Get best of the week

Epistemic standards of inventions

Finding answers to questions about why it took so long to invent a technology


Trying to study the history of progress, I constantly come across interesting examples of "ideas behind the times" - inventions that seem to have appeared much later than they could, such as a fiber separator or a bicycle. I put together a short list of such things and analyzed their occurrence.

Discussion of such issues often strangely turns into disputes in which people unsuccessfully exchange remarks with each other (although if people on the Internet can argue about how many days in a week, then, probably, you can argue about everything). Therefore, I would like to comment on our thoughts on such cases and standards of evidence.

To begin with, I will say about the need to observe a certain accuracy in the definitions. Take a steam engine. In antiquity there was such a thing, which was called a heron engine (turbine, ball), or eolipil . Some people, having seen it, come to the conclusion that “steam engines existed in the 1st century”, or that there are no reasons why the ancient Romans could not use such a thing in industry.



This is mistake. Eolipil does not at all resemble steam engines of the eighteenth and subsequent centuries: it is a turbine, that is, it rotates, and does not use the reciprocating motion of the piston, like the Newcomer steam engine . Why is it important? Because eolipil does not produce enough torque for practical use. In one analysisit is described that a Watt steam engine gave out a quarter million times more torque.


Newcomen steam engine

Even before I came across this analysis, I already had a suspicion that it was so - that’s why I guessed to search for the phrase “eolipil torque”, after which I came across this link on the first page of Google results. My intuition was based on several premises. Firstly, if such a simple and primitive turbine as eolipil could be put into practice, why no one invented it in the 18th or even 17th century? Why did Newcomen, Watt, and others focus on much more sophisticated piston engines? They were smart people, obviously actively working on this task - it is unlikely that none of them would have thought of such a simple solution. Secondly, the size of the eolipil is small - a sketch of Heron depicts it as a desktop device - however, Newcomen’s engine was huge,he had to build a separate barn for him. Why would you need to build such a huge engine, if only a tiny one were enough?

Again, an accurate understanding of each invention helps to reveal such important details. A simple concept such as “a machine (any) using steam (in any way)” is not enough.

This example also illustrates the second principle: practicality matters. A theoretically working device that, however, produces too little energy, is too inefficient, expensive or unreliable, from a practical point of view, doesn’t exist. It should work not only for demonstrations, but also for real, human, economic needs, in the context of the lives of consumers, industrial processes or business operations. Therefore, disagreements to the extent can turn into disagreements in the form of a device when the invention crosses the threshold of practicality.


Edison Light Bulb

By the way, that’s why the assertion that the bulb was invented in Edison’s laboratory, even though it was already before it, was absolutely accurate: they were either too expensive (for example, they used platinum threads), or they quickly burned out, because of which they often happened change. In my opinion, to say that someone invented the “first practical X” is redundant - it’s the same as saying that he invented X. To invent something means to invent a version that is applicable in practice. If your “invention” is impractical, this is just a demo, or prototype. It may turn out to be useful for testing ideas or explaining possibilities, but it is practical inventions that move history that remove obstacles to their widespread use.


The "propeller" da Vinci

Another example is a computer. The computer was invented by J. Presper Eckert and John Mokley of the University of Pennsylvania; their first model was ENIAC, completed in 1945. This was a breakthrough, as it was the first fully electronic computer, which made it run much faster than previous models, such as the IBM Automatic Sequence Controlled Calculator (ASCC or Harvard Mark I), which was electromechanical and used magnetic relays. If we compare the speeds of these machines according to Wikipedia, then ENIAC was 600 times faster than ASCC when dividing and more than 2000 times faster when multiplying. (Also, ENIAC was more than 2000 times faster than a person with a mechanical calculator that calculates a ballistic trajectory, which means that ASCC was not much faster than a person). ASCC was an interesting demonstration about which something was written in the newspapers; ENIAC was the computer that launched the computing revolution. Again, the difference in degrees becomes the difference in species. If you look further into the past, then computers such as the Atanasov-Berry machine or the Z3 Konrad Zuse also worked much slower than ENIAC, and also had other practical limitations. "A computer"Babbage was just an unfinished concept that could not be built with the help of modern technologies - therefore, despite my respect for his genius, I can not consider Babbage to be the inventor of a computer any more than da Vinci - the inventor of a helicopter.


Replacing the lamp in ENIAC

If you want to argue, saying that the invention could not have appeared before the occurrence of a “resolving” event, then I regard this story as an epistemological gold standard: strong economic motivation (and, in the case of computers, military motivation in the context of World War II) ; many previous attempts, including completed, well-working and fairly well-lit projects; the measured difference in a key parameter of practical importance (in our case, in speed); technology that allows to achieve a significant difference in this parameter (in our case, by about 3 orders of magnitude). Therefore, I confidently declare that the computer - again, I consider it unnecessary to write a "practical computer" - could not have appeared before the invention of the electronic amplifier tube in 1907.

It is likely, although it is not entirely obvious to me, that he could not appear long enough after that, when improved, more reliable electronic lamps were invented. It is likely that more than 17,000 lamps were used at ENIAC, and their reliability caused concern among engineers; it’s not obvious, because I don’t know examples of attempts to build a fully electronic computer on less reliable lamps, and because some statements by engineers say that reliability was not as serious a problem as previously thought, especially if the lamps worked without interruption without experiencing thermal loads. Therefore, perhaps, after 1907, engineers were stopped only by the lack of a key idea.


Dresin Karl Dres

On the other hand, if you want to argue that something could have been invented much earlier than this happened, you need to come up with some more serious arguments than just high-level concepts and components of the invention. It is necessary to carefully check every part, material, production process. Any, even a minor detail, can become critically important - especially if we recall that inventions should not only work, but be practical in terms of efficiency, reliability and cost. As an example, in my analysis of the invention of the bicycleI described the first proto-bike, the “trolley" or the Laufmaschine, made of wood, with wheels with metal tires - both of these technologies are quite ancient. However, as they later wrote to me, ball bearings technology was used in the trolley, which is much newer, and, possibly, critical for reducing friction.

In this regard, the question: how much do we need to be surprised to learn that the invention X took Y years after the invention of technology Z? Inventions do not appear immediately after they become possible. Ideas and information take time to spread, experiments are needed, financing must be found, laboratories organized, materials found; and as a result, all these actions are performed not by automata or clockworks, but by unpredictable human beings, with their own ideas, inspiration, hopes and fears, working in complex networks of teams, contracts, partnerships and other social structures. Even in the best case scenario, a gap of ten or more years between the advent of key technology and the commercial launch of the invention will not be surprising; if key technology was a scientific discovery,I’m not surprised for two or three decades. A case may intervene - a sudden illness, financial panic, war may confront you from the path to invention.


Cotton Gin (Cotton Fiber Separator) Eli Whitney

Overall, I think we should be more surprised at the long wait for inventions that have an obvious and predictable effect on large-scale industry. Therefore, I am more surprised by the later invention of the Cotton Gene and the weaving shuttle than the suitcases on wheels, role-playing games or bicycles, offering only convenience or entertainment. I think we can also expect a slower invention in places where fewer people live, they are less educated, there is less economic profit (funding research and development), fewer financing mechanisms (or they are less effective), less political stability, etc. .

I imagine innovation as a stochastic process in which some parameters are assigned by the environment and others by the very invention itself. The stronger the “pressure” to solve a problem (economic motivation), the more opportunities there are to solve it (educated, inventive people or organizations with free time, space, materials and financing in the context of well-functioning state institutions and political stability), the sooner you can expect a jump, and the smaller the time gap. In the limit, you can get instant inventions about which there are many stories (although some of them are exaggerated in the sense that was mentioned earlier - what is considered an "invention"). Some graduate students in economics may be able to defend a dissertation,formalizing this model and adjusting the parameters to the data - both to quantify the level of “ingenuity” in a given place at a given time, and to identify individual inventions that actually appeared “later than their time”.

More articles:


All Articles