Katharsis Author(s): Jonathan Lear Source: Phronesis, Vol. 33, No. 3 (1988), pp.

297-326 Published by: BRILL Stable URL: http://www.jstor.org/stable/4182312 . Accessed: 14/03/2011 04:17
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=bap. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Phronesis.

http://www.jstor.org

Katharsis
JONATHAN LEAR

I 1. Tragedy,says Aristotle, is a mimesisof a seriousand complete action, havingmagnitude,whichthroughpity and fear bringsabout a katharsisof such emotions.' But what Aristotle meant by what he said, in particular, what he meant in claimingthat tragedyproducesa katharsis,is a question which has dominatedwestern philosophyand literarycriticismsince the Renaissance.2 In the last hundredyearsit has been widelyacceptedthat by Now thereis a sense of the emotions.3 katharsis Aristotlemeanta purgation in which the interpretationof katharsisas purgationis unexceptionable: havingarousedthe emotionsof pity andfear, tragedydoes leave us with a feeling of relief; and it is naturalfor humansto conceive of this emotional process in corporealterms: as havinggotten rid of or expelled the emotions.4But at thislevel of generality,the interpretation is as unhelpfulas it is
See Poetics 6, 1449b22-28. See Baxter Hathaway, The Age of Criticism:The Late Renaissancein Italy (Ithaca: Comell University Press, 1962), pp. 205-300.
This is largely due to Jacob Bernays' influential Zwei Abhandlungen uber die aristote-

lische Theoriedes Drama (Berlin, 1880, first publishedBraslau, 1857). A chapterof this book has been translated as "Aristotle on the Effect of Tragedy" by Jonathan and Jennifer Barnes in Articleson Aristotle, v. 4 (J. Barnes, M. Schofield & R. Sorabji eds., London, 1979). Bernays' interpretation had a wider influence than on Aristotelian scholarship alone; for Bernays was Freud's wife's uncle and it seems that Freud and Breuer were awareof the interpretationand relied on it when formulatinghis conception of catharsisin the early stages of the formationof psychoanalytictheory. (See Bennett Simon, Mind & Madnessin Ancient Greece(Ithaca: Cornell University Press, 1978, pp. 140-143).) The katharsis-as-purge metaphoris used by Plato in the Sophist (230C-231E) where the Socratic elenchus is represented as purgingone of false beliefs. See e.g. SigmundFreud, The StandardEdition of the CompletePsychological Works 4 of Sigmund Freud, (London: Hogarth Press, 1981), X:233-34, XII:218-26; XIII:78-90; XIV:73-102, XIX:235-9; Wilfrid Bion, Learning From Experience (London: Karnac, 1984) and Second Thoughts, (London: Karnac, 1984); Melanie Klein, Narrativeof a Child Analysis (London: Hogarth Press, 1961) pp. 31ff, Contributionsto Psycho-AnPhronesis 1988. Vol. XXXIII/3 (AcceptedFebruary1988) 297

unexceptionable.For what we wish to know is how Aristotleconceivedof of a tragedy.Even if as it occursin the performance the processof katharsis in namingthis we acceptthat Aristotledrewon the metaphorof purgation reallythink did he is: know to want we what "katharsis", process emotional that this process was an emotional purgationor did he merely use the metaphorto name a processthat he understoodin some differentway?At the level of mere metaphorthereseems little reasonto choose betweenthe medical metaphor of purgationand its traditionalreligiouscompetitor, purification,not to mentionmore generalmeaningsof "cleansing","sepuse whichAristotlemakes of the aration"etc.5 In fact, the preponderant As faras I know, no discharge.6 is as a termfor menstrual word"katharsis" has suggestedthe model one in the extendeddebate about tragickatharsis of menstruation.But why not? Is it not more compellingto think of a naturalprocessof dischargeof the emotionsthan of theirpurging? It is only when we shift from the questionof what metaphorsAristotle mighthave been drawingon to the questionof whathe took the processof katharsisin tragedyto be that there is any point in choosing among the out whatAristotlemeantby variousmodels. Of course,the taskof figuring by a passing katharsisis made all the more alluring,as well as frustrating, the katharsis remarkwhichAristotlemakesin the Politicswhile discussing that music produces: "the word 'katharsis'we use at present without explanation,butwhenlaterwe speakof poetrywe willtreatthe subjectwith
more precision."7 We seem to be missing the section of the Poetics in which

Aristotle explicitlyset out what he meant.8

alysis (London: Hogarth Press, 1948), pp. 140-151, 303, Developments in PsychoAnalysis (Ldndon: Hogarth Press, 1952); W.R.D. Fairbairn,"Schizoid Factors in the Personality", in Psychoanalytic Studies of the Personality (London: Routledge and Kegan Paul, 1984); Richard Wollheim, "The Mind and the Mind's Image of Itself', in On Art and the Mind (Cambridge, Mass.: Harvard University Press, 1974), "WishFulfilment",in RationalAction (ed. Ross Harrison, Cambridge:CambridgeUniversity Press, 1979), The Threadof Life (Cambridge:CambridgeUniversity Press, 1984). s The idea that purgationand purificationneed not be treatedas contrariesis arguedby Humphrey House, Aristotle's Poetics (London: Rupert Hart-Davis, 1956), p. 104-111, and by Stephen Halliwell, Aristotle'sPoetics (Chapel Hill: Universityof North Carolina Press, 1986), pp. 184-201. 6 See e.g. Generationof Animals I.20, 728b3, 14; IV.5, 773b1; IV.6, 775b5; Historyof Animals VI.18, 573a2, a7; VI.28, 578b18; VII.2, 582b7, 30; VII.4, 584a8; VIII.11, to describe seminaldischarge, Genera587b2, b30-33, 588al. For the use of "katharsis" tion of Animals II.7, 747a19; for the discharge of urine: History of Animals VI.18, 573a23; for birth discharge:History of Animals VI.20, 574b4. 7 Poltcs VIII.7, 1341b37-39. only twice in Poetics:once, as we have seen, in the Aristotle uses the word "katharsis" 8

298

In this paperI will firstisolate a series of constraints whichany adequate interpretation of katharsismust satisfy. These constraintswill be derived from a consideration of Aristotle'sextendeddiscussionof the emotions, of the effect of tragedy, and of how tragedyproducesthis effect. The constraintsmay not be tight enough to delimit a single acceptableinterpretation, but I shall argue that they are strong enough to eliminate all the traditionalinterpretations. Second, I will offer an interpretation of tragic katharsiswhichsatisfiesall the constraints. 2. Let us begin with the suggestionthat a katharsisis a purgationof the emotions.To takethissuggestionseriouslyone mustthinkthat, for Aristotle, katharsisis a cure for an emotionallypathologicalcondition: tragedy helps one to expel or get rid of unhealthilypent-up emotions or noxious emotional elements.9The only significantevidence for this interpretation comes fromAristotle'sdiscussionof the katharsis whichmusicproducesin the Politics:10
definition of tragedy and once to refer to the ritual of purificationat which Orestes is recognized by his sister, Iphigenia, Poetics 17, 1455b15). 9 Bernays is explicit that katharsisis a cure for a pathological condition. '0 See Politics VIII.5-7. Bemays argues persuasivelythat to understandthe concept of tragic katharsis, we must look to Aristotle's discussion in the Politics of the katharsis whichmusicproduces;though, as we shallsee, he is less persuasivein his interpretation of that discussion. G.R. Else and, following him, Leon Golden have arguedthatone should not look outside the Poetics for the meaning of tragic katharsis. (G.F. Else, Aristotle's Poetics: The Argument (Cambridge: Harvard University Press, 1957) p. 439ff; Leon Golden, "Catharsis",Transactions and Proceedingsof the American Philological Association, 1962;and "Mimesisand Catharsis",ClassicalPhilology, 1969.) This, I believe, is a misapplicationof a principle from new criticism. The Poetics was not meant to be a self-containeduniverse;it was an integralpartof Aristotle's philosophy. If, for example, we were trying to determine what Aristotle meant by art (techne) or poetry (poiesis) in the Poetics, there would no plausibilityto claiming that we should completely restrict ourselves to the Poetics' discussion. Of course, Aristotle does use "poiesis"in a special way in the Poetics:it is to be translatedas "poetry"ratherthan as a "making"which is the appropriatetranslationin the Metaphysics.However, if we ignore all other Aristotelian works we remainblind to the philosophicallyimportantfact that, for Aristotle, poetry is a special type of making. There is no doubt that we must approachother texts with care, for, to return to our current concern, Aristotle's use of "katharsis"when discussing medical purging may be different in significantrespects from his use of the term when discussing tragedy. But such interpretive difficulties are not sufficient grounds for ignoring other texts altogether. (Indeed, Else's and Golden's stricture led them to formulatea highlyimplausibleaccountof katharsis,in which katharsisis not an effect on the audience of tragedy, but a resolution of the events in the play. This implausible interpretationdepends upon an even more implausibletranslationof Aristotle's definition of tragedy. For an excellent criticismof this interpretation,see Stephen Halliwell,

299

We accept the division of melodies proposed by certain philosophersinto ethical melodies, melodies of action, and passionateor inspiringmelodies, each having, as they say, a mode correspondingto it. But we maintainfurtherthat musicshould be studied, not for the sake of one, but of many benefits, that is to say, with a view to education, to katharsis(the word katharsiswe use at present withoutexplanation, but when hereafterwe speak of poetry we will treatthe subjectwith more precision) - music may also serve for intellectualenjoyment, for relaxationand for recreation after exertion. It is clear, therefore, that all the modes mustbe employed by us, but not all of them in the same manner. In education the most ethical modes are to be preferred,but in listeningto the performancesof others we mayadmitthe modes of action and passion also. For emotions such as pity and fear, or again enthusiasm, exist very strongly in some souls, and have more or less influence over all. Some persons fall into a religiousfrenzy, whom we see as a resultof the sacredmelodies when they have used the melodies that excite the soul to mysticfrenzy- restoredas though they had found healing and katharsis.Those who are influencedby pity or fear, and every emotional nature, must have a like experience, and others in so far as each is susceptible to such emotions, and all receive a sort of katharsisand are relieved with pleasure. The katharticmelodies likewisegive an innocentpleasureto men. Such are the modes and melodies in which those who perform music at the theatre should be invited to compete."I

It does seem that Aristotle distinguisheskatharticmelodies from those - and thus "ethicalmelodies"whichhelp to trainand reinforcecharacter that the point of katharsiscannot in any straightforward way be ethical education."2 But the only reasonfor thinkingthat katharsis is a cure for a is as exampleof katharsis pathologicalconditionis that Aristotle'sprimary a cure for religious ecstacy.3 However, even if we accept that religious is meantto apply ecstacyis a pathologicalcondition,the idea that katharsis to a pathologicalconditioncan only be sustainedby ignoringan important claim which Aristotle makes in the quoted text. Having begun his discussionof katharsis withthe exampleof those who are particularly susceptible to religiousfrenzy, Aristotle goes on to say that the same thingholds
Aristotk's Poetics Appendix 5, esp. pp. 354-356.) " Politics VIII.7, 1341b32-42a18. Here I have made a few changes in the revised Oxford translation:I use "ethical melodies" ratherthan "melodies of character"for "ta rather than translateit as "purgation";I ethika"; I use a transliterationof "katharsis" translate "pathos" as "emotion" rather than as "feeling"; and I translate "kouphidzesth/i meth' hedones"as "relieved with pleasure" rather than as "lightened and delighted". 12 Bemays makes this point. Halliwell interpretsthis passagesso as to diminishAristotle's apparent contrast between education and katharsis. For a criticism of this interpretation, see section 4, below. 13 See esp. PoliticsVIII.7, 1342a4-11.Bernaystakes religiousecstasyto be a pathological condition. 300

for anyonewho is influencedby pity and fear and, more generally,anyone who is emotionallyinfluencedby events.14In case thereshouldbe anydoubt that Aristotle means to include us all under that category he continues: "anda certainkatharsis andlightening withpleasureoccursforeveryone" .15 But everyoneincludesvirtuouspeople and it is absurdto supposethat, for Aristotle, virtuouspeople were in any kind of pathologicalcondition. Nor does the idea of a purgation seem like a plausibleanaloguefor tragic katharsis.In a medicalpurge, as the Aristotelianauthorof the Problems says, "drugsare not concocted - they make their way out carryingwith themanything whichgets in theirway:thisis calledpurging".16 The idea of a purgationseems to be that of the introductionof a foreign substance, a drug,whichlatergets expelledfromthe bodyuntransformed alongwiththe noxious substances.But the idea of a purgationas it is suggested by the commentators is of a homeopathic cure:we introducepity andfear in order to purgethe soul of these emotions. 1 The problemis thatthoughthe idea of a homeopathiccure was availablein Aristotle'stime, there is no evidence thathe wasawareof it andlots of evidencethathe thoughtthatmedicalcure was effectedby introducing contraries.18But once we abandonthe idea that for Aristotle a medicalpurgationwas a homeopathiccure, there seems to be little to recommendthe medical analogy. What foreign substance is introducedto expel what contrarynoxious substancein the soul? Why shouldone thinkthatthe virtuousmanhas anynoxiouselementsin his soul which need purging? Indeed, if we look to Aristotle'saccountof the emtions,they do not seem to be the sort of thingswhichare readilyconceivedas purgeable.Fear, for example, is defined as a pain or disturbancedue to imaginingsome destructiveor painfulevil in the future.19 That is, the emotion of fear is not
14

1342al1-13: tauto de touto anagkaionpaschein kai tous eleemonas kai tous phobetikous kai tous olos pathetikous. '5 "kaipasi gignesthaitina katharsinkai kouphidzesthaimeth'hedones" 1342al4-15; my translationand emphasis. (By the way, this statement seems to me to provide absolutely conclusive evidence againstHumphreyHouse's claim that, for Aristotle, a phronimosat the theatre would experience no katharsis. See his Aristotle's Poetics, op.cit., chapter VIII.) 16 Problems 42, 864a34. 17 See e.g. Franz Susemihl and R.D. Hicks, The Politics of Aristotle (London: Macmillan 1894, p. 651, n.1), who along with Humphrey House (op.cit. p. 110) quote Milton's preface to Samson Agonistes. Cp. Halliwell, op.cit., pp. 1924. 1 Halliwell is aware of this: op.cit., p. 193, n. 37. See Nicomachean Ethics 1104b17f, Eudemian Ethics 1220a36. 19 Rhetoric II.5, 1382a21ff.

301

exhaustedby the feeling one has when one feels fear. In additionto the the belief thatone is in dangerand feeling, the emotionof fearalso requires a state of mind which treats the danger as worthy of fear. All three conditionsare requiredto constitutethe emotionof fear.' If, for example, one believes one is in dangerbut one's state of mindis confidencein being An emotion,then, is not merely ableto overcomeit, one will not feel fear.2" a feeling, it is an orientationto the world. But if an emotion requiresnot merelya feeling, but also a belief aboutthe worldone is in and an attitude toward it, then it is hard to know what could be meant by purgingan an item for the emotion. An emotion is too complex and world-directed value. purgationmodel to be of significant 3. I do not wish to spend time on the idea that tragickatharsiseffects a since of the emotions,for thoughthisview has hadproponents purification the Renaissance,it is not seriouslyheld today.' The majorproblemswith the idea of purificationare, first, that virtuouspeople will experiencea certain katharsisin the theatre, but their emotional responsesare in no sense impure;second, it is not clear what is meant by purifyingthe emotions. One possibilitywas suggested by Eduard Muller: "Who can any longer doubt that the purificationof pity, fear and the other passions of consistsin, or at least is very closely connectedwith the transformation The fact thatwe do derive the pain that engenderedtheminto pleasure?"23 in a certainpleasurefromthe pitiableandfearfuleventsthat are portrayed tragedy is, I think, of the greatest importancein coming to understand tragickatharsis.However, it is a mistaketo think that, in tragedy,pain is into pleasure.Pityandfearare not abolishedby the tragedy;it transformed is just thatin additionto the pity andfearone feels in responseto the tragic events, one is also capableof experiencinga certainpleasure.Moreover, is to to conceiveof it as a purification even if there were a transformation, assume that the originalemotional responseof pity and fear is somehow clear polluted or unclean. But this isn't so. Aristotlemakes it abundantly responsesto a good tragicplot.2-The thatpity andfear are the appropriate
See Rhetoric l.1 and 11.5.In addition,Aristotlebelievesthere are certainphysiological changes which accompany an emotion. On the Soul, 403al6-19. 21 Rhetoric II.5, 1382b30ff. I See Hathaway op.cit. I Eduard Muller (Theorie der Kunst bei den Alten, Vol. 2, pp. 62, 377-88) quoted by Bemays, op.cit., p. 156. 4 See e.g. Poetics, 13-14where plots are evaluatedon the basisof the type of emotional response they tend to evoke in an audience.Those that do not producepity and fear, but, for example, disgust are rejected as inadequatefor tragedy.
I

302

pain of pity andfear is not an impurity whichneeds to be removed,it is the emotional responsewhicha virtuousman will and ought to feel. 4. Perhapsthe mostsophisticated view of katharsis, whichhasbeen powerfully argued for in recent years, is the idea that katharsisprovides an educationof the emotions.'5The centraltask of an ethical educationis to train youths to take pleasureand pain at the right sort of objects: to feel pleasurein actingnobly andpain at the prospectof actingignobly.26 This is accomplishedby a process of habituation:by repeatedly encouraging youths to perform noble acts they come to take pleasure in so acting. Virtue, for Aristotle, partiallyconsists in having the right emotional response to any given set of circumstances: feeling pain at painful circumstances,pleasureat pleasurable ones, and not feelingtoo muchor too little pain or pleasure, but the rightamount.27 Tragedy,it is argued,providesus with the appropriate objects towards whichto feel pity and fear. Tragedy,one mightsay, trainsus or habituates us in feeling pity and fear in response to events that are worthy of those emotions. Since our emotions are being evoked in the proper circumstances, they are also being educated, refined,or clarified.By being given repeated opportunitiesto feel pity and fear in the right sort of circumstances, we are less likely to experience such emotions inappropriately: namely, in response to circumstances which do not merit pity and fear. Since virtue partially consists in having the appropriateemotional responses to circumstances,tragedy can be considered part of an ethical education. There are two overwhelmingadvantagesto this interpretation which, I think, any adequate account of katharsisought to preserve. First, this interpretationrelies on a sophisticated,and genuinelyAristotelian, conception of the emotions. Tragedyprovides(a mimesisof) certainobjects towardwhich it is appropriate to form certainbeliefs and evaluativeattitudes as well as feel certain pains. Second, this interpretationoffers an accountof the peculiarpleasurewe derivefroma performance of tragedy.28
2 See Humphrey House, Aristotle's Poetics, Stephen Halliwell, Aristotle's Poetics, Leon Golden, "Catharsis"and MarthaNussbaum, The Fragilityof Goodness. Golden and Nussbaum speak of a "clarification"of the emotions.
2 2

Nicomachean Ethics II.

'

NicomacheanEthicsI1.6, 1106b6-28.This is Aristotle's famous doctrineof the mean. Aristotle is clear that one need not actually see a performanceon stage in order to experience the effect of tragedy;simply hearingit read out loud is sufficient. See Poetics 14, 1453b4-7;6, 1450b18-19,26, 1462all-12. For Aristotle's mention of the peculiarand appropriatepleasureof tragedy, see Poetics 14, 1453bl-14; 23, 1459al7-24;26, 1462bl2-

303

Aristotle,as is well known,believesin an innatedesireto understand, anda specialpleasureattendsthe satisfaction of that desire.29 If tragedyhelps to provide an ethical education, then in experiencingit we come better to understand the world, as fit object of our emotionalresponses,and better to understand ourselves,in particular, the emotionalresponsesof whichwe are capableandwhichthe events portrayed require.It is becausewe gaina deeper insightinto the humanconditionthat we derive a specialcognitive pleasurefrom tragedy. This interpretation does have a genuinelyAristotelianring to it: it is a position that is consonantwith much that Aristotle believed and it is a positionhe mighthave adopted.But I don'tthinkhe did. First,as we have seen, a virtuouspersonwill experiencea certainkatharsis when he sees or 30 Second,the hearsa tragedyperformed,but he is in no need of education Politics'discussionof musicclearlydistinguishes musicwhichis educative of the emotions and should be employed in ethical trainingfrom music which produces katharsis.31 The best attempt I have seen to meet this whichAristotleis contrastproblemis by arguingthatthe type of katharsis ing with ethical educationis only an extremeform derivedfrom orgiastic
music:
"Once attention is shifted to types of katharsis connected with more common emotions and with those who do not experience them to a morbidly abnormal degree (and both these conditions are true of the tragic variety), it is possible to discern that katharsismay after all be in some cases compatible with the process which Aristotle characterizesin Politics 8 as a matter of habituationin feeling the emotions in the rightway and towardsthe rightobjects (1340al6-18) . .. Simplyto identify tragic katharsiswith a process of ethical exercise and habituationfor the emotions through art would be speculative and more than the evidence justifies. But to suggest that these two things ought to stand in an intelligible relationto one another (as the phrase 'for education and katharsis'at Pol. 1341b38encouragesto see them), is only to arguethat tragickatharsisshouldbe capableof integrationinto Aristotle's general philosophy of the emotions, and of their cognitive and moral importance, as well as into the frameworkof his theory of tragedyas a whole."32

Of course, tragickatharsis and ethicaleducationmightstandin an "intelligible" relation to each other even if they served completely different purposes, but when one sees the phrase 'for education and katharsis' quotedout of context,it is temptingto supposethateducationandkatharsis
14, cf. 1462al5-17. MetaphysicsI.1. I Nor, contra Golden and Nussbaum, do his emotions need to be clarified. 31 Politics VIII.7, 1341b32-1342a18 (quoted above). 3 Halliwell op.cit., pp. 195-6.
29

304

are part of a single project. Unfortunately, the text will not support this supposition. Aristotle explicitly says that although one should use all the different types of melodies, one should not use them for the same function.33And when he says that music may be used "for the sake of education and of katharsis",34he is unambiguously listing different benefits that may be derived from music.35Nor is it true that, in this passage, Aristotle is only contrasting education with an extreme orgiastic form of katharsis. For although, as we have seen, he begins by talking about the katharsis of religious frenzy, he very quickly goes on to mention a certain katharsis had by everyone, and the fact that two lines before he explicitly mentions those who are susceptible to pity and fear suggests that he had tragic katharsis in mind.36Thus the contrast which Aristotle draws between ethical education and katharsis cannot easily be brushed aside. Moreover, Aristotle continues by saying that vulgar audiences will have vulgar tastes and that professional musicians ought to cater to those tastes, since even vulgar people need relaxation.37But if even some melodies are ethically educative, why doesn't Aristotle insist that the vulgar be confined to such uplifting tunes? The answer, I think, is that it's too late. Aristotle contrasts two types of audience: the vulgar crowd composed of artisans and laborers on the one hand, and those who are free and have already been educated on the other.38 In each case the characters of the audience have been formed and ethical education would be either futile or superfluous. Aristotle clearly thinks that tragedy is among the highest of art forms. Aside from the fact that tragedy is the culmination of a teleological development of art forms which began with dithyrambs and phallic songs,39and aside from the fact that Aristotle explicitly holds it in higher regard than epic, notwithstanding his enormous respect for Homer, Aristotle criticises certain forms of inferior plots as due to the demands of a vulgar audience.

1342al-2.

My translationof 1341b38. 3S This is made clear by 1341b36-38:... ou mias heneken opheleias tei mousikei dein alla kaipleionon charin(kaigarpaideiashenekenkai katharseos... .). But in case there is anydoubt, it is settled by "triton" at 1341b40: clearly,education,katharsis,andintellectual enjoyment are being listed as three distinct benefits obtainable from music. 36 1342bl1-15. 37 1342bl8-29. This passage is also cited by Bernays as part of his argument that katharsisis not meant by Aristotle to be morallyeducative. 38 ho men eleutheroskaipepaideumenos(1342bl9). Cp. also Poetics 26 (esp. 1461b2728) which suggests that tragedywill be appreciatedby a better sort of audience. 39 Poetics 4, 1449alO-15.

305

For example,Aristotlecriticizesthose allegedlytragicplotswhichend with the good being rewardedand the bad beingpunished.
"It is ranked first only though the weakness of the audiences; the poets merely follow their public, writingas its wishes dictate. But the pleasurehere is not that of tragedy . . s

This would suggest that a proper tragicplot would be appreciatedand enjoyedabove all by a cultivated person.It is hardto escapethe conclusion is for educated, that, for Aristotle, educationis for youths,tragickatharsis cultivatedadults. of katharsisought to The thirdreasonwhy the education-interpretation be rejected is that there is a fundamentalsense in which tragedyis not Shouldwe be spectators evokingthe properresponsesto eventsportrayed. to tragicevents whichoccurnot in the theatrebut in real life to those who are close to us, or to those who are like us, the properemotionalresponse wouldbe (the rightamountof) pity andfear. To take any kindof pleasure from these events would be a thoroughlyinappropriate response. Thus for the emotional thereis a sense in whichtragedyprovidesa poor training responsesof real life: first, we shouldnot be trainedto seek out tragedyin reallife, as we do seek it in the theatre;second,we shouldnot be trainedto find any pleasurein real life tragicevents, as we do find pleasure in the of the poets. Althougha mimesisof pitiable and fearful tragicportrayals to a must certain extent be like the real life events which they events respect represent,the mimesismust, for Aristotle,also be in an important unlike those same events. For it is precisely because the mimesis is a mimesis that a certaintype of pleasureis an appropriate response to it. Were it not for the fact that Aristotle recognizeda salient difference betweenmimesisand the real life events it portrays,Aristotlewould have had to agreewith Plato that poetryshouldbe bannedfromthe ideal state. AristotledisagreeswithPlato not over whethertragedycan be used as part of an ethical educationin the appropriate emotionalresponses, but over whethera mimesisis easilyconfusedwiththe realthing.Aristotle'spoint is that althoughthe properemotionalresponseto a mimesiswould be inapunlike the real event propriateto the real event, a mimesisis sufficiently that there is no dangerof it havingan impropereducationaleffect on the audience. From the point of view of ethical education alone, poetry is allowedinto the republicnot becauseit has anypositiveeducationalvalue, but becauseit can be shown to lack any detrimental effects. If poetry has positivevalue, it mustlie outside the realmof ethicaleducation.
4

Poetics 13, 1453a33-36.

306

"Thereis not the samekindof correctness in poetry,"Aristotlesays, "as in politics, or indeed any other art."' The constraintson the poet differ on the politician.The politicianis confrom the constraints considerably strainedto legislate an educationin which youths will be trainedto react appropriately to real life events; in particular, to feel the rightamountof pity and fear in response to genuinelypitiable and fearful events. The tragedianis constrainedto evoke pity and fear througha mimesisof such events, but he is also constrainedto provide a katharsisof those very emotions. It is in the katharsisof those emotions that the emotional response appropriate to poetrygoes beyondthat whichis appropriate to the reallife events.Thusin comingto understand corresponding whatkatharsis is, we will be approachingan understanding of the special contribution poetry makes to life. The final reasonwhy the educationinterpretation of katharsis ought to be rejected is that in the end it does not explainthe peculiarpleasureof tragedy.42 Of course, a properappreciation of tragedydoes requirea finely tuned cognitive appreciationof the structureof the plot and there is no doubt that the exercise of one's cognitivefacultiesin the appreciationof tragedydoes afford a certainpleasure. But the pleasurewe derive from tragedy is not primarilythat which comes from satisfyingthe desire to understand. In fact, there is little textualsupportin the Poeticsfor the hypothesisthat the peculiarpleasureof tragedyis a cognitivepleasure.The mainsupport comes from Poetics4, whereAristotleexplainsthe originsof poetry:
It is clear that the general originof poetry was due to two causes, each of them part of human nature. Imitation [mimesis]is naturalto man from childhood, one of his advantagesover the lower animalsbeing this, that he is the most imitativecreature in the world, and learnsatfirst by imitation.And it is also naturalforall to delightin works of imitation. The truth of this second point is shown by experience; though the objects themselves may be painfulto see, we delightto view the most realistic representationsof them in art, the forms for example of the lowest animalsand of dead bodies. The explanation is to be found in a further fact: to be learning somethingis thegreatestof pleasuresnot only to thephilosopher, but also to the rest of mankind, howeversmall theircapacity for it; thereason of the delightin seeing the picture is that one is at the same time learningand reasoning[sullogidzesthai]what each thing is, e.g. that this is that; for if one has not seen the thing before, one's pleasure will not be in the picture as an imitation of it, but will be due to the execution or coloring or some similarcause."43 Poetics 25, 1460b13-15. Here I am particularlyindebted to Giovanni Ferrari. 43 Poetics,4,1448b4-19,my emphasis. I have altered the revised Oxford translationof 1448bl4-15: sullogidzesthai ti hekaston, hoion hoti houtos ekeinos which is rendered
42

41

307

It is important to note that Aristotleis here concernedwith the originsof a processwhich culminatesin the developmentof tragedy. Childrenbegin learning by theirearlyimitationsof the adultsaroundthem, andin learning form of cognitivepleasure:but this is only an they derive a rudimentary explanationof how elementaryforms of imitationnaturallyarise among humans.It is not an explanationof the peculiarpleasureof tragedy. One must also be cautious in interpretingAristotle's claim about the pleasurein learning.Aristotle is tryingto explainwhy we take pleasurein viewingimitationsof objects that are themselvespainfulto look at. Now it is temptingto assimilatethis passage with Aristotle's admonitionin the Partsof Animals that one should not shy away "with childish aversion" from studyingblood and guts and even the humblestof animals:for the studyof even the lowest of animalsyields a pleasurewhich derives from discoveringthe intelligible causes of its functioningand the absence of chance."For Aristotlethere contraststhe cognitivepleasurederivedfrom causesfromthe pleasurederivedfrom an imitation: comingto understand
For even if some [animals] are not pleasing to the sense of sight, nevertheless, creating nature provides extraordinarypleasures for those who are capable of understandingcauses and who are by nature philosophical. Indeed, it would be unreasonable and strange if mimetic representations of them were attractive, because they disclose the mimetic skill of the painteror sculptor, and the original realities themselves were not more interesting, to all at any rate who have eyes to discernthe reasons that determined their formation.45

Aristotleis sayingthat there are two distinctpleasuresto be derivedfrom animals thatare in themselvesunpleasant to look at: a cognitivepleasurein their causes, and a 'mimeticpleasure'in appreciatingan understanding artist'sskill in accuratelyportrayingthese ugly creatures. It is this distinctively'mimeticpleasure'that Aristotleis concentrating on in Poetics4. Thereasonwhyhe focuseson the artistic representation of an uglyanimalis that he wants to be sure he is isolating the pleasure derived from the mimesis,ratherthanthe pleasureone mightderivefrom the beautyof the animalitself. In explainingthis 'mimeticpleasure',Aristotledoes alludeto the pleasurederived from learning.But that Aristotle has only the most rudimentary form of 'learning'in mindis made clear by his claim that this in pleasure learningis availablenot onlyto the philosophically minded,but
there as "gatheringthe meaning of things, e.g. that the man there is so-and-so". My translationis more literalwhich I think is importantto the interpretationof this passage. 44 Partsof Animals 1.5, 645a4-37. 45 Partsof Animals 1.5, 645a8-15.

308

to all of mankind however small theircapacityfor it. What one is 'learning' is

that this is that: i.e. that this (pictureof a dead mouse) is (an accurate representationof) that ([a] dead mouse). The 'reasoning'one is doing is confined to realizingthat one thing (an artisticrepresentation)is an instance of another. The pleasure, Aristotle says, is precisely that which wouldbe unavailable to someoneincapableof formulating this elementary realization:that is, to someone who had never seen a mouse.6 Such a person would not be able to recognizerepresentation as a representation, and thus his pleasure would be confined to appreciatingthe colors and shapes in the painting. Thus it is a mistake to interpretthis passage as suggestingthat the reasoningis in any sense a reasoning about causes. Poetics 4, then, is about the most elementary pleasures which can be derivedfromthe most elementaryof mimeseis.Althoughthis is a firststep towardsan understanding of tragicpleasure,it does not lend supportto the thesis that tragicpleasureis a speciesof cognitivepleasure. Now Aristotle does repeatedlyinsist that a good tragedymust have an intelligibleplot structure. There mustbe a reasonwhy the tragedyoccurs: thusAristotlesaysthatthe eventsmustoccurplausiblyor necessarily,47 that the events must occur on account of one another rather than in mere temporalsuccession,48andthatthe protagonist mustmakea certainmistake or error(hamartia) whichis responsible for andexplainshis downfall.49 And I thinkthere is no doubtthatthe propereffect of tragedyon an audienceis
4 Poetics 4, 1448bl7-19. Such a person, presumably,would not have heard a sufficient descriptionto recognize a mouse: the person Aristotle has in mind, I think, is someone who has no idea of a mouse: so he is in no position to recognize of any paintingthat it is a painting of a mouse. See e.g. Poetics 9, 1451a37-38;10, 1452al7-21; 15, 1454a33-36;16, 1455al6-19; 25, 4 1461bll-12. 4 E.g. Poetics 9, 1452a3-4;10, 1452a20-21. 49 Poetics 13, 1453a8-30.Nussbaumarguesthat the point of a hamartiais to render the protagonist sufficiently like us that we can identify with him to the extent required to experiencethe tragicemotionsof pityandfear (op.cit., pp. 382ff.). Her reasoningis based on her more generalinterpretation that, for Aristotle, the point of tragedyis to explore the gap which inevitably exists between being good and living well. I do not think that the generalinterpretationcanbe correct.AlthoughAristotle does acceptthatbeingvirtuousis not sufficientfor happinessand that externalmisfortunecan ruina thoroughlygood man (NicomacheanEthics1.10), it is quite clearthatAristotle does not thinkthatsuchan event could be the basis for a tragedy. Consider for example Poetics 13, 1452b30-36,where Aristotle says thattragedycannotportraythe fall of a good manfromgood to badfortune, for such an event does not arouse the tragicemotions of pity and fear but a thoroughly non-tragicemotion of disgust.Aristotle does reluctantlyadmitthat a virtuousman can be destroyedforno reasonatall, thatis, throughmisfortune,but he denies thatthisis the stuff of tragedy. Tragic events always occur for a reason.

309

broughtaboutvia the audience'scognitiveappreciation of the intelligible plot structure. The question,then, is notwhetheran audiencemustexercise its cognitivefaculties, nor whether it may find pleasure in so doing; the questionis whetherthis cognitiveexerciseand its attendantpleasureis the propereffect of tragedy.Is this cognitivepleasurethe pleasureappropriate and peculiarto tragedy?To see that the answeris "no", considerone of Aristotle'sclassicstatementsof the demandfor intelligibility:
"Tragedyis a mimesisnot only of a complete action, but also of fearful and pitiable events. But such events occur in the strongestform when they occur unexpectedly but in consequence of one another. For the events are more marvellous (thaumaston) when they occur thus than if they occur by chance . . "50

Aristotle'spoint is that a plot structurein whichthe events do not merely succeedeach other in time, but standin the relationof intelligiblecause to intelligibleeffect, albeit a relationin whichthe intelligibility only comes to lightwitha reversalandrecognition,is the best plot structure for portraying trulypitiableandfearfulevents. Whatit is to be a pitiableandfearfulevent is to be an event capableof inducingpity andfear in the audience.But pity and fear is clearlynot the propereffect of tragedy:it is merelya necessary step along the route towardsthe propereffect. For Aristotlesays that it is from pity and fear that tragedyproducesa katharsisof these emotions.5 Therefore, the audience'scognitive appreciationof the plot's intelligible structureand attendantpleasure are important,but they are causal antecedentsof the proper effect and proper pleasure of tragedy. when Aristotle does say that events are more marvellous(thaumaston) they occurunexpectedlybut in an intelligiblerelationto each other. And this fact is invoked by those who wish to argue that tragic pleasure is a cognitivepleasure.For in the Metaphysics andRhetoric,Aristotlelinksthe wondrousor marvellouswith our desire to understand.52 It is owing to wonder,Aristotlesays, thanmanfirstbeganto philosophize: the risingand setting of the sun, for example, provokesman'swonderand this wonder sets him on a journeyto explainwhy this phenomenonoccurs.53 Thus it is suggested that the wonder that is produced in a tragedy provokes the
Poetics 9, 1452a1-6[my trans. except for two phrases from Oxford]. Poetics 6, 1449b27-28.Literally, Aristotle says a "katharsisof such emotions" (ton toiouton pathematon), but Bernays has argued convincingly that "such" should be understood demonstratively,as referringexclusively to pity and fear. 52 Metaphysics982bl2ff, 983al2ff; Rhetoric 1371a31ff. 53 Metaphysics982bl2ff. IdiscussthisatsomelengthinAristotle: TheDesireto Understand (Cambridge, 1988).
50 51

310

audience to try to understand the events portrayedand the pleasurethat attendscoming to understand is tragicpleasure.54 If there were alreadya strongcase for thinkingthat tragicpleasurewas cognitive pleasure, then the link between the marvellousand tragedy,on the one hand, and with the desire to understand,on the other, would be suggestive. However, in the absence of a strong case, there are three reasonswhy Aristotle'sremarkson the marvellouscannotbe used to lend any significantsupportto the idea thattragicpleasureis cognitive.First,in the Poetics passagejust quoted Aristotle seems to be suggestingthat the relationbetweenwonderandunderstanding is preciselythe oppositeof that suggestedby the Metaphysics: it is by cognitivelygrasping that the events, though unexpected, are intelligiblylinked to one anotherthat wonderis producedin us. So while in the Metaphysics wonderprovokesus to understand, in the Poetics understanding provokes us to experiencewonder. Second, althoughin the quoted passageAristotle associatesintelligibility with wonder, towards the end of the Poetics Aristotle also associates wonder with irrationality.55 One advantageof epic over tragedy,he says there, is that it is better suited to portraying irrationalevents (to alogon). For since the audienceof an epic narrative does not actuallyhaveto see the irrationalityacted out on stage, it is less likely to notice it as irrational. However, Aristotle says, it is the irrational whichchieflyproduceswonder (to thaumaston).And he says that the experience of wonderitselfis pleasant.56 So in thiscase it cannotbe thatwonderprovokesunderstanding which is pleasant- for irrationality ultimatelyresistsunderstanding. And at the end of the Poetics, Aristotle suggeststhat the pleasureproperto epic and the pleasureproperto tragedyareof the sametype,57 even thoughtragedy is a higherform of the art. Yet if the pleasureproperto epic can be derived from a plot containingirrationalities, it hardlyseems thatthispleasurecan be cognitive. Finally,even if one grantsa-link betweenwonderand cognitive pleasure, this in itself does nothing to supportthe thesis that tragic pleasure is cognitive. For an anti-cognitivist like myself does not believe thatthere is no role for cognitionandits attendant pleasurein the appreciation of a tragedy;he only denies that cognitivepleasureis to be identified with tragic pleasure. For the anti-cognitivist, cognitivepleasureis a step that occursen route to the productionof the properpleasureof tragedy.
54
5 56 57

See e.g. Halliwell, op.cit., pp. 70-74.

Poetics 24, 1460a11-17.

Poetics 24, 1460a17. Poetics 26, 1462bl3-14. See the note on the passage in D.W. Lucas, Aristotle's Poetics (Oxford: Clarendon Press, 1968), p. 257.

311

The finaltext whichis cited in supportof the cognitivistthesis is Aristotle's claim that poetry is "morephilosophical" than history:
Poetry is more philosophicaland more serious than history:for poetry speaks more about universals, while history speaks of particulars.By universal is meant what sort of thingsuch a sort of person would plausiblyor necessarilysay or do - which is the aim of poetry though it affixes propernames to characters;by a particular,what Alcibiades did or had done to him.58

Of course, philosophyis an exercise of man'scognitivefacultiesand, as is well known, Aristotle repeatedly insists that it is universalswhich man understands.59However, even if we interpretthis passagejust as cognitivists would like us to - as suggesting an intimate link between the apprecia-

tion of tragedyand the exercise of our cognitiveabilities- nothingin this passagewould help us decide between the cognitivistand the anti-cognitivist theses. For, as we have seen, the anti-cognitivist does not deny that a cognitiveunderstanding of the plot is essentialto the properappreciation of a tragedy, he only denies that tragic pleasure can be identifiedwith the pleasurethat attendsunderstanding. But, more importantly,I don't thinkwe shouldinterpretthis passageas the cognitivistswouldlike us to. Thereis a certainplasticityin the idea of a universal whichfacilitatesthe transition frompoetryto cognition.The true objects of knowledge, for Aristotle, are essences and these essences are 'universal' in the sense that two healthyhumanbeingswill instantiatethe same essence: human soul. But the reason that essences are linked with knowledgeis that in coming to understanda thing'sessence we come to human understand what that thing is reallylike. In comingto understand essence, we come to understandwhatit is to be a humanbeing. Now when than historybecauseit Aristotle says that poetry is "more philosophical" deals with universals, it is tempting to read him as saying that poetry providesus withdeeperinsightsinto the humancondition.Thisis a temptation which ought to be resisted.' If we look to what Aristotle means by "universal" in the passageunderdiscussion,it is clearthathe does not mean 'universal whichexpressesthe essence of the humancondition',but somePoetics 9, 1451b5-11. At Metaphysics XIII. 10, 1087alO-25,Aristotle does qualifyhis claimthat episteme is of universals. See my "Active Episteme" (in Mathematicsand Metaphysicsin Aristotle: Proceedingsof the Xth SymposiumAristotelicum(A. Graeser ed., Bern, 1986) ) for an analysis of this passage. 60 Although I am certainlywilling to accept that Aristotle thoughtthat tragedyprovides deeper insight into the human condition than history does, I don't think that is the immediate point he is making in the passage under discussion.
59
58

312

thing much less grandiose:that poetry should refrainfrom describingthe particular eventsof particular people andinsteadportraythe sortsof things a giventype of personmightsayor do. Aristotlegivesan exampleof whathe means by the universalelement in poetrylater on:
The following will show how the universalelement in Iphigenia, for instance, may be viewed: a certainmaidenhavingbeen offered in sacrifice, and spiritedawayfrom her sacrificersinto another land, where the custom was to sacrificeall strangersto the Goddess, she was made there the priestess of this rite. Long after that the brother of the priestess happened to come; the fact, however, of the oracle having bidden him go there, and his object in going, are outside the plot of the play. On his coming he was arrested, and about to be sacrificed,when he revealed who he was either as Euripides puts it or (as suggested by Polyidus) by the not improbable exclamation, 'So I too am doomed to be sacrificed as my sister was'; and the disclosureled to his salvation.This done, the next thing, aftertheproper nameshave beenfixed as a basis for the story, is to turn to the episodes.6'

Aristotle'spointis simplythatpoetrydealswithtypesof actionsandtype of persons,even thoughthe poet, afterhavingconstructed the 'universal' plot
later assigns names to the characters.62 Aristotle does say that such a

universalplot is "morephilosophical" than history,but by this he did not mean that poetrygives us ultimateunderstanding of humanity.Rather,he meantthat it has emergedfromthe mireof particularity in whichhistoryis trapped and thus has taken a step along the way towards philosophy. Whetherfairlyor unfairly,Aristotle had a very low opinionof history(he seemed to hold historyin the sameregardas we hold newspapers) andthus somethingdoesn't have to be very philosophical to be more philosophical than history.63 What then is the point of Aristotle'srequirement that poetry deal with universalsif it is not to insistupon poetry'sultimatecognitivevalue? If we read Poetics9 throughto the end it becomes clear that Aristotle'soverall concernis with the formationof a plot that effectivelyproducespity and fearin the audience.' But in orderfor an audienceto feel pity andfearthey must believe that there is a certainsimilaritybetween themselvesand the character in the tragedy:andthe reasonthey mustbelieve in this similarity
6' Poetics 17, 1455b2-13(Oxford trans.). See also Aristotle's descriptionof the plot of the Odyssey at 1455b16-23. `2 Poetics 17, 1455b, 12-13; cf. 9, 1451b8-16. 3 Aristotle does not seem to have been familiar with Thucydides. One cannot but wonder how Aristotle would have changedhis mind about historyif he had carefullyread the History of the Peloponnesian War. ' As we have seen, that is why Aristotle says at the end of Chapter9 that the events in a tragedy should occur unexpectedly but on account of one another.

313

is that they must believe that the events portrayedin the tragedymight happento them. For a personto feel pity and fear he mustbelieve that he himself is vulnerableto the events he is witnessing.That is why Aristotle eventsthathavehappened,but saysthatthe poet'sfunctionis not to portray events that might happen - and that these possible occurrencesseem plausibleevents that The pointof portraying plausibleor even necessary.65 come to believe that these mighthappenis that the audiencewill naturally of events mighthappento them.And thisis a crucialstep in the production pity and fear in their souls. Poetry uses universalsfor the same purpose. butconcernsitselfwithtypesof Becausepoetryis not miredin particularity, events whichoccurto certainsortsof people, it is possiblefor the audience to appreciatethat they are the sort of people to whom this sort of event Aristotlehas in mindwhen he could, just possibly,occur.The universality of poetryis not as such aimingat the depth of talks about the universality the human condition, it is aiming at the universalityof the human condition.' Enough has been said, I think, to make it clear that the educationof what it is, mustbe rejectedasanaccount howeverattractive interpretation, But havingalreadyrejectedthe purgaAristotlemeantby tragickatharsis. allthe important we haveabandoned tion-andpurification-interpretations, It is accounts.What,then, didAristotlemeanbytragickatharsis? traditional to this question that I now turn. II 5. Althoughthe workso far has been largelycritical,I thinksomethingof fall positivevalue has emerged.For in seeing how previousinterpretations short, we have isolated a series of constraintswhich any acceptableinterpretationof katharsismust satisfy. These constraintsmay not be so butthey at least as to isolatea single,definitiveinterpretation, constraining set out a field in whichthe truthmustlie. In thissectionI wouldlike to state of katharsisand I would the constraintson any acceptableinterpretation whichfits those constraints. like to offer an interpretation is: on any interpretation One of the majorconstraints (1) There is reasonfor a virtuousmanto experiencethe performance of pity and fear.67 of a tragedy:he too will experiencea katharsis
6

Poetics 9, 1451a36-38,repeated again at 1451b4-5,just before Aristotle claims that poetry is more philosophicalthan history because it deals with universals(1451b5-7). I Among humans, that is.

314

Preciselybecause of (1), it follows that (2) Tragickatharsis cannotbe a processthatis essentiallyandcrucially corrective:that is, it cannotbe a purgation,insofaras purgationis of of some somethingpathologicalor noxious;it cannotbe a purification pollution;it cannotbe an educationof the emotions. Thisis not to deny that a cathartic experiencemaybe corrective.Aristotle, melodiescanhelp to restore as we have alreadyseen, thoughtthatcathartic those who are particularly susceptibleto religiousfrenzy; and one might could restorethose who are particsimilarlysupposethat a tragiccatharsis ularlysusceptibleto the tragicemotionsof pity andfear. Nor do I mean to experiencedenythata virtuouspersonmayexperiencereliefin a cathartic a relief that it is naturalto conceive of in terms of the release of pent-up emotions.However,the virtuousmanis not in a pathological condition,nor is he pollutedwithsome impureelementwhichneedsto be removed.Nor is he in need of anyfurthertraining of the emotions:indeed, it is becausehe is alreadydisposedto respondappropriately to the situationsof life, both in judgement,actionandemotion, that he is virtuous.The idea that provides an educationof the emotionssuffersfurtherfrom the fact: (3) What one feels at the performanceof a tragedyis not what one wouldor should feel in the real life counterpart. For althoughtragedyprovokespity and fear in the audience,it also elicits an appropriate pleasure:this pleasurewould be thoroughlyinappropriate to real life tragicsituations.But the fact that a good person (at least) feels pleasurein the performance of a tragedy,but would not do so in real life, suggests (4) A properaudiencedoes not lose sightof the fact thatit is enjoying of a tragedy. the performance in the Althoughthe audiencemay identifyemotionallywith the characters tragedy,this identificationmust remainpartial.Throughoutits emotional involvement,the audiencekeeps trackof the factthatit is an audience.For in a real life tragedya person would feel fear and, if he stood in the right relationto the tragicevent, pity, but he would deriveno pleasurefromthe tragicevent. This implies: (5) The mere expressionor releaseof emotionsis not in itself pleasurable.
67

See Politics VIII.7, 1342b14;and the numerousreferencesin the Poetics in which the plot of a good tragedyis distinguishedfrom that which will appeal to a vulgar audience: e.g. Poetics 13, 1453a30-36(cp. 9, 1451b33-1452aland 6, 1450b16-19)and Poetics 26, in which Aristotle seems to accept the principlethat tragedyis a higher art form than epic precisely because it appeals to a better audience.

315

pains.' The mereopportuniForAristotle, pityandfearare unadulterated ty to feel these painfulemotionsdoes not in itselfproviderelief:everything depends on the conditionsin which these painfulemotionsare to be felt. Those who have assumedthat a katharsis,for Aristotle, was a release or dischargeof pent-upor unexpressed emotionshave assumedthatthe mere experienceof emotions, even painfulones, has a pleasurableaspectto it. There is pleasureto be had in a good cry. Suchan idea may have a certain plausibility to it, but it is foreignto Aristotle. For him, it dependson what one is cryingabout. If one is cryingin the theatre, a certainpleasuremay ensue, but there is, for Aristotle, no pleasureto be had in cryingover real to be the mere life tragicevents. This is the problemwith takingkatharsis aboutexperireleaseof emotion. For Aristotlethereis nothingpleasurable encingpity and fear per se. These conditionsunderwhichwe can derivepleasurefrompity and fear and the conditions under which a katharsisof pity and fear occurs are intimatelylinked, for or it helps (6) Katharsis providesa relief:it is eitheritself pleasurable to explainthe properpleasurethat is derivedfrom tragedy.!' Constraints(3)-(6) together suggest that if we are to understandtragic katharsis,we shouldlook to the specialwaysin whichtragedyproducesits emotionaleffects. Aristotle, as we have seen, definestragedyin partby the effect it has on its audience:it is a mimesisof an action which by arousingpity and fear It mightseem odd to a modern producesa katharsisof those emotions.70 readerto see Aristotledefinetragedyin termsof its effect, for in a modem climatewe tend to thinkthat a work of art shouldbe definablein its own of whatevereffectit mighthaveon its audience.But it terms,independently to insistthatAristotlecouldnot have been defining wouldbe anachronistic tragedyin termsof its effecton the audience.Poetry(poiesis),forAristotle, is a type of making(poiesis), and the activityof any makingoccursin the For example,the person or thing towardswhichthe makingis directed.7" activityof the teacherteachingis occurring,not in the teacher,but in the studentswho are learning;the activityof the builderbuildingis occurring, not in the builder,but in the house beingbuilt. It standsto reasonthat, for
See RhetoricII.5, 8; cp. the account of anger as a composite of pain and pleasure: Rhetoric 11.2. 69 Aristotle, as we have seen, says that everyone undergoes a "certain katharsisand lightening with pleasure": Politics VIII.7, 1342b14-15. Poetics 6, 1449b24-28;see p. 297 above. 7 71 Physics III.3.
I

316

in an Aristotle,the activityof the poet creatinghis tragedyoccursultimately of the play.72 audienceactivelyappreciating a performance Not only does Aristotledefinetragedyin termsof its effect, he thinksthat various tragic plots can be evaluated in terms of their effects on an audience.
"We assume that, for the finest form of tragedy, the plot must be not simple but complex; and further, that it must imitate actions arousingfear and pity, since that is the distinctivefunction of this kind of imitation. It follows, therefore, that there are three forms of plot to be avoided. A good man must not be seen passing from good fortune to bad, or a bad man from bad fortune to good. The first situation is or piteous, but simplydisgusting.The second is the most untragic not fear-inspiring that can be; it has no one of the requisitesof tragedy;it does not appeal either to the human feeling in us, or to our pity, or to our fears. Nor, on the other hand, should an extremely bad man be seen fallingfrom good fortune into bad. Such a story may arouse the human feeling in us, but it will not move us to either pity or fear; pity is occasioned by undeservedmisfortune,and fear by that of one like ourselves;so that
there will be nothing either piteous or fear-inspiring in the situation."73

The important pointto note aboutthispassageis thatAristotleis evaluating plots not on the basis of feelings, but on the basis of the emotions. The reasonwe do not feel pity andfear in witnessingthe fall of a bad manfrom good to badfortune,is becausepityrequires the beliefthatthe misfortune is undeserved, fear requiresthe belief that the man who has suffered the
misfortune is like ourselves.74 (Presumably Aristotle assumed that the

properaudienceof tragedywouldnot believe themselvesto be sufficiently like a bad personto believe that the thingsthat befallhim (most likely as a consequenceof his badness)mightbefall them.) Similarlywith the disgustwe feel when watchinga good man fall from good to badfortune:suchdisgustisn'ta purefeelingwhichcanbe identified
on the basis of its phenomenological properties alone. Disgust requires the belief that there is no reason at all for this good man's fall. It is sometimes thought that Aristotle contradicts himself for he elsewhere seems to suggest that tragedy is paradigmatically about admirable men falling to bad forn

I say "ultimately"because there is a two step processinvolved: (1) the poet's creating the muthosand writingthe play, (2) the performanceof the play before an audience. I am using the word "performance" widely to cover both the enactmentof the play on stage by actors and the simple reading or recital of the play out loud. Aristotle is explicit that a tragedy can have its proper effect even when it is not acted out on stage: a person who merely hears the tragedy read out loud will experience pity and fear. See Poetics 14, 1453b3-7;6, 1450b18-19;26, 62all-12, a17-18. 7 Poetics 13, 1452b30-1453a8(Oxford trans. except that I use "disgusting"for "miaron" rather than Oxford's "odious").
4

Poetics 13, 1453a4-6.

317

But if we take the rest of Chapter13 as explicatingwhat Aristotle tune.75 means when he denies that the fall of a good man can be the basis of a properlytragic plot, I think we can see a consistentpoint emerging. In mustmakesome mistakeor tragedy,Aristotleinsists,the centralcharacter The hamartiais a mistakethat which leads to his fall.76 error (hamartia) rationalizesthe fall. So what Aristotle is excludingwhen he prohibitsthe fall:one thatoccursthroughno fault fall of a good manis a totallyirrational of the good man at all. Aristotlecertainlydoes allowthe fall of a good man to be the subjectmatterof tragedy:but not of a manwho is so good that he has made no mistakes which would rationalizehis fall. This distinction illuminateswhat is meantby disgust:disgustis an emotionthat is partially constitutedby the belief that there is no reason at all for the misfortune. Disgust is somethingwe feel in response to what we take to be a total absenceof rationality. Aristotlethinksthatthe merefact thattragedymustarousepityandfear the rangeof tragicplots. in the audiencejustifieshimin severelyrestricting
"It is not necessary to search for every pleasure from tragedy, but only the appropriatepleasure. But since it is necessaryfor the poet to producethe pleasure from pity and fear througha mimesis, it is evident that he must do this in the events in the plot. We should investigate, then, what sorts of events appearto be horrible or pitiable. In respect to such actions, it is necessary that the people involved be either friends with each other or enemies or neither of these. But if enemy acts on enemy, there is nothing pitiable about this- neitherin the doing of the deed, nor in intendingto do it - except in relationto the terribleevent itself (kat'auto topathos). The same is true when the people stand in neither relation. But whenever the terrible events occur among loved ones [friends, kin], for example if a brother should kill or intend to kill or do some other such thing to a brother, or a son to a father, or a mother to a son, or a son to a mother: we should search for these things."'n

Aristotle is clear that the peculiar pleasure of tragedy is produced by evoking pity and fear in the audience and that this is accomplished by constructing a mimesis of a special type of terrible event (pathos).Aristotle

both to signifya terribleevent, catastrophe uses the same word, "pathos", andto signifyemotion.When,for example,Aristotle or seriousmisfortune cites pathos as one of the three ingredients needed in a plot, along with reversal and recognition, in order to produce pity and fear, he is not requiringa certainmotion to be portrayedon stage, he is requiringthat act.78 So one mightsay that, for Aristotle,thereis an there be a destructive
"
76 "

See e.g. Poetics 15, 1454b8-13. Poetics 13, 1453a8-17. Poetics 14, 1453bl-22 (my trans.).

318

objectivepathosanda subjectivepathos:andthe two are related.Forwhat Aristotle is trying to do in this passage is delimit the precise type of objective pathos which is adequate to bring about a particulartype of subjective pathos - pity and fear - in response.79 The objectivepathosrequiredto producethe tragicemotionsis a terrible deed done betweenkinor loved ones. Thatis whythe greattragedians have correctlyfocusedin on just a few families,the familiesof Oedipus,Orestes, Medea etc., for these are the families that have been ripped apart by terribledeeds.' But what is it about the portrayalof a terribledeed done among kin that makesit particularly well suited to evokingpity and fear? Perhapsa start may be made in answeringthis questionby recognizing that at least a necessaryconditionfor the audiencefeeling pity and fear in responseto such terribledeeds is that they believe that such events could happento them. Forfearthisis obvious.Aristotle,as we haveseen, defines fear as a pain due to imaginingsome painfulor destructiveevent befalling one. And he furtherrequiresthat the fearfulevent be both imminentand capable of causing great pain.81For we do not fear distant pains, for example death, nor do we fear imminentbut minorpains.
"Fromthe definition it will follow that fear is caused by whateverwe feel has great

usin waysthattendto causeusgreatpain."I us, or of harming powerof destroying

Aristotle is explicit that we feel fear only when we believe that we are ourselvesvulnerableto an imminentand gravedanger:"we shall not fear thingsthatwe believecannothappento us".' A furtherconditionon fearis that we must believe that there is at least a faint possibilityof escape from the danger.8'
8 Poetics 11, 1452blO-11.For other 'objective uses of "pathos"in the Poetics, see e.g. 13, 1453b18, b19-20, b39, 54a13. See also Rhetoric 11.5, 1382b30; MetaphysicsV.21, 1022b20-21;Nicomachean Ethics 1.11, 1101a31. `9 It is tempting to speculate that, for Aristotle, there is also an objective as well as a subjective katharsis. For the katharsisreferredto in the definition of tragedy is clearly subjective- i.e. somethingthat goes on withinthe souls of the membersof the audience; while the katharsisat which Orestes is saved (17, 1455bl4-15) is clearly objective: viz. a ritual sacrifice. It goes beyond the evidence of the texts to construct a theory of the relation of objective to subjective katharsis. But it is worth noting in passing that if Aristotle believed that a subjectivekatharsisoccursin response to an objective katharsis, then the entire debate over where the katharsisis occurring,withinthe play itself or in the audience, would be idle. It would be occurringin both places (albeit in differentforms). ? Poetics 13, esp. 1453al7-22; 14, 1454a9-13. 81 Rhetoric11.5, 1382a22-30. 8 R

Rhetoric 11.5, 1382a28-30.

RhetoricII.5, 1382b31-31;cp. b28-1383al2.

319

At first sight, it appearsthat pity is the paradigm of an other-regarding emotion. We feel pity for others when they sufferwhat we believe to be However, Aristotle makesit clear that in orderto feel undeservedpain.95 pity for others we must also believe that the terrible event which has befallenthemmightbefallus or ourlovedones and, moreover,mightbefall us soon. Thusin orderfor us to feel pityfor others,we mustbelievethatthe similarto our own. One mightat firstthink others'situationis significantly that pity can be felt for those who are in some relevantrespect like us or age - even thoughwe do not believe eitherin socialstanding,character, we couldend up in theirsituation,but Aristotledeniesthis. We do feel pity for those who are like us, but the reasonwe do, Aristotlethinks,is because in such cases we think it more likely that the misfortunethat has befallen ThisexplainsAristotle'sotherwisepuzzlingremarkin them can befallus."6 the Poeticsthatwe fear for someonewho is similarto us.' Why,one might ask, shouldone fear for someoneelse - even if he is like us? The appropriate emotion to direct towardsanother, especiallytowardanotherwho is similarto us, should(we mightthink)be pity. Aristotle'spointis thatfearis person'smisfortune: an appropriate emotionto feel in responseto a similar we recognizethatwe standin the samedangerhe for throughhis similarity did. Likewisewith pity. Aristotle'sonly caveatis that the perceiveddanger cannot be too immediate:for in that case fear (for oneself) will driveout pity (for others)88Pity will also be drivenout of the souls of those who, alreadyruined, believe that no bad can furtherharmthem, and of those who believe themselves omnipotentand imperviousto harm.'
andhave suffered Thosewho thinkthey maysufferare thosewho havealready
safely escaped from it: elderlymen, owing to theirgood sense and theirexperience;

andalsoeducated to cowardice; meninclined people, for they weakmen,especially

are able to reason well."'

Aristotle clearlyrecognizespity as a reasonableemotion for an educated and thoughtfulperson: and since good tragedyis ideallyfor an educated
RhetoricII.5, 1383a5-8.Those who have lost all hope of escape grow resigned and callous. 11 Poetics 13, 1453a5;Rhetoric11.8, 1385bl4ff. 16 RhetoricI.8, 1386a24-27. 87 Poetics 13, 1453a5-6. 88 RhetoricI1.8, 1386a24-25. 19 Rhetoric 11.8, 1385bl9ff. 9 RhetoricIL.8,1385b23-27.Cp. PoliticsVIII.7, 1342b19,where an educated audience (hoi pepaideumenoi)is contrastedwith a vulgarone.
91

320

audience, it follows that, for Aristotle, the pity whichgood tragedyevokes is a reasonableemotionalresponseto the events portrayed. 6. It followsthata normal,educatedaudience,goingto a performance of a good tragedy, believes that the terrible events portrayed- infanticide, parricide, matricide, the tearing apart of the most primordialbonds of familyandsociety- couldhappento them. Had they lackedthatbelief they would, in Aristotle's eyes, be incapableof experiencingthe tragic emotions. This allows us to impose a further constraint, at least upon the emotions from which a tragickatharsisis produced: (7) The events whichin a tragedyproperlyprovokethe pity and fear from which a tragickatharsisoccurs must be such that the audience believes that such events could happento them. Beforeproceeding,I wouldlike to disposeof two objectionswhichmightbe raisedagainstthis conclusion.The most seriousobjectionis that the audience need not believe that the terribleevents could happento them: they are able to experiencethe tragicemotionsbecausethey are able to identify with the centralcharacter imaginatively andthusempathically feel whathe feels. WithinAristotle'sworld,it is clearthatthe objectionhasthe situation the wrongwayaround:for Aristotle,it is onlybecausewe thinkourselvesto be sufficiently like anotherthatwe canidentifywithhim.9" ForAristotle,we cannotidentifywiththe verybador withthe gods:it is preciselybecausewe are so distant from such beings that our emotions must retain a similar distance from theirs. That is why, for Aristotle, there is no important distinctionto be made between our feeling our fear and our feeling Oedipus'sfear. The verypossibility of ourimaginatively feelingOedipus'sfearis groundedin the recognitionthat we are like him: that is, it is groundedin the possibilityof our fearingfor ourselves.92 Moreover,this objectiondoes not take seriouslythe emotionof pity. We cannotfeel pity in imaginatively identifyingwith Oedipus:part of what makes Oedipussuch a remarkable andadmirable figureis his lackof self-pity,his willingnessto acceptresponsibility for his acts. But if our pity isn't an imaginativere-enactmentof Oedipus'sself-pity,then it must, as we have seen, be groundedin the belief that his fate could be ours.93
91

Since it is an incredibly complicated subject, I would like to reserve for another occasion a discussion of the general conditions required for emotional identification. 9 Poetics 13, 1453a5-6;Rhetoric11.5, 1383al0-13. 9 One might lamely try to keep the objection alive by saying that when we feel pity we are identifyingwith the chorus. But then the question arises:why should we identifywith the chorus? The only plausible answer is that the chorus is in some way expressing our

321

The less serious objection is that the audience doesn't come to the performancebelieving that the terrible events portrayedin the tragedy thatthisis so by the performance couldhappento them:they arepersuaded itself. The shortestanswerto this objectionis also the best: tragedyis not rhetoric, it is poetry. Because fear sets us thinkingabout how to escape from the perceiveddanger, an oratormay wish to persuadehis audience that they are in danger,' but a tragedydoesn'ttryto persuadeits audience of anything.The only effect on the audiencethata tragedyaimsto produce is a certain emotional response (the content of which we are trying to uncover). Of course, if tragedyis to succeedin this, it mustportrayevents But whichare convincing,plausible,events whichplausiblycould occur.95 Aristotle'spoint in insistingthat the poet constructplausible,convincing plots is not so that he maypersuadethe audienceof anythingbut so thathe may portrayan event whichthe audiencecan recognizeas one that could, just possibly, happento them. of a good Now if a normal,educatedaudience,goingto the performance tragedy,believes that the terribleevents to be portrayedcould, just posfactwhichis trueof them sibly, happento them, thereseemsto be a striking both beforethey enter andafterthey leave the theatre:they aremissingthe feelingswhichtogetherwith theirbeliefs wouldconstitutethe emotionsof pity and fear. One might like to say that they are cut off from their emotions, but that can't be quite right.Since, for Aristotle, emotionsare partiallyconstitutedby beliefs, it is more accurateto say that the distinct elements that conjointlyconstitutean emotion- belief and feeling- seem split off fromone another.Anotherway of puttingit is to say that normal andoutsideof the theatreseem to educatedpeople in normalcircumstances have certainbeliefs that they do not feel.' A misleadingway of putting an importanttruth is this: that when a of a good tragedy,he normal,educatedpersonexperiencesa performance to is able to unify certainbeliefs he has with feelings that are appropriate or he could commit those beliefs. He came to the theatre believingthat suffer terribledeeds. In the theatre he is able to feel those beliefs. But is a unification of belief and beforewe jumpto the conclusionthatkatharsis of the tragicemotions,let us stop to considerwhythis feeling, a unification
views. And if that is so, we are againled back to the conclusionthat we believe that what happened to Oedipus could happen to us. 9' RhetoricII.5, 1383a7-12. 9I E.g. Poetics, 9, 1452a36-38,b5-7, b15-19. 9' I use "outside the theatre" in the widest possible way: even the oral recitation of a tragedy counts for the purposes of this essay as going on "inside the theatre".

322

mode of expressionis misleading.It is misleadingbecauseit suggeststhat what we feel in the theatreis what we ought to feel in real life: that in real life the appropriate feelingsaresomehowkept at bayfromthe beliefswhich would rationalizethem. But this cannot be right. For constraint(1) requiresthat the virtuous person experience a katharsisin the performanceof a tragedy, but his emotional reactions are alreadyappropriateto the real life situationsin whichhe lives; and constraint(3) requiresthat our emotionalresponseto tragedyis not whatwe wouldor shouldfeel in responseto real life counterparts.Tragicpleasuredependscrucially on the belief thatone is emotionally respondingto a mimesisof tragic events.' Without this belief, tragic pleasureis impossible.Therefore,constraint (7) - thatthe audiencebelieve that tragic events could happen to them - must be interpretedin a way whichdoes not suggestthatthe virtuousperson,in not feelingpity and fear in ordinary life, is somehowcut off froma properemotionalresponseto his situation.It is completelyunAristotelian to supposethatwhatwe feel in the theatreis whatwe ought to feel in real life, but for some reasondo not. In real life the virtuousman feels just what he ought to feel. But, then, how could he believe that terrible,tragicdeeds could, just possibly, befall him and not feel fear and dread? Everything dependson the strengthof the modaloperator.The virtuous man believes that terrible,tragicevents could happento him, true, but the possibilityof those thingshappeningis, in his opinion, too remote for the actualfeelingof fearto be warranted.9Althougha tragicbreakdown of the primordial ties of humanlife is possible, the virtuousman also recognizes that this is less likely to happento him than almost anythingelse. That is why it is misleadingto say that tragedyrestoresthe appropriate feelingsto our alreadyexisting beliefs. Our belief that tragicevents could, just possibly, befallus alreadyhas the appropriate feelingattachedto it outsidethe theatre. No unificationis needed for, at least in the case of the virtuous person, there is no split that needs to be overcome. And yet the belief that tragicevents could, just possibly, happen to us does exert some pressureon our souls - even on the souls of us virtuous
9' See constraints(3)-(6). 9 If I may for a moment indulge my desire to be droll, let me put this in the languageof modal semantics: In the virtuous man's opinion (and thus: in truth) the worlds in which he kills his mother, is killed by his mother, etc. are possible worlds and thus stand in an accessibilityrelationto the real world. All tragicworldsare possibleworlds. However, all such tragicworlds are sufficientlyremoved from the actualworld of a virtuousperson (in ordinarycircumstances)that they do not fall within the set of legitimatelyfeared worlds.

323

people. This is preciselythe pressurewhichtakes us to the theatre. For in the theatre we can imaginativelybring what we take to be a remote possibilitycloser to home. As Aristotlehimselfsaid:
gestures,tones, ". . . thosewho heightenthe effectof theirwordswithsuitable in exciting successful actiongenerally,are especially and dramatic appearance, beforeoureyes, andmakethemseemcloseto us, pity:theythusputthe disasters or justpast."I justcoming

The tragicpoet, for Aristotle, playsa role in the worldof emotionssomewhat similarto the role of the skeptic within the world of beliefs. The skepticawakensus to the factthatwe ourselvesbelievein certainepistemic life we ignore:for example,thatwe couldbe possibilitieswhichin ordinary asleep, dreaming,or perhapsdeceivedby an evil demon. On the one hand, these possibilitiesare extremely remote, so we are justified in ignoring life;on the otherhand,theylendcontentto the ideathatin themin ordinary life we are living"insidethe plain":andthey fuel our desireto get ordinary outside the plain of everyday life and see how things really are,
absolutely."

The tragicpoet awakensus to the fact that there are certainemotional possibilities which we ignore in ordinarylife. On the one hand, these to ignorethem possibilitiesare remote,so it is not completelyunreasonable in ordinarylife; on the other hand, they lend content to the idea that in ordinarylife we are living "inside the plain": and they fuel our desire to experiencelife outsidethe plain. Even if tragedydoes not imaginatively we befallus, it goes to the root of the humanconditionthatit is a possibility must live with. And, even if remote, the possibilityof tragedyis not only much more imminent than the skeptical possibilities, it is much more areso designedthattheymake threatening.Forwhileskepticalpossibilities no differenceto the experienceof our lives, in tragedyour lives are ripped asunder. But there is a genuineproblemabout how to experiencetragicpossibillife is too remote of tragedyin ordinary ity. On the one hand,the possibility to justify real fear, on the other hand, it is too importantand too close to ignore. Tragic poetry provides an arena in which one can imaginatively of a play "capturesour experiencethe tragicemotions:the performance souls".01However,it is crucialto the pleasurewe derivefromtragedy,that
9

within the context of RhetoricII.8, 1386a32-35.Of course, Aristotle is here taLking rhetoricalpersuasion, but his point obviously carriesover to the theatre. ' See Thompson Clarke, 'The Legacy of Skepticism",Journal of Philosophy, 1972. 10 psuchagogei: cf. Poetics 6, 1450a33-36.

324

we never lose sight of the fact that we are an audience,enjoyinga workof of pityandfearwouldcollapseinto katharsis art. Otherwisethe pleasurable " Aristotle is keenly the merely painful experience of those emotions. between a mimesis of a seriousactionand difference awareof the important the seriousactionof whichit is a mimesis.The emotionalresponsewhichis - would be thorappropriateto a mimesis- tragicpleasureand katharsis to the real event. oughly inappropriate inappropriIt is thisexperienceof the tragicemotionsin an appropriately which, I think, helps to explainour experienceof relief in ate environment live life to the full, but we risk nothing.The the theatre. We imaginatively relief is thusnot thatof 'releasingpent-upemotions'per se, it is the reliefof 'releasing'these emotions in a safe environment.But to say that it is this is not to experienceof relief to whichAristotle gave the name "katharsis" characterize it fully:one needs also to knowthe contentof our relief, what our relief is about. Here I will only mentionbrieflycertainconsolationswhichare integralto Aristotle'sconceptionof tragedy.The worldof tragicevents must, Aristotle repeatedlyinsists,be rational.The subjectof tragedymaybe a good man, his fall."0The mere fall of a but he mustmake a mistakewhichrationalizes good manfromgood fortuneto badfortunefor no reasonat all, isn'ttragic, it's disgusting. 10The eventsin a tragedymustbe necessaryor plausible,and occur on accountof one another."0Insofaras we do fear that they must tragicevents couldoccurin our lives, whatwe fear is chaos:the breakdown of the primordial bondswhichlinkspersonto person.For Aristotle, a good tragedyoffers us this consolation:that even when the breakdownof the primordialbonds occurs, it does not occur in a world which is in itself ultimatelychaoticand meaningless. It is significantthat, for Aristotle, OedipusRex is the paradigmtragedy Forthe pointof tragedy,in Aristotle'seyes, is ratherthan, say, Antigone."0 not to portraya worldin whicha personthroughno faultof his own maybe and destructivedemands.In Arisirreconcilable subjectto fundamentally totle's conceptionof tragedy,the individualactor takes on the burdenof badness,the worldas a whole is absolved." And there is furtherconsola10

See constraints (4)-(6) above. Poetics 13, 1453a7-17;15, 1454b8-13. 104 Poetics 13, 1452b30-36. 105 Poetics 9, 1452a3-4; 10, 1452a20-21;15, 1454a33-36;16, 1455al7; 9, 1451a36-38. 106 Which was, of course, Hegel's choice. 107 See W.R.D. Fairbaim'saccountof "the moraldefense" in "The Repression and the Return of Bad Objects", PsychoanalyticStudiesof the Personality,op.cit.
03

325

tion in recognizingthat even when they are responsiblefor their misfortunes, humansremaincapable of conductingthemselveswith dignity and nobility.1' Even in his humiliationand shame, Oedipusinspiresour awe and admiration. In the RhetoricAristotle says that those who have alreadyexperienced great disastersno longer feel fear, for they feel they have alreadyexperiIn tragedy, we are able to put ourselves enced every kind of horror.109 imaginatively in a positionin whichthereis nothingfurtherto fear. Thereis consolationin realizing thatone hasexperiencedthe worst,thereis nothing furtherto fear, and yet the world remainsa rational,meaningfulplace in whicha personcan conducthimselfwith dignity.Even in tragedy,perhaps especially in tragedy, the fundamentalgoodness of man and world are
reaffirmed.110

Yale University

1" Aristotle makes a related (though different) point at NicomacheanEthics I.10: he reluctantlyadmits that even a virtuous person can suffer great misfortunehowever he offers the consolation that the virtuouspersonwill at least bear his misfortunesnobly and with greatness of soul. 109 Rhetoric11.5, 1383a3-5. 110 For another treatment of skepticism and its relationshipto tragedy see, of course, Stanley Cavell, The Claimof Reason (Oxford: ClarendonPress, 1979). I would like to thank Giovanni Ferrariboth for the many lovely evenings in which we translated and discussedthe Poetics together and for his criticismsof an earlier draft.

326