Change view: Change style to alternate Change style to default

Tubular Tablet Weaving

Identifying and Reconstructing the Hanging Cord from a 14th Century Embroidered Purse.


There's a pretty little 14th century purse in the Victoria and Albert Museum #8313-1863. Although its German brick stitch embroidery usually captures the most attention, the purse also happens to include a carrying cord and several loop braided drawstrings with tassels.

Linen embroidered in silk  German Bag, 14th century
Linen embroidered in silk, with couched metal thread
14.5 cm x 12.5 cm
Museum number 8313-1863
Victoria and Albert Museum, London

I had photos of this purse, but hadn't paid much attention to it until Timothy Mitchell (of German embroidery fame) wrote to a tablet weaving mailing list asking if anyone knew how the carrying cord was made. It looked potentially complicated for tablet weaving, as it had three rows of three colors traveling diagonally around the tubular band and reversing direction at irregular intervals. As it turns out, the nature of tubular tablet weaving itself made the most complex looking part of the cord's design (the spiral) into the simplest part.



Because tubular tablet weaving tends to spiral on its axis, the weaver only has to control the direction of the spiral rather than design it into the pattern itself. As Audrey Henshall noted in her article1 about a similar cord, the band will spiral if the direction of twisting in the warp cords is the same as the direction of twisting of the weft thread as it spirals its own way up the band.

While playing with this concept, I realized that the way the weft spiraled was influenced by two things in combination - the direction (right to left, or left to right) I passed the weft through the shed and whether I returned the weft (and connected the selvedges) over the band or under it.

I made a chart to help me keep track of the possible combinations and their effects. Assuming the weaving and cards are in front of you, and the unwoven warp stretches away from you beyond them:

For example: Z threaded cards turned forward (or S threaded cards turned backward), with the weft inserted from right to left and passing under the band, will create a Z twisting spiral to the band.

There are two options for reversing the direction of the spirals. You can either turn the cards backward or flip the entire band upside down. [Shortly after I sent off this article, I realized I'd left out the third option of flipping each card on its axis. But I guess that was just too boring to contemplate. - CM]

If you opt to turn the cards backward, don't bring the weft across the band after the last turn of the cards. Instead pass it back through the shed to the opposite side as you would if you were weaving something flat. Continue weaving by inserting the weft from this new direction, making sure to return either over or under the band as you were doing before. The most likely error is to forget to change the direction (left or right) the weft passes and only change the direction the cards turn. If this happens, the band will have sections that spiral alternating with sections that do not.

An alternative way to switch the spiraling direction is to simply flip the band over on its axis. This effectively changes the threading from S to Z, so you can reverse the twist by continuing to turn the cards forward. Because everything is upside down you can also continue to pass the weft through the shed in the same direction as before, but now must take care to make sure the tube curls in the same direction as before. The most likely error with this method is to forget (or not to notice) that the band is upside down and that the weft thread now needs to travel under the band where perhaps before it traveled over it. I'd dearly love to find an extant medieval tubular band where the weaver didn't notice this happened, and ended up turning the tube inside out.

Personally, I prefer the second method. Always turning the cards forward is easier to remember, and passing the weft from right to left is more natural for me because I'm right handed. Then all I have to pay attention to is whether I'm connecting my tube on the top or the bottom - and that's a matter of looking at it to see which way it's already curling. I've also noticed that the band just wants to flip over if the cards aren't kept under constant control.

I'd love to know which reversal method was used in a given textile, but without a tell-tale error I don't think it's possible to know. I like to consider all my weaving errors as gifts to the future.

Here's how to duplicate the carrying cord on the little German purse that started this investigation: Take 9 cards (4 holes each) and thread 3 cards in yellow, 3 in white, and 3 in red (or whatever the colors really are). Flip the cards until they're all oriented the same way - All S threaded, or all Z threaded. (An alternating SZSZ pattern wouldn't work out evenly when formed into a tube, and would resist the spiraling discussed earlier.)

Work tubular card weaving as discussed, pulling the weft tight to draw the selvedges together into a tube. In theory, this would create three colored stripes, each three cords wide, up the axis of the band. In practice, the structure of the band twists, and the vertical stripes end up spiraling around the band, duplicating the spirals apparent on the original.

Reverse the direction of the spirals several times, by whichever method you prefer, in accordance with the extant cord.

I'm not the first person to have looked at this particular item and determined it to be tubular tablet weaving, but hopefully this little article will be of use.


HENSHALL, Audrey "Five tablet-woven seal tags" in Archaeological Journal 121, 1964. pp. 154-162

wymarc logo A Stitch Out of Time
Timothy Mitchell (Master Richard Wymarc) has studied and re-created the above bag, and has full details on his website.