1 © Society for Medieval Archaeology 2011 DOI: 10.1179/174581711X13103897378311 Medieval Archaeology , 55, 2011 Anglo-Saxon Immigration and Ethnogenesis By HEINRICH HÄRKE 1 IT IS NOW widely accepted that the Anglo-Saxons were not just transplanted Germanic invaders and settlers from the Continent, but the outcome of insular interactions and changes. But we are still lacking explicit models that suggest how this ethnogenetic process might have worked in concrete terms. This article is an attempt to present such a model from an archaeological perspective, but with an interdisciplinaryapproach. The focus is on the role of the native British population and its interaction with immigrant Germanic groups. As a result, the model envisages two broad phases in the creation of the Anglo-Saxons: an ethnically divided conquest society in the 5th/6th centuries in which immigrants and their descendants practised a form of ‘apartheid’ in order to preserve their dominance; and a phase of increasing acculturationand assimilation of the natives in the 7th/8th centuries that laid the foundations of a common English identity. The last two decades have seen a lively debate on the srcins and the making of theearly English. The only earlier, sustained debate on this issue dates back more than acentury: the late 19th-century ‘Teutonist’ controversy had raged over the relative propor-tions of immigrant and native populations. 2 The present paper intends to suggest anarchaeological model using an interdisciplinary approach and paying particular attentionto the role of the native population of Britain (Fig 1). It is not the aim here to provide afull outline of the history of research on the topic, and the growth and contents of therecent debate; several such outlines are available elsewhere. 3 But it is necessary to providea very brief summary that emphasises the recent impact of scientific data and explainswhy the presentation of an explicit model is timely at this point.For most of the 20th century, archaeological and linguistic research seemed to con-firm the reports by Bede (Beda Venerabilis) and in the Anglo-Saxon Chronicle that severalgroups of Germanic immigrants from the Continent invaded post-Roman Britain andsettled there (Fig 2A). 4 They were thought to have brought with them their own cultureand language, creating their own ethnic kingdoms according to their respective continen-tal srcins (Saxon, Anglian, Jutish), and eventually replacing the native Britons in a processthat today would be termed ‘ethnic cleansing’. 5 The political process of incorporatingsmaller units into larger kingdoms described by Bede, as well as some of his terminology, 6 have long been taken by scholars to suggest the emergence, by the early 8th century, of a common identity that was then consolidated in the common fight against the Vikings. 1 Honorary Research Fellow, University of Reading; Honorarprofessor, Abteilung für Archäologie desMittelalters, Universität Tübingen. Pulvermühle 6, D-31863 Coppenbrügge, Germany . [email protected]
2 Biddiss 1979; MacDougall 1982. 3 Härke 1998a; 2004; Hills 2003; Laker 2008. 4 Anglo-Saxon Chronicle s a 446; Bede, Historia Ecclesiastica I.15; Leeds 1913; Collingwood and Myres 1936; Stenton1943; Myres 1986. 5 Cf the native perspective in Gildas, De Excidio 22, 24–5, about extermination, expulsion, enslavement andextinction by plague. 6 Hines 1994, 50–2, on Bede’s use of ‘gens Anglorum’. 2 heinrich härke Except for a few lone voices, 7 this narrative was accepted throughout most of the20th century, both within the scholarly community and outside, as the broad outline of early English srcins.The new debate in Anglo-Saxon archaeology, which began in the late 1980s, endedthis broad consensus. The debate had two main roots; one was the critical reassessmentof the written sources for the 5th and 6th centuries in Britain, leading to the realisationthat most of our sources for this period are unreliable in both chronology and content. 8 The other root was in Anglophone prehistory where the New (processualist) Archaeologyof the 1960s and 1970s included a substantial element of anti-migrationism; this providedan intellectual inspiration for a new generation of early medieval archaeologists. 9 Duringthe 1990s, a new perspective spread and took firm hold in the discipline and beyond: the Anglo-Saxon invasion, far from being a ‘folk migration’, consisted of only small groups 7 Lethbridge 1956; John 1966; Brown 1978. 8 Dumville 1977; Sims-Williams 1983a; 1983b; Yorke 1993. 9 For example Arnold 1984; Hodges 1989. fig 1Location map of sites andregions mentioned in the text. Created by M Mathews 3 anglo-saxon immigration and ethnogenesis fig 2Distribution of diagnostic 5th to 7th/8th centuries Anglo-Saxon and British evidence in England.(A) Anglo-Saxon burial sites of 5th to 7th centuries (after Alcock 1971, redrawn by N Griffiths). (B) Placenames in walh- (after Cameron 1980, redrawn by M Mathews). (C) Penannular brooches (after White 1988,Longley 1975, Fowler 1963, drawn by M Mathews). (D) Hanging bowls and escutcheons (after Brenan 1991,White 1988, Longley 1975). Drawn by M Mathews 4 heinrich härke of warriors and few, if any, families. This means that the numerous early Anglo-Saxoncemeteries should represent either the tiny Germanic elite who controlled the, as yetundiscovered, mass of the native population; 10 or they are actually the burial sites of majority Britons who had adopted the cultural attributes of the few continentalimmigrants. 11 Until the beginning of the 21st century, the debate was driven by new thinkingrather than new data, but, since then, biological data have had a major impact. Whileancient DNA has not yet fulfilled its earlier promise because of technical problems andcost, 12 modern DNA has been brought to bear on the question since the 1990s. But a dataset of modern mitochondrial DNA (inherited in the female line) that appeared to set anupper limit for post-Roman immigration from NW Germany had to be withdrawnbecause of concerns about their reliability. 13 In the first decade of the 21st century, atten-tion switched to modern Y-chromosome DNA (inherited in the male line), partly becausethe faster mutation rates of its microsatellites allow better estimates of the dates of pastpopulation events. Recent studies have suggested large-scale immigration in the post-Roman period and sparked a debate about early Anglo-Saxon social structures that mighthave contributed to shaping the modern genetic patterns (see below). 14 At about the same time, new projects started applying stable isotope analysis toearly Anglo-Saxon cemeteries with the explicit aim of studying mobility in the past.The few publications and conference papers so far appear to suggest a greater proportionof first-generation immigrants than apparently expected by the project teams. 15 Thus, the most recent approaches and methods seem to get us back to the oldestmodel, that of mass immigration and (at least, partial) ‘ethnic cleansing’. While thependulum appears to be swinging back towards more ‘migrationist’ ideas, the role of thenative Britons in the making of early medieval England is increasingly becoming a focusof interest. 16 At this point, an explicit model that addresses the new focus, integrates thescientific data and offers a falsifiable hypothesis for future research should help to take thedebate forward.The emphasis here on scientific data calls for a few comments on theory andmethodology. This is less necessary for the archaeological evidence as such, nor even theissue of ethnic identification from cultural evidence because this has been the subject of alot of discussion and critical reassessment. 17 Accordingly, it is assumed here that ethnicityis not a given, but a flexible and situational concept: ethnicity is ‘in the heart’, not ‘inthe blood’. Any ethnic identification based on ritual expressions and material remains(‘Anglo-Saxon’ burials, ‘Celtic’ artefacts, etc) would, therefore, represent the inferenceof a cultural statement of perceived group affiliation, and such a statement would haveseveral dimensions and layers of meanings, depending on the various audiences of thestatement. Material culture may also be used to create and reinforce group identities, andHines has suggested that dress played such an active role in the creation of regional 10 Higham 1992, 165. 11 Pryor 2004. 12 Richards et al 1995; Stoneking 1995; Renfrew and Boyle 2000. The only published analysis of mitochondrialDNA (mtDNA) from Anglo-Saxon bones did not produce any specific results for population movement orstability in the post-Roman period (Töpf et al 2006; 2007). A new project involving the analysis of ancient DNAfrom early Anglo-Saxon skeletal material is about to start at the Oxford Laboratory of Archaeology and theHistory of Art in 2009; C Edwards pers comm. 13 M Richards pers comm; cf Richards et al 1993. 14 Weale et al 2002; Capelli et al 2003. 15 Budd et al 2004; Montgomery et al 2005. 16 Härke 1998b; Higham 2007; Laker 2008. 17 Among others, Hodder 1982; Geary 1983; Hines 1994; Jones S 1996; Pohl 1994; 1997; Pohl and Reimitz1998; Härke 1999a; Brather 2000; 2004; Hakenbeck 2007; cf summaries in Hills 2003; Härke 2007b. 5 anglo-saxon immigration and ethnogenesis Anglo-Saxon groups. 18 Apart from dress items, it is, in particular, burial styles (cremation,dressed inhumation) and settlement types that are thought to be diagnostic Anglo-Saxonfeatures of the 5th to 7th centuries, partly because they can be traced to continentalsrcins, partly because they are markedly different from the preceding Romano-Britishculture. 19 A purely archaeological approach relying entirely on the ethnic identification of material remains would, however, suffer from an additional limitation: it is possible foronly one of the two main groups involved in the process of ethnogenesis discussed here.The collapse of Roman material culture some time in the early 5th century left a gapin the archaeological record that was quite rapidly filled by the intrusive Anglo-Saxonmaterial culture while the native population became archaeologically close to invisible. 20 It is, therefore, imperative to use other, additional indicators to distinguish between nativeand immigrant populations, and these are primarily biological data. The use of skeletaldata rests on the observation of significant skeletal differences between populations of thelate Roman and post-Roman periods in western and central Europe. 21 Such differencesare to be expected between separate, non-intermarrying populations, and they are likelyto persist in contact situations until they are levelled out by intermarriage and assimilation.More recently, population geneticists and physicists have demonstrated the potentialof using stable isotope and modern DNA data for elucidating questions of populationmovement and admixture in the post-Roman period (see above).It is, therefore, the combined use of archaeological and biological data that probablyhas the greatest potential for the study of ethnogenesis resulting from migration andpopulation contact. This is not to deny the methodological and practical problems of suchan approach. 22 Because of individual variations, skeletal data can only be used to describeand classify populations (in a statistical sense); they cannot be used to identify anindividual as ‘British’ or ‘Anglo-Saxon’. Stable isotope data have the potential to providesome information about the srcins of individuals, but this is only possible if there arecomparative geological data for suspected emigration areas. Modern Y-chromosome DNAcan be applied to historical questions at the population level only; the dating of demo-graphic events is only possible in a statistical sense and with wide margins of error; andinferences about the composition of past populations require explicit models and assump-tions about population dynamics in the intervening period. But these limitations do notseem to be the main reason why archaeologists still have misgivings about approachesinvolving biological data: 23 their reluctance appears to be a reaction against the unre-flected use, and deliberate racist misuse, of palaeoanthropology before 1945. The emer-gence in recent years of cross-disciplinary groups including archaeologists and scientistsworking on questions of mobility and migration is, therefore, a hopeful sign. 24 IMMIGRATION AND SETTLEMENTModern Y-chromosome DNA points to Anglo-Saxon srcins in Dutch Frisia, north-ern Germany and Denmark, which is in perfect agreement with archaeological, textual 18 Hines 1994; 1996. 19 Welch 1992; Lucy 2000. 20 Esmonde-Cleary 1989; 1993; Härke 2007a; but see now Cool 2000. Recent hoards and metal-detector findsshow that coin use and import did not stop abruptly at ad 410 (pers comm C Scull). 21 Rösing and Schwidetzky 1977. 22 Härke 2007b. 23 For example Mirza and Dungworth 1995; Hills 2003; 2009. 24 It does not help, however, if archaeologists hold the view that natural scientists are essentially ‘simplistic’,in contrast to ‘sophisticated’ social scientists, including archaeologists; cf Hills 2003, 71.