The following is a rough transcript which has not been revised by The Jim Rutt Show or by Michael Strevens. Please check with us before using any quotations from this transcript. Thank you.
Jim: Today’s guest is Michael Strevens. He’s a philosophy professor at NYU that’s New York University. Welcome Michael.
Michael: Thanks very much. It’s wonderful to be talking to you, Jim.
Jim: Yeah, this is really great. I read Michael’s recent book then reached out to him. Sometimes I do it the other way around, but in this case it was just part of my exploratory reading and he has a new book out called The Knowledge Machine, How Irrationality Created Modern Science. And today we’re going to focus mostly on that book and probe into his work on the history philosophy and sociology of science. I would like to highlight for listeners that in a field that sometimes produces some really dry books, this one is really enjoyable.
Jim: It’s full of lots of interesting stories, historical examples, fanciful rewritings of Shakespeare and all kinds of stuff. So don’t be afraid of the topic. This is a very readable book and a very enjoyable. Today we’re going to talk a lot about the great method debate. Let’s start with what is the great method debate? What do you mean by that?
Michael: It is a argument that philosophers, and scientists, and historians of science have been having for, I guess, well, over 100 years now, about how it is that modern science has been so successful since the time of the scientific revolution back in the 16oos. Science has, look around, totally changed the world. What did people start doing around them that really made the difference that allowed them to discover the way the world works to build machines that take advantage of the way the world works that have transformed our lives so completely?
Michael: What’s the secret of science, that special method that makes science so different from the kind of thinking of say the Greek philosophers or the Chinese philosophers, the philosophers of the middle ages, people who were thinking about how it is that the world works but who’s somehow never quite got the momentum going to create the modern world that we all now benefiting from and dealing with? So the great method debate in short is the question of what is it that modern science has been doing so differently for the past 300 years that has made such a difference to our capacity to figure out how the world works?
Jim: Yeah, isn’t it astoundingly interesting question. I’m one of those folks who continue to maintain that science is a fundamentally different way of knowing and anything that came before, and how that emerged. And also as we’ll talk about later in the book, why it took so goddamn long right? Why couldn’t Aristotle had done it? He was a smart dude, knew a lot of things. And we’ll talk about that later. The next thing we want to go to something you explore early in the book, which is you go into some detail on the two most commonly known perspectives on method, which are the Kuhnian and Popperian views.
Jim: I’ve read both their books way back yonder, and have my own views about them. Could you give us some background on both of them? And you can assume the audience may not know who Kuhn or Popper are, but they could probably understand their ideas. Tell us a little bit about their perspectives.
Michael: Yeah, I’m sure. So Popper, and Kuhn are both writing very roughly around the middle of the 20th century, trying to answer this question about, what makes science special? Popper was an immigrant or refugee from the Nazis, he was Austrian actually. For a while he spent a few years in my own home country of New Zealand. And that’s where he worked on many of these ideas and kind of blissful seclusion from the disasters that were going on around him. And his answer to that question of, what makes science special? Is that it is a kind of hyper-critical rationality.
Michael: So scientists are consumed with a kind of a, you might say a negative spirit, any theory they hear, they want to undermine, and they’ll do almost anything to dig up some little facts that that theory doesn’t get quite right and in order to show that the theory is wrong. So scientific progress unfolds according to Popper through this barrage of criticism. And the thing that makes scientists so effective is not some special logic or methods, so much as just the intensity of their devotion to refuting any theory that comes across their path. Kuhn is almost the exact opposite of that.
Michael: For Kuhn that secret of science is that scientists are so uncritical. So you couldn’t really have a greater contrast than this. Now when I say scientists are critical, that’s how science progresses. That’s, there’s something very intuitive and common sensical about that. Whereas Kuhn’s idea seems like a very bad way to start explaining how science could possibly be successful. Scientists are just disposed to believe that they received wisdom, what Kuhn referred to as the paradigm, that the accepted way of doing science, everything that the scientists learned in graduate school, Kuhn says they tend to take on uncritically.
Michael: And what they acquire is this enormous armory for doing science, but it’s for doing science exactly the way it’s always been done. So how is it that you ever get any scientific progress, if there is this kind of enormous ultimately ideological complacency that is the character of modern scientists? The answer is, Kuhn says that scientist is so confident that all of the ideas that they’ve learned, that they’ve been using uncritically will solve every single problem that they push, and push, and push to apply this framework, this paradigm to absolutely every issue that it could possibly be applied to.
Michael: And they start to run into problems because, of course, the paradigm is not perfect. In fact, the paradigm may be deeply mistaken. It might get the fundamental causal structure of the universe upside down, so that in the same way that, according to Popper, scientists uncover problems with the theory and the theory collapses and something new has to come along. Kuhn says, yes, that’s the way science operates.
Michael: Except the reason that scientists uncover these little refuting facts is not that they are so determined to undermine a theory, is that they are determined to show that the theory works perfectly for everything, and they push it too hard and it breaks down. There’s plenty more that Popper and Kuhn have to say, but that is the basic contrast between them in there are different stories about the way that science becomes a very discriminating machine for distinguishing truth from falsehood.
Jim: Yeah Michael and one other thing, which you didn’t quite hit on I don’t think, which is that Popperian view of science is then any statement which is not falsifiable or at least not apparently falsifiable shouldn’t even be counted as science. People talk about that a lot. And yet we also know it’s not necessarily how actual scientists work and we will talk about some examples of that later. Does that make sense to you?
Michael: Yes, that’s right. So Popper said… Popper had two modes. There was that, he could be very forthright and lay down these maxims and ways that made them seem like sort of somewhat ironically perhaps indisputable dogma, but he was a little bit more subtle about that typically later on in his… Or deeper in his writing. So on the one hand, he said, if this critical attitude that scientists have is going to do its job, then a theory has to be in a position to be criticized. It has to make predictions put itself on the line so that it can turn out to be wrong. And that’s something that makes a lot of sense about that.
Michael: That’s been perhaps taken a little too seriously by some scientists, but maybe we’ll end up talking a little bit more about that later. But it’s very important to both Popper and Kuhn that scientific theories have things to say about these very abstruse issues, exactly where the light of a particular star will be observed and so on. That they commit themselves to an enormous range of predictions in a certain sense that expose them to the danger of being shown to be wrong.
Jim: Yeah. It’s like you say that despite their seemingly great philosophical differences, they do have things to agree on. And one is that, one of the core things that drive science is the need and the reality of eliminating old theories that science proceeds by one fashion or another either through this paradigm shift or through falsifiableness of taking previous ideas and eventually getting rid of them.
Michael: That’s right. So the engine of scientific progress is reputation.
Jim: Yep. And the second thing that you agreed they were both right about, though in perhaps a little bit different ways is that in addition to the engine of science, the method itself, motivation is really important. And they have two different perspectives on motivation. Maybe you could talk a little bit about how the importance of motivation as a driver of science and differences between Popper and Kuhn’s view on that.
Michael: Yeah. So I think that the question motivation is the single most important question to answer in understanding why modern science is so superior as a method for figuring stuff out compared to the kind of natural philosophy that came before. And the reason is, that it turns out that these refuting facts, the facts that show the way, I mean we have as human beings are so great at coming up with ideas. We have so many fascinating theories. If you read the history of philosophy, you’ll see just how prodigious we are with our theoretical imaginations. To make progress we need to figure out which of those theories are wrong.
Michael: And that’s turned out, and I don’t think anyone foresaw this until a few hundred years ago, it’s turned out that the way to undermine a theory is to look at these extremely detailed predictions. And to look at those predictions in that kind of detail requires an enormous amount of commitment of time, sometimes of money, and always it requires incredible patience on the part of the scientific experiment. Maybe we’ll get to talk a little bit about exactly how much patience it takes.
Jim: Yeah. We’ll get to that when we get to your three rules, the tripod, right? And yet those of us have been close to real science, know it doesn’t really feel exactly like either Popper or Kuhn. And I think that’s where you start your exploration from the base that a lot of people know about to newer thinking. And you start out with a very interesting story. I’d love you to go into the story in some detail, which is the story of Eddington’s expedition to verify or refute Einstein’s theory of general relativity.
Michael: Great. Okay. This is really one of the most famous scientific experiments in the history of science in some ways, because it was so crucial for testing Einstein’s new theory of gravity. So for hundreds of years, Newton’s theory of gravity was reigned supreme, was accepted as true. And I think not long before Einstein was formulating his theory around in the very early 1900s was thought to be about as incontrovertible as any scientific theory that the human race had produced. But there were discrepancies exactly the kinds of things I was just talking about. Little facts that Newton’s theory didn’t get quite right.
Michael: For example, its prediction of the orbit of mercury around the sun was just a little bit off in ways that nobody could explain. So Einstein had this wonderful, very highfalutin idea about the way gravity might work, actually an idea in which there really isn’t no gravity. What feels like gravitational force is just kind of objects traveling through space and time along the straightest lines they can find. It was an idea that really appealed to some scientists because of its mathematical beauty, and also just I think it appealed to scientists who were looking for something really new and revolutionary.
Michael: Arthur Eddington was one of those scientists. And when Einstein published his theory, began to become available to the rest of the world around 1915 or 1916. So it was right in the middle of World War I. Einstein was in the German speaking world so he was in some ways the kind of person who would be considered an enemy by Eddington’s compatriots, the British. When the war finished, Eddington wanted to both find out whether Einstein was right. Well, the truth is Eddington was rather convinced already that Einstein was right. And also to bring about a kind of a rapprochement at the end of the war between the English speaking and the German speaking worlds of science to show that now the war was over the English speaking world and the German speaking world could work together.
Michael: So Eddington’s idea was to test a prediction of Einstein’s theory which hadn’t been, by looking at some data, which hadn’t really been examined before, because it was so difficult to collect, which is the degree to which a light would be bent by a really powerful gravitational field. Really the only available very powerful gravitational field given the technology of the time was the sun. So the idea was to look at star light, which was passing very close to the sun. In other words, look at the light of stars which were right next to the disc of the sun and see how that light was bent by seeing how the apparent positions of the stars would change compared to when they were being viewed when they were nowhere near the sun.
Michael: So in other words, you would want to look up at the night sky, record the positions of a certain group of stars, then wait 12 hours until the sun were right in the middle of that group of stars. And to the degree that light is being bent by the sun’s gravity. It will look like those stars have moved just a little bit further apart from one another. So the only problem with this suggestion is that is the sun is right in the middle of a field of stars, it’s so bright that you can’t see any of the stars. So what do you do? Answer, you wait for a solar eclipse. And there was going to be really a very suitable eclipse in 1919.
Michael: So right after the end of World War I. An eclipse that would in its totality, first of all, be a total eclipse, totally blurred out the sun, at a moment when the sun was in a part of the sky where there were some relatively bright stars, very close to it. So it was a great opportunity to go and perform this measurement. The measurement was incredibly finicky apart from the difficulty of just traveling to the editorials where the eclipse could be observed. And Eddington’s team actually went to two different places, one in the North of Brazil and one in Príncipe Island off the coast of Africa.
Michael: Apart from the arduousness of the travel, and the difficulty of setting things up, and the possibility that the weather would be terrible, the kinds of differences they were looking for were microscopic. They were so small. It’s still hard for me to believe that they could get any results at all. So what they were going to do is take photographs of these stars at the time of the eclipse and compare them to photographs taken of the same stars when the sun was nowhere near them, and compare them. And as I said, it would look like the stars had moved just a little bit further apart when the sun was in the middle of them.
Michael: But the distances on those photographic plates that they would apparently move were a fraction of a millimeter. And Eddington is doing this, of course it’s all before computers and so on. He and his team were basically just taking literally photographs and comparing the photographs to measure these little sub-millimeter distances, actually much less than a millimeter. And the exact amount of those differences would determine whether Newton or Einstein was right, or whether both were wrong. So that was the challenge that they were facing.
Jim: Yep. And as they went into it, they ran to all kinds of difficulties, and judgment calls, and things like that. I want you to tell a little bit more of that story when we get to then the punchline about the role of subjectivity in the analysis of the results.
Michael: Well, it was pretty in a way it was exactly the way that has science often work. So they had to sail there of course, this was 1919. They spent months getting there and setting up their equipment waiting for the big moment, really just a few minutes of totality with the solar eclipse. Eddington was in Príncipe that Island off of Africa. And when the eclipse came, it was cloudy. He got a few sun breaks, he got a few blurry photos, but that was all. Meanwhile, the team in Brazil had better luck with the weather, they did get some photographs. However, some of their equipment seem not to be working so well. In fact, they ended up getting two sets of photographs.
Michael: They had two telescopes with them, or in fact, they carried the lenses of the telescopes with them and then simply rebuilt the telescopes in the site where they were taking the photographs, but they were able to get some data from this. And so at the end of this great expedition, they had some data from one, Brazilian telescopes and data from the other Brazilian telescope, and some blurry data from the telescope in Africa. And they got, and this happens all the time in science, they got contradictory results. One telescope showed a pretty Einsteinian shift, that was one of the Brazilian telescopes.
Michael: One telescope showed the slightly smaller Newtonian shift. So that suggested that Newton’s theory was right. And the African telescope was kind of hard to interpret, but Eddington did a lot of fancy mathematics and argued that it also supported Einstein. So that was the situation they had spent a good part of a year collecting this data. And it seemed there was no clear result.
Jim: And so then what happened when they went to process this seemingly inconsistent and not particularly sharp date?
Michael: All right. Then we might look back and ask what Popper or Kuhn would have thought would happen in a situation like this.
Jim: Yeah, good idea.
Michael: So if Popper is right and then scientists are born to refute and Eddington however much he might’ve admired Einstein’s mathematics was sending out ultimately to see if he could falsify Einstein’s theory. That is to find results that would show Einstein is wrong. Of course, also to falsify Newton’s theory, what could make a Popperian scientist happier than falsifying two theories at once? Well, it seems though that Eddington was rather more interested in falsifying Newton’s theory and supporting Einstein’s theory. And he ended up arguing that,that one telescope in Brazil which gave data that actually fitted Newton’s predictions very well, had been malfunctioning.
Michael: Now, there was some reason to think that something had gone wrong. That although the weather was okay, the results were a little bit blurry. There must’ve been some explanations for that. There could have been a number of things. He was analyzing these results months later after the team returned. So no one really knew exactly what was wrong, or what had created the blurriness. The team had some guesses, but they were just guesses. They hadn’t done any independent checks to figure it out. Of course they were way too busy spending their precious few minutes simply taking the photographs.
Michael: So Eddington argued that what had happened is that there had been a kind of general expansion due to the sun’s heat in the apparatus and that particular telescope, although not for some reason, the one next to it. And as a result that had created a systematic distortion in the mirror of the telescope, and therefore the results were all systematically off. So in effect, he argued something went wrong as a result of which the photographic plate was just slightly off scale. And that just happened to be a change that made the results look like the results that Newton predicted.
Michael: Even though Eddington went on to say the actual positions of the stars or apparent positions were much more in line with Einstein’s prediction. So he said, we should listen to this one telescope, but not to this other telescope. And he was in a position actually, very good connections in the establishment. And he was in a position to essentially push through this interpretation of the evidence. However, a number of other scientists who were not immediately within Eddington’s sphere of influence, like some American scientists for example, were quite suspicious of the whole operation and thought it was, I mean, they didn’t know what had happened with… They didn’t know where the blurriness came from either.
Michael: But they thought it seemed entirely possible that it was not the one telescope but the other that had malfunctioned. Not that they were, they clearly there was a conflict in the data and they thought maybe the right conclusion as well, we just don’t have good enough data yet we have to go back and do it again. Unfortunately, that would have meant waiting years and years for another suitable eclipse to come along. So Eddington did not act in a Popperian way. He did not simply take his results to have falsified one hypothesis and not the other. He was a little bit more like a Kuhnian scientist in the sense he was very much attached to one theory and he did everything he could to make that theory appear to be true.
Michael: But he wasn’t really a true Kuhnian scientist because a true Kuhnian scientist has a certain kind of oblivious innocence when it comes to interpreting the data, they just can’t help but believe that their theory, their paradigm is right. And they see everything in the light of that paradigm. And without really knowing it, they impose their biases on the interpretation of data. But Eddington was a little bit more Machiavellian than that. I mean, I think he did genuinely believe that Einstein was right, but he went to work to use his social connections and so on to give Einstein, let’s say a leg up evidentially speaking in the scientific literature.
Michael: So it was a little bit more clear-eyed about than a Kuhnian scientist would be about the conflict and a little bit more of a manipulator. Maybe as I say, a manipulator who is ultimately interested in finding the truth, but who had some very strong ideas about what the truth was and was prepared to indulge in some arguments that struck some of his colleagues as rather dubious in order to push that truth through. So we have neither Popper or Kuhn is quite right. We neither have kind of a completely impartial refuter nor do we have a totally credulous idolatrous respect for the theory.
Michael: We have something else, something that actually looks a little bit more like human beings, as I think we all know, and mostly love them, somebody who has some genuinely good motivations. Eddington wanted to find the truth. And he also wanted peace, that a peace that would last between the English and German speaking worlds. But someone who is also willing to push things a little bit to pull some strings to indulge in a little bit of, end justifies the means operation behind the scenes in order to get what he wanted. Much more of a kind of a standard human operator, I would say.
Jim: Yeah, I often refer to that as the sociology of science, right? Scientists have a method, but they’re also humans, right? And they are driven by all the usual human constructs, jealousy, greed, loss, you name it. And so not at all surprising. He also introduced a very interesting supporting concept.
Jim: First time I’d heard about it, maybe it’s common in your field, which you called the theoretical cohort. Which is, analyzing the data is not always as straightforward as we think, because things around the data need interpretation. Could you talk a little bit about the idea of the theoretical cohort and how it was relevant to the Eddington project?
Michael: Yeah, sure. So the reason that these arguments are possible is that theories on their own never really make predictions. And a very simple and straightforward way to say this is to think about that, Eddington experiment, the data you have, the data that Eddington publish was these very precise, meticulously recorded differences in the position of the stars when they were close to the sun, the apparent positions, I should say, of course, and their positions when they were not close to the sun. So you have these little changes in the stars positions. And we like to say, I think I’ve said already, Einstein predicted a certain amount of change and Newton predicted a certain different amount of change.
Michael: And if that were true, we could simply look at the positions and say, okay, Einstein got it right here, and Newton got it wrong or vice versa. But it’s not really true that Einstein predicts that certain dots on a photographic plate are going to be in a slightly different position. To get that prediction, you have to make a whole bunch of other assumptions as well. For a start, you have to assume that your whole telescope set up and photographic setup is working the way it’s intended. That you don’t have one of these systematic distortions that Eddington thought he had with one of his Brazilian telescopes. You’re also typically making a bunch of assumptions about the way the world works as well.
Michael: So in this particular case, you’re making assumptions about the positions of the stars. I mean, you’re trying to measure them at the same time, of course, the mass of the sun and so on and so forth, that go into extracting from some big theoretical idea at a particular prediction. So the thing that’s really making the prediction is all of those assumptions together plus the theory and I call that the theoretical cohort, this big kind of body of assumptions, if you like, that may be traveling along with the theory and is needed to make that contact between the theory and the world. And whenever something goes wrong, whenever you get a prediction that looks wrong, like that Brazilian telescope data that seemed to show that
Michael: Einstein was wrong and Newton was right, you can always say, well, maybe it wasn’t the theory. Maybe it was one of these other assumptions, maybe the telescope wasn’t working properly, maybe some of our other parameters that we thought we’ve measured so carefully are a bit off and so on and so forth. And a lot of science consist in, because so much scientific data is often conflicted because scientific experiments are so complicated, a lot of science consist in a scratching your head and saying, well, I wonder if this went wrong. Well maybe this instrument wasn’t working quite right. Or maybe this assumption that we’ve all accepted is a little bit off and so on and so forth.
Michael: And this creates almost infinite scope for haggling, for arguing over the significance of data. Really quite the opposite I think of the picture of science that a lot of the general public have, that they’re often taught in high school I think. That once you have the data, you can simply bring a certain kind of scientific logic to bear to see which there is a support and which are undermined by the data.
Jim: Yep. And that could end up as an unresolvable mess which never converges. But then you point out that there’s a perhaps a saving algorithm, which you’ve named the iron rule, which constrains at least the formal discussion about both the theory and the theoretical cohort. So let’s introduce this first element of your tripod, the iron rule. What is it? What’s it say?
Michael: This rule says just one very simple thing. It says that scientists when they argue with one or other about this stuff, when they’re having these disputes, have to resolve their arguments by doing more experiments or making more observations. I think we’re so used now to the idea that science operates in this way that almost sounds too trite to be of any importance. But what really matters here is that it prevents scientists from arguing by philosophizing, by looking to the aesthetic properties of different theories. So Eddington, however, beautiful he thought Einstein’s theory was, couldn’t say Einstein has to be right.
Michael: It’s just such a lovely theory. That’s why we should believe one of our telescopes and not the other. In the end, what happens? Well, what happened with these observations is this, we will never know exactly what happens with those telescopes. And in particular that Brazilian telescope that gave the Newtonian results. So why is it that we now believe Einstein rather than Newton? Well it’s because scientists went and did more experiments. They couldn’t in the end resolve the question of which telescope was right, or which was wrong by going back and looking at the telescopes. Many observations of lights bending have been done since, in the intervening now, what is it?
Michael: That’s 100 years now. There have been more suitable eclipses. Measurements have been made using equipment that’s even more carefully calibrated that Eddington’s equipment. And it’s turned out in the end that it’s Einstein’s theory that makes the right predictions. So these disputes are resolved not by figuring out who is right and who is wrong about these early experiments, but simply by doing more experiments. And it’s the iron rule that says in the end do more experiments.
Jim: And then very importantly, and I think you make this point nicely in the book, is despite the conscious or unconscious biases that Eddington may have brought to the work, the iron rule basically constrained him from putting those formally into play in his publications. His publications were dry and proper within the iron rule. At least that’s my takeaway from your telling those stories, is that about right?
Michael: That’s exactly right. So what Eddington published, there’s a range of opinion about exactly to what extent Eddington was playing with the results was consciously doing a little trickery behind the scenes. But nobody disputes that Eddington in his publication, Eddington and his team presented a full, proper truthful record of what they saw in their photographic plates. So all the data was laid out there. In so far as there was anything Machiavellian going on, it was behind the scenes. So what you get in the scientific paper is something that has all of the interpretive framework if you like, stripped away just the evidence.
Michael: And what happens if scientists keep going back and doing more observation, more experiments than this dry, meticulous, but objective record of the evidence starts to pile up? And as more and more evidence comes in, then with a bit of luck, and history shows us typically we do have just that little bit of, we start to see that one theory is consistently getting it right most of the time, not all the time, because things are always going wrong, and other theories are starting to consistently get it wrong. And we get a kind of a convergence of opinion as scientists are more and more persuaded of whatever their biases that only one theory can really make sense of the evidence.
Michael: So the progress of science, I mean, Popper and Kuhn are right. That the progress of science hinges on building up this great inventory of extremely detailed and often it’s very difficult and expensive to obtain facts. And that the real question about what drives science is the question of what pushes scientists to keep going back and doing more measurements? I think that the answer ultimately is that science is set up for scientists. Is that kind of a bit of a game if you like, and the iron rule is the fundamental rule of the game, which says that the only legitimate move in the game is to do another measurement to make another observation.
Jim: Yep. I very much like that analogy because I do know a lot of scientists been involved in science governance, and some little bit of science myself over the last 20 years since I retired from business and the dry Popperian or Kuhnian perspective never resonated with me as the way scientists actually operate. They get their ideas and they have their opinions from all kinds of strange sources.
Jim: I know one guy who gets most of his ideas from doing hallucinogens, right? But it doesn’t really matter. According to you the iron rule so long as they only communicate via objective statements backed up by data or experiment, there’s no constraint on how they get their ideas, which I think is actually cool and is much more of the way humans might actually operate.
Michael: That’s right. And there’s no constraint on how they think about their ideas. So they don’t have to be ruthlessly critical like Popper. They can baby their ideas and argue for them, do as much special pleading for them as they like when they’re talking to their colleagues and their rivals. But at the same time, they might maintain quite a bit of distance from their ideas. What’s important is not so much their ultimate emotional attachment to the ideas, whether they desperate to refute everything as Popper says, or whether they’re hopelessly in love with their theories Kuhn says.
Michael: Probably for most scientists it’s something very much in between. What’s important though, is when they play the game, which means going into the lab and then publishing what they find in the lab, they’re constrained in this way, that forces them to always be generating more evidence and putting more evidence on the table. The thing that really matters is not so much their attitude to the theory whether it’s love or hate, but their compulsion to play the game, therefore to be a scientist ultimately by making measurements, by making observations.
Jim: Okay. Let’s maybe compare and contrast that a little bit with pre-scientific, whether you called natural philosophy. How is that different than Aristotle? And this might be a good spot to also probe a little bit on the seeming oxymoron in your subtitle, How Irrationality Created Modern Science. Talk about that a little bit. Why would Aristotle think this is kind of a crazy way to proceed?
Michael: So Aristotle has sometimes been called the first great scientist, and there’s a sense so much that’s a very apt title. So I think probably most people think of him as a great philosopher. And a lot of his work was philosophical in a narrower sense, metaphysics and all of that stuff. But he wrote an enormous amount that in its subject matter, and in its style, seems rather scientific. So his biological writings for example, contain an enormous amount of observation of the behavior of different lifeforms, animals in particular. And a lot of theory trying to explain the particular kinds of things that animals do, the features they have, why they’re built the way they are.
Michael: He was a great observational biologist. Likewise, when he was doing his astronomy, he was very much concerned to explain why the stars and the planets moved the way they did, or seem to move the way they did, the nature of various forms of weather, thunder and lightning, and meteor showers and so on. He was very much someone who was fascinated by the world around him and wanted to formulate theories that explained why the world had all of the character that it did. So he was coming up with theories that explained what was seen, what he saw with his own eyes, and what other people saw in both the natural world, the physical world, and also the biological world.
Michael: He was in that sense, he was a great scientist. But there’s another sense in which he wasn’t a modern scientist. Well, obviously he wasn’t modern because he was living almost two and a half thousand years ago, but he wasn’t doing things the way a modern scientist does them. And that is that he was not constrained by the iron rule, which is to say he didn’t think that all disputes or arguments about which theories are correct and what the evidence shows us about theories should be resolved simply by making more observations. Instead, he thought that philosophizing was a great way to resolve these disputes.
Michael: So what we should do is once we’ve formulated these different theories, which can explain the things we see around us, we then compare the theories for kind of philosophical coherence for appropriate connections to other aspects of our thinking. We focus on the big ideas once we’ve finished doing the observations. And in the end, we find the truth through philosophical disputation. Now that didn’t work out. And I’ll say a little bit about why it didn’t work out in ways that really Aristotle in no way could have foreseen. And then I’ll say a little bit about the irrationality and its role in creating modern science. So why didn’t the, if you like, the Aristotelian way of doing science workout?
Michael: Well it’s because Aristotle was not so much looking at those little details. So once he had gotten a sense of broadly or qualitatively how things moved, he then moved on to his theorizing and his philosophizing. It has turned out, and as I say, I don’t think this was not something that could be anticipated. It wasn’t anticipated. It turns out that philosophizing is not so useful for discriminating among theories. And that observing little details like Eddington’s quarter of a millimeter really matters. Aristotle for a start would not have thought that the quarter of a millimeter mattered very much, because he would have thought of those tiny little differences as being just kind of random noise.
Michael: So sure one time you do your observations and you get these numbers, and then you get slightly different numbers the next time around, but that’s just little jitters and buzzes and so on. And the data, you could never infer from a difference, that microscopic, that one theory is right and another theory is wrong. Instead it’s turned out to be completely wrong. What’s turned out to be the case, is that it’s precisely those little differences that give us the power to discriminate among theories.
Jim: Very interesting. In fact, I often will call, I think of Aristotle and all the earlier free modern people who were really smart people and did some very interesting work. I call it, they had the philosophers’ disease, no disrespect intended, in that they overestimated the power of metaphysics in particular.
Jim: Aristotle again and again comes back to essentially metaphysical claims about the nature of reality, which for instance, drive his physics of motion, which even the most trivial experiment, which could be done in two days, as Galileo indicated refutes it. But because he had this elaborate metaphysical argument, he never felt the need to do the actual work. He thought metaphysics was the most powerful tool in some sense.
Michael: I think there’s a lot of truth in that, you’re being just a little bit unfair to Aristotle in the sense that he thought he had done the work. So he really cared about having his theories, explain what he observed. And he did observe a lot. But he didn’t observe in the kind of quantitative detail that turned out to make the difference. So they have Galileo just about 2000 years later, sliding things down a plane and recording exactly how quickly they get to the bottom. Aristotle never would have thought that there would be much to be learned from doing that.
Michael: You get the exact rate at which they get to the bottom, depends on a bunch of complicated stuff and just a certain amount of kind of randomness or nature’s whim, if you like, but you wouldn’t have been able to discern the real principles of physics by looking at those numbers, so Aristotle thought. He was indeed wrong, but I would really like to emphasize not just overconfidence and philosophizing, though I think there was that. But also a failure to see how telling the small details ultimately could be.
Jim: Yeah, that’s a nice distinction. Let’s go onto the next step forward towards our modern science, which was Francis Bacon. What an interesting character. Tell us a little bit about him and the work that he did that moved us towards modern science.
Michael: Bacon is a really interesting character. He’s right there at the beginning of the scientific revolution. And he lays out a way of doing science that looks very much like what was needed as a cure for, if you like, for Aristotelianism. So what Aristotle needed to do, if you don’t mind me just backing up a little bit, what Aristotle would have needed to do is to have, if you like, taken on board the iron rule than a lot less philosophizing and not a lot more minute observation. He never would have done that. He would have thought that it was crazy to abandon philosophy. This is the irrationality if you like, and devote himself solely to observation.
Michael: But Bacon writing in the early 1600s lays out a program for doing just this. For simply now can be it’s possible to parody it a little bit. Maybe I’ll do that. Maybe a little bit of parody is not a bad idea in this situation. Bacon laid out a program for simply accumulating enormous amounts of facts. And he said, don’t start theorizing, simply observe, observe, and then do more observation, fill your store houses with facts, small and large. And only at the end of the day, when you’re ready, step back and look at this great treasury of data and figure out what theory is going to be capable of explaining it all. They’ll only be one, he thinks, and that theory is the one that’s guaranteed to be the truth. So he has this, unlike Aristotle, a very evidence driven and unphilosophical prescription for investigating nature.
Jim: Yeah. And he calls out in fact, what he calls idols that we have to discard. And again, I think this kind of gets back to your title, How Irrationality Created Modern Science, in some sense if people long believed in the efficacy of metaphysics, or the truth of revealed religion, or the authority of famous figures of the past, it does seem irrational to give up those tools because they have been thought to be of quality for thousands of years. And someone like Bacon is saying, nope, those tools are no good. You have to only use this one much seemingly narrower method.
Michael: That’s right. I think that Bacon lays out something that looks like exactly the prescription that we needed to do in modern science, but it is not in fact Bacon who was responsible for the advent of modern science. And the reason is that Bacon looked to many of his compatriots, not all of them, but the great majority, the same as he would have looked to Aristotle as a kind of a zealot. Harry is saying philosophy has no use, but he really had no arguments for that. He was simply prejudiced against philosophy. And he was saying, go out and collect all this evidence. But I must say although he did some of it himself, he didn’t in fact devote an enormous amount of time and effort, the way that true modern scientists devote time and effort to collecting evidence.
Michael: It’s very easy to say, devote your life to making these little measurements, meanwhile, I’ll be writing my big books about how science works, but to actually get scientists to do it requires a kind of motivational machinery that Bacon for all of his prescriptions, and his arguments, and his attempts at persuasion, did not succeed in creating. What we needed was something a little bit more precise. Well, I’ve used the notion of a game before, we needed to, as it were game for science. And Bacon did not quite do that. It was the creation of the iron rule that did that. Not too much later, but later.
Jim: Following on your discussion of Bacon, you talk about plausibility rankings and Baconian convergence. Let’s talk about that a little bit.
Michael: Right. This will take us back to the idea of a theoretical cohort. So all this evidence is being accumulated, but as I’ve said, no theory by itself makes predictions about the evidence. It’s always that a theory makes predictions when you add a whole bunch of assumptions. And scientists disagree about these assumptions. So Eddington was pretty confident that one telescope in Brazil that delivered the Newtonian looking data, he was pretty confident that, that telescope had malfunctioned in a certain systematic way. Some of his colleagues were not so convinced. They thought, well, maybe, but then again maybe not, maybe it was the other telescope that had something wrong with it.
Michael: I mean, if Keith can create these distortions, why not the other telescope? And these differences in opinion, what I call differences in plausibility rankings, that is the different scientists will find different assumptions, which are crucial to be differently plausible. So one of them will think this is very plausible. The other will think, not so plausible. And this is the source of all of the subjectivity you find in scientific reasoning. That the reason that one scientist like Eddington maybe will think that this experiment provided really powerful evidence in favor of Einstein, and another scientist will think that it was inconclusive.
Michael: The reason you get these subjective differences is because you have these differences in opinion about how plausible these various little assumptions are. So you might think, why doesn’t that subjectivity last forever? Well, the reason is that when the evidence begins to pile up, some of those assumptions start to look much more questionable. So back in 1919, when Eddington had just returned with his eclipse data, there wasn’t that much to go on. One of Eddington’s critics could quite easily and reasonably say, well, I think it was the other telescope that had something funny going on with it.
Michael: 100 years later when we have many, many of these measurements and the overwhelming majority of them have gone the Einsteinian way, plus we’ve done independent checks. The next time around, we take much more care that we don’t get some of the problems that Eddington suspected his telescopes had. It’s much harder to say, well, it could have gone either way. And so you get a kind of convergence of opinion, only the real diehards are going to say, no, it’s going to be the case that in every single one of these experiments that can looks like it favors Einstein, something went wrong with the telescope. It starts to look like conspiracy theorizing.
Michael: And well, as we know there are conspiracy theorists on the whole scientists are, although they’re overflowing with all of the usual human prejudices biases, one sidedness, there’re still at least somewhat reasonable people. So when enough evidence piles up, you do start to get a kind of convergence of opinion. I call this Baconian convergence, because it’s the kind of convergence Bacon thought we would have when we have these dreads store houses of fact and we look at them and we see only one side, when only one theory can really make sense of it.
Jim: Yeah I was kind of a little surprised that you used Baconian convergence after having previously said, well, Baconian science really isn’t going to work, but it’s an interesting coinage. I sometimes use the term inter-subjectivity or inter-subjective consensus for similar idea. I suspect that we’re talking about pretty much the same thing.
Michael: Oh yeah. Yes, we are.
Jim: The fact that a community of people who are working together reach a inter-subjective sense that enough evidence is accumulated to explain the anomalies and that, as you say if you’re always crankishly arguing, it’s always the camera dude, despite 125 different cases with 100 different cameras, then you’re falling outside the kind of inter-subjective consensus making, which is clearly an important part of science.
Michael: That’s exactly right. And I think Bacon, the sense of much Baconian science doesn’t work is not so much that once you get enough evidence, you can’t really begin to sift truth from false. So that’s exactly what we’re talking about here with this kind of convergence of opinion, this inter-subjectivity, which we all hope is not just inter-subjectivity, but a certain kind of objectivity a certain kind of convergence on the truth. The thing is though the biggest flaw in Bacon’s plan was he had no motivational mechanism for getting scientists to go out and just keep collecting stuff without thinking about it.
Michael: So I think the reason Bacon didn’t see that flaw, or didn’t regard it as a flaw is he thought that relatively little evidence would be needed to figure out the truth. I think he really thought maybe within his lifetime if enough scientists got to work, collecting the facts, then we would have pretty much have sorted out the main theories. Have the basic physics sorted out and so on. I in a way would I have done any better? I doubt it. I think it was a reasonable opinion that he had.
Michael: Now, of course, there hasn’t turned out to be the case, that was 300 years later, we were still overthrowing one theory of gravity Newton’s and crowning another Einstein’s. But the basic idea that with enough little observations, you’ll eventually see how it all works. That basic idea of Bacon’s is still an idea that lies at the foundation of science.
Jim: That’s a great transition to the third leg of your tripod. I call it your tripod, I don’t think you do. And that is the Tychonic principle. The story I remember quite vividly from my wonderful seventh grade science teacher who really brought my incipient love of science to life by telling amazing stories.
Jim: And he told us the story of Tycho Brahe, to Kepler, to Newton. An amazing example of how an astounding amount of tiny data can yield gigantic results. Maybe you could tell us that story in a little bit more depth than how you came to adopt the concept that the Tychonic principle.
Michael: Sure. Yeah. So Tycho was a Danish astronomer who was living, just again like Bacon, right at the cusp of the scientific revolution. He is famed for the incredible accuracy of his observations of the movements of the planets in particular, and of the night sky in general. But my inspiration here is extreme attention to detail in recording the positions of the planets. Now this is just a little bit before Galileo, one of the very first people to actually point a telescope at the heavens, at the night sky. And the reason Tycho was not pointing a telescope was that he didn’t have one.
Michael: So this is right before the new technology came along. Tycho’s observations were all naked eye observations. Nevertheless, he cared so much about getting the details right, that he built for himself, with some funding from the King of Denmark, an underground observatory. Now that sounds like a contradiction in terms. It was an observatory that had a clear view of the night sky, but it was built into the ground because Tycho was worried that vibrations in any structure he built like a tower, would be creating small and accuracies in the instruments he was using to line up his observation.
Michael: So he’d have a little kind of instrument and he would record the point at which some particular planet, for example, crossed a little pointer he had set up in his observatory. Vibrations just from street noise and so on, or the wind would create slight discrepancies. And his solution to that was to dig into the earth and use the planet earth as a solid foundation for his observations. Interestingly, hundreds of years later, a physicist Michelson and Morley who were making measurements that turns out to be crucial and the Einsteinian revolution physics did much the same thing.
Michael: They were doing their experiments as deep down in the basement of their building as they could to get away from these vibrations. Anyway, Tycho was one of the very first observers. So different from Aristotle again. Consumed with the thought of getting these numbers exactly right. And it was expensive, but also devoting so much time to these microscopic details that Aristotle would have thought were a complete waste of time to record.
Jim: Yep. And then this data then enabled sort of the next steps. Maybe talk a little bit about how Kepler used the data and then how Newton grew from that.
Michael: That’s right. So Kepler was Tycho’s assistant for a while and was able to use his data. Kepler was one of the pioneers of the idea that the planets are orbiting the earth. Something actually that Tycho was not convinced off. He was making all these measurements, but he, in his own thinking, he thought that other planets all went around the sun and then the sun orbited the earth. Sounds like a complicated and potentially dangerous arrangement. Anyway, but Kepler had the earth and the other planets orbiting the sun, just as Copernicus said decades before.
Michael: But one of Kepler’s great contributions was to be mathematically exact about these orbits and to figure out that they were not exact circles, but they were just ever so slightly elliptical. And that was possible because he had such great data from Tycho. So this is an example of where the very small numbers begin to make rather a big difference because these same ellipses that Kepler was computing and the same equations he wrote down, laying out the speed at which the planets went around the ellipses, which is a speed that is not exactly constant, that gets a little bit faster is that when the planets are a little bit close to the sun, those equations were then of enormous importance when Newton formulated his theory of gravity, which explained why all of these Keplerion laws of motion were correct.
Michael: So you have there, running from Tycho to Newton. First of all, a story that really illustrates the importance of the little details in finding out the big ideas that really matter. So my Tychonic principle, is simply this principle which I’ve been emphasizing actually the whole time we’ve been talking, that it’s the little details in the ends that drive scientific progress and not so much big philosophical ideas.
Jim: And of course, modern science, ratchets that’s games up to quite an extreme think about things like the CERN big physics experiment where we spend billions of dollars, build 50 kilometer tunnels, gigantic magnets to produce vast amounts of amazingly obscured data to try to prove one thing.
Michael: That’s right. The thing about these little details is that, they not only look unimportant and I’m sure and Aristotle would have thought they’re very unimportant and for that matter, maybe Bacon too, but they’re tremendously expensive to obtain. So you have to be really driven by the thought that they’re going to make a difference to make those measurements, build those structures.
Michael: And of course persuade all of the rest of us to pay for those structures. And in the case of something really expensive, like CERN or the LIGO observatories that detected gravitational waves.
Jim: Yep. And actually I was going to talk about this later, but let’s hop into it now. In some sense, it’s kind of curious that we can get people to commit their lives to the difficult painstaking accumulation of this dry data. In fact, I’m going to tell a personal story too, I’ve never actually talked much about before. I was actually a want to be physicist when I was an undergraduate at MIT back in the ’70s. And I discovered I was certainly good enough to have been an experimental physicist, but my math intuitions and skills weren’t quite good enough to be a theorist.
Jim: And I concluded that I personally didn’t have the personality type to, as I described it to myself, spend my adult life in a white coat in a damp basement. And I opted out of physics at that time. So I actually did not make the transition that obviously lots of other people do. And you talk a fair amount about that, about the acculturation that has to go on to get smart, ambitious young folks, especially in the experimental domains, to spend their life in excrutiatingly detailed experiments.
Michael: Yes. So this, you might think of modern science as being built on two achievements at the metal level. One is a more of an intellectual achievement, which is simply the recognition that these details, the way the truth that if we could only have this information, then we could make real progress on figuring out how the universe works. So as I say, that’s an intellectual realization and we can all say yes, it would be so great to have this stuff at our fingertips but the other side of it is actually getting people to go out and do it.
Michael: And there, I think science is great innovation is more a piece of social engineering than an intellectual hypothesis or preset. The kind of engineering, the kind of structure that gets scientists who after all are most often drawn to scientists by the excitement of dealing with big ideas, to get them to actually go out and day after day, week after week, turn up and deal with all of the frustrations of doing science. And this is where again, I’ll refer to, if you like, the gamification of science, the iron rules creation of a system where however much you love the big ideas, you’re not allowed to in fact wrestle with, you’re not allowed to argue in terms of the big ideas.
Michael: All argument has to be done by going and getting more information, making more measurements, every move in the game has to be an empirical move, another measurement, another observation. So however frustrating it may be, if you want to be a scientist, if you want to play the game, then you need to commit to doing this stuff day in, day out. And there I see the iron rule with it’s, what I take to be in a certain sense an irrationally narrow minded focus on observation that alone has created the kind of social structure in which relatively normal people can actually be productive scientists.
Jim: Yeah, very interesting. This is the one part of the book I’ll push back on a little bit, I’d love to get your response. Which is there are indeed theorists in areas of science where there’s lots of theorists who work on theory. But in fact, at our Santa Fe Institute, we pretty much focus on theory. We have no labs, we do no experiments other than computational ones. However, we have built a network where the theorists are in touch with the experimentalists and the data collectors.
Jim: And so that it’s more or less mandate this, frankly, some of the theorists would just like close the door and do theory all day. But our role is, in science governance, we make sure that they don’t, right? And that they essentially have a cyclical program between theory and experiment, but those are different people and they collaborate, but they’re not necessarily aligned as permanent teams. So my little pushback on the model is that there are pockets and nodes of pretty much pure theory, but for them to be effective, they have to be linked into a cycle with this more Tychonic and Baconic style of science.
Michael: Well, I didn’t take that to be really a disagreement with what I’m saying at all. The key is to look at, not at what those theorists are doing most of the time when their doors are closed, but what they do when they go out in public. And one thing they do, of course, when I say in public, I mean in public in science. So when they’re publishing their ideas in journals and so on. Now, one thing they do is of course they simply publish their ideas, but it’s critical. And I think here, the iron rule is playing that same role that you were playing at Santa Fe.
Michael: It’s critical that when they do so, they point to the ways, and in fact, they officially only care about the ways in which their ideas make a difference to what will be observed. So this is a sense of much Poppers’ idea that theories must be falsifiable is captures something very right about science. That theorists can do whatever they like in the privacy of their own office, but if they’re going to be a part of the argument, a part of the game, then there’s only one way to engage, which is to have your theory make some new prediction, or explain some phenomenon that’s well-known that nobody can explain.
Michael: It’s always got to be that contact with the observed facts that brings a theory into the conversation, that makes it part of the conversation and that gives it any kind of purchase in public scientific arguments. So the theorist may be sitting in their office saying, and I mean, theorists have this disposition, which is why we need the game, why we need the rule. People have this disposition, you’re sitting in your office thinking this is awfully wonderful.
Michael: It’s awfully beautiful. As Eddington thought about Einstein’s theory, it can’t be wrong. But the iron rule, the system of science is saying, no, one’s going to listen to you unless you open the door and go out and make those predictions.
Jim: Yeah, very good. That’s well said. I think you hit it right on the nail. You resolved the slight ambiguity for me. Next thing we want to go on to is the importance of Newton in the evolution of the science that we have. You make a pretty big deal out of this. And if you could tell us a little bit about Newton and we know that he wasn’t just a scientist. He was at least as much a religious fanatic and an Alchemist, but nonetheless, he was somehow able to be a pretty pure modern scientist in the part of his time which he dedicated to science,
Michael: Newton I think yes, in a way he was the first modern scientist. Now that will sound a little bit strange to people who know a little bit about the history of science in the 1600s, especially in Britain, there were other scientists like Robert Boyle, the chemist for example, who said things that were very much in the spirit of the character of the iron rule. We must focus on the evidence in the end only observation that is someone like Boyle’s very much continuing the Baconian tradition. But they didn’t necessarily actually carry through on what they said. I mean, they did some wonderful science.
Michael: In that time in the 16OOs, mid 1600s, let’s say, it was very unclear that anything like the iron rule could really step in and govern science with its iron grip. There was still a lot of philosophizing going on, a lot of philosophical arguments about atomism and so on. The thing that was really remarkable about Newton is it seems not because he was trying to invent a new form of inquiry or anything like that, but when he focused on his physics, he focused just on his physics. As you say, he was very interested in alchemy and did a huge amount of alchemical experimentation.
Michael: He had his own lab outside his offices in Cambridge, and he may have in fact have poisoned himself by experimenting with mercury. He was rather interested in philosophizing and wrote various diatribes against Descartes who never met because Newton is born a little bit later. This is really now in the latest 16oos that Newton is working. He was interested in scriptural interpretation. He was interested in predicting the end of the world. He was an amazing intellect and a very capacious and sprawling intellect.
Michael: Yet when he focused on doing his physics that is developing this theory of gravity in particular, that would explain the motion, the observed motions of the planets of falling object on earth, he, for some reason, which I think we will never fully understand, he focused purely on a mathematical theory that would get the details right. So it was as though the iron rule was standing over him saying, only observation matters, only experiment matters. Only getting the numbers right matters. Exactly what was going on in his head, I can’t really say. But he did in fact proceed that way.
Michael: And in proceeding that way, and here’s I think what really mattered because one idiosyncratic person is not going to change the world, in proceeding that way, he created this new theory, the Newtonian theory of gravity that was so accurate, and so marvelous that the rest of the scientific world couldn’t not take notice, not only of the theory, but of the way in which it was produced. And in the wake of Newton, who of all of these people, all of these would be scientists is thinking to themselves, maybe what I need to do is to think the way Newton thought.
Michael: Maybe what really matters here is that stop worrying about the philosophical foundations of my physics. And Newton was very explicit when he said, I don’t care about any of that. What’s important is that my theory makes the right predictions. All of these successes to Newton said, well, maybe I should stop worrying about the philosophy. I just want a theory that gives the right answers. And there, I think you have the birth of the iron rule as people start to become self-conscious about the incredible fruitfulness of this way of doing science.
Jim: Interesting. And perhaps if there hadn’t been a Newton, it might have taken quite a while for that phase change to occur.
Michael: I think that this is quite possibly true. It’s hard to know. Newton came along pretty quickly. After all we have the whole Copernican revolution, this new understanding of the organization of the solar system that you see in Kepler and Galileo around in the early 1600s. And then in the late 16OOs, about 50 years later, you have Newton figuring out his ideas. That’s not very much time. You might think, aah it was inevitable that somebody like that would come along sooner rather than later, but we only get to run history once. So we don’t really know.
Jim: It’s always the question, the big trends of history or the great men and women now of history? My usual answer to such dichotomous questions is no doubt some of both.
Michael: Right. I think that’s right. So on the one hand you need the person who just out of whatever individual weirdness it is, or in this case kind of weird narrowness decides, however, I’m so interested as Newton who’s so interested in the philosophy of space and time and has written all this stuff about it, which by the way, he never published. And yet somehow when he sits down to figure out how gravity works, just puts that all aside, ignores it completely. On the one hand you need a character who does that, and does it extremely well.
Michael: And on the other hand, you need the right kind of social milieu and to which people, seeing someone exercising that kind of narrowness, they can say, well, actually that seems like something that’s worth trying out. And so that individual germ actually spreads and you get a whole social institution that attempting to recapitulate Newtons’ success. So yes, some of both.
Jim: Yeah. And as we say, we’ll never know unless we get the time machine to run the experiment again, but it’s interesting to contemplate how important an individual might’ve been. Next topic. Despite the fact that we, in the real world, we do tend to adhere to the iron rule, even though I don’t know how well it’s been articulated in the past in the Baconic convergence and the Tychonic principle, there’s always counter tensions against that. Then you have an interesting chapter late in the book called, The War Against Beauty.
Jim: And unfortunately we’re getting late in time and I do want to get to why Western Europe? So maybe we can get a brief story about the continued emergence of beauty as an alternative argument on how to do science and maybe talk specifically about string theory where there’s a lot of people say, string theory isn’t even science goddamn it. And then the string theorists themselves say, well, it’s got to be true because it’s so beautiful. So the War Against Beauty.
Michael: Yeah, this is a great illustration of why we still need the iron rule, why we need not only the intellectual realization that in the end data matters, but we need a kind of a social structure to implement that thought in the day-to-day work of scientists. So many theoretical physicists are highly motivated by the idea that beauty is a guide to truth. That the final ultimate theory of why everything is the way it is going to be a powerfully deeply beautiful theory. Nevertheless, they’re forced by the iron rule to however much they’re personally motivated by beauty as Eddington was motivated by the beauty of Einstein’s theory, they’re forced to, as I was saying earlier, to bring their theory into contact with the world by making predictions, by having their theory say, predict certain of these little Tychonic details, if you like.
Michael: So the iron rule says you can’t just evaluate a theory, not just that you can’t just evaluate a theory based on beauty, but in the arena of science, in the scientific journals, you cannot argue for a theory based on its beauty. It has to be in terms of experiments and observation. And this is why it’s essentially, I think, been responsible for the last 300, 400 years of scientific progress that instead of going off and yelling at one another about how beautiful or philosophically coherent our scientific theories are, even when it seems like that’s such a great guide to truth, we’ve been forced to go out and do more experiments. That’s been so important. But now with string theory, we’ve run into a bit of an obstacle with that.
Michael: Some of the experiments that seem like they might actually test string theory are impossibly expensive or large scale to conduct. I mean, not just in the sense that no one would ever agree to pay for them, but in the sense that there simply aren’t enough economic resources in the world and never will be to build the kinds of vast structures that would be needed to perform these experiments. Well, some physicists have thought, well maybe we should revise the iron rule. Maybe it should be possible to argue for string theory on the grounds of its beauty and not because it makes certain predictions.
Michael: So in other words, we should renege on the kind of structure that has guided science through the last several centuries and try doing science a new way where we’ll start the journal of aesthetic physics. And we won’t worry about what string theory predicts because it’s too hard to test. Well simply try to develop more and more beautiful theories and argue with one another based on the perceived elegance of those theories. That’s what some people have called post empirical physics.
Jim: And your take on that is that would be a dangerous move.
Michael: I think it would be. So in so far is that you only ever applied it to theories that couldn’t be tested in any other way then, well, it probably to be honest, it probably doesn’t matter. Sure. Let people argue about the merits of different versions of string theory or alternatives to string theory and the journal of aesthetic physics. But I somehow doubt that it would be confined to that. Once it became possible to argue for theories on the grounds of their beauty, and in fact, once it became perhaps even kind of prestigious, if this were being done in an area of physics that is extremely glamorous, like string theory, then I think it would start to spread and people would start to think, well, okay, if the physicists are doing it or the fundamentalists are doing it, why not us?
Michael: Well, we’ll still go keep measuring and so on, but maybe instead of spending so much time at the lab this weekend, I’ll spend a little bit more time arguing eloquently for the mathematical inevitability of my hypotheses. And so this idea would start to eat away at that motivation to just throw everything into observation. That’s been so important.
Jim: Interesting. Yeah. And they talked about prestige and also resources and last time I really looked at it carefully was in the double odds. But back then the majority of new physics PhDs were going into string theory. And so if indeed string theory is a long journey to nowhere, as it might be, that’s in a gigantic waste of brainpower and no doubt all the salaries, and everything else, and the opportunity costs of other things those folks could be working on.
Michael: Yeah. I think this is a great illustration of the continuing glamour of big thinking. And I’m a philosopher, so of course I’m not opposed to big thinking. And in this book I’m thinking big about science. So I don’t mean to disparage it, but it has an extremely powerful draw. If we don’t keep scientists noses to the empirical grindstone, then they will very naturally for reasons I understand as well as anyone, be drawn to this kind of intellectual inquiry, this very broad based, very attractive kind of intellectual inquiry. We need to impose on them that what I think of as the irrationally narrow constraint of the iron rule, just so that they’ll have no choice other than to push harder and harder at the observational end of things however exciting and beautiful life looks at the other theoretical end.
Jim: All right, well, let’s move on from that one and we can go into it and I can tell some stories about people I know in the field, but we won’t because we’re running short on time. So the last topic we’ll talk about is why Western Europe in the 17th century? It’s a long damn time from Aristotle to say, Newton, what was going on in Europe that made that the place that it happened?
Michael: Yeah. Great question. So we talked a little bit before about whether it was just kind of good luck to that and Newton would come along right when he was needed, or whether there was a certain systematicity to it. Now you might think, I have to warn everyone that this is now historical speculation on a vast scale. You might think maybe there have been Newtons in the past. They’ve come along. They’ve been very concerned to simply to explain the quantitative details. They’ve spurns philosophy, they’ve even spurned this kind of aesthetic thinking, this idea that theories must be beautiful, but there’s been no uptake. People have looked at it and said, well, that doesn’t make much sense.
Michael: Maybe our records from antiquity from the time of ancient Greece are so sparse. Aristotle might well have had some colleagues, some rival working alongside and doing exactly this. And Aristotle looked over and said, well, that’s very interesting, but to simply say that the philosophical foundation of my physics doesn’t matter is absurd, it’s irrational. It makes no sense. And so it never caught on. Why would Newton’s ideas catch on? I think the idea has something to do with the political climate in Europe at that time in the wake of all of the Wars that had happened as a result of the Protestant Reformation. So there had been these terrible conflicts that by most estimates killed a higher percentage of the population in some parts of Europe than any war since over religion.
Michael: And then the resolution of those conflicts was a kind of uneasy truce in many different ways. People often had to agree that there would be certain spheres of obligation of duty that would be kept strictly apart. So for example a Protestant living in a Catholic principality would say, okay, if we’re going to have peace, then I need to be able to have my Protestant religion, which of course matters more to me than life itself since everlasting salvation is at stake, but I also, the Catholic Prince is going to insist that I obey the laws. So we need to divide up the rules that I’m going to follow into these two different areas.
Michael: On the one hand with civic matters, with taxation, and various other issues connected to the smooth running of the state, you’re going to follow one set of rules. And then when it comes to religious worship, I’m going to follow a completely different set of rules. Now, of course, everyone kind of knows that many reasons for these Wars is that religion is not entirely separate from politics. Religion has certain concepts, certain kinds of religion have consequences for politics. But what we’ll do is we’ll create this artificial separation simply so that we have peace.
Michael: Okay, we’ll all agree to act as though religion has no consequences for politics and politics have no consequences for religion, so that we’ll be able to live our lives in a way that does not lead to a rekindling of these Wars. So you have a kind of an agreement, to be narrow, and both of these fears to ignore the interaction between these fears, to pretend that doesn’t really exist in order to get something that is very valuable, peace. And this narrowness and separation of the spheres is the subject of a lot of philosophical writing. So it gets turned into really our conception of the modern liberal state, which is held up as of course a wonderful thing, and the only way politically to go forward.
Michael: Well, it’s exactly the same kind of artificial or narrow separation. Which the iron rule is imposing on inquiry into the structure of nature. It’s saying, nevermind the fact that you may believe that philosophy or beauty is an important consideration in theorizing. When you do science, impose upon yourself this constraint, this narrow constraint, this artificial constraint, according to which all arguments must be conducted only in terms, observation of the consequences of theories for what’s observed. A structure like that would have seemed so irrational and artificially narrow to Aristotle. Might’ve seemed equally artificially narrow to somebody living in the late 1600s in the wake of all of these Wars.
Michael: But that artificial narrowness now looks distinctively modern. It looks like the way forward in politics. So why not in science as well? My, really it’s just a speculation, but my speculation is that the political climate made a certain artificial even unreasonable or irrational narrowness seem viable and even kind of fashionable. And so really helped with the spread of Newton’s strange practice as the natural philosophers after him took on that way of proceeding and became modern scientists .
Jim: And very interesting. Well, that’s been a wonderful conversation Michael. Well this has been great. Again, I recommend the book to people that are interested in this topic. It’s actually a fun read, believe it or not, The Knowledge Machine, How Irrationality Created Modern Science by Michael Strevens. Thanks again Michael.
Michael: Well, thanks so much Jim. It’s really been great to talk about some of these ideas.
Jim: It really has been. I enjoyed the heck out of it.
Production services and audio editing by Jared Janes Consulting, Music by Tom Muller at modernspacemusic.com.