Journal / Signatures
Counting Signatures Before You Touch Anything
It is the least interesting step in any repair, and skipping it is the fastest way to make the second half harder.
A disbound novel from the 1930s sits on the bench in eleven loose gatherings, the spine glue long gone brittle, the old sewing cut in three places by whoever tried to fix it last. Before a single needle touches this book, before any decision gets made about thread weight or which folds still need support, someone has to count. Not read — count. Every signature, in the order it left the loose pile, checked against itself and against what should be there. It is the least interesting hour of any repair job, and it is also the hour that determines whether the next six go smoothly or turn into a slow unpicking of assumptions made too early.
The Step Everyone Wants to Skip
Collation has no craft romance to it. There is no needle, no thread, no satisfying click of a fold seating properly against a straight edge — just a stack of gatherings and the plain arithmetic of making sure they are all present, in order, and belong to the same book. It is easy to see why it gets rushed. A binder eager to start sewing looks at a pile of signatures and sees a book that is basically fine, just loose. But basically fine is exactly the assumption collation exists to test, and a repair built on top of an uncounted stack carries that uncertainty into every later step, where it is far more expensive to find.
What You're Actually Counting
A signature is a folded unit of pages — four, eight, sixteen, sometimes more — gathered together with its neighbors to form the block of a book. Older printed books usually mark each one with a signature mark, a small letter or number tucked into the bottom margin of the first page of the gathering, put there centuries ago for exactly this purpose: so the person assembling the book by hand could check the order before binding. Reading those marks in sequence is the fastest way to confirm nothing is missing, reversed, or borrowed from another copy entirely — a real hazard when repairing a book that has been rebound once already, sometimes with a stray gathering from a different printing mixed in. The same care that goes into reading a broken fold for its history applies here: the marks are recording something, and skipping past them means guessing at information the book was already offering for free.
A Collation Checklist, In Practice
On the bench, the process is unglamorous and repetitive by design. It goes roughly like this, gathering by gathering:
- Locate the signature mark, or count the leaves if the book predates or omits marking.
- Confirm the fold direction and grain match the neighboring gatherings.
- Check that no leaf is a later, tipped-in replacement with a visibly different paper stock.
- Note any gathering that is thinner or thicker than its neighbors — a sign of a missing or extra leaf.
- Record the running order on a slip before moving to the next gathering, rather than trusting memory across eleven or twenty of them.
None of these steps requires special tools. What they require is patience, and a willingness to stop and recount when a number looks wrong instead of talking yourself into it being close enough.
A gathering that collates "close enough" is usually a gathering that will need to be recollated later, after the sewing is already in and the mistake is buried under thread. The five extra minutes spent recounting a gathering that feels a leaf short almost always cost less than unpicking finished sewing to find out where that leaf actually went.
Collation does not make a book easier to repair. It makes the repair honest — you sew what is actually there, not what you assumed was there when you started.
When the Count Doesn't Match the Book
Sometimes the count comes up short, and the honest answer is to stop and figure out why before proceeding. A missing gathering might be sitting loose in the box the book arrived in, overlooked on first sort. It might have been removed decades ago and never returned. Or the book itself might be a composite — two partial copies bound together at some point to make one complete-looking volume, a practice more common in older working libraries than people expect. Each of those has a different answer, and none of them is solved by sewing around the gap and hoping it reads fine closed. The force a needle needs to pass cleanly through a fold depends on knowing exactly which folds are being asked to hold the block together — a question collation answers and guessing does not.
| Symptom | Likely cause | Next step |
|---|---|---|
| One gathering noticeably thinner | A leaf removed or lost during a prior repair | Check nearby storage for a loose leaf before continuing |
| Signature marks skip a letter | A full gathering missing from the block | Confirm against a second copy or bibliographic record if one exists |
| Paper stock changes mid-gathering | A later replacement leaf tipped in by hand | Note it and treat separately, not as original structure |
| Two gatherings share the same mark | A stray signature from a different printing or copy | Verify which belongs to this book before binding either in |
| Fold direction reverses partway through | A gathering bound in upside down at some earlier point | Reorient before sewing, not after |
The Tool We Built for the Boring Part
None of the judgment above gets any easier with software — reading a signature mark, spotting a mismatched paper stock, deciding whether a gathering belongs to the book, all of that still happens by eye and by hand. What software helps with is the part that has nothing to do with judgment: keeping a running tally across twenty or thirty gatherings without losing your place, especially when the phone rings or the count gets interrupted halfway through. That is the entire premise behind Signature Count, the small companion tool we built for this one unglamorous task — it does nothing more than hold the number steady while your attention is on the book, not on remembering where you left off.
This entry describes how we approach collation on the bench before repair begins — it is a record of a working process, not a substitute for hands-on training. Every book carries its own printing and binding history, and an unfamiliar case is always worth slowing down for.