Beyond these sources of typographic variability lies another factor, one that would best fit under the rubric of commerce. Historically, most printers have purchased, rather than designed, the typefaces they used. In the pre-digital world, this meant that printers would acquire matrices or punches from a designer or agent, and use these tools to cast type in their own shop. When, as invariably happened, one or more of these tools broke, pre-Industrial printers would rely on local craftspeople to cut replacement letters. As these re-cut letters filtered into a printer's typecases, the formal qualities of the typeface changed in subtle but sudden ways, often occurring in the middle of a book or even a single page. Replacement punches, after all, were not typically commissioned in idle moments, but in a state of urgent necessity; they were cut when a printer ran out of, say, the letter "f" half way through the printing of a book, found that his matrix was worn out, and, when attempting to strike a new one, broke the punch. In a panic, a local punchcutter would be called, shown a model from which to cut a new punch, and the book could be finished and shipped off to market.* Speed, more than aesthetics, is the engine of such moments, and a workable letter quickly engraved trumped any aspirations to strict formal fidelity. I believe this element of speed, whether in the initial cutting or the recutting of letters, was critical to the feel of pre-Industrial pages.
Since I began designing type I have tried to find ways to recapture some of this pre-Industrial liveliness in my typefaces. Short of cutting my own punches or creating large sets of extra characters, I have found that speed is one aspect of historical type design in which I can partake without risking the wonky anachronism that often results from strict historical mimesis. I have also spent a lot of time attempting to redraw historical letterforms in an effort to tap into the gestalt of pre-Industrial letterforms, a pursuit that is frustratingly and consistently futile. If I redraw the same letter five times, I end up with five distinctly different letters, none of which looks quite like the original I was trying to imitate. At best I create a passable likeness. So when considering how to approach the type for Hungry Bibliophiles, I decided that I would put aside any pretensions of genetic accuracy and begin from the presumption that I am acting as a replacement punchcutter. In the spirit of Tim Barrett's experiments with high speed papermaking, I set myself a daily goal of between ten to thirty letters (as if that many punches had broken since I was last at work) and a speed of ten letters per hour, allowing myself only one drawing and one revision per letter.
As my model typeface I chose the Canon Roman and Italic of Peter de Walpergen that was acquired by Oxford University Press in 1686. I consciously chose a typeface that was neither Italian nor Renaissance in origin, one whose workaday forms would have sent Nicolaus Jenson leaping from the Ponte Rialto as if he were a corporeal bit of Doves type. The Dutch typefaces that Bishop Fell acquired for Oxford are full of irregularities—in weight, slope, stress, alignment, and spacing—, and they delight in unexpected pairings and ambiguities. They can be disconcerting in their proportions but they sparkle with life, as if they were coursed through with a dappling light. From a practical standpoint, De Walpergen's types also have a proportionally large x-height, making them eminently readable at a small size. His are typefaces that are ready-made to slog through the various trenches of partisan pamphleteering, Calvinistic tracts, and Uncle Pieter's kugel recipe. Perfect, in other words, for a book like Hungry Bibliophiles.
_______
*Printers who design and manufacture their own type, such as Vicentino and Perugino, would obviously have re-cut a broken punch themselves.
