¹ Certain ideas that I have offered before are here mustered combatively.

² Dawe, R. D., The Odyssey, Translation and Analysis (Sussex, 1993)Google Scholar, notes on 24.147. In the course of his destruction, Dawe (82, 832) scorns Combellack and Erbse more than anyone else whom he takes seriously. To me, Combellack makes a good deal of fault-finding look like mischief-making, and Erbse tows a number of ‘analytical’ arguments to the junkyard. Dawe speaks elsewhere as an advocate against the detractors of Aeschylus and Sophocles, just the opposite of how he here speaks about Homer.

CQ's careful, generous referee suggested that I also assess the arguments of West, S., ‘Laertes Revisited’, PCPhS 35 (1989), 113–43.Google Scholar What did Aristarchus and Aristophanes Byz. mean in saying, and were they right, that Od. 23.296 was the end of the poem? The paper thoughtfully decides that the ‘epilogue’ is not so deeply embedded as the Cyclops episode, but would all the same have satisfied its patron well. Dawe notes that at least some of the epilogue was regarded as authentic by Plato and Aristotle. I allow that to my eyes a mutilated corpus of Homer would be a hateful sight. (There exists a first draft of ‘Lycidas’, but I would not look at it for the earth.)

In the woodland of the Odyssey some trees are exotic and others appear propagated from over yonder. Borrowing a chain-saw from Kirchhoff, the analysts fell whole stands, so as to reach the forest primeval. I myself see strange growths, and have my own glimpses of an Urwald, but would not condemn a sapling. It is wrong to think that a storyteller or dramatist, let alone his personnel, such as Odysseus and Clytemnestra, will tell the whole truth with detachment. I do not agree with Niese, B., Die Entwicklung der homerischen Poesie (Berlin, 1882), 47Google Scholar, see 38, that the poet must be ignorant of a matter if he does not speak of it when he might appropriately do so. Is Achilles unaware in Il. 11.608 and 16.55–9 that the return of Briseis had been offered in 9.274? I find no contradiction out of character and accordingly cannot follow G. P. Goold, ‘The nature of Homeric composition’, ICS 2 (1977), 10, in believing that books 11 and 16 must have been fixed before book 9 in its present form was composed. A storyteller or dramatist should ordinarily be regarded as a master whose art we are dull to appreciate. It was a long while before I saw what is asserted in my title, or realized that when Aegisthus arrives in Choe. 838 it is with his bodyguard, contrary to what had been indicated in line 771.

There is also the question whether anomalies are not to be found in authors the world over. Fifteen years ago Bryan Hainsworth wrote to me of one in Moby Dick: ‘How did they steer Pequod? A tiller (Ch. xvi with some picturesque detail, and Ch. cxxiii) or a wheel (Ch. lxi and Ch. cxviii)?’ And the eyes of Emma Bovary, are they blue or brown or black? (On this matter from Enid Starkie, and others from John Sutherland, see Zalewski, D. in Lingua Franca [Oct. 1997], 17–18.)Google Scholar Hamlet usually seems to be around eighteen, but we learn from the gravedigger that he is thirty: see Weigand, Hermann J., ‘Hamlet's consistent inconsistency’, in Schueller, Herbert M. (ed.), The Persistence of Shakespeare Idolatry (Detroit, 1964), 137–72.Google Scholar

³ This is the best evidence, if that is not too strong a word, for how the poetry of Homer became lettered. But whether the Iliad and the Odyssey are more truly oral poems or written ones is hardly to be decided. In favour of the former see Janko, Richard, ‘The Homeric poems as oral dictated texts’, CQ 48 (1998), 1–13.CrossRefGoogle Scholar Dawe would seem to agree at least partly ([n. 2], 873): ‘To the layman blood is a thick red liquid, best left undisturbed. Put it under a microscope and you find it is in constant motion, its constituent elements incessantly joining together and separating. For Homer, multiple oral versions of the same episodes, being fused and unfused, best account for the never-ending sequence of inconsistencies.’ I would add that after the poems had been written down, fragments may have been fused and unfused, not just by a guild of editors, but now and then by a single poet who modified his story as ideas came to him, and in doing so left a sequence of anomalies that are once in a while, though not always, the better for being at odds with what we expected.

⁴ Segal, J. B., The Hebrew Passover (London, 1963), 95–106Google Scholar, penetrates deeply into the ultimate sense of nD3.

⁵ E. Wüst in RE 19.1: 461 (with a reference to Benseler) offers this etymology among others;Rank, L. P., Etymologiseering en verwante verschijnselen bij Homerus (Assen, 1951), 66Google Scholar, does not find the others worth mentioning.

⁶ West, M. L., ed. Hesiod, , Works and Days (Oxford, 1978), 368–9Google Scholar, brings forward, from J. Chittenden and R. Carpenter, the idea that ργειφóυτης, an epithet of Hermes, means ‘dog-slayer’. The history could be: (i) that the epithet ργóς became the name of Odysseus’ dog, thanks to a truly Homeric inspiration; (ii) that ‘Argus’ consequently, or else independently, became the name of many another dog; (iii) that it afterwards meant ‘dog’ all by itself; and (iv) that ργειφóυτης with the sense ‘dog-slayer’ was created for Hermes, seeing that a god of thievery would kill any guard dog he encountered. In my opinion the first step is more than likely. The second is rather unlikely, and the third, highly unlikely. It is as if ‘Traveller', the name of the horse of the Confederate general Robert E. Lee, came to be the name, at least in Virginia, of other horses as well, and then came to mean simply ‘horse’. Step (iv) is remotely possible, but ask yourself, if you wanted to say ‘dog-slayer', would you use the word apyos at all?

⁷ In the Iliad and the Odyssey as they are, the basis for choice among the epithets—ξαυθòς Mευ⋯λαος or βo⋯υ ⋯γαθòς Mευ⋯λαος? με⋯λιυoυ ἓγχoς or δoλιχóσκιoυ ἒγχoς?—was not meaning but metre. For this elementary, though overwhelmingly important, observation the credit is owed to H. Düntzer, Die homerischen Beiwörter des Göiter- und Menschengeschlechts (Göttingen, 1859), 7: ‘So bildete sich die epische Dichtung fur besonders bedeutsame Gegenstande eine Anzahl verschiedener stehender Beiwörter, unter welchen der Dichter nach Bedürfniss des Verses und der Abwechslung frei wählte, ohne sonst auf den Inhalt der betrefFenden Stelle Rücksicht zu nehmen.’ What Parry—L'Epithete traditiomelle dans Homère (Paris, 1928), see esp. 50–1— added, also elementary though overwhelmingly important, was an emphasis upon the formula type: γλαυκπις ’Aθυη, λευκώλευoς ‘’Hρη, and δoλóεσσα καλυψώ can combine with the same predicate;θα γλαυκπις ’Aθυη θα λευκώλευoς ‘’Hρη and καλυψώ δα θεωυ can combine with another.

Where I regard Parry as mistaken is in saying that the epithets were more narrowly restricted each to a single person, and more truly apt in their contexts, early in the tradition than later.Heubeck, A. in Gymnasium 89 (1982), 420Google Scholar, agreed with me that Parry's sequence ‘particularized → distinctive → ornamental → generic’ ought for at least many instances to be exactly reversed. Parry (239) held that Diomedes, often described as ιππóδαμoς could just as well be υτθεo. That is untrue for the Iliad, though it might have been true for an earlier stage of the epic tradition, before Diomedes had become characterized as the breaker of horses par excellence.

⁸ Tayler, J. G., ‘Some notes on the Homeric shield’, CR 27 (1913), 222.Google Scholar