[Aristotle's Poetics: Introduction]

CLAS3152 FURTHER GREEK LITERATURE II:
Aristotle's Poetics

Lecture Notes

Parts of Animals 645a8-15: There are animals which are unattractive to the senses when one studies them; but even in these, nature's craftsmanship provides innumerable pleasures for those who can discern the causes. It would be unreasonable - in fact, absurd - if we got pleasure from studying pictures of these things, because then we are at the same time studying the art which crafted them (e.g. the art of painting or sculpture), but did not get even more pleasure from studying the actual products of nature - at least when we can make out their causes.

Basic assumption: Metaphysics 1.1, 980a20: 'All human beings by nature desire knowledge'. 982b12: philosophy begins from astonishment.

3. So biological research is rooted in astonishment and a desire to understand - as is study of human society and culture. This includes e.g. political theory (Politics), and poetic theory (Poetics): note that poetry had a prominent place in Greek culture (e.g. the Athenian dramatic festivals).

4. Understanding of poetics: not enough to make good poet (elements requiring natural talent: 55a30- 4 on empathy; 59a5-8 on metaphor); not necessary to make good poet (51a24 on Homer: art or instinct? 54a10f. on early tragedians: chance not art). But satisfies desire to understand; may enhance enjoyment of poetry (cf. Parts of Animals 645a8-15, above).

5. Did Aristotle get it right? Consider:
(a) chronology: 4th century analyst of 5th century tragedy (Sophocles and Euripides died 406 BC). Might 4th century tragedy (e.g. Aristotle’s friend Theodectes) have been different and therefore misleading?
(b) context: debate with earlier theorists, esp. Plato; might need to defend poetry against philosophical attacks distort Aristotle's apprach?

6. Aristotle's 'published' works lost; surviving works are private notes, so difficult to understand.

Preparation for next lecture:

• read Poetics chapters 1-4 (pp.3-9).

1.Normal view: poetry = verse (e.g. Gorgias fr. 11.9); for Aristotle’s view that verse is not definitive of poetry cf. ch. 1, 47b17-20 (Empedocles not a poet), ch. 9, 51b2-4. Note that Aristotle is marking out a limited technical sense for ‘poetry’ not evaluating Emepdocles’ verse (admiration for Empedocles: see Aristotle fr. 70).

2. Aristotle’s technical sense: poetry is imitation (mimêsis).

3. Imitation x (a) not a real x, and (b) like a real x (e.g. picture of dog not a dog, but looks like a dog). What kind of 'likeness' is involved in poetry? Possibly: evokes similar effect (murder in tragedy evokes shock, outrage, pity).

[On imitation see e.g. P. Woodruff 'Aristotle on mimêsis' in A. Rorty Essays on Aristotle's Poetics 73-96]

[Background to poetry as imitation (mimêsis):
(a) 5th c.: Aristophanes Thesmo. 156;
(b) Plato: Republic 3, 392-8 distinguishes narrative, imitative and mixed poetry; Republic 10 criticises all poetry as imitation (as things are imperfect copies of forms, so poems are imperfect copies of objects: cf. Rep. 2-3 on poets' falsehoods about gods and morality); Laws 2, 667c-8a.]

2.2 Poetry and human nature

1. Metaphysics 1.1, 980a20: 'All human beings by nature desire knowledge'.

2. 'Natural' implies 'enjoyable': enjoyment = unimpeded exercise of a capacity hence acting in accordance with one's nature is inherently pleasurable (Nicomachean Ethics 1174b14-5a21).

3. Human capacity for understanding and reasoning implies pleasure in understanding and reasoning (cf. Rhetoric 1371b4-10; but n.b. Rhetoric 1410b10-11: 'learning easily is naturally pleasant to all'). Evidenced in children’s delight in recognition - including recognition of imitations (which requires reasoning, to grasp that this object is an imitation of x).

4. Therefore poetry, as a form of imitation, is grounded in human nature and naturally pleasurable (cf. ch. 4, 48b4-19).

5. Poetry as imitation in verse (prose imitations: see ch. 1, 47a28-b18 on Socratic dialogues; but these were culturally marginal) explained by natural human pleasure in rhythm and melody: ch. 4, 48b20-22. (Is Aristotle's relative lack of interest in poetic language justified?)

2.3 The classification of poetry: medium, object and mode

1. Medium (ch. 1):

rhythm	rhythm melody	language rhythm melody	language rhythm melody (in parts)	language rhythm	language
(dance)	(instrumental music)	Lyric	Tragedy, Comedy	Epic	(prose)
		<-----------	poetry	----------->

2. Object (ch. 2): poetry imitates agents and actions: ch. 3, 48a1; ch. 4, 48b25-6; ch. 6, 49b24, 49b36-50a4 etc.

Agents better than us	(Agents like us)	Agents worse than us
Epic, tragedy		Comedy

3. Mode (ch. 3):

Narrative	(mixed)	Dramatic
Epic	Homer	Tragedy, Comedy

Homer's quasi-dramatic style: ch. 24, 60a5-11; cf. ch. 4, 48b34-49a2

4. Summary:

	Epic	Tragedy	Comedy
Medium	Language rhythm	Language rhythm melody	Language rhythm melody
Object	Agents better than us	Agents better than us	Agents worse than us
Mode	Narrative	Dramatic	Dramatic

2.4 Definition of tragedy

1. Ch. 6, 49b24-28:
(a) imitation: cf. ch. 1.
(b) of admirable action: cf. ch. 2 ('admirable' = spoudaios).
(c) of complete action with magnitude (= size): cf. ch. 7.
(d) in language made pleasurable: medium = language with rhythm and melody: cf. ch. 1 (and ch. 4 for natural pleasure in rhythm in rhythm and melody).
(e) each of of its species separated in different parts: i.e. song not used throughout: cf. ch. 1.
(f) performed by actors, not through narration: dramatic mode: cf. ch. 3.
(g) effecting through pity and fear the purification [katharsis]: first reference to emotional impact (only reference to katharsis).

2.4 Qualitative parts

1. For quantitative division of poetic text see ch. 12.

2. Qualitative analysis:
(a) dramatic mode implies visual dimension; hence (i) spectacle.
(b) medium is rhythmical language sometimes accompanied by melody; hence (ii) lyric poetry and (iii) diction (i.e. sung and spoken text).
(c) object of imitation is action; hence (iv) plot (sequence of events).
(d) action is performed by agents with moral and intellectual qualities; hence (v) character and (vi) reasoning (i.e. moral disposition and judgement of current situation, which together motivate action).

Preparation for next week’s lectures:

• read Poetics chapters 6-8 (pp.10-15);
• read Poetics chapters 9-11 (pp.16-19).

2. Plot is most important: cf. ch. 4, 48b4- 24: imitation discussed at greater length than rhythm (Aristotle as philosopher more interested in human capacity for reason? Cf. ch. 9, 51a36-51b11, poetry 'more philosophical and more serious' than history by virtue of universalised plots). Imitation (poetry is imitation in verse (ch. 1); tragedy imitates an action (ch. 6); plot = action imitated) is more fundamental than verse: cf. ch. 9, 51b27-32 (poet, unlike historian, actively forms plot).

3. History of poetry (ch. 4):

(a) Originates in improvisations, dividing into two streams: imitations of good/bad actions and agents (cf. ch. 2).

(b) First stage: hymns (praise) and invectives (blame); both lack narrative structure.

(c) Second stage: epic acquires narrative structure. Corresponding imitation of 'low' characters = Margites, burlesque narrative poem attributed to Homer.

(d) Third stage: Homeric narrative poems using quasi-dramatic mode (cf. ch. 3); anticipates drama.

(e) Fourth stage: tragedy and comedy: developed by adaptation of from improvisatory lyric performances (dithyramb, phallic songs).

3.2 The primacy of plot

1. Explicit arguments, relating theoretical premise to observed facts (are Aristotle's observations correct?):

(a) good/bad fortune depend on action, not just character (n.b. 'action' = praxis, what one does and how one fares; e.g. ch. 11, 52b10-13: suffering (pathos) is 'a praxis involving destruction or pain'). Good/bad fortune evoke tragedy's emotional response.

(b) tragedy can exist without character (motivation would be supplied by reasoning rather than moral disposition, i.e. is impersonal rather than concrete; cf. ch. 6, 50b4-12 on 'rhetorical' tragedy in the 4th century).

(c) emotional effect of tragedy is more dependent on the quality of the plot than on the quality of the portrayal of character and thought.

(d) devices of plot have greatest impact; see ch. 11 for reversal and recognition.

(e) excellence in plot-construction hardest to attain.

3.3 The well-formed plot

1. Structure of discussion of plot:

ch. 7-9: minimum criteria for well- formed plot;
ch. 10-11: simple/complex plots, and their components;
ch. 12: digression (quantitative parts of tragedy);
ch. 13-14: optimum criteria: the best kind of tragic plot.

2. Completeness (ch. 7): tragedy imitates a 'complete and whole' action, i.e. one which has 'beginning, middle and end'. Note:

(a) beginning, middle and end describe structure of action; i.e. the reference is to the plot, not the play.

(b) beginning, middle and end imply ordered structure, in two respects:
(i) a well-formed plot is a connected series of events: each thing follows on another as a necessary consequence;
(ii) a well-formed plot consists of a self-contained (or closed) series of events: first element is not a necessary consequence of something else; last element has no further necessary consequence.

3. Necessity and probability (ch. 7): i.e. what happens invariably (necessity) or usually (probability).

4. Correct magnitude (ch. 7): what is the right size for a plot?

(a) Upper limit: constrained by what an audience can grasp, i.e. memory.

(b) Lower limit: must allow scope for change of fortune (bad to good, or good to bad) consistent with requirement of necessary connection. (For change of fortune: cf. first argument for primacy of plot. How does change from bad to good fit with requirement of pity and fear?)

Where does the ideal size fall between these limits? Note the following claims:

(i) ch. 7: within the limits defined, the greater magnitude the better (51a10-11);
(ii) ch. 24: epic is subject to same maximum limit as tragedy, i.e. memory-span; (59b18-20);
(iii) ch. 18, ch. 24: tragic plots should be less extensive than epic plots (56a10-19, 59b17-18).

Implications of (ii) + (iii) seem inconsistent with (i). Also:

(iv) ch. 26: tragedy superior to epic because of more limited plot: greater concentration (62a18-b3);
(v) ch. 24: greater expansiveness of epic is advantage over tragedy: greater variety (59b22-8).

5. Unity (ch. 8): events may happen to one person without necessary connection; so plots with biographical structure defective. Cf. ch. 23: events may happen in one period without necessary connection: hence plots with historiographical structure also defective. Unity of plot depends on connected, closed series of events.

6. Why is poetry subject to these requirements? Consider Aristotle's teleological approach (i.e. in terms of purpose or function (telos) of poetry): right design for tragedy depends on what tragedy is for. (Cf. ethics: right way of life is one which enables us to achieve the goal of being human.) Function of tragedy must depend on, or be enhanced by, necessary or probable connection. What is function of tragedy? Provisional answer: 'effecting through pity and fear...' Why do pity and fear depend on necessary or probable connection? See ch. 9, 52a1-11; ch. 10.

7. Determinate structure (ch. 8, 51a30-5): if plot is series of events with necessary or probable connections, removal or transposition of any one element will disrupt whole.

Preparation for next lecture:

• read Poetics chapters 9-11 (pp.16-19).

Poetry 'tends to express universals' (cf. ch. 5, 49b5-9): more precisely, poetry makes particular statements which presuppose universals (e.g. 'killing his mother is the sort of thing that a person like Orestes would necessarily or probably do in such circumstances'); this presupposition is conditional on a necessary or probable connection between events.

This implies interest in universal patterns and structures, shared with philosophy; hence poetry 'more philosophical' than history: philosophy directly concerned, poetry indirectly concerned with universal truths; history is only concerned with particular truths.

Ch. 25: poetic plots may be based on falsehoods as well as truths (e.g. about gods).

2. Episodic plots (ch. 9, 51b3- 52a1): episodes without necessary or probable connection.

4.2 Astonishment, reversal and recognition

1. Astonishment (ch. 9, 52a1- 11): astonishment is good but not necessary: move from minimum to optimum criteria of plot. Note appeal to emotional effect as basis for judging superiority of one tragic plot over another.

Emotional effect maximised by: (i) astonishment and (ii) connection.

This explains why requirements of wholeness, unity etc apply to poetry: connection is linked to emotional effect, which is what tragedy aims at. (Is it true that emotional impact of causally intelligible events is greater than that of chance events?)

2. Example: Mitys' murderer killed by Mitys' statue. No connection; but illusion of connection increases astonishment - so real connection must do so also. Note:

(a) is this tragic? Pity (perhaps - unless murderer deserved to die: cf. ch. 13). Fear: dead man taking vengeance frightening?

(b) illusion of connectedness: is this significant? Ch. 24 (60a26-b2) on 'irrationalities' suggests requirement of necessary/probable connection is applied flexibly (cf. also ch. 15, 54b6-8 on irrationalities 'outside the play'; ch. 25, 60b23-6: irrationalities are acceptable if they (i) enhance effect of play or epic, and (ii) are concealed).

Is Aristotle to be seen as flexible and pragmatic or systematic and doctrinaire?

3. Simple/complex plots (ch. 10):

(a) simple: (i) events connected (= unified); (ii) change of fortune (cf. ch. 7, 51a11-15); (iii) no reversal or recognition.

(b) complex: as (a), with reversal and/or recognition.

4. Reversal/recognition (ch. 11):

(a) reversal (peripeteia) = 'a change to the opposite in the actions being performed, in the way stated', i.e. unexpectedly (ch. 9, 52a1-11).

(b) Recognition (anagnôrisis) = 'a change from ignorance to knowledge', if it affects good/bad fortune. Note 'close connection (= philia, including family) and enmity': cf. ch. 14.

5. Techniques of recognition (ch. 16):

(a) by 'tokens', i.e. physical marks (e.g. Odysseus' scar) or tokens (e.g. objects exposed with baby, as in Euripides Ion): these are accidental (cf. ch. 8, 51a25-6: wounding of Odysseus unconnected with Odyssey plot). Distinguish use: (i) to confirm information already public; (ii) to disclose new information: this may combine with reversal, hence more effective (cf. ch. 11, 52a32-3)

(b) by action or word contrived for recognition, i.e. not natural consequence of plot (e.g. Orestes' proof of identity in Euripides IT);

(c-d) based on memory and reasoning (interpretation difficult);

(e) arising from plot (e.g. recognition of Ipihgeneia in IT); cf. ch. 9, 52a4: astonishment plus connection enhances emotional impact.

6. Is the expectation in reversal that of character or audience? Ch. 9 implies audience (astonishment increases effect on their emotions); examples in ch. 11 may imply characters ('it seemed that he would...': seemed to whom?). Perhaps both: trend of action dictates characters' expectation; audience may feel emotional impact of reversal of trend even if knows outcome in advance.

7. Reversal and recognition both involve exposure of misjudgement; character misinterprets situation, and consequently acts inappropriately; cf. ch. 13 on tragic error (hamartia), ch. 14 on ignorance.

4.3 Complication and resolution

2. N.b. Aristotle discusses structure of plot, not play. Cf. Odyssey and Aeneid: both involve flashback to earlier events; are these events:

(a) part of plot, narrated out of order (ABC narrated BAC)?
(b) outside plot, so text contains things that are not part of plot?

Either implies text-structure independent of plot-structure. (Aristotle's theory is consistent with technique of Odyssey: see ch. 17, 55b15-23).

3. Hence complication/resolution parts of plot, not play: complication may include things 'outside the play'; all complication could precede play.

Difficulty of managing resolution well: 56a9-10: cf. ch. 15, 54a37-b8: resolution should arise from plot (prior events = complication), not arbitrary divine intervention.

4. What is legitimate use of gods? Conveying information about actions before or after play explicitly permitted; also 'irrational' divine interventions outside the play tolerated (cf. 54b6-7). But are divine interventions in course of play:

(a) implicitly rejected as irrational (e.g. S. Halliwell Aristotle's Poetics ch. 7); or:

(b) implicitly accepted if necessary or probable. Cf. ch. 25, 60b35-61a1, defending false theology in poetry; ch. 17, 55b17-19, on Poseidon in Odyssey.

Preparation for next week’s lectures:

• read Poetics chapters 13-14 (pp.20-24).

For the superiority of plots with reversal/recognition cf. ch. 6, 50a33-5.

[Quantitative parts of tragedy (ch. 12): i.e. segments of tragic text. Why is this section so unsatisfactory? Not important to Aristotle? Not by Aristotle? (N.b.: interrupts discussion of plot.)]

5.2 Best kind of tragic plot: first discussion (chapter 13)

1. Two variables:

(a) direction of change of fortune (good-bad or bad-good: cf. ch. 7);
(b) moral status of person(s) involved.

2. Yield four combination:

(i) virtuous person's fortune changes from good to bad: evokes disgust (Gk: miaron) rather than fear/pity;

(ii) virtuous person's fortune changes from bad to good: Aristotle does not consider this possibility - self-evident? (but see below on 'double' plots);

(iii) wicked person's fortune changes from good to bad: satisfies human feeling (to philanthropon: see below), but not fear/pity;

(iv) wicked person's fortune changes from bad to good: no fear/pity.

3. No fear/pity for wicked man assumes:

(a) fear is felt for someone like us: virtuous audience therefore will not respond sympathetically to a wicked person. (N.b.: fear in Poetics therefore sympathetic emotion: not fear for self evoked by human vulnerability; and not just 'shudder', though for this cf. ch. 14, 53b5.)

[Aristotle analyses pity and fear in Rhetoric 2.5 and 2.8; for a discussion of the tragic emotions in the context of other ancient evidence in M. Heath Poetics of Greek Tragedy (London 1987) 11-17.]

(b) pity is felt for someone who does not deserve misfortune.

'Agreeable' (to philanthropon - i.e. satisfies human feeling): cf. also ch. 18, 56a19-25. Opposite of outrage/revulsion.

[For a discussion: C. Carey '"Philanthropy" in Aristotle's Poetics' Eranos 86 (1988) 131-9.]

4. If all plots involving virtuous or wicked are ruled out, best must involve someone in between, i.e. moderately virtuous. Note:

(a) moderately virtuous, but in other respects outstanding: 'great esteem... great good fortune'. Cf. ch. 2: tragedy and epic concerned with spoudaios ('good': includes status); ch. 6: tragedy imitates spoudaios ('admirable') action.

(b) the fall into misfortune is not poetic justice: misfortune must be undeserved to evoke pity. (Question: if so, can the restriction to moderate virtue be justified? I.e. is Aristotle right to find suffering of wholly virtuous untragic? Note Plato's critique of poetry portraying virtuous people suffering: Rep. 3, 392a-b; is Aristotle too defensive?)

5. Cause of fall into misfortune is hamartia = 'error' in broad sense

(a) traditionally interpreted as moral flaw; impossible (a 'large-scale moral flaw' is wickedness!);

(b) more recent interpreters equate with intellectual errors, mistakes of fact. In favour of this, note connection with ch. 10 on reversal and recognition: both involve misinterpretation of situation.

(c) But:
(i) intellectual error may rest on moral error (e.g. negligence);
(ii) such an error may be blameworthy but understandable;
(iii) even with full knowledge, moral error may be understandable (i.e. there may be mitigating circumstances);
(iv) in this case, a moral error may be large-scale (i.e. disastrous) without implying wickedness.

Hence hamartia may include both errors of ignorance and mitigated moral errors. If so, the 'best kind of tragic plot' allows wide range of variants.

[See T.C.W. Stinton 'Hamartia in Aristotle and Greek tragedy' Classical Quarterly 25 (1975) 221-54 = Collected Papers (Oxford 1990), 143-85. For the further question of whether mischance can be included see M. Heath 'The universality of poetry in Aristotle's Poetics' Classical Quarterly 41 (1991) 389-402.]

6. Second best plot: good and bad characters end in opposite states. Is this wholly untragic (cf. 'more characteristic of comedy')? But n.b. pity for initial or threatened suffering of good.

5.3 Best kind of tragic plot: second discussion (chapter 14)

1. Proper sources of tragic effect:

(a) from plot, not (e.g.) costume;
(b) fear and pity, not shock, amazement, horror etc.

2. Argument from relations between (rather than moral status of) characters: emotional potency greater if characters 'closely connected' (philos: includes family). Then two variables:

(a) injury done/not done;
(b) knowing/in ignorance of the relationship.

2. For ignorance cf. ch. 10, ch. 13; note again plots involving ignorance superior: injury against philos planned or executed knowingly is repulsive (miaron), but pitiful if in ignorance.

3. Best if act averted: contrast ch. 13, best if ends in misfortune; how is this inconsistency explained?

(a) change of mind? But n.b. cross-reference from ch. 14 (54a9) to ch. 13 (53a18-22);

(b) exploration without commitment? (Only one kind of tragic plot can be best; but several kinds could be excellent.) But Aristotle does not say this.

1. Tragic pleasure: for the characteristic (distinctive, appropriate) pleasure (oikeia hêdonê) cf. ch. 13, 53a35-6 (opp. comedy); ch. 14, 53b10-13: 'the pleasure which comes from pity and fear... by means of imitation'.

2. Tragedy provides various pleasures, not all to be identified with 'the characteristic pleasure' distinctive to it.

3. Cognitive pleasure: ch. 4, 48b8-19: humans take pleasure in imitation because understanding an imitation ('this is so-and-so') involves the exercise of a natural cognitive capacity: learning is pleasurable. Tragedy is imitation (in light of chs. 7-9: to understand a tragic plot we must recognise the events as a connected sequence, corresponding to some general or universal pattern: 'this is what would happen necessarily or probably').

This is a pleasure of tragedy (and does come through imitation), but is not the distinctive pleasure:

(i) not distinctive to tragedy (in comedy also);
(ii) not from fear and pity: neutral as to accompanying emotions.

[For attempts to reconcile the cognitive pleasure with the proper pleasure see S. Halliwell Aristotle's Poetics ch. 2, esp. pp. 69- 77; E. Belfiore 'Pleasure, tragedy and Aristotelian psychology' Classical Quarterly 35 (1985) 349-61; L. Golden Aristotle on Tragic and Comic Mimesis (Scholars Press 1992)]

4. Pleasures of verbal text: ch. 4, 48b20-21: rhythm and melody 'natural' (therefore naturally pleasurable: cf. above) to human beings; cf. ch. 6, 49b25: 'in speech made pleasurable', i.e. using rhythm and melody (49b28-9). Cf. ; ch. 6, 50b16: song 'most important of the sources of pleasure' (hêdusmata); ch. 26, 62a15-17: one advantage of tragedy over epic is music, which gives pleasure - as does visible performance (spectacle): cf. ch. 6, 50b16-17.

But these are pleasures of tragedy, but are not the distinctive pleasure:

(i) not distinctive to tragedy (in comedy also);
(ii) not from fear and pity;
(iii) not through imitation - which implies through the plot.

5. Tragedy imitates admirable persons and actions (ch. 2, ch. 6, 49b24), and contemplation of admirable objects gives greater pleasure (Nicomachean Ethics 10.4, 1174b20-3): but this would be true without misfortune - so the pleasure does not depend on pity and fear.

6. Pleasure from pity and fear is paradoxical: pity and fear are species of distress (lupê): Rhetoric 1382a21-22; 1385b13).

7. Katharsis: cf. ch. 6, 49b27-8: 'effecting through pity and fear the purification (katharsis) of such emotions'; no further reference in (extant) Poetics, but Politics 1341b38-40 refers to a 'clearer' discussion in Poetics (in the lost second book? - Note discussion of comedy promised at 49b20-21 is missing.

[For a discussion of what Aristotle might have said on comedy see M. Heath 'Aristotelian comedy' Classical Quarterly 39 (1989) 344-54.]

Various uses of music: educative for children; relaxation and leisure for adults; also for katharsis, e.g. of 'enthusiasm' (hysterical or ecstatic frenzy, as e.g. in cult of Dionsysus); thus Politics 1342a4-15:

'The emotion which affects some minds violently exists in all, but in different degrees, e.g. pity and fear, and also enthusiasm; for some people are prone to this disturbance, and we can observe the effect of sacred music on such people: whenever they make use of songs which arouse the mind to frenzy, they are calmed and attain as it were healing and katharsis. Necessarily, precisely the same effect applies to those prone to pity or fear or, in general, any other emotion, and to others to the extent that each is susceptible to such things: for all there occurs katharsis and pleasurable relief.'

Note:

(a) Aristotle does not disapprove of emotions; contrast Plato Republic 10, esp. 603-6: poetry arouses emotion and so strengthens emotional tendencies, which should be controlled. For Aristotle it is right to feel emotions appropriately, e.g. courage is a mean between cowardice (fears too much) and recklessness (fears too little); cf. Nicomachean Ethics 1115b11-20 etc.

(b) hence katharsis cannot be a 'purging' of emotion, i.e. releasing or getting rid of it; we ought to feel pity and fear as appropriate;

(c) but katharsis brings 'as it were healing' and 'pleasurable [n.b.] relief'; so corrects some fault: a proneness to feel the emotions to excess? I.e. 'purging' relieves pressure to feel pity and fear inappropriately felt by someone with disordered emotional disposition.

(d) this is 'pleasurable' because restoration to a natural or healthy state is pleasurable (EN 1154b17- 19).

Implications:

(a) those most prone to excessive or inappropriate emotion will get most benefit and pleasure from katharsis ('healing' assumes something wrong!);

(b) not everyone has disordered emotions; otherwise, there would be no virtuous people, which is absurd;

(c) the best members of an audience are most responsive to best kind of tragedy: cf. ch. 13, 53a33-5: 'the weakness of the audiences' leads to preference for second-best kind of tragic plot; ch. 26: contrary to tragedy's critics, it does not appeal especially to vulgar audiences.

(d) hence the pleasure of katharsis is not the distinctive pleasure of tragedy, since (i) katharsis applies most to inferior members of the audience, and (ii) the distinctive pleasure applies most to superior members of the audience.

Opposing argument: definition (ch. 6) implies katharsis is essential, and is 'final cause' (i.e., function, purpose) of tragedy; hence:

(a) katharsis cannot apply most to worst members of audience;
(b) if function of tragedy is to produce the distinctive pleasure (cf. ch. 23, 59a20-1; ch. 26, 60b23-5) then pleasure of katharsis must be distinctive pleasure.

Conclusion: if katharsis is 'as it were healing', it is not function of tragedy; if it is the function of tragedy it is not healing. Does the definition of tragedy imply that it is; if so. do we accept implications of (i) Aristotle's language in Politics or (ii) the definition? If latter, how do we interpret katharsis?

[For some other recent interpretations see (e.g.) S. Halliwell Aristotle's Poetics ch. 6 (and Appendix 5, which gives a brief survey of approaches to the problem); J. Lear 'Katharsis' Phronesis 33 (1988) 297-326 = A. Rorty Essays on Aristotle's Poetics 315-340; R. Janko 'From catharsis to the Aristotelian mean' in A. Rorty Essays on Aristotle's Poetics 341-58.]

8. The reply to Plato: if my interpretation is right, katharsis is a beneficial effect which tragedy has incidentally on inferior members of the audience; hence Aristotle has a multi-layered defence against Plato's critique of tragic emotionalism:

(a) emotions should not be suppressed, but exercised appropriately;

(b) the proper pleasure is most available to the best members of the audience; i.e. tragedy does not pander to the emotionally disordered;

(c) members of the audience who are emotionally disordered are not harmed: because of katharsis the emotional experience of tragedy does not increase but mitigates their tendency to feel excessive emotions.

9. The distinctive tragic pleasure. [Warning: this is highly speculative! There is no reason to think Aristotle held this theory; but is it a theory which Aristotle would have been able to accept?]:

(a) people enjoy exercising natural or acquired capacities; so virtuous enjoy exercise of virtue (Nicomachean Ethics 1099a7- 20; Eudemian Ethics 1249a18-21);

(b) pity and fear are appropriate responses to pitiful and fearful events of tragedy;

(c) therefore responding to tragedy with pity and fear is (for virtuous) an exercise of virtue;

(d) as an exercise of virtue this response is (for the virtuous) pleasurable;

(e) in real life pity or fear are felt appropriately in painful and distressing circumstances, which interfere with that pleasure; cf. Nicomachean Ethics 1117a29-b22: in principle acting courageously is pleasurable to the courageous, but the unpleasant circumstances 'conceal' the pleasure;

(f) but tragedy is imitation, not real; so pity and fear are not not complicated by painful attendant circumstances, and the pleasure may be felt without interference.

(g) hence the appropriate pleasure of tragedy is distinctive to tragedy; comes from pity and fear; depends on imitation; and is most available to virtuous members of audience - all the criteria are satisfied.

Preparation for next week’s lectures:

• read Poetics chapters 15-19, 23-26 (pp.2-32, 38-48);
• read Aeschylus’ Oresteia.

[On character and reasoing: A.M. Dale 'Êthos and dianoia: "character" and "thought" in Aristotle's Poetics' in Collected Papers (Cambridge 1969) 139-55; M.W. Blundell 'Êthos and dianoia reconsidered' in A. Rorty Essays on Aristotle's Poetics 155-75]

2. Character: ch. 15: four requirements:

(a) morally good (except as necessary to plot): cf. ch. 2 (on moral character and status: spoudaios opposed to phaulos, applied in ch. 15 to slaves), ch. 13 (on moral character: virtuous opposed to ponêros, applied in ch. 15 to Menelaus in Euripides Orestes); hence:

(i) tragedy is about people of high status and morally good;

(ii) high-status characters can be morally bad, but (i) not if they are meant to excite pity (deserved suffering not pitiable), and (ii) only as required by plot (otherwise too much like comedy);

(iii) low status figures peripheral, and should be good of their kind.

(b) appropriate: e.g., woman should have only womanly virtues;

(c) like (homoios): meaning obscure. Cf. ch. 2, 48a12 (like us, i.e. not better or worse), ch. 13, 53a5-6 (fear about someone like us). Cf. ch. 15, 54b8-15: portraits may be 'like' and idealised. (Some translate as 'life-like': is that 'realistic' or 'faithful to the model'?)

(d) Consistent: entailed by necessary/probable connection (but n.b. recognition of 'consistent' inconsistency).

3. Reasoning (ch. 19): compare ch. 6, 50b4-12 (reasoning revealed in assertion, denial or generalisation) vs ch. 19 (includes speech arousing emotion or increasing/decreasing importance) = any persuasive speech: hence reference to Rhetoric.

Note parallel between speaker (explicitly arousing emotion through words) and poet (implicitly arousing emotion through plot): parallel resources useful to each.

4. Speech (chs. 19-22): how relevant?

ch. 20: general linguistic theory;
ch. 21: classes of non-standard discourse;
ch. 22: poetry (characterised by non-standard usage): not tragedy in particular.

5. Lyric poetry: no specific discussion.

On chorus see ch. 18, 56a25-32: (i) use chorus as actor (presumably = active engagement within acts), like Sophocles not Euripides; (ii) make act-dividing songs relevant to plot (i.e. not interval-song supplied by producer, as 4th-century practice), like Sophocles and Euripides not others.

6. Spectacle: no specific discussion; cf. ch. 6, 50b16-20: not part of poet's art (tragedy is tragedy even if not performed). N.b. also hostile critics cited in ch. 26 object to performance (and cf. Plato Republic 3, 392-8).

At ch. 26, 62a16 are 'music and spectacle' advantages, or (with some editors) just music?

Ch. 17, 55a22-9: poet must visualise action: tragedy is for performance, so avoid what would be incongruous on-stage (contrast freedom of epic poet: ch. 24, 60a11-17).

Ch. 14, 53b1-11: recognises impact of spectacle (objects to this as substitute for effective plot: but would he object to spectacle to enhance effective plot?)

7.2 Epic and tragedy

1. Ch. 5, 49b16-20: 'Some of the component parts are common to both...': epic lacks lyric poetry and spectacle - the least important elements of tragedy. Does it follow that 'anyone who understands what is good and bad in tragedy also understands about epic'?

2. Criteria for well-formed epic plot same as tragedy: ch. 23, 59a17-59b2 (for Homer's excellence at plot-construction see also ch. 8, 51a22- 7).

Magnitude: epic plots more expansive (ch. 18, 56a10-19). Plot of Iliad (selecting 'one part' of whole Trojan war story) uniquely concentrated - for epic. Are ch. 23, 59b2-7 (Il., Od. furnish material for one or two tragedies) and ch. 26, 62a18-11 (material of one tragedy stretched to epic length watery) consistent? Ch. 24, 59b18-22: optimum length epic approx. 4- 5000 lines: implications for Il. and Od.?

3. Is best kind of epic plot identical to tragedy? Identical aims implied by:

(a) Ch. 26, 62b12-14: 'tragedy surpasses epic... also in artistic effect (since they should not produce any arbitrary pleasure but the one specified)'.

(b) Ch. 25, 60b23-6: impossibility acceptable 'if it attains the end of the art itself (the end has been stated above): i.e. if it makes either this or some other part have greater impact.'

Possible difference implied by example of pursuit of Hector (Iliad 22): cf. ch. 24, 60a14- 17. Astonishment (in tragedy: cf. ch. 9, 52a1-11) has more scope in epic (not visible on stage, so violations of necessity or probability more easily concealed: ch. 24, 60a11-18; cf. ch. 17, 55a22-29 on visualisation of tragic action).

Homer as first tragedian: e.g. Plato Republic 10, 598d, 607a

Is the assimilation of epic and tragedy justified?

3. Ch. 24, 59b7-16: 'Epic must also have the same kinds as tragedy...'. Cf. ch. 18, 55b32-56a3 for kinds of tragedy: missing fourth kind restored from classification of kinds of epic: 'either simple or complex, or based on character or on suffering'.

Comparison of Iliad (simple, based on suffering) and Odyssey (complex, based on character):

Odyssey: recognition-scenes - hence complex (cf. ch. 10); antithesis of good/ bad - hence based on character. Ch. 13, 53a30-32: 'double' plot (ends happily for good characters); contrast suffering of the Iliad. (But is there really no reversal in Iliad?)

4. Ch. 24, 59b12-16: Homer praised for reasoning and diction. Ch. 24, 60a5-11: also for character. Homer's characters speak for themselves: cf. history of poetry (ch. 4, 48b34-6; ch. 3, 48a21-3; contrast hostile assessment of this technique in Plato Republic 3, 392-8).

3. Consolidated feedback from the seminar groups will be posted on the course Web pages.

8.2 Aristotle on Aeschylus

(i) ch. 4 (49a15-19): increased actors to two, reduced choral parts;
(ii) ch. 16 (55a4-6): on recognition in Libation Bearers;
(iii) ch. 18 (56a17): Aeschylus Niobe (?) illustrates good practice in not putting too much into plot;
(iv) ch. 22 (58b18-24): line from Aeschylus Philoctetes improved by Euripides.

Outside the Poetics: 10-line quotation (History of Animals 633a18); 1-line quotation (Rhetoric 1388a8); story about Aeschylus' life (EN 1111a10) - less than on Agathon, Theodectes.

2. On the story of Orestes:

(i) ch. 13 (53a20): people 'whose lot it has been to experience something terrible - or to perform some terrible action';
(ii) ch. 13 (53a36-9): reconciliation of Orestes and Aegisthus would be more appropriate to comedy;
(iii) ch. 14 (53b23f.): killing of Clytaemnestra by Orestes illustrates suffering inflicted on someone closely related.

3. Note no discussion of trilogy format (abandoned by tragedians after Aeschylus).

8.3 The Oresteia

Furies in Eumenides: cf. ch.14, 53b1-11; but Furies (invisible at end of Libation Bearers) esential to plot - spectacle reinforces plot (and staging is outside poet's control). Effect of horror (also from the ghost) is not instead of fear and pity.

2. Lyric poetry: note Aeschylus' use of long choral songs; ch. 4 (49a15-19) says he reduced the choral part in favour of spoken word - does Aristotle dislike long choral songs? But (cf. ch. 18, 56a25- 32) songs integrated with plot, chorus involved with in the acts (esp. in Eumenides!).

3. Diction: important in Aeschylus - e.g. imagery. Aristotle might approve: cf. ch. 22, 59a5-8: metaphor most important element of poetic language. But ch. 22, 58a18-20, on need to balance clarity and dignity (from non-standard usage): despite 58b18-24 (Euripides adds dignity to Aeschylean line by substituting poetic word) would Aeschylus' style be too difficult (cf. Aristophanes Frogs 923-30)?

4. Reasoning: Aeschylus generally less rhetorical than later tragedy; but n.b., in Eumenides trial-scene, Apollo's speech, Athene's subsequent attempt to win over Furies.

5. Character: In Agamemnon, note Watchman as good and appropriate slave. Clytaemnestra bad: has virtues (courageous, resourceful etc), but not those of a woman (n.b. Aga. 11, 351: 'like a man'). But this is necessary to plot, so justified (cf. ch. 15).

6. Plot: (i) reminder of some key concepts:

- plot not identical with play;
- change of fortune (good to bad, bad to good): ch. 7, 51a11-15, cf. ch.13;
- complication (up to change of fortune), resolution (from beginning of change of fortune): ch. 18, 55b24- 32;
- simple plot (change of fortune) vs complex plot (change of fortune plus recognition and/or reversal): ch. 10;
- recognition, reversal: ch. 11 (and n.b. on astonishment: ch. 9, 52a1-11);
- suffering: ch. 11, 52b9-13; may be averted: ch. 14;
- connection in accordance with necessity or probability: ch. 7, 51a12f. etc; handling of irrationalities: ch. 15, 54b6-8; ch. 24, 60a26-b2; ch. 25, 60b23-6
- closure (beginning/end): ch. 7, 50b23-34;
- magnitude (right scale of plot for tragedy): ch. 7, 50b34-51a15 (cf. ch. 18, 56a10-19; ch. 24, 59b17-20);
- best kind(s) of tragic plot: ch. 13, ch. 14.

7. Plot (ii) Agamemnon

Much material 'outside the play' (e.g. sacrifice of Iphigeneia, story of Atreus and Thyestes).

Change of fortune: powerful, rich, victorious king killed. No recognition; is there a reversal? (Not much of one? Can simple/complex be a matter of degree?)

Change of fortune begins with confrontation with Clytaemnestra: so complication up to this point, then resolution.

Does Agamemnon's change because of an 'error'? E.g. trusting Clytaemnestra; or (further back) sacrifice Iphigeneia - or is this wicked? (If an error is part of a necessary or probable series, should we insist on a single, crucial error, or recognise a series of errors?)

Is the concept of necessary or probable connection too rationalistic? Agamemnon implies indeterminate forces in the background; and do supernatural forces (gods, inherited curse) fit into Aristotle's system?

Can Aristotle account for (e.g.) scene with Cassandra (is it essential to plot?); or does concentration on abstracted plot rather than structure of text leave some sections of text unexplained?

Play is open-ended, pointing to sequel: trilogy format requires modification of 'closure'.

8. Plot (iii) Libation Bearers

Trilogy: plot of the Libation Bearers includes whole plot of Agamemnon.

Recognition of Orestes and Electra (not tied to change of fortune, so does not imply complex plot). N.b. ch. 16 (55a4-6) on this as example of recognition arising 'from inference': but rest of para. is about someone's identity revealed by their reasoning, not reasoning used to discover someone's identity. Is this in fact recognition using tokens (hair, footprint, embroidery) - the least artistic kind (ch. 16, 54b20-30)?

Second recognition: Clytaemnestra recognises Orestes (part of the change of fortune, implying complex plot). Recognition of someone closely related - by victim: so imminent violence not averted (contrast ch. 14): perpetrator already knew. Does this cause disgust? Or does fact Orestes is compelled to do this cause pity? Cf. ch. 13, 53a18-22.

Double outcome (cf. ch. 13, 53a30-39)? Wicked change to bad fortune, Orestes and Electra escape oppression (so complication includes events before play, reunion of Orestes and Electra, preparations for vengeance; resolution starts with the implementation of deception-plan): but Orestes, pursued by Furies, in fact falls from bad to worse fortune.

Hence play lacks closure: form trilogy has implications plot-structure not recognised by Aristotle.

9. Plot (iv) Eumenides

Change to good fortune (= Orestes' acquittal; so complication = plots of first two plays plus pursuit by Furies, appeal to Athene, setting up of the trial; resolution begins with trial).

After Orestes' acquittal Furies threaten Athens:

(i) this is necessary/probable consequence of Orestes' acquittal (Furies would react thus);
(ii) attention shifts of attention from Orestes to Athens as potential tragic victim: being about one person is not enough for unified plot (ch. 8, 51a16-30) - is it also unnecessary?
(iii) preference for concentrated plot (as Iliad) over 'single action of many parts' (ch. 23, 59a30- 59b2): does extended action of Eumenides become 'of many parts' (acceptable, but less good)?

[Aristotle's Poetics: Introduction]