¹ I should like to thank Professors Gordon Kirkwood, M. of Cornell University and Herbert A. Musurillo, S.J., of Fordham University for their many valuable and learned suggestions.Google Scholar

² In pre-Aeschylean authors ἄτη is found in the following passages: Homer, Iliad 1.412; 2.111; 3.100; 6.356; 8.237; 9.18; 9.115; 9.504; 9.505; 9.512; 10.391; 16.274; 16.805; 19.88; 19.91; 19.126; 19.129; 19.136; 19.270; 24.28; 24.480; Odyssey 4.261; 12.372; 15.233; 21.302; 23.223; Hesiod, Erga 216; 231; 352; 413; Theogony 230; Shield 93; Solon 1.13; 1.68; 1.75; 3.35; Alcaeus 43.12; Ibycus 1.8; Theognis 1.103; 1.119; 1.133; 1.206; 1.231; 1.588; 1.631; Pindar, Olympian 1.57; 10.37; Pythian 2.28; 3.24; Nemean 9.21.Google Scholar

³ These loci are: Persians 112; 653; 822; 1007; 1037; Seven 315; 601; 687; 956; 1001; Suppliants 110; 444; 470; 530; 850; Agamemnon 361; 386; 643; 730; 735; 771; 819; 1124; 1192; 1230; 1268; 1283; 1433; 1523; 1566; Libation-Bearers 68; 272; 339; 383; 403; 404; 467; 597; 825; 830; 836; 968; 1076; Eumenides 376; 982; Prometheus 886; 1072; 1078.Google Scholar

⁴ Odyssey 12.372; Erga 231; Alcaeus 43.12; Ibycus 1.8; Theognis 1.206; Pindar, Olympian 1.57; 10.37; Nemean 9.21.Google Scholar

⁵ The text and linear numeration here and in subsequent quotations from the Persians is that of Broadhead, E., The Persians (Cambridge 1960). The possible personification of ἄτη in 1007 has been raised. English editors (Paley, Smyth, Murray, and Broadhead) tend to capitalize ἄτη. German editors, on the other hand (Dindorf, Wecklein, and Wilamowitz), read ἄτα. It would seem (admittedly this is conjecture, since none of them states his case in his apparatus or notes) that those editors who want to personify ἄτη base their case on the verb δέδοϱϰ∊ν. But it does not necessarily follow that the verb δέοϰοµαι implies a personal agent. Homer uses the verb with animate, though not personal, agents; thus: Γοϱγώ … δ∊ινὸν δ∊ϱκοµένη and δϱάκων … σµ∊ϱδαλέον δὲ δέδοϱκ∊ν (Iliad 11.37 and 22.95). A similar usage is found in the Seven against Thebes: λ∊όντων ὡς Aϱη δ∊δοϱκότων (53). And Pindar employs the verb with inanimate agents: δέϱκ∊ται … ὁ µέγας πότµος, and δέδοϱκ∊ν φάος, and finally δέδοϱκ∊ν … φέγγος (Pythian 3.85; Nemean 3.84 and 9.41). Given such uses of δέϱκοµαι it seems better to read ἄτα, lower case, in line 1007, in agreement with Dindorf, Wecklein, and Wilamowitz.Google Scholar

⁶ These are: Sidgwick, Headlam, Wilamowitz, Rose, and Broadhead.Google Scholar

⁷ Thus, in Odyssey 12.372 it includes Odysseus and his men; in Erga 231 it affects farmers generally; in Alcaeus 43.12 the citizens of Miletus are its victims; in Ibycus 1.8 the Trojans are affected; in Theognis 1.206 it would affect a man's descendants; in Pindar's Olympian 10.37 the residents of an entire city are its victims; and in Nemean 9.21 it descends upon an entire army. The sole exception is Pindar's Olympian 1.57 where the victim of ruin is Tantalus.Google Scholar

⁸ Thus, in Odyssey 12.372 Z∊v πάτ∊ϱ ἤδ’ ἄλλ03BF;ι, µ03AC;03BA;α03F1;∊ς Θ∊οὶ in Hesiod's Erga 231 Z∊ύς, in Alcaeus 43.12 ‘Oλυνπίων τις and in Ibycus 1.8 διὰ Kύπϱιδα.Google Scholar

⁹ Aeschylean usage was quite consistent with respect to each of these three points. See Appendix A for a table which displays each instance.Google Scholar

¹⁰ As Smyth, H., Aeschylus (Cambridge, Mass. 1946), Rose, and Broadhead (again, the only editors or commentators to treat the passage) all agree, the meaning of ἄτη here is ‘disasters,’ ‘calamities.’ There is also a distinctive characteristic about the use of ἄτηισι in 1037. It is the first text in which the word is used in the plural in the sense of ‘ruin.’Google Scholar

¹¹ Persians 112.Google Scholar

¹² The text and linear numeration is that of Murray, Gilbert, Aeschyli septem quae supersunt tragoediae ² (Oxford 1960). Notice the similarity to Pindar, Olympian 10.37.Google Scholar

¹³ For a discussion of this whole problem, see Lloyd-Jones, H., ‘The Guilt of Agamemnon’ Classical Quarterly 12 (1962) 187–199, and A. Lesky, ‘Decision and Responsibility in the Tragedy of Aeschylus’ Journal of Hellenic Studies 86 (1966) 78–85.Google Scholar

¹⁴ The text and linear numeration here and in subsequent quotations from the Agamemnon is that of Fraenkel, E., Aeschylus: Agamemnon (Oxford 1950). Lines 768 and 769 are textually very difficult. As Fraenkel says: ‘After µόληι begins one of the deeply-rooted corruptions I have mentioned on 374ff. … It seems pointless to discuss the various attempts at restoration’ (op. cit., II, 352). Fraenkel is echoing the opinion of Wilamowitz who wrote: ‘The alterations [suggested by Wilamowitz] certainly are numerous though slight; the archetype must have been difficult to read, and the Byzantines must have tried to patch it up;… here the text of our MSS is so uncertain that corruptions may have taken place which make powerless any criticism which is more than mere trifling’ (Aischylos Interpretationen 197–198).Google Scholar

In view of such skepticism on the part of two distinguished philologists, it would be rash to try to vindicate any given reading for the two lines in question. Perhaps it is best to accept Fraenkel's reconstruction, with the same hesitation with which he proposes it, as a tentative working basis from which to discuss the two lines. Fraenkel writes: ‘I regard the following as a possible, though in detail by no means certain, arrangement of the text : ὅταν τὸ κύϱ03B9;03BF;ν µόληι φάος, κότ03BF;ν ν∊ώϱη, δαίmονα τίταν, ἄµαχ03BF;ν continuing with the MS reading’ (ibid.). With the passage already in such a condition, lines 770–772 present further textual difficulties concerning the phrase µέλαιναν … ἄταν, ∊ἰδοµέναν. The problem here is to determine the number and case. The MSS have µ∊λαίνας … ἄτας, with µ∊λαίνας ἄτας construed as a genitive singular. This has been followed by Triclinius, Hermann, Wecklein, Dindorf, Wilamowitz, and Thomson. Fraenkel follows Sidgwick in admitting that this is possible but ‘very harsh.’ Donaldson and Paley emended to µ∊λαίνα … ἄτα, ∊ἰδοµένα — i.e., all the words are duals. Sidgwick, Smyth, Lawson, and Denniston and Page emended to µ∊λαίνας … ἄτας, ∊ἰδοµένας, the accusative plural. Fraenkel correctly points out that both these emendations conflict with the whole train of thought beginning with 751, since there has been question of ‘one single child of ὕβϱις as well as of ὄλβος.’ He therefore tentatively adopts the suggestion of Auratus, viz., µέλαιναν … ἄταν (∊ἰδοµέναν) and remarks: ‘If we adopt it [Auratus’ suggestion], the result is (assuming the correctness of the very hypothetical arrangement of the preceding passage as suggested above) a satisfying articulation of the long series of accusatives that (768) follow on φάος all in apposition to ὕβϱιν; we get first two parallel elements, κότον ν∊ώϱν and δαίµονα τίταν, then the longer element ἄµαχον ἀπόλ∊µον ἀνί∊ϱον Θϱάσος, then after these preparatory periphrases the name of the daemon of evil … µέλαιναν Aταν, and finally the closing element ∊ἰδ. τοκ. which emphasizes once again the genealogical relationship by picking up the thought of 760. But this rearrangement cannot be regarded as certain’ (op. cit., II, 353).

¹⁵ Implicit confirmation is also obtained from the observation that the corresponding antistrophe δ (772–781) is a development of lines 761–762.Google Scholar

¹⁶ For a good discussion of the general context, particularly of the idea that in this scene Cassandra is a ‘messenger’ of the future to whom the past is also present, see Owen, E. T., The Harmony of Aeschylus (Toronto 1952) 84–86.Google Scholar

¹⁷ These four lines have been the subject of a good deal of discussion and emendation because, as Rose, H. J., ‘On an Epic Idiom in Aeschylos’ Eranos 45 (1947), remarks: the passage is one ‘which every modern editor declares more or less corrupt’ (98–99). None of the alleged difficulties, however, involves ἄτης λαΘϱαίον in 1230, and Fraenkel's text is quite acceptable since it adheres closely to the MSS and avoids many of the radical suggestions proposed.Google Scholar

The proposed emendations may be seen in four articles given here in chronological order: A. Y. Campbell, ‘Aeschylus, Agamemnon 1227–1230’ Classical Quarterly (CQ) 26 (1932) 45–51; J. C. Lawson, ‘Aeschylus, Agamemnon, 1227–32’ CQ 27 (1933) 112–114; G. Thomson, ‘Notes on the Oresteia’ CQ 28 (1934) 72–78; and A. Y. Campbell, ‘Aeschylus Agamemnon 1223–38 and Treacherous Monsters’ CQ 29 (1935) 25–36.

¹⁸ Wilamowitz transposed 1431–1433 to follow 1437 and argued: ‘Ad minas primum erat respondendum; numina autem testes facinoris advocantur.’ This transposition is rejected on stylistic grounds by Fraenkel, op. cit., II, 678, and by Kranz, W., ‘Zwei Lieder des “Agamemnon”’ Hermes 54 (1919) 314, because of the sense of the passage. Lines 1434–1437 are commented on by R. Böhme, Bühnenbearbeitung Aeschyleischer Tragödien (Stuttgart 1956), I, 30–32. For the connection between Θέµις and δίκη see B. Daube, Zu den Rechtsproblemen in Aischylos' Agamemnon (Leipzig 1938) 179–182. K. Reinhardt, Aischylos als Regisseur und Theologe (Bern 1949) 108–110, discusses the contrast between this and the following scene.Google Scholar

¹⁹ Op. cit., ad loc.Google Scholar

²⁰ There seems to be no compelling reason to personify the three abstract nouns in 1432–1433 as Fraenkel does in his text. δίκη and ἐϱινύς (particularly as echoed in the ἀλάστωϱ of 1501) are not necessarily personifications. Nor is ἄτη. Fraenkel's reference to Usener's article is misapplied here, for as Fraenkel himself admits, op. cit., II, 675, Usener ‘quotes Septem 42ff. but does not mention Ag. 1432ff.’Google Scholar

²¹ This suggestion has been followed by Dindorf, Wecklein, Paley, Smyth, Lawson, Thomson, Rose, Fletcher, and Murray.Google Scholar

²² On this point, see Paley, op. cit., 488.Google Scholar

²³ Op. cit., ad loc. Google Scholar

²⁴ This is the understanding of Smyth, Rose, and Fraenkel, op. cit., ad loc. Google Scholar

²⁵ Op. cit., II, 239.Google Scholar

²⁶ For an explanation of the variants from Murray's text, see Appendix B.Google Scholar

²⁷ This is the meaning given to ἄτη in these lines by Dindorf, Schütz (quoted by Dindorf), Smyth, and Thomson.Google Scholar

²⁸ It is interesting to note that in their suggested retort, πατϱὸ+ς αὐδάν the chorus justify Orestes' deed on precisely the same ground which Athena uses in the Eumenides.Google Scholar

²⁹ Cf. lines 306–307.Google Scholar

³⁰ The text and linear numeration here and in subsequent quotations from the Eumenides is that of Murray, G., op. cit. Google Scholar

³¹ Op. cit., ad loc. Google Scholar

³² Goheen, R., ‘Aspects of Dramatic Symbolism’ American Journal of Philology 76 (1955) 120, notes: ‘We may say, then, that the imagery of dark blood on the ground, which I believe gets its first visual statement in the carpet of the Agamemnon, helps to develop within the trilogy one of its thematic ideas — namely, since blood is irredeemable, bloodshed is not an adequate solution; legal process and a willingness to reach understanding offer more hope.’ For a possible historical allusion in this passage, see R. Livingstone, ‘The Problem of the Eumenides of Aeschylus’ Journal of Hellenic Studies 45 (1925) 129, and A. Podlecki, The Political Background of Aeschylean Tragedy (Ann Arbor 1966) 85–86.Google Scholar

³³ The text and linear numeration is that of Murray, G., op. cit. Google Scholar

³⁴ Thus, Paley, Henry, Smyth, Thomson, and Rose.Google Scholar

³⁵ The text and linear numeration is that of Murray, G., op. cit. Google Scholar

³⁶ Thus Paley: ‘infatuation, groundless panic’; Sidgwick: ‘panic’; Smyth: ‘panic’; Italie: ‘een verblinding.’Google Scholar

³⁷ References to Horace seem beside the point. Surely Aeschylus has something other in mind than the lyric poets' convention.Google Scholar

³⁸ Dindorf, op. cit., ad loc. Tucker, T., The Seven Against Thebes of Aeschylus (Cambridge 1908), ad loc.Google Scholar

³⁹ The MSS read ἄτα at line 131. But this has been universally rejected by editors on metrical grounds in favor of Hermann's emendation ἄyα. The same correction has been made at Suppliants 164.Google Scholar

⁴⁰ This is the meaning given by Rose, Fraenkel, Smyth, and Deniston and Page.Google Scholar

⁴¹ See Greene, W. C., ‘Dramatic and Ethical Motives in the Agamemnon’ Harvard Studies in Classical Philology 54 (1943) 25–33, especially 25–26.Google Scholar

⁴² See above p. 12.Google Scholar

⁴³ Op. cit., II, 340. Previously Fraenkel had indicated the metrical pattern for this section of the stasimon as follows: ‘Three clearly marked periods: (1) 2 glyc. + pherecr., (2) 3 dactylic ἡµι∊π (ἄνδϱα µ∊ ἔνν∊π∊ Mοσα), (3) 2 lecythia + priap. Every line of this stanza is a dimeter, acatalectic or catalectic. What we here call a dactylic colon occurs elsewhere as equivalent to a pherecratean, e.g., Ar. Thesm. 1139 παϱΘένον ἄζνγα κούϱην’ (op. cit., II, 328). The missing short syllable in the dactylic colon is needed for responsion to ἐν βιότον πϱοτ∊λ∊ίοις in line 720. Among editors only Wecklein retains the MS reading.Google Scholar

⁴⁴ ἄταισιν was given by Triclinius. Bothe first proposed [ἐν] ἄταις which has been adopted by Wilamowitz, Smyth, Thomson, and Fraenkel. Lawson, Fletcher, and Denniston and Page have [σὺ]ν ἄταις.Google Scholar

⁴⁵ Paley suggests ἄσαισιν, Hermann and Dindorf adopt ἄγαισιν, and Murray follows Schneider in reading µάταισιν.Google Scholar

⁴⁶ The importance of such an application was suggested by Knox's, B. M. W. article ‘The Lion in the House (Agamemnon 717–36)’ Classical Philology 47 (1952) 17–25. Further consideration of Knox's ideas will be given in the discussion of 735.Google Scholar

⁴⁷ Triclinius' reading is adopted because it is good Greek in itself and less awkward than either ἐν or σὺν ἄταις. This reading is strengthened, perhaps, in what follows. For the local force of ἐν or the accompaniment involved in σὺν are not always verified and would be positively misleading in some instances.Google Scholar

⁴⁸ Helen is the ‘principal’ figure represented by the lion, but not necessarily the only one as Knox, op. cit., has shown.Google Scholar

⁴⁹ The disastrous consequences to the Tyndarids are not expressed in terms of pastoral metaphor in the Agamemnon. Nothing at all is said about Helen's ultimate fate in this play. And Clytemnestra's ruin, like Aegisthus' (the only one of the Pelopidae thus omitted), is not central to this play but rather to the Libation-Bearers.Google Scholar

⁵⁰ Though clearly not attributable to Helen, the banquet of Atreus for Thyestes, strongly hinted at by the banquet mentioned in 731, is spoken of in pastoral imagery at 1590–1595.Google Scholar

⁵¹ There is an Aeschylean precedent for the use of ἄτη in the plural to signify ‘disasters’ or ‘ruins.’ It is Persians 1037. See p. 3 for the discussion of that passage.Google Scholar

⁵² Thus, Smyth, op. cit., ad loc. has ‘ruin’; Kaufmann-Bühler, D., Begriff und Funktion der Dike in den Tragödien des Aischylos (Heidelberg 1951) 62 has ‘Verderbens’; Denniston and Page, op. cit., ad loc. have ‘destruction’; Fraenkel, op. cit., ad loc. equivocates by translating ‘Ate.’Google Scholar

⁵³ See note 46 above for the precise reference.Google Scholar

⁵⁴ Page 22. Knox also develops the idea of antistrophe β with Menelaus (instead of Paris) and Helen as the lion. See Fraenkel, , op. cit., II, 342, for a possible source of this tale of the lion.Google Scholar

⁵⁵ Op. cit., 18.Google Scholar

⁵⁶ Ibid., 22. Paris, too, figures in the imagery, as Knox makes clear on p. 17.Google Scholar

⁵⁷ Connington, As, quoted by Fraenkel, op. cit., III, 737, wrote: ‘ἀϱαῑον … and πϱὸς ἄτα … are both so indisputable that it is hardly requisite to particularize them as departures from the reading of the MSS.’ Fraenkel, loc. cit., states: ‘κ∊κόλληται … πϱὸς ἄται is a drastic intensification of the Homeric expression in B.111 (= I.18) Z∊ὺς µ∊ … ἄτηι ἐνέδησ∊ βας∊ίηι, which (as Reisig saw) Sophocles takes over word for word in Oed. C. 525f.’Google Scholar

⁵⁸ This is the meaning given by all the editors and commentators who discuss ἄτη here. Thus, Paley, Smyth, Fraenkel, and Denniston and Page all understand ἄτη as ‘ruin,’ ‘calamity,’ or ‘perdition.’Google Scholar

⁵⁹ Cf. 386, 735, 771, and 1283.Google Scholar

⁶⁰ The text and linear numeration here and in subsequent quotations from the Libation-Bearers is that of Murray, G., op. cit. Google Scholar

From line 380 to line 585, with the exception of line 479 which is ascribed to Electra, the MSS give no indications of which actor is speaking at any given place. This has naturally led to a certain discrepancy among editors in assigning various passages. In the current instance, however, most editors agree that this strophe should be attributed to Orestes.

Thus, Portus, Wecklein, Wilamowitz, Smyth, Thomson, and Murray. Dindorf and Paley, however, follow the scholiast in assigning this strophe to Electra. This seems less satisfactory, however, in view of the construction of the entire kommos. For a good discussion of this point, see W. Schadewaldt, ‘Der Kommos in Aischylos Choephoren’ Hermes 67 (1932) 312–354.

⁶¹ All editors ascribe these lines to the chorus.Google Scholar

⁶² Goheen, R., op. cit., 119–120. He also remarks: ‘The image of blood on the ground, reiterated and modulated like a motif in music, leads from this point on [i.e., Agamemnon 1018–1021] through the entire Oresteia. In the Choephori it appears three times: 48, 400–2, 520–1; and in the most fully delineated of these the figure is adapted to crystallize the vengeful ethos of that play.’ Might it not be possible to add line 67 to this list?Google Scholar

⁶³ This is the meaning given by Paley, Klausen (quoted by Paley), and Thomson.Google Scholar

⁶⁴ Most editors, viz., Wecklein, Dindorf, Paley, Wilamowitz, Smyth, Thomson, and Murray, attribute these lines to the chorus. The scholiast assigned them to Orestes and Electra. Kirchhoff and O. Mueller attributed them to Orestes, Electra, and the chorus.Google Scholar

⁶⁵ See pp. 11–12 above.Google Scholar

⁶⁶ Thus, Blomfield (quoted by Fraenkel, op. cit., II, 319) understood ‘calamitas.’ Fraenkel has ‘bane,’ and Smyth translates ‘destruction’; both op. cit., ad loc. Fraenkel, op. cit. II, 320 sees δίλογχον as Homeric.Google Scholar

⁶⁷ See the excellent remarks of Lesky, A., Greek Tragedy (New York 1965) 13ff., on this point.Google Scholar