• Owaine ApDyfnwal

Viking Shield Project 2020/21

Around 14 years ago, as a group, we decided to try our hand at making a couple of planked shields utilising the knowledge available at the time and to a standard that we considered close to the originals. This was within an environment of the UK re-enactment arena that was almost exclusively using plywood blanks as a base. Using limewood for the planks and facing them with cow hide, this example with a black, white, and red design (an interpretation of the Isle of Man find) was the final result.

Fast forward to 2019 and I had started looking at all the new information coming from both experimental archaeologists and also from re-evaluations of the original finds. With this broader spread of knowledge, I embarked on gathering the material to make a revised version of the classic shield. Adam and Stuart had recently finished their reproductions of an insular shield type from Ballinaby, so with the advice of Adam from these experiments I set about gathering and constructing a new one. I first bought (rather than made this time) a shield blank from Steve at ‘The Longships Return’ (https://www.thelongshipsreturn.co.uk/). The planks were sawn, mostly tangentially, so this was a small compromise, and made from limewood. Whilst this wood is synonymous with shields, and is a very suitable wood type to use, it isn't actually the most common for the early medieval period as a whole, although there are examples from England in the period prior to the viking age. However, I was reasonably confident that is was a suitable and appropriate wood given what was available to me given it shares many properites with the poplar, willow, birch, scandinavian pine, and other woods used during the viking age.

The board was a little over 800mm diameter, and the planks were 10mm thick, however I wanted to try to alter the profile to match those of the originals found at both Gokstad and Trelleborg. To this end the outer edges were taken down to approximately 6mm whilst leaving the tops and bottom thicker to support the handle. I expected this to also improve both the weight and the handling of the shield in use. I also smoothed out the area around the hand hole so that it was more comfortable to use.

With the blank ready to use, I started looking at coverings. I wanted to use rawhide and managed to purchase some deer rawhides of appropriate size from Barrhead leather to cover both the front and rear. One was a little too small but with a little patching on the edge, it was usable. After a soaking in the bath for 24 hours, I then left them to dry out under weights for a couple of weeks so that they flattened out.













Then, when they were still malleable but reasonably dry, they were cut roughly to size. In future I wouldn’t cut the hand hole this early next time as it limited the manoeuvrability of the skins on the facings).













Next I used hot animal glue to attach them to the wood. This was quite tricky as you had to work reasonably fast to smooth out airbubbles before the glue cooled too much. They were then left to dry under some heavy weights for a couple of days.




















The first attempt at this was not a success, either I’d left the glue to go off for too long due to working a too large area, or the hides were still too wet. Adam suspected I had not used enough glue and worked it too slowly. Whichever it was when I released the clamps and weights, there was no adhesion and the hides just fell straight off. After some head scratching, I tried again but with lots more glue and working across the shield face in small areas. This time I left them for a couple of weeks under pressure and they seemed to have set solidly.

The handle was made from a piece of ash wood. At first this was twice the size of the final version, but I kept working it down until it felt right. In reality, it could still do with being a third smaller than it is now. Also, in hindsight, I wish I had trusted the strength and integrity of the finished shield and done a much shorter handle in the style of the Trelleborg find as this would have been more than adequate and would also have been a little more unique when it comes to interpretations of early medieval shields, where people seem to feel full length handles are nessecary for structural integrity, despite plentifull evidence contra to this from the broader early medieval corpus.

When deciding on a boss, I wanted something forged and a good representation of a 9th century type, such as those from Gokstad. In the end I ended up with something of a hybrid between a Rygh type 562 and 564, similar to the example from Grave 4 at Cumwhitton in Cumbria, which is within the area covered by our group, so it worked perfectly. Jason at Wieland Forge (https://www.facebook.com/wielandforge/) is ever helpful with any work of this kind and provided the finished item.


The next stage was the rawhide edging. I knew from the previous experiments that this needed to be approximately 10-15cm shorter than the size of the rim itself to enable it to shrink during drying and bind edge tightly. Too tight and it can warp the shield itself, too loose and it doesn’t shrink on correctly. Trying to get the length of rawhide needed in a consistent size and quality can be difficult, especially when buying on-line. The first attempt didn’t go well, the hide had hidden flaws in and it split when I tried to stretch it around the rim. I possibly soaked it for too long too and didn't allow it to dry enough after, which didn’t help.

After waiting for a couple of weeks for more to arrive from Medieval Craft in Poland (https://medievalcraft.eu/), I tried again. The ‘stretch’ on the new batch was more than expected and I ended up cutting it down 4 times and re-stitching it with linen thread before it sat well. Once fitted it was left to dry for a few days, this seems to have turned out OK











The shield had also maintained its shape and had not warped too much, so all in all I was reasonably happy. The handle and boss were attached with forged nails, another purchase from the guys at Medieval Craft.


The final stage of the construction was to stitch the rawhide to the boards around the rim. This was something that we hadn’t done on the first version many years ago as the hide had stretched to seal itself without the need, but this time I wanted to give it a go. At first, I tried drilling the holes but soon realised that using an awl gave a much better finish and enabled the stitching to glide through with a lot more ease. The whole thing is stitched using artificial sinew.

Once completed, we looked at the best finish and decided on a single colour. I wanted to try an authentic paint and we researched into making egg tempera paint. We also tried an oil-based version, both using charcoal as the pigment. The paint was mixed with on a slab with a glass muller.

The samples shown are; top - egg tempera made with the yolk, middle – egg tempera made with white and bottom – oil based made with Danish oil (a form of polymerised linseed oil in this case, as it was to hand). By far the best finish came from the yolk-based paint. Whether this was due to more mixing or just luck with the quantities, I don’t know but this was what I decided to finish the shield with. One yolk covered about a quarter of a shield in the mix I used.

I’m quite pleased with the finished article, it’s taken over a year to complete but worth the time. As Adam and I have discussed on many occasions, we are just apprentices at this, if I had another 20 of these to make, each one would get better. So, we can only assume that the originals were far more refined than we are producing today.


There’s a few things I’d do differently next time too, as well as the mistakes with the glue and edging and I’ve already mentioned the shorter handle. I think I would use shorter nails on the boss when its not going through the handle as the ones I’ve used are too long, I’d also thin the planks even more to reduce the weight. There is also some more work to be done with the paints, as this is not something I had tried before, and the final version is flaking more that I'd like, so there is some more experimenting to be done there with the mix, and surface preparation.

Will I ever fight with this? Not a chance! My thanks, as always, go out to Adam for all the conversations and advice. The work and mistakes are all mine but it would not have even got started without his help.

Mike/Anlaf