It is clear that Sappho's poems, so far as we can reconstruct them, are not ‘narratives’ in the ordinary sense, even if we do not wish to accept the argument that lyric and narrative are naturally opposed. There is an obvious difference between the kind of story being told in Sappho fr. 94, where the speaker asks a girl who is leaving to remember their gentle pleasures together, and the extended heroic narrative of the Iliad. But how are we to frame that difference in terms of narrative? More specifically, can a poem like fr. 94 (already compromised by its incomplete state) count as narrative at all?

Perhaps the easiest place to start in considering the place of narrative in Sappho is her use of named characters from Greek mythology. There are several instances of mythological narrative in Sappho's corpus, where parts of stories might be spelled out in detail or where the simple reference to a name will be enough to set a story in motion. When Sappho mentions Helen, for example, in fr. 16, or has a messenger recount Hector's return to Troy with Andromache in fr. 44, or evokes Tithonus in the New Sappho, we know that we are firmly in the world of Greek storytelling.

This chapter sets out to look for what was distinctive about Greek practice in the staging of tragic stories. There is no doubt that this new art form of the late sixth and early fifth centuries was a highly original experiment, with no obvious model in other cultures, but of course the early dramatists did not have to start from scratch. The epic and lyric traditions offered them a great range of serious narratives which had already been shaped for performance, whether by rhapsodes or by choruses, and without this precedent it would be hard to imagine Attic tragedy having developed as quickly as it did into a genre of such richness, sophistication and popular appeal.

Once the early dramatists had taken the crucial step of representing imagined characters (particularly famous figures from the mythical past) as present here and now, enacting their stories before an audience as if for the first time, the dynamics of narrative must have needed reshaping to suit the new medium, and this is where we should be looking for creative experimentation. For example, when we consider the forms taken by narrative in tragedy there is no need to see them as determined by, or as hangovers from, epic practice, or to single out ‘messenger speeches’ as having a specially privileged status.

Scholars who work on Greek historiography tend to focus on the differences: Herodotus is the charming one who passes on local traditions; Thucydides is cynical about human nature and leaves religion out of history; Xenophon writes autobiographical history and peppers his work with intriguing dialogues; Polybius produced a handbook for statesmen, but is torn between Greece and Rome; Diodorus Siculus is only as good as his source; and so forth. But when we focus on these differences we often forget that, in fact, the works of Greek historiography, at least until the first century BC, are much more similar to each other than to any body of historiography that came after. In this chapter I shall try to define those characteristics of Greek historiography from the fifth to the first century BC that made it a distinct genre. Some of these characteristics can be hard to spot because they started a trend which either persisted into Roman historiography or has been reignited in modern historiography, but the sum of these various characteristics is nevertheless what makes Greek historiography Greek.

Before we begin, some remarks about my choice of material are in order. This study focuses on the Greek historiographers of the fifth to the first century bc whose works are extant in substantial form; that is Herodotus, Thucydides, Xenophon, Polybius and Diodorus Siculus. I stop at Diodorus because this volume concerns itself with what is distinctive about Greek narrative, and after Diodorus Greek historiography merges with Roman historiography to such a degree that it becomes impossible to keep the two traditions apart.

‘The second task required of a writer of an historical work (the first task being to select a suitable subject) is to decide where to begin and how far to go.’ The fundamental questions of ‘where to begin’ and ‘where to end’ must be as old as literature itself. They automatically pose themselves to any would-be poet or author. Moreover, passages such as Od. 9.14 show that, from an early stage, poets openly addressed the questions ‘What then shall I recount first, what last?’ (albeit here in the voice of the character Odysseus at the beginning of his long narrative to the Phaeacians). In all likelihood these questions quickly became a literary topos (possibly before Homer), which in turn could playfully be modified by subsequent poets and authors (cf. Hunter, this volume). They thus prepared the ground for literary critics to treat the whole issue in their analyses. No doubt, they especially needed to address this issue because those compositions which were the earliest extant works and also the undisputed masterpieces of Greek literature, the Homeric epics, displayed a remarkable temporal structure and thus intrigued readers.

Behind this volume lies the hope that we will someday achieve a general view of the history of ancient Greek narrative (henceforth, for simplicity, often ‘Greek narrative’) – that is, that we will be able to present a meaningful narrative about how the practices of telling stories developed within Greek literature, and that this history will contribute to the understanding of both Greek literature and narrative generally. Before anyone can write a history, however, the historian needs to be certain that the field has been meaningfully defined, both temporally and spatially. A narrative requires a beginning and an end, and a historical narrative also requires a decision about what the subject is. These decisions will determine much of the history itself, and the choice of boundaries is difficult and ideologically weighted.

While the ideological implications of literary history are sometimes less obvious than those of the histories of nations, they are very real, and the boundary difficulties present themselves immediately. Originality conveys literary value, and literary value can be important for a nation's symbolic capital: this volume began as a conference in Edinburgh, with its monuments to Robert Burns and Sir Walter Scott nearby. There can be few cities where the significance of the cultural and literary inheritance, and its interactions with power and national identity, are so manifest.

Greek melic narratives are on the whole small-scale compositions and as a rule illustrate some aspect of the occasion for which they are composed, which is sometimes specified, sometimes implied and sometimes left totally unclear. In this chapter I focus on the occasion in order to explore the stimulus it offers for the composition of new narratives either for the same or for a related, subsequent occasion.

Pindar's songs for the Emmenids of Acragas will serve as my main test-case because (1) they constitute an interesting sequence of songs which call attention to their occasion, thus allowing assessment of the stimulus it offers; (2) the later songs allude to, rework and creatively integrate the occasions of the earlier songs; (3) the Third Olympian combines two distinct occasions, Theron's Olympic victory in 476 BC and the periodic celebration of the Theoxenia of the Dioscuri in Acragas; and (4) the Second Olympian is the model for Horace's Odes 1.12, which offers a precious comparandum for the catalytic role of Greek occasions on narratives.

In the following discussion I use the term ‘occasion’ in its broad sense, namely agonistic events, festivals, other celebrations and performance contexts. Epinician songs often treat contests and epinician celebrations as separate events. A characteristic example in this category is the much-discussed Tenth Olympian, which opens with a poetic apology for the long time that has lapsed between Hagesidamus’ Olympic victory and the composition of the epinician song. At the other end of the spectrum, there are songs that create a continuum between the victory and the epinician performance.