For anyone who saw the last post and wondered about how to improvise a Faroese shawl, here's what I did:
I measured one long bottom side of my shawl (that is, from the edge of the center panel out to one corner): in this case, something like 50". I reduced that to about 32", figuring that perhaps the correct proportion for this measurement is a-little-less-than-the-wearer's-height. [Error 1: my shawl was still too big for Diana, or at least that size done in bulkier yarn was overwhelming for her.] I then multiplied that by the gauge of the new yarn, about 5 sts per inch: 160 sts. [Error 2: The new gauge from which I calculated was based on the Koigu with which I cast on, but more of the shawl came from the DK-weight yarns than I'd anticipated, so the gauge was bigger than I expected, approximately 3.5 sts per inch.]
Took the same measurement and calculation for the bottom edge of the center panel. 7", reduced it (pretty randomly) to 6, 30 sts. [No errors here. At least, none that weren't covered in Error 1 and Error 2.]
Put a 5-st border at each side, for a total cast-on number of 360 sts.
Measured my shawl from cast-on edge to center back neck, compared that length to the distance from the back of Diana's neck to her tuches, reduced accordingly, and converted that to a row-count for the new yarn. 20" reduced to 16" times 10 rows per inch = 160 rows. [Error 3: I always assume the standard 2-to-1 ratio for garter stitch, i.e. if 5 sts = 1" then 10 rows = 1". But loose knitting and strong blocking, both so common in lacework, throw that off, until the vertical gauge can be much like that of stockinette (maybe 7 rows per inch in this example).]
Now, a word about how Faroese shawls are built: The obvious shape is a center-back panel that tapers upward (though not all the way to a point), flanked by two wide triangles. The usual method is to cast on at the bottom and then work upward while decreasing. The side triangles decrease (for the most part) in an easy and steady way: on every right-side row, a decrease is made at the beginning and end of each triangle. In the example here, that would mean 2 decreases every 2 rows in each 160-st triangle, so it should take 160 rows to reach the top.
The coincidence of this number matching the row calculation I reached above should look wonderful, but that would overlook the other, less obvious, shaping of a Faroese shawl: it has additional decreases in both the side and center areas, placed relatively near the top, which draw the "wings" in somewhat and make the shawl sit happily on the shoulders (instead of sliding off, as plain triangles tend to do).
An examination of my shawl and a few other shawl patterns gave me the notion that these supplemental decreases tend to start when about half the triangle stitches have been decreased, tend to occur about three times altogether, and tend to involve between 10% and 15% of the existing stitches. In this case, that would mean that there would be a row of extra decreases when there were 80 sts left in each triangle.
I (of course) have no notes about this process, so I have no idea what I was thinking, but examination of Diana's shawl indicates that I in fact worked 4 decrease rows, and that the first one decreased 6 sts, the second one 5, the third one 7, and the last one 4 sts.
Error 4: An astute observer might now remark that each of these decreases diminishes the number of rows needed to get to the top: thus, instead of the projected 160 rows, I'd be looking at more like 138. And indeed, Diana's shawl only has about 130 rows. But this error (which would make the shawl come out too short) actually compensates somewhat for Error 3 (which was going to make the shawl too long). The total length is about 16".
How does the center panel decrease? Many Faroese shawls have lace patterns up the center panel, and the requirements of the pattern dictate when and where the decreases occur. For this plain one, I worked the first decrease row at the same time as the first set of supplemental decreases at the sides, and I did a total of 6 decrease rows (three having 3 decreases each and 3 having 2).
To summarize the knitting thus far:
Cast on 360 sts. Knit 1 row, placing markers after the 4th stitch, the 165th stitch, the 195th stitch, and the 356th stitch. (If I can count, that should bracket the center 30 sts and set off the outer 4 on each side.)
Row 1: K4, slip marker, ssk, k to 2 sts before next marker, k2tog, k to next marker, slip marker, ssk, k to 2 sts before last marker, k2tog, k to end. 4 sts decreased.
Row 2: Knit.
Repeat rows 1 and 2 until 81 sts remain in each outer section.
Next row: K4, *slip marker, ssk, (k9, k2tog) 6 times, k to 2 sts before next marker, k2tog, slip marker,* (k6, k2tog) 3 times, k to next marker, repeat from * to * once more, k to end. 19 sts decreased: the usual 4, plus 6 in each triangle, plus 3 in the center. 73 sts remain in each triangle.
Knit 1 row.
Now go back to repeating rows 1 and 2 again until there are 63 sts in each triangle.
Next row: K4, *slip marker, ssk, (k8, k2tog) 5 times, k to 2 sts before next marker, k2tog, slip marker,* (k5, k2tog) 3 times, k to next marker, repeat from * to * once more, k to end. 17 sts decreased: the usual 4, plus 5 in each triangle, plus 3 in the center. 56 sts remain in each triangle.
Knit 1 row.
Repeat rows 1 and 2 twice more. Then work a row with the usual side decreases, plus 3 sts decreased more-or-less evenly throughout the center section.
Continue on until there are 48 sts in each triangle.
Next row: K4, *slip marker, k2tog, k3, (k2tog, k4) 6 times, k2tog, k to 2 sts before next marker, k2tog, slip marker,* (k6, k2tog) twice, k to next marker, repeat from * to * once more, k to end. 20 sts decreased: the usual 4, plus 7 in each triangle, plus 2 in the center. 39 sts remain in each triangle.
Knit 1 row.
Repeat rows 1 and 2 twice more, then work a row with the usual side decreases plus 2 in the center section.
Continue on until there are 33 sts in each triangle.
Next row: K4, *slip marker, k2tog, (k4, k2tog) 4 times, k to 2 sts before next marker, k2tog, slip marker,* (k4, k2tog) twice, k to next marker, repeat from * to * once more, k to end. 14 sts decreased: the usual 4, plus 4 in each triangle, plus 2 in the center. 27 sts remain in each triangle.
Back the the theoretical part:
From here it gets a little dodgy. My original shawl does this cool thing where the remaining stitches in the center section suddenly start to decrease by tapering inward from the sides, fast (like decrease-on-every-row fast). I assumed that this was typical, and duplicated it for Diana's. So after the most recent set of center-panel decreases, there were 13 sts remaining. I waited until I thought I was six rows from the end of the side triangles--which would be when there were about 7 sts left--and then began to decrease the center by working (ssk, k to last 2 sts, k2tog) between the center markers on every row.
Special note for the final decrease row in each section: when there are only 3 sts remaining in any section, work a double decrease (slip 1 st, k2tog, pass the slipped st over) to reduce them to one.
With any luck, this has brought everything to a point, and there are 4 sts outside the markers, plus 1 st in between each pair of markers: 11 sts. Option 1: rearrange the sts onto 2 needles and work a 3-ndl bind-off from outside edge to center, working the last 3 sts as "k3tog" instead of the usual "k2tog." Option 2: Rearrange the sts onto 2 needles and stare hard at them until you remember how to work a garter-stitch graft. (Hint: perhaps it would help to knit the first half of the row and then graft from the center out to the edge.)
After you have your second kid, you get a little less obsessive. (Or maybe "more chaotic" would be more accurate.) By the time I got to Eva's shawl, I'd noticed my own tendency to want to work more shaping and make things more complicated. This time, I only did 3 supplemental decrease rows in the outer triangles, and I decreased the center panel gradually but kind of randomly, working "ssk, k to last 2 sts, k2tog" every 4th row for awhile and then more-or-less every 2nd row until I ran out of sts at the same time as the side triangles ended. I also made what I think of as the "Landscape Shawl Modification": instead of decreasing on every right-side row, I decreased on every row, but only for the first half of the row. That is, I'd begin each row with "k4, ssk, k to 2 sts before next marker, k2tog," then knit the rest of the row plain. This keeps you from having to think about which is the right-side row, which can sometimes be a nuisance with garter-stitch. No one's every going to notice that the decreases on one side are offset by a row from those on the other.
Stripes and Yarn Requirements
All the stripes are in multiples of 2 rows (or 1 garter ridge), so that the color-changes are clean and the yarns can all be carried up one side edge. (A meticulous mommy might work a round of single crochet around the edges to hide all that, but I'm not that mommy. And even Eva, who complained about the ends sticking out, has nothing to say about the stranding up the side.) Diana's shawl was in 4-row stripes until I realized that neither the Koigu nor the Snuggly (which was, of course, the last skein in the world of a discontinued color) was going to be sufficient. Then I started mixing things up a bit, introducing the new green with a 2-row stripe and rationing the old green so that some of it continued all the way to the top. The next multi after the Koigu ran out was a skein of Regia (Ringel, I think, in a color that may have been called "Clown"), and I finished up with some Stahl Baby Color (is that stuff even made anymore?). There are 4 entire 50-g skeins in there, varying in yardage from about 175 to about 225, plus some more.
Eva's shawl used Debbie Bliss Baby Cashmerino (the raspberry), Dale Baby Ull (the turquoise), Peter Pan DK (the royal blue), and Fortissima Socka (the purple), all on US #4 needles, for an overall gauge of about 5 sts per inch and a cast-on number around 280. This time, I'd started with full skeins, and I knew what they were, so I was able to calculate the yardage: about 380 altogether. The colors always repeat in the same sequence, but the stripes are on a repeat of 3 (4 rows, 4 rows, 2 rows) to liven things up a bit.
The depth at back-of-neck on Eva's shawl is only 13", and the wings each measure maybe 25" from center corner to tip. There's a kind of pattern to the sizing here: I generally knit large for Diana, with the usual knitter's impulse to have her get the most wear out of it, and I often overdo it, as here. Then, when I knit for Eva, I overcompensate--not only because I don't want it to be too big again, but because I figure it doesn't need to fit for that long, since she'll inherit her sister's soon enough.
If these instructions seem too vague, take my advice: make a flying guess at a cast-on number and dig right in; figure the rest out as you knit. It's a shawl, after all, and exact measurements and shaping aren't crucial. Wing it. If you get stuck, come find me, and I'll bail you out. It's all a lot easier than it looks.