More than a thousand years after the Viking Age drew to a close, there’s still a lot we don’t know about these seafaring Norse warriors, who explored territory from the furthest reaches of Russia to the earliest settlement in North America and left a lasting mark on the lands and peoples they encountered.

Now, archaeologists are attempting to piece together a clearer picture of one of the darker aspects of the Viking world: slavery.

Historical accounts make it clear that when they raided coastal towns from the British Isles to the Iberian Peninsula, the Vikings took thousands of men, women and children captive, and held or sold them as slaves—or thralls, as they were called in Old Norse. According to one estimate, slaves might have comprised as much as 10 percent of the population of Viking-era Scandinavia.

While hard evidence in the archaeological record may be scarce, what seems clear is that slavery played an important part in the Viking way of life, as in many societies both before and since. In fact, the desire for slaves might have been one of the main reasons Vikings began raiding in the first place.

Many of these slaves came from the British Isles and Eastern Europe. In one historical account of Viking-era slavery, an early-medieval Irish chronicle known as The Annals of Ulster, described a Viking raid near Dublin in A.D. 821, in which “they carried off a great number of women into captivity.”

This is one of numerous written sources referring to slavery in the Viking world, which include historical chronicles produced within northern European monasteries—often by people who were the victim of Viking attacks. Other sources emerged from the Arab world, including the account of the 10th-century geographer Ibn Hawqual, who in A.D. 977 wrote of a Viking slave trade that extended across the Mediterranean from Spain to Egypt.

“These sources provide very clear indications that Viking raiding groups are engaging in slaving activity,” says Ben Raffield, an archaeologist at Sweden’s Uppsala University who is conducting research on the Viking slave trade as part of the Viking Phenomenon project.

In contrast to the wealth of historical and literary evidence for Viking-era slavery, actual archaeological proof remains relatively sparse. In an article published in the journal Slavery & Abolition in April 2019, Raffield detailed what has been discovered so far, starting with a collection of iron collars and shackles found at several sites thought to be Viking slave trading hubs, like Dublin (Ireland), Birka (Sweden), and Hedeby (Denmark).

Though it’s been suggested the objects could have been used for restraining animals, rather than humans, Raffield argues that their presence in these urban centers (rather than rural areas), as well as their concentration near the harbors tends to support their use on slaves. “They look strikingly similar to all kinds of restraints that have been used on humans throughout history, from antiquity to the early modern period,” he says.

Aside from the collection of restraints, researchers have discovered what may be evidence of slave quarters—an arrangement of smaller houses surrounding a large house at Sanda, a Viking site in Sweden.

“The few that have been excavated seem to have been used for crafting activities, things like textile making,” Raffield says. “They strangely look quite similar to what you see in the United States in the antebellum period.”

Scholars have long wondered why the Vikings suddenly emerged as a formidable raiding force in the late eighth century, starting with their attack on the Christian monastery of Lindisfarne, located on the northeast coast of England, in A.D. 793.

The answer might have been a need for foreign slave labor to help build their enormous fleets of ships and produce the textiles for their sails. Raffield and his colleagues see the desire to take slaves as a possible motivating factor behind the Viking expansion. “Fleets of hundreds of ships [were] sailing out of Scandinavia in the 9th century,” he says. “We wonder whether you would need a new labor force to produce the materials you need to do that.”

Slaves—who could also be traded at international markets—may have represented another type of resource for the Vikings, too. Evidence suggests Vikings often targeted women and girls in their raids, suggesting the existence of sexual slavery, as well as intermarriage. There are also indications that Vikings practiced polygamy, which in their highly stratified society would have meant that poorer unmarried men might have had limited access to women, and would have targeted female slaves as concubines (or even wives).

DNA mapping of the modern Icelandic population found that up to two-thirds of Iceland’s female founding population had Gaelic origins (either Ireland or Scotland) while only one-third had Nordic roots. The reverse was true for the male population, suggesting that many Nordic men in Iceland had children with women who were likely taken in raids from the British Isles.

It’s also possible that in addition to sexual motives, Vikings might have targeted women as slaves because of their specific value as a source of skilled labor. “Quite often in a slaving context, women are taken because in a lot of societies they are traditionally the people who produce high-value goods,” says Raffield. “A lot of people think if you wanted captives for labor, you would take men, but that's not necessarily the case. Textile working in Scandinavia, for example, is strongly associated with women.”

Whatever motivated the Vikings to start taking slaves, evidence suggests they were often brutal with those who had the misfortune to be captured. In one study, research Anna Kjellström of Stockholm University examined the skeletal remains of presumed Viking-era slaves found in graves in Norway, Sweden and Denmark, and found that they showed signs of abuse and decapitation.

In some cases, the slaves were buried alongside their masters, suggesting they may have ended up as human sacrifices, and included with grave goods to accompany powerful Vikings into the afterlife.

While written sources provide strong evidence of slavery in the Viking world, the slaves themselves—why they were taken, how they were transported, where and how they lived—left little trace on the archaeological record.

Raffield stresses the need to more fully excavate Viking sites where slaves are believed to have lived. Ultimately, there may be limits to how much we’ll ever know about forced labor in the Viking Age, beyond the evidence gleaned from written sources and archaeological digs.

“The thing about studying slavery and captivity is that these groups are often described in the archaeological literature as invisible, or unseen,” Raffield cautions. “Their movements are curtailed, they're denied of possessions, they're not always accorded formal habitation—places to sleep, places to live. They're really hard to identify in the archaeological record.”