Before I was overtaken by creeping Deadline Disease, I promised three things – some general principles on translating flat knit patterns to kntting in the round; a list of increase and decrease stitch equivalents for right and wrong side rows; and why someone would want to knit either way. That’s a bit much for one day.  Even for long-winded me.  Let’s start with the most subjective – why one would choose one style over the other.

To start off with I’l make some bold statements that kntting historians are welcome to debate.  Knitting garments in the round was far more prevalent during most of knitting’s history.  Think of socks, hats, and traditional sweaters – most of them were worked that way.  Seamed knitting (knitting in the flat) appears to have taken off around the time that knitting made the transition from traditonal garment to fashion wear – roughly about the same time when written patterns became more prevalent.  I believe it did so in order to conform better to fashionable clothing’s styles, tailoring and fit, and to allow greater reproduceability of results among an audience already familiar with sewn garment methods.

Why knit in the round?

• Simple, quick garment construction
• Simple pattern writing for boxy shapes
• Minimizes purling on stockinette pieces
• Eliminates (or minimizes) seaming
• Takes advantage of knitting’s inherent elasticity for fit
• Easier production of stranded colorwork
• Easier production of texture and lace patterns that require increase/decrease manipulations on every row
• Ideology (there’s a strong lobby of knitting purists that advocate it as the most natural or historically-connected way to knit)
• Fewer pieces to lose or match in gauge
• Flashing (color stacking of space dyed yarns) works better
• Easier to ravel back and add length or replace worn sections.
• If using one circ, fewer needles to use than knitting in the flat

Why knit in the flat?

• More complex and tailored garment shapes can be achieved
• Pattern writing for complex shaping is simpler for designers trained in garment construction theory
• Easier to adjust tif you need to produce complex garments in multiple sizes
• Seams can add structure and strength to a piece
• Smaller individual units can be more convenient to produce
• Intarsia colorwork is easier to do
• Eliminates cast on row join challenge, especially on fine gauge garments that employ large stitch counts
• Eliminates the at-join color jog problem (to be specific it substitutes a match stripes at seam problem)
• Can be easier to measure garments in progress to determine compliance with required dimensions
• In flight modification for fit can be simpler because problems can be spotted after one piece is made, and the entire garment does not need to be ripped back to make adjustments
• Easier to add width while the garment is in progress
• No scary DPNs or circs needed.  Fewer needles to loose compared to DPNs, no DPN juncture ladder problems
• Straight needles are less expensive than circs, and multiple lengths needn’t be purchased.  No need to have multiple diameters of the same size needle on hand to accommodate tubes of various diameters (required unless many DPNs, a two-circ or the oversized circ method is used)
• Easier to block pieces before final finishing stage.

I’ve probably left the reason why you chose one method or the other off these lists.  Feel free to add it in a comment.

Each method has its own strengths and shortcomings.  Each has styles for which it is particularly suited.  And each can be manipulated to do most of what the other style is better suited to do.  You can make faux seams on something knitted in the round. You can do stranded colorwork on something knitted in the flat.  You can add shaping to an in the round piece through planned and judicious use of increases and decreases to mimic the fabric manipulations of darts or tailored seaming. 

n many cases there’s logic in the choice of one method over another.  Dale’s Norwegian stranded pieces are perfectly suited for knitting in the round.  They employ strategies like steeking to place shaped collars or introduce other construction features.  I’m sure people have done it, but I wouldn’t want to translate one of them for knitting in the flat – there’s nothing to be gained by doing so.  Complex tailoring like this from a discontinued Berroco pattern would be a headache to render in circular knitting.  BUT logic doesn’t always prevail.  I have seen commercial patterns for stranded sweaters that ARE knitted in the flat.  I’d take a hard look at them to see if I could produce them in the round.   Likewise, I’ve seen all sorts of contortions and cutting done to circular knit patterns in the name of making them less boxy. Again I’d have to take a closer look to see if using an alternate approach was better.

While you’re far more likely to see in-the-round direction today than you were 25 years ago when I started knitting, you’ll still find that many fashion oriented magazines and yarn manufacturers booklets offer up more patterns for knitting in the flat than they do the other method.  I’m thinking pubs like Vogue Knitting, Adriafil’s Dritto& Rovescio and most of the modern European books, plus the old Phildar, Aarlan, and Pingouin books. 

As in so many things, ideology does play a part.  You can find books written in the 1940s through 1960s that sniff at knitting in the round, calling it "peasant work" or noting it in passing as the dreary ancestor of  more modern applications.  And you can find books written by knitting revivalists that excoriate the torture of imposing tailored seamed construction on a medium that has so many virtues in its most simple form.  I’m dogma-agnostic.  I use whichever method is best for me to produce the project at hand.  Which brings me to the real reason why I think patterns are written one way or the other:

• It’s what the pattern’s author/designer is most comfortable with

Overall though, the motivation to change something from flat to circular knitting is far more common than yesterday’s case.  Anything with a rectangular construction and minimal shaping  is a natural – especially sweaters with drop shoulders.  The exceptions might be sweaters in yarns that are prone to biasing or stretching, or ones in particularly flimsy yarns or knit in very open textures.  In those cases the structural integrity imposed by firm seams might be crucial to garment drape and longevity.  I’ll look at this in more detail in the next overly long post.

HolidayInnEloise saw yesterday’s post and sent in a question via eMail.  She’s not a fan of knitting in the round and wants to know if the Cabin Fever pattern I am using can be worked flat back and forth on two needles.  She’s a bit confused by how to change a top down in the round sweater into a more standard format.  I’ll try to answer.

First, while there are exceptions, most things that are knit in the round can be parsed to knit in the flat either in part or in their entirety.  Many but not all things that are knit flat can be knit in the round.  Like with everything in knitting, knowing the logic behind the design can help you make the transition.

To go from round to flat, there are two big things to take into consideration – the garment shape and the knitting texture or colorwork pattern used.  In order to transform a seamless thing into a knit-flat thing, seams must be introduced.  Sometimes figuring out where to put them can be a challenge.  The sweater I’m working on now is of very simple construction, but even it presents a challenge.

Following the original pattern, I started by making a tube for the collar.  After that there was a row of increases to add a bit more scope for the next step – symmetrical increases on either side of four diagonal lines from collar to armpit (the raglan lines).  The piece proceeded more or less like a poncho or capelet until the depth of the raglan increase lines accommodates a loose shoulder toarmpit fit (along the way one is directed to stop making increases otherwise the piece would grow too wide, but that’s a minor quibble).  After the capelet type shoulder yoke was done, the sleeve stitches were slipped off onto holders, and with the addition of a few extra stitches under each arm, the remaining stitches for front and back were worked as a big tube.  After the body tube was completed, the sleeve stitches were retrieved from their holders and worked out to the cuff, with some decreases at the bottom center to remove bulk.

Thinking through the logic top down we’ve got a challenge right off the bat.  To me, the turtleneck collar is a lost cause.  No one wants seams on the inside of an already bulky turtleneck.  That feature will probably need to be knitted in the round on DPNs.  After that we’ve got the capelet yoke area, formed by the faux raglan style increases.  The most obvious choice to flat-ify this part is to turn the unibody capelet back into four pieces – a front, a back and two sleeves – to be seamed together along the raglan lines as real raglans are. 

To do this, I’d probably take the stitch count from the after-collar increase row just before the faux-raglan feature starts, and deconstruct it back into those four pieces.  For example, I’d take the recommended stitch count for the front and add two selvedge stitches.  These will be eated up as seam allowance when the garment is sewn together.  Then I’d follow the instructions for the front area, working my increases as directed, but doing them TWO stitches in from the edge (one stitch to make the decorative line, plus the selvedge stitch for sewing up later).  At the post-raglan point where the piece is long enough according to the original pattern, (when the sleeve stitches were slipped onto holders), I’d cast on one quarter of the total stitches that are to be added at the left edge of my piece and half at the right, then continue knitting to the specified total length. (Remember the original just gives one number to be added between what is the front and back.  I need to divide that by four and put one quarter at each edge of my front.  The remaining stitches would be added to the back.)  I’d end up with an object shaped like a house with a Mansard roof.  The other pieces would be made the same way.   The back would be identical to the front.  The sleeves would be similar but predicated on a smaller initial cast-on.  I’d assemble the thing by first sewing the raglan seams, then the seam from lower hem to cuff.  Finally I’d pick up my provisional stitches at the neck and add the collar.

But I’m not bound to do this piece top-down.  I could also knit it in the flat bottom-up.  I could divide the ending body stitch count in half (adding a selvedge stitch at either side to make up for a seam allowance) and working on half the stitches – knit the front and back flat up to the undearm.  At that point, I’d cast off the stitches that were added in the original just after the sleeves were slipped off, removing one quarter at each edge.  Then I’d start the raglan shaping.  But in a bottom-up piece, that shaping will be formed by decreases rather than increases.  To preserve the simple yarn over detail of the original I’d need to do a bit of playing.  First I’d work an inch or two plain (in the original the raglan area ended before the piece was long enough to reach the underarm).  Then I’d work the left edge of my piece K1, SSK, YO, SSK; and the right edge K2tog, YO, K2tog, K1.  The K1s are the selvedge stitches.  The [decrease] YO [decrease} unit adds up to a net loss of one stitch, with one of the decreases making up for the decorative YO. This won’t be exactly analagous to the original because the stitches framing the YOs will be heavier, but it will be close enough to preseve the general appearance.  When I had the requisite number of YOs and my stitch count was equal to the post-collar neckline count (plus two for the selvedge stitches), I’d slip everything onto a holder and begin the next piece.  The back would be made the same way. 

The sleeves would start off with the final stitch count for the cuff, and along the way add a stitch at either edge right or left (I’d probably do M1 increases two stitches in from the edge at the ends of a row just to keep things neat when I seamed).  I’d knit the same length below the raglan line that I did on the front and back, then plunge into the same logic to make the decorative raglan line itself.  Once there were the same number of YOs in the raglan line of the sleeve I would guess that the sleeve’s upper dimension matched that of the front and back exactly (the open holes are more than decorative, they make the thing easier to count and measure).  I’d slip the remaining stitches onto a holder and seam all the raglan lines.  After that I’d sew the side seams and pick up and knit the collar in the round on DPNs.

Another alternate method would be a hybrid between flat and circular knitting.  I’d work the front and back up to the point where the extra stitches are cast off, and the sleeves until they met the body to cuff measurement.  Then I’d arrange them all on a circular needle, and finish out the couple inches of plain work followed by the raglan capelet yoke part in one big knit in the round piece.  I’d still have seams to sew under the arms and from the armpit to the hem, but that would be it.  (To do this, I’d eliminate the selvedge stitch that I added to the raglan edges in my first all flat knit alternative.

The second factor that might affect the transition from round to flat knitting is texture.  In this particular sweater it’s not an issue.  I’ve got miles of stockinette, a couple of rows of purl welting, and some K2, P2 rib.  That’s it.  In the round stockinette is "never stop knitting."  Flat stockinette is alternate rows of knit and purl.  No big translation problem there.  However, if I had a texture pattern and the original was knit in the round I’d need to do the mental shuffle, turning knits to purls and vice versa for the odd numbered wrong-side rows. Having a texture pattern in chart format makes the right-side/wrong-side translation easier.   I might even need to adjust the row on which the pattern starts so that the bulk of my increases and decreases end up on an easy-to-manipulate right side (knit) row.  Yes, they can be done on a wrong side (purl) row, but then the problem of which increase or decrease when done on the purl side mimics standard right side row stitches intrudes.  All exist, but many are puppy awkward to do (more on this tomorrow). 

Colorwork in the original can also present a challenge.  Many people find stranding easier in the round.  You always have a right-side row facing you, and it’s very simple to see the design build as you knit.  Stranding can be worked flat too, with every other row a purl row.  It’s harder to see the pattern on the purl side, and some folks don’t enjoy manipulating multiple strands while making purls.

So there you have it.  The first thing to do is to examine the original in the round pattern and see where seams would go.  The second is to look at the texture or color pattern used to see if it can be comfortably translated.  Once the individual pieces are determined, the cast on numbers can be derived from the original pattern (half for the front, half for the back, plus an optional selvedge stitch).  Then it’s just a matter of knitting and seaming.

Tomorrow I’ll look at going in the other direction, what the equivalents of standard increases/decreases are if you do them on the wrong side, translating flat knitting into circular knitting and why one might want to work in one style or the other.