Why the Nation That Invented the Computer Lost Its Lead

Last month I wrote about a piece that Lynn Conway wrote for IEEE Computer Magazine. (See my post The Conway Disappearance Effect.) I came across another article in the same issue, When Winning Is Losing: Why the Nation that Invented the Computer Lost Its Lead, by Marie Hicks, now at the Illinois Institute of Technology. It is based largely on her book, Programmed Inequality: How Britain Discarded Women Technologists and Lost Its Edge in Computing. When she wrote that in 2017, she was an Assistant Professor of the History of Technology at the University of Wisconsin—Madison.

I'll give you the two-line summary of the book, before going into more detail. During the Second World War, Britain invented the electronic computer at Bletchley Park, and there were a lot of women working there. After the war, most of those women were pushed out of the industry, and as a result, the US took the lead in computing.

Bletchley Park

If you want some details on Bletchley Park, then you can see my post British Computer Museums. I highly recommend a visit to the museum there, should you ever find yourself in London with half a day to spare. Bletchley Park is literally a couple of hundred yards from Bletchley station on the East Coast main line to York and Edinburgh.

Bletchley Park was the top-secret operation to decode German Enigma coding machine messages. A lot of what we would now call computer scientists worked there, most famously Alan Turing. They weren't called that back then because there were no computers to science, and this was still the era when a computer was a person with a mechanical adding machine doing calculations. Eventually, Bletchley Park got moved to Cheltenham and became GCHQ, roughly the UK equivalent of the NSA.

At first, the team at Bletchley Park created some mechanical machines called "bombes" that were used. The thing that made this feasible was that there were only a dozen choices of rotor for the Enigma machine (that why there are 12 columns in the picture to the right), and they knew how they were wired since the Polish had captured a set. Oversimplifying, the bombes allowed all the rotors to be searched at once. The Enigma machine had a fatal weakness, which was that a letter was never encrypted to itself. If you knew, or guessed, that somewhere in the message was the word "Berlin" or "cloudy", then that could be used to eliminate a lot of possible rotor settings. 

The second phase was to build what is regarded as the world's first programmable electronic digital computer, Colossus. This was obviously a lot faster than the mechanical bombes. It was actually designed by Tommy Flowers of the General Post Office (GPO, established 1660!), since that was still the era where post and phones were all in single organizations known as PTTs (post, telegraph, and telecommunications). The US is actually exceptional in that the telegraph did not end up as part of the Postal Service. But the Colossi were actually built and maintained by the Women's Royal Naval Service (WRNS, known colloquially, and even officially, as Wrens—my father was an officer in the Royal Navy, you never know when information you absorb as a kid might come in handy).

One interesting side note about Colossus is that it didn't have enough memory to store the message being decoded, which might have been as big as a few thousand bytes (I'm tempted to put a LOL in). The message was punched onto a paper tape, the tape was formed into a loop, and it was repeatedly run through the machine (at 5,000 characters per second, or nearly 30 mph). There was a second generation, Colossus Mark 2, that had shift-registers, and ran five times as fast.

If you are interested in more details of Bletchley Park, and how the codebreaking was done, then there are many books. There are also several books on Alan Turing (although as always with anything "based on a true story", you can't believe a lot of what happens in Hollywood versions—personally, I prefer to watch either original screenplays or documentaries).

After the War

After the war, Churchill and other high-ups considered the fact that Bletchley Park had broken the Enigma codes to be top secret, and ordered that all the computers be destroyed, along with the plans. All the computer scientists were warned that all the work they had done there was classified secret, and they could never talk about it for 30 years. So it was not until the mid-1970s that it turned out that several senior computer scientists in British universities had worked there, such as Brian Randall, the head of Newcastle University Computer Science Department. At Edinburgh, Donald Michie, head of the Artificial Intelligence Department, was another. (While I was doing my graduate work I remember him telling a story about going with Alan Turing to bury silver bars in a stream bed. Like most people, Alan believed that Britain might lose the war, and so he converted his money into silver. You'd think a couple of super-smart engineers would be able to triangulate a hiding place effectively, but after the war, they couldn't find the spot. Alan Turing basically lost most of his money. Literally lost it.)

If you go to Bletchley Park today, you will see a rebuild of Colossus. This immediately raises the question of how it was rebuilt, given that the originals and the plans were all destroyed. It turns out that there was a lot of information about it in engineers' notebooks—in the US where they were not all destroyed.

The women of Bletchley Park were told to masquerade as secretaries after the war, to maintain secrecy on what they had done there. As it happens, my great-aunt Kathleen worked at Bletchley Park. She must have been about 40. She never married since, as she told me once, "in the space of a year or two, every boy I knew was dead". She was part of the lost generation that came of age in World War I. She wasn't lost herself, obviously, since we didn't put women in the trenches, but her cohort of potential husbands was. I don't actually know what she did there since I only found out that she worked there when my father told me after she had died in her eighties. So I don't know if she would count as one of Marie's "discarded women technologists", or if she was doing something more like a clerk or a typist. If she'd told me when she was still alive, she'd have probably have honored her oath of secrecy and told me she had been a secretary, even if she'd been a technologist. That secret went to the grave.

To my mind, the decision to destroy everything and make everything top secret was an amazingly short-sighted decision. Britain had invented the key technology for the future but was all but forbidden to capitalize on it. Of course, you couldn't take all the ideas out of people's heads—it wasn't as if Britain didn't build any computers until 1975, but it was significantly harder to disseminate the knowledge once it was peace time. In Britain, computer science had a standing start.

The US took the opposite approach. The team that had built ENIAC ran seminars and invited academics and scientists working at places like Bell Labs or in military laboratories. People like Von Neumann, Mauchly, and Eckert became rapidly famous based on their war work. In the US, computer science hit the ground running. Later, in 1962, the newly formed ARPA funded four universities to create the first doctorate-level programs in computer science. They are still the pre-eminent computer science departments in the US, and perhaps the world: UC Berkely, Stanford, MIT, and CMU. You can read that story in my post "Lick" Licklider, Unsung Hero of US Computer Science (Lick was the ARPA program director).

Women in Computer Science

After the war, in the US and UK anyway, there was overt and covert discrimination against women for sure. Women were expected to be mothers, and perhaps do some part-time work in areas like primary education or nursing, where they had already carved out a significant niche. Nobody expected Rosie the Riveter to carry on riveting, and based on Marie's book, nobody expected female programmers to carry on programming.

I'm not sure how old Marie is, thus what year this anecdote early in her article refers to:

As the child of a computer programmer, this should have come as no surprise. But I had thought of my mother’s work on an individual basis, rather than as part of a trend. She got into computer programming after being pushed out of the graduate astronomy program at the same university where I was now a sysadmin, having been told that she was taking up a spot in the program that should have gone to a man. It had never occurred to me that the reason she failed as an astronomer but succeeded as a programmer hinged on computing’s gendered labor history.

To be honest, I don't understand the last part of that extract. It comes across of critical of computing, but in fact, computing took her, and it was astronomy that refused her because of her sex. I guess she is saying that in that era, computing was "women's work".

By the time of the 1960s, the stipulation to hire only men into computing was relaxed. One reason was that computing was in its high-growth phase and there were not enough men to be found to train for the new industry. But also society was changing, and the idea of women not being constrained to motherhood and limited female-okay occupations was changing, too.

I continue to believe that the main reason that the UK lost its lead in computing was twofold. First, it never had a significant lead in the first place. Digital computers were in the air, and there were developments in the US that were not far behind. Despite Colossus being a programmable digital computer, it was designed for code-breaking and was nothing like what we would consider general purpose. But the critical factor was the decision to make everything secret, and push everyone, men and women, out as part of demobilization. There were no computer science departments, or computer companies, waiting to absorb any of them. When the industry really took off, in the 1960s, someone at Bletchley who was 30 when the war ended would be 50, and presumably in another career completely. The people who moved into the industry came from all sorts of other subjects (Lick Licklider, who I mentioned above, was a psychologist). There were no computer science courses. Even by the time I went to Cambridge University in 1972, probably the top computer science department in the UK then and now, you couldn't do computer science as a full degree—it was just a one-year course you could take full-time in the third year (I studied mathematics for my first two years as an undergraduate).

Sign up for Sunday Brunch, the weekly Breakfast Bytes email.