I saw it once on Facebook in a machine-knitting group and then never saw it again, but it burned a hole in my imagination: a yoked all-over entrelac pullover in gradients of green and pink. It sparked my year of machine-knit entrelac, which occupied my knitting machine for a lot of 2020. I mentioned this pink-dominant entrelac work in my Ravelry group, and as usual, my brain trust provided me with links to videos that showed the method. One of them was a YouTube video in which the machine-knitting goddess Diana Sullivan demonstrated how to make an entrelac cover for a round cushion, with large entrelac diamonds decreasing to small ones that gathered up in a star shape in the center. Diana used scrap yarn to hold the live stitches on each entrelac rectangle, and it looked like a confusing mess to me. I wasn’t ready for that yet.

My other machine-knitting goddess Susan Guagliumi had a method for building entrelac without scrap yarn, which kept all of the stitches of the piece on the needles all of the time, and moved them over with a transfer tool in groups toward the rectangle being worked and joining each stitch of the already-knitted rectangle with the one in progress, one at a time. I used this method for the neck bands of ponchos I made for my sisters, but tears and blood were shed in the process. The method stressed my yarn and caused it to shred and break, so I had to wax it constantly as I worked. Also Susan’s method produces a flat rectangular fabric whose component rectangles are all the same size, whereas I wanted to make an unseamed yoked garment in the round shaped with expanding or contracting rectangles. Diana’s cushion cover showed me that it could be done. My google searches to find that video again led me to Diana’s website, where I found and bought her book and DVD set “Wear Your Diamonds.”

It’s a given that I am never, ever, ever going to make anything that my machine-knitting goddesses provide instructions for, exactly the way they make it. For one thing, I only have a standard-gauge machine and they usually are using mid-gauge or bulky machines, but even if their numbers applied to standard-gauge machines, I wouldn’t make what they make because I am bloody-minded and have a very different aesthetic than my idols have. But all the information I needed was in Diana’s “Wear Your Diamonds”, including information I didn’t know I needed. Aside from demonstrating the mechanics of forming interlocking rectangles with live stitches secured by waste yarn, Diana also explained ways to reconcile the significant discrepancy in gauge between entrelac and plain knitting, and she provided a useful ratio for that discrepancy, somewhere in the 60% range. Diana’s material contained instructions for a pullover with a plain stockinette body and an entrelac yoke, and instructions for a practice project, a hat that started with ribbing and then entrelac diamonds in decreasing sizes to shape the crown.

I decided that it would be prudent to start off with a hat, but I don’t like ribbing and Diana didn’t provide numbers for standard gauge machines, so I recalculated. Diana’s hat sidestepped the need to decrease stitches (a more complicated matter on knitting machines than it is for hand knitting) by starting with a 1×1 rib knitted at a larger gauge than the entrelac. That made it possible to use the same number of stitches that was already on the machine without having to carry out maneuvers to decrease that number, which she explained later when she got to the instructions for the sweater.

I didn’t want to do a ribbing for my hat, since I don’t like ribbing and am still a little traumatized by my early ribber experiences four years ago, so I e-wrapped stitches to make a rolled edge for 10 plain rows and then followed Diana’s procedure for reconciling the difference in gauge between stockinette and entrelac. I applied the ratio she used in her sweater recipe to make transitional triangles from stockinette to entrelac, which was something like 17 stitches for the start of the triangles and ultimately reducing it to 11 or 12 stitches. This reduction came at the end of the process. She made short, wide base triangles using all of the stitches inherited from the stockinette section, then did the decreases for the entrelac after the triangles were formed, before knitting the scrap yarn that holds the stitches of the completed triangle in place and waiting for their turn to be incorporated into the next layer of entrelac rectangles. Diana makes it look easy. I dropped all kind of stitches every which way when moving the doubled stitches onto the end needle to build the next layer. I stuck stitch markers into the dropped stitches to keep my fabric from dissolving into nothingness and hoped that practice would eventually make perfect.

Completed hat, pre-blocking and ends-eradication, but after repairing all the dropped stitches

I finished the hat with a good grasp of the mechanics of changing the size of entrelac rectangles, and the hat served as a gauge swatch that enabled me to calculate the numbers for my sweater. I decided that the body would be 20 11-stitch rectangles around and that I wouldn’t do any shaping until I got to the yoke. That way I could procrastinate on making any decisions until I had a better grasp on the amount of space the rectangles were going to take vertically and horizontally as the garment got bigger and heavier and stretched out under the weight of the superwash sock yarn I was using. I did a stash-dive and dug out an array of green-themed yarns and pink/purple-themed yarns and arranged them in a rough gradient. Then I cast on stitches for the rolled hem at the same ratio as the hem of the hat. I hoped I would get better at not dropping stitches in the decreases of the transitional base triangles over the course of 20 of them. I didn’t. My bottom layer was littered with green, orange, and pink stitch markers holding those dropped stitches in place until I could repair them at the end.

My idea for the color progression was to start with dark, moody colors and gradually lighten and brighten the gradient until I had spring greens and flowery pinks in the yoke. I had a bit of a problem in my planning because I didn’t know precisely what my gauge would be in the end. The hat wasn’t the guide I needed in order to project the size a 400-gram garment would end up being, because years of experience with Wollmeise have taught me that the gauge changes pretty drastically after about 250 grams. It’s superwash, and that’s what superwash does, it stretches. Not knowing exactly what size I was going to end up with, I decided that a longer body was better than too short a body, so I kept adding different dark multis to my sequence of colors. I wanted to create an overall woven effect that drew the eye in a diagonal direction rather than side-to-side in stripes, but I kept sabotaging that goal with color choices that had less in common with their predecessors than they appeared to have when they were sitting next to each other in cakes. Then I found I was stuck in turgid colors and everything I did to break out from the turgid and into bright and happy ended up also looking turgid. It was a conundrum. At some point I realized that adding dark gloomy color after dark gloomy color had earned me a tunic rather than a sweater.

I was feeling at this point as if I was not going to pull off this color experiment

As I got closer to the armpits, I had to figure out how I was going to join the future sleeves to the body. It would have been nice if I had had a pattern or some explanation of how it was done, but I didn’t, so it was just me and my little brain. I visualized various ways to join the rectangles of the sleeve with the rectangles of the body, or tried to visualize, but I decided that the visualization would go better when I had real-life pieces of knitting to fit together. So I postponed decisions. I finished the final body row the same way I had done every other row, and started the sleeves.

Diana’s method for changing the gauge for stockinette to the gauge for entrelac produces an invisible transition, as the decreased stitches are hidden in the join between the base triangles and the first row of rectangles, but it caused me angst whose metric was the number of stitch markers dangling from the base triangles row, preventing the many dropped stitches from unraveling the fabric. But what if I did the decreases when I started the base triangles, so that they were incorporated into the fabric of the entrelac from the very start, rather than doubling up stitches to be transferred from the waste yarn to the new rectangles and relying on my shaky ability to do it accurately? I was using the same rolled hem at the cuff as I did at the bottom hem of the garment, but this time I loosened the tension from 7.1 to 10. After measuring the width of the already-knitted rectangles of the body, I figured that six 11-stitch rectangles would give me a comfortable fit in the upper arm. I referred to the rectangles that shaped the crown of the hat to get a clue about how large the smallest rectangles should be, and decided that six stitches ought to do it. So: six triangles, six stitches each, 36 stitches. Even at the loose tension of 10, I was going to need more stitches at the cuff. I counted out 36 stitches on the hem of the body, which I had knitted at 7.1, then I counted out 48 stitches, and measured it. That was going to be a nice roomy cuff if I knitted it on tension 10.

Each triangle started with eight stitches. My memory is that I decreased the first stitch and the last stitch of each triangle before knitting the first row, but the photograph suggests that maybe I decreased the first stitch before knitting the first row and then decreased the last stitch after knitting that row. In fact, I might have forgotten to do it once after the first row and did it on the second row. I’m sorry, I should have written down exactly what I did. Anyway, it seems not to have made any difference in the result. There was also that rather alarming-looking gap between the triangles where the needles of the two decreased stitches were put out of work. That too made no difference in the end, because that extra length of yarn between the triangles got absorbed in the stitches I picked up to make the following layer of rectangles. So move along, folks, nothing to see here. It’s still worthwhile for me to learn how to do Diana’s method without dropping stitches, because my method wouldn’t work well if there were more stitches in the base triangles. If the ratio between entrelac stitches and stockinette stitches is much more than about 6:8, as it was here, the stockinette would pull in. In fact, the reason why it didn’t pull in here is that I loosened the tension of the stockinette.

Mind the gap! Or don’t mind the gap. The two out-of-work needles between the base triangles produced a gap that looked as if it might be a problem, but the elasticity of knitting absorbed the loose yarn between the triangles
See? No gaps!

I didn’t attempt to match the sleeves with the body. For one thing, I didn’t know how many entrelac sequences I would need to get the sleeve length I wanted, although I suppose I could have calculated it if I had wanted to, so I didn’t know if I would get the sleeve dimensions I wanted by replicating the colors of the body in rectangles that grew from small to large. But the main reason for doing something completely different was that I was really frustrated with the colors I had used previously. I felt trapped in this color logic in which everything had a relationship with what came before but wasn’t getting me to where I wanted to be, and that moment when everything came together hadn’t happened and might not ever happen. So I ditched the dark colors and started the sleeves with the lightest, brightest colors I had, which I had been saving for later as if they were dessert. The effect on my mood, and on the mood of the composition, was immediate and drastic. Placing the sleeves next to the body was like going out of a dark sealed-up house into the sunshine. It made my eyes hurt. In a good way. Finally! I had broken that dark spell.

This makes me think of the way log cabin quilts use drastic differences in light-dark value to emphasize and to change the emphasis of the shapes the parts of the pattern make

I still had hopes that I could build a gradient that was gradual enough to create the woven effect I was seeking, but even when I took care to choose what I thought were closely related colors and colorways, the entrelac seemed to emphasize the differences rather than the similarities, so I continued to have a horizontal striped effect. Well, all right, I like horizontal stripes. Now, at least, the colors were the bright contrast to the murky colors of the body that I wanted. I had been building the color sequences in the body toward lighter and brighter colors, and I now was building the color sequences in the sleeves so that they could meet up at the yoke join at pretty much the same place in terms of hue and intensity, coming from the opposite direction.

While I was knitting the body, I had started visualizing how to join the sleeves to the body at the armpits. What was needed was entrelac rectangles connecting the body with the sleeves, so the body and the sleeve would share these bridging rectangles, one half of the rectangle (or rectangles, I used two) on the body side and the other half on the sleeve side. The more efficient way to do it would have been to have left two rectangles unknitted at the right side of the body and two on the left side, and then filled in those holes during the final sequence of the sleeves by knitting the rectangles in the gap on the body and joining them with the stitches of the sleeve at the armpit. But I didn’t do it that way, because when I was at that stage of the body, I wasn’t yet ready to commit to doing that. I needed the pieces I was going to fit into each other to be physical realities rather than theoretical concepts so that I didn’t put the sleeves in the wrong part of the body or some other theory-crashes-into-reality miscalculation. So I knitted the final row of the sleeve with two fewer rectangles to form the hole into which the two bridging rectangles from the body would be sewn in. Sewing in the rectangles took a bit of extra time and effort, but the join at the armpit is indistinguishable from any other part of the fabric.

The green squares at the top of the sleeve (at the left) are the final sequence before the join to the body. The pink square that the arrow is pointing to is one of the two empty spaces into which green squares on the body at the armpit fit into
Here’s how the two pieces fit into each other
See the armpit seam? You don’t? Exactly!

Now, finally, the yoke! I had been working on this tunic, with interruptions, since September, and now it was January. I had to divide my attention between this everlasting project and January Swatchathon, but I had finally gotten this far and I was going to get all the way eventually, assuming I didn’t drop dead first. The motivator was that now I had reached the climax of the piece, the light bright happy colors that would ignite the dark dreary totalitarianism-in-America colors. Yes, my mind did go there, to that political place, when I tried to figure out why I had been so unable get out of the murky colors I kept using during that terrifying 2020 autumn. But first I had to figure out what to do about that final dark red sequence up at the top of the body, which threatened to draw a big dark red line right at the bust. What had I been thinking? Usually I have reasons for the choices I make, but this was a random non sequitur. I decided to wait to the end to see if I had a situation that needed correction, but I had a correction: to knit squares in a color that was in between the pink below and the purplish-pink above and sew them on top of the existing layer. I did one as an experiment and it looked perfect. But it was a whole lot of fiddly trouble to do it, and it might not be necessary, depending on how deep the yoke ended up being and where the stripe hit my body when the tunic was finished and blocked.

Spoiler alert: I didn’t cover up the dark pink row with a more muted pink, although I kept the rectangle I experimented with. See the stitch markers? Those are just a few of the stitches I dropped when I was trying to apply Diana’s method

I didn’t know ahead of time what size and shape would result from decreasing the size of the rectangles in each layer of the yoke, by one stitch per triangle. I was starting the yoke with 11-stitch rectangles, and I reckoned that the final triangles probably couldn’t be smaller than 5 stitches. Big or small, the number of entrelac units was fixed: 24 rectangles. That meant there would be at least six layers to get from 11 stitches down to 5 per unit. I knitted the first yoke sequence at 11 stitches per rectangle, and held the garment up against my shoulders. Its circumference looked enormous. I knitted the next sequence at 10 stitches per rectangle. Still enormous. I knitted the next two sequences at nine stitches per rectangle, and it was still pretty darn big around, but I was starting to have the length in the yoke that I needed for a comfortable fit, so I felt safe reducing the rectangles of each successive layer, down to five stitches at the light green row at the top. The pink finishing triangles were also five stitches each, and I short-rowed and scrapped off each triangle to keep the stitches live. My 24 5-stitch triangles gave me 120 stitches, which I put onto a circular hand-knitting needle to hand-knit short rows to shape and raise the back of the neck and to do a couple of decrease rows to complete the yoke shaping.

I had spent months, the entire duration of the long-drawn-out process of knitting the body and then the sleeves, trying to visualize how to make elongated rectangles at the back of the yoke that would fill the function of short-rowed neck shaping, but I couldn’t figure out how I would then decrease the size of the rectangles of the following row in order to shape the yoke. Maybe there’s a way to carry out these apparently mutually exclusive operations, but I couldn’t find it, so I punted. I waited until the very end to shape the neck and complete the yoke. At my level of machine-knitting skill, hand-knitting is much more nimble and efficient for decreasing in the round than machine-knitting, which requires moving all the stitches on the needle bed in order to do decreases on stitches in the middle of the array of needles and seaming the fabric knitted on a single bed. Theoretically, one can knit in the round by using the ribber, but I don’t know how to do that. But hand-knitting is my native language, so I used light, bright colors, pinks in stockinette and lines of garter stitch in greens to delineate the short-row shaping segments and to hide the final decreases that drew in the yoke and avoided draftiness of a too-open neck. Then I put half of the stitches back on the machine to knit a rolled edge, cast off (too tightly and had to redo it by hand), and did the same on the other half of the stitches. After I had a neck edge that would fit over my head, I seamed the two halves together, not invisibly and not accurately placed, but I’ll live with it.

When I created the light, bright color arrangement at the bottom of the sleeve, I placed the sleeve next to the body where the wrist would hang. That part of the body was where I had placed the darkest, murkiest of my color combinations, and the contrast between sleeve and body was literally stunning– I at least was literally stunned. The effect, for me, was like those randomly illuminated faces in a Rembrandt painting. Chiaroscuro was the organizing visual element of this piece, and in its current form the counterparts to Rembrandt’s illuminated faces were at the wrists and upper yoke of the garment. It needed one more spotlighted point: at wrist level on the body, in the form of a patch pocket of the lightest, brightest entrelac squares planted right in the middle of the darkest, murkiest entrelac squares. I had a bouquet of 1-ounce skeins of sock yarn dyed to gradate from one to the next, and I used four of them for the green rows. The pinks gradated from pinkish-orange to orangish-pink to salmon pink. The edging color was the same color as the edges of the hem, cuffs, and neck. This little square achieved what I had tried and failed to achieve anywhere in the garment, a gradient that was close enough to create a woven effect that made the eye move in a diagonal direction. Finally, after all this knitting, I had a 6.5″ square that looked like the image in my mind’s eye.

Compare and contrast, before and after the pocket.

As for the garment that inspired this project, I managed to find a picture of it a day or two ago, after this tunic was completely done. It was kind of like running into someone you used to have a crush on, after you’ve moved on and are no longer interested, and you discover that the person has none of the attributes that had made your heart go pitter-patter. The sweater was knitted from a variegated yarn that had a lot of hot pink as well as other colors that mostly did not include green and contained no gradients, and it resembled my tunic… almost not at all. Funny how memory works. That’s why I try to avoid looking at anything that inspires my work after I start working on my own project.

5 thoughts on “Chiaroscuro

  1. Amazing is an understatement – you awe me. The science, the engineering, the art, the creativity, the originality, the perseverance, the COLORS, the ingenuity, the process, the documentation of the process – and of course, the resulting creation.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.