Haplotypes and textual types: Interdisciplinary approaches to Viking Age migration and mobility

Migrations, settlement, and genetics

Despite popular associations of Vikings with colourful violence and excessive masculinity, this is a perception that has developed over time (Williams, 2020). A less sensationalist perception of the Viking Age (here defined as ca. 750-1100 CE) would be that it is characterised most of all by mobility, namely the movements of peoples, originating in Scandinavia, around quite extensive areas of the world, whether they were members of armies, traders, or settlers. Migration followed by permanent settlement was one of the most salient long-term effects of the Viking Age, at least in the western Viking world, here considered as the islands of Britain, Ireland, Faroe, Iceland, and Greenland. The settlement of Iceland in the Viking Age, from the late 9th century onwards, is an unusual example of human settlement in a previously uninhabited island in relatively recent, proto-historical, times. Although scholars are more and more recognising that migration is ‘a continuous process of varying duration’ (Furholt, 2018: 166), the settlement of Iceland is a clear example of a specific, identifiable and datable migration. However, it took place within a larger and more complex context of Viking Age movements throughout not only the western Viking world, but also as far as North America in the west and equally far away in the south and east (Price, 2020: 362–381). These movements suggest a range of questions. Were settlers drawn from the same groups that were raiding and trading, whether in the western Viking world or elsewhere? Did they come directly from Scandinavia (and if so which part?), or did some of them come from other parts of the world? Did settlers of different origins settle in different places? There is also a significant question regarding gender participation in Viking Age mobility. Unlike other places settled by Scandinavians during the Viking Age, Iceland was uninhabited and required the presence of women for the continuation of the population—were these women recruited from the same groups as the male settlers? In other areas, such as Britain and Ireland, there was already an indigenous population to ensure population continuation, so what role did women play in Viking settlement there? Similar questions could be asked about the social status of the settlers. Furthermore, there are important questions about the later effects of Viking Age settlement and migration. For example, what networks were established during the Viking Age which might have influenced population developments after the initial settlements?

The potential significance of genetics for answering some of these questions became clear in the early 2000s, when the deCODE project collected DNA from the whole population of Iceland (Gillham, 2011: 12–19). Studies by Agnar Helgason et al. (2000a, 2000b) of both Y-chromosomes and mtDNA aimed to determine the proportions of this population which descended from Scandinavian immigrants or other populations by comparing with the DNA of other modern populations. These papers proposed that more than two-thirds of the male population of Iceland at the time had Norwegian ancestry, while less than one-third could trace its origins to the ‘Celtic’ regions of Britain and Ireland, but that these proportions were reversed in the female ancestry. That there was a ‘Celtic’ element in the Icelandic population has long been known, and much discussed in the scholarly literature, based on older studies of various biological markers, linguistic evidence, and references in medieval Icelandic texts (Guðmundsson, 1997; Jacobsen, 2005; Mourant and Watkin, 1952: 23; Sayers, 1994; Sigurðsson, 1988; Williams, 1993). The genetic studies suggested that this Celtic element was more extensive than previously recognised and highlighted the importance of the gender dynamics of these migrations.

Gender was also a theme in one study which compared the Icelandic results with several regions in Scandinavian Scotland (Goodacre et al., 2005). The population of both the Northern and Western Isles showed a significant Scandinavian element, but in the former, the ‘Scandinavian’ proportion of both male and female populations was roughly equal, which the authors took to mean that the Scandinavian incomers arrived in family units. In the Western Isles, as in Iceland, there was a higher proportion of Scandinavian patrilineal ancestry compared to matrilineal and it was hypothesised that the Western Isles and Iceland were colonised by ‘lone males’ with female partners from the British Isles.

The connection between the Y-chromosome and surnames, both passed from father to son, was used to investigate subjects whose surnames were representative of the historical population of north-west England (Bowden et al., 2008; King and Jobling, 2009: 356). A substantial proportion of these subjects (≈50%) had a direct male ancestor with a Y-chromosome type most commonly found in Norway. This was taken to confirm the Viking Age settlement of the area, predominantly by people of Norwegian origin, as already suggested by place-names, archaeology, and some documentary sources, though the question of gender roles in this settlement remained underexplored (Jesch, 2015b). Teams led by Mark Jobling and Turi King continue to fine-tune the evidence that sub-lineages of Y-chromosome haplogroup R1a1 in modern populations provide for the demographic impact of Viking migrants in England (Lall et al., 2020). This most recent study is also able to compare the modern DNA with aDNA and finds that this ‘supports the interpretation that two sub-lineages of hg R1a1 spread with the Vikings from peninsular Scandinavia’.

The earlier genetic studies could not be the last word on the migrations of the Viking Age because of the well-known limitations of using present populations as proxies for past populations, the restricted informational value of linear descent through the Y-chromosome or mtDNA over a thousand-year period, and the often unknowable impact of genetic drift in small populations. As Sims-Williams (2012: 428) put it, ‘fashion swings between assuming population replacement and population continuity. Only under the latter hypothesis does the use of modern data become attractive’. So, for most geneticists and those who follow their work, it is aDNA which is the most promising body of evidence for the study of ancient populations and their movements (Furholt, 2018: 161; Racimo et al., 2020), without prejudging the issue of replacement vs. continuity.

This paper aims to bring this nascent use of aDNA to explain Viking Age migrations and mobility into dialogue with other evidence, particularly the Viking Age and medieval texts that allude to these migrations. Both texts and genetics have significant drawbacks, but for this reason may complement each other. In particular, both are unlikely to be fully representative as evidence for what actually happened in the past. Ancient DNA recovered from archaeological skeletons is indeed ancient but the datasets as used in archaeogenetics are small, still only in the low thousands worldwide (Racimo et al., 2020), and therefore may not be more representative than the large but problematic datasets of modern DNA. The epigraphical, historical, and literary texts which provide evidence for Viking Age migrations and mobility, though numerous, also suffer from doubts about how representative they are for a variety of reasons as will be outlined below. This common characteristic suggests that it would be fruitful to bring textual studies and archaeogenetics into dialogue with each other. This is not so that one discipline can ‘correct’ or ‘prove’ the results of another; rather, the aim should be to establish points of agreement or disagreement which can provide the basis for further research. To provide a focus for this interdisciplinary dialogue, it is proposed here that the evidence of both texts and aDNA should be considered through the framework of ‘diaspora’. In particular, the diaspora framework highlights issues of gender, language, and social status, and encourages Viking Age developments to be seen as part of a longer chronological process. After a brief introduction to the diaspora framework, this paper will consider a number of recent aDNA contributions to Viking Age studies to identify instances where the textual sources will aid a more nuanced and complex understanding of the population consequences of past mobility.

A diaspora framework

The idea of ‘diaspora’ is a relatively recent concept in Viking Studies, as developed by Abrams (2012) and Jesch (2015a, 2018). The term ‘Viking diaspora’ is now widely used, though not always with full consideration of its implications (Price, 2020: 555). However, it has significant explanatory power, especially with regard to the questions posed above. There are many different ways of understanding ‘diaspora’ (Jesch, 2015a: 68–80), but for present purposes it can be summed up as an explanatory model which posits a network of connections arising from migrations, which transcend the initial migrations in both time and space. Although scholars have been arguing for a more nuanced understanding of the processes involved in human mobility (Furholt, 2018: 165–171), many still work with quite simple models of migration and settlement. Unlike these, the important characteristics of the Viking diaspora as a basis for a comparison with archaeogenetics are that:

diaspora settlements like Iceland were part of a larger movement of peoples from Scandinavia to a variety of places in both east and west, a movement which had a substantial female element;

Iceland is fundamental to the model because the majority of the texts to be discussed emanate from there. However, the model can also be applied elsewhere in the Viking world, and can also extend beyond the chronological limits of the Viking Age, due to this common identity created by diaspora (Abrams, 2012; Šnē, 2017). It has been suggested that a diasporic approach reveals a long, broad Viking Age, defined as an extended period up to around 1500, when some of the processes typical of the Viking Age were still in operation (Jesch, 2015a: 55–56). This period has the benefit of including the post-Viking Age texts, as discussed below. Of course the culture of the Viking diaspora was not uniform throughout either the geography of the Viking world or the chronology of the Viking Age when these are conceived so broadly. Nevertheless, it is possible to find aspects of a common identity throughout both this geography and this chronology, and it is this common identity which challenges, or at least complicates, some of the results of archaeogenetics. Diaspora theory is most relevant when people of a particular origin migrate to several destinations, when the migrants in those various destinations are in touch with each other as well as with the homeland, and where those settlements endure for some time, all of which are the case for the Viking diaspora. It is thus the model that best explains these processes and this common identity.

Language, literacy, and the Viking Age

Historians have long grappled with the difficulty that those documents which are contemporary with the Viking Age were written from an outside perspective, as the Scandinavians themselves were largely pre-literate during this period. The arrival of literacy along with the conversion to Christianity in Scandinavia and in Viking settlements such as Iceland is one of the classical markers of the end of the Viking Age. This results in a wealth of texts which are in Old Norse, the language of the Vikings, but which are not contemporary with the Viking Age.

Despite Viking Age Scandinavia not being a fully literate society, there are two types of text which can with reasonable confidence be assigned to that period: runic inscriptions, which are contemporary written texts, and, less certainly, poetry, an originally oral form which later became a written one (Jesch, 2001). Runes are a form of writing which developed in Scandinavia before 150 CE (Barnes, 2012: 9–14), and which then continued to be used for some centuries after the Viking Age, not only in all of the Scandinavian homelands but also in regions settled in the Viking Age, especially Scotland, Iceland, and Greenland, in some cases up to about 1500 CE. Poetry also shows a remarkable continuity between c.300-1500 CE, as evidenced by some early runic inscriptions as well as later manuscripts (Jesch, 2008). The basic structures of Scandinavian poetry, especially its metres and diction, are maintained right through the major change from orality to literacy that happened between the late Viking Age and the 12th century. Although best attested in medieval Iceland and Norway, the geographical range of Scandinavian poetry also includes Sweden and Denmark, and the diasporic lands of England, Scotland, and Greenland.

All other surviving texts in a Scandinavian language date from after the Viking Age. This language is one of the main ‘shared cultural elements’ of the Viking diaspora (Abrams, 2012: 23) which crosses its chronological and geographical boundaries. Viking Age migrants took their language with them from Scandinavia to their new homes and it continued in use until either the settlement died out (as in Greenland in the 15th century) or the language and its speakers were gradually assimilated into the majority population and its language (as happened in England). In Faroe and Iceland, the language has survived (and developed) continuously since the Viking Age.

Old Icelandic texts

The maintenance of their language enabled Icelanders and others to keep up close contacts with both the Scandinavian homelands and other regions of the diaspora, aiding the development of medieval Icelandic textual culture. This culture is an example of the effects of diaspora, as much of the prose narrative literature of high medieval Iceland has the Viking Age past as its theme (Clunies Ross, 2010: 27–36, 72–80, 85–91). Explicitly historiographical works such as Íslendingabók (Book of the Icelanders) and Landnámabók (Book of Settlements) and some of the sagas about Norwegian kings began to be written in the 12th century, though many of the surviving versions were revised or composed in their current form in the 13th, and texts in this genre continued to change and develop into the late 14th century. The sagas of Icelanders (Íslendingasögur) were composed during the 13th and 14th centuries, yet many of them survive only in manuscripts of the 14th century or later. These texts, too, continued to develop. The sagas of ancient times or legendary sagas (fornaldarsögur) are a phenomenon that began in the 13th century, but the genre was long-lived, with a concentration of manuscripts in the 14th century and even later.

Despite these 12th- to 14th-century contexts of production, the sagas of kings, Icelanders, and ancient times were once taken by scholars to be accurate historical accounts of the Viking Age past. However, the tendency of saga scholarship in the later 20th century and until today has been to view all three genres as, at best, literary reconstructions of that past, or using the past to mirror the present, or even outright fictions (Jakobsson and Jakobsson, 2017). In particular, the interest these texts have in the past has been characterised as ‘cultural memory’, ‘fictions of memory’, and ‘imagined versions of the past’ (Hermann, 2020). While recognising that ‘the subgenres of the saga … fused imagination and reality’ (Hermann, 2020: 170), literary scholars have been more interested in the imagination than the reality. As the archaeologist Neil Price has pointed out (2020: 23), such skeptical literary scholars have not confronted the question of where all this imagined reality has come from. It is possible to argue that because the three saga-genres mentioned above, along with the historiographical works, take the Viking Age past as their main literary theme, and many of them base their narratives on Viking Age poetry of the type known as ‘skaldic’, a cultural memory perspective is not sufficient and a diaspora model has more explanatory power. In contrast to theories of cultural memory, which emphasise the present in which texts were produced, a diasporic understanding of these texts encompasses the whole period from the described past through to the time of writing. This enables us to see the sagas as neither a simple, transmitted memory of the past, nor as a thoroughly contemporised past, but as something in between (Jesch, 2018). The sagas do reach back into the Viking Age, but in a variety of complex ways that need to be unpicked.

The evidence of textual types

Thus runic inscriptions, various kinds of poetry, and the saga-genres mentioned above all tell, in different ways, stories about the Viking Age. But all are also unrepresentative of that period, again in a variety of ways. Most obviously this is a question of preservation and survival, which affects all historical evidence from this period – we cannot tell if the surviving evidence is fully representative of what once existed. The different kinds of texts also have individual characteristics influencing their representativity. Runic inscriptions, although contemporary documents, are short, formulaic, and overwhelmingly concentrated in eastern Scandinavia (Denmark and Sweden) in the Viking Age, though more common in the west (particularly Norway and Greenland) in the later period. With few exceptions, Viking Age poetry survives only in later manuscripts and assessing which texts have Viking Age origins is a complex matter. The survival of originally oral texts is likely to have been driven by the needs and predilections of the period (12th century onwards) in which they were written down. Similarly, the sagas or other prose narratives (sometimes using Viking Age poetry as a source) have a real but complex relationship with the Viking Age past that they describe which will vary between individual texts and genres. Also, each genre has its own favoured topics, and so does not give a complete picture of the Viking Age.

While, despite some honourable exceptions (Friðriksdóttir, 2020), literary scholars have shown little interest in the question of the sagas’ evidential value for the Viking Age, many historians and archaeologists have demonstrated that they both need and want to use sagas and other medieval texts to bolster their theories about the Viking Age (Price, 2020). While this is welcome, the lack of interest among saga scholars means there is little theoretical framework to draw on and laudable attempts to integrate sagas into other scholarship are still fairly superficial (Raffield, 2019: 685–686; Raffield et al., 2017a, 2017b). Many scholars, while recognising both the value of the sagas and the need for caution, nevertheless still see the potential of sagas largely as corroborating or supporting the more contemporary texts or archaeology, with little direct evidential value in themselves. Yet each of the kinds of medieval texts outlined above has something to offer interdisciplinary work on the Viking Age. Even the most problematic of these texts, the sagas, can provide evidence for Viking Age matters if analysed carefully by specialists who take full account of the sources, origins, transmission, and genre of the text (Dale, 2014; Jesch, 2014a, forthcoming; Larrington, 2008; Ruiter, 2020; Vidal, 2013; Vohra, 2008; Zilmer, 2008). Studies like these demonstrate the evidential value of sagas for Viking Age material culture, gender constructs, various social and legal concepts, kinship structures, and mobility patterns, and make a dialogue with archaeogenetics possible.

Slaves in northern Norway

The use of archaeological skeletons allows for the successful combination of aDNA with other currently common archaeological methods such as stable isotope analysis. Thus a study of 10 individuals from northern Norway (Naumann et al., 2014) focused on social class and concluded that individuals identified as ‘slaves’ in multiple burials had a similar diet (and therefore presumably a similar lifestyle) to individuals from the common population buried in single graves, whereas the high-status individuals with whom the ‘slaves’ were buried had quite a different diet and therefore different social significance.

The enslavement of people is a constant theme in Viking Studies (Brink, 2012; Karras, 1988; Price, 2020: 141–154), though the issue is complex and it is clear that the practice of and reasons for enslavement, the lives of enslaved persons, and the possibilities of manumission varied throughout the Viking Age and the Viking world. The vocabulary of enslavement is particularly diverse, suggesting that there were more complex social structures operating than implied by a simple concept of ‘slavery’, with different degrees of dependence on an authority (Brink, 2012: 121–168; Karras, 1988: 42–45). Legally, enslaved people were chattels, the possessions of their owner, but there is also evidence of categories of people who had responsibility and status.

Thus, describing an archaeological skeleton as a ‘slave’ on the grounds that they were buried with another skeleton (like the other possessions of that person) may be simplifying what was a more complex situation, so it is essential to combine the textual and scientific evidence to introduce more nuance into the discussion. It should also be remembered that slaves could be freed and runic inscriptions, laws, and Icelandic texts all indicate that this could happen (Brink, 2012: 92–120; Jesch, 2014b: 283–284; Karras, 1988: 123–131). The texts also suggest that many slaves were non-Scandinavians, whether taken to Iceland with the settlers, or back to Scandinavia to serve there. At the same time, what we know of reasons for enslavement suggests that not all slaves need be foreigners and the origins of enslaved persons in the Viking Age ought to be a fruitful avenue of enquiry for aDNA and isotopic studies. While many of these issues cannot be addressed through archaeological analysis alone, they need to be borne in mind when identifying the social status of archaeological skeletons.

Mobility in Sigtuna

Also combining genomic and strontium isotope analysis, but with a focus on mobility, Krzewińska et al. (2018) studied 23 individuals to analyse mobility in the late Viking Age Swedish town of Sigtuna and found a high degree of ‘cosmopolitanism’, reflecting ‘the whole network of the Viking world’.

The diversity of Sigtuna in this period can also be deduced from the runic inscriptions from the town, including one which mentions the name of the town in a context which probably refers to an immigrant and several which mention contacts with Frisia (Jesch, 2001: 116–117, 239–241). The study did not specifically address any links Sigtuna might have had with the western Viking world, yet the textual evidence adds this western perspective to the ‘cosmopolitanism’ of Sigtuna. There is contemporary evidence from the late Viking Age in the form of skaldic poetry which shows that two Norwegian kings and their Icelandic poets were familiar with the place, apparently as an important stopping-point on their return from the east (Jesch, 2001: 116). It is not inconceivable that some of the immigrants to the town came with these kings and this evidence demonstrates at the very least an elite network focused on the emerging monarchies. Such mobility probably included people of other social classes, such as enslaved persons. There is also the question of whether individuals who happen to have died and been buried in Sigtuna were permanent residents or just passing through, again something that the genomic and isotopic analysis cannot necessarily elucidate and yet is significant when assessing more general mobility in the period. This, combined with the small number of samples, suggests that the claims about ‘cosmopolitanism’ made in promoting the article were overstated (Källén et al., 2019: 82). At the least, these claims need to be nuanced, using other available evidence.

Westward mobility

In a substantial paper, Krzewińska et al. (2015) studied maternal lineages of 45 Norwegian individuals from the period 550-1050 CE, comparing them to previously analysed ancient Icelanders and various modern populations, with the aim of providing ‘a picture of the genetic variation and movements of the Norse people in the North Atlantic region during the Viking era’ with an important additional emphasis on ‘post-colonization population change’. This analysis suggested close affinities between ancient Norwegians and ancient Icelanders, but also that both of these had more affinities with other western Europeans than with modern Norwegians. This result the authors put down to the demographic changes in Norway caused by events such as the Black Death in the 14th century, whereas other populations, being larger, were better reservoirs of ancient lineages. The authors argue that for the ancient mtDNA lineages, which are shared between their dataset and data from Orkney and Shetland, to have survived in northern Scotland implies that women in Late Iron Age Scandinavia were actively involved in the settlement of new lands, as suggested by the modern DNA studies discussed above, and as can also be gathered from texts like Landnámabók (Callow, 2011; Jacobsen, 2005; Jesch, 2014b: 275–277). The study is notable in its emphasis on how post-Viking Age developments might have affected populations.

This study also emphasises the role of women, and this seems to be central to understanding the Viking phenomenon. Even the Viking armies that raided England in the 9th and 10th centuries are now recognised as being more than just an ‘army’ as we imagine it and more like a travelling community that included a variety of non-combatants, including women and children (Hadley and Richards, 2016: 54; Raffield, 2016: 314–316). Although we imagine the ‘British Isles’ component of Icelandic settlement to have come largely from Scotland and Ireland, further studies should perhaps look more closely at the possibility that some of the Icelandic settlers also had connections further south in Britain, or even on the European continent.

Making the Icelandic population

Also combining genome sequence data with isotope analysis, a study of 27 ancient Icelanders (Ebenesersdóttir et al., 2018) explored the questions not only of the origins of the migrants, but also the subsequent genomic development of the Icelandic population. The aDNA analysis in this paper found that the origins of these individuals were ‘a combination of Norse, Gaelic [i.e. from the ‘British-Irish isles’, excluding Orkney], and admixed’. Nineteen of these 27 individuals were male, leading the authors to conclude that this was an effect of sex differences in burial practices during the period, an aspect which, if true, has implications for other studies. Interestingly, the study also compared ancient Icelanders to modern Icelanders and found that the former were more similar to their source populations in Scandinavia and Britain and Ireland than the latter. This difference they ascribed primarily to genetic drift in the small population of Iceland. The authors also found some evidence that ‘settlers of Norse ancestry had greater reproductive success than those of Gaelic ancestry’ which they saw in relation to the fact that individuals of Gaelic ancestry came to Iceland as slaves, with restrictions on their ability to reproduce.

This aDNA study seems to replicate the earlier work on modern DNA done by Helgason et al. (2000a, 2000b) and provides an interesting contrast with the study cited above (Krzewińska et al., 2015). All of these highlight the gender aspects of migration, without fully explaining them. Also in both cases, the aDNA datasets are small, which undoubtedly creates problems in drawing wide-ranging conclusions about migration and settlement. As in the case of the northern Norwegian slaves discussed above, one should perhaps also nuance the conclusions that the ‘Celtic’ element in the Icelandic population is an effect of enslavement, and also consider that it might be the result of diasporic contacts with Britain and Ireland, both during and after the Viking Age. The two are also not mutually exclusive.

Population genomics of the Viking world

In recognition of the need for larger datasets, a wide-ranging and thorough study (Margaryan et al., 2020) presented an analysis of the genomes of the remains of 442 humans, 141 female and 296 male, with the rest unidentified. This dataset was not only larger in numbers than previous studies, but also broader in chronology (from the Bronze Age to the Early Modern period, though with a preponderance of samples from the Viking Age) and geography (with samples from Greenland in the west to Russia and Ukraine in the east). The ancient samples were also referenced against modern worldwide and other ancient populations.

Before considering some of the details of the study, it should be noted that the article itself is rather short (seven pages) and that much of the contextualisation, and indeed some of the conclusions, which are of interest to non-geneticists occur in the Supplementary Information (178 pages, though much of this is technical), which is ‘in the format provided by the authors and unedited’. A discussion that still needs to take place is how science-based papers should publish conclusions which are drawn from the scientific analysis but which involve the use of other types of evidence, and which are often placed in the non-peer-reviewed supplementary material (Jesch, 2019).

Genomics, population groups, and terminology

In elucidating Viking Age genetic structure in Scandinavia the study identified ‘ancestry components’ which were labelled ‘Danish-like’, ‘Swedish-like’, ‘Norwegian-like’, and ‘North Atlantic-like’, the latter being ‘possible individuals from the British Isles entering Scandinavia’. This last component is of great interest in any understanding of the Viking diaspora, as already noted above, and deserves further discussion. Genetic studies of the Viking Age have not yet grappled fully with problems of terminology, and particularly the nomenclature to use of observed groupings, as has at least been attempted for prehistoric studies (Eisenmann et al., 2018). The use of present-day nationalities for these groupings is potentially misleading, though the study does acknowledge that ‘the genetic data are structured by topographical boundaries rather than by the borders of present-day countries’. As to the individuals entering Scandinavia, would they have come as spouses, or as enslaved persons, or is it just their genes returning to Scandinavia through the back-migration of mixed-heritage individuals from Britain and Ireland? A diasporic model applied to the textual evidence reveals that there was just such back-migration from the western settlements to Norway (Jesch, 2015a: 74–75, 2018: 591), though it does not reveal the genetic aspect of this, nor how extensive this was. Similarly, the authors’ identification of ‘the first evidence of South European ancestry … in Scandinavia’ could be developed further in line with the point about the composition of Viking armies made above.

It is also not always clear what ‘genetic diversity’ might imply. The authors found the highest genetic diversity in Denmark and Gotland, which they linked to the role of these places as centres of interaction and trade. This role is certainly borne out by a wealth of archaeological and other evidence from these places and seems to reflect complex mobilities as discussed in the case of Sigtuna above. At the same time, the authors ascribed high diversity in inferred ancestry of northern Norway to a different cause, its mixture of ‘North Atlantic’ and ‘Norwegian-like’ ancestry. This seems to imply some kind of back-migration, or close ongoing social and cultural contact between northern Norway and the North Atlantic region, which is also visible in some saga and other textual evidence (Jesch, 2015a: 64–68). Once again, there appear to be multiple and complex possible explanations for observed genetic phenomena which deserve further investigation.

Genomics, language, and identity

The authors found that ‘gene flow within Scandinavia was broadly from south to north, dominated by movement from Denmark into Norway and Sweden’. This raises interesting questions in regard to the linguistic affinities of the Scandinavian languages, which was acknowledged in medieval times by the use of the term dǫnsk tunga, ‘Danish tongue’, for the common Scandinavian language. One should certainly be careful in assuming an identity between genetics and language which may seem obvious but is difficult to justify from a linguistic point of view, since any coincidence between genes and language could be due to geographical factors. However, it is also true that a shared language favours interbreeding and so genetic convergence (Sims-Williams, 2012: 430). This ties into a longstanding discussion in archaeology regarding migration of people versus cultural diffusion as explanations for cultural change (Furholt, 2019). In this context, ‘cultural diffusion’ can include language, since individuals can learn new languages while they cannot change their genes. Such questions are not resolvable on a grand scale, since there are undoubtedly a variety of possible explanations for language commonality, variety, and difference beyond just the migration of speakers. But on a smaller scale, closer attention to language use can help to shed light on the genetic results. For example, an alternative interpretation of dǫnsk tunga, a term which is already found in a contemporary Viking Age source (a skaldic poem of the early 11th century), is that it expresses a Scandinavian sense of group consciousness which arose as a result of the diaspora and the encounter with other cultures (Jesch, 2015a: 75–79, 169), rather than having anything to do with Danes and Danish. The use of the term in Iceland does not imply Danish-like gene flow into the North Atlantic since, according to the study, although ‘Danish-like’ components are widespread, they are not found in Iceland.

Place-names given by migrants in their own language can also provide evidence of their presence. An underexplored theme is that of how place-names in the overseas Scandinavian settlements correlate with genetic evidence. Some work has been done on this for the north-west of England by Eleanor Rye (2016; see also Jesch, 2015b). Both linguistic and modern population genetic evidence (Bowden et al., 2008; Lall et al., 2020) suggest a strong Norwegian element in this settlement. Yet the authors did not take account of this diversity within England, generalising that ‘A Danish-like ancestry is seen in present-day England, in accordance with historical records, place names, surnames and modern genetics’. They acknowledged the difficulties of identifying Danish input, but not that the Norwegian input seems clearer (Kershaw and Røyrvik, 2016: 1674). Scholarship has long recognised the possibility of both ‘Norwegian’ and ‘Danish’ input into the Viking Age settlement of northern England, but this hypothesis needs further testing and refinement, using the very latest onomastic and linguistic research alongside the findings of archaeogenetics.

Language is often closely bound up with identity and, just as new languages can be learned, so people can assume an identity that does not necessarily accord with their ancestry. While it can still be difficult to avoid the ‘conceptual blurring of genetic descent and cultural identity’ (Furholt, 2018: 170), the authors were in tune with recent research when they stated that ‘“Viking” identity was not limited to Scandinavian genetic ancestry’. However, in addition to the complication of appropriate nomenclature already noted, this generalisation needs some drilling down – what does a ‘Viking identity’ consist of and how is it manifested? How does this happen if there is no Scandinavian ancestry involved?

An example given by the authors involved four Viking Age individuals from three sites in Orkney with archaeologically Scandinavian links. These individuals were interpreted as having ‘native British’ ancestry which in the Orkney context suggests Pictish descent. The sample size was certainly small for generalisations about ‘Viking identity not limited to Scandinavian ancestry’. Further such local studies are needed, not only for their own interest, but to build up a bigger, and possibly more varied, picture (Furholt, 2019: 170; Racimo et al., 2020; Veeramah, 2018). Also, a throwaway comment that the genomes ‘contrast with the isotopic evidence’ was not developed and is a topic worth pursuing, showing that it is not only the textual disciplines but also other forms of archaeological science that archaeogenetics needs to develop its dialogue with.

The authors suggested that two of these individuals ‘are probably the first Pictish genomes published’. Of the 10 individuals with more than 80% of this UK ancestry, five are found in Norway and Sweden, which suggested to the authors ‘that Pictish populations were integrated into Scandinavian culture by the Viking Age’. This conclusion is vague, and why is it ‘by’ the Viking Age – surely ‘during’ is more likely? Despite the inscrutability of this evidence, the question of the integration or otherwise of an indigenous population in Orkney and Shetland with the incoming Scandinavians (how, when, how much?) is a long-standing matter of discussion in the scholarship which deserves further input (Jesch, 2015a: 78). More extensive genomic analyses could certainly contribute something, perhaps comparing Iceland, with no indigenous population, and the islands of Britain and Ireland, with an indigenous population some of whom might have reached Iceland or even Norway, and working out what the population flows were at the time.

This particular comparison also deserves to be given a longer-term chronological aspect (see also Veeramah, 2018: 86, for the desirability of studies over several generations). The examples discussed above suggested to the authors that ‘some North Atlantic-like individuals in Orkney became culturally Scandinavian’, but a big question is how long it took for this process to happen, or how it happened. Evidence from skaldic poetry and from Old Norse mythology suggests that some aspects of this cultural Scandinavianness took some time to develop (Jesch, 2015a: 146–150, 177–182). It has been suggested that the archaeological evidence points in the same direction (Griffiths, 2019: 474–475). It was not a given that West Norse culture came to the Northern Isles with the first immigrants. Rather, their interest in poetry and mythology developed over several centuries, through contact with both the Norwegian homeland and their fellow members of the diaspora in Iceland. This also applies to runic inscriptions, as the example of Greenland shows.

The authors concluded that ‘the Greenland Norse were an admixture between Scandinavians (mostly from Norway) and individuals from the British Isles, similar to the first settlers of Iceland’. Although this is not surprising, as it is borne out both by the archaeology of Greenland and by the textual evidence for the settlement of Greenland, there are questions for future research hidden in this generalisation. For example, the runic inscriptions of Greenland (which are medieval, i.e. post-Viking Age) show clear affinities with Norway, where there is also a wealth of runic writing from the 11th to the 14th century (Jesch, 2015a: 173–174). Iceland, on the other hand, has very little in the way of runic writing until very late in the medieval period. Undoubtedly there are questions of survival and preservation of evidence to be considered, since the Greenlandic inscriptions are archaeological finds and Icelandic conditions are not conducive to this kind of preservation. However, the diaspora model shows that the Greenlanders, although they originally most likely came from Iceland in the Viking Age, did not necessarily have Iceland as their main point of cultural and social contact later on; rather, they were better connected to Norway in the medieval period (Jesch, 2015a: 59, 173).

Genomics and connectivity

The authors concluded that ‘different parts of Scandinavia were not evenly connected’ and that ‘Scandinavia comprised a limited number of transport zones and maritime enclaves with external contacts’. Importantly, they noted that the trading and cultural networks that stimulated the movements of people ‘took time to affect the heartlands of Scandinavia, which retained pre-existing genetic differences into the Medieval period’. This strengthens the need for more detailed, longitudinal, local-focused studies. That such local studies will indicate variety in the intensity of connectivity is suggested by some findings in which the authors identified close relatives from sites that are geographically widely separated. A pair of second-degree male relatives were variously from Galgedil on Funen in Denmark and Oxford in England. Another pair of individuals related in the third or fourth degree came from different parts of Sweden, a male from the island of Öland and a female from Skämsta in Uppland. It would be interesting to explore how these examples of individual-level mobility map onto the different patterns of connectivity. It will also raise questions about who exactly was mobile in this period. The textual evidence, with its origins largely in elite circles, might suggest greater mobility of the elites (Sindbæk, 2007: 63–65, 70–71), as also in the example of Sigtuna discussed above. A focus on enslaved persons, on the other hand, might show a different kind of mobility, as in the example of the ‘Celtic’ migrants to Iceland. Yet the settlement of Iceland (and perhaps also parts of Britain) shows migration of large numbers of individuals, not all of whom could have been either elite or enslaved, but must represent a large middle layer of society.

In contrast to the close relatives found far apart, kinship analysis of the Salme burials in Estonia showed that four of the 34 individuals were brothers. This is an important topic for understanding possible overlaps between those Vikings who raided and those who settled. This result will also encourage a re-evaluation of runic inscriptions commemorating ‘brothers’, which have often been interpreted as being of ‘brothers-in-arms’ rather than siblings (Jesch, 2001: 222–223, 247).

Both archaeology and textual sources support the hypothesis that back-migration and interaction between the newly settled areas and Scandinavia were an important aspect of the Viking phenomenon (a theme throughout Jesch, 2015a). This aspect still needs to be fully integrated into the archaeogenetic analysis. As a promising avenue for future research, it follows on from the authors’ observation that ‘western regions of Scandinavia received ancestry from the British Isles’. While this accords with the large amount of insular archaeological material found in Viking Age graves in Norway (Mikkelsen, 2019), there is also a possibly longer chronological dimension to this which it would be interesting to see if archaeogenetics can explore.