It may refer to removal of bracketing, so all the corners are clean and sharp, rather than rounded, as many are when written with a broad pen.

Ah, that makes sense; thanks!

It does indeed make sense, but then I thought to double-check, and at first, I wondered.

The Didot typeface wasn't started until 1784. And even Baskerville was only designed in 1757 - but it did not have the unbracketed serifs characteristic of modern faces, even if it had other characteristics of theirs in incipient form.

However, the Romain du Roi dates from 1693. Whew!

It may refer to removal of bracketing, so all the corners are clean and sharp, rather than rounded

I think Bill is right – I’ve never understood it well myself, but now it seems to me a key passage. Fournier did this ‘removal of bracketing’ with his own capitals, especially the larger sizes that were used for drop initials or titling. It is not so noticeable in the text sizes of the romain du roi (for which the engraved ‘drawings’ all show bracketing), but here is an initial for casting on a body of 96 Didot points, from the specimen of the Imprimerie royale printed in 1760.

One also needs to reckon with the influence of the clean, sharp effect of big letters engraved as outlines on copper and then filled in, even if they then get bracketing added, as in this example from Diderot’s Encyclopédie.

That engraved line helps to shape the serif that eventually gets into both the capital and small letters of the new types. It is there in Fleischmann in the 1730s and the Didot types in the 1780s.

A note on dates. None of the punches of the romain du roi was cut before 1696, and the process of making it went on for decades. That big A was made by Grandjean’s successor Alexandre in 1730.

Given that reading, it's a bit curious that Smeijers does not return to that statement when, later in Counterpunch (pp143-5), he talks about the punchcutter's preference for avoiding sharp corners and straight lines.