Who were the Celts
There are three main strands of evidence to which we can look when searching for the ancient Celts. The first comprises the written accounts of the classical authors, many of whom encountered the Iron Age peoples of temperate Europe at first hand (2). Secondly, we have the linguistic links between the surviving languages of the modern-day 'Celtic fringe' and languages spoken across vast tracts of central and western Europe during the Iron Age. And thirdly, of course, we have the archaeology. Unfortunately, these three strands do not tie together as neatly as we might hope.
Greek and Roman references to the Celts occur from around the sixth century BC, and in 390 BC they were encountered uncomfortably close at hand when Celtic war-bands sacked Rome itself as well as rampaging through Greece and establishing control over parts of Anatolia. Following the Roman annexation of northern Italy and much of Mediterranean France from around 200-120 BC, classical writers were able to extend their first-hand knowledge of the Celts under rather more controlled conditions. Posidonius, who became, via later copyists, the source of much of our knowledge of the historical Celts, may have travelled in Celtic Gaul at around this time (3).
There was, however, little consistency among the classical authors.The early Greek sources ascribe various territories in western Europe, particularly parts of France and Spain, to a people known to them as the Celts ('Keltoi' in Greek). Later writers tend to be rather more specific: Caesar, writing in the first century BC, for example, states that the Celtae (or 'Galli' in Latin) were one of three peoples occupying the southern and central parts of what is now France. Strabo, writing at around the same time, suggests that it was the proximity of these Celtae to the Greek colony of Massilia (modern Marseilles), founded around 600 BC. that had led to their name being applied generically to the peoples of the region. Thus, if there ever was a prehistoric
1 Celtic languages are traditionally divided into two groups: Q- Celtic, the more linguistically conservative form, comprising Irish and Scots Gaelic, and Manx; and P-Celtic, in which the archaic 'qu' and 'k' sounds had been replaced by a 'p' sound, comprising Welsh, Cornish and Breton. Other known Celtic languages -Gaulish, Brit tonic, Cumbric, Pictislt, Lepontic and Celtiberian — were uniformly extinct before AD 800 and are shaded in green on the map.
2 The war-leader Calgacus addresses the Scottish tribes before the battle of Mons Graupitis. A traditional Celtic image.
people who called themselves by a name resembling Celtae or Keltoi, they were probably restricted to a relatively small area of south and central France. Indeed, no classical author ever referred to the British tribes as Celts: that equation was made in more recent centuries by the extrapolation of linguistic and artefactual evidence.
By contrast, numerous closely related languages, which have long since been classed together as Celtic, are attested to over substantial areas of northern and western Europe (see 1). Indeed it is with this second strand of evidence that the problems start, for the link between the Celtae/Keltoi of the classical authors and Celtic languages as defined by linguists is not straightforward. It was the recognition of a 'Celtic' language family, embracing living languages, such as Scots and Irish Gaelic, as well as vanished ones such as ancient Gaulish and Lepontic, which first saw the term Celtic applied to Britain and Ireland. This language family was defined at the beginning of the 18th century by the Welsh polymath Edward Lhuyd, building on the work of the Breton scholar Paul-Yves Pezron.The choice of the name 'Celtic' to describe this linguistic unity was essentially an arbitrary one. Lhuyd could as easily have called it Gallic or Gaulish, or else made something up from scratch. However, at the time Celtic seemed a neutral enough term with a respectable classical pedigree. What we have to remember, however, is that the linguistic basis of this definition does not necessarily bear any relation to the criteria by which the classical authors defined the Celts. Nonetheless, following from Lhuyd s work the idea of an ancient and widespread Celtic culture became established in the popular consciousness of the eighteenth and nineteenth centuries. Far from being limited to language alone, this wider vision of a Celtic past was extended to all aspects of culture and identity.
Whatever we might think of the appellation 'Celtic' to describe this language family, it is nonetheless clear that people all across substantial areas of Europe formed a recognizable linguistic, if not cultural, group. This basic linguistic reality has tended to be overlooked in recent attempts to write the Celts out of prehistory. For our present purposes it serves to remind us that the communities of Iron Age Scotland had intimate links, at least at the level of language, with their continental contemporaries. But unlike the Greeks and Romans, who were fully conscious of their own identity and ethnicity, there is no evidence to suggest that speakers of what we would now recognize as Celtic languages felt any such cultural bonds.
Our third strand of evidence comes from archaeology. By the time the discipline first began to emerge in Europe, during the course of the 19th century, the linguistically derived vision of an ancient Celtic past was already well-established. Two sites in particular were to assume particular importance and lend their names to the two successive periods of the European Later Bronze and Iron Ages: Hallstatt in Austria, a vast cemetery and salt-mining complex, and La Tene in Switzerland, a ritual centre, where masses of decorative metalwork and other objects were recovered from the drained waters of Lake Neuchatel. It was the study of artefacts from those sites that enabled the construction of the first detailed chronologies for the European Later Bronze Age and Iron Age.
The creation of these archaeological cultures, Hallstatt and La Tene, seemed to have produced the material remains of the Celtic peoples known to the Greeks and Romans. Until relatively recently, archaeologists were content to equate regular groupings of material remains, or cultures, with prehistoric peoples. Thus all three strands of evidence, the classically attested Celtae/Keltoi, the Celtic language family, and the artefacts and art styles of Hallstatt and La Tene, could be bundled together to recreate a lost European culture.
However, this culture-history approach to archaeology can be very misleading. Hallstatt and La Tene material culture and decoration may have penetrated vast areas of Europe but they fail to correspond to any regular changes in settlement, burial or other
3 At its height, around 200 BC, 'Celtic culture', as measured by the occurrence o f certain artefact types and art styles, spread across much of Europe and into Anatolia.
aspects oflocal culture. What has often been seen as the expansion of a Celtic people need have been little more than the spread of art styles and fashions. There is no necessary or even likely connection between Hallstatt and LaTene artefacts and Celtic languages, which probably had a rather wider social and geographical currency: Celtic-speakers in parts of Ireland, Spain and Portugal, for example, seem to have largely eschewed LaTene material.While people who used LaTene items were most probably Celtic-speakers, not all Celtic-speakers used La Tene items.
These and other weaknesses with the traditional vision of the ancient Celts have caused much disquiet among archaeologists in recent years, especially those working in Britain. Increasingly it has come to be recognized that the British Iron Age is characterized by great variation at the regional scale. Some areas have hillforts, some don't; a few areas have formal burial traditions, most don't. How, it is argued, can we hope to explain such disparate Iron Age societies by appealing to vague notions of a homogenous Celtic society, drawn from fragmentary written records based on peoples in far-off places who left archaeological remains very different from those we see in Britain? The ancient Celts, it is suggested, provide an easy but inappropriate template which has enabled archaeologists to avoid addressing the real distinctiveness of local Iron Age societies.
These problems are certainly valid and it is true that there has been a tendency for archaeologists to take the easy option in their interpretations of Iron Age society, applying the limited written evidence to societies manifestly different from those