Your SlideShare is downloading. ×
Poetics as system, by claudio guillen
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×

Saving this for later?

Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime - even offline.

Text the download link to your phone

Standard text messaging rates apply

Poetics as system, by claudio guillen

3,426
views

Published on

Published in: Education, Technology

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
3,426
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
9
Comments
0
Likes
0
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. University of OregonPoetics as SystemAuthor(s): Claudio GuillénReviewed work(s):Source: Comparative Literature, Vol. 22, No. 3 (Summer, 1970), pp. 193-222Published by: Duke University Press on behalf of the University of OregonStable URL: http://www.jstor.org/stable/1769913 .Accessed: 23/09/2012 13:38Your 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 ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.. University of Oregon and Duke University Press are collaborating with JSTOR to digitize, preserve and extend access to Comparative Literature.http://www.jstor.org
  • 2. xxii - volume summer I 1970 number 3CLAUDIO GUILLg-N Poetics as System HERE ARE a number of reasons for thinking that literature con- stitutes systems, or that it manifests itself as system. Let us leaveaside all nonhistorical acceptations of "literature," of the kind one islikely to encounter in aesthetics or in the theory of criticism. Let usassume that our concern is with the manifestation of systems in histori-cal time. There are, then, several ways in which one can speak ofliterary systems. They correspond to the following areas of study (though additions and further subdivisions could of course be made):poetics, that is, mainly the theory of genres and the appearance of sys-tems of genres; fundamental norms and materials, such as styles (the"three styles," etc.), rhetorical figures, themes, myths; structural rela-tions existing among the parts of an actual configuration or whole, likea movement, a period, a national tradition, or the establishment of acanon by means of anthologies; and, finally, the individual readingexperience. My original interest, as I first approached the subject, was in the reading experience. But it soon became clear to me that none of these categories could be treated independently from the others, and that it was necessary to confront the general problem of literary systematics. I found that this problem brought into focus, in turn, the need for a structural approach not merely toward the single poem but toward the 1 The main idea in these pages goes back to my article "Literatura como sistema(sobre fuentes, influencias y valores literarios)," Filologia Romanza, IV (1957),1-29. I have not been able to return here, for reasons of space and proportion, tomy earlier discussion of the reading experience in structural terms (which Iwould regard now as a sort of "intermediate" or "secondary" system, which medi-ates between the collective systems of history and the individual reader-whosepersonal literary "culture" or system is the real equivalent of "language" asstudied by structural linguists). Nevertheless, I have continued to use the Saus-surean model of the Spanish article and to develop my earlier thoughts, with noextraneous, last-minute attempts to incorporate the recent lessons of informationtheory, systems analysis, etc. It seemed to me more prudent and more honest topostpone these additions until a future study. 193
  • 3. COMPARATIVE LITERATUREbasic units and terms of literary history. It demanded, in fact, the elu-cidation of structures in history. In the present paper-or rather, pre-liminary notes for the longer study that will have to be written-I willattempt to show the significance and basic complexity of literary sys-tematics, with regard especially to one area: poetics. I will begin toindicate how the different cognate systems shift, change, become super-posed, and interact one with another. And I shall suggest that thehistory of literature-as distinct from language or society-is char-acterized not so much by the operation of full systems as by a tendencytoward structure or structuration. Thus it appears that the historian isled to evaluate, for every century or phase in the history of his subject,the precise scope of a limited, persistent, profound "will-to-order,"within the slowly but constantly changing domain of literature as awhole. The poetics prevailing in any period of European literary historycan be expected-though more or less explicitly, and with reference tovarious terminologies-to exhibit a tendency toward system. The firstwords of Aristotles Poetics-"Concerning the poetic art as a wholeand its species... "2-begin to formulate the dual purpose that wouldbecome traditional for literary theory: to elucidate the nature of thepoetic art "as a whole" (its goals, origins, validity, etc.) ; and to offera theory of genres or "species." In Aristotle, the discussion of the genus"art of poetry" underlies the analyses that follow of the more importantpoetic species (tragedy, epic, etc.) 8-and this essential conjunction hasremained traditional too. I do not allude here to the direct influence ofAristotles Poetics, but to general aspects of Greco-Roman thinking onthe subject of literature. The theory of genres to which an ars poeticaleads, more often than not, is not a simple exercise in classification. Itis a theory, insofar as it tries to organize the numerous facts at handaccording to principles that have been derived from an interpretationof the "poetic art as a whole," and beyond it, of its place in a largerscheme of knowledge. Theory and the tendency toward system, then,go hand in hand. In the European tradition, poetic theory, far frombeing simply critical or descriptive, has tended to propose and establishgenuine examples of intellectual order.4 It is not necessary to face here 2 I use Gerald F. Elses literal translation, where autes, "in itself," is under-stood generically, in opposition to the species or forms that are mentioned im-mediately afterwards, cf. Aristotles Poetics: The Argument (Cambridge, Mass.,1957), pp. 1-3. 3 See Else, pp. 4-5. 4 In a study of genre theory in China, James R. Hightower brings out the rela-tionship between the broad aims of literary theory and the classification of forms,with regard to the early Chinese theorists. "The Wen hsiian and Genre Theory,"in Studies in Chinese Literature, ed. John L. Bishop (Cambridge, Mass., 1965),194
  • 4. POETICS AS SYSTEMthe problems posed by the concept of genre itself (the extent to whichgenres, for example, were empirically adapted to actual differencesexisting between individual poems or plays). What matters in our con-text is the very existence, historically speaking, of such intellectualorders. As a rule, the author of poetics does not provide the practicing writerwith an endless spectrum of possibilities: he limits his options. Tolimit, subdivide, and distinguish, is to make order and system possible.These, in turn, become "dwelling-places" for the writer-envelopinghistorical situations-though the poet may be as vaguely aware ofthem as he is of the other historical situations-the political, social,economic systems-in which he lives from day to day. I should addthat I do not deny that significant relations exist between these variousstructural situations. In fact, it seems to me that it is precisely the ideaof literary system that makes cogent discourse possible on the subjectof the relations between social or economic or political history, on theone hand, and literary history, on the other. I shall return to this ques-tion later. For the moment, let us take notice of the fact that a theoryof genres supplies the writer-to use Ernst Robert Curtius term-with a sort of "ideal space."5Wherever such spaces exist, the isolationof an individual genre is more apparent than real. Its delineation, itscharacter, will depend on the place and the purpose which other genreshave been assigned, as well as on the connection existing between them. What does one mean, in this context, by system ? As far as the liter-ary historian is concerned, a system is operative when no single elementcan be comprehended or evaluated correctly in isolation from the his-torical whole (or "conjuncture": in Spanish, conjunto or coyuntura)of which it is a part. I assume that the term is to be used flexibly anddynamically, that it can refer to a stable order but also to a moment ina process of structuration. A system can be more or less open, loose,disjointed. We need have no "fearful symmetry" or other neatly pro-portioned arrangements in mind, for our model is neither mechanicalnor visual. Our subject is a certain type of mental order, characterizedby the functional importance of the relationships obtaining between itsvarious parts. A system is more than a combination or a sum of itscomponents. It implies a certain dependence of the parts on the whole,p. 513: "the conflictbetweenthe traditionaldidacticview of literatureas a vehiclefor moral instructionand the hereticalview of literatureas its own end is exempli-fied in the divergent lists of literary forms. It was the appearanceof the firstpurely literary genre, the fu, which precipitatedthe controversy . . ." 5 See Ernst Robert Curtius,EuropdischeLiteratur und lateinischesMittelalter (Bern, 1948), p. 16; or the excellent English translation by Willard R. Trask,EuropeanLiteratureand the Latin Middle Ages (New York, 1953), p. 18. I shallcite hereafter the Trask translation. 195
  • 5. COMPARATIVE LITERATUREand a substantial impact of the basic interrelationships. Our principalmodels, then, are linguistic and social. We are essentially indebted,above all, to Ferdinand de Saussures idea of linguistic system. System,not structure, as we know, was one of Saussures favorite words-forexample, on the subject of linguistic signs:Cest une grande illusion de considerer un terme simplementcomme lunion duncertain son avec un certain concept. Le definir ainsi, ce serait lisoler du systemedont il fait partie; ce serait croire quon peut commencerpar les termes et con-struire le systeme en faisant la somme,alors quaucontrairecest du tout solidairequil faut partir pour obtenir par analyse les elements quil renferme.6Similarly, it seems like a "grande illusion" to isolate the componentsin a system of artistic forms. A considerable reliance of the parts onthe whole, of individual definition on overall theory, has been operativein the area of poetics on a number of influential occasions, from Aris-totle to Northrop Frye. But the analogy between linguistic and literarysystems raises, to be sure, delicate and specialized questions, whichought not to be answered quickly. The social analogy, fortunately, isavailable and instructive in a less technical way. One need only recallordinary notions of social class. "Middle class," "lower class," etc.,imply not only membership in a group but nonmembership in others.The number of classes that constitute a particular system does notaffect this basic fact: class concepts are relative within a system; theysignify a mans position and immersion in a society; whether a certainclass is part, say, of a three-level or of a five-level structure, or even ofa relatively fluid scale, it refers necessarily to the existence and impor-tance of a system of class distinctions as a whole. As far as literary "classes" are concerned, in the history of poetics,the impact of certain closely related branches of learning and forms ofintellectual activity appears to have been crucial. I refer to the struc-tures of grammar, philosophy, or education, with whose influence onthe structures of literature we are familiar. Hermann Usener studiedmany years ago, in a useful article on the structural features of ancientphilology, the links between philosophical and grammatical systems inGreece and Rome: "die philosophische Bildung, welche die Voraus-setzung aller Systematik ist, hat an mehr als einem Orte ihre urspriing-liche Farbe noch bewahrt; sie stammte aus der peripatetischen Schule."7During the Hellenistic period a fragmentation of philosophy had occur-red, from which various branches of learning had arisen: grammar,rhetoric, poetics, philology, etc. Aristotle and his disciples had at- 6 Cours de linguistiquegenerale, 4th ed. (Paris, 1949), p. 157. 7 "Ein altes Lehrgebaude der Philologie," in Kleine Schriften (Osnabriick,1965; reprintof the 1912-13ed.), II, p. 303.196
  • 6. POETICS AS SYSTEMtempted to reconcile rhetoric and poetics with the substantial goals andorderly procedures of metaphysics and logic. They showed that philos-ophy could validate the poetic "imitations" that Plato had denounced."Aristotles teaching on the art of poetry," E. R. Curtius affirms,"must,then, be seen in connection with his entire system: as a parallel dis-cipline to ethics, politics, rhetoric, economics."8 We also know that Quintilians famous Institutio oratoria was ac-tually a treatise on education. I do not think that the essential continuityof literature is a product of its pedagogical uses. But its systematizationowes a great deal to pedagogy. As Curtius stresses in his indispensablebook, the fact that literature has been a school subject for nearly 2,500years has a great deal to do with the emergence and the survival of atheoretical order:Wherever literature is a school subject, we have elementsof a systematizedstudyof literature.We have the science of literature, in a form suitable for beginners.Anyone who read Homer as a school text could not but learn that the Iliad is apoem in narrative verse (epos) and that verse is a form of discourse subject torules.9Education has not only tended to foster literary theory: it has tradi-tionally built bridges between the sciences, or between the arts and thesciences, and promoted the application of methods of instruction de-veloped in one field to cognate branches of learning. (These educa-tional bridges have multiplied in our day, and it would be difficult tointerpret the achievement of Anglo-American literary criticism of thelast forty years, since I. A. Richards and the New Critics, without tak-ing into account the exigencies of teaching literature in a form "suit-able for beginners," as well as the prestige of such academic neighborsas linguistics and the social sciences. Today, the academy continues topromote "systematics.") In the early Middle Ages, grammar, rhetoric,and dialectic (or,logic) composed the curricular triad of the liberalarts. The range of rhetoric was fluid, but it often subsumed instructionin the different "styles" or stylistic levels-an obviously structuralteaching-or the presentation, as in Quintilian (X, 1), of a roster ofgreat authors. Grammar was originally the foundation of the entireeducational triad. Its study included instruction in both language andliterature, in the grammatical rules as such and in their exemplary usesby great writers; it could incorporate the figures of speech, and metrics.Walter of Chatillon (thirteenth century) assigns poetry to grammar:among the liberal arts that are called the trivium, he writes, "grammartakes precedence as the first foundation; under her serves the troop ofthose who write in verse": 8 European Literature and the Latin Middle Ages, p. 146. 9 Ibid., p. 247. 197
  • 7. COMPARATIVE LITERATURE Inter artes igitur, que dicuntur trivium, Fundatrix grammatica vendicat principium. Sub hac chorus militat metrice scribentium.10In one of the thirteenth-century artes poeticae published by EdmondFaral, the Laborintus of Eberhard the German, when Poetry ap-proaches, she is introduced as "Grammarsattendant": Grammaticae famulans subit ingeniosaPoesis.lAThe new Aristotle, Parisian Scholasticism, the quarrel between the oldsystem of the liberal arts and the proponents of philosophy and naturalhistory, between men of letters and scholars, altered the propaedeuticrole of grammar and rhetoric after the thirteenth century; but they didnot destroy their relations with poetics. Likewise, the theorists ofRenaissance Italy would take a fresh look at these relations, but notproceed to dissolve them. Giovanni Battista Pigna, in I romanzi(1554), classified poetry, rhetoric, and dialectic under logic. Pigna, aninnovator as a critic, had also a theorists respect for logical thought.Bernard Weinberg has pointed out that there is throughout the six-teenth century "a strong tradition that associates poetry with logic,grammar, rhetoric, and history as one of the discursive or instrumentalsciences. Poetry belongs with the others as a discursive science becauseit uses words (or discourse) as its means."12Other critics would finda different place for poetry in a general scheme of the arts and sciences.But the desire to classify and distinguish was general among bothmedieval and Renaissance theorists, and the tendency prevailed to makepoetics, in conjunction with the other cultural activities of man, assystematic as possible. The poetics of the Cinquecento, to be sure, did not coincide with thework or with the impact of any single theorist or theory. BernardWeinbergs History of Literary Criticism in the Renaissance chartsthe chronological course of three principal influences: Horace, Plato,and Aristotle. Renaissance theorists depended on widely differingprinciples of classification: poetry could be regarded as an ethical ac-complishment, akin to moral philosophy; it could be viewed as verbalconstruction, and hence a linguistic science, etc. Literary critics wouldattempt to relate their ideas to philosophical systems which they failedto comprehend fully, or which were not really amenable to artisticthemes. Too many theories, of course, had been revived, the humanists 10 Quoted by E. R. Curtius, p. 45. 11 Edmond Faral, Les Arts poetiques du XIIe et du XIIIe siecle (Paris, 1962),p. 345. 12 Bernard Weinberg, A History of Literary Criticism in the Italian Renais-sance (Chicago, 1963), I, 13. On "Poetics among the Sciences," see I, 1-37.198
  • 8. POETICS AS SYSTEMburden having become lembarrasdu choix. The Cinquecento, in Wein-bergs opinion, yields no simple explanations:One may ask, also, whether any change or progress or evolution in the theoriesof classification is perceptible through the century. This question is difficult toanswer. It should be clear that the major theories exist simultaneously throughoutthe century. Perhaps in the early years the tendency to occupy oneself with com-plete philosophical systems is more prevalent than in the later. Perhaps also theclassification as a discursive science or rational faculty is more prominent in thefirst part of the century, whereas the classification as a relative of moral phi-losophy is more frequent in the second part; the "rhetorical" approach gives wayto the "ethical."13 Systems existed simultaneously, and their coincidence made possiblethe famous theoretical quarrels of the second half of the century. Itmight be maintained that Minturno and Scaliger came close to beingrepresentative of the period, whereas Patrizi, for example, was aneccentric, and Campanella a reactionary. But Weinberg also demon-strates that all these theorists should be seen against the polemical back-ground of their time: "in a way, the history of poetic theory of theCinquecento might be organized as a series of quarrels and polemics,similar to those so prominent in practical criticism."l4 Cinquecentoliterary theory, in sum, tended to be highly systematic; and it became,in the process, and by virtue of its submission to different influencesand constantly shifting principles of classification, a running quarrelamong competing intellectual orders. To what extent, we could now ask, is a theory of genres a system?At this point in our inquiry, the following observations, or workinghypotheses, may be made: (A) The history of poetics exhibits the tendency not merely toenumerate but to organize norms, to join them as systems. Both analy-tical and synthetic procedures may be used: by submitting literaryworks to philosophical, grammatical, or social principles of organiza-tion; or by inferring certain characteristics which are then turned intoprinciples (such as styles), and ultimately, into alternatives, contrasts,polarities. (B) The history of poetics offers many examples of the failure toarrange poetic models into systems of norms. Often the attempt fallsshort of completion, and one is left with a half-way stage in a processof structuration. In other cases one encounters simple enumeration (asin Horace, perhaps, or Sir Philip Sidneys "sundry more special de-nominations"), while the tendency toward system manifests itself onother levels of artistic theory and practical criticism. 13 Ibid., I, 37. 14 Ibid., II, 809. 199
  • 9. COMPARATIVE LITERATURE (C) One should keep in mind the most evident but essential of facts:the extraordinary persistence through the centuries, in European litera-ture, of a limited number of generic models (it could be argued thatthe most persistent have been comedy and the short lyric forms)-and, to a certain degree, of their place within generic systems (or oftheir connection with other members of the same continuing order).Although changes and additions have taken place constantly, an over-all awareness has existed of the continuity of literary norms. Moreoften than not in the itinerary of European literature, one recognizesa certain consciousness of the relative stability that prevailed in therealms of culture and society. (The social analogy comes to mind, aswe know full well that an awareness of the relative continuity of socialconditions was for a long time one of the psychological bases for classdistinctions, and a reason for their perpetuation.) (D) The relationship between the class (or norm) and the systemto which it belongs is one of the conditions for the continuity of both.Thus, the concept of tragedy can survive a period in which a singledramatic genre, such as comedy, keeps the idea of the theater alive,etc. (For example, in Rome, where few tragedies were written: yettragedy occupied a prominent place in the Ars poetica of Horace.) Inthese instances, the system retains the class which has traditionally be-longed to it. On other occasions, a system will absorb a new work and legitimate it as a normative model, as a genre, on the strength of some structural connection between the new model and the existing classes. (The Byzantine philologist Photius-c. 820-897-would regard the Greek narrative romances by Heliodorus or Achilles Tatius as ex- amples of dramatikon.15).On still other occasions, a work or an author will be forced by the system into an inappropriate category. (Or con- versely, the original significance of a generic term will be forgotten: Wilhelm Cloetta pointed out long ago, in his Komodie und Tracgodie im Mittelalter, that a number of the most authoritative writers of the Mid- dle Ages, beginning with St. Isidore of Seville, had no understanding of what a theatrical performance, or theatrical genres, really meant; so that Averroes misreading of "tragedy," which Renan and Menendez Pelayo singled out, and which inspired Borges to write one of the best stories in El Aleph, was far from exceptional.) In other words: genre and system reinforce and perpetuate each other. (E) These mutual effects are forms of the interplay between con- tinuity and change in the history of literature. The novelty of an indi- vidual author may only lead, in generic terms, to the incorporation of 15Irene Behrens, Die Lehre von der Einteilung der Dichtkunst (Halle-Saale,1940), p. 38.200
  • 10. POETICS AS SYSTEMhis work into the existing order. Systems will tend, generally speak-ing, to absorb change and assimilate innovation. On the other hand, assimilation becomes a small step in a largerprocess of change. (The classic example is the famous sixteenth-century quarrel concerning Ariosto. The novelty of the OrlandoFurioso had to be ascertained, first of all, with reference to the pre-dominant norms of the epic. Several things could then happen: thesystem, or rather its proponents, could "reject" Ariosto; the romanzocould be absorbed by the norms of the epic; a special niche could befound in the system for a favola mista, etc. Ultimately, the ronwnzosassimilation caused a significant jarring and loosening of the system;and this became one of the conditions for the rise of the novel.) Imight also add that the relationship between the new literary work andthe prevailing system, which is essentially traditional, can be regardedsimply as an instance of the critical intelligence at work; the critic (orthe writer as critic) views a new work, as it were, "through" a system;he perceives, judges, and decides, for better or for worse, within thecoordinates of an available critical scheme. The critical intelligence"assimilates" and "accommodates" nearly in the sense that the psy-chologist Jean Piaget gives to these terms.16 (F) Literary theorists have generally dealt with different kinds ofclasses, of which only a certain kind could possibly serve in a systemof genres. It is often difficult for the modern historian to discoverwhether a certain class used in the past is comparable at all to what istoday a genre. Yet he will try to distinguish between systems of genresand the other cognate systems, not only because they are closely relatedbut because one of the latter could have shaped or even displaced theformer. To clarify and classify the different kinds of literary classesrequires a detailed, step-by-step history of the subject (that is, parallelhistorical studies of cognate systematics, or "intersystematics"). 16 Assimilation takes place whenever the individual incorporates the data ofexperience into a previous logical framework. Piaget distinguishes between threekinds of "assimilation," the third of which is more pertinent to our subject.Yvette Hatwell, "Propos des notions dassimilation et daccommodation dans lesprocessus cognitifs," in Psychologie et dpistemologie genetiques . . . Honnmagea Jean Piaget (Paris, 1966), p. 128: "lassimilation reproductrice, cest-a-dire larepetition simple dune action, qui assure en meme temps sa fixation; lassimila-tion recognitive, cest-a-dire la discrimination des objets pouvant etre assimilesa un scheme particulier; et lassimilation generalisatrice, la plus feconde sansdoute puisquelle conduit a elargir le domaine dun scheme donne et par la memea elargir la classe des objets pouvant lui etre assimiles." But assimilation alonewould bring no effective knowledge of the object, and it is countered constantlyby "accommodation," which is an active participation on the part of the subject:"accommoder, cest en quelque sorte prendre le parti daffiner et de modifier sescadres de pensee lorsque ceux-ci, se heurtant a une realite imprevue, saverentinoperants" (p. 129). 201
  • 11. COMPARATIVE LITERATURE I will limit myself to recalling three of the fundamental kinds ofclasses: (1) genres themselves, or what modern criticism recognizesas genres. For a long time these were thought of as "species" of a larger"genus," and for a very good and obvious reason: because models,tautologically enough, should be susceptible of "imitation." A genre Zcould not be so comprehensive that it might not be said of a particularwork, in the singular: this is a Z. For instance: this is a comedy, an ode,a novel. A genre makes possible, practically speaking, literary composi-tion, the actual assemblage of the work as a whole. In that sense, it isbroad enough. On the other hand, it should not be so vague that itserves more as a premise than as a model. A genre endures (and notto endure is not to emerge fully as genre) insofar as it continues to be aproblem-solving model-a standing invitation to the matching of matterand form. (2) Works classified with respect to versification alone: theGreek iambics, the elegy, etc. Prosody or external form is not a satis-factory basis for classification, as modern studies have shown concern-ing the ode or the sonnet, which, of course, are not only verse forms.17This point is historically important: despite the example of Aristotle (Poetics, 1447b), who pointed out that Homer and Empedocles hadnothing in common save meter, it has been a central tradition of Greco-Roman literary theory to identify poetic form with metrical form. Dio-medes (fourth cent.), whose ars grammatica was highly influential inthe Middle Ages, maintained that the essence of poetry was to be foundin versification.18This confusion between meter and form postponedthe emergence-until at least the Renaissance-of an effective idea ofthe lyric, or of lyric poetry in general, which one could not differentiatefrom dramatic or narrative verse according to meter alone. (3) Presen-tational modes, like "narrative," "dramatic," etc. There are of courseother modes or "universals," such as satire or allegory, which cutacross the differences between genres. I refer here to what NorthropFrye calls "radicals of presentation": "we have to speak of the radicalof presentation"-he writes--"if the distinctions of acted, spoken andwritten word are to mean anything in the age of the printing press."19 Fryes system is the tetrad "epos," "prose," "drama," and "lyric."He then subdivides these modes into "specific forms"-which are likethe older "species" of the artes poeticae and our own genres. I do notthink, unlike Frye, that these modes constitute the central principle ofall generic differentiation, and that the specific genres are forms ofthese modes. But I shall not argue my case here. My concern is with 17Karl Victor, "Die Geschichteliterarischer Gattungen,"in Geist und Form(Bern, 1952), pp. 291-309. is Curtius, p. 439. 19Anatomy of Criticism (Princeton, 1957), p. 247.202
  • 12. POETICS AS SYSTEMthe history of poetics, which shows that modes and genres have con-stituted different systems, and that one of the tasks of the historianconsists in observing the changing relations between the two. It seemsto me that modes have lent themselves to systematization, or have re-sulted from it, much more readily or frequently than genres have. Wehave had systems of modes, but as far as genres are concerned (per-haps because of the proximity of "specific forms" to poetry itself) whatwe often find is a mere will-to-structure. The now famous division into three modes-"narrative" (or"epic"), "dramatic," "lyrical"-was fully presented and defended forthe first time, according to Irene Behrens, by Francisco Cascales in hisTablas poeticas (1617).20 (One is puzzled by the apparent neglect ofAntonio Minturno, who sketched a similar view in his De poeta, 1559,21as well as by the possible motives for Cascales spirited advocacy of thetriadic system. In Spain as in no other country during the seventeenthcentury, to be sure, poetry, the drama, and the novel prospered simul-taneously: Gongora, Lope de Vega, and Cervantes were contempo-raries.) An obscure humanist of the period, Pedro Gonzalez de Sepul-veda, who taught rhetoric in Alcala, objected in a learned epistle tosome of Cascales ideas. The objections were published in the lattersCartas filologicas (1634). One of Sepuilvedascomments was the fol-lowing:El soneto, en la postreratabla, pag. 440, le reduce v. m. a la poesia lirica en con-secuencia de la antecedente division, que pone tres especies de poesia: lirica,scenica, epica. Si no son mas, de su bando me tiene v. m.; pero si no me engafiami juicio, no son tan pocas; porqueesas, se bien se mira, mas son diversos modosde que el poeta usa en sus narraciones,que diversas especies de imitaci6n. e Quiendira que la comedia y tragedia son una especie? e Por ventura no se diferencian 20Die Lehre von der Einteilung der Dichtkunst, pp. 128-129. See Tablaspoeticas del Licenciado Francisco Cascales (Murcia, 1617), p. 30: there arethree modos, "exegematico,dramatico,y mixto" (exegematico goes back probablyto the Ars grammatica of Diomedes, where exegematikon is an equivalent ofenarrativum,"the poet speaks alone"); and p. 38: "la Poesia se divide en tresespecies principales, Epica, Scenica, y Lirica." Though first published in 1617,it appearsthat the Tablas poeticas were already written in 1604: see Justo Gar-cia Sorianos Introductionto Cascales Cartas filologicas (Madrid, 1951), I, p.xx. It also seems that Cascales was a morisco (a converso or descendantof con-verts from the Islamic faith)-see Garcia Soriano, I, p. xiii: "jamas mencionanialude siquiera Cascales a sus padres ni a las circunstanciasde su nacimiento yfamilia. Tampoco lo hacen sus contemporaneosque le conocieron y trataron.Siempre se le da un solo apeliido"; and the words, quoted by Garcia Soriano, p.xii, n. 1, of the eighteenth-centuryscholar Jose de Vargas Ponce: "si a Cascaleslo dejaronmoro, o la fosa en que estaba el asiento de su crisma se la comieronlosratones o las curianas, o supli6 las necesidades de algun cura malandrin, queculpa tiene este infeliz comisionado?" 21 Irene Behrens is of the opinion that Minturnosidea "bleibt... stehen" (p.86), and so remains undeveloped. 203
  • 13. COMPARATIVE LITERATUREmas que en numero? d No hay mayor diferencia entre una comedia y tragedia queentre dos comedias ? d No la hay tambien mayor entre una lirica y ditirambica queentre dos liricas? Pues estas se diferencian en numero; luego la distinci6n deaquellas habra de ser especie; por donde las especies de poesia mas habran de serde tres.22To which Cascales responded:Y seguin esta divisi6n, no hay mas que tres especies, que son epica, lirica y scenica;que si bien la tragedia y comedia son en rigor diferentes, pero porque la una y laotra es dramatica, y se representan en el tablado, se habla de ellas como de unaespecie. Y cuando las digamos, como lo son, distintas, al prop6sito y fin que v. m.lleva, no importa. Pues el epigrama o soneto ne se quede reducir a la comediani a la tragedia, porque en nada, digo, esencialmente, convienen entre si, ya porqueestas son dramaticastotalmente,y el soneto no lo es, ya porquetienen accion quecelebrar,y el soneto no la tiene; pues la fabula del soneto es un conceptono mas,y no una accion, y por las mismas causas tampoco se puede reducir a la epica.Teniendo, pues, el soneto por alma de su poesia un concepto, como la lirica, y nocomprendiendo acci6n, como la heroica ni como la tragica ni como la comicacomprehende, a quien, sino a la lirica, podemos aplicar el soneto?23Sepulveda, the more conservative of the two, and the better Aristote-lian, considers tragedy and comedy to be especies (our modern genres).And he insists on the need to retain the traditional distinction betweengenres and modes. He recognizes, furthermore, that tragedies andcomedies have in common what Aristotle had described in the Poetics (1447a, 1448a) as a "mode of imitation." The various kinds of poetry,Aristotle had explained, differ from one another in three respects: themedium employed (melody, verse, etc.), the object of imitation (menbetter or worse than in real life, etc.), and the manner or mode of imi-tation. Concerning the latter, he had written (1448a) : "there is still athird difference-the manner in which each of these objects may beimitated. For the medium being the same, and the objects the same,the poet may imitate by narration-in which case he can either takeanother personality as Homer does, or speak in his own person, un-changed-or he may present all his characters as living and movingbefore us."24 This is one of the more obscure and disputed passages inthe Poetics, but it may be safely said that the basic principles at work (narration, direct speech, dramatic representation) are not relevantto the other two differentiae of the poetic art; and we are not allowedto forget this fact, so as not to confuse the modes of imitation with thethe actual poetic species. In Gerald F. Elses words: 22 "El maestro Pedro Gonzalez de Sepuilveda al Licenciado Francisco Cas-cales" (Epistola IX), in Cascales, Cartas filol6gicas, III, 217. 23 Ibid., III, p. 239 (Epistola X). 24 S. H. Butcher, Aristotles Theory of Poetry and Fine Art, 4th ed. (London,1911), p. 13.204
  • 14. POETICS AS SYSTEMHence one cannot simply identify the dramatic mode with drama, since the formermay appear outside of the latter, or the mixed mode with either epic or lyric,since it may appear in either. The differentiae do not simply run with the estab-lished genre-divisions or with each other: they cut across each other and so bringout significant differences. From one point of view, therefore, one might say thatdrama is anything that uses the dramatic method; but a more correct definitionwill draw on the differentia of medium as well and distinguish two actual dramaticgenres, tragedy and comedy.25 The "dramaticmode" and the "drama"failed to coincide, also, inso-far as the former implied not the conditions of staging-theatrical per-formance-but, rather, "direct" speech and dialogue. Much emphasiswas placed on the poets presentation or representation of a characteror a voice other than his own. In this context, "imitation" assumed anarrower sense than usual,26involving above all the poets participationin the portrayal of character. Thus the dramatic mode could be reducedto imitation in the sense of impersonation-mimesis in the sense ofmimicry: the illusion of speaking for, or even as, someone else. Duringthe Middle Ages it was often thought that several of Virgils Eclogues-those that offer only dialogue, between shepherds like Meliboeus,Tityrus, etc.-were among the purest instances of the dramatic mode.Now, these traditional concepts are echoed by Sepulveda-and theyretain one advantage: a fairly free and loose relationship between thesystem of modes and the roster of specific forms or genres. On theother hand, Cascales, the innovator, takes two decisive steps: he trans-fers the triadic "ideal space" to a system of genres; and he basicallyapplies one principle of definition to the drama and another to thelyric. He visualizes the stage, the scenic factor ("se representan en eltablado"), as the ultimate principle of all dramatic creation, whethertragic or comic. Moreover, he is in a position to test the validity of amodern-sounding category as the proper basis for a definition of thelyric. Aristotle had written that the object of all poetic imitation was"men in action" (1448a). Cascales maintains that the argument of asonnet is not an action but a concepto; and he suggests that this isthe essential feature of the lyric. (The dictionary published by Sebastiande Covarrubias in 1611, Tesoro de la lengua castellana, defines con-cepto as a discourse born in the mind, and later executed by eitherthe tongue or the pen: "el discurso hecho en el entendimiento, y despuesejecutado, o con la lengua o con la pluma.") 25 Else, op. cit., p. 91. 26 In the larger sense, mimesthai "means to make or do something which hasa resemblance to something else." Aristotle, Poetics, ed. D. W. Lucas (Oxford,1968), p. 55. The more restricted "impersonation" became confused with theother in the minds of several Renaissance theorists, as Bernard Weinberg ex-plains, op. cit., I, 60-63. 205
  • 15. COMPARATIVE LITERATURE The influence and prestige of Petrarch, of course, and of the ItalianRenaissance lyric in general, had led the theorists of the sixteenth andseventeenth centuries to seek a characteristic "object" for lyrical poetrythat would be clearly distinct from either narrative or dramatic action (and from meter as well). The desired notion was not difficult to find:it could be extracted, without looking any further, from Aristotlestraditional system of modes. (In his Tablas poeticas, p. 30, Cascalessaid: "El lirico casi siempre habla en el modo exegematico, pues hacesu imitaci6n hablando el propio, como se ve en las obras de Horacio, ydel Petrarca, poetas liricos.") The lyrical poet was the poet speakingthrough his own voice, for and as himself. The poet as his own "object,"needed to impersonate no one, to "imitate" no action. This identifica-tion of the poet with his creation would become for many centuries thecenter of gravity of the lyric. (From Petrarch until, by and large, thereversal of Mallarme; and later, Rimbaud, and in our day, FernandoPessoas heteronyms. Paradoxically enough, "je est un autre" is a sortof unavoidable impersonation.) In such a manner, Cascales (followingBishop Minturno) was able to transfer the old tripartite system-thetriadic "ideal space"-from the radicals of presentation, or "modes ofimitation" (which he named exegemactico,dramatico and mixto), tothe genres themselves (epica, escenica, lirica). One should note, inthis connection, that one of the Cartas filologicas-Epistola VI: "Sobreel nuimeroternario"-is an extensive, scholarly encomium of the num-ber three. Why, he asks, so many "triplicities"?-his pretext being therecondite signifiance of the Three Wise Kings:Y estos reyes magos eran tres, seglun san Augustin, san Leon, Ruperto y otros:llamabanse Melchior, Gaspar, Baltasar. Tres fueron las regiones de donde vinie-ron: Arabia, Saba, Tarsis; tres los dones que ofrecieron a Jes6s: oro, mirra,encienso. Pues i por que tantas triplicidades ? Porque adorando a Cristo, con quienpor via de concomitancia asistian al Padre y el Espiritu Santo, adoraban intrin-secamente la Santisima Trinidad; que no es posible que hubiesen venido tres paramenos que para simbolo de la divina Triada, la cual quiso Dios significar de milmaneras y en mil lugares.27 The debate between Cascales and Sepulveda is a brief chapter in thegradual shift in poetics from a triad of modes to a triad of genres.This shift, to be sure, cannot be documented within the limits of a shortpaper, and I will only add some comments on a few pertinent examplesand problems. (Most of the examples will be taken from the historyof literary systems, while most of the problems will derive from theirconnection with the other historical structures.) Aristotles system ofmodes is usually attributed to the influence of Plato, while others prefer 27 Cartas filolhgicas, I, 115.206
  • 16. POETICS AS SYSTEMto stress its original features. Among recent commentators, D. W.Lucas assumes its Platonic origin,28 while Gerald Else writes:We must recognize that this long and tenacious tradition [i.e., the three modes ofpoetry] is jointly Platonic-Aristotelian, fed partly by the Republic and partly byan Aristotelian work (almost certainly the dialogue On Poets), but that it hasbeen shaped more decisively by Aristotle than by Plato.29Plato had introduced in the Republic (392d-394d) an apparently origi-nal30arrangement of what we have been calling "radicals of presenta-tion." Socrates singles out three modes:You have conceived me most rightly-I said-, and now I think I can makeplain to you what I was unable to before, that there is one kind of poetry andtale-telling which works wholly through imitation, as you remarked, tragedy andcomedy; and another which employs the recital of the poet himself, best exempli-fied, I presume, in the dithyramb; and there is again that which employs both, inepic poetry and in many other places, if you apprehend me.31Elsewhere (595-601) Plato accuses the poets of imitating (in thebroader sense of mimesthai) appearances, and of remaining "at thethird remove" from truth. The moral and social functions of poetryare being questioned-its effects on the citizen of the ideal state. Thereal issue is pedagogical, and the division of poetry into three modesshould be understood in this light. The practical effects of poetry maybe such as to cause the student, or the citizen, to experience falsehoodand live lies. "If then the ruler catches anybody else in the city lying,any of the craftsmen Whether a prophet or healer of sickness or joinerof timbers he will chastise him for introducing a practice as subversiveand destructive of a state as it is of a ship" (389d). Stories about thetransformations of Proteus are objectionable (381d); for ". . . Godis altogether simple and true in deed and word, and neither changeshimself nor deceives others by visions or words or the sending of signsin waking or in dreams" (382e). Is the practice of impersonation aninvitation to mutability, formlessness, dispersion of character? Hencethe relevance of the two main criteria for the triad of modes. But whya triad? One should leave this question to Hellenists and experts inPlatonic philosophy, who may wish to ask whether the system of threemodes is significantly affected by the sociological and psychologicaltriads which the Republic introduces. Socrates explains at length that 28 Lucass edition of the Poetics, pp. 66-67. 29 Else, p. 99. 30 Lucas, p. 66: "as Plato explains himself with great care, actually puttingpart of Iliad i. 17-42 into narrative by way of illustration, it would seem that thisdistinction was unfamiliar." 31 The Republic, tr. Paul Shorey, The Loeb Classical Library (Cambridge-London, 1935), I, 231. I shall cite hereafter this translation. 207
  • 17. COMPARATIVE LITERATUREthe three elements of the soul, the Appetitive, Spirited and Rational,correspond to the three social groupings of his ideal state: the Eco-nomic, Auxiliary, and Guardian classes.32We have seen that the edu-cational role of literature-social or moral-is being seriously chal-lenged. The triad of modes may not be fully independent from the broadanalogy between the city and the soul that Socrates presents and praisesas a fundamental method of inquiry:But now let us work out the inquiry in which we supposed that, if we found somelarger thing that contained justice and viewed it there, we should more easilydiscover its nature in the individual man. And we agreed that this larger thingis the city, and so we constructed the best city in our power, well knowing thatin the good city it would of course be found. What, then, we thought we sawthere we must refer back to the individual and, if it is confirmed, all will be well.But if something different manifests itself in the individual, we will return againto the state and test it there and it may be that, by examining them side by sideand rubbing them against one another, as it were from the fire-sticks we maycause the spark of justice to flash forth, and when it is thus revealed confirm itin our own minds. Well-he said-, that seems a sound method and that is what we must do [434e-435a]. We meet the Platonic-Aristotelian triad again in the writings of thegrammarians, rhetoricians, and scholiasts of later Antiquity. The samedivision into "narrative" or "reportorial" (diegematikon, apangel-tikon), "dramatic" or "imitative" (dramatikon, mimetikon), "com-mon" or "mixed" (koinon, rmikton), occurs in Probus or Servius(fourth cent.), Proclus (fifth cent.), the anonymous prolegomena toHesiod, etc.33 Proclus scheme, we might note, is part of his highlyinfluential Chrestomathia or grammatical textbook based on literaryexcerpts and selections. This was not a chance encounter.Having grown up under the aegis of Greek philosophy [Curtius summarizes atone point] literary science came of age in the form of Hellenistic philology. Itwas called upon to classify the matter of literature-"studiorum materia," asQuintilian puts it (X, 1, 128)-in accordance with two different principles: bygenres and by authors. The selection of authors presupposes a classification ofgenres.34Now these observations are most pertinent to our theme: doubtlessthe normative systems or "ideal spaces" of literature have often beenformed around a cluster of "great authors,"each of whom would occupythe position that was ordinarily assigned to a particular genre. A his-tory of anthologies, of how the canons of authors and authorities arose 32 R. C. Cross and A. D. Woozley, Platos Republic: A Philosophical Com-mentary (New York, 1964), pp. 112-115. 33 Else, p. 98; Behrens, pp. 25-32. 34 Curtius, p. 248.208
  • 18. POETICS AS SYSTEMand coalesced through the centuries of late antiquity and the MiddleAges, would have much to contribute to the study of literary systems.But a number of previous questions would have to be answered too:one would wish to know what the anthologists used, built upon, orcorroborated. How coherent, how schematic, were the "classificationsof genres" that Curtius mentions? Did they reflect another sort oforder ? In some cases, to be sure, an anthology would offer a mere suc-cession, a meaningless series. In other cases, the process of selectionderived from orderly criteria, which facilitated the choice of "repre-sentative" works, fragments of works, or (as in ancient grammars)shorter quotations. Generally speaking, I would reiterate my earlierpoint: although the Platonic-Aristotelian triad was prevalent, no widelyvalid system of genres emerged from the literary theorists of Greeceand Rome. Under the circumstances, two attitudes appear to have beencharacteristic. There was the example, first, of Aristotles Poetics andits successors: the different genres were referred, as examples or illus-trations, to other systems and principles of classification, such as modes;while at the same time, no genre was considered in isolation: the epic,for example, was confronted with tragedy (1449b, 1459b, etc.), andthe Poetics ended with a last comparison of the merits of the two.Second, there were grammarians, rhetoricians, and literary theoristswho named-and described-the genres one by one: we enlcounterthese series in the highly authoritative grammar of Dionysius Thrax (second cent. B.C.), the rhetorical works of Cicero (De opt..gen. orat.,1), or the Ars poetica of Horace, who glides over the epic, the elegy,and the iambic-largely in terms of meter (I. 73-82)-and then ex-patiates at length upon the three theatrical genres: tragedy, comedy,and the satirical play. The tone and method of an epistle, of course, areinformal; but Horace indicates along the way that he is familiar withthe traditional division into dramatic and narrative modes (line 179:"Aut agitur res in scaenis aut acta refertur"). The status of the lyric,in this connection, clarifies the influence that anthologists exerted ontheorists. Horace alluded to the lyrical poets in a number of his compositions:"he wrote poems"-C. M. Bowra states,-"to be read in the study andnot to be sung in company, but in obedience to his Greek models heassumed a convention that they should in theory be sung to the lyre. He did not mean this to be taken literally; his purpose was rather to claim his spiritual and artistic descent from Alcaeus, Sappho, and Pindar."35The latter point is important: Horace never did elucidate what he meant by a lyrical poem. If the relation with the lyre was not 35 C. M. Bowra, Greek Lyric Poetry, 2nd ed. (Oxford, 1961), p. 1. 209
  • 19. COMPARATIVE LITERATUREto be taken literally, if it really was symbolical of a poetic quality, onewonders whether a genre or a mode was being denoted at all. Likeother ancient theorists, Horace was deterred from framing a clear con-ception of the lyric by a number of factors: the emphasis on meter;the modal system based on events and actions ("aut res ... aut acta");the wavering association of the different "melic" or "lyrical" formswith music, etc. The scope of these terms was fluid, and the sum of thepoems regarded as either melic or lyrical by Alexandrian philology,36radically heterogeneous. The lyrical was not merely the nondramaticand the nonnarrative. There were narrative elements in the odes ofPindar. The dithyramb, which Aristotle mentioned, was originallychoral and half-dramatic-yet Simonides and other lyricists wrotedithyrambs. The distinction between choral song and monody was notabsolute: "a hymnos, or hymn to the gods, could be sung by a choir,as in Pindars Hymns, or by a single person, as it must have been insome Hymns of Alcaeus and Sappho."37The elegiac poem was com-posed for the flute, and therefore not considered lyrical in the strictsense. Archilocus of Paros was a highly personal poet, but as an authorof elegiac and iambic verse, he was not accepted as one of the lyricalpoets by the Alexandrian critics: "he put the self into poetry"-Bowraexplains,-but he himself preferred an art which was closer to speechthan to song."38 Surely a difference existed between a poem that wasactually sung and another that was merely written for musical accom-paniment: the latter was spoken or recited, though the tune would forcea certain metrical pattern on the poem. Demetrius (first cent. A.D.), theauthor of De elocutione, was a passionate admirer of the "divine Sap-pho," but he did not think all her works were lyrical; in fact, he wroteabout some of her wedding songs: "these poems of hers are... better suited for use in conversation than for singing. They are by no means adapted for a chorus or a lyre-unless indeed there is such a thing asa conversational chorus."39 Rather than to poems, the term "lyric" became most clearly attachedto an elite of poets. The appearance of the lyrical mode is a good ex- ample of the impact that a system of authors can exert on the other orders of poetics. Alexandrian criticism refined a "classical" concep- tion of literature, consisting of the great writers from Homer to Menander, for the use of grammarians and philologists in the class- 36 Behrens, p. 17; John Edwin Sandys, A History of Classical Scholarship(New York, 1958 rept.), I, 43. 37 Bowra, op. cit., p. 6. 38 Ibid., p. 14. 39 Aristotle, Longinus, Demetrius, The Poetics. On the Sublime. On Style,ed. W. Hamilton Fyfe and W. Rhys Roberts, The Loeb Classical Library(Cambridge-London, 1932), p. 407.210
  • 20. POETICS AS SYSTEMroom. These few writers were the select, the aristocracy, "the admitted" (egkrinomenoi). (Curtius brings out sharply the analogy betweensocial elite and anthology.40) Aristophanes of Byzantium (c. 257-c.180 B.C.) and Aristarchus of Samothrace (c. 220-145 B.C.) drew up alist of model authors of various kinds. There were nine lyric poets:Alcman, Alcaeus, Sappho, Stesichorus, Ibycus, Anacreon, Simonides,Pindar, and Bacchylides. Irene Behrens remarks that the word melikoswas applied in Alexandria to the noncanonical poets, while lyrikosqualified almost exclusively the Nine.41 These would be the novemlyrici whom Horace would strive to join (Carm. I, 1, 35-36)-withevery success, according to Quintilian, who had declared that Pin-dar was the first, by far, of the nine lyric poets (X, 1: "novem verolyricorum longe Pindarus princeps"). Later pantheons and pleiadesof lyric poets, of course, would follow. And I need not stress the dis-tinctions between a canon of great writers, such as the novem lyrici,and an anthology of actual works. The former implies an adjectivalkind of comment-lyricus as quality or mode-while the latter maypresuppose a species, a genre, a subgenre. Meleagers famous Garland (beginning of first cent. B.C.42),on which the Greek Anthology wasbuilt, was a collection of epigrams, the lightest and briefest of poeticforms. Some of the poets picked by Meleager were second-rate--or atleast, subordinate to the genre and to the anthology as a whole. One may ask: is it a fact that the choice of either classical "canons"or anthologies played a central role in the rise of the "lyrical" modesor genres? This is a question which-like many of the important ques-tions of literary theory-can only be answered by comparatists trainedin the Oriental literatures. The oldest of Chinese poetic monumentswas a vast collection, the Shih Ching (or Shih King), or Classic ofSongs, containing the words of 305 poems composed and sung between1000 and 700 B.C. In James R. Hightowers words: "the Classic ofSongs is after all an anthology of poetry, and its four-fold division intoFeng, Big and Little Ya, and Sung, may have been an attempt by itscompiler to establish different categories of song."43 Another ancientcollection was the Chu tzu, which preserved the special "elegiac" 40 Curtius, p. 249: in Aulus Gellius, classicus is, metaphorically, a first-classauthor, meaning not a tax-payer, unlike the proletarians. 41 Behrens, p. 7. 42 The Greek Anthology. Hellenistic Epigrams, ed. A. S. F. Gow and D. L.Page (Cambridge, 1965), "Introduction," I, xvi: "it is not necessary to con-sider here the sources from which Meleager compiled his Garland beyond notingthat the belief once held that he was the first anthologist is no longer tenable.Papyrus scraps of anthologies of sufficiently early date make it plain that he hadpredecessors . ." 43 "The Wen hsiian and Genre Theory," 515, n. 7. 211
  • 21. COMPARATIVE LITERATUREverses known as sao. But a more precise attempt on the part of an-thologists to delimit the various provinces of poetry took place muchlater: from the days of Chih Yii, who compiled "the first knownanthology of diverse genres,"44and died around 310 A.D., to the sixth-century Wen hsiian, with its thirty-seven different types of writing."The development of genre theory in China"-Hightower sums up-"has been closely associated with anthology making."45 As for Arabic literature, the influence of anthologists and philologists,like Abf Tammam and al-Buhturi, was considerable during its goldenage, from the eighth century to at least the eleventh.46"The report thatal-Walid II (d. 744) caused to be made a collection of the dzwan ofthe Arabs... is thought to be quite plausible."47Since the eighteenthcentury, and Sir William Jones translation (1783), European read-ers have known about the Muallaqat or "Seven Suspended Odes"-the illustrious poems chosen in the eighth century by Hammad al-Rawiya, "the Transmitter," and regarded for hundreds of years asmodels of the all-important art of the qasida.48 I spoke earlier of the effects on poetic theory of the grammariansturn of mind, which was schematic and normative: language was di-vided-or shattered-into parts of speech, and then put back togetheragain by means of orderly rules. These structures were bequeathedto the European Middle Ages by the scholars of imperial Rome.Among these, the grammaticus Diomedes (fourth cent.) was influen-tial, for he had something to say on the subject of poetics. His grammardistinguishes three modes: a genus activum vel imitativum, in whichthe poet does not intervene and the characters "act alone"; a genusenarrativum, where "the poet himself" speaks; and a genus commune,combining the other two. Moreover, Diomedes decides to assign tothis traditional triad a number of subordinate "species" or genres. Eachof these becomes a submode. For example: there are four species ofthe first mode-tragica, comica, satyrica, mimica. Although they appearto be theatrical forms (including mimes and satyr plays), the emphasisis on direct speech: Diomedes fits into this category the first and nintheclogues of Virgil. The second mode embraces, among other genres,didactic writing (Empedocles, Lucretius, the first three books of theGeorgics), which Aristotle had considered alien to poetry (1447b). 44Ibid.,p. 515. 45Ibid.,p. 512. 46Francesco Gabrieli,Storia della letteraturaaraba (Milan, 1962), pp. 29, 165-166, etc. 47 A. J. Arberry, The Seven Odes (London-New York, 1957), p. 17. 48 The study of the seven odes by A. J. Arberry, cited in the previous note,devotes a great deal of space to the history and impact of their Europeantransla-tions.212
  • 22. POETICS AS SYSTEMLucretius is placed among the poets who speak "for themselves." Butnot Horace, curiously enough, as lyric poetry is subsumed under thethird or mixed mode. There are two species of the genus commune:heroica species (Homer) and lyrica species (Archilocus and Horace,explicitly; and one assumes, the remaining eclogues of Virgil).4°Diomedes system included three modes and nine species. Obviously,it was with some difficulty that he constructed such a tidy scheme. Buthis contribution to the history of poetics was far from negligible, forhe was one of the first who tried to insert a fixed number of poeticforms into a system of presentational modes, that is to say, to proposea system of genres. This was the strained but orderly solution whichwould begin to prevail during the Renaissance, and ultimately makepossible the modern triad of genres. The old scheme, Platonic and Aristotelian, retained its position formany centuries still. Julius Caesar Scaliger had no other division tooffer in his highly respected Poetices libri septem (1561).50 But ofcourse he was not interested in the great vernacular writers, nor didhe attempt to adjust the traditional systems to the masterpieces of hisown age. Other theorists, however, were more sensitive to the needfor different models and "ideal spaces." I have already mentionedAntonio Minturno and Francisco Cascales. Minturno attended theCouncil of Trent, and became Bishop of Ugento and Cretone. His Latintreatise, De poeta (1559), was followed by a rather more didacticdialogue in Italian, Larte poetica (1563).51 Both books offered theterms of a new triad: "Vespasiano: Quante adunque sono le parti dellapoesia? Minturno: Tre generali: luna si chiama Epica, laltra Scenica,la terza Melica, o Lirica, che dir vi piaccia."52This formal scheme wasbuttressed with numerous other triads: there were three kinds ofimitation, three types of character, three species of the epic, three"causes" of poetry, three kinds of drama, etc.53 But Minturno did notmake clear that the principal types were "species" instead of "modes."Cascales did, as we have observed. There is little we know about thisinteresting transitional figure except that he was a morisco fromMurcia: though his ancestors had held, in other words, that Allah wasOne, he was a most zealous believer in la divina Triada. So were otherChristian Humanists of the sixteenth and seventeenth century. Irene 49 On Diomedes see Curtius, pp. 439-441; Behrens, pp. 25-30. 50 Book I, Chapter3, titled "Poematumper modos divisio, et eorum ordo," p.6 of the facsimile edition by August Buck (Stuttgart-Bad Cannstatt, 1964). 51 BernardWeinberg, "The Poetic Theories of Minturno,"in Studies in Honorof Frederick W. Shipley (Freeport, N.Y., 1968 reprint), pp. 101-129. 52Larte poetica del signor Antonio Minturno (Naples, 1725), p. 3. There themodes are called parti. But a marginal note reads: "tre maniere di Poesia." 53Minturno, pp. 2-8. 213
  • 23. COMPARATIVE LITERATUREBehrens has told in her very useful book the rest of the story-thegradual diffusion of the tripartite system in poetics from the seven-teenth century to the nineteenth century: through Gravina (1708),1Abbe Batteux (1746), the rise of a broader conception of the lyric,despite the weakness and the delaying effects on that score of theFrench neoclassical theorists, and the triumph in Germany around1800 of the Dreiteilung der Dichtkunst, the three-fold division of theliterary art. Roman Jakobson and Nikolai S. Trubetzkoy started to develop morethan forty years ago a structural conception of language as a set ofbinary oppositions, especially from a phonological point of view: thephonetic events themselves were not stressed as much as the system ofoppositions and correlations which they manifested. This was incor-porated into Jakobsons "distinctive feature system," where all featuresevidence binary structures (voiced-unvoiced, tense-lax, strident-mel-low, continuant-interrupted, etc.). In the Fundamentals of Language,written with Morris Halle, these structures were reduced to twelveoppositions, "out of which each language makes its own selection."54In the social sciences too, the British anthropologist A.-R. Radcliffe-Brown explored the varieties of forms of primitive social life in orderto infer, by means of the comparative method, certain universal struc-tural principles such as the union of opposites: Heraclitus had saidthat "Polemos (strife) is king, and rules all things"; and the Yin-Yang systems of ancient China (the union of male and female, day andnight, summer and winter, producing the organized totality or tao: thecouple, the day, the year, etc.) corresponded probably to social cus- toms like the pairing of rival clans by marriage.55 Radcliffe-Brown, however, abstracted his "structural forms" from the variations of par- ticular instances, or from the proven continuity of certain social struc- tures in historical time.56 Claude Levi-Strauss, in an exceptionally thought-provoking series of studies, went on to postulate logical models in anthropology on the basis of structural analogies between the human mind and the world which it interprets and organizes. In Le Totemisme aujourdhui (1965), for example, each group of relations is treated as 54Roman Jakobson and Morris Halle, Fundamentals of Language (The inHague, 1956), p. 29; see also Jakobson,"Retrospect," Selected Writings (TheHague, 1962), I, 631-658; and Halle, "In Defense of the Number Two," inStudies Presented to Joshua Whatmough,ed. Ernst Pulgram (The Hague, 1957),pp. 65-72. 55A.-R. Radcliffe-Brown,"The ComparativeMethodin Social Anthropology,"in Method in Social Anthropology (Bombay, 1960), pp. 91-108. 56Meyer Fortes, "Time and Social Structure: An Ashanti Case Study," inSocial Structure. Studies Presented to A.-R. Radcliffe-Brown, ed. M. Fortes (New York, 1963), p. 54.214
  • 24. POETICS AS SYSTEMa particular instance of a broader system of thought that is either realor potential, so that all instances may be derived or explained throughthe appropriate rules of transformation; thus, the so-called totemismof primitive tribes is but a single way of expressing the basic oppo-sitions that are also manifested in myths, customs, and beliefs. Turning to poetics again, we might venture to ask: do ternary sys-tems function on all levels? When the ideal spaces of poetics are tri-partite-on the explicit level of genre theory, at least, where tertiumdatur-are the fundamental dichotomies at work too? Surely a prin-ciple as essential to most conscious or unconscious frameworks asopposition (logical, psychological, formal, etc.) plays a role in theformulation of the choices and the alternatives of which the systems ofpoetics are made. One can readily discover in these systems the jointoperation of binary and ternary forces-both tending, simultaneouslyor on different levels, toward systematization. With these questions in mind, let us return once more to theinvestigations of Hermann Usener into the intellectual systems ofGreece and Rome. Usener documents the impact of the orders ofgrammar on the other branches of culture. The basic grammaticalform was the tetrad (as in the division of Roman textbooks into lectio,enarratio, emendatio, iudicium). (Tetrads, of course, may unfold andresolve an original set of polarities, as in the "natural" dualism ofwinter and summer, barrenness and fertility, spreading out in the fourseasons of the year. In our day, the "naturaltetrad" reappears in FryesAnatomy of Criticism.) Aristotle had referred in several of his worksto the "table of opposites" of the Pythagoreans (Met. 986a: "Limit-Unlimited, Odd-Even, Unity-Plurality, Right-Left, Male-Female, Rest-Motion, Straight-Crooked, Light-Darkness, Good-Evil, Square-Ob-long"). In the Physics he explained that the concept of opposites wascharacteristic not only of the Pythagoreans but of their predecessors:"for they all, even when they assumed them without due process ofreasoning, nevertheless taught that the elements and what they calledfirst principles were opposites, as if under the compulsion of the truthitself" (Phys.. 188b). The allusion was to Empedocles, Heraclitus, andthe search on the part of Pre-Socratics for interacting archai or firstprinciples.57Useners emphasis is on the prevalence of both triads andtetrads in Greece: among the grammarians, Asclepiades of Myrlea em-ployed a ternary system, in the manner of the writers on medicine,architecture, or city-planning, of the tragedian and philosopher Ion ofChios (fifth cent. B.C.), reputed to have written a Triagmos or dis- 57 On the Pythagorean opposites see J. A. Philip, Pythagoras and Early Py-thagoreanism (Toronto, 1966), pp. 44-59. 215
  • 25. COMPARATIVE LITERATUREquisition on the number three, of Platos Republic,58 and the variousfollowers of Aristotle.59 Aristotle himself singled out the three partsof the syllogism, and the idea that virtue is a mean between oppositeexcesses (as in "cowardice-courage-temerity"). In his Politics heapplied the same doctrine of the mean to the three classes of society:"in all states there are three elements: one class is very rich, anothervery poor, and a third is a mean" (Pol. 1295b). And in the De Caelohe laid stress on the number three with regard to the motion and thedimensions of natural bodies:Magnitudedivisible in one directionis a line, in two directions a surface, in threedirectionsa body. There is no magnitudenot includedin these; for three are all,and in three ways is the same as in all ways. It is just as the Pythagoreanssay,the whole world and all things in it are summedup in the numberthree; for end,middle, and beginning give the number of the whole, and their number is thetriad. Hence it is that we have taken this numberfrom nature, as it were one ofher laws, and make use of it even for the worship of the gods.60Usener goes on to illustrate the diffusion of the tetrad in Rome, par-ticularly through the influence of Varros linguistic works.61 This pointis confirmed by Curtius with respect to the different "levels of style":he sees a passage, in this area, from a Hellenistic trichotomy to a tetradand other systems in classical and late Rome. The earliest evidence wehave of the notion of the three styles is in the Rhetorica ad Herennium (c. 86-82 B.c.) and in Cicero. But Quintilian, while discussing the samethree, suggests that intermediate styles are legitimate too; and thetetrad of styles was proposed by Hermogenes and Demetrius (firstcent. A.D.), whose four types were the "plain," the "elevated," the"elegant," and the "forcible."62 (Curtius extends his observation to theMiddle Ages, a period during which, in his opinion, the boundaries be-tween not only the styles but poetry and prose were most fluid: "thedivision of the ars dictcaminis [into meter, rhythmical or rhymed prose,and regular prose] had as a result the replacement of the dyad poetry- 58 I mentionedearlier in this article, with regard to the three modes, this aspectof the Republic; cf. Ernest Barker, Greek Political Theory, 3rd ed. (London,1947), pp. 215, 263, etc. 59Usener, art. cit., especially 272-276. On the city-planner Hippodamus ofMiletus, and the three classes of his ideal state (craftsmen, farmers, soldiers), cf.Barker, op. cit., p. 81. On Ions Triagmos see Pauly-Wissowa, Real-Encyclopddie.(Stuttgart, 1916), I, 1864-1865. 60 On the Heavens, tr. W. K. C. Guthrie, The Loeb Classical Library (Cam-bride, Mass.-London,1939), p. 5. 61 Usener, pp. 303-307. 62 Demetrius, op cit., II, 36, p. 323; and E. R. Curtius, "Die Lehre von dendrei Stilen in Altertum und Mittelalter (zu Auerbachs Mimesis)," RomanischeForschungen, LXIV (1952), 57-70.216
  • 26. POETICS AS SYSTEMprose by a triad or a tetrad. The boundaries between poetry and prosethus became more and more blurred."63) Evidently, the doctrine of the mean implies the existence of binaryoppositions, in ethics, society, or literature: "in Aristotles own ethicaltheory the aim of ethical conduct"-writes J. A. Philip-"is presentedas some one point that is median, but not necessarily the middle, in thewhole locus or area lying between contraries."64Similarly, the secondor median term in the old trichotomy of styles implies the polarity thatlogically precedes it-indeed it is meaningless without it (the stylecalled mediocre in St. Isidore refers solely to the contraries grandi-loquum and humile)-even as the medieval rhymed prose is dependenton a broader framework, and the prose poem in the nineteenth century,on Baudelaires regular prose and regular poems. As I close this paper, my concern is with the ways in which the"natural"binary opposition yields, for different historical reasons, andin connection with various linguistic and social systems, to more com-plex "cultural" structures, of which the most important, in the field ofpoetics, has been the triad. I shall only add some suggestions or topicsfor further study. It would be useful, I think, to read the basic textsin the history of poetics from the point of view of the combined effectof binary and ternary systems, and of the conditions underlying thesesystems. The archetypal example, of course, would be the Poetics ofAristotle. There are also conflicting structures in the poetics of ro-manticism that would be deserving of similar study. Now, Gerald F.Else has shown that the procedures of diaeresis (Plato, Phaedrus,265d-266a: concepts are split into two until genuine classes are dis-cerned) are applied by Aristotle in the Poetics-but to the partsof a triad: "within the tripartite division... Aristotle does proceeddiaeretically."65For example, the three differentiae of poetry are de-fined-"medium," "object," and "mode." Under the heading of each,then, a dichotomy is introduced. The resulting oppositions stand outmost forcefully in the historical observations on the origins of poetry1448b-1449a) : poetry originally "split up" according to the characterscorresponding to it, either superior or worthless men; it producedeither invectives or heroic hymns and encomia; as time passed, invec-tive gave way to comedy, and heroic or epic verse to tragedy: "thelampooners became writers of comedy, and the epic poets were suc-ceeded by tragedians, since the drama was a larger and higher form 63 European Literature and the Latin Middle Ages, p. 149. 64 Pythagoras and Early Pythagoreanism, p. 49. 65 Else, op. cit., p. 16; cf. also pp. 67-68, 91-101, etc.; and among Platos otherdialogues, Philebus 16-18, the Sophist 253d, etc. 217
  • 27. COMPARATIVE LITERATUREof art."66 Clearly, in Aristotles opinion, though forms and genres maychange, a basic polarity remains; and the central genres-tragedy, epic,comedy-have their places in a continuing system of binary oppo-sitions.67 "Woe to the cognoscenti who love their systems more than beauty"-Friedrich Schlegel warned his brother August Wilhelm in a letter of1796-or to the theorist who "must destroy history" in order to "sup-port his system" !168 It was characteristic enough of the romantics toboth detest and pursue all kinds of intellectual order. Friedrich Schlegelhimself elaborated the idea that the history of Greek literature wasexemplary insofar as it comprised three successive stages-first, epic,then lyric, and finally dramatic-that is, the notion that triadic peri-odization reflected the rise of the three genres. This was a decisiveattempt to reconcile the historical imagination with the traditional senseof system. It involved a temporal unfolding, as it were, of the archetypal"ideal space."69 August Wilhelm Schlegel, in his 1801 lecture series (Vorlesungen iiber schbne Literatur und Kunst), applied the Fichteanscheme of "thesis-antithesis-synthesis" to the principal genres. Platosoriginal division was invalid-the notes began; and Schlegel proceededto apply a contemporary philosophical system to poetic theory:Einteilung der Gattungenbeim Plato... Ungiiltig. Kein poetischer Einteilungs-grund.-Episch, lyrisch, dramatisch; These, Antithese, Synthese. Leichte Fiille,energische Einzelheit, harmonischeVollstandigkeitund Ganzheit... Das Epische,das rein Objektive im menschlichen Geiste. Das Lyrische, das rein Subjektive.Das Dramatische, die Durchdringungvon beiden.70 The success of such tripartite conceptions in the nineteenth centurycoincided with a decidedly dualistic view of literary history, of which 66 Poetics, tr. Butcher, p. 17. 67Averroes famous confusion of tragedy with panegyric, and of comedy withsatire, which MenendezPelayo consideredabsurd (Historia de las ideas esteticasen Espana, Madrid, 1909-1912,II, 137) preserved at least the basic polarity inAristotle, which became central in the rhetorical tradition (the topoi conveyingthe idea of an ideal man, an ideal landscape, etc.) with which Averroes was insympathy. A poetics based on satire and eulogy was developed in the sixteenthcentury by Antonio Viperano (Weinberg, I, 209). 68Friedrich Schlegels Briefe an seinen Bruder August Wilhelm, ed. Oskar F.Walzel (Berlin, 1890), p. 263: "Wehe dem Kenner, der sein System mehr liebtals Sch6nheit,wehe dem Theoristen, dessen System so unvollstandigund schlechtist, daB er die Geschichtezerstoren muB,um es aufrecht zu erhalten!" 69 This began with historical periodization itself: the triads like ancient,medieval,and modern,or Altertum, Mittelalter, Neu-zeit,were first used fully byCellarius (1634-1707), who wrote concerninghistoria antiqua,historia medii aevi,historia nova-though the idea goes back to the Renaissance: cf. Ernst Bernheim,Lehrbuchder historischenMethode und der Geschichtsphilosophie, ed. (Leip- 5thzig, 1908), p. 78. 70A. W. Schlegel, Kritische Schriften und Briefe, Vol. II: Die Kunstlehre,ed.Edgar Lohner (Stuttgart, 1963), pp. 305-306.218
  • 28. POETICS AS SYSTEMthe most famous expression was probably Schillers essay on "naive"and "sentimental" poetry (1795). Poets, playwrights, critics foughtand lived for a radical distinction between two attitudes and traditions-the "classical" and their own. The tendency of the romantic was topolarize and divide writers and writing into opposite camps and mili-tant movements, as Francesco de Sanctis remarked long ago: "studiatetutte le concezioni romantiche, e vi troverete in fondo unantitesi."71 By and large, the nineteenth century would often witness the co-existence of binary and ternary systems, from Hegel and Marx to Taineand Freud.72As far as poetics are concerned, the best example I knowis Victor Hugos Preface to Cromwell. "Mettons le marteau"-he de-mands-"dans les theories, les poetiques et les systemes." Elsewherein the same manifesto, nevertheless, he expatiates fully upon the threeages of poetry-"dont chacun correspond a une epoque de la societe:lode, lepopee, le drame."73 (Again, the triad is "historical.") AndHugo goes on to acclaim and proclaim an ultimate or "natural" po-larity between the grotesque and the sublime in literature. Poetry willadvance and thrive in the future, if only the ideas of "la critique phi-losophique" can be understood:Elle [la poesie] se mettra a faire comme la nature, a meler dans ses creations, sanspourtant les confondre, lombre a la lumiere, le grotesque au sublime, en dautrestermes, le corps a lame, la bete a lesprit; car le point de depart de la religion esttoujours le point de depart de la poesie. Tout se tient.74 These last words, "tout se tient," and the conception that they ex-press, have been applied to language by modern linguistics, since Saus-sure, with conspicuous success. It remains to be seen whether theyare -equallyrelevant to literary history. We have just observed that onthe level of poetics, systematization has been remarkably explicit, per- 71 Saggi critici, ed. Luigi Russo (Bari, 1952), III, 3 . 72 I have in mind, concerning Marx, the insistence on the essential class conflictbetween a ruling and an oppressed class, while on other occasions he maintainsthe idea of a middle class as part of a trichotomous structure. See the importantbook by Stanislaw Ossowski, Class Structure in the Social Consciousness (NewYork, 1963) ; and Pierre L. van den Berghe, "Dialectic and Functionalism: Towarda Theoretical Synthesis," American Sociological Review, XXVIII (1963), 695-705. Taines "trois forces primordiales" (race, milieu, momeent) are well known:see his cause-and-effect schemes in the Introduction a lHistoire de la litteratureanglaise, ed. Gilbert Chinard (Princeton, 1944), VII, p. 25: "il y a ici des couplesdans le monde moral, comme il y en a dans le monde physique, aussi rigoureuse-ment enchaines et aussi universellement repandus dans lun que dans lautre. Toutce qui dans un de ces couples produit, altere ou supprime le premier terme, produit,altere ou supprime le second par contre-coup." In Freud, trichotomous divisionsof personality give way in the late Abri/3 der Psycho-analyse to the fundamentalconflict between Eros and the death instinct (Destruktionstrieb or Todestrieb). 73 Theaitre complet, ed. R. Purnal, J.-J. Thierry, and J. Meleze (Paris, 1963),pp. 434, 422. 74 Ibid., p. 416. 219
  • 29. COMPARATIVE LITERATUREsistent, and self-perpetuating. But we also know that Baudelaire wrote,as any poet might: "un systeme est une espece de damnation qui nouspousse a une abjuration perpetuelle."75 How does the poet confrontthe fact that he writes, and exists as a writer, within a system of signs,a network of forms, that is as traditional and perhaps more "conserva-tive" than the community or nation in which he lives? In what waysis literature for him a "potential" system? A grammar of models? Notmerely a musee imaginaire, but une cite imaginaire? Or rather, hisown personal version of that city, which is like a "secondary" or "inter-mediate" system ? For the moment, I think, our need is to clarify the terms of theproblem. The systematics of genre theory offer one of several possiblepoints of departure. But there are others, and the relations betweenthe various theoretical levels demand a great deal of research. In thispaper, I have discussed briefly the historical interaction between genretheory and the other orders-such as the systems of "great authors,"and the "three styles." Regarding the latter, we have the exceptionalexample of Erich Auerbachs Mimesis, which is based on a highlysimplified and effective model of the relationship between systems ofgenres and systems of styles. Information is available concerning genresand their connection with rhetorical classes (the five divisions ofrhetoric, and the three kinds of eloquence76), or with the schemes ofancient psychology and ethics.77 I have only mentioned in passing theinfluence of social systems. This is a question that should not beanswered simplistically, or dismissed without second thoughts. His-torians of literature are reasonably well acquainted with such systemsas the medieval "Virgils wheel," and its application of a triadic socialspace (actually, a circle divided by three radii) to poetic themes andforms;78 the framework of numerous triads (social, linguistic, etc.) 75 Euvres, ed. Y.-G. Le Dantec (Paris, 1961), p. 995. 76The five divisions were inventio, dispositio, clocutio, memoria, actio; thethree kinds of eloquence, the judicial, the deliberative, and the panegyrical orepeideictic (judiciale, deliberativum, demonstrativum) the three functions of ;oratory, according to Cicero, docere, delectare, movere. Cf. Curtius, op. cit., pp.68-69; and CharlesC. Baldwin, MedievalRhetoric and Poetic (Gloucester, Mass.,1959), p. 64. 77 Behrens, op. cit., pp. 124, 174. It was traditionalto mention correspondencesbetweenthe three modes of poetry and the three faculties of man-memory, rason,imagination. But this sort of parallelism has remained operative until our day.Cf. Friedrich Schlegels referencesto Geist, Seele and Korper in "Gesprachfiberdie Poesie" ("SchluBfteil zweiten Fassung"), in Charakteristiken Kritiken der undI, ed. Hans Eichner (Miinchen-Paderborn-Wien, 1967), p. 356. 78The rota Vergilii can be found in the Poetria of John of Garland (XIIIthcent.); cf. Faral, Les Arts poetiques du XIIe et du XIIe siecle (Paris, 1962),p. 87; the full text was publishedby Giovanni Mari, "Poetria magistri Johannisanglici de arte prosayca metrica et rithmica," Romanische Forschungen, XIII220
  • 30. POETICS AS SYSTEMin Dantes De vulgari eloquentia ;79 the hardening of social analogiesin certain literary theories and hierarchies, markedly "monarchical,"ofthe Renaissance and Baroque; the division of poetic modes in ThomasHobbes according to "the three regions of mankind, court, city andcountry,"80etc. We have also seen that Aristotles three social classesowe a great deal to his doctrine of the mean, and to the logical andethical roots of this doctrine. It is itself a theory-insofar as the phi-losophers mind, in Wittgensteins words, "is held captive by a pic-ture."81The idea that the modern novel is a product of the bourgeoisierests on a notoriously convincing analogy not between two sets of harddata, but between two "pictures," and two systems. Class distinctions,like theories of genre, mediate between the events themselves and theconsciousness of individuals and groups. There are two ways of confronting the problem of literary historythat I cannot recommend. The first consists in dissolving the processesof literature into "general history." The immediate aim becomes a suc-cession of totalities, each of which assumes absolutely the unitary prin-ciple: not only tout se tient but tout est tun. Should the historian recordmore than the efforts of men to devise and establish, vis-a-vis the one-ness of the physical world around them, orders that are comparablyinterrelated and coherent? But "one may suspect"-quietly writesJorge Luis Borges-"that there is no universe in the organic, unitarysense of this ambitious word." ("Cabe sospechar que no hay universoen el sentido organico, unificador, que tiene esa ambiciosa palabra."82)The second attempts to find meaningful relations between the singlepoem or work and entire social or economic or intellectual systems.The incongruity is obvious, and the problem, thus posed, absurd. (It isnearly as absurd as the attempt to relate a single economic fact to anentire literary period.) My interests-to propose a final triad-lie some-(1902), 900. The three parts correspondto Virgils Eclogues, Georgics,or Aeneidrespectively. The A eneid stands for the gravis stylus, the soldier as a type, Hectorand Ajax as heroes, the horse as an animal, the sword, the city or the war-camp,the laurel and cedar trees, etc. 79 To the ydioma triphiarum (the languages of oc, oil, and sd, reflecting thetriad of the classical languages: Latin, Greek, Hebrew) correspond the threesocial classes in Sicily (I, xiii), the three kinds of beings (angels, men, beasts),the three parts of man, the three kinds of action (II, ii), the three styles and thethree parts of the canzone, etc. I refer to the divisions in the De vulgari eloquentia,ed. Aristide Marigo and Pier Giorgio Ricci (Florence, 1957). 80 "The Answer of Mr. Hobbes to Sir William Davenants Preface beforeGondibert," in The English Works of Thomas Hobbes (London, 1840), IV, 443. 81 Cited by Avrum Stroll, "Statements," in Epistemology. New Essays on theTheory of Knowledge, ed. A. Stroll (New York, 1967), p. 192. 82 "El idioma analitico de John Wilkins," in Otras inquisiciones (Buenos Aires,1960), p. 143. 221
  • 31. COMPARATIVE LITERATUREwhere between these two extremes. Literary systems, like social orlinguistic ones, exist; whereas the "universe," as Borges suspects, maynot. Our task is to identify the careers of these different systems inhistorical time, to discover those that prevailed, and to listen to thedialogue between them. University of California, San Diego222