Polychrome Echo Weave on Eight

When Alice Schlein presented her "Echo Weave" in Weaver's 32, it was the polychrome effects she achieved on 16 shafts that really caught my attention: two or more warp colors and the weft color, each predominating in distinct regions. (On eight shafts, she only got two colors to show.) The polychrome effect was made even more appealing to me by an enigma in the history of network drafting. The original 1938 network drafting monograph (Brandon and Guiguet's Méthode des Initiales: Un aspect mathématique du tissage à lames) has attached to its cover a color fabric illustration, showing three distinct color regions. So far as I can tell, the monograph itself gives no hint how this fabric was woven.

The challenge

Inspired by Schlein's article, I wove a polychrome scarf on sixteen shafts, and was proud enough of it to send a snapshot to Madelyn van der Hoogt, the editor of Weaver's. She replied that she liked it enough that it might well run in the scarf issue. She cautioned, however, that it might not, because she was trying to strictly limit the number of 16-shaft projects "and it'd be hard to reduce yours to fewer shafts or offer alternatives to those who have fewer shafts." That was very nearly the perfect prod to cause a structure/pattern weaver like me to come up with the eight-shaft version. (It would have been a sure thing had she used the word "impossible" -- but even "hard" did the trick.)

Before explaining my solution to this challenge, I'll slip in two more general remarks. First, regarding nomenclature, Bobbie Irwin points out in Weaver's 33 that the name "Echo Weave" was already taken. I therefore nominate "corkscrewed networked twill" as an alternative -- making my subject the ungainly "eight-shaft polychrome corkscrewed networked twill." Second, the back of these polychrome fabrics (whether 16-shaft or eight) is quite different from the front. There are still regions where one or the other warp color predominates (the opposite one, of course). But where the front has the weft predominating, the back has a blend of the two warp colors. Thus in a scarf, or any project where both sides can show, it is possible to get two quite different color effects. In return for this bonus, you need to pick warp colors that blend interestingly. I chose copper and turquoise, which produce a corroded copper effect on the side where they blend. (Admittedly this blend is not quite so nice as it was on the 16-shaft version.) Given this choice of warp colors, I used deep purple for the weft to provide a third dimension. (A photo should appear in Weaver's 44 that is better than the snapshot I have.)

Drafting the basis twill

Before I could echo a twill, I needed a twill to echo. Thus my first step was to design an ordinary eight-shaft networked twill. There was a hitch, however, caused by the interaction of three factors. First, my design called for a telescoped snail pattern. Second, my idea for how to make a polychrome echo weave on eight shafts required that I use a three-shaft initial. Third, Schlein's book says that only non-telescoped turtle patterns are possible if the the number of shafts isn't a multiple of the initial height. (See her Network Drafting: An Introduction for the nomenclature and explanations.)

Since I couldn't do a telescoped snail if I followed the conventional rules, I needed to make up some new rules. Note that these don't really have anything to do with polychrome echo weave per se, but rather with gaining additional freedom in drafting eight-shaft networked twills with three-end straight initials. If you want to use polychrome echo weave with a non-telescoped turtle, you can skip ahead to the next section.

My first new rule is that you initially plot your design on a large network, and then telescope it to the available number of shafts. (The conventional approach is to telescope first and network afterward.) If you analyze the threading draft in Figure 1, you'll see that the two ends on shaft 1 near the left edge aren't really on the network -- they would need to be on shaft 3 or 6. The explanation is that they were on shaft 9 of the original large network and got telescoped to shaft 1. This trick is not limited to such minor telescoping, and can be used with turtles and fleas as well as snails. As Figure 1 demonstrates, you can do this with ordinary one-warp networked twills, not just echo weave, provided you use a suitable tie-up. However, the threading is less flexible than those that rigidly stick to the network. You can't weave the entire width of the warp as continuous 1/2 or 2/1 twill, for example. If you try, you'll get overly long floats. For my purposes, however, this style of threading works fine.

The second new rule is that to use a snail pattern, the length of the draft shouldn't be a multiple of the initial width. Instead, the pattern should stop mid-initial so that the last end winds up on the last shaft after telescoping. The example draft is only 50 ends long -- not a multiple of three -- so that it can end on shaft 8. Again, this can be used for normal one-warp networked twills but risks extended floats (at repeat boundaries). As before, this isn't a problem with the tie-up shown.

Dovetailing the echoes

To weave the threading from the previous section as an ordinary echoless networked twill, I'd use a 2/1/1/2/1/1 tie up, as shown in Figure 1. However, that tie-up doesn't provide enough of a range of interlacements to produce a good result if interleaved with an offset version. So, for the echo weave I used a 2/3/1/2 tie up instead. As you can see in Figure 2, this would be completely impractical with a single warp, as there would be some very long weft floats. However, when the two warps are interleaved, each ties down the other's long weft floats. You can see the result in Figure 3. The threading from Figure 2 has been interleaved with an echo version that is offset by three shafts. For example, the echo of shaft 4 is shaft 7. This echoing wraps around, or, to say it differently, is telescoped. For example, the echo of shaft 7 is shaft 2. (We add the offset of 3 to 7 and get a nominal shaft number of 10. Since 10 exceeds 8, we subtract 8 and get shaft 2.)

Project directions

This scarf, based on the draft in Figure 3, is a snap to weave because the treadling is a straightforward advancing twill and only a single shuttle is used. The finished size is 12-1/4" x 67". Given the light, supple hand of the fabric, this size is perfect for bundling your whole head up in.

8-shaft loom, 14" wide; 15-dent reed; 1 shuttle.

Warp and weft: 18/2 Zephyr wool/silk (5040 yds/lb, JaggerSpun), 2 oz each turquoise and copper for warp, 2 oz deep purple for weft.

Wind a warp of 420 ends 3 yds long alternating copper and turquoise.

Sley two ends/dent (one of each color) in 15-dent reed, 30 epi; center for 14".

Thread using draft in Figure 3 as follows from right to left. Start with a copper floating selvedge. Then thread the leftmost 18 ends of the of the draft, followed by four full repeats of the draft. Finish with another floating selvedge -- it should be turquoise.

Weave 16 ppi using the tie-up and treading shown in Figure 3. After first couple inches are woven, hemstitch. Weave to 72" length and hemstitch again.

Finish by cutting from the loom and tying fringe in extra long bunches of ends with overhand knots to protect during washing. Soak in cold water with shampoo. Rinse, spin, and tumble until damp. Iron dry. Trim fringe to 2-3/4", cutting off knots in process.

This article by Max Hailperin is copyright © 1999 by Max Hailperin. For copying and re-distribution permissions and definitive copy policy, see http://www.gustavus.edu/~max/weaving/policy.html. The definitive copy is http://www.gustavus.edu/~max/weaving/scarfart/. This is revision of 1999/06/29 19:52:35.