The Model of the Text: Meaningful Action Considered as a Text Author(s): Paul Ricœur Source: New Literary History, Vol. 5, No. 1, What Is Literature? (Autumn, 1973), pp. 91-117 Published by: The Johns Hopkins University Press Stable URL: http://www.jstor.org/stable/468410 . Accessed: 31/08/2014 18:13 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. . The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to New Literary History. http://www.jstor.org This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions The Model oftheText: MeaningfulAction Consideredas a Text PaulRicoeur MY AIM IN THIS PAPER will be to testan hypothesis.I assume thatthe primarysenseof the word "hermeneutics"concerns the rulesrequiredforthe interpretation of the writtendocumentsof our culture. In assumingthis startingpoint I am remaining faithfulto the conceptofAuslegungas it was statedby WilhelmDilthey; whereas Verstehen(understanding,comprehension)relieson the recognitionof what a foreignsubject means or intendson the basis of all kindsof signsin whichpsychiclifeexpressesitself(Lebensiiusserungen), Auslegung(interpretation, exegesis)impliessomethingmorespecific: it a coversonly limitedcategoryof signs,thosewhichare fixedby writing, includingall the sortsof documentsand monumentswhich entail a fixationsimilarto writing. Now my hypothesisis this: if thereare specificproblemswhich are raised by the interpretationof texts because they are texts and not spoken language, and if these problemsare the ones which constitute hermeneuticsas such, then the social sciencesmay be said to be hermeneutical (I) inasmuchas theirobject displayssome of the features constitutive of a textas text,and (2) inasmuch as theirmethodology the same kind of proceduresas those of Auslegung or textdevelops interpretation. Hence the two questionsto which my paper will be devoted: (i) To what extentmay we considerthe notionof textas a good paradigmfor the so-calledobject of the social sciences? (2) To what extentmay we as a paradigmforinterpretause the methodologyof text-interpretation tion in generalin the fieldof the social sciences? I. The Paradigm of Text In order to justifythe distinctionbetweenspoken and writtenlanguage I want to introducea preliminaryconcept,that of discourse. It This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions 92 NEW LITERARY HISTORY is as discoursethat language is eitherspoken or written.Now, what is discourse? We shall not seek the answerfromthe logicians,not even fromthe exponentsof linguisticanalysis,but fromthe linguiststhemselves.Discourse is the counterpartof what linguistscall language systemsor linguisticcodes. Discourse is language-eventor linguisticusage. This pair of correlativeterms-system/event, code/message-has played a basic role in linguisticssince it was introducedby Ferdinand de Saussure and Louis Hjelmslev.The firstspokeoflanguage (langue)-speech (parole), thesecondofschema-usage. We can also add competenceperformancein Chomsky'slanguage. It is necessaryto draw all the epistemologicalconsequencesof such a duality,namely,that the linrulesfromthe linguisticsof language. guisticsof discoursehas different It is the French linguistEmile Benv6nistewho has gone furthestwith thisdistinction.For him, thesetwo linguistics are not constructedupon the same units. If the sign (phonologicalor lexical) is the basic unit of language,the sentenceis the basic unit of discourse.Thereforeit is the linguisticsof the sentencewhich supportthe theoryof speech as an event. I will retainfourtraitsfromthislinguistics of the sentencewhich will help me in a littlewhile to elaboratethe hermeneuticof the event and of discourse. Firsttrait: Discourseis always realizedtemporallyand in a present, whereasthe languagesystemis virtualand outsideof time. Emile Benv6nistecalls thisthe "instanceof discourse." Second trait: Whereaslanguagelacksa subject-in thesensethatthe question"Who is speaking?"does not apply at itslevel-discourse refers to itsspeakerbymeans ofa complexsetofindicatorssuch as thepersonal pronouns.We will say thatthe "instanceof discourse"is self-referential. Third trait: Whereas the signsin language referonly to othersigns withinthe same system,and whereaslanguage thereforelacks a world just as it lacks temporalityand subjectivity,discourseis always about something.It refersto a world which it claims to describe,to express, or to represent.It is in discoursethatthe symbolicfunctionof language is actualized. Fourthtrait: Whereaslanguageis onlythe conditionforcommunication,forwhich it providesthe codes, it is in discoursethat all messages are exchanged. In thissense,discoursealone has not onlya world,but an other-another person,an interlocutor to whom it is addressed. These fourtraitstaken togetherconstitutespeech as an event. It is remarkablethat these four traitsappear only in the movementof effectuationfromlanguage to discourse. Every apologyforspeech as an This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 93 is significant event,therefore, if,and onlyif,it makesvisiblethe effectuation by which our linguisticcompetenceactualizes itselfin performance. But the same apology becomes abusive as soon as this eventwhere it is valid, to undercharacteris extended from effectuation, standing.What is it to understanda discourse? thesefour traitsare actualized in spoken Let us see how differently and writtenlanguage: (I) Discourse,as we said, existsonly as a temporaland presentinin living speech and in stance. This firsttrait is realized differently writing.In livingspeech,the instanceof discoursehas the characterof a fleetingevent, an event that appears and disappears.That is why thereis a problemof fixation,of inscription.What we want to fix is what disappears. If, by extension,we can say thatone fixeslanguagesyntacticalinscriptioninscriptionof the alphabet,lexical inscription, it is for the sake of that which alone has to be fixed,discourse. Only discourseis to be fixed,because discoursedisappears.The atemporal systemneitherappears nor disappears;it does not happen. Here is the place to recall the mythin Plato's Phaedo. Writingwas given to men to "come to therescue"ofthe "weaknessof discourse,"a weaknesswhich was that of the event. The giftof the grammata-of that "external" thing,of those "externalmarks,"of that materializingalienation-was just that of a "remedy"broughtto our memory.The Egyptiankingof Thebes could well respondto the god Theuth that writingwas a false remedyin that it replaced true reminiscenceby materialconservation and real wisdom by the semblance of knowing.This inscription,in spiteof its perils,is discourse'sdestination.What does writingfix? Not the eventof speaking,but the "said" of speakingwherewe understand by the said that intentionalexteriorizationconstitutiveof the aim of discourse thanks to which the sagen-the saying-wants to become Aus-sage-the enunciation,the enunciated. In short,what we write, what we inscribe,is the noema of the speaking. It is the meaning of the speech event,not the eventas event. What does writingfix? If it is not the speechevent,it is speech itself in so faras it is said. But whatis said? Here I would like to propose that hermeneuticshas to appeal not only to linguistics(linguisticsof discoursevs. linguisticsof language) as it does above, but also to the theoryof the speech act such as we findit in Austin and Searle. The act of speaking,accordingto these authors,is constitutedby a hierarchyof subordinateacts which are distributedon threelevels: (I) the level of the locutionaryor propositional act, the act of saying; (2) the level of the illocutionaryact or This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions 94 NEWLITERARY HISTORY force,that which we do in saying; and (3) the level of the perlocutionaryact, that which we do by saying.When I tell you to close the door, forexample, "Close the door!" is the act of speaking. But when I tell you this with the forceof an order and not of a request,this is the illocutionaryact. Finally,I can stirup certaineffects, like fear,by the fact that I give you an order.These effectsmake my discourseact like a stimulusproducingcertainresults.This is the perlocutionary act. What is the implicationof these distinctions for our problemof the intentionalexteriorization by whichtheeventsurpassesitselfin meaning and lends itselfto material fixation?The locutionaryact exteriorizes itselfin the sentence. The sentencecan be identifiedand reidentified as being the same sentence.A sentencebecomes an enunciation(Austo othersas beingsuch and such a sentence sage) and thusis transferred with such and such a meaning. But the illocutionaryact can also be exteriorizedin grammaticalparadigms (indicative, imperative,and subjunctivemodes, and otherproceduresexpressiveof the illocutionary and reidentification. force) which permitits identification Certainly, in spoken discourse,the illocutionaryforce leans upon mimicryand gesturalelementsand upon thenonarticulatedaspectsof discourse,what we call prosody. In thissense,the illocutionary forceis less completely inscribedin grammarthan is the propositionalmeaning. In everycase, its inscriptionin a syntacticarticulationis itselfgatheredup in specific paradigmswhichin principlemake possiblefixatiodby writing.Without a doubt we must concede that the perlocutionaryact is the least inscribable aspect of discourseand that by preferenceit characterizes action is preciselywhat is the spokenlanguage. But the perlocutionary least discursivein discourse.It is the discourseas stimulus.It acts, not by my interlocutor's recognitionof my intention,but energetically, by directinfluenceupon the emotionsand the affectivedispositions.Thus the propositionalact, the illocutionaryforce,and the perlocutionary action are apt, in a decreasingorder,forthe intentionalexteriorization which makes inscriptionin writingpossible. Thereforeit is necessaryto understandby the meaningof the speechact, or by the noema of thesaying,not onlythe sentence,in the narrow sense of the propositionalact, but also the illocutionaryforceand even action in the measurethat thesethreeaspectsof the the perlocutionary are codified, gatheredinto paradigmswhere, consequently, speech-act and be identified can reidentifiedas having the same meaning. they ThereforeI am here givingthe word "meaning" a verylarge acceptationwhich coversall the aspectsand levelsof theintentionalexteriorization which makes the inscriptionof discoursepossible. This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 95 The destinyof the three other traitsof discoursein passing from discourseinto writingpermitsus to make more precisethe meaning of thiselevationof sayingto what is said. trait (2) In discourse,we said-and thiswas the second differential of discoursein relationto language--thesentencedesignatesits speaker and personality.In spokendiscourse, by diverseindicatorsof subjectivity this referenceby discourseto the speakingsubject presentsa character of immediacythatwe can explainin the followingway. The subjective intentionof thespeakingsubjectand themeaningof the discourseoverlap each other in such a way that it is the same thingto understand what the speakermeans and what his discoursemeans. The ambiguity of the French expressionvouloir dire, the German meinen, and the Englishto mean atteststo thisoverlapping.It is almostthe same thing to ask "What do you mean?" and "What does that mean?" With writtendiscourse,the author's intentionand the meaningof the text cease to coincide. This dissociationof the verbal meaningof the text and the mental intentionis what is really at stake in the inscription of discourse.Not thatwe can conceiveof a textwithoutan author; the tie betweenthespeakerand thediscourseis not abolished,but distended and complicated.The dissociationof the meaningand the intentionis stillan adventureof the referenceof discourseto the speakingsubject. But the text'scareer escapes the finitehorizonof its author. What the text says now mattersmore than what the author meant to say, and every exegesis unfolds its procedureswithin the circumferenceof a meaning that has brokenits mooringsto the psychologyof its author. Using Plato's expressionagain, writtendiscoursecannotbe "rescued"by all the processesby which spoken discoursesupportsitselfin order to be undertood-intonation,delivery,mimicry,gestures. In this sense, the inscriptionin "external marks," which firstappeared to alienate of discourse. Henceforth,only discourse,marksthe actual spirituality the meaning "rescues" the meaning, withoutthe contributionof the physicaland psychologicalpresenceof the author. But to say that the is the meaning rescuesthe meaning is to say that only interpretation the of which its can for weakness discourse author no "remedy" longer '"save.5" (3) The eventis surpassedby the meaninga thirdtime. Discourse, we said, is what refersto the world,to a world. In spoken discourse this means that what the dialogue ultimatelyrefersto,is the situation This situationin a way surroundsthe common to the interlocutors. dialogue, and its landmarks can all be shown by a gesture,or by pointinga finger,or designatedin an ostensivemannerby the discourse This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions 96 NEW LITERARY HISTORY itselfthroughthe oblique referenceof those other indicatorswhich are the demonstratives, the adverbs of time and place, and the tense of the verb. In oral discourse,we are saying,referenceis ostensive. What happens to it in writtendiscourse?Are we sayingthat the text no longerhas a reference?This would be to confoundreferenceand world and situation. Discourse cannot fail to be about demonstration, In something. sayingthis,I am separatingmyselffromany ideology of an absolutetext. Only a few sophisticatedtextssatisfythis ideal of a textwithoutreference.They are textswherethe play of the signifier breaksaway fromthe signified.But this new formis valuable only as an exceptionand cannot give the key to all othertextswhich in one manner or anotherspeak about the world. But what, then,is the subject of textswhen nothingcan be shown? Far fromsayingthatthe text is then withouta world,I will now say withoutparadox thatonlyman has a world and not just a situation. In the same manner that the text freesits meaning from the tutelage of the mental intention,it freesits referencefromthe limitsof ostensivereference.For us, the world is the ensembleof referencesopened up by the texts.Thus we speak about the "world" of Greece, not to designateany more what were the situationsfor those who lived them, but to designate the nonsituationalreferenceswhich outlive the effacementof the firstand which henceforthare offeredas possiblemodes of being, as symbolic of all For me, thisis the referent dimensionsof our being-in-the-world. of references ostensive Umwelt of the the no dialogue, literature; longer of everytextthat but the Welt projectedby the nonostensivereferences we have read, understood,and loved. To understanda text is at the same time to lightup our own situation,or, if you will, to interpolate which make among the predicatesof our situationall the significations a Weltof our Umwelt. It is thisenlargingof the Umweltintothe World which permitsus to speak of the referencesopened up by the text-it would be betterto say thatthereferences open up theworld. Here again itself of discoursemanifests the spirituality throughwriting,which frees and limitationof situationsby openingup a world us fromthe visibility for us, that is, new dimensionsof our being-in-the-world. In thissense,Heideggerrightlysays-in his analysisof verstehenin Being and Time-that what we understandfirstin a discourseis not anotherperson,but aiproject,thatis, the outlineof a new being-in-theworld. Only writing,in freeingitself,not only fromits author,but fromthe narrownessof the dialogical situation,revealsthisdestination of discourseas projectinga world. In thus tyingreferenceto the projectionof a world, it is not only This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 97 Heideggerwhom we rediscover,but Wilhelmvon Humbolt,forwhom the great justification of language is to establishthe relationof man to the world. If we suppressthis referentialfunction,only an absurd remains. game of errantsignifiers of (4) But it is perhapswiththefourthtraitthatthe accomplishment discoursein writingis most exemplary. Only discourse,not language, is addressedto someone.This is the foundationof communication.But it is one thingfor discourseto be addressedto an interlocutorequally presentin the discoursesituation,and another to be addressed,as is the case in virtuallyeverypiece of writing,to whoeverknowshow to read. The narrownessof the dialogical relation explodes. Instead of being addressedjust to you, the second person,what is writtenis addressed to the audience that itself creates. This, again, marks the spiritualityof writing,the counterpartof its materialityand of the alienationwhichit imposesupon discourse. The vis-a-visof the written is just whoeverknows how to read. The copresenceof subjects in a dialogue ceases to be the model for every"understanding."The relation writing-readingceases to be a particular case of the relation speaking-hearing.But at the same time,discourseis revealed as discourse in the universalityof its address. In escaping the momentary characterof the event-the bounds lived by the authorand the narrowness of ostensivereference-discourseescapes the limitsof being face to face. It is no longera visibleauditor. An unknown,invisiblereader has become the unprivilegedaddresseeof the discourse. To what extentmay we say thatthe object of the social sciencesconformsto the paradigm of the text? Max Weber definesthis object as sinnhaftorientiertesVerhalten,as "meaningfullyoriented behavior." To what extentmay we replace the predicate"meaningfullyoriented" derived from the by what I would like to call readability-characters four criteriaof our to of the text? Let us try apply precedingtheory what a text is to the concept of meaningfulaction. a. The FixationofAction Meaningfulaction is an object forscience only under the condition which is equivalentto the fixation'of a disof a kind of objectification course by writing.This traitpresupposesa simpleway of being meaningfulwhich is similarto the dialogical situationas regardslanguage. Meaningfulaction may be graspedand understoodwithinthe process of interaction,which is quite similarto the processof interlocutionin This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions 98 NEW LITERARY HISTORY the fieldof discourse. It is at thisstrategiclevel that the so-calledphithinkers. losophy of actions operates among post-Wittgensteinian G. E. M. Anscombein Intention,A. I. Melden in Free Action, and Richard Taylor in Action and Purpose require no other conceptual frameworkfor theirtheoryof action than the one which is at work in ordinarylanguage. Science is another "language game" based on semanticrules. It is one thingto speak of actions,purquite different poses,motives,agentsand theiragency,and it is somethingelse to speak of movementsas happening,of mental events (if thereare any), or of physical or mental causes. The duality of linguisticgames, that of ordinarylanguage and that of the behavioral and the social sciences, is inseparable. As is known,the main discrepancybetweenboth lanof motive,conceivedas "reason guage games concernsthe irreducibility in to cause Humean terms as an antecedentevent for," interpreted distinct and linked from, to, itsconsequent.But is logically contingently it truethat a scientificapproach mustnecessarilyexclude the character of meaningfulness and that ordinarylanguage alone preservesit? Is therenot a scientificlanguage for which action would be both "objective" and "meaningful"? The comparisonbetweeninterlocutionand interactionmay help us at thisstageof our analysis. In the same way thatinterlocution is overcome in writing,interactionis overcome in numeroussituationsin which we treataction as a fixedtext.These situationsare overlooked in a theoryof action forwhich the discourseof action is itselfa part of the situationof transactionwhich flowsfrom one agent to another, exactly as spoken language is caught in the process of interlocution, or, if we may use the term,of translocution.This is why the underlevel is only "knowledgewithout standingof action at the prescientific observation,"or as G. E. M. Anscombe says, "practical knowledge" in the sense of "knowinghow" as opposed to "knowingthat." But this in the strongsense which understandingis not yet an interpretation deservesto be called scientificinterpretation. My claim is that action itself,action as meaningful,may become an object of science, withoutlosing its characterof meaningfulness, by similarto the fixationwhich occurs virtueof a kind of objectification in writing.By thisobjectification, action is no longera transactionto which the discourseof action would stillbelong. It constitutesa delineatedpatternwhich has to be interpreted accordingto its innerconnections. This objectificationr is made possibleby someinnertraitsof the action which are similarto the structureof the speech act and which make This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 99 "doing" a kind of utterance. In the same way fixationby writingis made possibleby a dialectic of intentionalexteriorization inherentin the speech-actitself,a similardialecticwithinthe processof transaction preparesthe detachmentof the meaningof the action fromthe event of the action. First,an actionhas thestructureof a locutionaryact. It has a propositional contentwhich can be identifiedand reidentified.This "propositional" structureof the action has been clearly and demonstratively expounded by AntonyKenny in Action,Emotion and Will. The verbs of action constitutea specificcomplexof predicateswhich are similar to relationsand which,like relations,are irreducibleto all the kinds of predicateswhich may followthe copula "is." The class of action predicates in its turnis irreducibleto the relationsand constitutesa specific setof predicates.Amongothertraits,theverbsofactionallow a plurality of "arguments"capable of complementingthe verb, rangingfromno argument (Plato taught) to an indeterminatenumber of arguments (Brutus killedCaesar, in the Curia, on the Ides of March, with a ..., with the help of. . . .). This variable polydicityof the predicative structureof action-sentencesis typical of the propositionalstructure of action. Anothertrait which is importantfor the transpositionof the concept of fixationfrom the sphere of discourse to the sphere of action concernsthe ontologicalstatusof the "complements"of the verbsof action. Whereas relationshold betweentermsequally existing (or nonexisting),certainverbsof action have a topical subject which is identifiedas existingand to which the sentencerefers,and complementsof which do not exist. Such is the case with the "mental acts" (to believe,to think,to will, to imagine,etc.). AntonyKenny describessome othertraitsof the propositionalstructure of actions derivedfromthe descriptionof the functioningof the verb of action. For example,the distinctionbetweenstates,activities, and otherperformances can be statedaccordingto the behaviorof the tensesof the verbsof action which fixsome specifictemporaltraitsof the action itself.The distinctionbetweenthe formaland the material betweenthe notionof all object of an action (let us say the difference inflammablethingsand this letterwhich I am now burning) belongs to the logic of action as mirroredin the grammarof the verbs of action. Such, roughlydescribed,is the propositionalcontentof action which gives a basis to a dialecticof event and meaningsimilarto that of the speech-act. I should like to speak here of the noematicstructure of action. It is the noematicstructurewhichmay be fixedand detached fromthe processof interactionand become an object to interpret. This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions I00 NEW LITERARY HISTORY Moreover,thisnoema has not only a propositionalcontent,but also presents"illocutionary"traitsvery similar to those of the complete classes of performativeacts of discoursedespeech act. The different scribedby Austinat the end of How to do Thingswith Wordsmay be taken as paradigmsnot onlyforthe speech acts themselves, but forthe A actionswhichfulfillthe corresponding acts. typologyof action, speech the model of therefore is acts, illocutionary following possible. Not only a typology, but a criteriology, inasmuchas each typeimpliesrules,more rules" which,accordingto Searle in Speech-Acts, precisely"constitutive allow the constructionof "ideal models" similarto the ideal typesof Max Weber. For example,to understandwhat a promiseis, we have to understandwhat the "essentialcondition"is accordingto which a given action "countsas" a promise.This "essentialcondition"of Searle is not far fromwhat Husserl called Sinngehalt,which coversboth the "matter" (propositionalcontent) and the "quality" (the illocutionary force). We may now say that an action,like a speech act, may be identified not only accordingto its propositionalcontent,but also according to its illocutionaryforce. Both constituteits "sense-content."Like the speech act, the action-event(if we may coin thisanalogical expression) developsa similardialecticbetweenits temporalstatusas an appearing and disappearingevent,and its logical statusas having such and such identifiablemeaning or "sense-content."But if the "sense-content"is what makes possiblethe "inscription"of the action-event, what makes it real? In otherwords,what correspondsto writingin the field of action? Let us returnto the paradigm of the speech-act.What is fixedby writing,we said, is the noema of the speaking,the sayingas said. To what extentmay we say that what is done is inscribed? Certainmetaphorsmay be helpfulat this point. We say that such and such event leftits mark on its time. We speak of markingevents. Are therenot "marks" on time,the kind of thingwhich calls for a reading,rather than for a hearing? But what is meant by this metaphorof the "imprintedmark"? The three other criteriaof the text will help us to make the natureof thisfixationmoreprecise. b. The Autonomizationof Action In the same way that a textis detached fromits author,an action is detached fromits agent and develops consequences of its own. This autonomizationof human action constitutesthe social dimensionof This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 101 action. An action is a social phenomenonnot only because it is done by several agentsin such a way that the role of each of them cannot be distinguished fromthe role of the others,but also because our deeds escape us and have effectswhichwe did not intend. One of themeanings of the notion of "inscription"appears here. The kind of distance which we found between the intentionof the speakerand the verbal meaning of a text occurs also between the agent and his action. It is this distancewhich makes the ascriptionof responsibility a specific We not do raised "Who smiled?" "Who hand?" The his ask, problem. doer is presentto his doing in the same way the speakeris presentto his speech. With simple actions like those which require no previous action in order to be done, the meaning (noema) and the intention (noesis) coincide or overlap. With complex actionssome segmentsare so remotefromthe initialsimplesegments,which can be said to express the intentionof the doer, that the ascriptionof theseactionsor actiona problemas difficult to solve as thatof authorship segmentsconstitutes in some cases of literarycriticism.The assignationof an authorbecomes a mediateinferencewell-knownto the historianwho triesto isolatethe role of an historicalcharacteron the course of events. We just used the expression"the course of events." Could we not say thatwhat we?call the courseof eventsplaysthe role of the material thingwhich "rescues" the vanishingdiscoursewhen it is written?As we said in a metaphoricalway, some actions are eventswhich imprint theirmark on theirtime. But on what did theyimprinttheirmark? Is it not in somethingspatial that discourseis inscribed? How could an eventbe printedon somethingtemporal? Social time,however,is not only somethingwhich flees. It is also the place of durable effects, of persistingpatterns.An action leaves a "trace," it makes its "mark" when it contributesto the emergenceof such patternswhich become the documentsof human action. Anothermetaphormay help us to delineatethisphenomenonof the social "imprint": themetaphorof the "record" or of the "registration." introducesthis metaphor John Feinberg,in Action and Responsibility, in another context,that of responsibility, in order to show how an actionmay be submittedto blame. Only actions,he says,which can be "registered"for furthernotice, placed as an entry on somebody's "record," can be blamed. And when there are no formal"records" (like those which are kept by institutionslike employmentoffices, schools,banks, and the police), thereis still an informalanalogue of theseformalrecordswhich we call reputationand which constitutesa basis for blaming. I would like to apply this interesting metaphorof a record and reportingto somethingother than the quasi-judicial This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions 102 NEW LITERARY HISTORY situationsof blaming,charging,creditingor punishing. Could we not say that historyis itselfthe record of human action? Historyis this quasi-"thing"on which human action leaves a "trace," puts its mark. Hence the possibilityof "archives." Before the archives which are writtendown by the memorialists, thereis thiscontinuous intentionally processof "recording"human actionwhich is historyitselfas the sum of "marks,"the fate of which escapes the controlof individual actors. Henceforthhistorymay appear as an autonomous entity,as a play with playerswho do not know the plot. This hypostasisof historymay be denounced as a fallacy,but this fallacy is well entrenchedin the processby which human action becomes social action when written down in the archivesof history.Thanks to thissedimentationin social time, human deeds become "institutions,"in the sense that their meaningno longer coincideswith the logical intentionsof the actors. The meaningmaybe "de-psychologized" to thepointwherethesinnhaft the work In the termsof P. Winch, resides in itself. (meaningfulness) The of Social the the social sciences is a in Idea of Science, object "rule-governedbehavior." But this rule is not superimposed;it is the meaning as articulatingfrom within these sedimentedor instituted works. Such is the kind of "objectivity"which proceeds from the "social fixation"of meaningfulbehavior. c. Relevanceand Importance Accordingto our thirdcriterionof what a textis, we could say that a meaningfulactionis an actionthe importanceofwhichgoes "beyond" its relevanceto its initialsituation.This new traitis verysimilarto the way in which a text breaks the ties of discourseto all the ostensive references.Thanks to this emancipationfromthe situationalcontext, whichwe called a "world," references discoursecan developnonostensive in the sense in which we speak of the Greek "world," not in the cosmologicalsense of the word, but as an ontologicaldimension. What would correspondin the field of action to the nonostensive referencesof a text? We juxtaposed,in introducingthe presentanalysis,the importance of an action to its relevance as regards the situation to which it wanted to respond. An importantaction, we could say, develops meaningswhich can be actualized or fulfilledin situationsother than the one in which this action occurred.To say the same thingin differentwords,the meaning of an importantevent exceeds, overcomes, This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 103 transcendsthe social conditionsof its productionand may be reenacted in new social contexts. Its importanceis its durable relevanceand, in some cases,its omnitemporalrelevance. This thirdtraithas importantimplicationsfor the relationbetween culturalphenomenaand theirsocial conditions.Is it not a fundamental traitof the great worksof cultureto overcomethe conditionsof their social production,in the same way a textdevelopsnew referencesand constitutesnew "worlds"? It is in this sense that Hegel spoke,in the Philosophyof Right, of the institutions(in the largestsense of the word) which "actualize" freedomas a second nature in accordance with freedom.This "realm of actual freedom" is constitutedby the deeds and workscapable of receivingrelevancein new historicalsituations. If this is true, this way of overcomingone's own conditionsof productionis the keyto the puzzlingproblemsraised by MarxismiconThe autonomyof supercerningthe status of the "superstructures." infrastructures has its their to own structuresas regardstheirrelation A work does text. of not paradigm in the nonostensivereferences a within which bears it only mirrorits time, but it opens up a world itself. d. Human Actionas an "Open Work" Finally, according to our fourthcriterionof the text as text, the meaning of human action is also somethingwhich is addressedto an indefiniterange of possible "readers." The judges are not contempoist Weltgericht. raries,but, as Hegel said, historyitself. Weltgeschichte That meansthat,likea text,human actionis an open work,the meaning of which is "in suspense." It is because it "opens up" new references and receivesfreshrelevance from them that human deeds are also which decide theirmeaning. All sigwaitingfor freshinterpretations nificantevents and deeds are, in this way, opened to this kind of practicalinterpretation throughpresentpraxis. Human action, too, is opened to anybodywho can read. In the same way that the meaning of an eventis the sense of its forthcoming the interpreinterpretations, tationby contemporarieshas no particularprivilegein this process. will be the This dialectic betweenthe work and its interpretations that we shall now consider. topic of the methodologyof interpretation II. The ParadigmofText-Interpretation I want now to show the fruitfulness of this analogy of the text at the level of methodology. This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions I 04 NEW LITERARY HISTORY The main implicationof our paradigm,as concernsthe methodsof the social sciences,is that it offersa freshapproach to the questionof the relation between erkldren(explanation) and verstehen(understanding,comprehension). As is well known,Dilthey definedthis relation as a dichotomy. For him, any model of explanationis borrowedfroma different regionof knowledge,thatof the naturalsciences with their inductive logic. Henceforth the autonomy of the socalled Geisteswissenschaften Is preservedonly by the recognitionof the irreduciblefactorof understandinga foreignpsychiclife on the basis of the signsin which thislife is immediatelyexteriorized.But if verstehenis separatedfromerkliirenby this logical gap, how can the social sciences be scientificat all? Dilthey kept wrestlingwith this paradox. He discoveredmore and more clearly,mainly afterhaving are read Husserl'sLogicadtInvestigations, that the Geisteswissenschaften of of lifeundergoa kind objectificasciencesinasmuchas the expressions tion,whichmakespossiblea scientific approachsomewhatsimilarto that of the naturalsciences,in spite of the logical gap betweenNatur and Geist, factual knowledgeand knowledgeby signs. In this way the mediation offeredby these objectifications appeared to be more important,for a scientificpurpose,than the immediatemeaningfulness of the expressionsof life for everydaytransactions. My own investigationstartswith this last perplexityin Dilthey's thought. And my hypothesisis that the kind of objectification implied in thestatusof discourseas textprovidesa betteranswerto the problem raised by Dilthey. This answerrelieson the dialecticalcharacterof the relationbetween erkliirenand verstehenas it is displayedin reading. will be to show to what extentthe paradigmof readOur tasktherefore ing, which is the counterpartof the paradigm of writing,providesa solutionforthe methodologicalparadox of social sciences. The dialectic involved in reading expressesthe originalityof the relationbetweenwritingand reading and its irreducibility to the dialogical situationbased on the immediatereciprocitybetweenspeaking and hearing.There is a dialecticbetweenexplainingand comprehending situationdevelops a problemof its own because the writing-reading which is not merelyan extensionof the speaking-hearing situationconstitutiveof dialogue. It is here thereforethat our hermeneuticis most criticalof the Rowhich took the dialogicalsituation manticisttraditionin hermeneutics, as the standardforthe hermeneutical operationapplied to the text. My contentionis thatit is thisoperation,on the contrary, which revealsthe meaningof what is already hermeneuticalin dialogical understanding. This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 105 Then, if the dialogical relationdoes not provideus with the paradigm of readingwe have to build it as an originalparadigm,as a paradigm of its own. This paradigm draws its main featuresfrom the statusof the text itselfas characterizedby (I) the fixationof the meaning, (2) its dissociationfromthe mental intentionof the author, (3) the displayof nonostensivereferences,and (4) the universalrange of its addressees. These fourtraitstakentogetherconstitutethe "objectivity"of the text. From this "objectivity"derivesa possibility of explaining,which is not derived in any way from another field, that of natural events,but which is congenial to this kind of objectivity.Thereforethere is no transferfromone regionof realityto another,fromthe sphereof facts, let us say, to the sphereof signs. It is withinthe same sphereof signs that the processof objectification takesplace and gives rise to explanait within thissphereof signsthat explanation And is toryprocedures. and comprehensionare confronted. I proposethatwe considerthisdialecticin two different ways: (I) as from to as and proceeding (2) proceeding comprehension explanation, from explanation to comprehension.The exchange and reciprocity betweenboth procedureswill provideus witha good approximationof the dialecticalcharacterof the relation. At the end of each half of thisdemonstration I shall tryto indicate brieflythe possibleextensionof the paradigmof reading to the whole sphereof the human sciences. a. From Understandingto Explanation This firstdialectic--orratherthisfirstfigureof a unique dialecticmay be convenientlyintroducedby our contentionthat to understand a textis not to rejoin the author. The disjunctionof the meaningand the intentioncreates an absolutelyoriginalsituationwhich engenders the dialectic of erkliirenand verstehen.If the objective meaning is somethingother than the subjectiveintentionof the author, it may be construedin variousways. The problemsof the rightunderstanding can no longerbe solved by a simplereturnto the alleged intentionof the author. This constructionnecessarilytakes the formof a process. As E. D. Hirsch says, there are no rules for making good guesses. But there This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions 106 NEW LITERARY HISTORY are methodsforvalidatingguesses.1This dialecticbetweenguessingand one figureof our dialecticbetweencomprehension validatingconstitutes and explanation. In this dialectic both termsare decisive. Guessing correspondsto what Schleiermachercalled the "divinatory,"validation to what he called the "grammatical." My contributionto the theoryof this dialectic will be to link it more tightlyto the theoryof the textand textreading. Why do we need an art of guessing?Why do we have to "construe" the meaning? Not only-as I tried to say a few years ago-because language is metaphoricaland because thddouble meaningofmetaphoricallanguage requiresan art of decipheringwhich tendsto unfoldthe severallayers of meaning.The case of metaphoris onlya particularcase fora general theoryof hermeneutics.In more generalterms,a text has to be construedbecause it is not a mere sequence of sentences,all on an equal footingand separatelyunderstandable. A text is a whole, a totality. The relationbetweenwhole and parts-as in a work of art or in an animal-requires a specifickind of "judgment" for which Kant gave the theoryin the Third Critique. Concretely,the whole appears as a hierarchyof topics,or primaryand subordinatetopics.The reconstruction of the text as a whole necessarilyhas a circularcharacter,in the sense that the presupposition of a certainkind of whole is implied in the recognitionof the parts. And, reciprocally,it is in construingthe detailsthatwe construethewhole.There is no necessityand no evidence what is essenconcerningwhat is importantand what is unimportant, tial and what is unessential.The judgmentof importanceis a guess. in otherterms,if a textis a whole, it is once To put the difficulty more an individual like an animal or a workof art. As an individual it can only be reached by a processof narrowingthe scope of generic conceptsconcerningthe literarygenre,the class of text to which this text belongs,the structuresof different kinds which intersectin this text.The localizationand the individualizationof this unique textare stillguesses. Still anotherway of expressingthe same enigma is that as an individual the textmay be reachedfromdifferent sides. Like a cube, or a volume in space, the textpresentsa "relief." Its different topics are I Validity in Interpretation(New Haven, 1967): "The act of understandingis at firsta genial (or a mistaken) guess, and there are no methods for making guesses, no rules for generatinginsights.The methodical activityof interpretationcommences when we begin to test and criticize our guesses" (p. 203). And further: "A mute symbolismmay be construedin several ways." This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODELOF THE TEXT 107 not at the same altitude.Thereforethe reconstruction of the whole has a perspectivist It to that of is similar always possible aspect perception. to relate the same sentencein different ways to this or that sentence consideredas the cornerstoneof the text. A specifickind of onesidedness is implied in the act of reading. This onesidednessconfirmsthe "guess" characterof interpretation. For all thesereasonsthereis a problemof interpretation not so much because of the incommunicabilityof the psychic experience of the author,but because of the very nature of the verbal intentionof the text.This intentionis somethingotherthan the sum of the individual meaningsof the individual sentences. A text is more than a linear successionof sentences.It is a cumulative,holisticprocess.This specific structureof the text cannot be derived from that of the sentence. Thereforethe kind of "plurivocity"which belongsto textsas textsis somethingotherthan the polysemyof individualwordsin ordinarylanguage and the ambiguityof individual sentences.This plurivocityis typicalof thetextconsideredas a whole,open to severalreadingsand to severalconstructions. As concernsthe proceduresof validationbywhichwe testour guesses, I agree with Hirsch that theyare closer to a logic of probabilitythan is to a logic of empiricalverification.To show that an interpretation more probable in the light of what is known is somethingotherthan showing that a conclusion is true. In this sense, validation is not verification.Validation is an argumentativediscipline comparable to the juridical proceduresof legal interpretation.It is a logic of uncertaintyand of qualitative probability.In this sense we may give an and acceptable sense to the oppositionbetween Geisteswissenschaften to the without alleged dogma concedinganything Naturwissenschaften of the individual. The method of conveyance of of the ineffability indices,typicalof the logic of subjectiveprobability, gives a firmbasis for a science of the individual deservingthe name of science. A text is a quasi-individual,and the validation of an interpretation applied to it may be said, with completelegitimacy,to give a scientificknowledge of the text. Such is the balance betweenthe geniusof guessingand the scientific characterof validation which constitutesthe modern complementof the dialecticbetweenverstehenand erkliiren. At the same time,we are prepared to give an acceptable meaning to the famous concept of a hermeneuticcircle. Guess and validation are in a sense circularlyrelatedas subjectiveand objective approaches to the text. But this circle is not a vicious circularity.It would be a This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions I O8 NEW LITERARY HISTORY which, cage if we wereunable to escape the kindof "self-confirmability" relation between threatens this to Hirsch guess according (pp. 165 ff.), and validation.To the proceduresof validationalso belongprocedures of invalidationsimilarto thecriteriaof falsifiability emphasizedby Karl The of falsification is played here role in his Popper Logic of Discovery. An conflict the between by competinginterpretations. interpretation must not onlybe probable,but more probablethan another.There are criteriaof relativesuperioritywhich may easily be derived from the logic of subjectiveprobability. In conclusion,if it is true that thereis always more than one way are equal and of construinga text,it is not truethatall interpretations may be assimilatedto so-called"rulesof thumb" (Hirsch, p. 203). The text is a limitedfieldof possibleconstructions. The logic of validation allows us to move betweenthe two limitsof dogmatismand skepticism. It is always possible to argue for or against an interpretation,to confrontinterpretations, to arbitratebetween them, and to seek for an agreement,even if this agreementremainsbeyondour reach. To what extentis thisdialecticbetweenguessingand validatingparadigmaticfor the whole fieldof the social sciences?That the meaning of human actions,of historicalevents,and of social phenomenamay be construedin several different ways is well knownby all expertsin the social sciences.What is less knownand understoodis thatthismethodois foundedin thenatureof the object itselfand, morelogical perplexity that it does not condemn the scientistto oscillatebetween dogover, matism and skepticism. As the logic of text-interpretation suggests, thereis a specificplurivocity belongingto themeaningof human action. Human action,too, is a limitedfieldof possibleconstructions. A traitof human action which has not been emphasizedin the preceding analysismay provide an interestinglink between the specific plurivocityof the textand the analogical plurivocityof human action. This traitconcernsthe relationbetweenthe purposiveand the motivational dimensionsof action. As many philosophersin the new fieldof action theoryhave shown,the purposivecharacterof an action is fully recognizedwhen the answerto the question what? is explained in the termsof an answer to the question why? I understandwhat you intended to do, if you are able to explain to me why you did such and such an action. Now, what kinds of answer to the question why? make sense? Only those answerswhich afforda motiveunderstoodas a reasonfor... and not as a cause. And what is a reasonfor... which is not a cause? It is, in the termsof G. E. M. Anscombeand A. I. Melden, an expression,or a phrase,which allows us to considerthe action as thisor that. If you tellme that you did this or thatbecause This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT IO9 of jealousy or in a spiritof revenge,you are asking me to put your action in the light of this categoryof feelingsor dispositions.By the same token,you claim to make sense with your action. You claim to make it understandableforthe othersand foryourself.This attemptis particularlyhelpfulwhen applied to what G. E. M. Anscombe calls the desirability-character of wanting.Wants and beliefshave the propertynot only of being forceswhich make people act in such and such ways,but of makingsense,by virtueof the apparentgood which is the correlate of their desirability-character. I may have to answer the what as want this? do the On basis of these desirabilityyou question, charactersand of the apparentgoods which correspondto them,it is possible to argue about the meaning of an action, to argue for or against thisor that interpretation.In this way the account of motives already foreshadowsa logic of argumentationprocedures. Could we not say that what can be (and must be) construedin human action is the motivationalbasis of this action, i.e., the set of desirabilitycharacterswhichmay explainit? And could we notsay thatthe process of arguinglinked to the explanationof action by its motivesunfolds a kind of plurivocitywhich makes action similarto a text? What seems to make legitimatethis extensionfrom guessing the meaningof a text to guessingthe meaning of an action is that in arguingabout the meaningof an action I put my wantsand my beliefsat a distanceand submitthemto a concretedialecticof confrontation with of view. of This action at a distance opposite points way puttingmy in orderto make sense of my own motivespaves the way for the kind of distancingwhich occurswithwhat we called thesocial inscription of human action and to which we applied the metaphorof the "record." The same actions which may be put into "records" and henceforth "recorded" may also be explained in different ways according to the of the to their motivational plurivocity argumentsapplied background. If we are correctin extendingto actionthe conceptof "guess" which we took as a synonymfor verstehen,we may also extend to the field of action the concept of "validation" in which we saw an equivalent of erkliiren.Here, too, the modern theoryof action providesus with an intermediarylink between the procedures of literarycriticism and thoseof the social sciences. Some thinkershave triedto elucidate theway in whichwe imputeactionsto agentsin the lightof thejuridical proceduresby which a judge or a tribunalvalidates a decision concerninga contractor a crime. In a famousarticle,"The Ascriptionof Responsibilityand Rights" (Proceedingsof the AristotelianSociety,49 [1948-49], I71-94), H. L. A. Hart shows in a very convincingway This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions I 10 NEW LITERARY HISTORY that juridical reasoningdoes not at all consistin applyinggenerallaws to particularcases, but each time in construinguniquelyreferring decisions. These decisionsterminatea careful refutationof the excuses and defenseswhich could "defeat" the claim or the accusation. In saying that human actions are fundamentally"defeasible" and that juridical reasoningis an argumentativeprocesswhich comes to grips with the different ways of "defeating"a claim or an accusation,Hart has paved the way fora generaltheoryof validationin which juridical reasoningwould be the fundamentallinkbetweenvalidationin literary criticismand validation in the social sciences.The intermediary function of juridical reasoningclearlyshows that the proceduresof validationhave a polemicalcharacter.In frontof the court,theplurivocity common to textsand to actionsis exhibitedin the formof a conflict of interpretations, and the finalinterpretation appears as a verdictto which it is possibleto make appeal. Like legal utterances,all interpretationsin the fieldof literarycriticismand in the social sciencesmay be challenged,and the question"What can defeata claim?" is common to all argumentative situations.Only in the tribunalis therea moment when the proceduresof appeal are exhausted. But it is so onlybecause the decisionof the judge is implementedby the forceof public power. Neitherin literarycriticismnor in the social sciencesis theresuch a last word. Or, if thereis any, we call that violence. b. From Explanationto Understanding The same dialecticbetweencomprehensionand understandingmay receive a new meaningif taken in the reverseway, fromexplanation to understanding.This new Gestaltof the dialecticproceedsfromthe functionof the text.This referential natureof the referential function, as we said, exceeds the mere ostensivedesignationof the situation common to both speaker and hearer in the dialogical situation.This abstractionfromthe surroundingworld gives rise to two oppositeattitudes. As readers,we may eitherremain in a state of suspenseas reworld,or we may actualize the potential gards any kind of referred-to nonostensivereferencesof the text in a new situation,that of the reader. In the firstcase, we treatthe text as a worldlessentity;in the second,we create a new ostensivereferencethanksto the kind of "execution" which the art of readingimplies. These two possibilitiesare equally entailed by the act of reading,conceived as their dialectical interplay. This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT III strucThe firstway of readingis exemplifiedtoday by the different tural schools of literarycriticism.Their approach is not only possible, but legitimate. It proceeds from the suspension,the epoche, of the ostensivereference.To read in thisway means to prolongthis suspension of the ostensivereferenceto the world and to transferoneselfinto the "place" wherethe textstands,withinthe "enclosure"of thisworldless place. Accordingto this choice,the textno longerhas an outside, it has only an inside. Once more, the veryconstitution of the text as textand of the systemof textsas literaturejustifiesthis conversionof the literarythingsinto a closed systemof signs,analogous to the kind of closed systemwhich phonologydiscovered at the root of all discourse,and whichde Saussurecalled "la langue." Literature,according to thisworkinghypothesis, becomes an analogon of "la langue." On the basis of thisabstraction,a new kind of explanatoryattitude may be extendedto the literaryobject,which,contraryto the expectation of Dilthey,is no longerborrowedfromthe natural sciences,i.e., from an area of knowledgealien to language itself.The opposition betweenNatur and Geist is no longeroperativehere. If some model is borrowed,it comes fromthe same field,fromthe semiologicalfield. It is henceforth possibleto treattextsaccordingto the elementaryrules which linguisticssuccessfully applied to the elementarysystemsof signs that underliethe use of language. We have learned fromthe Geneva school, the Prague school, and the Danish school that it is always possibleto abstractsystemsfromprocessesand to relatethesesystemswhetherphonological,lexical,or syntactical-to unitswhich are merely definedby the oppositionwith other units of the same system.This interplayof merelydistinctiveentitieswithinfinitesets of such units definesthe notion of structurein linguistics. It is this structuralmodel which is now applied to texts,i.e., to sequences of signs longerthan the sentence,which is the last kind of unit that linguisticstakes into account. Claude L6vi-Straussformulatesthis Id his Anthropologie structurale, in the workinghypothesis followingway in regard to one categoryof texts,that of myths: "Like everylinguisticentity,the mythis made units.These constitutive unitsimplythe presenceof up of constitutive those which generallyoccur in the structuresof language, namely phonemes,morphemes,and semantemes.Each form differsfromthe one which precedes it by a higher degree of complexity. For this reason we will call the elements,which properlybelong to the myth units" (p. (and which are the mostcomplexof all): large constitutive means of this the are which large units, 233). By workinghypothesis, This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions I 12 NEW LITERARY HISTORY at least the same size as the sentenceand which, put together,form the narrativeproperto the myth,can be treatedaccordingto the same rules as the smallestunitsknownsto linguistics.It is in order to insist on this likenessthat Claude Levi-Straussspeaks of mythemes,just as we speak of phonemes,morphemes,and semantemes.But in order to remain within the limitsof the analogy between mythemesand the lower-levelunits,the analysisof textswill have to performthe same sort of abstractionas that practiced by the phonologist.To him, the phonemeis not a concretesound,in an absolutesense,withits acoustic quality. It is not, to speak like de Saussure, a "substance" but a "form,"that is to say, an interplayof relations. Similarly,a mytheme is not one of the sentencesof a myth,but an oppositivevalue attached to several individual sentences forming,in Levi-Strauss' terms, a "bundle of relations." "It is onlyin the formof a combinationof such bundles that the constitutiveunits acquire a meaning-function"(p. is not at all what the 234). What is here called a meaning-function mythmeans,its philosophicalor existentialcontentor intuition,but the arrangement,the dispositionof mythemes-in short,the structureof the myth. We can indeed say thatwe have explained a myth,but not that we have interpreted it. We can, by means of structuralanalysis,bringout the logic of it throughthe operationswhich relate the bundles of relationsamong themselves.This logic constitutes"the structurallaw of the mythunder consideration"(p. 241). This law is preeminently an of and not at all of in the sense of a reciting object reading speaking, wherethe powerof the mythwould be reenactedin a particularsituation. Here the text is only a text, thanks to the suspensionof its meaning for us, to the postponementof all actualizationby present speech. I want now to show in what way "explanation" (erkliiren)requires "understanding"(verstehen)and bringsforthin a new way the inner dialecticwhich constitutes"interpretation" as a whole. As a matterof with a of fact,nobodystops conception mythsand narrativesas formal as thisalgebraof constitutive units.This can be shownin different ways. of mythsby Lvi-Strauss, First,even in themostformalizedpresentation the units he calls "mythemes"are still expressedas sentenceswhich bear meaning and reference.Can anyone say that theirmeaning as such is neutralizedwhen theyenterintothe "bundle of relations"which alone is taken into account by the "logic" of the myth? Even this bundle of relations,in its turn,must be writtenin the formof a sentence. Finally,the kind of language game which the whole systemof This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 113 oppositionsand combinationsembodieswould lack any kind of significance if the oppositionsthemselves,which, according to Levi-Strauss, the mythtendsto mediate,were not meaningfuloppositionsconcerning birth and death, blindnessand lucidity,sexualityand truth. Beside these existentialconflictsthere would be no contradictionsto overcome, no logical functionof the mythas an attemptto solve these contradictions.Structuralanalysisdoes not exclude, but presupposes, the oppositehypothesis concerningthe myth,i.e., that it has a meaning as a narrativeof origins. Structuralanalysismerelyrepressesthisfunction. But it cannot suppressit. The mythwould not even functionas a logical operatorif the propositionswhich it combinesdid not point toward boundarysituations. Structuralanalysis,far from gettingrid of this radical questioning,restoresit at a level of higherradicality. If thisis true,could we notsay thatthe functionof structuralanalysis is to lead from a surface-semantics, that of the narratedmyth,to a that of the boundarysituationswhich constitutethe depth-semantics, ultimate"referent"of the myth? I fullybelievethatifsuch werenot the functionof structuralanalysis, it would be reduced to a sterilegame, a divisivealgebra,and even the mythwould be bereftof the functionwhich LUvi-Strausshimselfassigns to it, that of making men aware of certain oppositionsand of mediation.To eliminatethisreference tendingtowardtheirprogressive to the aporias of existencearound which mythicthoughtgravitates would be to reduce a theoryof mythto the necrologyof the meaningless discoursesof mankind. If, on the contrary,we considerstructural analysisas a stage-and a necessaryone--between a naive interpretation and a critical interpretation, between a surface-interpretation and a depth-interpretation, then it would be possibleto locate explanation and understandingat two different stages of a unique hermeneutic arc. It is this depth-semantics which constitutesthe genuine between object of understandingand which requiresa specificaffinity the readerand the kind of thingsthe textis about. But we must not be misled by this notionof personal affinity. The of the text not what is the author intendedto say, but depth-semantics what the textis about,i.e., the nonostensivereferenceof the text. And the nonostensivereferenceof the text is the kind of world opened of the text. up by the depth-semantics Thereforewhat we want to understandis not somethinghidden behindthe text,but somethingdisclosedin frontof it. What has to be understoodis not the initial situation of discourse,but what points toward a possibleworld. Understandinghas less than ever to do with This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions I14 NEWLITERARY HISTORY the author and his situation. It wants to grasp the world-propositions opened up by the referenceof the text.To understanda textis to follow its movementfromsense to reference,fromwhat it says to what it talks about. In this processthe mediatingrole played by structural both the justification of thisobjectiveapproach and analysisconstitutes the rectification of the subjectiveapproach.We are definitely prevented fromidentifying withsome kind of intuitivegraspingof understanding the intentionunderlyingthe text.What we have said about the depthsemanticswhich structuralanalysisyieldsinvitesus ratherto thinkof the sense of the textas an injunctionstartingfromthe text,as a new way of lookingat things,as an injunctionto thinkin a certainmanner. Such is the referenceborne by depth-semantics. The text speaks of a possibleworld and of a possibleway of orientatingoneselfwithinit. The dimensionsof thisworld are properlyopened up by, disclosedby, the text. Disclosureis the equivalentfor writtenlanguage of ostensive referenceforspokenlanguage. we preservethe language of Romanticisthermeneutics, If, therefore, when it speaksof overcomingthe distance,of making"one's own," of appropriatingwhat was distant,other,foreign,it will be at the price of an importantcorrective.That whichwe make our own-Aneignung in German-that which we appropriate,is not a foreignexperience, but the power of disclosinga world which constitutes the referenceof the text. This link betweendisclosureand appropriationis, to my mind, the cornerstoneof a hermeneuticwhich would claim both to overcome the shortcomings of historicismand to remain faithfulto the original intentionof Schleiermacher'shermeneutics.To understandan author betterthan he could understandhimselfis to display the power of disclosureimplied in his discoursebeyond the limitedhorizon of his own existentialsituation.The processof distancing,of atemporalization, to which we connected the phase of Erkliirung,is the fundamental forthis enlargingof the horizonof the text. presupposition This second figure,or Gestalt,of the dialectic betweenexplanation and comprehensionhas a strongparadigmaticcharacterwhich holds forthe whole fieldof the human sciences. I want to emphasizethree points. First,the structuralmodel, takenas a paradigmforexplanation,may be extendedbeyond textual entitiesto all social phenomena because it is not limitedin its applicationto linguisticsigns,but applies to all kindsof signswhich are analogousto linguisticsigns.The intermediary link between the model of the text and social phenomena is consti- This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT I 15 tuted by the notion of semiologicalsystems.A linguisticsystem,from the point of view of semiology,is only a species withinthe semiotic genre,althoughthis specieshas the privilegeof being a paradigmfor the otherspeciesof the genre. We can say thereforethat a structural model of explanationcan be generalizedas far as all social phenomena which may be said to have a semiologicalcharacter,i.e., as far as it is possibleto definethe typicalrelationsof a semiologicalsystemat their level: the generalrelationbetweencode and message,relationsamong the specificunits of the code, the relationbetweensignifierand signified, the typical relation within and among social messages; the structureof communicationas an exchangeof messages,etc. Inasmuch as the semiologicalmodel holds,the semioticor symbolicfunction,i.e., the functionof substituting things signsfor thingsand of representing the of means by signs,appears to be more than a mere effectin social life. It is its veryfoundation.We should have to say, accordingto this generalizedfunctionof the semiotic,not only that the symbolicfunction is social, but that social realityis fundamentallysymbolic. If we followthissuggestion,then the kind of explanationimpliedby the structuralmodel appears to be quite differentfrom the classical causal model, especiallyif causation is interpretedin Humean terms as a regularsequence of antecedentsand consequentswith no inner logical connectionbetweenthem. Structuralsystemsimplyrelationsof a quite different kind,correlativeratherthan sequentialor consecutive. If this is true, the classical debate about motives and causes which has plagued the theoryof action theselast decades loses its importance. If the search for correlationswithinsemioticsystemsis the main task of explanation,thenwe have to reformulate the problemof motivation in social groupsin new terms. But it is not the aim of this paper to develop this implication. The second paradigmaticfactor in our previous concept of textinterpretation proceeds fromthe role we assigned to depth-semantics betweenstructuralanalysisand appropriation. This mediatingfunction of depth-semantics must not be overlooked,since the appropriation's losingits psychologicaland subjectivecharacterand receivinga genuine epistemologicalfunctiondepends on it. Is theresomethingsimilarto the depth-semantics of a text in social that tend to I the for correlations search should say phenomena? within and between social phenomena treated as semiotic entities would lose importanceand interestif it would not yieldsomethinglike a depth-semantics.In the same way that linguisticgames are forms of life,accordingto the famousaphorismof Wittgenstein, social struc- This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions 116 NEW LITERARY HISTORY tures are also attemptsto cope with existentialperplexities,human predicaments,and deep-rootedconflicts.In thissense,thesestructures, dimension.They point toward aporias of social too, have a referential existence,the same aporias around which mythicalthoughtgravitates. And this analogical functionof referencedevelops traitsvery similar to what we called the nonostensivereferenceof a text,i.e., the display of a Welt which is no longer an Umwelt,the projectionof a world which is more than a situation.May we not say thatin social science, to critical interpretations, too, we proceed from naive interpretation fromsurface-interpretations to depth-interpretations throughstructural which But it is gives meaning to the analysis? depth-interpretation whole process. This last remarkleads us to our thirdand last point. If we follow the paradigm of the dialecticbetweenexplanationand understanding to its end, we must say that the meaningfulpatternswhich a depthwants to grasp cannot be understoodwithouta kind of interpretation similar to that of the reader who grasps the commitment personal of the textand makes it his "own." Everybodyknows depth-semantics the objectionswhich an extensionof the concept of appropriationto the social sciencesis exposed to. Does it not make legitimatethe intrusion of personal prejudices, or subjective bias into the field of scientificinquiry? Does it not introduceall the paradoxes of the hermeneuticalcircle into the social sciences? In other words, does not the paradigmof disclosureplus appropriationdestroythe veryconcept of social science?The way in which we introducedthis pair of terms withinthe frameworkof text-interpretation providesus not only with a paradigmaticproblem,but with a paradigmaticsolution. This solution is not to deny the role of personal commitmentin understandinghuman phenomena,but to qualifyit. As the model of shows, understandinghas nothing to do with an text-interpretation immediate graspingof a foreignpsychic life or with an emotional identificationwith a mental intention. Understandingis entirely mediated by the whole of explanatoryprocedureswhich precede it and accompany it. The counterpartof this personal appropriationis not somethingwhich can be felt,it is the dynamicmeaning released by the explanationwhich we identifiedearlier with the referenceof the text,i.e., its power of disclosinga world. must be applied The paradigmaticcharacterof text-interpretation This means that ultimate the conditionsof down to this implication. an authenticappropriation,as theywere displayedin relationto texts, are themselvesparadigmatic.Thereforewe are not allowed to exclude This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions THE MODEL OF THE TEXT 'I7 the finalact of personalcommitmentfromthe whole of objectiveand explanatoryprocedureswhich mediate it. This qualificationof the notion of personal commitmentdoes not eliminatethe "hermeneuticcircle." This circle remainsan insuperable structureof knowledgewhen it is applied to human things,but this qualificationpreventsit from becominga vicious circle. Ultimately,the correlationbetweenexplanationand understanding, betweenunderstandingand explanation,is the "hermeneuticcircle." UNIVERSITY This content downloaded from 144.92.188.16 on Sun, 31 Aug 2014 18:13:31 PM All use subject to JSTOR Terms and Conditions OF PARIS