Why I Like to Set Type by Hand

Figure 1.

Sample page hand type-set by Barbara Henry. Page spread from Specimens from Harsimus Press, 2002. Designed and printed by Barbara Henry; reproduced by permission of the author.

It is almost impossible to tell from looking at a page of letterpress whether the type has been set by hand. A hand-set page may have subtle qualities of spacing or design to set it apart—a flaw or two, perhaps—but the only truly identifiable clue is the use of a typeface unavailable in a machine; few people would know. Still, it is something I love to do. And there is nothing more satisfying than to do it properly.

What do I mean by setting type by hand? I mean holding a steel composing stick in my left hand and picking up pieces of type, one by one, from a compartmented job case (a shallow tray in which the letters are arranged according to long tradition and committed to memory). An accomplished compositor can set a thousand ems an hour: about two and a half pages of twelve point type. Nineteenth-century newspaper compositors could do better—there were races held, and in 1870 George "The Velocipede" Arensberg set 2,064 ems an hour of minion (seven point type), a record that was soon surpassed before being rendered meaningless by the invention in 1886 of the Linotype machine. I don't care about speed; in fact, I set type slowly, considering word spacing and design problems as they come up. When a line of type is complete, when all the letters have been set in the stick, the line must be justified. That is a lovely term which means that the line tension, the degree of snugness of the pieces of metal within the line, must be precisely correct and equal in every line. Each line must be justified before proceeding to the next. When the composing stick is full, and it can only hold a few lines (for one thing, it gets heavy fast, for type is made of an alloy of lead) the type must be slid out onto a galley (a shallow metal tray with one open end) for storage. The composing stick guarantees the uniformity of line length; thus, the type form is a perfect rectangle. At the end of the day, or at the end of the job, the type is wrapped with ordinary cotton string, the end of which is forced under the layers of wrapped string and tightened up to a corner to secure it. It is not knotted. It can be loosened by simply pulling on the end of the string. This is done because the type form consists of hundreds or thousands of very small loose pieces and can easily fall into pi. Pi is the term for a jumble of type that takes a very long time to straighten up. This brings us to the biggest problem with hand composition: distribution.

Before 1886 all type was set by hand. There was no other way to do it. As presses improved and were able to turn out more and more impressions per hour, hand composition began to seem like a bottleneck in the process—hence, the races. Most of the nineteenth century was spent in the quest to mechanize composition. Mark Twain, who found hand composition a bore, lost a fortune investing in the wrong machine. Ottmar Mergenthaler finally came up with a machine that assembled matrices, or molds, from a keyboard and cast solid lines of type from a reservoir of molten lead; when the type was printed and no longer needed it was returned to the pot and melted for re-use. The type did not have to be laboriously returned to its place, or distributed, to be used again. A hand compositor in the nineteenth century was not paid for distribution. He simply had to do it or he could not do his job. Distribution is a meditative exercise; unlike composition, not a lot of thinking is required; the mind is free...