Archaeology is the study of people and their actions, preserved through the physical traces they left behind. How, then, could we not appreciate the detailed insights into individuals, communities, and populations provided by their very DNA? On its own, though, aDNA can only tell us so much, and it is the combination of genetic data and archaeological material that gives us the best new view of the post-Roman period. Artefacts, sites, and even entire landscapes represent the material remains of the past, and they are also contingent on specific events in time. It was people who dug graves, who dressed their dead, who placed them in the ground, and who looked for comfort when sharing their grief. What can we learn from the archaeological echoes of these communal rituals?
In the east of England, we find richly furnished burials dating between the 5th and 8th centuries. These are often described as early Anglo-Saxon cemeteries, and it has often been inferred that they were created by the new settlers described in near-contemporary sources (as John Hines explores from p.20 of this issue). But how many settlers were there, how did they build their communities, and how did they integrate with the pre-existing population? To understand these questions, let’s delve into the 11 early medieval cemeteries featured in our study in more detail. These sites are primarily concentrated on England’s south and east coasts, stretching from Worth Matravers in Dorset to West Heslerton in North Yorkshire. From this sweep of geography, we took 235 samples. Strikingly, each site has its own unique story to tell. We have already explained how DNA evidence for these sites was analysed (see the preceding article, from p.26), but – for convenient reference here – a quick reminder that, when referring to ancestry, CNE stands for Continental Northern European and WBI for Western Britain and Ireland. This distinction gives us a way of comparing the data from cemetery to cemetery. For example, on the south coast there is markedly less CNE DNA compared to the sites we studied elsewhere. The most-westerly cemetery, Worth Matravers, has the least, with just 6%. Rookery Hill, at Bishopstone in East Sussex, had 15%; and, although Apple Down (West Sussex) appears to have a more-significant proportion of CNE DNA at 53%, this figure is still one of the lowest in our study. By contrast, the Kentish cemeteries have relatively high proportions of CNE ancestry, with Polhill showing 92%, Buckland, near Dover, 81%, and Updown, near Eastry, 60%. Meanwhile, over on the east coast the proportions of CNE ancestry are consistently high: Lakenheath has 92%, West Heslerton 88%, Oakington 81%, Ely 79%, and Hatherdene Close 72%. This is not surprising, as these sites sit closest to the North Sea and were therefore highly exposed to immigration from the Continent. Nonetheless, every one of these cemeteries contained individuals of apparently mixed ancestry, albeit with these people present in different proportions, providing us with evidence of clear but regionally distinctive patterns to early medieval genetic ancestry. It appears that post-Roman migration must have been experienced differently from place to place.
By analysing these patterns, we can explore regional difference in various ways – but first of all we must understand the sample itself. Out of the 11 cemetery sites in our study, nine have well-documented associations between artefacts and grave contexts – a body of evidence encompassing 210 individuals. Within this group, perhaps the most-important observation is that there was no significant difference in the presence of CNE or WBI ancestry between men and women. Individuals of both ancestries were also well represented both in prominent and/or furnished burials and in unfurnished ones. This is key, because it implies that the Anglo-Saxon migration (and its subsequent genetic impact) was not driven by population movement specific to sex or status. In other words, this was not predominantly an elite male migration, as has been previously suggested.
Drilling down into statistical patterns, there are other illuminating insights to be found. In the cemeteries that we sampled, women with 50% or more CNE ancestry were the most likely to be found with grave goods, and particularly with brooches. This trend did not continue for men, though: the graves of males with predominantly CNE ancestry were no more likely to contain grave goods or other artefacts than those with mainly WBI ancestry. At Updown in Kent, for example, the male occupant of Grave 37 – who had almost 100% WBI ancestry – was interred with a seax under a barrow surrounded by a ring ditch. This later 6th- or early 7th-century individual had been accorded a prominent, highly visible weapon burial, suggesting that he was someone that the grieving community wanted to place in the foreground of their cemetery space (see the following feature for more on this site). By way of contrast, 300 miles to the north at West Heslerton, all eight of the male weapon burials that we sampled displayed near 100% CNE ancestry. This was not a rite reserved only for individuals with specific origins.
One of the most-fascinating aspects of the genetic data is that these results have highlighted several graves that – based on their skeletal characteristics – had been assigned to the wrong biological sex during previous investigations. Notably, some weapon graves had been identified as female and subsequently discussed in the wider literature as evidence of a complex, multi-layered cultural relationship between sex and gender during the early medieval period. At West Heslerton, for example, Grave 184 contained a shield boss, spearhead, knife, and buckle, and Grave 144 held a spearhead, knife, and buckle. Such objects are traditionally associated with male graves, but the skeletons found with these objects had been osteologically identified as female. Following analysis of these individuals’ DNA, both have now been reidentified as genetically male. Such reassessments have also gone the other way: at Buckland, Grave 383 contained beads; Grave 350b had beads, a loop, a buckle, and rings; and a disc brooch, beads, a pair of small square-headed brooches, a buckle, a knife, rings, an armlet, and tweezers were found in Grave 281. Going by grave goods alone, all three would typically be interpreted as female burials, but skeletal characteristics had led to them being interpreted as male. Following aDNA analysis, all of these graves have now been reclassified as genetically female.
That does not mean that there were no graves where artefacts and genetic sex did not tally as expected – we found a single example of such a difference (representing 0.5% of the sample population of 210 people) in the form of Grave 122 at West Heslerton. This burial contained the remains of a juvenile who had died at the age of 12-15 and who had near-100% CNE ancestry. This individual had been buried with an equal-armed brooch, beads, and a knife, leading to suggestions that this was a young female, but genetics reveal otherwise. Moreover, Janet Montgomery’s earlier study of skeletal isotopes had already flagged this individual as unlikely to have been local to the area. The combination of information is wonderful: the equal-armed brooch gives us an early date, and taken together the archaeological and genetic evidence identifies this person as a Continental migrant who arrived into East Yorkshire early in the post-Roman period.
The fact that our sample includes three cemeteries from East Anglia and three from Kent also allows for some limited investigation of regional differences. Take, for example, the wider pattern mentioned above, of mainly CNE ancestry women being the most likely to be found with grave goods. This trend is not followed by our Kent sites, but it is evident in East Anglia (though not in the case of brooches, which were as likely to be found with mainly WBI ancestry women as they were with mainly CNE). At Hatherdene Close in Cherry Hinton, for example, there seems to have been a particularly strong association between artefacts and ancestry. There, individuals with CNE ancestry predominate in furnished graves, such as Skeleton 225, an adult woman aged 26-44, who was found with a great square-headed brooch, a cruciform brooch, and a small long brooch, as well as a knife, a buckle and beads. Indeed, of the four near-100% WBI burials at this site, only one was found with any artefacts: a 26- to 44-year-old man who was buried with a knife and buckle.
There were some eye-catching East Anglian exceptions to this rule, however – none more so than Grave 80 at Oakington, just eight miles from Hatherdene Close. Aged 20-25, its female occupant had predominantly WBI ancestry, but she had a remarkably well-furnished burial, accompanied in the grave by silvered disc brooches, beads, a knife, a girdle hanger including a chatelaine – and a fully articulated cow (see CA 270 for more on this burial). The loss of this animal would have represented a considerable sacrifice for her community, suggesting that this woman was of some social standing. Adding to this picture, she was also probably interred under a small barrow that was surrounded by satellite graves, and so visually was one of the most-prominent 6th-century burials in the cemetery.
Integration and isolation
Above all, the story emerging from our research is one of how local communities adapted to change, and how this varied between regions. Perhaps the two sites with the greatest contrast in this respect are Buckland and Apple Down. At the latter site, it is a story of segregation. There, the graves can be classified into two distinct groups (A and B) according to their alignment, location, and quantity of grave goods. Group A graves contained more children and older individuals, and were located more towards the middle of the cemetery; these individuals were also more often associated with CNE ancestry. Group B graves, by contrast, contained the skeletal remains showing the most signs of wear and tear acquired in life; they were more commonly located towards the fringes of the site, were aligned north–south, and were more likely to contain WBI ancestry. At this site, then, there were significant differences in the treatment of individuals which correlated with their ancestry – in which case it is even more striking that the mean CNE ancestry at Apple Down was 53% of those sampled. This site appears to have been fairly evenly mixed in terms of population, making the evidence of separation between two distinct groups even more stark.
The situation was completely different at Buckland. This was the most-comprehensively sampled cemetery in our study, and as a result we can observe the mixing of genetic and cultural identities at a family level. Although mean CNE ancestry was 80%, within the cemetery population we see much less obvious differentiation between ancestries than at Apple Down. For example, we found a kinship group spanning at least three generations with unadmixed CNE ancestry – but down that line, a woman with near-100% WBI ancestry was integrated into this group. She was buried in Grave 304, and her two daughters (Graves 290 and 426) were of mixed ancestry. WBI ancestry entered this family again a generation later, visible in near 50/50 mixed-ancestry grandchildren who were identified in Graves 414, 305, and 425. As for how these individuals were treated in death, grave goods including brooches and weapons were found on both sides of this family tree, and the mixed younger generations were found with weapons, beads, and pins.
At Buckland, an interesting pattern of grave location emerges over the course of four generations. Grave 426, that of a woman with a buckle, was buried to the north of the main concentration of genetically linked burials, but her child who died youngest – a boy of 8-10 years – was buried within the cluster, which also included his mother’s parents, grandparents, and their close family. His two siblings, though – who had both lived to adulthood – were buried close to their mother. Perhaps the occupant of Grave 426 had laid her young son in his grave, safely tucked in among her family, but her adult children were the ones who had buried her. We might then imagine a scenario in which they chose a space for her close to their father (although he remains unidentified) and his relatives; it is in this space that they too would eventually be buried.
DNA allows us to dig into family histories, but what we find there is not simply the minutiae of the past: these family histories are the story of migration and of cultural change in the post-Roman period. When combined with archaeological evidence, DNA data can be used at multiple scales which include national, regional, and local levels. Although furnished early Anglo-Saxon cemeteries were a phenomenon of the post-Roman east of England, they were also deeply rooted in their locality. The architecture and history of these cemeteries reflect the people who made them, who dug the graves, and most importantly who buried the dead. These people adopted different strategies depending on their family histories, integrated with or segregated from the local population, and that decision was made at a community level and varied from place to place. What is interesting is that grave goods seem to have played only a very limited role in the signalling of different ancestries – assuming that was what was intended – and where it is seen, that signalling was dependent on biological sex.
All images: Duncan Sayer, unless otherwise stated