The Tragic Saga of James W. Paige's Marvelous Typesetting Machine. By Richard L. Hopkins. Hill & Dale Private Press and Type Foundry, 2014. 15pp. Paperback. $20.00 (postpaid)

The following review appeared 24 October 2014 on the Mark Twain Forum.

Copyright © 2014 Mark Twain Forum
This review may not be published or redistributed in any medium without permission.

Reviewed for the Mark Twain Forum by:
Kevin Mac Donnell

The word marvelous is a bit of a spoiler, and tells you right away what author Richard Hopkins thinks of James W. Paige's invention. Most Mark Twain scholars familiar with the oft-repeated story of James W. Paige and Mark Twain would agree with Tragic Saga but instead of Marvelous, they would substitute words like unreliable, infernal, and insane, and none would expect to have their minds changed by a privately printed 15 page pamphlet. But Paige's machine was marvelous and this is no ordinary pamphlet. Rather than a "Brief Note" for a 15 page paperback, this work merits closer scrutiny.

Richard Hopkins is a seasoned printer who did not merely set the type for this booklet, but cast his own fonts, set the type, and printed it himself. The double-page illustration of Paige's machine, reproduced from a 1901 issue of Scientific American, is printed via color offset and the text is letterpress, and the front wrapper reveals the title through a die-cut oval. Hopkins's typography and press work are impeccable, and so are his credentials. Hopkins, who established his press in 1963 and his typefoundry in 1971, owns over 3,000 fonts of typefounding and Monotype matrices, seven operational casters (he can cast type from 5 1/2 to 72 point!) and among his printing presses are a Vandercook Universal, a Heidelberg platen press, an antique Golding Pearl platen press, and some smaller presses. Twainians without printing experience may not fully appreciate the forgoing, but anyone with a printing background may be excused for drooling in admiration. Hopkins is also an experienced Linotype machine operator (the compositor that defeated Paige's compositor). He keeps all of this old machinery working in order to produce beautiful work like the present booklet and to keep the old traditions associated with letterpress printing and metal type-casting alive. Who better to reexamine what we think we know about Paige's achievement?

The familiar story of Paige's compositor is that Paige embarked on a hopeless quest to mechanically replicate the complicated art of human type-setting, and that Twain, because of his own youthful experience as a "type-sticker" and his perpetual dreams of riches, was lured into this scheme that resulted in a mechanical monstrosity that was horribly unreliable and was justly defeated by the far superior Mergenthaler Linotype Compositor, which in turn led to the bankruptcy of Twain's publishing firm, Charles L. Webster & Company, and to Mark Twain's personal financial ruin. A well-documented narrative of the history of the Paige compositor by Corban Goble ["Mark Twain's Nemesis: the Paige Compositor," The Journal of the American Printing History Association, 36 (1998) pp. 2-16] is the best account published so far, and corrects some of these misconceptions but Goble did not provide the kind of in-depth technical analysis of how the machine actually worked that is provide by Hopkins. This lack of understanding the machine itself has made Twain's faith in Paige's machine seen like one more example of his self-delusion when it came to investments. A critical (but misinformed) account of the venture reported by John S. Thompson in The Inland Printer in February 1902 added credence to this now familiar history of the Paige episode in Twain's biography. Hopkins ably debunks some of the key elements of this story, and explains how Paige's machine was actually a technical triumph.

In 1916 an encyclopedic review of printing technologies, Typographical Printing Surfaces, was published by Lucien A. Legros and John C. Grant. This book contained a detailed description and analysis of the Paige compositor, with diagrams, which debunked the misinformation presented by Thompson fourteen years earlier. Legros and Grant interviewed Paige's assistant and his patent attorney as well, and applied their own expertise to their analysis of his machine. Hopkins begins with this account and explains the complicated technology, the true genius of Paige's accomplishment, in layman's language. Paige produced only two working machines. Hopkins gives a brief history of the fate of those two machines that is, unfortunately, inaccurate but was widely accepted as fact for years. The University of California Press's Autobiography of Mark Twain, Volume 2, 2014, (p. 505, note 80.19) has helped set the historical record straight. Paige's 1894 machine was loaned to Columbia University where it was later disassembled and discarded, possibly melted down for a scrap metal drive in World War II. His 1887 machine was loaned to Cornell where it was displayed from 1898 to 1921 and then returned to Mergenthaler Corporation who in 1957 donated it to the Mark Twain House in Hartford where it can be seen today.

Paige's typesetter weighs 7,550 pounds and has about 18,000 parts. It was designed to assemble type, justify that type, and redistribute the type. The distribution system worked independently (and simultaneously) from the rest of the machine. Rather than make its own type like a later Monotype machine, or cast lines of type like its chief competitor, the Linotype machine, Paige's machine actually composed foundry type just like a human type-setter. The Linotype machine did not require a redistribution system since the lines of type it cast were melted and used again to cast new lines of type. However, it should be remembered that Paige's redistribution system operated automatically and independently from the rest of the machine, requiring no time or effort by the operator.

Paige's machine included an innovative keyboard unlike the familiar QWERTY keyboard, and Hopkins explains it well. It was designed with certain letters arranged in close proximity, based on frequency of use, and it allowed some words to be set in correct order even if the keys for those letters were all pressed at the same time like a chord. The operator could also insert "pi" characters (non-standard type like italics) without impeding the smooth operation of the machine. This led to incredible type-setting speeds even in the hands of inexperienced type-setters, routinely equal to the speed of twelve of fifteen hand compositors, and twice the speed of an experienced Linotype operator.

Hopkins also describes Paige's brilliant way of handling justification (spacing out the words in a line to maintain even right and left margins). As the operator set type, an arm dropped down every time a space key was tapped, and when the operator tapped a key to signal the end of a line, the machine would calculate the amount of space needed in each space, and the arms would retract as the correct spaces would drop into place, within a tolerance of .005". Leading (spacers) between lines was automatic as each line of type was fed into the receiving galley. In addition, as pages of type were completed, receiving galleys full of type could be removed and empty galleys put in their place without disturbing the operator. Likewise, dead matter (type that had already been used to print or to make a printing plate) could be slid into the channels and automatically redistributed with no effort from the operator. Every piece of type was nicked like a key and fed into the correct channels by itself. Foul, stuck, or damaged type was automatically ejected through an ingenious system, and the machine had an automatic "stop" mechanism if undue strain was placed on any part of the machine, and an indicator dial to tell the operator where the problem was located. There were no carrier belts, or gravity devices, and redistribution channels fed from the bottom to avoid jams. The whole machine was powered by a 1/12th horsepower motor using a 1/4 inch belt that turned a fourteen inch pulley that could be rotated with one finger.

Hopkins points out that the 1894 test at the Chicago Herald produced astounding results. Paige's machine, even with "all delays counted against it, delivered more corrected live matter to the imposing stone, ready for the forms, per operator employed, than any of the thirty-two Linotype machine in operation…" This is a most revealing statement. Anyone who has operated a Linotype machine (Richard Hopkins and this reviewer have both done so) will remind you that those machines are notoriously balky and prone to jamming and breakdowns, and that they require a good deal of practice to become proficient. The earliest models of the 1880s and 1890s--Paige's competition--were no doubt even less reliable than the newer (and still balky) models operated by anyone in the last fifty years. The Paige machine broke down no more or less often than its balky competition and set more type, and set it faster, even when inexperienced operators sat at its keyboard. So, why did it fail? As this statement makes clear, there were already thirty-two Linotype machines in operation when Paige's machine arrived for testing.

Hopkins attributes the failure of the Paige compositor to two key factors. Paige's constant delays in allowing his machine to be brought to market allowed his competition to gain huge footholds. By the time that test took place in 1894, Linotype machines were in common use in all major newspaper offices, and a few years earlier, Ottmar Mergenthaler and John R. Rogers, Paige's chief competitors, had settled a patent infringement lawsuit which removed Rogers' machines from the American market. An entire 500pp. book set by Linotype machine (Henry Hall's Tribune Book of Open Air Sports) had been published in 1887, within a year after the first Linotypes were installed at the Tribune. Although Paige began his work in the 1870s, he was the last to bring his product to market.

If being late to market wasn't bad enough, the second reason proved fatal. Paige stated the cost of his machines to be $20,000 each. The Mergenthaler Linotype machine could be purchased for $3,500. Even Mark Twain could do that math, and he calculated that in order for Paige's machine to be price competitive with the Mergenthaler machine, it would have to run flawlessly, twenty-four hours a day, seven days a week, and no Mergenthaler or Paige machine could do that.

Thompson's article had indicated that the filing of the patent had delayed things, and went so far as to claim it took eight years, that two patent office workers died during the process (one driven insane by the paperwork) and that the patent attorney who filed the case died in an insane asylum. But Legros and Grant were able to track down Paige's patent attorney, D. H. Fletcher, alive and well and apparently unshackled and sane, who confirmed that there were no special delays associated with the filing, but rather the usual "rather orderly give-and-take interaction with the Patent Office prior to gaining approval." Fletcher called Paige's machine an "intellectual miracle."

It was indeed a tragic saga, and had Paige been able to bring his invention to the market sooner and at a competitive price, the saga might have ended differently. Paige certainly made a mistake in trying to imitate human type-setters instead of thinking of a new and less complex method of type composition like Mergenthaler, and this no doubt contributed to the delays and enormous expenses, but he did indeed produce the "intellectual miracle" recalled by his attorney. Hopkins gives us the most expert analysis of Paige's technical achievement yet, and leaves no doubt that Paige's invention was a technical success that was a financial failure because it arrived past deadline and over budget.

A note on ordering copies of this booklet: This finely printed booklet is printed in small folio format that would have required 30pp. in normal octavo format. Only 150 copies were printed and it is a handsome piece of printing that should hold or increase its value over time as a Mark Twain or printing history collectible. Only about 40 copies remain as of this writing, and can only be ordered directly from the Hill & Dale Press, 169 Oak Grove Road, Terra Alta, WV, 26764. A check ($20) should be enclosed with the order.