Literature in the Greek and Roman Worlds (OUP), edited by Oliver Taplin, appeared last year. Written by 'new' classicists, their bibliographies concentrating on the criticism, theory and scholarship of the last two decades, the book has a mission: to redefine the trajectory of classical literature by introducing at the heart of things, as an element in the actual creative process, the audience (civic or sympotic), the 'receptor'. The newness of the bibliographies is itself instructive: classical studies have been reborn, not only because of new theories, but because of new texts which the Egyptian dunes have generously yielded over the last century, papyruses which adjust our sense of what has come before and which must affect our attitudes not only to the classics but to our own culture's partial and often wrong sense of them.

The second essay in the book, written by Leslie Kurke, Professor of Classics and Comparative Literature at the University of California, Berkeley, is entitled 'The strangeness of song culture : Archaic Greek poetry' (she contributes an excellent essay on Herodotus and Thucydides, too). The strangeness she discovers underscores the utter otherness of the poetry of ancient Greece and, by inference, of other ancient cultures, an otherness which five centuries' scholarship has been slow to identify. She makes clear how ingrained habits of (mis)reading falsify in order to assimilate the unfamiliar to our own sense of historical and cultural order: our need to identify 'modern' subjectivities at the heart of lyric poetry, for example. If she overstates her case, it is worth overstating.

The poetry of Hesiod, Archilochus, Sappho, Alcman, Alcaeus, Anacreon and others is didactic, not always in that it teaches fact or precept, but in that it creates voices and scenarios which a singer or reciter can try on, different circumstances and identities, if you like; and the audience too considers and appraises them. Much of the poetry is specifically about initiation or devised for initiation rituals, formalising the transition from childhood to adolescence, from adolescence to adulthood. 'Paedophile' sympotic verse may begin to assume quite a different aspect when read as a poetry of recognition of the (usually noble) subject reaching adolescence, alive with manifest promise such verse may purpose to draw the subject into a social group, to socialise him. Indeed, the civic function may be the very element we overlook when we consider such work as essentially erotic.

Various challenges to modern readers and writers follow from Kurke's essay. What is hardest to understand, and at the same time most suggestive, in her argument is that we cannot expect the qualities from ancient lyric poems that we do from the modern. She insists on difference. 'Because Greek lyric poets are found speaking for the first time (almost) in the first person in their texts, there is a great temptation to assimilate them to modern notions of lyric subjectivity. Thus, it is often said that with archaic Greek poetry, the individual I first emerges onto the stage of history, with lyric subjectivity inexorably succeeding the objective form of epic as the Greek spirit develops. Both of these claims (lyric as the invention of the self, and the organic development from epic to lyric) are romanticising modern projections that fail to take account of the cultural specificity and difference of ancient Greek poetry.'

The 'I' is not an individual but a characteristic voice; such detail as a poem gives is characteristic rather than specific. The presence of lyric poems in the context of the epic, 'harvest song, wedding song, paean to Apollo, threnos or mourning song', add strength to the view that lyric prosody is as old as or older than epic hexameter. 'What suddenly enabled the long-term survival of the lyric in the period under discussion was, paradoxically, the same technological development that ultimately ended the living oral tradition of epic composition-in-performance: the invention of writing.' Poetry was sufficiently important to be recorded in a medium more durable than popular memory. As soon as there were physical mnemonics, the need to remember receded.

Early lyric poetry survives thanks to writing, but it was an oral medium. 'Greece down through the fifth century has aptly been described as a song culture ,' she insists. There were occasions for song, they were 'embedded' in a culture which was formalised and to some extent ritualised. Lyric was not a personal outpouring but a song to be made in a specific place within the Greek day or night. The place, the 'occasion' of the poem, was culturally, not geographically or psychologically specific. It is this sense of the characteristic experience, metaphor or image that it is hard for readers not steeped in Renaissance or neo-Platonic culture to accept as at once poetically valid and original. The most suggestive passage in Kurke's essay is this: 'embedded poetry was one medium (along with, for example, family, military service, and forms of commensality) for constructing individuals as social subjects. This formative process applied to both the singers and the audience of early Greek poetry, since, throughout most of this period, the singers would have been non-professionals and members of the same community as their listeners, whether that community be the entire city or a small group of companions at the symposium.'

By contrast, modern poetry for the most part deconstructs 'the individual as social subject'. The idea of a normative literature, with specific social occasions and functions, is now remote: the last poet to have practised such an art may have been William Cowper, struggling back from the wild waves of madness and self-doubt to the terra firma of common experience. He was a not negligible translator of Homer.

The idea of 'voice' in poetry has undergone an alteration in modern times, to such an extent that it means quite the opposite of what it did even in Wordsworth's not-so-distant generation. It is no longer a question of 'a man speaking to men' or the characteristic voice of a certain type of person speaking in a way that defines a type; it is the individuated and apparently individual voice with mannerisms, eccentricities and irreducible subjectivities that we look for, and if we fail to find it we declare the poem unoriginal.

At the end of the archaic period the changes that occur are not unlike those which mark the stark transition in any culture between the period of space, purpose, creative confi- dence, and the formalised space of criticism, analysis and anatomisation. Kurke says: 'Timotheos' Persians was a tour de force, a kind of compendium of the whole tradition of lyric rhythms, styles, and generic forms. As such, it helps us understand the ways in which the New Music contributed to the end of a living performance culture. This was not because (as Plato and Aristotle would have it), this New Music was morally decadent and corrupting; it was rather that, in its very virtuosity, it required performance by professional musicians.'

The amateur is excluded. Adolescents moving into adulthood no longer perform the poem of transition: instead, a professional is employed to perform it. The sense of cultural professionals meant that 'song culture' was destined to fail. There was a change in educational style. The rise of the Sophists led to educational specialisation: poetry, no longer communal, was 'increasingly privatised and professionalised'. Systematisation, the categorisation of areas of knowledge: the grammarian and the pedant had their day. Education became a matter of privilege. And writing was not a friend of living memory.

Poetry was displaced less by drama than by the 'rhetorical displays of the law courts and assembly Political and forensic oratory became the sites for the negotiation of speaker and audience, for the collaborative construction of ideal community that had been the ideal of choral poetry in performance.' Public oratory was in conflict, then, with private literary production. Lyric poetry lost its social role and ideological function.

What is to be learned from the new classicism and its altered purchase on the form and character of classic poetry? Less about the structure of the lyric, always altering and alterable, more about the responsibilities of a poet, about constitutive rather than alienating art. Something about the ways in which audience can enable (and disenable) high art. Kurke is not alone in insisting that the civic and sympotic audiences which enable an art are informed and trained by that art, recognise originality and variation and can with equal assurance identify a fault, a short-cut, an artistic deception, whether it be prosodic, thematic, textural or in narrative.

It is no accident that the earliest poets Orpheus, Arion, Musaeus and Linus are mythical; we know and can know next to nothing of the lives of the poets through to the fifth century BCE. But apparent specifics do not matter: what matters is the universality of theme and application of a poem. Homer's poems do not adopt a single diction, ideology or geographical focus; the specifics we find in Hesiod hardly root the poem: Hesiod is not Wordsworth in the Lakes or Hardy in Dorset. He has more in common with the didactic generalities of Thomas Tusser and Erasmus Darwin. Poetry can find bearings in such functional verse; the modern lyric might usefully look to the pre-Hellenic classics to see how a tradition can be serious and central. In that classicist Anne Carson's phrase, we are looking for 'unexpectedness': what more unexpected place to find it than at the forgotten beginning of our journey?