17 October 2017

Double-Faced Designs on a 7-Block Grid with Radial Symmetry

© 2000 Carolyn Priest-Dorman

The last time I wove some motifs from the 13th century Scottish seal tag described in Audrey Henshall’s “Five Tablet Woven Seal Tags,” I was struck by some design properties they all shared. The motifs are 28 tablets wide; they were all designed in blocks of four tablets in a turning sequence of seven sets of two. This can be represented in shorthand fashion on a square grid consisting of seven rows of seven squares each. Further, all the Scottish motifs shared radial symmetry, and all the lines were orthogonal rather than the more typical diagonal lines of tablet woven square motifs (e.g., the Snartemo V motifs). The orthogonal lines show up very neatly due to the warp being entirely threaded in one direction rather than alternating from tablet to tablet.

Recently, I was looking for something in Brigitta Schmedding’s Mittelalterliche Textilien in Kirchen und Klöstern der Schweiz when I rediscovered several other textile examples of square motifs with radial symmetry and orthogonal lines. There is something about these motifs that looks extremely ancient to me! With the TWIST sample exchange in mind, I began doodling to test the limitations of the design. Since the sample exchange was supposed to be for relatively narrow bands, I kept my designs to the proportions of the original and used a 49-square grid. Soon I think I will have to try reproducing some of those elaborate Swiss designs from the fourteenth and fifteenth centuries; it will take a grid of at least 15x15 blocks.

On a seven-by-seven grid using radial symmetry, it is fairly hard to avoid repeating the Scottish designs or designing out-and-out swastikas, but I did my best. They’re far too basic to be “original,” of course, although I wasn’t looking at any other sources while I designed them. Here are the four designs I liked the best; I call them the Whirlpool, the Cross-and-Boss, the Running T, and the Well.

The motifs need to be squared carefully (my samples are not all carefully squared, due to unexpected time constraints). Also, my experience weaving these types of motifs shows that using smaller yarns minimizes the effect of the feathery weftwise color changes. They’re especially breathtaking in Size A sewing silk, which makes them pretty much identical in size to the Scottish piece of my original inspiration..


Henshall, Audrey. "Five Tablet-Woven Seal-Tags." Archaeological Journal 121 (1964), pp. 154-62.

Schmedding, Brigitta. Mittelalterliche Textilien in Kirchen und Klöstern der Schweiz. Bern: Schriftern der Abegg-Stiftung, 1978.

* * * * * * * * * *

This document was originally prepared to accompany my work in a tablet weaving sample exchange in Spring 2000 sponsored by TWIST (Tablet Weavers' International Studies and Techniques). It is also available at my old website which is currently frozen. (I am considering my options for recreating and updating that site.)

The original samples were woven of 20/2 Czech linen.

12 October 2017

Viking Age Tablet Weaving: Kufic or Not?

Please read the important update at the bottom of this post.

Well, here on the Left Coast of the USA I woke up to a Viking textile controversy this morning, and no mistake! Maybe I'm an idiot, and maybe I'm not, but here's my take on it.

Today's controversy involves the Birka tablet-woven brocade bands. According to Annika Larsson (about whom more below), the geometric designs actually depict Kufic inscriptions saying "Allah" and "Ali." We already know there was plenty of contact with Muslim culture of the period given the large quantity of Persian silver and Eastern silk found in Viking Age contexts. That contact is not at issue, not even remotely. But this particular reinterpretation of the bands has me really steamed, and here's why.

Larsson's "discovery" is predicated on unfounded extensions of pattern, not on existing pattern.

If you consult any of the current crop of articles about this topic, you'll see a photo of a graph on a page with a mirror next to it. (Here's a link to the Heritage Daily one.) The pattern graph is quite clearly Band 6 from the Birka finds, an artifact found in a tenth-century woman's grave, Grave 965, which was published in Agnes Geijer's 1938 Birka III: Die Textilfunde aus den Gräbern. (See photo below, which is taken from Abb. 20, "Muster der Brettchenbänder," on page 82 in Geijer.)

If you look at the pattern Larsson is postulating, it shows nine additional pattern units at each side of the band, for a total of 18 additional tablets' worth of width. In Larsson's photo you can tell the additional pattern units apart from the original pattern units printed in Geijer because the additional units indicating the brocade weft at the two sides of the graph are slightly lighter than the ones in the central part of the graph; they are also printed 90 degrees off from the direction of the original unit graphics. This unexplained extrapolation practically doubles the width of the band, and here's why that's a problem.

According to Geijer, Band 6 was woven in a technique common to almost every piece of tablet-woven brocade at Birka. Each pattern tablet was threaded half with silk, half with linen and offset by one-quarter turn from the tablet next to it. "Stave borders" of warp twining one tablet wide marked off the selvedges; they are threaded entirely with silk. The tablets were alternately threaded and turned continuously forward. The band was woven with a structural weft that is hidden inside the band as well as a supplementary metallic brocading weft that floats on top of the band to make the pattern. When linen is "up" during the weaving, it's always covered with metallic brocading weft; when silk is "up," it's often (but not always) visible as a tie-down point. This technique is very economical, as the resulting band looks like it's woven with 100% silk when it is much less expensive to weave than 100% silk as about half the warp is linen rather than silk.

If you consult Tafl 17:1 in Birka III for a photo of Band 6, you can clearly see the continuous metallic weft of the band turning at each selvedge to enter back in the other direction. If Larsson were correct that Band 6 was originally significantly wider, you would not see those turning loops; you'd see a series of discontinous single passes of brocading weft with cut or broken ends at each edge.

Annika Larsson is currently with the Institute for Archaeology and Ancient History at Uppsala University in Sweden. The bio at her web page there says "[t]he first degree I took at the Textile Institute, where I trained myself as product designer and pattern designer for the industry. But I am also a craftsman and collaborate with various craft crafts in textiles and throws." She is not trained in archaeology as nearly as I can tell.

You may remember how she offered a radically disruptive theory in 2007 that Viking Age women's oval brooches were worn at the level of the nipples. (You can read more about her idea here.) I have yet to read a textile archaeologist's endorsement of that particular version of costume history, which doesn't surprise me since on the face of it it seems incompatible with much of the published archaeological material on women's clothing from Birka, let alone the rest of the Viking cultural milieu.

Again, I have nothing against the theory that these patterns are Kufic. I would welcome additional overt evidence for Persian influence at Birka, since I already believe the mix of cultures there is too rich and thorough to gloss Birka as representing a single and "Viking" cultural context. But Larsson's theory flies in the face of what we know about Band 6; it doesn't pass my sniff test.

UPDATE 17 October 2017: I see the Guardian has called me a "textile archaeologist." This is incorrect. I have never claimed to be a textile archaeologist, although the discipline of textile archaeology is my chief intellectual interest. Please do not hold me responsible for anybody's failure to describe me correctly.

UPDATE 1 November 2017: The Enköping Museum has put out a statement about this controversy.

The criticism is directed towards a part of the exhibition, but not to the whole. In anticipation of clarifications from the researcher at Uppsala University, the museum now chooses to pause this part of the exhibition until further notice. [via Google Translate]

The statement further makes it clear that the issue centers on interpretation of a Birka band, not a band from some other site.

19 July 2017

Know Your Vintage Knitting Yarns

Earlier today Carrie Schutrick made me aware of the University of Southampton's digital project, Knitting Reference Library, which has been added to the Internet Archive. This project brings some of the late knitting historian Richard Rutt's personal library to the Internet, particularly several nineteenth century works representing the earliest knitting books in English.

This sparked some discussion about knitting vintage patterns and using vintage yarns, which led me to scrape together the useful references I've squirreled away for identifying and approximating vintage yarns. I'm posting that information here as well, so I'll be able to find it again easily when I want it.

Here are a few links to sources for information on the exact sizes of various vanished yarns mentioned in vintage patterns.*

Vintage Yarn Wiki -- There's a Vintage Yarn Wiki!

Yarn and Thread Conversions -- at SandyJ's blog for exploring Mrs. Beeton's Book of Needlework

Misc. Discontinued Yarns -- "a list of older yarns, arranged by weight of yarn, by brand name, with fiber content and yardage (where available)."

Vintage Yarns -- "historical yarn names and modern gauge/needle size equivalents. Also some suggestions on possible modern yarns." Mostly from the period before the 1930s. SCA readers may recognize our own Countess Ianthé's fair hand here.

Discontinued Yarn Chart -- " a guide only to help you find comparable yarns today to substitute for yarn specified in vintage patterns." This list is itself "vintage," having been first compiled in 1965.

*Today's posting is brought to you by the letter V.