(Pre-)shrinking linen

Some fabric dealers say their linen will shrink up to 10-15 % over the first few washes. Now, most extant linen garments have probably been washed several times, so if we use their precise measurements to sew an exact copy from unwashed linen, we'll end up with an item that will be smaller than the original once it's been washed—and probably disproportionate too, as fabrics shrink mostly in the warp direction. Did linen always shrink this much?
 
In the book "Seventeenth-Century Women's Dress Patterns 1", edited by Susan North and Jenny Tiramani, there's a plain 17th century kerchief; just a hemmed linen square. And it is almost square, at 27" x 26 3/8". Assuming that it was cut as a perfect square, that's just a 2.3 % difference in shrinkage between warp and weft. (Obviously we don't know if it really was cut perfectly square, nor how many times it was washed, but still...)

A washing experiment

I came across a good washing guide for linen at the blog In deme Jare Christi. I soaked and washed some of my linen fabrics according to these instructions, washing at 60 degrees Celsius - i.e. 140 F, or "hot" (I substituted an enzyme-free detergent though, as that's how I've been taught to wash linen). To my surprise, the shrinkage varied a lot.

My washload contained some newly bought linen (with today's fairly open weave of tightly twisted threads), as well as some vintage linen yardage I'd found at the thrift store (closely woven tabby weave, likely intended for bedsheet). Neither had been washed before. The new linen shrank 6 % lengthwise; the vintage linen just 2.5 %. (Both fabrics will probably shrink a little in the next wash as well, though much less than in the first wash.)

I think it's really interesting that the shrinkage of my closely woven vintage linen is so close to the hypothetical shrinkage of the 17th century kerchief! But what might be causing the different shrinkage rates?

Left—vintage linen; right—contemporary linen.
Note the difference in weave!

Linen production today

The process of turning flax into linen has changed quite a bit over time.

For the past several decades, if not more, linen thread has usually been spun from 'cottonized' fibers. This is a chemical process that makes the flax fibers more cotton-like (thinner and shorter), so they can be spun on cotton machinery. Possibly this process also affects the fiber's tendency to shrink; while there's documentation about cottonization online, I haven't been able to find anything on how long it's been done, or how it affects fabric shrinkage.

The threads in todays fabrics are spun very tightly compared to vintage linen. This is a consequence of cottonization—shorter fibers have to be spun more tightly to make the thread strong enough.

Tightly spun threads are also the reason why today's linen fabrics seem to be stiffer and have a more open weave than vintage linen does. Could it be that the threads in a traditional, closely woven, tabby are so tightly packed that they simply don't have room to shrink as easily as in today's more open tabby?

The shrinkage rate could be related to the bleaching process. In the past, linen yardage was bucked as part of the bleaching process, before it was sold. Today's bleaching seems to use hydrogen peroxide (bucking used lye, which is sodium hydroxide). It's possible that there's a difference in how much heat is involved, or to which degree the fabric is agitated in the process, that might have caused more of the shrinkage to happen already during the bleaching back then.

At some point the fabric is also stretched and calendered, so it looks neat and straight and smooth. Could it be that the fabric is stretched more today, causing it to shrink more later?

As far as I know, any of these process stages might affect shrinkage, to some extent.

Bobbin lace shrinks a lot

The classic textbook on Swedish bobbin lace making, "Knyppling" by Sally Johansson, says that lace always shrinks more than fabric. She estimated that a medium to fine weight bobbin lace has to be about 5 % longer than the edge it will be sewn to, to compensate for this (5 to 10 % for coarse lace). Therefore, she suggests preshrinking lace in this way (my translation):

Place the lace on a soft ironing surface. Cover it with a thin, wet, cloth, and hover a hot iron over so it touches the cloth lightly. The steam that is formed shrinks the lace. Treat the lace in this manner a couple of times, then place it on a flat surface to dry.

This method produces a lot of steam right around the lace; if bits of the lace extend beyond the cloth you actually see them moving as the lace shrinks. I doubt that a steam iron would work as well.

So, lace shrinks a lot more than fabric. I think this, too, points to that the openness of the textile plays a large role in shrinkage.

What about the sewing thread?

For the future, I'll continue to pre-shrink new linen by washing it once before cutting and sewing. Then the finished garment should shrink about as much as a garment from un-shrunk vintage linen would.  But will I have to shrink the tread too before sewing? Maybe, maybe not.

Sharon Burnston, who's got a great set of webpages about 18th century shifts in the US, says on her site that when she stitched with linen lace making thread, it shrank slightly in the wash, leaving her hand sewn seams puckered. Ouch! She mentions using running backstitch and slip stitch for her seams. (Running back stitch was commonly used on linen in the 18th century US, and also in France, as it's described in Garsault's L'art de la lingere. I haven't used this stitch myself, as it isn't documented to Northern European linen sewing as far as I know.)

Based on her experience, my sewing preshrunk linen with un-shrunk Bocken's linen lace making thread should spell disaster. Strangely though, it has worked fine for me. But that may depend on my choice of stitches; it's the same type of thread as in those bobbin laces that shrink 5 to 10 % more than linen fabric, so I know the thread in itself does shrink! But I've been using hemming stitches and the counter-hem seam described in "Seventeenth-Century Women's Dress Patterns"; a felled seam where both lines of stitches are done in hemming stitch. The zig-zag nature of the hemming stitch means there's a bit of 'give' in the seam, which running backstitch doesn't have. Also, you can choose to leave the stitches a tiny bit looser (no loops, just don't pull them quite as tight as you want the finished seam), to allow for shrinking. And for some reason I've got into the habit of waxing my thread thoroughly (running it over the wax 2 or 3 times, not just once), possibly that might counteract shrinkage?

If you want to try preshrinking linen, I suggest you start by preshrinking a small square (say 10 cm or 4"), test sew with your chosen seam and thread on it, and then wash it again to see how it behaves. It would also be possible to preshrink linen thread with steam as described above. You could cut several pieces of a convenient length, steam them all together a few times, and when they're dry you just braid them loosely and tie a thread around each end to keep them in order. Then you can pull threads out as you need them.

BTW, it's worth remembering that linen and cotton shrink when washed at e.g. 40°C (105 F or "warm") too, it just takes more washes before it adds up to the same shrinkage as when it's washed at a higher temperature.