Graphotyping a Newspaper

1. What

The article reproduced here, "Chicago Experiment[:] An Epic in Cold Type" by Carl R. Kesler, appeared in The Inland Printer, Volume 124, Number 6 (March, 1950): pp. 37-40. It describes how during a two-year printers' strike in the late 1940s the editorial staff of The Chicago Daily News used several makeshift techniques, including composing "col" relief type on the Graphotype, to produce the newspaper.

2. How

Kessler's article was written for professional printers, and assumes a familiarity on the part of the reader with the technologies used to produce newspapers at the time. These were perhaps never as widely known outside the industry as they should have been, and are now almost forgotten. I'll try, therefore, to summarize the technical situation that the article describes.

The newspaper was printed using relief-printing techniques ("letterpress" techniques) rather than offset lithography. (Offset lithography would displace relief printing in newspaper production within a couple of decades of this time, but had not yet done so.)

Ordinarily, the newspaper was set and cast into type using Linotype machines (or the very similar Intertype machines manufactured by a competitor). These machines allowed the composition of the newspaper at the keyboard. They did not set pieces of pre-existing type type in imitation of hand typesetting, but instead cast brand new type each time, interactively, under keyboard control, from a pot of molten typemetal. Moreover, they did not cast these as individual pieces of type (one per letter), but rather they cast fully justified lines of type as "slugs," one slug per line ("Linotype" = Line o' type). Very large headlines and some advertising work probably would have been cast using a complementary machine, the Ludlow, which was slower but well adapted to very large type. (A very different competing machine, the Monotype, did compose and cast text as assemblies of individual types, but the Linotype and Intertype dominated newspaper work. They were fast, and slugs were easy to handle.)

While it is quite possible to print directly from a column of Linotype slugs, newspaper work generally introduced another step into the process. The pages of the newspaper would be set up using Linotype slugs (and perhaps Ludlow slugs and photoengraved relief plates and other bits and pieces). Then a relatively thick paper mold would be taken of the page, under high pressure. This mold, called a "matrix" (or "mat") would then be used to cast a copy of the page as a solid plate of typemetal. (Yes, molten typemetal was cast against paper.) The printing plate so produced was called a "stereotype." (The Greek root "stereo-" means "solid." This use of the word "stereotype" was the original meaning of the word, signifying a "solid" printing plate (vs. one made up of many individual pieces of type). Our current use of the word "stereotype" is a derivative one which comes from the practice in printing of making and distributing many copies of the same text as stereotype plates.) The actual printing of the newspaper was done from these stereotype plates. There were two advantages in stereotyping: first, many plates could be made from the same original, thus extending the number of impressions possible, and second (and more importantly) the stereotype mat could be bent into a curve and a semi-cylindrical plate cast from it (type is flat, whether hand-set type or Linotype slugs). This allowed the use of rotary printing presses and therefore permitted high-speed printing. A fast rotary press without a stereotype department is about as useful as a fast car without a road.

The strike situation described in this article is called a "printers'" strike, but that doesn't necessarily mean what you think it does. Composing room staff (Linotype operators, for example), were called "printers," while the people who actually operated the printing presses were called "pressmen." Therefore, since apparently the pressmen were not on strike, they could actually run the presses to print the newspaper. What they could not do in the stike conditions was compose and cast the Linotype slugs. The situation therefore was one where in order to produce the newspaper the management and editorial staff had to find some way to supply their stereotyping department with raised printing surfaces for each page of the newspaper, to be stereotyped and then printed, without the composing and linecasting abilities of the Linotype.

Handsetting type probably wouldn't have been fast enough, and in any case there probably weren't enough skilled compositors left by that time to even try. The solution had to involve some kind of keyboarding of information by the editorial staff (and semi-skilled temporary workers).

The solution devised relied upon a technology (and nonstriking department at the newspaper) generally used for making relief printing surfaces of individual images, not of entire pages: photoengraving. This process could take a pure two-dimensional input image and produce from it a relief printing surface. In normal practice this photoengraving was then mounted on a lead-alloy base and placed alongside Linotype slugs to produce the complete printable page ready for stereotyping. In the makeshift process described here, photoengraving was used to produce the entire printable page. This stretched the limits of this technology.

To produce the two-dimensional input for photoengraving, they combined two technologies.

(1) "Straight" text matter was typed on Varityper machines. These were typewriters which could also proportionally space and justify text. The output from the Varitypers was two-dimensional paper with (very nice) typing on it.

(2) The Varitypers couldn't do the larger type sizes, though, so regular story headlines were done using "phototype" (it isn't so identified, but what is described sounds very much like the "FOTOTYPE" brand product). These were small tabs of cardstock with high-quality letters printed in reverse (just as type is reversed) on one side (and right-reading reference letters printed on the back). You composed them into a line in a sort of a composing stick, face down. Then you ran double-sticky tape over the back and transferred them, face-up, to your page.

The combination of these two techniques, together with laborious paste-up by hand, produced a two-dimensional page which was then photoengraved and turned into a three-dimensional relief printing surface. This process was a relatively complex one involving not only photography and engraving, but also mechanical steps such as sawing the plate, routing out unwanted areas, mounting the photoengraved plate onto a lead-alloy base, and planing it perfectly flat.

Even the "phototype" for headline input had its limits, though, and apparently couldn't be used for very large "banner" headlines. If these were to be present, they were "set from fonts of flat engraved individual letters" (p. 38) taped down to the lead photoengraving base along with the photoengraved plate, in holes routed in the photoengraved plate.

This potentially composite relief printing surface was then stereotyped and the resulting cast, curved stereotype plates used to print the newspaper on rotary presses.

This system had one serious drawback (aside from excessive labor): it was slow. In particular, since the photoengraving process was being pushed beyond its normal limits, it took too long (well over an hour) to allow for the introduction of last-minute breaking stories. (Remember, this was in a day when newspapers did more than simply reprint corporate and government press releases as if they were news.) What they needed was a way to insert material into the already photoengraved relief plates just before they were stereotyped and printed (stereotyping was, by that time, a very rapid process taking only minutes).

The Graphotype provided their solution. Breaking news stories were keyboarded as relief type on regular Addressograph/Graphotype plates. These were then cut to size and backed up to appropriate height to match the photoengraved plates, and inserted into holes routed in the photoengraved plates.

So, for at least a while, albeit under duress, the Graphotype was used as a substitute for the Linotype.

3. Forgotten Fame

As an interesting historical aside, this strike occurred during the 1948 Truman/Dewey presidential election, where another Chicago newspaper, the Chicago Daily Tribune produced the now famous headline "Dewey Defeats Truman" Apparently the Tribune was also using the Graphotype in strike conditions, and one of the Graphotype "slugs" in the text of this story was printed upside-down.

A look at online images of the newspaper (generally from sellers of surviving copies) reveals that this inversion occurred in the story in the far right-hand column of the front page. (It is also interesting to see the appearance of a Graphotyped story.)

4. Compounding an Error

As a sort of an ironic compounding of errors, even today online accounts purporting to give the "real" history behind the "Dewey Defeats Truman" Tribune issue printing error seem at times to have been done without any knowledge of the production techniques used. A 2008 article in a respectable online publication, for example (which I'll leave unidentified here because it is in general a good article and I don't really wish to be too critical), asserts that the "replacement staff" were using a "typewriter" for this story and that the inverted material in question was "typed upside down." Although this article is laudable in its intent, this contains one potential error in the understanding of management and production and two certain errors in the understanding of technology.

First, although the Daily News (at least, and probably also the Tribune) hired temporary Varitype operators, it isn't clear that the late-breaking stories would be set by them. The whole point for these stories was speed. While they might have been keyboarded by an inexperienced temporary operator, it is more likely that they were keyboarded directly by the editorial staff, and indeed in the story reprinted here, Kesler writes of Graphotyping stories himself.

Second, and most important for the technology discussed here, the material was not produced on a "typewriter" but on a Graphotype.

Moreover, you can't type anything "upside down." Even if the story in question had been typed on the Varityper it would have been typed right-side up. The inversion happened during page makeup, when the slug made of the Graphotyped plate was inserted into the electrotyped page upside-down (possibly seconds prior to stereotyping, and minutes prior to printing, given the pace of newspaper production). The idea that something might have been "typed upside down" seems to me to come from a 21st century technological mindset where text appears on the computer more or less as it is typed, and printing is just securing a hardcopy of the screen. Because printing is just a version of the screen, to the 21st century mind the appearance of inverted text in print implies that somehow it must have been typed upside-down in the first place. Yet to think that somehow a 1948 newspaper went directly from a keyboard (whether of a Linotype or typewriter) to printed paper is to erase from our historical memory the processes and professions involved in composition, makeup, electrotyping, stereotyping, and for that matter printing itself.

This may not seem important; they're just obsolete printing machines, after all. But our civilization is built on and with machines, and the period from the late 19th through the mid 20th century was shaped largely by the (machines of the) printed media. If erasing from our memories the principles by which these machines operated seems unimportant, consider what you might think of a similarly deep erasure from memory of the political situations which they reported - "Orwellian" comes to mind. If you can believe in upside-down typing, you can believe in upside-down logic.

5. Source

Here's the story from The Inland Printer