Johns Hopkins Magazine -- November 1999
Johns Hopkins 
     Magazine Home

NOVEMBER 1999
CONTENTS

AUTHOR'S NOTEBOOK

RELATED SITES

Mention the name of alumnus John Mauchly and you're apt to be met with a blank stare. How is it that one of the 20th century's great inventors is now largely forgotten?

P U B L I C    P O L I C Y    A N D    I N T E R N A T I O N A L
A F F A I R S

The Story That
Doesn't Compute

By Dale Keiger
Illustration by Ferruccio Sardella
Photos courtesy John W. Mauchly Papers, Rare Book and Manuscript Library, Univ. of Penn.

Scott McCartney likes to tell a story, by way of Washington Post writer Bob Levey: A mother in Alexandria, Virginia, was quizzing her 9-year-old son on inventors, and the kid was rolling. He'd named Thomas Edison as deviser of the light bulb, and Eli Whitney as responsible for the cotton gin. Then Mom asked, "Who invented the computer?" The boy thought for a moment, then responded: "Radio Shack."

McCartney is a reporter for The Wall Street Journal. A few years ago when he was covering high technology, he realized that no one he knew could give a better answer than that 9-year-old in Virginia. The computer is ubiquitous as a tool, as a driver of economies, as a metaphor. It has been around for little more than 50 years. Yet, says McCartney, "I was shocked that here I was, covering the industry, and I couldn't tell you who had invented the computer." So he decided to dig into the technology's history. What he found was the bittersweet story of John Mauchly, Hopkins PhD 1932.

Mauchly and research partner Presper Eckert deserve most of the credit for inventing the modern electronic computer, credit they almost never get. In McCartney's recent book, ENIAC: The Triumphs and Tragedies of the World's First Computer (Walker, 1999), he has tried to right what he considers an injustice and give Mauchly and Eckert their due. He says, "You know, to me the story of the first computer doesn't compute. How come the two guys who built the first one and founded the first computer company and had the patent...how could they already be forgotten? It doesn't make sense."

The Second World War provided the impetus for research into a new way to do rapid calculations. In 1942, the U.S. Army was desperate for a faster means of creating firing tables. To aim a cannon, a gunner had to adjust the gun to several settings. That required tables to account for all the variables that affect the trajectory of an artillery shell: wind speed and direction, humidity, temperature, elevation, even the temperature of the gunpowder. A 155-millimeter "Long Tom" cannon required a table that covered 500 different sets of conditions. It took a team of women punching numbers into push-button mechanical calculators more than a month to produce just one complete table. These women, known at the time as "computers," were so far behind in their work that guns were arriving at the western theater of war that could not be aimed for lack of firing tables. The Army asked the University of Pennsylvania to find a way to calculate tables faster. Penn's Moore School of Electrical Engineering had a special defense-related program in electronics, and on its premises were Pres Eckert and John Mauchly.

As a boy, Mauchly was on call with his neighbors in Chevy Chase, Maryland, whenever they had trouble with the wiring in their houses. He was more dependable than the electric company. When work crews dug trenches for new water lines in the community, Mauchly laid wires in the ditch to create an intercom system for his friends. When he was 11, he liked to read in bed late into the night. His parents wanted him to get some sleep, and from a certain spot on the stairs they could see if light was coming from under his door. So young Mauchly secretly wired a sensor into one of the lower steps. Whenever his mother or father put a foot there, a light would blink off in Mauchly's room, giving him time to douse his reading lamp. He was a straight-A student who thrived in Chevy Chase, which at that time was home to many scientists and engineers.

In 1926, Hopkins awarded him an undergraduate engineering scholarship. He seems not to have attracted a lot of attention as student. Recalls William Banks '29, "He was almost a hermit because he was trying to do four years of work in three. He was buried in the catacombs most of the time, in the engineering building working on different stuff. He certainly didn't mingle socially with the rest of the class. He didn't have time." Other of Mauchly's engineering classmates don't remember him at all. He shows up in one photo of the 1929 Hullabaloo yearbook, which mentions that he was on the staff of the student humor magazine The Black & Blue Jay. Says Banks, "It seems a paradox that a guy who was so serious about everything that he never had time to associate with anybody was interested in writing jokes."

Mauchly quickly became bored by engineering. "He became very disillusioned," says Kay Mauchly Antonelli, who became his second wife in 1946. "He always said it was 'just cookbook stuff.'" Before his junior year he switched majors, going straight into the Hopkins graduate physics program. In a videotaped interview from 1978, Mauchly recalled, "Physicists -- those were the boys who were going to have fun." He studied molecular spectroscopy, calculating the molecular energy of gases by endlessly pulling a large handle on a mechanical calculator. Though he never completed work on a bachelor's degree, he had a PhD by 1932.


Mauchly envisioned a machine with no moving gears or wheels or levers, just moving electrons.
He moved on to teach at Ursinus College in Pennsylvania, where he became a one-man physics department, before heading to Penn. He'd been thinking all the while about how to calculate faster. His hobby was weather forecasting, and he was frustrated by how long it took to calculate meteorological data by mechanical means. He thought there might be a way to apply electricity to vacuum tubes to create a faster calculator. When he began working with Eckert at Penn, they theorized about an electronic calculator. Mauchly envisioned a machine with no moving gears or wheels or levers, just moving electrons. If it could be made reliable, it would be extremely fast.

In August 1942, Mauchly summarized his ideas in a seven-page proposal titled "The Use of High-Speed Vacuum Tube Devices for Calculation." At that time, everyone who was anyone--at Harvard, MIT, Bell Labs--was working on mechanical, analog computational devices. Mauchly's superiors at Penn ignored his memo.

Herman Goldstine, a PhD mathematician turned Army lieutenant, did not. Goldstine had been put in charge of the Army's project to produce gunnery tables. One day, a grad student at Penn asked him if he had heard of Mauchly's idea for an electronic calculator. He hadn't but wanted to. He asked to see Mauchly's memo, which had to be reconstructed from a secretary's shorthand notes. Goldstine grasped Mauchly's central idea. Making a gear wheel stop or start in one-millionth of a second was tough, but it was easy to start or stop an electron in that span of time. Goldstine took Eckert and Mauchly to the Aberdeen Proving Grounds in Maryland to pitch the Army brass on their idea in April 1943. Among those listening were Colonel Leslie Simon, director of the Army's Ballistics Research Laboratory, and Oswald Veblen, a renowned mathematician. Partway through the presentation, Veblen interrupted and said, "Simon, give Goldstine the money." Thus was born, with $61,700 in funding, the project to create the Electronic Numerical Integrator and Computer--ENIAC.


ENIAC's electronic brain contained nearly 18,000 vacuum tubes.
A dozen researchers began work in July 1943. Mauchly was the senior at age 35. Later, Eckert would say, "Had John and I been five years older and that much more experienced, we might have 'known' a true electronic computer would not be built." But they didn't know, so they set to work. They decided their machine would have three parts: one to handle the math, one to be the memory, one to be a controller. As much as possible, they avoided anything that would slow the machine. Mauchly envisioned making the desktop mechanical calculator electronic, then hooking up many of them so they could work simultaneously on parts of difficult calculations. Such "parallel processing" is the basis today of modern supercomputers.

Mauchly wanted to use vacuum tubes but had to work around their unreliability. ENIAC was supposed to have 5,000 of them, operating at 100,000 pulses per second. That meant that each second there would be 500 million chances for a faulty tube to screw up. Televisions of the day frequently needed repair, and they contained but 30 tubes. As the Army kept increasing its demands, plans for ENIAC went from 5,000 tubes to 18,000. That meant 1.8 billion chances to fail...each second.

Eckert solved the problem by finding the best available tubes, made for the telephone company for transatlantic cable, then running them at a much lower voltage than they'd been designed for. Worried about rodents chewing on the insulation of ENIAC's wiring, he got hold of some lab mice, starved them for a few days, then put scraps of different insulation in their cages. The type they liked least became what he specified for ENIAC's wiring.

The ENIAC team worked at Penn seven days a week. Whenever they became frustrated, the engineers would refer to ENIAC as MANIAC. They relieved tension with pranks. Eckert fell asleep one night on a cot beside the growing machine. Two technicians picked up the cot, carried him to the second floor in an elevator, and placed him in an identical but empty room. When he awoke, he thought ENIAC had been stolen.


The computer weighed 30 tons, but was 1,000 times faster than mechanical calculators.
Mauchly was insecure about his position on the ENIAC team, and worried about whether his work was appreciated. But he was the calm foil to Eckert's exacting, temperamental personality. Antonelli, before her marriage to Mauchly, was one of the original "computers" who worked on firing tables. She had been selected to program ENIAC, and recalls, "Pres was volatile. He could explode all over the place, and Mauchly could walk into the middle of the situation, get everybody's point of view, and straighten it out." Mauchly was a talkative chain smoker who liked nothing better than discussing ideas. "He was utterly charming, in a way," Antonelli says. "He had read so much and knew so much stuff. He was never dead serious or depressed." Around the ENIAC project, he was famous for being easily distracted. Said Goldstine: "He was a nice guy, very smart, but probably a space case." Antonelli agrees: "He was a dreamer. Without Eckert he would never have built a computer. But we often said that without Mauchly, Eckert wouldn't have thought of it."

ENIAC was ready to begin work just as the war ended. Antonelli remembers seeing it for the first time, late in 1945: "The overall length of the machine was 80 feet. It was eight feet tall and black with little twinkling lights." It weighed 30 tons, occupied 1,800 square feet of floor, and was 1,000 times faster than any mechanical device. What took 20 hours on a desk calculator--computing a trajectory--now could be done in 30 seconds. The computer contained 17,486 vacuum tubes, 500,000 soldered joints, 70,000 resistors, and 10,000 capacitors. Such circuitry now fits on a lapel pin. The Apple iMac on which this story was written packs more computing power. But ENIAC proved that electricity could think.

In 1946, the only ENIAC that would ever be built was disassembled, pulled out of a hole knocked in the building at the Moore School, and taken to Aberdeen, where it worked on hydrogen bomb problems, predicted Russian weather patterns to map nuclear fallout (the Russians, says McCartney, actually inquired about buying an ENIAC and were turned down), aided wind tunnel design, and created ballistic tables. The machine was not retired until 1955. Part of it is now on display at the Smithsonian in Washington. Says McCartney, "We now throw out computers after two or three years. That one lasted eight."


An engineer checks a fundamental electric signal on an ENIAC oscilloscope.
The public has grown accustomed to high-tech innovators piling up vast fortunes. Bill Gates, Steve Jobs, Mark Andreesen, Paul Allen...everyone can name a few computer billionaires. John Mauchly made their fortunes possible. But he realized little financial gain himself, and not much credit. Before ENIAC went out the hole in the Moore School wall, Eckert, Mauchly, and renowned mathematician John von Neumann of Princeton began designing a second computer, called EDVAC. The Army again funded the effort, and in June 1945, Goldstine issued a mimeographed document titled First Draft of a Report on the EDVAC, by John von Neumann. In the report, von Neumann skillfully explained the development of the new technology but did not explicitly give credit to Eckert or Mauchly for certain crucial ideas. A team effort suddenly appeared to be von Neumann's show.

At the time, Eckert and Mauchly regarded First Draft as simply that, a draft for internal use that crystallized current EDVAC thinking. But Goldstine sent 24 copies to friends and colleagues of the mathematician. Soon several hundred were in circulation. Universities began building machines using the von Neumann draft as a blueprint, and press articles credited him-- not Eckert and Mauchly--with concepts that shaped the machines. The basic design, still much in use, became known as the "von Neumann architecture." When Time magazine published a special issue last March, "100 Scientists and Thinkers of the 20th Century," it said, "Virtually all computers today, from $10 million supercomputers to the tiny chips that power cell phones and Furbies, have one thing in common: they are all 'von Neumann machines,' variations on the basic computer architecture that John von Neumann, building on the work of Alan Turing, laid out in the 1940s." Nowhere is there mention of Eckert or Mauchly.

In 1946, Eckert and Mauchly still hadn't filed for a patent on their ENIAC inventions, but they had made a standard deal with Penn before the project started: They could file for patents, and Penn and other schools would be granted a license to build and use computers for noncommercial purposes. But Penn later changed its mind and argued that it should be the entity filing for patent rights because the researchers had done their work under the university's auspices. When Irven Travis, Penn's new supervisor of research, insisted that Mauchly and Eckert sign an agreement giving Penn the patent on any future machines, they left to found the Eckert-Mauchly Computing Corporation, the world's first computer company.

In the midst of setting up the company, Mauchly and his first wife, Mary, decided they needed a vacation. They drove to Wildwood Crest on the New Jersey shore. One night, they impulsively dashed into the surf naked. Mary was knocked down by the water and swept out. She drowned before John could save her. They had been married 16 years. Says McCartney, who interviewed many of Mauchly's associates and family members, "It was emotionally devastating, but he didn't have a lot of time to feel sorry for himself. He was a single dad trying to start the first computer company. He was doing everything he could to put food on the table and take care of his two kids. He was not a brooding guy. I didn't get the sense that there was a long period of hibernation or depression. But at the same time, he and Mary had been very close and it was very tough on him."


Mauchly (above) and Eckert put off filing a patent on ENIAC, an oversight that would come back to haunt them years later.
Distracted by events and careless about business matters, Eckert and Mauchly didn't complete their ENIAC patent filing until June 26, 1947, three years after they had started working on the application. "You know," says McCartney, "they were not the best at documenting their work. They did not publish a lot. It wasn't part of their character. They didn't take the time to sit down and write the papers. The patent fell into that. The application was a huge, complicated document, and they procrastinated." In their application they claimed more than 100 inventions. After various delays, they finally received a patent in 1964.

Sperry Rand, a player in the emerging computer business and at the time Mauchly's employer, had obtained the right to the patent from the two men, and entered into a secret deal with IBM. The two companies would exchange technology at low cost, but demand exorbitant licensing fees from anyone else who wanted to use the ENIAC architecture.

In 1967, a competitor, Honeywell, challenged the Eckert-Mauchly patent in court. Honeywell's investigators came across an old acquaintance of Mauchly's named John V. Atanasoff , an engineer whom Mauchly had visited in 1941 at Iowa State University. Atanasoff had tried to create an electronic computer of his own, without much success. He moved on to other endeavors, became a millionaire, and never tried to claim credit for any part of ENIAC--until urged by Honeywell's lawyers. Now he claimed that Mauchly had stolen critical concepts from him and used them to create the first computer. If Honeywell could convince a judge of this, the Eckert-Mauchly patent would be invalid.

Mauchly had documentation that supported his claim to have been experimenting with vacuum tubes and digital devices in the 1930s, before he even knew Atanasoff. But according to McCartney, Mauchly was a bad witness, defensive, forgetful, and evasive. He claimed to have spent only 90 minutes with Atanasoff; actually, he'd spent several days. Students testified that Mauchly had nosed around the circuitry of Atanasoff's machine, and Atanasoff had let him read the design documents. What really damned Mauchly in court was a letter he wrote to Atanasoff on September 30, 1941, in which he said: "A number of different ideas have come to me recently about computing circuits--some of which are more or less hybrids, combining your methods with other things, and some of which are nothing like your machine. The question in my mind is this: Is there any objection, from your point of view, to my building some sort of computer which incorporates some of the features of your machine?"

After sifting through the evidence, McCartney says, "I think Atanasoff has claim to have possibly been an influence on Mauchly. As any inventor will tell you, there were lots of influences. Mauchly had looked at the work George Stibitz was doing at Bell Labs, and had looked at the Monroe desk calculator, and physics experiments at Swarthmore, and been influenced by experiments at Hopkins. But Mauchly said he never learned anything from Atanasoff because his stuff wasn't all that hot, and I think that's probably right. The Iowa machine never worked. It was prone to errors and not fully electronic. There's nobody I know of, outside Iowa, who says that's the first computer. What really sank Mauchly was his tendency to be gracious. After the visit to Iowa he wrote Atanasoff a cordial letter thanking him for his time and telling him that he'd learned so much. That came back to haunt him. I think Mauchly was just trying to be nice."

In 1972, seven months after the last courtroom testimony, Judge Earl Larson ruled against Eckert, Mauchly, and Sperry on a technicality: Because application for the patent had not been filed within one year of the first public use of ENIAC, the patent was invalid. Larson stunned the inventors by concluding that the first problem run by the Army on ENIAC in 1946 constituted public use, thus starting the patent clock. "Eckert and Mauchly had no control over that," says Antonelli. "It was not public use." McCartney is convinced that Larson did not really aim his ruling at Mauchly, but at Sperry, nailing the company for restraint of trade. The judge was breaking a potential monopoly. His decision broke Mauchly's heart. "He took it very badly," recalls Antonelli. "Very badly. He was completely devastated by the whole thing. The idea that he had stolen the idea from this crackpot...it was unbelievable to him." Eight years later, in a videotaped interview, Mauchly fought tears as he discussed Atanasoff.

Mauchly struggled in his last two decades. He had left Sperry in 1960 and formed a company named Mauchly Associates; it failed. In 1966, he founded Dynatrend Inc. It failed as well and Mauchly went broke, nearly losing his house. Says Antonelli, "He was a terrible, terrible businessman. He really didn't have any sense whatsoever about money. He didn't understand about contract law and things like that." He finally went back to Sperry in 1973 as a consultant.


The judge was breaking a potential monopology. His decision broke Mauchly's heart.
Two years later, he was among the first to order from Radio Shack a TRS-80 home computer. He used two of them to pursue his old hobby of weather prediction. "As early as 1960," Antonelli recalls, "he was saying that computers were getting smaller and faster all the time and there was no reason why we wouldn't get to the point that each person could have a little desk computer in his own home." Mauchly also envisioned e-mail , encryption for computer financial transactions, and portable computers, ideas far in advance of the available technology.

John Mauchly died on January 8, 1980, at age 72. When Pres Eckert heard about his friend's passing, he cried all night.

Antonelli concedes her bias, but says, "Everyone who ever worked with him was impressed. He would infect people. He would throw out a very small idea, see what the person could do with it, then just keep pushing at it. He made them think for themselves. He did this with everybody and everything. I just came across the eulogy that Pres Eckert gave when John died, and he said, 'He was without a doubt the most brilliant man I ever met.'"

Before he died, Mauchly did achieve a redemption of sorts as the personal computer came to the fore. "There sprang up everywhere these little personal computer societies," Antonelli recalls. "Every city had one. John became the darling of them. They would invite him to give talks and treated him with the utmost respect."

Says McCartney, "There's no doubt that these guys were brilliant and did some amazing things. At the same time, they failed at a lot and fell victim to some unbelievable circumstances. I work for The Wall Street Journal and we see financial markets throwing enormous sums of money at anybody with a computer idea. Eckert and Mauchly had to beg and borrow and scrape and scratch, and got doors slammed in their faces. Their story, ultimately, is tragic."


RETURN TO NOVEMBER 1999 TABLE OF CONTENTS.