shove it up your ass’’; but as he hears it again and again, repetition immediatizes the conflictual dynamic interpretants, until that implicative final interpretant too is completely conventional. Immediate element (immediatized) linguistic meaning (immediatized) maxims

180

Implicatures (performative uptake) Dynamic element dynamic context emotional/behavioral context: cooperative? verbal/conversational context: cooperative? dynamic reference Final element conversational implicature Immediate element (immediatized) linguistic meaning (immediatized) maxims immediatized context immediatized reference Dynamic element dynamic context emotional/behavioral context: cooperative? verbal/conversational context: cooperative? dynamic reference Final element conventional implicature

This constitutes, in fact, a performative schematization of language change, based on the conventionalization of conversational implicature, to put it in Gricean terms, or, to put it in my modified Peircean terms, the immediatization of dynamic intendants and interpretants. Linguistic novelties born of verbal creativity are ‘‘implicative’’ in the broadest sense possible: they mean things that have to be worked out, because they have never been expressed in those terms before; they use language in strikingly new ways that at first resist or thwart interpretation. They do so, however, through a successful (final) manipulation of the clash between immediate and dynamic intendants, producing successful final interpretants: interpreters work out the conversational implicatures involved. So successful is the linguistic novelty that gradually its use spreads; as more and more people are exposed to it, it becomes easier and easier to construct a successful final interpretant, until its dynamic element increasingly begins to be immediatized. Everybody recognizes it ‘‘immediately.’’ Everybody knows what it means and how it is used. Nobody has to think about it. Nobody has to plan a final intendant for it, work out a final interpretant for it. These final elements are created subliminally, by the ideosomatic unconscious. Eventually, it appears in dictionaries: the immediate element is, after all, typically thought of as the ‘‘dictionary meaning’’ of a linguistic sign. The sign’s once novel but now routine meaning and use ^ its ‘‘conventional implicature’’ ^ is so patently obvious to everyone, so ‘‘immediate,’’ that it comes to seem like its ‘‘property,’’ one of its ‘‘semantic fields.’’ It begins to seem, in fact, ‘‘abstract’’: etymologically, after all, what is abstract is ‘‘pulled away’’ from real-world contexts. The immediate element seems to have no contextual variability, no

Conventional implicature and language change 181 situational surprises: it has become so worn down through repetitive use that it seems to inhabit the constative ‘‘null context.’’ It has become what appears to be a mentalist fiction, purely transcendental and ideal.61 But it started as conversational implicature. It started as a performative novelty that people had to struggle to understand. And the traces of that novelty are still there, in the sign’s ideosomatic substrata, in the archeology of its double-voicing; and the performative linguist, inclined to see things in terms of actions, will want to uncover those traces, unbury them.

13 Conversational invocature

‘‘Implicature’’ in Grice’s definition implies that unstated information is conveyed to an audience that is able to work out what is being said by reference to cultural/linguistic maxims that are being flagrantly flouted. As I have been suggesting all through Part III, this is too narrow a definition; certainly it is not invariably what happens when people say things in a new way, so that their listeners do not immediately understand them. I have suggested that we need to expand Grice’s theory to include illocutionary and perlocutionary implicature, doing things implicitly with words, and metalocutionary implicature, learning by doing things implicitly with words (Chapter 10), and a more complex analytic of dialogue (Chapters 11 and 12). Here in Chapter 13, I want to offer a substantially different kind of expansion. A more general description of the broader category that I’ve been exploring here under the expanded term ‘‘implicature’’ might be that people tend to make reference to the new and the old simultaneously: something new, requiring for interpretation the (re)construction of something old. In Grice’s theory, conversational implicature arises when something someone says (the new) doesn’t ‘‘fit’’ what we know or surmise about the context and we decide that the person must be flouting a conversational maxim (the old) in some controlled (‘‘cooperative’’) way. Conventional implicature, as I showed last chapter, is what happens when the ‘‘new’’element in conversational implicature grows old. But there are other kinds of newness in speech, and other kinds of oldness as well. In this chapter I want to explore another implicative new^old axis, what I want to call ‘‘invocative conversational implicature’’or simply ‘‘conversational invocature’’: implicating something new by invoking or alluding to something old, something familiar, something that the speaker assumes the listeners know. This is an old idea, of course; reframing it in the ‘‘new’’ context of Gricean implicature, however, will yield us more insight into language change.

I remember Let me set this new expansion of Grice up slowly by building62 a series of miniature short stories, or fictional vignettes, based on the ‘‘I remember’’ line

Conversational invocature 183 from the Que¤be¤cois example that Anthony Pym took from Annie Brisset back in Chapter 3, the Que¤be¤cois actor speaking on the Montreal stage the lines that Michel Garneau translated into Que¤be¤cois French from Act IV of Shakespeare’s Macbeth: 1. ‘‘Remember last night?’’ he said, stepping close to her, putting both arms around her, burying his face in her neck. They had made love. This was a ritual they had developed a year or two after they had been married: every time they made love one of them would ask, some time later, whether the other remembered ‘‘last night’’ or ‘‘this morning’’ or whenever. And the other would reply that s/he did. She said it now. ‘‘I remember,’’ she said, and smiled warmly up into him. 2. ‘‘Remember last night?’’ he said, looking across the breakfast table at her. He tried to put remembered warmth into his words, but they felt flat to him. Not horribly wrong; just slightly off. Maybe a bit mechanical. Had they repeated this ritual too many times? Was it going stale? Maybe their marriage was. Their lovemaking the night before had been kind, gentle, but perhaps just a bit distracted. She paused a few beats before looking up from her paper. She smiled wanly. ‘‘I remember.’’ 3. ‘‘Remember last night?’’ he said, with a mischievous smile. They had had too much to drink, and had gotten silly. They had taken their clothes off and dressed up in goofy costumes, he in swim fins and a swim mask and snorkel, she in a cow costume they had bought for her one Halloween, udders over her belly, and tried to make love, but they had been too drunk, had been laughing too hard. They had fallen asleep half-costumed in each other’s arms. She shook her head and laughed a little, giving him a fond I can’t believe us look.‘‘I remember,’’ she said with perhaps a bit too much emphasis on remember. But her eyes were shining. 4. ‘‘Remember last night?’’ he said. They had had friends over, a couple whose marriage was going bad. Their guests had sniped at each other all evening, until finally their acrimonious exchanges had escalated into a fullscale battle right there in their living room, the guests beyond caring that their friends were hearing all this: detailed accusations of infidelity on both sides, detailed descriptions of both partners’ sexual inadequacies, all at the top of their lungs. She pulled her mouth back in an eek grimace. ‘‘I remember,’’ she said. 5. ‘‘Remember last night?’’ he said sardonically, in response to her cold accusation that he never wanted to make love to her any more. They had made love the previous night, but it had not been pleasant for either of

184

Implicatures (performative uptake) them. They had not looked each other in the eyes. Their touch had concealed worlds of bitter disaffection. Her eyes widened in angry disbelief that he should invoke their old ritual now, of all times. ‘‘I remember,’’ she snapped back.

Taking these five little vignettes as successive snapshots of the married life of the same couple, we note first off that they form an example of what Jacques Derrida calls iterability: the ‘‘same’’ question and answer are repeated over and over in this couple’s life together, in a variety of contexts (which in Peircean terms means significant alterations in the dynamic element), and with a resulting transformation of overall meaning (as Peirce would say, the final intendants and interpretants of the utterances vary from dialogue to dialogue). I start with a highly ritualized dialogue in (1) because the repetition of the same words in almost the identical context with almost identical reference (the last time they made love) and with almost identical feelings seems to give it the appearance of a formal system: the ritualization of performative creativity makes actual speech seem to approach the kind of idealized laboratory conditions (the ‘‘null context’’) that constative linguists favor. In Peircean terms, in (1) the dynamic element seems almost totally subordinated to the immediate element, giving the final element the appearance of decontextualized or at least decontextualizable stability. But then, of course, vignettes (2^5) stand as reminders that the dynamic element is still very much present in (1) as well: change anything in the dynamic element ^ the reference (what is remembered, lovemaking or a fight) or the feelings accompanying the remembering (warm love, routinized love, laughing love, bitterness at the end of love) ^ and the dialogue means different things. In order for the couple to go on generating ‘‘the same’’dialogical interaction over and over, they have to keep performing it in the same way, in the same contexts, with the same emotional accompaniment. The sameness of the dynamic element is something that they have to work to maintain: keeping it the same is a performative act, not a default setting. Take away even a little of the somatics of loving memory, as in (2), and the dialogue begins to feel flat, mechanical, like an empty ritual. Mechanical repetition of empty ritual is somewhere near the constative null-context ideal, of course; enough repetitive sameness in our speech and it becomes much easier to pretend that context can usefully and even realistically be ignored. In this series of vignettes, however, it is clear that the mechanical repetition is still performed ^ staged, voiced, therefore dynamic ^ and thus also somatized, felt: the couple can feel the discrepancy between what they once felt when they performed the dialogue and what they feel when they perform it after ho-hum going-through-the-motions lovemaking. And, of course, when the dynamic shifts are even more obvious, as when what is remembered is not the same old act of lovemaking but drunken horseplay (3) or another couple’s fight (4), or when the loving feeling in both the remembered sex and the remembering itself has metamorphosed into something much uglier (5), this performative quality becomes inescapable.

Conversational invocature 185 Indeed, the tensions between the dynamic and immediate intentions/interpretations in (3^5) point us to an important observation in (1): the immediate element in (3^5) that the couple plays off dynamically in using the ritualized dialogue to refer to drunken horseplay or another couple’s fight or the incipient breakdown of their marriage is not the abstract ‘‘meanings’’ of the words in the ‘‘null context,’’as a constative linguist would want to insist; it is rather the specific expression of love that we saw in (1) (and saw beginning to waver in (2)).63 As my tentative model of linguistic change at the end of Chapter 12 suggested, through a series of repeated dialogues the couple immediatizes the dynamic element in their exchange, incorporating more and more of the love they are expressing into the immediate element, so that for both of them the immediate intendant and interpretant of saying ‘‘Remember last night?’’ and ‘‘I remember’’ become no longer the linguistic meaning of those sentences but rather something like ‘‘I love you’’ and ‘‘I love making love with you.’’As should by now be clear from the chapters on iterability, somatics, double-voicing, and language change, language use is always accretive: because we perform language with our bodies, and our bodies store our performances somatically, even the immediate element is only relatively stable, based on so many somatically stored performances that it comes to possess a cumulative inertia that resists change. But that inertia is itself the product of change: it must be built up.64 The dialogue ‘‘Remember last night’’/‘‘I remember’’ would not mean ‘‘I love you’’ or ‘‘I love making love with you’’ for every couple; that final intendant/interpretant is ritualized by this particular couple. And ritualization means specifically, in Peircean terms, that the repetition of the same final element gradually begins to immediatize it, turn it into an immediate element, which thus becomes available for dynamic modification, as in (3^5). It is possible, for example, to imagine the further ritualization of (5): 6. ‘‘Remember last night?’’ he said sardonically, invoking the new form their old ritual had recently assumed, as the love and harmony in their relationship gradually crumbled. They had had a terrific fight the previous evening. This morning, as always before, their bodies felt bludgeoned by cruel words; they were too battered and exhausted to fight any more, but unwilling to leave themselves vulnerable by forgiving and forgetting. She made a disgusted mouth, gave her morning paper a quick snap. ‘‘I remember,’’ she said tightly. The dialogical ritual has changed, here, taken on a new ritual form. It still serves basically the same function in the marriage, helping the spouses to remember something that they don’t want to forget ^ how wonderful their lovemaking is, at first (1^4), how bad their fights are, later ^ but the bitter final element in (5) has now seeped down through an oft-repeated dynamic context until it has been transformed into a new immediate element. Like a Hegelian synthesis becoming the thesis in a new dialectic, so too with sufficient social

186

Implicatures (performative uptake)

repetition does a Peircean final element become the immediate element in a new triad: ‘‘the meaning of a word,’’as Wittgenstein (1953/1984: no. 43) wrote in the Philosophical Investigations, ‘‘is its use in the language,’’ and this specifically means every use. If some group, no matter how small, uses a word or a phrase in a certain way enough times, for the members of that group that particular idio- or sociosyncratic usage gradually sinks down into ‘‘immediacy,’’ takes on, as we saw at the end of Chapter 12, the characteristics of what comes to feel like ‘‘dictionary meaning.’’ This ritualization or conventionalization or immediatization of in-group usages is the source of the phenomenon familiar to all of us whereby families or close groups of friends or coworkers come to use common words in a way that outsiders do not understand, or invent words to mean things for which there seems to be no commonly accepted term. This is, of course, the origin of slang, argot. Inside jokes and references have a similar origin: through frequent repetition a story based for its effect on the (dynamic) element of surprise becomes so well-known (unsurprising) to the members of a group that they can use it, within the group, as an immediate element in a wide range of other dynamic contexts. These new immediate elements can in fact display a remarkable persistence, surviving for generations, even eventually being transformed into in-group legend or myth. The ‘‘distortions’’or alternations emerging from this process patently resemble those found in the game of Telephone or Gossip, where the ‘‘same’’ message is whispered from one person to another down a long row, and the last person in the row reports what s/he thinks s/he heard, in order to compare it with what the first person said ^ and everyone laughs at the changes the message underwent. Unlike that game, however, the changes in the series of vignettes (1^6) are rung through persistence: once some dynamic element of the dialogue is ritualized, conventionalized, somatized as the ‘‘immediate’’element, it remains fairly stable, only gradually succumbing to variations if they are repeated enough times. This is basically the same model I developed in Chapter 12; the only real difference so far is that the socius for whom the linguistic novelties in this series of examples are ‘‘immediatized’’ is far smaller than the one in Chapter 12, a single married couple rather than the entire language community. I am narrowing the socius down in this manner partly in order to pinpoint the performative process involved: it is much easier to trace the immediatization of dynamic elements in the conversational history of a socius of two than in an entire language community. In addition, however, I want to focus on the immediatization or ritualization of ‘‘insider allusions’’ in order to begin to extend my argument’s analytical trajectory: instead of stopping at the conventionalization of linguistic novelty, as I did in Chapter 12, so as to illustrate the performative origins and iterative/ somatic channels of apparently abstract or mentalist (constative) semantics, I want to move on to the implicative manipulation of these conventionalized references in what I’m calling conversational invocature. In Chapter 12, I suggested that what Grice called the conversational maxims are, in fact, just

Conversational invocature 187 immediatized rules derived from repeated dynamic exchanges, and that (as Grice says) conversational implicature emerges out of the controlled manipulation of those rule-based expectations; here I want to argue that shared ‘‘insider’’ knowledge too is the immediatized traces of repeated dynamic exchanges, and conversational invocature emerges out of their controlled manipulation. Where the Gricean conversational implicator flouted immediatized expectations, however, the conversational invocator simply alludes to them. This makes it just as possible as in implicature to mean things without spelling them out. Let’s see how that works. In all six vignettes given above, for example, what can we imagine being ‘‘remembered’’? On the locutionary surface, obviously, the husband is asking the wife whether she remembers what happened last night: a simple factual question. But unless she is suffering from an advanced case of dementia praecox or Alzheimer’s disease, surely he assumes she remembers what happened. Below the ‘‘factual’’ surface, then, what is just as obviously happening is that he is encouraging her to remember what happened: the apparent question is an illocutionary speech act, performing the action not just of seeking information or confirmation but invoking a memory, a memory of an event and, perhaps, of the feelings surrounding that event. Because he does not invoke the memory outright ^ doesn’t tonalize his utterance as a command, say: ‘‘Remember last night!’’ ^ this might be called implicative invocature or invocative implicature. But so far the ‘‘invocature’’ involved is of a very tame variety. It invokes a memory, but doesn’t do anything with it. The husband means something he doesn’t spell out, but only in the most technical and nit-picky sort of way. Where the concept of invocative conversational implicature takes on significance is at a deeper level of remembrance: remembering not just what happened last night, but also what ‘‘we’’always say the morning after; invoking the memory not only of the specific nonverbal event, in other words, but of the verbal ritual as well. In vignettes (1^2), for example, the couple invokes a dialogical ritual in order to reaffirm their love for each other; invoking that ritual means invoking its history, obviously. What the husband wants the wife to remember is not just that they made love the previous night, and not only that last night’s lovemaking was just the most recent in a long string of such loving encounters; but also, and more importantly for a performative linguistics, that this dialogue he is replaying with her is itself just the most recent in a long string of such loving dialogues. This deeper implicative/invocative level is even more obvious in vignettes (3^5), where the specific previous-night event whose memory the husband is invoking is significantly different from the event of previous remember-lastnight dialogues. In (3), the husband and wife do not make love, but they play out their little question-and-answer ritual the next morning anyway. How does that work? What significance does the ritual have for them? If the ritual were purely conventional, if such verbal rituals could only work conventionally, using it to refer to a non-lovemaking event would render it meaningless. The conversational invocature in (3) involves an old element, remembering the

188

Implicatures (performative uptake)

linked histories of their lovemaking and their ritualized dialogues about their lovemaking, and a new element, remembering an event that seemed destined for lovemaking but didn’t quite make it there ^ yet which left an emotional aftertaste that was similar to that invoked in past dialogues referring to lovemaking. The conversational invocature in (3) might be identified as a way of extending the couple’s ritualized/dialogized (‘‘conventionalized’’) memories of their own lovemaking to include a making of love that ended up being not particularly sexual. By invoking the dialogized history of their lovemaking and their drunken clowning in the same words, the husband can be imagined as conveying a message to his wife without spelling it out: that he loves her as much when they are silly as he does when they have sex; that silliness can for him be a kind of sex. The old remember-last-night dialogue, ritualized or immediatized for reference to lovemaking, is conversationally invoked in order to implicate something novel, something new; presumably, given the convergence of their memories of events, emotional charges, and verbal rituals, she is‘‘immediately’’able to work out this invocative implicature as well. And basically, the same thing happens in vignettes (4^6), too. In (4) what is implicated through the invocature might be something like ‘‘aren’t we glad our marriage is better than theirs; aren’t we glad that our love is still strong, our sex still loving and harmonious?’’ (What is actually implicated cannot be predicted perfectly: the husband may mean one thing by it, and the wife may understand another from it. This happens all the time, as anyone who has ever been married, or in a long-term relationship, knows well. As I showed at the end of Chapter 11, there is no guarantee that the final interpretant will correspond to the final intendant. And since performative speech acts are always intensely local, steeped in the actual context and everything that speakers bring to that context ^ their own personal experiential histories ^ any implicature that I predict in this series of hypothetical vignettes is pure speculation.) In (5) the invocature is somewhat more complex. She has accused him of never wanting to make love any more; he responds by reminding her that they just made love the previous night. Invoking the memory from last night, then, might be imagined as generating a first-level implicature on the order of ‘‘you’re wrong and I’m right.’’ But that implicature only encompasses the actual event, the sheer fact that they made love. If what is invoked in the new reiteration of the old dialogue is (at least) triple ^ what happened, how they feel about what happened, and how the history of their past iterations of this dialogue suggests that they should feel about what happened ^ then the obvious and painful truth that their lovemaking did not feel particularly loving might generate a second-level implicature on the order of ‘‘sure, we made love last night, but it was pretty awful, wasn’t it, and if it was that bad, why on earth are you bitching at me for not wanting it more often?’’And if the invocation of the whole history of their verbal ritual makes them feel all the more strongly the discrepancy between what they now feel at its iteration and what they used to feel, what indeed the ritual was designed to make them feel ^ loving toward each other ^ then the husband’s invocation of the old ritual begins to feel a bit

Conversational invocature 189 kneejerk: she accuses him of not wanting to make love any more because, say, she wants to underscore just how bad their relationship has become, how far it has deteriorated, how little hope they have left; he reacts half-unconsciously only to the most obvious level in what she’s saying, missing her deeper implicature, and invokes their old dialogue in a way that serves to underscore her point, that things are bad, very bad, especially in pointed comparison to how they were back when that verbal ritual meant ‘‘I love you’’ and ‘‘I love making love with you.’’ Now it still means that at the immediate level, perhaps, but something like the exact opposite at the dynamic level, ‘‘I hate you,’’ or perhaps ‘‘I hate making love with you,’’ or something a little milder but no less bleak: the husband’s invocative implicature is not really worked out, not quite planned, based on a very vague and underprojected final intendant. It just pops out of him, and reveals, perhaps, more about how he feels than he really wants it to. And so, when she responds, she does so with some exasperated heat: ‘‘Her eyes widened in angry disbelief,’’ I wrote, ‘‘that he should invoke their old ritual now, of all times.’’ She remembers, as her words explicitly say, not only that they made love last night, but also, implicitly, invocatively, how bad it was last night, and how good it used to be, and how this cute little dialogue used to go, and how insurmountable the differences have become between the sweetness of those earlier times and the anger and bitterness of this. In other words, the invocative implicature of ‘‘I remember’’ might involve something like ‘‘I’m right and you’re wrong, yes, of course, we did make love, but how, how did it feel to make love, just as bad as I’m saying it is in general these days between us, so that, in fact, you’re absolutely right to invoke what we used to say and feel in this same situation, because invoking it shows us just how wrong we both are to keep holding on to something that was once but is no more. . . .’’

J’m’en souviens It should be easy enough to work out the complexities in (6) along similar lines; it is, in any case, more or less an immediatization or ritualization of the bad feelings in (5) ^ assuming that the couple decide to stay together, for the kids, for financial reasons, for whatever other reason, but cannot resuscitate the love they once felt for each other. Instead of spending more time on that, let me start to bring this chapter to a close by adding one more vignette to the series, this one returning us full-circle to the translation example we started with back in Chapter 3: 7. ‘‘‘Tu t’souviens tu d’la nuit dernie're?’ he said,’’ she said, glancing over at Claude to see how he was taking this. He was lying back on the bed, his hands laced behind his neck. He seemed fine. After her divorce from her first husband, she had not wanted to have anything to do with men for a long time. Then she met Claude, and melted. He was

190

Implicatures (performative uptake) a French-speaking Quebecker, and a different breed of man than her anglophone ex, soft, intuitive, passionate, sensitive. She had had four years of French in high school and two at university, and they had spoken French from the start, she her Parisian variety, he his Que¤be¤cois; but gradually, over the year they had been together, living with him now in Montreal, she had begun to pick up more and more of his Quebeckisms, and to love them. She had also begun more and more to identify with the francophone cause, and to associate anglophone Canada, of which she too was undeniably a representative, with her ex. And in some strange way this retelling of the verbal ritual she and her ex-husband had played out so many times was invoking for her something of the Que¤be¤cois nationalist sentiments in which she found herself immersed: she was telling Claude about the time her ex had invoked the ritual after a particularly unpleasant round of sex, and his question of whether she remembered it seemed like the ultimate verbal assault. ‘‘Qu’est-ce que tu re¤pondais?’’And what did you say? ‘‘‘J’m’en souviens.’’’ I remember. ‘‘Pis tu t’en souviens, non?’’And you do, don’t you? ‘‘Oui, j’m’en souviens.’’ I do.

Here the concept of conversational invocature helps us see how this Que¤be¤cois French translation (for which thanks to Manon Bergeron) of an English dialogue can be a translation and still perform a new implicative action in a new (Que¤be¤cois nationalist) context. Conversational invocature implicates things by invoking shared knowledge, knowledge that the speaker knows or assumes the listener possesses. At the immediate level, ‘‘J’m’en souviens’’ is an accurate translation of an immediatized (conventionalized, denotative) ‘‘I remember.’’ But in the dynamic context of speaking French to a francophone Quebecker in Montreal, the translation takes on the additional reference of the nationalist slogan stamped onto all Que¤bec license plates, and it accretes new invocative/implicative complexity: Immediate element Linguistic meaning ‘‘I recall the event to which your question refers’’ Maxim only tell your new lover stories that cast him or her in a better light than the old lover, and your present with him or her in a better light than the past with his or her predecessor Dynamic element Dynamic context speaking Que¤be¤cois French with a francophone Quebecker in Montreal telling a new lover a story about an ex-husband that casts the ex in a negative light (obeying the maxim)

Conversational invocature 191 Dynamic reference in the translated quote: ‘‘I recall the event to which your question refers’’ outside the translated quote, in the untranslated present: ‘‘I still remember the dialogical event I’m quoting, and the feelings I had while I was living it’’ Final element first-level implicature: she has enough French, and enough distance on her marriage, to translate her ex accurately second-level implicature: she was hurt by her ex, and still harbors some bitterness about it (‘‘remembers’’) third-level implicature: francophone Quebeckers too have been ‘‘hurt’’ by anglophones, and still harbor bitterness about it (they too ‘‘remember’’) The key point for a study of conversational invocature in translation, here, is that the dynamic reference in ‘‘I remember’’ works on two different planes: as a quoted past utterance, which she is translating, and as a simple statement in the present. She quotes herself (in French) saying ‘‘I remember’’ in the past (in English); and she is also (taken by Claude to be) saying (in French) that she remembers in the present. Quotation marks cannot normally be heard in speech; in the complexly double-voiced dynamic intendant of this dialogue, therefore, she is both quoting/translating that she remembered and stating that she remembers, and Claude’s dynamic interpretant must encompass both voices, both dynamic references. His actual response,‘‘And you do, don’t you,’’ suggests that he is recognizing her quoting (signaled by the transitional rhetoric of and) but mostly focusing on her words as a present-time statement to him, to the effect that she does remember all that now: he wants, we can imagine, to express sympathy for her past pain by recognizing that it continues into the present. His response does not make clear whether he also incorporates into his final interpretant the third-level implicature of Que¤be¤cois nationalism (in the vignette that level was explicitly built into her final intendant). In any case, this analysis does explain how Michel Garneau’s ‘‘m’en souvenir’’ can be an explicit translation of Macduff ’s ‘‘I remember’’ and at the same time an implicit invocation of Que¤be¤cois nationalism, and thus, harking back once again to Pym’s analysis in Chapter 3, both a translation and a performative speech act. Translations are speech acts: they are aimed by a speaker or writer at an audience; they perform whatever actions that particular dialogical context seems to require. If those actions add something to the ‘‘original’’ text, or shift the performative ground presumably occupied by the original text, that does not make the translation any less a translation; the ideal according to which a text is only a translation if it changes nothing in the original is a constative illusion that needs to be dispelled by a properly performative analysis. And in any case, the complexity of invocative (and other) conversational implicature that I have been tracing here in speech is also found in

192

Implicatures (performative uptake)

writing: conversational invocature is the channel through which literary allusion is designed to flow, and as every professor of literature knows, it doesn’t always flow. In the classroom, in fact, it hardly ever does. Literary allusions are typically constructed by the author (though often as unconsciously as by the husband in (5)) for an audience presumed (again, often unconsciously) to have read and remembered the same texts that the author has read. Training in literary interpretation, especially when it is directed at a single text or author, is in large part a process of accreting that kind of shared knowledge, accumulating a knowledge of what the author one is studying had read and how s/he would have been likely to allude to it. It is, in fact, a process similar to the one Claude is going through in getting to know the anglophone woman he is in love with: learning to know what she knows, so he can better understand her allusions. And just as indifferent lovers never bother to learn that stuff, neither do indifferent readers about their authors. Hence the tearing of hair and gnashing of teeth that literature professors indulge in with regard to how little their students have read, how minimal their ‘‘cultural literacy’’ is. But this is a bit unfair: it takes a lifetime to get to know another person well; why should it be any different with a good book? The indifferent lover doesn’t get detailed lectures on the required material, of course; but the lectures directed at indifferent readers by literature professors are often dry and off-putting and too little oriented toward the student’s own active construction of ‘‘literary meaning’’ to have much of a motivating effect. The indifferent lover also doesn’t have to take tests on it (at least none explicitly designated as ‘‘tests’’), and doesn’t have only fifteen weeks to prove to have learned a respectable fraction of what has been taught. Otherwise, the skills are very similar. The literature student who protests ‘‘but why didn’t he just say this stuff outright?’’ ^ which is to say, why all this implicative invocature ^ either has never been in a long-term relationship where working out such invocature is critical to success, or has never made (perhaps never been encouraged to make) the connection between interpreting the speech of someone you love and interpreting the writing of a book you love (or maybe the student just didn’t love the person or the book).

14 Metalocutionary implicature and cross-cultural misunderstandings

Grice’s maxims and cultural relativity Let’s pause here at the beginning of this last chapter to take a look at what we’ve got so far, and what we’ve done with it. In Chapter 8, I launched a two-pronged critique of Grice on conversational implicature, one prong new, the other borrowed from previous critics: Grice’s maxims are culturebound (the old critique), and our conversations, no matter how complexly implicative, are not necessarily grounded in either rationality or cooperation (the new critique). In Chapter 10, I slotted Grice’s notion that we implicate information conversationally into Austin’s tripartite locution/illocution/ perlocution framework, showing that what Grice theorized was basically ‘‘locutionary’’ implicature, and for a fuller analysis of complex implicative speech acts we needed a theory of ‘‘illocutionary’’ and ‘‘perlocutionary’’ implicature as well. This movement spawned a new possibility, that of ‘‘metalocutionary’’ implicature, which thwarts easy understanding in order to bring about a new selfreflexive understanding, a conceptual readjustment that helps the speaker and/ or listener to reframe the purposes and conditions and assumptions underlying the current conversation and his or her own role in it. Metalocutionary implicature was, in fact, the first instance, as I noted at the beginning of Chapter 10, of my apparently counterintuitive suggestion that ‘‘breaking the rules creates the rules’’ ^ that we first formulate rules, including implicatures and the maxims they seem to be based on, out of specific interpretive needs, and then project them backward into the prehistory of the current conversation as a preexisting framework that conditioned the speaker’s apparently transgressive conversational behavior. As I suggested there, the metalocutionary process by which we generate ‘‘rules’’ or ‘‘maxims’’ out of our need to impose some sort of interpretation on confusing conversational behavior might be imagined as moving through the following four steps, a kind of hermeneutical movement in the interpreter’s head: a.

subjective puzzlement (generating the need for)

194 b. c. d.

Implicatures (performative uptake) a subjective explanation (generating the need for) a projective objectifying explanation (yielding an impression of ) objective understanding

When we don’t know how to proceed, when we need to understand but find ourselves at a loss for how to construct a tentative working image of speaker intention that will allow us to move forward in the conversation, we are brought to an uncomfortable stop, plunged into a ‘‘stall,’’ as Bob Ashley says, and find ourselves forced to regroup. We remain locked in that stall until we come up with a hypothesis of what the speaker might mean, or even what the speaker must mean, which enables us to proceed: based on that hypothesis we then construct an objectifying explanation that will, in our imaginations, prepare the ground for the speakers saying what they said, behaving conversationally as they behaved. They must be flouting a maxim, and the maxim must be thus-and-such. In Chapter 11, then, I slotted Grice’s model of implicature into Charles Sanders Peirce’s interpretant triad in order to show analytically not only the different levels on which the conflicts between what is said and what is meant operate, but also the ways in which the ‘‘immediate’’ or ‘‘preexisting’’ or ‘‘given’’ aspect in any conversational encounter is, in fact, the accretive (iterative) byproduct of thousands of previous dynamic speech acts. To put it simply, all analytical models ^ those like Grice’s, based on the assumption that a solid foundation of rational cooperative behavior actually exists and must exist for complex communication to be possible, as well as those like my own postfoundational version of Grice and Peirce ^ reflect the apparently ‘‘objective understanding’’ (4) that is generated metalocutionarily out of the hermeneutical movement from subjective puzzlement (1) through subjective explanation (2) to a projective objectifying explanation (3). This is where analytical models come from. This is where rules, maxims, schemata, formulae come from. Something doesn’t make sense, some conversational force stops or ‘‘stalls’’ us, and we pull back and try to figure it out. Once we’ve come up with a hypothesis, we project our explanation back in time to some imaginary period before the puzzling behavior we are trying to understand, and adjust the (ana)chronology so that the explanation becomes an objective pattern that preexisted not only our formulation of it but our interpretive need to formulate it. The etiology of an axiology is typically an anachronology. At the end of Chapter 12 and Chapter 13, then, I slotted Grice into the Derridean framework of iteration: the theories of conventional implicature in 12 and conversational invocature in 13 were both grounded on the notion that implicatures become ‘‘familiar’’ and thus both conventional (12) and available for in-group back-reference or allusion (13) through a process of iterative immediatization. The more a group repeats an implicature, the more it shifts from the dynamic element in my expanded Gricean/Peircean framework ^ the

Metalocutionary implicature and cross-cultural misunderstandings 195 ‘‘new’’ ^ to the immediate element, the ‘‘old.’’ This explains how, after a group has iterated a given implicature enough times, its members can imply things not only by flouting maxims but by invoking familiar past speech acts; and, further down the line, after many more iterations, how its members begin to use these implicatures unconsciously, conventionally. This slow unpacking of the ‘‘new’’ problem I identified in Grice ^ how it is possible to imagine conversational participants making sense of complex implicated speech acts without a stable foundation in rational cooperation and shared maxims ^ may have been analytically necessary, but it also led us on a four-chapter detour from the ‘‘old’’ problem, the cultural relativity of Grice’s maxims. In this last chapter, I want to double back and pick up that thread again, reviving both the problem and the tentative solution offered by metalocutionary implicature in the now expanded analytical context provided by the Gricean/Peircean model of Chapters 11 through 13. In a way, of course, the example I developed for Chapter 12, with the thirteen-year-old boy and his grandmother, itself involved considerable cultural difference. The boy was born after the advent of MTV, which the grandmother may never have watched; the vast complex mindscape of internet culture ^ web surfing, chat rooms, instant messaging, email listservs, on-line video games, and so on ^ is a familiar part of the boy’s life and but a vague rumor to the grandmother; he and his friends watch movies and comedy shows on HBO and Showtime and learn turns of phrase that the grandmother is likely never to come in contact with. But they are still speakers of the same language, products of the same national culture ^ just different temporal instantiations of that culture.The boy may be puzzled by his grandmother’s obsession with square-dancing, the grandmother in turn likewise by her grandson’s obsession with video games and cable television and the internet, but they both recognize the temporal differences that separate them: the boy, that his grandmother represents an earlier version of the culture that produced him, the grandmother, that her grandson represents the latest version of the culture that produced her. Still, a full analysis of the interpretive process by which the grandmother strives to make sense of her grandson’s suggestion that she ‘‘fold the squaredancing four ways and stick it where the moon don’t shine’’ would need to delve more deeply into the difficult metalocutionary process of exploring one’s own ignorance, one’s own lack of explanatory tools for understanding this or that apparently insurmountable puzzler in ordinary conversations. In many ways we all come from different cultures, because life nudges and prods us in so many different ways; even siblings close in age grow up with cultural differences due to birth order and gender, and those cultural differences often widen as the siblings leave home, enter different professions, live in different regions, etc. The ‘‘maxims’’ we bring to any given conversational context will be different enough that we can never count on a shared ‘‘axiological’’ foundation (though of course we most often do assume a shared foundation, and are repeatedly baffled when our assumptions are defeated). Whatever maxims we need in order to move toward an interpretation (or a sense of our own intention) will have to be

196

Implicatures (performative uptake)

formulated metalocutionarily in each problematic conversational context ^ in each‘‘stall’’ brought on by confusion, by misunderstanding. So, for example, the grandmother in my scenario in Chapter 12 might be imagined as forced to give metalocutionary consideration not only to what maxims her grandson might be flouting, but also, and in many ways more important, what maxims she herself is operating by, and whether they might differ from those her grandson lives by: Maxim 1: cooperate behaviorally (do whatever the important adults in your life ask you to do) Maxim 2: cooperate emotionally (be friendly, speak in a friendly tone) Maxim 3: cooperate verbally (tell your parents and grandparents (etc.) whatever you know without trying to hide anything) Maxim 4: don’t use lower-class English Maxim 5: don’t refer to taboo body parts Maxim 6: don’t encourage people to perform actions that might cause unnecessary physical discomfort Most people, perhaps, would not enter a conversation like the one I imagined in Chapter 12 with a clearly articulated list of maxims like this; they would have only a very vague sense of what they expected or assumed, and would only formulate it as six distinct maxims, if at all, in response to a strong interpretive need. Even if we imagine this particular grandmother as someone who knows exactly what she believes, however, and who could spell these maxims out for you at any time without careful rumination, there still remains for her the question of what her grandson knows of such maxims. If she is locked into her own cultural values, of course ^ and almost all of us are almost all of the time ^ this will not be a high priority for her. It may not even occur to her to ask such a question. These maxims are (must be!) universals, as they were for Grice. But if the grandmother is still intellectually alive, and lets herself be stalled by the very strangeness of her grandson’s suggestion to her ^ fold it four ways and stick it where the moon don’t shine? ^ and lets the stall nudge her into metalocutionary reflections on what this whole situation implies about her own assumptions and his, and the differences between them, then she might well begin to consider whether her grandson shares her taboo on the naming of body parts or the use of lower-class English. She may reflect that she has on numerous occasions heard him say ‘‘butt,’’ as in ‘‘we kicked their butts.’’ This is mildly shocking to her. And while she has never heard him use third-person don’t before, or ain’t, or double negatives, he does say (we can imagine her now reflecting) ‘‘This is me’’ instead of ‘‘This is I,’’ ‘‘I wished I woulda gone’’ instead of ‘‘I wish I had gone,’’ ‘‘I laid down on the couch for a rest’’ instead of ‘‘I lay down,’’ ‘‘So I’m like ‘Did you go?’’’ instead of ‘‘So I said,’’and so on. Does he share the same culture with her? Can she assume that he and she would bring the same conversational maxims to bear on any exchange? If she perceives him as violating one of her maxims, must she assume that this is a maxim violation for him too? Or is it

Metalocutionary implicature and cross-cultural misunderstandings 197 simply normal speech for him? Gricean analysis of conversational implicature requires an assumption of cultural universals, or at the very least a shared culture in the current conversation; and indeed we do proceed on something like that assumption in most of our conversational behavior, especially when dealing with people from our own group, whatever that is ^ language group, age group, gender, race, class, profession, etc. But this assumption is constantly letting us down ^ as when men complain that women are impossible to understand, or women that men are terminally stupid, or parents that their kids are aliens from another planet, or kids that their parents are insane.

Conversations with foreigners It should be clear that this sort of conversational failure due to cultural difference is exacerbated a hundredfold in conversations with native speakers of other languages.65 A culturally alert and intellectually alive grandmother would probably come to the conclusion, after much fruitful thought, that her grandson was, in fact, aware of most or all of the maxims she has constructed as a foundation for their conversation, and could reasonably be assumed to be deliberately flouting the ones she perceived him as violating. He may talk quite differently with his friends, bring a whole different set of conversational maxims to bear there; but at least when he talks to his grandmother he knows that he must be more careful to adapt his behavior more or less to her cultural expectations. This would not be the conclusion reached by culturally alert and intellectually alive expatriates about the native speakers of the language(s) in the midst of which they live. The only people who live abroad (or work closely with foreigners) for any length of time and go on assuming that everyone shares the same universal cultural values, that everyone believes basically the same things and acts in basically the same way, are idiots ^ or, to put that more circumspectly, people blinded by the uncritical universalism and objectivism that is rampant in society anyway into thinking that they themselves are somehow functionally perfectly aligned with universal objective reality. Look for example at the following dialogues, taken from Storti (1994):66 1.

COL. GARCIA:

Yes, we know that, Colonel Wilson. This battalion has not been doing as much as it could. COL. GARCIA: Yes, yes. COL. WILSON: I’ve told Sergeant Diaz that if we don’t get a few projects started before the end of the year, we’ll be falling behind some of the other units. COL. GARCIA: Yes, some units have fewer projects. (40) COL. WILSON:

2.

MR. JONES:

It looks like we’re going to have to keep the production line running on Saturday. MR. WU: I see.

198

Implicatures (performative uptake) MR. JONES:

Can you come in on Saturday? Yes. I think so. MR. JONES: That’ll be a great help. MR. WU: Yes. Saturday’s a special day, did you know? MR. JONES: How do you mean? MR. WU: It’s my son’s birthday. MR. JONES: How nice. I hope you all enjoy it very much. MR. WU: Thank you. I appreciate your understanding. MR. WU:

(52) 3.

MS. COOPER:

The new tracking procedure hasn’t worked, has it? There were some small problems. MS. COOPER: Whose idea was it anyway? MR. WONG: We need to learn from this lesson. MS. COOPER: Yes. It came from Mr. Tung’s division, didn’t it? MR. WONG: Many people worked on the proposal. MR. WONG:

(55) In each of these dialogues, the speakers on both sides of the cultural divide face baffling behavior that may or may not cause them to go into a ‘‘stall’’ to figure out what could possibly explain the other person’s strangeness. In (1), Col. Garcia is presumably puzzled by Col. Wilson’s insistence on competition between units (we’re all in the same army, aren’t we?), and Col.Wilson is almost certainly stumped by Col. Garcia’s response to his call for increased interunit competition,‘‘Yes, some units have fewer projects.’’ Col. Wilson not only assumes that interunit competition is a good thing; he assumes that it is universally recognized as a good thing, and that Col. Garcia must therefore share his assumption. And he assumes all this unconsciously. It operates in his cultural value system at such a deep level that he is not aware of assuming it. If Col. Garcia were to say to him,‘‘You know, I’m not so sure that interunit competition is such a productive way to motivate the troops,’’ he would be surprised, but at least he would then be able to think through his own unconscious attitudes toward competition. Given his unconscious assumptions, Col. Garcia’s ‘‘Yes, some units have fewer projects’’ is a complete non sequitur. In (2), Mr Wu must be constantly amazed at Mr Jones’s insensitivity to the repeated plain-as-day verbal cues he is sending that he does not want to come in on Saturday: ‘‘I see,’’ ‘‘I think so,’’and then, when both of those cues get ignored, the story about his son’s birthday. Mr Wu says ‘‘I see,’’ signaling clearly that this is not a welcome development for him, and he would rather not hear anything more about it, and then Mr Jones blatantly goes ahead and asks him outright whether he could come in! And then when he replies politely, so as to save Mr Jones’s face, but just barely politely, again signaling his deep reluctance, Mr Jones seems to take this for agreement! When Mr Jones finally understands, saying ‘‘How nice. I hope you all enjoy it very much,’’ then Mr Wu brings out the full force of his gratitude: at last this Westerner has understood that he can’t

Metalocutionary implicature and cross-cultural misunderstandings 199 possibly come in during his son’s birthday, and excused him! Mr Jones, for his part, may be very surprised on Saturday when Mr Wu doesn’t show up; or, if he is slightly more sensitive to conversational implicature, he may realize that with ‘‘Thank you. I appreciate your understanding’’ Mr Wu has taken him to be giving permission to stay home and celebrate the birthday, and wonder how he could possibly have reached that conclusion. (This dialogue revolves around interpretive guesses at illocutionary implicature: implied permission to stay home or implied agreement to come in.) In (3), finally, we have another American^Chinese misunderstanding: Ms Cooper naturally wants to pinpoint the problem so as most efficiently to be able to fix it; Mr Wong naturally wants to protect the face of all involved. From Mr Wong’s Chinese point of view, Ms Cooper is surprisingly and disturbingly aggressive, punitive, relentless in her quest for someone to blame; from Ms Cooper’s American point of view, Mr Wong is exasperatingly evasive and uncooperative, equally relentless in his quest for bland faceless euphemism. Each of them is implicating important cultural values: Ms Cooper, that efficiency is much more important than the feelings of individuals, and that the best way to solve a problem is to trace it back to its source; Mr Wong, that the public humiliation of people who made some mistakes is going to cause more problems than it solves. Neither is reading the other’s implicatures with any kind of accuracy.

The failure of Grice’s original model With these three dialogues, in other words, we seem to reach the almost total breakdown of Grice’s model as he himself envisioned it. (And, I should note, it is much more American of me than, say, British, let alone Chinese, to state this failure so baldly. Like Ms Cooper, I have the quintessential American belief in directness, in the value of getting quickly and surgically to the bottom of a problem.) Grice insisted on a foundation of rational cooperation and universal maxims for the analysis of conversational implicature; in these three dialogues all of the speakers are speaking rationally and cooperatively within the axiology of their own cultural values, but are perceived by their counterparts as irrational and uncooperative. Here are three situations in which conversational implicature is rife, and presumably capable of being worked out; but the implicatures are not read as such, but rather are taken as non sequiturs, puzzlers, deliberately or perhaps unconsciously stupid failures to cooperate. Of course, one could argue that all we need here is better knowledge. If all six of these speakers had more cultural knowledge, deeper insight into the cultural values of the foreigners they’re dealing with, the dialogues would proceed smoothly. All six speakers would recognize that their counterparts were bringing divergent conversational maxims to the table, and adjust accordingly. This is the kind of idealized context that a Gricean constative linguist might want to posit: adequate knowledge for all speakers involved. If the cultural universalism that Grice seems to have assumed for his model doesn’t work,

200

Implicatures (performative uptake)

and if in order to be properly constative the model must rest on some structurally stable foundation, then we must seek some new foundation, such as multicultural awareness. Even if the conversational maxims are not universal, in other words, even if members of different groups will bring different maxims to each conversation, all we need to assume in order for Grice’s model to continue working constatively is that all speakers in any conversation know all of the maxims that they themselves and the other speakers are employing. If the misunderstanding in (1) is based on differing cultural assumptions about the value of competition, then the obvious solution is for both armies to train their officers (say) in the different cultural values brought to conversations by their counterparts from the other country. If the misunderstanding in (2) is caused by Mr Jones’s insensitivity to Mr Wu’s subtle verbal cues, then Mr Jones needs to be trained in the subtleties of Chinese politeness. If the misunderstanding in (3) is caused by differing conceptions of corporate efficiency, then general efficiency will be enhanced by a training program that covers these conceptual divergencies. Then everyone involved will be equipped to read conversational implicature accurately, and communication will be improved. And there is certainly something to this. People who live abroad for many years do mostly learn these things about the foreign culture in which they live, and find their lives greatly eased by their growing knowledge. People from different cultures who work or live together (bilingual communities or workplaces, cross-cultural marriages) usually do develop increased sensitivity to many areas of cultural divergence, and learn to make the necessary intentional and interpretive adjustments to facilitate communication. But, of course, even adequate knowledge is rare in the real world. Most of us never even come close to knowing all we need to know to get by. No matter how long you live in a foreign country, the locals are going to keep surprising you with new cultural differences. No matter how long you are married to a person from another country, s/he is going to keep coming out of left field with new ‘‘bizarre’’ conversational behavior. No matter how thorough the intercultural communication training program you undergo before assuming that management position in another country, there are going to be things in almost every conversation you have with the locals that just flatout don’t make sense. This means that the implicatures in these dialogues will need to be worked out, as if they were conversational implicatures, even though for the speakers themselves they should probably be considered conventional implicatures. This is the interesting problem we face, both as conversational participants and as performative speech-act analysts: to interpreters who do not know the conventions, foreign speakers’conventional implicatures become conversational implicatures. As we saw in the ‘‘moon don’t shine’’ example in Chapter 12, an implicature can be conventional in the culture at large and still seem conversational (novel, needful of being worked out) to an interpreter who doesn’t know the phrase. The consequence of this seeming in cross-cultural conversations is that the interpreter is not given much help. According to Grice, conversational

Metalocutionary implicature and cross-cultural misunderstandings

201

implicators know they are implicating something novel and signal to their listener not only that they are doing so but how it is to be taken; conventional implicators implicate things all unawares, drawing on dead metaphors, fossilized implications, implicatures that were once dynamic but have long since become immediatized through long repetition. Conventional implicators do not even notice they are implicating things ^ that they are leaving anything unsaid, anything for interpreters to work out. When the speaker and the listener are both from the same culture, this is not usually an insurmountable problem: both have typically internalized more or less the same immediatized implications; neither has to work anything out. In a cross-cultural conversation, it becomes a significant problem. Conventional implicators’ lack of awareness of their own implicature means not only that they see no need to give the interpreter any help; they unconsciously assume that the listener will understand, and are therefore totally unprepared for misunderstanding. Furthermore, in the three dialogues above (and indeed in most problematic cross-cultural conversations) each speaker is doing this: each conversational participant is not only failing as listener to understand the other speaker’s implicatures, but failing as speaker to recognize that s/he is implicating things, and failing therefore also to signal these implicatures to the other participant. This not only doubles the problem; it multiplies it exponentially. But of course, even as problems proliferate, things are not all that bad. There are four factors that prevent conversation from unraveling into maximum mutual incomprehension: each conversational participant’s (1) knowledge of the other participant’s culture, (2) knowledge of the other participant’s personal conversational habits, (3) knowledge of and sensitivity to human behavior and motivation in general, and (4) determination to make sense of what the other participant is saying. In the absence of one or more of these, the others have to compensate:67 (2/3/4) If you don’t know the other person’s culture very well, but have worked with the person long enough to recognize some recurring patterns in his or her conversational and other behavior, then a general awareness of how human beings behave and feel (a sensitivity to pride, face, and hurt feelings, say) and a determination to work through all problems may carry you through. (3/4) If you and the other person are strangers to each other and each other’s culture, but you both have a fairly good sensitivity to feelings and a determination or strong pragmatic need to figure out what is being said and done with words, there will almost certainly be numerous unresolved problems (each person’s sense of what may or may not damage the other’s face or hurt his or her pride may be way off due to cultural differences), and communication may take longer than in (1/2/3/4 or 2/3/4), but it should be possible to move toward some sort of minimally acceptable understanding. (4) If, in addition to being strangers to each other and each other’s culture, you are both insensitive monsters, who don’t know and don’t care about pride or face or hurt feelings, but have a strong enough pragmatic need to figure out

202

Implicatures (performative uptake)

what is being said (you have been told outright by your boss that you either learn to work with this person or lose your job), communication may be exceedingly difficult, but it will not be impossible. (^) Without any kind of knowledge of or sensitivity to the other person cultural values and/or emotional needs, or desire or pressing pragmatic necessity to overcome these lacks, communication may still work, but its success will be completely hit-or-miss. So the question is, how do we work out conversational implicature when it is not designed as such by the speaker? What analytical processes do we (or might we) undertake in order to pick our way around cross-cultural conversational impasses?

Cognitive dissonance We might make a tentative beginning at the same place Grice begins: with the experience of cognitive dissonance. We expect to get a certain kind of response and get another, one that thwarts our expectations in some way. We expect to be told about a job candidate’s ability to do philosophy, and get told instead about his regular attendance in tutorials. Does this person want to give us the information we need, or not? We warn a colleague that ‘‘if we don’t get a few projects started before the end of the year, we’ll be falling behind some of the other units,’’and, expecting to hear some sort of concern for this ‘‘falling behind’’ in reply, instead hear ‘‘Yes, some units have fewer projects,’’ which sounds vaguely complacent. Has this person heard our warning, or not? Is he taking it seriously? We ask pointed questions ^ ‘‘Whose idea was it anyway?’’ and ‘‘It came from Mr Tung’s division, didn’t it?’’ ^ and get evasive replies: ‘‘We need to learn from this lesson’’ and ‘‘Many people worked on the proposal.’’ Does this person know the answers to our questions, or not? The cognitive dissonance experienced by the American speakers in the three cross-cultural dialogues I cited above is similar to that experienced by the receptors of conversational implicature in Grice’s examples, which suggests that Grice’s original model might well be a good place to start. Grice’s solution to the problem of uptake when a speaker seemed to be flouting conversational expectations, you’ll recall, was to posit shared conversational maxims that the speaker can be taken be flouting, thus only indirectly cooperating with the recognized purpose of the exchange. The writer of the apparently uncooperative letter of recommendation provides too little information, but did write the letter, so he can be assumed to be cooperating. Therefore, the recipient of the letter surmises, this writer must be trying to tell me something indirectly, trying to imply or insinuate something. The missing information concerns the student’s proficiency at philosophy; by omitting that, the writer must be implying that the student is no good at philosophy. Now, I’ve argued that in these cases of cross-cultural misunderstanding the implicature is conventional and therefore largely unconscious; the utterers of

Metalocutionary implicature and cross-cultural misunderstandings

203

the puzzling remarks should not be taken to be deliberately flouting maxims, but rather to be obeying maxims from a different cultural set. But let us put that consideration aside for the time being and see whether it might not be possible to build a satisfactory way of proceeding from within the interpreter’s cultural set ^ since that is where the interpreter’s response must in any case come from. After all, if the interpreter knows enough about the speaker’s culture to adjust in advance (or even after the fact) to alien conversational maxims, then we don’t face a cultural misunderstanding; then there is no problem to be worked out. And it does seem as if this procedure might, in fact, work with dialogues (1) and (3). Col.Wilson warns Col. Garcia that ‘‘if we don’t get a few projects started before the end of the year, we’ll be falling behind some of the other units,’’ and Col. Garcia replies: ‘‘Yes, some units have fewer projects.’’ Using Grice’s original model, we might imagine Col. Wilson reasoning somewhat as follows: ‘‘Col. Garcia seems to be accepting the fact that our unit has fewer projects than other units. I am warning him about falling behind, and he is reframing that ‘falling behind’as a natural state of affairs ^ perhaps not even as a falling behind, but as a kind of natural distribution. Some have more projects, others have fewer. Maybe he’s implying something? Maybe he has a different philosophy of group activities that he wants to convey to me indirectly? He may not see the uneven distribution of projects as a dynamic matter of ‘getting ahead’ and ‘falling behind,’ but rather as, what, something ordained by God, or fate. Maybe he’s a fatalist? Or maybe this apparent fatalism is a cover for laziness? Either way, how am I ever going to build a fire under him? But has he shown signs of fatalism or laziness in other spheres? Not really. If anything he’s something of an eager beaver. So it must be something else. Could it be that he doesn’t like the idea of units competing? Is that what he’s implying here, that my emphasis on ‘falling behind’ is counterproductive?’’ And so on. Here Col. Wilson is proceeding more or less as Grice’s interpreters did, with the exception that unlike them he has no access to shared conversational maxims or any universal assumption of rationality or cooperation. (It is possible, for example, that Col. Garcia is shirking, simply trying to evade pressure to work harder. This would be uncooperative and, from Col. Wilson’s point of view, almost certainly irrational.) Rather than referring to universals, therefore, Col. Wilson refers his judgments about Col. Garcia’s possible motivations for thwarting his conversational expectations to what he knows of Col. Garcia’s personality from previous encounters: he’s not lazy, so he must be cooperating, and ultimately must be implicating something. Col. Wilson may, in fact, be wrong about this ^ Col. Garcia may not be implicating anything consciously, may not be trying to ‘‘signal’’anything to Col. Wilson indirectly, may, in fact, be reacting out of bafflement at what Col. Wilson is trying to say ^ but the assumption that the reply ‘‘Yes, some units have fewer projects’’ is implicative is certainly more conducive to mutual understanding than the assumption (a kind that is very easy to jump to in cross-cultural encounters) that it is simply a stupid irrational uncooperative non sequitur.

204

Implicatures (performative uptake)

A similar procedure could yield equally effective results in (3): MS COOPER:

The new tracking procedure hasn’t worked, has it? There were some small problems. MS COOPER: Whose idea was it anyway? MR WONG: We need to learn from this lesson. MS COOPER: Yes. It came from Mr. Tung’s division, didn’t it? MR WONG: Many people worked on the proposal. (55) MR WONG:

Here too Ms Cooper could start from the observation that Mr Wong is not giving her what she wants, what she is insistently asking for: ‘‘I say ‘it hasn’t worked, has it,’ a yes or no question, and he says ‘there were some small problems.’ I ask whose idea it was and he says we need to learn from it. The correct answer to my question is a name and he talks generally about learning from mistakes. I ask for confirmation that it was Mr Tung’s division and he spreads the problem out over ‘many people.’ Each of his responses sounds to me like misdirection, as if he is trying to lead me away from the true culprits; but maybe it’s just, what, redirection? Maybe he’s trying to tell me something here but isn’t spelling it out. Maybe he would consider it rude to blurt out his true meaning; he wants me to guess at it, and is giving me hints. So what are the hints? I say ‘it hasn’t worked’ and want him to confirm it, and he talks about ‘some small problems.’ Sounds like euphemism; he’s dragging my rather blunt characterization of the situation back into a euphemistic middle ground. Maybe he’s softening it? Why? My question about whose idea it was is about the past; his reply that we need to learn from it is about the future. Maybe he’s signalling to me that my approach is the wrong way to handle this sort of matter? Don’t make a big deal about a failure; don’t assign blame; downplay the whole thing and look to the future. He would never come right out and tell me that my approach is wrong; he’s trying to imply it, teach me indirectly. Could that be it?’’ Again, as in (1), my imaginary analytically alert Ms Cooper may be ascribing far more implicative awareness to Mr Wong than he actually experiences; he may be reacting ‘‘instinctively,’’ which is to say conventionally, to what he intuitively (ideosomatically) experiences as a witch hunt. He may even be conveying his take on the situation as openly as he can: this may be direct conversational behavior in Chinese society. Ms Cooper’s (by her lights quite generous) willingness to take his utterances as implicative, as deferential indirection, may simply arise out of her ignorance of the conventions that make those utterances signify ‘‘openly’’ in Chinese culture. Still, by giving him the benefit of the doubt, by refusing to take the intellectually lazy person’s refuge in contemptuous dismissal (‘‘these Orientals are so inscrutable, you can never figure them out’’), by entering into a Gricean process of working out his conversational implicature, she moves toward a better understanding of his culture, and thus also, most likely, toward a more effective administrative interface with her Chinese colleagues.

Metalocutionary implicature and cross-cultural misunderstandings

205

Metalocutionary implicature But it is perhaps utopian to expect even fairly obvious misunderstandings like these two to yield to such implicative analysis. Certainly, the inner analytical monologues I’ve assigned to Col. Wilson and Ms Cooper, leading as if inexorably to the correct cultural ‘‘read’’ of the foreigner’s verbal behavior, have a contrived feel to them. Most people working out conversational implicature in a monocultural context do so quickly and easily; people who have repeated difficulties in figuring out what others are implicating conversationally are relatively rare, and their condition is often considered a form of social retardation. (‘‘She’s so literal; she never gets the slightest irony!’’) Working out the cross-cultural ‘‘conversational implicatures’’ in (1) and (3) is considerably more complicated; indeed the ability to carry such a process to a successful end may conceivably even be as rare as the inability to do so in a monocultural context. Most people in a complex and problematic cross-cultural conversation do not even notice, perhaps, that there is a problem to be solved; of those who do notice, perhaps, most do not assume that a solution is even possible. And, in any case, the misunderstandings in (1) and (3) are considerably easier to ‘‘solve’’ than in (2), because the ‘‘non sequitur’’ in (1) and the obvious evasions in (3) pose immediate stumbling blocks to understanding. The apparent uncooperativity of Col. Garcia’s and Mr Wong’s replies is a clear signal to their American interlocutors that something is wrong and must be worked out; and because these remarks seem uncooperative but probably are not, it is at least conceivable that their ‘‘implicit’’ meanings might be worked out along Gricean lines. In (2), however, Mr Wu says nothing whatever that offers the slightest resistance to Mr Jones’s conventional American expectations. It is entirely possible that Mr Jones will come to work Saturday and be surprised not to find Mr Wu there ^ that the two men will walk away from the quoted conversation with utterly divergent assumptions about what has just been agreed to. And, in fact, this utter failure even to notice a problem, let alone solve it, is far more common in cross-cultural communication than the working out of just what is going on, just what is being ‘‘conversationally implicated.’’ I want to argue, therefore, that what is needed to overcome cross-cultural misunderstandings is no simple analytical procedure for ‘‘working out’’ conversational implicature a' la Grice, but a more radical turnabout, a ‘‘stall’’ that brings about a transformative metalocutionary reconceptualization of what one knows and what one doesn’t know, what one can safely assume and what is open to question. It doesn’t matter, perhaps, ultimately, what the trigger is for this sort of metalocutionary rethinking: whether, say, Mr Jones does manage to spot some subtle strain of cognitive dissonance in Mr Wu’s conversational behavior on this one occasion; whether he arrives at work on Saturday and Mr Wu is not there, and when he asks him about it later is thanked profusely for permission not

206

Implicatures (performative uptake)

to come in; whether he only begins to develop a sense of some larger pattern emerging from many such encounters. No matter how subtle the cognitive dissonances arising from cross-cultural misunderstanding, eventually, given enough time and sufficient conversational exposure, they will sink in. What matters, then, is not what provokes the metalocutionary ‘‘stall’’ but how one responds to it: with what degree of attentiveness to and respect for cultural difference, with what degree of acceptance of one’s own relative powerlessness in the face of such difference. I suggest that the key metalocutionary discovery in cross-cultural contexts is not just that the foreigner’s words may have different meanings, but that they have a different history. It’s not just, in other words, that Mr Jones needs to adjust his thinking about Mr Wu’s verbal cues so that he comes to appreciate that the noncommittal caution reflected in ‘‘I see’’ and ‘‘Yes. I think so’’ means something quite different in Chinese culture than in his own, but that those different meanings are the byproduct of an iterative history. Meanings, synchronic meanings in a Saussurean purview, may seem arbitrary; meanings as the most recent products of an ongoing iterative process are highly motivated. It is easy, dealing with nonnative speakers of your own language, to slip into a suppressed-annoyance mode that says (sub voce, of course): ‘‘Well, that’s a strange way of doing it, but if that really is how they think about it, I guess I have to make some kind of adjustment to it in order to get any work done around here at all.’’ This attitude seems to me to reflect a static conception of the alien usages’ arbitrariness, and thus unmotivated ‘‘weirdness.’’ Your own usages feel deeply motivated because you learned them slowly, iteratively, through thousands upon thousands of repetitions, supported and complexly retonalized by many different members of your speech community; the alien usages feel unmotivated, therefore arbitrary, therefore bizarre, because you encounter them without the history. They are just isolated objects, like bits and pieces of crockery among thousand-year-old ruins. You pick them up and turn them over and over in your hands, wondering what they could possibly have been used for. Take, for example, Mr Wu’s remark ‘‘I see.’’ This can be used in English to express grave reservations, in certain contexts, but would not necessarily be taken to be expressing such reservations, especially by an American accustomed to a more direct form of address. You got a problem? You spill it. End of story. We can imagine Mr Jones getting his nudge into metalocutionary reprocessing from a fellowAmerican who has been living in China for years: ‘‘He said he was coming to work Saturday, and then he didn’t show. He’s normally so reliable! When I confronted him about it later, he just acted confused. I don’t know what to think.’’ ‘‘So what did he say when you asked him to come in?’’ ‘‘He said fine, sure, whatever. I’ll be there.’’ ‘‘He said that, in so many words?’’ ‘‘Not in so many words, no. But he clearly implied it.’’

Metalocutionary implicature and cross-cultural misunderstandings

207

‘‘You think he implied it.What did he actually say?’’ ‘‘Well, I told him we were going to have to keep the production line running Saturday. Since that’s his department, that meant he would have to come in. And he said okay.’’ ‘‘That was the word he used,‘okay’?’’ ‘‘No, but it was something like that.’’ ‘‘Think.What did he say?’’ ‘‘Uh . . . I think he said ‘I see.’ ’’ ‘‘Oh!’’ ‘‘What do you mean,‘Oh!’?’’ ‘‘That’s not the same thing as ‘okay’!’’ ‘‘What is it, then?’’ ‘‘It’s basically the same as you saying ‘Oh shit.’ ’’ ‘‘Really?’’ ‘‘Really.Very strong statement.’’ ‘‘You call that strong!’’ ‘‘He would.’’ ‘‘How could he?’’ ‘‘What do you mean, how could he? He does, that’s all.’’ ‘‘But it’s English. We don’t use ‘I see’ that way in English. I mean, if we were speaking Chinese . . . ’’ ‘‘But his culture is steeped in caution. Nobody wants to rush in and blurt something out, because someone might lose face. He’s not going to unlearn thousands of years of Chinese caution just by learning a foreign language.’’ ‘‘Well, he’d better start trying to unlearn it, because we just don’t do things that way in English.’’ ‘‘We don’t?’’ ‘‘No.’’ ‘‘What about this. You’re at your mother-in-law’s for dinner, and she’s cooked up some new thing she found in a magazine, and it’s awful. She’s obviously very proud of it, and is looking for praise.What do you say?’’ ‘‘I don’t know. I try not to say anything.’’ ‘‘But she looks you square in the eye and asks you how you like it.‘Greg, do you like it? It’s good, isn’t it?’ What do you say?’’ ‘‘I say, uh ^ well, I might say, for example,‘It’s interesting.’ ’’ ‘‘Exactly! See?’’ ‘‘See what?’’ ‘‘You replied cautiously. That’s precisely the same kind of caution Mr Wu was exercising.’’ ‘‘But that’s totally different!’’ ‘‘How was it different?’’ ‘‘That was my mother-in-law. She’s a very insecure person. She can’t handle directness. This is me. What am I going to do, collapse in tears if he tells me straight out that he can’t come in?’’

208

Implicatures (performative uptake)

‘‘So he’s cautious in more contexts than you are. That doesn’t change the fact that it’s the same kind of caution. It’s not that the English language won’t allow him to use ‘I see’ as he did; it’s that you’re not used to reading social situations in a Chinese way.’’ ‘‘Hm.’’ ‘‘I bet, in fact, there are situations where you might use ‘I see’ in exactly the same way with your mother-in-law.’’ ‘‘Really? Hmm.’’ ‘‘Her dear friends from college and their children and grandchildren are coming to stay with them for two weeks, and they only have room for the friends, not the children and grandchildren. She’d like them to stay with you, because you have such a big house and all. And the grandchildren are so tiring for people of their age; you’re younger, better able to put up with all that. But she doesn’t want to ask you outright; she wants you to offer. So she explains and explains, hoping you’ll get the drift and say ‘Well why don’t we put them up at our place? We’ve got plenty of room.’ You get the drift but don’t want them.What do you say?’’ ‘‘‘I see.’ I get it.’’ ‘‘Right.’’ ‘‘But you know, there’s still something I don’t get. If I said ‘I see’ to my mother-in-law in that context, I would back it up with body language. You know, a certain tightness or something that would signal to her that I wasn’t particularly pleased with her pressure. I wouldn’t just evade her pressure; I’d put counterpressure on her. We’d have this whole unstated thing going on, through indirection and body language. But Mr Wu gave me none of that. No body language at all. His ‘I see’ seemed, well, interested. Politely, calmly interested. As if encouraging me to go on. How was I supposed to guess from that that he was throwing on the brakes?’’ ‘‘Well, he would certainly never be so crass or offensive as to bracket his ‘I see’ with a major body bristle. He would never be obvious about it, because that would cause you to lose face.’’ ‘‘Then how am I supposed to figure this stuff out?’’ ‘‘The Chinese figure it out because they’ve had extensive training in it their entire lives. Just as you’ve had training in similar complex forms of conversational indirection or even misdirection all your life, so that you can send just the right kind of polite-but-unyielding signal to your mother-inlaw vis-a'-vis the friends’ grandchildren. But the conventions are different, backed up by massively different histories. The Chinese are always going to want to hone the cognitive dissonance between their polite exterior and their negative cues to the narrowest possible razor’s edge, because the more clearly they signal their actual reluctance, the more face is lost. But that’s a lot of work, too much really, so certain politely negative cues have become conventionalized, and everybody recognizes them. Same thing in America, in fact. But the system there works differently. You have to retrain all your social instincts here.’’

Metalocutionary implicature and cross-cultural misunderstandings

209

And so on. With this kind of help (or without), Mr Jones begins to build bridges between the foreign usage and his own ^ and they are specifically iterative bridges, aimed at showing him that the alien usage might well have a similar iterative trajectory to usages he is already familiar with. This is not to say that the best way to overcome cultural difference is to assimilate foreign usages to your own; just that finding some kind of parallel in your own usage helps give you a sense of the iterative history that must undergird the foreign one. You can feel the iteratively accreted ‘‘rightness’’ of your own usage; starting there, you can begin to construct a somatic model of the foreign one as well, a‘‘felt’’ projection into the alien speech act. Let us imagine a whole class of cautious, noncommittal evasions or euphemisms in English ^ ‘‘How do you like the rice?’’ ‘‘It’s . . . interesting.’’ ‘‘It’s different.’’ ‘‘Do you like my dress?’’ ‘‘It’s very . . . lively.’’ ‘‘It’s you.’’ ^ and ask not just how they work, but how they came to work. What iterative process brought them to the particular social efficacy they have now. I’m thinking specifically of the fact that most of these evasions are now instantly recognized as evasions, and tend to evoke responses (from friends or very direct people, anyway) like‘‘You hate it, don’t you?’’ To which you then have to say something awkward like ‘‘No, no! I don’t hate it, not at all, I just ^’’And the other person nods sadly, mouthing hate it, while you protest uneasily and in vain. I suggest that the iterative trajectory of such phrases in fact rehearses the major theoretical stages of my discussion of Grice here in Part III: 1. strategic concealment 2. conversational invocature 3. conventional implicature In other words, the first time such an evasion is used, it is designed to maintain a high degree of politeness or friendliness in a context that seems to require such ‘‘positivity,’’ and thus to conceal whatever negativity the speaker may actually feel. In the Peircean terms I developed in Chapter 11, it is a final intendant designed to block an accurate final interpretant. You don ’t want your friend to know what you’re actually thinking about the food or her dress, or whatever. The phrase arises in a dynamic context and is intended for strategic concealment in that context. However, if the phrase is clever enough, it may catch on. Another member of your group, hearing you say it, or hearing that you said it, may like it so much that he or she may use it again in another context where you are present, as a kind of in-joke ^ or what I called in Chapter 13 ‘‘conversational invocature.’’ What is thus invoked is a complex memory of the whole earlier context, including your awkwardness at being put on the spot and your cleverness in extricating yourself from it. If the person who asked you the original ‘‘do you

210

Implicatures (performative uptake)

like it’’question is also present, and the people involved are good enough friends for this sort of banter, the in-joke may also invoke the questioner’s pathetic need for affirmation, for purposes ranging from ridicule to affectionate recognition of and support for human emotional frailty. As the phrase catches on in the wider culture, it shades by imperceptible degrees from conversational invocature to conventional implicature (or perhaps the correct term would be conventional invocature). It is difficult to say at just what point in a potentially infinite sorites series of immediatizations a conversational implicature or invocature becomes conventional; but at some point we would almost certainly want to start saying that the implicature or invocature has become‘‘unconscious’’and thus no longer what Grice calls‘‘conversational.’’ We might imagine, for example, a highly charged social situation in which it is so important to make a good impression that your nervousness about screwing up doesn’t permit you to relax into either easy direct honesty or carefully conscious implicature. (This is the sort of situation that comedies tend to revolve around.) All you have mental energy left over for is conventional implicature/ invocature. Thus when you praise the hostess’s dress as ‘‘lively’’ or the main course as ‘‘interesting,’’ you have no sense at all of concealing more negative opinions; you are swept up by the euphemistic tide and carried along almost unaware of the iterative history that precedes such phrases. And ultimately such conventionalized or immediatized meanings may even entirely displace the earlier meanings, as seems to have happened with the word ‘‘gay,’’ for instance. No man would ever unself-consciously call himself ‘‘gay’’ when he is happy any more, because the new meaning, a noncommittal euphemism for homosexual that over the last few decades has become immediatized as the primary term, and now takes precedence. Similar tendencies may well be under way with a word like ‘‘interesting,’’ which has been so heavily ‘‘contaminated’’ with euphemistic overtones like the ones we’ve been discussing that it is‘‘losing force.’’Calling a book or a movie‘‘interesting’’these days is almost tantamount to damning it with faint praise.‘‘How’d you like X’s new novel?’’ ‘‘Interesting.’’ ‘‘That bad, huh?’’ ‘‘No, not bad. Just . . . interesting.’’ To make an ‘‘interesting’’ book or movie sound truly interesting you have to add intensifiers like ‘‘really interesting,’’ ‘‘incredibly interesting,’’ ‘‘surprisingly interesting,’’ or, better, step up the somatic signals you wrap it in: tonalize it with excitement, lean forward with eyes wide, gesticulate earnestly, etc. But this may be a rear-guard action; in fifty years the primary dictionary definition of ‘‘interesting’’ may be something like‘‘bland, flat, humdrum, of no great or enduring interest.’’ And it is this kind of iterative history, I suggest, that metalocutionary implicature in a cross-cultural conversation must ultimately make available. Not necessarily consciously; not overtly, in a well-articulated series of iterative steps, such as I have presented here. But something like this. Some sense that ‘‘I see’’ or ‘‘I think so’’ as Mr Wu used them come out of such a history ^ that they were once novel, dynamic concealments, and then gradually, through many varied repetitions, became conventionalized parts of the native speaker’s unconscious repertoire.

Metalocutionary implicature and cross-cultural misunderstandings

211

Maxim-formation and reimmediatization I argued in Chapter 12 that the maxims don’t exist as maxims until they are ‘‘activated’’ dynamically in the act of intending or interpreting ^ that it is a specific dynamic need for a maxim as an imaginary back-story for implicature that brings each maxim into being, gives it what active analytical form it has (ever had). In other words, I continued there, the maxims are only ‘‘immediate’’ in the sense of being an unconscious or ideosomatic potential that is activated by the intender or interpreter needing to build or explain a dynamic conflict. They do not exist in consciously articulated form prior to their dynamic activation. This revised ‘‘iterative’’ theory of Grice’s conversational maxims may have seemed counterintuitive in a monolingual and monocultural context: of course (you may have felt) the maxims exist prior to their being formulated by specific utterers and interpreters in specific speech situations; the mere fact that someone hasn’t yet recognized their existence doesn’t mean they don’t exist! (At any rate I felt a little uneasy arguing that classic poststructuralist position there, because I knew that its apparent counterintuitiveness made it a difficult sell.) In a cross-cultural context, however, this theory makes perfect sense. For Mr Jones, the conversational maxims by which Mr Wu might be imagined to be structuring his verbal behavior, and which he will most likely need to learn (perhaps even formulate) in order to understand MrWu, really do not yet exist.68 It seems to Mr Jones as if Mr Wu is either following his (Mr Jones’s) own maxims, but badly, or else has none at all. In order for Mr Wu’s ‘‘alien’’ usages to become comprehensible to an interpreter like Mr Jones, the latter is going to have to construct new maxims for him ^ to reperform his speech acts interpretively as arising out of a different or novel set of immediatized conversational maxims. And here is the important metalocutionary point: this will involve not only constructing the new maxims, dynamically, but immediatizing them ^ not just learning that such maxims might exist, not just memorizing them as alien cultural artifacts, but reimmediatizing our own sense of utterances like ‘‘I see’’ and ‘‘I think so’’ in terms of these new maxims. Slowly, iteratively, through many repetitions, restructuring his immediate interpretants, coming to see those phrases immediately in the restructured terms. In fact, the metalocutionary process might be imagined as moving through the following steps: a. b. c.

divergent immediate elements, (perhaps repeated) dynamic misunderstanding the ‘‘stall,’’ the inability to continue that pushes the user of language past old assumptions into a reassessment dynamic adjustment to ‘‘alien’’ immediate elements, dynamic (but slow and erratic) understanding

212 d.

Implicatures (performative uptake) reimmediatization of own immediate element in terms of the ‘‘alien’’ one, immediate understanding

We can illustrate this process schematically, using the expanded Peircean diagram I developed in Chapter 11, reproduced here as Figure 14.1. In (a), this (clearly idealized) model fails: the divergence is too great to overcome and misunderstanding results, as in Figure 14.2. In Figure 14.2 each interlocutor’s box is grayed to indicate an imaginary conversational engagement, an interchange with an imagined conversational participant ^ a real person, certainly, but with imaginary characteristics, projections of one’s own unconsciously ‘‘local’’ or relativistic cultural assumptions. Thus, for example, Mr Jones thinks Mr Wu is agreeing to come in on Saturday, and Mr Wu thinks Mr Jones is giving him permission to stay home. Each assumes that the other is working with the same ‘‘immediate’’ meanings, but, in fact, it is precisely the immediate meanings that diverge here, due to divergent histories of iterative immediatization. In similar situations, over many decades, perhaps centuries, Chinese and American utterers and interpreters have repeatedly constructed divergent dynamic intendants and interpretants (which is to say, imagined the situations in divergent ways), and built different final interpretants and intendants around those divergences, with the result that the differences have become immediatized, conventionalized. The consequence of this divergent immediatization is that each conversational participant thinks he is dealing with the ordinary dictionary meanings of the words, with a stable semantic structure that can somehow be

Figure 14.1 The intendant/interpretant model.

Figure 14.2 The intendant/interpretant misunderstandings.

model

with

adjustments

for

cultural

Metalocutionary implicature and cross-cultural misunderstandings

213

absolutely or universally relied upon to convey ‘‘the usual’’meanings ^ whatever feels right at the time ^ but the semantic structures on which he relies are only relatively stable in his culture, not universally. They are relatively stable because they have been reiterated as stable, repeatedly reperformed as semantic structures, immediatized, conventionalized, by actual people in a specific culture in tens or hundreds of thousands of speech situations. The ‘‘stall’’ in (b) is the key to the metalocutionary process, the withoutwhich-not, but I haven’t figured out an effective or instructive way to diagram it, so I’ll skip it schematically. The dynamic shift achieved in (c) is reflected in Figure 14.3. Here the divergence of immediate elements is recognized dynamically, in once-off situations: Mr Wu, for example, perceiving that Mr Jones just doesn’t get it, breaks out of his own cultural assumptions and raises the problem directly: his son’s birthday is so important in Chinese culture that it is going to be very difficult for him to come in on Saturday. Col. Garcia tells Col. Wilson his reluctance about interunit competition. Mr Wong tells Ms Cooper that her American assumptions about the importance of tracing the problem to its source publicly are not going to work in China. The ‘‘working-out’’ of the puzzling utterances as if they were conversational implicatures that I discussed above ^ Col. Wilson figuring out that Col. Garcia is implicating a dislike for or philosophical disapproval of interunit competition, Ms Cooper figuring out that Mr Wong is implicating a better way of handling screw-ups in the company ^ would fit in here as well. The strategic metalocutionary reiteration or reimmediatization of these semantic or syntactic or pragmatic elements in (d), finally ^ what I would argue is the true process of foreign language learning and cultural adaptation ^ would look something like Figure 14.4. This figure can be read in two ways: 1. The Utterer and the Interpreter are the same person, say Mr Jones, learning through numerous iterations to reimmediatize both his immediate intendants when speaking to Chinese people and his immediate interpretants when trying to understand their speech. In this visualization, the series

Figure 14.3 The intendant/interpretant model with adjustment for dynamic recognition of immediate divergence.

214

Implicatures (performative uptake)

Figure 14.4 The intendant/interpretant model with adjustments for the iterative reimmediatization process in a cross-cultural context.

of intentional and interpretive reiterations running up and down the slanted middle line can be thought of as supporting each other, each IIrp putting stepped-up pressure on the intentional side to generate a corresponding IInd, and vice versa. 2. The Utterer and the Interpreter are two di¡erent people, say Mr Jones and MrWu. Thus visualized, the diagram can be read from either direction separately, or from both at once. We can, in other words, imagine either Mr Jones working to reimmediatize Chinese conventions as his own, or Mr Wu working to reimmediatize American conventions as his own, or both at once. The most common phenomenon, of course, is for the foreigner to adapt to the local ways ^ in this case, for Mr Jones to do the reimmediatizing, the retraining of his own immediate interpretants (if we put him at the top, as an Interpreter of Chinese speakers’ words) and/or of his own immediate intendants (if we put him at the bottom, as an Utterer to Chinese listeners). Notice here also, however, that Mr Wu is doing a lot of adapting as well: he is, after all, speaking English, which is presumably not his native language. A fuller analysis of all three of these cross-cultural misunderstandings, in fact, would need to engage the fact that in cross-cultural conversations somebody is always speaking somebody else’s language nonnatively. The end result of Figure 14.3 is, ideally, something very like the original model in Figure 14.1, where the immediate intendants and interpretants line up so neatly and apparently so stably that they do appear to be ‘‘immediate,’’ abstract preexisting givens, solid and reliable structures that are just there. The advantage of the explorations of immediatization we’ve been through in the last few chapters is that now we know how they got there.

Metalocutionary iterability But how can this work in practice? It takes decades at the very least for a phrase to become immediatized in just this way; sometimes centuries. How can a foreigner coming in from the outside ever ‘‘learn’’ this history,

Metalocutionary implicature and cross-cultural misunderstandings

215

rehearse it metalocutionarily, fast enough for the process to be of any practical value to him or her? The fact is, living or working with nonnative speakers of your own language is remarkably like growing up in your own speech community: you are constantly bombarded by slightly different iterations of the same. The trick is to direct your attention to those iterations, register them as iterative repetitions, learn to notice the subtle shadings and alterations that native speakers are constantly reperforming them with. The metalocutionary turnabout that I have spoken of is fundamentally a turning toward these iterations, an opening up to them, an opening up specifically to their ongoingness, the traces in them of having arisen out of a past and being on a trajectory into the future. It entails, in fact, a transformative recognition that we are always learning language ^ that language learning is not something we do once with our native language in early childhood and then maybe once or twice again in school or college with foreign languages, but a constant part of our lives. We are never not being iteratively bombarded with new and old usages, never not dynamically adjusting to and then gradually learning to immediatize the new, or else gradually phasing it out. (In this sense there is no significant difference between first-language learning and second-language acquisition, except insofar as the immediatized conceptual structures of the first language have hardened into a brick wall that will not let new dynamic iterations through.) New usages bamboozle us at first, in our own language at home or at work, in our own language abroad, in foreign languages ^ that is what newness is, an unfamiliarity that poses a problem to be solved, a hurdle to be gotten past. But linguistic novelty is never completely new; it is always complexly grounded in creative reiterations of the old. Ultimately, I suppose, what I am describing is a process by which speakers of a language who have been lulled by its apparent ‘‘sameness’’ or ‘‘stable structures’’ into thinking that they have learned their language and can just coast from now on open themselves up to the vital dynamic performative nature of all language ^ a process, to put it only slightly more tendentiously, by which constative linguists become performative linguists. Constative linguistics may be useful back home, where everything seems cozy and stable; when you wander outside that safe haven into the world of cultural difference, you are going to need performative linguistics, a sense that language is people performing actions with words, not (just) a stable set of structural patterns. I think, in fact, that it is no coincidence that two of the most fertile breeding grounds of all for performative linguistics have been anthropological linguistics and intercultural communication (ICC) studies: when you venture out into the wider world beyond the normative bounds of your own group’s conventions, out where you can’t take anything for granted, the Saussurean notion of stable synchronic structure comes to seem increasingly like a conceptual straitjacket preventing you from adapting flexibly to massively novel iterative histories, and you begin to develop more person-centered, act-centered, situation-centered theories of how language works.

216

Implicatures (performative uptake)

It is also no coincidence that the other ‘‘field’’ or ‘‘subdiscipline’’ of performative linguistics that I’ve been tracking throughout this book is translation studies. For the translator is the cross-cultural communicator par excellence, the professional mediator across the cultural barriers that thwart the six speakers in dialogues (1)^(3), pp. 197^8. The problems that to new cross-cultural communicators seem so insurmountable, and which ICC scholars have put at the center of their discipline ^ and I have put at the center of this chapter ^ are the translator’s bread and butter.69 Not just understanding foreign usages; not just building bridges between foreign and domestic usages. Rather, reiterating or reperforming both the foreign and the domestic, the alien and the familiar, the distant and the local in dynamic ways that bring them closer together ^ each in terms of the other, in fact: reperforming the foreign as domestic and the domestic as foreign, the alien as familiar and the familiar as alien. My performative/iterative take on intercultural communication, in fact, the vision I have been developing for it here in this final chapter, is quintessentially a translator’s view. It is based on the assumption that different languages can always be performatively brought into closer conceptual proximity, can always be reperformed as different in complexly linked ways. Those links do not necessarily preexist the translator’s or other cross-cultural performative linguist’s bridge-building; this is emphatically not an argument for the essential sameness or similarity of all languages. Rather, two languages can always be linked by human beings who experience both iteratively, through close repetitive contact with them.

Conclusion

The constative crisis As I write, at the beginning of the twenty-first century, constative linguistics is in crisis. This is not, perhaps, a universally accepted fact. But evidence of it is, it seems to me, everywhere. From the late 1950s to the early 1970s linguistics was the wunderkind of the humanities and the social sciences. In the 1950s and 1960s Claude Levi-Strauss’s applications of Saussurean structuralism to anthropology spawned a proliferating chain of such applications, in philosophy, political science, psychology, literary theory, historiography. In the 1960s and 70s Chomskyan transformational-generative models were if anything even more popular; they seemed to scholars in a wide variety of fields to have a magical power to explain everything. Today those dreams are all but dead. Constative linguistics is increasingly perceived (even by many linguists! see the critiques of Firth, Halliday, Sampson, Harris) as unable to explain even its own object of study, language, let alone the rest of the world. (It’s telling, also, that when Sampson and Harris first began to lodge their vehement criticisms of constative linguistics, most dismissed them as cranks; now they are received with a shrug: so what else is new?) Enrollments in linguistics programs are dropping; that attrition corresponds roughly to rising enrollments in cognitive science programs, which seek to explain many of the same phenomena constative linguists claimed to explain, but along new disciplinary lines that are increasingly performative. And as constative linguists become more and more peripheralized in the academy, their definitive theoretical and methodological assumptions have begun to crack wide open, their abstract positivism and formalism more and more giving way to large-scale psychosocial and cultural explanatory models that have been flooding into linguistics from systems theory (Halliday and the systemic-functionalist linguists), cognitive psychology (Lakoff and other cognitive linguists), literary theory and cultural studies (Fairclough and other critical discourse analysts), and other erstwhile ‘‘enemy’’disciplines. Let me quote once again the passage from Mona Baker (2000: 22) that Iquoted in the introduction: During the 1980s and 1990s, a new generation of translation scholars with training in linguistics began to carry out more sophisticated descriptive

218

Conclusion studies. These focused more on exploring what actually happens in translation than on exploring what should happen or what can potentially happen. The starting-point for much of this work was the idea that meaning is diffuse, in the sense that it is not located in the word or grammatical category but is signalled by a variety of means which cross the traditional boundaries of word, phrase, clause, sentence, and even text. It is also unstable, in that it is only realized in text, and a ‘‘dictionary’’ meaning may even be negated within a specific textual environment. Moreover, meaning is now understood to be culturally constructed and all language use is seen as mediated (culturally, ideologically, cognitively). Language, both generally and in translation, has come to be seen as intimately linked to the social and cultural context in which it is produced.

In other words, constative linguists are beginning to grope their way toward precisely the performative approaches that they have until now rigorously excluded from their methodological and theoretical purview. This is perhaps still only circumstantial evidence of a paradigm crisis. But look again at Baker’s paragraph: 1. ‘‘The starting-point for much of this work was the idea that meaning is diffuse, in the sense that it is not located in the word or grammatical category but is signalled by a variety of means which cross the traditional boundaries of word, phrase, clause, sentence, and even text.’’ What does it mean, exactly, for meaning to be ‘‘diffuse’’? Baker’s next clause begins ‘‘in the sense that,’’ which seems to promise an explanation of that diffusion, but the explanation itself is so diffuse as to be virtually useless. It’s not located in traditional linguistic objects or categories, but it’s signalled by a variety of means. If it is not located in the old place constative linguists believed they had found it, where is it located? Or should we conclude that it is not located anywhere? And what does it mean for meaning to be ‘‘signalled’’? What variety of means? Are we still talking about formal markers here? The simple fact that the various means cross traditional constative boundaries doesn’t tell us much about what they are; and the fact that constative linguists are thinking about these new means that signal meaning in fairly timid terms (they cross the old boundaries but seem to leave those boundaries more or less in place) that are increasingly vacuous does point fairly strongly to an explanatory crisis of mounting proportions. The old models are under fire; new ideas are flooding in; there is no sense yet of what new models might work, might make sense of the long-excluded data without totally cracking open the old disciplinary frameworks. 2. ‘‘It is also unstable, in that it is only realized in text, and a ‘dictionary’ meaning may even be negated within a specific textual environment.’’ This gives meaning three attributes, one a negative adjective that tells us only that it is not stable (but what is it?), the other two passive constructions (‘‘is realized,’’ ‘‘may be negated’’) without agents, which is to say without any

Conclusion

219

sense of the human beings who might be in a position to perform the ‘‘realization’’ or ‘‘negation’’ of meaning. The depersonalization of language on which constative linguistics is theoretically and methodologically based survives well into its crisis, even into descriptions of the ‘‘new’’ linguistics in which meaning is ‘‘culturally constructed’’ (by whom? as what?), ‘‘mediated (culturally, ideologically, cognitively)’’ (by what mediatory channels? if not the somatic, then which? the intellectual? how does that work?), and ‘‘intimately linked to the social and cultural context in which it is produced’’ (what kind of intimacy are we talking about here? who links it? who produces it?). The constative paradigm was not, however, yet in crisis when J. L. Austin casually tossed out into it his suggestion that language isn’t merely a mechanical vehicle for transmitting information, like a radio: people perform social actions with it, do things with it. Austin must be considered a revolutionary scientist in Thomas Kuhn’s (1970) sense, it seems to me; but when he delivered his lectures in 1955 the time was not yet ripe for a paradigm shift. Nor was it in 1962 when they were published, or in 1969 when his revolutionary ideas were assimilated back to the constative paradigm by the aggressively normal scientist John Searle. The 1960s were, after all, the halcyon days of constative linguistics. H. Paul Grice presented his revolutionary ideas at Harvard in 1967, and they were circulated in mimeograph form for almost a decade before ‘‘Logic and Conversation’’ was first published in 1975: that people use language in ways that puzzle others, forcing them to work out what’s being said; that doing things with words creates novelty, brings change; that gradually that novelty is habitualized, conventionalized through repetition, hardens into structure. The time was not yet ripe for a performative revolution then either, clearly: Grice’s followers have almost without exception pushed him too back into the constative mainstream, beginning in 1983 with Stephen Levinson’s book Pragmatics. Jacques Derrida’s deconstruction of Austin in his 1971 article ‘‘Signature Event Context’’ might be taken as the first signal attempt to consolidate Austin’s revolutionary ideas into a new linguistic paradigm, based on the performative concept of iterability. But Derrida was not a linguist seeking to revolutionize linguistics; he was a philosopher perceived (by Searle) as ‘‘attacking’’ Austin from the outside (outside of the Anglo-American philosophical tradition, and outside of the constative linguistic tradition into which Searle had assimilated him). Mikhail Bakhtin’s revolutionary anti-constative work from the late 1920s and mid-1930s had been rediscovered in Russia in the early 1960s and was just beginning to be translated into English and French and other Western languages in the early 1970s; he too began to be recruited for a new linguistic paradigm throughout the 1970s and 80s (though Western attention initially focused on Bakhtin’s retheorization of carnival in the Rabelais book, and bypassed what I have called his performative linguistics). All of these new paradigms, I have been arguing,

220

Conclusion

shared a common focus: they wanted to build all of their models on people doing things. Of course it is difficult to break all of an intellectual tradition’s old mental habits at once, especially when those habits have been iterated all the way down into the ideosomatic substrata of our ‘‘common sense’’; and all of these would-be revolutionary scientists retain a powerful lingering loyalty to their constative foundations. Austin and Grice too remained powerfully attached to constative ideas: Austin theorized the performative, drawing attention to the ways in which utterances perform actions, but could not find his way around the unsettling perception that actual performances of performative utterances may render them ‘‘nonserious’’ or ‘‘parasitic.’’ Grice theorized rule-breaking behavior, but could not explain our ability to make sense of that behavior without positing a rule-governed environment for it to be interpreted in ^ rules governing not just how the rules can be broken but what certain kinds of transgression mean. Still, however mixed, however nostalgic for constative stability, Austin’s and Grice’s ideas have remained subversive enough for a constative project that their texts keep insistently (though perhaps only subliminally) pushing us to ask: what if we were to base the entire linguistic enterprise on the fact that people do things with words? What if all language could be explained from this performative perspective? My task in this book, then, has been to rally support from all these divergent and centrifugal theoretical interventions for a ‘‘new’’ (actually only newly formulated) linguistic paradigm; to bring together counterhegemonic strands of ‘‘linguistic’’ thought to offer an overarching performative explanation of speech acts in particular, and also, secondarily, such social interactive byproducts of speech acts as linguistic structures, cognitive categories, and norms and ideologies: . . . . . .

where they come from; how they become (relatively) stable; how they change over time; how they di¡er geographically; how they are transmitted from person to person and thus also from generation to generation; how they connect up with the many interactive dynamics of social intercourse (power and authority, intimacy and distance, mimicry and innovation, etc.).

The basic theoretical answer to all of these questions within a performative explanatory framework is that language, cognition, culture, ideology are all forms or products of iterative saturation, the somatic accretion of double-voiced speech acts. In my working out of this model in Part II, Jacques Derrida’s theory of iteration provides the temporal dynamic, my theory of somatics the neural channel, Mikhail Bakhtin’s theory of

Conclusion

221

double-voicing the dialogical model. A nutshell summary of the benefits gained through each: . .

.

Iterability derives linguistic, cognitive, cultural, ideological stabilities from repeated social actions; explains both stability and change. Somatics explains human actions in terms of actual human neural function; o¡ers a plausible ‘‘ergonomic’’ explanation of iteration, how ‘‘language’’ or ‘‘cognition’’ or ‘‘personality’’ or ‘‘habitus’’ can retain regularity through iterations, and how it can change through those repetitions; o¡ers a personto-person and generation-to-generation channel of transfer through experientially acquired ‘‘somatic markers.’’ Double-voicing not only revolutionizes our normative and rather problematic individualistic/atomistic approach to speech acts, by grounding analysis in the dialogical socius; it makes possible a step-by-step, dialogue-by-dialogue (also post hoc) analysis of the iterative process that led to the current stabilities and regularities in language and cognition, and their ongoing destabilizations; enables the analysis of additive voicing in translation, making translation an ordinary part of language use.

The application of these theoretical principles to Gricean implicature, then, with the addition of the doubled Peircean schema of dialogical intention and interpretation, yields the analytical models of Part III: illocutionary, perlocutionary, and metalocutionary implicature; the expanded model of linguistic change based on the immediatization of dynamic elements, or, in Gricean terms, the conventionalization of conversational implicature; conversational invocature (long known as ‘‘allusion’’ but here for the first time theorized as an act of conversational implicature); the metalocutionary reimmediatization of ‘‘structure’’ (including conversational maxims) in foreign-language learning and adaptations to foreign cultures. Each of these models builds explanatory coherence out of the performance of social actions.

The next thing? I originally envisioned this book as an expansion of linguistics. As I mentioned in Chapter 1, it has bothered me for a long time that the scholars of language who I find least interesting are called ‘‘linguists’’ and the scholars of language who I find most interesting are called something else; I wanted to enlarge the realm of linguistics so that it would encompass the scholars I liked. Since the exclusion of my favorite language scholars from ‘‘linguistics’’ was based on their use of a radically different theoretical paradigm, this meant setting up an internal agon within linguistics, two rival linguistic camps, the constative and the performative; but even that seemed preferable to the existing situation, where scholars exclusively interested in a fantasy edifice called la langue were entitled to the name ‘‘linguist’’ and scholars interested in the things people

222

Conclusion

do with words were called ‘‘language philosophers’’ and ‘‘literary theorists’’ and whatnot else. Still, as I put the finishing touches on the book I can’t help but wonder whether the internal agon I have imagined might not be simply a transitional stage in the evolution of a single linguistic discipline ^ whether the disciplinary crisis I have described in constative linguistics might not in fact be foreshadowing a massive theoretical and methodological shift to something like what I have called performative linguistics (and would certainly be called other things by other theorists). Certainly the performative models I have borrowed and developed here seem to fill in the gaps quite nicely in the rather vague descriptions offered by Mona Baker of the recent new directions taken by constative linguists: meaning is not so much ‘‘diffuse’’ as iteratively created and recreated, performed and reperformed in every dialogical exchange. Calling it ‘‘diffuse’’ implies a nostalgia for a lost stability that from a performative standpoint never actually existed, was only dreamed of. Nor is meaning so much ‘‘signalled’’ as it is controlled and regulated by utterers drawing on the regulatory pressures exerted by the speech community and internalized by utterers and interpreters alike in the form of somatic markers. Those somatic markers teach utterers to construct their utterances as effectively as possible for interpretation, of course, which means using constructions that interpreters are likely to recognize as meaning more or less what the utterer intends; this is the only sense in which meaning is ‘‘signalled.’’ But human beings are too restless to perform speech acts in strict accordance with collective regulatory impulses; they always inject their own personal voice into whatever they say, whatever they do, and that changes things, and makes interpretation problematic. Indeed it makes intention problematic as well, even for the intender: most of us, most of the time, do not know exactly what we mean when we say things. What Daniel Dennett calls the ‘‘pandemonium’’ of speech act-potentials constantly surfacing within us makes the supreme executive control of intention (a.k.a. ‘‘rationalism’’) impossible. The relative stability of linguistic structures, the bedrock of constative linguistics, remains an undeniable fact, of course. The structures that are simply passively ‘‘inherited’’ by the speakers of a language, not situationally ‘‘performed,’’are legion. Nor is there any reason to believe that the study of such relatively stable structures will or should ever cease. What I am suggesting in envisioning a paradigm shift from the constative to the performative, therefore, is emphatically not that linguists should start exclusively studying language as act and leave off studying language as structure. This would be as dogmatically narrow as the former constative determination to ignore language as act. A much more likely form that a paradigm shift might take would entail significant change not in what linguists study or even how they do study it but rather in how linguists explain what they study. Constative linguists would go on studying relatively stable structures; but, perhaps, they would no longer see them as naturally stable, or as objective things existing in the null context. It would be increasingly clear that whatever stability linguistic structures have is produced,

Conclusion

223

performed by human societies, and that that stability can be usefully illumined through performative analyses of the history of its production. Thus, for example, just as John Searle’s book Speech Acts should be read as an attempt to impose a constative explanatory model on Austin’s nascent performative linguistics, so too can we read George Lakoff ’s 1987 book Women, Fire, and DangerousThings (especially his long case studies on ‘‘over’’ and ‘‘there’’ in Book II) as an attempt to impose a performative explanatory model on the erstwhile constative discipline of syntax. Just as it has been possible to subsume performative models under a constative aegis, so too is it possible and perhaps desirable to subsume constative models under a performative aegis: to reexplain linguistic structure in more dynamic, more psychologically and socially and politically aware, more philosophically credible, and even (this is a big claim) more scientific ways.70 In fact, it could be persuasively argued that the explanatory paradigm shift I’m envisioning is already underway. In the United States, cognitive psychology has already built a thick and fluffy nest deep in the heart of Chomskyan linguistics; M. A. K. Halliday’s (1989) systemic-functionalist alternative to formalist linguistics is increasingly powerful in Australia and Europe; Roy Harris (1990, 1996, 1998, etc.) and his colleagues in England have been working hard to spread integrationalist linguistics; Norman Fairclough (1996, 1997) and other critical discourse analysts are looking hard at the social and political regulation of linguistic structures once perceived as just ‘‘naturally’’ stable. Increasingly linguists are looking at the socially contingent construction of structure, and so moving decisively past the formalist paradigm legislated for the discipline by Saussure, according to whom abstract stable objective structure quite simply was language, the only aspect of language worth studying linguistically. It may also be, however, that I’m reading too much into these tentative signs of change in the air. It may be that constative linguistics has simply adjusted slightly, covered its tracks philosophically, and will remain fundamentally constative ^ that no significant move into what I’ve been calling the performative camp is imminent. It may be that this prediction of a paradigm revolution in linguistics is just an idle wishful fantasy of mine, born out of the clash I felt in college between my love of language and my hostility toward (constative) linguistics. If only, I thought then, and continue to think now, linguistics could be transformed in some significant and beneficial way, so that it dealt with language as what people do with words rather than as an idealized Platonic realm of pure structure! Whatever happens to linguistics as a discipline, however, it seems clear to me that the language theorists that I have here called performative linguists have much to offer students of language as social action. Anyone who is as impatient as I have been with the naive idealized objectivism of constative linguistics will certainly want to reread Austin and Grice more closely, more performatively; and if they don’t already know Jacques Derrida on iterability or Mikhail Bakhtin on double-voicing, they will want to go right out and

224

Conclusion

get those books under their belts. There are many other exciting performative theorists whose work I did not manage to incorporate into the book; the one whose near-total exclusion I regret most is Ludwig Wittgenstein, whose Philosophical Investigations is, after Austin’s How To Do Things With Words, my second-favorite philosophical joke book of all time, and one of the most radical rethinkings of language philosophy that we have. Kenneth Burke should have been here in a large and imposing way. At the last minute, trying to streamline an overlong book sufficiently to satisfy the publisher, I cut a long chapter on George Lakoff and cognitive linguistics. The Nietzschean tradition in performative linguistics, churning through Martin Heidegger and Hans-Georg Gadamer into such varied poststructuralist thinkers as Michel Foucault, Jacques Lacan, and (the one Nietzschean performative linguist I did showcase) Jacques Derrida, is far more exciting and productive than I have been able to show here. What I have made here is a hesitant and incomplete but I hope suggestive start. There is much more to do. My project in this book has been to reiterate linguistics ^ and also, I suppose, to reiterate some other interesting ideas about language as linguistics. As Derrida and the other performative linguists I have been reiterating would insist, that is all we ever do: reiterate other people’s words, other people’s ideas. Our new iterations inevitably bring something new to the old words and old ideas; iterability always entails originality, even if only relative originality. And that is precisely the quantum of originality I want to claim for myself here: that I said some old things in some new ways; that I brought other people’s ideas (and some of my own earlier ideas) together in a new framework. It also follows from the theory of iterability, however, that you too, in reading this book, have reiterated linguistics in a new way: interpreting is as much an iteration of language as uttering is, reading and listening as iterative as writing and speaking. This places a certain responsibility on readers: how have you reiterated linguistics in the reading of this book? What actions will proceed from that reiteration? It is my hope, of course, that readers will not do to my notion of performative linguistics what many readers of Austin did to his notion of performative utterances ^ reobjectify it in constative terms ^ but will pursue the performative project at least to the extent of pointing out the places in which I too remain reliant on outdated constativist assumptions. All that, of course, is entirely up to you.

Notes

1. The problem faced by constative linguists of translation, it seems to me, is that their orientation to the object of their study is by definition objectivist, i.e. based on the assumption that what they study is a stable object, with certain specific stable properties.This severely limits the range of what they can study, and how they can study it. If they are going to study translation, for example, they can’t simply set about studying any text somebody calls a translation; they have to study a text that has certain properties that will justify its presentation as objectively a translation. And traditionally the one property that has always been offered as evidence that a text is a translation is equivalence. To study translation, therefore, it has seemed to constative linguists that they must study texts that demonstrably ‘‘contain’’ the property of equivalence. From this conventional ‘‘scientific’’ point of view, the performative methodology I will be arguing for throughout this book will seem like cheating ^ like moving the pieces in plain view of your opponent. From a performative point of view, the constative methodology seems like tying your hands and feet. Such are the differences divergent methodologies bring. 2. One way in which Catford’s 1965 book was not Firthian, of course, was that he paid so little attention to the Malinowskian ‘‘context of situation’’ that Firth so insistently stressed. In ‘‘Personality and Language in Society’’ (1957: 177^89), for example, Firth draws a distinction that is clearly aligned with that I’m drawing here between constative and performative linguistics: Saussureans, he says, regard linguistic structures as in rebus: ‘‘The structure is existent and is treated as a thing’’ (181). Firth himself follows Malinowsky in treating speech events as the critical‘‘things’’or objects of inquiry: Malinowsky’s context of situation is a bit of the social process which can be considered apart and in which a speech event is central and makes all the difference, such as a drill sergeant’s welcome utterance on the square,‘‘Stand at ^ ease!’’ The context of situation for Malinowsky is an ordered series of events considered as in rebus. My view was, and still is, that ‘‘context of situation’’ is best used as a suitable schematic construct to apply to language events, and that it is a group of related categories at a different level from grammatical categories but rather of the same abstract nature. A context of situation for linguistic work brings into relation the following categories: A. The relevant features of participants: persons, personalities. (i) The verbal action of the participants. (ii) The nonverbal action of the participants. B. The relevant objects. C. The e¡ect of the verbal action. (182) 3. This is not in fact Austin’s example, but it should have been. Austin spoke of saying ‘‘I do’’ in the wedding ceremony, which doesn’t really work the way he wants it to.

226

Notes

In an editorial note to Austin’s first mention of this particular example, in fact (p. 5n), J. O. Urmson writes ‘‘Austin realized that the expression ‘I do’ is not used in the marriage ceremony too late to correct his mistake. We have let it remain in the text as it is philosophically unimportant that it is a mistake.’’ There is of course a problem in referring generally to ‘‘the’’ marriage ceremony, as if there were only one; every religion and every civil tradition has a different one; within every religion every denomination will have a different one; within some Protestant denominations there are great variations from region to region. In most Protestant denominations in the U.S., for example, ‘‘I do’’ is used in the marriage ceremony, so Urmson’s disclaimer seems a bit strange. What is philosophically important about Austin’s mistake in citing ‘‘I do’’ as a performative utterance, however, is that he repeatedly describes the speech act that it performs as ‘‘marrying’’: ‘‘When I say, before the registrar or altar, ‘I do,’ I am not reporting on a marriage: I am indulging in it’’ (6),‘‘To marry is to say a few words’’ (7), ‘‘Here we should say that in saying these words we are doing something ^ namely, marrying, rather than reporting something, namely that we are marrying’’ (13), etc. This is obviously not the case. To say ‘‘I do’’ is not ‘‘to marry’’; it is to promise to take this man or this woman as your lawful wedded spouse if the minister or registrar or other officiant pronounces the true performative utterance that marries you,‘‘I now pronounce you man and wife.’’ To say ‘‘I do’’ is to promise to marry (specifically, in fact, to promise to stay married), not to marry. 4. Cf. Derrida, too, in ‘‘Limited Inc a b c . . . ’’: ‘‘In view of this parasitism, effected in and by a discourse on parasitism, are we not justified in considering the entire chapter, the entire discussion of Austin, as only an exercise in parody designed to cause serious philosophical discourse to skid toward literary play?’’ (84). 5. I take these titles from Bartley’s (1972/1985:144) biography of Wittgenstein.We will also see John Searle working to distance himself from the open-ended performativity of Wittgensteinian language games in Chapter 4. 6. Here is the relevant passage from Lecture III: If somebody issues a performative utterance, and the utterance is classed as a misfire because the procedure invoked is not accepted, it is presumably persons other than the speaker who do not accept it (at least if the speaker is speaking seriously). What would be an example? Consider ‘‘I divorce you,’’ said to a wife by her husband in a Christian country, and both being Christians rather than Mohammedans. In this case it might be said, ‘‘nevertheless he has not (successfully) divorced her: we admit only some other verbal or nonverbal procedure’’; or even possibly ‘‘we (we) do not admit any procedure at all for effecting divorce ^ marriage is indissoluble.’’ (27) The example points in the direction that I am taking this point, toward cultural relativity, the fact that each group decides what will count as a successful performative within its social sphere; Austin doesn’t expand on this relativity in his discussion, but it seems clear to me that he is aware of it. Otherwise why repeat and emphasize ‘‘we’’? ‘‘We do not admit any procedure at all for effecting divorce.’’ For us ‘‘marriage is indissoluble.’’ This would be a small group of fundamentalists, clearly, within the Christian culture that Austin is contrasting with the Islamic culture that might permit a husband to divorce his wife by uttering the performative ‘‘I divorce you’’ three times: thus is the social and specifically group origin of the performative subtly underscored. 7. It is useful in this context to remember V. N. Voloshinov’s (1930/1973: 74^5) concept of the ‘‘alien word,’’ the foreign or archaic discourse that is universally (he says) used by priests to mystify their authority over their lay followers. These examples of

Notes

8.

9. 10.

11.

227

disagreements between communities over what counts as a marital performative should make it clear that the defining characteristic of this ‘‘alien word’’ for each religious or other community is not merely its alienness, but its adherence to one particular form of alienness. It is not enough for a priest to import any old alien discourse to mystify his authority; he must use the alien discourse prescribed by his specific community, handed down to the contemporary members of that community by their ancestors and clung to by them as the only true expression of divine authority. All other ‘‘alien discourses,’’ the authoritative words of the power discourses recognized by other communities, will not seem authoritative at all; they will be pure mumbo-jumbo. When these communities come into conflict, as over the performing of an inter-faith marriage, or, as in the case of my student, when the accepted clerical discourse in one community seems unclerical to an outsider who is an integral party to the proceedings, the universalism cherished by each community is challenged and undermined. Of course, the liturgical use of Latin did little to demystify performative magic among the vernacular-speaking laity, who developed the potent magic incantation hocus pocus from the Latin Mass: Hoc est corpus ‘‘this is the body [of our Lord Jesus Christ].’’ For a history of this performative sacramentalism in the early church, and its roots in the ancient mystery religions that it supplanted, see Robinson 1996:Chapter 1, pp.73^8. Other books by the prolific Harris include: Synonymy and Linguistic Analysis (his 1971 doctoral dissertation, published in 1973), Approaches to Language (1983), Developmental Mechanisms of Language (1985), Linguistic Thought in England 1914^45 (1988) Language, Saussure, and Wittgenstein: How to Play Games With Words (1990), Signs of Writing (1995), The Language Connection: Philosophy and Linguistics (1996), The Origin of Language (1996), and RethinkingWriting (2000). In fact, Harris himself admits this in Introduction to Integrational Linguistics (Harris 1998: 1), his first full-length attempt to spell out just what an integrational linguistics might entail: ‘‘T. H. Huxley once observed that it is ‘the customary fate of new truths to begin as heresies and to end as superstitions.’ Integrationism is still at the heresy stage. Which means that most of its intellectual energies are still directed towards challenging an orthodoxy already established.’’ I’m not so sure that this last causal chain (‘‘Which means that . . . ’’) necessarily follows. Not all heretics are as obsessed with the orthodoxy they challenge as Harris. My own sense, in fact, is that Harris would rather remain a heretic than begin to build the kind of alternative system that could eventually slide toward uncritical belief and superstition: he would rather tear down the old linguistics than to build a new one. This is not to take away from his monumental achievement in the former task; but I think it is important to note that he and his colleagues have not made a great deal of progress with the latter ^ especially important since that is precisely the task I have set myself here. Here is a fuller statement of Harris’s rhetorical strategies in pursuing his integrationist program, from his postscript (Harris 1998:149^50): This Introduction to Integrational Linguistics has introduced its subject mainly by way of comparison and contrast with another approach ^ orthodox (segregational) linguistics ^ which not only focusses on words but specifically on spoken words as the essential building-blocks of language. This contrastive strategy takes advantage of the fact that orthodox linguistics is more likely to be already familiar to the majority of students. But it is a strategy with certain risks attached and it perhaps invites certain questions. Is integrational linguistics, then, an alternative to orthodox linguistics? Or does it merely point to the need for something extra to be added to the orthodox

228

Notes programme? Or, if neither of these is the case, what exactly is the relationship between integrational and orthodox linguistics? Integrational linguistics certainly does not seek to supply alternative or better methods of achieving the same goals as orthodox linguistics; for, in the integrationist view, the main goals of orthodox linguistics, particularly with regard to the decontextualized description of ‘‘languages’’ (their phonology, grammar, etc.), are in any case either misleadingly formulated or else impossible of fulfillment. Nor, by the same token, does integrational linguistics merely seek to fill a lacuna that orthodox linguistics has overlooked. What integrational linguistics attempts to provide is an alternative perspective on language and the whole process of linguistic inquiry. This includes, by definition, the kind of inquiry conducted under the aegis of orthodox linguistics. Consequently, it provides the basis on which orthodox linguistics can be subjected to critical scrutiny.

And that is about it: ‘‘it provides the basis on which orthodox linguistics can be subjected to critical scrutiny.’’ It doesn’t provide a basis for a new linguistics; it just articulates a very persuasive critique of the old one. I would agree that this relentlessly negative attack strategy is fraught with risks; the main question Harris’s strategy arouses in me, however, is not about the either^or or parasite^host relations between segregational and integrational linguistics. It is, rather, what on earth is integrational linguistics? This seems to me a much more devastating problem in this strategy: Harris has been working for two decades to formulate this alternative theory of language, and now has written this entire book-length introduction to it, and all he can say, over and over, is that it’s not this and it’s not that. Yes, ‘‘This contrastive strategy takes advantage of the fact that orthodox linguistics is more likely to be already familiar to the majority of students’’; but at some point it must occur even to Harris that his ‘‘contrastive strategy’’ only serves in the end to perpetuate the disciplinary host to which his attacks are parasitically attached. Perhaps if he had been building a practical methodological alternative to segregational linguistics, rather than this shadowy hypothetical one, the majority of students would now no longer be familiar with orthodox linguistics? Well, of course, an orthodoxy of the magnitude of ‘‘segregational’’ or ‘‘constative’’ linguistics is extraordinarily difficult to topple. This book too undertakes that task, and as a result will also no doubt be received as heretical. But I think the time has come for us heretics to stop hurling stones at the reigning dogmas. The dogmas are already crumbling; they cannot sustain their own weight. It’s time for us to get on with the project of creating a new linguistic discipline. 12. This claim is, in fact, at least open to question. How, first of all, do we even know what opens a meeting? It seems to me that a convincing (performative) argument could be made for subjective audience response as the only sign or signal that a meeting has been opened, that it now is open: you’re sitting there, you hear the chairperson call the meeting to order, and you stop talking to your neighbor and start paying attention. Then, for you, the meeting is open. If this model is accepted, then the interpretation of ‘‘I declare the meeting open’’ into another language might very well be accepted as a performative utterance opening the meeting for monolingual speakers of that (target) language. Of course, things are not that simple. It may also happen that a meeting is called to order when you aren’t paying attention; you suddenly realize that the meeting already is open, and you missed the opening. Something happened, somebody spoke some words, and you weren’t aware of it. It seems likely that the

Notes

229

performative act of calling a meeting to order is far more complicated than Pym is here assuming; that the act takes different forms for the different actors (including the audience) who participate in it. Still, it is hard for me to imagine that conference organizers would go to the trouble of hiring a consecutive interpreter to report on the opening of a meeting; to constate to the target-language speakers in the audience that the meeting has just been opened. Surely the significance of the ritual interpretation of the words that open the meeting is to open it in a succession of languages? The only reason Pym has to assume that such is not the case, I am guessing, is the constative disposition that assumes for events a single stable unchanging essence that is independent of participants’ perceptions of it. From a performative standpoint, no social event is ever independent of participants’ perceptions. 13. Pym does not leave his discussion of performatives in translation here; he offers a tentative formulation of the problem and uses that formulation to move on to a consideration of the performative as a key to modes of translational discourse, specifically what Christiane Nord calls ‘‘documental’’and ‘‘instrumental’’ translation. But this is as far as I am going to be following Pym’s treatment. 14. Cf. Livingston’s (1997: 86) summary: Recent accounts of literary irony have identified two terms and associated these with inter- or intrasubjective positions. D. C. Muecke aligns a ‘‘lower’’ and an ‘‘upper’’ term with an innocent ‘‘victim of irony’’ and an ironist or ‘‘ironic observer’’ (1969) [this would be my first category, two different people, one who gets it, another who doesn’t]. Paul de Man’s irony splits off a ‘‘fictional self ’’ from an ‘‘empirical,’’ ‘‘actual,’’ or ‘‘world-bound self ’’ (1983) [this is roughly parallel to my second category, where the ‘‘fictional self ’’ that doesn’t get it is projected onto another person, imagined as intellectually inferior]. Anne Mellor aligns ‘‘creative’’ and ‘‘de-creative’’ positions with Friedrich von Schlegel’s ‘‘two opposing drives: one [which] seeks order and coherence (to become being), while the other seeks chaos and freedom (to be becoming or to become nonbeing)’’ (1980: 8) [this inner split is roughly my third category]. 15. Note that Searle (1969: 57n) also follows Austin in ‘‘contrasting ‘serious’ utterances with play acting, teaching a language, reciting poems, practicing pronunciation, etc., and . . .‘literal’ with metaphorical, sarcastic, etc.,’’ and excluding the nonserious and nonliteral acts from his purview. 16. In the letter to Derrida that structures Derrida’s ‘‘Afterword’’ to Limited Inc, Gerald Graff asks: At the end of ‘‘Limited Inc . . .,’’ in response to Searle’s invocation of speech act rules, you say that ‘‘there is always a police and a tribunal ready to intervene each time a rule [constitutive or regulative, vertical or not] is invoked in a case involving signatures, events, or contexts’’ (p. 105). Could you elaborate on this statement? It seems to say that any specification of linguistic rules and conventions plays into the hands of the police, or that there is something politically suspect in the very project of attempting to fix the contexts of utterances. (1988:131; italics in original). In his answer to this question Derrida clarifies that while he did mean that theory of any kind is always political, he certainly did not mean that theorizing is politically suspect, or that the implicit existence of police and tribunals in the theoretical act of distinguishing between ‘‘normal’’ and ‘‘deviant’’ cases, or ‘‘serious’’ and ‘‘nonserious’’ speech acts, somehow implies a repressive regime, a police state.

230

Notes

He simply meant that attempting to separate ‘‘marginal, fringe, and partially defective’’ speech acts off from ‘‘ideal’’ ones is in effect laying down a law, and laws imply the possibility of transgression, and that possibility suggests the need for enforcement, and enforcement means police to catch the transgressors and tribunals to try them. In his 1983 review of Jonathan Culler’s book On Deconstruction in the New York Review of Books, Derrida reminds us, Searle wrote that Culler ‘‘also mistakenly supposes that the theory of speech acts seeks some sort of precise dividing line between what is and what is not a promise. But, in fact, it is a consequence of the theory that in real life there can be all sorts of marginal cases within each family of speech acts’’ (quoted in Derrida 1988:124^5). Derrida tries to be nice about this, but it really is an astonishing claim for a respected philosopher to make. Searle does, of course, seek ‘‘some sort of precise dividing line between what is and what is not a promise.’’ And not only does he do so right in this very passage (in setting ‘‘real life’’ and its ‘‘marginal cases’’ off to one side); he has to do so in order to philosophize about promises at all. In that passage from Speech Acts Searle (1969: 54^5) himself argued, at least implicitly, that philosophical analysis is about seeking precise dividing lines between things: ‘‘But this insight into the looseness of our concepts . . . ,’’ he wrote, ‘‘should not lead us into a rejection of the very enterprise of philosophical analysis; rather the conclusion to be drawn is that certain forms of analysis, especially into necessary and sufficient conditions, are likely to involve (in varying degrees) idealization of the concept analyzed.’’ Derrida agrees with this wholeheartedly: yes, it is difficult or impossible to make hard-and-fast distinctions stick, but yes, making those distinctions is still ‘‘the very enterprise of philosophical analysis,’’ and yes, finally, the only thing that makes that enterprise possible is idealization. And Searle’s selfappointed task as a language philosopher, as he made amply clear in that same passage, was to seek some sort of precise dividing line between the ‘‘idealized’’ promises of abstract or formal (‘‘constative’’) theory that he wanted to study and the ‘‘marginal, fringe, and partially defective promises’’ of ‘‘real life’’ that he wanted to exclude from study. It is truly, as he wrote there, a consequence of his constative theory of speech acts that real-life marginal cases get split off from ‘‘the center of the concept of promising’’ ^ which is to say that promising is ideally defined in terms of that center and real-life examples of promising that differ from that ideal definition are a priori excluded as not-really-promising and thus ideally incapable of serving as the grounds for a refutation of the model. Is this not differentiating? And, Derrida asks further, is this differentiation not a kind of legislating? These cases over here will count as ‘‘the center of the concept of promising,’’ those others over there will count as ‘‘marginal, fringe, and partially defective promises,’’ the sort of thing you can find in ‘‘real life’’ but that will by definition have no power to derail Searle’s constative analysis. The possibility that some real-life counterexample might somehow leap the barrier from the margins where it belongs into the center where Searle is about his important constative business implies the need for vigilance: it’s not just, as Searle somewhat ingenuously states, that ‘‘Their existence does not ‘refute’ the analysis’’; it’s that their existence must not be allowed to refute the analysis. Hence the implicit existence of a theoretical or philosophical ‘‘police’’ and ‘‘tribunal’’: the marginal, fringe, and partially defective speech acts of real life that performative linguists concern themselves with must in a constative linguistics be forcibly kept out. But of course since force of any kind, indeed action of any kind, is specifically a performative concern, the constative linguist must work very hard to make it seem as if no action is being taken, no force being wielded: as if, in fact, the marginal real-life counterexamples had no real desire to break into the center of the concept of promising and so did not have to be kept out; as if the two were just sort of calmly different, essentially different, occupying different spaces that had no tendency or inclination to intermix.

Notes

231

Derrida also goes on to theorize the possibility of transgression as the very essence of iterability: the outside chance, which Searle has to guard against while seeming not to be worried about it, that a marginal or fringe or partially defective promise might leap the guardrail from the margins into the center and thus come to serve as a counterexample or refutation of his constative analysis is itself an example of iterability. Since I haven’t presented Derrida’s theory of iterability yet in the text, I won’t go into that here; suffice it to say that Searle’s examples and analyses of promises, ‘‘central’’ and ‘‘marginal’’ alike, are new iterations of past promises. These iterations are not ideally protected against misuse, or against divagations from the straight and narrow path. They can go any way. To make them go where you want them to go, you have to exert force, which as Derrida says, since the exertion of force is no guarantee against transgression either, implies the police and tribunals. Note that Petrey (1990:145) argues that in this 1983 review of the Culler book Searle has in effect abandoned the rigid constativism of Speech Acts 14 years earlier: Perhaps more importantly, Searle’s polemic with Derrida leads him implicitly to abandon one of his major repudiations of Austinian principles, his assumption that speech acts are best apprehended by constructing abstract models rather than by observing concrete societies. As we saw in Chapter Four, Searle’s Speech Acts ignores Austin’s distrust of the simple situations envisaged by logical theory and defines illocution through the neat set of reliable trains provided by ‘‘a simple and idealized case’’ (1969; p. 56). In his review of Culler’s On Deconstruction, Searle sets idealized simplicity aside to validate ‘‘indeterminate’’ phenomena that necessarily include ‘‘marginal, diverging cases’’ (1983b; p. 78). Searle now sees a ‘‘complex network of linguistic and social practices’’ that ‘‘neither require nor admit of rigorous internal boundary lines and simple mechanical methods’’ (79). Marginalization, indeterminacy, and complex social practices are just what Speech Acts sought to eliminate, and Searle’s later position is incomparably closer to Austin. 17. I will be quoting from three different pieces by Derrida on Austin, all collected by Gerald Graff in 1988 under the title Limited Inc: 1. In 1971, Derrida wrote a paper called ‘‘Signature Event Context’’ (Derrida 1988: 1^23) on Austin’s book and the problem of the parasitic speech act for a communication conference held in Montreal in August of that year by the Congre's international des Socie¤te¤s de philosophie de langue francX aise. The paper appeared in the Congress proceedings, and was reprinted the next year in Derrida’s own essay collection Marges de la philosophie. In 1977, the journal Glyph published in its first volume an English translation of it by Samuel Weber and Jeffrey Mehlman, leading John Searle, professor of philosophy at the University of California at Berkeley and, as author of Speech Acts, the world’s leading philosophical proponent of speech act theory, to write a brief response, ‘‘Reiterating the Differences: A Reply to Derrida.’’ 2. This in turn led Derrida to write a crushing 100-page deconstruction of Searle’s reply called ‘‘Limited Inc abc . . . ’’ (Derrida 1988: 29^110). Both Searle’s and Derrida’s papers appeared in the second volume of Glyph, also in 1977. This exchange came to be known as the Searle^Derrida debate; it was then and remains now of significant interest as Derrida’s fullest engagement with the Anglo-American philosophical tradition. 3. About ten years later Gerald Graff edited the debate for Northwestern University Press, including the two essays by Derrida and, since Searle refused to have anything to do with the project, and would not allow Graff to reprint his reply

232

Notes to Derrida, Graff ’s brief (two-and-a-half-page) summary of Searle’s reply. In addition, Graff sent a letter to Derrida with questions about the debate, especially in light of later things Searle had written about the issues broached in the debate, such as his 1983 book Intentionality and a review published in The New YorkTimes Book Review that same year on Jonathan Culler’s book On Deconstruction. Derrida replied to these questions at length, and his response, entitled ‘‘A fterword’’ (Derrida 1988: 111^60), also appeared in the book in Samuel Weber’s English translation.

18. ‘‘Metaconcept’’ because, as Derrida shows at length in Limited Inc, iterability is what makes conceptualization possible: to conceptualize something you have to repeat it, reiterate it, and specifically reiterate it in a new and significantly altered context, so that the ‘‘thing’’ reiterated changes, becomes something new ^ becomes, in fact, a concept, which continues to carry with it the baggage of previous (‘‘preconceptual’’) iterations, so that the‘‘new’’concept is never purely a concept nor ever purely new. 19. If this is confusing, cf. Derrida’s simpler definition of iterability in the ‘‘Afterword’’ to Limited Inc: I hasten to add, for I am not certain that I said it clearly in ‘‘Limited Inc. . .,’’ that the concept of iterability itself, like all the concepts that form or deform themselves in its wake, is an ideal concept, to be sure, but also the concept that marks an essential and ideal limit of all pure idealization, the ideal concept of the limit of all idealization, and not the concept of nonideality (since it is also the concept of the possibility of ideality). Let us not forget that ‘‘iterability’’ does not signify simply, as Searle seems to think, repeatability of the same, but rather alterability of this same idealized in the singularity of the event, both the rule and the event, concept and singularity. There is thus a reapplication (without transparent self-reflection and without pure self-identity) of the principle of iterability to a concept of iterability that is never pure. There is no idealization without (identificatory) iterability; but for the same reason, for reasons of (altering) iterability, there is no idealization that keeps itself pure, safe from all contamination. (119) 20. Derrida pauses in his long deconstruction of Searle to raise a similar form of iterability in connection with the rather blithe conception of ‘‘intention’’ on which Searle bases all meaning:‘‘hearers,’’ Searle wrote in ‘‘Reiterating the Differences,’’ ‘‘are able to understand this infinite number of possible communications simply by recognizing the intentions of the speakers in the performances of the speech acts’’ (1977: 208). It’s that simple! Elsewhere: ‘‘a meaningful utterance is just a standing possibility of the corresponding (intentional) speech act’’ (1977: 202), ‘‘understanding the utterance consists in recognizing the illocutionary intentions of the author and these intentions may be more or less perfectly realized by the words uttered, whether written or spoken’’ (ibid.). In the passage from ‘‘Limited Inc abc. . .’’ I am referring to, Derrida quotes Searle in Speech Acts to the effect that in promising something a speaker ‘‘must be aware of or believe or know, etc.’’ (quoted in Derrida 1988: 75, Derrida’s emphasis) that the thing promised is something that ‘‘the hearer wants done, or considers to be in his interest, or would prefer being done to not being done, etc.’’ (ibid.), and comments: This description ultimately excludes every criterion other than the distinct, determining, and determinable consciousness of the intentions, desires, or needs involved. The rigorous distinction between promise and warning and threat, for instance, is established only by this expedient. Yet what would happen if in promising to be critical I should then provide everything that Sarl’s

Notes

233

Unconscious desires, for reasons which remain to be analyzed, and that it does its best to provoke? Would my ‘‘promise,’’ in such a case, be a promise, a warning or even a threat? Searle might respond that it would constitute a threat to Sarl’s consciousness, and a promise for the unconscious. There would thus be two speech acts in a single utterance. How is this possible? And what if Sarl desired to be threatened? And what if everything that is given to please or in response to a desire, as well as everything that one promises to give, were structurally ambivalent? What if the gift were always poisoned (gift/Gift) in a manner so as to prevent any simple logic (desire/nondesire, for example) from being able to decide, i.e. to distinguish between the two or to determine their meaning univocally? And if, now, I were unable to know (to ‘‘be aware of or believe or know,’’ as Speech Acts puts it) what it is that Sarl as speaker consciously or unconsciously desires, would I not be incapable as speaker either of promising or of threatening to criticize? What is the unity or identity of the speaker? Is he responsible for speech acts dictated by his unconscious? Mine, for instance, might well wish to please Sarl by gratifying the wish to be criticized; or it might want to cause Sarl unhappiness by refusing to be critical; or to please Sarl by being uncritical and cause pain by being critical; or to promise Sarl a threat or to threaten with a promise; also to offer myself as a target for criticism by taking pleasure in saying things that are ‘‘obviously false,’’ inviting Sarl to delight in my weakness or to enjoy the exhibition from above, etc. (Derrida 1977: 75) Need I add at this point that all of these definitional problems (though never the complexities!) disappear completely once we surrender the constative desire to determine what kind of speech act a given utterance is? A performative linguistics is situationally and dialogically pragmatic: a speech act is constructed in contexts by the participants in those contexts (including every later ‘‘interpretive’’or‘‘theoretical’’ context that raises an older context for new study). What is interesting is not what the speech act is in some stable ontological sense (the metaphysical quest that keeps getting constative speech-act theorists into definitional hot water) but how it is constructed, how it is constructivistically performed. 21. In his ‘‘Afterword’’ he quite cheerfully admits that iterability pretty much covers everything: ‘‘A s the last condition of possibility and of impossibility, with all the paradoxes to which this last formula constrains us, iterability retains a value of generality that covers the totality of what one can call experience or the relation to something in general’’ (1988:129). 22. For a slightly less globalized version of basically this same position, see Andre¤ Lefevere’s (1992) concept of translating, editing, criticizing, annotating, etc. as ‘‘rewriting.’’ 23. I should note that Damasio’s first published presentation of this somatic-marker hypothesis was not his (popular) 1994 book, but a (scientific) 1991 article; and that in the book he also cites Shallice and Burgess’s (1993) scientific article ‘‘Supervisory Control of Action and Thought Selection.’’ He also stumbled upon like-minded work after he had finished his book: Ronald DeSousa’s The Rationality of Emotion (1991), and an article by P. N. Johnson-Laird and Keith Oatley (1992) which argues that‘‘basic emotions help manage actions in a rational way’’ (Damasio 1994: 201). Damasio also cites similar work from 1987 that I was not aware of when I wroteTheTranslator’s Turn back in the winter of 1987^88, Lakoff (1987) andJohnson (1987). In addition, my 1991 elaboration of a very similar theory was not exclusively based on introspection, but also on the more speculative/philosophical work of Friedrich Nietzsche (1887/1956, the inculcation of ideology or ‘‘morality’’ through physical pain), William James (1890/1908: 1.245ff., the feeling theory of meaning),

234

24.

25.

26.

27.

28.

29.

Notes

Kenneth Burke (1966, the somatic origins of the negative), and Akhter Ahsen (1984: 52^6, somatic responses to imagery as the effective channel of psychotherapy). Which is not to say, I should add, that linguists are necessarily better at teaching students to write. The kind of systematic sensitivity to linguistic structures that linguists develop is a good first step in perceiving problems in student writing, but those structural problems in a student paper are usually only the tip of the iceberg. This social regulation of language use has been recognized by other linguists, of course. Edward Sapir (1955: 12), for example, writes that words are associated with ‘‘whole groups, delimited classes . . . rather than with single experiences’’. ‘‘My ‘notion’of this house (the word) must be merged with notions that all other individuals who have seen the house have formed of it. The particular experience that we started with has now been widened so as to embrace all possible impressions or images that sentient beings have formed or may form of the house in question’’ (13). Sapir also later talks about writers who attempt to demonstrate ‘‘the origin of linguistic elements within the domain of feeling’’ (39). I assume, in fact, that the negative reactions many skeptical readers had to the somatic theory in The Translator’s Turn came out of something like this trained ignorance of somatic response. We are not taught about such things, how our bodies signal experiential preferences to us, how we store what we’ve learned for quick reference in somatic response, and so when our bodies do that work we don’t notice it happening ^ and talk about it seems to us mildly ludicrous. The ideological prejudice against the body remains very strong in our civilization, and conditions much of our assumptions about how we operate. It is only in the past decade or two, in fact, that neurologists themselves have begun to study such things, and to trace their neural functioning. Reasoning works with images corresponding to knowledge about the past, present, and future (knowledge about the consequences various past choices had for actions in their future), which are activated and brought into focus for processing. Damasio (1994: 196^8) argues that this process is governed by three factors, somatic markers, working memory, and attention; the significance of the somatic markers is specifically that they guide the operation of working memory and attention. Which is not to say that such bodies do not exist; most professional translator organizations and unions have ethics committees that attempt to legislate translator ethics, and the civil codes of many countries also specify norms for translation. But nobody imagines that these legislated codes have the power to change professional norms for translation that are already in place in the working world; they are rather attempts to codify for novices what professional translators know ‘‘intuitively,’’ from repeated structured/structuring interactions with clients, project managers, editors, and other translators. I should also note that I am not taking Toury to task for actually believing that translational norms are literally negotiated; only for using the metaphor without apparently giving much thought to its tenor. For ‘‘mosaic habitus’’ see Simeoni’s remarks in the on-line translation studies colloquium organized at the Autonomous University of Barcelona in early March, 1997 (this particular post from March 5, 1997, and stored at http://cc.uab.es/
iuts0/ 0503.html). Simeoni is here referring to the opening papers by Anthony Pym on the translator’s interculture and me on the translator’s disaggregated agency (from the book project, then underway, which became Robinson 2001): I am currently working on the very same notion of habitus as applies to the translator (conceived as a‘‘single human being’’). Just as Doug Robinson refers to disaggregated agency, I came out recently with the notion of a ‘‘mosaic habitus’’. I found the term useful to express: (i) the particular brand of habitus required of

Notes

235

the human being a.k.a. translator. All social agents have more or less ‘‘mosaic’’ habituses but the translator must cultivate this pluriidentity and modulated submissiveness, or at least make do with it willingly. This feature may provide a bridge for Anthony Pym’s notion of an intercultural space or ‘‘interculture’’ defining the peculiar position of the translator, although it is still not clear to me how an interculture could stand off in a balanced way between regular cultures. The prefix does not quite evoke the astounding complexity of the domain; (ii) the tension felt while translating (not only intellectual but physical); (iii) the faculty of adaptation which is a distinguishing trait of the profession. In this no doubt biased and partial and summary reading, the two constructs ^ disaggregated agency and mosaic habitus ^ strike me as fairly compatible. Perhaps the former is less affirmative than the latter, due to the deprivative morpheme dis. But again, what matters is the way they can (and ought to) be made to function in case studies, to enlighten descriptions of intercultural transfer from the point of view of the agent. 30. I will also be drawing, more peripherally, from some essays Bakhtin wrote in the last decades of his life (he died in 1975 at the age of 79): ‘‘Toward a Reworking of the Dostoevsky Book’’ (1961), and several pieces collected in English in the volume Speech Genres and Other Late Essays, especially ‘‘The Problem of Speech Genres’’ (1952^3), ‘‘The Problem of the Text in Linguistics, Philology, and the Human Sciences’’ (1959^61), and ‘‘Response to a Question from the Novy Mir Editorial Staff ’’ (1970). 31. For an extended discussion of Dennett’s theory of consciousness in connection with Jacques Lacan and translation, see Robinson 2001: Chapter 5. 32. The ‘‘superaddressee’’ (nadadresat in Russian) is a Bakhtinian notion that includes such speech acts as imploring God (seriously or humorously) to do this or that (‘‘Look at the crosses I bear, Lord!’’), invoking an idealized listener or reader (posterity, say) or other imaginary audience (‘‘Can you believe this, ladies and gentlemen?’’). See Bakhtin 1986: 126^7, and Morson and Emerson 1990: 135^6. Holquist (1990: 39) comments: ‘‘As the need to posit a category such as ‘super-addressee’ outside the present moment makes clear, conditions for creating meaning in the present moment are not always the best. A dialogic world is one in which I can never have my own way completely, and therefore I find myself plunged into constant interaction with others ^ and with myself.’’ 33. Heteroglossia is the accepted English translation of the Russian word raznorechie, meaning something like ‘‘different-speechedness,’’ different ways of speaking: ‘‘many ‘languages,’’’ as Morson and Emerson (1990: 140) put it,‘‘reflecting the diversity of social experience, conceptualizations, and values.’’ 34. In Russian dat 0 is ‘‘to give,’’zadat 0 ‘‘to set, put, assign’’ (a task). Bakhtin is playing off these verbs in saying that a unitary language is not dan (‘‘given’’ in the philosophical sense, an existing thing that we can take for granted from the start) but zadan (‘‘given’’as a task is given, assigned, set). Zadanie and zadacha both mean task, the former more toward job or assignment, the latter more toward problem or mission. 35. It bothers me especially that ‘‘I’m like’’and ‘‘she’s like’’are present-tense verbs used only for past-tense reference, to mean ‘‘I said’’ and ‘‘she said.’’ It bothers me so much that I do what I swore I would never do, do precisely what my father did when I was their age, make fun of the usage to my daughters:‘‘So is that really what you’re like?’’ 36. Since this Gricean ‘‘misperformative’’ or miscommunicative perspective will be my focus throughout Part III, I should pause to note here that the Gricean pragmatic tradition is not the only linguistic subdiscipline to have tackled the problem of miscommunication. Another is represented in Coupland, Wiemann, and Giles (1991), who speak of the ‘‘theoretical and empirical exile’’ to which the study of miscommunication has been relegated by the idealizing methodologies of what I am

236

Notes

here calling constative linguistics: ‘‘A key reason why we have traditionally failed to embrace the concept of miscommunication, its diversity and its implications,’’ they write,‘‘can be called the ‘Pollyanna’ perspective that language and communication research has tended to adopt: Researchers have looked for the ‘good’and ignored the ‘bad’; communication problems were treated as aberrant behavior which should be eliminated’’ (1). Their own approach is grounded in the social psychology of language. 37. Like Austin, Grice neglected to publish his lectures, but in the years between the lectures and publication they were circulated in mimeograph form and built for their author an enormous reputation. When the eponymous lecture ‘‘Logic and Conversation’’ was published in Syntax and Semantics in 1975, for example, it almost instantaneously generated an entire linguistic discipline called ‘‘pragmatics.’’ Unlike Austin, Grice saw his lectures through the press, under the title Studies in theWay of Words ^ the source to which my page citations refer ^ but like Austin, Grice died (in 1988) before the printed lectures were released. 38. I’m condensing Grice’s careful philosophical argument considerably, here. He says specifically that there are four ways in which ‘‘a participant in a talk exchange may fail to fulfill a maxim’’: 1. He may quietly and unostentatiously violate a maxim; if so, in some cases he will be liable to mislead. 2. He may opt out from the operation both of the maxim and of the Cooperative Principle; he may say, indicate, or allow it to become plain that he is unwilling to cooperate in the way the maxim requires. He may say, for example, I cannot say more; my lips are sealed. 3. He may be faced by a clash: He may be unable, for example, to fulfill the first maxim of Quantity (Be as informative as is required) without violating the second maxim of Quality (Have adequate evidence for what you say). 4. He may flout a maxim; that is, he may blatantly fail to fulfill it. (30) I would add: 5. He may structure his conversational contribution carelessly, violating a maxim accidentally. 6. He may violate a maxim partially, so that fulfillment and violation overlap. 7. He may have a different sense of the maxims governing conversation than the other participants, because he comes from a different culture (country, social class, age group, gender, educational level, professional group, etc.). 8. Etc. The rest of Part III will be an extended exploration of the et ceteras involved here. 39. Grice specifically begins his lecture on ‘‘Logic and Conversation’’ with a brief disquisition on the ‘‘commonplace of philosophical logic that there are, or appear to be, divergences in meaning’’ between the formal devices of symbolic logic and‘‘what are taken to be their analogues or counterparts in natural language’’ (1988: 22). He notes that philosophers who have wanted to argue that there are no such divergences ‘‘have been subjected to some pretty rough handling’’ (22). Those who agree that there are divergences between symbolic language and natural language, he says, fall into two camps: the formalists and the informalists. For the formalists ‘‘the proper course is to conceive and begin to construct an ideal language, incorporating the formal devices, the sentences of which will be clear, determinate in truth value, and certifiably free from metaphysical implications; the foundations of science will now be philosophically secure, since the statements

Notes

237

of the scientist will be expressible (though not necessarily actually expressed) within this ideal language’’ (23). The fact that natural language deviates from this ideal language is not a strike in favor of natural language; rather, for the formalist, the divergences between the two are precisely what gives the ideal language its great advantage over natural language: the very fact that natural expressions are not as logically pure and stable as their counterparts in the ideal language ‘‘is to be regarded as an imperfection of natural languages; the elements in question are undesirable excrescences’’ (23). To this, Grice notes, the informalist might reply that ‘‘Language serves many important purposes besides those of scientific inquiry; we can know perfectly well what an expression means (and so a fortiori that it is intelligible) without knowing its analysis’’ (23).‘‘So there must be a place for an unsimplified, and so more or less unsystematic, logic of the natural counterparts of these devices; this logic may be aided and guided by the simplified logic of the formal devices but cannot be supplanted by it. Indeed, not only do the two logics differ, but sometimes they come into conflict; rules that hold for a formal device may not hold for its natural counterpart’’ (24). So far this distinction sounds superficially like the one I’ve based this book on, the formalist being a constative linguist, the informalist perhaps a performative linguist. Clearly, at any rate, the preference for formal symbolic logic over natural language is one of the definitive features of constative linguistics, as it is for what Mikhail Bakhtin calls ‘‘theoretism’’ (Chapter 7). What Grice describes as informalism doesn’t really conform very closely to what I’ve been calling performative linguistics, but the parallels are striking enough to make it quite tempting to map the one distinction onto the other. But then, tellingly, Grice upsets the apple cart: he has, he says,‘‘no intention of entering the fray on behalf of either contestant. I wish, rather, to maintain that the common assumption of the contestants that the divergences do in fact exist is (broadly speaking) a common mistake, and that the mistake arises from inadequate attention to the nature and importance of the conditions governing conversation’’ (24). He does not spell out just how that common assumption is mistaken; he goes on immediately to define implicature, calling his remarks an inquiry ‘‘into the general conditions that, in one way or another, apply to conversation as such, irrespective of its subject matter’’ (24). This sounds like a performative project, concern with conversation, with what people do when they converse, which is, in fact, the burden of this lecture; but he sets about that project with the explicit though unargued intention of showing that there are no divergences between symbolic logic and natural language! The only way this makes sense philosophically is if he means that symbolic logic controls conversational uptake: that the CP and its maxims are themselves a symbolic logic governing the understanding of meaning in natural language. If this is in fact what he is trying to say, clearly, he is grounding what I would want to call a fundamentally performative project in a fundamentally constative set of assumptions. Unfortunately for symbolic logic and generally constative projects, it doesn’t work; as I will show, Grice is unable to show that the CP and its maxims actually control conversational uptake. They outline a direction that the interpreter might take, without ever tying his or her interpretive hands.The divergence between symbolic logic and natural language that he wants to deny still exists, clearly. 40. As Louis Althusser remarks in ‘‘Ideology and Ideological State Apparatuses,’’ ‘‘Linguists and those who appeal to linguistics for various purposes often run up against difficulties which arise because they ignore the action of the ideological effects in all discourses ^ including even scientific discourses’’ (245n). It is important to note that linguists, including linguistically oriented translation theorists, are now beginning to learn these lessons from the Nietzschean tradition as well, whether

238

Notes

in the form of political theory (Althusser), philosophy (Derrida), psychoanalysis (Lacan), or sociology (Bourdieu). Hatim and Mason (1990: 86), for example, write: ‘‘Bourdieu (1982) offers a crucial insight in pointing out that illocutionary force is invested in an utterance not by the words themselves or by any particular combination of them but by the system of social relations which influences the production and reception of utterances in particular situations.’’ 41. For a flip side to this articulation of the maxims as dominant masculine episteme, see Jean-Jacques Lecercle’s (1990) discussion of Deleuze and Guattari’s ‘‘agonistic’’ approach to language in A Thousand Plateaus. Calling Grice’s maxims ‘‘irenic,’’ he contrasts Deleuze and Guattari’s approach by outlining an agonistic revision of the maxims: Thus, an agonistic maxim of quantity would run like this: say as much, or as little, as is necessary to reduce your opponent to speechless rage or to force him to abandon the field. Sometimes a flow of invective, and sometimes ahaughty silence, will achieve this end. Or, again, an agonistic maxim of quality would state that you must say not what you believe to be true, but what is most detrimental to your opponent. Sometimes truth, and sometimes falsity, will do the trick. (43) 42. Grice did of course write an entire series of lectures for the 1967 WilliamJames series, and they were published at book length in 1989; but most followers of Grice agree that the meat of his innovation is all in the single lecture/article entitled ‘‘Logic and Conversation.’’ 43. For other applications of speech-act theory to the study of literature, see Ohmann (1971, 1973), Levin (1976), Pratt (1977), de Man (1979), Fish (1980: 197^245, 1982), Culler (1982), and, for an overall introduction to such applications, Petrey (1990), which the author describes as an ‘‘inquiry into what Austin teaches literary critics’’ (145). This narrow focus makes Petrey chary of discussing the Searle-Derrida debate, because in that debate both Searle and Derrida distance themselves from Austin: Petrey is not interested so much in theorizing literature along speech-act lines himself, or in following interesting leads in new directions, as in summarizing how other theorists have applied Austin to literature, period.This is probably also why Altieri (1981) doesn’t appear in Petrey’s book: Altieri uses Grice, not Austin. 44. Though of course Venuti is also wrong to think that for Grice the flouting of the maxims means ‘‘that language is much more than cooperative communication’’: for Grice the flouting of the maxims is able to do the conversational work it does because of cooperation, because the interlocutor assumes that the maxim-flouter is still cooperating. 45. For an application of Sperber and Wilson’s relevance theory to translation, see Gutt (1991). Interestingly, while Sperber and Wilson derived their theoretical framework from Grice ^ the move from four maxims to one of relevance ^ the Gricean origins of this approach have dropped out of the equation entirely for Gutt, who mentions Grice only once in his book, and that in a rhetorically distancing context that seems to portray Grice’s theory of implicature as from outer space: ‘‘and, for example, the Gricean account (Grice 1975) of such uses assumes that it involves the violation of certain norms’’ (Gutt1991: 33). For discussion, see also Fawcett 1997:135^6. 46. Somewhat surprisingly, Malmkjaer too invokes Venuti in her article (1998: 36) in order to argue that translators do not always seek equivalence, and that Grice’s theory is therefore inadequate as a theory of translation: Besides, the exact conveyance of writers’ original intentions, even at the level of literal meaning, is not necessarily what translations are intended by their creators to do, as, for example,Venuti’s (1995) accounts of his own efforts as

Notes

239

a translator clearly show. But this cannot be said to mean that Venuti’s translations are not translations, unless we wish to return to the stalemate which Toury (1980) points to. When this difficulty is seen in conjunction with the argument which I have just put forward to the effect that the Gricean mechanism cannot account for all micro-instances of translation, I think it is clear that it is not a satisfactory theory of translation. It lacks full descriptive and explanatory power. Another conclusion to be drawn from this, of course, is that the problem is not in Grice’s theory but in the (hegemonically) limited imaginations of those who would apply it to translation. Their conservatism will not allow for the possibility that his theory can explain approaches likeVenuti’s. 47. Let me note, in order to head off a constative critique of this made-up example, that ever since Wittgenstein (on ‘‘perspicuous examples’’) the role and status of the example in performative linguistics has been significantly different from its counterpart in constative linguistics. The constative linguist assumes that linguistic forms exist, in some stable objective state; and many constative linguists assume further that to study those forms empirically, one must have detected them in actual language use. These latter scholars don’t really care about the actual use; their concern is for the ontological origins of the forms they theorize. If the form can’t be shown to figure in specific real-world usages, then it is probably imaginary or hypothetical rather than real. This is a critique lodged quite emphatically against Noam Chomsky’s imagined examples like ‘‘Mary is easy to please/Mary is eager to please’’ and ‘‘Colorless green ideas sleep furiously.’’ From a performative point of view, however, these matters look quite different. Performatively speaking, the ‘‘reality’’ behind language is not form, but performance; and performance is viewed differently by the different participants in it, including its observers (including scholars). There is, therefore, no ‘‘objective’’ reality behind language use, and thus no possibility of protecting a claim’s appearance of objectivity by finding it an actual speech situation (or, worse from a performativist’s point of view, a corpus). There are only performative acts, including the interpretive act of the observer. It doesn’t matter whether you video- or audiotape a speech situation, or take copious written notes, or transcribe your tapes into a corpus, you are always going to have to interpret what you find there, and interpretation is always a performative act ^ another performative act, not simply a stable reliable reproduction of the first. Another way of putting this: the performative act of ‘‘construing as real’’ is exactly the same for a ‘‘real’’ example as it is for a ‘‘made-up’’ one; and, in fact, it is often difficult to tell the difference. If I had written up my examples in this book as real-life first-person anecdotes, it would have been virtually impossible for anyone to challenge their veracity; who’s to say they didn’t actually happen? Because the performative participant in a speech situation (including those of story-telling and example-providing, in linguistics books as in other real-life cases) will always construct a sense of the reality behind the story, subjectively there is often no difference at all between a ‘‘real’’ and a ‘‘made-up’’ example ^ and we have no reliable access to the objective facts that might make an empirical determination possible. This means that in performative linguistics the made-up/found-example debate becomes meaningless. All examples are ‘‘found,’’even made-up ones; and even the kind of ‘‘found’’ example that a constative linguist will value is always ‘‘madeup’’ in the sense of having to be constructed-as-real interpretively by an observer. In a performative linguistics, therefore, the crucial issue is not whether an example is made-up or found, but how it is used in context ^ how well it persuades the reader or listener in the speech act that is the scholarly book or article. Each use of a

240

Notes

word or a sentence or a dialogue is a new use, a new iteration, with its own situational dynamics. And while any example can be persuasive or plausible to the interpreter, and thus a useful and valid example ^ even a phrase as decontextualized as ‘‘Flying planes can be dangerous’’ ^ a performative linguistics will gravitate toward highly contextualized examples, with a situation and a mood spelled out, and some sense of the backgrounds and orientations of the participants. The performative linguist may speculate as to how each participant will ‘‘see’’ or ‘‘read’’ the emerging dialogue, may attempt to get inside the participants’ heads, but only one at a time. The temptation to construct a god’s-eye view, giving you everyone’s thoughts at once, will be resisted. A good performative example works like a situation in a short story: it gives you a fully human sense of what’s going on, from someone’s point of view. Whether what you are reading is a short story, or a journalistic essay, or a page out of a history book or memoir, or an example in a linguistics book, becomes one of the issues for paratextual negotiation ^ not an ontological given; not an a priori guarantee of theexample’s real-world legitimacy. 48. See the list of idealized provisos for conversational implicature in ‘‘Logic and Conversation,’’ for instance: The presence of a conversational implicature must be capable of being worked out; for even if it can in fact be intuitively grasped, unless the intuition is replaceable by an argument, the implicature (if present at all) will not count as a conversational implicature; it will be a conventional implicature. To work out that a particular conversational implicature is present, the hearer will rely on the following data: (1) the conventional meaning of the words used [what Peirce would call Firstness], together with the identity of any references that may be involved; (2) the CP and its maxims; (3) the context, linguistic or otherwise, of the utterance; (4) other items of background knowledge; and (5) the fact (or supposed fact) that all relevant items falling under the previous headings are available to both participants and both participants know or assume this to be the case. (50) The tricky one here is, of course, (5). No one can ever be sure that ‘‘all relevant items falling under the previous headings are available to both participants and both participants know . . . this to be the case,’’ which is why Grice hedges and introduces ‘‘(or supposed fact)’’ and ‘‘or assumes.’’ In an earlier piece entitled ‘‘Meaning’’ (1948, 1957) and other lectures from the Harvard series ^ ‘‘Utterer’s Meaning and Intentions,’’ ‘‘Utterer’s Meaning, Sentence-Meaning, and Word-Meaning’’ ^ all now conveniently collected in Studies in the Ways of Words, Grice is even more willing to assume the simultaneous transparency of utterer’s intention and listener’s interpretation to the philosopher. 49. This unGricean terminology is shorthand for ‘‘no implicature’’ (literal) / ‘‘implicature present’’ (implied). The distinction might be rendered ‘‘explicit’’/‘‘implicit’’ as well, were it not for the clash that would create with the Austinian terminology of ‘‘explicit’’and ‘‘implicit’’ performatives. 50. This too draws on private correspondence from Bob Ashley, who writes: Since linguistic failure abounds, there is a good chance that the hearer will not register the exact attitudinal texture the speaker ‘‘thinks’’ she is attempting to implant or to ‘‘incite,’’ if it’s a subjective transgression. But, to me, it remains ‘‘experimental’’ to the core, for the speaker has no guarantee of which transgression type will stall the hearer. And, ironically, the

Notes

241

experiment could be, by sheer luck, perfectly successful, the hearer without hesitation, without stall, catching her transgressive drift immediately ^ meaning that the success is actually a failure. The experiment must not be ‘‘too’’ successful, otherwise it fails to make the hearer ‘‘dwell’’ upon the utterance. And it contravenes the principle of originary linguistic failure. I think it seeks deliberate initial failure of comprehension in the hearer (subjective puzzlement), but also deferred success, a component of which is the initial failure. Recognition of the failure ignites the somatic laughter, part of which must be directed at the hearer’s own other somatic marker for puzzlement. Tension and resolution. And, perforce, she must tinker, but it is certainly not ‘‘ex nihilio’’, for she has an available inventory of conventional approaches, which she has had socially imprinted on her. She, and her interlocutor, if these know one another well, will probably experience more successes or partial success based on interpersonal experience with each other. It might be harder for the husband in a 50-year marriage to make his wife laugh, her knowing the habitus of her husband’s guile. But he may still be a riot at parties. His success as a riot is partly a function of his ability to incite transgressive failure. To incite the stall. His success with his wife, though, becomes maladapted, because the delivery is too frictionless. 51. Another example: ‘‘You know you’re not supposed to climb up on the roof!’’ the parent cries, knowing full well that the child has never climbed up there and so has never been forbidden to do so. If the child ‘‘knew’’ the rule, it was only through the iterability of the often repeated ban on potentially dangerous actions ^ which is to say that it might have been possible (and this no doubt is what the parent had been counting on, or what s/he thinks in retrospect that s/he must have been counting on) to predict the formulation of a rule against climbing on the roof by extracting out of earlier rules against dangerous actions a principle of iterability, a pattern. Still, narratively speaking, breaking the rule creates the rule. 52. I should also note that Peirce’s theory of ‘‘abduction’’ undergirded my Derridean reformulation of Grice in the first run-through of the interpreter’s four-step move toward understanding: noting that inductive and deductive reasoning are operations performed on old knowledge that are incapable of generating new knowledge, Peirce postulates a logical operation called abduction, which makes an intuitive leap from confusing data to a tentative hypothesis. This would be the move in my Derridean model from subjective puzzlement to a subjective explanation. The concept of abduction anticipates Kuhn’s notion of revolutionary science as well. For a very different application of Peirce to the study of language and translation, see my Becoming aTranslator (Robinson 1997a: Chapter 4). 53. For an elaborate formalization of this interpretant triad (a reading that from my point of view makesThirdness a sham) seeJappy (1984). 54. This is roughly, in fact, the charge that Peirce leveled against his former student William James, in dissociating himself from what he considered the intellectual laxness of Jamesian pragmatism. 55. For further discussion of Lindsey and ‘‘popular apocalyptic’’ predictions, see Robinson (1985: 39^41). 56. An abstraction is etymologically something that has been ‘‘dragged away,’’ in this case something that has been dragged out of specific dialogical contexts and given a transcendental stability ^ and in that sense I suppose we could call the programmed somatic responses that ground our understanding of language ‘‘abstract.’’ It might be more accurate to say that they are not abstracted but implanted, since they are conditioned into us at an early age and work to predetermine our responses to specific dialogues; what is then abstracted is the mentalist image or record of this implant in dictionaries, for instance.

242

Notes

57. Constative theology was grounded in the assumption that the immediate intendant and immediate interpretant of the Bible were identical; because constative theologians from Paul to Hal Lindsey could not conceive of a theology that did not posit that identity, or of a society that was not grounded in that theology, it was essential for many centuries that the assumption be protected politically ^ through a careful legislation and penalization of heresy, for example. The radical innovation of ‘‘modern’’ or ‘‘liberal’’ or ‘‘hermeneutical’’ theologians from Schleiermacher on was often perceived by traditional theologians as a willingness to tolerate slippage between immediate elements, but in fact it wasn’t; it was only an attempt to open up some elbow room for the dynamic interpretant. Could twentieth-century ‘‘negative’’ or ‘‘existentialist’’ or ‘‘death-of-God’’ theologians be read as positing a noncoincidence between the divine immediate intendant and the human immediate interpretant? Probably; in any event, it would be an interesting case to make. 58. For a discussion of the performative status of the made-up example, see p. 239, n. 47. 59. For another analysis of miscommunications between generations, see Ochs (1991). 60. Grice himself hints at this possibility, in fact, but doesn’t develop it; his interests lie elsewhere: To speak approximately, since the calculation of the presence of a conversational implicature presupposes an initial knowledge of the conventional force of the expression the utterance of which carries the implicature, a conversational implicatum will be a condition that is not included in the original specification of the expression’s conventional force. Though it may not be impossible for what starts life, so to speak, as a conversational implicature to become conventionalized, to suppose that this is so in a given case would require special justification. So, initially at least, conversational implicata are not part of the meaning of the expressions to the employment of which they attach. (39) In other words, while he passingly recognizes that conversational implicatures may become conventionalized, he is not really interested in the origins of conversational implicatures, or for that matter in the nature of conventional implicatures; he simply wants to distinguish conversational implicatures clearly from everything else they might be taken to be. 61. Notice that my diagram suggests the origins in immediatization of ‘‘linguistic meaning’’ as well ^ the process that I intimated at the end of Chapter 7 on Bakhtin’s double-voicing, that all linguistic structures, including semantic ones, are the products of this sort of iterative conventionalization or immediatization. 62. For a discussion of the performative status of the made-up example, see p. 239, n. 47. 63. Note here a strange, almost Austinian moment in Chapter 3 of Searle’s Speech Acts. Searle is looking for an example of his fifth condition for the successful promise, which is ‘‘It is not obvious to both S and H [hypothetical promisor and promisee] that S will do A [the promised act] in the normal course of events,’’ and suggests ‘‘It is out of order for me to promise to do something that it is obvious to all concerned that I am going to do anyhow. If I do make such a promise, the only way my audience can interpret my utterance is to assume that I believe that it is not obvious that I am going to do the thing promised. A happily married man who promises his wife he will not desert her in the next week is likely to provide more anxiety than comfort’’ (1969: 59). No doubt! Taken out of context, of course, the locutionary act ‘‘I promise I won’t leave you in the next week’’ seems a bit unrealistic, perhaps; why would any imaginary husband say that to his imaginary wife, except as a joke, or as a rather far-fetched example in a book on speech acts?

Notes

243

But just entertaining such a possibility is a very Austinian gesture, rare in Searle’s book; and it has the interesting effect, at least if we stop and take the trouble to work out the consequences of this example, of undermining Searle’s own constative principles. If in fact, as he says just two pages later (1969: 61),‘‘the meaning of a sentence is entirely determined by the meaning of its elements, both lexical and syntactical,’’ just where does the wife’s anxiety come from? ‘‘I promise I won’t leave you in the next week’’ may be a defective promise, in Searle’s terms; but if its meaning is entirely determined by the meaning of its lexical and syntactical elements, there is nothing even in its defectiveness to cause her anxiety. The meaning of the locutionary act’s lexical and syntactical elements is completely harmless: he’s not going to leave! He’s promising not to leave! And if he specifies that he will not leave in a week’s time, so what? What is there to worry about in the specification of a week? If he said he was going to leave in a week, that would be one thing; but he says he won’t leave in a week. No problem! Of course, the fact is that the meaning the wife assigns this locutionary act will be determined by considerably more complex psychosocial considerations than just lexis and syntax. The obvious question she is going to ask herself, as she tries to figure out what he’s trying to do in or by saying this, is why he needed to come right out of the blue to deny this imaginary leaving in the first place. You don’t have to have read Freud on negation to figure out that denying something that no one has accused you of implies some sort of desire to do that very thing, followed then (publicly) by the denial. Suppose the woman has just had a medical scare, a lump in her breast, say, and her doctor has promised her the test results in a week’s time. To this her husband responds with a nervous little giggle and the remark ‘‘Well, I promise I won’t leave you in the next week then, ha ha!’’ The horrified wife, hearing this, would be a fool not to consider the possibility that her husband has been thinking about leaving her ^ not to read into his locutionary act a good deal more than the lexical and syntactical meanings that according to Searle determine the utterance’s entire meaning. Studying an example like this one, it should come as no surprise to us that constative linguistics has never been much use with literary language. The only real surprise is that constative linguists could seriously claim that such examples should be excluded from linguistic analysis, because they fall outside the bounds of ‘‘ordinary’’ language! 64. Bob Ashley protests, in personal correspondence: ‘‘I disagree that all language is ‘always’accretive. All one needs do is pick up an Anglo-Saxon dictionary to find that much of the lexicon has ‘eroded,’ exfoliated into oblivion. Lexical meanings narrow or widen. E.g. Anglo-Saxon ‘meat’ has narrowed to the flesh of animals for us, but for them meat was practically anything you could stuff into your mouth. Similarly, in my own life, my adolescent ‘cool’ jargon has all but vanished.’’ But this ‘‘vanishing’’ of previously current usages is an accretive process too. Not only do new usages accrete and displace the old ones; the process by which certain usages become obsolete involves the accretive adding of neglect and disuse, each building block of neglect being added to the last to push the increasingly obsolete usage further out. 65. The linguistic study of cross-cultural miscommunication is still relatively new; for one fairly rudimentary analysis see Gass and Varonis (1991), who are concerned to build a binary tree diagram of types of ‘‘problematic’’ communication, opposing ‘‘miscommunication’’ to ‘‘nonengagement,’’ under miscommunication opposing ‘‘misunderstanding’’ to ‘‘incomplete understanding,’’ and under incomplete understanding opposing ‘‘nonunderstanding’’ to ‘‘partial understanding.’’ They also offer a hermeneutical flowchart of nonnative-speaker miscommunication. More extensive work on cross-cultural miscommunication has been done in the field of

244

66. 67.

68.

69.

70.

Notes

intercultural communication (ICC), where the focus tends to be less on linguistically mediated social interaction and more on cultural difference. See e.g. Gudykunst, Ting-Toomey, and Nishida (1996). Most of the interesting work that has been done in this area is grounded in the anthropology and sociology of speaking; see e.g. Geertz (1966), Gumperz (1971, 1982), Hofstede (1980), Hymes (1962, 1974), and Katriel (1986). The classic (though from my point of view clearly constative) Gricean work in this field is Brown and Levinson (1987). For a discussion of the performative status of the made-up example, see p. 239, n. 47. In addition, of course, to be more or less accurate this formulation would need to be fractalized for different distributions of these conversational readinesses between or among the participants. Participant A might for example be knowledgeable about B’s culture but humanly insensitive, and might therefore apply his or her knowledge more or less coldly and mechanistically; B might be relatively ignorant about A’s culture but sensitive enough to read verbal/emotional cues effectively. One conversational participant’s cultural knowledge and emotional awareness and sensitivity might well compensate for the lack of both in one or more other participants. One conversational participant’s obtuse insensitivity and indifference to cultural and other differences can derail a conversation entirely. And so on. The criss-crossing permutations of knowledges, skills, sensitivities, and compensations are potentially infinite. It strikes me now that in a constative purview this cross-cultural argument must seem equally counterintuitive: the maxims by which Mr Wu is structuring his speech must (seem to) exist in some imaginary ‘‘objective’’ state, and Mr Wu recognizes them, but Mr Jones does not, yet. So be it. From a performative perspective, in any case, it is important to establish what each conversational participant knows (and thus what seems to exist to him or her) and how that informs or shapes his or her conversational behavior. For a more programmatic discussion of how the disciplines of intercultural communication (ICC) and translation studies (TS) might dovetail, TS picking up where ICC leaves off, ICC filling in the gaps in the cross-cultural learning curve between monolingual communication and TS, see Robinson (1997a: 231^2). This is a big claim, of course, because constative linguistics has craved ‘‘scientificity,’’ and has built its most debilitating formalisms around this desire to be taken as a science like physics. To say that performative linguistics is potentially more scientific than constative linguistics, then, is to beard the lion in its own den. The theoretical shift I am envisioning might in fact usefully be compared to analogous theoretical shifts in another science, that of biology: the shift a century and a half ago from a folk conception of disease as a collection of externally observable symptoms to a biological conception of disease as caused by microbes (a shift from physical causes evident to the naked eye to physical causes invisible to the naked eye), and, later, in our own time, to a psychobiological conception of disease as accelerated by psychosocial states like stress (a shift from physical to nonphysical causes). According to the older folk theory, disease is a symptomatology: the patient’s skin gets hot, s/he is wracked with hot and cold shivers, s/he complains of aches and pains, and the doctor says fever, prescribes plenty of rest and hot liquids to sweat the fever out. And it works! In a day or two the fever breaks, and a day or two after that the patient is feeling strong enough again to be up and about. Given the patent success of the older assumptions (except of course in minor passing cases like the bubonic plague and other epidemics), how the doctors of the time ridiculed the newfangled theories of microbes! ‘‘Oh, sure, we can’t see ’em, but there are tiny invisible bugs that fly around in the air and go inside us and make us sick, right, uh huh, tell me another! How am I going to avoid them if I can’t even see them?

Notes

245

I’m going to walk around in absolute screaming terror the rest of my life! And these little bugs: where do they get all this supernatural strength they’re supposed to have? I weigh 200 pounds and a bug I can’t even see is going to make me sick!’’ But gradually the microbiological theory took hold. It explained the whole field, including epidemics, which the older theory had to ignore. Doctors came to believe in it, and told their patients about it. It was taught in biology classes in school. Eventually it was iterated into the folk theory of disease, ideosomatized as ‘‘just sheer common sense,’’ so that today a child coughing or sneezing without covering his or her mouth makes everybody feel uneasy (and, if it’s a stranger’s child, angry at the stranger for not teaching the child to cover his or her mouth ^ the somatic markers for uneasiness and anger signaling the operation in these individuals of significant collective regulation). But then researchers increasingly started talking about the interactions of psychological and biological states: happy people stay healthier longer; unhappy people get sicker faster. Stress has a deleterious effect on the immune system. And again this seemed counterintuitive to doctors, who had been taught all their lives, first by the folk theory, then, more complexly, in medical school, to think along the lines of the prevailing folk theory according to which disease is caused by microbes, little bugs that invade us from the outside. Now all of a sudden our emotions can make us sick? And the ridicule began again:‘‘Oh, sure, stress can give me cancer, right, my life is constant stress and I’m healthy as a horse, tell me another!’’ And much the same thing, I suggest, is now happening in the study of language. The folk theory of language, iterated by constative linguists whose theories influence mother-tongue and foreign-language teachers, whose teaching in turn influences the next generation of constative linguists, says that language is made up of ‘‘things’’ like nouns and verbs and adjectives and adverbs, and we ‘‘build’’ sentences out of them by lining them up in a nice neat row, according to clearly defined linguistic rules. The ‘‘things’’ exist, and the rules exist. There are a few unclear questions, perhaps, but that’s what linguists are for: to clarify those few remaining areas where we can’t tell exactly where the divisions between linguistic objects go, and what objects go in which category. Instead of looking for causal chains in human neural, cognitive, and social functioning, the obvious place to look for the origins of any human social activity, constative linguists look at symptomatologies: speakers of languages seem to group together words for people, places, and things and call them nouns, so we’ll take ‘‘noun’’ to be an objective category and see how it works. And then one fine day some crank language theorist comes along saying that objectively speaking language doesn’t even exist ^ that it has no objective existence that can be studied scientifically! The very idea! Why, everyone knows that language exists, that its categories have an objective reality that renders them available for scientific study! If language doesn’t exist, how do we learn it in school? How do linguists study it? How do textbooks get written about it? Of course language exists! Performative linguistics seems likely to generate a more scientific approach to language than constative linguistics because it more realistically defines the empirical sources and etiologies of the linguistic facts with which the constativists work ^ at the broadest level, structural stabilities and linguistic change.

Bibliography

Adams, Hazard (1971/1992) Critical Theory Since Plato. Revised edition. New York: Harcourt BraceJovanovich. Ahsen, Akhter (1984) Trojan Horse: Imagery in Psychology, Art, Literature & Politics. NewYork: Brandon House. Althusser, Louis (1971) ‘‘Ideology and Ideological State Apparatuses (Notes Toward an Investigation)’’. In Althusser, Lenin and Philosophy and Other Essays, 127^86. Translated by Ben Brewster. NewYork and London: Monthly Review Press. Altieri, Charles (1981) Act and Quality: A Theory of Literary Meaning and Humanistic Understanding. Amherst, MA: University of Massachusetts Press. Asad, Talal (1986) ‘‘The Concept of Cultural Translation in British Social Anthropology.’’ In James Clifford and George E. Marcus, eds., Writing Culture: The Poetics and Politics of Ethnography,141^64. Berkeley and Los Angeles: University of California Press. Austin, J. L. (1961) Philosophical Papers. Oxford: Oxford University Press. Austin, J. L. (1962/1975) How To Do Things With Words. Second edition. Edited by J. O. Urmson and Marina Sbisa'. Oxford: Oxford University Press. Baker, Mona (1992) In Other Words. London: Routledge. Baker, Mona (2000) ‘‘Linguistic Perspectives on Translation.’’ In Peter France, ed., The Oxford Guide to Literature in English Translation, 20^26. Oxford: Oxford University Press. Bakhtin, Mikhail (1929/1984) Problems of Dostoevsky’s Poetics. Theory and History of Literature, vol. 8. Edited and translated by Caryl Emerson. Minneapolis: University of Minnesota Press. Bakhtin, Mikhail (1934^35/1981) ‘‘Discourse in the Novel.’’ In The Dialogic Imagination: Four Essays, 259^422. Edited by Michael Holquist. Translated by Caryl Emerson and Michael Holquist. Austin: University of Texas Press. Bakhtin, Mikhail (1965/1984) Rabelais and His World. Translated by He¤le'ne Iswolsky. Bloomington: Indiana University Press. Bakhtin, Mikhail (1979/1986) Speech Genres and Other Late Essays. Edited by Caryl Emerson and Michael Holquist. Translated by Vern W. McGee. With an introduction by Michael Holquist. Austin: University of Texas Press. Bartley,WilliamWarren, III (1973/1985) Wittgenstein. LaSalle, IL: Open Court. Bassnett, Susan (1980/1991) Translation Studies. London: Routledge. Bates, Elizabeth and Brian MacWhinney (1982) ‘‘Functionalist Approaches to Grammar.’’ In L. Gleitman and E. Wanner, eds., Language Acquisition: The State of the Art,173^218. Cambridge: Cambridge University Press.

Bibliography

247

Benveniste, Emile (1996) Proble' mes de linguistique ge¤ ne¤ rale. Paris: Gallimard. Bloom, Harold (1975) A Map of Misreading. NewYork: Oxford University Press. Brown, Penelope and Stephen C. Levinson (1987) Politeness: Some Universals in Language Use. Cambridge: Cambridge University Press. Burke, Kenneth (1945/1962) A Grammar of Motives. Berkeley: University of California Press. Burke, Kenneth (1950/1969) A Rhetoric of Motives. Berkeley: University of California Press. Burke, Kenneth (1966) ‘‘A Dramatistic View of the Origin of Language.’’ In Burke, Language as Symbolic Action: Essays on Life, Literature, and Method, 419^79. Berkeley: University of California Press. Bybee, John and Carol Moder (1983) ‘‘Morphological Classes as Natural Categories.’’ Language 59.2: 251^70. Catford, J. C. (1965) A Linguistic Theory of Translation: An Essay in Applied Linguistics. London: Oxford University Press. Chamberlain, Lori (1988/1992) ‘‘Gender and the Metaphorics of Translation.’’ Signs 13 (Spring 1988): 454^72. Reprinted in Lawrence Venuti, ed., Rethinking Translation: Discourse, Subjectivity, Ideology, 57^74. London and NewYork: Routledge. Cheyfitz, Eric (1991) The Poetics of Imperialism:Translation and Colonization from The Tempest toTarzan. NewYork: Oxford University Press. Chomsky, Noam (1957) Syntactic Structures. The Hague: Mouton. Chomsky, Noam (1965) Aspects of theTheory of Syntax. Cambridge, MA: MIT Press. Clyne, M. (1981) ‘‘Culture and Discourse Structure.’’Journal of Pragmatics 5: 61^66. Coupland, Nikolas, John M. Wiemann and Howard Giles, eds. (1991) ‘‘Miscommunication’’ and ProblematicTalk. Newbury Park: Sage Publications. Culler, Jonathan (1982) On Deconstruction: Theory and Criticism After Structuralism. Ithaca: Cornell University Press. Damasio, Antonio R. (1994) Descartes’ Error: Emotion, Reason, and the Human Brain. NewYork: Putnam. Damasio, Antonio R., D. Tranel and H. Damasio (1991) ‘‘Somatic Markers and the Guidance of Behavior: Theory and Preliminary Testing.’’ In H. S. Levin, H. M. Eisenberg, and A. L. Benton, eds., Frontal Lobe Function and Dysfunction, 217^29. NewYork: Oxford University Press. Deleuze, Gilles and Felix Guattari (1986) Kafka: Toward a Minor Literature. Translated by Rita Bensmaia and Dana Polan. Minneapolis: University of Minnesota Press. Deleuze, Gilles and Felix Guattari (1987) AThousand Plateaus. Capitalism and Schizophrenia, vol. 2. Trans. Brian Massumi. Minneapolis: University of Minnesota Press. de Man, Paul (1979) Allegories of Reading: Figural Language in Rousseau, Nietzsche, Rilke and Proust. New Haven: Yale University Press. Dennett, Daniel (1991) Consciousness Explained. Boston: Little, Brown. Denzin, Norman (1982) ‘‘The Paradoxes of Play.’’ In John Loy, ed., The Paradoxes of Play, 13^24.West Point, NY: Leisure Press. Derrida, Jacques (1985) ‘‘Des tours de Babel.’’ In Joseph F. Graham, ed., Difference in Translation, 209^48. Ithaca, NY: Cornell University Press. Derrida, Jacques (1988) Limited Inc. Edited by Gerald Graff. Chicago: University of Chicago Press. D|az-Diocaretz, Miriam (1985) Translating Poetic Discourse: Questions on Feminist Strategies in Adrienne Rich. Amsterdam: John Benjamins. Even-Zohar, Itamar (1979) ‘‘PolysystemTheory.’’Poetics Today 1.1^2: 283^305.

248

Bibliography

Even-Zohar, Itamar (1981) ‘‘Translation Theory Today: A Call For Transfer Theory.’’ PoeticsToday 2.4 (Summer-Autumn):1^7. Even-Zohar, Itamar (1997) ‘‘The Making of Culture Repertoire and the Role of Transfer.’’Target 9.2: 373^381. Fairclough, Norman (1996) Language and Power. Glenview IL: Addison-Wesley. Fairclough, Norman (1997) Critical Discourse Analysis: The Critical Study of Language. Glenview IL: Addison-Wesley. Fawcett, Peter (1997) Translation and Language: Linguistic Theories Explained. Manchester: St. Jerome. Felman, Shoshana (1983) The Literary Speech Act: Don Juan with J. L. Austin, or Seduction in Two Languages. Translated by Catherine Porter. Ithaca: Cornell University Press. Fenton, Sabine (1997) ‘‘The Role of the Interpreter in the Adversarial Courtroom. In Silvana E. Carr, Roda Roberts, Aideen Dufour and Dini Steyn, eds.The Critical Link: Interpreters in the Community, 29^34. Amsterdam and Philadelphia: John Benjamins. Firth, J. R. (1957) Papers in Linguistics, 1934^1951. London: Oxford University Press. Fish, Stanley (1980) IsThere aText in this Class? Cambridge: Harvard University Press. Fish, Stanley (1982) ‘‘With the Compliments of the Author: Reflections on Austin and Derrida.’’ Critical Inquiry 8: 693^721. Foucault, Michel (1961/1982) Madness and Civilization: A History of Insanity in the Age of Reason. Translated by Richard Howard. NewYork: Pantheon. Gadamer, Hans-Georg (1960/1989)Truth and Method. Revised edition. Translated by Joel Weinsheimer and Donald G. Marshall. NewYork: Crossroad. Gass, Susan M. and Evangeline M. Varonis (1991) ‘‘Miscommunication in Nonnative Speaker Discourse.’’ In Coupland,Wiemann, and Giles 1991:121^45. Geertz, Clifford (1966) Person,Time, and Conduct in Bali. New Haven: Yale Southwest Asia Program. Gentzler, Edwin (1993) ContemporaryTranslationTheories. London: Routledge. Godard, Barbara (1989) ‘‘Theorizing Feminist Discourse/Translation.’’ Tessera 6 (Spring): 42^53. Goffman, Erving (1959/1980) The Presentation of Self in Everyday Life. Harmondsworth: Penguin. Grice, H. Paul (1975/1989) ‘‘Logic and Conversation.’’ In Studies in the Way of Words, 22^40. Cambridge, MA: Harvard University Press. Gudykunst, William B., Stella Ting-Toomey and Tsuka Nishida (1996) Communication in Personal Relationships Across Cultures. Thousand Oaks, CA: Sage. Gumperz, JohnJ. (1971) Language in Social Groups. Stanford: Stanford University Press. Gumperz, JohnJ. (1982) Discourse Strategies. NewYork: Cambridge University Press. Gutt, Ernst-August (1991) Translation and Relevance: Cognition and Context. Oxford: Basil Blackwell. Halliday, M. A. K. and Ruqaiya Hasan (1989) Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective. Oxford: Oxford University Press. Hanks, William F. (1999) Intertexts, Writings on Language, Utterance and Context. Denver: Rowman and Littlefield. Harris, Brian (1981) ‘‘Observations on a Cause Ce¤le'bre: Court Interpreting at the Lischka Trial.’’ In Roda P. Roberts ed., L’interpre¤ tation aupre' s des tribunaux. Ottawa: University of Ottawa Press. Harris, Roy (1973) Synonymy and Linguistic Analysis. Oxford: Blackwell. Harris, Roy (1977) On the Possibility of Linguistic Change. Oxford: Oxford University Press/Clarendon Press.

Bibliography

249

Harris, Roy (1980) The Language-Makers. London: Duckworth. Harris, Roy (1981) The Language Myth. London: Duckworth. Harris, Roy, ed.(1983) Approaches to Language. Oxford and NewYork: Pergamon Press. Harris, Roy (1986) The Origin of Writing. London: Duckworth. Harris, Roy (1987) Reading Saussure: A Critical Commentary on the Cours de linguistique generale. London: Duckworth. Harris, Roy (1987) The Language Machine. London: Duckworth. Harris, Roy, ed.(1988) LinguisticThought in England, 1914^1945. London: Duckworth. Harris, Roy (1988) Language, Saussure, and Wittgenstein: How to Play Games with Words. London and NewYork: Routledge. Harris, Roy (1990) The Foundations of Linguistic Theory: Selected Writings of Roy Harris. Edited by Nigel Love. London and NewYork: Routledge. Harris, Roy (1993) Language and Philosophy: The Controversial Interface. Oxford and NewYork: Pergamon. Harris, Roy (1995) Signs of Writing. London and NewYork: Routledge. Harris, Roy (1996) The Language Connection: Philosophy and Linguistics. Bristol: Thoemmes. Harris, Roy, ed.(1996) The Origin of Language. Bristol: Thoemmes. Harris, Roy (1996) Signs, Language, and Communication: Integrational and Segregational Approaches. London and NewYork: Routledge. Harris, Roy (1998) Introduction to Integrational Linguistics. Kidlington and Oxford: Pergamon. Harris, Roy (2000) Rethinking Writing. London: Athlone. Harris, Roy and Charles James N. Bailey, eds. (1985) Developmental Mechanisms of Language. Oxford and NewYork: Pergamon Press. Harris, Roy and Talbot J. Taylor, eds. (1989) Landmarks in Linguistic Thought: The Western Tradition from Socrates to Saussure. London and NewYork: Routledge. Harris, Roy and George Wolf, eds. (1998) Integrational Linguistics: A First Reader. Kidlington and Oxford: Pergamon. Hatim, Basil (1997) Communication Across Cultures: Translation Theory and Contrastive Text Linguistics. Exeter: University of Exeter Press. Hatim, Basil and Ian Mason (1990) Discourse and theTranslator. London: Longman. Hatim, Basil and Ian Mason (1997) TheTranslator as Communicator. London: Routledge. Hermans, Theo (1999) Translation in Systems: Descriptive and System-Oriented Approaches Explained. Manchester, UK: St. Jerome. Hofstede, Geert H. (1980) Culture’s Consequences. Beverly Hills, CA: Sage Publications. Holmes, James (1988) Translated! Papers on Literary Translation and Translation Studies. Amsterdam: Rodopi. Holquist, Michael (1990) Dialogism: Bakhtin and His World. London and New York: Routledge. Holz-Ma«ntta«ri, Justa (1984) Translatorisches Handeln. Helsinki: Finnish Academy of Science. Hymes, Dell (1962) ‘‘The Ethnography of Speaking.’’ In T. Gladwin and W. Sturtevant, eds., Anthropology and Human Behavior, 13^53. Washington, D.C.: Anthropology Society of Washington. Hymes, Dell (1972) ‘‘On Communicative Competence.’’ In J. B. Pride and Janet Holmes, eds., Sociolinguistics: Selected Readings, 269^93. Harmondsworth: Penguin. Hymes, Dell (1974) ‘‘Ways of Speaking.’’ In Richard Bauman and Joel Sherzer, eds., Explorations in the Ethnography of Speaking, 433^51. Cambridge: Cambridge University Press.

250

Bibliography

Jacquemond, Richard (1992) ‘‘Translation and Cultural Hegemony: The Case of FrenchArabicTranslation.’’ In LawranceVenuti, ed., RethinkingTranslation,139^58. James, William (1890/1908) The Principles of Psychology. 2 vols. Reprinted New York: Henry Holt. Jappy, Anthony G. (1984) ‘‘Peirce’s Third Trichotomy and Two Cases of Sign Path Analysis.’’Semiotica 49.1/2:15^16. Johnson, Mark (1987) The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason. Chicago: University of Chicago Press. Kant, Immanuel (1790/1951) Critique of Judgement. Translated by J. H. Bernard. New York: Hafner/Macmillan. Katriel, Tamar (1986) Talking Straight: Dugri Speech in Israel Sabra Culture. Cambridge: Cambridge University Press. Katz, Jerrold J. (1977) Propositional Structure and Illocutionary Force. Hassocks: Harvester. Krings, Hans P. (1986) Was in den Ko« pfen von U«bersetzern vorgeht: Eine empirische Untersuchung zur Struktur des U«bersetzungprozesses an fortgeschrittenen Franzo« sischlernern. Tu«bingen: Narr. Krings, Hans P. (1987) ‘‘The Use of Introspective Data in Translation.’’ In Klaus Faerch and Gabrielle Kasper, eds., Introspection in Second Language Research. Clevedon: Multilingual Matters. Krontiris, Tina (1992) Oppositional Voices: Women as Writers and Translators of Literature in the English Renaissance. London and NewYork: Routledge. Kuhn, Thomas S. (1970) The Structure of Scientific Revolutions. Chicago: University of Chicago Press. Kussmaul, Paul (1991) ‘‘Creativity in the Translation Process: Empirical Approaches.’’ In Kitty Ven Leuven-Zwart and Ton Naaijkens, eds., Translation Studies: The State of the Art. Proceedings of the First James S Holmes Symposium on Translation Studies, 91^101. Amsterdam: Rodopi. Kussmaul, Paul (1995) Training the Translator. Amsterdam and Philadelphia: John Benjamins. Labov, William (1966) The Social Stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics. Labov, William (1972a) Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Labov, William (1972b) Language in the Inner City. Philadelphia: University of Pennsylvania Press. Labov, William (1994) Principles of Linguistic Change. Internal Factors, vol. 1. Oxford: Basil Blackwell. Lakoff, George (1987) Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. Chicago: University of Chicago Press. Lecercle, Jean-Jacques (1990) TheViolence of Language. London and NewYork: Routledge. Lederer, Marianne (1994) La traduction ajuourd’hui. Paris: Hachette. Lee, David (1992) Competing Discourses: Perspective and Ideology in Language. Harlow: Longman. Lefevere, Andre¤ (1992) Translation, Rewriting, and the Manipulation of Literary Fame. London and NewYork: Routledge. Leuven-Zwart, Kitty M. Van (1990) ‘‘Translation and Original: Similiarities and Dissimiliarities, II.’’Target 2.1: 69^95. Levin, Samuel R. (1976) ‘‘Concerning What Kind of Speech Act a Poem Is.’’ InTeun A. van Dijk, ed., Pragmatics of Language and Literature,141^60. Amsterdam: Elsevier Science.

Bibliography

251

Levine, Suzanne Jill. (1992) The Subversive Scribe: Translating Latin-American Literature. St. Paul: Graywolf. Levinson, Stephen C. (1983) Pragmatics. Cambridge: Cambridge University Press. Linell, Per (2001) Approaching Dialogue:Talk, Interaction and Contexts in Dialogical Perspectives. Amsterdam and Philadelphia: John Benjamins. Livingston, Ira (1997) The Arrow of Chaos: Romanticism and Postmodernity. Minneapolis: University of Minnesota Press. Lo«rscher, Wolfgang (1991) Translation Performance, Translation Process, and Translation Strategies: A Psycholinguistic Investigation. Tu«bingen: Gunther Narr. Lotbinie're-Harwood, Susanne (1991) Re-belle et infide'le: la traduction comme pratique de ree¤ criture au fe¤ minin/The Body Bilingual: Translation as a Re-Writing in the Feminine. Montreal: Editions du Remue-menage. Loveday, L. J. (1982) ‘‘Conflicting Framing Patterns: The Socioeconomics of One Component in Cross-Cultural Communication.’’Text 2.4: 359^74. Maier, Carol (1980) ‘‘Some Thoughts on Translation, Imagination and (Un)academic Activity.’’Translation Review 6: 25^9. Maier, Carol (1984) ‘‘Translation as Performance: Three Notes.’’ Translation Review 15: 5^8. Maier, Carol (1989): ‘‘Notes AfterWords: Looking Forward Retrospectively at Translation and (Hispanic and Luso-Brazilian) Feminist Criticism.’’ In Herman Vidal, ed., Cultural and Historical Grounding for Hispanic and Luso-Brazilian Feminist Literary Criticism, 625^53. Minneapolis: Institute for the Study of Ideology and Literature. Malmkjaer, Kirsten (1988) ‘‘Cooperation and Literary Translation.’’ In Leo Hickey, ed., The Pragmatics of Translation. Clevedon, UK: Multilingual Matters. Melby, Alan and Terry Warner (1995) The Possibility of Language. Amsterdam and Philadelphia: John Benjamins. Meschonnic, Henri (1973) Pour la poe¤ tique II, E¤piste¤ mologie de l’e¤criture poe¤ tique de la traduction. Pris: Gallimard. Morson, Gary Saul and Caryl Emerson (1990) Mikhail Bakhtin: Creation of a Prosaics. Stanford: Stanford University Press. Neubert, Albrecht and Gregory M. Shreve (1992) Translation asText. Kent OH: Kent State University Press. Nida, Eugene A. (1964) Toward a Science ofTranslating. Leiden: Brill. Nida, Eugene A. (1985) ‘‘Translating Means Translating Meaning: A Sociosemiotic Approach toTranslating.’’ In Hildegund Bu«hler, ed., Der u«bersetzer und seine Stellung in der O«ffentlichkeit/Translators and their Position in Society/Le Traducteur et sa place dans la societe¤ , 119^25.Vienna: Wilhelm Braunmu«ller. Nida, Eugene A. and Charles Taber (1969) The Theory and Practice of Translation. Leiden: E. J. Brill. Nietzsche, Friedrich (1887/1956) ‘‘The Genealogy of Morals: An Attack.’’ In The Birth of Tragedy and The Genealogy of Morals, 147^299. Translated by Francis Golffing. New York: Doubleday/Anchor. Niranjana, Tejaswini (1992) Siting Translation: History, Post-structuralism, and the Colonial Context. Berkeley and Los Angeles: University of California Press. Nord, Christiane (1997) Translating as a Purposeful Activity: Functionalist Approaches Explained. Manchester: St. Jerome. Ochs, Elinor (1991) ‘‘Misunderstanding Children.’’ In Coupland, Wiemann, and Giles, 44^60.

252

Bibliography

Ohmann, Richard (1971) ‘‘Speech Acts and the Definition of Literature.’’ Philosophy and Rhetoric 5:1^19. Ohmann, Richard (1973) ‘‘Literature as Act.’’ In Seymour Chatman, ed., Approaches to Poetics, 81^107. NewYork: Columbia University Press. Panofsky, Edwin (1951/1971) Gothic Architecture and Schlolasticism. New York: Meridian Books. Parret, Herman (1984) ‘‘Regularities, Rules, and Strategies.’’ Journal of Pragmatics 8: 569^92. Peirce, Charles S. (1958) Selected Writings (Values in a Universe of Chance). Edited by Philip P. Wiener. NewYork: Dover. Pergnier, Maurice (1993) Les Fondements socio-linguistiques de la traduction. Lille: Presses universitaires de Lille. Petrey, Sandy (1990) Speech Acts and LiteraryTheory. London: Routledge. Pratt, Mary Louise (1977) Toward a Speech-Act Theory of Literary Discourse. Bloomington: Indiana University Press. Pym, Anthony (1991) ‘‘Limits and Frustrations of Discourse Analysis in Translation Theory.’’Fremdsprachen 2^3: 29^35. Pym, Anthony (1992) Translation and Text Transfer: An Essay on the Principles of Intercultural Communication. Frankfurt/Main, Berlin, Bern, NewYork, Paris,Vienna: Peter Lang. Pym, Anthony (1993) ‘‘Performatives as a Key to Modes of Translational Discourse.’’ In linga Klaudy et al. (ed.),Transferre necesse est . . . Current Issues inTranslationTheory, 47^62. Szombathely: Pa«dagogische Hochschule Berzsenyi Da¤niel. Pym, Anthony (1998) Method inTranslation History. Manchester: StJerome. Rafael, Vicente L. (1980/1993) Contracting Colonialism: Translation and Christian Conversion in Tagalog Society under Early Spanish Rule. Ithaca, NY: Cornell University Press. Revised edition. Durham, NC: Duke University Press. Reinelt, Janelle G. and Joseph R. Roach, eds. (1992) Critical Theory and Performance. Ann Arbor: University of Michigan Press. Reiss, Katharina and Hans Vermeer (1984) Grundlegung einer allgemeine Translationstheorie. Tu«bingen: Niemeyer. Robinson, Douglas (1985) ‘‘Intentions, Signs, and Interpretations: C. S. Peirce and the Dialogic of Pragmatism.’’Kodikas/Code/Ars Semeiotica 8.3^4 (1985):179^93. Robinson, Douglas (1986) ‘‘Metapragmatics and its Discontents.’’ Journal of Pragmatics 10: 359^78. Robinson, Douglas (1991) The Translator’s Turn. Baltimore: Johns Hopkins University Press. Robinson, Douglas (1992) Ring Lardner and the Other. New York: Oxford University Press. Robinson, Douglas (1995) ‘‘Theorizing Translation in a Woman’s Voice: Subversions of the Rhetoric of Patronage, Courtly Love, and Morality by Early Modern Women Translators.’’TheTranslator 1.2 (November 1995):153^75. Robinson, Douglas (1996) Translation and Taboo. DeKalb: Northern Illinois University Press. Robinson, Douglas (1997a) Becoming a Translator: An Accelerated Course. London and NewYork: Routledge. Robinson, Douglas (1997b) Translation and Empire: Postcolonial Theories Explained. Manchester, UK: St. Jerome. Robinson, Douglas (1997c) Western Translation Theory From Herodotus to Nietzsche. Manchester, UK: St. Jerome.

Bibliography

253

Robinson, Douglas (1998) ‘‘Twenty-Two Theses on the Study of Translation.’’ Journal of Translation Studies (Hong Kong) 2 (June): 92^117. Robinson, Douglas (2001) Who Translates? Translator Subjectivities Beyond Reason. Albany, NY: SUNY Press. Sampson, Geoffrey (1986) Schools of Linguistics. Stanford, CA: Stanford University Press. Sampson, Geoffrey (1997) Educating Eve: The ‘‘Language Instinct’’ Debate. Cassell. Sapir, Edward (1949) Selected Writings of Edward Sapir. Edited by D. Mandelbaum. Berkeley: University of California Press. Sapir, Edward (1955) Language: An Introduction to the Study of Speech. NewYork: Harvest. Schultz, Emily A. (1990) Dialogue at the Margins:Whorf, Bakhtin, and Linguistic Relativity. Madison: University of Wisconsin Press. Searle, John (1969) Speech Acts. Cambridge: Cambridge University Press. Shallice, T. and P. W. Burgess (1993) ‘‘Supervisory Control of Action and Thought Selection.’’ In A. Baddeley and L. Weiskrantz, eds., Attention: Selection, Awareness, and Control,171^87. Oxford: Clarendon Press. Silverstein, Michael (2001) The Life and Metalife of Language: Discursive Semiosis and Sociocultural Formations. Manchester: St. Jerome. Simeoni, Daniel (1998) ‘‘The Pivotal Status of theTranslator’s Habitus.’’Target 10.1:1^39. Simon, Sherry (1996) Translation and Gender. London and NewYork: Routledge. Snell-Hornby, Mary (1988). ‘‘The Role of Text-linguistics in a Theory of Literary Translation.’’ In: R. Arntz ed., Textlinguistik und Fachsprache ^ Akten des Internationalen u«bersetzungswissenschaftlichen AILA-Symposiums. 13^16 April 1987, 433^48. Hildesheim: Olms. Sperber, Dan and Deirdre Wilson (1986) Relevance: Communication and Cognition. Oxford: Basil Blackwell. Thomson, G. (1982) ‘‘An Introduction to Implicature forTranslators’’. Notes onTranslation1. Tirkkonen-Condit, Sonja (1990) ‘‘Professional vs. Non-professional Translation: A Think-aloud Protocol Study.’’ In M. A. K. Halliday, John Gibbons, and Howard Nicholas, eds., Learning, Keeping and Using Language: Selected Papers from the Eighth World Congress of Applied Linguistics, Sydney, 16^21 August 1987. Amsterdam and Philadephia: John Benjamins. Tirkkonen-Condit, Sonja, ed. (1991) Empirical Research in Translation and Intercultural Studies: Selected Papers of theTRANS-SIF Seminar, Savonlinna 1988. Tu«bingen: Narr. Toury, Gideon (1980) In Search of a Theory of Translation. Tel Aviv: The Porter Institute for Poetics and Semiotics. Toury Gideon (1995) Descriptive Translation Studies and beyond. Amsterdam and Philadelphia: John Benjamins. Toury, Gideon (1998) ‘‘A Handful of Paragraphs on ‘Translation’ and ‘Norms.’’’ Current Issues in Language and Society 5.1^2:10^32. Venuti, Lawrence (1986) ‘‘TheTranslator’s Invisibility.’’ Criticism 28 (Spring):179^212. Venuti, Lawrence (1992) Untitled review of Andrew Benjamin, Translation and the Nature of Philosophy, Eric Cheyfitz, The Politics of Imperialism, and Douglas Robinson, The Translator’s Turn.Textual Practice 6.2 (1992): 316^24. Venuti, Lawrence (1995) TheTranslator’s Invisibility. London and NewYork: Routledge. Venuti, Lawrence (1998a) The Scandals of Translation: Towards an Ethics of Difference. London: Routledge. Venuti, Lawrence (1998b) ‘‘Introduction.’’TheTranslator 4.2:135^44. Venuti, Lawrence and Mona Baker (2000). A Translation Studies Reader. London and NewYork: Routledge.

254

Bibliography

Voloshinov, V. N. (1930/1973) Marxism and the Philosophy of Language. Translated by Ladislav Mateija and I. R. Titunik. NewYork: Seminar Press. von Flotow, Luise (1997) Translation and Gender: Feminist Theories Explained. Manchester: St. Jerome. Wechsler, Robert (1998) Performing Without a Stage: The Art of Literary Translation. North Haven, CT: Catbird Press. Whorf, Benjamin (1941/1959) Language, Thought, and Reality. Edited by J. B. Carroll. Cambridge, MA: MIT Press. Wittgenstein, Ludwig (1953/1984) Philosophical Investigations. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell.

Index

addressivity (Bakhtin) 104 ‘‘Afterword’’ (Derrida) 55, 68^9, 229, 232^3 Ahsen, Akhter 234 alien word (Voloshinov) 226^7 Althusser, Louis 83, 237^8 Altieri, Charles 132^3, 238 Approaches to Language (Harris) 227 Aristotle 79; on hexis/habitus 91 Asad,Talal 15 Ashley, Bob xi,158,194, 240, 243 Augustine 17, 29, 34,79, 81 Austin, J. L. xii, 1, 99, 223^4, 236, 238, 242; and the constative/performative distinction 3^6, 16, 18^20, 23^8, 29, 30^4, 43^5, 219, 225^6, 240; and the locution/illocution/perlocution distinction 21, 24, 131, 143, 146, 150, 160; on metaphor 125; and the parasitic speech act 44, 57^63, 123, 137, 160, 220, 229, 231; and the problem of separate individuals 95^9 Baker, Mona 11^13,134^7,140,156, 217^19, 222 Bakhtin, Mikhail 3^4, 6,12, 69,131,138, 164, 219^20, 223, 235; and dialogism 8, 40, 93, 95,163^4, discussed 99^108; and double-voicing 20, 53, 63,70, 81, 86,179, 223, 242, discussed 108^15; and language change 80,115^20,171; and the superaddressee 235; and theoretism 100^1 Bartley,WilliamWarren, III 226 Bassnett, Susan 14 Becoming aTranslator (Robinson) 241 Benveniste, Emile 3,18, 25^6, 34, 44, 95 Bergeron, Manon xii,190 Bourdieu, Pierre 69, 82, 91^2, 99, 238

Brisset, Annie 46^7, 51, 53, 57, 66^8,183 Brown, Penelope 156, 244 Burgess, P.W. 233 Burke, Kenneth 3, 24, 69, 234 Castoriadis 91 Catford, J. C. 7,14,15, 225 Catullus 114^15 Chamberlain, Lori 15 Cheyfitz, Eric 15^16 Chomsky, Noam 3, 22, 217, 223, 239; and the performance/competence distinction 26,103, and translation 7, 13^14 Clyne, M. 135^6 Communication and Language (Harris) 35 Consciousness Explained (Dennett) 110 constative linguistics, defined 4^6; recuperation of the performative for 24^6; in the study of translation 7^9,11^14 conventional implicature 21,143, 209^10; for foreigners 200^1; produced through immediatization 130,179^81 conversational implicature 20^1, 40, 68, 116, 209^10, part 3 (passim); and translation 18 conversational invocature 21, 54,143, 159,187^92,194, 221 Coupland, Nikolas 235^6 Culler, Jonathan 230^2 Damasio, Antonio 19, 39, 53, 69,72^6, 83, 94, 233 Davidson, Donald 137 Davis, Hayley 35 Davis, J. 84^5 Deleuze, Gilles 139,149

256

Index

Deleuze/Guattari, on Grice’s maxims 238 de Man, Paul 229, 238 Dennett, Daniel 110^13,168, 222, 235 Derrida, Jacques 3^4, 6,12, 37, 44, 45, 55, 94, 99,126,131,138,163,164, 219, 223, 224, 238, 241; on (in)stability 117; on iterability 39, 60^9,70, 81, 86,116,123, 153,160,183,194, 226; and the Searle^Derrida debate 229^33; on translation 114^15 Descartes’Error (Damasio) 72 DescriptiveTranslation Studies and beyond (Toury) 14, 82, 84 DeSousa, Ronald 233 ‘‘Des tours de Babel’’ (Derrida) 114^15 Developmental Mechanisms of Language (Harris) 227 dialogism (Bakhtin) 40, 95,100^10; as pandemonium (Dennett) 110^13 Dialogism (Holquist) 100^6 Diaz-Diocaretz, Myriam 15,17 Discourse and theTranslator (Hatim/Mason) 16, 41,134 Discourse in the Novel (Bakhtin/Emerson/ Holquist) 99,103 Dostoevsky, Fyodor 20, 99,106 double-voicing (Bakhtin) 20, 53, 62,70, 86,138,179,185, 220^1, 223, 242; active 110^13,114; of dynamic intendant 191; and language change 115^19,171; passive unidirectional 108^9,114; passive varidirectional 109,114; and somatics 81; in translation 113^15 Dryden, John 88 Dyer, Don xii dynamic intendant 167^70,171,176^80, 212^14 dynamic interpretant (Peirce) 164, 167^70,171,176^80, 212^14; defined 161^2; double-voiced 191; as site of interpretive conflict 173^4 Elias, Norbert 91 Emerson, Caryl 99,100^6,114, 235 Even-Zohar, Itamar 14, 86^7 face (Brown/Levinson) 156^7 Fairclough, Norman 217 Faust (Goethe) 34 Fawcett, Peter 7, 41^2,154, 238 Felman, Shoshana 18, 27^8

Fenton, Sabine 155 final intendant 167^70,171,176^80,183, 186, 212^14 final interpretant (Peirce) 167^70,171, 176^80,183,186,191, 212^14; expressive 175 Firth, J. R. 7,14, 22, 217, 225 Fish, Stanley 238 Foucault, Michel 128, 224 Freud, Sigmund 27 Gadamer, Hans-Georg 34, 224 Garneau, Michel 46^53, 54, 66^8,138, 183,191 Gass, Susan M. 243 Geertz, Clifford 7 Gentzler, Edwin 13,14 Giles, Howard 235^6 Gnostics 34 Godard, Barbara 15,17 Goethe, JohannWolfgang von 34 Gothic Architecture and Scholasticism (Panofsky) 91 Graff, Gerald 68, 231^2, 229 Grice, H. Paul xi^xii, 3, 6,18, 20^2, 40, 44, 53^4, 68,70,78, 80, 99,116, part 3 (passim), 219^21, 223, 236; and miscommunication 235; on translation 132^3 Grundlegung einer allgemeinen Translationstheorie (Reiss/Vermeer) 14 Guattari, Felix 139 Gudykunst,William B. 244 Guille¤n, Claudio 114 Gumperz, JohnJ. 7, 244 Gutt, Ernst-August 238 habitus theory (Bourdieu/Simeoni) 53, 82, 84,179; discussed 91^3; mosaic (Simeoni) 234^5 Halliday, M. A. K. 7, 22, 217, 223 Hanks,William F. 7 Harris, Brian 154 Harris, Roy 8,18, 35^9, 217, 223, 227^8 Hatim, Basil 16, 41,134^5,140,142,154^5, 156, 238 Hegel, G.W. F. 34,162,164 Heidegger, Martin 12, 34, 224 Hermans,Theo 14,15 heteroglossia (Bakhtin) 235 Hofstede, Geert H. 244 Holmes, James 14 Holquist, Michael 99^101,114

Index Holz-Ma«ntta«ri, Justa 14,17 Homer 114^15 Ho«nig, Hans G. 42 HowTo DoThingsWithWords (Austin) 4,18, 23, 26, 30, 45, 224 How to Play GamesWithWords (Harris) 227 Huxley,T. H. 227 Hymes, Dell 7, 244

1,

‘‘Ideology and Ideological State Apparatuses’’ (Althusser) 237 ideosomatics (Robinson) 76^9, 82, 105^6,116,128,166 idiosomatics (Robinson) 76,79^80, 82, 90,116,166; in Bakhtin 106 illocutionary acts (Austin) 95^9,143; of translation 18 illocutionary implicature 21,143,144^9, 151^2,160,182, 221 immediate intendant 77,166^70,174, 176^80, 212^14 immediate interpretant (Peirce) 77, 166^70,174,176^80, 212^14; defined 161^2 immediatization 130,143,185^90,194, 221; reimmediatization 211^16, 221 implicature 53^4,171,182,193, 221; conventional 21,143,171^81, for foreigners 200^1; conversational 20^1, 68,116, part 3 (passim), of translation 18; expressive (Altieri) 133,175; illocutionary/ perlocutionary 21,143,144^9,151^2; invocative 21,159,187^92,194, 221; metalocutionary xi, 21,143,151^3, in cross-cultural misunderstandings 195^7, 205^10, and translation 154^8 In OtherWords (Baker) 134 In Search of aTheory ofTranslation (Toury) 14 integrational linguistics (Harris) 19, 34^9 Integrational Linguistics (Harris/Wolf) 35 intendant triad 166^70 Intentionality (Searle) 232 internal dialogism (Bakhtin) 63,103^4, 106^10,163,194, 241 interpretant triad (Peirce) 103,159, 161^70 interpreting, conference 229; court, 154^5

257

Introduction to Integrational Linguistics (Harris) 35, 227 invocature 21, 54,143,159,187^92,194, 221 irony, theories of 49 iterability (Derrida) 39,70,116,123,138, 183,185,194, 220^1, 223, 224, 231; and addressivity 104; covers all experience 233; defined 19^21, 61^9, 232; and double-voicing 115^19; and immediatization 194; iterated politically 153; and somatics 81, 93^4; in translation 18, 66^8 Jacquemond, Richard 15 Jakobson, Roman 3 James,William 233, 241 Jappy, Anthony G. 241 Jesus, as constative linguist Logos 29 Johnson, Mark 233 Johnson-Laird, P. N. 233

32^3; as

Kabbalism 34 Kafka (Deleuze/Guattari) 139 Kant, Immanuel 75 Katriel,Tamar 244 Katz, Jerrold 18, 25^6, 29, 34, 58, 95,163 Kinbote (Pale Fire) 109 Krings, Hans 14 Krontiris,Tina 15 Kuhn,Thomas 63, 220, 241 Kussmaul, Paul 14, 42 Labov,William 7 Lacan, Jacques 224, 238 LadyWelby 161^2,164 Lakoff, George 72, 217, 223^4, 233 Landmarks in LinguisticThought (Harris/Taylor) 35 Language Connection,The (Harris) 227 language games (Wittgenstein) 18, 28^9 Language Machine,The (Harris) 35 Language Makers,The (Harris) 35 Language Myth,The (Harris) 35 Language, Saussure, and Wittgenstein (Harris) 227 Lardner, Ring 109 Late Great Planet Earth,The (Lindsey) 165 Lecercle, Jean-Jacques 127^8,139^40, 161, 238 Lederer, Marianne 11

258

Index

Lefevere, Andre¤ 14,15, 233 Leuven-Zwart, Kitty M.Van 42 Le¤vi-Strauss, Claude 217 Levin, Samuel R. 238 Levine, SuzanneJill 15,17 Levinson, Stephen C. 124,156, 219, 244 ‘‘Limited Inc abc . . .’’ (Derrida) 62, 226, 229, 231, 232 Lindsey, Hal 165, 241^2 linguistic approaches, anthropological linguistics 7, 22; cognitive 223; context-of-situationalism (Firth) 7, 22, 225; critical discourse analysis (Fairclough) 223; integrational (Harris) 19, 34^9, 223; metalinguistics 104; remainder (Lecercle) 128,160; segregational (Saussurean) 35^9; sociolinguistics 7, 22; systemic-functionalism (Halliday) 7, 22, 223 Linguistics Inside Out (Wolf ) 35 LinguisticTheory ofTranslation, A (Catford) 7 LinguisticThought in England, 1914^45 (Harris) 227 Literary Speech Act,The (Felman) 27 Livingston, Ira 9^10, 229 ‘‘Logic and Conversation’’ (Grice) xi, 20,121,124, 219, 236, 240 logos 29^30; contrasted with davhar (Bloom) 34 Lo«rscher,Wolfgang 14 Lotbiniere-Harwood, Susanne 15,17,142 Loveday, L. J. 135 Macbeth (Shakespeare) 46^53, 57, 66^8, 183 Macduff (Macbeth) 46^53, 68,191 Madness and Civilization (Foucault) 128 Maier, Carol 15^16 Malinowsky, Bronislav 225 Mall Rats (Smith) 119 Malmkjaer, Kirsten 134,137^8, 238 Map of Misreading, A (Bloom) 34 Marges de la philosophie (Derrida) 231 Marx, Karl 34 Mason, Ian 16, 41,134^5,140,142,154^6, 238 maxims (Grice) 20^1,78,124^6; accidentally violated by foreigners 202^4; applied to Venuti 142^3; and cross-cultural misunderstandings 159;

as culture-bound 128^30,136,143,193, 195^7, 200; Deleuze/Guattari on 238; dynamically activated 176^8,194; failure to fulfill 236; produced through immediatization 178^9; and reimmediatization 211^16, 221; Venuti on 140^1 ‘‘Meaning’’ (Grice) 240 Mehlman, Jeffrey 231 Melby, Alan 71^2, 85 Mellor, Anne 229 Meschonnic, Henri 13 metalocutionary implicature xi, 21, 143,151^3,182, 221; in cross-cultural misunderstandings 195^7, 205^10; and translation 154^8 Mikhail Bakhtin (Morson/ Emerson) 100^6 Morson, Gary Saul 100^6, 235 mosaic habitus (Simeoni) 234^5 Muecke, D. C. 229 Nabokov,Vladimir 109 ‘‘Neglected Argument, A’’ (Peirce) 163, 168 Neubert, Albrecht 134^5,140 Newmark, Peter 15 Nida, Eugene A. 11,13^14,15,17^18 Nietzsche, Friedrich 34, 83, 224, 233, 237 Niranjana,Tejaswini 15,16 Nishida,Tsuka 244 Nord, Christiane 14,15 norm theory, in translation (Toury) 82^91 Oatley, Keith 233 Of Grammatology (Derrida) 37, 68 Ohmann, Richard 238 On Deconstruction (Culler) 230, 232 On the Possibility of Linguistic Change (Harris) 35 Origin of Language,The (Harris) 227 Pale Fire (Nabokov) 109 Panofsky, Erwin 91 parasitic speech acts (Austin) 19^20, 44, 45, 68,123,137; in Derrida 231; and performance 220; as ‘‘remainder’’ 160 Paul 32^3, 242 Peirce, Charles Sanders 22, 99; on abduction 241; expanded dialogical model 148,153,171,172,176,186, 212^14, 221; the interpretant triad 21,

Index 103,131,143,159,161^70; and language change 80 performance 4^5,10; and Austin 23^4; and competence (Chomsky/ Katz) 25^6,103; of constatives 43^6; as constitutive of the performative 51^2; and conventions 20^1; of failure 28; as iteration 67^8; as metaphor for translation 41; as metaphorical field 39; and the parasitic 19^20, 52^3, 57^60, 220; as primitive magic 30; as (re)iteration 61^6; and somatics 80^1 performative utterance (Austin) 23^4; in animistic religion (‘‘spell’’) 30; in Bible 32^3; defined by community 30^2; recuperated for constative linguistics 24^6; in translation 42^53 ‘‘Performatives as a Key to Modes of Translational Discourse’’ (Pym) 42^3 Pergnier, Maurice 11 perlocutionary effect (Austin) 24, 95^9, 143; of translation 18 perlocutionary implicature 21,143, 144^9,151^2,160,182, 221 Petrey, Sandy 231, 238 Philosophical Investigations (Wittgenstein) 186, 224 Piaget, Jean 102 ‘‘Pivotal Status of theTranslator’s Habitus’’ (Simeoni) 71 Plato 29, 34,79,100 polyphony (Bakhtin) 63,123 polysystem theory (Even-Zohar) 14 Pope, Alexander 114^15 Port Royal 29 Possibility of Language,The (Melby) 71 Pound, Ezra 114 Pragmatics (Levinson) 219 Pratt, Mary Louise 238 Problems in General Linguistics (Benveniste) 25 Problems of Dostoevsky’s Creative Art (Bakhtin) 99 Problems of Dostoevsky’s Poetics (Bakhtin/Emerson) 20, 99,106, 111^12 ‘‘Problem of Speech Genres,The’’ (Bakhtin) 104, 235 ‘‘Problem of theText in Linguistics, Philology, and the Human Sciences, The’’ (Bakhtin) 235

259

‘‘Problem of theText,The’’ (Bakhtin) 99 Propositional Structure and Illocutionary Force (Katz) 25 Pym, Anthony xii,15^16, 57, 85,162,183, 229, 234^5; on translation as performative utterance 19, 42^53, 66^8,138,191 Rafael,Vicente 15^18 Rationality of Emotion,The (DeSousa) 233 Reading Saussure (Harris) 35 Redefining Linguistics (Davis/Taylor) 35 Reiss, Katharina 14 ‘‘Reiterating the Differences’’ (Searle) 231, 232 ‘‘Response to a Question from the Novy Mir Editorial Staff ’’ (Bakhtin) 235 RethinkingTranslation (Venuti) 138 RethinkingWriting (Harris) 227 Rhetoric (Aristotle) 91 Richard, I. A. 7 Robinson, Douglas 15,19^20, 53, 69, 99, 217, 227, 234, 235, 244 Sampson, Gregory 35, 217 Sapir, Edward 7, 234 Saussure, Ferdinand de 3^5,103, 217, 223, 225 Scandals ofTranslation,The (Venuti) xi, 139 Schlegel, Friedrich von 229 Schleiermacher, Friedrich 138 scientificity, and constative linguistics 12, 244^5 Searle, John 42,125, 226, 242^3; and the constative recuperation of Austin 33, 34, 58^60, 95, 219, 223; and the Searle^Derrida debate 62^3, 229^33 Shakespeare,William 46^53, 66^8,183 Shallice,T. 233 Shreve, Gregory M. 134^5,140 ‘‘Signature Event Context’’ (Derrida) 19^21, 61, 219^20 Signs, Language and Communication (Harris/Wolf ) 35 Signs of Writing (Harris) 227 Silverstein, Michael 7 Simeoni, Daniel 19, 53, 69, 99,179, 234^5; and Toury’s norm theory, 82^91; and Bourdieu’s habitus theory 91^4 Simon, Sherry 15 Smith, Kevin 119

260

Index

Snell-Hornby, Mary 42 somatic theory of language (Robinson) 19^20, 53, 69,70^81,102, 123,138,179,185, 220^1, 233^4; in Bakhtin 105^6; in Derrida 115; of iterability 93^4; and norm theory 82^91; and Peirce 166^7; somatic markers (Damasio) 19, 39^40,70^81, 83, 233 Sophists 34 Speech Acts (Searle) 33, 223, 231, 232^3, 242^3 speech-act theory (Austin) 23^8; constative applications of to translation 41^2; and iterability 61^6,194; and the parasitic 57^61, 68,123 Speech Genres and Other Late Essays (Bakhtin) 99, 235 Sperber, Dan 124,142, 238 stall, the (Ashley) 158^9,161,194 Steiner, George,72 Studies in theWay of Words (Grice) 236, 240 superaddressee (Bakhtin) 235 ‘‘Supervisory Control of Action and Thought Selection’’ (Shallice/ Burgess) 233 Synonymy and Linguistic Analysis (Harris) 227

Translation andTextTransfer (Pym) 42^3, 48, 50 Translation asText (Neubert/Shreve) 134 Translation Studies (Bassnett) 14 Translation Studies Reader (Venuti/ Baker) 17 translation studies, various approaches 14^17; cultural-studies approaches 11^13 Translator as Communicator,The (Hatim/ Mason) 156 Translatorisches Handeln (HolzMa«ntta«ri) 14 ‘‘Translator’s Invisibility,The’’ (Venuti) 138 Translator’s Invisibility,The (Venuti) 138 Translator’sTurn,The (Robinson) 19, 69^73,75^6, 80, 83^5, 90, 99,114^15, 233^4 Truth and Method (Gadamer) 34

Talbot,TaylorJ. 35 telementation, model of communication (Harris) 8, 37 theoretism (Bakhtin) 100^1 Theory and Practice ofTranslation,The (Nida/ Taber) 13 Thomas Aquinas 34,79 Thompson, John B. 92 Thomson, G. 135 Thousand Plateaus, A (Deleuze/ Guattari) 238 Ting-Toomey, Stella 244 Tirkkonen-Condit, Sonja 14 tonalization (Bakhtin) 119; as somatic 105^6 Toury, Gideon 14,15, 82^91, 234 ‘‘Toward a Reworking of the Dostoevsky Book’’ (Bakhtin) 235 Toward a Science ofTranslating (Nida) 13 Translation and Language (Fawcett) 7

Vermeer, Hans 14,15 Violence of Language,The (Lecercle) 161 Voloshinov,V. N. 226^7 von Flotow, Luise 15

Urmson, J. O. 226 ‘‘Utterer’s Meaning and Intentions’’ (Grice) 240 ‘‘Utterer’s Meaning, Sentence-Meaning, and Word-Meaning’’ (Grice) 240 Varonis, Evangeline M. 243 Venuti, Lawrence xi,15^18,76,138^43, 238^9; as constative rationalist 149^50 139,

Weber, Samuel 231, 232 Wechsler, Robert 41 ‘‘What Pragmatism Is’’ (Peirce) 163 Whitman,Walt 57,125 Whorf, Benjamin Lee 7 Wiemann, John M. 235^6 Wilson, Deirdre 124,142, 238 Wittgenstein, Ludwig 3,12,18, 28^9, 58^9, 63,186, 224, 226, 239 Wolf, George 35 Women, Fire, and DangerousThings (Lakoff) 223 Zukovsky, Louis and Celia 114^15