#672 1884 · Ottmar Mergenthaler · Printing / publishing
Mergenthaler realized the bottleneck in typesetting wasn't setting type, it was putting it away afterward, and cast it fresh instead
the problem
Typesetting machines sped up composing type but not distributing it back
background
By the 1880s hand compositors set newspaper pages one pre-cast metal letter at a time from wooden cases, then, after the page was printed, spent nearly as long sorting every individual piece back into its correct compartment for reuse -- a step called distribution. Several inventors, including the famous and financially disastrous Paige Compositor backed heavily by Mark Twain, built machines that mechanized only the setting half of the job, arranging pre-cast type faster than a human hand could.
None of them touched distribution, so their real productivity gain was small: a machine that composed type twice as fast still needed the same slow manual sorting afterward, and often broke down trying to also automate that half. The industry kept optimizing the visible, timed half of the job -- composition speed -- while the unglamorous second half quietly ate back whatever time the machine had saved.
what everyone would do
Every rival inventor, including the famously expensive Paige Compositor, built machines to arrange pre-cast type faster -- automating the half of the job that was visible and timed, on the assumption that speeding up composition was the lever that mattered.
what they saw
The real bottleneck wasn't setting type, it was putting it away afterward. Every earlier machine still needed the same slow manual distribution once a page was printed, which quietly erased whatever time composition had saved. If the type itself were simply melted down and recast fresh each time instead of reused piece by piece, there would be nothing left to distribute at all.
the move
Ottmar Mergenthaler, a German-born instrument maker working in Washington, D.C., abandoned the idea of arranging pre-cast type altogether. His Linotype machine let an operator assemble a line of reusable brass letter molds (matrices) from a keyboard, then cast that entire line as a single solid metal slug on the spot; the matrices dropped back into their magazine automatically, ready for the next line, while the printed slug itself was simply melted down and recast later. There was nothing left to distribute by hand.
why it works
By casting a fresh line of type from reusable brass matrices instead of arranging thousands of individual reusable metal letters, the Linotype removed distribution from the workflow entirely: the matrices return to their magazine automatically after casting, ready for the next line, while the cast slug itself needs no sorting -- it is simply melted down after use. Because the labor Mergenthaler eliminated (distribution) was roughly equal in time to the labor everyone else was still trying to speed up (composition), the net productivity gain was far larger than any composition-only machine could deliver, which is why the Tribune's 30 machines could replace the output of roughly 90 hand compositors.
the payoff
The first working direct-casting machine ran in July 1884; the New York Tribune set part of its July 3, 1886 edition on one, and had 30 machines in daily use within a year, saving a reported $80,000 annually versus the roughly 90 hand compositors that same output would have required. A single Linotype operator could set 5,000-7,000 characters an hour.
where it breaks
The trade cast metal for labor rather than eliminating cost entirely: each line used fresh metal that had to be melted down after use, an ongoing material and energy expense that only pays off when labor is expensive relative to machine time and raw material -- true of 1880s American newspaper publishing but not universal. It also only works when the item being 'put away' can be cheaply destroyed and remade each cycle; recasting is a poor substitute for distribution when the thing being sorted has to be reused intact rather than melted down.
what came after
The Mergenthaler Linotype Company had manufactured roughly 16,000 machines by 1910 and 70,000 worldwide by 1954; the Linotype remained the dominant technology behind newspaper and book composition for over 125 years, until phototypesetting and digital systems replaced it in the late 20th century.
references
- [1]Ottmar MergenthalerImmigrant Entrepreneurship (German Historical Institute / Business History Conference), 2015immigrantentrepreneurship.org
- [2]Linotype Invented and Built in BaltimoreAmerican Printing History Association, 2014printinghistory.org
- [3]Machine for producing linotypes — US Patent 436,532 (granted 16 September 1890)United States Patent and Trademark Office (via Google Patents), 1890patents.google.com