The politics of invention
By Michael Kanellos
Staff writer, CNET News.com
February 13, 2006, 4:00 AM PST
From a technological perspective, the Electronic Numerical Integrator and Computer was an unqualified success. But the story behind ENIAC--its development and demise--is a classic illustration of how scientific brilliance all too often falls victim to squabbles of ego and finance.
First, ENIAC's inventors, J. Presper Eckert and John Mauchly, had a falling out with a prominent mathematician over his failure to credit them for their contributions. John Von Neumann of the Institute of Advanced Studies in Princeton, N.J., had discussed ENIAC and computers in 1944 with Eckert and Mauchly and then published "The First Draft of a Report on Edvac" in 1945.
The document laid out the basic elments, including stored programming, of modern computers. But Von Neumann did not identify Mauchly and Eckert, who were not allowed to publish anything about their work because of security, as contributors. The omission rankled them for years.
"He (Von Neumann) gets a lot of credit, but the inventions were ours," Eckert said in a 1989 interview.
The two inventors then argued with the University of Pennsylvania. Eckert and Mauchly said they had a tacit agreement with Penn that they would own the patents, but the new head of the Moore School, Irvin Travers, had different ideas. The two were forced to leave the university, an again, animosity festered for decades.
Only a few years ago, University of Calgary historian Michael Williams was interviewing Eckert and told him he was off to visit Travers, who lived a few blocks away. "Eckert said, 'I haven't spoken to that bastard in 40 years.' They wouldn't walk across the street to spit on each other," Williams recalled. "Everybody hated everyone else, and it got worse over time. Don't forget, there was a lot of money involved."
The inventors formed the Eckert-Mauchly Computer Corp. after leaving Penn but did not find vindication there either. The company sold their own computers, UNIVACs, to the Atomic Energy Commission and the Census Bureau, but it struggled financially, often breaking budgets and deadlines.
A white knight came in the form of entrepreneur Harry Straus, who owned American Totaliser, maker of machines for race tracks. Straus lent money to the professors and stayed out of management. But Straus was killed in a plane crash on Oct. 25, 1949, and Totaliser sold the company to Remington Rand, which marginalized computers. The company later became Sperry Rand.
"They were brilliant engineers, and brilliant engineers aren't necessarily great businessmen," said Margaret O'Mara, assistant professor of history at Stanford University, adding that the state of Pennsylvania lost out too. "If we are all sitting around here in 1946 and I asked you to predict what would be the electronic capital of America, a lot of people would have said Philadelphia. It was the place to be."
The years of conflict reached a crescendo with a lawsuit in 1967. That year, Honeywell alleged that Eckert-Mauchly's patents were invalid because the ABC computer had pre-dated the ENIAC. If that were true, Honeywell wouldn't have to continue paying royalties to Sperry Rand. They wouldn't have to pay anyone, in fact, because ABC inventor John Atanasoff had never filed patents of his own. Mauchly and Sperry countered that the ABC machine--long ago dismantled, virtually forgotten and never used on any significant projects--was based on an entirely different architecture.
A judge invalidated the Eckert-Mauchly patents, but Penn history professor Nathan Ensmenger said historians generally agree that Mauchly didn't take ideas from the ABC computer. "The real triumph of the ENIAC was in engineering. The concept of a large-scale electronic computer is fairly simple, but the engineering is difficult," Ensmenger said.
Rumors, controversies and disputes continue to this day.
Mauchly asserted that ENIAC engineer Arthur Burks, now a Honeywell consultant, was blackmailing the professors by offering not to testify against them in the case if they put Burks' name on their patents.
Hear ENIAC programmer Jean Bartik discuss her personal experiences on the project.
Listen now... (2.6MB mp3)
"I knew Arthur Burks as a nice, mild-mannered man in the style of the Mr. Milquetoast depicted in the New Yorker cartoons. Apparently he married a woman with magical powers who turned him into this vicious, all-knowing author," wrote Jean Bartik in a reader review of a book by Alice and Arthur Burks on Atanasoff. Bartik, a Mauchly defender, was one of the many female mathematicians working on the project.
Not surprisingly, Burks strongly contests the accusations. "We do not hold that Mauchly intended to stead Atanasoff's inventions when he first saw and learned about them, but, over time, having used the concepts in a more complex device, he came to claim it all for himself," Burks wrote on behalf of himself and his wife.
Eckert survived the turmoil and went on to become a vice president at Sperry Rand, which later became.
Mauchly did not fare as well. Labeled briefly as a communist sympathizer in the McCarthy era, he left Sperry in the mid-1950 to become a consultant.
Mauchly, broke, died in 1980. Eckert died in 1995, six years after retirement.
Their places in technology history, however, remain strong.
"More than anything else, they built a machine that didn't fail," said professor Alex Randall of the University of the Virgin Islands, a longtime Eckert acquaintance. "If it had failed, people would have said, mechanical computers are the way to go."