CENTRV. flflfleas []f THE r:l~~~es[]tfl

Transcription

1 CENTRV i\[ak~~ fffeas [f THE r:~~~es[tf ~E~TEA f[a f[)uf~~eu STU~ES ~~ f~f E, STYE, ~ ~TEAfAY THE[AY Centrum: Working Papers of the Minnesota Center for Advanced Stud1es in Language, Stye, and Literary Theory is pub1shed tw1ce a year, in Autumn and Spring, by the M1nnesota Center for Advanced Studies in Language, Stye, and Literary Theory. Subscriptions: $2.00 an issue ($2.50 outside the United States). Subscribers wi be bied upon their receipt of each issue. U[';'E ~ f 11 Minnesota Center for Advanced Studies in Language, Stye, and Literary Theory 1977

2 EDTOR Michae Hancher Engish EDTORAL BOARD F. R. P. Akehurst French and taian Danie V. Bryan Communication and Humanities Keith Gunderson Phiosophy Martin Steinmann, r., Robert Brown, uie Carson,. Lawrence Mitche, Donad Ross Engish Gerad A. Sanders Linguistics EDTORAL POLCY Centrum pubishes papers somehow bearing upon the theory of anguage, stye, and iterature, incuding computeraided anaysis of discourse, especiay papers with an inter-discipinary approach. Each issue is copyrighted in the name of the Minnesota Center for Advanced Studies in Language, Stye and Literary Theory; contributors are given permission to reprint their contributions upon request and without charge. Every manuscript submitted shoud be doube-spaced throughout (incuding quotations and, foowing the text, footnotes), conform to some estabished stye (e.g., that of Language or The MLA Stye Sheet, second edition) and be accompanied by a stamped, sefaddressed enveope. Every manuscript reporting research shoud be prefaced by an abstract of no more than 200 words. Each contributor is sent without charge five copies of the issue in which his contribution is pubished. COMMUNCATONS A communications--manuscripts, enquiries, books for review, subscriptions--shoud be sent to Centrum, 207 Lind Ha, University of Minnesota, Minneapos, Minnesota ACKNOWLEDGEMENT Pubication of Centrum is made possibe by a grant from the Graduate Schoo, and by editoria assistance furnished by the Department of Engish. The typescript was prepared by Caro Merin and Diane Yarosh. ohn Hagge and Ann T. Taranto are Editoria Assistants.

3 Lt~TRr: CENTRVM Lt~TRr: CENTRVM VOLUME NUMBER 2 FALL 1975 Prefatory Note "Speech Acts and Literature" (Forum) : STANLEY FSH Speech-Act Theory, Literary Criticism, and Corioanus MARTN STENMANN, R. Perocutionary Acts and the nterpretation of Literature BARBARA HERRNSTEN SMTH Actions, Fictions, and the Ethics of nterpretation E. D. HRSCH What's the Use of Speech-Act Theory? 121 Discussion KENNETH BURKE Words as Deeds BRUCE FRASER Warning and Threatening ELZABETH BRUSS Manufactured Signs: the Automobie Books Received Query Contributors Semiotics and

4 PREFATORY NOTE The papers in this issue by STANLEY FSH, BARBARA HERRNSTEN SMTH, MARTN STENMANN, and E. D. HRSCH are versions of their contributions to the forum, "Speech Acts and Literature," organized by the editor for the Seventeenth Annua Meeting of the Midwest Modern Language Association hed in Chicago in November A transcript foows of the discussion among the paneists and with the audience. (References by paneists to page numbers, and some references to detais of argument, refer to the origina versions of the papers predistributed to the paneists and audience. An expanded version of Professor Fish's paper appears in the Centennia ssue of MLN [91:5, October 1976; pp under the tite~ow to Do Things with Austin and Seare: Speech Act Theory and Literary Criticism.") Forthcoming issues of Centrum wi incude the foowing: BRENDA DANET / "Speaking of Watergate" ROBN LAKOFF / review of Edwin Newman, Stricty Speaking MARORE PERLOFF "Symboism/Anti-Symboism" KENT BALES review of Richard H. Brodhead, Hawthorne, Mevie and the Nove and Ruth Bernard Yeaze~anguage and Knowedge in the Late Noves of---genry ames DONALD ROSS review of Roger Fower, ed., Stye and Structure in Literature: Essays-r the New StYist1CS STEPHEN BEHRENDT ~review of Ronad Pauson, Embem and Expression: Meaning in Eng1sh Art of the E1ghteenth Century EARL MNER "AssaYn"gthe Goden Word of Engish Renaissance Poetics" WLLAM KEACH/ "'Pagiarism' and Eizabethan Poetry" PETER HUGHES / "Ausion and Originaity in Eighteenth-Century Writing" NORMAN FRUMAN "Originaity, Pagiarism, Forgery, and Romanticism" ELZABETH BRUSS / review of Arthur H. Robinson and Barbara Bartz Petchenik, The Nature of Maps

5 STANLEY FSH Speech-Act Theory, Literary Criticism, and Corioanus n the second scene of the second act of Corioanus, the tribune Sicinius makes a prediction that proves uncann1y accurate. He is specuating as to what wi happen when the hero petitions the peope, as the ceremony of the state demands. Corioanus, says Sicinius, wi "require them,/ As if he did contemn what he requested shoud be in them to give." How does Sicinius know this; what is the source of his understanding of Corioanus' future behavior? One answer, and an answer not atogether facetious, is that Sicinius may have been reading ohn Seare's book Speech Acts; for on page sixty-six of that book, he woud have found an anaysis of requesting that ceary shows why Corioanus woud be ikey to have troube with that act. The reevant rues or conditions are two: the preparatory condition, S beieves H is abe to do A; and the sincerity condition, S wants H to do A. Together they te us (what at some eve we aways knew) that if Corioanus were to "discharge the custom of request" (, iii, 148), he woud be acknowedging the right and the abiity of the citizens to certify his merit. Corioanus' point is that his merit needs no such certification; it is sef-vaidating and shoud be acknowedged as one acknowedges any natura phenomenon. n short, he refuses to accept the pubic and stipuated procedures by which the state determines vaue, and substitutes for them a standard of vaue he aone can recognize because he aone is its embodiment. t is from this confict that everything ese foows: on the one hand, the community cings to its conventions and to the obigations and commitments those conventions encode; and, on the other, Corioanus tries desperatey to hod himsef aoof from those same obigations and commitments by decining to recognize or participate in the community's conventions. That is why he is unabe or unwiing to request; it is aso why he can not easiy accept praise, for to do so \voud be to admit the right of someone other than himsef to determine what in his actions was creditabe. Corioanus' iocutionary behavior is consistent and monoithic: he is aways in the act of decaring his independence from the system of conventiona speech acts; he is aways [Centrum, 3:2 (Fa 1975), )

6 108 STANLEY FSH doing things (with words) to set himsef apart. He is the Greta Garbo of the Roman state, saying again and again, want to be aone. He says it twice in the cimactic scene, first when he refuses even to greet: " woud not. check my courage for what they can give,; To have't with saying 'Good morrow"' (, iii, 90-93). Greeting is the bottom ine of civiity; for a greeter is committed to nothing more than being a member of the (speech act) community whose conventiona means of expressing courtesy he invokes. f we te someone " wi never say heo to you again," it is understood that we wi have nothing whatever to do with him. Corioanus says that to the Roman citizens, and when, in the very next instant, they banish him, they merey say it back. The banishing then is anti-cimactic: it merey ratifies, with one iocutionary act, the message of Corioanus' many iocutionary acts. That message is deivered again, in thundering tones, when he utters his famous decarative: banish you. With this act he reocates the state and its power, pacing them where he has aways beieved them to be in the first pace, in his own breast, where they are sustained and nourished and recognized by a virtue that exists independenty of any pubic or communa system of vaues. The tragedy of Corioanus is that this "word esewhere," this word totay apart from speech act communities and their constraints, does not exist, at east not on this side either of madness or sainthood. The point of my exercise is that the course of this tragedy emerges step by step in the course of a speech act anaysis; that is to say, the events of the entire pay can be seen as the unfoding of the necessary consequences of the hero's iocutionary behavior: 1) n his inabiity to make a request or accept praise, he decares his independence of conventiona (that is, pubic) procedures for certifying merit or desert. 2) When he goes so far as to refuse to greet, his setting himsef apart from the community is compete, and he stands aone. 3) By banishing him, the citizens simpy ratify and confirm what he has aready done; by banishing them, he makes expicit his rejection of the community and his intention to stand aone, as a society of one, as a state compete in himsef, independent of a externa supports and answerabe ony to the aws he himsef promugates. n short, he decides to be a God. 4) As a God, he demands absoute obedience to his word (the sacred text), estabishing his promises and pedges as the aw against which no other considerations or oyaties

7 SPEECH-ACT THEORY, LTERARY CRTCSM, AND COROLANUS 109 can stand ("Thou shat have no other Gods before me"). 5) By going back on his pedge, he stands against it and is struck down accordingy. Dying, he acknowedges invountariy his necessary invovement in the community from whose conventions he sought to be free. n the end, the fiercey private man exists ony by virtue of the words of others ("He sha have a nobe memory."). must confess that ike this anaysis, but it is ony preiminary to a cautionary tae. The fact that Corioanus is avaiabe to a speech act description shoud not be generaized into a program for iterary criticism. The case is a very specia one, and we can see just how specia it is by ooking at some of the recent attempts to make iterary use of the theory. These attempts a fai in the same way, by trying to extend the theory into areas that are propery cosed to it. When, for exampe, Richard Ohmann decares that King Lear's apportioning of his kingdom is an infeicitous act, he confuses an iocutionary with a perocutionary criterion. n terms of the conditions on iocutionary acts, what Lear does is perfecty feicitous; the fact that it turns out bady is not a fact the theory can dea with or even recognize except to point out that it marks the boundary of its competence. Whatever speech-act theory is, it is not a rhetoric. Neither is it a psychoogy. n the anayses of Ohmann and others we occasionay hear of the speech acts of amenting, rejoicing, assuming, hoping, wishing. None of these is the name of an iocutionary act; that is, one does not conventionay perform them, athough one can report on their performance, which, in terms of the theory, takes pace off stage, in the interior recesses of the heart. This too is an area into which speech-act theory does not reach, athough this imitation is ignored when every verb is assumed to be the name of an iocutionary act. A simiar misunderstanding underies the attempts of some to make the theory a theory of narrative by rewriting as speechact rues the conventions of story-teing; but if these conventions are rues, they are reguative; that is, they are imposed on an antecedenty existing form of behavior. The rues for iocutionary acts, however, are constitutive; they do not reguate behavior, but enumerate the procedures which define it. Again the mistake is an honorabe one; it is made in the service of trying to get the theory to do something we woud a ike it to do; but it remains a mistake, and to persist in it does nothing more than add another set of terms on top of those we aready have.

8 110 STANLEY FSH An even more serious mistake is committed when the iocutionary/ocutionary distinction is made into a distinction between stye and meaning. Whatever the difference between a command and a promise or a request and an order may be, it is not a difference in stye; and it is hard to see what treating it as such coud ead to; in practice it seems to ead to the counting and arranging in statistica patterns of kinds of acts; but once this operation is performed the resuting data are uninterpretabe, or what is even worse, they are interpretabe in any direction one ikes. The one area in which the whoesae transference of speech act to iterary terms seems to have paid off invoves the drawing of a distinction between fiction and nonfiction; but this enterprise aso fais, and on three counts: First of a, the distinction is typicay drawn as one between assertions; that is, it is restricted to those speech acts to which one can reasonaby put the question, is it true or fase? Nothing is said (nor coud be said, think) of the difference between fictiona and no~fictiona promises, requests, questions, etc. Second, in practice, as ohn Seare admits, the distinction speech-act theory is abe to support is not between fiction and nonfiction, but between fictiona and nonfictiona discourse;2 works of fiction contain both, in an unprinciped mix; the same can be said of works of nonfiction, which can be made up of any ratio of hypothetica to seriousy intended utterances. The distinction survives, but its iterary interest does not, because there is nothing it can hep us to do. And third: even as a distinction it is not so crucia as we have been ed to beieve by those who mistake it for a distinction between anguage that has a grasp on the rea word and anguage that operates with diminished responsibiity to that word, and is therefore derivative (Austin's word is parasitic). The difference, think, is much ess sharp. To be sure, the conventions of nonfictiona discourse incude a caim to be in touch with the rea word, and aso incude procedures for determining whether that caim is justified in a particuar instance. However, those procedures do not examine the caim from a perspective externa to the conventions, but operate rigorousy within them. That is to say, the reaity nonfictiona discourse has a grasp on is the reaity its own conventions stipuate; the difference is finay not between discourse in touch with the rea and discourse that is not, but between discourse with reaistic pretensions, and!

9 2 SPEECH-ACT THEORY, LTERARY CRTCSM, AND COROLANUS 111 discourse without those pretensions. unprofitabe and miseading question. Which comes first is an ndeed, there don't seem to be very many questions eft that speech-act theory can answer; it is not a rhetoric, nor a psychoogy; it does not yied a theory of narrative or of stye; it can't distinguish between fiction and nonfiction or between what is rea and what is not. What can it do? We, it can provide us terms and descriptions for discussing Corioanus, and it can do this because Corioanus is ~speechact~ By this don't mean that it is fu of speech acts (a condition no piece of discourse escapes), but that it is about speech acts, the conditions on their performance and the commitments one enters into or avoids by invoking the conventions that constitute them. ndeed, as we have seen, it is just these conditions and commitments that the characters discuss, so that it seems at times that they are exampes in a textbook that Austin and Seare are writing. What this means is that the future of speech-act theory as a too of iterary criticism is severey circumscribed; it wi depend on the care its proponents exercise in choosing texts which, ike Corioanus, do what it does; and when such a text is found, the reationship between it and the theory wi not be one of object to description, but of description to description, of one account of how to do things with words to another. NOTES "Literature As Act," in Approaches to Poetics, ed. Seymour Chatman (New York, 1973), p "The Logica Status of Fictiona Discourse," NLH, 6 (1975), 332. Department of Engish The ohns Hopkins University

10 MARTN STENMANN, R. Perocutionary Acts and the nterpretation of Literature The centra question that 'm interested in--and don't think that everyone ese at this tabe is interested in this question--is this: what does a competent reader of iterature (fiction) do--not what shoud he do, but what does he do--in interpreting a iterary text, and how does he do it? A theory of iterature or of iterary interpretation is, take it, an answer to this question. A theory of iterature is not a critica method--not an instrument for interpreting texts. Like any other theory, it has as its soe function expanation and prediction of phenomena. t does not te readers how to interpret texts; it simpy makes expicit what they do when they competenty interpret them and what they have to know in order to do it. After forty years of the New Criticism, this view of iterary theory is hard to understand, et aone accept. For the New Criticism assumed that the chief function of iterary criticism is to interpret texts and that the specia expertise of the iterary critic is necessary for competent interpretation of them. A competent reader must, in other words, either be a iterary critic or sit at the feet of iterary critics--reading iterary criticism, taking courses in the interpretation of iterature, and so on. Hy assumption is quite different. assume that for every work of iterature--indeed, for any piece of discourse whatsoever--there is (or was) a body of competent readers and that they acquired their competence, not by any kind of forma instruction, but in the way that we a earned how to speak our native anguage. ust as there are native speakers of a anguage, there are (or were) native speakers of Eizabethan tragedy, seventeenth-century epics, eighteenth-century satires, Romantic yrics, and stream-of-consciousness noves. f Shakespeare's contemporaries were competent readers or viewers of his pays--pmla, of course, is fu of artices by critics who beieve that they were not--how did they acquire their competence? Surey not by reading or taking courses in dramatic criticism. Today, of course, there are no native speakers of Shake- Lc en t rum 1 3 : 2 (Fa ) ! ) 3! j

11 ! PERLOCUTONARY ACTS AND THE NTERPRETATON OF LTERATURE 113 spearean drama (or of any other antique texts) ; and, if we want to acquire competence in reading or viewing it, we must do it in the way that we acquire competence in Od Engish or Latin or Sanskrit: by instruction or sef-study based upon someone's theory (or grammar). But this necessity in no way affects my basic assumption. The function of a theory of Shakespearean drama is to make expicit what Shakespeare's contemporaries did when interpreting his pays and what they had to know in order to do it. Let me briefy sketch the fundamentas of speech-act theory; the distinction between four kinds of speech act. First, the utterance act or, as Austin caed it, the ocutionary act--which consists simpy of the utterance of one or more sentences in some anguage. Second, the propositiona act--which consists of an act of reference and an act of preaication. The writer refers to something and predicates a property of it, or refers to severa things and predicates a reation of them. By referring and predicating, the writer expresses a proposition. But the propositiona act--expressing a proposition--must not (as it sometimes is) be confused with asserting a proposition, which is an iocutionary act. Third, then, the iocutionary act--which is the act of do1ng something with the proposition expressed. Asserting that this proposition is true is one kind of iocutionary act; asking whether it is true is another; requesting that it be made true is a third; and so on. Finay, the perocutionary act--which is the effect that a combined utterance-propos1t1onar=ocutionary act has upon the reader, once the reader has understood it. The first three kinds of speech act--the utterance act, the propositiona act, and the iocutionary act--are ruegoverned. Moreover, as Staney Fish has pointed out, they are, ike games, governed by constitutive rues--rues that constitute or define the very acts they govern. Utterance acts, for instance, are constituted and governed by the grammatica rues of the anguage the writer uses. Foowing these rues--uttering sentences conforming to them--guarantees successfu performance of utterance acts. But the fourth kind of speech act--the perocutionary act- is not rue-governed; there are no rues governing the performance of perocutionary acts. n performing a combined utterance-propositiona-iocutionary act, doubtess the writer aways intends to perform a certain perocutionary act, to have a certain effect upon the reader: to get the reader to beieve an assertion, for exampe, or to answer a question or to grant a request. But there are no rues that the writer can

12 114 }~RTN STENMANN, R. foow that guarantee performance of the intended perocutionary act. The writer aways, of course, performs some perocutionary act--has some effect upon the reader--but he never knows in advance what it wi turn out to be. So far 've been taking about speech-act theory as it appies to what Barbara Smith cas natura discourse or to what ohn Seare cas serious or nonfictiona discourse. Now, what about fictiona discourse--noves, poems, pays? Among those who beieve that speech-act theory can make a principed distinction between fiction and nonfiction, there is considerabe agreement. Let me tick off some of the points upon which they agree. First, fiction, ike nonfiction, is governed by constitutive rues. Second, the utterance acts of fiction, ike those of nonfiction, are governed by the grammatica rues of the writer's anguage--in the broad sense in which grammatica rues incude semantic rues; and the sentences of fiction are meaningfu in precisey the way that the sentences of nonfiction are. t has been a mistake--a very persistent mistake--to ook for a genera distinction between fiction and nonfiction at the eve of utterance acts, grammatica rues, and sentences. So far as am aware, no one has ever been abe to find one there. Third, there is, as Barbara Smith says, a rue that in effect suspends the propositiona-act and the iocutionary-act rues of nonfiction- not just the iocutionary-act rues for making assertions, but a propositiona-act and iocutionary-act rues. This rue-suspends, for exampe, the propositiona-act rue that for the writer to refer to something that thing must exist or must once have existed. When ane Austen performed the utterance act of writing the first sentence of Emma--"Emma Woodhouse, handsome, cever and rich with a comfortabe home and happy disposition, seemed to unite some of the best bessings of existence.. "--this rue permitted her to use the referring expression Emma Woodhouse, as we a perfecty we know, without referring to anything--without making any commitment to the existence of anything. Finay, there is a rue that, in performing an utterance act, the writer of fiction must somehow pretend to perform a propositiona act and an iocutionary act as we or--a better way of putting it--what the writer must pretend to do is to report that someone ese performed an utterance act, a propositiona act, and an iocutionary act. n writing "My Last Duchess," for instance, Browning pretended to report a monoogue deivered by the duke. This way of distinguishing between fiction and nonfiction has, think, at east two advantages. One is that it makes the utterance acts of fiction just as meaningfu, and meaningfu!! 1 i )

13 PERLOCUTONARY ACTS AND THE NTERPRETATON OF LTERATURE 115 in just the same way, as those of nonfiction. t ends the futie search for a distinction at the eve of utterance acts, grammatica rues, and sentences. The other advantage is that it recognizes what, think, we fee intuitivey: that fiction is parasitic or rides piggy-back upon nonfiction--that it is make-beieve or mimetic. Mimesis is an ancient notion and, beieve, a correct one. Skipping over some other things, et me distinguish four kinds of meaning often confounded with one another in iterary theory. First, there is intended meaning--the meaning that the writer intends to represent or encode by performing a speech act or that the writer pretends that a character intends to represent by a speech act that the writer pretends to report. Second, there is conventiona meaning--the meaning that, by virtue of the reevant constitutive rues, the writer's speech act or a speech act that the writer pretends to report does have. Third, there is betrayed meaning--the meaning that, by virtue of natura aws (physica, bioogica, psychoogica, socioogica), the writer betrays--unintentionay reveas--or that the writer pretends that a character betrays. And, finay, there is perocutionary meaning--the effect that the writer's speech act or a speech act that the writer pretends to report has upon the reader: tragic effects ike crying, for instance, or comic effects ike aughing. Two comments about these four kinds of meaning. First, if meaning is what the reader gets through interpretation of speech acts, and if interpretation is inferencemaking, then perocutionary meaning is not meaning at a, and getting it is not interpretation at a. The reader does not infer the effect that the writer's or the character's speech act has upon him. The effect, rather, is simpy something that happens to the reader once he has inferred meanings of the other three kinds. count perocutionary meaning as meaning ony because it is often spoken of as that ("the tragic meaning of Lear")--which is a right (there are many meanings of meaning)--and because it is so frequenty confused with the other, inferred kinds of meaning--which is not. The dogma that every change in form entais a change in meaning, for exampe, rests upon this confusion or equivocation. Second, the reader infers different kinds of inferred meaning in different ways. Making an inference aways depends upon knowedge of some principe of inference. To infer conventiona meaning, the reader must knot,v constitutive rues. To infer intended or betrayed meaning, however, the reader must know natura aws. Let me iustrate this often-ignored distinction by a hypothetica exampe from nonfiction.

14 116 MARTN STENMANN, R. Suppose say to you, "Peope are foowing me a the time." Now, if you know the grammatica rues of Engish, the propositiona-act rues, and the iocutionary-act rues for making assertions, you infer that have asserted that beieve that peope are foowing me a the time. This inference you make upon the basis of your knowedge of the reevant rues. But you might aso make a further inference--namey, that 'm paranoid. This inference you make upon the basis of your knowedge, not of rues, but of psychoogica aws. Peope who go around saying that peope are foowing them a the time are probaby paranoid--uness it's the case that peope are indeed foowing them a the time, which has occurred quite often in recent times. Department of Engish University of Minnesota s 3 1

15 BARBARA HERRNSTEN SHTH Actions, Fictions, and the Ethics of nterpretation My paper is not concerned with appying or evauating speechact theory proper, of which am neither a practitioner nor an advocate. share with Mr. Fish both fascination with much that is practiced in its name and skepticism regarding its vaue for iterary criticism and theory. But my paper is concerned with speech acts and iterature, or rather with the distinction between what refer to as natura discourse and fictive discourse--a distinction which, by one set of terms or another, has received increasing attention from inguistic and iterary theorists of various persuasions, incuding, as it happens, ohn Seare, who has recenty deat with it as the distinction between "serious speech acts" and "nondeceptive pretended speech acts." n fictive discourse the conventions, assumptions, and expectations that normay govern the transactions between speakers and isteners do not atogether obtain. They are, rather, suspended or quaified or repaced by another set of conventions. A fictive utterance is understood to be not a verba act but the representation of such an act. Thus, when the French instructor iustrates a point of grammar by uttering the words "e m'appee acques," it is understood that those words represent, but do not constitute, the verba act of giving one's name in French, and do not have the force- make the caims, require the response or entai the responsibiities--that they woud have in a natura verba transaction. There are many varieties of fictive discourse serving many functions, from the verba payacting of chidren--"'m a wicked witch and 'm going to ock you in a dungeon"--to the exempary sentences of ogicians--"a swans are white," "The cat is on the mat"--to the yrics of popuar songs-- "Let's fa in ove," or, more recenty, "Let's do it in the road." Fictive discourse can be handy or diverting and it can be composed for specificay aesthetic ends, that is, as verba artworks: poems, stories, noves or the scripts of pays. [Centrum, 3:2 (Fa 1975),

16 118 ACTONS, FCTONS, AND THE ETHCS OF NTERPRETATON Fictive discourse obviousy incudes much that we do not speak of as iterature, and many works that we do speak of as iterature were not composed as fictive utterances. ~y concern in the present paper, however, is not with the reation between fictive discourse and iterature, but with the distinction between natura and fictive utterances, particuary the impications of the distinction for the ethics of interpretation. And my centra point is that there are impications: that the very nature of interpretation is different with respect to each, and that the socia assumptions and obigations that govern our interpretations of natura utterances do not obtain and cannot be invoked for the interpretation of fictive utterances. The fundamenta set of assumptions that governs natura verba transactions and makes anguage an effective instrument of communication for both speakers and isteners may be stated as foows: 1) That the determinants of an utterance--the speaker's motives in speaking, and the historica particuar conditions, circumstantia and psychoogica, to 'vhich his verba act is a response--that those determinants are impied by the fact and form of the utterance in accord with the reevant rues and practices of the inguistic community; and 2) That those determinants wi be inferred accordingy by the istener. n other words, the istener assumes that the speaker means what he says, and the speaker assumes that the istener wi take him to mean what he says, both in accord with the conventions of their shared inguistic community. f these assumptions are vioated in natura discourse, if the speaker's motives and circumstances are not propery refected in his utterance or not correcty inferred by the istener, we regard the speaker as inept or a iar, and we regard the istener as inept or--there's no word, interestinqy enough, for the converse of iar, but misconstrua, ike misstatement, can obviousy be maicious or sociopathic as we as dim-witted or soveny. neptitude of speech or interpretation is unfortunate but forgiveabe. Lying and wifu misinterpretation are, w~ say, unethica. Both dissove the network of mutua expectat1ons and responsibiities that maintain the good heath of the verba community and the effectiveness of anguage for a its members, both speakers and isteners. n fictive discourse the assumption that the speaker means what he says and t~at. the istener wi take him at his word is not vioated; 1t S suspended. t is understood that the fact and form of the! ' '!

17 BARBARA HERRNSTEL~ SMTH 119 fictive utterance do not impy the author's or performer's persona motives and circumstances, or any historicay particuar determinants. t is understood that we cannot and shoud not infer that the French instructor's name is acques, that the chid beieves hersef to be a witch, or that there is a particuar cat or mat to which the ogician was responding. Since the suspension is patent and presumaby understood, we do not regard the instructor, chid, or ogician as either inept speakers or iars. The question at the center of my paper is how we are to regard the istener or audience of a fictive utterance, particuary if the utterance S a verba artwork, and my quarre, insofar as the paper is poemica, is with those who woud invoke the ethics of natura discourse, particuary the assumptions and expectations of responsibe interpreting, to direct and constrain the interpretation of iterary fictive discourse. Professor Hirsch has maintained that the ethics of anguage hod good in a uses of anguage, that no viabe, reevant distinction has been or ever coud be made between different uses of anguage, that a interpretation shoud be governed by the intentions of the author, and directed toward the recovery of what he cas "oriqina meaning." My argument briefy is as foows: 1) That there is a reevant distinction; 2) That interpretation under any circumstances is ethicay governed not by the intentions of the author, but by the conventions of the inguistic community; 3) That with respect to a fictive utterance, meanings are not equivaent to authoria intention, and authoria intentions presumaby incude among other thinas the intention that the work be taken and interpreted as fictive, not natura, discourse; 4) That an utterance taken as fictive cannot be interpreted as natura discourse, because by definiton its form does not bear the same expressive or impicative reation to its author's or per former's motives and circmstances as does a natura utterance; 5) That the particuar historicay determinate meanings of a fictive utterance cannot be recovered because the fictive utterance, so regarded, does not have particuar historicay determinate meanings. Hhat woud---se-such meaninqs in a natura utterance are here historicay indeterminate and must be suppied by the reader as part of his experience of the work; and finay 6) That to insist that the interpretation of iterary fictive discourse be governed by the same socia considerations and criteria of correctness as the interpretation of

18 1 120 ACTONS, FCTONS, AND THE ETHCS OF NTERPRETATON natura discourse is utimatey to deprive iterature of some of its most vaued functions and characteristic effects, among which is the avaiabiity of iterary works for constructive, rather than inferentia, interpretation, that is, for suppying or creating meaning rather than recovering it. None of this impies that there are no ethica considerations in iterary interpretation, and sha in concuding mention two of them. The first arises from the fact that the effectiveness, and thus continued production, of a art, incuding verba art, depends upon the audiences who are abe and wiing to experience the effects that individua art works were designed to produce. What foows here is that we shoud strive to make ourseves and our students skifu, informed, and responsive readers capabe of appreciating and experiencing iterary art as such. The second concerns not the reationship of the interpreter to the author of the iterary work, but to his own audience, that is, those to whom he offers his articuated interpretations, which are of course, or presumaby, natura discourse, and thus subject to the same ethica constraints as any natura utterance. And here the primary constraint is the fundamenta one, namey, that the interpreter te the truth, which woud mean, amonq other things, 1) That he not caim to have identified as specific authoria intention or origina meaning what there is good reason to beieve that the author did not specificay intend or mean, or no good reason to beieve that he did; and 2) That the interpreter not caim to have estabished the uniquey vaid interpretations of the meaning of the work, when by the nature of iterary interpretation, no such thing is possibe. Department of Engish University of Pennsyvania s 1! s

19 E. D. HRSCH What's the Use of Speech-Act Theory? wi not waste much time in a preambe to my argument. Professor Hancher has given each of us a maximum of eight minutes to summarize our views, originay to be distributed in advance in written form before this meeting took pace. When faied to meet the deadine, incurred the task of presenting my argument and its summary both at once in eight minutes. Since have four points to make, can devote an average of two minutes to each point by using the famiiar technique of gross oversimpification, a vice prefer to overcompication. Point number one tries to answer the foowing question: What is the iocutionary force behind speech-act theory? Why did it come into being? The point want to make in answering the question is that the strengths and imitations of the theory as we as its vaue for iterary criticism have something to do with its Oxford origins. A reading of Austin's work up to How to Do Things with Words persuades me that one of Austin's intentions was to demoish the mentaistic concept of proposition as enunciated by Frege and impicit even in ogica positivism. Austin from the start took an extremey nominaistic view of propositiona statements, we iustrated by the foowing anecdote reated by saiah Berin; "Austin asked: 'f there are three vermiion patches on this piece of paper how many vermiions are there?' rone,' said. ' say there are three,' said Austin, and we spent the rest of the term on this issue." By taking a very skeptica stand on the status of propositions, which is to say on the idea that sentences represent cass concepts which are true or fase about the word, Austin was drawn into a highy interesting study of the nonpropositiona uses of actua anguage. He turned his attention first to performatives ike " promise you wi go," because these utterances are obviousy not propositiona statements about reaity, but rather statements acting out their own reaity. After pondering performatives, he reached the more genera and to iterary students more interestinq doctrine of iocutionary force. On this doctrine the meaning of a sentence is never merey the propositiona meaning as determined by its words under the rues of the anguage. This ocutionary or propositiona meaning is incompete unti we know the unspoken intention--the iocutionary force behind the ocution. Despite [Centrum, 3:2 (Fa 1975),

20 122 E. D. HRSCH the arguments which sti go on over the proper statements of Austin's doctrine, his basic principe is one that beieve to be an absoutey genera and correct observation about anguage use. Austin's vocabuary in describing this and other features of anguage was new. The very term iocutionary (which coud be transated as transpropositiona} was a phiosophica orientation to a ingu1st1c observat1on. The observation itsef woud have surprised no empirica student of anguage from von Humbodt to the present. For a the vaue of Austin's specuations, they remain narrowy tied to their origins in certain probems that concerned Oxford phiosophers in the thirties, and remain quite aoof from the traditiona, empirica study of anguage. To this provinciaity Austin's theories owe some of their interest and originaity, and aso some of their faws, which wi mention shorty. But my second point is a positive one. The interest that Austin's theories have generated now guarantees that we wi have a usefu counterpoise to the iterary theory which hods that a text is sef-sufficienty avaiabe to a reader. From a structura standpoint this iterary doctrine is equivaent to the phiosophica assumption that a particuar sentence states a particuar proposition. They are not the same doctrines, of course, but they have in common the assumption that one needs at hand ony a text and a knowedge of the anguage. f Austin is right, this postuate of the sef-sufficiency of the iterary text is wrong. An unsaid intention must be invoked to compete the said meaning--give it an accurate shape and scope. ocutionary force is required to compete ocutionary meaninq. Since the very same text might be governed by different kinds of iocutionary force, we need to posit a particuar kind of iocutionary force in order to understand the text's meaning. Austin presents the doctrine of iocutionary force as appying to a speech acts, whether spoken or written, and if this universa doctrine of iocutionary force in some form or other is correct, as beieve it is, then the present interest in Austin wi ead to a genuine cognitive advance over the theory of textua sef-sufficiency. Point three: if were asked to describe a faw in Austin's theory, even as now amended by various phiosophers, woud hazard a criticism of the kind Austin himsef often iked to make, based on empirica observation and experiment. Strawson and Seare have both raised the objection that the ine of demarcation between ocutionary and iocutionary meaning is neither cear nor universa. One of Austin's defenders, Ferguson, concedes the point that the various eves in Austin's theory, the phonic, rhetic, ocutionary,,and ~ocutionary, may be ooked upon as abstractions. He says, t m1ght for some purposes be ) 3! 1 3 1

21 WHAT'S THE USE OF SPEECH-ACT THEORY? 12 3 better to say that in the tota speech act, one compex intention is operative. Whether we speak of one intention or many seems to me mere taxonomica preference." But this is a generous concession. One of the big uses made of Austin's theories is the taxonomic one of cassification. The vocabuary of the various forces is, apparenty, indispensabe to the appication or vugarization of the theory. f we admit that a singe compex intention may wipe out the cear demarcations between the various semantic forces, we may decide that the terminoogy of speech-act theory shoud be regarded with as much nominaistic skepticism as Austin aotted to the word vermiion. Moreover, even the mode of the different forces requires some improvement as a mode. Psychoinguistic research has shown that the various ayers of speech interact with one another in different directions, not just one. On Austin's mode, the phonic act gives rise to the phatic act, to the rhetic act, thence to the ocutionary and the ocutionary. The sound gives us the word, the words give us sentences, the sentences yied ocutionary meaning, which must be competed by iocutionary meaning. But inguistic experiments have shown that the highest eve is aso the owesti that iocutionary force can ater not ony ocutionary meaning, but aso even sounds. Austin did concern himsef with such inguistic probems, as we know from the foowing anecdote of George Pitcher'si Austin said to him: "Suppose ask you, 'f cod water is iced water, what is cod ink?' and you repy 'ced ink.' You woud have uttered the sounds ' stink' but not the words ' stink' i rather, in saying the words 'iced ink' you uttered the sounds ' stink'." Austin might have been deighted to find out that he was wrong, and that the whoe matter is a ot more compicated than that. My fina point is about the use of theories ike Austin's in the practice of iterary schoarship. Athough have spoken more respectfuy of Austin's very Oxford anti-mentaism than privatey fee, have nonetheess fet constrained to present a more sympathetic account of the uses of his theories than had originay intended when agreed to come on this pane. For Austin reay is a usefu antidote to certain idoatries and mysteries in iterary theory, whether emanating from the od dispensation or from the new one ed by Derrida. But wi sti be oya to my origina intention in this ast point, where wish to question the vaue of speech-act theory as a guide to critica method. Austin's theory is, after a, a whoy genera theory about anguagei that is its phiosophica virtue, and competey to its credit as a theory. But in'some respects generaity in a theory is a handicap to anybody who wants to use it as a methodoogica too in criticism. The more genera a theory is, the ess usefu it is ikey to be as a methodoogica too. This principe of increasing useessness

22 124 E. D. HRSCH can be proved a priori for a cases where a decision has to be made in criticism or schoarship--for instance, where one of two interpretations is to be decided on. Obviousy, whichever interpretation you choose wi be equay sanctionea by the genera theory of speech acts. f that weren't so, the theory woud be fase as a genera theory about speech. n my experience one of the most usefu precepts to be earned from a study of genera theories is to beware of them, particuary if they are both genera and true. t is deusive to think they can directy sponsor a method of criticism or any concrete judgement about a iterary text. Department of Engish University of Virginia 5 s 1 s

23 Discussion [See prefatory note, p FSH: n genera find mysef in compete agreement with everything that Professor Hirsch has said; disagree in part with what Mr. Steinmann has said; and disagree whoy with everything that Professor Smith has said. My agreement with Professor Hirsch extends to his two main points; and that is, the significance of Austin, Seare, and whatever--the significance of speech-act theory as a theory which gives us a kind of principed way of arguing the thesis that you can't get everything out of the anguage, that is, you can't buid up everything simpy out of the words and out of syntactic or semantic rues. offer as an exampe an exampe often use in cass. f were waking down the street with a gir and she ooked up at me as we approached a rather arge and brawny individua who was standing in a menacing fashion in our way and she said to me, "He doesn't ook at a that strong," we coud specuate as to exacty what she meant, but we woud a agree, think, that one of the things she didn't mean is "He doesn't ook at a that strong." t seems to me that in the sense that we know immediatey- or woud know immediatey--that this is a dare, or a chaenge, or some other speech act, the taxonomica specifications of which we coud argue about--in that sense speech-act theory heps to [Gap between tapes. if the end of his paper warns against the iterary abuse of [Centrum, 3:2 (Fa 1975),

24 126 DSCUSSON this theory--except in those rare cases, ike Corioanus where conditions as genera as those it tries to describe are ~he overt subject of the work in question. Otherwise--that is, if you use speech-act theories as an a-purpose interpretive key--it wi either have nothing to te you or it wi te you very trivia things. Now, my disagreement with Professor Steinmann reay can be boied down to, think, one basic issue from which some others coud be extrapoated. t seems to me that he is mistaking an anaytic account of the breakdown of the aspects of a fuy-aunched speech act with a genetic account, or with a possibe genetic account. And this is a mistake that see often in the iterature--that is, the assumption (to take just Seare's terminoogy) that if you distinguish between iocutionary force and then, in Seare's parentheses, propositiona act, that what you have is a ki11d of recipe which invoves starting with a pure and basic ~ropositiona act to which you add, as you might add an intensifier to a piece of meat, an iocutionary force. n other words woud endorse again Professor Hirsch's suggestion on the top of page four of his paper that the matter is not simpy a buid-up matter from a kerne of basic and atomic significance or meaning, to these higher forces, but rather,! woud, 'm afraid, endorse a buid-down theory from an interpretive or intentiona gesture to in fact those aspects which can ony be picked out afterwards. n other words am denying that there is--and here come to my disagreement with Professor Smith-- am denying that there is a use of anguage which hooks up with the rea word (as opposed to denying that there is a use of anguage which caims to hook up with the rea word.) Let me just quicky run through, keeping the pages of Professor Smith's paper in mind, some of the points of disagreement find between us. 'm on page two of her paper, in which she is arguing that someone who has asked us to take the poem perhaps written by him or her in such a way--asks us to regard a poem as a reveation of his persona state. Someone who so asks us is somehow vioating the way in which poems are constituted and presented. t seems to me that on page two and on other pages in her paper Professor Smith is decreeing, giving us aws about what wi and wi not count as a poem, and offering us, think, no reason for these aws, except their own decreeing. That is, assume, that to the counterexampes of phiosophica poetry ike Pope's Essay on Man, or Eizabeth Barrett Browning's Sonnets from the Portugese, 1 s 1 '

25 DSCUSSON 127 which were intended, we might assume, at one point as direct messages to Robert Browning--we must assume that a theory ike this one woud disquaify them as poetry at that point, a move which has actuay been made in both cases, in one case by Ohmann and in another by Eias Schwartz, Schwartz denying the tite poetry to the Essay on Man, Ohmann to Sonnets from the Portugese, at east when they were used to send Robert Browning a message. t seems to me that it is hard to see what one does in aying down rues ike this except aying them down. On page six on the bottom Mrs. Smith writes, "t is understood, however, that the appropriate consequences of a poet s act in composing and offering a poem are not a matte~ of securing our services or affecting our behavior." My answer to that is, sometimes they are, and sometimes they re not. That is, with some exampes that we coud both chose, that is the case (that it is not a matter of trying to secure our services or affect our behavior) ; in some other cases--! coud tak for days on the exampe of Miton--it seems to me to be patenty so, or at east, certainy not disprovabe that it s so. Now it seems to me again that everything that Professor Smith says in the paper stems from the notion that we do something as readers and interpreters differenty when we read poetry than when we engage in natura discourse. And r m going to deny that difference, but from the other end. wrote down on a scrap of paper whie Mrs. Smith was taking one of her sentences: "There are many varieties of fictiona discourse." My counter-statement is yes, and nonfictiona discourse is one of them. This, think, is the core of what have to say; that in fact that we necessariy do what we do when reading or istening to anything. Professor Smith asks, "Do the ethica imperatives that we acknowedge in ordinary discourse hod good in a uses of anguage?" (asking it with an increduous tone); but am going to answer to that increduity, "yes"; athough do not mean by this that the prime mora obigation is the discovery of origina meaning. Were this the time and the pace, might argue with Professor Hirsch about the status or possibiity of origina meaning; but don t think his argument in Vaidity in nterpretation necessariy goes aong with this point. Now, et me make it cear what m suggesting. And that is that the imperative of reading--or istening--and say when reading a newspaper, or a nove; or istening to me, or

26 128 DSCUSSON to an opera--is to assume that we are deaing with intentiona beings. With that assumption we can operate in a number of ways depending on the reationship we think obtains between what is said and what is meant. t seems to me that that assumption, however, hods good aways. But it is not aways the case that we are operating on the assumption that the reationship between what is said and meant is aways the same. woud make two points: one, those reationships between what is said and what is meant do not distribute themseves neaty as to any fiction/nonfiction distinction; and two, no one of them is basic or prior to the others. Finay, on page five of Professor Smith's paper, second paragraph, she begins to describe moray reprehensibe behavior on the parts of isteners and speakers. Taking iteray what we mean payfuy, persisting in interpreting atenty what we offer patenty, for instance, "'m being foowed a the time," extracting reveations or unfattering unconscious motives from our casua remarks, so forth and so on. t seems to me that this is precisey what happens in natura or norma discourse and it's aso precisey what happens in iterary discourse, insofar as we are readers, or interpreters, or both. The norma speech situation, which Professor Smith sets up, or aways hods up impicity in her paper, argue does not exist. t's a tremendousy ideaized situation, or if it does exist, once in a whie, it exists ony as one very specia corner of the discourse universe. As when, to use Seare's favorite exampe (and here Seare woud be on Professor Smith's side) : Seare says if the waitress asks, "What' 11 you have?" and repy, "A hamburger, hod the reish," that's basic communication, that's rea. t's not so much that fictiona discourse doesn't have the properties Professor Smith ascribes to it, but that nonfictiona discourse, to a great extent, has those same properties, except in the idea, noise-free situation which she must posit as the norm, but which do not think is the norm, and which think rarey occurs. concude by repeating a warning that have sounded before, and repeat it in the confidence that it wi have to be repeated again: ordinary anguage, as it usuay appears in these debates, is an impoverishing notion. SMTH: curiousy enough, have no quarre with Mr. Fish; ony admiration for a paper that manages at one stroke--though in more than eight pages--both to embrace and banish speech-act theory whie simutaneousy eaborating a perfecty dazzing reading of Corioanus. s

27 E DSCUSSON 129 Since argue with Mr. Hirsch in my paper, wi not do so to any great extent now, though woud ike to say that don't think that speech-act theory wi fight the battes for which he seems incined to impress its services. ocutionary force, if understand it correcty, is not equivaent to unsaid intention. And the iocutionary force of an utterance is appreciated, not by divination, but precisey through that knowedge of anguage--mr. Hirsch speaks of "the text" and "knowedge of anguage"--but [in any case, it is appreciated through the knowedge of anguage which incudes not ony knowedge of a exicon and a grammar, but aso of the conventions that govern iocutionary meanings. One of the signa contributions of speech-act theory, to my mind, is its drawing attention to how much more of verba behavior is rue-governed than was sometimes sppposed, and how much more is conventionay impied by a verba act. aso think that speech-act theory wi not come to Mr. Hirsch's aid in fattening out the distinctions among uses of anguage, since athough Austin quite propery did not make a distinction between inscribed and spoken speech acts, he nevertheess recognized and took pains to point out that certain verba structures--among them poems and pays, if remember the passage- do not have the iocutionary force or indeed [the genera status of speech acts. want to make some comments on Mr. Steinmann's paper and then, if have a minute, perhaps say some things about Mr. Fish's comments on mine. For obvious reasons found much to agree with in its genera drift, that is, Mr. Steinmann's paper. Since, however, had to question so many of the individua points of which it is composed, coud not find mysef utimatey at ease even with the genera drift. t seems to me that Mr. Steinmann is cavaier about some difficut matters, such as the meaning of meaning and the nature of iterary interpretation, and that he overcompicates reativey simpe matters, such as the means by which a fictiona character is created. Maybe that's not so simpe, either, but he overcompicates it quite a bit. And aso the number of pretences invoved in fictiona discourse. Aso, find troube in his easy equation of fictiona discourse and imaginative iterature by way of a parenthesis and, a aong the ine, his use of the terms, concepts, and categories of speech-act theory as though they were scriptura. ' just expand a few of these points. Mr. Steinmann says that to interpret a text is to discover

28 130 DSCUSSON its meaning, and then provides four senses of "meaning." He says these are often confused in comments about interpretation of iterature. That's probaby true. Nevertheess, don't think that those four are exhaustive, and aso think at east one of them, as he seems to acknowedge, isn't propery regarded as a meaning at a--that is, perocutionary. Of the second, betrayed meaning, we, there are those for whom the primary interest of iterature is probaby the discovery of betrayed meaning; there are others--! wonder what Mr. Hirsch woud fee about this--who woud regard the search for betrayed meanings as a betraya of the author's intention. There are intended meanings, what the writer intends his encoding to impy, and conventiona meanings, what his encoding actuay impies when interpreted by reevant rues. As Steinmann suggests, where. there is a discrepancy between those two (that is, intended and conventiona), it's hard to say how we determine the first of them (that is, intended) or, it might be added, how we know that there is a discrepancy. Which eaves conventiona meaning itsef: what the text impies when interpreted in accord with inguistic convention. This is certainy the primary object of textua interpretation; but it is aso the primary object of the interpretation of any speech act, inscribed or voca, and indeed of any symboic or conventiona action. don't see that Steinmann's discussion of texts and meanings tes us anything about iterary interpretation or has any reevance to his own distinction, earier in the paper, between fictiona and nonfictiona discourse, a distinction which he seems to have forgotten here, where he treats texts as iterature and iterature as texts, much as Mr. Hirsch does, both of them ignoring, or obscuring (or feeing that something doesn't exist) the ways in which the meanings and interpretation of fictiona discourse are distii1ctive. Mr. Steinmann says that a writer creates a character by uttering sentences containing first-, second-, and third-personreferring phrases. But uttering such sentences is something we a do a the time, without thereby creating characters, except in the sense that Mr. Fish might wish to pursue, namey that the distinction between characters and rea persons is as tenuous as a distinctions between fiction and reaity and that in speaking of peope we actuay create them. t's a point with which can have some sympathy, but not just now. To return to Mr. Steinmann's observations, it is not, think, that a noveist creates characters by uttering sentences with person-referring phrases, but that he impies characters by composing, fashioning, fabricating such sentences. f the characters are actuay created by anyone it is by the reader, who, in suppying the appropriate fictive persons for the given s s

29 DSCUSSON 131 fictive references, gives body, and the sembance of vitaity, to what woud otherwise be an inert array of words on a page. On the number of pretences invoved in a fictiona discourse: there is ony one pretence, and Mr. Steinmann's Chinese boxes are, think, a fancifu soution to a gratuitous probem. The probem may have been created in part by his use of the term "pretend" to characterize the reation of fictive to nonfictive speech acts. prefer the term "represent" for a number of reasons, among them the fact that it does not presuppose an agent, and therefore does not present a probem about who's doing the pretending. n a three exampes, the passages from two noves and Browning's poem, it is not that the authors are pretending to perform speech acts, but that they have, in a seriousness, composed verba structures that represent speech acts. ane Austen has composed a sentence that represents a description of a woman named Mrs. Bennett. Anything that Mrs. Bennett is described as saying in the nove is part of the represented description of her. The pretence attaches not to what Mrs. Bennett says, but to the statements by which she is described as saying it. Henry ames has composed a sentence that represents the expression of a sentiment concerning New York. There are no boxes within boxes in the nove, ony the singe box, or frame, that contains the entire structure of represented sentiments and reports of which the nove consists. Browning has composed some sentences that represent a string of speech acts performed, in this case, by a highy individuaized speaker in a vividy impied dramatic context, but here too there is ony one representation or pretence, namey the one produced, or engaged in, by Browning in composing fictiona discourse. n each case it is the entire work, poem or nove, that is fictiona, the entire verba structure that constitutes a representation of discourse. This is a point which has euded not ony Mr. Steinmann, but, might add, aso Mr. Seare, who, in the recent artice on fictiona discourse mentioned earier, manages by the end to fudge matters atogether by reintroducing into fictiona works a certain number of aowabe serious speech acts. suspect that the roots of the probem here are the specificay phiosophica commitments and concerns of speech-act theory which is aso, suspect, why speech-act theory may be finay of distincty imited vaue for iterary theory. On Mr. Fish's comments on mine there are a number of points woud want to concede; perhaps don't even have enough time to concede them. think that his suggesting that my descriptions of natura discourse are highy ideaistic is we-taken.

30 132 DSCUSSON Let me just mention a few points, if can remember them. He observes that on page two, regard as unethica the feow who gives us a poem and asks us to take it as something other than a poem. But that wasn't my exampe. My exampe is that someone offers us something as a poem, says it's a poem, and somehow we get the sense that there's something ese that's being asked of us; and what said there is that he's trying to have it both ways: that in offering it as a poem, he's simpy suggesting, impying, that we regard it~ith certain attitudes, but at the same time he's suggesting that there are other attitudes that he reay wanted. So think that that question shoud reay be resoved by my having made the point cearer to begin with. Status as poetry is not the same as status as fictive utterance. There's a tremendous probem with terminoogy which can't say that have soved. want to use the word "poetry" in order to impy much that we usuay impy by it; but don't want to say that fictive discourse is equivaent to poetry, or the other way around. And in an artice which wrote, actuay, after wrote some of the beginnings of this, caed "On the Margins of Discourse" [ Critica nquiry,,. 4 (1975), , devote mysef entirey to considering precisey those contrary cases that Mr. Fish suggests are there; for exampe, what happens with the poem that is orig.ina.y designed for a specific audience and reay as natura discourse. don't mean to categorize-- in fact, the whoe notion of fictive discourse means that one can't categorize, you might say upon inspection, any particuar verba structure--that which particuar way we regard it is a combination of many things, sometimes specific statements, but then we can disregard them too. And here is a point made that perhaps Mr. Hirsch wi take up, and that is, suggest that (and here disagree with Seare) it is not the author's intention necessariy that defines something as fictive. We can ca something fictive and take it that way. Anyway, many of the arguments, or, et's say, sef-defenses, against Mr. Fish's points, which can see are produced by this paper, hope argue better where have more space, in the artice that just mentioned. STENMANN: A paradox of this discussion is that we find ourseves agreeing with our harshest critics; and, just as Mrs. Smith found most of her sympathies with Mr. Fish, _ find most of my sympathies with Mrs. Smith. Let me begin by making a coupe of genera observations that think distinguish my genera view--quite apart from the detais of speech-act theory; whether that's right or wrong- but distinguish my genera view from that of Mr. Fish and that s! )

31 E DSCUSSON 133 of Mr. Hirsch; and, for a know, from that of Mrs. Smith ( can't be sure). My first genera point is this: that a theory of iterature, or of iterary interpretation, is not a critica method, or methodoogica too or too of iterary criticism--that's not what a theory is supposed to be. A theory of anything whatsoever is supposed to be an account of it that expains some phenomenon that occurs. There are various kinds of theory that one might propose in connection with iterature; and et me briefy distinguish some of those. One kind of theory that one might propose in connection with iterature is a theory of competence, a theory of iterary competence; this theory woud attempt to answer the question, what does a writer need to know in order to compose a nove (et us say; it doesn't matter--any other genre you want to think of) and what does a reader need to know in order to interpret the nove once it has been composed. This is a theory of competence, of the knowedge that the writer and the reader must share in order to go about their business of,writing and reading. Knowedge of iterary conventions, knowedge of inguistic conventions, knowedge of speech-act rues, particuar historica facts (that there was a man named Napoeon and there was the batte of Austeritz, and so forth and so on). That's one kind of theory, and that's a kind of theory that am at the moment interested in and was trying to set forth in my paper. Mr. Fish seems to beieve that was trying to do a different kind of theory from that. This might be caed a theory of production or genesis--what are the steps that a writer must go through, what are the processes (if you ike that word, and don't) that a writer must go through in order to compose a work and what are the processes that a reader must go through in order to interpret it. know nothing about that and don't think anybody ese does either. don't mean 'm not interested in it, it's simpy that know nothing about and don't pretend to know anything about it. And don't think that was trying to do anything of that sort. My first genera point, then, is that a theory is something that pretends to account for some data, to account for some phenomena; it may do this or it may not do that, but that's its aim. Now, what a theory of competence tries to do is to (insofar as it's deaing with rues speech-act rues, or rues of iterary genres, or what not), what a theory of competence tries to do is to make expicit or formuate these rues. Not to ay down rues but to formuate rues that are aready there. And Mr. Fish accused Mrs. Smith of aying down rues. We, she wasn't aying down rues, she was, think, trying to formuate the rues that we a know when we interpret works of

32 134 DSCUSSON iterature. And she may be wrong in her formuations of them, but she's certainy not trying to egisate in the matter, ay down aws. We, the soe function of a theory, as see it, is expanation and prediction of phenomena. n the past forty years iterary criticism and schoarship has argey consisted of interpretation of texts, of finding and using critica methods or methodoogica toos. This regard as a major disaster. think that by and arge the ast forty years of iterary criticism in America have just been a waste, in that what they have devoted themseves to is at best a task of engineering and not a theoretica task at a. For one thing don't think that any methodoogies are needed for interpretation. assume that there are native readers of iterature in the same sense that there are native speakers of a anguage, that peope earn how to read noves, on the whoe, in the way they earn how to do anything ese, not by forma instruction in schoo. How did Shakespeare's audience earn how to interpret his pays? Did they have teachers who had Ph.D.'s in interpretation that tod them how to do it? They picked it up as they picked up their knowedge of the anguage, and woud assume that they interpreted the pays correcty. The assumption of the ast forty years has been that you can't read a work of iterature uness you've got a methodoogy for reading it, and that you can't reay have a methodoogy uness you have a Ph.D. in some reevant discipine. Now there's a grain of truth in this in that we a have students who are not native readers of iterature. And one of the tasks of ~ eementary courses ought to be to provide them with the competence--try to provide them with the competence--that they don't pick up in the norma way. But surey this is not schoarship- this is a kind of minor pedagogica task. Perhaps we need a methodoogy when we're deaing with very antique or antiquarian works. But surey not for modern works. think it's preposterous--the journas (MLA's the worst offender) fied with new interpretations of works by Faukner. Everyone ought to know how to read Faukner if he's read a ot of noves, and [doesn't need any experts to hep him. A few particuar points, if have just a moment. First, Mr. Fish's paper. As said disagree with him. A speech-act theory doesn't pretend to be a too of anything whatsoever and therefore is useess as a too; and nothing against it that it's not. think that he confuses, in one of the points he makes in the paper, utterance acts with iocutionary acts; he says that one's intention to make a promise is identica with his being responsibe for the promise that he makes and think that s 3!

33 DSCUSSON 135 we do have intentions that are quite independent of anything that we do say and we a know ways of making guesses about what other peope's intentions are. We of course never know anyone ese's intention for sure, but this is just a specia case of not knowing anything ese for sure either, that is a matter of empirica fact. The ony things we know for sure are tautoogies. Anything ese we can be mistaken about. But as a matter of fact we have a great dea of expertise in judging what other peope's intentions are. A point or tvo about.hrs. Smith s paper. On the whoe fee mysef, as say, in sympathy with it, so don't wish to criticize it very much. think in some of her exampes she does confuse betrayed meaning with conventiona meaning; [she starts out taking about one but ends up giving an exampe that reay appies to the other. But think she's essentiay correct in taking about fiction as suspending the norma rues of ordinary discourse. Mr. Hirsch, just a word about his. He asks what is the iocutionary force behind speech-act theory. don't reay care about the genesis of speech-act theory. The question to ask about the speech-act theory as deveoped by Austin or eaborated by Seare is whether it's correct or not--does it indeed account for the phenomena that it's supposed to account for. f it doesn't it's bad. What motivated Austin in doing this, what Austin's antecedents were, what got him to thinking this way, this is of no reevance so far as can see. agree with Mr. Hirsch that one of the good things, certainy, that speech-act theory has done is to ay to rest the notion that a you have to have is the text in order to interpret; and some peope have taked about the text as if you didn't even have to know t.vhat the anguage was in order to interpret it. That's enough. HRSCH: We're amost at the end of our aotted time. 'm going to be brief in the hope that there can be some audience participation; and aso 'm going to take a sighty different tack since it's the end of the session, and ' not refer in detai to each of the individua papers, but rather to some of the topics that have been raised in the course of the discussion. First of a want to take up the actua reaity of fictiona speech and the actua reaity of so-caed ordinary speech or norma speech situations. The distinctions that have generay heard in this discussion seem on the whoe to discriminate ora conversation between two persons from the situation that obtains when a person is composing, or a person is construing a text, a piece of written anguage. One of the points to be made

34 136 DSCUSSON about fictionaity in both of these circumstances is that a certain degree of fictionaity obtains in each. One sees for exampe in some books on communication theory, coege textbooks, and the rest, diagrams which show ordinary conversation going on something ike this (it's a pre-griceian version of what Grice has himsef said in a more compex way) : You have oe and Bi taking. n oe's mind (pictured in a coud) is oe's idea of Bi as requiring him to present his (oe's) utterance in a certain way, and simutaneousy oe's idea of Bi's idea of oe, which is required because oe wants to know what kind of uptake is going to be received. Now on the corresponding right-hand side in ordinary discourse you have Bi's idea of oe in making such an utterance and aso Bi's idea of oe's idea of Bi in making such an utterance. think we're famiiar enough, whether consciousy or unconsciousy, with these situations in ordinary speech to recognize these constructions, and these constructions are ceary fictiona. You cannot engage in ordinary speech without such fictiona notions of audience and one's own projected sef as an author or speaker. ~ow the situation one gets in written speech, whether it's fictiona or poetic or nonfictiona (historica), seems to me to correspond in genera to that same kind of structure. The difference is a kind of difference that Chaim Pereman, the Begian phiosopher, has stated most expicity--[it is that in a genres of writing the audience and the speaker tend to be more highy fictionaized--in a genres of writing--because for one thing, in writing one is presenting a discourse to somebody whom one doesn't see, and generay in some future time and in some unknown situation. There's a tendency therefore to generaize and make vague the audience, so the audience is aready a fictionaized projection, and in reation to that fictionaized projection one needs a fictionaized author who aso takes on a roe of a certain kind when he's engaging in a certain kind of utterance. So, do see that in the actua reaity of speech [there is a kind of roe-paying or fictionaity ridded through the act of speech itsef; and, therefore, to come to my next point (the question of intention) it seems to me that the whoe issue of whether something is meant to be taken as directy referentia or meant to be taken as fictiona or imagined, is obviousy intended. t is normay part of the authoria intention that it sha be taken that way. was very interested in the exampes of our not paying attention to authoria intention in cases of fiction. Normay, conventionay, in those situations, the authoria intention is that authoria intention sha not be taken as though the author were a biographica person, not-a5 though n~s daiy habits [were germane to the text that he presents us. The intention is to disregard biographica intention. s! )

35 DSCUSSON 137 ust a point, very briefy, about the term "convention" and the term "rues." Speech-act theory is not monoithic in its attitudes towards these two points. We've heard a Seareand-Austin view of speech-act theory here. A Strawson-Grice account of speech-act theory woud be much more intentiona; and my own feeing is that the Strawson-Grice description of speech-act theory as primariy intentiona-conventiona and not merey conventiona is a more accurate account of what reay goes on in speech. The best exampe of that sort of thing to my mind is in a book caed Convention by David Lewis, where he presents a series of situations that get conventionaized and one wonders how they got conventionaized in the first pace. For exampe if you get cut off in a two-way teephone conversation--the ine gets disconnected, and goes dead--who's the person who' hang up and try again? On what convention do we make that inference? We, obviousy some process of divination is invoved. No doubt the convention, after a number of times--particuary between two peope to whom this happens often enough--the convention gets firmy estabished and it becomes rue-governed. But at first there's a process of divination that's going on, and normay one woud guess, "we he caed me, introduced the ca the first time; ' hang up and et him try again"- that sort of thing. You may be wrong and of course after a certain pause if it were ong enough, you wi assume the other convention, "he caed me the first time and now it's my turn." t coud work either way. But ceary something very simiar is what happens in the way conventions are originated in anguage and that's the kind of point Strawson made about the importance of intention. Now finay-- don't want to go on overong but do want to come down to what seems to me [to be the centra point in the chaenge that was thrown out to me about wanting to go back to origina meaning or authoria intention. t's thrown out most ceary in the ast footnote in Professor Smith's paper: "Hirsch tends to speak not of readers but of interpreters; e.g. 'we who interpret as a vocation.' One does not suppose, however, that he woud \ish to maintain a strict distinction between the ethica obigations of readers at arge and of those readers who speciaize professionay in the pubic articuation of their interpretations. Or woud he?" And see there the probem between us because the answer is yes, woud make distinction between the professiona enterprise of interpretation and the ordinary uses to which one can independenty, or for one's own purposes, put an interpretation, ~r use an interpreta~ion. And the reason for that is simpy that f1rst of a--as th1nk reay Professor Smith granted [this

36 138 DSCUSSON hersef--that the probem is not whether we sha use interpretations independenty of the intended meaning but whether we sha abe what we are doing, particuary if we're doing it pubicy and in print; and with that can whoeheartedy agree. On the whoe, though, think what we need as a discipine (insofar as we consider ourseves a cognitive discipine, which we may not do) is to have some sort of a common goa to which we contribute something which woud stand as knowedge; and the ony common goa in interpretation that seems as yet satisfactory is the goa of recovering that which was meant to be conveyed by the text. There are other goas,.. and if one abes them propery have no quarre with them, but there does seem to me to be a distinction between the uses of iterature and the discipine of interpretation. So perhaps that's a basis for a reconciiation, as ong as that misunderstanding has been ceared up. That's a have to say. Q: This is directed to Professor Hirsch athough Professor Fish may want to comment on it. My preambe is a reference to work by a psychoanaytic anthropoogist, George Devereux, who, in taking about the probems of interpretation--specificay in a situation where the interpreter is in actua human interchange with the peope he's interpreting (an anthropoogist in the fied, or a psychoanayst in an anaytic situation with an actua patient with its give and take)--he uses the mode of transference and counter-transference--which see as anaogous to what you were taking about when you started taking about the way in which it is necessary in a given communication for both parties to have fictiona constructs of the other and of the other's constructs of [the one. Obviousy this is not the case when we're deaing with the iterature of an author who's dead and there is no human interaction [that has arisen as in the case of the anthropoogist in the fied where the tribe is reacting to his presence and he's reacting to the tribe in certain ways. But without dismissing what Devereux says want to work from it for a moment to see what we can get to. The situation perhaps that we have in interpreting iterature, that is midway between the ive interaction between human beings and the noninteraction of a physicist who's observing events at the obser[vatory for exampe, the Crab Nebua through the teescope at his side--a right, iterature's somewhere in between there. There is no interchange, no communication interchange between physicist and what he's reacting to. There is a ive interchange between the anthropoogist and the tribe he's reacting to. A right. n the case of Miton and his poems, to propose something! s ) 3

37 DSCUSSON 139 that Professor Fish might want to think about, where there is a ive author and a ive audience--at whatever remove from whatever interpositions of the media of pubication, or circuation--you have a situation coser to Devereux's exampe of the anthropoogist and the tribe, where there is give and take, there are ivey fictions [? of one and the other. But somehow what's there (and Professor Fish has given a ot of energy to thinking about this probem} carries over even after an author has died and the origina pubic is gone, cutura change has taken pace--there's some kind of force of communication, or structure of communication, that is preserved, as there is in the exampe Father Ong brought up ast night, the etter that's written by someone who dies before the etter's received. What happens--'m trying to reate this to what you said about the creation of fictions--what happens to your notion of intention--and the word "unconscious intention" was dropped, the word "betrayed meaning" coud be transated possiby to unconscious intention--what happens to your notion of intention when you get to the business of a transaction taking pace between an author and a reader? want you to put it in those terms rather than a text and a reader because texts don't engage in communications, peope do, a right? FSH: Even dead peope? Q: Even dead peope, yes. You react to Shakespeare. You react to Miton. Like it or not. We betray ourseves in the terms we use--we "ove Shakespeare." What happens to your notion of intention when you put it in this compicated give-andtake situation of the constructions of fictions about two sides of a communication? t's a big order. [Gap between tapes. HRSCH:... Suppose we discovered that the historica figure uius Caesar was as fictiona as the Pitdown Man, that actuay uius Caesar never existed, but historica evidences of various kinds were put in a ater period into the historica evidence by peope who wanted to perpetrate a hoax. What has happened in that case, once we made that discovery, to this figure, this dead historica figure? Does he suddeny cease to exist? And the answer, it seems to be, pretty obviousy, that we sti have the same evidence about the existence of that historica figure. What we now think is something ese about

38 140 DSCUSSON the reaity of his existence. The contours of his personaity, of his actions, and a the rest, remain exacty what they were. There are a number-- presume one coud find a number of exampes--of what we thought to be historica figures who ater on turned out to be not historica at a either because they'd been misinterpreted or because they didn't exist. don't see why that is in essence different in fact from what happens when we engage in ordinary conversation, because we find out from peope we're deaing with, that our ideas of them get disabused, our constructs that we make, our notions of who they are and of what kind of persons they are, get disabused. They change. don't know whether this answers your question--by giving you a coupe of exampes--but it does seem to me that in a the cases you've presented there is a continuum, a continuity, of constructs. You have a great dea more evidence of course, with somebody you're directy taking to. But don't see that you've removed the hypothetica character of another person. That exists in principe in ordinary ife just as it does in the historica past. That's a ' say about that because don't want to usurp more time. Q: wanted to ask a question about the nature of theory as it's perceived, guess, by a of you; think at one point it was said that theory was an attempt to account for something [descriptive and that it shoud not yied any kind of methodoogy. We, it seems to me that-- don't quite understand how you can say that about theory, because it seems to me that theory creates the possibiity of taking about things that might not even have been conceived before, first of a; and second of a, if you take something ike the theory of iterature there is aways that kind of rhetorica edge to it--the Week-and Warren thing--that is positing a preferentia mode of approach to a work, as opposed to some other kind of mode. Even if it ony discriminates it usuay takes a stand. 'm thinking about Poanyi's notion of a "guiding interest" or of a "persona knowedge"--that you reay can't tak in purey descriptive terms about theory. wonder if you'd agree with this or if you want to kick that around awhie. STEN~~N: We, a theory of course is not a historica description of actua particuar events. Rather it consists of the formuation of some genera principes: these may be empirica aws, in the case of physics, chemistry; these may be formuations of constitutive rues, in the case of the grammar of a anguage, in the case of speech-act rues, in the case of genre rues. And the two sides of the coin of the theory are expanation of historicay given phenomena and prediction of any phenomenon that comes aong. So that a formuation of the speech-act rues for-- s! )

39 DSCUSSON 141 the iocutionary-act rues for--asserting or making a statement, for exampe, are intended to account for what historicay peope have regarded as being statements and to predict what they wi in future regard as statements. Q: Woud you say that one of the tests of a good iterary theory (is being abe to predict what forms of iterature are on the horizon? STENMANN: No, woudn't. A fact that we have to dea with with any human institutions constituted by sets of rues is the matter of change. Languages change; this is a matter of rues changing. Perhaps some speech acts change. think there're some fundamenta ones ike assertion that do not change. So the theory is static in the sense that a grammar of a anguage is static; it pretends to say what at the moment wi be counted--maybe the rues wi change some more--it wi not predict that. Q: (Severa words missing phiosophy of anguage behind speech-act theory. The question-- refer, bibiographicay, to Water Cerf's criticism of Austin's speech-act theory ["Critica Notice of How to Do Things With WOrds," Mind, 77 in a symposium, Sand's coection, there; we, Water Cerf think has a very exceent set of criticisms of Austin's phiosophy of anguage. He raises the issue, for instance, of how paced within the Greek tradition of homoogum ethum the Austinian view of speech acts is; that is, that we sti have a very Western view of anguage--a very Western phiosophy of anguage, that man as too-maker and too-manipuator has this too which he uses through a system of acts in order to convey something and we're ed by the vocabuary of this description into a whoe set of metaphors about the use of anguage so that the question of speech-act theory reies on a certain phiosophy of anguage that tends to naturaize and mechanize the way in which we perceive the processes of anguage and their operation in the word. For instance, et's imagine trying to appy speech-act theory to, et's say, African use of anguage in poetry to account for the--it doesn't transpose very we, et's put it that way--at east have some difficuties transposing speechact theory from one cuture to another and suppose that if you broadened it enough so as to say, we, there's aways a hearer, there's aways a sender, you can transpose then your ideoogica furniture on top of a phiosophy of anguage; but it seems to me that that issue can be raised; and another issue is: quibbe with the word "act"-- woud prefer say, for instance, the word "event." You know? But agree with Don

40 142 DSCUSSON Hirsch that the reveation of iocutionary force is a very important thing and it's important to know authoria intention; but woud wonder whether it doesn't change the game a itte bit to speak of the event that takes pace between text and reader, rather than the speech act--which aways takes us back, then, to the movement of the speaker in creating the text and draws us away from the hermeneutica probem of interpreting the text itsef, so that--we, that's overstating it, but at east woud say that the term "event" is a superior term to "act" and incudes it. FSH: The second part of the question strayed into things that 've aways been interested in, but 'm going to forbear returning to that interest because agree so much with the impications of the objection--or at east the observation--made in the first part; and that is simpy that speech-act theory is an ideoogy; and reay think it's an ideoogy--it's an ideoogy to which we might in some moods even want to put the word "bourgeoise." That is, when Austin says on page ten of How t.o Do Things With Words, "a man's word is his bond," you can see-rhat ega jurisdictiona contract moraity camping down with a vengeance. And it does. That is, if you ook at speech-act theory carefuy, and especiay if you ook at the atest taxonomy of speech acts which is one that's been pubished by Seare in the Minnesota series ["A Taxonomy of ocutionary Acts"-,_you' find out.that a of the speech acts that he distinguishes as major ones have to do with things that are actionabe in aw, and are ways of enforcing obigation. Now for instance, when on page ten of [How to Do Things With Words Austin makes his famous, "we ook, if a man's word isn't his bond, then we're at the mercy of somebody who says with moraistic piety, 'my tongue uttered but my heart forswore'"--a right, that's very forcefu there. But of course mysef have spent most of my ife reading and writing about ohn Hiton, whose theory of everything was exacty that one; that is, he had a revoutionary ideoogy, more revoutionary think than those Christian humanists who have tried to appropriate him sometimes admit, and the basis of that revoutionary ideoogy was intentiona in the sense that at east the Austin Seare ine of speech-act theory woud not aow. That is, Miton, ike others, competey woud side off from the form of any utterance to the spirit or intention behind it. So that everything changed depending on the perspective from which something was uttered. Now it's quite cear that since Miton, for instance, aways took the tack--when opposed by critics saying "we, ook: you've done this, and criticized me ~or doing the same thing"--he woud aways come back and say but did it, and am authorized in doing it and the conventiona forms upon which you are reying are so much dross." Now 'm s!!

41 DSCUSSON 143 not going to argue for either one of these views, but merey to point out with this genteman here that they are views, and ideoogies, which impose upon us certain, as-natura--that is, as-not-made--certain ways of ooking at the word. think speech-act theory is an ideoogy. n a briiant artice in [The New York Review of Books of Apri 1972 Seare both criticized and heraded the end of the Chomskian revoution when he pointed out, finay, that in Chomsky there is reay no way, as he said, to get from the syntax to the semantics or from the physics to the semantics. And he pointed out that for Chomsky, in some curious way, man was a syntactica anima. We, woud now ike to extend that generous criticism and say that for Seare and Austin, at east, man is a ega anima. And that one view is as partia and ideoogica as the other. Q: My question reates to what Mr. Fish has just said, but it's directed to Mr. Hirsch. have a feeing that might know the answer to the question but want to be proved wrong. Your grudging approva of speech-act theory was given in terms of its effectiveness as an antidote to two "dispensations": not atogether surprisingy, "the od" and "the new." was wondering if you woud te me what in your understanding is the new dispensation ed by Derrida. HRSCH: We, of course this is my interpretation of Derrida. As understand Derrida, there is no end point of interpretation--there's no object to be interpreted. don't caim to be even a carefu reader of a Derrida's works, but the one thing that seems to me to be a recurrent theme is that every end point is just a sign for another sign; and that being so, what coud possiby be a common goa in the enterprise of interpretation? That's why thought that--since Austin has one sort of corrective force for the textua-sufficiency theory-,- and--we need a different sort of corrective force for the idea that there is no object or no meaning there for us commony to interpret. FSH: Peope who have read what Don Hirsch and write might be surprised to find us in agreement, and don't want anyone to go away disappointed. t seems to me that Hirsch's agreeing with me in no way--that is, don't take what he says to be the resut of having agreed with me. But his notion that--what was it--''continuing construction"--it seems to me that if you take the notion of continuing construction seriousy you're going to end up in the Derrida position. The difference between Mrs. Smith and mysef and Hirsch on one hand is the difference which finds its way into Seare's speech-act theory

42 44 DSCUSSON of reference. There are two ways to have a theory of reference, and curiousy enough in his book Seare has them both. One is a theory of reference in which successfu reference occurs when you identify something for a speaker in a situation. The other is a theory of reference [in which successfu reference occurs when you identify something. Now on that distinction hangs everything, because if you have a situtation-centered theory of reference, then the existence outside of the situation of the thing to which you are referring is bypassed--the question of that existence is bypassed. On the other hand if you think of referring as in a sense picking out an unique and distinct object, then of course you're putting an extraordinary burden on your anguage. Not ony are you putting an extraordinary burden on your anguage, but you're putting an extraordinary burden on the history of science, and on those peope who were unfortunate enough to ive, et's say, in the days when the Ptoemaic astronomy was beieved in. once asked Seare on this point, ''do you then mean that those 'peop'e weren't referring," and he said nyes," as he necessariy had to, given that position. Of course Seare invokes for fictiona discourse the notion of shared pretence. What of course want to do is expand the notion of shared pretence much in the way that Mr. Hirsch suggests, to a discourse, but if we so do, and admit that our references are aways to things which have been identified in the context of situations and depend on beiefs and opinions hed by speakers rather than reaities, it woud aso seem that we woud have to give up as a possibiity the determination of origina meaning; athough sti think it is the attempt aways to reach that impossibe goa that keeps us going. So that we woud in fact--can keep on taking about this origina meaning and aways do it in the confidence that we' never have to stop. Q: May speak two words in agreement with you and to the theorem--about your understanding of the Derrida position? think you aso suggested, contrary to what you were saying, that you too thought that was the position. t seems to me if one directed one's attention to Derrida texts ike the end of his /essay on Nietzsche, or the chapter in the Grammatoogy caed "The Exorbitant Question of Method," or his interview that has been transated as "Positions" in Diacritics [2 (Winter 1974), pp one woud find that in fact Derrida is not arguing- this isn't some sentiment one finds a over the pace--because we in America are used to thinking of the French 1n a sort of scapegoat fashion in a certain way. But don't think Derrida is suggesting that there is no object, just as much as he is not suggesting that there is no subject. think he is saying!

43 DSCUSSON 145 something ike Heidegger and Nietzsche which simpy is examining a very famiiar conventiona, comfortabe beief, and, ifting the id, one sees in fact a can of worms--which is that the subject is constituted--the subject is aways provisiona--it is constituted by its presence as much as by its absence, and the end--the object--is aways provisiona--one must in fact find an object, have an object, one must endeavor. But we know this practicay, that the object too--in fact, for the act of interpretation not to become atogether fascistic--the object too is provisiona; the object, as much as the subject, is constituted by its presence as much as by its absence; and this aows us to ive and breathe, as much as it aows us not to forget the fact that breathing is not ony being aive but aso deferring the fact of our death. don't want to speak any more, but it seems to me that the presence ange of Derrida is often simpy occuded because we want to beieve that he is in a position of absence when he's not. Q: First to Professor Smith: just want to carify the act of identifying a text as a text that we wi not interpret intentionay, and just want to make it cear in my own sense whether it is--this speaks to the bottom of page six of her essay--whether it is a categorica act, et us say--that we impose a category upon the text, a priori, which then enabes us to suspend an interest in intention; or whether it's a particuar act, an act that is sensitive to the particuar caims of this text on our interest in intention. SMTH: spoke--if get the direction of your question--what 'd have to say is that certain texts seem to end themseves to that sort of categorica treatment; so there's no way in which the category can be arrived at a priori; in other words, don't see any specific features in the text (aside from, you know, announcements of how it wants to be taken) that necessariy constrains us. However, there are certain texts, as say, that seem to offer themseves for the suspension that have spoken of. Does that answer the question? Q: These are certain particuar texts which we recognize uniquey as having this caim? SMTH: We, it woud work differenty for different peope, different times of your ife, certainy different ages. A text which offers itsef in the seventeenth century as a statement of someone's reigious beiefs or the construction of the word may offer itsef in the twentieth century for a very different sort of reading. Q: And 'd ike to ask Professor Hirsch about the very ast page of his essay, think. He points [out that the very

44 146 DSCUSSON generaity of the speech-act theory may pose a probem because the very particuar cases in dispute, or the particuar options in dispute, might be sanctioned by the genera theory. guess want to ask first of a whether or not the generaity of the theory might not hep us carify the particuar options that are open to us in particuar interpretative cases, choosing among particuar iocutionary acts, for exampe, that might be operative in a given text; and aso want to ask if he's not attracted to the speech-act theory as a generic theory of speech, one that might have some reation to his notion of "extrinsic genre" and, finay, "intrinsic genre." HRSCH: The ans\er is yes. think that the kinds of iocutionary forces that are deat with in speech-act theory correspond to what we have normay caed in iterary interpretation, "genres." But 'm aso Crocean on the matter of genres and think that there aren't any cear boundary ines between them- which means that Croce is on the side, say, of Strawson and Grice in speech-act theory--it's the same sort of a correation. 've forgotten your first question. Q: Whether or not the genera theory might hep identify particuar probems--not sove them, but just identify them. HRSCH: Oh yes, think so. t can identify probems, that's right; but don't think a genera theory can show you how to make actua choices when you have empirica questions that need to be answered. think that's done on evidence.

45 KENNETH BURKE Words as Deeds. L. Austin. How to Do Things With Words: The Wiiam ames Lectures Deivered at Harvard university in nd ed. Ed.. o. Urmson an~mar1na Sb1sa. Cambridge, Mass.: Harvard Univ. Press, Pp. viii Paperback, $2.95. Theories of anguage invove two kinds of specuation that are quite different yet by no means mutuay excusive. One might be caed "scientistic" because it gravitates about anguage as a mode of knowedge; the other "dramatistic" because it approaches anguage in terms of action. The distinction is presented in the opening remarks of David Burne's An nquiry Concerning Human Understanding: "Mora phiosophy, or the sc1ence of human nature, may be treated after two different manners. for the one considers man chiefy as born for action... the other species of phiosophers consider man in the ight of a reasonabe rather than an active being, and endeavour to form his understanding." The distinction probaby gets its utimate grounding in the shifts of approach between "ontoogy" and "epistemoogy." But woud confine the terms to a terministic emphasis. This summary formuation, on the first page of the ast chapter of Austin's How to Do Things With Words, is in itsef enough to indicate why woud cass the book as "dramatistic" in its approach: "The tota speech act in the tota speech situation is the ony actua phenomenon which, in the ast resort, we are engaged in euc1dat1ng." True, his theory of words as deeds is itsef a contribution to knowedge, as per his opening sentence: "the ony merit shoud ike to caim for it is that of being true, at east in parts." But the systematic choice of a dramatistic approach to his subject impies that the pursuit of knowedge in such matters is best guided roundabout via specuat1ons about anguage as a mode of action. The book being notes for ectures rather than a definitive text, it is presented in the spirit of a "work in progress," even to the extent that it undergoes en route a "fresh start" (p. 121), with corresponding "sea-change" (p. 150). At the [Centrum, 3:2 (Fa 1975),

46 148 KENNETH BURKE time when Austin began his specuations (which he dates from 1939), had aready been consideraby infuenced by the dramatisticay santed essay, Bronisaw Mainowski's "The Probem of Meaning in Primitive Languages," pubished as a suppement in Ogden and Richards' basic voume, The Meaning of Meaning. had got so deepy interested in it, began buiding with it in sympathetic ways of my own. Austin begins with two terms, ony one of which ("performative") is dramatistic. The other ("constative") is ceary designed to be on the "scientistic" (T-F) sope. He finay abandoned his "beief in the dichotomy of performatives and constatives in favour of more genera famiies of reated and overapping speech acts" (p. 150). But my terministic ine-up had made it impossibe for me, from the start, to find it as cear as he had seemed to do. Yet the very views had on the subject aso get in my way when try to see the issue as think he wants us to, in his cuminating chapter. His way of conducting us through the transformations from his initia dramatistic anecdote (presenting "performatives" in terms of such "speech acts" as vows and decarations) to his grand finae surveying "casses of utterance, cassified according to their iocutionary force" and "quite enough to pay Od Harry with two fetishes which admit to an incination to pay Od Harry with, viz. (1) the true/fase fetish, (2) the vaue/fact fetish," is itsef a performance of appeaing virtuosity. t is one ong ingeniousy sef-imposed aporia. vhen trying to decide on a name for his "performa ti ves," he says: One technica term that comes nearest to what we need is perhaps "operative," as it is used stricty by awyers in referring to that part, i.e. those causes, of an instrument which serves to effect the transaction (conveyance or what not) which is its main object, whereas the rest of the document merey "recites" the circumstances in which the transaction is to be effected. [p. 7 particuary ike that kind of anecdote because it ends itsef so we to a drarnatistic extension. That is to say: f one thinks of such a ega transaction as iteray an "act" (in the sense of "speech act"), then the recita of the circumstances" in which the transaction is to be effected defines the "scene" in which it is to be "active," or "enacted." When the pattern is appied to such "enactments" as the U. S. Constitution, the reationship between "scene" and "act" wi be found to revea what Austin might ca an "infeicity" )! ) ) )

47 WORDS AS DEEDS 149 of this sort: The Constitution, as a ega "instrument," was "enacted" in a "scene," or situation, quite different from the circumstances in which the Constitution, with its amendments and often innovative judicia interpretations, is operative as an enactment now. n this respect can the same utterances, as a speech act, have the same "meaning" they had when first decared? Austin's term, "constative," take it, woud cover the description of the circumstances prevaiing at the time of the origina enactment or their modified nature now. This is a drarnatistic consideration intend to take up esewhere, in connection with the terms "meaning" and "significance" ("text" and "context"), when reviewing The Aims of nterpretation, by E. D. Hirsch, r. There are some interesting simiarities and differences between Mainowski's and Austin's informative (better, "preformative"?) anecdotes in conformity with which to form their anaytic nomencatures. Both authors choose cases which ceary invove what Mainowski cas "context of situation"; on page 140, Austin refers to "circumstances of situation," esewhere to "circumstances," "situation," or "context." Both stress the conventiona aspects of the verba behavior they are starting from, and of the human reations impicated in those acts. Mainowski's two anecdotes are much the more rudimentary, being concerned with iiterate tribesmen in group activities: Take for instance anguage spoken by a group of natives engaged in one of their fundamenta pursuits in search of subsistence--hunting, fishing, tiing the soi; or ese in one of those activities, in which a savage tribe express some essentiay human forms of energy- war, pay or sport, ceremonia performance or artistic dispay such as dancing or singing. The actors in any such scene are a foowing a purposefu activity, are a set on a definite aim; they a have to act in a concerted manner according to certain rues estabished by custom and tradition. n this, Speech is the necessary means of communion; it is the one indispensabe instrument for creating the ties of the moment without which unified socia action is impossibe. Having described in some detai the operations of some fishermen, Mainowski sums up thus: A the anguage used during such a pursuit is fu of technica terms, short references to surroundings, rapid indications of change--a

48 150 KENNETH BURKE based on customary types of behaviour, we-known to the participants from persona experience. Each utterance is essentiay bound up with the context of situation and with the aim of the pursuit, whether it be t:1e short indications about the movements of the quarry, or the expression of feeing and passion inexoraby bound up with behaviour, or words of command, or correations of action. The structure of a this inguistic materia is inextricaby mixed up with, and dependent upon, the course of the activity in which the utterances are embedded. The vocabuary, the meaning of the particuar words used in their characteristic technicaity is not ess subordinate to action. For technica anguage, in matters of practica pursuit, acquires its meaning ony through persona participation in this type of pursuit. t has to be earned, not through refection but through action. [ take this ast sentence to be deaing with what have treated in terms of my scientistic-dramatistic distinction, though obviousy, if there weren't the makings of many T-F distinctions impicit in the tribesmen's "technica terms" and the ike, there coudn't have been a iving in that way of ife... The consideration of inguistic uses associated with any practica pursuit, eads us to the concusion that anguage in its primitive forms ought to be regarded and studied against the background of human activities and as a mode of human behaviour in practica matters... The manner in which am using it writing these words, the manner in which the author of a book, or a papyr~s or a hewn inscription has to use it, is a very far-fetched and derivative function of anguage. Mainowski then turns to a different anecdote that brings out a different aspect of anguage--and his name for this is "phatic communion," not to be confused with what Austin cas a "phatic act." Here "we turn our attention to free narrative or to the use of anguage in pure socia intercourse; when the object of tak is not to achieve some aim, but the exchange of words amost as an end in itsef." Verbaizing as so denominated invoves "a type of speech in which ties of union are created by a mere exchange of words... The whoe situation consists in what happens inguisticay. Each utterance is an act serving the direct aim of binding hearer to speaker by a tie of some socia sentiment or other." These two anecdotes struck me as amost cassic in the simpicity and suggestiveness of their reevance to the subject. And it made good sense to me that the strategic instrument!

49 WORDS AS DEEDS in so major an activity as the gathering of food (a cooperative function that can equay we serve to the ends of competition) shoud be, we might say, enjoyed for its own sake; for the typicay symbo-using anima might be expected to exercise its prowess as the typicay symbo-using anima, as fish take to swimming and birds to fying. And in the offing there were the makings of a third anecdote, suggested by ~-1ainowski 's re.ference to contrasting written texts which woud be at east comparativey "sef-contained and sef-expanatory." When first read his essay, had been taking notes on various socia and iterary "devices," and aong those ines had written a book around the contrast between "yea-saying" and "nay-saying" ("acceptance" and "rejection," with intermediate twists). Accordingy, the first essay in my coection, The Phiosophy of Literary Form (pubished in 1941, composed of things done in the Thirties), begins thus: Let us suppose that ask you: "What did the man say?" And that you answer: "He said 'yes.'" You sti do not know what the man said. You woud not know uness you knew more about the situation, and about the remarks that preceded his answer. Critica and imaginative works are answers to questions posed by the situation in which they arose. They are not merey answers, they are strategic answers, styized answers. For there S a difference in stye or strategy, if one says "yes" in tonaities that impy "thank God" or in tonaities that impy "aas!" So shoud propose an initia working distinction between "strategies" and "situations," whereby we think of iterature.. as the adopting of various strategies for the encompassing of situations. These strategies size up the situations, name their structure and outstanding ingredients, and name them in a way that contains an attitude towards them. This point of view does not, by any means, vow us to persona or historica 151

50 152 KENNETH BURKE subjectivism. The situations are rea; the strategies for handing them have pubic content; and in so far as situations overap from individua to individua, or from one historica period to another, the strategies possess universa reevance. Situations do overap, if ony because men now have the same neura and muscuar structure as men who have eft their records from past ages. We and they are in much the same bioogica situation. Furthermore, even the concrete detais of socia texture have a great measure of overap. And the nature of the human mind itsef, with the function of abstraction rooted in the nature of anguage, aso provides us with "eves of generaization" (to empoy Korzybski's term) by which situations greaty different in their particuarities may be fet to beong in the same cass (to have a common substance or essence). Consider a proverb, for instance. Think of the endess variety of situations, distinct in their particuarities, which this proverb may "size up," or attitudinay name. But a written text is ike the parry to a thrust. The thrust comes from the "context of situation" out of which the text arose. And whereas the thrusts of history keep undergoing changes of position and direction, the written text acks corresponding immediate piancy. The written text makes its particuar parry statuesquey permanent, whereas the context of situation into which it has survived may have put forth a thrust that cas for a quite different parry. So waver a bit concerning the extent to which we can recover the origina "sef-contained" meaning of a text that arose from a "context of situation" greaty different from ours. But a tod, we have (1) words as iiterate instruments in a cooperative act, itsef motivated by a "context of situation"; (2) words as the exercise of iiterate human prowess with words in a "context of situation"; (3) words as a itera text that is a context for its own words (with a somewhat probematica reation to "context of situation" in Mainowski's sense'. n (3) the exercising of (2) "phatic communion" (in Mainowski's sense) can be deveoped internay to the extent where the mere socia reaxations of gossip can be transformed into such profound exercising of our symboic prowess as the

51 WORDS AS DEEDS 153 "tragic peasure" discussed in Aristote's Poetics. For just as experts in the soving of crossword puzzes wi demand the most exacting chaenges, so the fuest exercising of our aptitudes in the way of "symboicity" wi have a gratification in and for itsef. But besides the fact that both Hainowski's and Austin's ("preformative") anecdotes concerning words as modes of action invove a great stress upon the roe of circumstances in the meaning of an act, we confront important dramatistic considerations to do with the reation between acts and the attitudes of agents. And as read Austin, he has offered some remarkabe contributions to that aspect of the subject. As see the issue, he has introduced a weath of accurate discriminations into an area which is impicity ambivaent. To point up the ambivaence in the concept of "attitudes," in my section on "'ncipient' and 'Deayed' Action" (Grammar of Motives) contrasted two exceent works buit around the term "attitudes":. A. Richards Principes of Literary Criticism and George Herbert aead' s Phiosophy of tie Act. Reduced to its buntest the situation woud be thus=---- f fet an attitude of sympathy towards someone in troube, may sincerey express my sentiment. Or my attitude of sympathy might ead to an overt act of seeking to modify the conditions that are the cause of the distress. A phianthropic intermediate stage might be, say, as with a coin tossed to a eper, whereby coud give hep, but at a distance. Quite as the acts of agents are grounded in scenes, so those acts are associated with corresponding attitudes on the part of the agents, attitudes that sometimes accompany the acts, sometimes ead 1nto the acts, and sometimes serve as surrogates for the acr-tas you might write a poem voicingsympathy for the unjusty treated, and fee that you had done enough--a turn that coud aso be characterized as "dancing an attitude" of sympathy--and you might do so persuasive a job that, if ushers passed pates, a tidy sum coud be coected for the cause). As a kind of "work in progress" on my part, want to see what might be done by a view of "iocutionary force" as a synonym for "attitude," and to ask what quaifications might be needed for treating from this point of view the five casses of "iocutionary acts" isted in Austin's ast chapter (actuay two further "-ives" are mentioned there). But first et's review some of the main steps en route.

52 154 KENNETH BURKE A notaby "preformative" aspect of Austin's anecdote is its contrast with the coective iiteracy of the speech acts Hainowski chooses as his way-in. And above a, there is the whoy different approach inherent in the ingenious roe that is payed by an intrinsicay reated aspect of his anaysis: his cautiousy quaified use of the "first person singua:r present active indicative form" as a grammatica test for traces of his "performative" verbs. (There are references on pages 56, 62, 63, 64, 67, 150.) n particuar note the shrewd test of "asymmetry" he discusses on pages Here he expains why a sort of expicit expansion is needed, to guard against taking a performatve utterance "in a non-performative.way." Austin's sef-admonitions of this sort were argey responsibe for his decision that the constative-performative dichotomy had to be abandoned. For such expicit expansion indicated that, by the same token, a performat1ve might be impicit in what woud otherwise ook ike a "non-performative." The 1ssue is ceary iustrated in the various references to a "bu in the fied" (pp. 33, 59, 62, 74). n out-and-out "dramatistic" terms, the point might be summed up thus: mpicit in what ooks ike a constative "description" there may be Si.ch a.kind of speech act as, if expicity expanded, woud amount to saying: "On the basis of ths description you had better do, or not do, such-and-such." That is to say, the mere woud-be description, or report, woud not just be saying what the situation is. The utterance woud aso impy an attitude (such as a warning) towards the situation. Thus, utterances that on their face have the form of a constative ocution might aso have the iocutionary force of a performative attitude (which might even eventuate in an act having perocutionary consequences). am here tentativey anticipating, on the assumption that these pages are being written for readers aready quite famiiar with Austin's book- and this seemed to be a handy pace for indicating my direction. n any case, the ambiguity of "constative" and "performative" is ceary stated on page 33: "'There is a bu in the fied' may or may not be a warning, for might just be describing the scenery." Here, in advance of the reevant passages in Austin's text, are the equations am tentativey working with: A ocution's roe as a warning makes it a performative; this performative function is synonymous with its iocutionary force--and with regard to the reation between speech acts and attitudes, woud equate an utterance's iocutionary force

53 WORDS AS DEEDS with its roe in expicity or impicity symboizing an attitude. n abandoning his two-term start, Austin (p. 95) works with "three rough distinctions between the phonetic act, the phatic act, and rhetic act," which reate t;hus: The phonetic act is merey the act of uttering certain noises. The phatic act is the uttering of certain vocabes or words, i.e. noises of certain types, beonging to and as beonging to, a certain vocabuary, conforming to and as conforming to a certain grammar. The rhetic act is the performance of an act of using those vocabes with a certain more-or-ess definite sense and reference. Two pages ater: The same "pheme, e.g. sentence" (that is, the same "phatic act") ".may be used on different occasions of utterance with a different sense or reference, and so be a different rheme" (that is, a different "rhetic" act). 'd take this triad to be summing up the fact that the mere sounds of words, in their nature as mere sounds, are in the ream of nonverba motion. Such sounds are in the ream of a speech act when interpreted as words, i.e. "phatic." nsofar as such words function with reference to contexts (either a context of situation or the text itsef as a context) they become "rhetic." And insofar as the same sentence can have different impications in different contexts, its nature as "phatic" takes on different "rhetic" identities. But though Austin does not reject this intermediary ine-up of terms, he moves on (p. 98) because "they do not so far throw any ight at a on our probem of the constative as opposed to the performative utterance." t's a good basic ine-up, though not the best for his purposes. And particuary want to end by considering more cosey the reation between the mere "noises" of words (as "phones") and their nature as "phemes," that is, when their function in the proper symbo-system transforms them into the possibiity of speech acts as "rhemes." For the mere noise of words (phones) is reated to their roe as speech acts (phatic acts, rhetic acts) as the ream of nonsymboic motion is to the ream of symboic action. And this reviewer's "work in progress" shoud end on matters of that sort, in ine with concerns that Austin ceary touches upon in Chapters V and X, as per for instance the footnote on page 112: "f we suppose the minimum physica act to be movement of the body.. " etc. A tod, a dramatistic 155

54 156 KENNETH BURKE study of Austin's dramatistic book fas into three phases: (1) The reation between speech acts and the circumstances of an act; (2) the roe of iocutionary force (equatabe with "attitude"?) in the speech act; (3) the grounding of a speech acts ("symboic action") in the ream of wordess "nonsymboic motion." By no means do commit mysef to a promise that the issues wi be setted. But do dare hope that the confronting of a our quandaries may contribute towards a better coordination of our future efforts. There woud be good grounds for starting my discussion precisey where we now are. And tried to do so. But obviousy too much woud be ost. For though haven't done justice to the engaging virtuosity of Austin's professiona scrupes in working out the ingeniousy anayzed impications of his "first person singuar present active indicative" formua, trust that have at east paid tribute to his methodoogicay grammatica excursions around such prior approaches, as with German metaphysicians who buid from an Absoute Ego (as diaecticay paired with a Non-Ego)--with perhaps Descartes' cotito, ergo sum utterance or pheme or ocution as a hafway stage for do incine to judge that Descartes' cogitatio-extensio poarity was impicity much coser to a purey emprca dstinction between "symboic action" and "nonsymboic motion" than his terms' metaphysica trappings permitted to be expicit). But to proceed. Austin next hits upon the triad that. he wi sette for: "three kinds of act--the ocutionary, the iocutionary, and the perocutionary" (p. 103). "Our interest in these ectures is essentiay to fasten on the second, the iocutionary act and contrast it with the other two." So et's copy out, as our basic text, his summarizing presentation of the three: (E.1) Act (A) or Locution He said to me "Shoot her!" meaning by "shoot" shoot and referring by "her" to her. Act (B) or ocution He urged (or advised, ordered, &c~) me to shoot her.

55 (E. WORDS AS DEEDS Act (C. a) or Perocution Act (C. b) 2) Act Act Act Act He persuaded me to shoot her. He got me to (or made me, &c.) shoot her. (A) or Locution He said to me, "You can't do that. (B) or ocution He protested against my doing it. (c. a) or Perocution He pued me up, checked me. (c. b) He stopped me, my senses, &c. [pp he brought me to He annoyed me. The quiety joting thing about that aignment is in the fact that the verb "persuaded" is not in the same 11 cass as the verbs "urged,.. advised," and 11 ordered. 11 This is a radica matter. Traditionay, "persuasion" is the term of terms in texts on rhetoric. (ncidentay, the word 11 rhetoric 11 does not appear once in this book. Though we might find a faint etymoogica trace of it in Austin's term 11 rhetic," his "rhemes" differ as much from 11 rhetoric 11 as his 11 phatic" act differs from Hainowski's 11 phai:ic communion." Perhaps in the ight of traditiona contests between "rhetoricians" and "grammarians," Austin's basic investment in the "first person singuar present active indicative" formua automaticay turns his attention away from the traditiona formuations of rhetorica theory.) Except for a brief section on "inartificia" proofs (which woud incude such "nonconventiona" means of persuasion as torture), Aristote's Rhetoric is whoy concerned with the art of persuasion by the resources of diction. n this respect the verb "persuade" woud definitey fa under the head of cass B (the "iocutionary" act), whereas Austin puts it 157

56 158 KENNETH BURKE in a cass of "perocutionary" acts that invove "consequences" outside the ream of speech acts, as the act of murder woud obviousy be, even if the murder had been a consequence of purey verba persuasion. Austin's reasons for his procedure here are ceary stated: We must systematicay be prepared to distinguish between "the act of doing x," i.e. achieving x, and "the act of attempting to do~ " n the case of iocutions we must be ready to draw the necessary distinction, not noticed by ordinary anguage except in exceptiona cases, between (a) the act of attempting or purporting (or affecting or professing or caiming or setting up or setting out) to perform a certain iocutionary act, and (b) the act of successfuy achieving or consummating or bringing off such an act. [pp n the ight of Austin's cear distinction, we reaize that a whoy accurate tite for Aristote's text woud be something ike "The Art of Trying to be as Persuasive as One can." To this end it surveys various conventiona "topics" which the orator can expoit in his efforts to convince an audience by verba means, though the acts they do as a consequence of expert "deiberative" oratory may invove matters of "ways and means" as ceary in the nonverba (or more-than-verba) ream as a coonizing expedition or a miitary operation. The traditiona position woud amount to something ike this: "Here are the ways to be persuasive. Sometimes you may use them successfuy, sometimes not." And in that spirit, going back to the opening anecdote of Austin's book, we might say that, just as a marriage vow may turn out "happiy" or "unhappiy," so an attempt to be persuasive may turn out "happiy" (if successfu), "unhappiy" (if not). "And in the course of using speech acts to be persuasive, to the best of my abiity woud so design my speech as to form, in the audience, the kind of ATTTUDES that woud ead them to adopt the poicies or judgments am advocating." So now we're ready for the "famiies" in the ast chapter. We are by no means importing the term ATTTUDE into Austin's

57 WORDS AS DEEDS 159 text. See, for instance, the definition of "Behabitives" (p. 160): "Behabitives incude the notion of reaction to other peope's behaviour and fortunes and of attitudes and expressions of attitudes to someone ese's past conduct or imminent conduct." On page 152 he says they are "a very misceaneous group, and have to do with attitudes and socia behaviour." Page 155: "to bame is a verdictive, but 1n another sense it is to adopt an attitude towards a person and is thus a behabitive." Page 157: "such exercitives as ' chaenge,' ' protest,' ' approve,' are cosey connected with behabitives. Chaenging, protesting, approving, commending, and recommending, may be the taking up of an attitude or the performing of an act." With regard to Permissives (p. 159): "Behabitives commit us to ike conduct. Thus if bame, adopt an attitude to someone ese's past conduct, but can commit mysef ony to avoiding ike conduct." To sum up, we may say that the verdictive is an exercise of judgment, the exercitive is an assertion of infuence or exercising of pmver, the commissive is an assuming of an obigation or decaring of an intention, the behabitive is the adopting of an attitude, and the expositive is the carifying of reasons, arguments, and communications. [p. 163 When Austin says, on page 152, "t coud we be said that a aspects are present in a my cases," submit that the essentia eement they have in common is their common nature as speech acts in genera. nsofar as a speech act has "meaning," it invoves an attitude of some sort--that is why Austin compains that his "Behab1tives" are too misceaneous. The "rhetic" act of persuasion is designed to so perform an "iocutionary" act that "a certain effect is achieved" (p. 116). This "effect must be achieved on the audience if the iocutionary act is to be carried out." can understand his major triad but in these terms: f consider the utterance simpy as an utterance, it is a "ocution." f consider it in its attitudina nature (its "force" in expressing or shaping an attitude) it is an "iocution." f consider it in terms of its consequences (that is, if note how the attitude that it ernbod1es as a purey verba effect eads to an action in accordance with the attitude it aroused) am considering it as a "perocution." Presumaby if it aroused an unintended effect, it woud sti be a "peroctuion," but maybe for me an "unhappy" one. But if my oration produces three different responses in

58 160 KENNETH BURKE an audience, does Austin's use of "consequences" "perocutionary" act mean that my one rhetorica acting the oration is three perocntionary acts? as a test of the act in speech- n any case, must "dramatisticay" take it that a speech acts either as uttered or as responded to, are intrinsicay attitudina. Thus Austin's five famiies (in the chapter he aso refers to "permissives" and "descriptives," the atter of which woud presumaby run the same risks as in the case of "constatives") woud be inter-reated in ways whereby the attitudina Behabitives woud be the ancestors of the ot, with differentiations emerging sowy. ('d go aong with Austin's notion that anguage probaby deveoped out of hoistic utterances.) n any case, since the iocutionary act (with its inherent attitudinizing as its "iocutionary force") encounters a boundary beyond which a speech act cannot pass without ceasing to be a mere speech act, Austin's penetrating remarks on the subject of "consequences" bring us to our third consideration. n Austin we confront to perfection the step from the Metaphysica "" to the Grammatica "." But we confront the possibe need of a further step, aong ines that Austin ceary touches upon. The very nature of the reationship between a speech act and the circumstances (or "context of situation") in which any such act takes pace forces us (in utimatey dramatistic terms) to ask how a such "conventions" (such verba or "symboic" action) must reate to the nonconventiona, nonverba, nonsymboic ground or context that, by the very "diaectica" nature of the case, must be there somehow. For don't the diaectica conventions of speech itsef force us to recognize that our aptitude with words emerges from a ream of wordessness? Such nonverba, nonconventiona, nonsymboic ground woud be a ream of sheer MOTON in the sense that, if a verbaizingy active animas were erased from the word (as they 1n a rke1hood some day wi be) despite the absence of such speech acts there ~aud sti be the motions of the winds and tides, of the earth's revoutions about the sun, the processes of geoogy, astronomic unfodings in genera, etc., a going their way without benefit of verba cergy here on earth. And inasmuch as anguages are conventiona symbo-systems, by sheer diaectica necessity 'd propose to ca such a ream devoid of speech acts a ream of nonsymboic motion. Whatever the uncertainties of the metaphysica or grammatica "" might be, such an out-and-out dramatistic statement of the case woud give us a purey empirica principe of

59 [ WORDS AS DEEDS 161 individuation to buid from; namey: the human body in physioogica motion, each with the centraity of its particuar nervous system whereby, however its peasures and pains might resembe the peasures and pains of other such bodies, it immediatey experiences ony its own. Hence there woud be a drast1c quaitative difference between a state wherein it rather than some other physioogica organism immediatey experienced some particuar peasure or pain. And whatever may be the continuity between such organisms and the environment of which they are a part, the centraity of each one such particuar organism's nervous system woud be born and woud die as that individua. ts "" in the socia sense woud be deveoped insofar as the organism somehow deveoped the abiity and need to earn conventiona symbosystems which, being in their very essence modes of "socia behaviour," endowed those particuar wordess ("infant") organisms with the ingenious capacity to deveop speech acts that coud discuss speech acts. And so, in sum, we have the ream of symboic action that is quaitativey different from the ream of nonsymboic motion. Though Mainowski's and Austin's anecdotes have a great dea in common (since both ay great emphasis upon the "conventiona" nature of verba utterance) had to admit that, since had approached the subject from the standpoint of!1ainowski' s anaysis and of my buiding on it (first in Phiosophy of Literary Form and then more definitivey in Grammar of Motives) had some troube with Austin's way-in. For the gen1us of his own term, "speech act," was guiding his specuations from the start, hence demanding that the difference between "constatives" and "performatives" dissove 1nto their common identity as speech acts, whereby utterances that expicity ooked ike constatives coud impicity have the force of performatives. Then have tried to show why, since a speech acts invove attitudes, Austin's anaysis of "behabitives" as attitudes coud be extended to his other reated "famiies" of "-ives." argued that the function of the attitudina weighting or santing they a performed (though in varying ways) coud be equated with what Austin cas their "iocutionary force." might have borrowed here an expression that Bentham uses with regard to what (in his Tabe of the Springs of Action) he cas "censoria appeat1ves,~h1ch "havethe force but not the form of an argument." We were here confronting the gradua encroachment of rhetorica concerns ("persuasion") upon an inquiry that was buit about a grammatica device (Austin's ingenious heuristic use of the "first person singuar present active indicative"

60 162 KENNETH BURKE formua). We must now consider why this stage in our anaysis invoves a basic methodoogica concern with the probem of "drawing a ine" between the reams of "motion" and "action." The choice of these particuar terms is not important, but the distinction is. n Austin's text the issue comes up with regard to his distinction between the "iocutionary" and "perocutionary." The probem comes to a head in Lecture X, from which ' quote severa representative passages: We have then to draw the ine between an action we do (here an iocution) and its consequences. Now in genera, and if the action is not one of saying something but a non-conventiona "physica" action, this is an intricate matter... We can, or may ike to think we can, cass, by stages, more and more of what is initiay and ordinariy incuded or possiby might be incuded under the name given to "our act" itsef as reay ony consequences, however itte remote and however naturay to be anticipated, of our actua action in the supposed minimum physica sense, which wi then transpire to be the making of some movement or movements with parts of our body (e.g. crooking our finger, which produced a movement of the trigger... which produced the:death of the donkey). [ pp Note that if we suppose the minimum physica act to be movement of the body when we say " moved my finger," the fact that the object moved is part of my body does in fact introduce a new sense of "moved,"... The ordinary use of "move" in such exampes as " moved my finger" is utimate. We must not seek to go back behind it to "puing on my musces" and the ike. [Footnote, pp We do not seem to have any cass of names which distinguish physica acts from consequences: whereas with acts of saying something, the vocabuary of names for acts (B) seems expressy designed to mark a break at a certain reguar point between the act (our saying something) and its consequences (which are usuay not the saying of anything). [p. 112

61 WORDS AS DEEDS [For drawing the ine we seem to derive some assistance from the specia nature of acts of saying something by contrast with ordinary physica actions: for with these atter even the minimum physica action, which we are seeking to detach from its consequences, is, being a bodiy movement, in pari materia with at east many of its-r:mmed1ate and natura consequences, whereas, whatever the immediate and natura consequences of an act of saying something may be, they are at east not normay other further acts of saying something [ assume that, as per page 117, he'd ca them "responses.. or 11 Seques 11?, whether more particuary on the speaker's own part or even on the part of others. So that we have here a sort of natura break in the chain, which is wanting in the case of physica actions, and which is associated with the specia cass of names for iocutions. [p. 113 [Footnote, same page: This in pari materia coud be miseading to you. do not mean.. that my.. moving my finger 11 is, metaphysicay, in the east ike 11 the trigger moving 11 which is its consequence, or ike 11 my finger's moving the trigger... But 11 a movement of a trigger finger 11 is in pari materia with 11 a movement of atrigg"er. We coud quote many other reevant passages. But these shoud be enough to point up the underying dramatistic quandaries which specuations aong these ines just naturay confront, and which Austin considers with his characteristic acuity. His probem of drawing the ine between iocutionary 11 Seques 11 or 11 responses 11 and perocutionary 11 Consequences, 11 hence a fortiori between speech acts and 11 minimum physica actions" or bodiy "movements,.. invoves us in various considerations that come to a head on pages , which pease consut. We confront one accidenta probem of nomencature here. 163

62 164 KENNETH BURKE The terms "action,. and 11 motion 11 are not generay differentiated as must be the case for out-and-out dramatistic purposes. We refer to the 11 actions 11 of a motor, for instance, whereas in the strict usage required for our purposes one woud have to say the 11 motions 11 of a motor. Or we refer to the 11 movements 11 of a symphony, whereas we'd have to ca them different ''acts,.. 11 as we refer to a drama in five acts... n this ight, consider (p. 114) the 11 Uttering of noises, 11 which is a physica movement,.. yet can be caed a phonetic act... Confining ourseves to the stricter use of the two terms, we shoud state the case thus: The utterance, in being caed a speech act, woud thus be denominated exacty right for our purposes. (Since such a medium is a conventiona symbo-system, a synonym woud be 11 syrnboic act. 11 ) But the term "phonetic act 11 requires some protective quaification. For the 11 act 11 of speaking (and of interpreting an utterance) is made possibe ony by its grounding in two aspects of motion; namey: (a) such physioogica motions as the neura processes invoved in speaking, hearing, interpreting, and the ike; (b) such environmenta motions as the vibrations in the air which carry the words from speaker to hearer. (As per Austin, footnote page 114, we are 11 Confining ourseves, for simpicity, to spoken utterance... Written words woud depend on visua rather than auditory kinds of environmenta motion, Braie on motions invoved in touch.) "Vv'hen "iocutionary force" (which woud equate with "attitudina santing 11 ) is viewed thus, a notabe principe of duaity enters the case. Austin expicity deas with an 11 aspect of it, when on page 117 he says: Many iocutionary acts invite by convention a response or seque. Thus an order invites the response of obedience and a promise that of fufiment. The response or seque may be 'one-way' or 'twoway': thus we may distinguish arguing, ordering, promising, suggesting, and asking to, from offering, asking whether you wi and asking 'Yes or no?'" Note the underying difference he has hit upon here. The "seque" to an "order" invites a "response,. on the part of the hearer. But the "seque" to a "promise" invites 11 fufiment" on the part of the speaker. When first presenting his grammatica as test (see page 57), he said: "We need not waste our time on the obvious exception of the first person pura, 'we promise. 'we consent,' &c." But a rhetorica "two-way" consideration has entered. So it's conceivabe that the speaker who says on page 117 " promise" is saying in effect not " promise that ) 1

63 WORDS AS DEEDS 165 wi deiver the goods" (as per Austin's opening anecdotes to iustrate performatives) but " promise that if we a do what am advocating we' a get fufiment." n either case, the "response of obedience" obviousy invoves rhetorica matters of persuasion. And the "two-way " aspect of the case is centra~ Whatever attitude my speech act might utter, note that such utterances can work two ways. may fee an attitude, say, of resentment. No, et's make it an attitude of friendiness; there's enough meanness et oose these days aready. fee so friendy, my utterance manifests my friendiness. Or, if you wi, -'m not friendy. at a, but fake an attitude of friendiness. Or suppose that, whether 'm reay friendy or not, woud evoke in you an attitude of friendiness. Aristote's Rhetoric gives you many tips on how to go about trying to be persuasive thus, whether or not you happiy attain the consequences of the attitude that you aimed to estabish in your audience. The mere fact that have a pronounced attitude towards something and want you to share it with me is no guaranty at a that my way of expressing the attitude wi be the best way of getting you to share the attitude. But whatever the differences between a one-way or a two-way kind of attitudinizing in my speech act, in either case my speech act invoves correative behavior in the ream of motion. Regardess of what attitude any speech act "symboizes," it can be enacted ony insofar as there are corresponding neura motions of the body (whatever they may be) And whatever attitude (response) such a "rhetic" structure of utterances (ocutions) may evoke in a hearer, they wi necessariy be paraeed somehow in the ream of wordess (nonconventiona, non-symboic) motion. Behaviorism woud be pointing in the right direction here, except for its basic methodoogica error at the start, its assumption that the distinction between the reams of action and motion is but a matter of degree, rather than a difference in kind. And having thrown out phiosophy (which 'd want to ca Logoogy), Behaviorists can be phiosophicay (Logoogicay) obtuse in their projects that by impicit definition, reduce the ream of action to terms of motion. And run-down Technoogism being what it is, the more methodoogicay obtuse they persist in being aong those ines, the more grants they' get. But by a dramatistic reinterpretation, much of their work can be of great use, in heping to suggest the proper admonitions when we are attempting to sum up just what is invoved in our being the

64 166 WORDS AS DEEDS kind of symbo-using, speech-acting anima we are, as viewed in terms of MOTON, ACTON, and ATTTUDE. Towards summing up, think of these sampe histories: (1) A drug, introquced in-eo the bood stream, produces physioogica effects. The drug functions, et us say, as a kind of irritant. ( go back to a mean exampe, since my anecdote is cearer thus.) t makes me so irritabe that my speech act refects my attitude of irritation. By the same token, such woud be the iocutionary force of my reated utterances. Thereby a condition in the ream of physioogica motion wi have surfaced as a speech act which can have iocutionary force with hearers ony insofar as the motions of their bodies and in the conditions of the environment make it possibe for those famiiar with the conventions of my utterance to receive it and interpret it as the kind of speech act (symboic action) it is. (2) Or et's take another route. Mine enemy referred to me in utterances (a rhetic act) the attitude, or iocutionary force, of which found quite irritating. My response to his attitude had a physioogica counterpart in the ream of motion by such "behavior" as increased bood pressure, acceerated puse beat, secretion of adrenain, without which body symptoms his attitudinizing coudn't have had such iocutionary force so far as my response to his speech act was concerned. (3) Or might address you in a way designed to buid up in you an attitude of irritabiity that woud induce you to sympatheticay join with me against mine enemy. n any case, whether you went aong with me or not (whether or not my iocutionary act attained the consequences that entited it to be caed a perocutionary act) whatever your response was it necessariy invoved physioogica and environmenta motions of one sort or another. Basicay, the two-way ambiguity ("asymmetry") is this: The speaker's iocutionary uttering of his attitudes is not directy equatabe with the hearer's response to such an iocutionary force, the "preformative" aspects of speech, be they expicit or impicit.--xtid in any case, even if there are no "consequences" in the sense that, when A said to B "shoot her," B shot her, there must be the physioogica and environmenta motions that serve as the "materia cause" of the speech act. Though obviousy neither Mainowski nor Austin woud have any reason to deny such a grounding of speech in a context

65 KENNETH BURKE 167 of speechessness (symboic action grounded in nonsymboic motion), it is my caim that ony an out-and-out, formay dramatistic nomencature sets the conditions for inquiries into the sheer bodiy equivaents of the speech acts' attitudes, a ream of quantification such as the corresponding quaitative nature of the speech act decidedy is not. Given a performer's expertise in speech acts, for instance, his body may be "behaving" in ways that are quite heath-giving, or maybe in the ways of psychogenic iness. One dreamer's "brain waves" may be a to the good, another's may be in bad need of repair. A citizen's behavior is one thing. His body's behavior is someth1ng ese. n the speech department he may be anticipating or remembering. But with his nonsymboic body, as with a dancer's symboizing body, everything is NOW. As interpret Richards (and his pa Ogden) aong Behaviorist ines, somehow our speech acts are utimatey cathartic insofar as the symboic "resoution" of our conficts digs down even 1nto its tota physioogica counterpart. And we are eft with this: n Mainowski's anecdotes there is the featuring of iiterate "they's." n Austin's story, there is featured a highy iterate grammatica "." Now propose to end on what want for a starter in sympathy with them a; namey (now brace yoursef for this outand-out dramatistic statement of the case): We start from such things as trees, and draw an absoute ine between trees and the word "trees." And never the twain sha meet, despite ung's wavering batte to persuade himsef that they hoisticay can. After a this is over, we confront this summarizing ine-up: Speech acts are iocutionary attitudes; a such are grounded in the speechess behavior (motions) of the body; speech acts are not reducibe to terms of the body's motions, but we can inquire into their modes of motion, which are quantifiabe in such terms as brain waves and endocrine secret1ons, though as acts they are whoy quaitative. There are utimate behavioristic ("behabitive"7) correspondences of this sort, to be confronted in our quandaries concerning the reation between speech acts (as pubic, socia, "conventiona") and their grounding in the individuated wordess motions of each communicant's physioogica organism (aong with the environmenta motions needed to carry the "noises,"

66 168 WORDS AS DEEDS or "phones," from speaker to hearer). The speech act woud be in the coective ream of "cuture. But it woud be grounded in each user's individua physioogica "nature. Andover, New ersey!

67 BRUCE FRASER Warning and Threatening As the interest in semantics expanded during the eary 1970's, it became obvious that a carefu, critica anaysis of individua words or sets of words comprising a semantic fied was necessary if the program of semantic compositionaity set out initiay by Katz and Fodor (1964) was to be vaidated. think there is now an anaogue within the deveoping fied of pragmatics, namey that, criven the basic notions expressed in Austin (1962), and further eaborated in Seare (1969), we must turn to an anaysis of certain crucia speech acts if the entire program is not to be eft in the programmatic stage. The foowing is a short attempt to do this with two reated acts: warning and threatening. We wi first consider the act of warning in detai and then discuss threatening, which we wi treat as a specia type of warning. Let us estabish some reevant distinctions. First, we can warn someone by either verba or non-verba means. might ye to you that the tree you are cuttinq down is beginning to fa towards you, thereby verbay warning you; or might just sienty point, mouth agape, at the faing tree. n either case, can have warned you. We wi focus our discussion ony on verba warnings but intend the anaysis to be appicabe to the non-verba as we. Second, a warning can be either intentiona or unintentiona. might, for exampe, say to you that have just seen your boss coming into the buiding, intending that this count as a warning that you shoud appear busy. On the other hand, my intention in saying this might have been soey as a report of my observation, say for exampe, if had not seen your boss for severa weeks and thought him to be out of the country. You might have interpreted my utterance as a warning, though had not meant it to be taken as such. And third, foowing Austin's distinction of speech-act types, a warning can be made directy as a function of what one says--warning as an iocutionary act--but aso as a consequence of what one does--varninq as a perocutionary act; whereby the hearer becomes warned as a resut of [Centrum, 3:2 (Fa, 1975),

68 170 BRUCE FRASER the speaker having, say, caimed something. iocutionary acts of warning first. We wi consider We wi say that ocutionary Warning is successfuy performed when2 C C2 C3 the speaker intends to convey a proposition, p, which specifies a future action; the speaker intends that p be taken as representinq an action which a. is ikey to occur (perhaps conditionay); b. has consequences disadvantageous for the hearer; the hearer recognizes the intentions C and C2 of the speaker. Condition C requires that the time of any proposition used to warn be in the future. Obviousy, cannot warn you about a past event. nsofar as might warn you, for exampe, by teing you of what happened to the ast person who did what you are doing, must convey a proposition about a future action invoving you, even though this proposition may ony be impied. Condition C2 requires that the speaker intend the hearer to view the action specified in the conveyed proposition as being both ikey to occur and having consequences unfavorabe to him. We incude the notion of conditionaity in C2(a) because the ikeihood of the action at issue may be contingent on the hearer's own behavior. For exampe, to say "if you sit down hard on that chair, it wi break under you" is to warn of potentia harm to you which wi resut from the chair breaking under you, but to indicate that the ikeihood of this happeninq rests with how hard you seat yoursef. t is not necessary that the hearer take the effects of the action as, indeed, being unfavorabe. The hearer may, for reasons of fact or differing opinion, view the action as being quite in his favor, as being quite desirabe. However, as ong as the speaker hods the unfavorabe view (in some cases even after becoming aware of the differing opinion of the hearer), condition C2(b) has been met. These two conditions refect a qenera characteristic of a iocutionary acts, namey, that they are aways intentiona acts; one cannot warn (or request, promise, caim, etc.) unintention- 1

69 W,ARNNG,...ANO T-REATENE-G 171 ay athough a hearer can misunderstand a speaker's intention and take him to be warning (or requesting, etc.). Condition C3 is simpy the condition required of a iocutionary acts, namey that the successfu performance of any iocutionary act requires that the hearer recognize both the reevant proposition the speaker intends to convey and the way in which the speaker intends the proposition to be taken. Faiure to recognize one or both of these speaker intentions resuts in the faiure of the successfu performance of the intended iocutionary act. n such a case, the speaker can be said to have intended to warn, but not succeeded in doing so. On this anaysis, warning is a kind of Representative Act.3 There are severa Associated Conditions on the iocutionary act of warning: conditions which we can expect to be met but are not necessary for the successfu performance. A A2 A3 A4 The speaker beieves that the effect of the action specified is disadvantageous to the hearer. The speaker beieves that the action is ikey to occur. The S?eaker beieves that the hearer is capabe of doina something to avoid the consequences of the action. The speaker beieves that the hearer does not view the action as ikey. Condition A indicates that it is usua, though not necessary, that the speaker beieves the "warning" proposition. can, for exampe, successfuy warn you that your car engine is making strange noises but be quite aware that the probem is not at a serious--it is ony severa eaves caught up in your fanbet puey. An insincere warning, but a warning nonetheess. To vioate condition A2, might warn you, for exampe, that the bu is about to charge, when know that the bu is chained to his sta. Again, an insincere warning. Simiary, might issue an insincere warning by announcing that there is an avaanche rushing down on you, though beieve that neither you nor anyone ese can do anything to avoid caamity. The fourth associated condition specifies that, under ordinary conditions, the speaker beieves the hearer to be unaware that the action is ikey to occur. But here again, it need not be the case, though to warn someone of what he aready knows, is to issue a defective (though

70 172 BRUCE FRASER not insincere) warning.4 Let us now consider the notion of a Perocutionary Warning. The term "warning" is one of the reativey few instances which serves both to name an iocutionary and perocutionary act, though these acts are not identicar:- We have, on the one hand, acts such as promising, requesting, suggesting, authorizing, and caiming, which are ony iocutionary acts: they are directy reated to what can be performed ony as a function of what the speaker is saying, they are aways intentiona, and there is a conventiona way of guaranteeing their successfu performance. We have, on the other hand, acts such as convincing, surprising, annoying, aarming, and overwheming, which are ony perocutionary acts: they can be performed ony as a consequence of the speaker having done something in speaking, they may unintentionay be carried out (as we as intentionay), and there is no conventiona way of guaranteeing their successfu performance.s n introducing the notion of a perocutionary act, Austin (1962) wrote that: Saying something wi often, or even normay, produce certain consequentia effects upon the feeings, thoughts, or actions of the audience, or of the speaker, or of other persons... We sha ca the performance of an act of this kind the performance of a perocutionary act or perocution.... t wi be seen that the consequentia effects of perocutions are reay consequences, which do not incude such conventiona effects as, for exampe, the speaker's being committed by his promise. [pp. 101, 102 As the quote indicates, the essence of a perocutionary act is that the certain effects on the feeings, thoughts, or actions of the hearer come about as a consequence of the speaker having said and therein done something. The perocutionary effect of a request, for exampe, might be that the hearer decided to carry out what was asked or ony to get angry and stomp away. For a warning, intentiona or not, the reevant effect is that the hearer comes to be aware of some ikey action, the effects of which the hearer takes to be disadvantageous. We must stress that on ths anaysis, any effect of fear, anxiety, or the ike produced by a warning is a consequence, not integray part of it. As Austin was quick to point out, there is no conventiona formua which coud assure the speaker of the successfu

71 WARNNG AND THREATENNG 173 performance of a given perocutionary effect (as there is the performative formua for iocutionary acts), nor any way to assure the speaker that the hearer wi not become warned without the speaker intending it. There is nothing more precise can say, at east at this point, about the performance of the perocutionary act of warning. We turn now to the notion of a threat, which we see to be a specia type of warning, abeit a rather uncharitabe one. We wi say that an ocutionary Threat is successfuy performed when: C C2 C3 The speaker intends to convey a proposition, p, which specifies a future action; The speaker intends that p be taken as representing an action which a. is going to occur (perhaps conditionay); b. has consequences disadvantageous for the hearer; c. the speaker wi assume responsibiity for carrying out (though not necessariy directy doing it) ; The hearer recognizes the intentions stated in C and C2. n short, a threat is a warning when the speaker takes on the responsibiity for bringing about the disadvantageous action. t foows from this anaysis of threatening that any warning of no matter how disadvantageous proportions to the hearer, for which the speaker assumes responsihi i ty, wi- count as a threat. f, for exampe, shoud te you that "f you do not get here before 8:00 p.m., we wi start the party anyway," then, under the assumption that you fee strongy about being present at party-beginnings, have threatened as we as warned you. Some native speakers of Engish may disagree that such an utterance shoud count as a threat. f so; then the aternative is to set threats apart from warnings by two criteria: speaker responsibiity and some criteria for the degree of disadvantage to the hearer after \-7hich a warninc:r

72 174 BRUCE FRASER becomes a threat. This issue marks an interesting, though yet unexpored area. The Associated Conditions for a threat are somewhat different from those of a warning. A A2 A3 The speaker beieves that the effect of the action specified is disadvantageous to the hearer. The speaker beieves that he is capabe of bringing about the action specified. The speaker intends to bring about the action threatened. The first condition is the same as that for a warning, but the others are different, particuary A2, since inherent to a threat is speaker invovement in the impending action. Conditions A3 and A4 on a warning--that the speaker beieves the hearer to be capabe of avoiding the consequences of the action--are not reevant to a threat. Many threats are issued in the form of a conditiona which expicity sets forth the conditions for avoiding the action; for exampe, when say to you that 11 f you ca up and bother me one more time, ' have you arrested... On the other hand, can perfecty we threaten you by saying ony 11 am going to have you arrested for what you have been doing to harass me,.. with the impication that there is nothing now that you can do about it, and independent of whether or not you are aware that fee harassed. Condition A3 specifies that it is reasonabe to assume, a other things being equa, that the speaker intends to carry out his threat. This may not be true, as is the case with what we ca an 11 ide 11 or 11 empty 11 threat. Anaogous to the perocutionary warning is the perocutionary threat. As a resut of performing one or more iocutionary acts, the speaker can, intentionay or not, cause the hearer to fee threatened. But here again, there is an important distinction to be made: success in threatening, whether it be by iocutionary or perocutionary means, does not entai that have thereby aarmed you, made you afraid, or anything of the sort. That these are often the intended effects of threatening is an interesting and important point, but not germane to the anaysis of the act of threatening, per se. t has been suggested that a threat is reay a type of negative promise. To promise invoves specifying a future act whose effect is advantageous to the hearer; a threat 1 1

73 WARNNG M{D THREATENNG 175 invoves an act with disadvantageous effects. But there is an important difference: whereas in promising the speaker intends the utterance to count as the undertaking of an obigation to carry out the action, a threat invoves no such commitment. There is the impication in both cases that the speaker intends to act as he has indicated, but one has not obiged himsef to act by threatening. We cannot chaenge the person who has threatened to bankrupt us, for exampe, by asking a court of aw to require him to compy. We can, of course, do this with a promise, which is seen as a type of contract with its rights and obigations. There is one very interesting point that have eft unti ast: f a threat is, indeed, an iocutionary act, why is it then that one cannot use threaten as a performative verb?6 The sentence " threaten to f1re you," for exampe, is grammaticay we-formed; however, it is what one woud answer to the question, "What do you aways do when come in and comoain that want a raise?" ts utterance cannot count as a threat. Other forms of the compement of threaten (e.g., threaten you with being fired; threaten that wi fire you, etc.) are aso qrarnmatica, but none of these can be used directy to perform a threat. 7 Katz (1976) argues that exampes such as " threaten you and your city with destruction," "We () hereby threaten you for the ast time" are not pecuiar if the speaker is someone with power and has the abiity to annihiate the city. The genera point he makes is that the acceptabiity of sentences such as these may depend on the identity of the speaker in the context of utterance. (This is an interesting and potentiay important extension of what factors contribute to sentence-acceptabiity, but one which we wi eave unaddressed at this time. n this regard, see McCawey, 1976.) Katz argues further that, since threats are serious matters, and the use of the reativey forma perforrnative form focuses more on form than on the content of the threat, "the threatener must try to reduce the forma aspects of his or her iocutionary act as much as possibe in order to maximize the significance of its message and hence its menacing aspect" (p. 218). Generas, presidents, and the ike, whose threats must be performed in a more rituaized, forma situation, are presumaby permitted to use the perforrnative form. The argument invoving the more forma situation of utterance seems to me to be simpy wronq. To fire, forbid, condemn and hex someone a seem pretty serious matters, particuary to the hearer, yet, they each can be used performativey. For exampe,

74 176 BRUCE FRASER " (hereby) hex you with the Mundini curse," " (hereby) put the Mundini curse on you," " (hereby) condemn you to a year of studying structuraism," " (hereby) forbid you to read that book." There seem to be severa aternatives here: (1) that threatening might not reay be an iocutionary act, ony a perocutionary act; or that threatening is an iocutionary act, but (2) there are good conversationa-reasons why threaten is not used performativey, or (3) threaten is just an "uncooperative" iocutionary verb. can see no justification at the present to reject threatening from the ranks of iocutionary acts. t meets those conditions which distinquish such an act from others, namey, that the act is successfu just in the case the intended proposition and its force is recognized by the hearer. Moreover, our anaysis above indicates that threaten can be anayzed as a specia type of warning, an incontestabe iocutionary act. Assuming, then, that threatening is an iocutionary act, is it the case that threaten is just an uncooperative iocutionary verb? We do find iocutionary verbs such as report, say, and criticize which are found far more acceptabe when preceded by some moda, for exampe, " can report that ohn is home now," " must say that have never reay iked that quy." But no moda faciitates the performative use of threaten. n an effort to determine if this apparent uncooperativeness was pecuiar to Engish ony, inquired about the performative use of threaten in a variety of other anquaqes. n none of the anguages inquired about coud threaten be used performativey. A of the foowing exampes are grammatica, but not usabe with the performative sense. Engish: German: Spanish: French: apanese: Russian: Hebre'.v: threaten you with ch bedrohe Sie mit Yo te amenaco que. e vous menace de. Boku wa kimi ni... to kyoohakasuru Ya grozhu vas ubit M'ayeme aecha... This suggests stronqy to me that the ack of oerformative use in Engish is not an idiosyncracy of the verb itsef, but invoves s 1 1

75 WAfu~NG AND THREATENNG 177 some principe(s) of conversation. Are there, then, certain principes of conversationa behavior that precude one predicating directy that he is threatening? Harnish (1975) has presented a convincing argument of why one cannot acceptaby caim " am urking here beside your window," based on an anaysis of the verb urk which requires that the urkee not be aware of the urker, est the urking be thwarted. Simiary, one might argue that to utter " hereby boast that am the greatest" is to predicate of mysef outstanding positive auaities, something one just does not do, whether or not they are justified. Thus, the above sentence, whie grammatica, is not used performativey. t may we be the case, then, that the unperformative nature of threaten resides with the fact that to threaten invoves not USt a persona warning to the hearer but aso requires the speaker to assume an orientation that wi guarantee (or presumaby quarantee) hearer disfavor. We find that there are severa acts which invove,a simiar speaker orientation, namey criticize and admonish, each of which appears to be an iocutionary act but ne1ther of which has a performative use: criticize you for doing that; admonish you for not heping her. f these verbs are unperformative because of this reason, we have here a case of where socia rather than inguistic forces determine acceptabiity. Schoo ~f Education Boston University NOTES i woud ike to acknowedge the comments and criticisms of Benoit de Cornuier and Donad Freeman on earier versions of this paper. 2poowing Austin, Seare (1969) distinauishes hetween a successfu and non-defective iocutionary act, and a successfu but defective one. He writes: There are various kinds of possibe defects of iocutionary acts but not a of these defects are sufficient to vitiate the act in its entirety. n some cases, a condition may indeed be intrinsic to the notion of the acts in question and not satisfied in a given case, and yet the act wi have been performed

76 178 BRUCE FRASER nonetheess. n such cases say that the act was defective! (p. 54) According to his anaysis, an insincere promise is a successfu abeit defective promise. A speaker who indicates that he intends to create the obigation to do something can be said under the appropriate circumstances to have promised, irrespective of whether he intends to carry out his obigation. 3 am using the term as suggested by Seare (1976) to designate a those acts such as stating, caiming, teing and agreeing, in genera, acts of decaring. 4Though the term "insincere" has been used to refer to the type of defectiveness when such necessary conditions have been vioated, the term has, itsef, never been carefuy defined. 5 other acts with dua citizenship incude apprising, informing, notifying, reminding, aerting, teing, congratuating, criticizing, baming, condoning, and praising. t seems to be a genera rue that those iocutionary acts which have a nominaized form which occurs in the frame' ~ was " (e.a., was apnrised of the fact; was criticized for saying that) have perocutionary correates. These acts of dua citizenship raise an interesting question. Assume, fvr exampe, that intend to warn you but eave unspecified how wi do this. f utter " warn you that S," then have ceary warned you by an iocutionary act. At what point, however, does my warning cease to be a function of the meaning of what say and certain conversationa principes, thereby ceasing to be an iocutionary act, and become a perocutionary act? The boundary ine between these two types of acts needs to be carefuy carified, even if we must concude that there is no sharp distinction. 6There are some constructions in which either the nominative threat or warning (or the associated adjectiva form) can co-occur, but not the other. For exampe, we have a warning but not a threatening track at the edge of the baseba fied; one may hear a warning signa but not a threatening signa; a biboard with information about drinking can serve as a warning but not a threat to a drivers; and we can put someone on warning but not on threat; and one issues a warning but not a threat. Conversey, an iceberg can serve as a threat but not a warning to a sea traveers (though the tip can serve as a warning, not a threat); and we make a threat to someone but we do not make a warning. Except for the make/issue difference, these a appear to foow from the distinctions between threats and warnings which we have discussed above.! 1

77 E WARNNG AND THREATENNG 179 Why one can issue a warning but not a threat (or why one can make a threat but not a warning) is uncear. One issues demands, requests, responses, recommendations, but one can make them as we; to "issue.. something smacks of more formaity and, perhaps, more power on the part of the speaker than to simpy "make" the same. These need not be ceremonia acts such as judging or procaiming. However, one issues instructions but does not make them. On the other hand, one makes promises, sug~,,tions, caims, and criticisms but one does not ordinariy issue them. f the formaity distinction hods up, then it woud foow that a warning impies more formaity than does a threat. This is contrary to what Katz caims as we discuss beow. 7 The situation is even "worse 11 than the fact that threaten has no performative use: there is no simpe expicit way to perform a threat at a. To be sure, can utter " am taking it upon mysef to see that you fired," and amost everyone woud interpret it as a threat. But, whie have made expicit my assumption of responsibiity, there is nothing in the meaning of the sentence which guarantees that intend the action of firing you to be taken as disadvantageous. There is certainy a strong connotation associated with the act of firing that woud permit the inference of disadvantage, but nothing conventiona about the meaning of "to fire" that invoves disadvantage. coud, of course, go further and tack on to the above sentence "which action see to be to your distinct disadvantage," thereby presumaby fuy expicating the meaning of the verb to threaten but at an extraordinary verba cost. 8 Austin (1962) found threaten acceptabe as a performative verb; he wrote, 11 he can use the performative ' warn you that' but not ' convince you that' and can use the performative ' threaten you with' but not ' intimidate you by'; convincing and intimidating are perocutionary acts.. (p. 130). BBLOGRAPHY Austin,. L., 1962, How to Do Things With Words~ Massachusetts: Harvar~University Press. Cambridge, Fraser, B., 1976, On Requesting: An Essay in Pragmatics_ To appear. Ditto, Boston University. Harnish, R. M., 1975, "The Argument from urk." Ditto, University of Arizona. Katz,., 1976, Propositiona Structure: A Study of the Contribution of Sentence Meaning to Speech Acts. Tobe pub1shed by Crowe, New York.

78 180 BRUCE FRASER Seare,., 1969, Speech Acts: An Essay~ the Phiosophy of Language. Cambr1dge: Cambridge Un1versity Press. 1975, "A Taxonomy of ocutionary Acts." Minnesota Studies in the Phiosophy of Science, 7. )!

79 ELZABETH W. BRUSS Manufactured Signs: Semiotics and the Automobie Semiotics becomes probematic, it seems, just when it becomes most interesting, when it significanty aters od habits of study and at the same time unsettes our famiiar conceptions of the sign. The diversity of objects and approaches now aying caim to the discipine threatens to overwhem it, and there is no rea agreement among those who ca themseves semioticians {or semioogists) about the imits of egitimate semiotic research. n a recent survey of "the state of the art," Thomas Sebeok is forced to ask:... whether such generaizations, represented by the gossematicians {and again in such works as Barthes', e.g. 1968), merey express methodoogica attitudes, at best providing a uniform frame of reference for some or a of the behaviora sciences, or incorporate genuine empirica insights. 1 Whether or not one uses Sebeok's empirica criteria, his question is indeed crucia for semiotics. What motivates the decision to treat somethinq as a sign, and do some treatments have better motivation than others? On one side, we face the risk of overextending a theoretica term, making "sign" so eastic that it is vacuous. Can any interesting generaizations be made about a domain which incudes objects, events, and actions, conscious as we as unconscious processes, both anima and human behavior? But the dangers of imprecision, contradiction, and anthropomorphism--great as they are--must be weighed against the equa ~anger of tautoogy. f semiotics can embrace ony what is most "anguage-ike," then it merey dupicates the efforts of other discipines--or bases itsef on a circuar definition. Semiotics cannot rest on the very assumptions about the nature of communication it was designed to formuate and test. nstabiity is probaby inevitabe in a theory which is capabe of taking itsef as an object of study, and the structure and use of meta-anguages is certainy a potentia area for semiotic inquiry. [Centrum, 3:2, (Fa 1975),

80 182 ELZABETH W. BRUSS Severa authors, Sebeok among them, have pointed out one common source of confusion in semiotics, a tendency to confate it with the study of structure pure and simpe, particuary those structures which somehow resembe grammatica constructions.2 But a reguar distributiona pattern is not in itsef proof that a structure is significant. Thinkers disagree about whether syntax is necessary to semeiosis, but it is ceary not sufficient. The aternative is to define the sign in terms of its semantic or communicative function. Whie this is more intuitivey satisfying, it raises the new probem of defining the definition, stating precisey what we mean by "meaning" and "communication," or at east justifying our choiae of certain paradigmatic exampes.. At first gance, "significance" in an automobie woud seem remote from verba sense, the exchange of information invoved in anguage and in commerce too disparate for any but the most tenuous or reductive anaogies. But perhaps our usua notions of meaning and communication are too anguage-bound, serviceabe ony because we can take so much for granted from our position inside the system. n anguage, as Wittgenstein has put it, "one is ~ rushing ahead and so cannot observe onesef rushing ahead."3 f the suggestion that automobies are aso signs chaenges our opinions, it might ead us to re-examine semiotic practices which have become invisibe through famiiarity. Broad semiotic generaizations wi seem queer as ong as we abor under the misconception that meaning exists independenty ~ of the sign and its use, in the form of menta or physica objects or in prefabricated sets of synonyms and dictionary definitions. (Saussure's famous distinction between "signifier" and "signifiant," which has achieved a new currency in European semiotics, 1s dangerousy miseading in this respect.4) When efforts to ocate the referent or the synonym for some nonverba sign fai, one may ~ judge the whoe project a faiure. But whie we do use anguage (and some other sign systems) for ocating and describing features of our environment or for expressing our subjective experiences, it is not the objects or the feeings themseves which determine the names they wi receive or provide criteria for judging when an act of naming is correct. t is not what signs express or name but how they are used, the way they are manipuated and the standards which must be met, which make them meaningfu. "Meaning," "sense," and "significance" are shorthand terms which concea the compex operations which are performed in and through signs--operations which need not incude "naming," or which may "express" correations other than that between emotions and sentences. What makes the category of "sign" so sippery is the fact that amost anything can be used as a correate for amost anything ese, according to ~ very different criteria of emission and interpretation, and with equa variety in the ways criteria ar~ estabi~hed, ~nforc~d, a~d transmitted. Peirce's cataogue of s1gn reat1ons, 1ncud1ng h1s important distinction between "icon," "index," and "symbo," provides

81 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE 183 a hint of the number and kind of semiotic systems which are ogicay possibe. f there are in fact empirica imits to semeiosis, then one coud argue that semiotics shoud remain within them. But one might aso argue that the study of potentia sign systems is vauabe in itsef, or that it eads to striking specuations on the way bioogy, psychoogy, and socia structure constrain the use of signs. When a phiosopher of anguage ike Wittgenstein proposes provisiona rues for verba meaning, he aso throws ight on the parameters that distinguish one sign system from another.... how am to. know that the word "s arne" describes what recognize? Ony if can express my recognition in some other way, and if it is possibe for someone ese to teach me that "same" is the correct word here. 5 The conditions here put forward for using "same" correcty may work for anguage without appying equay to other semiotic processes. For exampe, one woud be hard put to find instances where one anima deiberatey instructs another in the proper use of a mating cry. Forma instruction is equay rare in human nonverba communication, incuding training in the meaning of artifacts and manufactured goods. (Wittgenstein's exampe is, of course, too simpe and atomistic by far to expain anguage acquisition; it merey iustrates the inguistic fact that words can be taught and earned, that the function of a word in an instructiona context is one of the many functions that together constitute its meaning.) When there is training in nonverba behavior, it is customariy accompanied by and mediated through anguage--with the resut that the verba goss is often mistaken for "the meaning" of the sign. Even so insightfu a writer as Roand Barthes has suggested that anguage aone is truy "inteigibe," pausing in the midst of his study of French fashion to ask "y a-t-i un seu systeme d'objets, un peu ampe, qui puisse se dispenser du angage articue?"6 t is not hard to see why Barthes goes wrong, when anguage users themseves so often treat their synonyms and definitions as if they were the source of meaning in speech, rather than impicit transations and comparisons. But when Wittgenstein audes to the abiity to "express my recognition in some other way," he painy has in mind an exchange of one speech act for another--not the anchoring of words to somethinq more substantia. An incomparabe sign, sharing no features of form or function with any other, woud indeed approach meaningessness. Yet anguage is not the ony sign system which permits comparisons and substitutions. Music is a prominent case of a nonverba

82 184 ELZABETH W. BRUSS system with precise transation vaues. One can distinguish fine gradations of pitch and, to some extent, timing, athough other reations are ess exact and usuay confined to the thematic structures of a particuar score.? The parameters of precise comparison are far more numerous and compex in anguage. With its broader range of uses and more highy ordered semantic fieds, anguage aows much more subte cacuations of synonymy, contrariety, and reative scope. t is a richer instrument and this no doubt makes it "fee" more meaningfu than most nonverba systems, athough the semiotic principe is much the same. t is the capacity for substitution and exact distinction which makes the automobie system a particuary good subject for semiotics. The advent of mass production, giving rise to standardized parts and procedures, estabished new eves of absoute identity and inter-changeabiity among manufactured objects. Whie mass production restricts the idiosyncratic features of an artifact and imposes imits on its reevant dimensions of vaue and use, it simutaneousy creates a base for more precise distinctions in stye. n any semantic fied, according to Urie Weinrich, "the more popuous and amorphous a paradigmatic set of eements, the ess certain is the organization of their contrastive features."8 By refining and restricting the paradigm, mass production impoverishes individua products, making them transparent, tokens of a singe type. But it thereby increases the meaning-potentia of those variations that do occur. Thus annua mode changes, distinctivey "sporty" or "famiy" cars, aong with the deicate rankings of "uxury" or "economy," must await a we-estabished system of mass production. Without such standard parts and identica methods of assemby, each automobie remains argey a thinq unto itsef, grossy simiar to others, yet unpredictaby and immeasuraby different as we. The impications of owning one car rather than another are too vague to serve as the basis for a system of communication. Whie each car is recognizabe as a car, the detais of its construction and even its cost are a matter of private knowedge or pure conjecture. The code is too poor for more than the crudest dispays of capita investment; articuate degrees of "taste" are quite beyond it. Thus in the primitive transactions which characterized automotive commerce at the turn of the century, the car figures ony as "a rich man's toy." The system coud not as yet effect a transation of weath into other, subter terms.9 But the rise of mass production and mass consumption created the conditions for truy significant automotive stye- stye without idiosyncracy or rea aesthetic function, a stye of pure information, egibe to an entire society (if the standards were set by reativey few members of that society.) ndeed since the 1920's, when Ford's hegemony in mass production gave s s

83 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE 185 way to "Soanism" and the Genera Motors strategy of mode proiferation, stye has repaced the automobie itsef as the principa commodity of the American automotive industry. Athough the so-caed "stying strategy" was a response to saturated markets and infexibe capita, it coud not have succeeded, might not have even suggested itsef, had the prior semiotic base been acking.o The over-zeaous absence of variety in Mode-T production made it a the more ikey, no doubt, that the next cyce of consumption woud consist in reieving the artificia monotony Ford had induced--a monotony of work as we as acquisition, fast becoming the pattern for the rest of American economic ife as we. The Mode-T is the epitome of the automobie-as-token, a ceebration of reduced presence and subimated energies. t was-actuay a simpification of earier car design, not what one might imagine given our habit of thinking that technoogica history is aways progressive. n making his engine and gear-system ess compex to suit the Mode-T to a wider group of buyers, Ford created an important precedent: rather than encourage a more educated pubic, the manufacturer woud make a ess difficut product.2 t is interesting to note that both the eary Ford and its cosest counterpart, the German Vokswagen, were subjected to pubic ridicue in the form of nicknames, caricatures, and so forth, and that in both cases the manufacturer was eager to accept that ridicue. Humiiation is a correate of humiity in the product, and activey embracing it suggests sefmortification, abstinence rather than poverty. Absence, the zero degree, is charged with meaning in such a context. The stripped-down Mode-T takes on an aura of miseriness, its uniform back finish no onger itera but a figurative ausion to Puritan sobriety and thrift. Abbreviated production time, simpified mechanica structure and unimaginative design--a come to impy that there is a secret hoard of unused "man-hours" and untried inteectua energies hed in reserve. The sheer voume of production is offered as dazzing proof that nothing has been ost, ony dispaced and mutipied ten-fod on another eve. A faacious argument, of course, the faacy of composition to be exact. But whie it confates eves and ignores important distinctions between the efficiency of one group and the accumuations of another, it is not without a certain superficia pausibiity. But the romance of quantity is as fragie as it is facie, especiay as the market becomes congested. Pubicity photographs of cars pouring off the assemby ine ike a Danaean shower of god, of ots crammed fu ike coffers, are noticeaby ess frequent when the rhetoric of communa penty gives way to that of personaized consumption. Stying and advertising strategies are therefore intimatey connected. But changes in the anguage of advertisements and

84 186 ELZABETH W. BRUSS pubicity (as we as changes in the way visua media are expoited) are the counterpart--and not the cause, as Barthes woud have it--of fuctuations in the manufactured siqn itsef. Revoutions in stye embrace the whoe automotive syst~m, even the purey "digita" ream of industria statistics, which turns from recording the voume of production and the tota hours of empoyment to registering the income from saes and the age and sex of car owners.3 Both in its standardization and subsequent proiferation of types, the automotive sign system shows the same'asymmetry of communication which is characteristic of a manufactured codes. Unike anguage, where the roes of sender and receiver are reversibe, the requisites of automobie manufacturing create a situation where access to the roe of sender is severey imited. As in eary stages of iteracy, encoding is entrusted to a sma group of speciaists.4 The automotive corporation, however, divides the responsibiity for sty'ipg among severa different departments, engineerinq, saes, production, and pubic reations, with the modes that finay appear refectinq a compromise between diverqent interests and kinds of expertise. Thus, even within a singe corporation, there are constraints on how esoteric a code can become. Then too, the probem of raising sufficient capita debars a but a few riva senders, so that there is no need to mystify the code itsef in the way that primitive writing systems were often mystified to protect scriba interests. Most importanty, the automotive siqn must be comprehensibe to a broad consuming pubic which acks speciaized training. Athough advertising and deaer networks work to some extent to seect "quaified" receivers, the highways reroain open to a and it is there that most consumer-dispays take pace, usuay at high speed. The buyer must be confident that his vehice can be interpreted quicky and unambiguousy. Even if he does not create the stye, he must be abe to use it, just as he must be abe to operate an engine the compexities of \.Yhich he may not fuy understand. Because it functions on more than one eve of exchange, the same automotive siqn has more than one vaue and may even have conficting vaues. An automobie has different impications when the context of use changes from "buying a car" to "buying a means of transportation," to "buying a consumer good," or to "producinq a com!nodi ty." t is the ast context w hich is r~evant to the manufacturer's cacuations, where production cost market probems, and investment potentia determine the ~acceptabe" transformations a vehice may undergo. The constituents of the automobie--"body she and frame" as opposed to "engine, transmission, and suspension systems"--are thus arbitrary and reative to these transformationa constrai~ts. Such divisions had their origin in the eariest phases of the 1ndustry, s

85 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE 187 when entrepreneurs estabished pants for assembing cars from machines and carriage-bodies which had been manufactured esewhere.s Larger companies graduay acquired their own "subassembies" and body-works in order to assure rapid deivery and guaranteed inventory, but they retained the former body/ engine distinction in order to save the cost of buiding new pants and production machinery. Contemporary American automotive corporations use the distinctions to shunt the cost and risk of technoogica innovations onto their suppiers, whie they concentrate on assemby methods, marketing techniques, and "package design" as those areas which are most reevant to the immediate constituents of the car. The structure of the code is reinforced by the oigopoistic nature of the industry: For a firm faced by oigopoistic rivas, stying changes are not ony a way of quickening the repacement pace of its "oya" customers, but aso a way of attracting ne,., customers away from other firms. With price competition argey eschewed, rivary for customers has centered on a stying strategy rather than on a major technoogica strategy for the same reasons as were advanced for a stying strategy to speed repacement. Detroit is basicay convinced that it is stying that ses cars.6 The articuation of the automobie into stabe and unstahe components is a matter of economic strategy; components are increasingy variabe as they approach the outer surface and maximum pubic visibiity, so that externa stampings change every year, whie the body she changes ony every two to three years. (One coud specuate a bit on the way a tacit anatomica anaogy to skeeton, muscuature, and skin has worked its way into purey economic considerations, however.) The other major expression of this strategy is dividing the continuum of a possibe automobies into discrete "makes" (Ford, Hercury, and ~incon; Chevroet, Pontiac, Buick, Odsmobie, and Cadiac), which in turn divide into "ines" (Chevette, Chevy, Camaro, Chevee, Corvette, and fu-sized Chevroet) which themseves subdivide into "series" (Chevroet Biscayne or Ford LTD) with a number of "modes" avaiabe in each series (two o four-door sedans, hardtops, station wagons, and convertibes.) 7 The vertica hierarchy of "makes"--"a car for every purse," to quote Afred Soan--compicates but aso coincides with the cateqorica distinction between body and engine, stying and technoogy, cited above, since the same body-she (minus exterior sheetmeta work and interior fittings) may serve more than one make if each car is fitted with a different power train. Thus body

86 188 ELZABETH W. BRUSS is opposed to engine as the sign of "novety," the site of diachronic variations in the code, whie the power pant reqisters the more or ess changeess distinctions of make which continue from year to year. Even with periodic "upgradings" of ess expensive makes by means of new "uxury" modes or onger whee-bases or added "optiona" equipment mimicking the standard features of a neighboring, higher make, the ranking system as a whoe continues undisturbed, with new features- continuay emerging to protect against erosion of the paradigm. Stying contributes to make recognition, but in itsef is too fuid, too easiy copied to sustain permanent distinctions. ndeed, impermanence and dupicity are the essence of automotive stying, which must be the "sweet cheat" of unquenchabe desire and an instigator of endess "carpe diem." The scae of makes heps to rationaize a scae of prices which are no onger commensurate with the cost of production.8 ("Lines" and "series" were ate additions to the system, deiberatey ambiguating estabished categories in the hope that a buyer entering a deaer's showroom in search of an inexpensive car might be convinced to purchase a mode costing as much or more than a "more expensive" make.) There is a stronger correation between a vehice's power and its price, athough even this is far from exact. "High performance" modes are generay the most expensive vehice in any ine or series, but other variabes such as driving comfort and interior appointments aso enter into the equation. Power is never naked; it contributes to the identity of a make much as an abstract distinctive feature contributes to the identity of a phoneme. But more important, in the automotive system power is inherenty ambiguous. There are too many competing ways of measuring it-- from the number and the capacity of the cyinders, to the precision of the machining, to the ratio of potentia energy to the car's tota weight. Power may be cacuated in terms of speed--but is it speed in acceeration or average maximum speed?--or in terms of strength--but strength for cimbing his or strength for hauing oads? Too many ways to measure and none of them sufficient, for power is utimatey normative, the mana of the machine, irreducibe to quantitative measures outside the "ernie" framework of the code. There is no absoute transation, not even money, for automotive power. Power excites cupidity because it is a vaue that can ony be grasped within the circe or permutations and combinations which comprise the automotive system. Thus there is the massive power of the Cadiac and the festive power of the Corvette, the reiabe power of a Buick and the purey nomina power of a Hustang. Some common semantic feature can be factored out of this array, but it can ony be possessed and expressed in combination with other features. Power has been a distinguishing trait of automobies from the beginning, even before there was a system beyond t~e primitive opposition of cars to other vehices. The automob1e aone offered s

87 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE a way of turning one's own simpe gestures into a super-human capacity for movement. Unike the rairoad and the streetcar, the automobie did not carry passengers as inert objects, but instead refected and transformed the active energies and desires of an individua driver. Unike the horse, it offered no brute resistance to human inteigence, and with its machinery wehidden and its effortess operation, it gave a dream-ike quaity to movement acking in the bicyce. This much is impicit in the thrust of eary vehice improvements, which gave the driver powers of acceeration as we as braking, centraized the contros around a singe individua, and moved the steering coumn off distinctivey to one side, where one and ony one occupant coud use it.9 A series of such deveopments made automotive power ever more intimate, ever more the immediate expression of a private impuse. The singuar, "autobiographica" function of the automobie is evident even in the pacement of mirrors and windows; extensive rear and side windows fitted to the gaze of other occupants were ony an afterthought, and seem to be receding once again. The pecuiar impications of automotive power, as the code deveoped in the United States, hep to expain our otherwise mysterious resentment over a petty task ike shifting gears-- a resentment which contributed to the success of Henry Ford and his simpified panetary gear system. For Americans, the automobie was, and in great part remains, an index of the wi to power, brooking itte or no mediation. As Danie Boorstin puts it: ndividuaism--defined in terms of privatism, freedom of choice and the opportunity to extend one's contro over his physica and socia environment--was one of the important American core vaues that automobiity promised to preserve and enhance in a changing urbanindustria society.20 Long before annua stying changes and mode proiferation, then, the morphoogy and meaning of the automobie was arbitrary- even those features which we have come to see as entirey "functiona." The interna combustion engine, for exampe, was not inherenty superior to steam or eectricity, but it coud be more quicky adapted which made it the "natura" choice in an economy of intense, short-term competition.21 Tn addition, the gas engine offered greater power-for-weight, whie its rivas exceed in the ess meaningfu areas of ceaniness, sience, frugaity, and fexibiity.22 And the exposive principe, the simutaneous vioence and containment of interna cornbusion, was a better icon for the phenomenoogy of driving. Aong with being a semantic prime in the automotive code, however, power has its instrumenta 189

88 190 ELZABETH W. BRUSS vaues as we. The visibe drama of it was far greater than in the case of thrift or purity, far more amenabe to free pubicity based on races, endurance runs, even accidents, and a better source of advertising spectace. Moreover, speed can aso be interpreted in terms of savings--abeit '"i th ess obvious costs than mass production--creating vast, potentiay hedonic stores of reified time. n fact, just as the first stages of mass production invented a new mode of profit based on extensive voume rather than arge unit earninqs, so the economy of gas engines was aso a matter of voume. Ony heavy hauing or ong and frequent trips coud offset the high initia investment.23 The consequences of a commitment to voume in both production and operation became cear ony ater, when the market and the road were both threatened with congestion--but fe~ v consumers or manufacturers have abandoned an economic rationae so intimatey bound up with the basic semiotic cateqories of the svstem. Whie they occupy different eves of the communication matrix, consumers and manufacturers often converge in the v1ay they use the code, and particuary in how they rationaize their own participation. There is far ess evidence of convergence between other eves, between designers, engineers, factory workers, and deaers,, or b8twee~ this group as a whoe and the entrepreneurs invoved in seina an~ consuming. (For in an artificay acceerated market, the buyer specuates on resae vaue just as the manufacturer specuates on the future success of a design.) Between the sencer and the receiver- who re-transmits the code on another eve--there are those who "hande" it but do not use it for communication and those who first must transate it into their mm, quite different terms. Even the designer, who woud seem cosest of a to the meaninas he is making, differs markey in that he focuses on the token rather than the type, seeing the quaities of an individua form vhere those who use the code see ony the marker of more abstract vaues. Moreover, the speciaized roes of designer, engineer, etc., are posterior to the code.itsef. They refect and absorb its underying contradictions, reconciing motion and possession, profit and vaue, through a series of successive per.qutations not unike those mediations Levi-Strauss notes in successive renderings of a myth. From drawings to cay scae-modes to transparent pastic modes (offering a view of the interior), there ar~ inevitab~ tensions and discrepancies even in the design process itsef..4 Designing is a the more probematic because it reverses ~he hierarchy of the working vehice, with motive power and dr1ver on the inside thrusting out, and grants priority to ines and surfaces over inhabitabe voume. The inversion may, however,!!

89 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE 191 refect the ogic of desire, in which the automobie functions as both "ego idea" and "idea ego," merging on the one hand with the sef-as-subject and on the other with the sef-as-object. There is a spit between the wi-to-be at the root of driving and the wish-to-be-seen at the root of acquisition. The "stying strategy" expoits and exacerbates this spit, putting design ahead of engineering and icensing the designer to create phantasms in two dimensions. The competed design is sent to engineering, yet here it reverts to raw uninteigibiity which engineers must put into a new subcode with the capacity of cacuating weight distribution, points of structura stress and impediments to mobiity. Notorious cases of mistransation, of cars which ater prove unwiedy or too dangerous to drive, have occured and wi continue to occur as ong as engineering is constrained within the imits of anterior designs. n an index of recent papers by the Society of Automotive Engineers, for exampe, one finds a patchwork of ingenious remedies for emission, coision, and passenger discomfort--a the epicyces of an infexiabe system.25 There are further strains at ater stages where professiona subcodes must be transformed into dies and jigs or time-and-motion studies and production quotas--discrepancies invoving, at east, the reduction of workers and tieaers into units in the cacuus of production and distribution.26 t is ony by becoming consumers that empoyees may "speak" the code that otherwise speaks them. Henry Ford's high wages first made the roe change possibe, and simutaneousy extended the automobie market whie gaining a coaborative oyaty for the process of production. Members of an assemby-ine shoud, above a others, recognize the artificia constructions they have produced. Yet even they "naturaize" the sign by joining in the genera process of exchange: proof of affuence or highway dominance for money, as ong as one agrees to treat the imits of the system as the natura imits of aspiration. The trade is not incomprehensibe. The automobie system promises carity, a chance to formuate private fantasies and to communicate a vision of onesef to others, a secure environment of identity and meaning. t is therefore extremey we adapted to a word of brief and anonymous encounters, where identity is dispersed between widey separate work, home, and recreation--a word that mass private transportation originated. The "stying strategy" is more than the incidenta correate of mass car ownership; it is amost a compensation for it.. the automobie has diffused and eveed and stirred and homogenized a continentciviization. t has spread the freedom to trave among a casses and at the same time

90 192 ELZABETH W. BRUSS has heped to remove the very differences between parts of the country, between kinds of andscape which were once an incentive to trave. The automobie has brought farmers to the city and the city to the farm; it has siphoned city-dweers to the suburbs and has made suburbs more city-ike. More than any other device, it has been responsibe for transforming arge tracts of America and the dweing paces of most Americans into an environment neither urban nor rura, not propery to be caed andscape. Dominated by superhighways and motes and drive-ins and parking areas, much of America can now be caed motorscape.27 t is understandabe that such a cuture woud foster styistic differences to repace those it has ost, and why a hunger for-,.'lnaw -"inf~rmation.,, mightuaeveop where there is s_o much entropy in the entire socia system. Moreover, egaitarian principes in the cuture make it necessary to pay down and offset a semantics of cass hierarchy by a semantics of novety. t is worth notin~ that changing fashions, of dress or transportation, become prevaent ony after hierarchica distinctions and inherited ranks are threatened. Signs of status can ony "wear out 11 when there is no monopoy on who can acquire the:u1.28 What makes hierarchy toerabe in the automotive system is the fact that the code, as a whoe, is open to a. Credit buying makes it possibe to acquire status as a Cadiac owner far more easiy than in other cutura systems--of home-ownership, occupation, education, and so forth. But free access aso makes the code potentiay vacuous, and thus in need of constant--if superficia--reorganization. No doubt the stying strategy is narrow and manipuative, but the consumer is aware of what he is buying even if he fais to know its origin or consequences and fais to consider other codes for articuating other aspirations. A producer must aso know what he is seing, aside from whatever he knows about profit margins. f his notions of what a car is, how it works and what it means, become too distant from his buyers', he risks marketing what no one wi desire or recognize. Market research and advertisina are meant to bridge such gaps, but they too can misfire through-excessive sophistication. A provisiona contrast with another cutura system, such as etiquette, demonstrates some of the semiotic features pecuiar to automotive communication. As a code of acceptabe behavior,! s

91 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE 193 etiquette expresses certain patterns of reationship between interacting parties; the usua semantic primes are aqe, rank, kinship, and reative famiiarity. Automobies seem argey insensitive to such features of reationship or identity. n part this may refect the genera eveing of many of these distinctions in this century, but the acunae aso indicate how itte automotive communication has to do with socia integration of any kind--how great the roe of persona expression and how few the adjustments for conditions of interaction. Litte wonder that extrinsic traffic aws are needed. Distinctions of strength and speed which the code does mark wi not suffice, since they are frequenty vague and hyperboic--intoerabe if they reay functioned for determining riqht of way. (t is true that some power-markers, when combined with we-defined agressive driving tactics-- 11 tai-gating 11 or "revving up" at stopight--may win occasiona and informa road privieges.) The near autism of the system is especiay cear where marks of power become most expicit, in the "sporty cars" introduced in the mid-1960's, with sef-consciousy anima names ike Mustang and Cougar and entirey civiized Ford and Mercury enqines. t is not by chance that the industry designated these empty muscuar dispays as "persona cars."29 The automobie code is coser to etiquette in matters of formaity and informaity, since the distributions of modes does create a partia taxonomy of automotive situations and functions. A station wagon woud be as inappropriate at a grand ba as Cinderea's rags. At present, the sinqe ''unmarked.. mode appears to be the sedan, with its broad synchronic range of uses and its ong diachronic ongevity (since the 1930's when it gained pre-eminence over smaer cosed cars ike the coupe.)30 There is aso a reguar correation between the number of passenqers a car carries and its acceptabe uses, the unmarked case again beinq the same four to six-passenger "famiy car" which has figured as standard for over three decades. Before this, the code was more eveny distributed between arger touring cars and business or sport runabouts, whie more recent trends (post-1968) indicate that famiy cars are again osing ground to more speciaized vehices. A "fu-sized" automobie now frequenty seats five with eaborate attention to the needs of the front-seat passengers which is not dupicated for the "optiona" riders in the rear. The recent vogue in "andau" roofs, featuring ony a tiny simuacrum of a side-window, set ike a jewe in a mock-convertibe top--impinges on the hard-won visua equaity of the back-seat passengers. The amost unrecognizabe ausion to the carriagetrade in these roofs sti carries the vague sugqestion of eegance, whie aso making the car effectivey a tvo-seater--wi th the weight, size, and interior room of a sedan.

92 194 ELZABETH tv. BRUSS The functions and idea passenger groupings the code prescribes have no more than a margina reation to the chassis of the car. Once again it seems that we are faced with a set of argey expressive categories--modes which encode character rather than perform tasks, which serve to communicate one's habits and associations, to indicate that one is a "famiy man," or has pastora rather than urban aspirations. Thus the exhuberance of a convertibe is ess a matter of its warm weather settings than its opposition to other, permanenty cosed vehices which cannot risk seasona and occasiona speciaization. The distribution of the code expains why some of those who once bought convertibes now are buying campers and vans, despite superficia differences in form. To own a speciaized "recreation vehice" is to procaim the same functiona insouciance, a ike capacity to dispense with the safety of an "a-purpose" car. Automobies therefore provide a way of forming and pubicizing a version of identity, without rea socia interaction or communa responsibiity. (The socia duties bound up in acquiring the purchase price are buried in the iusion of free consumer choice.) The narcissistic and a-socia impications of the automobie code as a whoe distinguish it from architecture, which it might otherwise resembe. The automobie estabishes a point midway between territoria and persona space, a pace that is at once portabe and fitted with fixed was. But ther~ is no "high art" of automobie design to parae that of architecture. Privatism, the rhetoric (if not the reaity) of functionaism, and rapid cornmercia devauation bar eisurey contempation of the automobie as an artifact vauabe in itsef. Yet since automotive stye aso defies a purey practica interpretation, it is difficut to reduce cars to simpe toos of transportation. n fact, the system shuns any itera identification with abor: the truck, the jeep, the ambuance, and even the poice car are kept segregated from private vehices. Even cothing, which has the same strict separation between occupationa uniforms and private garments, is more mundaney instrumenta. Cothes are tied to times of day and to the change of seasons--ties which are no doubt arbitrary enough, but nonetheess articuated. The strong gender demarcation which, unti recenty, has been the most striking feature of the garment system is extremey weak in automobies. The neutraity may be ony superficia, since one singes out "women drivers" and "teenaqe motorists" in contrast to the unnamed norm of mae aduts. The apochrypha first owner of any.desirabe~u~ed car is a timid spinster who never dared to face the fu sexua impications of her machine. Automobie stying is either mascuine or non-mascuine, the. atter cass embracing femininity, immaturity, and a who reect, 1

93 MANUFACTURED SGNS: SEHOTCS AND THE AUTOMOBLE 195 ignore, or prove unabe to afford the standard automobie and its associated vaues. Consider how recent compact and subcompact cars stand opposed to the standard both in their visibe mascuarity and their usua casses of owners. To divide the automotive code into mascuine and feminine woud eiminate too many potentia buyers, but more important, a feminine car woud be an oxymoron in a cuture which inks power to mascuinity and vaues automobies principay for their power. t is tempting to suppose that the diminutive proportions of the compact naturay emascuate it. But a brief qance at automobie history shows that this is arbitrary as we, since formery it was the arger cars which were associated with women drivers and passengers.31 As in any true semiotic system, the meaning of the vehice is mediated hy the rues and categories the code makes avaiabe. One must know the fu array of producers, makes, and modes to determine the vaue of an individua car. And size itsef is reative, since what at one stage is "compact" may become the standard at another. But the mediation offered by the automotive code remains far from the compex grammar and morphoogy of a anguage. Truy discrete and arbitrary categories are fevt, rues of exponence are highy inexact, and the pm,1er to anayze, re-combine, and expand constituents amost nonexistent. The abstract categories of make, ine, series, and--to a esser extent--mode are aone discontinuous. The visua impact of a car is goba, on the other hand, athough certain features such as cab, front and rear projections, and various subdivisions ike doors, roof, cow and trunk, gri, fenders, front and rear bumpers can be isoated on an ad hoc basis. But these are reative distinctions, inferred from the overa design, and it is often extremey difficut to te where to dravr the ine between running board and fender or to say when a curve is extreme enough to create a separate back deck rather than the rear end of the cab. True, each automobie is iteray an assembage of parts, but since the stying strategy's inception there has been a persistent trend towards cars that present a singe, unbroken fow of body design to the eye. ncised ines traverse the entire side of a car from front to end, for exampe, or head ights graduay merge into the fenders and gri-work. ~he fuid contours mask an anachronistic assemby technoogy behind an imitation of modern extruded forms.32 The mask is aso a repression of whatever guity knowedge one has of the tedious abor of assemby, and makes the artifact a more innocent object of desire. n reading the vehice, the outer perimeter serves as a frame deimiting the fied of attention and orienting perception, aowing one to measure the indentations or nrotuberances, reative degrees of ight or shadow, and to judge the overa

94 196 ELZABETH W. BRUSS distribution of mass.33 By such attention, refection, and inference, the features which identify the car graduay emerge. But these features may have itte reation to the way the car was partitioned for assemby. One seizes instead. on partia, evocative detais and discontinouous eements: a prominent hood ornament and a particuar ange in the roof which together indicate "Thunderbird." Thus unike anguage, the minima distinctive units, the "morpho-phonoogy," if you wi, of the automobie system are ony reative, athough discriminations between makes, etc., at a higher eve of anaysis are absoute. Transformationa grammar has shown us that in anguage, too, the deepest grammatica reations may be but dimy and dis- continousy marked at the surface eve, but the automobie code has none of the precise mapping functions for reating deep and surface structure which one can construct for natura anguages. The comparison is most vauabe for what it reveas about the distance between a manufactured sign system and a fu anguage. f the primitive grammar of the automotive code is a mixture of motivated and unmotivated eements, digita and anaog, the merger is even more pronounced when we turn to semantics. Arbitrary paradigms constrain but do not entirey contro the semantic interpretation of a vehice. One reads W a car indexicay and iconicay as we. The fu vaue of ~ stying is usuay not apparent uness one treats it as an ausion to something ese--to other artifacts ike carriages or missies, for exam~e. t is through iconic vaues that the W code most frequenty expresses novety and innovation. The ~ resembance invoked is doubtess artificia and schematic, but it does te us something about what counts as "new" in the cuture and when it has ost its novety, as was the case with both torpedoes and rockets after ony a year or two of automotive mimickery. Even schematic icons are interesting for nhat they refect about the identifying pecuiarities and criteria features of their originas. Thus automobie stying iustrates how jet propusion caught the popuar imagination in terms of a marked thrust from the rear and a burst of (metaic) fame-ike contrai. Transportation is not the ony source of inspiration--cars have resembed weapons, insects, or simpy other cars (in their roe as pure co.'ttro.odi ty.) The proiferation of a once popuar design (or rather, the design of a popuar car) is sometimes so starting as to suggest a kind of superstition among automobie manufacturers, a beief in sympathetic magic. There are aso cases of manufacturers imitating their own designs, harking back to an od favorite or eevating one make by impicity comparing it to another. Thus the Ford becomes a Thunderbird in the 1950's--within the safe imits of a simie--and a Ros Royce in Both consumers 1

95 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE 197 and manufacturers seem to exercise a "wiing suspension of disbeief" in matters of resembance, reaizing perhaps that the itera version of the icon woud invove risks and sacrifices neither coud afford. There is a trade-off: the consumer wi not chaenge the gross inadequacies of a "Rocket- 88" in return for which the producer wi induge him in his timid fantasies of fight. Creduity goes far deeper when it comes to indexica.. vaues. But then it is difficut to isoate the indicies pecuiar to automotive signs from very broad beiefs about sequence, cooccurrence, ogica and empirica necessity which one must hod to function propery in most aspects of the cuture. Whie particuar expressions of "the new" may be iconic, the yeary withering of each new automobie stye woud seem far more artificia without a prior matrix of annua cyces of new growth and decay. Manufacturers were quick to expoit this association by seeking a uniform introduction date for a new modes. The eary automobie shows took pace in the spring, but ater the introductions were moved up to the fa to induce a syncopated pattern of buying: pitting harvest ag;:~inst the reverdie. Buying and restying thus take pace within the protective penumbra of seasona necessity. But nature is not the ony or even the richest source of indexica interpretations. Cuture provides another kind of motivation for the automotive sign. Since every cuture paces restrictions on the array of possibe forms--restrictions which maintain inteigibiity--any departure from these is ikey to appear unnatura or impossibe. Literate cutures, which record diachronic variations in the array, foster a more critica attitude and make additiona arguments necessary if a form is to appear motivated. One of the most ancient and respectabe is the teeoogica argument. Not that automobie design ever seems "providentia," but there are certainy covert appeas to the humber teeoogy which fits an instrument to its use. The match between form and function is even enhanced when the instrument is man-made, since the purposiveness is conscious and human, therefore putativey more corrigibe. Yet the form which has remained the index of the automotive function is archaic, dating from the eariest experiments with interna combustion engines. The front-engine "syst~me Panhard" was introduced before the turn of the century, at a time when increasing the distance between the engine and the whees by a ongitudina crankshaft, rather than seating the engine beneath the seat and directy above the whees, represented a significant increment in power. The change made a dramatic difference in the car's overa appearance, turning it from a box-ike affair to a predominanty horizonta shape with a sunken cab. Athough it was at first ony the consequence

96 198 ELZABETH W. BRUSS of interna rearrangements, the new outine soon came to signify "high performance" in itsef.34 So strong was the association that manufacturers ike Henry Ford were even moved to insta dummy hoods on cars that retained the od transmission system.35 Obviousy there have been aternatives to the systeme Panhard since 1891, but the ong, ow sihouette has retained its symptomatic vaue, overriding power systems as good or better. Departures do occur, as in the thirties, when aerodynamic curves on a higher sihouette vied for approva. But after each departure, the Panhard has returned with.redoubed strength, the outine even ower, the profie even onger. Thus the forties saw car foorboards sunk around and under the crankshaft and trunks turned into back decks extending as far behind the car as the hood extended in front.36 The reason that the "air-fow" modes did not sustain pubic confidence may be that an untrained pubic coud not interpret them propery. Aerodynamic features seem to have been read iconicay, as an imitation of the burnished surfaces and windbown contours produced by high acceeration rather than as a means for overcoming wind-resistance. The ogica outcome of this was the exaggerated and fancifu taifins of the ate 1950's, features so purey mimetic that the functiona aerodynamic threat they presented (some cars actuay ifting from the ground on a sharp turn) was i-perceived. One coud aso specuate, on the other hand, that the faiure of the origina air-fow designs was the resut of strong but inappropriate iconic vaues; that the broad and vouptuousy rounded body had too much in it reminiscent of a Rubens nude. Or perhaps it was what the design eft out--the occasiona sharpened edges and overa atera thrust of a projectie which comprise a covert iconography of danger that the anguage of advertising and pubicity woud not dare to make expicit. There are genuine indexica readings of automobie structure and design, but these are rarey part of the code itsef. For exampe, stying has a direct correation with drafting techniques, and one can see that the contours of automobies changed markedy after the introduction of new procedures for manipuating and specifying more compex curves.37 Advances in hoography and computer graphics, which permit new surface and vom~e projections, cannot but have as great an impact on future automobie designs. But to read these causes in their effects requires specia information or training. The ordinary consumer might see a difference, but the appearance of innovation woud be more reevant to his semiotic praxis than the remote and (for him) useess techniques that produced it. Despite the arbitrary nature of the scae of makes, the tendency to treat cars as natura symptoms of weath, cutivation,!

97 MANUFACTURED SGNS: SEMOTCS AND THE AUTOMOBLE 199 and authority is widespread. The ogic of this is fauty, since even weath is neither necessary nor sufficient to owning a Lincon Continenta. Rapid and transient dispays make the system as a whoe more credibe, however, since there is ordinariy itte opportunity for disproving prima facie impressions. Moreover, the automobie itsef is not a statement. Unabe by virtue of its syntax to predicate, it cannot reay ie. The "fascination" of the artifact is in its goba appea, with no way of isoating topic from comment or argument from attribute. There is none of the descriptive precision of a anguage, but aso none of the vunerabiity to objection and reproof. t is difficut to compain about the pretensions of a particuar bumper or demonstrate that a certain wheebase is a fase promise of power.38 What criteria coud one invoke, what verification procedures or feicity conditions? The vehice as a whoe is the sign, and its symptomatic vaue is bound up with its position vis a vis every other car within the svstem--to reject the bandishments of one (even of a part of one, if it coud be isoated) is to reject the entire system. Moreover, the process of interpretation is indirect; the driver of a Lincon Continenta simpy eaves it to his observers to formuate their own propositions about "the sort of peope who drive a car ike that." t is they who ascribe properties and infer from what the system entais in the way of monetary vaues and contexts of use to what is not stricty entaied. The weter of ausions, fugitive resembances and causes which figure in the stying of a car, aso bearing on this indirect process, make interpretations even more difficut to specify and criticize. Despite--or perhaps in part because of--the ack of precision, iconic and indexica vaues are important to the a'.tomobie code. They hep to make its rapid changes more readiy accessibe. And in a system committed to continuous but superficia novety, they hep to guard against the appearance of stagnation, on the one hand, or triviaity, on the other. Even the deeper distinctions and omissions at the heart of the code come to appear ess arbitrary, more famiiar or functiona- "masquer a nature systematique et semantique des enonces. en transformant 'equivaence en raison," as Barthes remarks.9 Yet it is not ony consumers who come to see the code as natura. Once fixed, a manufactured code ceases to be arbitrary for its originators, and is instead the index and soe protection of a certain organization of abor and capita. Radica innovations in either method or product become "unthinkabe" if corporate identity is to be preserved. The assemby technoogy, marketing networks, and the type and turnover of vehice desiqn which characterize the automobie industry are mutuay reinforcing, each seeming efficient and necessary in terms of the others.