Since only small fragments of serks have been found, the reconstruction of the serk proves to be more difficult that the type of garment itself. With medieval clothing one can turn to tapestries, paintings and other types of illustrations. When it comes to the Viking Age, hardly any illustrative evidence can be found.
If we instead turn to the archeaological findings we have a few Scandinavian excavations that have given a little more insight on the matter. Based on the majority of serk fragments found during the excavations in Birka during the 1870’s, I’ve constructed a plain linen serk. The serk evidence from Birka is very fragmentary but the numerous amount of fragments can however indicate both construction and cut if analysed together.
Analysis of the linen fragments from Birka indicates that the majority were left undyed, which is why I choose an unbleached, greyish linen fabric in a tabby weave. As with the Haitbau serk (which I’ve reconstructed here) there is little to no evidence of the length of the serks. Archaeology doesn’t present any fabric fragments other than those preserved by metal artefacts, which usually weren’t placed at the lower parts of the garments.
The Hunnestad Monument in Ystad, Sweden, has an image stone (DR284) depicting the giantess Hyrrokin riding a wolf, wearing a serk that comes no lower than to her knees. It’s not impossible to imagine the serks being short to function as work clothes and not get soiled at the hems but at the same time the sub-arctic climate in Scandinavia (even though it was a few degrees warmer a thousand years ago) should suggest one would’ve worn longer dresses than thigh length ones.
One theory is that Hyrrokin was drawn with an extra short serk to highlight her indecency. Many other stone images (e.g. Bota in Garde parish, Tjängvide in Alskogs parish and Tängelgårda in Lärbro parish, all on the island Gotland, Sweden) from the Viking Age show both women and men wearing knee and ankle length garments. If this length is applicable to both over and under garments one might speculate. I’ve chosen to make my Haithabu serk ankle length to provide as even layers as possible between my under and over garments in order to make them resilient against the Swedish weather conditions; to isolate and keep the cool inside during warmer days and to retain the heat during colder days.
Most dress patterns that apply to recreating Viking clothing are based on findings from all over Scandinavia. Serk fragments from Pskov, Køstrup and Adwick-le-Street are linen. Fragments from Birka have been linen mixed with wool, making the wool fragments difficult to identify as part of the serk or part of a e.g. smokkr or dress. However, garments found at Haithabu show that pieces had been cut in shapes that closely follow the body instead of the geometrical shapes found at other grave sites. To be able to dress and move in such narrow garments, they’ve must consisted of both slits and gores.
Since the small amount of serk fragments can’t make up for a full pattern, most re-enactors turn to other findings for inspiration. The Skjoldehamn tunic and the Kragelund tunic are examples of reference material I’ve used. The Skjoldehamn tunic was found in a bog in Andoya, Norway back in 1936 and has been re-examined and re-evaluated several times until Dan Halvard Løvlid’s radio carbon dating has appreciated the tunic to be from about 995 – 1090 BC. The earliest examination, just a year after the discovery, was conducted by Gutorm Gjessing who made the following line drawing of a possible reconstruction of the tunic (Gjessing 1938; 38).
The Kragelund tunic was discovered in 1898 and has been dated to circa 1100 BC. The tunic has two gores in the front, two in the back and two on each side; all of them gathered into the points. The front and back gores can have been left split (which would be logical considering the thigh length of the garment) but no seams are preserved well enough to tell. The neck is pointed in back and front, creating a V-shaped slit about 7 cm long.
Since I wanted my serk to be closed in the front, to cover the tattoo on my chest, I chose to split the front part in two and use the seam allowance to create a keyhole neckline which I can close with a small clasp. What has been found in Birka, though, are linen fragments pinned to the back of round brooches which indicates that the brooches have been used to close a slit connected to a keyhole neckline.
I made the serk ankle long to prevent it from dragging in mud and dirt but still protect my legs from the outer wool garments. I reduced the gores into four slimmer ones to prevent the serk from becoming too wide. The arms are full length and narrowed towards the wrists to hold them in place above the hands. The seams and hem are felled with a half bleeched linen thread which I purposely left visible on the right side to really give the impression of a hand made garment. The small lining part inside the neck is attached to the seamy side with small overcast seams that are nearly invisible on the right side.
Although serks are generally left un-decorated I choose to embroider the hem-line and edges with a darker brown linen yarn in a fishbone pattern based on the Mammen embroideries.
I’ve also used the image of Hyrrokin on the Hunnestad Monument as a reference to the relatively slim rectangular arms and straight armholes. To allow movement I added square gussets, as suggested by Hägg based on the Birka findings.
Gutorm Gjessing – Skjoldehamndrakten, en Senmiddelaldersk Nordnorsk Mannsdrakt, published in Viking, Tidsskrift for Norrøn Arkeologi. #2, (pp. 27-81) 1938
Margareta Nockert – Bockstenmannen, Och Hans Dräkt. Halmstad och Varberg, 1985
Per Holck – Myrfunnet fra Skjoldehamn — Mannlig Same Eller Norrøn Kvinne?, published in Viking, Tidsskrift for Norrøn Arkeologi. #51, (pp.109-115) 1988
Inga Hägg – Die Textilfunde aus dem Hafen von Haithabu, K. Wachholtz, 1984