Ancient divination had something of the status of modern science. Diviners, like scientists, were expected to offer public comment on all manner of subjects, from strange meteorological phenomena to proposals for military campaigns, from shooting stars to Star Wars. In both cases this public responsibility is based upon a supposed expertise in prediction. And the arguments of those ancient philosophers who attempted to justify such a status for divination run in some ways remarkably parallel to those of modern philosophers who have attempted to justify such a status for science. Thus we are told both by Stoics and by modern Scientific Realists that the world forms a single and unitary physical system, with nothing outside it to prevent the course of events from unfolding in accordance with the adamantine regularities which are the laws whereby nature operates. This determinism, and its consequence that the entire future of the world is already prefigured in the current state of things, makes it possible for us now to tell what will happen later. And that possibility is realised. Scientific predictions, we are told, may not be uniformly correct, but they are correct too often for that to be due to chance. The predictions of diviners, one must likewise concede, are not uniformly correct: nonnumquam ea quae praedicta sunt minus eueniunt (Cicero, De Divinatione, 1.24).

quae tuo ueniunt ero,
quanta gaudia, quae uaga 110
nocte, quae medio die
gaudeat! sed abit dies:
prodeas noua nupta.

In these lines, addressed by the poet to the wedding-couch of Junia and Manlius, a difficulty presents itself: uaga nocte does not cohere with medio die, since the epithets characterizing this polar pair of nouns belong to different categories. This was pointed out by R. G. M. Nisbet (PCPS n.s. 24 (1978)99) who saw medio die as the culprit on three grounds: (i) the incoherence of the pair of epithets, (ii) the inappropriateness of a reference to ‘sex in the afternoon’ in an epithalamium and (iii) the pointless repetition of dies in 112 after die in 111. As a solution he suggested emerito for medio, ‘when the day has done its service’, repeating the essential sense of uaga nocte (another reference to night rather than a change to day) but with a contrasting nuance of epithet: uaga would suggest the vagabond flightiness of night, emerito the solid service of the sun during the day. In a reply to Nisbet (PCPS 25 (1979) 69-70), Roland Mayer answered (i) by asserting the principle of an ‘illogical doublet’ (the incoherence between epithets is real but acceptable), regarded (ii) as mistaken and did not deal with (iii). More, I think, remains to be said, and elements of both these views may contribute to a different solution.

Two journeys are implied by the existence of Tristia 1: one, by a poet, a from Rome to the gates of the Black Sea; the other, by a book, from the gates of the Black Sea back to Rome. Each of these journeys is explicitly, and prominently, discussed in Tristia 1; and each makes its presence felt in various ways throughout Tristia 1. Leaving for another day the outward voyage, described especially in the second, fourth, tenth and eleventh poems, I am going to deal in this essay with the return trip of Ovid's book to Rome, as anticipated at some length in the very opening poem of the collection. And (because that is still a somewhat unwieldy topic) I am going to focus on the final destination of Tristia 1 within Rome, as specified in the last twenty lines or so of this first poem: viz: the bookcase in Ovid's Roman home. In these programmatically charged lines, the personified first book of exile poetry finds itself face to face with the poetry books written by Ovid before his exile. I want in the ensuing pages to take a closer look than is usually taken at some details of this and other encounters with Ovid's past writings in the first poems from exile; and my hope is that this analysis will tell us a few things along the way about how the poet is trying here to relate his literary present to his literary past.

This essay aims to discover and to illustrate one of the leading themes of Horace's first book of Epistles. This theme, good manners, despite its importance, is nonetheless neglected, or at any rate huddled up within the general praise of Horace's tact. Tact there certainly is, and we shall watch it in operation. But the poet would be setting aside that uerecundia which his patron praised (Epist. 1.7.37) if he were merely blowing his own trumpet. The display of tact and good manners perhaps serves other ends, and the suggestion of reasons for this choice of theme will form the substance of my remarks.

Some preparation of the ground is necessary. This essay completes an argument begun in an article entitled ‘Horace's Epistles I and philosophy’ which will appear in the American Journal of Philology for 1985. In order that this essay may be fairly complete in itself it will be useful to set out very briefly the considerations which induced me to look more closely at good manners. To do this most compendiously, I shall offer some quotations from E. C. Wickham's English commentary on the first epistle and explain why they appear unsatisfactory. Wickham is not chosen as a whipping-boy for any other reason than that he neatly expresses a common view; in all respects he is a serious and helpful student of the poems.

A paper on the herms needs no excuse. There is no proper treatment in English of the origins of the herm or of the erection of herms by Hipparkhos. The Eion herm monument has been well studied by Jacoby and Harrison, but the recent treatment by Clairmont demands their reconsideration. The scholarly literature on the mutilation has been very narrowly focused on the questions of who mutilated the herms and why. These questions are clearly closely inter-related, and yet the attempts to answer the question ‘why?’ have been singularly unsatisfactory. Some have arbitrarily chosen to emphasise a single feature of the herm, and with little evidence have made wild claims: thus Farnell could write that ‘the mutilation of the phallic Hermai of Athens produced… in the Athenians the despondent sense that the luck of the state was gone and the divine power of fertilisation impaired’, Crome that ‘with this mutilation must be connected the religious world of the phallus cult of the old native population which had certainly always been strange to the oligarchs’. Others have ignored the fact that it was specifically herms that were mutilated and have reduced the activity of the hermokopidai to a massive act of general impiety, as Dover does when he argues that the mutilation was politically significant because as an act of impiety it would seem to bring down the wrath of the gods on the whole community, as a conspicuous act of vandalism it would make the number of people involved seem large, and as an offence against nomos it would prompt concern for the fact that a body of men had set themselves beyond the law.

The reader I have in mind is a poet. My immediate interest is the example he provides of a writer's relationship with her or his reading. My aim is double: to suggest both that Homer illuminates the work of the later poet and that the later poetry can function as an interpretation of Homer which offers even to a scholar valuable ways of reading the epics, especially the Odyssey. Accordingly, I shall usually offer translations both of the modern and of the ancient Greek, since not all classicists know modern Greek intimately and those who study modern Greek do not always know the ancient language well.

Let us begin by reading one of Seferis' best-known poems. He wrote it in the Thirties and many contemporary poetic influences, both French and English, are at work in it. But I want to read it now from a special perspective, which I shall argue was crucial to Seferis through all his work. I shall read it as a search for a significant but bearable relationship in his own poetry with Homer and, through Homer, with the whole ancient poetic tradition.

It remains an open question what sort of structure, or what kind and degree of unity, the reader may legitimately expect to find in the Homeric epics, or in oral poetry in general. More specifically, when critics identify long-range correspondences, symmetrical patterns and precise correlation of speeches, scenes, similes and events, they are often confronted with considerable scepticism. Here it may be that excessively schematic attempts, such as those of the late Cedric Whitman, have shown such patent weaknesses that the whole approach has fallen into undeserved disrepute. Yet it is not unreasonable to suppose that any story-teller, oral or literate, improvising or premeditating, should desire and be able to connect his ending with his beginning, to develop themes in a recognisable and coherent sequence, and to draw attention to related scenes by direct repetition or allusive resemblances. Homeric scholarship has not been blind to such considerations, but they have yet, I think, to be systematically and sensitively applied. Much remains to be done, especially perhaps for the Odyssey.

In this paper I intend principally to argue that there is a significant resemblance (embracing equally important differences) between the finale of Odyssey 8 together with the opening of book 9, and the finale of book 21 and opening of 22. The situation in the later scene is modified, indeed reversed and distorted; the earlier scene presents the norm, the later scene the unusual and extraordinary; and these differences are stressed in such a way as to reinforce the development of major themes which run through the poem: themes such as hospitality, feasting, self-revelation, and the hero's status as both man of action and poet.