No CrossRef data available.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 11 February 2009
In Poetics chapter 9, Aristotle claims that the poet's function differs from the historian's. The historian should describe what has happened, but the poet should say ‘what sorts of thing might happen, that is, the things possible according to likelihood or necessity’ (1451a36–8). The difference is not between fiction and non-fiction. Some past events happened according to likelihood and are thereby candidates for poetic representation (1451b29–32). Rather, the poet differs from the historian with respect to the level of abstraction at which he considers the actions and experiences of agents. The historian should engage in accurate and thorough ἱστορ⋯α—‘research’ or ‘fact-gathering’—by carefully recording his or others' observations of particular events. The poet, on the other hand, looks for causal relations among fictional or non-fictional events, for he cares whether his composition has a plot with events that happen because of other events and not merely after them (cf. 1452a20–l). The poet may discern in the historian's materials some causally related events fit for dramatic or epic representation, but that is not the historian's concern. As Aristotle says in chapter 23, the historian should report ‘whatever befell one or more people during a particular period of time, each of the events relating to the others by chance’ (1459a23–4). The last clause is an overstatement; we have just noticed Aristotle's admission that some past events were likely to happen. Also, he elsewhere says that ‘future events will for the most part be like past events’, presumably because they share a similar causal structure (Rh. 1394a8). The remark's point is clear enough, though: the historian should report what happened whether or not the events exhibit explanatory coherence.
1 Translations from the Oxford Classical Texts are my own. Abbreviations of the titles of Aristotle's works are from Liddell, H. G., Scott, R., AND Jones, H. S. (edd.), A Greek-Engfish Lexicon (Oxford, 1940; 9th edn, repr. 1992), p. xixGoogle Scholar
2 Aristotle's concern for accuracy appears in his criticism of Herodotus at H.A. 523a 15; cf. Herodotus III.101.
3 Aristotle compares the aggression of past Persian kings with what to expect from the present Persian king (Rh. 1393a32–b4).
4 How could Aristotle seem to have got Herodotus, his sample historian, so wrong (cf. 1451b2–4)? Herodotus tells stories, after all. Unfortunately, Aristotle's remarks on historiography are scant, but we can hazard a response based on the difference of function (
, 1451a37): the historian's is to make an accurate report. If he also aims to tell a good story, he functions as a poet.
5 Indeed, even in chapter 9 comedy is used as an example of composing at the level of the universal instead of the particular (1451b8–15)
6 Metaph. 993b 19–20; cf. Top. 105b30–l. This contrasts with Aristotle's narrow conception of philosophy as the science of being quabeingMetaph. IV
7 E.N. 1094a28, Top. llla 37–
8 E.N 1099a23, 1146a 15, 1098a9, 12, 14, 1137 M, 5, 10, 1152a 21; Top. 152a8, 9, 28; E.N. 1170b 19, 1176b 19, 1174b22, 25
9 The close connection also appears at 1448b34 (read in the context of 1448b24–27), 1449blO, and 1449b24
10 See, for example, Janko, R. (trans.), Aristotle, Poetics (Indianapolis, 1987), p. 12, and Halliwell, S. (trans.), The Poetics of Aristotle (Chapel Hill, 1987), p. 41.Google Scholar
11 Besides the greater excellence of the objects of a science, de Anima 402al–3 claims that the marvellousness of a science's object and the greater precision of a science in investigating its object are also reasons for holding one science to be more fine and honourable than another
12 Bywater, I., Aristotle on the Art of Poetry (Oxford, 1909), p. 189; Gudeman, A., Aristoteles Poetik (Berlin/Leipzig, 1934), p. 207; Janko (n. 10), p. 92Google Scholar
13 Halliwell (n. 10), p. 106.
14 Halliwell, S., Aristotle's Poetics (London, 1986), p. 106. Although Halliwell in a later article says that poetic universals are not ‘straightforwardly formulable moral truths’ (‘Pleasure, understanding, and emotion in Aristotle's Poetics’, in Rorty, A. O. [ed.], Essays on Aristotle's Poetics [Princeton, 1992], pp. 241–260, at p. 251). In the same article he also denies that they are ‘abstractions’ (p. 249), saying, ‘The poet does not deal in abstracted universals, as the philosopher does’ (p. 250). But while poets and philosophers do not deal with universals in the same way, for example, poets do not attempt deflnitions of them, poets do deal with universals in the form of plots and the incident- and character-types that constitute them. See below.Google Scholar
15 Woodruff P., ‘Aristotle on mimesis’, in Rorty, A. O. (ed.), Essays on Aristotle's Poetics (Princeton, 1992), pp. 73–95, at pp. 86–8.Google Scholar
16 Lucas, D. W., Aristotle, Poetics (Oxford, 1968), p. 120.Google Scholar
17 Ibid., p. 119.
18 Butcher, S. H., Aristotle's Theory of Poetry and Fine Art (New York, 1907; 4th edn, repr. 1951), pp. 163, 194.Google Scholar
19 Else, G. F., Aristotle's Poetics; The Argument (Cambridge, MA, 1957), p. 305.CrossRefGoogle Scholar
20 Heath, M., ‘The universality of poetry in Aristotle's Poetics’, CQ 41 (1991), 389–402, at p. 389.CrossRefGoogle Scholar
21 Woodruff (n. 15), p. 88.
22 Although lacking an article, it would seem that the xadoXov
of T2 is a singular substantive serving as the antecedent of the singular relative 
at the beginning of the phrase translated as ‘which is what poetry aims at’. It also parallels the singular 
of the next sentence. Although I prefer the singular reading of 
in T2, it is possible that 
refers to the whole idea of the preceding clause and that each of the sorts of thing that a certain sort of person says or does is what Aristotle here calls a universal. In this case, a unified plot representing ‘one action’ would be a structure of universals causally linked by likelihood or necessity. I thank CQ's anonymous referee for this observation.
23 It is possible that Aristotle sometimes substitutes 
for the more common 
and 
to signify the particular proposition, i.e. one that predicates or denies a property of some members of a class (cf. A.Pr. 24a 17–19). However, the scant evidence for this possibility is sometimes ambiguous between prepositional and ontological readings of the terms (cf. Rh. 1357b 1–3, 1359a24). Independent of this issue, however, remains the fact that the claim about Alcibiades in T2 is not technically a particular proposition
24 For another use of an event-type as a universal, see A.Po. 94a36–b8 (cf. Herodotus V.97–102).
25 The ordering of the parts is explicitly mentioned as a requirement in Aristotle's formal definition of ‘whole’ at Metaph. 1024al–3. If the order of a thing's parts does not matter, then that thing is only a ‘totality’ and not a whole.
26 Or, in this case, by necessity or by ‘what happens for the most part’ (1450b30). I take Aristotle's use of ‘naturally’ in this passage to be ambiguous between necessity and likelihood. In later passages, Aristotle drops ‘naturally’ as the description of how one incident in a plot ought to follow another and uses ‘by likelihood or necessity’ instead
27 Thus Aristotle has a principle of action individuation different from the one Bittner uses in his interpretation of Aristotle's ‘one action’. Bittner claims that ‘an action stands out as an action … [b]y being meaningful. The doings of a person coalesce into individual actions discrete by virtue of separately making sense. One action is what you get if you cut a person's doings into the smallest meaningful parts’ (‘One action’, in Rorty, A. O. [ed.], Essays on Aristotle's Poetics [Princeton, 1992], pp. 97–110, at p. 99). With this principle in hand, Bittner argues that the unity of an action comes ‘too cheap … or else it is incomprehensible’ (p. 103). But he nowhere argues that this is how Aristotle individuates actions, and Aristotle explicitly endorses another principleGoogle Scholar
28 See, for example, 1451a12–13, 27–8, b 13, 31, 1452a 20, 24, 1456b 4
29 In the case of an actual epic, however, Aristotle says that Homer was justified in omitting mention of Odysseus' wounding on Parnassus and his pretence of insanity during recruitment since these events are not linked by necessity or likelihood to the events of the Odyssey (Po. 1451a22–8).
30 E.g. A.Po. 1.30; Ph. U.S. Also relevant is Aristotle's claim that action is the subject of deliberation, and we do not deliberate about matters which could not be otherwise, i.e. those that occur by necessity (Rh. 1357a23–7; E.N. 1140a35–bl).
31 Contra Butcher (n. 18), who claims, ‘The 
of [T3] denotes the broad outline, the bare sketch of the plot, and is wholly distinct from the 
of [T2], the general or universal truth which poetry conveys’ (pp. 193–^4).
32 The latter sort of connections are mentioned at 1454a33–7 and in T2
33 A passage in Poetics 6 might be understood as evidence against this conclusion. There Aristotle argues that plot is the most important part of tragedy, even more important than character. ‘The tragedy is a representation not of human beings but of actions and life’, Aristotle explains (1450a16–17). As an apparent consequence of this, he says that ‘tragedy cannot exist without action, but without characters it may’ (1450a24–5). It is difficult to understand what this could mean, and Aristotle's examples do not clarify matters. He says that the tragedies of his own day often are without character and that the paintings of Zeuxis, in contrast to those of Polygnotus, contain no character at all (1450a25–9). We, however, do not have any ‘characterless’ tragedies from the mid-fourth century, nor do we have any of Zeuxis' or Polygnotus' paintings. We might, however, guess that by ‘character’ here Aristotle means the rather narrow notion of ‘that which reveals decision [
]’ (1450b9). In that case, it would not be accurate to say that the characters of the tragedy have no personality traits, but rather that they are not shown making choices that reveal deliberation about the right action to take (E.N. III.2–3).
34 Earlier versions were presented at the 1997 Eastern Division Meeting of the American Philosophical Association in Philadelphia, at the 19th Annual Workshop in Ancient Philosophy at Texas A&M University, and to the Euthyphrones at the University of Texas at Austin. I sincerely thank Julia Annas, Thomas Christiano, Christopher Colvin, Daniel Graham, Hankinson R. J., Martha Nussbaum, Gregory Scott, Joseph Tolliver, Stephen A. White, Paul Woodruff, and an anonymous referee of Classical Quarterly for all of their very helpful comments.