It's the details thnt matter

For all of its obviousness, “thnt” is not a one-time mistake in this book: we see it here, on sig. C1v, but it shows up first on B2v and then again on G2v and H2v. Nor is it the only error that appears in the headline:

The substitution of “thei” for “that” is also a mistake that appears more than once, showing up on A1v, D1v, and F1v. (There’s also a slight variant on that error on F3v, which reads “they” for “that.”)

Both “thei” and “thnt” are obvious mistakes. So why are they interesting? For a starter, “thnt” is interesting in part because it shows what happens when you set type by touch. Just like proficient pianists and typists know where the keys are without looking at their fingers, compositors knew where the sorts in the case were without having to hunt through the letters. You reach for the box where the “a”s are, you put the sort you grab in your composing stick, and you move on to the next letter. But if an “n” had been accidentally mixed in with the “a”s, then you could end up with “thnt” instead of “that” and never notice while you’re setting your type. In theory, that mistake would be picked up and corrected when you proof your work. But in practice, I would bet that headlines are proofed less diligently than the text proper.

Headlines are, by intent, not particularly noticeable; they’re there when you need them, but otherwise fade to the background. They typically stay the same for long stretches, and the elements that change do so in a regular pattern. In this book, the headlines differentiate between the different sections of the work: they announce the first part of the things they bring from the West Indias from A1v through H4r, the second part from I1v through Y2v, and the third from Y4v through 2E1v. ¹ In other words, for the entirety of the “first part,” the headlines should read the same for all 63 pages (“The first part of the things that” on the verso; “they bring from the West Indias” on the recto) or, to think in book-making terms, the same for all 16 formes.

And here we have to pause for a moment to understand what “formes” are. As you already know, books were printed on sheets of paper, with multiple pages laid out on each sheet. (See my earlier posts on signature marks and printing mistakes for a refresher.) Sheets, of course, are printed one side at a time, and the locked-up type for each of those sides is referred to as a forme—the outer forme is the side of the sheet with the recto side of the first leaf (A1r) and the inner forme is the reverse side (A1v). Often the outer forme is printed first, then the inner forme, then the outer of the next sheet, that that inner, and so on. In a quarto like our book, that means that the first pages to be printed would be A1r, A2v, A3r, and A4v; the next set of pages to be printed would be A1v, A2r, A3v, and A4r; and then B1r, B2v, B3r, and B4v. Once the outer forme of A is done being printed, the forme would be stripped, the type distributed back into the cases, and the next forme set and imposed. The trick is to be efficient about this work, so that the press is never idle. As soon as one side is done, the next should be ready to go. It’s likely, then, that while the inner forme of A is being printed and the outer forme of A is being stripped, the outer forme of B is already being set up.

What does this mean for headlines? Take a look at what the outer and inner formes of a sheet would look like from the standpoint of working with headlines. What you see below are the two formes with the main text shown as a grey box (it’s not relevant to our current concerns) and the headlines shown as they would be laid out on a printed sheet (each page is identified by its signature marks, although the signature marks in parenthesis wouldn’t have actually been printed).

As you can see, the headlines stay in the same location whether it’s the outer forme or the inner. So, when you’re putting the text of the grey box away, there’s no point in stripping the headline—you’d just have to set the exact same type again. These elements of a page that remain the same from one page to the next make up the skeleton forme: the wooden and metal pieces that hold the type in place, the headlines, any ruling or decorative elements that are reused on each page get left in place while the text is changed.

So what happens if there’s a mistake in the skeleton? It shows up again and again: “thnt” shows up not only on the third forme that was set but on the penultimate forme of the section. The only explanation is that the same series of sorts was used repeatedly—it doesn’t make sense that there would be multiple “n”s mixed in with the “a”s and that the only place they would show up is in that one word. And how would such a mistake escape proofreading? It’s not noticed because it’s not seen. Headlines aren’t part of the text that we typically pay attention to.

In theory one might be able to trace the order in which the formes were printed, if they’d been done so in a regular pattern, but that doesn’t seem to be the case with this book. There are actually six recognizable variants in the headlines for the first part: “thei,” “thnt,” and “they” all stand out as errors, but “first” also appears with both a long-s and a short-s and spelled as “firste.” But the variants don’t, to my eye, indicate any obvious pattern. However, with the exception of the single occurence of “they,” all the variants in the headlines show up in the first four formes, and they all show up in the last six formes of this section. In other words, they get set up as the headlines early on in the book’s production and they continue to be used as long as the headline doesn’t change to a new section.

The “thnt” and “thei” errors disappear once the first part of the book is done; presumably they were noticed when the headlines were reset for the second and third parts. But other errors show up in the headlines, particularly around the transitions between sections:

And my favorite, which obviously happened when some of the sorts fell out when the forme was unlocked and then were shoved back in without much attention:

So if “thnt” was the first mistake in the crocodile, what was the second? The foliation number is wrong.

The page shown on the right of the opening is the 10th leaf, but is marked as the 12th. That mistake is harder to spot—Bob MacLean noticed it by observing that the book looked as if it was a quarto and that the signature mark (C2) suggested it was the 10th leaf (4 leaves in the A gathering plus 4 in the B plus 2 in C); another reader worked it out by spotting the bleed-through of the inked 9 on the previous leaf. It’s also not quite as interesting, alas. As you might expect, the 12th leaf is misfoliated as the 10th—clearly the numbering was reversed when it was imposed, although the pages themselves are in the right position. Both C2r and C4r are on same forme, so it would be easy enough to accidentally put a “12.” where the “10.” should go, and vice versa. Again, it’s something that could have been caught in the proofing, but wasn’t.

What’s the take-away from this? For me, it’s a reminder that the most banal parts of a book sometimes have their own stories to tell!