let me explaiKnit

Expanding on a theme

One of the aspects about increasing that many people find tricky is doing it in a pattern, and incorporating the new stitches into the pattern stitch. We'll talk about how it works generally, and then look at the specific cases of ribbing and seed stitch since those come up frequently.

It's going to simplify the following discussion if we first look at how successive increases affect the structure of the fabric. Typically, patterns will call for successive decreases to be done at the same place in relation to an edge of the fabric -- that is, you'll be asked to work 1 or 2 stitches, do an increase, and then work to the same distance from the other edge and do an increase, on each of several increase rows. This means that your second increase is going to be between this fixed edge stitch and your first increase, and so on. This is going to affect your choice of stitch for each of the increased stitches as you incorporate it into the pattern.

A second concept that's helpful to understand is that you will be temporarily out of pattern as you work your initial increases. A well-crafted pattern will have a number of increase rows that's a multiple of the number of stitches in your stitch pattern; for instance, a pattern where you're increasing in seed stitch, a 2-stitch pattern, should have an even number of increase rows, while a pattern in K2P2 ribbing should have a multiple of 4 increase rows. This will enable you to work your way back into the pattern, but being off during the interim is inevitable.

So, let's look at how this works, using seed stitch as an example. Let's say that we're working on a seed-stitch sleeve, and our pattern tells us, "Increase 1 st on each end every 8th row, 6 times." Let's further say that the pattern is well-written, and therefore we have a stockinette selvedge of 1 stitch (to make seaming clean and easy), and an even number of stitches in seed stitch (so that after doing our easy seam, the pattern will flow smoothly across the seam, rendering it invisible); therefore, the RS row directions are "k1, (k1, p1) to last st, k1," and the WS directions are "p1, (p1, k1) to last st, p1." Now, as we've already discussed, a KFB type of increase is the best choice for seed stitch, and it can be placed on the second stitch in from each edge; as to the specific type, we'll choose between KFB and PFB based on the type of the stitch that the increase will be done on, and do a mirrored version for the increase at the left edge so that the new stitch appears before the stitch where the increase is done and not between that stitch and the edge stitch.

Let's briefly review the mirrored front-and-back increases. These are done like so: slip a stitch knitwise and put it back on the left needle with its left leg in front; work (in either knit or purl, as appropriate for the original stitch in pattern) into the left leg and let the old stitch go; pick up the old stitch again using the tip of the left needle; work (in either knit or purl, the same as used for the first stitch) into the right leg of the stitch. This gives you your twisted stitch with the bar around it first, and your untwisted stitch with no bar second, which is the exact reverse of the standard front-and-back increase.

Looking at the first increase row, we'll knit the selvedge stitch, then do the increase on the following stitch, which is a knit, so that will be a KFB. We'll then work in seed stitch as established, doing (p1, k1) across until we reach the 2nd stitch from the end; this one is a purl, and we want to have the new stitch appear first, so we'll do mirrored-PFB here, and then knit the final selvedge stitch.

Now, let's look at the first WS row after this increase, which is where you have to make the decision for the first time on how to work that new stitch into the pattern. We'll purl the selvedge stitch, then purl the first stitch of the seed-stitch pattern, and the next one is the new stitch. Subsequent increases are going to come in between these two stitches, separating this new stitch from its "parent" (I'm going to use that as the term for the stitch that the increase was done into), but it's always hereafter going to be adjacent to the stitch that follows it. Therefore, you want to work the stitch as the type of stitch that should be adjacent to that following stitch, rather than the type that the pattern says should be adjacent to the parent stitch. The following stitch is going to be knit on this row, and therefore the new stitch should be purled; this does, temporarily, give you two purls in a row, but that's okay. You'll then (k1, p1) until you have 3 stitches left in the row, which are the other new stitch, its parent, and the other selvedge; by the same reasoning as above, you will knit the new stitch and also knit its parent, and then purl the selvedge stitch as always. The other WS rows between this row and the next increase row will be worked in exactly this same manner; RS rows will be "k1, k2, (p1, k1) to last 3, p2, k1" (yes, I could have said "k3" at the start -- but I think this makes it clearer). You are slightly out of pattern for the 7 rows between the two increase rows. However, since what you have is a 2-column strip of garter stitch on the edge of your seed stitch, it's not at all obtrusive; you can take my word for that, or you can swatch it up and see what it actually looks like.

Now, let's see what happens on the second increase row. Again, you're going to place the increases on the second and second-to-last stitches of the row, and since those stitches haven't changed, these are going to be a KFB and a mirrored-PFB respectively. Therefore, this row will be "k1, KFB, (k1, p1) to last 2 stitches, m-PFB, k1." On the following WS row, you'll purl the selvedge stitch and purl the parent; the stitch that follows the new stitch is the same one that was new during the first increase, and was purled on the WS rows, and therefore the new stitch for this row should be knitted on the WS to be compatible with that, and just like that, you're back in pattern. Continue on across the row, and end by purling the new stitch between the two knits over there. You're now completely back in pattern until the next increase row comes along. You'll continue in this fashion, getting out of pattern on the odd increase rows, and back into it on the evens, until you've done all the increases.

Increases in ribbing are similar. Let's look at a 4-stitch pattern instead of a 2-stitch one, and just to make it a bit different, let's suppose we're doing a ribbed sleeve in the round. Let's say that our rounds are set up so that the beginning of the round is in the middle of a knit rib, so that our ribbing directions are "k1, p2, (k2, p2) until 1 st remains, k1," and that our increase directions are "Inc 1 st at beginning and end of round, every 6th round, 8 times." Since we're working in the round, there's no need to maintain a selvedge for seaming; an increase can be done in the very first or very last stitch of the round, and therefore we can do a KFB in the first stitch, and a mirrored-KFB in the last stitch, on each increase round. As with the seed-stitch example above, each successive increase will appear between the "parent" stitch and the prior increase, so the new stitches should be worked in pattern with the stitch outside the increase. Therefore, on the first and second increase rounds, the new stitches will both be knits, and you'll have a band of first 4 and then 6 knits up the inside of the sleeve. On the third increase round, however, the new stitches will be purls, so you'll have a single purl rib separating your new 2-knit ribs from the original center rib, and then after the 4th increase round, you'll have a 2-purl rib on either side of the original center knit rib, and you'll be back in pattern; the 5th through 8th increase rounds will just be a repetition of the first four.

So far, we've only looked at cases where the parent stitch is always a member of the same column, so the second and subsequent increases are spawned between the parent and the prior increase. This is by far the most common placement. You can, however, have cases where the parent column changes from one increase row to the next, with subsequent increases being done from the column which was new in the prior increase, and this has two effects which are different from what we've looked at so far: first, each increased stitch will stay adjacent to its parent, rather than being pushed away by the following increase, so it will need to be "in pattern" with respect to its parent, rather than with respect to the stitch beyond; second, because the parent column is different every time, the type of increase used will vary, rather than staying constant as it's done in our prior examples. This is a somewhat related concept to the concept of strong and feathered decreases, which either have a single surviving column or a change of surviving column each time.

A place where these "feathered increases" might be used to good effect would be in the thumb gusset of ribbed mittens or armwarmers, so let's look at some very simple knitted armwarmers as an example -- these will be just a tube with a thumb, and since they're totally lacking in fingers, the thumb can be placed right at the side, with no adjustment for left and right hands. Let's suppose a fingering-weight yarn, and a cast on of 48 stitches in k2p2 rib. After working to the desired length for the wrist, our first increase row will be "mKFB, KFB, place marker, p2, (k2, p2) to end"; subsequent increases will be "increase in 1st st of round and last st before marker, every 3rd round, until there are 18 sts before marker."

Here's how this works out. On the first increase round, you're doing a mirrored KFB into the first stitch of the round, which is a knit stitch, and a regular KFB into the second stitch, which is also a knit stitch, and then you're placing a marker, and finishing the round. On the following 2 work-even rounds, the stitches that were increased will be purled, since a purl stitch is what should be adjacent to each of the parent stitches. On the second increase round, you're going to be doing the increases into the columns that were new last time, and since those are both purls, the new increases will be a mPFB and a PFB. Those stitches will also be worked as purls on the work-even rounds, since you want a 2-purl rib on either side of the 2-knit rib you started with, and the increase on the 3rd increase round will also be mPFB and PFB. For the work-even rounds that follow this increase, you'll be knitting the new columns, and then the increases in the 4th increase round will be mKFB and KFB.

By now you should be seeing how this works; you'll have a total of 8 increase rounds, increasing 2 purls, 2 knits, 2 purls, and 2 knits, on either side of the original 2-knit rib you started with. This produces a very interesting gusset, with the new ribs starting sort of from nowhere rather than seeming to spawn off of another column as they would if you did strong increases.

If you'd like to actually knit up some armwarmers like this, I've written up a pattern for them, and you can find that right over here. If you're on Ravelry and want to add it to your queue, here's the link for that.

That should give you a pretty good grasp on the basics of incorporating increased stitches into a pattern. I do have a bit more to say on the topic -- in particular, I want to talk about how one works increases into a lace pattern -- but I think I want to do it in a slightly more generalized way, looking at the decreases also, so I'm going to do that separately. Stay tuned!

An expanding view

We've gone over the various types of single increases, and I've shown you some of the pattern language that generally serves as a trigger that a certain type of increase is expected by the pattern writer. Now, let's talk for a bit about choosing increases when you don't have that guidance -- maybe the pattern just says, "increase X stitches in the next row," or "increase at each end every Nth row X times," or perhaps you're working up your own design, and not sure what kind of increase you'd like to use.

In general, it's best to suit your increases to your fabric. Stockinette, for instance, is such a smooth fabric that the little bar of a KFB-type increase really stands out. Sometimes this can be a design feature, but most of the time, it's probably a better idea on stockinette to use an M1-type or a lifted-stitch type of increase; no increase is truly invisible, just as no decrease is, but these fade into the background of the fabric more cleanly. As to the choice between M1 and lifted-stitch increases, that's largely personal preference; however, in those cases where the increases are matched against decreases (for instance, when you're doing waist shaping), I'd suggest the lifted-stitch increases, because they're the closest match to the decreases -- they have an "emerging from underneath" appearance that matches the "disappearing underneath" look of the decreases well.

On a textured fabric, such as seed stitch or ribbing, the texture of the KFB-type increases can blend in very nicely. In particular, where increasing out of ribbing into stockinette, the KFB is beautifully suited; if placed on the last stitch of the knit rib, the barred stitch blends into the purl rib almost invisibly (alternatively, mirrored-KFB can be placed on the first stitch of the knit rib, with like results). However, again, if you're matching increases to decreases, the lifted-stitch increases are certainly workable in a textured stitch, and they do match the decreases well.

Now that you've picked a type of increase, where do you put it? The two places where you see this kind of low-guidance instruction to increase are where you're increasing at one or both edges of a piece, or where you're just given a total number of increases to distribute within a row. If the placement is more complicated than that, you're generally told exactly where to do it.

Even when you're told to increase "at the edge," you will almost never want to place an increase directly on the edge of a piece, as this will give it a choppy appearance. Furthermore, if the edge is to be seamed, it's generally best to leave the edge stitch untouched to provide a clean selvedge for seaming. Leaving it untouched has slightly different implications for the different types of increase. For instance, if you're doing an M1, this type of increase tends to exert influence on the stitches on either side of it, and therefore it's probably best to place it two stitches away from an edge to be seamed, in a sequence such as "k2, m1, k to last 2 sts, m1, k2" (although you will see directions calling for only 1 edge stitch, at times; considerations such as the ease of incorporating stitches into a pattern may outweigh the desire to keep a clean edge for seaming). A KFB increase, on the other hand, only affects the stitch it's done into and to a lesser degree the stitch that's beyond the new stitch; it doesn't affect the stitch that remains adjacent to the original stitch. Therefore, a sequence such as "k1, kfb, k to last 2 sts, mirrored-kfb, k1" works nicely with this type of increase; a similar sequence works well with lifted increases, when a left lifted increase is done a stitch away from the right edge and a right lifted increase done near the left edge, but if you wish to use the right lifted increase on the right edge and vice versa, then two untouched stitches are advisable if you're going to be seaming.

When you're distributing increases evenly across a row, the biggest challenge for most people is to remember edge stitches. For instance, if one is distributing 4 increases across 20 stitches, it's easy to think, "Aha, I'll just divide by 4," and so one does "k5, m1, k5, m1, k5, m1, k...." and about this time, realizes that one is going to run out of stitches without having done the last increase. The trick is, when distributing X increases across Y stitches, divide Y by (X+1) instead of by X -- in this example, divide 20 by 5, not 4, so the sequence would be "k4, m1, k4, m1, k4, m1, k4, m1, k4" and the decreases would work out perfectly. If the number of stitches isn't evenly divisible by the number of increases, round down, do the division, and then add half the excess to each end. If you'd like that math done for you, Lucia Liljegren has a very nice calculator which will help you out.

Now, having given you a nice little rule of thumb, I'm going to point out some exceptions to it. First, if you're working in the round, the need to add edge stitches disappears; the initial example, with 5 stitches between increases, works out fine if that's a round of 20 stitches instead of a flat row, although it does put the final increase right at your round marker. Lucia's calculator can still be used, but will work best for work in the round if you use the unbalanced rather than the balanced calculation. Second, this calculation (and Lucia's calculator) assumes an increase that doesn't use up one of your existing stitches; if you're doing KFB increases, which do, you'll need to subtract one from each group of stitches except the last, to account for the fact that you'll use one up when you do the increase -- the sequence in the example would become "k3, kfb, k3, kfb, k3, kfb, k3, kfb, k4." Third, you may want to modify this distribution if you're working in a pattern stitch, to something that will mesh more nicely with the pattern. For example, as I earlier pointed out, KFB increases done on the last knit of a knit rib will blend very nicely into ribbing; making sure that each increase ends up in this placement is therefore probably more useful than making sure the placement is exactly even. This is probably easier to do by hand, by making a sketch that shows your number of repeats and then marking which ones you're going to do the increases on, rather than using a calculator and then modifying the output. With all these exceptions, you can see that there's as much art as science in evenly distributing increases, but this at least gives you a place to begin.

Next up, we'll talk about working increases into a pattern stitch, particularly ribbing and seed stitch since those frequently give people trouble.

Expanding our horizons

We've spent a while talking about decreases -- how about some increases for a while? Let's look at the general types, some of their advantages and disadvantages, and where they can and cannot be freely substituted. We'll also talk about how increases are specified within patterns, which is more often vague than the way decreases are specified.

I'm going to classify the basic single increases into three groups: those which are done between existing stitches, those which are done into a stitch in the current row, and those which are done into a stitch in the prior row. Within these groups, they behave fairly similarly, and in most cases you can freely substitute one member of the group for another member of the group without doing serious violence to your end result.

The easiest of the between-stitches increases is the YO. Of course, easy doesn't necessarily mean simple, or I wouldn't have written quite a bit about them already, but fundamentally, it's just a loop of yarn over the needle. In common with the other increases in this group, the YO makes 1 stitch out of 0, not using up any of the existing stitches. Unlike the other stitches in the group, it has no directionality once completed; it has an orientation as it sits on the needle, but lies flat once you've worked it in the following row. And, of course, it leaves quite a noticeable hole; for this reason, it's really not interchangeable with the other decreases in the group. Pattern instructions are generally quite clear when this increase is desired, and will call specifically for this or one of the British terms for specific varieties of YOs.

The remaining between-stitches increases are close variants of one another, and with a few minor caveats which I'll get to as I discuss each of them, may be freely substituted. Typically an instruction of "make 1" or "m1" is an indication that this type of increase is expected, and you may use whichever one of these you prefer. Each of these does have a left-leaning and a right-leaning version, and in situations where you're doing a series of paired increases (for instance, both sides of a sleeve, or either side of the shoulder join on a raglan), it's usually best to use them in mirrored pairs. There are three variants: the lifted bar, the backwards loop, and the twisted YO.

The lifted-bar M1 will probably be the most familiar to you; most people learn it quite early in their knitting careers. This increase is done by picking up the bar that connects the next stitch on the left needle to the stitch just worked, placing that bar on the left needle as if it were a stitch, and then working into it (in knit or in purl, as appropriate) so that it's twisted, preventing a hole. The left-leaning variety is performed by putting the bar on the needle in standard Western orientation and working into the back leg; for the right-leaning variety you put the bar on the needle in Eastern orientation and work into the front leg. Amy at KnittingHelp has some nice photos of these; they may be found at the third knitted diamond down from the top, and are called M1R/M1F and M1L/M1B respectively.

The backwards-loop M1 may also be familiar; Elizabeth Zimmermann was quite fond of it and recommended it frequently. This increase is done by simply making a backwards loop of the yarn and placing it on the needle, just like the backwards-loop or e-loop cast-on. For the left-leaning version, the yarn is twisted clockwise, so that the working yarn points away from you; for the right-leaning version, it's twisted counterclockwise, so that the working yarn points towards you. Again, KnittingHelp offers photos; the first and second diamonds on the same page mentioned above show these increases, labeled as M1A and M1T, respectively. The first diamond also shows what happens if you don't use mirrored increases -- it's a less dramatic effect than what you get with unmirrored decreases, but you can still tell a difference. These increases differ from the lifted-bar M1s in two respects: they use up a bit more yarn and pull a bit less on the neighboring stitches, and the twisted stitch at the bottom of the new column appears one row later. If you have issues with this increase leaving a hole, you might try substituting the lifted-bar M1 instead; because it "steals" its yarn from the stitches on either side, instead of being given a stitch worth of yarn to itself at the time it's created, it tends to leave a smaller footprint. Conversely, if you have issues with the lifted-bar M1 causing pulling or puckering, this one is a nice substitute that usually doesn't have that problem. The one-row difference usually doesn't cause an issue, except that if you're counting stitches to see how many rows you've worked, you'll want to remember whether the twisted stitch is part of the increase row, or part of the row before that.

The twisted-YO M1 shares some of the attributes of each of the other types, and is done by creating a YO (either backwards or forwards) on the current row, and then working into it so it's twisted on the following row; a forwards YO is worked through the back leg to create the left-leaning version, while a backwards YO is worked through the front leg to create the right-leaning variety. This increase feels like a slightly looser lifted-bar M1 when you're working it, but like the backwards-loop M1, the twisted stitch appears on the increase row; the amount of yarn used is in between the other types. I favor this one, finding it to strike a nice balance between ease of working and neatness in appearance, but because it's really not completed until the row following the increase, I use it with caution where the row immediately following the increases is a complicated one.

That covers the increases done between stitches; now, let's look at those done into an existing stitch. These generally fall into two groups: those where you work the same type of stitch into both the front and back, in some order, and those where you both knit and purl into the same stitch, in some order. Unlike the between-stitches increases, these increases use up 1 existing stitch to make 2 stitches. Many patterns are explicit about these types of increases, with them spelled out within the line of instructions or clearly indicated as an abbreviation, but others are more obscure. When you are told to increase "in" or "into" a stitch, generally this type of increase is intended; that's fairly clear, since it is done within a stitch. However, often but not always, a pattern instruction that says merely "inc 1" is also looking for this type of increase; this seems to be a little more prevalent with British authors. If a pattern uses both "m1" and "inc 1" then you can be fairly confident that the "inc 1" is looking for this type of increase; if it uses only "inc 1", then be careful. Often the stitch counts will provide a clue: if a pattern says "k1, inc 1, k to last 2 sts, inc 1, k1" then you know that the "inc 1" is intended to use up a stitch, and therefore must be this type; if it says the last stitch instead of the last 2, then they must mean for you to use a between-stitches increase, because if you used this type there wouldn't be a stitch left for the k1. And with that out of the way, let's start looking at the mechanics of the individual increases.

First, we'll look at the front-and-back increases, and let's start with a fundamental question: why front and back, instead of just working twice into the same part of the stitch? The answer is that you can't work twice into the front without an intervening step. If you knit into the front of a stitch, and then immediately knit into the front of it again, you actually undo the anchoring of the working yarn into the stitch, and end up with a single stitch that's just wrapped around the needle in the middle, like what you get when you do a double YO. When you work into the front and then the back, or the back and then the front, a loop of the existing stitch wraps around the working yarn and secures it, so you can make that separate second stitch. This bit of yarn from the existing stitch winds up wrapped around the base of the second stitch worked, forming a little horizontal bar of yarn on the knit side of the stitch (you can see a good photo of this on that same KnittingHelp page, 5th diamond down); intriguingly, it's always the 2nd stitch and always the knit side of it that shows the bar, whether you do front first or back first, and whether you do knits or purls. Because of this little bar of yarn, these increases are sometimes called bar increases; however, that's susceptible to confusion with the lifted-bar M1 increase, so I think it's probably better to avoid the term and just be specific with KFB or PFB.

KFB is by far the most common of these, and is done by knitting a stitch through the front leg but retaining it on the left needle, and then swinging the tip of the right needle around to the back and inserting it in the back leg to work the stitch a second time; this gives you one normal-looking knit stitch, and a second to its left with a little collar twisted right up around its base. If you work into the back first and then the front, you don't get the reverse of this -- you get the front stitch obviously twisted, and then the second stitch looking just about the same as with KFB. I mention this primarily so you'll realize this doesn't give you a mirror-image; I'm not sure when you'd actually use it, except possibly if you were increasing in a twisted rib. PFB is done similarly, working into the front and then swinging the needle around to work into the back; again, the back-leg-first version just leaves that first stitch twisted, rather than forming a mirror.

You may have gathered that I think mirroring is kind of a Big Deal with increases of this type. Actually, I think it's important in any situation where you've got increases in pairs -- things just look funny if the fabric isn't increasing symmetrically -- but with these increases, the mirrored option is non-obvious. When you increase within a stitch, you'll have one column that appears to be the continuation of the column the stitch was originally in, and one column that springs up on the row where you do the increase; with each of the 4 increases I just mentioned, even if you reverse the order of the front and the back, the continuing column will always be the one on the right, and the new column always the one on the left. One common way to adjust for this is to move the leftmost increase one column to the right -- for instance, if you want your increase points to be 2 stitches from the edge, on the right edge you'd do "k1, KFB", but then on the left edge, you'd do "KFB, k2", to put the increased column as the 3rd column in from both edges. That's not bad; you can tell a difference if you have stacked increases, but it's not as dramatic as it would be with decreases, so you can get away with it. A slightly improved option is to do KFB on one edge on the RS rows, and PFB on the other edge on the WS rows -- you might do "k1, KFB, k to end" on the RS, and then "p1, PFB, p to end" on the WS, for instance. PFB still puts the increased column to the left, but since you're working it on the WS, it'll appear to the right on the RS; there's a one-row offset, but if the edges aren't brought close together, the eye won't pick that out, and the slant of the columns within the fabric will match nicely.

Until quite recently, I'd have said that was as good as things were going to get. However, very clever knitter Jenn recently pointed out to me a way to do a truly mirrored KFB, where every single thing is switched around, and the nice untwisted continuing stitch ends up on the left and the newly created column ends up on the right. To do Jenn's Mirrored KFB, you slip the stitch knitwise, then replace it on the left needle in Eastern orientation; knit it through the front and let it go, and then pick up the left leg again (still in Eastern orientation), and knit it through the back. It's a perfect match to the KFB -- isn't she smart? (It's quite possible, even likely, that someone else had already thought this up -- but I've never seen it before, and Jenn figured it out independently, so I'm giving her credit, although I'll be happy to provide equal mention to anyone else who's also documented it.) With that to go on, I've worked out a Mirrored PFB to match: slip a stitch knitwise and return it to the left needle, then purl it through the front and let it go; pick up the left leg, which will be crossed over the front of the stitch, into Eastern orientation, and purl it in the back. So, there you have it -- nicely mirrored options to match each of these decreases. I'd suggest using these any time you're doing KFB/PFB in pairs.

Now let's turn to the other group of stitches that falls in this category -- those where you knit and purl into the same stitch, in some order. The major difference between these and the front-and-back variety is that instead of the existing stitch getting wrapped around the working yarn to form a bar, the working yarn gets wrapped around the top of the existing stitch, forming a little nub sort of like half a stitch worth of moss or seed stitch. Because of this, this increase is sometimes called a moss increase; referring yet again to that handy page at KnittingHelp, you can see this one on the third diamond counting up from the bottom. Despite the name, however, this probably isn't the best increase to use when working moss or seed stitch. That's because this increase tends to open up the stitch it's worked into, leading to a small hole under the increase point; I think it's really best suited to the one case where that's a real advantage, which is when you're working into a YO, to make a large hole -- there, it gives a cleaner, smoother result than working into the front and back (and, in fact, a knit and a purl in some order are almost always indicated when working into the two loops of a double-YO, which is structurally a single stitch). The double-increase version of it has some other uses, which I'll get into when we talk about double increases generally, but the single increase really has just that one place where it's clearly the superior choice. It doesn't matter much whether you knit first and purl second, or vice versa; it shifts the slant of the nub a tiny bit, but so little that it's probably going to be overwhelmed in the texture of the yarn anyway.

What if you combined these, and did, for instance, a purl into the front and then a knit into the back of the same stitch, or the other possible combinations? Well, you can do that, and it doesn't give a bad result. It's a lot like KFB/PFB, with whichever stitch you do first and its twist or lack of twist being the most clearly presented; you get both the working yarn wrapping around the top of the stitch and the stitch wrapping around the base of the second stitch, so the type of stitch and whether it's twisted or not is pretty well obscured. It might be a little snugger than KFB/PFB, which would be all to the good, but it's a bit more of a pain in the neck to work, so I'd suggest experimenting and seeing if you think it's worth the bother. As with the KFB/PFB, you can do this in a mirrored fashion, but I'm going to leave that one as an exercise for those of you who're curious.

Before I move on, let me now clarify something that I deliberately left a bit obscure when I was talking about pattern instructions that indicate an increase to be done into a stitch: which kind do you do? In almost all cases, you're going to want to pick the KFB/PFB type; as I went into above, the knit-and-purl sort is best suited to those cases where you want an open result. In addition, your instructions may indicate which to choose as between the KFB and PFB by telling you to increase knitwise or purlwise. I have never seen a pattern indicate a mirrored KFB/PFB or the combined form, but I think I've given you sufficient guidance to understand when those might be called for.

Finally, there's our third class of increases, those done into a stitch in the prior row. There are only two of these, and they are mirrors of one another, so there's not really any question of substitution within this group. To perform these increases, you pick up a stitch from the prior row and knit into it; for that reason, these are often called "lifted" increases, but make sure you don't confuse them with the "lifted bar" type of M1 increase. The first of these is done by working into the stitch immediately below the next stitch on your left needle; you will pick up the right leg of this, place it on the left needle, and knit into it, then let it go. It leans right, away from the column where it was done, and is usually called a right lifted increase, or RLI. The second lifted increase is done by working into the stitch immediately below the last one that has already been knit, which will be two rows below the last loop on your right needle; you pick up the left leg of this, place it on the back of the left needle, knit into it, and then let it go. It leans left, away from the column where it was done, and is called a left lifted increase, or LLI. Amy at KnittingHelp has photos of this also, on the 4th diamond from the top, but calls them KLR and KLL, respectively. Meg Swansen suggests a variation on these, which is to knit them so that the old stitch is twisted; it's tighter and counteracts the tendency for a hole to develop where a stitch has multiple stitches worked into it. These increases are the best match for single decreases; they have nearly as strong a lean, and an appearance of stitches emerging from the side of a column which complements the way they seem to disappear under a decrease. This is particularly true when the lifted stitch is twisted, and is part of why I like this variation even though in other places (like the "improved" SSK) twisted stitches make me twitch. Usually instructions will be clear about calling for these if it's the designer's intent, but because they don't affect the count of the stitches waiting to be worked, they can be substituted for the M1 increases if that seems desirable to you.

Next up: we've covered the instructions that are usually used to indicate a particular intended increase, but it'll be useful to talk a bit about those cases where no intent can be determined, and how the fabric you're working in can affect the appropriate choice of increase there. I'd also like to cover the double/multiple increases, but I'm not sure if that'll be the same post or a different one, and then after that, with both increases and decreases in our toolbox, maybe I'll spend a bit of time on increasing and decreasing in pattern, particularly in ribbing and seed stitch.  Stay tuned!

Two forward, one back

And while I'm on the topic of decreases again, here's a quick little nugget that I think is really clever, and thought I'd throw out:

I have a commercially-knit cardigan that has a section of 2x2 rib that merges into a 5x2 based pattern, so that they needed to get rid of one of the stitches in a "k2, p2, k2" sequence. Now, one of the design issues with decreases is that they do slant visibly -- you can minimize the impact by not lining them up, but you can't really get rid of it. However, what the designers of this cardigan did was to use "k1, SSK, YO, k2tog, k1" across these 6 stitches, and by using two mirrored decreases on either side of a neutral increase, they produced a net decrease of 1 stitch with no resulting slant. Of course, this does produce a hole, which is rather more visible than an increase, but is also rather more easily incorporated as a design element, and indeed that's what was done here, as there's a lace pattern in the upper part of the cardigan. One could use an M1 to avoid the hole; there's a slight slant due to the twist in this increase, but it's less prominent than the slant in a decrease, and the result would still be more symmetrical than a single decrease. I find this to be really interesting as a lesson not to think of my increases and decreases as single units, but in terms of the net increase or decrease to the row overall, and I'm busily thinking up intriguing ways to incorporate it into some of my designs.

A left-leaning bent

We've talked before about why decreases lean, and mirrored decreases, but there's one thing I didn't get into back then, and that's the niggling little way that our left-leaning decreases never look quite as good as our right ones. Those k2togs are nice and crisp, but SSK, SKP... they're tough, and usually the top stitch of the decrease ends up being much more prominent than the top stitch of the right-leaning decrease, which is annoying when you're trying to do them in pairs. I'm hardly the first to note this; for instance, a while back Nona discussed this on her blog, and came up with a comparison of several left-leaning decreases. I touched on this while talking about neater ribbing, but I thought I'd go into it in a bit more depth; I don't necessarily have a conclusion here, but I do have some more thoughts on why we see this phenomenon, and some ideas for you to explore to address it in your own knitting.

First, why does this happen? At first glance, there doesn't seem to be a structural reason for it; a loop drawn left-to-right through two stitches with their right legs in back (an SSK) ought to produce the exact mirror of a loop drawn right-to-left through two stitches with their left legs in back (a k2tog). That leads to the idea that maybe the difference is executory, and perhaps is caused by putting too much tension on the loops when they're being slipped and/or passed over. However, if that's the case, there's an obvious solution; on the prior row, wrap the two stitches the opposite way so they're already sitting with their right legs in back, and you won't have to do any slipping. It turns out, though, that while this certainly does speed up the decrease, it really doesn't do as much as you might expect for the overall look of it. This surprising result got me looking more closely into the mechanics of forming the decrease.

It turns out that there is a structural difference in the formation of left and right decreases, after all. June Hemmons Hiatt, in her illustrious The Principles of Knitting, pointed me in what I think is the right direction, when she noted that the rightmost stitch of a 2-stitch decrease will always be bigger, and that the reason the k2tog is neater is that the bigger stitch is hidden under the other one, whereas in a left-leaning decrease it's necessarily on top. That got me thinking about why it would be bigger, and therefore looking into exactly what really happens to the stitches as they're worked. Work some plain knit stitches, and take a look at how the yarn moves as you complete one stitch and pull the needles apart. You'll notice that as you separate the needles to insert the tip of the right one into the next stitch, the strand between the stitch you've just worked into and the one you're about to work is pulled a bit and that this tugs on these two stitches; it's not much, but it actually pulls the one on the left slightly open, and the one on the right closed. When working even, you don't notice this, because each stitch goes through the process from both directions -- enlarged when it's the next one up, and then snugged down again when it's finished. In a decrease, though, the rightmost stitch of the decrease goes through the first half of the process, pulling it open, but it gets insulated from the latter part where it ought to get snugged down again -- the leftmost one gets snugged, but this one doesn't. This explains, too, while the first-blush solution of avoiding the need to slip can kind of work, if you do it carefully -- if you work it up on the tips, you can minimize the "pulling open" part of the process -- but mostly doesn't, because you still miss out on the following tug. It even fits with the fact that some people get better results with an SKP than with an SSK -- if you just slip the first stitch, then it does get tugged down when you separate the needles to knit the second one; you've got to open it back up a bit to get it over the knit stitch and off the needle, but if you're careful you can minimize that.

A very careful SKP is one at least partial solution. Ms. Hiatt offers a couple of suggestions: if one is knitting flat and the decreases are fairly separated, she suggests using a SSP (slip 2 stitches individually knitwise, return to left needle, purl together through the back) decrease on the WS instead, so that the stitch which ends up on top on the RS is the second stitch in the decrease, which gets its tug in due course; if the decreases are adjacent, so that the 1-row deviation would be apparent, or one is working in the round so that the WS is not available, she suggests achieving a match by loosening the right-leaning decrease rather than tightening the left-leaning one, using the technique of knitting a stitch, returning it to the left needle, passing the following stitch over, and then returning the worked stitch to the right needle (one may abbreviate this as k-return-pass-return, or KRPR), and I'll add the additional step of being very careful and delicate when working the following stitch, so that that top stitch doesn't get tugged down too much and defeat the purpose of these maneuverings. And then there's the so-called "improved" SSK, where one slips the first stitch knitwise but the second purlwise, and then knits both through the back; this twists the second stitch, which has the effect of tightening it and making it pull slightly on the first stitch, thus giving it at least some of the tug it would otherwise miss out on. I'm afraid this one makes me twitch -- it's clever, but it leaves a twisted stitch sitting there. Yes, there, right there, can't you see it? Well, I can. Ugh. But if you can live with it, since it is, admittedly, mostly tucked up under the other stitch of the decrease anyway, well, it's your knitting.

And then there are "brute force" solutions, which are just to tug that excess yarn back out of the problem stitch. At the time of doing the decrease, sometimes I will just use my fingernails to grab the back of the second stitch and give a little yank on it, which pops the first stitch into place nicely. By pairing this with the KRPR right-leaning decrease, I get quite nice matching, even in inelastic fibers like cotton; in wool, a careful SKP may do it, especially after blocking. If there's any looseness remaining, I've been known to use a crochet hook to grab the bar between the loose stitch and its neighbor to the right, and just pull it a bit, which gets the looseness into the bar, hiding on the back of the work, instead of in the stitch on the front. It's not elegant, but it is effective.

I'd certainly suggest that you try out a few of these on your own knitting, to learn what works for you; although there are some structural effects, execution definitely plays a large part here, and that means that individual idiosyncracies come into play. You may find, as I do, that a combination of techniques gives you the best overall result, or you may find that one technique is a homerun for you, and in any case, you'll learn a lot about how your knitting works.

When good gauge goes bad

Sometimes, you get a gauge swatch that is not lying, but is instead insisting upon telling you an uncomfortable truth. Despite having carefully planned and knit, washed your swatch, attempted to block it to gauge -- all the right steps -- you find that you just can't get your yarn to match the gauge for your pattern. Or perhaps you can match either stitch gauge or row gauge, but not both. This can often happen with substituting yarn, and sometimes can happen when a designer achieves a gauge toward the limits of a yarn's natural range or otherwise has rather quirky gauge results (Alice Starmore's spectacular colorwork designs are notorious for having hard-to-match gauges, even in the recommended yarns, for instance). When this happens, you'll need to make pattern adjustments so that you end up with fabric of the right size and shape.

Before we get started, here are a few considerations to make sure you're not headed down the wrong path. First, if you are persistently short on row gauge, you may want to make sure you're not twisting stitches. Twisted-stitch fabric draws up and is shorter than untwisted fabric, as well as being less flexible; more than one person has discovered they were accidentally doing combination knitting from exactly this factor. If that's the case with you, you'll be better off addressing this directly, so you only twist stitches when you mean to, rather than adjusting your gauge to compensate for the shorter fabric. Second, if you're substituting yarns, you may want to consider just one more time whether the characteristics of your substitute truly match the characteristics of the original; is it possible that the same factors that are throwing your gauge off will also end in less than desirable results in the finished product? For instance, are you achieving a longer row gauge because your yarn is a heavier fiber with more drape to it, and will that result in a saggy sweater? You may decide that you still think it's a good substitute, or even that the change of characteristics is part of what you're trying to achieve, but it never hurts to stop and make sure you've fully thought things through.

Once you've decided that, yes, compensating for a gauge that's off really is what you want to do, how do you go about that? We'll start with the simpler case, where you can match either stitch or row gauge but not both, and then move to the slightly more complex case where you don't match either.

When you can match only one of stitch or row gauge, you generally want to match your stitch gauge. As a rule, this leads to fewer and simpler adjustments to be made. This may not be the case if you're doing something where row gauge is the greater determinant of fit -- for instance, for a sweater knit side to side -- but in most cases it's going to be true. So for now, we'll assume that you've matched stitch gauge, and you just need to get your row gauge to work.

The easiest case for this is where you have a pattern with no shaping. For instance, look at Jenna Adorno's cute little Tempting sweater. It's form-fitting due to the stretchiness of ribbing, but there's not an increase or decrease in the piece. This pattern also illustrates a common pattern convention, telling you to work until the piece measures a certain length. When you're adjusting for row gauge, what you want to do is figure out how many rows that length translates to -- take your row gauge (if you've measured it over 4", divide it by 4 to give you the gauge over 1", but do not round off yet), and multiply it by the specified length in inches, and then round to the nearest whole number since you can't knit partial rows, and then work that many rows. Pretty easy, and nicely accurate, as long as your row gauge was properly measured. In fact, I'm going to go out on a limb and say that even when you do match row gauge, you're better off making this conversion and knitting to the specified number of rows, especially if you're going to be trying it on as you go and it may be getting stretched out, or if you need to make two pieces that will be seamed together; you're better off doing the math and relying on your carefully controlled measurements from your swatch, rather than hasty measurements on a piece that may be getting temporarily distorted as you work.

It's only a little more work if the pattern gives you a number of rows to work, instead of a measurement. Let's say that instead of saying to knit 14 inches, the pattern told you to work 94 rows, at the pattern gauge of 27 rows/4", and let's say that you're getting 24 rows in 4" instead. You're going to knit (94/27*24) rows -- pattern rows times actual gauge divided by pattern gauge  -- which is 83.5; you can't knit half a row, so round up to 83 or down to 84. Half a row, or even a full row, doesn't make a difference in the scheme of things; it's a pretty small fraction of an inch and the eye can't pick that out. This pattern's in the round, so it doesn't matter which way you round off; if you were knitting flat, then whether you wanted to end with a RS row or a WS row might indicate which way to round.

If you have shaping, then you may need to do a little bit more work just on the shaped portions (straight-sided portions can be handled as above) and change the frequency of your shaping rounds so your shaping doesn't end too late or too early. This is not necessarily so -- if the shaping is fairly minimal, or the difference in row gauge isn't very big, then you may decide to ignore the difference. I'd let the decision rest on whether it's big enough to see -- if it's only throwing things off by a single row, maybe even two rows in a fine gauge, then I'd probably ignore it, because you won't see that the slope of the line breaks a row early or continues a row further into the next step, but if it's more than that, it may be visible, and adjustments become worthwhile.

You make these adjustments about the same way you make the adjustments for length in an unshaped area. Take your number of shaping stitches and the number of rows they're occurring across to figure out how many total rows you have in the shaped section, convert that to inches, convert the inches to the rows in your actual gauge, and then redistribute the shaping so it occurs over that number of rows instead. For example, let's say that you're doing a decrease every other row 12 times, and you're using a pattern with a specified gauge of 6 rows/inch. However, you've substituted a cotton that you know from your swatching will shrink lengthwise to 8 rows/inch. Following the pattern, your 12 decreases should be occurring over 24 rows (don't forget to count the plain row after the last decrease), which should occupy 4 inches. Four inches in your actual gauge, however, gives you 32 rows, and if you did the decreases over the first 24 rows, that would leave you with 8 rows, or an entire inch, with no decreases; that's definitely going to be visible, and you definitely need to compensate, which means distributing the decreases over the entire 32 rows. However, 12 doesn't go into 32 evenly; you'd need to do a decrease every 2.7 rows, which is obviously impossible. Fortunately, this is another place where not being able to see that the knitting's a row off one way or the other helps us out, and a line where the decreases are sometimes every 2 row and sometimes every 3 will still look pretty even. Therefore, you could do the decreases with 2 plain rows twice and then a single plain row once, so you're decreasing on rows 1, 4, and 7 of each group of 8, and that would work out okay.

That's a dandy solution if you're working in the round, but if you're working flat, instead of having all your shaping on the RS rows, now you've got some of it on the WS rows. You can do that; this article explains in some detail which decreases match which RS counterparts, but the short form is that you'd use p2tog on the same line where you'd use k2tog, and SSP on the same line where you'd use SSK/SKP. Another possible solution is to let some of the decreases fall still further apart, perhaps using 3 plain rows once and a single plain row twice, so you're decreasing on rows 1, 5 and 7 of each group of 8, but that's more likely to be visible to the eye than the other distribution.

Another possible choice is to break the decreases up into two (or more) groups, with different rates of decrease in each. For instance, you might do your decreases every 3 rows 8 times, and then every 2 row 4 times; another choice would be every 4 rows 4 times, and then every 2 rows 8 times; still another choice would be every 4 rows 3 times, every 3 rows 4 times, and then every 2 rows 4 times. These patterns give you different slopes in the different sections, but because the slope within a section is constant, it'll look intentional and not like an error, which is a possible danger with the more uneven distribution in the paragraph above. It's an aesthetic call, though, and therefore entirely up to you; there's no one "right" way to do this.

That pretty well covers the situation where you're just changing row gauge, so let's turn to those situations where the entire gauge is off. You're going to do this in two steps, first adjusting your stitch gauge, and then doing the same compensation for row gauge outlined above.

For unshaped sections, where the piece is effectively a rectangle, adjusting the stitch gauge is pretty easy, and quite similar to adjusting the row gauge when you're given a specific number of rows. The number of stitches you need is equal to the number of stitches in the pattern, times your actual gauge, divided by the pattern gauge. To go back to the Tempting sweater for an example, let's say you're knitting a size small, which has 160 cast-on stitches, and instead of the pattern's 18 stitches per 4", you've substituted a finer yarn with which you're getting a gauge of 22 stitches per 4". Your revised cast-on number, by the equation, would be (160*22/18), or 195.5. You can't do partial stitches, so you'd need to round up or down, to 195 or 196.

However, there's a potential additional quirk, which is that you may need to have a particular number of stitches for the pattern repeat. In the case of the Tempting sweater, you need a multiple of 4 for the 2x2 ribbing, so 196 it is. You can run into the need to adjust your corrections for the sake of a pattern repeat with row gauge, too, but it's less common; many patterns are only a one row or two row repeat anyway, and even for those that are longer, it's often easy to stop midway through the rows of a repeat.

Once you've adjusted your stitch gauge, you'll also adjust your row gauge, in the same way you'd do it if you were adjusting row gauge alone, and you're ready to knit up this section.

When adjusting stitch gauge on shaped sections, you want to focus on make the same change in the width. There are two ways to go at this: you can calculate the stitches you need for the width at the beginning, then the stitches you need for the width at the end, and subtract the difference; or you can add up the number of shaping stitches, and then figure out what the width of those would have been at the pattern gauge, and then convert that width to the number of stitches in your actual gauge. These should give you the same result, and you might find it useful to calculate it both ways, as a check against arithmetic errors.

Once you've got your number of shaping stitches to do, then you need to do your row gauge conversion, so you know how many rows to distribute your shaping over, and then you want to see if you can evenly distribute them. This step can actually turn out to be pretty easy -- if your stitch gauge and row gauge are both off from the pattern's gauge in the same proportion, then it may well turn out that while you're doing more (or fewer) shaping stitches, you're also doing them over more (or fewer) rows, and the interval at which you need to do them is the same. As an example, let's say you've got a pattern for worsted weight yarn, with stitch gauge of 5st/in and row gauge of 7r/in, and you're converting it to a DK weight yarn at 5.5st/in and 8r/in, and that you've got a shaping section where you do 10 decreases, every other row. Your change in width is (10*5.5/5), which is 11 stitches. The original length over which this occurred was 20 rows; the new length is (20*8/7), or 22.9 -- using the principle that one row either way doesn't make a difference, you can call this 22 rows, and that means that your 11 decreases can also be done every other row, which is a pretty simple change. It won't always be this easy, but often it will, because the proportions of knit stitches tend to be fairly constant even when you change yarn weight, especially when your fiber type is the same.

However, the caveat about stitch patterns may still apply -- for instance, if you're working in seed stitch, you really need to decrease by even numbers to stay in pattern. Applying this to our example above, if it was in seed stitch you'd really like to have either 10 or 12 decreases, and in this case, I'd recommend going with 12, and going up to 24 rows (again, one row either way doesn't matter), so you can still do them every other row, and you stay in pattern. A couple of stitches more or fewer doesn't make a big difference in terms of width, so that's a perfectly fine adjustment to make. However, if you're making several of these adjustments, you want to be sure that together they don't add up to a more significant discrepancy. The easy way to do this is to make sure that if you make one adjustment in the direction of giving you a bigger item, then the next adjustment is made in the direction of giving you a smaller item, and vice versa. For instance, if I rounded up to get an odd number of stitches to do seed stitch in the round, which would tend toward giving me a larger item, then when calculating my decreases I'd want to round toward more decreases, not fewer, which would tend to give me a smaller item, and if I made a third adjustment, I'd lean towards a larger item again, and so on.

Hopefully, this really helps to demystify making these adjustments. Yes, you have to do some planning, and yes, there's some math, but it's just grade-school arithmetic, and you can even use a calculator. You can definitely handle this, and you'll love the results of making your gauge work for you.

The fabric of our lives

In the previous posts on gauge, I've mostly focused on gauge over stockinette swatches. There's a reason for that, of course; that's by far the most common fabric over which gauge is measured. This is the case even for many projects which aren't worked in stockinette; designers will quite often make the assumption that if you match gauge in stockinette, you'll also match it in the pattern stitch. It's not an unreasonable assumption. However, sometimes the gauge is given over the pattern stitch, or you'd just like to check your gauge in pattern against the finished size, or perhaps you're designing something of your own and want to predict the finished size of the full item. We'll talk about a couple of common situations: measuring gauge over ribbing, and measuring gauge over cables, lace, and other irregular textures.

Counting stitches over ribbing is pretty easy, and the stitches are fairly uniform in size; also, the fabric is balanced, so you don't need borders on your swatch to keep it from rolling up (although putting perhaps one or two garter-stitch stitches on each edge, so the first and last ribs aren't edge stitches, might be helpful, and the other guidelines such as using the same method and taking pre- and post-wash measurements would still apply). The issue with this fabric is how stretched or relaxed it's supposed to be when you take the measurement -- many patterns specify "slightly stretched" or some similar language, which is subject to all kinds of interpretation. This is a fairly subjective area, with individual taste coming into play, but hopefully I can provide some useful guidance.

The key to evaluating gauge over ribbing is to evaluate the fabric as a whole: do you get gauge when the fabric is stretched to a degree that will be attractive when it's in use? There are two approaches to this -- either stretch the fabric to a state you like and then count stitches, or stretch it to the measurements you want and then consider whether you like the fabric in that state. You can use whichever approach you prefer; I like the first one when I'm designing and trying to decide what my own gauge is, but find the second somewhat easier when I'm trying to match gauge for a pattern. We'll look at the mechanics of both.

For either of these methods, you're going to want something to pin to and enough pins to hold down your fabric without scalloping the left and right edges. For the evaluate-then-measure method, pin down one side edge, and then stretch the fabric out to a degree that looks good to you, and pin down the other side. Don't pin the top and bottom edges. Measure across 4" in the middle of the fabric, and count the number of stitches (including fractions). Ribbing generally contracts a bit vertically when it's stretched horizontally, so you'll need to look at that; measure vertically toward the center of the fabric, away from both pinned edges. Unpin the fabric, and measure vertically in the relaxed state; this state is how the fabric will appear when you're knitting, but it will pull up to the prior measurement when worn -- account for this if the pattern asks you to knit to a specific length.

For the measure-then-evaluate method, it's useful to have a paper template of a 4"x4" square. Pin down one edge of your fabric, line the paper template up with that edge, and then stretch the fabric so that the appropriate column of stitches lines up with the other edge of the template, and pin it down. Your fabric now matches the gauge -- do you like it? Is it overstretched, too loose, or just right? If you like it, you're done; if it's overstretched, it's too small and you should go up needle sizes, while if it's too loose, it's too big and you should go down. As with the other method, you'll want to evaluate any drawing up as a result of the horizontal stretch.

You can see that there's a pretty strong subjective element with both of these methods; unlike a stockinette swatch, a ribbed swatch doesn't have one right answer to "what's the gauge?". It's entirely possible, for instance, that one knitter could decide that a swatch looks great when stretched to 22 stitches over 4", and another knitter could decide that a swatch of exactly the same size is too stretched at that gauge, and that bigger needles are needed. However, in the end, both of these knitters should end up with finished items that make them happy.

Measuring gauge over cables, lace, and other textures is slightly less subjective than measuring over ribbing, but it's still got some tricky aspects. One of the complicating factors is that stitches are often not the same size consistently throughout the swatch -- some are stretched out, some are drawn in, some patterns make the rows slant up or down, and that makes just counting them a difficult proposition. Another factor, which does contain a more subjective element, is that these patterns are often intended to be blocked, but the degree of blocking is subject to the knitter's preferences.

A method for dealing with the inconsistencies between stitches in the textured patterns is to use an average gauge for one or more repeats, and convert that to the equivalent of 4" worth of stitches, rather than trying to measure 4" on the swatch. As an example, let's say that you're working on a lace pattern, and the gauge given is 16st/4" in the blocked pattern, but your pattern repeat is 12 stitches. You can't do a pattern and a third, and you really wouldn't want to try. What you want to do instead is cast on about 32 stitches, which gives you enough for a 3-stitch border plus one stockinette stitch, on either side of 2 pattern repeats in your lace pattern. Work around 4 inches of your pattern (and preferrably at least 2 vertical repeats of the lace), using bottom and top borders as with any other swatch. Measure your gauge by measuring from the inside edge of one stockinette edge stitch to the inside edge of the other one (if your pattern varies the number of stitches per row, or if the sides aren't straight, then you'll want to mark the point in the center of the swatch where you go from one pattern repeat to another, and take your measurement there), and then divide 24 by that number to get stitches per inch, and then multiply by 4 to get stitches over 4 inches. This gives you a pre-blocking measurement, useful for comparing as you knit. Now, since you're trying to match a particular gauge, you'll want to block to size -- 16st/4" is 4 stitches per inch, and 24 pattern stitches divided by 4 stitches per inch gives you 6 inches, so that's the width you want to block to. If you were instead designing with the lace, you'd want to block to the degree that seemed good to you, and then measure the resulting gauge and convert it to a 4" equivalent just as you did with the pre-blocking gauge.

What about vertical gauge? This can really be tricky in some lace patterns, where offset increases and decreases cause the fabric to bias, and makes it difficult to tell if you're measuring between the same number of rows at all points. I'd suggest placing a line of sewing thread, in a contrasting color, through all the stitches of the first row of the lace pattern, and again in all the stitches of the first row of a subsequent repeat, at least 4 inches and 2 pattern repeats away (if you only want to knit 2 repeats, knit just the first row of a third repeat as the final row of your swatch). This will let you measure the vertical distance between points that you know to be the same part of the repeat, and over a known number of rows, so you can take an average measurement as you did with the width.

Finally, a note about measuring the overall length of an item when a lace pattern has caused the bottom edge to scallop, and a pattern directs you to knit to a given number of inches. As a general matter, you're going to want to measure from the highest point of the scallop to the needle; measuring from the lower point is going to give you fabric that's too short in places, leading to unintended peekaboo effects.

Unfortunately, the rules for measuring gauge over these specialty fabrics just aren't, and can't be, as clear as those for measuring over stockinette; personal judgment plays a much larger part. You can see why designers are often happier to provide a stockinette gauge swatch! However, it's my hope that this will at least give you some guidance and a place to begin, on which to exercise your judgment in a way that's helpful to you.

Gauge swatches don't lie

No, they don't, and I don't care who told you otherwise. However, I'll concede that they will happily help you lead yourself down the garden path, if you aren't wise to their tricksy little ways. It's easy to get a little fudge factor here, and a little fudge factor there, until at the end you've got enough fudge to open a candy shop, and a swatch that bears no useful resemblance at all to what's actually going to happen with your knitted item.

Here's the game plan: we'll talk about the common ways in which fudge creeps into your swatching, and then we'll go through a really rigorous approach to swatching that will remove as much of the uncertainty as possible, and then we'll talk about when you really do need to go through the whole rigorous process and when it's safe to do something less (and how much less).

Here are some common places where distortion can creep in:

• The swatch is too small
• Edge stitches are distorting the measurement
• You use needles and/or a method different from that you'll use for the project
• You push or pull on the swatch in the measurement process
• The fabric grows or shrinks when washed
• The fabric grows lengthwise from its own weight when worn

You may notice that the first three of those relate to actually making the swatch, and the last three more to what happens with it after that; things can go astray in either the swatching itself or in the measurement after. I'm sure someone can think of some other way swatching can go wrong, but those are the usual suspects, and if you eliminate those, your chances are pretty darn good of winding up with a finished item that's the size you had intended it to be. So let's get on with how to make a dependable swatch and how to measure it in a reliable way.

Make your swatch big enough. It's not so much that a little-bitty swatch causes distortion as that a bigger one averages it out. Look at it this way: if you make a 1/4-stitch error in measuring across 4" worth of stitches, then in a 40" sweater, that's going to add up to 2.5 stitches of difference, and your chances are fairly good that it won't matter enough to notice (that's about half an inch in a worsted-weight sweater); if, on the other hand, you make a 1/4-stitch error across only  1" worth of stitches, then in a 40" sweater you've got 10 stitches of difference, and that could matter a lot. The bigger your measurable area, the more accurate your results are likely to be.

Give it edges, and get away from them. Speaking of measurable area, it's not a great idea to cast on exactly the number of stitches that are supposed to add up to 4". Go look at the edge stitches on some of your items -- you'll notice that often they're a little off in size from the rest of the piece. If you include those in your measurement, you're likely to be off a bit. Also, remember that stockinette curls. What do you think your chances are of measuring fabric without pulling or pushing on it if you're trying to flatten it down so it doesn't roll up? To address both of these issues, cast on enough stitches to make up the number you want to measure across, plus 3-4 for each side to give you a garter or seed-stitch edging, plus another 2-4 for each side to get away from the edging; work 3-4 rows of garter before starting your stitch pattern, and work a few more rows than you intend to measure across, before ending with 3-4 rows of garter again. And bind off the swatch; measuring with it still on the needles can propagate error quite a way down into the fabric. This gives you a reliable measurement area, and makes the swatch behave itself while you measure. Yes, it takes longer to make a swatch like this. An adult sweater in worsted weight averages around 40,000 to 50,000 stitches; you tell me if skimping on a couple hundred of them at the outset sounds like a smart idea.

Use the needles and the process for your project. It's fairly common to find that your gauge shifts just a little bit when you go from one needle material to another; some of us tighten up when working on slippery metal, for instance, while others relax as they get into a speedy flow. Not everybody will experience this, but until you know that you don't, it's probably wise to at least do the majority of your swatching with the needles you'll use for the majority of your project, and then perhaps a smaller swatch with any other needles you might be using to be sure you don't have a visible shift. It's even more important to make sure you're swatching flat for projects knit flat, and in the round for projects in the round. When you knit stockinette flat, you're alternating rows of knits and rows of purls; when you knit stockinette in the round, you have only rows of knits. If you have a difference between the size of your purls and the size of your knits, and very many people do, you will likely have a gauge difference between the two methods. If you want to see whether you've got a difference, and by how much, make two swatches: do the first with about 20 stitches and 30 rows of garter stitch, all knits, and then do another, the same width and length, but this time purl every stitch of every row. The swatch that's larger will be the stitch that's larger for you, and how much larger will give you some indication of how big the difference can be. Unless there's essentially no difference at all, it's probably a good idea to do two swatches for a project that includes both round and flat knitting (for instance, a sweater that's knit in the round to the underarms, and then split and knit flat for front and back), and make sure you're matching gauge in both methods. You may find it helpful to find the needle size that gives you gauge for the round portion, and then change only the needle for the purl rows for the flat portion, so that you're holding the knit stitches the same size as the round portion and matching the purls to those, rather than matching average gauge against average gauge, which could still give you a visible break.

How do you do a round gauge swatch, anyway? There are several possible methods. One is to just knit a tube (Elizabeth Zimmermann recommended making a hat as a swatch for a sweater), but in order to have a 4" flat measurable area on a tube, you need to make it at least 10" around, so that it can be about 5" when folded and you can measure 4" without getting into either fold, or you need to make it more like 6" and then cut it and lay it out flat, which works fine but does give you issues with reusing the yarn. Another choice is to do a faux in-the-round method, which simulates the knit-every-row aspect of working in the round, but allows you to make a flat swatch. The most common of these is the "giant i-cord" method: using DPNs or a circular, you work a row, and then instead of turning, you slide the stitches to the other end of the needle, loosely strand the working yarn across the back of the piece, and work another row in the same direction. With this method, you need to be sure to leave the back strands loose enough that they don't restrain the piece, and edge stitches are absolutely essential, as the end stitches will definitely have odd tension things going on and you don't want to measure anywhere near them. The method I prefer to use is a variant on this. Again, using DPNs or a circular, you work a row, and then slide the stitches to the other end of the needle; instead of drawing the yarn across the back of the piece, however, you leave a big loop (for a 6" wide swatch, you'll want this to be at least 2 feet of yarn) and shift your hold so you're feeding yarn to the needles from the big loop instead of from the ball; you'll be leaving the strand that goes to the ball attached to the right edge of the fabric. Work this row, using up most of the yarn from the loop, and then slide the stitches to the other end again, and work a row normally using the yarn from the ball which you left at the right edge. When you finish, you'll have a flat swatch with a series of loops on the left edge; if you like, you can tie each of these in an overhand knot snug against the edge of the fabric, to make sure the edge stitches behave themselves. There's a certain temptation with this method to try to finely adjust the amount of yarn in the big loop so you run out as you reach the left edge; resist it. If you do that, you'll necessarily be having to change the way you hold the yarn so it's different from the way you normally hold it, and this is likely to cause more gauge shift than what you're avoiding by working a faux-round swatch in the first place. This method does use a bit more yarn than the giant i-cord method, since the final loops are generally a little bit longer than strands straight across the back would be, but it's a little bit nicer looking as a finished product, and keeps the edge stitches under better control, which is why I like it.

Observe your swatch in its natural habitat. Er, I mean, measure your swatch without distorting it in the process. There are a few schools of thought on how to go about this. First, do you use a measuring tape or a hard ruler? Some people like a tape, because with a ruler it can be a bit tempting to just sort of... nudge things a little. I like the straight edge of a hard ruler, though, and I trust myself to resist that temptation; if you do use a tape, you want to hold it snugly extended, rather than letting it lay on the surface of the fabric where it may ripple a bit. The one thing I don't like is the kind of needle gauge that has a window in it: in the first place, these are usually only 2" long, and we already discussed how smaller areas tend to multiply error; in the second place, it just seems way too easy to play tricks on yourself with these, angling the edge to scoot the fabric in or spread it out just a bit. The next question is whether you mark the stitches to be measured or not, and if so, with what? Some authorities advocate placing pins at the points to be measured; others think that this tends to cause some distortion of its own, and I'm inclined to agree. I like to mark off the measurement area, but I use contrasting sewing thread, loosely basted in place, instead of pins. And then, if you mark, exactly what are you marking? Again, there's two ways to do this: you can mark each edge of the group of stitches that should be 4" wide and tall, and then measure to see if it is, or you can mark the left and bottom edge, and then measure to the right and up 4", mark those, and then count. I like to do the first version when I'm matching someone else's gauge, and the second when I'm designing and trying to determine what my own gauge is. Whichever method you determine, it's important to make sure you don't distort the swatch when you measure it. Don't carefully smooth the swatch out, patting it into shape -- you're not going to have someone following you around smoothing your sweater all day. Instead, pick it up and give it a good shake or three, and then drop it flat on the table and measure it as it naturally falls; if you basted in marking strings, you can pull those straight if they got a little rumpled, but don't fuss with the swatch itself. Put your ruler on it, lining the edge up properly, and then let go, so you're not putting pressure on the stitches, and count. Remember what I said about getting away from your borders -- take your horizontal measurement away from the top and bottom borders, and your vertical measurement away from the side ones. Count partial stitches as well as full ones; quarter stitches can make a difference over an adult-sized sweater, and half stitches definitely do. It's probably not a bad idea, either, to pick the swatch up and shake it a second time, and measure it again; you should be able to get the same numbers twice running. Finally, write the numbers down, on something you're fairly sure you can hang onto.

Give it a good wash. Do you really have to wash your swatch? Well, do you intend to wash your finished item? It does you no good at all to make a sweater that fits you perfectly right off the needles, if it's going to grow or shrink several inches the first time you wash it, and it's a whole lot better to find this out with a swatch that took you an hour to make than with a sweater on which you've proudly lavished several weeks of your free time. Some yarns, of course, are more prone to change than others; cotton, for instance, is likely to shrink, and superwash wools are notorious for growing to an astonishing degree. You want to wash your swatch as you plan to wash your finished garment; it's not that informative to know how the swatch behaves with gentle hand-washing and flat drying, if what you really mean to do with the sweater is toss it in the washer with your jeans and dry it in the dryer on hot. Block your swatch once it's washed, if you mean to block the finished item (if it's intended to be a wash-and-wear item, like a kid's cardigan that you know will never get used if you have to do anything more than washer-to-dryer-to-drawer, then skip straight to the post-washing measurements -- those are the ones you're going to have to live with). With some yarns, especially wool, you can play a few little tricks of your own with gauge in this step, blocking a piece to the measurements you want, but you need to use care so you don't introduce distortion here: block the piece and let it dry or cool, but don't measure it in that state -- pick it up and shake it out well and drop it flat, just as you did at the pre-wash stage, and then measure it. This gets rid of any "bounce back" that's going to occur from tension put into the fabric in the process of blocking it out; the measurement you get after this will be one you can rely on. Write these measurements down too.

Let it all hang out. This step won't be necessary in all situtations, by any means, but sometimes it's a good idea to test out what's going to happen to your swatch under the influence of gravity. Thicker cotton yarns, in particular, are notorious for growing under their own weight, and some of the tape and ribbon yarns are also notably bad about this. Designs with enough ease that they're going to hang free of the body are also more prone to stretching; those that lie close to the body also get some support from it. To check out how gravity is going to affect your swatch, hang it up. (Be sure you did the "native state" measurements first!) Clip the top of it to a hanger, or run a knitting needle through the edge, and let it hang; if it's going to be a particularly heavy item (like a cabled cotton cardigan with an intended baggy fit), you may want to weight the bottom as well -- try running another needle through the bottom edge. See how much vertical difference this makes, after leaving it several hours or overnight, and also check to see if there's any horizontal change (there may or may not be). Again, write the numbers down.

Ta da! Once all the washing and the blocking and the hanging are done, you've got a set of measurements that you can rely on. If your final numbers match the gauge you're supposed to get, and you knit to the pre-wash gauge without variation, then after you wash and block your finished item (and if it had a growth issue, let it hang up overnight), barring pattern or knitting mistakes, it should be exactly the advertised size.

Okay now, really, do you have to do all of that every time you knit anything? Nope. That's just the all-out no-holds-barred method. A lot of times it's safe to do less; the trick is in knowing when, and how much less. First, you'll notice that I've referred a lot of times to the making of adult sweaters, and there's a reason for that. Sweaters are both fairly large in scale and fairly sensitive to fit. That means that there's more scope for small errors to be multiplied, and that it's more likely to matter if they are. For a lot of other items, this isn't going to be the case -- for instance, a blanket is big, but does it really matter if it's 6 inches wider than you meant it to be? A hat is somewhat fit-sensitive, but it's generally only about half as big around as a sweater, and has enough negative ease that it may overwhelm an error. It may be okay there to do a smallish swatch, or one that's knit flat even if you're going to work in the round. On the other hand, I wouldn't skip the washing step even for a small item, if it's one of the fibers that tends to shrink or grow a lot -- I have a friend whose superwash wool tam, intended to be blocked on a dinner plate, grew so dramatically it would have fit her biggest Thanksgiving turkey platter. For baby sweaters, you may want to go ahead and do a sleeve, and wash and block that -- it's not so much larger than a decent-sized swatch anyway, and if it turns out the right size, then you're one sleeve ahead of the game. For socks (small and with a lot of negative ease), I'll readily confess that I do a mini-swatch, maybe 20 stitches worth and only a couple of inches of fabric and I knit it flat, although I do knit back backwards rather than turn and purl, which I know to give me personally a closer match to my round gauge. However -- and this is absolutely the key to successful swatch skimping -- I do not have issues with denial, and I do have a willingness to admit that an item really isn't the right size and to rip right back to the start. If you aren't willing to do that, then you'd better check your gauge properly in the first place. For sweaters, the point at which it becomes apparent that there's a problem is further in than I'm willing to risk (and that's not even counting any post-washing issues), so I really do go through that entire process. Yes, it takes some time and effort -- but if I'm going to have an issue, I'd a lot rather find it out after using half a ball of yarn and a day or two, than 15 balls of yarn and a month or two.

What about those cases where I suggested you might want to do two swatches, such as one with circs and one with DPNs, or one in the round and one flat -- do you need to do the full process with both? I don't think so. Usually in this case, what you're trying to be sure of is that you don't have a visible break or measurable size change between the two methods. Therefore, I do my full swatch process with the method that's going to be used for the majority of the project, and then for the secondary method I do a smaller swatch, and check it for apparent compliance with the size and appearance of the larger swatch; if it matches visually in the pre-wash stage, then I assume that it will behave similarly in the post-wash stage, and so on.

One swatching shortcut that I haven't talked about yet is the multi-size swatch. This is a tool to use when you know that you may need to go through several sizes to zero in on the appropriate needles, especially if you expect a size change in the washing, and you want to get all your swatching over with in one session. To do a multi-size swatch, you'll start with a normal swatch, but then instead of binding off, do a short border section, and then change needle sizes, and do the next section of swatching. Your border section can be the same as the lower border stitch, but I like to do a row of drop-stitching (k, YO,YO, across, then drop the YOs on the next row) -- the loose row thus created makes sure that the smaller or larger stitches of one section don't influence the next section. Repeat this process until you've tried out all the needle sizes you wanted to try, and then bind off normally, and wash and block the whole swatch strip. One place where this technique really comes in handy is when you use a particular yarn a lot, at a variety of gauges; you can prepare a "master swatch" this way, and if you've kept careful notes, when you go to use the yarn in a project, you can just consult your master swatch and know what needle size to use for the project at hand. Some people, when doing a multi-size swatch like this, like to build right into the swatch an indicator of what needle size was used -- for instance, when swatching with size 6 needles, on the first row they might do (k2tog, YO) 6 times, leaving 6 eyelets, to indicate that number. Personally, I just use paper price tags attached after the washing process, but do whatever lets you keep it straight.

There is one other place where gauge goes astray that I haven't talked about, and that's gauge drift in the course of a knitting session. Some people experience this as a regular thing -- they may start off tight with every session, and gradually loosen up over the first 15 minutes or so, for instance. For those people, I'd suggest keeping something really simple to start each knitting session with, such as a garter-stitch dishcloth or a simple afghan square; switch to your real project (and do your real swatching) after you've gotten into the groove. Other people may have gauge drift only occasionally, probably due to emotional factors -- a friend once noted that she could tell exactly which parts of a scarf had been knitted during the chase scene of the movie she was watching, and Elizabeth Zimmermann recommended keeping one tight-gauge project, like a sock, expressly for stressed-out knitting. If this is a big factor for you, you probably want to try to do your swatching when you're not stressed, and maybe even make a second, unwashed swatch, to check your knitting against as you work, to make sure you're not experiencing a drift.

Finally, what do you do with your swatch once you've made and measured it? I suggest keeping it, at least while you're working on the project. Having it on hand can be reassuring, if you're not certain that your fabric is behaving as intended, or if you need to remind yourself that it's going to be fine once you wash it. If you need to scavenge the swatch yarn to finish the project, you can always do that at the end. If you don't need to, then saving the swatch gives you a nice record of the project, and if you do another project with the same yarn, a head start on that project.

The measure of our days

Okay, I know it's only been a couple of days since the last post; don't anybody keel over in shock. I spent most of this week in meetings that required minimal attention from me for large parts (though intense attention in the remainder), but in which I couldn't knit, websurf, check email, or read -- but I could, under the guise of "taking notes," write out the outlines for several posts. So we'll have a nice little spate of them while I work through this backlog, and then, I'm afraid it's likely to be back to normal around here. At least until the next round of meetings.

So, let's talk gauge! Now, I don't want to hear any panicking. So many knitters seem to consider gauge a bane, but they shouldn't; if you understand it, it's a really useful tool, leading to things such as sweaters which magically fit even after they've been through the first wash, and socks which stay up and actually fit the intended recipient and not some larger or smaller relation. So, let's talk about why you need to use it, and then we'll get into how to do it right.

Do you have to check gauge? No, not if you don't care if your items turn out the size they were intended to be. That's a bit of a flippant statement, but there's a little more to it than first appears. For some things, it really doesn't matter if they turn out exactly the intended size. It's hard for a scarf, and darn near impossible for a blanket, to be the wrong size; if you knit hats for charity, rather than for specific recipients, it's likely that your output will fit someone who will be happy to have it. For these things, you might do a tad of swatching to check that you like the resulting fabric, but you probably really can safely skip your gauge check, or at least do a cursory one on a minimal swatch instead of going all out. However, I think it's important that you first understand how gauge works and how to check it properly, so you can honestly evaluate the level of risk in skipping or skimping your gauge swatch, and avoid unpleasant surprises down the road.

Avoiding unpleasant surprises, after all, is what checking gauge is all about. It does take time to do a gauge swatch, and even more time to wash and block it. However, it takes quite a bit less time to do that than to knit an entire sweater only to find that "close enough" turns out to be a deal further off than you thought, and that the sweater intended for your 6-year-old is sized more for his 16-year-old cousin -- except that the latter has rather gotten over a taste for pullovers with dinosaurs.

There are two statements about gauge that I particularly hate to hear. One is the knitter insisting that she doesn't need to swatch, because she "always gets gauge," and the other is a knitter lamenting that she can't understand why her project doesn't fit, because after all, she used exactly the yarn and the needles called for in the project.  There is no such thing as a knitter who always gets gauge, and even if you were one, how on Earth would you know that the pattern writer was?  What if they're a loose knitter, and have gone down a size from the average -- if you follow them exactly, are you going to enjoy the results?  No, of course you aren't. The needles specified in the pattern are a suggestion, a place to start your swatching, but that's all they are. It's the pattern gauge that you need to match, not the needles. If you know that you're about average, and the pattern gauge matches the ball-band gauge for your yarn, then starting with the recommended needles is probably a good place to begin, but that doesn't mean you don't need to check.

I'm going to set aside, for the moment, how to do a really accurate gauge swatch and get a good accurate measurement from it. Instead, I'm first going to assume that you've done a swatch and measured it accurately, and it's off. How do you use this information?

There are only two cases for how it can be off. The first case is where you have too many stitches in your 4" measurement area -- for example, it's supposed to be 20st/4", and it's 22. Too many stitches means that the individual stitches are too small -- you need to make them bigger, so you need a bigger needle. In this case, where you're about half a stitch per inch off, one needle size might do it; if you instead had 23 or 24 stitches, so you were off closer to a full stitch per inch, then skipping up two sizes would probably short-circuit some of the iteration.

The other case, probably obviously, is when you have too few stitches -- for example, you've only got 18, or 16, when you're supposed to have 20. Too few stitches means the individual stitches are too big -- you need to make them smaller, which means a smaller needle.

By the way, if you've got the idea that there is anything wrong or strange or odd about having to move several needle sizes to get gauge, get over it. There's not, and it's not in any way a reflection on your knitting abilities. Even if a variation seems unusual, it may reflect characteristics of the yarn you're using, or it may reflect characteristics of the pattern writer or test knitter. As an example, I am a fairly tight knitter; I have a friends who's a fairly loose one. We've knit socks from the same pattern, using the same yarn, but I used a needle that's four sizes bigger than the one she used -- and our socks came out the same size, because we both matched the pattern gauge, using the needles we personally needed to do that. And we know knitters who are looser than she is, and tighter than I am, who would have even more variability. There's nothing wrong with that. And to clear up a common misconception, our fabric is not any different -- hers is not more dense, mine is not more airy. Our fabric looks and feels exactly the same, because our stitches are exactly, precisely the same size: that's what matching gauge means. It's the size of the stitch that matters, not the size of the needle used to produce it.

One more comment on adjusting needle size: sometimes it can be tricky to match both stitch and row gauge, particularly if you're at a gauge that's open or dense for the yarn. When this is the case, work toward the appropriate stitch gauge first; when you've gotten there, try moving one more needle size to see what that does to your row gauge. Sometimes it'll change the row gauge without changing the stitch gauge appreciably; needle diameter affects row gauge more directly than stitch gauge, which is to a large part determined by the thickness of the yarn itself. If you still can't match both, generally it's easiest to match stitch gauge and adjust for row gauge; the exceptions would be things like sideways knit sweaters, where row gauge is the dominant measurement. We'll get to how you make those adjustments a couple of posts from now, so just hang tight on that one. If you're not too far off, it may be possible to block to the correct gauge, and you'll want to test that by blocking your swatch and seeing how it performs; we'll get to that in the very next post, when we talk about how to actually do the swatch.

And that's a good note on which to close: next up is the making and measurement of a fool-proof gauge swatch.

Curl and bounce

There's one lesson that a very large number of knitters learn the hard way, fairly early in their knitting careers: stockinette curls. There's nothing quite like watching your laboriously produced scarf roll right up in a tube, is there? A fair number of knitters, however, learn this lesson like rote memorization, and don't really understand why it happens, or what if anything can be done about it. So let's take a look at the whys and wherefores of curling stockinette, and then just for grins, we'll look at why the springiness of ribbing is a related phenomenon.

These phenomena have their root in the shape of the knit stitch. Knit stitches are a little bit longer across their front faces than across their back faces in the horizontal direction, and a little bit longer across their backs than their fronts in the vertical direction. If you want a visual aid, take two pieces of paper that are the same size and shape, and cut a sliver off the top of one and the side of the other, and then tape them together along the edges, with the shorter and wider piece in front. See how the sides roll back, and the top and bottom roll forward? When you place a whole bunch of knit stitches together, that's how the resulting fabric wants to behave.

Stockinette is the fabric you get when the RS of your knitting is all knit stitches; if that statement doesn't make sense to you, you'd probably better look back at the post about knitting stitches as they appear, or perhaps the one about reading charts and how that relates to fabric. It aligns all the bigger front faces side-to-side, and all the bigger rear faces top-to-bottom, and it curls right up.

So, is there anything you can do about it? In a way, but to a certain extent this is an inherent property of the fabric, and so what you need to do is work with it or work around it, because you're not going to be able to make it go away.

One way to work around it is to substitute out a balanced fabric. This is a fabric where the distribution of knits and purls as viewed from the RS is fairly even, with about as many of one as the other. Some examples of balanced fabric are seed stitch, garter, and ribbing, and that's a reason that these are often used at the edges of a project. There are plenty of other choices for balanced fabrics, though, and a stitch dictionary that uses charts will make these easy to pick out. One source for these is the Walker Treasury Project -- this is a very laudable effort (in which you can yourself participate!) to produce nicely photographed color swatches of every stitch pattern in the Barbara Walker Treasury books, and to catalog them in ways that make them easy to find. Those patterns listed as being similar or the same on the wrong side will usually be balanced fabrics, since in order for it to look the same, it has to have the same stitches showing.

Another way to work around curling is to note that it only occurs on the free edges, and to eliminate those edges. For instance, you've probably not noticed horizontal curling on sweaters -- it's not that they're not curling, but that there's no unrestrained edge where the curling can manifest, so each stitch just flows placidly into its neighbor. If you want a stockinette scarf, one way to do it is to knit the scarf as a tube and then flatten it, so you've got a double thickness with stockinette on each side; it'll still curl at the ends, but you can seam those together and then they'll be restrained also.

What about just sort-of restraining the edge, as with a border? Well, that can have some success. I repeatedly see the suggestion to slip the first stitch on each row; I've never noted that as having any significant effect, but I can't stop you trying it for yourself. Adding a border of several stitches is somewhat more effective, but it's not a slam-dunk guarantee; a fabric that's strongly curling will just roll the border right into the middle of the tube. Here's a lovely example of a seed stitch border being totally overwhelmed, even though each side of the border is nearly a third as wide as the center of the scarf.

What about blocking? Again, it's of some limited utility. It works the best in wool, since wool has enough memory to hold a shape fairly well. It's more of a temporary solution, though, better for making something behave long enough to get your seams done than to make a scarf lay flat when worn; since the curl is an inherent property of the fabric, it'll eventually reassert itself. A possible exception is acrylic that's been "killed" with heat -- heated close to its melting point, so that a permanent change in the fiber shape is effected. This changes the entire character of the fabric, though, making it limp, and it's a permanent change, so it's not exactly the same as blocking, and often the change is not wholly desirable. It's definitely something that should be practiced on a swatch before applying it to a whole project.

One thing that does make a definite difference is that looser fabrics curl less. Tightly knitted fabrics have more energy, more tension in the stitches so that each acts more strongly on its neighbors. At one extreme, the scarf I showed above was knitted quite tightly for the yarns used, and that certainly contributed to the way it overwhelmed the border; at the other extreme, stockinette based lace, which is knitted exceedingly loosely for the yarn size, lies mostly flat, especially after blocking.

You may find success by combining several of these notions. For instance, a fairly loosely knitted scarf with a moderate seed-stitch or garter border, especially if it's made of wool and blocked, would probably lie fairly flat for a decent while, although you might need to periodically reblock.

Okay, so what's that I was saying about how ribbing is a related phenomenon? I did just mention that ribbing was one of the balanced fabrics, and it is. Ribbed fabric as a whole lays flat, and it doesn't curl up top to bottom at all, if it's a fairly even ribbing (knit ribs and purl ribs close to the same width). It does, however, draw in horizontally, and it does so in an elastic way, rebounding to at least some degree after being stretched out. It's this rebounding effect that's related to the curl of stockinette.

Ribbing is, fundamentally, adjacent strips of stockinette and reverse stockinette. What gives it its texture is the way these stitches act at the transitions -- the strips of stockinette want to curl backwards, while the strips of reverse stockinette want to curl forward, and they wrap around each other at the interface (I think of this as being like a lock seam in metal roofing, but this may prove nothing more than that I'm a geek in areas other than knitting, too). This gives you the texture with the knit ribs popping forward and the purl ribs receding, and it's also what gives you the elasticity, as the curl flattens out under tension and reasserts itself when tension is released.

This interaction helps explain why 2x2 ribbing is the most elastic for most knitters -- it gives the most scope for this interface to act on the stitches. In 1x1 ribbing, every stitch is trying to engage in this curling relationship on both sides, which sets up competing forces -- the same stitch is trying to curl to the right and to the left. In 2x2 ribbing, by contrast, every stitch has one side engaging in this, but the other is next to a stitch of the same type, in a relatively stable arrangement, so the curling side can curl as far as it wants to in just one direction. To go toward the other extreme, 3x3 ribbing has two-thirds of its stitches in this same type of relationship, but the center stitch of each strip is just sitting there, not trying hard to curl in either direction, so the fabric as a whole has less energy than the 2x2 ribbing has, and this tendency only increases as the ribbings get even wider. There are individual variance in stitch formation that can affect this, so there may be some knitters for whom a ribbing other than 2x2 (possibly one of the slightly unbalanced ones, like 2x1 or 2x3) is their stretchiest, but for most of us, that's the springiest.

The interaction theory also explains why corrugated ribbing (2-color, as discussed when we got deep into stranded knitting) is less elastic -- it's the curling between the columns that causes elasticity, and the stranding behind the purl columns pops them forward and doesn't let them recede as they would in a single color, nor let the knit stitches curl back as far as they'd like. This keeps it permanently somewhat flattened, and you lose a lot of the bounce.

The spring of ribbing, being a related phenomenon, can be affected by some of the same factors as curling. For instance, just as tight stockinette curls more, tight ribbing bounces more and pulls in more effectively. Blocking can also flatten it, and that's generally something you want to avoid when blocking items with ribbed edges -- either don't flatten it in the first place, or re-wet it after blocking the rest of the piece and scrunch it in nice and tight and let it dry that way.

I encourage you to play around with some swatches -- learn how the different factors affect the curl of stockinette, and find out which ribbing draws in the most for you.