Would you notice if science died?

Science is a big deal. It would be worth knowing if it stalled, regressed, or died out, whether the body of knowledge or the techniques for generating more knowledge. You could practice by reviewing history and looking for times and places where it stumbled. In this exercise you have the advantage of hindsight, but the disadvantage of much less direct access to the raw data of the scientific practice of the time. But regardless of how it compares to the real task, this is practice. This is an opportunity to test theories and methods before committing to them. There is a limited amount of history to practice on, but it’s a lot more than the real event, the present.

Many say that they would notice if science died because engineering would grind to a halt or even regress. What does this heuristic say when applied to history? Does it match other criteria?

Many say that the Greeks were good at science and the Romans at engineering (perhaps also the Han vs the Song). This is not really compatible with the heuristic above. What options do we have to draw a coherent conclusion? Either science did not die, or engineering did not advance, or science is not so necessary for engineering; Either we are bad at judging science from history, or we are bad at judging engineering from history, or engineering is not a good heuristic for judging science. None of these are comforting for our ability to judge the future. The third is simply the rejection of the popular heuristic. The first two are the rejection of the exercise of history. But if we cannot judge history, we have no opportunity to practice. Worse, if we are unable to judge history, the present may be no easier.

One recourse is to posit that the past is difficult because of sparse information and that the future we experience ourselves will be easy to judge. But many people lived through the past; what did they think at the time? In particular, how did the Romans think they compared to the Greeks? Did they think that there was progress or regress? Did they agree with modern hindsight? They thought that the Greeks were good at science. Pop science books by Pliny and Seneca are really accounts of Greek knowledge. Similarly, Varro’s practical book of agriculture is based on dozens of Greek sources. And the Romans were proud of their engineering. Frotinus urged his readers to compared the Roman aqueducts to the idle pyramids and wonders of the Greeks. Maybe he should be discounted for his professional interest. But Pliny describes the Roman aqueducts as the most remarkable achievement in the world in the midst of account of Greek knowledge. Indeed, the modern conventional wisdom is probably simply copied from the Romans. Did the Romans endorse the third claim, that science was a prerequisite to engineering? I do not know. Perhaps the they held that it was necessary, but could be left to Greek slaves.

I think that this example should make people nervous about the heuristic about science and engineering. But people who don’t hold any such heuristic should be even more nervous.

I think I know what the answer is. I think that engineering did regress, but the Romans did not notice. They were too impressed by size, so they made bigger aqueducts, without otherwise improving on Greek techniques; and they failed to copy much other Greek technology. Perhaps the heuristic is fine, but it just passes the buck: how much can you trust your judgement of the state of engineering? On the other hand, I think that science regressed much more than engineering, so I do not think them as coupled as the heuristic suggests.

Would you notice if science died? How would you notice? Have you tried that method against history?


Some historical test cases: the transition from Greece to Rome; Han vs Tang vs Song; the Renaissance.

New Comment
40 comments, sorted by Click to highlight new comments since: Today at 6:47 AM

First, I think that the chain looks like this: science produces theories/concepts/explanations/recipes, then engineering takes them and makes practical products on that basis, and then people use these products. So if science stalls, the pipeline will be empty and, basically, there will be nothing much for engineering to do except polish the existing products.

Second, looking at history is a bit iffy at the moment -- the reason is speed of progress. In our times (technological) progress is very very fast by historical standards. That makes it easy to notice if science dies. But that does not apply to, say, the Middle Ages when the progress was so slow it took many generations to produce an appreciable change. During one lifetime things (technologically) did not change much if at all.

Second, looking at history is a bit iffy at the moment -- the reason is speed of progress. In our times (technological) progress is very very fast by historical standards. That makes it easy to notice if science dies. But that does not apply to, say, the Middle Ages when the progress was so slow it took many generations to produce an appreciable change. During one lifetime things (technologically) did not change much if at all.

If you are correct, then historical figures have an excuse for not noticing technological stagnation. But that has nothing to do with my exercise, that we look at history. We can look at change over centuries and are not limited to the perspective of a single lifetime. You give an excuse for Pliny's error, although it doesn't explain why he thought differently of science and engineering. But we can look at actual history and not just trust Pliny.

Maybe stagnation is invisible, but regress can be quite visible, if only because of the artifacts left behind. Anyone can tell that Giotto had invented a technique of perspective just by looking at his paintings. All the Italian painters of 1400 knew that their art had regressed. We know that they knew because so many of them made failed attempts at perspective, unlike those before Giotto.

I am not sure what are you driving towards.

Did we have periods of technological stagnation? Of course, there is a reason a period of European history is called the Dark Ages.

Did we have technology regress? Of course, see above (or e.g. a story of how British Navy lost the knowledge of scurvy in more recent times).

Can you see the stagnation "in real time"? Of course, if you are using the same tools and technologies that your grandfather used, it's a pretty good sign of technological stagnation.

An interesting question might be "How quickly can you recognize that in your society science has stalled?", but I think the answer got to be "It depends" and here the (recent) speed of progress is very relevant. What you notice is deviation from the baseline of expectations.

If it is so easy to tell the difference between progress and regress, why do people living today disagree on whether Roman engineering progressed or regressed from Greek engineering? about whether the Renaissance was a period of scientific progress or regress?

If people disagree about the past, they may well disagree about the future.

why do people living today disagree on whether Roman engineering progressed or regressed from Greek engineering?

Because our information about the engineering of those times is very incomplete. We can look at surviving artifacts, but the available engineering texts from that era are a very small random sample of their knowledge.

about whether the Renaissance was a period of scientific progress or regress?

I was not aware that this is a contested issue. Links?

they may well disagree about the future

The question of whether the science has stalled is a question about the present (and immediate past). If you want a prediction of whether science will stall in the future, that's an entirely different thing.

I already gave you an example of Renaissance regress. Here is a longer essay.
Added: Here is someone even more negative about the Renaissance, although I'm not sure where he expressed that opinion except in his book (one book; two titles).

The example of Giotto is an anecdote of art technique, not regression of engineering.

And I don't think the essay claims actual regression (unless I missed it), it basically says that if you define Renaissance narrowly (1453-1564) then the intellectual achievements during that century are curiously lacking. As the author puts it, "everything came to a stop", though he spends a lot of time looking at literature, art, and philosophy. The Black Death is mentioned as a likely contributing cause.

But so what? Technological progress is not a smooth process, it moves in fits and jerks. Besides, it's multifaceted -- some chunks of technology develop faster, some slower, some are stuck for a while in the same place, etc. Even if you are having problem figuring out whether a particular century deserves a "Progress" or a "Regress" label, why does it matter?

P.S. The little note saying (emphasis mine) "For an account of the history of ideas free of anti-medieval prejudice, see my..." at the bottom of the essay is not a good sign.

Sure, if numbers are close to zero, it can be hard to tell if they are positive or negative. But I claim that there is substantial disagreement. If metrics live in a high dimensional space, the judgement can be sensitive to choice of projection, but neither is that responsible for the disagreements.

P.S. The little note saying (emphasis mine) "For an account of the history of ideas free of anti-medieval prejudice, see my..." at the bottom of the essay is not a good sign.

You asked me to document a controversy. That quote is a good sign for your purpose. I happen to believe that the Franklin is correct and the conventional wisdom is incorrect, but that is another matter. I added a link to Hannan, who has an even more extreme view that I do not endorse.

You asked me to document a controversy.

True, I accept that. The reason I said that is not a good sign is because it indicates he has an axe to grind and might be more interested in proving his point than in figuring out what actually happened.

On the basis of a brief look at your links, it seems to me that the controversy is really about, ahem, status of different historical periods. The people you quote think that Renaissance is overvalued, overexposed, overestimated, etc. while the late Medieval period is ignored and vilified. That might well be so, I have no opinion on the topic, but it's rather peripheral to your main topic of technological stagnation.

But I claim that there is substantial disagreement.

I suspect that a large part of that disagreement concerns what people are looking at. Once you give hard-edged definitions to the metrics you're interested in, much of it will go away. But if everyone is handwaving and implicitly or explicitly defining things in the way most advantageous to them, there could be much ado about nothing.

But if we run with the abstraction that science is input to engineering (and engineering is description of production (processes)) then it could very well be that all the technological progress we see could result from a very long back-log created by earlier science for engineering to catch up to and it could very well be that science has stalled to produce results (at least those that can be used as input to engineering) without us noticing anything wrong just looking at new products. Could.

Could.

Yes, and..? I am not sure what point are you making.

That we have only tentative evidence that science hasn't stalled from engineering productiveness.

Why would you care about that when you can go and look at science directly, without trying to proxy it with engineering success?

When you can reliably determine whether science has stalled then this argument isn't relevant. Some might disagree on that. I point out that in absence of evidence of science stalling or not we can't rely on current engineering output as a proxy. That's all.

Would you labels Google's project of AlphaGo "science" or "engineering"?

Probably engineering -- it is mostly about creating a working "thing" and not about discovering new underlying principles. But the boundary between engineering and applied science can be very fuzzy and there are often feedback loops between the two.

[-][anonymous]8y20

AlphaGo is absolutely science (as well as engineering -- all experimental science involves some engineering). It involves fundamentally new constructions...

It involves fundamentally new constructions

Like what?

[-][anonymous]8y00

The simple construction of using evolutionary learning to refine heuristics that were extracted from deep learning neural networks trained on expert data.

Two previously known and well understood components, put together in a new and novel way that expands our knowledge of what is possible. That is science.

Two previously known and well understood components, put together in a new and novel way that expands our knowledge of what is possible. That is science.

Interesting. I think that is pretty clearly engineering :-)

Of course, this is all a matter of definitions.

[-][anonymous]8y20

It's not an either-or. Some reasonable working definitions: Science is a process by which we expand human knowledge. Engineering is using extant human knowledge to construct artifacts, sometimes repetitive, sometimes novel. Doing some mindless engineering task is not science. But doing something innovative and new makes available new knowledge, which if processed in the correct way is doing science. So you can do both.

You are basically saying that the creation of s'mores was science ("previously known and well understood components, put together in a new and novel way that expands our knowledge of what is possible").

My idea of science is more narrow.

[-][anonymous]8y20

The first person that created a s'more? Yes. Culinary science is a thing.

Both. AlphaGo is a major engineering achievement in itself, and a pretty significant step in the empirical science of reinforcement-learning systems.

Does that interpretion suggest that the model of science first producing theories/concepts/explanations/recipes and engineering then using them is falsified?

Not strictly. It could very well be that

  • there is parallelism (think technology graphs from games)
  • that science feeds of from intermediate technological results

Would you label the LHC "science" or "engineering"?

I think the science/engineering-distinction used by Douglas Knight and Lumifer provides no good model, so you have to ask them.

It's both. I think the distinction can be reasonably clean - science aims at understanding via explicitly modeling the process (not necessarily mathematically but often) and then testing the model. The process of building the LHC was engineering, the experiments themselves are part of science.

The LHC is multiple things

  • a set of theoretical results describing what might happen under what physical circumstances
  • an application of said theory to a certain realizable sub-set of technological reality and the prediction of what happens then
  • an engineering effort to build a complex experimental apparatus

(and also a social process driving the people to do all this)

It would help a lot to start with a definition of the terms science and engineering before talking about the relationship between them.

Do you have an example where it would make a difference?

Somewhere in his book Thomas Kuhn equates science as being a field that makes progress in understanding it's problem domain. If someone manages to learn knowledge to build better aqueducts than could be build in the past, he's therefore engaging in science.

The relevant distinction:

science is about accumulating (edit: and systematizing) knowledge;

engineering is about building things, possibly but not necessarily using the knowledge accumulated by science.

Then the accumlation of knowledge that a YCombinator startup does when it goes out and interviews users of their product is science?

Possibly, I suppose that depends on how one would classify the "butterfly-collecting" aspect of science.

Was Roman engineering really based on Greek science? And by the way, what is Greek science? If I understand correctly, the most remarkable scientific contributions of the Greeks were formal geometry and astronomy, but empirical geometry, which was good enough for the practical engineering applications of the time, was already well developed since at least the Egyptians, and astronomy didn't really have practical applications.

It is a lot easier to document that the Greeks had cutting-edge engineering than to prove that it was based on theoretical knowledge.

Greek aqueducts and post-Greek Roman aqueducts were much better than pre-Greek Roman aqueducts. The process of building them may not have been better, but the choice of what to build was more sophisticated. Before the Greeks they just had water run downhill, requiring tunnels and bridges, afterwards they also ran water uphill. So the Romans definitely learned something from the Greeks. Some people think that they must have understood something about water pressure to do this, which would count as science. But there is no record of how they did it, neither theory, nor rules of thumb developed by trial and error. It is a great mystery that the surviving books by Roman aqueduct engineers don't seem adequate for running the aqueducts, let alone building them.

(By "the Greeks" I mean the Hellenistic period of 300-150BC.)

A better documented connection between theory and application is that Archimedes wrote a book on the theory of simple machines and invented the screw pump. However, that history is also controversial.

Thanks for the information.

The modern world and classical antiquity may have different relationships between science and engineering.

Perhaps the engineering achievements of antiquity were simple enough to be created empirically, without being backed by a scientific theory. You could build an aqueduct without mathematical physical theories, so you did, and science regressing (or not existing in the first place) didn't hold back engineering much.

But today engineering is creating greatly more complex things that are impossible without a formal scientific theory backing them. You can't make a transistor or engineer bacteria to make insulin without a lot of science. So today engineering is bounded by scientific progress.

Even in the second scenario, I think it's likely that a lot of things which can already be understood scientifically have not yet been done by engineers, so engineering could progress for a while in certain fields after science stopped.