Archaeologist Simon James has been called an English nationalist and been accused of perpetrating genocide. Why? Because he dared to argue that the Celts never actually existed. Barry Cunliffe sorts fact from fiction

When, earlier this year, archaeologist Simon James declared Celtic civilisation a myth, an idea invented for political reasons in the 18th century, the cry of fury that went up from nationalists was hardly unexpected. After nearly 300 years, just as Wales and Scotland were about to be given a degree of autonomy, the very root of the nationalists' identity, based on a carefully nurtured belief in their Celtic otherness, appeared to be under attack.

For, whether it be a myth or not, "Celtic identity" is passionately felt. Every small-town book shop in Brittany has shelves of Celtic books and at the many summertime fest noz, held in Breton communes, Galician dancers, Irish bands and Scottish pipes are to be seen (although the arrival at one gathering of Morris men on bicycles was slightly incongruous).

But how "Celtic" are the Welsh, the Scots, the Irish or, for that matter, the Cornish, Bretons or Galicians? The debate reignited by James has stirred surprising passions. Some academics believe that something that is properly the province of academe has been hijacked and distorted beyond recognition over the past three centuries by nationalists, mystics and others in need of a useful fiction.

³ÉÈËVRÊÓƵ

Certainly, the Celtic label is being exploited by publishers. One book-club flysheet offers, under the title A Celtic Odyssey: the Beauty and Wonder of a Lost Civilisation, 19 books and two videos, most with the word "Celt" or "Celtic" in their titles. It is an eclectic mix from solid tomes by archaeology professors to what can only be described as fringe literature with such titles as The Celtic Wisdom Tarot Pack and Celtic Tattoos (complete with transfers and body paint - real woad, do you suppose?).

The interest lies in what these outpourings represent - a resurgence of Celtomania. But where did it all start and how much academic evidence is there that a Celtic culture and people ever actually existed?

³ÉÈËVRÊÓƵ

The first wave of Celtomania began in 1703 when a Breton cleric, Paul-Yves Pezron, published his Antiquities of Nations. Pezron's work impressed the Oxford scholar Edward Lluyd, whose Archaeologia Britannica (1707) laid the basis for modern Celtic comparative philology, the study of different "Celtic" languages, including Welsh, Breton and Gaelic.

This "Celtic" paradigm was taken up by 18th-century antiquaries keen to produce a context to explain enigmatic prehistoric monuments such as Stonehenge and Avebury. It was decided that these were the work of our Celtic ancestors - the preserve of druids, whose activities were explained by classical authors such as Livy and Polybius. Thus, when the British antiquary William Stukeley decided to write a prehistory of Britain, it was understandable that he should call it the History of the Ancient Celts. He never completed the work, but a volume on Stonehenge was published in 1740 and one on Avebury in 1743.

Mute monuments, gleanings from the classics and a lively imagination, however, were not enough. "Les Celtomanes", as they were scathingly styled by a late 19th-century archaeologist, wanted their heroes to be literate and so the search for Celtic texts began. Many genuine Irish and Welsh works, preserved in medieval manuscripts, were brought to light and, where a literature did not survive, it was invented using scraps of oral tradition.

The most notorious case was the famous poems of Ossian manufactured by the Scot James Macpherson in the 1760s and admired throughout Europe. The Welsh fared little better when, in 1819, the honourable ceremony of the Eisteddfod was augmented by an "ancient tradition" called the Maen Gorsedd - entirely invented by a Welsh stonemason in 1792.

To this rag-bag of fact, fiction and fantasy, archaeology increasingly contributed. Excavation at Hallstatt in Austria, at La T ne in Switzerland and in the Marne region of France produced an array of artefacts spanning the seven centuries before the Roman invasion. A breakthrough came at an international conference in Bologna in 1871, when it was agreed that the similarity between artefacts from north of the Alps and those recently discovered in the Po valley in Italy could be explained in terms of the migration of Celts described by Roman writers.

By this time Celtomania had developed a distinct nationalistic flavour. In Brittany, La Villemarque, who had earlier gathered together Breton oral traditions, called the first "Interceltic Congress" in 1867, inviting "compatriots" from Wales, "brothers" from Cornwall and "cousins" from Ireland and Scotland.

³ÉÈËVRÊÓƵ

This was one of many attempts in which "Celticness" was used as a metaphor for "other" - "we are not French, we are not English, we are Celtic". By emphasising the "Celtic brotherhood" of the Atlantic provinces, these remote regions created an identity that gave strength to their desire for self-government.

Thus described, the evidence for the existence of Celts seems slight. Yet many Greek and Roman writers of the first millennium BC believed that there were Celts occupying parts of Europe. For Ephorus, writing in the fourth century BC, the Celts were one of the four great barbarian peoples (alongside Scythians, Persians and Libyans), but earlier writers, such as Hecataeus and Herodotus, gave a more specific geographical focus, implying that there were Celts from the Atlantic coasts of Portugal to the eastern Alps.

³ÉÈËVRÊÓƵ

These anecdotes, for that is all they are, suggest that Mediterranean observers recognised a similarity between some of their northern neighbours, which they chose to express by placing them together in the ethnic category Celts. Whether these disparate people recognised themselves as part of the same group is debatable but of those living in Gaul in the mid-first century BC, Caesar is specific: "IWe call them Gauls, though in their own language they are called Celts."

Any educated Greek or Roman would have been able to offer a description of a Celt - fond of war, exuberant, prone to drunkenness, superstitious, indulging in human sacrifice - the stuff of ethnic stereotypes. We are presented with the Celt as the antithesis of the civilised Mediterranean citizen - his otherness no doubt contorted into an easily understandable model.

The Celtic language presents a different set of problems. Philologists generally agree that a language group, which they have chosen to call "Celtic" since the early 18th century, was, by the first century BC, widely spoken in western Europe from Ireland to Iberia and fromthe Atlantic to the eastern Alps and beyond. How, when and where the language arose is undecided. The traditional belief is that massive folk movements were involved in its spread, but an alternative view - that the languages developed largely in situ and are of greater antiquity - is beginning to gain sway.

So much for the linguistic and classical arguments - what about the archaeological evidence? So-called "Celtic art'' - a distinctive style of decoration found across much of central and western Europe - was used to ornament weapons and jewellery. Yet the spread of "Celtic art" cannot be claimed to represent Celts. All that can safely be said is that the desire for an item decorated in this way implies sympathy with the belief/behaviour system of which the object is a symbol (like the appearance of the Chinese teacup in 17th-century England).

So where do these different perceptions of Celtic - the Romans' view of the Celt as "other", the philologists' coining of Celtic as the name for a language group and the use of "Celtic" to describe an art style recognised in the archaeological record - leave our Celts of today?

What we can say is that today's Scots, Welsh, Irish, Cornish and Bretons are the direct descendants of peoples who occupied the same lands in prehistoric times. Remarkably, they still speak a version of the ancient language of their ancestors. These are things we should not lose sight of in the smoke from the academic grape-shot.

³ÉÈËVRÊÓƵ

Barry Cunliffe is professor of European archaeology at the University of Oxford and author of The Ancient Celts, Penguin paperback.