sign in sign up
<<
Home , Intertextuality

INTERTEXTUALITY: ORIGINS AND DEVELOPMENT OF THE CONCEPT Author(s): María Jesús Martínez Alfaro Source: Atlantis, Vol. 18, No. 1/2 (Junio - Diciembre 1996), pp. 268-285 Published by: AEDEAN: Asociación española de estudios anglo-americanos Stable URL: http://www.jstor.org/stable/41054827 . Accessed: 14/10/2011 19:37

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 [email protected].

AEDEAN: Asociación española de estudios anglo-americanos is collaborating with JSTOR to digitize, preserve and extend access to Atlantis.

http://www.jstor.org ATLANTISXVIII (1-2) 1996

INTERTEXTUALITY: ORIGINS AND DEVELOPMENT OF THE CONCEPT*

MaríaJesús Martínez Al faro Universidadde Zaragoza

Theanalysis of the concept of intertextual itycarried out in this essay begins with a surveyof thevarious ways in whichthe subject appears before Kristeva's introduction ofthe term as such.Then it concentrateson Bakhtin'sand Kristeva'srole as thefirst contributors to the developmentofa theoryof intertextuality.The detailedsudy of the wide variety of perspec- tivesfrom which the phenomenon and its increasing relevance have been approached by later criticsconstitutes the last part of the essay.

1. INTRODUCTION Intertextualityas a term was firstused in 's "Word, Dialogue and " (1966) andthen in "The Bounded Text" (1966-67), essays she wrote shortly after arriving inParis from her native Bulgaria.2 The concept of intertextuality thatshe initiated proposes thetext as a dynamicsite in whichrelational processes and practicesare thefocus of analysisinstead of static structures and products. The "literary word", she writes in "Word, Dialogue,and Novel", is "an intersectionoftextual surfaces rather than a point(a fixed meaning),as a dialogueamong several writings" (1980, 65). DevelopingBakhtin's spatializationof literary language, she argues that "each word (text) is an intersection of otherwords (texts) where at least one other word (text) can be read"(1980, 66). Thereare always other words in a word,other texts in a text.The conceptof intertex- tualityrequires, therefore, that we understandtexts not as self-containedsystems but as differentialand historical, as tracesand tracings of otherness, since they are shaped by the repetitionand transformationof other textual structures. Rejecting the New Critical principleof textual autonomy, the theory of intertextuality insists that a textcannot exist as a self-sufficientwhole, and so, that it does not function as a closedsystem. Fromthis initial approach, there have appeared a widerange of attitudestowards the conceptof intertextuality andwhat it implies,to suchan extentthat it is practicallyim- possibleto deal withit without considering other related subjects or withouttaking into accountthe various contributions made by a largenumber of literarycritics. One of the mostimmediate consequences of such a proliferationofintertextual theories has been the progressivedissolution of the text as a coherentand self-contained unit of meaning, which has led,in turn,to a shiftof emphasis from the individual text to theway in whichtexts relateto one another.

1 Theresearch carried out for the writing of this paper has been financed by the Spanish Ministry of Educationand Science (DGICYT, PS94-0057). 2 In Desirein Language, "Word, Dialogue and Novel" is datedin 1966and "The Bounded Text" in 1966-67;both essays appeared in herfirst volume of essays Recherches pour une sémanalyse in 1969.

268 INTERTEXTUALITY:ORIGINS AND DEVELOPMENT OFTHE CONCEPT 269

Thoughintertextuality as a term appeared some three decades ago, and thetwentieth centuryhas proved to be a periodespecially inclined to itculturally, intertextuality is by no meansa time-boundfeature: the phenomenon, in some form, is at leastas old as recorded humansociety (Worton and Still 1990, 2). Unsurprisingly,therefore, we can findtheories ofintertextuality wherever there has been discourseabout texts, from the classics, like Plato,Aristotle, Horace and Longinus, to Bakhtin,Kristeva and other twentieth-century theoristssuch as Genette,Barthes, Derrida and Riffaterre, among others. Goingback to the classics and beginning with Plato, it mustbe said thatin spiteof his oppositionto poetryon moral,and hence political, grounds, certain aspects of his theory havemuch in commonwith some modern approaches to intertextuality.Bakhtin himself locatesin theSocratic dialogues one ofthe earliest forms of whathe termsvariously the novel,heteroglossia, dialogism - whatKristeva will christenintertextuality. The dia- logues,Plato's typicalcreation, are usuallymeandering and inconclusivediscussions lackingoverall unity and characterized by their digressive and playful tone. There is am- bivalencenot only in thediversity of ideologies evoked, but also in thecentral im age of Socrates,the wise fool, sometimes sympathetic or affectionate,sometimes ironic or even savagelysatirical. This serious truth-seeking bymeans of a pluralityof voices obviously recallswhat Bakhtin will celebrate in thedialogic novel. In additionto theform of the Socraticdialogue, intertextual relations are highlighted inother aspects of Plato'stheory, suchas hisnotion of texts as subliminalpurveyors of ideology that can influenceand alter thesubject, as wellas inhis view of . Neither Platonic nor Aristotelian imitation is tobe understoodas imitationof nature. In thecase ofPlato, the "poet" always copies an earlieract of creation, which is itselfalready a copy.For Aristotle, dramatic creation is the reduction,and hence intensification, ofa mass of textsknown to thepoet and probably to theaudience as well.These texts vary from other written works of literatureto theoral traditionof myths,stock characters or social codes of conduct.Though Bakhtin and Kristevasaw Aristotelian logic as relatedto themono logic pole of discoursedue to its emphasison unifiedand universaltruths, later critics have arguedthat the Aristotelian accountof composition as drawingfrom a varietyof sources can be consideredclose to the notionsof polyphony and dialogism (Worton and Still 1990, 4). If Aristotleholds that we learnthrough imitating others and that our instinct to enjoy worksof imitation is an inborninstinct, both Cicero and Quintilian will emphasize later on thatimitation is notonly a meansof forging one's discoursebut also a consciouslyin- tertextualpractice that contributes tothe definition of theindividual. For them, imitation is notrepetition but the completion of an actof interpretation. Thus, imitation as theoryand practicepresupposes a virtual simultaneity and identification ofreading and writing, but it alsoimplies and depends upon a processof transformation. Present literary practices have producedeven copies without an original(Baudrillard's notion of simulacra),that underline theproblematic relationship between physical and semiotic reality. However, in itsearlier phase,imitation presupposes reference toa pre-existentreality which is concreteas wellas textual.In anycase, thestylistic exercise of imitationreveals the idea of proliferation, variation,regeneration, which is a centralfeature of language. TheMiddle Ages, with their multi-levelled interpretation oftexts, also showa similar beliefin the possibility to expandthe meanings of a work,even if such meanings should fit intolimited and predetermined categories. Church fathers and medieval theologians made currentthe view that the created world in its radiant order and hierarchy should be regarded as God's symbolicbook. If thiswas true,the objects which com posed the world were a kindof dictionary of God's meanings.Thus, when God himselfwrote a verbalbook (the Bible)the words in itpointed, at a literallevel, to the objects in His otherbook, the Book of Nature,but, in addition,the things signified by thosewords had a spiritualsense as

ATLANTIS XVIII (1-2) 1996 270 MaríaJesús Martínez Alfaro well,since they had been invested with God's meanings.Somehow, the interpretation of theBible already depended on anintertextual practice and, at a timein whichliterature was subordinatedtoTheology, what was trueof religious texts was also madeextensible to secularones. All literaryworks were seen as goingback to the Bible and all couldbe read likeit, a viewthat can be regardedas a medievalversion of what contemporary authors like Borgeshave conceived as theGreat Book of and the circular memory of reading. If God's twoBooks were intended to have more than one sense,it was just natural for individualwriters to disregardoriginality and to try and emulate God's techniqueinstead. Yetit was Renaissance literature that showed, perhaps for the first time in Western culture, a consciousawareness of discourse as open,unfinished, and subject to an infinitenumber ofinterpretations. Thetextual past is alwayspresent through or in the workof such writers as Bacon,Shakespeare, Montaigne, Ronsard, Du Bellay,etc. What theseauthors perceive is theglobality and infinite potentiality ofthe culture in whichtheir owndiscourses are inscribed, rather than their debt to previouswriters, whose merit is seento lie in theirpower of expressionand not in anymonopoly (or evenoriginality) of thought.Accordingly, the way in whichsuch writers deal withthe work of preceding authorsis based on theirbelief that their possibilities of imitation,understood as interpretationand re-writingof the Urtext,are limitless.Only by multiplyingand fragmentinghis/her models can theindividual writer assert and maintain his/her indepen- dence.This is thetheory proposed by Montaigne, for instance. He believedthat the "self is tobe foundin a distancingof the reading and writing subject from the anterior "other" (a viewmuch in consonancewith the Bloómian concept of "anxietyof influence")and defendsa sortof boastful forgetfulness as the best means of escaping the tyranny of past masters(Worton and Still 1990, 7-9). Thewriter's efforts to detachhim/herself from the work of previousauthors as wellas to proclaimhis/her own creative space received a newimpulse from the mid-eighteenth centuryonwards. But, in contrastwith the lack of interestin originalitythat, as we have seen,dominated the literary world during the classical period, the Middle Ages and the Renaissance,the eighteenth century brought with it a revaluationof as theonly truesign of an author'sgenius. It wasthen that the concept of influence arose, thus bearing fromthe very beginning the seeds of a methodologyrelated to, but ultimately different from thatof intertextuality.From an intertextualperspective, there is no wayof considering originalityas a traitto be cherishedby either authors or readers. T. S. Eliotwas perhapsthe firstto statethe fact that the most individual parts of an author'swork may be thosein whichhis/her ancestors are morevigorously present (1971, 784). Whilethe notion of traditionhad quite often led tointerpreting an author's work in the light of those that had precededit, it was also Eliot's innovation to assertthat influence moves in twodirections: whenstudying a work one must consider what has come before it, but one must equally be awareof the fact that the work of the dead poets changes and enriches its meaning in the lightof what has been written by later authors. In spiteof his undeniable influence on the ideasof the New Criticism,T. S. Eliot,an outstandingfigure in thecontext of modernist poetry,qualified in thisway the New Critics'view of the work of artas a self-sufficient whole,a completesystem based on therelation between images, rhythm, sounds, etc., whichdetermine its structure. Such an approach,which goes back to the romantics and is relatedto the ideas expounded by Symbolism and , exemplifies the position that latertheories of intertextuality have tried to undermine. It can thereforebe saidthat Eliot's quasi-intertextualideas about the simultaneity ofall worksof literatureand the perpetual processof re-adjusting the relations among them are surprisingly up-to-date. In thelate twentieth century, temporally sequential habits of thinking and reading have beenparticularly questioned. Robert Darnton has contended that "we constantly need to be

ATLANTISXVIII (1-2) 1996 INTERTEXTUALJTY:ORIGINS AND DEVELOPMENTOF THE CONCEPT 27 1 shakenout of a falsesense of familiarity with the past, to be administereddoses of culture shock"(in Kiely1993, 19). But we also needto be shakenout of a falsesense of our abilityto see thepast "as it was". Accordingly,Borges points out that "the present has somethinghard and rigid about it. But as to thepast, we arechanging it all thetime" (in Kiely1993, 17). In thissense, the contemporary preoccupation with inter textuality tends to questionthe usefulness of previous critical narratives of unified progression, in orderto suggest,instead, a viewof literaryworks as crowdedwith layered images of multiple reflectionsand unexpected relationships. Robert Kiely, for instance, sees postmodern literatureas engagedin "an effortto rupture the myth of a coherenttra dition ... It maybe perfectlyobvious (and easy to prove) that Beckett learned from Dickens and Joyce. But this does notcontradict the fact that Beckett teaches us (perhapsforces us) to reador reread Dickensand even Melville in particular ways" (1993, 19). Whileall authorsre- write the work of predecessors, many contemporary writers con- sciouslyimitate, quote, plagiarize, ... extensively.As HeinrichF. Plett(1991, 27) putsit, ré-écriture dominates écriture in twentieth-century literature: the image for writing haschanged from original inscription to parallelscript, and writers think less of writing originallyand more of re- writing. Even if, as we havesaid, intertextual ityis byno meansa time-boundfeature, itis obviousthat certain cultural periods incline to it mòre than others and thatour centuryhas alreadywitnessed two such phases.In the modernistera, intertextualityis apparent in everysection of culture: literature (Eliot, Joyce), art (Picasso, Ernst),music (Stravinsky, Mahler), photography (Heartfield, Haussmann), etc., even if it is interpretedindifferent ways. shows an increaseof this tendency which nowincludes films (e.g. Woody Allen's Play it Again, Sam) andarchitecture (e.g. Charles Moore's Piazza d'Italia, New Orleans) (Plett 1991, 26). Riskingsome degree of oversimplification,onecould say that the pretexts of themodernist work are normative (thesepretexts come from a widerange of epochs and cultures but the privi leged ones are alwaysthe canonized and classical texts). The postmodernist work, by contrast,has as its veryaim the levelling down of all traditionaldistinctions between high and low: pastand present,classic and pop,art and commerce,they all are reducedto thesame statusof disposablematerials. Anyway, the production of art and literature during our century has becomean act of creationbased on a re-cyclingof previouslyexisting works. This developmenthas not occurred in a theoreticalvacuum; it has actually been accompanied by a particulartheory legitimizing and re-defining the status of texts and their producers: the theoryof intertextuality. Thus, the typically New Critical textualism of the early twentieth century,and the close readings which showed tensions and multiplicity reconciled within thesingle text, have been succeeded by an approachwhich expands criticism beyond the individualwork in order to consider it in relation to the whole literary system as wellas to culture,history and society, no longerregarded as objectiveentities over and againstthe textbut as partakingof the same textuality as literature. Thecrucial step that separates every previous approximation tointertextuality from the notionthat Kristeva initiated taking Bakhtin's theories as a pointof departure is theview of theexterior of the text as a system(or an infinity) of other such textual structures. Thus, if a textrefers to "all othertexts", these are seen,in turn,as convergingwith history and reality,both existing only in textualized form. Likewise, both in Bakhtinand Kristeva, the subjectis conceivedas composedof discourses, as a signifyingsystem, a textunderstood ina dynamicsense. No wonder,then, that, once this step was taken,intertextuality came to be definedas nothingbut "l'impossibilité de vivrehors du texteinfini" (Barthes 1973b, 59). Fromthis point of view,it is easyto understandthe fact that the development of the theoryof intertextualitywould constitute in itselfa complexintertextual event. If Plato,

ATLANTIS XVIII (1-2) 1996 272 MaríaJesús Martínez Alfaro

Aristotle,Cicero, Quintilian ... can be consideredthe intertexts of Bakhtin'stheory, Kristeva's dialogue with the texts of Bakhtin, in whichshe initiallyuses theterm inter- textualité,is carried out, in turn,with the mediation of theworks of Derridaand Lacan, amongothers. The ideasexpounded by all theseauthors are equally present in anyof the latercontributions tothe subject (Barthes, Genette, Riffaterre . . .), andthey will be inthose approachesstill to come. In whatfollows, I willtry to explainthe main aspects of somefundamental views on thesubject of intertextuality,beginning with Bakhtin as thecrucial mediator between twentieth-centuryintertextual theories and thosetraceable antecedents (in theclassical period,the Middle Ages, the Renaissance, etc.), already commented on. The nextlink in thechain will be theone constituted byJulia Kristeva, who, as we havenoted, introduced theterm in herinterpretation of Bakhtin's work for the Western public, adding to his theorieselements that she had taken from such fields as formallogic, psychoanalysis and deconstruction(Worton and Still 1990, 16). To endwith, I willanalyze some of thelater theoriesof intertextualityas they have developed from its firstintroducers. Since this centuryhas seen such a widevariety of positions on thesubject that concerns us, I willnot attemptto cover the material exhaustively, but rather to give a briefanaly sis of someof the mostimportant and influential theorists after Bakhtin and Kristeva.

2. THE ORIGINSOF THE TERM: MIKHAILBAKHTIN AND JULIAKRISTEVA AlthoughBakhtin started publishing as earlyas 1919, for a varietyof reasons (personal,political . . .) itis onlyover the last twenty or so yearsthat he hasslowly come to be recognizedin criticalcircles outside Eastern Europe, His theoryof language (everyday dialogism)and of the poles of literature (the monologic "and the dialogic) can be takenas a powerfulprecursor of and influenceon the developmentof later approachesto intertextuality. Whatcan onlyuneasily be called"Bakhtin's philosophy" is a pragmaticallyoriented theoryof knowledge, one amongother modern epistemologies that seek to grasphuman behaviourthrough the use we makeof language. Bakhtin's distinctive place among these is specifiedby the dialogic concept of language that he proposes as fundamental. Dialogism's immediatephilosophical antecedents are to be foundin the attempts made byvarious neo-Kantians to overcomethe gap between"matter" and "spirit".Moreover, Kant's argumentthat there is an unbridgeablegap betweenmind and worldis the conceptualrock on whichdialogism is founded.One ofits basic aims is to framea theory of knowledgefor an age whenrelativity dominates physics and cosmologyand, thus, whennon-coincidence ofone kind or anotherraises troubling questions and does away withthe old conviction that the individual subject is thesite of certainty, whether the subject so conceivedis namedGod, the soul, the author ... (Holquist1990, 17,19). For Bakhtin, the"self is dialogic,it lives in a relationof simultaneity with the "other": consciousness is othernessor, more accurately, itis thedifferential relation between a centreand all thatis nota centre.The self,then, may be conceivedas a multiplephenomenon of essentially threeelements: a centre(I-for-itself), a not-centre (the-not-I-in-me), and therelation betweenthem (Holquist 1990, 29). Dialogismis thename not just for a dualism,but for a necessarymultiplicity inhuman perception.We arein dialoguenot only with other human beings and with ourselves, but also withthe natural and the cultural configurations we lumptogether as "theworld". In sum,dialogism is basedon theprimacy of the social, and the assumption that all meaning is achievedthrough struggle.

ATLANTISXVIII (1-2) 1996 UNTERTEXTUALITY:ORIGINS AND DEVELOPMENTOF THE CONCEPT 273

Anyanalysis of Bakhtin's theory about literary discourse must begin with his view on language,since the former is a directconsequence of thelatter. Just as he triesto avoidan essentialistunitary conception of theself, Bakhtin gives language the same treatment. His projectis nota linguisticsbut, to use his word,a "metalinguistics",a view of language withina socialand historical frame. This metaposition, however, is notso mucha move towardstranscendence as it is a battlestance, a polemicalinsistence upon situating theories of languagewithin the constraints of theirparticular social and historicalperiod. He systematicallyquestions and subvertsthe basic premisesand argumentsof traditional linguistictheory. Thus, while Saussure is interestedin languageas an abstractand ready- madesystem, Bakhtin is interestedonly in the dynamics of living speech. Where Saussure speaksof passive assimilation (in relationto languageas opposedto speaking),Bakhtin sees a processof struggleand contradiction.And whereasSaussure dichotomizes the individualand the social, Bakhtin assumes that the individual is constitutedby the social, thatconsciousness is a matterof dialogue and juxtaposition with a social"other". Insteadof working with the old dualism of system and performance, he posits commu- nicationand not language as thesubject of hisinvestigations. Language, as conceivedby mostlinguists, would embrace grammar, lexicon, syntax and phonetics;discussions of wordcombinations wouldn't include a unitmore comprehensive than the sentence. All the- se featuresplay a roleas wellin Bakhtin's metalinguistics, butas dynamicelements in con- stantdialogue with other features that come into play only in particularacts of commu- nication.So, tothe list of topics appropriate tothe study of language should be addedthose appropriatetothe study of communication as well: utterances, Bakhtin's fundamental unit ofstudy, and speech , the conventions by which utterances are organized. ForBakhtin, unity or plenitude in language can only be an illusion.Literary authors can attemptartificially tostrip language of others' intentions, a unifying project which he calls monologismor poetry.On theother hand, at certainhistorical moments, writers have artisticallyelaborated and intensified heteroglossia, creating what he calls the(dialogic) novel.As Wortonand Still point out (1990, 15), it is importanttonote that these categories do notcorrespond to traditionalones, for example, Heine's lyricverse is includedby Bakhtinin the category of "novel", whereas Tolstoy's prose is presentedas monological. Monologismhas, according to Bakhtin(1981, 271), beenencouraged or imposedby hierarchicalorcentralizing socio-linguistic forces such as AristotelianorCartesian po etics, themedieval church's "one language of truth", or Saussurean linguistics. The novel, on the otherhand, has been shaped by iconoclastic, even revolutionary popular tradi tions, among whichthe carnival plays a veryimportant role. It is in popularlaughter, in theparody and travestyof all highgenres and lofty models that the roots of the novel are to be sought. Thisapplies in particularto the Socratic dialogues and also toMenippean satire, in which laughteris usedas a weaponagainst authority and the established hierarchies. Whereas the epicworld, and the world of high literature ingeneral, is absoluteand complete, closed as a circleinside which everything is past and already over, low genres and, in a broadsense, thecommon people's creative culture of laughter,deal withcontemporaneity, flowing, transitory,"low", present (Bakhtin 1981, 20). Undersuch an influence,the novel appears as uniqueamong all genresdue to its ability to change and de velop.Moreover, every time thenovel, in any of its manifestations, has triedto adoptand keep a stable,fixed form, it hasbeen heavily attacked and criticized(Bakhtin 1981, 6). This is whathappened, for instance,with the chivalry romance and thesentimental novel. However, it is fromthe parodyof previousnovelistic traditions that some of themost important works in the historyof literature have taken their force. From this perspective, nothing conclusive can be said aboutthe novel,which must always be seen as open, free,and in continuous development.

ATLANTIS XVIII (1-2) 1996 274 MaríaJesús Martínez Alfaro

The novel constitutes,for Bakhtin, the highest incarnation of the dialogical play that characterizesall discourse.No utteranceis devoid of dialogical dimension.The only dis- tinctionthat can be drawnin thisregard is notbetween discourses endowed with dialogism and thosedevoid of it,but between the two roles, one weak and one strong,that dialogism can be called to play. Yet, at times,he is temptedto inscribeit into a single opposition wherethe "dialogical" utterancewould face a "monological"one, as is the case withthe relationthat he establishesbetween prose and poetry(Todorov 1984, 63-8). Fromas earlyas thefirst edition of Problemsof Dostoevsky'sPoetics (1929), and es- pecially in the chapterentitled "Discourse in the Novel", prose, which is dialogical, is opposed to poetry,which is not. In poetry,language is conceived as unitaryand, conse- quently,the utterance appears as self-sufficient,with no relationto otherutterances external to it. The poet fullyassumes his/herspeech act and regardseach wordas his/herown, as a pureand directexpression of his/herintentions. The novelist,on the otherhand, does not exclude the intentionsof others as they are presentin every utterance,and allows heteroglossiato enterhis/her work, thus creating a unique artisticproduct out of such a diversityof voices (Bakhtin 1981, 297-98). Prose and poetryare, then,the resultof opposedtendencies: one is centrifugaland leads to pluralityand variation(the novel), the otheris centripetaland is associatedwith the unitary and thesingle (the poetic genres). This de-centralizingimpulse that characterizes the novel accounts,from the very beginning, for its parodiequality and its oppositionto any kindof authority.While poetrycontributes to thecultural and politicalcentralization of theverbal and ideological worlds,the novel has, fromits origins,a de-estabilizingfunction that opposes it to officiallanguage. Irreverent, like the carnival, the novel turnsout to be the subversive and liberating par excellence. Intimatelylinked with this subversive and liberatingquality is the novel's capacityto questionitself and its own conventions(Bakhtin 1981, 39). Consideringthe ways in which novelisticdiscourse may overtly appear as self-reflexive,Bakhtin (1981, 412-13) mentions thoseworks centred on whathe calls "the literaryman", who sees life throughthe eyes of literatureand triesto live accordingto it. Don Quixoteand Madame Bovary are the best knownexamples of thistype, but the "literary man" and thetesting ofthat literary discourse connectedwith him can be foundin almostevery major literarywork. There are also those which introducean authorialfigure in the , someone thatcomments on his/hercreative task and on the process of writing(a "laying bare of device", in the terminologyof the Russian formalists).In thatway, we have not only the novel in its propersense, but also fragmentsof a sortof "novel about the novel". TristramShandy is perhapsthe best example and an importantprecedent of theself-reflexive quality central to Britishpostmodernist literature. Before finishingthis surveyof Bakhtin's ideas on language and the novel, it would perhapsbe convenientto call attentionto two aspects of his theorywhich may be mis- leading.The firsthas to do withthe positionof the authorin relationto his/herwork. As has alreadybeen noted,in the monologicalgenres the authorsubordinates all the voices presentto his/herown intentions.In a differentway, the prose writerdoes not impose a single criterionto thereader but, on the contrary,allows each voice to keep its own in- tegrityand independence.However, Bakhtin does notgo so faras to arguethat the author is absentfrom his/her text: thereis no unitarylanguage or stylein thenovel. But at thesame timethere does exista centerof language(a verbal-ideologicalcenter) for the novel. The author(as creatorof the novelistwhole) cannot be foundat any one of the novel's language levels: he is tobe foundat thecenter of organization where all levelsintersect. (Bakhtin 1981, 48-9)

ATLANTIS XVIII (1-2) 1996 INTERTEXTUALITY:ORIGINS AND DEVELOPMENTOF THE CONCEIT 275

Thisassertion separates Bakhtin from poststructuralist critics, with whom he has beenas- sociatedon account of his defence of the polyphonic character of language in general and of thenovel, in particular. The otherpoint I wouldlike to commenton goes backto whathas alreadybeen said aboutthe novel's link with the present, a present which is alwayschanging and moving to- wardsan equallyinconclusive future. In spiteof this,the novel can have,and often has, thepast as itscentral subject. But even when this is thecase, the present and its open cha- racterwill always be thebasis on whichthe portrait of the past is structured.According to AídaDíaz Bild (1994, 140),this direct contact with contemporary reality has important consequencesfor the novelistic discourse. One of them has to do withthe author's ability tomove freely within his/her field of representation (something unthinkable in thecase of theepic), which makes possible the introduction ofone ofthe novel's basic features: its literaryself-consciousness. In addition, this new temporal orientation puts it in contact with extra-literarygenres, that is, witheveryday reality and ideology. This fact enhances even morethe open quality of thenovel, since literary evolution brings with it notonly the introductionof changeswithin existing limits but also themodification of suchlimits: "Afterall, the boundariesbetween fiction and nonfiction,between literature and nonliteratureand so forthare not laid up in heaven" (Bakhtin 1981, 33). Accordingto Bakhtin, everything stands under the same : the sign of plurality. Our livesare surrounded by the echoes of a dialoguethat undermines the authority of anysingle voice,a dialoguethat takes place within the text, but which is, at the same time, a dialogue withall thevoices outside it. Unlike the third eye ofTibetan Buddhism, which gives those whopossess it a visionof the secret unity holding creation together, Bakhtin seems to have hada thirdear that permitted him to hear differences where others perceived only sameness, especiallyin the apparent wholeness of the human voice (Holquist 1983, 307). * * * Oneof the most important, and earliest interpretations of Bakhtin's work for a Western publicwas theone byJulia Kristeva. In thelate 1960s,Kristeva subscribed to the Tel Queliannotions of textual, i.e. cultural,revolution. She saw Bakhtin'sconcept of dialo- gismas quintessentiallydynamic, even revolutionary. In her view, what it triedto revolu- tionizedynamically was not only the static structural model but cultural politics in general. Bakhtin,in propagatingthe relativity of each singleposition, the self-criticism of each word,the undermining of all dogmaticand officialmonologism, the carnivalesque profanizationofall thatis sacredand the subversion of all authority,was fightingagainst theincreasing rigidity of post-revolutionarySoviet cultural politics and thedoctrinary canonizationof SocialistRealism. He was,in fact,continuing the revolutionary struggle againstrepression. ( It wasthis revolutionary potential of Bakhtin's criticism of ideological monologism that fascinatedKristeva and other writers of the Tel Quel circlein thelate sixties, and they, in turn,employed Bakhtin's concept of dialogism in their own struggle against the bourgeois ideologyof theautonomy and unity of individualconsciousness and theself-contained meaningof texts. Kristeva tried to achievethis objective by fusing ideas from philosophy (Husserl/Derrida),political science (Marx/Althusser) and psychoanalysis(Freud/Lacan) withthe procedures of structural linguistics (Chomsky) and formal logic. For Kristeva,Bakhtin represents the possibility of openinglinguistics to society: "Bakhtinsituates the text within history and society, which are seen as textsread by the writer,and into which he insertshimself by rewritingthem" (Kristeva 1980, 65). Fol- she also calls lowingBakhtin, Kristeva attempts to transformsemiotics into something the "translinguistics"(1980, 37), a methodof analysis that allows her to confront literary

ATLANTIS XVIII (1-2) 1996 276 MaríaJesús Martínez Alfaro workon theformal and the social levels simultaneously. At thesame time, however, she transformsBakhtin's concepts by causing them to be readin conjunction with ideas about textualitythat were emerging in France in the mid-sixties. For instance, she slips "text" into a paraphraseof Bakhtin: "each word (text) is an intersectionofwords (texts) where at least oneother word (text) can be read"(1980, 66). Thoughthe parentheses imply that Kristeva is onlysupplying a synonym,or at most,a neutralexpansion of Bakhtin's concept, this textualizationofBakhtin changes his ideas, changes them just enoughto allow thenew conceptof intertextuality toemerge. Kristeva's term became quite popular but it did not do so, however,because of its own coherence: the face of "intertextuality", as a new master term,is less a simple,single, precise image, a bronzehead by Rodin,than something shattered,a portrait bust by an avidexponent of analyticcubism too poor to afforda good chisel(Clayton and Rothstein 1991, 11). In "Word,Dialogue, and Novel" Kristevaintroduces Rabelais and His Worldand ProblemsofDostoevsky's Poetics. However, her reading of Bakhtin is mediatedby other textsand other critics' theories, not to mentionthe political motivation behind her work: "thereis no equivalence",she writes,"but rather, identity between challenging official linguisticcodes and challenging official law" (1980, 65). In thelast resort, hers is a polit- ical conceptwhich aims at empowering the reader/critic to oppose the literary and social traditionatlarge. Amongthe authors that mediate Kristeva's reading of Bakhtin, Derrida plays a crucial role.When she characterizes Bakhtin's "conception of 'theliterary word' as an intersection oftextual surfaces rather than a point(a fixedmeaning), as a dialogueamong several writings"(1980, 65, italics in the original), one cannot help but notice Derrida' s critiqueof voicebehind this slight shift towards a dialogueof "writings", not "utterances", particularly sinceKristeva cites Of Grammatology on the first page of her essay. A Derrideanview of writingsupplies a dimensionthat was notpresent in Bakhtin originally, the dimension of indeterminacy,of différance, of dissemination.Although Bakhtin's notions of "heteroglossia"or "hybridization"might seem near equivalents to thepoststructuralist concepts,Bakhtin's emphasis on thehistorical uniqueness of the context of every utterance distanceshis terms from the endlessly expanding scope of intertextuality.In Kristeva's usage,the intersection oftextual surfaces in a literaryword can neverbe circumscribed,it is open to endless dissemination.In fact,Derrida's readingof Saussure in Of Grammatologymust be regardedas a crucialintertext of mosttheories of intertextuality. The wayin whichhe subordinatesdifference between the signifier and signifiedto the differencebetween one signifier and another, his notion of the general text (// n'y a pas de horstexte),as well as his definitionof iterability,which leads to a view of textsas inevitablyquoting and quotable,provide ample space withinthe object of studyfor a multitudeofintertexts. In muchthe same way,Lacan functionsas a largelyunacknowledged intertext for Kristeva'saccount of Bakhtin.She notesthat Bakhtin's claim that the language of epic is univocalcannot withstand a psychoanalytic approach to language. It is psychoanalysisas wellas thesemiotics she cites (the theory of Benveniste), that reveal dialogism to be in- herentin every word, as thetrace of a dialoguewith oneself (with another), and a writer's distancefrom him/herself (1980, 74). Thispsychoanalysis is Lacan's andhis ideas may havelain behind Kristeva's choice of the term "ambivalence" to describe certain forms of dialogism(Clayton and Rothstein 1991, 19-20). Fromthis and other modifications of Bakhtin,then, there emerge Kristeva's several "definitions"ofintertextuality:

ATLANTISXV'''{'-2) 1996 INTHRTEXTUALITY:ORIGINS AND DEVELOPMENT OFTHE CONCEPT 277

anytext is constructedas a mosaicof quotations; any text is theabsorption and transforma- tionof another.The notionof intertextualityreplaces that of intersubjectivity,and poetic languageis readas atleast double. (1980, 66) Dialogueand ambivalence lead me to conclude that, within the interior space of thetext as wellas withinthe space of texts, poetic language is a "double".(1980, 69) Thewriter's interlocutor [ ... ] is thewriter himself, but as readerof anothertext. The one whowrites is thesame as theone whoreads. Since his interlocutor is a text,he himselfis no morethan a textrereading itself. The dialogical structure, therefore, appears only in thelight ofthe text elaborating itself as ambivalentin relationto another text. (1980, 86-87) Forthe practising critic, Kristeva's conception of intertextuality opens several lacunae thatdo notappear in Bakhtin(Clayton and Rothstein 1991, 20). The firstone involvesa vaguenessabout the relation of the social to the literary text. Kristeva does not discuss what happensto a fragmentofthe social text when it is "absorbed"and transformed byliterature, nordoes sheaccount for how specific social texts are chosen for "absorption". A second problem,the inability to constructa convincing literary history, follows from the first. Kristeva(1980, 71) claims,for example, that a breakoccurred at the end of the nineteenth centurythat clearly marks off the dialogism of Joyce, Proust, and Kafka from the dialogical novelsof the past, including Bakhtin's principal examples, Rabelais and Dostoevsky. But herconception of intertextuality generates no meansof distinguishing the modern novels fromthose earlier polyphonic works. In spiteof all this,the notion and the concept introduced by Kristeva soon spread all overEurope, and flowed across the Atlantic into the discourses of American criticism as a major,shaping influence on literaryhistory and theoryfrom the mid-seventies until the present.But like any influential idea, it changed as itwas assimilatedwithin different cul- turaland intellectual contexts. Theorists, of course, try to regularize it, but adaptations of intertextualityhave refused orthodoxy, often encouraging multiple interpretations.

3. LATER THEORIESOF INTERTEXTUALITY: MAIN LINES OF DEVELOPMENT As Claytonand Rothstein (1991, 21) pointout, later approaches to intertextuality have keptand further textualized Kristeva's strong basis in :the science of signsde- velopedby Saussure and still intact in Barthes' Elements of Semiology (1964) is treated lessas a sciencewith objective aims and verifiable results and more as a modeof interpre- tation.Theories proceeded from this point in three directions (Clayton and Rothstein 1991, 21): (1) thedeconstructive path of aporia and the reader's puzzlement or play; (2) whatcan be regardedas an attemptto limit the endlessly expanding intertextual space suggested by theprevious deconstructive approach, in an effortto find out the basic crite ria of a method thatcan throw some light on the practical analysis of intertextuality in literature; and (3) the socialor political path taken by cultural materialism or newhistoricist criticism. Barthes willserve to exemplify (1); MichelRiffaterre, Jonathan Culler and GérardGenette will illustratesome of thebasic aspectsof (2); and Rezeptionand Foucaldiancriticism will appearperhaps as thebest representatives ofthe last approach (3). Even beforeKristeva's 1965 presentationof Bakhtinin Barthes'seminar, Roland Bartheswas evokingsomething like intertextuality under the name of crytographie.In Le degrézero de Vécriture (1972) he uses the term thus: Sin dudapuedo hoy elegir tal o cualescritura, y con ese gestoafirmar mi libertad, pretender unfrescor o una tradición; pero no puedoya desarrollarlaen unaduración sin volverme poco a pocoprisionero de las palabrasdel otroe inclusode mispropias palabras. Una obstinada remanencia,que llegade todaslas escriturasprecedentes y del pasadomismo de mipropia

ATLANTISXVIII (1-2) 1996 278 MaríaJesús Martínez Alfaro

escritura,cubre la vozpresente de mispalabras. Toda huella escrita se precipitacomo un elementoquímico, primero transparente, inocente yneutro, enel quela simpleduración hace aparecerpoco a pocoun pasado en suspensión, una criptografía cada vez más densa. (Barthes 1973a,25) Thisis intertextualityin the sense that a textmay appear to be thespontaneous and trans- parentexpression of a writer'sintentions, but must necessarily contain elements of other texts.Barthes provides an extraordinaryexample of this in S/Z where he picksout some of the quotationswithout marks, some of the referencesto culturalcodes, stereotypes,received wisdom and so on inBalzac's Sarrasine. LikeKristeva, Barthes holds that the limitations ofthe linguistic-structuralist approach have to be overcomeby meansof a meetingof differentepistemes, namely dialectical materialismand psychoanalysis. This new method will produce a newobject that we call textand which is intertextualbydefault: other texts are alwayspresent in it,at varying levelsand in more or less recognizable forms (Barthes 1987, 39). Barthes'vision of intertextuality alsohighlights the frequent anonymity ofthe "sources" of intertextualquotations. This idea was implicitin Kristeva's discussionof the "absorption"of social texts,because the social maybe thoughtof as the networkof anonymousideas, commonplaces, folk wisdom, and clichés that make up thebackground ofone's life.Whereas traditional influence studies primarily hunted for allusions to cele- bratedworks of the past, Barthes, however, makes the commomplace central: "the whichgo tomake up a textare anonymous, untraceable, and yet already read" (1990, 160). The "alreadyread" in Barthesencompasses more than the idea thatwe all possess conventionalknowledge whose sources we cannotrecall. It extendstowards a notionof the subjectas constitutedby the texts of his/her culture, the subject as alreadyread: "This T whichapproaches the text is alreadyitself a pluralityof othertexts, of codes whichare infiniteor, more precisely, lost" (1974, 10). Kristevaherself has consistentlyargued, in accordancewith new French psychoanalytic theory, for this re-definition ofthe subject as alwaysalready cleft asunder or evenradically dispersed. The fracturingof thereading subjectis inevitablyassociated with the dissolution of the author, or death of the author as Barthesputs it. This impliedrejection of authoritydoes notcorrespond exactly to the politicalor even revolutionary thrust which Kristeva emphasizes in Bakhtin. Barthes tends tosound rather neutral in his sense as heseems ever ready to politicize matters of taste, but alsoto aestheticize political issues. Valuableas Barthes'account of intertextuality is,it does not provide the critic with a particularlyeffective tool for analyzing literary texts. Clayton and Rothstein(1991, 23) pointto the fact that Barthes' radical intertextuality foregoes the possibility of rigour in the discussionof individualtexts, so muchso thatto attemptsuch a rigorousdiscussion, he mustretrench on the theory.This theory,however, has a real heuristicor, at least, iconoclasticvalue in unsettling customary ideas about the author, the work, and the repre- sentationof reality.

The secondpath, that in whichintertextuality is used to achievegreater interpretive certainty,has beentaken by critics who have applied it rather effectively totheir practical criticism,such as MichelRiffaterre, Jonathan Culler and Gérard Genette. While the latter privilegesthe literary text in itsnarrower sense, the former concentrates on theact of readingand Culler relies on the linguistic method and its analogies with literary discourse in orderto carry out his critical task from an intertextual perspective. However, in spiteof the differences,their approaches are equallybent on establishingcertain limits to the intertextualscope of every particular text.

ATLANTISXVIII (1-2) 1996 INTERTEXTUALITY:ORIGINS AND DEVKLOPMKNTOF TI IE CONCEPT 279

Takinginto account the relevant role of the reader, Riffaterre approaches intertextuality notonly from the point of viewof all thepossible relations among texts but as themain, fundamentalcharacteristic of(literary) reading. He definesthe literary phenomenon as not onlythe text, but also its reader and all thereader's possible reactions to the text (1983, 3). Althoughhe hassometimes been called a structuralist,thislabel needs qualification, for herejects the structuralist search for a deepgrammar in literature,as wellas thenotion that all literaryworks of a giventype share the same structure. On thecontrary, he believesthat theonly significant structure ina literarywork is thatwhich the reader can perceive. Yet he mustalso be distinguishedfrom the reader-response critics, in thathis work is basedon a concernwith textual elements that readers are obliged to recognize. He distinguishestwo stagesof reading(1980, 625-27). The firstone is a naive, "mimetic"reading which yields what he calls the"meaning" of a work,the linear, word- by-worddecoding of themessage in accordancewith an assumptionthat language is referential,that words relate directly to things. In thecourse of thisreading, however, one encounters"ungrammaticalities" - difficulties, obscurities, undecidable moments, figu- rativelanguage - anywording so unacceptablein a mimeticcontext that it promptsthe readerto look elsewhere for the "significance" ofthe work. This emerges only in a second stageof reading,no longerlineal but comparative.Riffaterre considers two possible, thoughnot exclusive, ways of reading comparatively: retroactive reading and intertextual reading.The former refers to the way in whichthe reader keeps reviewing and comparing backwards,recognizing repetitions and variations upon the same structure(s). Intertextual reading,on theother hand, is theperception ofsimilar comparabilities from text to text; or it is theassumption that such comparingmust be done even if there is nointertext athand wherein tofind comparabilities. (1980,626) Ambiguityexists only as a stagein the reading process and serves to alertthe reader to thepresence of an intertextthat will resolve the work's difficulties. These function as traces leftby the absent intertext, as signs of an intertexttobe completedelsewhere. Such "clues" areenough to setin trainan intertextualreading, even if the intertext is not yet known or hasbeen lost with the tradition itreflected (1980, 627). Rejectingthe poststructuralist dispersal of meanings, Riffaterre claims that there is only onecorrect reading and that it is theintertextual method that guides the reader in his/her interpreting.According to him, the ability to recognize gaps and ungrammaticalities arepart of everyreader's linguistic competence and it does not requiremuch erudition or "preternaturalinsights" (1987, 373). Yet hisown interpretations ofpoems and novels are fullof learnedallusions and draw on an encyclopaediccommand of French and English .Anyway, what is relevantin histheory is hisbasic concern with the effect on thereader of a textualpresupposition: readers presuppose that there is an intertextwhich givesstructural and semantic unity to thework, but the success or failureto locatethat intertexton thepart of the reader is, in a sense,irrelevant to theexperience of intertextual reading.Analogous if not identical with Kristeva's assertion that every text is underthe jurisdictionof other discourses, Riffaterre's thesis is thatliterary reading is possibleonly if thereader recognizes that the text articulates a presupposition ofintertext, tosuch an extent thatthe text can be considerednot simply a sequenceof words organized as syntagmsbut a sequenceof presuppositions (1980, 627). Theconcept of presupposition is also a centralone in the theory of Jonathan Culler. The exampleof Riffaterre illustrates the logical independence of intertextualityfrom many poststructuralistassumptions. Culler also arguesfor the constrainingpower of in- tertextuality,although not so absolutelyas Riffaterre,for whom, as we havenoted, there is

ATLANTISXVIII (1-2) 1996 280 MaríaJesús Martínez Alfaro onlyone properinterpretation of a text,the one thatis reached throughthe intertextual method.In "Presuppositionand Intertextuality"(1976), Culler suggeststhat we startthe studyof intertextualityby consideringits linguisticdimensions and in termsof two kinds of presuppositions:logical and pragmatic.The formerare best thoughtof as the presup- positionsof a sentence.Thus, the question"Have you stoppedbeating your wife?" pre- supposesthat one previouslymade a habitof beatingone's wife (Culler 1976, 1389). On theother hand, a sentencelike "Once upona time",though poor in logical presuppositions, is extremelyrich in pragmaticones sinceit relates the story that follows to a series of other stories,identifies it with the conventions of a genre, etc. (Culler 1976, 1392). The linguisticanalogy suggests, according to Culler,two ways of approachingintertextuality: Thefirst is to lookat thespecific presuppositions of a giver)text, the way in whichit pro- ducesa pre-text,an intertextualspace whose occupants may or maynot correspond to other actualtexts ... The secondenterprise, the study of rhetoricalor pragmaticpresupposition, leadsto a poeticswhich is less interestedin theoccupants of thatintertextual space which makesa workintelligible than in theconventions which underlie that discursive activity or space.(1976, 1395) The practicalvirtue of Culler's proposalis, then,that it limitsthe set of possible intertexts to thosewhich are eitherlogically or pragmaticallysuggested by thework one is studying, and thatit does so withoutexcluding the anonymous,already read discourseof thesocial text,which is usuallyignored in influencestudies (Culler 1976, 1383). Fromthis point of view, Culler also calls attentionto the complex quality of the relationshipthat exists betweeninfluence and intertextuality.When he proposes to follow the linguisticmodel (and, in particular,the notionof presupposition),he does so as a means of avoidingthe dangerof settingout to studyintertextuality and focusing,in theend, on a text'srelation to specificprecursors, something more in consonancewith influence studies. Accordingto Culler, this is precisely the sort of mistake that Kristeva makes in her analysis of Lautréamont'sPoésies, a disappointingreading for anyone under theimpression that the wholepoint of intertextualityis to take us beyondthe study of identifiablesources (Culler 1976, 1384-85). Like Michel Riffaterre'sand JonathanCuller's proposals,Gérard Genette' s approachto the subjectof intertextualitycan be consideredas an attemptto delimitthe definitions of intertextualityput forwardby Kristeva,Derrida, Barthes, etc., as theyhave been found difficultto applyto thepractical analysis of texts.In contrastwith Bakhtin's and Kristeva's wide interests,which are not only linguisticbut also social, political,philosophical ..., Genetteconcentrates basically on theliterary text in thestrict sense of theword. Reading Kristeva'snotion of intertextualityas referring to theliteral and effectivepresence in a text of anothertext, he assertsthat intertextuality is an inadequateterm and proposesin itsplace , by whichhe meanseverything, be it explicitor latent,that relates one textto others.Therefore, though he centreson theparticular literary text, he acknowledgesthat it can no longerbe studiedin isolation. El objetode la poética[ ... ] no es el textoconsiderado en su singularidad[ ... ] sinoel ar- chitexto. . . Hoyyo diría, en unsentido más amplio, que esteobjeto es la transtextualidado la transcendenciatextual del texto. (1989, 9) In Palimpsestes,his "last word"upon intertextuality,Genette insists on theglobality of thenotion of transtextualityand offersfive subcategories (1989, 10-15): 1. Intertextuality:the relation of co-presencebetween two or more texts,that is, the ef- fectivepresence of one textin anotherwhich takes place by meansof , quotationor .

ATLANTISXVIII (1-2) 1996 INTERTEXTUAUTY:ORIGINS AND DEVELOPMENT OFTHE CONCEPT 28 1

2. Paratextuality.the relations between the body of a textand its title, subtitle, epigraphs, illustrations,notes, first drafts, and other kinds of accessory signals which sur- roundthe text and sometimes comment on it. 3. Metatextuality:therelation, usually called "commentary",which links one textwith anotherthat comments on itwithout quoting it or, even, without mentioning itat all.It is thecritical relation par excellence. 4. Archtextuality:thegeneric category a textbelongs to. The textmay not recognize its genericquality, which should be decidedby its readers, critics . . . However,this genericperception determines to a greatextent the reader's "horizonsof expectation",and, therefore, the work's reception. 5. Hypertextuality:therelation between the latecome text (hypertext) and its pre-text (hypotext).He defineshypertext as every text derived from a previousone by meansof director indirecttransformation (imitation), but not throughcom- mentary.In theformer, direct or simpletransformation, a text B maymake no - explicitreference to a previousone A, butit couldn't exist without A. For in stance,The Eneyd and Ulyssesare, in differentdegrees, two hypertexts of the samehypotext, The .Imitation is a morecomplex kind of transforma- tion,since it requires the constitution ofa genericmodel. In spiteof this complex and detailed classification, the five categories established by Genettetend to overlap when it comes to the practice. For example, the may also contributetodetermining the generic quality of the text, thus merging with archtextuality. Hypertextualityseen as thepresence of one textin anothertext does notseem to be very differentfrom intertextuality. Only by restricting the latter notion to plagiarism, quotation or allusion,and the former to parody,travesty and pasticheis he able to keep themapart. However,both categories fall together again when he acknowledgesthat nohay obra literaria que, en algún grado y segúnlas lecturas,no evoque otra, y, en este sentido,todas las obras son hipertextuales ... (1989, 19) In such a generalstatement as this,one could equally say "intertextual"instead of "hypertextual". Despitethe complex terminology and the unavoidable overlapping, Genette' s concepts maybe takenas a usefulpoint of departureto be sufficientlydefined later on by each individualcritic/theorist, and,above all, they contribute tounderlining the complexity of the notionof intertextuality. Each textis trappedin a networkof relations,between the differentparts that constitute it, between that text and those which precede it, or those that comeafter it, or even thosewhich never were (Borges' pseudo-textuality).In turn, all thoserelations can be saidto existin as faras theyare perceived by the reader, who may discoveran echo,but may equally silence it. Genette himself goes as faras toassert that the of its hypertextnecessarily gains in someway or anotherfrom the reader's awareness signifyingand determining relationships with its hypotexts (1989, 494), thusconfirming LindaHutcheon's assertion that most discussions of intertextuality endup consideringthe roleof the reader, no matterhow formalistic they have attempted tosound (Hutcheon 1986, 232). As Wortonand Still (1990, 23) pointout, Genette's analysis of individual texts may be less sustained,less close than Riffaterre's, but his work does insistently remind us that well as memorycan be actively"revolutionary" only so long as it is creativeas commemorative. * * *

The thirdpath, that of puttingintertextuality at the service of politicaland historical andcritics associ - projects,has becomeidentified with two schools, Rezeptions-Ästhetik

ATLANTIS XVIII (1-2) 1996 282 MaríaJesús Martínez Alfaro atedwith Michel Foucault.The formerhas triedto charthistorical development by looking at the ways in whichthe intertextual connections that a textevokes change over time.Its leadingproponents, Hans RobertJauss and WolfgangIser, do not relyextensively on the termintertextuality, but theirinvestigations into continuityand change employ related notions.For Claytonand Rothstein(1991, 26), Jauss' Gadameriannotion of "horizonof expectations"that a readerbrings to a workresembles intertextuality, because thereader's horizonis constructedby an inheritedsystem of normsand conventions.To the studyof thisintertextual field, Jauss' receptioncriticism adds a historicaldimension by tracingthe ways in whichdifferent readers' horizonsdiverge from one anotherover time.As far as Iser's reader-responsecriticism is concerned,it employsa similarintertextual concept, the notionof "repertoireof thetext", a repertoirethat exists only in thereader and is activated by "referencesto earlierworks, or to social and historicalnorms, or to the whole culture fromwhich the text has emerged"(in Claytonand Rothstein1991, 26). Foucault's pathleads towardsa conceptionof intertextualitythat emphasizes the role played not only by discursivebut also by non-discursiveformations such as institutions, professionsand disciplines.Unlike Barthesand Derrida,with their boundless visions of textuality,Foucault highlights the forces that restrict the free circulation of thetext. Among themhe citesthe author principle, that of commentaryand thatof discipline: We tendto see, in an author'sfertility, inthe multiplicity ofcommentaries and in thedevel- opmentof a disciplineso manyinfinite resources available for the creation of discourse. Perhapsso, but they are nonetheless principles of constraint, and it is probablyimpossible to appreciatetheir positive, multiplicatory role withoutfirst taking into consideration their restrictive,constraining role. (Foucault 1972, 224) To those "principlesof constraint"he also adds theconditions under which discourse may be employed.Although every textpossesses countlesspoints of intersectionwith othertexts, these connections situate a workwithin existing networks of power,simulta- neouslycreating and discipliningthe text's abilityto signify.Foucault insiststhat we ana- lyze therole of powerin theproduction of textualityand of textualityin theproduction of power.This entailslooking closely at thosesocial and politicalinstitutions by whichsub- jects are subjected,enabled and regulatedin formingtextual meanings. Even if he regards thetext as a siteof "anonymity"and theauthor as a "role function"played out in thetext, Foucaultdoes notagree with Barthes' isolationof thetext from history and ideology. His concept of culture as intersectingdiscourses representsa form of the concept of intertextualitythat emphasizes the production of ideology. Foucault's neglectof genderissues has oftenbeen notedand historicistcriticism in the eightiesand ninetieshas generallyattempted to correctthis lacuna in Foucault's project,so much so as to suggest thathistoricist critics should begin by hyphenatingrace-class- gender.In line withthis, and in relationto theproblematic question of literarycanons, Paul Lauter(1993, 242) assertsthat an adequate theoryof criticismcan only be developed by fully considering the art produced by women, by working people and by national minorities. Oppositionalcriticisms have also adoptedand criticizedthe conceptof intertextuality. Justas AnnetteKolodny thoughtthat feminist theorists must revise Bloom's notionof influence,1feminist and criticsof colourhave begunto rethinkthe notion of intertextuality.

1 Thereare no femalefigures against which a womanwriter can reactin an attemptto demarcate herselffrom them, since not only Bloom's but any literary canon among those traditionally pro- posedinclude no womenwriters. In addition,Bloom's theoryreproduces very specifically the

ATLANTISXVIII (1-2) 1996 INTHRTKXTUALITY:ORIGINS AND DEVELOPMENTOF THE CONCEIT 283

Claytonand Rothstein (1991, 28) mentionthe example of BarbaraJohnson, for whom "questionsof gender may enrich, complicate and even subvert the underlying paradigms of intertextualitytheory" (Johnson 1987, 124). It hardlyneeds to be said thatthe work of decenteringmale-centred culture as it is expressedin language,literature, art and institutionalconfiguration has alwaysbeen a majorconcern of feminist criticism. For more thantwo decades now,feminist scholars have been reactingagainst the apparently systematicneglect of women's experience in the literary canon, neglect that takes the form ofdistorting and misreading the few recognized female writers and excluding the others. Moreover,the predominantly male authors in the canon have dealt with the female character andthe relations between the sexes in a waythat both reflects and contributesto sexist ideology(Robinson 1993, 213). The feministalternatives to the male-dominated membershipand attitudes of the accepted canon have contributed towidening and enriching theintertextual space through the recovery of lost works by women,and the restoration of thevalue of disdained genres. Even if, as LillianS. Robinsonasserts (1993, 214), feminist criticismhas tended to concentrate on writingby women, it hasalso emphasized alternative readingsof the tradition, readings that re-interpret women's character, motivations, and actions,and that identify and challenge sexist ideology: "from this perspective, Milton may comein forsome censure, Shakespeare and Chaucer for both praise and blame,but the clearintention ofa feministapproach to these classic authors is toenrich our understanding - ofwhat is goingon inthe text, as wellas how- forbetter, for worse, or forboth they haveshaped our literary and social ideas" (Robinson 1993, 214). Feministcritics' recognition ofthe individual as a sitecrossed and modelled by the dis- coursesthat surround him / her, their moving from the margins of culture an entirelitera- turethat was previouslydismissed and theiralternative approaches to traditionalworks nowpresented ina newlight, ultimately tend to support the already acknowl edged relevan- ce ofquestions of gender within the realm of intertextuality. To mention a lastexample, mostfeminist criticism has questioned the overall anonymity that surrounds the figure of theauthor in themain discourses of intertextuality. This is, forinstance, what has hap- penedin the American critical scene. Certainly many American critics have used the term in itspure French form, but in general the transplanted concept has resistedthe erasure of the writer,i Thus, N. Miller'smethod in her work "Arachnologies" is a deI berateblending of Barthesiannotions of thetext as "textile"or "web" witha clashingAmerican feminist insistenceon theimportance of theauthor. But whereBarthes' text is an infiniteweb - seeminglyspinning itself Miller insists on re-introducingthespider as author,as subject, as agent,as genderedbody, as producerof the text (Friedman 1991, 158). Manyother approaches to the subject of intertextuality could be addedto theones men- tionedhere. Such a proliferationof theories underlines the inaprehensible quality of the conceptanalysed in thisessay. Yet it is also a proofof the increasinglyrelevant rob intertextualityhasbeen playing in contemporary . Linda Hutcheon, among others,has called attention tothe fact that this particular change in criticismhas come about theway most critical changes do - thatis, primarilybecause of a changein theliterature

Oedipalconflict between sons andfathers, which makes it ultimatelyinapplicable to thecase of women. 1This does notexclude the existence of a similarreaction on thecontinent. Friedman (1991 : 176) citesas examplesWittig's reference toan intertextualityinwhich the author is stillclearly present in "TheMark of Gender" and a Britishvolume entitled A Dictionaryof Modern Critical Terms, edited byRoger Fowler, in whichintertextuality is defined under the general category "Creation" in the contextof Marxistcriticism. For Fowlerand Wittig,the author'sagency is assumedin the "practice"of intertextuality.

ATLANTIS XVIII (1-2) 1996 284 MaríaJesús Martínez Alfaro itself(Hutcheon 1986, 231). Accordingly,a literature constructed on theprinciple of intertextuality,as seems to be thecase withmost , asks for intertextual readingsas wellas intertextualmethods of interpretation. I hope that the survey developed inthis essay will be usefulto answerthe demands imposed on us by a literaturenot less complexthan the period in which it has been produced: the postmodern.

WORKSCITED Barthes,Roland 1973a (1972). El gradocero de la escritura,seguido de Nuevosensayos críticos.Trad. Nicolás Rosa. Madrid: Siglo Veintiuno. 1973b.Le plaisirdu texte. Paris: Seuil. 1974(1970). S/Z.Trans. Richard Miller. New York: Hill & Wang. 1983(1979). TextProduction. Trans. Terese Lyons. New York:Columbia Univer- sityPress. 1987(1981). Theory of the Text. Untying the Text: A Post-StructuralistReader. Ed. RobertYoung. London: Routledge and Kegan Paul. 31-47. 1990 (1977). FromWord to Text. Image-Music-Text.Trans. Stephen Heath. London:Fontana. 155-64. Clayton,Jay and Eric Rothstein, eds. 1991. Influenceand Intertextuality inLiterary His- tory.Madison: University ofWisconsin. Culler,Jonathan 1976. Presupposition and Intertextuality. Modern Language Notes 91: 1380-396. Eliot,T. S. 1971.Tradition and theIndividual Talent. Critical Theory since Plato. Ed. HazardAdams. Trans. S. H. Butcher.San Diego: Harcourt.784-87. Foucault,Michel 1972 (1971). The Discourse on Language. The Archeologyof Knowledgeand the Discourse on Language. Trans. A. M. SheridanSmith. New York:Harper & Row.215-37. Friedman,Susan Stanford 1991. Weavings: Intertextuality andthe (Re)Birth of the Author. Influenceand IntertextualityinLiterary History . Eds. JayClayton and Eric Rothstein.146-80. Galván,Fernando, Tomás Monterrey, Aída Díaz Bild and ManuelBrito 1994. Ensayos sobremetaficción inglesa . La Laguna: Secretariadode Publicacionesde la Universidadde La Laguna. Genette,Gérard 1989 (1962). Palimpsestos:la literaturaen segundogrado. Trans.C. FernándezPrieto. Madrid: Taurus. Holquist,Michael 1983. Answering as Authoring:'s Trans-Linguistics. CriticalInquiry 10, 2: 307-319. 1990.Dialog ism: Bakhtin and his World. London: Routledge. Hutcheon,Linda 1986. LiteraryBorrowing ... and Stealing:Plagiarism, Sources, In- fluences,and Intertexts. English Studies in Canada XII.2: 229-39. Johnson,Barbara 1987. Les Fleursdu Mal Arme:Some Reflections on Intertextuality.A Worldof Difference. Baltimore: Johns Hopkins U. P. 116-33.

ATLANTISXVIII (1-2) 1996 INTERTEXTUALITY:ORIGINS AND DEVELOPMENTOF TI IE CONCEIT 285

Kiely,Robert 1993. ReverseTradition. Postmodern Fictions and theNineteenth Century Novel.Cambridge: Harvard U. P. Kristeva,Julia 1980 (1977). Word, Dialogue, and Novel. Desire in Language: A Semiotic Approachto Literature and Art . Ed. Leon S. Roudiez.Trans. Thomas Gora et al. New York:Columbia U. P. 64-91. Lauter,Paul 1993(1991). Caste,Class andCanon. Feminisms. An Anthology of Literary Theoryand Criticism.Eds. Robin R. Warholand Diane Price.New Brunswick: RutgersU. P. 227-48. Plett,Heinrich F. 1991.Intertextualities. Intertextuality. Ed. Heinrich F. Plett.Berlin: De Gruyter.3-29. Riffaterre,Michel 1980. Syllepsis. Critical Inquiry 6, 4: 625-38. 1983(1979). Text Production. Trans. Terese Lyons. New York:Columbia Univer- sityPress. 1987.The Intertextual Unconscious. Critical Inquiry 13, 2: 381-85. Robinson,Lillian S. 1993.Treason Our Text. Feminist Challenges to the Literary Canon. Feminisms.An Anthology ofLiterary Theory and Criticism.Eds. RobinWarhol andDiane Price. New Brunswick: Rutgers U.P. 212-26. Todorov,Tzvetan 1984 (1981). MikhailBakhtin. The Dialogical Principle. Trans. Wlad Godzich.Manchester: Manchester University Press. Worton,Michael and JudithStill, eds. 1990. Intertextuality:Theories and Practices. Manchester:Manchester University Press.

ATLANTIS XVIII (1-2) 1996