Nerman’s Weaponry

Professors tend to have a few pet issues that they emphasise time and again over their careers as researchers and supervisors. This is quite clear with two 1960s-70s professors in my field. In Bertil Almgren’s case, one such pet issue was the source-critical quality of archaeological information. In Mats Malmer’s case, one was clear and exact verbal definitions of terms.

I agree with both of these imperatives. But there’s one case where an adherence to Almgren’s priorities over Malmer’s was clearly not the right way to go.

Birger Nerman’s monumental folio-format work Die Vendelzeit Gotlands, about portable objects from Gotland in the period AD 550-800, appeared in two parts. The illustrations in 1969, two years before Nerman died. The text in 1975, four years after he died, with heavy input from his daughter Agneta Lundström. The two books place every well-preserved object found before 1969 in one of five phases.

In 1983, Bo Annuswer wrote his third-term paper in Uppsala about a part of Nerman’s work, with Almgren as his supervisor. Annuswer looked at all the find combinations with weaponry in them, and classified them according to source-critical quality. He concluded that the weapon chronology with its five phases has extremely weak source-critical footing.

You can criticise Annuswer for not using all available sources of information. A good deal of the find combinations are not at all as poorly supported by archival information as he claimed. The whole thing was a bit of a hatchet job on Nerman, written to flatter Almgren’s source-critical agenda. But that’s not my main issue with Annuswer’s paper.

The thing with Nerman’s weaponry chronology from a malmerian perspective is that it doesn’t properly speaking exist at all. There are no type definitions, no seriation, no identification of diagnostic types. It’s really just a lot of pictures divided into five sections and some extremely brief supporting text from Nerman’s posthumous editor. Die Vendelzeit Gotlands analyses nothing, it just postulates a chronology out of the blue.

You don’t have to spend weeks examining the source quality to take a hatchet to Nerman’s weaponry chronology. Because there never was a scientifically argued chronology to begin with.


New Dates for the Bronze Age

When I was an undergrad in 1990 we were taught that all six periods of the Scandinavian Bronze Age were 200 (or in one case 300) years long. The most recent radiocarbon work shows that they all had different lengths and were more likely 130-280 years long. And the periods with the most abundant metalwork finds, II and V, are the two shortest. So their previously known status as metal-rich eras looks even more pronounced now, and the intervening periods look even poorer.

Per. I. 1700-1500 cal BC (200 yrs)
Per. II. 1500-1330 (170 yrs)
Per. III. 1330-1100 (230 yrs)
Per. IV. 1100-950/20 (165 yrs)
Per. V. 950/20-800 (135 yrs)
Per. VI. 800-530/20 (275 yrs)

Each of these periods translates to a list of artefact and monument types that are commonly found together. Their relative ordering through time has been known since the 1880s. Current work looks at the absolute dates at which these typological laundry lists were current. It uses a new technology, radiocarbon dating of cremated bone, and new applications of Bayesian statistics, which allow us to constrain the uncertainty of the radiocarbon results using stratigraphical observations. The latter means that if we know that grave B was later than grave A because one sat on top of the other, then we can tell the software to disregard parts of the probability distributions that gainsay this observation.

Hornstrup, K.M et al. 2012. A New Absolute Danish Bronze Age Chronology As Based On Radiocarbon Dating Of Cremated Bone Samples From Burials. Acta Archaeologica 83. Copenhagen.