Chicago, June 1948: the calculating-machinery concern acquired in the last number stages its demonstration at last, before the government's buyers and the press, and the author is made a founder by a typist's grammar. He was minted once before, in a desert; the reader is invited to watch the die come down a second time, and to note whose hand is nowhere on it. —Ed.
The concern Sennett had bought us a piece of did business on West Adams Street, in a former piano warehouse, under a name I will improve to the Marquette Computing Machines Corporation, the original being too easy to look up.1 The front room was mine by charter, and on its desk the little eagle-stamped gyroscope off a certain railhead did duty as a paperweight; visitors took it for a piece of the machine, and I let them. Behind the partition stood the machine itself, half an acre of switchboard on four thousand tubes, warm as a bakery and about as temperamental; my whole legal standing toward it was my father-in-law's sentence over the chop: don't touch the machinery.
Sennett's two professors were Dr. Ottley, seventy and deaf on the left side, who had designed the logic, and Dr. Kessler, fifty, who had built the circuits and did the worrying for both. Under them ran nine engineers; the one that matters here I will call Stan Zaremba, twenty-six, off the South Side via a signal battalion, who owned the arithmetic unit the way a man owns his wrist. Kowalczyk's soldering iron had left me radio enough to follow him at a distance. "You ask good questions," he said my first week, "for a man who's not allowed to touch anything." I said the not touching was the hardest work I did, and we got on.
My duties were to be presentable at the money end and to produce a demonstration. I began it whistling into the Loop in autumn '47 and delivered it the following June; the months between went to tubes ordered and not shipped, the multiplier catching fire quietly on a Sunday in March, the Bureau of Standards twice postponing, and Sennett wiring HASTE. I spent the winter learning the vocabulary, pulse and gate and acoustic delay, rehearsed as I had once rehearsed cyclotron: not to understand, but to say half a beat ahead of any man who might check.
The demonstration was set for the eighth of June, and the announcements went out of Sennett's Washington office, drafted from his wire by a girl taught that a telegram is English with the money squeezed out. The wire read: DEMONSTRATION JUNE 8 ELECTRONIC CALCULATING MACHINE DEVELOPED CHICAGO ARRANGED PRESENTED KLIPSPRINGER. What the mimeograph put on every interested desk in Washington was "a demonstration of the electronic calculating machine developed, arranged, and presented by Mr. J. Klipspringer of Chicago."2 The Bureau's acceptance came back inside the week, thanking us for the chance to inspect "the Klipspringer machine."
I ought to report that I reached for the telephone. I did not, having been minted once already and knowing the feel of the press coming down. In '45 the papers made me a hero of the atomic desert on no evidence but proximity and a clean uniform; my own wedding notice ran under HERO OF THE ATOMIC DESERT WEDS, and I corrected nothing then either, a correction being the only story no editor will print twice. In Philadelphia I had watched better men lose the computing age off a mimeographed cover: credit rides the cover, and the cover belongs to whoever the typist believes it does. My second gift was standing next to the right man wearing the right face; the amendment of 1948 was small: the right man was a machine.
The eighth came in at ninety degrees, and the machine room, tubes lit, made a fair Turkish bath. Folding chairs for a dozen: Dr. Renner of the Bureau of Standards; Mr. Pell of the Census, who clicked his mechanical pencil when thinking; a colonel from Ordnance; and four pressmen who had discovered the electronic brain the winter before and had it playing chess, forecasting weather, and replacing Congress; not one of them ever asked whether it could add, adding being a claim a man could check.3 Kessler wore the expression of a man whose daughter is marrying badly, Ottley beamed from the wrong distance, and Zaremba was behind the racks, out of sight.
I opened with the payroll race, Miss Dunn from the front office at her comptometer against the machine on a week's wages for twelve hundred men. The machine threw a card feed and Miss Dunn beat it flat, apologetic in her flowered dress; I took the laugh, declared a moral victory for Illinois womanhood, and had the machine rerun the table clean in under two minutes, which set them murmuring. The colonel asked about fire control, and Kessler got out "The question is not well posed" before I was in front of him promising to compute anything the government cared to bring us.
Then Pell clicked his pencil and dictated a multiplication of his own, two six-figure numbers off his pad, one ending in a seven, the other in a three. I felt nothing; multiplication was the machine's bread and butter. Zaremba set it on the switches, the room leaned, the teletype woke and put up its twelve digits, and Pell looked at his pad and said, in the mild voice statisticians keep for disaster, "Your last digit is a four. Seven times three, gentlemen. It should end in a one."
Then, while Kessler was still assembling a sentence, the hum went down — not out, but down a half-tone, four thousand tubes remembering they were mortal, and every lamp on the monitor board sat dark. I have stood in a desert waiting to learn whether the air itself would catch fire, and I was calmer there than on West Adams Street that morning. The sweat came through my shirt in one motion, my legs went to wet rope, and I stood smiling at the government of the United States with my mind fixed on a water stain on the ceiling shaped like Florida. Behind the racks a panel came off; I heard Zaremba say one word to God about the multiplier, then his trouble light going rack to rack.
So I did the only work I was ever fitted for: I stepped in front of the dead board the way you step in front of a man whose trousers have failed, and I talked. I gave them the machine checking its own arithmetic before publishing. I gave them the 1950 census, a hundred and fifty million Americans coming at Mr. Pell like weather; payroll, ballistics, insurance; a machine that never slept, never struck, and never asked for a raise, and whether that was thinking I would leave to the editorial page. I had covered a German's politics with this voice and run a bridge at Juárez on it; four minutes of dead machinery was long, but it was level ground. The colonel was laughing, Renner taking notes. Behind me a tube came out of its socket like a cork, and Zaremba's voice came flat through the racks: "Ask it again."
"Gentlemen," I said, "the machine will now publish," as though the whole silence had been mine, and nodded at the racks like a conductor. The hum came up, the board lit, the teletype ran its twelve digits over, and the last of them was a one. Pell worked the product longhand, a full minute, then said "Correct," and the government of the United States applauded a repaired tube. Flash bulbs went off. Ottley bowed, hearing his own guess of the occasion.
The wire-service photographer wanted one more, and posed it himself. "Mister, put your hand on it. Like you're listening to it." I hesitated, the record may show, having kept the one commandment for ten months. Then I laid my palm flat on the main frame, warm as a horse, and looked modest; and that is my entire contribution to American computing machinery, honestly reported: one hand, four seconds, for the flash. The Tribune man read the circular back with his pencil up. "Developed, arranged, and presented by. That's you? Developer?" Kessler was behind the partition with the dead tube; Ottley was pumping the colonel's hand. "The machine speaks for itself," I said, and he wrote developer.
The Tribune gave it four columns, CHICAGO ELECTRONIC BRAIN DOES A DAY'S FIGURING IN FOUR MINUTES, and under the photograph: "Jay Klipspringer, developer of the electronic brain, with associates." The associates were Ottley's left shoulder and a fire extinguisher. The trade weekly did worse, or better: two pages on "the Klipspringer calculator," with a diagram Kessler had drawn and no Kessler anywhere on it. Ten days on came the Bureau's letter of intent, tests and funds contemplated, and the machine named in it the Klipspringer computer. The misattribution was setting like concrete.
Zaremba came into the front room with the weekly rolled in his fist like a man after a wasp. "Developer," he said. "You couldn't wire a doorbell." As it happens I could, Kowalczyk saw to that much, but a man in my position learns which corrections to skip. "Write to them," he said. "One paragraph. The logic is Ottley's, the circuits are Kessler's, the arithmetic unit is mine, and you're the man who books the room. They'll print it. It's true." I was fond of him, so I gave him what Philadelphia had cost me a year to learn: paper carries names, not truth, and this name was set; the Bureau had asked for Mr. Klipspringer, and the money liked what the Bureau liked; his letter would run cut to two sentences under an advertisement for solder. He wrote it anyway.4 He stayed on too, which is the sorrow of the trade: the credit men can always afford to rob an engineer, because the engineer can't leave the machine.
Kessler I expected more trouble from, and got fifteen minutes of it, German at the edges, until the Bureau's letter lay on the table and he had read it twice and sat down. "They will send money to this Klipspringer machine. They would not send it to an Ottley-Kessler machine?" I said Washington had met neither of those gentlemen at dinner, and that was the whole fight. A government buying what it cannot understand does not want a schematic; it wants a face it can seat at a hearing, and hires the understanding afterward, by the day, to be overruled. Ottley never learned there was anything to forgive; he read nothing wider than a column of figures, and died in '54 believing the papers had been kind to us all.
And there is the craft, free to the apprentice, the patent having run out with all the others that year. Never claim the machine; a claim can be checked, and checking is fatal. Arrange to be the man the room already knows, standing nearest the works in a good suit, saying modest things, and let typists and captions and letters of intent do the claiming. When the wrong version goes up, taking it down is nobody's job, least of all yours. I never once, in fifty years, said I built anything. I never had to.
I wrote the board of one its two true sentences, the demonstration went well and Chicago is hot, and Tuesday's three lines came back: the lake flat, the roses blighted, Daddy reading the clipping twice at breakfast. Then, under the signature, a fourth line, the first in three years of Tuesdays: "Mind the machinery, darling." I have not finished construing it. Sennett wired: TRIBUNE SEEN. WASHINGTON WARM. FOUNDER SUITS. DO NOTHING. I complied with the whole of my strength. Since 1943 I had kept my name off paper as a religion; now it stood in the government's own type, in four Tribune columns, on a trade weekly's diagrams, and not a word of it in my hand. Paper testifies for whoever holds it last, and by July the whole country was holding mine.
So the ledger. One title, founder, conferred by a typist and ratified by the Bureau of Standards. One block of founder's stock, worth rather less than the certificate it was printed on. One framed letter of intent in a front room where I touched nothing. One photograph, palm on frame, four seconds of contact, the whole engineering career of the father of the computer age; the costume was only basted that June; the proper fitting came later, on television.5 The caption has never come out of the record; ask the hero of the atomic desert, we are the same man. I rode home along the shore adding it up, twenty-five years old and founded, and if my celebrated nose itched all that fine summer, I put it down to tube heat. It was not tube heat. Away east, on the other set of books where I was also entered, men were finishing an arithmetic of their own, and the good years had begun to run out. That bill is the next number's business; it came addressed to every man who had ever stood near the first bang.
The concern is identifiable from the minute-book of Halsted Instrument & Export (No. 11, note 4), which dates the purchase to the autumn of 1947 and the demonstration to June 8, 1948; the author renamed it, as he renamed by lifelong habit everything he meant to keep, and the editor sees no need to undo it. The commercial setting is as he gives it: in 1948 the market for electronic computation was the federal government almost entire. The Census Bureau had contracted through the National Bureau of Standards for the Eckert-Mauchly UNIVAC in 1946; Engineering Research Associates of St. Paul built for the Navy; the universities built for themselves. A Chicago concern seeking custom would deal with the Bureau of Standards exactly as described, and a demonstration before its examiners was the standard gate to funds. —Ed.
Three documents survive in the Sennett office files. The wire of May 1948 reads as the author quotes it, eleven words without punctuation. The mimeographed circular typed from it carries the phrase "developed, arranged, and presented by Mr. J. Klipspringer of Chicago." The Bureau's acknowledgment, adopting "the Klipspringer machine," followed within six days, and the phrase recurs in every federal paper on the machine thereafter. Three documents, three authors, none of them the author of these papers; the files show no correction sought by him or by anyone, which is the fact the reader is asked to carry forward. —Ed.
"Electronic brain" fastened itself to computing machinery at the ENIAC unveiling of February 1946 and could not afterward be got off; the press of the period had the machines composing music, playing chess, and forecasting the weather, usually in the same column. Whether a machine could think was asked at nearly every demonstration on record. The author's account of what was not asked is consistent with the surviving coverage. —Ed.
The engineering record of the machine survives, and the editor sets down what it shows. The logic design belongs to the elder professor and the circuit design to the younger; the arithmetic unit, and the demonstration-day repair (a failed tube in the multiplier, found and replaced in under five minutes), belong to the engineer the author calls Zaremba, then twenty-six. The author's name appears on no drawing, no logbook, and no patent paper of the corporation, an absence the reader will by now recognize as his preferred seat. Zaremba's letter to the trade weekly survives in carbon, five paragraphs, naming every man who built the machine and not mentioning the author; the weekly's issue of July 1948 prints two sentences of it, at the foot of a page, beneath an advertisement for flux. The robbery was not local: the decade assigned whole machines to the names on covers and captions while the builders went to the appendices (No. 9, notes 1 and 4). The difference here is only that the name on the caption had built nothing at all. —Ed.
"Father of the computer age" does not attach to the author in print until the election night of November 1952, the business of a later number; what 1948 conferred was "developer," and no document in the record ever withdrew it. As to his summer's arithmetic: the demonstration of June 8, 1948 stood fourteen months and three weeks from the morning of August 29, 1949, when the Soviet Union fired its first atomic device, announced by the President on September 23. The good years he counts at the close were therefore fourteen months and some days. His nose, the record shows, was on schedule. —Ed.