The Origins of the British: The New Prehistory of Britain (61 page)

In the absence of any genetic evidence, one useful message to be drawn from the Norman conquest is in its name. In spite of Rollo’s descendants changing their language from Scandinavian to French, they were nevertheless following a well-established trail, forged since the Neolithic, of men from the north coveting, invading and plundering Britain’s natural riches. For the Normans, riches derived from the wool trade, as measured previously in Danegeld, were probably high on the list.

In this concluding chapter I have pointed out that the last two invaders of England from the nearby Continent, the Vikings and the Normans, shared their northern origins with those who had invaded before, the Angles and Saxons. Furthermore, they both had existing cultural affiliations with the English targets of their invasion and used them to effect. William, Duke of Normandy, descended from Rollo, Jarl of Orkney, invaded England on the pretext that King Harold had sworn him an oath of allegiance. The Angles and Saxons, the previous invaders and rulers, were acknowledged even by Bede and Gildas to have been invited by British King Vortigern. The bulk of the 30% genetic intrusions, which help make England genetically so similar to the countries across the North Sea, had, however, arrived long before the Romans, yet even for these unnamed English there is ample evidence from the archaeological record of cultural links and
trade across the water. In the Early Mesolithic, that shallow strip of water was of course a landbridge, and East Anglia was merely a western continuation of the Northern European Plain.

The most important message of my genetic story is that three-quarters of British ancestors arrived long before the first farmers.
¹
This applies in varying proportions to 88% of Irish, 81% of Welsh, 79% of Cornish, 70% of the people of Scotland and its associated islands and 68% (over two-thirds) of the English and their politically associated islands.
²
These figures dwarf any perception of Celtic or Anglo-Saxon ethnicity based on concepts of more recent, massive invasions. There were later invasions, and less violent immigrations; each left a genetic signal, but no individual event contributed even a tenth of our modern genetic mix.

There certainly is a deep genetic division between peoples of the west and east coasts of the British Isles, particularly between the English and the Welsh, but this does not merely reflect the Anglo-Saxon, Viking and Norman invasions. These were only
the most recent of a succession of waves of cultural and genetic influx from north-west Europe, going back to the first farmers and before. Even the first settlers to come up from the Basque Ice Age refuge left different genetic traces on the east and west coasts of Britain. That difference was merely added to by subsequent migrations across the North Sea.

The second main issue, which should be resolved before any further discussion of British heritage, relates to how Celts are and were defined in terms of history, language, culture and genetic heritage. While most people of the British Isles hold an unambiguous, community-derived cultural identity, there are a number of different popular modern perspectives on Celtic national identity. Popular history books and the media now have more effect on these views than do orally transmitted folk memory and original written sources. Fashions among archaeologists and historians in the past couple of hundred years have in turn mis-informed the media. Fashions change like a pendulum, and the growing ‘Celto-sceptic’ view among archaeologists and others argues for a rejection of the term ‘Celtic’ as hopelessly corrupted and too vague in classical sources to be meaningful.

My view, argued in this book, is that Celtic ethnicity is a valid concept, both modern and ancient. I think the sceptics go too far and risk losing the baby with the bathwater. I do agree that the term ‘Celtic’ has become corrupted, but only to the extent of my rejection of the archaeological orthodoxy that ‘Celts’ as a people arose from somewhere in Central Europe, and during the Iron Age swept en masse across Western Europe and into the British Isles. There is no genetic or convincing historical evidence for
this conviction, and it is the main source of corruption of Celtic perceptions. The picture of ‘Iron Age Bohemians’ came from the archaeologists – not, as the sceptics seem to imply, from early antiquarian philologists. The latter correctly linked the insular-celtic languages of Brittany and the British Isles to languages spoken during Caesar’s time among Gauls living south of the Seine. Such celtic languages were later found inscribed on metal and stones in Italy, France and Spain from the late first millennium
BC
.

To the earliest historians of classical times, Celts were a real, defined Continental nation, not just ‘western barbarians’. I disagree with the Celto-sceptic view that the classical historians were collectively so vague and contradictory on Celts to the point of having nothing useful to say to us. They certainly differed in their degree of inclusion and exclusion, in the same way as many British do today when using such broad terms as ‘Asian’. But we can find sufficient cross-checks in classical sources spanning a thousand years to derive definitions better than those found, for example, in the older Webster dictionaries.

Celts in the British Isles have real cultural and linguistic connections to former Continental Celts. There is sufficient corroborated evidence in the classical texts to place Celts in south-west rather than Central Europe at an early stage – specifically, in France south of the Seine, in Iberia and Italy. Gaulish, Celtiberian and Lepontic inscriptions have been found in these same regions, dating from a few centuries
BC
and showing a clear linguistic relationship with insular-celtic languages. In short, the inferred location and languages spoken by people the ancients called ‘Celts’, need not be arcane or confusing, so long as the Central European Celtic homeland is seen for what it is – a modern archaeological myth.

Languages geographically associated with the ancient Celts and modern insular-celtic languages all have a common southwest European origin. I have used the recent literature in several disciplines, and my own re-analysis, to ask when celtic languages moved from the European mainland to the British Isles, with which culture, and carried by how many people. My answers are (1) Neolithic, (2) Neolithic and (3) not many.

Barry Cunliffe suggests that celtic language developed along the Atlantic fringe during the first four millennia of maritime trade, spanning the Neolithic and the Bronze Age, and was carried north into Ireland and Wales with Maritime Bell Beakers, by early metal prospectors from the south from 4,400 years ago. This suggestion has support from the material culture as shown by archaeology and in extrapolated dates for the legendary Irish king lists. Also, there appears to be at least one clear genetic colonization event in Abergele, north Wales, to match the archaeological evidence of an early copper mining colony there from 3,700 years ago. However, the genetic evidence, including Abergele, is more consistent with a Neolithic date for ‘Celtic’ arrival.

Following the Mesolithic, the gene flow up the Atlantic fringe to the British south coast dates mainly to the Neolithic, nearer the
beginning
of Cunliffe’s long period of maritime contact. Following new attempts to date the break-up of insular from Continental celtic, Peter Forster’s speculative estimate of 5,200 years for the fragmentation of Gaulish, Goidelic, and Brythonic from their most recent common ancestor is also in the Neolithic. Other date estimates from New Zealand argue strongly for the Neolithic being the driving force behind the spread of the entire Indo-European language family, which includes celtic. Given
this evidence stretching back to the first farmers, sceptical archaeologists should not be so ready to pull the plug on the idea of a common heritage between those ancient Celtic areas of south-west Europe and the celtic-speaking western side of the British Isles.

Gildas’ Dark Ages horror story of the Saxon invasion has generated the view that Celts were somehow
the
aboriginal population of the British Isles. The view that languages in Britain were 100% celtic when the Romans invaded is part of this false assumption. If ‘aboriginal’ means Neolithic immigrants speaking celtic languages who had replaced the former inhabitants of Britain, then nothing could be further from the truth. The genetic evidence does not support this at all. There was no Celtic replacement, any more than there was an Anglo-Saxon replacement.

The arrival of celtic languages and associated gene flow could hardly be classed as evidence for the establishment of a Celtic replacement of a former unknown British population on genetic grounds. The highest single rate of Neolithic intrusion from the Mediterranean route in the British Isles was in Abergele at 33%. But in Ireland, such Neolithic intrusion was only around 4%, while it was 2% in Cornwall, 6–9% in the two Welsh peninsulas, and 8–11% in the Channel Islands and southern England (
Figures 5.6
and 5.8). For England and the Channel Islands, the Neolithic contribution from the East via the northern route, just across the North Sea, was the same or greater than for the Atlantic coastal source (Figure 5.7).

In other words, Ireland and the Welsh peninsulas – which, on the basis of recent history and language, might be thought to be Celtic bastions – have less evidence of Neolithic genetic
intrusions, let alone from the Bronze or Iron Age, than anywhere else in the British Isles. Of course, the flip side of this is that their descendants are
truly
aboriginal and genetically represent the most conservative parts of the British Isles, retaining respectively 88% and 89% of their pre-Neolithic founding lineages (
Figure 11.5a
). And where do those founding lineages come from? They come from the same part of Europe, the southwest, but more specifically they match the equally conservative region of the Basque Country.

Ultimately ancestors for the modern Irish population, male and female did come from the same region as those ancient celtic inscriptions, but thousands of years before celtic languages. But then every other sample in the British Isles shows at least 60% retention of those pre-Neolithic aboriginal male founders, reflecting the very conservative nature of the British Isles after the Last Glacial Maximum.

Translating all this back to question the assumption that ‘Celts’, however defined, were the aboriginal peoples of the British Isles, we can see new perspectives, which depend on how that definition is applied. First, if Celts were to be defined by their languages, the small proportion of associated gene flow would make them an invading cultural elite with no stronger claims to aboriginal status than the Anglo-Saxons. If we focused more specifically on those 2–10% of immigrating southern Neolithic, Bronze or Iron Age genes as identifying people rather than language, they would be even less ‘aboriginal’ in Ireland and Wales than in the rest of the British Isles.

I think we should take Cunliffe’s gradualist concept of the
Longue Durée
of the Atlantic cultural network as a paradigm for the genetics, as Irish geneticists Brian McEvoy and Dan Bradley
of Trinity College Dublin, with English colleagues Martin Richards and Peter Forster, have done. Rather than being on the fringe of a celtic-speaking Neolithic revolution, the Atlantic fringe countries of Ireland and peninsular Wales then become the genetic aboriginal strongholds of post-LGM and Mesolithic gene flow from the Iberian glacial refuge, now best represented in south-west Europe by the equally conservative genetic profile of the Basque Country. The rest of Britain and the northern isles off Scotland then become more or less aboriginal with rates varying from 60% to 80% of ‘indigenous’ male markers (
Figure 11.5a
). In a sense, this is similar to the position taken on the Y gene group markers of ‘the indigenous population of the British Isles’ by geneticist Cristian Capelli (see
Chapter 11
), only my estimates for indigenous survival are much higher.

I feel that the genetic picture, both male and female, best reflects the broader picture of cultural and migration influences repeatedly moving up along the Atlantic coast to the western British Isles since the last Ice Age, so well described by Barry Cunliffe. Celtic languages need not be left out as irrelevant latecomers from this deep view of British ancestry, since they are part of the spray of repeated cultural waves along the Atlantic coast. We can also see that even if celtic languages did arrive on the western coasts of the British Isles as long ago as the Neolithic, there is no reason to assume that they spread universally in Britain, let alone remained universal during the Neolithic. What about languages arriving from across the North Sea before the Dark Ages?

The third question I have posed in this book is how old are the genetic and cultural divisions between eastern and western Britain. An even greater flow of Neolithic-dated genes (10–19% of modern gene lines) were impinging on eastern and south-east Britain from Scandinavia and north-west Germany, from across the North Sea, over the same Neolithic period (Figure 5.7). The male genetic evidence shows that even before the advent of farmers, the first pioneers were making a choice of east or west when they arrived at the mouth of the English Channel. Where this differs from the Atlantic coastal tin trade in the west is that we have no cultural clues with which to argue that the people in eastern Britain were, by default, speaking celtic languages like those in the west. Rather the opposite, since Neolithic cultural influences were arriving on the east and south coasts of Britain from north-west Europe, along with genetic markers of clearly north-west European rather than Iberian origin. The fact that Britain was one island is of no help in arguing for one language, since the highways of influence were maritime, not land-based.