Cancels and Sophistication - Bibliographica

Bibliographica

SECTION V: CANCELS & SURGERY

Everything until now has treated the book as it left the printing house, an object whose structure we could read off the paper and the marks. But a book is not printed and then frozen. It is corrected on press, mended by its printer, rebound by its owners, repaired by its dealers, and, now and then, quietly improved by someone who stands to profit from our not noticing. This section is about that second life, and we will divide it cleanly into two halves: the honest surgery the printer performed on his own book, and the dishonest, or at least undisclosed, surgery performed on it later by others. The first is called cancellation. The second, when it crosses a line, is called sophistication. Learning to tell an honest scar from a disguised wound is, in plain commercial terms, the difference between a fair price and a fleecing.

The cancel, and why a careful printer cuts his own book

Start with the printer's own knife, because it is the innocent case and it teaches the vocabulary. Suppose a sheet has been printed and the whole edition run off, and only then does someone notice an error on one leaf: a libellous sentence, a patron's name misspelled, a date wrong, a coarse joke the author's second thoughts regret, a page of text that a censor or a lawsuit now forbids. Reprinting the entire sheet would be ruinous. So the printer does something far cheaper and rather surgical. He sets and prints a small correction, usually a single leaf, then has that one offending leaf sliced out of every copy near the spine and the corrected leaf pasted onto the stub that remains. The leaf removed is the cancellandum, the leaf that is to be cancelled; the leaf pasted in is the cancellans, the leaf that does the cancelling. The Latin here is worth keeping straight because the whole trade uses it, and the gerundive cancellandum literally means 'the thing needing to be cancelled', which is a small grammar lesson smuggled into bibliography.

Now recall what we know about conjugacy, because it is the entire tell. In a gathering, leaves come in conjugate pairs joined at the fold, each leaf the twin of another. When the printer slices out a cancellandum, he cannot also remove its distant twin without disturbing the whole gathering, so he cuts close to the spine and leaves a narrow stub, and onto that stub he pastes the single, twinless cancellans. The result is a leaf that has no conjugate. It is an only child in a family built entirely of twins. That is the tell of a cancel, and it is why the forensic exercise in Section III had us hunt not for the leaf that looked wrong, but for the leaf with no partner. A cancel is detected less by what the substitute says than by the stub it stands on and the twin it lacks.

Cancellation gives us, occasionally, the rarest of book-historical treasures: an uncancelled copy. Now and then a copy escapes the workshop before the correction was made, or a binder was lazy, and the original cancellandum survives in place, carrying the very sentence the printer paid money to suppress. These uncancelled states are prized precisely because someone once wanted them gone, and the collector who can recognise a cancellandum still present, when every other copy shows the tame cancellans, holds a leaf that is both rarer and often far more interesting than the corrected text that replaced it. The suppressed reading is the one everybody wants, which is a very human fact about books and about people.

How do we write it down, now that Section III gave us the grammar? A cancel is a minus and a plus at the same address, the old leaf out and the new leaf in, so a gathering of eight leaves with its third leaf cancelled is written A⁸(±A3). And here is the subtlety: because the printer did this to every copy of the edition, the cancel belongs to the edition, and so it belongs in the formula itself. Contrast a leaf merely torn out of a copy by a clumsy reader, which is damage to the copy alone and is recorded as a note against that copy, not in the edition's formula. Same physical absence of a conjugate, utterly different meaning, and the whole art is in the reading, not the seeing.

The many faces of the small repair

Before turning to villainy, we should be fair, because most of what looks like tampering is innocent and the suspicious mind ought to be a fair mind or it is merely paranoid. A great deal of surgery in old books is honest housekeeping. A plate or a map, printed separately on different paper, was often pasted onto a stub or a guard from the start, and so it too has no conjugate, and so it too can alarm the beginner who has just learned the cancel's tell. A torn leaf mended with a slip of old paper, a corner restored, a wormhole filled, a fragile leaf reinforced along the gutter, these are conservation, not deception, and a book that has been cared for across three centuries has every right to bear the marks of that care. The point of this section is certainly not to breed suspicion of every repair. It is to make repairs legible, so that the honest ones can be recognised and set aside, and attention reserved for the few that are pretending to be something they are not.

De re hortensi libellus: the title page here is a modern facsimile, supplied to replace a leaf wanting in this copy.

A close-up of the facsimile title page to the left.

Genuine letterpress bites into the paper, leaving an impression we can feel; here the ink instead stands proud of the surface, and the characters break up into a fine pattern of dots. Both are tells of a photographic or digital reproduction.

For comparison, a detail of the genuine title page of Seminarium sive plantarium earum arborum, where the ink is pressed into the sheet.

Sophistication

Now the dishonest half, and the word for it is one of the most elegantly sinister in the trade. A sophisticated copy is one that has been altered to appear more complete, more perfect, or more valuable than it truly is, in a way meant not to be noticed. The word carries its old double sense, sophistication as refinement and sophistication as adulteration, and in the book world the two meanings collapse into one: a copy made more refined precisely by being adulterated. The sophisticator's whole craft is that his work should not read as work at all.

The commonest form is the made-up copy, and its logic is pure arithmetic of scarcity. A perfect copy of a rare book is worth a great deal; a copy lacking three leaves is worth a fraction. So someone with a defective copy and access to a second defective copy, or to loose leaves, or to a skilled facsimilist, makes the missing leaves good. Leaves are supplied from another copy, or expertly pen-and-ink facsimiled, or photolithographed on matched paper, or washed and inserted from a broken copy of the same edition. A title page, the single most valuable and most often missing leaf, is married in from elsewhere. Sometimes two incomplete copies are cannibalised into one apparently complete one, which the trade calls, with grim accuracy, a made-up copy. The finished object presents itself as whole. It is a chimera.

Sophistication is defeated by the ordinary disciplines we have already learned. Supplied leaves betray themselves by paper that does not match: hold the questioned leaf and a leaf known to be original up to the light and watch whether the chain lines run at the same spacing and the watermark belongs to the same stock. A leaf from another copy will often be a shade too large or too small, its margins not quite agreeing, its wormholes failing to line up with the tunnels in the leaves on either side, since a worm bores through a closed book in a continuous track and a stranger leaf breaks the track. Facsimile leaves fail under magnification, where printed type shows its bitten, three-dimensional impression and a photographic or drawn imitation shows a flat, dead surface. A made-up copy frequently carries a catchword that does not chain, or a signature in the wrong sequence, or a press figure that belongs to a different campaign of work, because the sophisticator matched the sense of the text but not the humble bookkeeping beneath it.

The most dangerous sophistications are the ones performed by people who love books. The great nineteenth-century collectors and their binders routinely 'perfected' copies, washing and pressing and remargining and supplying, often recording nothing, in an age that valued the appearance of completeness over the documentation of intervention, and much of what a modern cataloguer must patiently detect was done not by a crook but by a connoisseur with a clear conscience. This is why provenance bears so directly on this problem: a copy that passed through certain famous hands should be examined more carefully, not less, because those hands were the most skilled improvers of their day. Reputation is not a certificate. It is sometimes a warning.

Remboîtage

One special case deserves its own name because it fools even careful people. We tend to trust an old binding as evidence of an old book. But bindings and text blocks are separable, and a binding is itself a saleable, transferable thing. Remboîtage is the practice of putting a text block into a binding that was not made for it, typically a handsome old binding placed around a book to lend it borrowed dignity or a false provenance. The trap is precise: here the binding is truly period, fine, and genuinely old, and every instinct says trust it, yet it is a lie about this book, because it belongs to another. The tells are again physical. The text block's edges do not quite match the binding's dimensions; the sewing stations do not align with the wear inside the covers; an armorial or a bookplate speaks of an owner the text's own provenance never mentions; the spine's original title, if it survives, names a different work. Remboîtage is the reason the central rule of this course must hold even against the evidence that feels most trustworthy: the object is the sole authority, but every part of the object must be cross-examined against every other part, because a book is an assembly, and any part of an assembly can have been swapped.