The migration question: re-grounding Anglo-Saxon archaeology

Was the arrival of the ‘Anglo-Saxons’ in England a limited elite takeover, an invasion force, or a mass movement of migrants? Perceptions of the immediate post-Roman period have shifted and evolved over the centuries. John Hines places the new genetic data explored in this issue in the context of research history.


One of archaeology’s great strengths is that it is rooted in concrete reality: it uses material evidence to uncover what happened in the past, especially in terms of how people lived at different times and in different situations. In this way, it also contributes vitally to understanding how humankind interrelates with its material resources and diverse environments. This is why the greatest step-changes that advance archaeological understanding are always empirical – based on observable data or experience – rather than developments in theory. These advances are driven by the discovery of new evidence, an unceasing process that, since the 1980s, has been greatly amplified by the growth of metal-detecting and the Portable Antiquities Scheme, both of which have changed the very ground on which archaeology stands, both in terms of the quantity and the distribution of known finds.

John Speed’s map of c.1611 depicts the kings who reportedly conquered parts of southern Britain and founded the earliest Anglo-Saxon kingdoms there in the 5th and 6th centuries – as imagined in the early 17th century. Here we see Hengist of Kent, Ælla of Sussex, and Cerdic of Wessex. IMAGE: Cambridge University Library

The framework of ideas within which we interpret and understand these raw facts is also constantly developing, thanks to a huge array of precise, rigorous data delivered by laboratory sciences. The range of new information now at our fingertips is vast and very exciting. The radiocarbon revolution which has transformed dating methods in recent decades (see CA 259) is still benefiting from refinements that make radical differences to our understanding of the past. Meanwhile, analysis of residues in metalworking crucibles has enabled us to reconstruct detailed production processes in ancient craftworking of quite unanticipated kinds. And, most recently, biomolecular studies of human, animal, and plant remains have uncovered deep wells of evidence that could only be dreamt of even a few years ago. Now we have robust archaeogenetic data on which we can draw to help illuminate the population changes involved in the transition from the Roman to the Anglo-Saxon period in the south and east of England. (See CA 366 for a fuller discussion of the appropriateness of using ‘Anglo-Saxon’ as a historical label, and how the inhabitants of early medieval England viewed themselves.)

This is not to dismiss theory as mere babble: it has an essential and complementary place in this discussion. After all, nobody can free themselves from perceiving the world around them through the lens of their own experiences and ideas, and while that does not necessitate or excuse fanciful subjectivity, thoughtful and careful self-awareness can help to counteract such limitations. Many theoretical positions have broadened and sharpened our understanding without requiring absolute ideological commitment: consider, for example, how feminism has led the way towards a subtler appreciation of the complex relationships between biological sex and cultural genders, or how Marxism has compelled awareness of the mutual creation and reinforcement of social and economic power relations. More recently, agency theories have helped us to see how material culture itself stimulates human activity, rather than being purely instrumental in human hands.

Ultimately, the role of theory is to help us interpret facts and assess their significance. Sooner or later we reach the limit of what we can securely deduce from the available evidence, and at that point theoretical approaches may identify where we might seek more evidence to help clarify matters. When reconstructing past events – using archaeological evidence to write history – we often reach a point where we can realistically hypothesise beyond what we can see, but even well-informed and plausible speculation is not solid knowledge.

ABOVE & BELOW: Metal-detectorist discoveries recorded by the Portable Antiquities Scheme have transformed the number of finds available to scholars. This early medieval selection comprises a 5th- to early 6th-century cruciform brooch from Lincolnshire; an early 6th-century disc brooch from Kent; and a late 7th-century sceatta from Suffolk. IMAGES: Portable Antiquities Scheme/Kent County Council/Suffolk County Council

Images of invasion

In cases where both historical and archaeological evidence have to be accommodated, few questions have proved as fractious as how Roman Britain was succeeded by Anglo-Saxon England. Controversy has particularly blazed over the past half century – arising from a desire to apply new trends of thought in prehistoric studies to this historical context.

Throughout human history, people have defined themselves in opposition to those perceived as ‘Other’. In Europe, ethnographic ideas that language, customs, culture, and group-identity largely coincide can be traced at least as far back as ancient Greece. These views tend to be coloured by contemporary concerns: in the world of medieval Christianity, ideas of identity and the Other were shaped by the major religious divisions of the time, while by the 15th and 16th centuries, Renaissance Humanism – and responses to unfamiliar experiences, discoveries, and challenges in the ‘New World’ – ushered in a long reappraisal of how Europeans saw themselves, and how they categorised people seen as outside this group.

The Enlightenment and the 19th century saw the advent of the sciences of humankind – physiology, psychology, sociology, and anthropology – which sought to define systematic categories of difference. Come the second half of the 20th century, the post-war generation of archaeologists inherited a subject strongly shaped by a ‘culture history’ paradigm that assigned distinct cultures to specific peoples and tended to explain major archaeological shifts in terms of conquest and population replacement. For a striking and widely available example of that worldview, see Christopher and Jacquetta Hawkes’ Prehistoric Britain, a Pelican paperback of 1943 in which recurrent references to inundations of ‘warlike races’ now look quaint and antiquated, to put it as kindly as possible. In a 1966 edition of the journal Antiquity, though, Grahame Clark published a seminal article attacking ‘the invasion hypothesis in British prehistory’. His arguments stood on empirical grounds, although Clark also nailed his ideological colours to the mast by referring to an ‘invasion neurosis’ and ridiculing the earlier generations as ‘archaeologists of the era from Kipling to Churchill’.

That is not to say that Clark denied that any major cultural transformations had followed large-scale population movement; he accepted that this was the case for the spread of the Beaker phenomenon at the dawn of the Bronze Age in Europe (see CA 338), the Romanisation of Britain, and the Anglo-Saxon and Norman invasions. Within just 15 years, though, referring to the onset of the Anglo-Saxon period in terms of population change was being treated as ludicrously passé. The alternative minimalist position in demographic terms, shifting the explanation for this change to impersonal material ‘processes’ and transmission of ideas rather than people, is usually labelled ‘elite takeover’ and contrasted with ‘mass migration’.

ABOVE & BELOW In 1943, Prehistoric Britain imagined an invasion of ‘warlike races’ – something vigorously refuted by Grahame Clark in his influential Antiquity paper of 1966.
IMAGE: Antiquity

Migration matters

Those who dissent from ideas of a limited elite takeover are often accused of relying credulously on a handful of near-contemporary textual sources, but the situation is not so simple. As our colleague Kazutomo Karasawa, Professor of English Philology at Rikkyo University in Tokyo, has shown through his own recent research (see ‘Further reading’ at the end), it was many centuries before any Anglo-Saxon writer presented an account of history in which Britons had been numerically overwhelmed and almost totally displaced by his own ancestors (see CA 346 for more on written sources for this period). The only ‘invasions’ specifically recorded in early medieval sources like Gildas, Bede, and the Anglo-Saxon Chronicle are notably small-scale and restricted to the south-east coast. Much vaguer, although consistently alluded to, is a further wave of colonists that followed these initial incursions.

Writing in the 6th century, the Christian British scholar Gildas emphasises the wolfish ferocity of the Saxons (whom he names once) in his polemical sermon On the Ruin and Conquest of Britain. Gildas had a rhetorical reason for such language, but when Bede composed his Ecclesiastical History of the English People in 8th-century Northumbria, he too stressed the strength of the newcomers, while using additional sources to extend the story to encompass the now-familiar triad of Angles, Saxons, and Jutes. For the first Anglo-Saxon Chronicle, produced in 9th-century Winchester, its compilers needed to try to reconfigure multiple accounts relating to southern Hampshire into a simpler tale of the origins of Wessex. Within this story, though, none of the names of the kingdom’s earliest leaders are definitely Germanic, several are identifiably British, and two derive from local Latin place-names established during the Roman period.

It was Geoffrey of Monmouth, a cleric from the Welsh borders writing in the 12th century, who made the story of overwhelming numbers canonical. In his History of the Kings of Britain, he depicts the Germanic warlord Hengist explaining to Vortigern – a British ruler who had hired him and his brother Horsa as mercenary leaders – that surplus population numbers in their homeland repeatedly drove the Saxons to seek new territories. Geoffrey ends this account with a further ‘innumerable multitude’ of Saxons occupying land from Scotland to Cornwall. Interestingly, following the 16th-century Dissolution of the Monasteries and the dispersal of their manuscript collections, when it would have been possible for Elizabethan historians like Camden and Holinshed to examine authentic early sources, their approach was to paraphrase and collate Gildas, Bede, and overseas records rather than determinedly pushing back against the narratives presented by Geoffrey and disseminated through 13th-century Anglo-Norman retellings of his version of events (known as the Brut histories).

The first surviving Anglo-Saxon Chronicle’s annals for the period from AD 507 to 522 records the earliest invasions and conquests of Cerdic and Ceawlin and the ‘West Saxons’. IMAGE: Parker Library, Corpus Christi College, Cambridge

Over time, critical representations of the mass-migration position have become exaggerated to the point of claims that it supposes that the Anglo-Saxon period was ushered in by genocide and ethnic cleansing of the previous population. The bombastic and opinionated, but influential Victorian historian Edward Augustus Freeman is often cited here. Although many of his political and social prejudices are shocking to modern ears, and he had a tendency to romanticise a ‘Teutonic’ character typical of his time, it is essential to critique his work on early medieval England accurately and fairly. In his six-volume History of the Norman Conquest, Freeman dedicates half of the text to discussing conditions within Anglo-Saxon England before 1066, mostly focusing on the final quarter-century spanning the reigns of Edward the Confessor and Harold Godwinson. There, Freeman also clears the ground for a straightforward comparison of pre-/post-Conquest administrative and juridical practices by establishing the earliest ‘English’ (Freeman disliked the term ‘Anglo-Saxon’) kingdoms on a tabula rasa. After reviewing the evidence, he concluded that no institutions from Roman Britain survived, and took that to imply the eradication of indigenous men who would have run things – although he commented that ‘the women would doubtless largely be spared’, to serve as slaves and concubines. That casual chauvinism – other references to ‘healthy barbarism’, ‘the distinct and glorious inheritance of the English’, and more reinforce the tone – should make us grimace, although Freeman does thereby posit the survival of half the indigenous adult population, to say nothing of the young.

New data

Setting aside misrepresentations, though, a ‘mass-migration’ standpoint is a reasonable one in modern scholarship, in the form of arguments that major population movements across large areas of Europe and Asia were key features of what was going on between the 4th and 6th centuries AD – long called ‘the Migration Period’. This position is also prepared to accept that – for all their limitations – early medieval sources can reflect the reality of what was, to them, a relatively recent past.

It is not only in England and within the Migration Period that longstanding models incorporating major population movements are now supported by aDNA data like those discussed within this special issue. Large-scale prehistoric relocations are now proven as a factor in the Neolithic farming revolution (CA 290), the introduction of Indo-European languages, the spread of the Beaker Culture, and the diffusion of a Celtic cultural archetype at the dawn of the Iron Age (CA 307). On the Continent, as the Roman Period became the early Middle Ages, it is perhaps the movements of Langobards through Hungary into Italy, then of Avars into the Hungarian plain, that have most conspicuously been confirmed in this way.

It is understandably difficult to avoid some feeling of vindication for those who have been willing to trust in a plausible interpretation of archaeological evidence and historical traditions which includes large-scale colonisation of southern and eastern Britain by Germanic-speaking groups in the 5th and 6th centuries AD. This is not simply a case of ‘we told you so’, primarily because nobody for several generations has been confidently telling the world precisely what the situation was in such terms, but also because the secure picture now available, while vitally enlarged and substantiated, leaves many open questions.

Longstanding models of population movement from other periods are also supported by recent aDNA data, such as the spread of the Beaker Culture; here we see a Beaker burial excavated at Trumpington by Cambridge Archaeological Unit. IMAGE: Cambridge Archaeological Unit

In post-Roman Britain, it has been revealed that a large swathe of eastern England, from the south-east to Teesside, saw a dramatic change in the genetic profile of the population to one in which as much as 74% could represent newly settled immigrants from Continental Northern Europe. As we would expect, there is considerable variation from site to site, though that, it must be emphasised, is in the context of a sample that, for pragmatic reasons, was focused on classic ‘culturally Anglo-Saxon’ burial sites, mostly of the 5th to 9th centuries AD. Nor is it ignored that at least a quarter of those populations overall appears to be descended from indigenous ancestors. We categorically need more genetic information than we have at present for the Romano-British population in what was to become England – not least to ascertain how far the cross-Channel influence evident in populations in Kent and Sussex might have had deeper and earlier roots. Such data might eventually reveal important patterns of difference around the principal urban centres and the most highly militarised areas of Britannia, while for post-Roman Britain we need much more data from the west: both southern Wessex and the Upper Thames, and if possible the south-west Midlands. There is good reason to anticipate those regions revealing a gradient with increasing prominence of Western British and Irish haplogroups as we move further west.

Another line of potential investigation involves comparing specific individuals with the collective genetic profiles. As I noted at the end of an In Our Time discussion about St Cuthbert of Lindisfarne early last year, the bones of Cuthbert and the skull of King Oswald are both interred in Durham Cathedral. It would be genuinely revelatory to have information on the genetic background of these two men who played such key roles in the development of the Christian kingdom of Northumbria.

The Anglo-Saxon period: a fresh start

In respect of the establishment of the Anglo-Saxon Period, it might come as a surprise to read a concluding declaration here that these new data explain nothing – but that is because it is vital to appreciate that whatever happened subsequently in the evolution of an ‘English’ nation, it cannot be treated simply as the inevitable and necessary consequence of the demographic transformations of the 5th and 6th centuries revealed by genetic research. It represents no change in a ‘migrationist’ position to insist that the relocation and displacement of populations was only ever one key parameter in a mass of relationships, causes, circumstances, and effects that lie behind this early medieval narrative.

It appears, too, that the earliest accounts of this period can no longer be dismissed as nothing other than fabrications adapted to immediate goals and readerships in the early Middle Ages – something that calls for even greater critical care and respect in their use. Like it or not, Freeman correctly recognised that a high level of population change over a major part of Britain in the 5th and 6th centuries must be a crucial variable in comparative historical studies with Roman and Late Antique contexts elsewhere in Britain or western Europe.

The results we will go on to discuss in this special issue mean we can move on from one level of high uncertainty – but, as ever, only to the next set of questions. These genetic data turn an uncompromisingly clear spotlight on to arguments that have sought to marginalise or even erase the substance of the changes that affected southern and eastern England at the start of the Anglo-Saxon period, when a largely new material culture was introduced – now undeniably accompanied by the regionally dominant presence of settlers who (and whose ancestors) were at home with that culture. That was the root of a process that, step-by-step, coalesced into a series of transformed and continually transforming identities – local, regional, and national – including, eventually, the idea of being ‘Anglo-Saxon’, and finally ‘English’, in ‘England’.

The most important conclusion of all, though, is that over the whole range of possibilities – from mass migration and population replacement to long periods of stasis and increasing homogenisation – there is nothing in Britain’s entire demographic history that should be supposed to be either normal or unprecedented.

Further reading
John Hines (1995) ‘The becoming of the English: identity, material culture and language in Early Anglo-Saxon England’, Anglo- Saxon Studies in Archaeology and History 7: 49-59.

Kazutomo Karasawa (2022) ‘Historical origins of a mythical history: the formation of the myth supporting Anglo-Saxonism reconsidered’, in Karen Jolly and Britton Brooks (eds), Global Perspectives on Early Medieval England (Woodbridge: The Boydell Press), pp.171-189.

Rory Naismith (2021) Early Medieval Britain c.500-1000, Cambridge: Cambridge University Press.