In bookbinding, the term “swell” describes the thicker area of a book block at the spine due to the addition of sewing thread. It depends on factors detailed below, but a binder or conservator generally does not have control over all these variables. Knowing how to estimate the amount of swell that will develop is one of the most important aspects when planning to sew a book, since it corrolates to the degree of round and the shape of the shoulders that a book will end up with.
Different binding styles need different amounts of swell. Too much swell creates a textblock that is unstable, squiggly, and difficult to back. Too little swell and there are insufficent shoulders which are necessary for some styles of binding. In all case binding structures, there is more leeway with amount of swell, and as long as it is not excessive it will be sucessful; with bound books the tolerances are much tighter. For example, if you are resewing an existing text block it is critical the new shoulders exactly fit the original boards.
Softer, multi-ply threads afford much more control of swell while sewing as compared to a hard modern thread. Currently, I primarily use threads sold by Colophon Book Arts, including the Colophon Best Blake Thread and the Londonderry Linen Lacing Cord #4. Both can be deplyed if desired to make them thinner. I also prefer to support smaller, specialist bookbinding supply companies. If you are using standard threads, 25/3  is a reasonable starting place.Â
Many early binding structures—even up to the late 18th century—manipulate the shape of the boards to fit the spine, rather than the modern fine binding tendency to fit the boards to a 90 degree shoulder.  This makes it easier to fit a wider range of boards to a given swell. Obviously, this is not an option for many binding structures.
There is no formula, instead these ten aspects need to be considered:
1. THICKNESS OF THREAD. Thick thread (or more plys) = more swell. Thin thread (or fewer plys) = less swell. Although not ideal, the thickness of thread can be changed during sewing if too much or too little swell develops.Â
2. HOW HARD OR SOFT THE THREAD IS. Hard thread does not flatten in the signatures = more swell. Soft thread flattens in the signatures = less swell. A compressible thread gives more control. It is often advisable to untwist hard modern threads a bit to make them softer by running them through your fingernail and thumb, and let them relax. Â Waxing thread also makes it harder, so I generally avoid it if possible. Sometimes excessive kinking and twisting comes from using too small of a needle. Softer thread can fray more during sewing, though.Â
3. THICKNESS OF THE TEXT PAPER. Thick paper absorbs more thread = less swell. Thin paper absorbs less thread = more swell.
4. HOW HARD OF SOFT THE PAPER IS. Soft paper absorbs more of the thread = less swell. Hard paper absorbs less thread = more swell. It is easier to control swell with softer paper. Guarding the spine will increase swell. Washing and resizing can also affect how much swell develops. Swell can also be adjusted before sewing by beating or otherwise compressing the sections.
5. HOW MANY LEAVES ARE IN EACH SIGNATURE. More leaves can absorb more thread = less swell. Fewer leaves = more swell.
6. HOW MANY SIGNATURES THERE ARE. More signatures = more swell. Fewer signatures = less swell. Some binders like to visualize this by wrapping the thread around a pencil the same number of times as there are signatures.
7. SEWING STYLE. All-along, two-on, three-on, etc. All-along produces the most swell, more “-on” sewing styles = less swell. Packed sewing produces more swell due to a small overlap of thread. This can be controlled, to produce naturally packed sewing, which has one length of thread on the cords for each signature.
8. SEWING SUPPORTS. Tapes, cords, thongs. Tapes produce the least swell, cords and thongs slightly more since the thread can overlap slightly inside the signature. Supports also differ in the amount of adjustment that can be done after sewing, ie. how much the thread can move on the supports during consolidation and backing. A professional sewing frame, such as the Nokey makes this easier.
9. HOW MUCH CONSOLIDATION IS PERFORMED DURING SEWING. More consolidation during sewing= less swell. I have often observed students sewing identical text blocks, with identical thread, end up with significantly different results. A loaded stick, or knocking down stick can help with compression, although some people prefer to use a bone folder or wedge shaped piece of wood.
10. TIGHTNESS OF SEWING. Tighter sewing makes a thinner book before pressing. Looser sewing can develop due to improper tensioning or too large of a needle. A book sewn too tightly can develop a “banana” shape, thinner at the kettle stitch. Even tension is crucial.
Best practice: sew with the thickest thread possible.
Peachey Conservation 
Nice article! May I reproduce it in my group at. FB? Thank you very much!
Sure, thanks for asking and I’d appreciate a link back to this blog. Happy sewing, Jeff
Excellent summary.
One more factor: shoulder shape. If a book is backed to nominal 45 degrees (actually 135 degrees) it can accommodate far more swell than a book backed to 90 degrees. With grooved-joint styles, the 45 degree shoulder is actually more natural than 90 degrees. With tight joint styles, the spine edges of the boards can be shaped to match the backing angle. The angle can be anywhere between 90 and 45, but when you get too low, say nominal 30 degrees (actual 150 degrees), many new problems occur. Using this effect, I’ve often managed to deal with over 50 per cent swell on books with many thin sections and a lot of mending.
With 45 degree joints, the joint height can be fine-tuned to the exact board thickness by putting the book and the boards together into a press for a while, after the book is backed but before the board edges are shaped or the spine linings are put on. This would crush 90 degree joints, but with lower joints it just eases them down a bit, say from 45 nominal to 40 or 35 nominal.
And another variable to consider: thin thread slices into the sewing stations, especially with a spine that moves a lot. For this reason I never use thread thinner than 30/3, and use 18/3 as my medium. Hard thread is more apt to attack the sewing stations than soft. But the amount of swell can be fine-tuned by adding a bit of beeswax to the thread, one or two pulls over the lump, giving intermediate between 30/3 and 18/3, and between 18/3 and 12/3, and a bit more thickness beyond 12/3.
There is no particular difficulty in getting the swell right, just a brutal fact to swallow: measure the swell when the book is sewn. If it is not right for the style, pull the book immediately, before your courage wanes, and resew from scratch. Resew two or three times if you must. Good sewing is the necessary, inescapable basis for good forwarding.
Thanks for the compliment, Tom. It means a lot coming from you, since you wrote what I consider the best article on the movement of the book spine:
http://cool.conservation-us.org/coolaic/sg/bpg/annual/v06/bp06-01.html
I agree that often (especially for pre- board shear era books) the shoulder is less than 90 degrees, and mention making the boards fit the book, rather than the modern working method of making the shoulder fit the 90 degree board. This is one of the considerations for how much swell you might need.
I rarely wax thread, and pay attention next time I do, and see if it acts as a thinner one. Since it becomes harder, I imagine it might cancel out the thinness?
Point well taken about resewing, thankfully that hasn’t happened to me in quite some time!
In my experience waxing the thread makes it act thicker. I think the unwaxed thread squishes down more easily when the text block is consolidated, while the waxed thread is harder and stays closer to round. But that’s just how it works for me; this kind of detail tends to vary from binder to binder with variations in other factors (for instance I consolidate heavily, usually by pressure rather than impact; with less consolidation the added wax might have different effects.)
I find that if I pull the thread more than twice over the wax, the extra wax just crapes off on the first few sewing stations. On the other hand, it I cut a length of thread too long and it starts to fray when sewing near the end of the length, a pull or two of wax will bring it back to the hardness it had near the beginning.
Many thanks for the plug. I value your opinion highly.
Yep, this makes sense in terms of waxing causing the thread to retain the round and not flatten or unplug as easily. What doesn’t really make sense is why waxing started in the first place. I should recheck my 19th c. manuals, but offhand think it started to be mentioned in the later 19th c. And certainly most early sewing thread does not seem to have any evidence of wax. Possibly a response to harder machine made papers with more size?
I suspect that waxing thread may have seemed so obvious that no one mentioned it until very late, even though they did it; rather like using leather dressing for maintenance. It was used in a number of crafts. Certainly, saddlery and harness sewing was done with very heavily waxed thread, yet here again I don’t know how far back there are mentions of waxing. According to Salaman p. 120, Holme’s Academy of Armoury (1688) talks about shoemakers keeping wax in cold water in summer to keep it hard enough to use. I wouldn’t trust myself to be able to tell the difference between waxed and unwaxed in a sewn book, even one I had sewn myself, not even the difference between the wonderful completely unwaxed Swedish linen that Bookmakers used to sell and the heavily waxed Barbour’s of the eighties, with Colophon somewhere in between but definitely waxed. Despite the way the Swedish unwaxed linen would disappear into the section, it was a pain to use since it frayed and fluffed very quickly, so that I found I had to use lengths about two-thirds of Colophon, or even less.
This may be true, waxing as too obvious to mention. Yet in these objects which are exposed to elements such as shoes, gloves, harnesses, sails, etc, the waxing does help repel moisture in the finished product, and there is (hopefully!) less of a need for this in books.