For the first 350 or so years of printing as a commercial business, its technology changed slowly. It remained on a craft scale, typical in an era of guilds in the 1400s to 1700s, even as it grew in importance in reaching an ever-more-educated emerging middle class.
Demands on the profession pulled harder and harder, as people and businesses required a faster dissemination of information to keep commerce and society functioning.
After the Industrial Revolution kicked in, starting around 1760, it took only a few decades for it to have a deep impact on printing. Paper was formerly made as single sheets but could suddenly be produced in continuous rolls. Human- and water-powered presses shifted to steam power, then electricity, vastly increasing rates of output. The frailty and limits of humans in setting type by hand was shifted into high gear with typesetters clanging away on the keyboards of hot-metal composition systems that shot boiling lead into type matrices to create pages of letterpress type.
The pace of the 19th century was unreal, and reading contemporary accounts, it was thrilling and challenging for print shops and periodicals to keep up with capital costs, technological upheaval, and increasingly larger differences of scale between remaining cottage-sized businesses and enormous printing plants. But the 19th century was just preparation for the 20th.
Nearly all the developments from Gutenberg, in the mid-1400s, through the 1950s were thrown out largely in the space of 20 years. Conceptually, printing remained the same: putting words and images made of ink onto paper. Yet typesetting and presswork changed in nearly every way, even the ink changed substantially.
It was a technological displacement that many people today who work outside the graphic design and printing world or who came to it only since the 1990s may have thought happened first with desktop publishing (DTP), which ushered in digital type and full-page output. But that was just the culmination of decades of preceding work.
If metal-based letterpress was heavy, its replacement was literally light. If DTP was a coming-to-Jesus moment, photolithography and phototypesetting were its John the Baptist predecessors.
Letterpress is a relief-printing method. Something you want to print via letterpress has to be created in a mirror image in metal, wood, and other materials that all measure exactly the same height, 0.918 inches, and are all receptive to a brushing of ink. That surface, once inked, comes into direct contact with paper under pressure to create an impression for the final printed piece. Each color requires a separate impression.
Every part of that process is heavy, cumbersome, expensive, prone to breakage and wear, and requires massive physical ecosystems to produce the raw elements, which are refined and manufactured to meet tolerances. Millions of tons of lead, steel, iron, and wood combined into presses and type.
Then, a technology that began a very weak challenger to relief printing took hold. Lithography was invented around 1798 and perfected over decades. It doesn’t rely on relief to separate inked and non-inked areas, but rather uses materials that are ink-attracting (oleophilic) and ink-rejecting (oleophobic). This makes lithography a flat or planographic method. Some people invert the description of the effect and use the terms for surfaces that attract and repel water: hydrophilic and hydrophobic.
The original use of lithography was for artistic prints, relying on a stone (“lithos” in Greek). Artists and craftsman would draw directly onto a stone surface with a greasy medium, a waxy crayon, oil, or even fat! The surface would then be treated with some combination of acid, rosin, talc, and gum arabic (a tree sap), which would etch the greasy material in place to make it accept ink and cause the nongreasy area to repel it. The surface could then be inked to apply paper and make a printing pass. The stone could be scraped and reused many, many times. If you’ve ever wondered why art prints are typically numbered in the order made, it’s because the quality of the reproduction degrades slightly over time.
Lithography wasn’t suitable for large quantities and wasn’t compatible with that era of type. However, printers had already worked out a method to make relief plates, which allowed printing an entire surface at once of the same material, so they were prepped for the change that came later. The definitive history of printing, A Short History of the Printed Word, by Warren Chappell and revised by Robert Bringhurst, notes that in 1727, a Scottish printer figured out how to take a “forme”—varied material ready to go on a letterpress—and create an impression in a soft material that could then be used to cast the plate in metal. This wasn’t a popular idea among other printers for a load of reasons, including quality and a potential reduction of employment.
The technique didn’t catch on until French printer Firmin Didot pushed it forward in 1794 to create what he called a “stereotype,” and others then took up. In 1837, another method, called “electrotyping,” supplemented stereotyping. Instead of casting metal into an impression, a wax mold was dusted with graphite; then, through an electrical process, metal would precipitate into the matrix as a shell, which could then be cast to make a printing plate. (Stereotype in this use derives from Greek words for “firm” or “solid” and “impression” and gave rise to the metaphorical use of a fixed notion of someone’s characteristics. The word “cliché,” thought to be onomatopoeia from the sound of making stereotypes, is used in the UK as a synonym for stereotype or electrotyping and also produced a figurative meaning — an oft-repeated phrase.)
Mass production of type only started in the early 1800s. Mechanical aids to make the elements that create type molds or matrices didn’t advance until the 1880s, when hot-metal typesetting systems also arose. This kept type expensive and sometimes scarce.
Stereotyping and electrotyping solved the problem of scarce type by creating plates that could be used while the original type was distributed for other projects, as well as reducing the wear on type. It led to rotary presses, in which plates were created flat and then bent to work on a cylinder, which, among other advantages, allows paper to pass through continuously from a roll for a newspaper.
The next step from relief plates was a logical one. Photography was a flat medium, like lithography, and its modern form emerged as the printing world sped up and increased mechanized typesetting and production. But the photographic process was one of making one-off prints from a negative or direct exposures. How could the manifold nature of printing be married to the flatness of photography?
It took a while for those two to create the right pairing. When they did, everything changed.
Take a Picture: It’ll Last Longer
It was logical for inventors to look at lithography and printing and try to marry them together. Some efforts to speed up lithography had taken place in the 19th century, but they didn’t challenge letterpress. However, in 1904, a printer named Ira Rubel provided the impetus for changing that. He created an “offset” method of lithography using metal plates created with “photolithography.” Better yet, it was accidental! Let me break these terms down.
Instead of drawing and etching to create an image on stone or other substances for lithography, photolithography relies on coating a metal plate with oleophilic material that can be exposed to light. Where exposed, it hardens. The unexposed portions are washed away with solvents, leaving those portions of the plate receptive to water.
That’s one part. Offset printing doesn’t contact paper directly with the item to be printed. Instead, there’s an intermediary step, the offset, in which a rubber blanket on a roller receives the ink from the plate, transferred in perfect fidelity, and then rolls onto paper. Offset printing allows the very thin layer of ink to transfer with fidelity while also reducing wear on the metal plate, making it last far longer than a direct contact.
On an offset press, the plate is secured around a cylinder. When engaged, the press rotates the plate against dampening rollers, which use a combination of water and other chemicals to coat the ink-resistant areas, and inking rollers, where the exposed parts pick up ink. The plate then contacts the impression roller, a rubber roller which then offsets against paper being fed between it and other rollers. That step deposits the ink.
It took until around 1950 for offset (as it’s known for short) to start to take off, as materials and methods and presses all had to be improved and come together. While photolithography started replacing relief printing, a similar change was happening on the type and image side.
Toward the end of the 19th century, creating a design for a typeface and then executing it in a form that resulted in usable type was a whole series of interrelated crafts — potentially hundreds of different skills were required. This started from the crisp offices of a designer, rather like an architect of a project drafting heady plans, through the people who transferred and perfected drawings, cut them as templates in metal, and worked through stages that resulted in matrices for hot-metal systems or individual pieces of cast type for hand-set letterpress. Fewer stages and less complexity was required to make wood type and etch images and illustrations for reproduction, but they were still quite involved.
The practical development of offset printing, which required a photographic step to make plates, pulled phototypesetting into existence. As John Seybold wrote in his seminal 1984 work, The World of Digital Typesetting, that by 1950, “The need for a photocomposition device was readily apparent.” With pages still being composed in relief, photolithography required a printer to go through all the stages of letterpress printing, except making multiple copies. Composed material was printed in a single copy as a proof, which was then photographed to make a plate.
Phototypesetting systems relied more or less on taking the same sorts of original drawings created by type designers and perfected by craftspeople to start the stages of cutting in metal. But instead of producing physical templates, the designs were turned into photographic masters that could be reduced in size to be transparent type windows in a dense black film background.
Shine a bright, focused light on something that often resembled a filmstrip or a piece of microfiche, pass the result through a lens, and have it land on photosensitive, high-contrast, resin-coated (RC) paper, a black-only medium that provided sharp details for photography. The type could be made optically larger or smaller through manual, mechanical, or computer control, allowing the use of a single master filmstrip to produce many sizes of types. A printer no longer needed one “font” of a type, but rather picked among styles (like Roman, italic, and bold) that were usable for a range of sizes.
It didn’t appear all of a sudden. As Briar Levit, director of the 2017 movie about phototypesetting, Graphic Means, told me in an interview, “People had been working on it for decades, and finally it was ready.” A combination of factors, including strong union control of typesetting and printing, delayed it. And it was welcomed. “A lot of designers preferred the look of cold type,” Levit says, as phototype was referred to. “It could be output so beautifully.”
But when it hit, all the metal and wood required for composing a page went out the window, sometimes literally. Designer Erik Spiekermann says in a biography that, in the late 1960s, printers in the Kreuzberg part of West Berlin threw cases of metal type right out the window, where he collected them.
Pasting Up Is Pasting Down
When I started in the printing industry, around 1985, offset was fully perfected, and making an offset plate had many, many steps. You start with a right-reading layout that you assemble from flat material and paste onto a board with wax or rubber cement — a “layout” or a “paste-up.”
You assemble type output from RC-paper-printed galleys, which are runs of text in columns, often on narrow strips of paper; from headlines, often printed or typeset separately; photostats, or photographic black-only artwork shot from hand illustrations or other assemblages; and halftones, which turned continuous shades of gray into dots of varying sizes that fool the eye into perceiving grays (or shades of color with full-color reproduction). You might also put down tape with lines (“rules”) to make a box, or cut shapes from Zip-a-Tone or Letratone to add flat tones of black or color to areas. We could make invisible notes on board using pens that used a particular shade of color — nonreproduction blue — which the next stage of photography didn’t pick up.
That layout board is photographed precisely flat and square onto high-contrast negative film that doesn’t see tones at all: Everything on the board has to be either black or white. The resulting film is opaque where it was white on the board and transparent where there was black. After “stripping,” in which the film may be cut up or masked to be perfectly arranged for positioning relative to the plate, this composite is placed in a vacuum frame against an unexposed plate. The vacuum ensures alignment and removes light leakage and shadows. After exposure to high-intensity ultraviolet light, the plate — which is also right-reading — goes through a chemical bath that removes all the unexposed areas, and it went on press, as described above. Paste-up is right-reading as you look at it; a negative mirrors that; the plate mirrors that; the offset blanket flips it again; the final stage, on paper, thus winds up in the correct direction!
But this phototype and paste-up era was short-lived, even as offset persists. It was a quick transition that went through stages.
The last stage combined computers and lasers into an imagesetter. Software like Aldus PageMaker and QuarkXPress allowed a full page to be composed onscreen — or even entire imposed sets of pages to be printed on a single sheet on press — and sent for output to something like an Apple LaserWriter (1985) relying on Adobe’s size-independent PostScript type. That output, while low-resolution at 300 dpi, could be photographed and used for inexpensive print jobs.
That evolved into high-resolution devices, like the Linotype Linotronic 300, which could paint pixels at 1,200 dpi on RC paper — and then directly to film.
When I ran my university’s imaging center in our printing plant, in 1991 and 1992, where we did job work like small books, posters, letterhead, and the like, we moved from a lot of RC paper imagesetting to direct film output. The plate specialist used that film directly.
But the imagesetter period, whether to RC paper or to film, wasn’t the end stage in development, either.
In modern plants and even small shops, computer-to-plate (CTP) digital technology also takes the film out of the process. A CTP device uses a laser to either expose or burn away plate material. Some may require a baking or exposure stage afterward or can handle that part inside a single device.
This weirdly takes us full circle, almost back to the days of the stereotype and the electrotype, where a printer in their own shop takes a design and produces a metal reproduction that goes on press. It’s enormously faster, cheaper, crisper, and all the rest. But the evolution of typesetting and printing has taken us much closer to 1850 than 1950.
In a wonderful bit of irony, lasers are helping revive letterpress in the 21st century, making it more feasible as a means of production than it has been in decades.