I was sick all last week, so the last knitting of the swirl went very slowly, but it finally got done. I employed lots and lots of safety pins to the one seam so I could finally try it on and see how it actually fit. How it actually fit was okay, but not great, and what was the most frustrating was that it was impossible to tell until I got this far (ie: nothing but the finishing to do).
In Gulliver's Travels, Jonathan Swift has Gulliver encounter a society called the Laputians, who like to do everything by the "scientific" method (whether or not their methods are in fact scientific is up for some debate). One of their tailors measures Gulliver's thumb in order to ascertain his size for a new suit of clothes. While the thumb size does enable the tailor to create a suit which Gulliver can put on, the new clothes are hilariously ill-fitted, not having taken into account the different quirks of his body which were not reflected in his thumb. The thumb measurement wasn't a bad start for determining Gulliver's overall frame size (bigger thumb, bigger man and vice versa), but would not have revealed, say, his waist size.
The swirl patterns suffer from much the same issue. The pattern book sizes swirls by yoke size, that is, the measurement from one armpit, across the back of the neck, to the other armpit. Lots of different body sizes can have the same yoke size, since it doesn't take into account things like bust or waist measurements.
And that's the problem: yoke size isn't the only measurement where sizing matters. Yoke size determines whether or not one will fit into a swirl, but not how the fabric emanating from the yoke will cover the rest of the wearer's body. The swirl above fits in an A-line on me reminiscent of a cutaway 19th-century riding jacket, which sounds good except that the fabric flares out past my body at the lower edge and just sort of sticks out into space at the back, while providing insufficient coverage in the front unless I stretch the fronts and hold them in place with a pin. Not terribly flattering. The fabric hand has too much density to drape, and not enough ease to flow. It's a cliche, but it does not look like it does in the photos, even taking the much-thinner model's body into account.
You might be thinking, "Well duh, you should have checked your gauge!". And so I did, but as you can see from the photo, the entire garment is knitted in a rib-type fabric that stretches a lot horizontally. This is supposed to be a good thing, so that gravity can create an A-line while keeping the same-number-of-rows collar narrower. It means that different parts of the garment have different effective row gauges, and while the knitter can logic out which parts will have a stretched/hung gauge and which will have an unstretched gauge, the ribby fabric and the weight of the garment itself make it hard to say just how stretched or unstretched things will get in wearing.
In the preface to the book, designer Cat Bordhi praises the Knit, Swirl! author Sandra McIver for her precision, mentioning that during the development of the swirl, McIver kept spreadsheets of fabric stretch ratios. When I was trying to figure out what was going on with my swirl — that is, why it wasn't swirling — I came across that note and it gave me a chill. McIver mentioned that the yarn she used for this particular pattern (Karabella 8) has a tendency to "grow". My yarn is Patons Classic Wool, and as you might guess from a name like that, it's a very versatile yarn that's on the springy side. The pattern just calls for worsted weight yarn that will meet the gauge. Interestingly, the manufacturers' descriptions for the two different yarns are very similar, but the fabric hand I got compared to how the sample looks in the book photos are poles apart.
Meanwhile, the armhole depth is low, very low, and no measurement/adjustment info was given in the pattern. The lower edge of the armhole is nearly at my waist. The sleeves are almost leg-o'-mutton, very generous to about the elbow and then stretched flat over the forearm. The entire torso shifts out of place when I raise an arm. I'd rather the standard armhole depth for set-in sleeves: low enough to be comfortable over a shirt, but not so low that the sides of the garment get pulled up when the wearer moves their arms.
The sleeve length was the only place where the pattern provides instructions on how to lengthen (though not shorten), but then no base sleeve length is given — only the cuff-to-cuff measurements across the back. To get an idea of sleeve length, the knitter would have to subtract off the cross-back measurements and then divide the remaining measurement in half. It's an odd measurement to omit, since the actual cross-back measurement in the pattern is pretty fixed by size. I didn't make any changes, and the length is fine, but there was no good way to tell that prior to the knitting-up. I could do some back-of-the-envelope math from the gauge and cross my fingers, but that was about it.
The only solution provided in the book for fixing the diameter/draping issues I mentioned is to go up a needle size and make a larger swirl, but, as the book goes on to warn, this increases all of the measurements, including that armhole depth that's way too generous. So I'd have to work in a bigger needle size and take a gamble with one of the other patterns in the book, one that has a higher armhole depth and bigger torso diameter. As I said before, both those dimensions are never given.
The book does a lot of hand-waving about fit. The garments are versatile, different parts of the jacket are "fitted" or "slightly tapered" or "flow around the body" (except they don't). It's maddening since A-line jackets tend to look good on me, if they close without visible strain at the fronts and have a flattering sleeve shape. The swirl fails here. And it would be all right if it failed, except there was no way to know until I'd knitted the whole thing. I've adjusted just about every sweater pattern I've ever knitted, going right back to lengthening the sleeves on the very first cardigan I knitted when I was thirteen, but because of the way swirls are constructed, a lot of the familiar adjustment-points are moved or missing. It would be nice to have them noted to give the knitter a fighting chance, instead of having to blindly follow a pattern which may be leading them down the wrong path.
I keep thinking that these swirls would make a lot more sense if they were knitted in reverse. Make the sleeves, then make the upper back. Cast on for the collar, and start working in the round to create the collar/lower back. Work two extra rounds for a border and then cast off when the desired diameter is reached.
I think I just talked myself into frogging this one, which is a pity, because I still like the intended shape, if not the shape I got. I still want to try other patterns from the book. It's just I'd much rather I had more control over the outcome before I knitted it. The way the book works, you're knitting blind far too much of the time. I'm going to have to reverse-engineer a few swirls before I can make one with any confidence. I'm not sure it's worth the bother.