Life and death in a 5th-century ringfort

This aerial view of the Sandby borg ringfort from the west shows its location on the south-east coast of Öland island, Sweden. Stone ramparts enclose an oval area, but excavations have revealed that these defences could not protect the people living within the settlement. [Image: Sebastian Jakobsson]

It is not unusual for archaeologists to find caches of artefacts stashed in the ground, but their owners rarely remain nearby. Excavations on the island of Öland are revealing traces of a ringfort’s violent end. Ludvig Papmehl-Dufay, Helena Victor, and Clara Alfsdotter explain what a community’s demise can tell us about Sweden in the 5th century AD.

It all started in the spring of 2010, when geophysical surveys were carried out in a Migration period (c. AD 400-550) ringfort on the island of Öland, as part of a PhD project at Stockholm University. During the fieldwork, suspected looting pits were discovered, which led to a decision by the local authorities to subject the entire site to a metal-detector survey. This revealed five magnificent jewellery caches, containing large gilded silver brooches, along with items such as finger-rings, beads, bells, and smaller brooches. These remarkable finds, tucked away in different houses within the central block of the fort, were the catalyst for a small excavation the following year. It was not long, though, before the glamour of these discoveries took a macabre turn, when human remains displaying traces of lethal trauma appeared in the trenches.

This fine relief brooch was found together with a finger-ring and spiral beads in one of the jewellery caches at Sandby borg in 2010. [Image: Jan-Henrik Fallgren]

As the number of skeletons increased, it became apparent that our excavations were exposing evidence for a brutal massacre in the late 5th century AD. During this episode, large numbers of people were slaughtered and their bodies left where they fell. Of the ringfort’s 53 houses, only three have so far been excavated in their entirety. A total of about 15 individuals still lay within them, while a similar number are accounted for by human bones scattered across the street outside the houses. The dead span all ages, from infants to the elderly, but the sex range seems rather more restricted. Where determinations have been possible, so far only men are known to be present. Several of the skeletons show traces of fatal injuries, mostly to the head, which were inflicted by both sharp and blunt weapons. Many of these blows seem to have been struck from above or behind, while wounds to the victims’ forearms suggesting an attempt to defend themselves are conspicuously absent. In short, the evidence points towards a massacre rather than a battle.

A sensational discovery

The excavations in the Sandby borg ringfort are led by Kalmar County Museum and have been carried out annually since 2011. The ringfort itself is one of several on Öland, and lies close to the coast, protected by a stone rampart that traced out an oval area about 5,000m². Initially, the excavations were limited in scale, but over the last few years they have become somewhat more extensive. In total, less than 10% of the site has been excavated, but the findings are quite extraordinary. The contrast between shiny jewellery and sudden, violent death has made this one of the most startling archaeological discoveries in Scandinavia over recent years.

The slaughtered lambs under excavation in House 40. As they were aged between three and six months, it is likely that the Sandby borg massacre occurred between late spring and early autumn. [Image: Kalmar County Museum]

We refer to the three excavated buildings as Houses 4, 40, and 52. Within, we discovered a simultaneously varied and bleakly unambiguous picture of what transpired within the ringfort. House 40 and 52 both lie in the central block of buildings, and contained jewellery caches. House 4, by contrast, did not yield any high-status artefacts, and is located in the northern part of the fort. All three buildings contained human remains, ranging from two to nine individuals. House 40 contained six complete bodies and parts of another three, including several children, lying on the floor. In the back of this house, we found animal bones from at least eight slaughtered lambs, all of which died between the ages of three and six months, suggesting that the massacre was perpetrated sometime between late spring and early autumn.

Various finds from House 52 hint at it having enjoyed special status. These include the gold hoard found outside the building’s north end. [Image Daniel Lindskog]

In House 4, the skeleton of a 5- to 7-year-old child was found lying just inside the entrance, while the innermost part of the house contained the partially scattered skeleton of an older man. Close to the child and near the entrance, we encountered the remains of a young teenager who had been decapitated, as well as bones from another adult.

House 52 contained a child’s arm bone, but only a single complete body: an elderly man who was found face down across the central fireplace. He lay outstretched with his legs crossed. Fire damage to the man’s pelvic area shows that the fire was lit when he fell, meaning that he must have been unconscious or dead when he tumbled into the flames. The northern portion of this house features an unusual, beautifully rounded gable, where we found several spectacular objects just a few metres from the old man’s body. These included a jewellery cache, as well as shards of Roman glass and even a small gold hoard, which included a Roman gold coin. On the street, just beyond the rounded gable, were sherds from finely decorated ceramic vessels, of a type commonly associated with drinking ceremonies. These artefacts, together with the unusual architecture, suggest that this house (and possibly also the elderly man) may have been associated with religious and cultic activities – presumably high-status – held in the northern part or ‘hall’ of the house.

Sherds from at least two beautifully decorated ceramic vessels were also found outside the north end of House 52. Such receptacles are associated with drinking ceremonies. [Image: Kalmar County Museum]

Fabel, the archaeology dog

The unusual circumstances at Sandby borg present exceptional scientific opportunities, but also considerable challenges. Given that this was a period when cremation was the favoured burial rite in Sweden, the human remains represent an invaluable source of information. They can reveal much about not only the massacre, but also the prior health and demography of the victims, opening a window onto the reality of life during this period on Öland island and in southern Scandinavia more generally. Naturally, accessing this information is dependent on finding and recovering skeletons at the site. But how could we pinpoint where unburnt human remains lay underground?

Archaeobotanist Jens Heimdahl in his field laboratory, investigating plant remains recovered during the June 2018 excavations. [Image: Daniel Lindskog]
A gilded bronze pendant found outside Sandy borg in 2014. The object measures about 2.5cm in diameter and seems to date to some centuries after the massacre. Clearly people were still visiting the site, but why?

At Sandby borg this question found a novel answer, when Sophie Vallulv, an archaeologist and dog-trainer, embarked on solving the conundrum. For several years, she trained Fabel, her German Shepherd, to catch the scent of buried and unburned human bones. Remarkably, tests in the laboratory and at the ringfort uniformly suggested that the method worked, so in 2015 Fabel received his formal archaeology dog certificate! Since then, we have used Fabel on several occasions to detect where skeletons might be found, and in most cases the archaeological results have proved him right.

This is an extract from an article featured in issue 96 of Current World ArchaeologyClick here for more information about subscribing to the magazine