The Smithsonian Institution’s computer historian, Henry Tropp, has written:
We had the technical capability to build relay, electromechanical, and even electronic calculating devices long before they came into being. I think one can conjecture when looking through Babbage’s papers, or even at the Jacquard loom, that we had the technical ability to do calculations with some motive power like steam. The realization of this capability was not dependent on technology as much as it was on the existing pressures (or lack of them), and an environment in which these needs could be sympathetically brought to some level of realization.46
It took the exceptional circumstances of a second global war, topped by the threat of a nuclear one, to nudge governments and managements in the advanced nations into providing or at least tolerating briefly the kinds of environments where ‘these needs could be sympathetically brought to some level of realization’: highly informal settings where machines like the Colossus were built (by Post Office engineers for the British code-breaking center at Bletchley Park, to break German high-command codes – but scrapped and erased from the official record immediately afterwards47), and the ‘Electronic Numerical Integrator And Computer’ (ENIAC), designed in Philadelphia for calculating gunnery tables, and completed in 1946.
Colossus was the world’s first true, programmable, digital electronic computer, and it owed its existence to suspension of ‘business as usual’ by the threat of military defeat, which also briefly overshadowed the normal regime of homophobia and snobbery. Bletchley Park’s codebreaking genius, the mathematician Alan Turing, was tolerated while hostilities lasted despite his homosexuality and awkwardness. Colossus was built by a team of five working-class General Post Office (GPO) engineers who would never have been allowed near such an important project in normal times (they very nearly weren’t anyway, and certainly weren’t as soon as the War ended).
The GPO team was led by TH (Tommy) Flowers, a bricklayer’s son from the east end of London. While in his teens, Flowers had earned an electrical engineering degree by night while serving a tough engineering apprenticeship during the day. By 1942 he was one of the few people in the world with a practical and theoretical knowledge of electronics, and an imaginative grasp of its possibilities. Management referred to Flowers as ‘the clever cockney’, and tried to get him off the project, but he and Turing got on well from the first, and Turing made sure the project went ahead despite the opposition and sneers. One of Flowers’ team, SW Broadhurst (a radar expert who had originally joined the GPO as a laborer) took it all without complaint. Computer scientist Brian Randell recorded this impression from him in 1975:
The basic picture – a few mathematicians of high repute in their own field accidentally encounter a group of telephone engineers, of all people… and they found the one really enthusiastic expert in the form of Flowers, who had a good team with him, and made these jobs possible, with I think a lot of mutual respect on both sides. And the Post Office was able to supply the men, the material and the maintenance, without any trouble, which is a great tribute to the men and the organization.48
Turing’s biographer and fellow-mathematician, Andrew Hodges, records that Colossus was built with extraordinary speed and worked almost perfectly, first time: ‘an astonishing fact for those trained in the conventional wisdom. But in 1943 it was possible both to think and do the impossible before breakfast.’49 The GPO team worked so fast that much of the first Colossus ended up being paid for, not by the government, but by Flowers himself, out of his own salary.50
But as soon as the War was over these men were sent back to their regular work and could make no further contribution to computing, or even talk about what they had done, until the secret finally emerged in the 1970s. As Flowers later told Randell: ‘It was a great time in my life – it spoilt me for when I came back to mundane things with ordinary people.’
…BUT TOOK COMPUTING DOWN THE WRONG PATH
The ENIAC was nearly a dead end for almost the opposite reason. Its designers, Presper Eckert and John Mauchly, were so intent on being successful capitalists that they nearly buried the project themselves. Unlike Colossus, ENIAC gained public recognition but, as the Dutch computer scientist and historian Maarten van Emden has argued, the rapid commercialization that its makers had in mind could have turned it into a ‘revolution that didn’t happen’51 had it not been for the fortuitous and somewhat unwelcome involvement of the Hungarian mathematician John von Neumann who (by further fortuitous connections, including discussions years earlier with Alan Turing in Cambridge) was able to relate what he saw to other, apparently quite unrelated and abstruse areas of mathematics and logic.
Eckert and Mauchly seem not to have understood von Neumann’s idea which, nonetheless, von Neumann was able to publish, to their annoyance, free of patent restrictions, in the widely circulated report on ENIAC’s successor, the EDVAC. This report effectively kept computer development alive, and out of the hands of normal capitalist enterprise, which (as Van Emden argues) would then have smothered it. He writes that:
Without von Neumann’s intervention, Eckert and Mauchly could have continued in their intuitive ad-hoc fashion to quickly make EDVAC a success. They would also have entangled the first stored-program computer in a thicket of patents, one for each ad hoc solution. Computing would have taken off slowly while competitors chipped away at the initial monopoly of the Eckert-Mauchly computer company. We would not have experienced the explosive development made possible by the early emergence of a design that, because of its simplicity and abstractness, thrived under upheaval after upheaval in electronics.
Von Neumann’s idea – the ‘von Neumann architecture’ – specified a central unit for doing arithmetic; a memory store shared by the program instructions, the data to be worked on and intermediate results; and a control unit to initiate each step of the program, copying data and instructions alternately from and back to memory in a ‘fetch-execute cycle’, as well as receiving input from a keyboard (or other input) and passing it back to a printer (or other output). The system is robust and comprehensible because it does just one thing at a time, in step with a timing pulse or ‘clock’. This turned out to have surprisingly expensive consequences, as computers began to be applied to tasks never envisaged in the 1940s: taking photographs and movies, playing music, and so on… as we shall see later on, in Chapter 11.
Von Neumann’s design is still the basis of nearly all modern computers, and the whole computer revolution might not have happened, had it not been for his freakishly broad interests, his unwanted intervention in Eckert and Mauchly’s business, and then his airy disregard of commercial propriety in circulating his specification. This proved a lucky break for capitalism, despite itself.
More and more powerful machines became possible thanks to von Neumann’s disruptive presence, but capitalist firms still resolutely had nothing to do with their development unless all of the costs were underwritten by governments. As for using computers themselves, they had to be coaxed endlessly, like recalcitrant children, before they would even try what was good for them. Computer development remained utterly dependent on government support for decades.
In his 1987 study for the Brookings Institute, Kenneth Flamm estimated that in 1950 more than 75 per cent of US computer development funding had come from the government, and any commercial investments were largely made in anticipation of lucrative defense contracts. One such contract financed development of IBM’s 701 machine – originally known as ‘The Defense Calculator’. IBM’s commitment to computers was built on guaranteed returns from military projects. A decade