A case of bloody speculation...

... or why scientists and historians need mutual understanding to avoid a fight over genetic heritage.

The brutal Viking stereotype has a convincing pedigree. Their culture venerated ideas and images of ruthless aggression. Their victims' view differed only in attitude. In 1014, Wulfstan, archbishop of York, wrote a specific account of atrocities meted out by the Danes, including enslavement, plundering, arson, murder and rape.

The Viking campaigns began as a protective response to internal needs and external pressures that threatened their religious and cultural traditions as the growing church in the West hoarded the treasure on which Scandinavian social networks depended.

But in tackling their economic needs, the Viking adventurers proved progressive. Reaching out through Russia to Baghdad to trade, developing urban markets and improving merchant shipping, they bequeathed to northern Europe the strong economic infrastructure it has profited from since.

Recent scholarly developments have recruited the Vikings to link human genetics with history. The mapping of gene types in modern populations is primarily done for medical research. But given their hereditary nature, it is inevitable that historical explanations should be proposed for the distinct geographical patterns discovered. And geneticists have proved remarkably enthusiastic about using their science to reconstruct the past.

These investigations pose exciting but serious challenges to historians and scientists. The white heat of contemporary science does not automatically reveal facts that revolutionise our perceptions of the past in the way dendrochronology and radiocarbon dating have done. Oversimplification in proposing historical explanations of scientific data is always a danger. Sadly, good science and poor history provide only poor history.

Two recently published scientific papers have focused on the Viking contribution to the genetic make-up of the population of the Orkneys. The islands were under Scandinavian rule from the 9th to the 15th century and became Norse-speaking. There has been keen debate over how much demographic change occured in this period.

From a historical perspective, the less convincing of the two papers appeared in the Proceedings of the National Academy of Sciences in April. This featured in The Times under the headline "Viking blood still flows in Orkney today", with the results ostensibly endorsed as "exciting" by Colin Renfrew. This research is now the cornerstone of a BBC-sponsored project covering all of Britain, under the garish title The Blood of the Vikings .

The study considered male lineage represented by the Y chromosome. A sample of 71 males with surnames recorded in the Orkneys before 1700 was examined. The intention was reasonable - to get as close as possible to a "historical" population. But subtler questions were apparently unasked. Would a representative proportion of the population have had its surnames recorded before 1700? Do we know how stable surnames have been since? Most importantly, how many different surnames and thus families were involved?

Two genetic "haplotypes" were identified as diagnostically Scandinavian, accounting between them for 60 per cent of the Norwegian control sample and per cent of the Orkney sample. The remainder of the Orkney sample was made up of three haplotypes, which provided 32 per cent of the Norwegians. If the two distinctive haplotypes represent colonists from Norway, by implication a matching proportion of the non-distinctive groups should too. This puts the proportion of the Orkney sample of putative Norwegian origin at just over 40 per cent.

The conclusion drawn is that Viking invasions "left a significant paternal genetic legacy". But significant in what sense? Historically, the idea that nearly half the population of the Orkneys in the late 17th century had a Scandinavian male ancestor has little revelatory value. We cannot extrapolate from this the levels of Scandinavian colonisation and native survival in the Viking period, as the gene pool could have accumulated throughout the Middle Ages and indeed beyond.

It is interesting to compare this figure with the 35.5 per cent female Scandinavian input to the Orkney population calculated by Agnar Helgason et al ( American Journal of Human Genetics , March 2001) from mitochondrial DNA. Disconcertingly, the team's comparative mtDNA analyses produced results so varying, they entitled their article "Genetic evidence for different male and female roles during cultural transitions in the British Isles".

Sooner or later, some cynic will coin the phrase "lies, damned lies and genetics". But it will be unpardonable if this whole area becomes a battleground for historians and geneticists. These parties need to understand one another and collaborate throughout research projects. Above all, historians need to think about what they will do with the new demographic information being revealed. As in the Orkney case, it may not be startling. But it will offer historians new scope to contextualise cultural transition against demographic facts.

As for the Vikings, let us keep a proper view of their activities. They came to take, not to give. They did leave a legacy of global importance, but we risk missing that if we try to accommodate them too comfortably within ourselves.

John Hines is professor of medieval archaeology and history at Cardiff University and editor of Medieval Archaeology .