… and now it turns out there’s a mistake in the Big Sack Sweater in the Stitch ‘n BitchM book.
The Sack was based on Banff, at least in the sense that it was intended to be an oversized sweater to wear on days when you needed to be warm and cozy. To that end, it was designed with the same philosophy, in only two sizes: big and cozy (49″), and bigger and cozier (57″). Banff is in those two sizes as well, but besides the raglan styling, the two sweaters are quite different in the hem, collar, and sleeve styles, not to mention the central cable panel.
Apparently, the editors didn’t agree with the two-sizes-fits-most thinking. They deleted the large size, and replaced it with three other sizes, making the small only 42″. The instructions were adjusted accordingly, and without my knowledge. That’s fine, I assigned the copyright, and if that’s what they wanted to do, that was their business. But when they came to the adjustments for the front neck, they missed a teensy bit of logic behind the original instructions.
See, when you knit cables, it takes more stitches to cover the same width than it would in stockinette. So, at the beginning (hem) and the end (neckline) of the cables, the instructions have you increase, then decrease, four stitches to prevent flaring in the hem and collar. Those decreases were spaced out at the center front neck, since you don’t want to bunch all the decreases all together. However, when they wrote the instructions for the smallest size, they were too clever in phrasing the instructions: they try to get you to pass a stitch on the right-hand needle over a ssk you just made, when that stitch on the right-hand needle doesn’t actually exist.
So when you get to this point in the instructions, don’t fret about them; just make sure that as you bind off the central 14 stitches in the smallest size, that you decrease 4 times as you go, spacing them out by one or two stitches. Then continue on to the other side of the neck.
The moral is: technical editors should edit. If they start redrafting, they’re not editing, and that’s not good.
[Comments: I explained that the pattern in SnB came to be in this way: "I submitted the design ideas, they were approved, I knitted up the samples and sent in the patterns (the sample Sack sweater is in what is the second-largest size in the book). The technical editor asked me one question to confirm some instructions in the LBT pattern, but that was it. Perhaps they didn't want to bother me by asking me to review their proposed substantial changes to the patterns. I didn't see what happened until the book was published."
Megan raised a question about the diagram dimensions vs. the instructions. The text says to knit 12(13,13,14) inches in stockinette for the back, but the diagram is labeled 11(12,12,13). Since I had given away my copy of the book in a fit of pique, I suggested that the discrepancy was due to the rolled hem. You knit for 12 inches, but when the hem rolls it won't be more than 11".
And then I reported that the medium size of the Big Sack is flawed, too. It seems that the raglan seam of the back is longer than the raglan seam on the sleeve. That's not good. And the schematics of the sleeve suggest that the width of the very top of the sleeve will vary by size; that's not true, because the pattern has you bind off 9 stitches no matter what. I reported this problem to the editor, but I told her that I would not be providing corrections to those errata, because I felt it should be left to the technical editor who created the problem.
Then Reverend Kate asked whether there'd be any issues in replacing the original cable with the DNA cable, since they have the same stitch count, and whether she could just work the same 4 decreases at the neckline to correct cable splay. I started writing a piece on cable substitution for publication later, as a result of this question, but quickly responded that:
"The replacement cable itself should be fine, but in general when you want to sub cables, it's not a question of matching stitch count--it's a question of whether the cable takes up the same physical width as the original. That can be determined by swatching.
That being said, on such a loose-fitting sweater, any discrepancy probably won't be problematic.
Now, the decreasing bit, you're right. Just decrease the four stitches (or whatever) along the cable panel, and that's good. The whole reason for those decreases is to correct for flare when cables are begun/ended. If you feel like you need to increase/decrease more or fewer stitches at the beginning/end, that's just fine."
And then Tanya asked for suggestions for modifying the pattern at the raglan shaping, because her stitch gauge is on but her row gauge is off (she has fewer rows per inch). That question happened to dovetail nicely with my planned article for Knitty's spring 2005 issue, which I work on periodically. Now I know what to use as an example (Banff, not Big Sack). I responded:
"As for the vertical gauge/armhole question: You've got stitch gauge and the decreases on the armholes are more-or-less consistent, so the easiest solution is to (1) calculate how many rows *you* need to make the intended armhole depth; (2) redistribute the decreases so that you accomplish the correct number of decreases (i.e. what the pattern says, because your stitch gauge is on target) in *your* number of rows. I suggest increasing the decrease frequency at the beginning of the shaping (i.e. the part that's closer to the underarm), then reducing the decrease frequency at the upper part of the armhole so it's back to every other row. It helps to just chart or sketch this part out for yourself."
Oh, and either in this post or in the LBT post I explained that I couldn't republish the original patterns as I had intended, because I assigned the copyright. I'm not foreclosed from designing loose-fitting, basic raglan sweaters or simple sleeveless tops, I just can't publish those particular patterns myself.
This post has resulted in the largest number of comments to date, by the way. Although I suppose if I opened the Alice Chronicles to comments, that would get pretty vocal.]