A short article appeared at the end of 2021 called “The End of the Electromechanical Era” by Elliot Williams. The idea of a distinctive electromechanical era of technology has stuck in my head, and I return to it every so often. There are a number of websites that identify a electromechanical era from 1840 to 1940, though I’m not familiar with this from any history book. The hybrid technologies of the electromechanical era imply the existence of “pure” (non-hybrid) technologies before and after, with electromechanical technologies being preceded by purely mechanical technologies and followed by purely electronic technologies. That’s something of an idealized sequence, since any technology has to have some mechanical component by which it interacts with the world, if only it’s a button for a human being to push. But even as an idealized sequence it has some utility for organizing our thoughts.
An obvious example of a purely mechanical technology is the diesel engine, which was purely mechanical in its original form (with the first working diesel engine built in 1897)—really, a wonder of engineering that produced an internal combustion engine with no electrical parts. We can see this as being, in a sense, a culmination of purely mechanical technology. The first internal combustion engines with an ignition magneto (invented by Robert Bosch) were being built just after the first diesel engine, being used in engines on a zeppelin in 1900 and on a Mercedes racecar in 1901. Internal combustion engines had been in slow, steady development for a century, much as steam engines had been in slow steady development for almost a century prior to James Watt. Both technologies, then, gasoline and diesel engines, became available right at the dawn of the twentieth century, and both forms of internal combustion engine dominated the industrial development of the twentieth century much as the steam engine dominated the industrial development of the nineteenth century.
Although the diesel engine began as a purely mechanical technology, as electronic control systems increased in sophistication, diesel engines began to use them, and a contemporary diesel engine is as trussed up by its wiring harness as any internal combustion engine with an ignition coil. Technologies are now encrusted with electronics, which surround and envelop the technologies that are by subsumed by their wiring harnesses, as is a contemporary diesel engine. I wrote about this in newsletter 274, in which I discussed this electronic encrusting of the surfaces of our world by assimilating technologies, and in newsletter 316 I formulated a quinquepartite taxonomy of technologies consisting of mechanical technologies, electromechanical technologies, electrical technologies, electronic technologies, and solid state or integrated technologies (essentially, silicon wafer technology).
There are several interesting side branches of the main technology tree that defined industrialization during the twentieth century. For example, technologies like the Wankel engine—used in Mazdas for the mass market—are newish, but not really revolutionary. It is an internal combustion engine, but mechanically it is quite different from the conventional internal combustion engine. Still, despite its advantages—smaller, lighter, more uniform torque—the Wankel did not come to dominate the transportation market, but remained marginal in comparison to conventional engines. Similarly, there were several gas turbine cars, including the Chrysler Turbine Car (only 55 were built), the General Motors Firebird (four prototypes built), and the British Rover. The gas turbine cars didn’t make it to the mass market. Technologically, there are a lot of problems—for example, they aren’t fuel efficient, and they require an enormous amount of air, which means enormous exhaust pipes—but it is, I would argue, more revolutionary than, say, a Wankel engine. The attempt to build a gas turbine car is a kind of exercise in bragging rights, since it’s a difficult and demanding technology, and in fact most of them were only built as prototype for auto shows as demonstrations of what the manufacturers were capable of doing. None of these developments initiated a new technological revolution, and none of them redirected the development of the industrial revolution. There are degrees of technological innovation and change. Even the gas turbine powered car had a lot of continuity with conventionally powered cars, and the earliest cars had a lot of continuity with late horse-drawn carriages. Carriage makers repurposed themselves to build automobile bodies.
Once electromagnetic technologies began to replace purely mechanical technologies they rapidly became as complex as had the purely mechanical technologies before them. The three-phase squirrel cage induction motor was developed a decade before the first diesel engine; it, too, dominated the industrialization of the twentieth century, alongside the internal combustion engines, and it, too, was the result of many incremental steps of development. There’s a fascinating discussion of the design of General Motor’s Futurama exhibit in James Mauro’s 2010 book Twilight at the World of Tomorrow: Genius, Madness, Murder, and the 1939 World's Fair on the Brink of War, which goes into some detail about the electrical control systems of the attraction:
“Designed by James Dunlop and built in part by Westinghouse, ‘the Polyrhetor,’ as it was called, was a twenty-ton contraption that pulled sound from twenty-one individual strips of movie film revolving around an eight-foot steel drum. Seven photoelectric beams divided the strips into one hundred and forty-seven units of sound, and each was picked up and transmitted to two cars at a time via seven corresponding trolley tracks that ran beneath the cars... one hundred and forty-seven separate bits of narration were delivered simultaneously throughout the ride, on a continuous loop since the cars themselves never stopped revolving around the circuit. Each segment was calculated to the exact second and the precise amount of time two cars took to pass a certain point. Incredibly, despite the fact that the conveyor occasionally slowed down to accommodate large crowds, the system never fell out of synch.” (p. 177)
We can imagine this being built today with off-the-shelf technology, cheaper, and with greater functionality, but the fact that such a control system was possible at all using the electromechanical technologies of the late 30s is amazing in its own way. The most sophisticated technologies are custom-built at first, like those of Futurama, and only later become standardized. The role of standardization in technology—or, more specifically, for engineering a technology for mass production—is described in Simon Winchester’s The Perfectionists: How Precision Engineers Created the Modern World (discussed in newsletter 82 and a PS to newsletter 271).
I wonder if there is a similarly interesting story behind the electrical control systems of elevators in early skyscrapers. I’ve been (mildly) curious how they managed to run the elevators efficiently, say, calling the nearest one when a button is pushed on a given floor (though not curious enough to do any actual research). The Empire State Building has 73 elevators in it. That would be a pretty complex system. It always seemed a little laughable to me to have an elevator operator, but in early buildings with a 100 floors and 73 elevators, it would be justified to have human elevator operators. They would have stood in for what today would be done by a computerized elevator control system. I could argue that electronic control systems that require a human being to remain “in the loop” for them to function represent a distinct class of technologies, and a class moreover distinct from technologies that keep a human being “in the loop” for reasons of judgment rather than reasons of functionality. The growth of sophisticated computerized and AI control systems has been steadily chipping away at the mechanisms that require a human being for functionality, and I could posit a threshold at which this human presence is entirely eliminated. We are, at present, clustered below this threshold, as multiple systems exceed this threshold even as many other systems still represent a level of complexity not yet reliably automated.
In a blog post of many years ago I coined the term “anonymization” in an attempt to describe production processes that verge on mass production, but without industrialization, as I was trying to describe some of the production processes of the Roman Empire, when large factories were turning out things like oil lamps effectively for mass consumption, but without the technology to engineer mass production. This is admittedly an unlovely word, and I’d like to come up with a better word to describe the same kind of production (which we could think of as clustering below the threshold of industrialization), as there is something of anonymization in the custom fabrication of advanced mechanisms in the early stages of a technology, as these technologies cluster below the threshold of the next technological advancement. So I could say that electromagnetic technologies clustered below the threshold of true electronic technologies, and as electronic technologies rapidly developed, driven by the demands of the Second World War, they in their turn rapidly came to cluster just below the threshold of solid state technologies, which soon became available even as enormous computers (like ENIAC) were being built with vacuum tubes.
It sounds like a contradiction to talk about custom fabrication being something like anonymous production, and this shows the extent to which I haven’t yet arrived at a good formulation of what’s going on (or what I think is going on). My own thoughts are clustered below the threshold of the next conceptual advance I can make in understanding technology, especially as it relates to the production processes of mature societies with several layers of what Austrian economist Eugen von Böhm-Bawerk called round about production processes, which is what Adam Smith was describing in the famous passage of The Wealth of Nations when he wrote a long paragraph detailing all that goes into the manufacture of a simple sailor’s coat.