A Female Viking Warrior Interred at Birka

In archaeology, we distinguish osteological sex from artefact gender. Osteo-sex is with very few exceptions (odd chromosomal setups) the same thing as what your genitals are like. Artefact gender is the material correlate of a role you play according to the conventions of your time: e.g. whether you keep your genitals in Y-fronts or lacy knickers. We judge these two parameters from separate source materials. Your skeleton can’t tell us anything about your gender, and your grave goods can’t tell us anything about your osteo-sex. They are in principle able to vary independently.

Nevertheless, 1st millennium Scandinavians seem to have been quite conventional about this: mismatches between osteo-sex and artefact gender are extremely rare. The graves are clearly divided into osteo-female jewellery graves and osteo-male weapon graves. If you exclude cremated bones and poorly preserved inhumations that can cause misdeterminations, the number of mismatches shrinks even more. And when you do see a mismatch it’s typically partial: e.g. a male skeleton buried with a full set of weaponry and horse gear, plus a single ladylike brooch. I was until recently not aware of any well-preserved and richly furnished Scandy inhumation of the 1st millennium with a complete mismatch between osteo-sex and artefact gender. But now we have one.

Birka’s grave 581 is one of the famous chamber inhumation graves where this Swedish Viking town’s 10th century elite buried their dead. It has loads of high-quality weaponry and two horses. It has no hint of any female attire. And it has the skeleton of a person whose funny bent position suggests that, like in many other chamber graves, the individual was buried sitting on a chair and then keeled over inside the chamber.

In the 1970s, the skeleton had become disassociated from the artefact finds, and an osteologist (sadly uncredited in the paper discussed below) quietly identified it as female. In 2014 osteologist Anna Kjellström identified the bones as belonging to Bj 581, the famous weapon burial, and agreed that the skeleton is female. Certain archaeologists have replied that they don’t believe this because of the weapons. Others have suggested more diplomatically that maybe the bones represent two individuals, or that a male body was removed while still articulated. Others again have simply dismissed the whole issue with reference to 19th century sloppiness in keeping the Birka bones correctly labelled grave by grave.

Now a team of researchers, of whom I am proud to count half as my professional buddies, have sequenced the genomes of the bones. Yes, plural. To test if the skull and one arm are from the same person. There is only one person there, and just as Kjellström said, she’s biologically a woman. I am extremely happy with this investigation, because it gives us our first real female Viking, and it shows that osteologists can indeed judge osteo-sex correctly on well-preserved ancient skeletons. Very commendably, the paper is available online in full for free: Open Access.

Here’s a few notes.

  • The grave was selected for analysis because of the controversy over its osteo-sex. It is not a randomly chosen weapon burial that happened to prove female. If you pick a random Birka inhumation, this is not the result you are likely to get.
  • Assuming that burial furnishings speak directly about a person’s role in life (which is always debatable), we don’t know if the dead person was perceived as a cross-dressing woman, or just as a man. In other words, we have no way to tell if she was “out”. There are examples of both from later centuries, where for instance Joan of Arc never tried to pass as a man despite wearing armour and commanding an army.
  • The plan of the grave shows which bones were well preserved. This should be enough to counter the charge that maybe the skeleton currently labelled Bj 581 is not in fact the one found in this weapon grave. This the authors should have written a few sentences about. I take their silence to mean that having already published her arguments about this elsewhere, Kjellström considers the issue uncontroversial.
  • We still can’t rule out the early removal of an articulated male body. But such an argument ex silentio would demand that we place similar female bodies in all other weapon graves as well. We can’t just create the bodies we want in order for the material to look neat.

The “Discussion” section hasn’t been properly copy-edited.

  • I don’t know what “The archaeological material provides a reference for the Viking Age” means.
  • Because of the odd phrasing, I don’t know what the authors are trying to say about earlier scholarship here: “Although not possible to rule out, previous arguments have likely neglected intersectional perspectives where the social status of the individual was considered of greater importance than biological sex. This type of reasoning takes away the agency of the buried female.”
  • “Grave Bj 581 is one of three known examples where *the* individual has been treated in accordance with prevailing warrior ideals lacking all associations with the female gender” : “The” here should be “a female”.

Hedenstierna-Jonson, Charlotte et al. 2017. A female Viking warrior confirmed by genomics. American Journal of Physical Anthropology 2017. DOI: 10.1002/ajpa.23308

I discussed the issue of shield maidens in 2013, the year before Anna Kjellström went public with her identification of the female skeleton with Bj 581.


  1. #1 John Massey
    September 13, 2017


    I read that, but can’t say that I (a) particularly enjoyed it or (b) am much the wiser for it. OK, I get the point, actually already had it, that there were not dozens or hundreds of fierce shield maidens terrorising various parts of Europe during the Viking Period. But, from what you say, there are three known examples of females buried with all of the accoutrements and trappings of war, and lacking any grave goods normally associated with females; so, rare, but not unique or unknown.

    So, Bj 581 is not unique, which tends to puncture the “mixed up remains” idea at least somewhat.

    I am still fixated on the remains themselves, which are now clearly demonstrated to be unequivocally female, the DNA confirming the osteology, and that they are of a single individual.

    A quick online check (which I should have made earlier) suggests that she was about the same height as the average male Viking – so, for a woman, she would have been considered to be quite strikingly tall; well above the mean for Viking females. I had assumed that the mean height of Viking males in their prime years would have been taller, but evidently not.

  2. #2 Vilhjálmur Örn Vilhjálmsson
    Denmark, Iceland and http://fornleifur.blog.is/blog/fornleifur/
    September 16, 2017

    Is this interesting comment by Fedir Androshchuk really not worth mentioning?

  3. #3 Vilhjálmur Örn Vilhjálmsson
    Denmark, Iceland and http://fornleifur.blog.is/blog/fornleifur/
    September 16, 2017

    … and not to forget Judith Jesch’s important contribution, might it not now be in order to ask for a second opinion with a thorough revision of the material used for the DNA sequencing of Bj 581? See http://norseandviking.blogspot.dk/2017/09/lets-debate-female-viking-warriors-yet.html#links

    DNA-scientists’ despect for everything except DNA is not a new problem. Because of bad statistics in connection with DNA studies, Icelanders are now allegedly descendants exclusively by “Celtic” women (females from the British isles) and Norse men, although this cannot be detected in traditional osteology, indexing of limb proportions, etc.

  4. #4 John Massey
    September 16, 2017

    Vilhjálmur@3 – To be fair, in this case the DNA testing was done simply to cross-check and confirm the osteological findings. And determining an XX individual from an XY individual is about the simplest DNA determination there is – no statistics required. But then, the osteology should be pretty clear also. Whether the remains of the individual tested are properly associated with the grave goods in grave 581 is not something I can comment on with any knowledge.

    I share the frustration with the sensationalising of the finding, extending into hyperbole, but unfortunately there are always people who will do that.

    In the case of Icelanders that you refer to, those DNA studies sound very dodgy to me, because geneticists cannot differentiate ‘Celtic’ from others. My understanding is that ‘Celts’ were never a genetically distinct population, and certainly there is no reference population of ‘Celtic people’ that geneticists can refer to – not as such. They can differentiate post-Roman British inhabitants from Anglo-Saxons, pretty well, but that is about all.

    They can differentiate “British/Irish” as a single population (they cannot separate those two populations) from “Scandinavians” as a separate population. They cannot differentiate between French and German as two distinct populations, so my Prussian great grandfather whose mother was actually half French and half Spanish just shows up in my ancestry as “French/German” (one group) and “Iberian” (another group).

    So you are right to be dismissive of those studies of Icelanders, in my opinion – they definitely sound very suspicious to me.

  5. #5 Vilhjálmur Örn Vilhjálmsson
    September 16, 2017

    John Massey, thank you very much for your comment. I should have added that the studies of the maternity of first settlers in Iceland bases solely on the mitochondrial DNA in modern Icelanders, as studied by Iceland-based genetic-enterprise deCode (Íslensk Erfðagreining). No ancient DNA was studied. The DNA researchers didn’t bother to mention the results of the most resent results of metric studies on human bones from Icelandic Viking Age burials. The main responsible researcher knew them well. Such a study was initiated by me and Human biologist Dr. Hans Christian Petersen of Odense University (University of Southern Denmark) in the first part of the 1990s. H.C. Petersens’ results clearly indicated a British / Insular influx in the first Icelandic population, while the bulk of the first settlers, whose bones were measured, showed the closest relationship to values measured on Norwegian Iron Age and Medieval populations. The settlers, who matched best the British / Insular indexes were about 15 % of the measured individuals. 10 to 15 % of the individuals showed values and traits similar to Saami ones. Thus, Petersens’ results didn’t contradict the information which possibly can be extracted from the Landnámabók and other written sources of the 12th century about the origins of the first settlers in Iceland. Thus, the British Isles / Celtic / Insular origin of the Landnam-women of Iceland claimed by DNA-sequencing of modern Icelanders lacks all support from osteological observations of the bones of the first settlers themselves.

  6. #6 John Massey
    September 17, 2017

    Vilhjálmur – Thanks.

    Just using the mtDNA to identify origin of first settlers is questionable, because they are not identified uniquely with any particular population.

    To use an example, my mtDNA is from a female associated with Pleistocene hunter-gatherers in Europe, which is pretty rare among modern European populations, occurring at a frequency of only about 5%. But I cannot tell anything at all about where in Europe that ancient ancestor might have been located. She might have been Saami, for all I know, but there is simply no way to tell.

    From my brief reading of Icelandic history, it seems the first settlers did take some ‘Gaelic’ people with them (as serfs or slaves or whatever), but Gaelic is a language group, not a genetic population, and I have never seen mtDNA haplogroups uniquely assigned to any Gaelic ethno-linguistic group. It seems very unlikely to me that *all* modern Icelanders would be descended from Gaelic-speaking women, even if the mtDNA of such women could be uniquely identified.

  7. #8 Birger Johansson
    September 17, 2017

    I don’ t really see why so many get worked up about warriors/soldiers. I mean, they exist in all agricultural societies, and sometimes women got work as archers.
    But there was nothing remotely romantic about it, not then and not today.

  8. #9 John Massey
    September 18, 2017

    90% boredom and 10% terror, for normative soldiers.

  9. #10 Martin R
    September 18, 2017

    90% girls and 10% boys, for heteronormative soldiers. 😀