let me explaiKnit

I'll take the combo

Since we've been discussing different styles of knitting, this seems like a good place to talk about combination knitting, especially as it will also tie back to the recent discussion on decreases.

First, what is it? In the common Western styles of knitting, stitches sit on the needle in a particular orientation, which I discussed when we were talking about decreases; the leading leg, which is the one further to the right and closest to the point of the needle, sits in front of the needle, while the trailing leg, further to the left and further from the needle point, sits behind the needle. In most Eastern styles of knitting, stitches sit in the opposite orientation, with the leading leg behind and the trailing leg in front. In combination knitting, stitches sit in the Western orientation after being knit, and in the Eastern orientation after being purled, and it is this combination of orientations that gives it its name.

Combination knitting is done by wrapping the yarn around the needle clockwise (as viewed with the tip of the needle pointing at the viewer), rather than counterclockwise, when purling. This changes the orientation of the stitches thus produced -- it basically gives them a little half twist.  Then, on the subsequent row, the stitches are knit through the back leg, and that gives them a half twist in the opposite direction, resulting in stitches that are untwisted and back in the Western orientation.

Why would someone want to do this? Well, most of the practitioners of combo knitting started off as Continental-style knitters (and some think they still are, which we'll get to in a moment).  However, Continental purling can be kind of tricky; the reason for this is that the pull of the yarn, forward and upward to the left, is exactly the direction that makes the yarn want to jump away from the tip of the needle instead of going neatly through the stitch being purled. Because a combo knitter wraps the other way, coming between the right and left needles instead of in front of the right needle, the pull brings the yarn against the right needle instead of away from it, making it easier to scoop through, and many knitters find it faster and more comfortable to their wrists and/or fingers. Because this is a primary reason to do it, most combination knitters do carry the yarn in their left hands, like Continental knitters do; although there's no reason why a knitter with a right-hand carry couldn't wrap clockwise, there's less of a reason for them to do so as a matter of course.

The motion of the combination purl is a very close analogue to the motion of the Continental knit stitch, which also has the yarn coming between the two needles. Many people who experience "rowing out" (a discernable difference in size between rows of knits and rows of purls) find that when knitting combination, this issue disappears.  Priscilla Gibson-Roberts, in her book Knitting in the Old Way, suggests that this is because the path taken by the yarn in the Continental purl is inherently longer than the knit stitch, whereas the combination purl is exactly the same.  I don't agree; I think the difference is executive, not structural, and the fact that some people's purls are tighter, not looser, certainly would seem to indicate that the difference isn't inherent.  However, that doesn't mean it's not real, and this method can in fact make a difference and produce tighter purls; it's just why it makes a difference that I disagree with.

Now, I mentioned that some combination knitters think they're Continental knitters.  Combination method is actually extremely easy to do by accident, especially among self-taught knitters.  Wrapping the purls clockwise looks like the way knits are wrapped, and makes the purl stitches easier to form, so it seems right, and because the open orientation of the Eastern stitches actually facilitates knitting through the back loop, many accidental combo knitters produce untwisted knitting, which cannot be differentiated after the fact from knitting produced by Western methods. The difference in orientation is subtle, and not easily noticed if you don't know to look for it.

Accidental combo knitting tends to reveal itself in a few common ways, depending on whether the knitter is compensating for the orientation on the knits or not.  If the knitter is not compensating, and is therefore producing twisted stitches, she may discover she's knitting this way when she compares the texture of her stitches to somebody else's, when she tries to figure out why her work biases (slants) in one direction persistently (this is especially noticeable on ribbing), or when she tries to figure out why she's persistently short on row gauge, since the twisted stitches pull up shorter than untwisted ones. If she is compensating, it's trickier, because the finished work looks exactly like the finished work of a Western-method knitter, in almost every particular.  Generally, an accidental combination knitter of this type will figure it out when she does a project in the round and suddenly begins having twisted stitches where she never did before, or when she first moves to projects where the direction of the decreases matter, and her decreases are slanting a different way than she expects. In other words, she figures it out when she reaches one of the points where the combination knitter needs to compensate, and she doesn't know to compensate because she doesn't realize she is a combination knitter (or often even that there is such a thing).

That brings me to the discussion of the areas where a combination knitter does need to compensate for her style of knitting, and the fact that patterns and instructions are written for the Western knitter whose stitches are always in Western orientation.

Combination knitting works extremely well in any stitch pattern where stitches are purled on one row or round, and then knit on the next, or vice versa. This obvious includes flat stockinette, but it also includes flat ribbing, and garter and seed stitch in the round. It does not work so well for stitch patterns where the same stitch column is knit (or purled) on consecutive rows; this includes stockinette and ribbing in the round, and garter and seed stitch flat. The reason it does not work as well in this circumstance is that you don't have the two compensating pieces working together. Purl stitches sit in the Eastern orientation, and are still sitting that way when you come to purl them on the next row; knit stitches begin in the Western orientation, and must be worked through the front, not the back, to keep them untwisted. For knit stitches the compensation is fairly simple; simply knit through the front loop instead of the back one. For purl stitches, there are a couple of choices -- you can either purl in the Western manner, with the yarn wrapping counterclockwise, or you can keep your clockwise wrapping but purl through the back loop on subsequent rows, which is actually not as hard as it may sound, since the back loop in this orientation is the right one.  In fact, that's really the key -- to keep your stitches untwisted, always work through the right, or leading, leg, regardless of whether it's located in front of or behind the needle.

The other major area of compensation for a combination knitter is decreasing. The combination knitters knit-row decreases slant the opposite way from those of the Western knitter, and we'll take a look at why that is. When a combination knitter prepares to knit two stitches together, she is starting with two stitches, each of which is sitting on the needle in the open, Eastern orientation, and she is going to knit them together through their back legs. I hope that sounds familiar from the first decreases article -- it's exactly what the Western knitter has when she does an SSK. And, of course, this decrease leans to the left. If the combination knitter wants a right-leaning decrease, she needs to slip two stitches as if she were going to purl them through the back loops, replace them on the left needle, and then knit them together through the front, which is sort of a combo SSK, and the same structure as a Western k2tog. It's reasonable to think of this as "flipping" the two types of decreases, and many combination knitters do.

The fact that the combination k2tog is exactly the same structurally as the Western SSK gives rise to a neat little trick which the Western knitter (whether Continental or English style) can use to make her SSKs neater, with some advance planning; on the row before the SSK will be worked, work the two stitches which will be involved in the decrease with a clockwise wrap (whether knitting or purling), so they sit in the Eastern orientation, and then they can simply be knit together through the back loops without slipping, which will give you a tighter SSK because there's less manipulation tugging on the stitches. And, of course, the combination knitter could employ the same trick to set up her right-leaning decrease; work the two stitches with a counterclockwise wrap, and they'll be in the Western orientation and set up for a Western k2tog.

There are a few other, less common, areas of compensation, where the difference that the two methods make is subtle.  For instance, to make a twisted stitch, the Western knitter works through the back leg of the stitch. A combination knitter can make a twisted stitch by working through the front loop, but the twist will be in the opposite direction.  This will rarely matter, but she should think about whether it does. As another example, when doing yarnovers, the combination knitter may wish to change their orientation to match the orientation of the other stitches on the row; if she does not, she'll want to make sure that she works the stitch on the subsequent row through the right leg, whether it's in front or in back, so the yarnover doesn't twist closed. Combination knitters need to have their heads in the game, and work to develop an understanding of why stitches are executed in the way they are and whether their method makes a difference to that execution, for the best experience. I'd highly recommend Annie Modesitt's book, Confessions of a Knitting Heretic, for learning more about the structure of knit stitches and knitted fabric, for both combination knitters, and Western knitters who simply want to learn more about why knitting works the way it does; it's a great resource.

Out in left field

Despite my passionate defense of the right of lefties to knit mirror fashion if they want to, that actually isn't what I teach lefties who come brand new to the craft. Instead, I teach a variant on unmirrored Continental, which puts even more of the motion into the left hand than the standard variety. Today I'm going to talk about how that works, and then I'll talk a little bit about why this is what I teach.

In this variety of Continental, the role of the right hand is to (a) hold the right needle steady, and (b) ... there is no (b). This is a completely left-dominant method, and the right hand is just along for the ride. It's so left-dominant that I usually begin by demonstrating it with the end of the right needle held in the crook of my elbow, and my right hand completely off the needle.

The actual variation from standard-variety Continental is pretty small, and just consists of moving the left needle to achieve the same relative positions of the needles, rather than moving the right one. It's subtle, but the movements of knitting are for the most part subtle; it makes a substantial difference.

To form a knit stitch, hold the right needle steady, pointing to the left and angled slightly upward. Move the left needle so that the first stitch is hooked over the tip of the right needle, which should, of course, enter the stitch from its left side in front, and emerge from the right side in back. Bring the yarn across the tip of the right needle as usual for the Continental method. Then, instead of moving the right needle tip to pull the new stitch through, twist the left needle back slightly to pop the old stitch over the tip of the right needle and the yarn held against it, which forms the new stitch. Pull the left needle tip out of the old stitch.

To form a purl stitch, hold the right needle steady, again pointing to the left but more straight across than angled. Move the left needle so the first stitch is hooked over the tip of the right needle, which should enter on the right and emerge on the left, in front of the left needle. Bring the yarn in front of and over the right needle tip, as usual for the Continental method. Then, instead of drawing the new stitch through with the tip of the right needle, twist the left needle forward slightly to pop the old stitch over the tip of the right needle and the yarn held against it. Pull the left needle tip out of the old stitch.

See, not a big difference. But try it, and you'll feel that your left hand is doing all the work.

The other place where you'll want to adjust which hand does the work is for the long-tail cast-on. Again, you're going to hold the right needle still and move the left hand; instead of going up the thumb, down the index finger, and down the thumb, you'll hook the outer thumb strand over, and then bring the index-finger strand up from beneath and lift the entire thumb loop over.  That sounds much more complicated in writing than it actually is to do; in practice it's easy, and it's fast. It's actually faster than moving the needle, in fact, because the steps of bringing the index-finger strand and the thumb loop up can be done in a single rocking motion; I always do my long-tail this way.

The foregoing explanation is probably not enough for an absolute novice who's never knit before to understand this method; it's not really meant to be. It's directed more at the knitter who is preparing to teach a left-handed friend, or the mirror knitter who is interested in experimenting with unmirrored knitting, either of whom will fully understand what knit and purl stitches are intended to look like, how a long-tail cast-on works, and so on. If you're a brand new knitter and want to experiment with this, contact me, and I can provide you with a few more resources and some more involved explanation.

One note about this method: it's not quite as fast as standard even-handed Continental. I think that's inherent, however, in any method where the work is done by one hand, rather than splitting it between hands, because simultaneous motion cuts down on the duration for the sequence of motions to be completed. Personally, I find it slightly slower than even-handed Continental, but slightly faster than English, where I have not quite mastered not letting go of the right needle with I throw.

Now, why do I teach this instead of mirrored knitting? There are several reasons. As we previously discussed, there are certain cases where mirrored knitting requires modification, and it can take some analysis to figure out exactly where those changes are applicable. While I don't feel that figuring that out is a monumental task, it is an additional task that the unmirrored knitter doesn't have to undertake. Also, there is the don't-knit-backwards reaction that pretty much every mirrored knitter seems to encounter at one point or another; skipping the opportunity to engage in such confrontations seems to have its benefits. And finally, there's a more subtle point about the way knitters tend to learn the craft these days; instead of being taught by our mothers and grandmothers, we're mostly picking it up from a variety of disparate sources, a bit from a book, some from online, perhaps a class or just being shown something in a sit-and-knit session at our LYSes. I think mirrored knitters face a little bit more of an obstacle in learning this way, and may have some difficulty getting help, as potential helpers look at the way they knit and say, "Oh, I can't help you with that," instead of taking the time to analyze the problem and see whether the different method even impacts the answer. Again, it's by no means an insurmountable obstacle, but it is an obstacle, and I believe in offering new knitters the choice of whether or not to confront it.

I am not dogmatic about teaching left-dominant Continental; I present the two options, explaining the differences, and what I see as the potential ups and downs of each, and then suggest that we start with the left-dominant Continental, and see how that works out. If I had a student who strongly wanted to try mirrored knitting first, I'd teach it; so far, that hasn't happened.

Left to your own devices

Today let's talk about left-handed knitting. Specifically, let's talk about what I'm going to call "mirrored" knitting, which is the type of knitting where stitches begin on the right needle and are worked off onto the left one.

"Mirrored" is a pretty good term for this, because it is exactly what you'd see if you help up a mirror to the work of a knitter who knits from the left needle onto the right one; in fact, many knitters who knit this way learn new techniques by holding a mirror up to a picture or video. I think calling it "left-handed" knitting is potentially confusing, because many people use that term for knitting in the majority direction using the Continental method, which is popular for left-handed knitters but also employed by many right-handed ones. Calling it "knitting backwards" is also potentially confusing, because many people use that term for the method used to purl back without turning the work, which is also done working from the right needle onto the left one, but forms the stitches slightly differently (I like to call this "knitting back backwards" myself, as I feel that's most evocative of how this method is used); it's also offensive to some, to the extent that "backwards" implies something wrong about the method. "Mirrored" lacks this connotation, and avoids ambiguity.

Let me preface the main discussion by stating that I am not left-handed. I'm a nearly-ambidextrous righty. I can knit mirror fashion, however, even though it's not my preferred style, and in any case I don't think my personal handedness is actually relevant to my ability to analyze the issues faced by mirrored knitters. I simply believe in full disclosure.

Many mirrored knitters feel somewhat attacked by other knitters, having been told at one time or another that they are knitting "backwards," or "wrong," or that they will have to reverse and re-write every single pattern ever, or some similar assault.  These criticisms are just plain wrong. I very rarely say that about something in knitting, but I am saying it now. Mirrored knitting will occasionally require some slight adjustment to an unmirrored pattern, but frankly, lefties are used to adapting to a right-handed world; most of them can handle the challenge of adaptation very well, and all of them are entitled to choose whether they wish to take it on, without being insulted or ridiculed for the choice.

And that brings me to my main topic.  I'm not going to show you how to do mirrored knitting; there are other resources for that. This page has some great illustrations of the steps for the basic knit and purl stitches, although she does call it "knitting backwards," which as we've already discussed, I think tends to be confusing. Boye produces a learn-to-knit book, called "I Taught Myself Knitting," which contains a good basic section on mirrored knitting; there are other books out there as well. And there is always the mirror method; holding a mirror up to videos, such as those on KnittingHelp, works fine if you learn well from videos. Instead of teaching you how to do the method, what I'm going to explain is the various adaptations that may be required when using unmirrored patterns.

Many patterns will require no modification at all. That's because many patterns, especially for garments, are symmetrical: armholes come in the same distance on each side, necklines are centered, sleeves are identical, and so on. Following the instructions stitch by stitch produces a garment that is technically reversed, but you can't tell, because everything on one side is balanced by the other side.

What about decreases? That seems to be a sticking point for the don't-knit-backwards crowd; I recently had someone respond to my advice to a mirrored knitter by vehemently insisting that all decreases needed to be reversed, under all circumstances. That just isn't true. If a pattern uses paired decreases, and is otherwise symmetrical, it can be knit without modification.

Let's look at why that is. As a hypothetical, we'll consider a stockinette sweater with feathered decreases at the armholes. A typical instruction for the upper front might thus be "k1, k2tog, k until 3 sts rem, ssk, k1." When an unmirrored knitter works this row, she begins at the right edge of the fabric, knits a single stitch, makes a right-leaning decrease, knits until three stitches are left, makes a left-leaning decrease, and then knits the final stitch at the left edge. What happens when a mirrored knitter knits the same row? Well, when a mirrored knitter makes a k2tog, it leans to the left, and when she makes a SSK, it leans to the right; I've already given you the tools to figure out why that is. Therefore, she starts at the left edge, knits a single stitch, makes a left-leaning decrease, knits until three stitches are left, makes a right-leaning decrease, and knits the final stitch at the right edge. Looky there -- both knitters have a right-leaning decrease on the right edge, and a left-leaning decrease on the left edge. Looks good to me.

What about unpaired decreases? Well, that doesn't fit the "otherwise symmetrical" criteria, and there you may wish to make a change. Let's consider a hat pattern, where evenly-spaced k2tog decreases are used to create a spiraling pattern at the crown.  When an unmirrored knitter follows such a pattern, her hat top will spiral to the right; when a mirrored knitter follows it without modification, hers will spiral to the left. If she wants her hat to match that of the unmirrored knitter, she'll have to change each k2tog to SSK or another right-leaning (for her) decrease. On the other hand, does it necessarily make a difference which way the hat spirals? If you don't care, there's no need to make a change.

That's going to be the basic rule for anything that's asymmetrical: if the reverse shaping doesn't bother you, there's no need to change.  For some things it's going to matter; cardigans with buttons, for instance, ought to button right-over-left for women, left-over-right for men, so you'll want to make sure you're putting the buttonholes on the correct side.  If they're picked up and knitted after, that may be a very simple change; if they're knitted as a piece with the fronts, it may be more extensive.

The easiest  way to reverse an asymmetrical pattern to match your direction of knitting is simply to go through and write out each line backwards, from the last instruction to the first.  However, this will cause your decreases to be reversed as well, and as we saw from the symmetrical sweater example above, that's not desirable; therefore, you'll need to replace the unmirrored right-leaning decreases with one of the decreases that's right-leaning for you, and the same for the left-leaning decreases.  The first time or two that you do this, or for very complicated pieces, it may be helpful to draw a couple of sketches; mark one with the shaping for the unaltered pattern, and then mark a second with where the shaping will fall with your alterations.  This will help you to produce the intended effect.

A second area that's tricky for mirrored knitters is cables. In order to produce a cable where the stitches held in front slant to the left, an unmirrored knitter will hold the first-half stitches in front while knitting the second-half stitches. A mirrored knitter, however, approaches the cable from the left, so will need to hold the first-half stitches in back while the second-half stitches are knit, to produce the same slant. The reverse is true, of course, for cables that slant to the right.  The general rule, then, is to hold the first-half stitches in the opposite direction of that given in the instructions. However, be careful about applying this too slavishly! Let's look, again, at the case of the symmetrical sweater, and this time we'll have a cable at each side, spiraling in opposite directions. Our directions might look like this: k12, p2, C6F, p2, k to marker, p2, C6B, p2, k12. As the unmirrored knitter follows this, her first cable, near the right edge, is held in front, and spirals to the left, while her second, near the left edge, is held in back, and spirals to the right. As the mirrored knitter follows it, and comes to the first cable, the general rule would say that she should hold her cable in back; however, keep in mind that she's coming from the other direction! The first cable she comes to is the one near the left edge, and she wants it to spiral to the right.  Therefore, she's going to need to hold it in front, just as the unmirrored knitter did. The mirrored knitter working with cables needs to spend a few minutes thinking about what the pattern is calling for and why, and analyze whether adjustments are necessary to produce the same effect.

One more area that can be tricky for mirrored knitters is charts.  On some level, charts are great for mirrored knitters; they show the design as it should appear when knitted, and can be followed in the direction of knitting, left to right on the RS rows, and right to left on the WS rows. This is true, up to a point, but there are a couple of potential pitfalls.  First, if you're going to work in the direction of knitting, you'll want to adjust the key slightly. If the key shows a decrease symbol slanting to the right, calling for a k2tog, the mirrored knitter working in her direction of knitting will not want to do a k2tog there; for her, that slants to the left. Instead, she'll want to use a decrease that slants to the right for her, such as an SSK, wherever she encounters that symbol. The same is true for cables; the chart symbol corresponding to a C4F, for instance, should clearly show a cable that spirals to the left, and a mirrored knitter, who will approach that cable from the left, will wish to hold the stitches in back to produce that effect. The second pitfall is the use of patterns that are partially written out and partially charted.  For both charts and written directions, it's possible to do either of two things: follow the directions in the order written for unmirrored knitters, or reverse the order and read them backwards/left-to-right.  Either of these is fine, although as we've already discussed, adjustments may be needed in some cases.  However, you should generally make the same choice on both the charted and the written part; if you're going to read the chart left-to-right, you should also read the written directions backward, with necessary adjustments, and if you're going to read the written directions forward, you should work the chart right-to-left as unmirrored knitters do.

As we've just seen, mirrored knitters do face a few challenges that unmirrored knitters don't.  However, I don't think any of them are anywhere close to insurmountable. In fact, because they lead the knitter to analyze her work, and to understand what she's being asked to do and why, I think they ultimately lead to being a better knitter. And that can only be a good thing.

Leaning in the right direction

Most knitters who've moved beyond the very basics probably have grasped the general point of why you'd want to use both right-leaning and left-leaning decreases, which is so that you get a similar appearance on both sides of a symmetrical piece of knitting.  However, many knitters are much less sure about which direction of lean you want where, and what difference it makes, so let's take a look at that.

When using paired decreases, there are basically two choices available to you.  You can use decreases which lean into the center of the piece, or decreases which lean away from it.

When you use, for instance, a right-leaning decrease, the appearance of the two stitches decreased is that the leftmost stitch is on top, leaning toward the right, and the rightmost stitch is neatly tucked up beneath it.  If you place this right-leaning decrease on the left edge of a piece of fabric, and do a series of decreases which are each an equal distance from the edge (such as, for instance, knitting until 3 stitches remain, and then doing k2tog, k1), you will get a very clear line coming up the left edge of the fabric, as each of the leftmost decreased stitches lines up on a diagonal.  The same is true, of course, of a left-leaning decrease placed on the right edge of a piece of fabric.  This type of decrease placement, where decreases lean toward the portion of the fabric which is getting smaller (usually, but not always, the center), is called a "strong" decrease.  It's commonly seen on places like sock toes and gussets, where the strong line produced is a part of the design.

When you use, instead, a left-leaning decrease on the left edge, the decreased stitches don't align visually.  With this type of decrease, the rightmost stitch is on top, and lies at an angle to the line of the fabric edge, so there's no continuous feature. Again, the same is naturally true of a right-leaning decrease on a right edge. This type of decrease placement, where the decreases lean away from the portion of the fabric that's getting smaller, is called a "feathered" decrease.  It's commonly seen on armholes and sleeve caps, especially on commercial knitwear, and can be used anywhere that the strong diagonal line isn't desired.  It's still visible -- there really aren't any invisible decreases, in contrast to increases -- but it's much softer.

[Edit: It occurs to me that I have a bit more to add here, about why the strong and feathered decreases form. When decreasing, one column of stitches appears to survive, and one to tuck beneath it and disappear. Strong decreases form because the decreases are placed so that the same column is always the "survivor"; with feathered decreases, in contrast, the column that disappears in one decrease is the one that was the "survivor" from the prior decrease. In the example above, where "k2tog, k1" is done at the left edge of the fabric, the second column from the edge is always the "survivor" from the k2tog decrease, and thus forms a strong decrease; if "SSK, k1" were done at the same location, the 3rd stitch from the edge would be the survivor, and would become the 2nd stitch from the edge, to be eaten up in the following decrease.]

As for which right-leaning or which left-leaning decrease to use, that's up to you. As discussed when we were talking about the single knit decreases, SKP and SSK are structurally the same, and so are k2tog and KRPR; the difference, if any, comes from your execution, so use whichever you're the happiest with.

What about the centered double decreases? Again, it's a matter of whether you want the strong line. The S2KP produces a very prominent line of single stitches, which appear elevated from the surface of the fabric.  Sometimes this is a great design feature and can be used as an accent.  If that's not desirable, then the softer, more feathered, SK2P should be used.

I hope you've found this interesting; if so, you may want to take a look at Eunny Jang's wonderful piece on the structure of lace, part of her incredible Majoring In Lace series.  She shows a number of the beautiful effects that can be achieved by the way decreases are aligned, and it's fascinating.

Somebody to Lean On

Last time, we discussed the single knit decreases, and what happens to the stitches that creates the lean. This time, we're going to look at the single purl decreases, and the double decreases.

The simplest of the purl decreases is p2tog. The needle enters through the right side of the first stitch and emerges from the left side of the second stitch; the yarn is pulled out through the right side of the rightmost stitch, resulting in the rightmost stitch lying on top. This is, therefore, a left-leaning decrease. However, there's a catch! This decrease is usually worked on the WS of the fabric, and therefore, while it slants to the left on the side where it's worked, it slants to the right on the RS. It's most often used at one end of a WS row to balance a left-leaning decrease on the other end of a RS row.

A more complicated purl decrease is the SSP, which is done by slipping two stitches, one at a time, as if to knit (not as if to purl), putting them back on the left needle, and then purling them together through their back loops. Let's look at what that does. Slipping the stitches in this manner changes each of them to the open orientation, but leaves the leftmost on the left and the rightmost on the right. This means that you're still working through the right legs of the stitches when you purl them together through the back loops, so you aren't twisting them. You enter from the left side of the leftmost stitch, so the new stitch is pulled through the right stitch first, and then the left stitch, which leaves the left stitch on top, and therefore creates a right-leaning decrease; however, as with the p2tog, this stitch is usually worked on the WS, which leaves you with the appearance of a left-leaning decrease on the RS.

Another possible purl decrease is the p2tog-tbl.  As with the SSP, this decrease is done by coming into the back loops of the stitches from the left, and produces a decrease that's right-leaning on the WS and left-leaning on the RS.  However, because the stitches aren't reoriented before they're worked together, they are twisted; this decrease resembles SSP in the same way that k2tog-tbl resembles SSK.

Now, let's look at the double decreases.  There are double-decrease variants on the single decreases, such as k3tog and SSSK; these lean the same way as the single-decrease varieties they resemble, although the lean is generally even more pronounced because more stitches are involved.  Since we've already discussed the mechanics of the single decreases, there's no need to discuss these types of doubles.  However, there are also a couple of centered double decreases, in which the outer stitches lean together into the middle stitch.

The first of these, slip2tog-knit-p2sso, or S2KP, produces a prominent single stitch in the center, with the other two stitches tucked neatly underneath it; let's look at how that works.  The first step in this decrease is to place the needle into the first two stitches together knitwise, as if you were going to do a k2tog, and then slip them to the right needle. This twists the stitches around into the open orientation, but does it as a group; the rightmost stitch ends up further to the left, and vice versa. This is a definite contrast to the one-at-a-time slipping in SSK, for instance. The third stitch is simply knit. Now, the two slipped stitches are passed together over the new stitch; this means that the stitch which was originally on the right ends up in the middle of the stack, and the stitch that was originally in the middle ends up on top, and that's what gives you that prominent central stitch.

The other centered double decrease is slip-k2tog-psso, or SK2P (note the different position of the 2 in this abbreviation). In this decrease, the first stitch is slipped knitwise, putting it into the open orientation.  Next, the second and third stitches are knit together, and just like a regular k2tog, this puts the leftmost stitch on top.  Finally, the slipped stitch is passed over, which leaves it at the top of the stack. The center stitch is on the bottom, and recedes, giving this decrease a softer appearance than S2KP.

I hope these last couple of posts have given you some useful tools to understand what's going on with your knitting, and to assist you in knowing which way your decreases are going to lean.  Next up, we'll discuss briefly how to choose which decrease to use when.

Lean on Me

Let's talk about decreases! All single decreases, where two stitches become one, lean to one side, either the left or the right.  Double (or greater) decreases, where three (or more) stitches become one, may lean to one side, or the disappearing stitches may lean into the center. What we're going to look at today is why the lean happens, and how to figure out and remember which decreases lean which direction.

Let me make it clear what "left leaning" and "right leaning" mean with respect to decreases. What this means is that the tops of the two (or more) existing stitches appear to lean to the specified direction.  Therefore, for a left-leaning decrease, the tops of the two decreased stitches will be displaced to the left, with respect to the bases of those same stitches.

Also, this is one of the fairly rare posts where the manner in which you knit makes a difference: I am specifically discussing standard Western knitting (not combined), done with stitches beginning on the left needle and being worked onto the right. If you knit combined, or you knit from the right needle onto the left one, this discussion will not fully apply to the way you knit, and I'm not going to make it more convoluted by trying to specify how it's different.  You may, however, still get something out of this -- I hope you will -- at least as a conceptual framework you can then apply to what you do.

If you don't know how to do some of the decreases I mention, or would just like to see some knitted-up examples or a video of how to do them, check out this page of decreases from KnittingHelp.com. Videos of the centered double decreases,  but not knitted examples, can be found on the abbreviations page, listed as cdd and sk2p.

We'll start with the simplest decrease, k2tog. For a k2tog, you insert your needle in two stitches together, knitwise, and simply knit both as if they were one stitch.  This is a right-leaning decrease. Let's look at what you actually do in detail, and it will become apparent why that is. When you put your needle into the two stitches, you are starting on the left side of the leftmost stitch, and the needle tip emerges from the right side of the rightmost stitch. Your working yarn is wrapped around the needle, and then drawn back through the two stitches.  The end result is that the new stitch emerges from the left side of the leftmost stitch, and the two stitches are stacked up around the base of it, with the rightmost stitch underneath and the leftmost stitch on top of it. Because the new stitch wants to come straight up, this means that the tops of the prior two stitches are pushed to the right, and that's what gives you the right-leaning appearance.

Let's look at some of the left-leaning decreases. The simplest of these to execute is k2tog-tbl, where the needle is inserted through the back loops of two stitches together, and they're worked as one. In this case, the needle goes from the right side of the rightmost stitch and emerges from the left side of the leftmost stitch; the new stitch is pulled out the right side, and the stitches stack up with the leftmost one on the bottom and the rightmost one on top -- leaning to the left is exactly what you'd expect, and in fact what happens.  This is also true for the SSK decrease, which is the same as k2tog-tbl except that the stitches are individually twisted first.

It looks rather like we're building up a rule where the direction the needle is inserted from gives you a decrease that slants in the opposite direction. But wait -- what about sl1-k1-psso (SKP)? There, you're slipping the first stitch, and knitting the second, and the needle is inserted from the left; shouldn't that give you a right-slanting decrease?  Well, no.  To see why, let's look at what actually happens to the individual stitches as you execute the SKP.

First, let me back up just a trifle and talk about stitch orientation a little bit.  In normal Western knitting, stitches sit on the needle so that the leg further to the right is in front, and the leg further to the left is in back.  I tend to think of the Western orientation as "closed", with the "face" of the stitch -- the part that will become the surface of the fabric -- turned away from the tip of the needle.  The function of slipping a stitch knitwise, then, is to "open" it, turning its face toward the right; slipping it purlwise leaves it closed.

So, let's look at exactly what the mechanics of the SKP are. You slip the first stitch knitwise; that leaves it sitting on the right needle in the "open" orientation, with its left leg in front. Now, you knit the second stitch normally, which means you're going into the front, or through the right leg, as usual. You're coming in from the left, but you've only got one stitch involved, so it lies straight, not slanting either way.  Now, you're going to take that first stitch, and pass it over the new stitch you just made. This gives us the reason for the "opening" step -- it allows the stitch to lie flat when passed over the next stitch.  If it was still in the "closed" orientation, the right leg would cross over to the left side of the stitch, giving you a twisted top stitch, with a loopy appearance similar to that of k2tog-tbl (and the closed orientation of both stitches in k2tog-tbl is, of course, why it has that loopy appearance). Now, you've got two stacked stitches sitting around the base of the new stitch, and the right stitch is on top, so that if you consider the two decreased stitches as a group, the new stitch is emerging from the right side of the group.  And that is the key to remembering which decreases lean which way -- they lean away from whichever stitch ends up on top.

Now, just as a point of interest, let's compare what happened with the SKP to what happens in an SSK.  With the SKP, you ended up with the leftmost stitch face up, and the rightmost stitch, also face up, on top of it, around the base of the new stitch. When you do an SSK, you slip two stitches knitwise, one after the other.  This opens each of them up, while keeping the one that was further to the right still on the right, and the one on the left still on the left. You then knit them together through the right legs, which are now in the back.  The new stitch is drawn first through the leftmost stitch, and then through the rightmost stitch, so that they both end up laying face up around the base of the new stitch, with the rightmost stitch on top.  Doesn't that sound familiar?  That's right -- structurally, SSK and SKP are identical.  So why do they sometimes look a little different? Execution.  It's easy in an SKP to stretch out that passed-over stitch, which makes it a little looser and a little more prominent.  Working on the tips of the needles, and concentrating on not tugging too much as you pass the stitch over, will improve your SKP.

Speaking of improvement, what about the so-called "improved" SSK, where the second stitch is slipped as to purl?  That means that the second stitch stays in the closed position, which means it's worked through the left leg rather than the right one, and that twists it. This is why some people achieve a flatter result; the twisting tightens the lower stitch, which pulls on the upper stitch and keeps it neat and small. However, the twisted lower stitch is sometimes visible, particularly in very smooth yarns like mercerized cotton; I've even found it to sometimes cause a small hole under the decrease, when knitting a smooth yarn at a gauge that's tight for the yarn.

And, just as a point of interest, what about the knit-return-pass-return decrease, which is shown at KnittingHelp as another right-leaning decrease? As you may suspect, it bears the same relation to k2tog that SKP bears to SSK -- in other words, it's structurally identical, but the manner of execution tends to loosen the passed stitch a bit and make it a little more prominent.

I think that's actually enough for a single post.  I'm going to stop here, and we'll pick up next time with a look at purl decreases, and double decreases.