^{page 47 note 1} Death or intense pain (11. 1452^bII, cf. Rhet. 2. 1382²22, 5^b13).

^{page 47 note 2} The untragic effect of (I) is seen most clearly in Euripides' Orestes.

^{page 47 note 3} Commentary, pp. 224–5.

^{page 47 note 4} Those who do not believe with Jaeger, (Aristotle, p. 238)Google Scholar that the E.E. is an earliesr work of Aristotle's, may still be interested to note the relation of the views which it contains to the psychology of drama. for the sake of convenience I shall allode to the author(s) of both Ethics as Aristotle.

^{page 48 note 1} 1224^b21–9 ὥοτε τ μν βᾳ κτερον ϕάναι ποιεῖν ἒϰει λόγον κα δι τν ρεξιν κα δι τν οογισμν κτερον ἄκοντ ποτε πράττειν κεϰωρισμένα γρ ντα κτερα κκρούεται ὑπ' γγγων. ὅθεν κα π τνὅγην μεταϕέρουσι Ψυϰήν, ὅτι τν ν Ψυϰῇ τι τοιοτον ρσιν. π μν οὖν τν μορων νδϰεται τοτο λγειν δ' ὅλη κοσα Ψυϰ κα το κρατος κα το γκρατοις πράττει βίᾳ δ' οὐδ τερος, γγ τν ν κενοις τι, πε κα ϕύσει μϕότεραἔϰομεν. The less accurate way of speaking is retained in Rhet. I. 1369^b5 βᾳ δ, ὅσα παρ' πιθνμίαν ἤ τοὺς λογισμοὺς γγνεται δι' αὐτν πραττντων.

^{page 48 note 2} 1225^a17 οὓτω γlρ ναγκαξμενος κα βᾳ πρξειἢ οὐ ϕύσει, ὃταν κακν γαθοἓνεκα ἤ μεξονος κακο πολύσεως πράττῃ, κα ἂκων γε ο γρ ϕ' αὑτῷ τατα.

^{page 48 note 3 a}20 δι κα τν ἒρωτα πολλο κοσιον τιθέασιν κα θυμοὺς νους κατ ϕυσικά,ὂτι σϰυρυ κα ὕπρ τν ϕύσιν' κα σνγγνώμην ἔϰομεν ὡς πεϕυκότα βιάξεσθαι τν ϕύσιν

^{page 48 note 4 a}25 τύ γρ ϕ' αύτῷ, ενγεται ὅλον, τοτ' στιν ὅ αὐτο ϕσιμ οἵα τε ϕρειν. ὅ δ μ οἴα τε μηδ' στ τςκενον ϕσει ρξεως ἢ λογισμο, οὐκ μϕ' αὕτῷ.

^{page 48 note 5 a}28–30. Cf. 8. 1248^b33 where some πιθυμία is itself a divine νθουσιασμός.

^{page 48 note 6 a}31 ὥστε κα δινοια τινες κα πθη οὐκ ϕ' μῖν εἰσν, ἢ πρξεις αἱ κατ τς τοιατας διανοας κα λογισμος, λλ' ὥσπερ Φιλόλαος ἒϕη εἶνα τινας λγους κρείττους μν. Cf. 8. 1246^b14 ἄν ἰσϰυρυἦ πιθυμία, στρέΨει κα λογιεῖται τναντα.

^{page 49 note 1} 1110^a23 π' νοις δ' ἒπαινος μν οὐ γνεται, συγγνώμη δ', ὅταν δι τοιατα πρξῃ τις ἃ μ δεῖ, ἅ τν νθρωπίνην ϕύσιν ὑπερενει κα μηδες ἂν πομεναι. Cf. 7. 1150^b6–9 (Philoctetes). Failure to resist pleasure or pain is a form of κρασα, but Aristotle denies that motives Compel merely because their source is external to the agent, 1110^b1–3, 9–11. Anger, though painful, is not a real exception (5. 1135^b26 οὐ γρ ἄρϰει θυμῷποιν, λλ ργσας).

^{page 49 note 2} 1110²26 ἕνια δ' ἴσως οὐκἕοτιν ναγκασθναι,λλ μλλον ποθανετέον παθόντι τ δειντατα κα γρ τν Eὐριπδον' Aλκμαίωνα γελοῖα ϕαίνεται τ ναγκσαντα μητροκτονσαι. Cf. 5. 1136^a13. Alcmeon seems to have been put on trial for his life at psophis and to have pleaded his father's command and curse of καρπα γς κα τεκνα if he should disobey, and various other less cogent excuses, Eur. fr. 67–70 (Nauck).

^{page 49 note 3} 1112^a32–3, Cf. 5. 1135^a32 τ δ γνον,ἢ μ γνοομενον μν μ π' αὐτῷ δ' ν, ἤ βᾳ, κοσιον. πολλ γρ κα τν ϕσει ὑπαρϰν των εἰδτες κα πρττομεν κα πσϰομεν, ὧν οὐθν οὔθ' κοσιον οὔτ' κοσιν στιν, οἷον τ γηρν ἤ ποθνσκειν.

^{page 49 note 4} Cf. 1117^a4, 7. 1147^a16, 24 ff.

^{page 49 note 5} 7. 1149^b4 ἔτι ταῖςϕυσικαῖς μλλον συγγνώμηκολουθεῖν ρξεσιν πε κα πιθυμαις ταῖς τοια ταις μλλον ὅσαι κοινα πσι, κα ϕ' ὅσον κοινα δ θυμς ϕυσικώτερον κα ϰαλεπότης τν πιθυ μων τν τς ὑπερβολς κα μ ναγκαων.

^{page 49 note 6} 7. 1148^b15–9^a20. Cf. 1149^b27 αἵ μν (sc. πιθυμαι κα δονα) νθρώπινα εἰσι κα ϕυσικα κα τῷ γνει κα τῷ μεγθει, αἵ δ θηριώδεις, αἵ δ δι πηρώσεις κα νοσμΡτΡ.

^{page 49 note 7} If an angry man wrongly thinks that he has received an injury (7. 1149^a32–4), or if he strickes too hard and kills the assailant he intends merely to repel, he acts γνον but δι πθος not δι'ἄγνοιαν. If the assailant is his unknown father, he acts δι' ἄγνοιαν, since his ingnorance is independent of his anger, or even the cause of it.

^{page 49 note 8} 1111^a1 ν τοτοις γρ κα ἕγεος κα συγγνώμη. γρ τοτων τι κουσως πρττει. Cf. ^a16 τοτων τι γνοσας ἅκων δοκεῖ πεπραλναι. Whether this excludes all ignorance due to πθος is not clear, but it excludes any for which the agent is himself to blame, 1113^b24, 30 ff. κα γρπ' αὐτῷ τῷγ7nu;οεῖας κολξουσι, ν αἵτιος εἶναι δοκῇ τς and unnatural πθος are also excluded, see note 1, p. 52 below.

^{page 50 note 1} E.N. 5–7 = E.E. 4–6. I take a view to be ‘later’ when it does not follow directly from opinions expressed in the earlier parts of E.E. but seems partly a development, partly a modification of the changes in the earlier parts of E.N.

^{page 50 note 2} See note 8, p. 49 above and 1, p. 52 below. Acts springing from abnormal irdOos may be involuntary apart from being done in ignorance, cf. 1135^a32, 7. 1148^b31 ff.

^{page 50 note 3} The need for Aristotle's insistence is shown by the sophistic composition Dialexeis 3. 9 (Diels, , Vors.⁵ ii. 410)Google Scholar ϕονεύεν δ τὼς ϕɩλτάτως δίκαɩον. πεì καί ʼΑλκμαίων. καί θεός ἓϰρησε δίκαɩα αὐτὼς ποɩσαɩ.

^{page 51 note 1} Apart from his rule for the use of the μηϰανή (Poet. 15. 1454^b3–6) Aristotle is silent on the role of the gods in tragedy. Cf. 6. 1449^b36 πεì δ πράξεώς στɩ μίμησɩς, πράττεταɩ δ ὑπ τɩνν πραττόντων … πϕυκεν αĭτɩα δύο τν πράξεων εἰναɩ, δɩάνοɩαν καί θος, καί κατά ταύτας καί τυϒϰάνουσι καί άποτυϒϰάνουσɩ πάντες.

^{page 51 note 2} Soph, Track. 545Google Scholarτ δ' αὖ ξυνοɩκεῖν τῇδ' μο τίς ν ϒυνή

δὐναɩτο, κοɩνωνοσα τν αὐτν ϒάμων;

^{page 51 note 3} Eur, . Med. 1339Google Scholarοὐκ ἒστɩν ἣτɩς τοτ' ἂν 'Εληνìς ϒυνή

ἒτλη ποθ' ….

1342

λέαɩναν, οὐ ϒυναῖκα, τς Τυροηνίδς

Σκὑλλης ἔϰουσαν ϒρɩωτέραν ϕὐσɩν.

Cf. Rhet. I. 1375^a6–8 καί τ θηρɩωδέοτερον δίκημα μεῖζον … καί ὂ οί κούονταɩ ϕοβονταɩ μλλον ᾓ λεοσɩ.

^{page 51 note 4} Eur, . Med. 1078Google Scholar, Hipp. 380. But even Medea thinks she will need to ‘forget’ in the moment of action, 1246–8. Both these heroines are suffering from the physical effects of πάθος. Cf. Gorgias' theory of the emotions, Hel. Enc. 15–19 (Diels, Vorsf.⁵ ii. 293–4), esp. 19 εἰ δ' στìν (sc. ἔρως) νθρώπɩνον νόσημα καί ѱυϰς ϒνόημα, οὐϰ ὣς μάρτημα μεμπτέον λλ' ὣς άτύϰημα νομɩστέον.

^{page 51 note 5} E.N 7. 1147^b6 πς δ λύεταɩ ἣ ἂϒνοɩα καί πάλɩν ϒίνεταɩ πɩοτήμων ό κρατής, ό αύτός αύτός λόϒος καί περì το καί καθεδοντος καί ούκ ἴδɩος τούτον το πάθους, ν δεî παρά τν ϕυσɩολόϒων άκούεɩν. In drama this must be observed by the spectator, if the earlier ayvoia is to be convincing. Cf. 3. 1110^b18 τ δ δɩ' ἃϒνοɩαν ούϰ έκούσɩον μν ἃπαν οτίν, κούσɩον δ τ πίλυπον καί ν μεταμελείᾳ.

^{page 51 note 6} <οὐ> το δοκεῖν μοι, τς δ' άληθείας μέλεɩ the only surviving fragment of this play.

^{page 52 note 1} In the account of involuntary action δɩ' ἂϒνοɩαν in E.E. 2. 1225^b it was said that a failure to use knowledge might sometimes count as ignorance in the required sense (^b12–14). In E.N. we seem to be faced with a choice between acts done strictly δɩ' ἂϒνοɩαν and the involuntary but unpardonable acts done in ignorance from abnormal πάθος (5. 1136^a5–9), which are also the basis of the new theory of άκρασία (7. 1147^a11–17 ff.). We might have expected Aristotle to pardon an impulse which, though exaggerated and misdirected by temporary illness or insanity, is in itself natural (e.g. revenge on an enemy) and followed, after the return to knowledge, by pain and repentance. From the physiological point of view, however, there is perhaps little to choose between the various kinds of abnormality. The poet, by attributing the crisis to direct divine intervention, separates it from its physical or moral antecedents in the life of the individual, and causes it to appear as a misfortune which might have happened to any human being, Neither moralist nor physiologist can accept this explanation, cf. Hippocrates, π. ἱερς νούσου.

^{page 52 note 2} E.N. 5. 1135^a28 νδέϰεταɩ ε τόν τυπτόμενον πατέρα είναɩ, τν δ' ὃτɩ μν άνθρωπος ᾓ τν παρόντων τɩς ϒɩνώσκεɩν, ὃτɩ ε πατήρ ϒνοεῖν. Cf. Soph. O.C. 992–4.

^{page 52 note 3} Poet. 7. 1451^a12–15, 10.1452^a16, etc.

^{page 52 note 4} E.N. 1. 1100²5 πολλαί ϒρ μεταβολαί ϒίνονταɩ καί παντοῖαɩ τύϰαɩ κατ τόν βίον, καί νδέϰεταɩ τόν μάλɩστ' εὐθηνοντα μεϒάλαις συμϕοραῖς περɩπεσεῖν πì ϒήρως, καθάπερ ν τοῖς Τρωɩκοῖς περƖ IIρɩάμου μυθεύεταɩ. (Cf. 1101²6–13.) These are not discussed in E.E. (cf. 1. 1219^b5–8), where εὐτυϰία is treated as an alternative cvirpayia depending on ὃρεξɩς ἃλοϒος (8. 1246^b37 ff.).

^{page 53 note 1} If Aristotle had meant what By water thinks (commentary, p. 215) it would have been much clearer if he had said δɩ' ἂϒνοɩαν, not δɩ' μαρτίαν.

^{page 53 note 2} E.N. 6. 1142^a21 (ϕρόνηοɩς) ἒτɩ μαρτία περì τ καθόλου ν τῷ βουλεύσασθι περì τό καθ' ἓκαστον. Opposed to ρθότης, ^b10. E.E. 2. 1226^a36 δɩϰ ϒɩνομένης τς ἃμαρτίας (ἤ ϒάρ λοϒɩζόμενοɩ άμαρτάνομεν ἤ κατά τήν αίσθησɩν αὐτ δρντες) ν μέν τῇ ἰατρɩκῇ μϕοτέρως νδέϰεταɩ καί πρξɩν. Cf. M.M. I. 1189^b21 ff., Phys. 2. 199^a33, Poet. 25. 1460^b15, 17, Probl. 19. 919^b25 μαρτɭα δέ έστι το ϰεɭρονος πρξɩς.

^{page 53 note 3} δɩά τν τοιαύτην μαρτίαν ἄδɩκοɩ κοί ὃλως κακοί ϒίνονταɩ.

^{page 53 note 4} Cf. Rhet. ad Alex. 4. 1427^a35 τ δ δɩ' ἄϒνοɩαν βλαβερόν τɩ πράττεɩν μαρτίαν είναɩ ϕατέον.

^{page 53 note 5} E.N. 7. 1148^a3 μν ϒαρ κρασία ѱέϒεταɩ οὐϰ ὡς άμαρτɭα μόνον λλά καί ώς κακία τίς ἤ πλς ούσα ἤ κατά τɩ μέρος, τούτων δ' οὐδείς. Cf. Rhet. 1. 1374^b7 μαρτήματα δ όσα μή παράλοϒα καί μή π πονηρίας, δɩκήματα δ όσα μήτε παράλοϒα π πονηρίας τ' στίν· τά ϒάρ δɩ' πɩθνμίαν π πονηρίας. Cf. E.E. 1. 1223^a37, 8. 1246^b13.

^{page 53 note 6} ϒίνεταɩ δ τν μαρτɩν μν ὃτɩ <> οὐ δεî (SC. ϕοβεῖταɩ), ἤ δ ὃτɩ ὃτϰ ὡς δεῖ, ἣ δ ὂτɩ οὐϰ ὂτε, ἣ τɩ τν τοɩούτων.

^{page 53 note 7} Cf. Phys. 2. 197^a21–5. The basis of the distinction between μαρτήματα and μαρτήματα is the same, E.N. 5. 1135^b16–19 ὃταν οὖν παραλόϒως βλάβη ϒένηταɩ, τύϰημα … (μαρτάνεɩ μν ϒρ ὃταν άρϰή ν ὐατῷ ἣ τς αἰτίας, τυϰεῖ δ' ὂταν ἓξωθεν), i.e. an act is an τύϰημα only when injury was neither intended, nor the probable consequence of what was intended or of any impulse (e.g. not when an angry man strikes too hard, or when boiling is chosen as a means to rejuvenation, or when a dubious love-charm leads to death). The co-operation of external causes may make the tragic act itself hard to classify, but it seems probable that for Aristotle τό δɩ' μεταβάλλεɩν εἰς δυστυϰίαν means, not the effect of the single act, but a certain sequence of events distinct from chance on the one hand and purposive action on the other, cf. Poet. 9. 1452^a3–6. Except in the δɩπλος μῡθος, which Aristotle expressly repudiates (13. 1453^a12–16, 30–3), ναϒνώρɩσɩς and περɩπέτεɩα must arise from μαρτία. (Cf. μαρτία in nature, Phys. 2. 199^a33–^b4, and the injudicious diet of the carabus which leads to a περιπέτια, Hist. An. 8. 590^b12–18.) History also remarks these sequences, e.g. the misfortune of the hitherto prosperous and prudent Chians, Thuc. 8. 24. 5 εί δέ τɩ δ τοῖς νθρωπείοɩς το βίου παραλόϒος σϕάλησαν, μετ πολλν οἳς ταύτά ἒδμοξε, τ τν 'Αθηναίων ταϰύ ξυναɩρεθήσεσθαɩ, τήν μαρτίαν ξυνέϒνωσαν.

^{page 54 note 1} The repetition of the tragic motive may show human ἦθος, but it shows divine προαίρεσις. Aristotle cannot have been unaware of these implications, cf. Poet. 9. 1452^a7 πε κα τν π τχης τατα θαυμασιώτατα δοκεῖ ὂσα ὢσπερ πίτηδες ϕαίνεται γεγονέναι. For the impression of an over-ruling destiny conveyed by περιπέτεια see Lock, C.R. ix. 252.

^{page 54 note 2} Cf. E.E. 1. 1214^a22–5; in the account of involuntary action (2. 1225^a28) νθουσιασμός is merely a psychological parallel, cf. Pol. 1342^a7. Contrast the Pythagorean version, Stob. Ecl. 1. 6 (Diels, Vors. ⁵ i. 478) περ δ τύχης τάδ' ἒϕασκον εἶναι μέν τι κα δαιμόνιον μέρος αὑτς γενέσθαι γρ νίοις π τ βέλτιον ἦ π τ χεῖρον κα εἶναι ϕανερς κατ' αὐτ τοτο τοὺς μν εὺτυχεῖς, τοὺς δ τυχεῖς.

^{page 54 note 3} Cf. lambl. V.P. 137, 174 (Vors. ⁵ i. 468.)

^{page 55 note 1} Men. Περικειρομένη 44–50.

^{page 55 note 2} e.g. Eur. Ion, I.T. The whole situation may nevertheless be the consequence of earlier divine action.

^{page 55 note 3} Cf. E.N. 3. 1112^b16–19. μαρτία leading to δυστυχία may be κατ τ ναγκαῖον either as divinely compelled (βίαιον), or as an indispensable means (ξ ὑποθέσεως νάγκη Phys. 2. 199^b34 ff.), or as a natural process which is not interfered with (ibid.^b26, E.N. 7. 1147^a30). The events which lead to εὐτυχία may be κατ τ εἰκς, but as far as the ναγνώρισις we are more concerned with the probability that they will lead to disaster.

^{page 55 note 4} ἒοικε γρ τοιατα οὐκ εἰκῇ γενέσθαι (Poet. 9. 1452^a9). So even τύχη can wear a divine look ^a6, cf. Phys. 2. 196^b6).

^{page 55 note 5} Cf. Eur. Ion 1510–17, I.T. 870–2.

^{page 55 note 6} Rhet. 2. 1385^b13 π ϕαινομένῳ κακῷ, not π γενομένῳ. 1386^a34 γγὺς γρ ποιοσι ϕαίνεσθαι τ κακν πρ μμάτων ποιοντες, ἢ ὡς μέλλον ἢ ὡς γεγονός κα τ γεγονότα ἂρτι ἢ μέλλοντα δι ταχέων λεεινότερα δι τ αὐτό. Cf. Phys. 2. 197^a27–30.

^{page 55 note 7} Rhet. 2. 1382^a21 κ ϕαντασίας μέλλοντος κακο. It can be roused by signs and associations (^a31, cf. pity, 1386^b2).

^{page 55 note 8} E.N. 3. 1111^a1, 12. Plut. de es. carn. 2. 998 e (Nauck ap. ad Eur. fr. 456) σκόπει δ κα τν ν τῇ τραγῳδίᾳ Μερόπην π τν νἱν αὐτν ὡς ϕονέα το υἱο πέλεκυν ραμένην κα λέγουσαν ‘ὠνητέραν δ τήνδ' γὼ δίδωμί σοι ∣ πληγήν’ ὂσον ν τῷ θεάτρῳ κίνημα ποιεῖ συνεξορθιάζουσα ϕόβῳ, κα δέος μ ϕθάση τν πιλαμβανόμενον γέροντα κα τρώσῇ τ μειράκιον. If the story is known, fear must be a kind of κρασία, cf. οἱ μελαγχολικοί who are κολουθητικοί τῇ ϕαντασίᾳ (E.N. 7. 1150^b28) and require strong pleasure as λύπης ἰατρεία (1154^b11–14). Cf. de Div. per Somn. 464^b1 ff.

^{page 56 note 1} Poet. 4. 1448^b13, Rhet.I. 1371^b5–10.

^{page 56 note 2} Poet. 9. 1452^a2–5, cf. 6. 1450^a33, 24. 1460^a12–17. Wonder is the desire to understand, Rhet. 1. 1371^a31–4 (cf. ^b5–11), Met. 1. 982^b17–19.

^{page 56 note 3} Poet. 14. 1454^a, 16. 1455^a17.

^{page 56 note 4} Rhet. 1. 1371^b10 κα αί περιπέτειαι κα τ παρ μικρν σῴζεσθαι κ τν κινδύνων (sc. δύ) πάντα γρ θαυμαστ τατα.

^{page 56 note 5} Through the realization of the persons on the stage and through memory, cf. Rhet. 1. 1370^a27 ff.

^{page 56 note 6} i.e. π λέου κα ϕόβου δι μιμήσεως δονή (Poet. 14. 1453^b12). κουϕίζεσθαι μεθ' δονς (Pol. 1342^a14) is only a part of this pleasure. If the theory of κάθαρσις had implied the unhappy ending, Aristotle would not have needed to appeal to the evidence of actual stage performance in Poet. 13. 1453^a27. The general explanation in Pol. 1342^a4–15, written before the Poetics and perhaps before the E.N. (Jageer, Aristotle, pp. 266 ff.), may, however, belong to the earlier period when Aristotle still preferred the unhappy ending.

^{page 56 note 7} To refer μαρτία to a family curse, or to the agent's behaviour in a previous existence (cf. Plat. Rep. 10. 619^b6–^d1), extends but does not solve the religious problem, which arises from the direct relation of μαρτία to a scheme of activity beyond the control of the individual.