Patent System Applied to Computers

by philg

Just finished Who Invented the Computer, a recent book by Alice Rowe Burks.  For those of us who assumed that the modern digital computer, having been primarily funded by the U.S. and British governments, was a public domain idea, there are some eye-opening facts.  Here’s the chronology:



1939-42:  John V. Atanasoff, a physicist at Iowa State College, works with Clifford Berry, a master’s student, to build a digital vacuum tube computer (Atanasoff-Berry Computer or “ABC”).  The project has only paltry funding and is abandoned due to a problem with the punch card input/output system.  Atanasoff and Berry wander off to tackle projects related to America’s war with Germany, Italy, and Japan.


1941:  John Mauchly, who has been working on weather calculations using analog computers, drives out to Iowa to visit Atanasoff and gets a good look at how digital circuits can be built from analog components.  He reads Atanasoff’s big design document and goes home after five days of continuous meetings with Atanasoff and Berry.


1943:  vacuum tube Colossus code-breaking computer starts operating at Bletchley Park in England, workplace of Alan Turing, the father of Computer Science


1945:  Eckert and Mauchly’s ENIAC is operational, using many of Atanasoff’s ideas without credit


1947:  Eckert and Mauchly apply for a patent on fundamental ideas in electronic computing (“ENIAC patent”)


1964: ENIAC patent is issued by the U.S. Patent Office, running through 1981.


1965: Sperry Rand, the owner of the Ecker-Mauchly patents, demands approximately $1 billion in royalties from other computer vendors.  IBM is exempt from these demands due to an earlier cross-licensing agreement with Sperry.


1971: Honeywell sues Sperry, asking the Federal court to invalidate Sperry’s patents on the grounds of Atanasoff’s prior art and the more than one year of delay between the customer deliveries of ENIAC and its patent application.


1973:  Judge Earl R. Larson rules that the ENIAC was derived from Atanasoff’s innovations, that Sperry’s patents are unenforceable, and that IBM has been violating the Sherman Antitrust Act (i.e., that IBM is a monopoly).  Honeywell’s legal costs are estimated to be $3.5 million or roughly 700 times the cost of building the ABC computer.


The story has a number of interesting lessons.  First it shows the terrible consequences of being a bit too early.  Nobody was interested in Atanasoff’s project in the late 1930s and it did not seem worth funding to the point that minor obstacles such as the punchcard reader problem could be overcome.  If Atanasoff had only been a few years later he might have played a major role in massive federally funded projects.


A second lesson from the story is that there isn’t all that much true innovation in the engineering world.  People working independently at disparate sites often came up with similar solutions (except for Mauchly, of course, who had the opportunity to stand directly on the shoulders of Atanasoff).  Only one team, however, can get a patent and it turned out that this one was mired at the Patent Office for about 20 years following the design of the ENIAC.


Imagine how different the world would be if Sperry had been able to control, through licensing, whether or not the first microprocessors could be built (note: 1968-70 the Air Force funded a fairly powerful 20-bit microprocessor project for the F14A fighter jet, which worked but was kept secret; Intel introduced the 4004 4-bit microprocessor in 1971 for use in desk calculators).


The last half of the book is devoted to an academic bitchfest in which Burks talks about all of the hacks who don’t credit Atanasoff.  It is interesting mostly for its discussion of exhibits at the Smithsonian and PBS documentaries.  Sperry funded these projects and the non-profits obligingly ignored or downplayed Atanasoff’s contributions in favor of Eckert and Mauchly.  The only publication that couldn’t be bought and wouldn’t be intimidated by threats of legal action was … Car and Driver magazine.  Atanasoff was a hero to Car and Driver writer Patrick Bedard because he testified that a late night high-speed drive in 1937 in a V8 Ford and a few drinks in a roadhouse in Illinois had inspired a couple of the critical designs in the ABC.  Bedard speculated “Atanasoff didn’t get nearly the credit due him because the [court] decision was issued just one day before the Watergrate-inspired ‘Saturday Night Massacre’ and it lacked the combination of inconsequentiality and putrescence necessary to compete for media attention.”

9 thoughts on “Patent System Applied to Computers”

  1. You write: “1968-70 the Air Force funded a fairly powerful 20-bit microprocessor project…which worked but was kept secret” Can you supply some more references for this? It sounds like an interesting story.

    Ironically, all of Intel’s successful microprocessors are direct descendants of the 4004. The attempts Intel has made to deviate from that linage (the i432, 860, Itanium) have flopped.

  5. The book “ENIAC” by Scott McCartney examines the same period of history. McCartney acknowledges that Mauchly was influenced by Atanasoff, but raises interesting questions and finds different conclusions.

    McCartney finds documented evidence that Mauchly tried to work with Atanasoff and encouraged him to apply to the Moore School in 1941. Atanasoff did not take action on this invitation. Atanasoff also tried to get his computer working for the Navy in competition to ENIAC. The Navy rejected Atanasoff’s work because it didn’t work. Atanasoff also rejected ENIAC and went in a different but unsuccesful direction in further development. The net of this is that Atanasoff had some good concepts and ideas, but Eckert and Mauchly were the first people to get a digital computer truly working for real applications and customers. Moreover, it isn’t clear at all that Mauchly did not independently develop the same ideas of Atanasoff. Moreover, Atanasoff’s work is questionable from a prior art perspective because it didn’t work.

    Atanasoff made his own attempts in 1940 to obtain a patent on his computer and to work with RAND and IBM. Atanasoff also attempted to get a working programming system for the ABC. All of these attempts were unsuccessful. Atanasoff’s patent application was rejected because his computer did not work. Mauchly kept up communication with Atanasoff and gave him progress reports on ENIAC. Mauchly asked Atanasoff what had happened to his computer, but Atanasoff said he abandoned it. Atanasoff has further gone on historical record admitting he abandoned his work and never saw it through to completion.

    Atanasoff had no programming system. He tried for a patent and was told his work was too incomplete for a patent. Eckert and Mauchly achieved a complete workable computing system. Atanasoff did not, and he declined to work with Mauchly to complete his ideas.

    Your point about timing should be turned around to a point about perseverance. Atanasoff should not have given up so easily and should have accepted efforts to work with Mauchly.

    Your point about innovation is wrong. ENIAC was a great innovation by virtually any definition. It was the world’s first digital computer. It was Eckert and Mauchly’s invention. All inventions and research are influenced by others. That does not mean that it is not the province of a particular group of individuals to complete the theory or invention and make it work. Innovation requires that the idea work, not to just have an idea, fail at it, and then try to take credit for it. Just as Edison built on scientific ideas to invent the lightbulb, so too did Eckert and Mauchly.

    Interesting to the point of Atanasoff not getting credit, McCartney makes a case that Mauchly and Eckert are often ignored and not given proper credit for computers; IBM, Bill Gates, and Steve Jobs apparently seem to get all the credit :-).

  6. Uh, Bryan, there remains the awkward question of how these two great innovators, without whom we could never have had electronic computers if we are to believe your comment, managed to lose so badly in court. As for how original the innovation was… the British seem to have stumbled upon most of the same ideas two years earlier at Bletchley Park.

  7. I was watching a Discovery Channel documentary on the how great the NSA is. They mentioned the great computers they had to do codebreaking for the British in the second world war. This computer was named Colossus. They failed to mention how they got this computer in the first place. It all sounded as if the British had no idea what they were doing and needed the Americans to help them out. Sigh…

    Interestingly also is the failure of mentioning in any documentary on the allied code breaking efforts that the allied’s own codes were so terrible that intercepting transmission was a walk in the park for the Germans…

  8. Colussus was not a complete general-purpose computer. It was not possible to reprogram it for general computation, as it was built for breaking codes. ENIAC achieved a programming system because you could program it with repluggable cables. Certainly there are ideas in common between the systems we are discussing, but Colussus was not a complete digital computer.

    Eckert and Mauchly created a general-purpose digital computer with the basic essentials, including a programming system. Atanasoff, by his own admission, failed to do this. Atanasoff’s history of rejection and that he admits his computer did not work ends the discussion of whether his computer was first. The prior art claim is suspiscious on these grounds. As for why the patent was overturned, it is hard to say, we can only speculate. The decision was never appealed, despite Eckert and Mauchly’s wish to do so. It was not their decision. Whether it was a social good that the patent was overturned, given the industrial monopoly at the heart of the issue is a separate question, upon which I have no comment.

    Were there other ideas and work that were relevant to what Eckert and Mauchly did and the evolution of comptuer science than ENIAC? Yes. Did someone else create a working general-purpose digital computer with a prorgamming system before ENIAC? No. That was Eckert and Mauchly’s invention and innovation.

Comments are closed.