This document is provided as is without any express or implied warranties. While every effort has been taken to ensure the accuracy of the information contained, the author assumes no responsibility for errors or omissions, or for damages resulting from the use of the information contained herein.
Permission is granted to make and distribute verbatim copies of this document for non-commercial private research purposes provided the copyright notice and this permission notice are preserved on all copies. Website mirroring is permitted by express prior arrangement. Permission will only be granted if the document is posted in its entirety and the content and format of the document remain completely unaltered by the mirroring site. The definitive version of this document can be found at http://www.shelaghlewins.com.
Copyright © 2008 Shelagh Lewins.
Grace Crowfoot gave a detailed description of the fragment of tablet-weaving found in St John's Cricket Field, Cambridge (Figure 1). She dated it as being Saxon, late pagan, which would be around the fifth or sixth century A.D. The fibres appear to be bast and the dimensions and cross markings of the fibres indicate that they may be linen.[1]
Figure 1: The Cambridge fragment.
The fragment is still attached to a broken piece of bronze that strongly resembles anglo-saxon belt tongues found in nearby sites (Figure 2). The tongues are simply made from two plates held together with a single rivet. My reproduction is riveted to a similar tongue that was made by Alister Perrott (Figure 3). This has performed very well in use and shown no signs of fraying at the end. Because no buckle was found with the fragment, I made a loop at the other end by wrapping the ends together with silk. This hasn't been so sturdy and I may try rewrapping it at some point.
The weave structure gives the band several unusual properties. Both sides of the band are identical, although the pattern is offset on the reverse. The band is very thick and firm. It makes an excellent belt but would be less suitable for use as garment trim because of its stiffness.
Figure 2: A complete bronze strap end from Fakenham.
The weaving technique is unusual in that the tablets are threaded differently to each other and the pattern is not created by changing the turning direction. In each pick, every other tablet is turned. The only other band known in this exact technique is mediaeval and has a chevron pattern (it's apparently in Norwich Museum, or was when Crowfoot wrote in 1950). There is some suggestion that this band may also be mediaeval but the find was originally identified as Anglo-Saxon and is generally regarded as such.
I have examined the original band, which is held at the Museum of Archaeology and Anthropology in Cambridge. It's not on display but the curator kindly got it out of store for me. It's about 11 mm in width and it is possible to see the colours of the linen. Although linen apparently often takes on a blue or green hue from contact with metal, I think the threads must have been dyed because the colour clearly follows the pattern. The inner diamond is a paler blue: the border is very pale and neutral in colour, and the background is a darker blue. It is very unusual to see evidence of dyed linen but I don't see how else the colours could follow the pattern of the weave so exactly. I used the most similar colours I could find in my reproduction, which is linen and about 10 mm in width (very slightly narrower than the original).
Figure 3: Reproduction of Anglo-saxon tablet-woven belt. Click for a larger picture.
Figure 4 shows how to thread the tablets. The numbers along the top identify the tablets: the letters signify the holes in the tablets. The first three tablets should be threaded with four dark threads, the fourth with four light and so on.
The 19th tablet is shown in Crowfoot's drawing of the restored weave but not in the threading diagram, and the weaving instructions ignore it. I usually include it for symmetry. For reasons I haven't fully understood, the 19th tablet (if used) has to be handled differently to the others when weaving.
I thread the centre tablet, number 10, first. The remaining tablets can be threaded in pairs using a cut-down continuous warp method. Tablets 9 and 11 can be dropped of on either side of tablet 10, and then tablets 8 and 12 can be added and so on. This halves the threading time and ensures even tension. It's actually easier if the holes in your tablets are not labelled: just thread the right colours in each tablet, you can arrange them correctly after they are threaded.
Once the tablets have been threaded, flip them to be alternately S and Z threaded. Figure 4 shows which way to orient each tablet.
Next turn each pattern tablet (6 to 14) so that the colours are correctly oriented. Looking at the left-hand side of the pack of tablets, the holes are labeled clockwise:
A B
D C
Figure 4: Threading and arrangement of tablets.
Weave by turning the odd-numbered tablets a quarter turn forwards and throwing the weft. The second pick is woven by turning the even-numbered tablets a quarter turn and throwing the weft. Repeat these two picks. It's a bit strange but if everything is set up correctly, the diamond pattern will appear!
Tablet 19 must be counted as an even tablet rather than an odd tablet because otherwise the w eft does not catch it.
I find it easiest to slide tablets 1, 3, 5 etc. forward to make a separate pack just separated from the remaining pack that contains tablets 2, 4, 6 etc. Leave tablet 19 with the even pack. It is then possible to turn the odd pack, beat and pass the weft, and next turn the even pack, beat and pass the weft.
As is usual with a warp-twined band, the warp threads will gradually become twisted. At the half-way point, it is possible to reverse the weaving schedule and undo the twist by the following procedure.
Again, this sounds cryptic but does work. The pattern continues and the warp twist gradually unwinds.