Paul Nash: “How printing types were made in the hand press period”, 21 January, 2011

Book of Hours (Paris: Franciscus Regnault, 1534). Bodleian Library, Gough Missals 177, fol. C2, recto (detail)

from Martha Repp

The first in the 16th series of the Oxford Seminar on the History of the Book, convened at All Souls College by Professor Ian Maclean, was given on 21 January 2011 by Dr. Paul Nash, of the Bodleian Library, on the topic of “How printing types were made in the hand press period”.

The main topic of Dr. Nash’s talk was the method by which individual pieces of moveable type were produced for most of the hand press period. This was illustrated throughout by extremely informative video clips from You Tube of the self-taught American type founder Stan Nelson at work; links to these can be found in the Resources section of the Codrington Library’s ‘Resources’ web-page, There was also an opportunity to see and handle some of the materials used, and examples of type at each of the stages of its production.

The process of making type was traced from beginning to end, and broken down into three main stages. The first of these was the creation of the initial punch, where a rod of steel is taken and the mirror-image of the required character drawn on it in ink, graphite or soot, a gauge being used to make sure the image is of the correct size and depth. After this, the character is cut, using tools similar to those used by jewelers or watch-makers for engraving, and the punch hardened and tempered to give it strength. Punch-cutting was the most laborious and time-consuming stage of the whole process; the creation of a full set of type could involve making anywhere between 80 and 300 individual punches, depending on the language and the number of characters with accents, contractions and ligatures to be used, and it is unlikely that even a skilled punchcutter could produce more than one punch a day.

The punch is then taken and placed face down on a block of softer metal, usually copper, and struck while being held as upright as possible in order to leave an impression of the character in the copper. The resulting impressed copper block is referred to as a matrix.

Finally, the matrix is placed in a mould and held in place with a spring. Molten metal is then poured into the mould through a funnel shaped opening at the top, and allowed to set. The preferred metal is an alloy of lead, tin and antimony, as it has the necessary properties of melting at a low temperature, flowing easily and freezing quickly while maintaining a consistent volume. The mould would be shaken as the metal sets in order to ensure the even distribution of the metal; the precise nature of the shake required would depend on the character being cast.
This process was certainly in place by 1500, and changed little until around 1800. Our evidence for it comes largely from the survival of original materials, and from the account provided by Joseph Moxon in his Mechanick exercises on the whole art of printing of 1683-4. Initially, printers would have undertaken the whole process themselves, in order to safeguard what was still a relatively new technology. As the demand for type grew, separate workshops for making punches and matrices developed, and finally separate type foundries.

Other questions considered in the talk included what came before and after this method of producing type. The question of what preceded it is necessarily shrouded in some ambiguity, and here the scholarly debate surrounding the precise nature and method of production of Gutenberg’s early types was touched on, and various different theories considered. Later industrial methods of producing type such as the pantograph and the pivotal typecasting machine were also explained and illustrated.

A stimulating final discussion broadened the topic to include issues such as the design and production of typefaces for Greek characters, the production and proof reading of the very small types required for, for example, the 32mo volumes produced by the Elsevier Press, and the link between typefounding and cryptography, in that both depend on an analysis of the likely frequency of particular characters in any given language.

See pictures of a printing class taking place in the Bodleian Library’s Bibliography Room.