I got a wonderful present from my son who was visiting from NYC. It was Sanzo Wada's book called "A Dictionary of Color Combinations".

A Dictionary of Color Combinations by Sanzo Wada

It it a lovely, small book derived from the original 6-volume set compiled between 1933 and 1934. I immediately wanted to weave something using some of the color combinations in the book. I thought of double weave as possibly the best way to isolate the colors and see them in juxtaposition with the other colors, but I wanted to do towels or napkins both because I always gravitate toward useful textiles, but also because I have the widest range of colors in 8/2 cotton and cottolin in my stash.

I started thinking about Polychrome Summer and Winter. I had woven it once before as a sample for a Meta Weave class.

Close up of Summer & Winter Polychrome border with 2 different threading variations

I wove it as a border for a towel using 16/2 cotton (warp and tabby) with 8/2 as pattern weft. I kept the body of the towel plain.

It worked well as a border because it is a dense weave and not particularly drape-y. It is also a slow, 3-shuttle (min) weave. Being a glutton for punishment, I decided to see if I could make a series of napkins with over-all patterning and using many different color combinations from the book. I developed a graphic motif from a modern Finnish design I had seen online.

Graphic for napkin design

I marked the blocks and created a draft.

Graphic for napkin design with blocks marked

I used a standard Summer & Winter threading, but I knew that the trick with polychrome is to make the pair of pattern picks usually deployed in S&W behave a little differently. 

Drawdown of standard Summer & Winter design with tabby in orange and using the alternating treadling

In standard S&W (alternating style) the pattern picks in a block alternate between tie-down shaft 1 and tie-down shaft 2. They are typically not alternating in color and are tied up to appear as "pattern" (weft floats appearing on the surface of the cloth) in some threading blocks and "background" (weft floats appear on the back of the fabric) in other threading blocks. 

But from what I understood about Polychrome, the patten picks can be tied up on opposites - if you want pattern pick 1 to appear on the surface in some threading blocks and pattern pick 2 to appear on the surface in the opposite blocks – thereby achieving the polychrome effect of getting multiple colors to appear across the warp "at the same time" ish.

Drawdown of Summer and Winter Polychrome using my first tie-up

You can see in the polychrome version there are 2 colors used as pattern and the tabby is in the following order: tabby A, pattern color 1, pattern color 2, tabby B, pattern color 1, pattern color 2, etc. In standard S&W pattern picks alternate with tabby picks: tabby A, pattern, tabby B, pattern, etc. 

This was my understanding of how Polychrome worked, but I did do a little research. I discovered that different books arranged the tie-downs differently—and none of them explained why. While I found scads of info on S&W variations, Polychrome info seemed sparse. And a lot of what there was focused on Taquete or other "no tabby" variations that are used for heavy rugs and wall hangings. I really wanted to figure out how to deploy this neat trick for a napkin or towel. So I jumped in as I usually do by winding an 8-yard warp:))

I figured that a fine warp relatively loosely sett would be a good starting point for attempting to get a lighter-weight result. I had some white 20/2 cotton from Lunatic Fringe on my shelf, so that is what I used. It is about 8,400 ypp so I decided to sett it at 30 epi. I drafted it like this (I changed the tabby in the drawdown to mint green so you could see it better), and it looked pretty good theoretically.

Drawdown of my graphic woven using the tie-up in the close-up below

Close-up of my first napkin drawdown

But there were some surprises when I wove it. The beat wasn't even close to the expected 60 ppi, and I had to use a ribbon template to place my color and block changes correctly because my predicted pick count was way off. I was getting about 45 ppi. I cut the first napkin off, and washed it. The colors were more muted than expected, and there was some strange weft shifting in the solid areas where there are block changes (look at the charcoal area), but the hand is actually pretty nice for a napkin.

First napkin off the loom

Once I had woven a few napkins in different color combinations (initially thinking that the muted colors were the color interactions I was choosing rather than something with the structure) I started to think about the beat problem. I had a suspicion about what the problem was and I had an idea about how to fix it. However, I was worried that making this correction/adjustment will get the beat closer to normal (twice the sett or 60 ppi) and get the colors to pop more will also result in a denser and less napkin-y napkin.

In order to test my theory, I went back to the books to see if my suspicions were correct, and yes they were. It was the way I had arranged my tie-downs. My tie-downs were tied up in a traditional alternating sequence (see tie-up above. But in many of the drafts in the books they are tied up like this:

The corrected tie-up

with each pattern treadle pair using the same tie-down and then alternating with the next pair, i.e. tie-down shaft 1 for treadles 3,4; tie-down shaft 2 for treadles 5,6 and so on. I had missed this!! I also looked back at the Meta Weave towel draft in Polychrome and saw that I had  my tie-downs arranged in Dukagang order. 

Summer and Winter Polychrome for Meta towel using Dukagang style for the tie-downs

Both of these arrangements will have longer warp floats, and if I am correct in my hypothesis, will allow the pattern weft picks to move together as more of a unit for more vibrant color, closer to expected ppi and probably a denser cloth. I also noticed that in my Meta Weave draft I didn't arrange the tie-ups as opposites. And I remember why now. For that design, I needed to have some background areas (areas where neither of the colors were on the surface) to create the spaces between the graphic "threads". So as you can see in the drawdown, Threading block A has neither the pink or the green pattern weft on the surface. The down side of this arrangement is that the back isn't simply a reverse of the front the way it is in the "opposites" arrangement, but something else entirely. 

I have re-tied my loom, and I will weave the next napkin in the revised tie-down arrangement. I will report my findings in the next blog post: Summer & Winter Polychrome Rabbit Hole Continued.