Eduard Fraenkel (ed.), Aeschylus: Agamemnon, Vol. 3: Commentary on 1056–1673, Appendices, Indexes
APPENDIX CCho. 991–1006
The speech of Orestes, Cho. 973–1006, contains a difficulty of the first order. It will be expedient to leave aside, at any rate for the moment, some minor points and concentrate upon the crucial issue. What is the meaning of 997 τί νιν προσείπω? As the text stands in the MS, any reader must needs refer pg 810νιν to the person who is the subject of the whole preceding period 991–6, Clytemnestra. Only when we go on do we realize with a shock that Orestes seems to be talking no longer about his mother but about the fatal garment, the 'net'. The difficulty was of course noticed long ago,1 and for more than a hundred years many critics have resorted to transpositions or other alterations of the text. Before, however, considering any such changes, we must give the arrangement in the MS the benefit of the doubt. Our first task, therefore, will be to examine the views of some of those scholars who thought they could understand the unaltered text.
In Gilbert Murray's edition we read the following note on 997: '997–1004 post 982 ξυνωρίδα traiecit Scholefield, sed vide Verrall ad loc.' It might have been better to refer to Conington instead of Verrall, for the latter confines himself to quoting Conington's note and expressing complete agreement with it. This is what Conington says: 'Or. wishes to find a name for his mother, without saying anything that ought not to be said, νιν is Clyt., as is evident from what goes before. He proceeds to identify her with the net, the instrument of her crime, enlarging on its villainous uses, as if he had nothing else in his mind, till in v. 1005 he at last returns to her. This identification is doubtless a symptom of the frenzy which is beginning to work on him, at the same time that it has its own imaginative truth. Precisely the same identification is made by Cassandra, Ag. 1114 foll. ἒ ἔ, παπαῖ παπαῖ, τί τόδε φαίνεται; Ἦ δίκτυόν τί γʼ Ἅιδου; Ἀλλʼ ἄρκυς ἡ ξύνευνος, ἡ ξυναιτία φόνου. Thus it would be worse than useless to follow Meineke (anticipated by Scholefield) in transposing this and the seven following lines so as to insert them after v. 982' etc. After what I have said already in the commentary on the Cassandra scene, there is no need to reject once more the identification which Conington assumes there. The subject of the phrase ἄλλʼ ἄρκυς ἡ ξύνευνος κτλ. cannot be anything but the garment, the net, as it appears to Cassandra in her vision. The whole character of the scene with its clear distinctness of each separate vision makes it impossible to imagine that even for a moment Cassandra could confuse the apparition of the garment with that of the woman. The rejection of the parallel from the first play of the trilogy would not, however, be fatal to Conington's main thesis if his interpretation of the speech of Orestes carried conviction in itself. But I think it does not: Conington, whose sympathetic understanding of Aeschylus often proves most admirable, has erred here. It is fundamentally wrong to believe that at any moment during Orestes' first speech in the last scene, 973–1006, 'the frenzy is beginning to work on him'. The poet clearly marks the transition from a perfectly sane, if highly excited, mood to mental disturbance. In his third speech, at 1.1026, Orestes says ἕως δʼ ἔτʼ ἔμφρων εἰμί. He is obviously aware of the imminent danger of madness, but not yet in its grip. At 1048 a fit of frenzy sets in:2 he bursts into the wild shout ἆ, ἆ (cf. the corresponding interruption of the trimeters Ag. 1214); now he sees the Erinyes, though no one else can see them (1061), they are unmistakable and real, no deceptive illusions (1053 f.). The difference between the stage at which a tormented man is afraid lest he may go mad, while the working of his mind is still perfectly pg 811clear, and the later stage at which frenzy has actually taken hold of him was as well known and as important to Aeschylus as it was to Shakespeare.1 We are not entitled to correct the poet's conception by confusing what he kept separate. Neither is it permissible to resort to psychological subtleties and argue that Orestes is mistaken about his own state of mind and that, in fact, he has already been raving a considerable time before he protests that he is still sane but fears madness. Such a presentation of a decisive turn would not be in the manner of Aeschylus. Moreover, according to Conington, Orestes would 'identify Clytemnestra with the net' in that very speech whose central theme (980 ff.) is the real net, the garment, and, what seems to me quite intolerable, would do so a moment after that garment has been unfolded with elaborate care before the eyes of the people on the stage and of the audience. No one who is aware of the dominating position held by the fatal cloak throughout the Oresteia, no one who appreciates the concentrated energy with which the poet in every speech of this culminating scene of the Choephoroe dwells on one primary issue, will be prepared to accept Conington's 'identification' and the substitution implied in his hypothesis.
Another attempt to interpret 997 ff. without altering the order of the lines in the MS was made by Wilamowitz. His view is fully expressed in his commentary (published 1896) on the Choephoroe and in his German translation with its stage-directions; a brief summary is given under the heading 'Actio' in his editio maior of Aeschylus.2 Wilamowitz assumes that in 983 f. Orestes bids his servants spread out the garment and carry it round in a circle ('entfaltet das und tragt es rings im Kreis herum'), 'from παρασταδόν we learn that the servants approach each man in the crowd separately; the horror of the onlookers as they one by one inspect the murderous cloak must have greatly enlivened the scene' (commentary on 983). Then, according to Wilamowitz, the garment, after being carried round the whole circle, would be returned to Orestes at 997. 'Orestes uses the mere pronoun νιν, which was quite intelligible to the spectators although the reader here and at 1004 requires a παρεπιγραφή' (ibid.). 'νιν was quite intelligible to the spectators.' Was it indeed? The light, purely anaphoric pronoun, after the long diatribe against Clytemnestra? In his translation Wilamowitz goes round the obstacle by rendering τί νιν προσείπω: 'das da -… wie bezeichn' ich's recht?' That would be τόδε rather than νιν. But we need not elaborate this point. For Wilamowitz's whole construction is based on a misunderstanding of 983 κύκλωι παρασταδόν. Whether we accept with Wilamowitz the reading of M παρασταδόν [cf. S. Trach. 194 f. κύκλωι … παραστάς] or write περισταδόν, the meaning of the clause is clearly: 'spread it out and stand near by (or "round") pg 812in a ring … to make it possible for Helios to see it.' The ἄπειρον ἀμφίβληστρον is a garment of uncommon size (cf. on Ag. 1382); to unfold the huge thing properly and show it to the Sun, the attendants have to stand round in a circle, 1 each of them holding a piece of the edge. There is nothing in the text of Aeschylus to suggest that the cloak should be carried round.
Other critics2 have resorted to transpositions. Scholefield, 3 and, independently, Meineke (see Hermann's note), placed 997–1004 after 982. This was accepted by Hermann and many others. At first sight it seems very attractive, for it rids us of the two main difficulties in this speech by giving the νιν of 997 its proper relation to the garment and removing the awkward abruptness with which, according to the MS text, Orestes at 1005 returns to Clytemnestra. Still we should hesitate to acquiesce in this alteration. For, as Weil observed, 'ita Orestes alloqueretur velum nondum explicatum'. In the text as it stands in the MS the large cloak is first unfolded in an action which takes a considerable time (see below) and only then does Orestes turn to it with his excited questions τί νιν προσείπω κτλ. We shall presently be in a position fully to appreciate this arrangement, a master-stroke of Aeschylean stage-effect. But even now we see enough to reject an alteration in consequence of which Orestes would first expatiate on the enormous size of the voluminous garment (997–1004) and afterwards have it unfolded so that the spectators may have the full view of it. It is obvious that it is the actual sight of the evil thing when it is widely spread out that provokes the outburst τί νιν προσείπω κτλ.
Weil himself inserted 997–1004 in the next speech after 1013,4 Blass and A. Y. Campbell, C. Q. xxix, 1935, 31 n. 4, followed him. The prejudice that it was desirable to make the three speeches of Orestes about equal in length had a share in Weil's transposition. The change implies grave consequences. In the first place, it is obviously much more appropriate that the questions τί νιν προσείπω κτλ. should belong to that part of the dramatic action which has as its centre the spreading out of the cloak. Secondly (and this I regard as even more vital) Weil seems completely to miss the keynote of the intense excitement in Orestes' second speech (1010 ff. ἔδρασεν ἢ οὐκ ἔδρασε; κτλ.). In this speech it is one issue and one alone that matters: the matricide, weighed down by the consciousness of his pollution (1017), asks from the depth of a tortured mind 'was she really guilty?' Only in so far as it provides an answer to this horrible doubt does the garment play a part in this brief speech. Any diversion from the burning problem would be intolerable here. It should also be pg 813noticed that the whole thought of 997–1004 shows that when Orestes speaks these sentences he feels no doubt whatsoever as to his mother's guilt. It may perhaps be tempting that, if we followed Weil, the προσφωνῶν in 1015 would seem to hark back to 997 τί νιν προσείπω. But this apparent harmony is deceptive, for προσφωνῶν points in fact to the following words ἀλγῶ μὲν ἔργα κτλ.: the προσφωνεῖν of the murderous cloak finds its expression in the ἀλγεῖν.
After finding ourselves forced to discard several ingenious attempts, we may now be disposed to do justice to a suggestion of a different type. That ruthless obelizer W. Dindorf applied his formidable scalpel to the speech of Orestes. But he went much too far, cut out twenty lines (987–1006), and thus killed the patient straight away. Of his excellent observations on certain passages of the speech something must be said later on. More cautious than he, Wecklein contented himself with bracketing 991–6 and 1005 f. The result of this operation seems to be very satisfactory. Before we go into details, it will be useful to examine the general consequences. Let us then provisionally accept Wecklein's deletion as a working hypothesis. What is left after the removal of 991–6 proves to be a consistent and highly effective piece of dramatic construction. At 983 Orestes gives the order to unfold the garment. Several attendants are wanted to hold the large drapery: they step round in a circle, grasp the border, and then slowly lift the thing and spread it out until it is fully displayed. This action, performed not in a hurry, but in the dignified manner suitable for a tragic climax, takes time:1 only at the end of 1. 990 is it completed. It is worth noting that the whole period from 983 to 990 is governed by only one primary clause: ἐκτείνατʼ αὐτὸ καὶ … δείξατε. This command is the pillar that supports the whole structure; all that follows is, both in syntactical form and in matter, subservient to the central thought and action. Therefore, if Orestes after 990, i.e at the moment when the spectators have obtained a full view of the instrument of crime, continues (997) τί νιν προσείπω κτλ., there is no break or abruptness whatsoever.
It is possible that at least some of the obscurities and the clumsy phrases which in 993–5 have puzzled the critics are to be accounted for by the assumption that this section was not written by Aeschylus but patched together (see below) by a botcher. To begin with 993, who or what is the subject of φαίνει? 'Nobody knows' (A. Y. Campbell, C. Q. xxix, 1935, 32). Many editors, e.g. Paley, Tucker, Headlam, Blass, regard τέκνον or τέκνων βάρος as the subject; Sidgwick 'as she shows'; Conington 'φαίνει seems to be impersonal … "as things show" '. Wilamowitz in his first edition connected ὡς φαίνει κακόν, which seems unbearable; afterwards he gave this up, without, however, indicating what he thought to be the subject. In 994 f. the construction is utterly baffling. As it stands (with θιγοῦσαν), its disentanglement seems hopeless, as may be seen from the determined, but unsuccessful, attempt of pg 814Wilamowitz (p. 242 f. of his commentary). The majority of the editors accepted Robortello's reading of θιγοῦσαν as θιγοῦσʼ ἂν, but then we should certainly expect μὴ δεδηγμένον (which here is metrically impossible) instead of οὐ δεδ. (cf. Wilamowitz)1 the scholiast, with disarming naïveté, paraphrases: καὶ τὸν μὴ δηχθέντα. Moreover, many editors have rightly objected to ἄλλον (Blass's excuse is not helpful, his treatment of the whole sentence is violent and unconvincing). I doubt whether any amount of critical ingenuity would be sufficient to rid these lines of their muddle, for this seems to be a case where incerta haec si tu postules ratione certa facere, nihilo plus agas quam si des operam ut cum ratione insanias. As for the content, I gladly subscribe to the verdict of Sidgwick: 'these violent and almost grotesque words'. In 996 there is nothing confused, but the sense, especially in connexion with the monstrous snake, strikes me as extremely weak and the phrasing as commonplace. The lines 991–2 are of better quality; it would not be surprising if this clause, or part of it, had been taken over from another tragedy.
But it is the general turn of thought in 991–6 and 1005 f. that, more than any detail, ought finally to shake our belief in the genuineness of these lines. Here W. Dindorf, despite his rashness, deserves high praise for his thorough insight into the character of the Aeschylean Orestes. He says (in the preface to his 5th edition of Aeschylus, 1870, p. xcvii f.): 'Perpetrata matris caede Orestes orationem suam his finit verbis [980–6], quae nullum produnt vehementiorem animi motum, sed solam recte convenienterque oraculo facti conscientiam, quod ut omnibus patefiat, Solem … invocat, ab conviciis vero in matrem iaciendis prorsus abstinet. Nam Orestes, licet apud Aeschylum pariter atque Sophoclem et Euripidem matris facinus aversetur, tamen cum ea et viva et post caedem, quam solo Apollinis iussu ab se invito et haesitante patrari identidem dicit, satis moderate pro criminis ab ea commissi gravitate agit, ut vel ex versibus huius fabulae 892–930 videre licet.' The last remark is particularly valuable.2 It may be added that the same noble restraint can be perceived in the words with which Orestes, Eum. 458–61, describes the murder of Agamemnon, and also in his moving allusion to his own crime, Eum. 611 (on the meaning of these words cf. my note on Ag. 1171, p. 534 n. 1).
From these observations the following conclusions may be drawn with a fair degree of probability (there can be no certainty in such a matter). The noble simplicity of Orestes' speech would not satisfy the coarser taste of a later generation; some people might complain that 'there were no sallets in the lines to make the matter savoury'. A clever producer, sailing with the wind of the aura popularis, and perhaps working in the interest of an actor who wanted a piece of rant to display the power of his voice, thought he could improve on Aeschylus by putting in a few lines of his own. In doing so, he freely stole feathers from the old master's plumage. As Dindorf saw, μύραινα (994) is a variation of the ἀμφίσβαινα in Ag. 1233 (there the expression is as suitable to the ecstatic horror of Cassandra as μύραινα is revolting in the mouth of the son), and ἔχιδνα comes probably from the very fine passage Cho. 249. The question 994 τί σοι δοκεῖ; seems to have been inspired3 by 997 pg 815τί νιν προσείπω …; I regard this very repetition as a strong argument against the genuineness of 994 ff., for Aeschylus would hardly, within a few lines, have capped the question 'what is my mother (or, as others understand the text, 'her deed') like?' with the question 'what name shall I give to the garment?'
So anxious was the amplifier to have Orestes abuse his mother more than once that he did not even refrain from spoiling the conclusion of the original speech by adding 1005 f. As long ago as Scholefield's edition it was realized that 1005 does not attach to the preceding lines. But the fact that there is no connexion in thought is not the only objection to 1005 f. It should also be noticed that the sentence 1001 ff. τοιοῦτον … φρένα (with most modern editors I regard Lobeck's correction as necessary) constitutes the conclusion of the whole speech: for the opening of the tail-end of a ῥῆσις by τοιοῦτον cf. on Ag. 613 f. It is intolerable that the formal ending τοιοῦτον κτλ. should be duplicated1 by the addition of 1005 f. τοιάδε κτλ. As regards the thought of 1005 f., it is not only vulgar, it is revolting.2 Fancy this Orestes at this juncture considering possibilities of marriage!
So much about the main issue. I want to add a brief remark on a point of minor importance. It seems to me almost certain that 1028 πατροκτόνον μίασμα καὶ θεῶν στύγος was not written by Aeschylus, but concocted by the same man who added 991–6 and 1005 f. It is the only other passage in which Orestes applies abusive terms to his mother. It destroys the fine structure of the preceding sentence, which culminates in the momentous οὐκ ἄνευ δίκης, exactly as the verdict on Aegisthus at 990 culminates in the concluding ἔχει … δίκην. 1028 is patched together from 1015 πατροκτόνον θʼ ὕφασμα and Ag. 1645 χώρας μίασμα καὶ θεῶν ἐγχωρίων.
1 See, e.g., Butler: 'non premam suspiciones meas versiculum in quo diserte vestem nominaverat Orestes hic [i.e. after l. 996] intercidisse.'
1 I am referring to the third act of King Lear, where Lear says (iii. 2. 67): 'My wits begin to turn', and again (iii. 4. 21 f.): 'O, that way madness lies; let me shun that; No more of that'. Then, iii. 4. 28 ff., he speaks of the 'poor naked wretches' in deeply moving lines full of humanity and wisdom, without any tinge of mental disturbance. A short time afterwards the words 'Hast thou given all to thy two daughters?' etc., and later on 'What, have his daughters brought him to this pass? …' (iii. 4. 48 f. and 62 ff.) and his following outbursts, reveal the first fit of unmistakable madness.
2 The scanty later remarks of Wilamowitz, Interpr. 215, can hardly be regarded as a substantial contribution. It may, however, be noticed that the line of psychological approach which he followed there had been anticipated, even in the details, by Paley.
1 Cf. P 391 f. δεξάμενοι (the ox-hide) δʼ ἄρα τοί γε διαστάντες τανύουσι κυκλόσε (κύκλωι Zenodotus).
2 Tucker maintains the arrangement of the MS, but gives no substantial explanation. From his comment on 997 τί νιν προσείπω, together with his critical note ('Weil strangely transposes' etc.), it appears that he is not even aware of the real difficulty. His commentary on the Choephoroe, published in 1901, shows no sign of any acquaintance with the commentary of Wilamowitz published in 1896.
3 In his second edition of Aeschylus, 1830, Scholefield speaks with remarkable reserve: 'Neque tamen diffiteor, olim me in suspicionem incidisse, versus hos a recto ordine detrusos fuisse. Certe facilius procederet oratio, si post 982 legerentur vv. 997–1004. Deinde v. 1005 ad Clytaemnestram optime referretur. Sed huiusmodi suspiciones omnino ipsae non sine suspicione sunt excipiendae, ne forte pro ipso Aeschylo exhibeamus "disiecti membra poetae".'
1 It can often be observed that a dramatic poet calculates the length of a speech or a section of a speech in accordance with the time required for a particular action on the stage (for S. Aj. 579–95 see T. v. Wilamowitz, D. dramat. Technik des Sophokles, 60). The speech of Clytemnestra Ag. 958 ff. affords a very striking example. At 957 the king begins slowly to walk from the place where the chariot stands (probably in the centre of the orchestra) towards the door of the palace, at 972 he is still outside, for the words ἀνδρὸς τελείου κτλ. are obviously meant for his ear, whereas the following outburst Ζεῦ Ζεῦ τέλειε κτλ. would be inconceivable if he could hear it. It is therefore clear that between 972 and 973 he enters the house.
1 Tucker speaks of 'the emphatic οὐ (instead of the possible μὴ), "though not bitten".'
2 This aspect of the scene had already been emphasized by Otfried Müller (Aesch. Eum. § 98, p. 195), who gave a fine general appreciation of the figure of Orestes in the Choephoroe.
3 The author of the ridiculous lines E. Or. 588–90 (the thought is even worse than the grammatical blunder which shows that the man was as unfamiliar with the difference between γαμεῖν and γαμεῖσθαι as the interpolator of E. Med. 262) got the cue for his insertion ὁρᾶις κτλ. from 591 ὁρᾶις. [Cf. Dem. 9. 6–8 εἰ μὲν οὖν … εἰ μὲν οὖν ….]
1 There is no such duplication in the apparently similar ending of Clytemnestra's speech Ag. 312–16, for there τοιοίδε … νόμοι concludes the description of the beacon-post, whereas the following τέκμαρ τοιοῦτον links the whole speech up with, and makes it explicitly an answer to, the question of the coryphaeus (272) ἦ γάρ τι πιστόν ἐστι τῶνδέ σοι τέκμαρ;
2 For the manner in which the speaker passes on to a wish that concerns him personally cf., e.g., E. El. 948 f. (Electra facing the body of Aegisthus) ἀλλʼ ἔμοιγʼ εἴη πόσις μὴ παρθενωπός, ἀλλὰ τἀνδρείου τρόπου.