The first (or last) word on the trickster-spider

This book is unashamedly presented as a new contribution to the familiar black-bound series of Penguin Classics. Yet it is not a compilation of written stories but of written-down stories. This anthology, in other words, is a collection of narrative tales that were not originally intended to be textual. It is, thus, a "classic" contribution to that most paradoxical of categories, "oral" literature.

The assembling of these stories in a single large volume thus raises interesting questions. Walter Ong once remarked that thinking about an essentially oral heritage as literature is like thinking about horses as automobiles without wheels.

Our starting point is that of the already literate; we cannot easily imagine a world without the written word. The purely oral recounting of the stories of the past is, by its very nature, a fluid event - eclectic and mobile - rather than adherent to any strict rules and regulations. But once monumentalised and made available in print or recording so that it can be repeated and referenced in one particular version, that version tends to influence all subsequent retellings.

This is not an isolated observation - the same is true of anthropological monographs. Revisiting a society that already has a substantial ethnography written about it can be a frustrating experience. The published account of any ritual often becomes a reference work for all future re-enactments. The intangible has been templated; it has been rendered, in a sense, tangible.

Ritual appears to have become what the Freudians saw it is as being: a form of scripted play-acting.

Myths such as those presented in this book can in their turn end up more like bedtime stories than fundamental accounts of the circumstances in which any given society - or even humanity itself - was created. This book poses these problems, and the selector of this anthology is aware of them.

It is hard, in such circumstances, to know how to approach these "texts".

Are we to analyse them as we would literature in the conventional sense, when any verbal subtlety is at several removes from the source? The narrative is usually that of a translator whose words have been reworked by the compiler, as here; certainly, the actual words used are no longer those of the original (now anonymous) teller of the tale.

We might want to read the stories as moral fables; but almost all of them have little overt moral purpose to which the general reader - to whom this book is largely addressed - could relate. Indigenous meanings, to the extent that they are recoverable, are locked up not just in the stories themselves but also in the wider ethnographic and historical detail of the cultures from which the tales come.

Wider research is needed by the specialist who might wish to contrive a convincing interpretation. Latter-day structuralists might be tempted to treat the tales as having some underlying conceptual scaffolding awaiting discovery. But how to do that when, usually, only one version is privileged above all the possible variants that might, in combination, prove revelatory?

All that, however, is the negative "reading". Where, we might reasonably ask, would the world of literature be without Aesop and Chaucer, or without the later transcribers of Homer? There is now barely a culture anywhere whose interactions are articulated in a purely oral mode.

Stephen Belcher, the compiler of this volume, is clearly aware of these issues, and where possible he does give alternative versions of some tales.

And, in any case, the tales in this volume are all drawn from existing written sources, most to be found in published sources dating from the first half of the 20th century - they are not being transcribed for the first time. For those worried about how to approach a volume of this sort, there are introductory materials designed to help.

Apart from an informed general introduction, Belcher has assembled more than 70 stories according to several broad themes - stories about hunters, about cattle herding and about the ubiquitous tricksters who inhabit such tales worldwide: examples quoted include an errant but ingenious child (among the Zulu), a trickster-spider (among the Zande on the Sudan/Congo borders and the Asante in Ghana) and the deity (or orisha) Eshu (Yoruba, Nigeria).

This thematic presentation is followed by a series of geographically arranged narratives that reveal the foundation myths of different kingdoms and chieftaincies that share a regional proximity and thus, potentially, a common environmental niche and possibly interconnected histories. Each story or related body of stories has a short scene-setting preface.

By their nature, the stories are all, in one way or another, concerned with questions of origin. Inevitably, given their oral sources, many will have been concerned in their oral retelling with questions of legitimacy. In the process, a stress on common origins and common identity emerges. Yet they also represent in one guise indigenous speculations about the world in which they and we live.

The principle that has been applied is to try to find the earliest recorded version of any story. Given what has been said above, this is clearly a sensible approach, for it promises to offer the version from which subsequent retellings derive. To that extent, it seeks to recapture the moment at which the purely oral slipped into the referential.

The very earliest of these are very old indeed, notably a series of mythic accounts from Egypt and Ethiopia, the two areas that escape the otherwise uniform focus on sub-Saharan Africa. Given the circumstances of their recording, however, most stories are unlikely to be versions with a time depth much earlier than the second half of the 19th century. Many will be familiar to Africa specialists; others, central to some well-studied cultures, will be missed in a wide-ranging survey such as this.

But an anthology of these stories, conveniently assembled in a single volume, will be useful to those with wider interests who lack either the time or the library access to allow them to be tracked down in the original sources.

John Mack is professor of world art studies, University of East Anglia.