¹ I am grateful to D. Browning, H. Granger, S. Waterlow, and the editors of CQ for their comments on earlier versions of this paper. They cannot be blamed for its faults, but they have helped to improve it.

² Compare Nicomachean Ethics 1105b21–3, Rhetoric 1378a20–3, but then notice Magna Moralia 1186a12–15.

³ Woods, M., Aristotle's Eudemian Ethics, a translation and commentary (Oxford University Press, 1982), especially pp. 109–10Google Scholar.

⁴ See, for example, Prior Analytics 43b30 ff., Posterior Analytics 87b18–26, 96a7–19.

⁵ W. W. Fortenbaugh's discussion of our passage implies that we do not need to favour either alternative: Aristotle, wants to maintain both (‘Arius, Theophrastus and the Eudemian Ethics’, pp. 208–12Google Scholar in Rutgers Studies in Classical Humanities, 1). But the arguments I shall offer against Woods's interpretation (2) show not only that it is not sufficient, but also that it runs counter to too much that Aristotle holds. Hence embracing both interpretations is mistaken.

⁶ See, for example, the passages mentioned in note 2.

⁷ Woods does not suggest (nor should he) that on those occasions in which pleasure or pain is not felt, something else is. Pleasure or pain is what is felt here if anything is, according to Aristotle.

⁸ One minor qualification here. De Somniis 460b32–461 a3 makes room for not noticing, say, the pain of anger because, say, the pain of a nail in one's foot is so overpowering. However, this does not help Woods's position. For it will remain the case that this pathos is accompanied by pain.

⁹ I observed at the outset that the disadvantage of this interpretation (1) compared to Woods's (2) is the loss of a foolproof criterion for species of pathē. We should notice that this point was not lost on Aristotle, as the dropping of epi to polu from the characterization of ta pathē in the Rhetoric and the Nicomachean Ethics would suggest. See also Nicomachean Ethics 1104b 13–15.

¹⁰ How many other pathē are without feeling is unclear. By the characterization in the Rhetoric loving and kindness are not painful or pleasant, though realizing or failing to realize their end may be. Since desires are considered pathē in the Eudemian Ethics, we might also want to include boulēsis. (For a fuller discussion of the different uses of ‘ta pathē’ as employed in the Ethics versus the Rhetoric and the force of ‘accompanied by pleasure and pain’ see my paper ‘Aristotle and the Emotions’, Phronesis, September 1982.)

¹¹ Elsewhere I have tried to explain why Aristotle might have made this move. (See note 14 of my article, mentioned above.) A very different and interesting explanation is offered by Fortenbaugh (ibid.). I should add that a bodily nature seems to be a very central component to a pathos (cf. Nicomachean Ethics 1128b 10–15).