Planted by his wife's money at the founding of the inventors' company, the author takes up the minute-book again and spends a year recording a business he does not understand, in rooms where the age's chief property is argued away. The conference he reports from April 1947 is in the record, down to the ruling; his seat at it is not, for the reason he gives. —Ed.
Sennett's wire beat my shirts back from the hotel laundry: MONEY CURIOUS NOT COMMITTED. NEW COMPANY FORMING. SIT IN. BE USEFUL, BE QUIET. REPORT TUESDAYS. The company was Eckert and Mauchly's, promised over coffee, and it opened that summer on an upper floor at 1215 Walnut Street: two rooms, a drafting table with one bad leg, and the Census Bureau's quiet man already writing letters about delivery dates. I attended as the Chicago money's observer, a trade without duties, and when the first meeting asked who would keep the record, I said "I will" before the echo died.
I commend the move to any young man traveling on manufactured papers: take the minutes. The talkers watch each other; no man alive watches the pencil, and you walk out of every meeting holding the only account in existence. It served me at Los Alamos; it served in Philadelphia inside a month.
What I recorded, heaven knows. Eckert dictated circuits at the speed of a tobacco auction: tanks of mercury, pulses, a thousand words of storage, the words being numbers, which was explained to me twice. I wrote down "mercury" and "delay" and trusted them to be adjacent by morning. "You don't follow a syllable of this, do you," Mauchly said once, kindly, across the drafting table. "I follow the minutes," I said, and he laughed and let it stand.
The report, meanwhile, went on breeding. I walked my borrowed copy back to the Moore School and found Goldstine snowed under with correspondence, a stack at his elbow wearing foreign stamps. "Tuesday's mail," he said, and dealt them off like a bridge hand. "Cambridge, England. Manchester. A man in Sydney, Australia. They all want the report, and they all get it. Knowledge is not sugar, Mr. Klipspringer; you don't ration it against a shortage. Ten years of war work went into safes. This one thing went to everybody, and it is the proudest thing I ever did with rank on my shoulders."
"And it travels under whose name?" I said, handing it over.
"Under the name of the man who wrote it down." He said it without a flicker, and I looked hard for one. He believed it entire, and believed it was generosity; I have been lied to by the best men of the century and slept sound after, but sincerity I had no tools for.1
It was Eckert, late one night over the bad-legged table, who gave me the pedigree straight. "You want to know where the report came from? Meetings. Two winters of meetings in the back room. Johnny came down from Princeton once a week. Everybody talked, the blackboard filled: my storage in my own chalk, Mauchly's programming notions in his, a dozen men's work on one slate. Then Johnny wrote it up in longhand, on trains, and mailed the pages to Goldstine, and Goldstine had them typed. When it came back mimeographed there was one name under the title. One. I had some of it in writing in January of '44, before the man ever saw the building; the memorandum is in a drawer at the university, which unlocks it for nobody, least of all me."2
Mauchly was at the dark end of the room with his spectacles pushed up. "Johnny doesn't steal," he said, mild as milk. "Johnny arrives. It's the men who hand him the room I wonder about." Then he looked out at the street and said no more, and I wrote the date at the top of a fresh page.
That winter the money sent me to Princeton, to size the Institute's machine for Chicago, and on the local out of Broad Street I found the consultant himself in the facing seat, pencil going like a sewing machine. I knew him from the mesa, where the whole Tech Area walked faster the weeks he came through. He saw the mimeographed cover on my lap before he saw me.
"You are reading me," he said, delighted. "Is it good?"
"I understand one word in ten."
"Then you are ahead of the Congress and level with several deans. Have we met?"
"Los Alamos. Liaison. The kind nobody checked on."
"The best kind," he said cheerfully. "There were more liaisons on that mesa than in a French novel." And we were off. Past Trenton I tried him with a costing question Halsted's people had cabled me, machine-hours against clerk-hours across a Census decade, and he gave the figure while I was still laying out the units, then begged my pardon for the speed.
I fished, fishing being my trade. "The two in Philadelphia believe the report picked their pockets."
He put the pencil away, which I later learned was his fullest form of attention. "Eckert is the best engineer in this country, and the mercury storage is his, and I have said so in print. But make the distinction. The tubes, the tank, the solder: engineering. A man may own engineering; let him patent it and grow rich. What a computing machine is, the logic of it, the order underneath the metal, that is mathematics. Can you own a theorem, Mr. Klipspringer? I put the mathematics in order. A man who puts theorems in order does not send a bill."
"The name went on the cover," I said.
"A name goes on every cover," he said, and shrugged, pleasant as morning. "It is how the mail knows where to go."
He took me to lunch, and I drank his cocktails and watched him work a tableful of professors the way I work a boardroom, only forward: giving credit away by the double handful and gaining stature by the pound, a trick I have never dared try. He was the genuine article, the fastest head I ever stood beside, and I have stood beside them all. The whole difference between us was that when his name settled on other men's thinking, the thinking improved. Mine never improved anything it landed on.3
The lawyers had been circling since the resignations, and in April of '47 Ordnance called the parties to the Moore School to settle whose the stored program was. I smelled the ruling two days out, the way I smell fires, and for once the fire was somebody else's; the company wanted its own record kept, the Army's minutes being the Army's, and there was no discussion of who would hold the pencil. I had become furniture, which is the summit of my profession.
They met on the second floor, ten men at a long table. Eckert and Mauchly sat on one side with their patent attorney. Von Neumann came down from Princeton with Goldstine beside him, the captaincy traded for the Institute. The university sent a lawyer of its own, the institution having discovered at the reading of the will, as institutions do, that it had been the widow all along. And Ordnance sent three patent men in suits the color of weak tea, men who produce nothing and outlive everybody, and I never saw property destroyed more courteously than they destroyed it that afternoon, or closer to an hourly rate.
The senior one had a voice like a form being read aloud. "One question of fact, gentlemen, and then the law. The document of June 30, 1945. How many copies went out, to whom, and under what restriction?"
"Twenty-four at the start," Goldstine said. "Project staff and the interested services. Afterward, whoever wrote and asked. Cambridge wrote. Manchester wrote. Everyone wrote."
"Marked confidential? Restricted?"
"It was never classified. It was knowledge."
"Then as of its distribution it is a publication within the meaning of the statute." No cruelty in it, and no interest either; he might have been certifying a boiler. "The law allows an inventor one year from publication to make application. That year ran out in June of 1946. Everything the document discloses has stood in the public domain since. The stored program cannot be patented by Dr. von Neumann. It cannot be patented by Mr. Eckert or Dr. Mauchly, or by anyone on earth. It is public property, gentlemen, the same as the alphabet."4
The radiator ticked. I have heard verdicts read that took longer and killed less.
Eckert came forward over the table, speaking slowly for the first time in my acquaintance. "The cover says First Draft. A draft. Working paper, for the project's own use. My storage is in that report. Two years of our meetings are in that report, in everybody's chalk, and it went into the mail in another man's envelope, and you sit there and tell me the postage made it the world's?"
"The law does not read covers, Mr. Eckert. It counts copies and it counts months."
He looked across at Goldstine. "You ran the mimeograph yourself, captain. I have seen the ink on your cuffs."
"I sent knowledge to men who could use it," Goldstine said. He was white, but he was level. "There are machines growing out of that report on three continents. Everything else we did in the war went into a safe. One thing went to everybody, and I am asked to repent of it. I decline."
Mauchly took his spectacles off, polished them with the end of his tie, and did not put them back on. "We were building it, you see," he said, to nobody in particular. "While it was being described, we were building it." That was the whole of his case, and the table let it lie where it fell.
The senior lawyer turned last to Princeton. "Dr. von Neumann. Do you assert a claim?"
"No." He sat easy in the hard chair, the only man in the room whose afternoon it was. "I wrote the report because the design seemed to me science, and science is published. If the law has decided the design belongs to everybody, then the law has arrived at my own opinion by a longer road, and I congratulate it on the journey."
"The credit," Eckert said, very quietly, "was not before the committee."
"Credit, Mr. Eckert," said the lawyer, squaring his papers, "is not a commodity the Ordnance Department regulates."
For once in my life I understood every word of a meeting. Since 1943 I had kept my own name off paper as a religion, on the reasoning that paper is loyal to whoever holds it last; here the collection ran the other way entire. The machine had gone to whichever men could keep its tubes alive. The property had just gone to everybody, which is to say to nobody. And the invention, the thing that would be named and taught and remembered, had gone into the mail two years before, riding a cover with one name on it, and no ruling on earth could call the mail home. I ought to report that I grieved for the builders. What I felt was my pulse going, the way it goes at a good race, and I wrote the minute out in a fair round hand and kept no copy, needing none.
Tuesday was two days off, but I wired that night, which Sennett would read correctly as news: EDVAC PATENT DEAD ALL HANDS. REPORT RULED PUBLICATION. ENIAC APPLICATION SURVIVES, COMPANY PROCEEDING. CREDIT UNTOUCHED. CREDIT RIDES THE COVER. ADVISE MONEY MIND WHO SIGNS WHAT.
Bunny's letter was at the hotel, Tuesday's three lines, the lake, her mother's roses, yours, Bunny. I have reported to boards of fourteen and told every chair a different story; the board of one in Lake Forest got two true sentences a week, it being poor practice to bluff a player whose cards you have never seen. Thursday brought Sennett's answer: SO NOTED. REPORT READ IN ENGLAND ALSO. TEDDINGTON BUILDING MACHINE ON PAPER, GOVERNMENT MONEY. CROSSING ARRANGED. LOOK, LEARN, ACQUIRE. NAMES FOLLOW.
I telephoned Princeton, on the theory that the shortest road into anything is through a great man's lunch hour. "Teddington," he said, around something he was enjoying. "The National Physical Laboratory. The ACE; faster than ours, on paper. That will be Turing."
"What is he?"
"He was here before the war. I tried to keep him; he went home to their war instead. Mr. Klipspringer, attend, this concerns you professionally. The whole idea, the universal machine, the report before all the reports, is in a paper he published in 1936, which nine men read.5 We are all of us building his machine. Mine is the one with my name on it."
"I'll give him your regards."
"Do. He will not receive them warmly; he does nothing warmly. He does everything honestly instead, and you will find, I think, that this is worse."
I wrapped the eagle-stamped article back into its grey sock against the salt air, and went up the gangway a fully educated man: I knew what the age paid, where it kept the funds, and whose signature moved them. Teddington would see about my education.
The document is the First Draft of a Report on the EDVAC, one hundred and one mimeographed pages, dated June 30, 1945. Von Neumann composed it in longhand, much of it on trains between Princeton, Washington, and Los Alamos, and mailed the sheets to Goldstine at the Moore School, who had them typed and ran off the first two dozen copies; the title page carried a single name, and the blanks left for references to other men's work were never filled in. Copies were made from copies at every institution the original reached, and the logical plan it sets out has been called the von Neumann architecture from that day to this. —Ed.
The meetings are documented: the EDVAC design conferences at the Moore School, 1944 into 1945, attended by Eckert, Mauchly, von Neumann, Goldstine, Burks, and others, and minuted at the time. Eckert's memorandum of January 1944, "Disclosure of Magnetic Calculating Machine," describing calculation with instructions and data held in an alterable store, predates von Neumann's first visit to the project by some eight months; the mercury delay-line storage was his, as von Neumann said in print. —Ed.
John von Neumann, born Budapest 1903, professor at the Institute for Advanced Study from 1933, was by the common consent of the men entitled to an opinion the fastest mathematical mind of the age. He consulted at Los Alamos on the implosion problem, the mathematics of the lenses that fired the Trinity device, which is where the author would have seen him; the computer project he began at the Institute after the war issued its engineering reports to well over a hundred institutions with no patent reserved, on his stated principle that the work belonged in the public domain. The author's account of the conversation is consistent with every memoir of the man, down to the apology for answering too quickly. —Ed.
The conference took place at the Moore School in April 1947, Army Ordnance patent counsel presiding, with Eckert, Mauchly, von Neumann, and Goldstine among those at the table. Counsel ruled that the First Draft, distributed without restriction from June 1945, was a publication under the patent statute, and the statutory year having run, the stored-program design it disclosed was in the public domain and unpatentable by any party. The ENIAC application, covering the earlier machine, was filed that June and issued in 1964 as United States Patent 3,120,606; a federal court found it invalid in October 1973, in part for derivation from the earlier work of John V. Atanasoff, so that the builders ended at law with neither the stored program nor the machine. The conference record lists the principals by name; the man keeping the company's minutes is not a principal, and appears in no list, which the reader will recognize as his preferred seat. —Ed.
The paper is "On Computable Numbers, with an Application to the Entscheidungsproblem" (1936), in which A. M. Turing, then twenty-four, defined the universal machine on which every stored-program computer stands. Turing took his doctorate at Princeton in 1938, declined von Neumann's offer of a post as his assistant, and returned to England; his design for the Automatic Computing Engine went before the National Physical Laboratory's executive committee early in 1946, citing the First Draft on its way past it. The man himself is the business of the next number. —Ed.