[Aristotle's Poetics: Introduction] [Aristotle's Poetics: Seminar Notes]


Notes on Homer's Iliad and Odyssey

1. What does Aristotle say?

Aristotle makes the following references to Homer:

(i) ch. 2 (48a11-12): object: Homer imitates men 'better than we are';
(ii) ch. 3 (48a21-22): mode: narrative, but with impersonation of the characters;
(iii) ch. 4 (48b34-49a2): Homer's excellence he uniquely achieves pre-eminence in both the serious and comic traditions of poetry, and his quasi-dramatic style points the way to tragedy and comedy in the strict sense;
(iv) ch. 8 (51a22-30): Homer's excellence: unity of plot (contrasted with defective plots based on a single person);
(v) ch. 15 (54b1-2): an inappropriate use of divine intervention in the Iliad;
(vi) ch. 15 (54b14-15): Homer's portrayal of Achilles;
(vii) ch. 16 (54b25-30): the use of the scar in the recognition of Odysseus by the nurse (combined with reversal) and the swineherds ;
(viii) ch. 16 (55a2-4): the recognition of Odysseus in Alcinous' palace;
(ix) ch. 18 (55b15-23): episodes in epic and the plot of the Odyssey;
(x) ch. 23 (59a30-b7): excellence in plot-construction (contrasted with defective plots constructed like a work of historiography and with plots that are well-formed but have ‘many parts’);
(xi) ch. 24 (59b12-16): the Iliad and Odyssey compared;
(xii) ch. 24 (59b12-16): the excellence of Homer's quasi-dramatic style;
(xiii) ch. 24 (60a18-26): a Homeric model for the handling of irrationalities: false inference. (xiv) ch. 24 (60a34-b5): Homer's ability to conceal irrationalities by other good qualities.

Some general comments on epic:

(i) ch. 5 (49b9-20): a general introduction;
(ii) ch. 18 (49b9-20): tragedy and epic: the difference in scale;
(iii) ch. 24 (59b16-60a2): differences between tragedy and epic;
(iv) ch. 24 (60a11-18): astonishment in epic;
(v) ch. 26 (62a18-b11): epic length may result in dilution and loss of unity.

In addition, the discussion of problems and solutions (ch. 25) is mainly concerned with points from the Homeric poems. Aristotle discussed this kind of material more extensively in his Homeric Questions, of which only a few fragments survive; a selection of these fragments is available in the collection of supporting texts.

2. What do Aristotle's theories imply?

Homer's excellence

1. One thing that Aristotle admires in Homer is his use of direct speech (see especially ch. 24, 60a5-11). Epic is a narrative form, but by allowing his characters to speak for themselves Homer makes his epic narrative approximate to drama: he thus foreshadows the later development of drama which he regards as a superior form of poetry (ch.4, 48b34- 49a6). Why? The passage in praise of Homer's use of direct speech gives two pointers. First, the poet speaking in his own voice as narrator 'is not what makes him an imitator'. This may seem to contradict Aristotle's claim that all poetry - including epic - is by its very nature imitation; but it is in fact perfectly consistent to say that epic and tragedy are both forms of imitations, but that tragedy is more genuinely imitative. Secondly, Aristotle comments that all of Homer's characters 'have character'. Since speech is, for Aristotle, an important (perhaps, the most important) vehicle for the imitation of character (ch. 6, 50b8-12), Homer's use of direct speech opens up possibilities for the imitation of character absent in pure narrative. In particular, letting the characters speak for themselves makes it possible for the poet to make clear, not just what happened, but why it happened: what attitudes and dispositions motivated the person to act like that. (Thus, for example, at the beginning of the Iliad we do not just learn that Agamemnon refused Chryses’ request for his daughter’s freedom: because we hear Agamemnon’s own words we see how that refusal is rooted in the kind of person he is.)

2. Aristotle also admires Homer's skill in plot construction - something which is of course central to his view of what is important in poetry (though Aristotle’s enthusiasm for Homer’s use of direct speech to convey character shows that his arguments for the primacy of plot do not imply that character is unimportant). There are two levels to this. First, Homer has avoided the mistake made by many other epic poets of giving his epic the structure of a biography (telling the story of a single hero: ch. 8, 51a16-30), or of a historical narrative (telling the story of everything that happened in a given period of time: ch. 23, 59a17- 30). A single person's life, or the history of a single period of time, will contain lots of unconnected events; so the plot in such epics will not consist of a series of events linked to each other in accordance with necessity or probability.

3. But necessary and probable connection is only a minimum requirement for a well-formed plot; so it is not saying very much that Homer satisfies this condition - and Aristotle recognises that there are other poets who pass this test. What makes Homer unique (ch. 23, 59a30-59b7) is the concentration of his plot on a very short sequence of events: most epic poets construct plots that are connected, but have many parts; Homer selects in a single section (the anger of Achilles and its consequences) out of a larger connected story (the story of the whole Trojan war) and makes his plot out of that. We will see below how Homer goes about this.

4. Aristotle goes on to comment on Homer's ability to handle all the elements of poetry (different kinds of plot, character, reasoning, diction) well. This passage (ch. 24, 59b7-16) includes a comparison of the two Homeric epics: the Iliad is simple and based on suffering, the Odyssey is complex and based on character. This formulation is (of course) simplistic, and leaves much unsaid; but it is obviously not intended as a full analysis of the two epics (the question to ask is not whether Aristotle has produced a satisfactory account of the two poems, but whether it is posible to do better in one short sentence). The characterisations of the two poems are worth looking at more closely.

Odyssey: 'complex and based on character'

5. 'Complex', in Aristotle's terminology (ch. 10, 52a12-21), means that the plot has recognition and/or reversal. In the case of the Odyssey, as Aristotle says, 'recognition pervades it': Telemachus identifies himself to Nestor (book 3) and is recognised by Helen in Sparta (book 4); Odysseus makes himself known to the Phaeacians (book 9), and to Telemachus (book 16); he is recognised by the old nurse Eurycleia (book 19); and he reveals his identity to the herdsmen (book 21), to the suitors (book 22), to his wife (book 23), and to his father (book 24).

6. We may note in passing that none of these recognitions are, according to Aristotle's classification, technically very sophisticated. The recognition of Odysseus among the Phaeacians is an example of a kind ranked relatively high ('by means of memory': see ch. 16, 54b37-55a4; Aristotle is referring to the way Demodocus' song about the sack of Troy moves Odysseus to tears, prompting Alcinous to enquire about his identity). But Eurycleia and the herdsmen recognise Odysseus by the scar - an example of the least artistic kind of recognition, using physical tokens; but Aristotle points out that in the case of the nurse the recognition is combined with a reversal, since Odysseus' cover is unexpectedly blown by his own choice to be bathed by the old woman (ch. 16, 54b20-30).

7. 'Based on character' cannot mean that characterisation is absent or unimportant in the Iliad - Aristotle’s enthusiasm for the way that Homer gives all his characters character by letting them speak proves that. The most likely explanation lies in the discussion of the best kind of tragic plot the Odyssey is the example of the second best kind - that with a double outcome, in which the good characters end happily and the bad unhappily (ch. 13, 53a30-33). For in this kind of plot it is moral character that determines the outcome.

8. Just as saying that the Odyssey is based on character does not mean that there is no character in the Iliad, so too the contrast with the Iliad does not imply that there is no suffering in the Odyssey (one need only think of the fate of Odysseus' crew, to say nothing of the suitors). Aristotle is only claiming that character and suffering are key elements in their respective poems in a way that they are not in the other.

Iliad: 'simple and based on suffering'

9. 'Suffering' is a technical term for Aristotle: 'an action that involves destruction or pain' (ch. 11, 52b9-13), such as death, wounding and physical agony. There are many deaths and woundings in the Iliad. One of the striking features of Homer's technique is how even minor figures are brought into focus at their death in a way designed to create a sense of the significance and pathos of their death. This example - the first to come to hand - illustrates the point well (Iliad 5.152-8): Diomedes 'went in pursuit of Xanthos and Thoön, sons of Phainops, both children late-born and loved: but he was worn by cruel age, and could father no other sons to leave over his possessions. So then Diomedes killed them, and took the dear life from them both, leaving lamentation and cruel sorrow to their father, when he did not welcome them alive back from the battle: and distant relatives divided the inheritance'. And of course there are other deaths, more elaborately prepared and narrated, with a more central place in the plot - notably the deaths of Patroclus and Hector.

10. The deaths of Patroclus and Hector - like the death of Xanthos and Thoön and many others - have an impact on those left alive. Is their grief and distress included within Aristotle's concept of suffering? The focus of his examples on physical suffering might suggest not: and that might be thought be a major shortcoming. Alternatively, you could point out that Aristotle's examples end with an 'and so on' of undefined scope: can we get emotional suffering in through that opening? (It would be surprising if Aristotle did not see the importance of Achilles; grief and sense of guilt.) Another possibility is to see the physical suffering and its emotional impact as a unity: death is tragically significant (in part) because of the emotional impact on the survivors; the representation of their grief and distress is what gives tragic significance to the merely physical facts. In this sense, we might say that while Aristotle's term refers to physical suffering, it does not exclude the emotional suffering of the other characters, since that is bound with the physical suffering.

11. Aristotle says that the Iliad is simple, i.e. that there is no recognition or reversal. It could be been objected that the death of Patroclus does involve reversal: Achilles' actions lead, contrary to his expectation, to the very thing he least wanted - the death of his dearest comrade. Is this analysis right? Achilles’ actions do achieve what he aimed at — the restoration of his position within the Greek army. Note too that, immediately before Patroclus set out for the battle, Achilles specifically warned him not to follow up his success too far (Iliad 16.87-96, cf. 16.685-7: 'if he had kept to the instruction of the son of Peleus he would have escaped the vile doom of black death'; 18.13f.). So the death of Patroclus is not exactly an unexpected outcome of Achilles' actions: on the contrary, he foresaw and tried to prevent the danger to Patroclus. It is true that Patroclus' death is a consequence of Achilles' withdrawal from the fighting (had it not been for that, the situation in which Patroclus exposed himself to such danger would not have arisen); but that is bound to be the case if the plot consists of a series of connected events. A reversal must be more than just a by- product of someone's actions, however unwelcome. So it is could be argued that Aristotle was right on this point. (On the other hand, Aristotle suggests that the recognition of Odysseus by his nurse does involve a reversal: see 54b29-30. This ‘reversal’ seems to be no more than unexpected and undesired side-effect of Odysseus’ reluctance to be washed by one of the younger servants.)

Homer and the best kind of tragic plot

12. The fact that the Iliad has a simple plot means that it is does not fully match up to the ideal tragic plot of ch. 13, since Aristotle starts his analysis from the observation that the best tragedy will have a complex rather than a simple plot (52b30-32). In this respect, the complex plot of the Odyssey must be technically superior to that of the Iliad. On the other hand, we have already seen that the double outcome of the Odyssey makes it an instance of the second best kind of tragic plot; so the Iliad might be superior in this respect - if it corresponds to the other requirements of the best kind of tragic plot. According to Aristotle's analysis in ch. 13 (52b34-53a22), the best kind of tragic involves a change from good fortune to bad fortune on the part of someone who is neither outstandingly virtuous nor wicked, who falls into misfortune not because of wickedness but because of an error (hamartia). Can this pattern be applied to the Iliad?

13. We have already seen that Patroclus' death is the result of his failure to heed Achilles' warning. This would certainly seem to qualify as an error.

14. The series of decisions which brings Hector to the fatal confrontation with Achilles is also worth considering closely. In book 6, Andromache asks him to stay with her on the city wall (6.431f.); he refuses, as he must - it is his duty to fight. In book 14 Polydamas advises against an assault on the Greek camp; Hector overrules him. In book 18 Polydamas (knowing that Achilles will rejoin the fighting the next day) argues that the Trojans should withdraw inside the city and defend the walls; Hector again overrules him. In the next day's fighting the Trojans suffer a heavy defeat. They retreat inside the city; Hector is the last man left outside the walls; his parents, on top of the walls, beg him to come inside; but conscious that the disaster was due to his own misjudgement, he feels obliged to stay outside and face Achilles. He too, then, falls into misfortune as a result of an error - the rejection of Polydamas' advice in book 18. (Polydamas plays the role, common in Greek literature, of the adviser whose warnings are ignored.)

15. Hector is fighting on the wrong side; but that is not his fault (he has little sympathy for Paris' misdeeds) - he is simply doing his duty in defending his people. He is pious (consider how he goes back to the city in book 6 to arrange sacrifices to appease the gods); he is a good soldier; he resolutely refuses the chance to shirk his duty while in the city; and yet seeing him with his wife and baby shows him as a good husband and father as well. So he seems to be a virtuous man; does that mean that Hector is too virtuous to be a tragic figure by Aristotle's standards? Yet his refusal to listen to Polydamas' advice is not just an intellectual error (a miscalculation of the tactical situation); it expresses a shortcoming in his moral character - over- confidence and recklessness. This does not mean that he is a bad man; but he is not entirely blameless either. So this case supports the view that Aristotle's concept of error can include moral errors that fall short of depravity as well as intellectual errors.

16. A question that might be considered in passing is what bearing it has on our view of Hector that he is Trojan. In fifth-century tragedy the Trojans are often presented sympathetically (likewise Aeschylus' Persians is able to see the war from the perspective of the enemy who had sacked Athens not many years before, as a tragedy for them as well). Most would see Homer as equally able to treat both sides sympathetically (after all, in book 24 Priam and Achilles can in some measure transcend the fact that they are enemies). In the centuries after Aristotle, commentators on Homer often took the view that Homer was pro-Greek, and presented the Trojans (including Hector) in a bad light; but there is nothing in the Poetics or the fragments of the Homeric Questions to suggest that Aristotle took a similar view.

17. Of course, we must not forget Achilles: he too undergoes a change to bad fortune (cf. 19.321f.: 'There could be no worse suffering more me, not even if I heard of the death of my father...') as an indirect consequence of his quarrel with Agamemnon, which in retrospect he regards as folly. It is obvious enough that Achilles is not morally perfect; but Aristotle, at least, seems to regard him as a good man - Homer's portrayal of Achilles is his example of how one can incorporate shortcomings in an idealised portrait (ch. 15, 54b8-15: but note that the text of the reference to Homer and Achilles is confused, and we cannot be certain what Aristotle said at this point).

18. It seems, then, that the pattern which Aristotle prescribes for the best kind of tragic plot recurs several times in the Iliad. It is interesting that ch. 13, although it is not overtly concerned with epic, gives an account of the two Homeric poems - the Iliad as the best kind of tragic plot, the Odyssey as the second best kind.

19. The analysis of tragic plots in ch. 14 has less relevance to Homer. One point to note is that the deaths of Patroclus and Hector are examples of violence inflicted by enemy on enemy, 'there is nothing pitiable either in the action itself or in its imminence, except in respect of the actual suffering in itself' (ch. 14, 53b17-18). Does the example of the Iliad call this judgement into question?

Plot and poem

20. The plot of the Iliad develops briskly in book 1: the quarrel occurs, Achilles withdraws and ensures, through his mother's intervention with Zeus, that the Greeks will do badly in his absence; accordingly Zeus sends a dream to Agamemnon at the beginning of book 2 to trick him into giving battle; the consequential next step in the development of the plot is a Greek defeat. What in fact happens? Agamemnon puts the army to the test, with results that are almost disastrous; but in the end (thanks to divine intervention) nothing comes of this. Then we have the Catalogue of Ships and the corresponding Trojan Catalogue. Then Paris challenges Menelaus to a duel; this could lead to the settlement of the war; but in the end (thanks to divine intervention) nothing comes of this either - Paris is rescued and the truce broken. Then we have a proper battle (ending with another duel, between Hector and Ajax: but nothing is at stake in this duel, and it is curtailed without a result). Contrary to what we might have expected, the Greeks are extremely successful in this day's fighting - so much so that the Trojans send envoys offering to settle the dispute; but these overtures are rejected, so nothing comes of this either. It is not until book 8 that we get the Greek defeat which the plot demands. Similarly in later parts of the narrative. The nocturnal raid on the Trojan camp in book 10 does not advance the plot. In the next day's fighting the Trojan advance is reversed by the intervention of Poseidon, aided by Hera's seduction of Zeus; but since Zeus eventually notices what is happening and orders Poseidon out of the battle, nothing comes of this short-lived Greek success.

21. It seems, then, that in the Iliad there is a relatively small core of events which are essential to the progress of the plot, and round this is built a much larger mass of events that are not essential to the development of the plot. Here we come back to Aristotle's analysis of the unique excellence of Homer's plot construction: having selected a single section out of a larger connected story, he uses other material as 'episodes' to give the poem variety (Aristotle cites the Catalogue of Ships as an example: ch. 23, 59a35-37). Note that, in distinguishing the episodes from the core of the plot, Aristotle is not dismissing them as irrelevant or unimportant; the fact that they add to the poem's variety means that they make a positive contribution to the quality of the poem (Aristotle regards its greater variety as a point in which epic has an advantage over tragedy: ch. 24, 59b26-31). It is obvious from the summary above how much richer and more interesting the poem is as a result of those parts of the narrative which do not advance the core plot. We have criticised Aristotle often enough for his concentration on the plot in abstraction from the poem; but here we see that he was also aware of the poem as something which is more than the vehicle of a plot, and as something whose quality does not depend on the plot alone. (On the other hand, we might want to say that much of the material that is not essential to the plot is thematically important: this is something to which Aristotle pays little or no attention in the Poetics: is that a serious flaw? Does it mean that he is oblivious of something that is essential to literature?)

22. Compare the discussion of the Odyssey in ch. 17 (55b15-23). Here too we find the distinction between the limited core of plot and the expansive episodes.

Epic and tragedy

23. We have seen how the discussion of tragic plots in ch. 13 can be applied to the Iliad and Odyssey. Aristotle seems to think that, broadly speaking, epic and tragedy aim at the same kind of effect, so it is not surprising that the analysis of tragic plots is relevant to epic as well. But there are differences, as well as similarities, between epic and tragedy.

24. The differences include the different medium - narrative vs drama; we have already seen that Homer comes closer to drama than most epic in this respect. There is also epic's greater scope for variety, mentioned above, and the greater for astonishment and irrationalities that comes from the fact that events are not enacted before the audience's eyes (ch. 24, 60a11- 17).

25. Perhaps the most important difference is in length.

(a) Epic works on a larger scale than tragedy (ch. 5, 49b12-14). This is true even of the Iliad, despite its highly concentrated plot: we have seen how a typical tragic plot-pattern recurs several times. Aristotle comments that the plot of any epic, including the Iliad, contains a multiplicity of stories suitable for tragedy (ch. 18, 56a10-19); the greater length of the epic text gives scope for developing this multiplicity of stories on an appropriate scale. On the other hand, when he is emphasising the concentration of the plot of the Iliad, he suggests that unlike other epics it would yield only one or two tragedies (ch. 23, 59b2-7).

(b) In ch. 24 (59b18-22) he suggests that the optimum length for epic would be 'shorter than those of the ancient epics', matching 'the number of tragedies presented at one sitting'. If this refers to the length of the tragic text, it is a veiled criticism of Homer: the Iliad has more than 15,000 lines, the Odyssey more than 12,000 lines, and the recommended length would be 4-5,000 lines. (Most of the other early Greek epics we know of were much shorter than the Iliad and Odyssey, and thus closer to what Aristotle recommends.) An alternative view might be that he is referring to the scale of the epic plot, not of the epic text; this would mean that the ideal is not a certain number of lines of verse, but a plot that can be decomposed into not more than three tragic plots - so that the Iliad (yielding one or two plots) is better than non-Homeric epics like the Cypria and Little Iliad, just as Aristotle has said in ch. 23. (But why, if that is what he means, does Aristotle not name Homer as exemplifying his ideal?) Or perhaps there is a stronger sense of contrast with the following sentence: one way of ensuring that the whole plot can be taken in at once is to keep the epic text shorter than those of Homer, but epic also has resources unique to itself that make it possible to achieve expansiveness without going beyond what can be kept in memory. (This has the advantage of being consistent with the enthusiasm shown in the previous chapter for the way Homer expands a plot that concentrates on a concise series of events.)

(c) There is other evidence of ambivalence in Aristotle's views on this matter. In ch. 7 (51a9-11) he said that, within the upper limit imposed by the requirement that it should be possible to hold the whole plot in memory at once, the larger the plot the better. Yet in ch. 26, when he is arguing for the superiority of tragedy over epic, he claims that tragedy has an advantage in being shorter, more concentrated and more unified. Is this consistent? (Some of the things Aristotle says in ch. 26 seem more like debating points than serious arguments; and he does not mention the feature of tragedy which seems to be the basis of its superiority to epic in ch. 4 - its dramatic medium. But we cannot simply say that this chapter is an independent fragment tacked onto the end of the Poetics: since there was probably a second book, this is the middle, not the end, of the work as a whole.)

(d) At the very least, we can say this: the greater expansiveness of epic, though it has the potential advantage of making the poem more impressive and grandiose, involves risks: if you increase the number of events so as to fill up the narrative text available, you risk making the plot too complex for the audience to keep in memory at once; if you keep the number of events moderate, you risk boring the audience by spending too much time on each event. Homer’s solution - a plot that concentrates on a very small number of events, plus a lot of very varied ‘episodes’ to expand the text - is the best possible solution for an epic poet.


26. At first it seems that Aristotle has a very strict theory of plot: only events that have a necessary or probable connection meet the minimum requirements for an acceptable plot. But later in the Poetics it emerges that he is more flexible than that implies. Tragedy, for example, can get away with irrationalities if they are outside the play (54b6-8, 60a27-30). Epic has more scope: because the events it narrates are not seen on stage, it is easier to conceal irrationalities - for example, the pursuit of Hector (60a11-17). Realistically, the lone Trojan outside the walls with the whole Greek army running round and round with one man in pursuit makes no sense. But because we do not see this happening, we do not find it obtrusively absurd; and Homer gains many advantages by doing it this way (and not, e.g., simply ensuring that Hector and Achilles meet each other in the course of the battle): for example, Hector is isolated (achieving pathos), the one-on-one confrontation is emphasised (achieving intensity), and Hector’s death is the end of the battle (we don’t have to get the rest of the Trojan army out of the way). In ch. 25, on how to handle ‘problems’ in poetry, the general principle is: something that is mistaken or irrational is acceptable if it makes part of the poem more effective (60b23-26).

27. So if a poet can get some advantage from departing from necessity and probability, and if he can conceal the resulting irrationality from his audience, he has Aristotle’s support. Homer, in particular, can get away with things that other epic poets could not get away with: for example, Odysseus’ arrival in Ithaca (60a34-b5). It is absurd that the Phaeacians do not wake him up, but simply dump him on the shore; and it is absurd that he does not wake up when this is done. But this gains many advantages (the pathos of Odysseus’ initial uncertainty and anxiety, the encounter with Athene, the recognition of home), and is made unobtrusive because Homer’s unique abilities as narrator distract us.


28. Some relevant discussions:

This page is maintained by Malcolm Heath, and was last updated on 4 May 2001.

[Aristotle's Poetics: Introduction] [Aristotle's Poetics: Seminar Notes]