Actors recreate King Harold's 1066 march south after his victory at Stamford Bridge. Photo by Ian Forsyth/Getty Images


August 25, 2020   7 mins

In the early 5th century the Roman legions abandoned Britain, and the sceptered isle fell off the pages of history. When it reemerges two centuries later Celtic Britain had become the seedbed for the nation-state of England. The Christian religion, newly-established on the island at the time, had given way once again to paganism. Brythonic Celtic speech was ascendant only on the fringes. A cacophony of German dialects spread out across the fertile south and east, radiating out of the “Saxon Shore”.

This ethno-religious transformation of the island occurred under the shadow of semi-history, allowing for the development of an imaginative romantic tradition exemplified by the Arthurian Cycle. But this Dark Age also became a bone of contention between the English who saw themselves as deeply rooted in the land, and those who declared that they were a German folk who had won their new home through conquest and blood. The dominant view at any given time reflected social and political events of the 20th century more than facts. The propaganda value of myth meant more than the conjectures of scholars.

While in the early 20th century the dominant position was that the English were a people akin to German Saxons, a race apart from the Welsh, by the early 21st century serious scholars assumed that the spread of Anglo-Saxon culture occurred through imitation rather than replacement.

Today we can say with some confidence that neither stark view is correct, and that the middle path between is far more interesting and complex. Large numbers of Saxons, Angles and Jutes did in fact cross the North Sea — but the preponderance of England’s heritage still draws from the Celtic-speaking peoples. It is not coincidence that the earliest rulers in Alfred the Great’s lineage bear Celtic names, not German ones.

Those who argued for the erasure of the Celtic people did not do so without any basis. St. Gildas, a 6th century British Celt, recounted in On the Ruin and Conquest of Britain the defeat and destruction of his people at the hands of the Saxons. More recently, 19th century philologists observed that the number of Brythonic Celtic loan words in English is extremely small; in fact, there may be more Celtic loan words from Gaulish, due to the later Norman French influence. Finally, the collapse of institutions like the Roman Christian Church and the total decay of urban life indicates incredible disruption of the social hierarchy which characterised post-Roman Britain.

A contrast here exists with Gaul, which absorbed a German-speaking elite but retained Roman language and religion. Some of the nobility of southern and western France even traced their descent from Romans, not Germans. On a more demotic level, British archaeologists have also observed that the arrival of the Saxons seems to have been associated with a transformation of the layout of rural farmsteads. In most societies, farmers have customs and traditions which they hew to, and they are often quite stubborn and set in their ways. Such a change indicates new people, not just practices.

But by the late 20th century such views of cultural and demographic disruption were in bad odour. The dominant ethos is that people did not move, their customs and traditions did. Hengist and Horsa may have existed, but rather than a folk migration the Anglo-Saxon conquest was one of a small number of German mercenaries who were engaged in elite capture of the post-Roman peasantry.

The Welsh historian Norman Davies observed in his 1999 book The Isles that “blood price” in 8th century Wessex differed between whether one was Saxon or British, the implication here being that there were many Celtic Britons living in the Anglo-Saxon lands, even if our documentary evidence is from the Saxon elite; this would tally with the 6th century ancestors of Alfred the Great having names such as Ceawlin, Cynegils and Cerdic, all of which have a distinctive Welsh flavor.

Genetics has untangled the Gordian knot of this semi-historical mystery, although illumination has not come at once, and only in fits and starts. One of the primary reasons is that the genetic difference between “Celtic” and “German” peoples is very small. Most Northern Europeans separated from each other very recently. Ancient DNA from between three and eight thousand years ago shows that Northern Europe underwent several mass migrations which transformed the genetic landscape.

First, the blue-eyed dark-skinned hunter-gatherers who descend from Ice Age Europeans disappeared and were absorbed by brown-eyed pale-skinned farmers who moved north out of the Near East. Then, these agriculturists were themselves overwhelmed by a people who migrated out of the Eurasian steppe into Europe 5,000 years ago. These pastoralist people probably brought Indo-European languages, and 4,500 years ago they arrived in Britain as the Bell Beaker culture. Within a few generations there was 90% genetic turnover, as the farmers who first erected Stonehenge disappeared, and were replaced by people who seem to have arrived from what is today northern Germany, possibly prefiguring the later Anglo-Saxon migration.

The problem from the perspective of genetics in understanding the proportion of Anglo-Saxon ancestry in the modern English goes back to the reality that Germans and Celts themselves had only been separated for 3,000 years, at most. These are genetically very close populations, and the technology of the early 21st century could not resolve the questions being asked.

UCL geneticist Steve Jones did attempt such a thing in his 2003 book Y: The Descent of Men. Jones observed that the distribution of two Y chromosomal lineages exhibits a sharp break at Offa’s Dyke. A far higher proportion of Welsh men are R1b, which is very common across the Atlantic facade of Europe, while more English men carry R1a, which is found in higher frequencies in Germany and Norway. In contrast, Professor Jones observed that there was no difference in the maternal heritage of the Welsh and English, suggesting that the ethnic change was due to the impact of men. Jones’s UCL colleague Mark Thomas later developed an “apartheid model” to explain why the genetic difference between the English and Welsh was so striking.

But the true understanding of the situation could only be obtained by looking across the whole genome, not simply the paternal and maternal lineages. This was done by the Peopling of the British Isles Project, which published a paper in 2015 that drew from analysis on hundreds of thousands of genetic markers from 2,000 British individuals who were sampled from all across the United Kingdom.

They estimated that 10-40% of the ancestry in central and southern England was Anglo-Saxon — that is, DNA segments more similar to the Germans than the Welsh. Another paper from 2016, utilising ancient as well as contemporary DNA, estimated that 38% of the ancestry in the “East English” — people from East Anglia and the East Midlands — is derived from the Anglo-Saxons. These researchers actually found DNA from Dark Age-era graves identified as Anglo-Saxon, and some of these individuals were far more like the Germans in their DNA than the modern English; they differed from earlier Iron Age samples, proving beyond a doubt that a significant number of Germans did cross the North Sea in the 6th century.

Where does this leave us in relation to the question of whether the transformation of Dark Age Britain to early medieval England was one of genes or memes? The clear answer seems to be both. The emergence of a new style of farming, pottery and the collapse of urban Roman civilization and Christianity in eastern Britain was not simply due to the prestige and power of a small number of German warlords. Whole villages must have transplanted themselves across the North Sea, creating the nucleus of a new people, and absorbed the remaining British Celts. The lack of Celtic loanwords and the adoption of Saxon peasant culture may indicate the self-confidence of the newcomers. If St. Gildas is correct, the British elites moved to the west of the island, leaving the common people to their own devices.

But though the southern and eastern fringe of England has a substantial Anglo-Saxon demographic imprint, that fades out as one moves to the west, including to the lands that once comprised the kingdom of Wessex. There is far less German genetic influence in Hampshire, Berkshire or Wiltshire, let alone Devon. We know from early medieval records that Celtic language speakers did exist as late as the 8th century in these domains (and much later in Devon) but by then Old English, which is for all purposes a purely Germanic language, was dominant.

The genealogy of the House of Wessex may offer a clue as to what occurred in broad swaths of western England. In the 6th century Celtic names imply that this elite lineage was identified with British culture, and looked west, but by the 7th German names became common, and the kings were pagan. Though the Saxons may have imposed their way of life through sheer numbers in the east, explaining the light impact of British Celtic culture upon their folkways and language, their expansion beyond the Saxon Shore seems to have been due to the adoption of the German identity by native British. The killing of a Celtic-speaking individual under the Saxon system of blood price was far cheaper than for a German speaker, serving as a clear inducement to assimilate.

What science makes clear then is that both extreme scenarios presented in the 19th and 20th centuries were wrong. The English are not a race apart from the Welsh. The modern English are genetically closest to the Celtic peoples of the British Isles, but the modern English are not simply Celts who speak a German language. A large number of Germans migrated to Britain in the 6th century, and there are parts of England where nearly half the ancestry is Germanic.

These folk served as the focus of a cultural revolution that transformed the British Isles. It was not a passive affair: the cities, churches, and hamlets of the previous inhabitants were blotted out, and what had been one of the provinces of the Roman Empire became a backwater pagan land. Though the original Romano-British elites had some knowledge of Latin, and patronised the Christian Church, the patina of civilization was clearly thin upon them, and the loosely Christian Celtic warlords of Dark Age western Britain transformed seamlessly into the pagan kings of Anglo-Saxon England.

The initial founding of the Saxon Shore was surely based on a level of brutality that Christian priests, if any had lived to tell the tale, would have recorded with foreboding. But the transformation of vast swaths of western Britain into the core of what had become England by the Viking Age occurred consensually, so seductive had the Saxon society become to the Celts, highborn and low.

The lesson that history and genetics teach us that cultural change is a complex phenomenon, and a single factor does not explain the whole story. Today we live in an age of migration, and native peoples fear being replaced, while immigrant communities fear being assimilated. Numbers matter, but the Saxons tell us that numbers are not everything.


Razib Khan is a geneticist. He has written for The New York Times, India Today and Quillette, and runs two weblogs, Gene Expression and Brown Pundits. His newsletter is Razib Khan’s Unsupervised Learning