The Computeriter revolution

A Utopian fiction

Our Spring 1963 issue included the only piece of science fiction CJR has ever published. Reporter Edward Edelson imagined with a mix of prescience and parody what computer automation might mean for the news business. The revolution began, dear reader, on a September day in 1967, when residents of a small Pennsylvania town awoke to find that their paper contained “the first non-human newspaper stories ever written.” The below paragraph originally ran above the piece.

The following is a work of the imagination; it may be one of the first items of science fiction written about the newspaper business. Its author is Edward Edelson, a reporter for the New York World-Telegram & Sun, who has been following closely the advances of computers into newspaper printing.

By the middle of the 1960’s, the experts were so certain of what would happen to American newspapers that there were no longer any debates about it. There were going to be fewer metropolitan newspapers, suburban papers would flourish, competition would fade sadly away, and television would continue to elbow newspapers into the tightest of corners. Within twenty years, it was freely predicted, every newspaper would either buy its way into a monopoly or be bought by a chain. With the subject becoming too hackneyed for hand-wringing, there was little reason to look for any revolutionary change in the nature of newspapers. It was just at this time that the revolution began.

Its father was George Hilburton Hargraves, who inherited the Bartlesberry (Pennsylvania) Clarion from his father early in 1965. The Clarion was a 7,200-circulation daily that had declined somewhat from its barely perceptible peak, reached early in the forty-three-year reign of the elder Hargraves. It was a stodgy, respectable sheet, not much worse than anything else in its class and firmly profitable. Hargraves père had paid little attention to anything but its net profit and his son, then barely 30, might have been expected to let things rock along as before.

But the younger Hargraves, a Harvard man, already a world traveler and something of a scientific dilettante, restlessly began making changes. Roaming through the Clarion’s rambling, outmoded plant, he asked unexpected questions, taught himself the printer’s trade and then began spending for new equipment. After six months the Clarion was making 20 percent more profit on the same advertising revenue. Now Hargraves read the paper thoroughly for the first time.

He had taken some interest in the Clarion’s news columns, as much as his schedule allowed, but he had few ideas about style and was indifferent to content, once he satisfied himself that the community was getting what it wanted. What impressed him most, as he read the collected back issues of the Clarion page by page, was “the sheer repetitiveness of the day-today items that made up the bulk of the news we printed,” as he wrote later. After a month of reading and classifying, he had the broad outlines of a new species of newspaper firm in his mind.

“The notion came to me,” Hargraves wrote, “as I read through the women’s pages of the Clarion. I read column after column of wedding, engagement and tea party stories, most of them very nearly identical except for names. It struck me one afternoon that there was no reason for a reporter to spend his time writing out the immutable formulas for births, engagements, weddings and deaths. Surely, here was the ideal place for science to save me money.”

The idea of turning machines to work on words was not unknown at the time. Computer translations of scientific papers were a commonplace, and some experiments in related fields were under way. Teletypesetting—the non-human setting of type using punched tape—was already common in the newspaper business. But though a method had been developed that allowed a reporter to punch a Teletypsetting tape as he wrote his story, automation for the most part had bypassed the news room. It was Hargraves’s vision of a new form of newspaper and his firm grasp of economics that were to change this.

He turned for help to Quincy Snyder, a resident of Bartlesberry who had just left his engineering job with Sperry Rand to experiment with unorthodox ideas about computers. Snyder, nearly as young as Hargraves and possibly a more zealous worker, became a perfect collaborator. During the weeks that followed, the two men held long, intense discussions in Hargraves’s cramped office, hammering out solutions to the problems that arose from Hargraves’s suggestions. Snyder proved to have as solid an understanding of newspaper economics as did Hargraves.

“It would have been relatively easy to automate almost immediately,” Snyder has written. “But cost, more than anything, was the complicating factor. I had to remember at every turn that we were trying to compete with the relatively low per-hour price of newspaper labor. It was necessary to cut every possible corner to bring the project in at a practical cost.”

It was three months before the theoretical work was done and they could start building the machinery. Here a stubborn refusal to admit the complexity of their job was an asset for Hargraves and Snyder. If they had accepted the then current scientific opinion of the project’s chance of success, they might never have begun. Historians of technology have compared their leap ahead of their times, using what most experts would have called inadequate means, to that of the Wright brothers. The first Hargraves-Snyder computer was a ludicrous failure; the second, although it already incorporated much of the radical and ingenious equipment designed by Snyder, was not much better. As his expenses mounted Hargraves was tempted to sell the idea to obtain financing. At Snyder’s insistence he went into debt to pay for succeeding machines. More than a year after work had begun, after many false starts, they prepared for a practical trial of Number Six, the first workable Computeriter—a name we owe to Hargraves.

The trial came on the night of September 17, 1967. We have Hargraves’s description of the scene—scoffing reporters standing by as a punched tape was fed into the four-foot cube that held Number Six, printers lounging in from the composing room to watch, the distraught women’s editor wringing a handkerchief at her desk, and the reflective silence as the final tape was fed into a linotype machine to produce, with a clatter, its news stories. The test was a success, of course, and the Clarion’s readers were given, the following morning, the first non-human newspaper stories ever written.

For all its historic importance, Number Six was a disappointment to Hargraves. Barely two columns of type were written and set by the Computeriter on its first run, and hardly more than that on succeeding days. Hargraves was able to reduce his staff somewhat—the women’s editor went—but the savings he had hoped for eluded him. With the pressure of his debts increasing day by day, Hargraves, over Snyder’s protest, sold his share of their patents to the Morgan Corporation, with the proviso that Snyder would keep his rights to the Computeriter.

It was a bad move. Hargraves sold himself out just before the dust of a stampede toward the Computeriter appeared over the horizon. Morgan soon scrapped its original cautious plans and began turning out Computeriters at top speed in a hastily built assembly line constructed, under Snyder’s supervision, in a vacant ice cream plant near Bartlesberry.

The need was there and the Computeriter filled it. Smaller newspapers were the first buyers; most large papers delayed, appalled by the crude copy that came from the first clumsy Computeriters. But the Computeriter was improved steadily, and publishers all over the country leaped to put the new machines into operation. One evidenee of the trend was the situations-wanted section of Editor & Publisher, which nearly doubled in size within the year. The amount of Computerited copy (the term was denounced as a barbarism but became common usage) increased at an exponential rate.

With the Computeriter turning out 60 percent of the nation’s obituaries, far-sighted men began to see greater possibilities for it. Frank Jaspers, publisher of the Haven (New Mexico) Sun-Star, wrote: “Why should I pay a reporter $75 a week when a Computeriter can do the same job and give me a depreciation write-off to boot?” It was not that simple, however. It soon became obvious that only the wire services could afford to pay the high costs of Computeriting any appreciable amount of hard news copy. (The New York Times announced tersely that, aside from obituaries, it would not use the Computeriter—a resolve it maintained for six years.) The Associated Press signed with RCA, United Press International turned to IBM, and both were soon able to announce great advances. But both found their finances strained.

The sheer number of facts to be programmed in all but the most routine news stories made for slow, expensive work. Old Number Six had three programs for wedding stories alone. One began, “Announcement has been made…”; the second: ________ (Church) was the scene __________ (day) of the wedding…”; and the third: “______ (Name of bride) became the wife ________ (date)…” Number Six had four programs for obituaries, two for engagements. The wire services, on the other hand, found that, aside from the simplest and most predictable news events—religious holidays, highway death counts, and the like—the number of alternate leads to be programmed was almost prohibitively large. Both AP and UPI pressed ahead, however, and by early 1972 32 percent of AP copy was Computerited.

The only answer given their more vocal critics by the wire services, and a truthful one, was that insistent demands from publishers were responsible for the headlong pace of Computeriterization. Indeed, publishers and their trade organizations continued to call for more Computerited copy despite pleas by the wire services for more time to eliminate crudities.

The successes of Computeriting were so great that many observers looked ahead to the reporterless newspaper. We find Hargraves (who bad gained the same sort of fame as had, in another era, John Augustus Sutter) predicting that no reporters would be needed after 1980. But it was found that some reporters would always be needed. Someone had to gather facts; not all stories could be Computerited; and some human work was invariably needed on Computerited copy. The demand for reporters stabilized, after the initial panic, at slightly above half its former level. Of the many labor disputes caused by the Computeriter, the most significant was the 310-day New York strike that broke the power of the Newspaper Guild, removing the last barrier to the fullest use of the Computeriter.

The problem of the “obsolete reporter” (a phrase coined by a Presidential Task Force on Journalism) was solved rather quickly, considering the tumult it aroused. Many reporters found work in neighboring fields, since only the largest public-relations companies could afford the Computeriter and it was found almost useless in advertising.

Schools of journalism felt the effects of the Computeriter almost as drastically. It was a curiosity of the times that the journalism schools, at the moment when there was the least use for their graduates, were flooded with contributions from publishers made affluent by Computeriting. The New York University School of Communication Arts, in an almost unmatched action, refused a $500,000 gift from Quincy Snyder with the explanation that it found no reasonable use for the money.

The revolution, by then, was very nearly over, and evolution went on more placidly. There were some chronic complainers, but most professional students of journalism took the balanced view. The typical newspaper office of today, after all, with its quiet bank of Computeriters facing a small island of desks, may easily be preferred to the rubbish-littered news room of pre-Computeriter days. There were some excesses; few defend the programmed editorials eagerly adopted by many small newspapers, and many miss the more varied style of individual journalism, although only the most nostalgic regret the passing of the typographical error. The hand-written news story is in the same category as the hand-made pair of shoes: admirable, but a trifle too expensive for everyday use. American newspapers, everyone acknowledges, are on a firmer financial foundation than ever before, and they are printing more news more quickly than ever.

Ten years after he had his original vision, George Hargraves was asked by an interviewer if he regretted his role in the Computeriter revolution, “I only regret not having enough money to see it through,” Hargraves answered. “We have the greatest newspaper system in the world, and the Computeriter is the heart of it. The Computeriter is American journalism’s finest hour.”

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Edward Edelson was a reporter at the New York World-Telegram & Sun. He later was science editor at the Daily News, and president of the National Association of Science Writers.