The Interior of My Brain: A Knitting and Fiber Arts Blog

When I previously wrote about using the machine knitting method Diana Sullivan calls “sew as you go”, which joins a new piece of knitting to the edge stitch of a previously knitted piece, I had made three variations on a sweater formula I’d developed and a pair of socks with a wavy shape built into the panels. The three sweaters gave me numbers and a working method for constructing this kind of garment, and the socks were a swatch for the next generation of paneled sweaters that were starting to take shape in my mind.

‘Twas the season to knit birthday sweaters for my daughters, whose birthdays are three weeks apart at the end of the fall. Prima’s sweater was the crocheted heart sweater and knitted kittens hat that I described here; Secunda’s sweater was going to be a machine-knit paneled garment that contrasted the warm and cool colors of my plant-dyed yarns in a geometric arrangement, using the increase/decrease shaping that I had tested out in the socks. I was envisioning a rectangular-shaped garment with a triangle in the center of the front piece, reaching its peak at the base of the neckline and flanked by two side panels, one mostly yellow and the other mostly orange, that increased at the same rate that the triangle decreased, to form the rectangular shape of the front of the garment. The back would be knitted in two panels joined at the center, with stripes of the cool colors alternating with stripes of the warm colors in an offset arrangement between the two sides.

When I devised the numbers for the size and shape of the garment, I referred back to my notes for previous sweaters I have made for Secunda to get a general idea of how many stitches I would need for width and how many rows for length, then I looked for the simplest possible numbers to achieve the slope of the triangle and the placement of the neckline within the parameters of the width and length I needed for fit. Secunda was satisfied with the fit of sweaters knitted in similar yarn that were 160 stitches wide and about 270 rows long. In order to calculate the shaping of the triangle and the placement of the armholes and the neckline, I took 10 rows for the folded hem out of the count, for 260 rows. For the sake of my sanity, the rate of increase/decrease needed to be the same number of rows throughout the slope, and three rows worked the best.

Here’s my thinking: starting from 160 stitches at the base of the triangle, I wanted to reduce to 6 stitches at the start of the neckline, and I needed enough rows left over for the depth of the neckline, for a total number of rows long enough to cover my daughter’s torso. Since I was decreasing a stitch on either side of every decrease row, I divided the number of stitches by two (80 stitches down to 3 stitches per side) and subtracted 3 from 80 to figure the number of times I would need to decrease from the base of the triangle to its top, which was 77 times. For the number needed to create the slope of the triangle, I needed to multiply 77 by a number that would end the triangle shaping at a good place for the neckline on a 260-row garment. It didn’t take a lot of math to calculate that 3 was the only option, since 77 x 2 would end the triangle at 154 rows, which would put the base of the neckline somewhere on the tummy, and 77 x 4 would end the triangle 48 rows past the total garment length of 260. But a multiple of 3 would end the shaping at 231 rows, for a nice, comfortable neckline that was 29 rows deep.

The center triangle went fairly smoothly, if not 100% accurately. 77 decreases, every three rows, is a lot of numbers to keep track of. For the side panels, I was so focused on the shaping at the triangle edge of the front piece that I made a miscalculation at the armpit that changed the design. I was modeling the shape of this garment on a template I developed a few years ago for my daughter, with slightly dropped shoulders set into a square armhole. The numbers were for 160 stitches for each side, binding off 20 stitches at the base of each armhole, front and back. Somehow I tangled my numbers and didn’t notice that I had 10 stitches more than I was expecting when I got up to the shoulder shaping. Instead of binding off 20 stitches at the base of the armhole, I had only bound off 10. But nope, I was not frogging the stripes I had knitted between the bound-off stitches and the start of the shoulder shaping. Knitting stockinette on the machine is a lot faster than in hand-knitting, but frogging and fixing mistakes is a whole lot slower. I didn’t even have to summon the Fiber Goddess to decide this was just going to have to be a new design feature.

The back piece was numerically much simpler than the front. I divided the piece into two equal panels, 80 stitches each. The color group chunks would be slightly varying lengths, mirrored by the opposite color group for the same length as the other side. On the first side, I played with putting a lot of contrast into my color group chunks, because I had done some of that in the triangle panels of the front and liked the effect. But when I looked at the first panel, I decided I had overdone it because the colors were reading as individual stripes rather than as components of a color group chunk. I hoped I could achieve a course-correction on the second side, where I would put more stress on keeping the components of the color chunks closer together in value, so that they read as color blocks and could move the eye to similar colors on the opposite side and give a more unified effect to the whole. This tactic was somewhat helpful, and if it wasn’t completely successful, stripes are nice and the colors are pretty. I made the sleeves in panels, with the cool colors on the front, flanking the warm colors of the side panels, and the warm colors are on the back side of the sleeves.

For my color use, I focused on the blues, greens, yellows, oranges, and browns that I have dyed over the years, mostly using onion skins, the soaking water of black rice and black beans, coreopsis, phragmites, common beggar’s tick, Jerusalem artichoke, spiderwort, walnuts and their leaves, and rudbeckia, all of which I can forage in my neighborhood. I have been using my plant-dyed yarn in knitted garments for several years, and it seems to me that the dyes from these plants hold up to wear, sunlight, and time better than other plants I’ve experimented with. However, I have learned that all of these foraged dyes will change with time, and I have come to avoid wearing my garments outside on sunny days, although my husband still goes outside for walks in the striped sweater I made him a few years, and it still looks great, if less vivid than it did when it was fresh off the needles. I have given family members sweaters made of yarn I dyed with foraged materials, and I always tell them that the colors will change with use, but they were made to be used, so they should use them as they want to and accept the changes.

I finished Secunda’s birthday sweater in time to mail it to her in time for her birthday, and– very importantly– in time to have a photo shoot with Samantha, a super-cute knitting friend who agreed to model for me. I was really happy to see how well it fit Sam.

I haven’t had a chance to photograph Secunda to my satisfaction in her birthday sweater, but she did send me a photo that gives me some good information about the fit. She seems to rattle around in it a bit, but comfortable and relaxed is a look I personally appreciate.

When it came to making a version of this shaped-panel sweater for myself using the same yarn, I took comfortable and relaxed to the extreme. This one would be the expansion of the socks with the wavy panel that I had made as a feasibility study, with panels of contrasting colors that increased when the adjoining panel decreased to form a zigzag line down the middle. I borrowed the general size, shape, and numbers from the template I developed for the first three paneled sweaters I made (two of which now belong to my sisters in Philadelphia), but minus the side panels because I second-guessed the way they make the garment flare out when viewed from the side. I thought it might make me look fat, which just shows how I have been indoctrinated into a foolish and destructive social construct. Well, anyway… the total number of stitches for the body of the first three paneled sweaters was 432, so I rounded up a bit. The new design was to consist of four panels, two front and two back, each 110 stitches, for a total of 440. Each panel would be knit straight up to the shoulder shaping so that the joining with the already-knitted piece would end at the armpit of the dropped shoulder.

My original plan for the socks was for the panels to be zigzags rather than waves, but the small number of stitches and rows for socks meant that I only had room enough to do five increase/decrease sets over five stitches and 10 rows, which cut off the zigzagging before it could really get started. But 20 increase/decrease sets over 40 rows would give enough space between direction changes so that the join between the panels would read as a zigzag line. A full zigzag, to the right and then left to get back to the starting point, would be 80 rows. Three full zigzags would be 240. That’s long on my ever-shortening body, but it was already oversized at 440 stitches around, and the numbers were simple. I’m all for letting the simplicity of the numbers determine the design decisions. In that spirit, I decided the neck shaping would be 20 rows deep over 30 stitches per side, which left 80 stitches for shoulder shaping. I like the slope I get from short-rowing the shoulder at 8-stitch intervals, so that gave me 10 shoulder slope decreases over 20 rows. The 30-stitch neck shaping could be done in 10 3-stitch short rows on the alternate rows. Hooray for mathematical simplicity!

I wanted the two sides of the front to be symmetrical, with each side of the zigzag going 20 stitches from the center on either side of the front halves. Therefore, the center of the front piece started the first zigzag at the midway point of the diagonal line, 10 stitches in. So the direction changes were at row 20, row 60, row 100, row 140, row 180, row 220, and ending at row 240. That seemed self-evident enough that I didn’t bother writing it down and checking my direction changes for accuracy as I was knitting the piece. What could possibly go wrong? Cue the ominous music.

I was using most of the same colors I had used in Secunda’s sweater, but this time I tried to make mostly smooth gradients, with a few places where the stripes were contrastive with their neighbors. It’s hard to see how the composition is growing on a knitting machine, because the right side faces away from you when it’s on the needles and you can’t hold it up against your body, but I tried to make sure that the darkest colors of the two color groups weren’t placed next to each other. I wanted the eye to travel diagonally over the piece rather than be drawn to one part of my body and get stuck there.

I did the blue/green side first, and when I finished with the yellow/orange side, I took the completed piece off the machine and saw what I had missed when I examined the piece after finishing the blue/green side: the numbers for the direction changes weren’t as self-evident as I thought. Exactly at the halfway point of the 240 rows in the zigzag, at row 120, I had reversed the direction of the shaping, decreasing when I should have kept on increasing, which threw off the zigzag for the rest of the piece. And the fatal error was on the first piece, where the joining with the second piece was buried under the new knitting.

This meant that frogging the bad knitting would be no simple matter of rip-it, rip-it, but rather detangling the entire strand of blue/green yarn from the multiple places at every join where the yellow/orange yarn wound around the blue/green yarn that it was connecting to. That’s a lot of yarn barf to detangle. Fortunately, detangling yarn barf is one of my superpowers.

When I finished detangling each length of frogged yarn, I wound it up in a little ball and stuck it on a circular needle in the order of usage so that I could reconstruct the color arrangement when I reknitted.

Then I Russian-joined all of the little balls together when I rewound them into the cake that machine-knitting requires.

It took about six hours of work to get back to where I thought I was when I took the completed front piece off the machine.

The back was based on the same idea as the back of my daughter’s sweater, but this time the color blocks were all the same number of rows, five blocks per side for 260 rows, which came to 52 rows each. I arranged the color blocks in fairly close gradients and reused the same color arrangements on the opposite side, more or less in reverse order. This time I wanted there to be no mistaking the offset checkerboard pattern. For the sleeves, I divided them into separate halves of yellow/orange and blue/green and placed them so that the blue/green sleeve half was next to the yellow/orange zigzag on the front and the yellow/orange half was next to the blue/green zigzag, and placed the opposite color grouping on the other side of the sleeve.

Done at last! I showed my Ravelry group a modeled photo of the front and back, and my co-mod gave me the most gratifying feedback:

It’s gorgeous! Putting the join at the center front and giving it the shape of a zigzag makes the design so much different in effect from your daughter’s. It’s almost shamanistic – I’m expecting that a wise woman with extra powers would need such a design – as though you are looking in two directions at once or are able to see between this world and another one. The back of the sweater shows a different kind of dichotomy, and the coloration enhances the idea of landscape, as Johanna has noticed.

Shamanistic! A wise woman with extra powers to see between the worlds! Moi??? And here I thought my superpowers were the ability to detangle yarn and the fortitude to repair mistakes.

But wait, there’s more! Melissa has been hosting a trunk show with a local dyer, Avalon Springs Yarn, and there were two colorways that caught my eye, one a solid Schiaparelli shocking pink and the other a multi of reds and pinks. Immediately I knew how to put them together. I would strip down the idea of the zigzag pullover to its basic numbers, no shaping, no fancy stuff, just pink on one side and red multi on the other and let the colors be the design feature. It would be quick and easy for a change, what a concept!

Unfortunately, quick and easy brings the tendencies to complacency and inattention, which are fatal in machine knitting. In machine knitting, mistakes happen in a split second. Fixing them takes hours. I think I spent as much time frogging as knitting. First picking my knitting up off the floor and back on the needles because I forgot to shut the yarn feeder gate, just as if I was a beginning machine knitter. Then forgetting to stop joining the pieces together when it was time to keep them separate for the dropped-shoulder sleeves. Then going on autopilot and shaping the back of the neck like the front neck that I’d just made. I was spacing out, cursing, fixing the error, and spacing out again. Rinse and repeat. I asked my daughters if they thought I was losing my marbles. Prima would want to reassure me, even if it wasn’t true, but Secunda would give it to me straight, and they both said my marbles were still intact. Well, maybe, but the repeated dumb mistakes were unnerving.

I am temperamentally unable to allow vast swaths of stockinette in solid colors remain pristinely undecorated, and I added pockets to each side of the front in the opposite color. I’m seeing the plain knitting as a canvas for finally fulfilling a neglected 2021 goal. One of last year’s Swatchathon experiments was embroidery on knitting, and it never happened during 2021, at least not yet. Now the 2022 Swatchathon is almost upon us, but in final week of 2021, there will be embroidery on this knitted item.

Here it is, unadorned, and it’s a nice wearable garment exactly as it is.

But if there’s a lily to be gilded, bring on the gilt!