Episode 116: The Knowledge Machine

• Links to this episode: Spotify / Apple Podcasts
• This transcript was generated with AI using PodcastTranscriptor.
• Unofficial AI-generated transcripts. These may contain mistakes. Please check against the actual podcast.
• Speakers are denoted as color names.

Transcript

[00:00:00]  Blue: Hello out there. This week, on the Theory of Anything podcast, we’ve got another deep dive into epistemology. Bruce goes headfirst into Michael Strevin’s book, The Knowledge Machine, which is an attempt to describe how science is a self -correcting system designed to create knowledge based on explanation. The book is somewhat critical of Papyrian falsification, though the reading of Popper presented may be a superficial reading. Bruce describes how Strevin’s iron rule of science, or his idea that we should settle science based on empirical tests, overlaps with what Bruce calls Popper’s ratchet, or the idea that we should strive to move our theories in an empirical direction and avoid the dreaded ad hoc saves designed to make our theories less testable. I enjoyed listening to Bruce as I always do, and I hope others do too.

[00:01:11]  Red: Welcome to the Theory of Anything podcast. Hey, Peter. Hey, Bruce. How are you doing today? Good. We are going to talk today about Michael Strevin’s book, The Knowledge Machine, How Irrationality Created Modern Science. And I’m going to be kind of asking the question, which came first, science or a culture of criticism, at least according to this book? So this is

[00:01:41]  Blue: kind of a Coon versus Popper thing?

[00:01:45]  Red: No. Not bad. You know what? No, no, that’s a good question, though. So Strevin’s, I know you’re not finished reading the book. You’re like halfway through or something in the book?

[00:01:59]  Unknown: Yes.

[00:02:00]  Red: So he does, his point of view could be summarized as that he disagrees with both Coon and Popper.

[00:02:11]  Blue: Okay.

[00:02:11]  Red: But he believes something that takes some of both.

[00:02:15]  Blue: So he’s using their debate as a starting point and then going his own direction?

[00:02:18]  Red: Yes.

[00:02:20]  Blue: Okay.

[00:02:21]  Red: So let me just say that this book is not likely to be popular with crit rats, which is too bad because I do think there’s some really important ideas in this book that crit rats could learn, right? On the other hand, I get why crit rats would just really hate this book. He is constantly critical of Popper’s epistemology throughout. He also claims to have his own replacement of epistemology that he claims improves upon and thus falsifies Popper’s epistemology. He argues throughout, at least in the language of induction, Beaconian induction no less. I can’t find anywhere where he actually says induction is true. So it’s maybe a stretch to call him an inductivist, but in terms of the way he situates things, in terms of his lingo, he comes across as an inductivist. Let me just say that upfront, okay? He does go over the problem of induction and the problems that and how Popper solved it. And he doesn’t necessarily seem to directly disagree with any of that. So he lives in some weird world where it’s a little unclear. I’m going to call him an inductivist and that’s maybe unfair, but I just feel like his language is so inductive that calling him an inductivist probably makes sense, at least as he’s a semi -inductivist or a pseudo -inductivist or something like that. And he’s very pro something called plausibility rankings. Now, what are plausibility rankings? Is that maybe that sounds like Bayesian probability rankings. And in fact, even though he never brings up Bayesian probability rankings in the main text of the book, in the footnotes, he explicitly talks about Bayesian probability rankings as being one possible way to understand his concept of plausibility rankings. So it does have a tie to Bayesianism.

[00:04:31]  Red: OK, and we’re going to talk about that as well. Also, it should probably be noted, and this might surprise a lot of critical rationalists, particularly the Deutsche ones. But if you were to read, I’m currently reading E.T. Jane’s book, Probability Theory and the Logic of Science, which is one of the Bible of Bayesian reasoning, if you will, one of them anyhow. And he calls Bayesian reasoning a plausibility calculus instead of a probability calculus. So Bayesians have no problem with thinking of their ideas as a plausibility calculus. The reason why I say that might surprise like Deutsche and Crit Ratz is, Deutsche will go out of his way to cleanse probability out of his language. And then because he still needs conceptually exactly what probability means, he’ll refer to a theory as more or less plausible. And he’s trying to use the word plausible to remove the word probable. But those words mean the same thing. And to Bayesians, they mean the same thing. So there’s a lot of interesting things here that aren’t the easiest to deal with. And since one of the things I’m trying to do is I’m trying to make sense of Bayesian reasoning right now and how I think it fits with critical rationalism or doesn’t fit with critical rationalism. Reading Strevin’s book made some sense to me. And I felt like even though I cringed at certain points, I felt like it was a really interesting book that made some really interesting points that really critical rationalism should take seriously. So that’s why we’re going to probably in depth cover a lot of the ideas in this book. Both in terms of our agreements and disagreements with my agreements and disagreements with it.

[00:06:24]  Blue: One question. Can you can you review for me what the difference between regular induction and Baconian induction is? Something I think I knew, but I forgot or something. Okay.

[00:06:35]  Red: So those are my terms. So it’s not like you could have read about this somewhere else. It’s my way of trying to express myself. Okay. So my argument is this. Okay. So think back to our episode where we talked about Popper’s refutation of induction. And then we talked about Tom Mitchell’s refutation of the futility of unbiased learning. And one of the things I showed is that they were the same refutation. Okay. But Popper saw that as refuting induction. And Mitchell saw that as teaching us something about induction. Okay. Because machine learning is quote unquote induction. And my point was this, that we shouldn’t get too caught up in the word induction. The word has a long intellectual history. It’s been used all over the place to mean different things. It should not surprise us in the slightest that machine learning is referred to as induction. Okay. Does that then disprove Popper who said induction is a myth? No. Because Popper is using the term induction differently than the way Tom Mitchell is using it. And in fact, if you understand what they’re both saying, they are not disagreeing with each other. Tom Mitchell is in essence saying there is no form of induction that works unless you start with a set of theory as to what your solution needs to look like. You have to have a bias toward an inductive bias. What’s what it’s actually called in machine learning theory. You have to have some sort of inductive bias, which is really a theory about what your solution needs to look like. And so this is not something that’s incompatible with Popper’s views, right? In fact, I would argue that it completely corroborates Popper’s views.

[00:08:24]  Red: And yet, Mitchell would call that induction where Popper wouldn’t. So when I say Baconian induction, what I really mean is the kind of induction that Popper was against. But one of the things we’re going to find is that it’s not even obvious that Bacon was in favor of Baconian, what I’m calling a Baconian induction. And for historical reasons, we ascribe it to Bacon. But like Bacon’s overall theory was closer to Popper’s epistemology than Popper seems to have realized. This is something that I got actually off of the Popperian podcast. One of the students of Popper said that he felt that Popper, as most people at the time did, misunderstood Bacon. And so these are all complicated things, right? Like, we kind of draw these army lines between I’m against induction and I’m, you know, you’re in favor of it. And really, that’s not what we should be doing. We should be digging down to the concepts. We should be trying to understand conceptually what is true and what isn’t. And then what we actually call it doesn’t matter as much. So that’s what I’m really trying to say here is that when we talk about the induction of Bacon, typically, even though this may not ultimately turn out to be true, Bacon is accredited for this idea that science used induction. Induction was this idea that you could start with observations and that you could generalize from those observations into general scientific theories, right? That’s what I kind of mean by Baconian induction. It’s the kind of induction that Popper said, no, that’s just a myth.

[00:10:03]  Blue: Okay. Well, I’m glad you’re doing that. And someone’s got to, right?

[00:10:11]  Unknown: Yeah.

[00:10:11]  Red: And how did you figure, find out about this book?

[00:10:14]  Blue: One more question there.

[00:10:16]  Red: You know, I don’t remember. You know how I just read a lot of books, right?

[00:10:20]  Blue: Yeah.

[00:10:21]  Red: So if for anyone who wonders what happens to the donations to this podcast, the 100 % of the donations to the podcast get spent on something related to the podcast, usually me buying a book, which I then do a podcast on.

[00:10:39]  Unknown: Okay.

[00:10:41]  Red: So I think what happened was, is I just saw the book and it sounded interesting. And so I bought the book, I listened to it. And I thought, you know what, that book actually, even though it was very inductivist, it genuinely clarified some critical rationalist ideas that I hadn’t quite figured out. This isn’t an uncommon thing, right? You can genuinely learn from the other side. You can, right? The other side almost always has some truth to it.

[00:11:14]  Blue: And he’s got some really interesting ideas that at least deserve some discussion, right?

[00:11:19]  Red: Without truly endorsing his ideas, which I’m going to disagree with in a number of ways. So let me just say that if you are a crit rat that kind of sees critical rationalism in terms of what I’ve called the crit rat war on words, this book’s going to bother you because he uses all the wrong words across the board, right? And it’s going to just catch you up and you won’t be able to continue. However, if you’re more like me, if you’re a popper’s ratchet kind of critical rationalist, then yes, this book’s a little cringy at times, but it’s got a lot to like, too. And it has some very direct relevance, if his theory is true, which I’m not necessarily saying his theory is true. Like I definitely think some of it’s true, but like I think a lot of it’s not. But if his theory is true, it has direct relevance to the question of static societies versus dynamic societies, and it poses one of the strongest challenges and adjustments to Deutch’s theory of static societies. If it’s true, which it may not be. Okay. So it’s got direct relevance to epistemology in general. I feel like it has at least some interesting contributions to epistemology that I think at least in some cases may actually improve upon popper.

[00:12:39]  Blue: Okay.

[00:12:40]  Red: And I think it has something interesting to say that represents a challenge to some of Deutch’s points in his static society versus dynamic society theory. And therefore, it fits into our series on static versus dynamic societies also. So it serves several purposes at the same time. Let me also say that we’re going to probably spend more than one podcast on this, on this today’s episode. I’m going to summarize his main ideas and I’m going to try to situate them in such a way that hopefully I won’t, I’ll show that critical rationalists shouldn’t just run for the hills the moment they hear these theories that there’s some overlap with popper’s theories worth talking about. Okay. But also some things that maybe really do disagree with popper and etc. But then what we’re going to do in future episodes is we’re going to look in depth at his criticisms of popper. And then we’re going to ask, is he just misunderstanding popper or does he have a valid criticism? And I’m going to ask Peter to help me with that, right? Because I don’t actually think it’s an easy thing to work through. So Strevin’s book has several big ideas that I’ll summarize first just so that it’s easy for you to wrap your mind around his theory. Okay. So modern science, so this is his theory in a nutshell, is modern science’s power stems from its what he calls iron rule of explanation, which is that science is unreasonably, scare quotes, closed -minded about a code of argument that dictates all arguments must be carried out with reference to empirical evidence only.

[00:14:25]  Red: So Strevin sees this as irrational and even closed -minded because it refuses to consider many other sources of explanation and criticism that are non -empirical, such as philosophy, theology, and aesthetics. Okay. So page 96, this is his own words now. Here then in short is the iron rule. One, strive to settle all arguments by empirical testing. Two, to conduct an empirical test, to decide between a pair of hypotheses, perform an experiment or measurement. One of whose possible outcomes can be explained by one hypothesis and accompanying cohort of ideas, but not the other. Okay. So when I read that, the second part of that is just Popper’s concept of a crucial test. So this shouldn’t maybe sound too far off for critical rationalists. This is very close to Popper’s demarcation criteria and then the idea of a crucial test. So Strevin’s believes that the iron rule involves four methodological innovations. One is a shallow notion of explanatory power culturally agreed upon by all scientists that focus on causal derivations from observable phenomena, regardless of underlying philosophical or metaphysical nature, which Strevin’s believes science isn’t that interested in. Number two, a distinction between public scientific argument and private scientific reasoning. Public arguments follow the iron rule, private reasons do not need to. That’s actually the part of his theory that I find the most interesting and that I actually think is the most likely to be something that Popper didn’t say, but that I agree with. Three, a requirement for objectivity in scientific argument achieved through sterilization, which excises subjectivity as much as possible, even if it makes arguments incomplete or blunted. And four, a

[00:16:33]  Red: prohibition on non -empirical considerations, philosophical, religious, aesthetic in official scientific argument, insisting that only empirical testing counts, even if other considerations, philosophical or aesthetic, let’s say, went into the development of the theory. And then Strevin’s argues that this methodology leads to two things. The first is what he calls Baconian convergence, speak about Baconian inductive sounding language, a term bound to make crit rats see red. But it’s the idea related to, it’s the idea that observations, it’s related to the idea that observations eliminate competing theories until there’s only a few left. And the idea that, and this idea isn’t that far from what most crit rats claim they believe. The second is what Strevin’s calls the Tychonic principle, the idea that the above methodology creates an irrational drive, scare quotes on their words irrational, towards increasingly small observable empirical details, testing theories via eking out another few decimal points, so to speak. Okay, so does any of this sound familiar? Yes, hopefully if you’re a regular listener of this podcast, you might recognize that this at least has a passing similarity to how I’ve framed Popper’s epistemology in terms of non ad hoc theory exhaustion. Okay, which might be distinct from how other critical rationalists understand Popper’s epistemology. So that’s, that’s by the way is a coincidence. I had formulated my version before I ever came across Strevin’s book. And then when I read it, I suddenly realized, oh, he’s picked up on some of the same things I’ve picked up on. But I saw it as coming from Popper, and he saw it as not as being the as at odds with Popper, which I found interesting that we read Popper so differently.

[00:18:34]  Red: As we get into the details though, my views and Strevin’s, they will start to depart from each other. But I at least agree with him on the idea that what makes science work so well is that it concentrates on the kinds of theories that can be empirically tested and falsified, and that it caches out all arguments with actual empirical tests, because they are objective criticisms that anyone can check. Okay, that is my version of Popper’s epistemology in a nutshell. And that is, I think, what Strevin’s is trying to argue for. I believe that this epistemology, whether it’s Strevin’s or Popper’s up to you to decide, forces out or at least reduces subjective criticisms, which has the advantage of, at over time at least, removing subjective feelings about theories, and that the result is rapid progress because epistemological sinks are removed from consideration, because they have to be put in terms of actual non ad hoc empirical tests. One thing I think Strevin’s misses, and to me this is a big miss, is that he seems to have no real concept of the no ad hoc rule, or what I’ve called Popper’s ratchet, where we not only disallow non empirical theories, but we require all theories to have potential falsifiers distinct from the problem that we were trying to solve. And also that you can only save a theory if you increase the overall testability of the collective theories. He does mention this in passing, but he dismisses it, which I think is unfortunate because I think it’s a necessary part of making sense of his own theory.

[00:20:20]  Red: So this is a huge missed opportunity to make a necessary distinction between merely having any empirical test, including just the problem that you invented the theory to solve, which could then still be ad hoc, versus having an independent empirical test, i.e. a test that is unrelated to the problem that you’re trying to solve, that virtually forces your theory, if it passes and is corroborated by the test, that now the competing theories have a problem they have to solve. And it really does show that your theory is non ad hoc. So Strevin’s also seems to have no clue why his iron rule works so well, outside of a vague idea that it somehow forces Baconian convergence. And to him, it’s less a methodology, that’s kind of my wording, and it’s more a speech code. Other disagreements between me and Strevin’s is that I believe that this epistemology is Popper’s epistemology, whereas Strevin sees this as at odds with Popper’s epistemology. However, Strevin seems to know Popper relatively well, as amongst people that I’ve read that tried to quote Popper and then criticize him, I would rate him as understanding Popper better than most. That’s probably the most I can say at this point. We’ll have to get into the actual details at future podcasts. But I feel like he got Popper at least close enough in enough places that maybe he deserves some attention. Even when I feel he’s misreading Popper, it seems to me that it’s kind of a subtle misread. In fact, Strevin’s would almost assuredly argue with me that I’m the one misreading Popper instead, and that I’m reading into Popper something that wasn’t there.

[00:22:07]  Red: Which many crit rats would agree with him on, because many crit rats do not agree with me on my read of Popper. What makes this difficult is that honestly, Popper says a lot of questionable things, if you just try to take statements out of context. And I hinted at this in the previous podcast, where I showed quotes from Popper that aren’t really technically correct, and leave false impressions that he’s a naive falsificationist. Whereas if you read him more broadly, you can see that that wasn’t really what he meant. So we have a similar problem here. Any time I feel Strevin’s is misreading Popper, I can’t help but notice that it’s often around issues where Popper really does say things that really are kind of misleading. And so, or even he just misexplains them. And so I think that there’s often this kind of unfortunate choice of words that Popper uses that make me kind of go, I don’t think Strevin’s is reading Popper right, but I can think of a quote where Popper said something, where I can understand why Strevin thinks Popper thinks that. I don’t know if that makes sense or not, right? Where I know Popper well enough that I can say, I can look at Strevin’s, where he’s misreading Popper, and I can think, oh, but I know why he’s misreading Popper. It’s because Popper actually did say X, and it sounds exactly like what Strevin’s is saying. So we also have, Strevin’s and I also have a disagreement over if science is irrational or not. Strevin’s thinks it’s irrational because it gives up on other sources of criticism, such as philosophy, theology, and aesthetics, in favor of only in public anyhow, empirical testing results.

[00:23:47]  Red: Whereas I see that as precisely the rational thing to do for the reasons that I explained in the previous podcast episode. Plus, I’ve argued that Popper’s epistemology, which is really his scientific epistemology, is counterintuitive, which maybe isn’t that far a leap from what Strevin’s is calling irrational, okay? So Strevin’s argument is that science, with its iron rule of explanation, seems like a terrible idea, and argues that that is why science is so hard to discover, is because it forces you to give up what seems like a rational approach in favor of what seems like a less irrational approach. So Strevin’s argues that this is why it took nearly 2,000 years from philosophy to be invented for science to really be discovered. And by the way, he places science as really starting with Newton. I mean, it’s a little more complicated than that, but he places it more at starting with Newton rather than the scientific revolution. So I don’t know if that is a correct theory or not, okay? The theory here being that science is so counterintuitive, or maybe even if you take Strevin’s point of view irrational, that it’s difficult to discover because it seems like a terrible idea. I don’t know if that’s entirely true or not. I will at least admit that it seems very counterintuitive to me. But if he’s right about that, to know if he’s correct, I’d probably need more time to dig into his examples and criticize them better.

[00:25:23]  Red: But if he’s right, that is a deeply intriguing theory about why humans got stuck in static societies for so long, is because one of the ways you get out of static societies, according to this theory, is that you have to make an irrational sounding leap, okay? Where you give up on philosophy and aesthetics and other things that in some way really are decent theories and forms of criticism, and that you have to constrain yourself down to only empirical theories and empirical criticisms.

[00:25:56]  Blue: What was this distinction you made between the Newton and the Scientific Revolution?

[00:26:01]  Red: So they’ve got this idea of the Scientific Revolution, which is like a century or too long. It sort of ends with Newton. So you’ve got Scientific Revolution. So more like

[00:26:12]  Blue: Galileo and Copernicus.

[00:26:14]  Red: Yes, okay. So the end of the Scientific Revolution is Newton, and the beginning of the Enlightenment is also Newton. So Newton is like the transition character between Scientific Revolution and the Enlightenment. So he places modern science as starting with Newton. I’m going to read some of his quotes where he talks about it. Decide for yourself why I’m interpreting him right.

[00:26:40]  Blue: I see.

[00:26:40]  Red: But he sees the true invention of science, not with the beginning of the Scientific Revolution, but with the end of the Scientific Revolution with Newton.

[00:26:50]  Blue: I see. Okay.

[00:26:51]  Red: Okay. Well, I disagree with Strevin’s that science is irrational. I have to admit that Strevin’s point is better than it first appears when taken as a whole. He’s not really arguing that science does not use philosophy, theology, and aesthetics. He gives ample examples of all three of those, yes, including theology, helping drive science forward. Rather, he argues that non -empirical considerations are cleansed from public communication, i.e. in scientific communications, in scientific papers, in scientific journals. Okay. And that the final result is, as much as possible, purely objective empirical results, even though they made the subjective criticisms, may have played a role in developing these results. They just get cleansed out at the end. So really, it’s the cleansing that he argues is a quote -unquote strategically irrational. But ultimately, he argues that ultimately, the secret of science is rapid progress compared to philosophy, theology, and aesthetics. By the way, I thought that that was interesting. If you were to look at our episode 97 for Popper’s own thought on the lack of progress in music, one of the things that Strevin’s is trying to explain is why those other fields seem to just get bogged down. It’s difficult for them to make progress compared to science. And that’s an idea that Popper agreed with. Popper felt that I don’t have the quote handy, but Popper did make a quote that science was really the only progressive field like that. So this isn’t something that I think Popper and Strevin’s are necessarily disagreeing with, that these other fields just tend to get bogged down because they don’t have the iron rule of explanation, whereas science does.

[00:28:43]  Red: Now, it seems to me that the fact that the iron rule of explanation, as Strevin’s calls it, I would think that it would be, for Popper, it would just be called the no ad hoc rule. But I think the fact that the idea actually works, which Strevin’s agrees, it actually works, and that it’s wildly successful, I would think that’s identical to saying it’s rational, but I digress. So while I disagree with Strevin’s that science is irrational, I have to admit Strevin’s point is at least makes some sense to me. Okay. So we need to discuss Strevin’s claim that science uses shallow explanations first and how this relates to two different popular critical rationalist views. So this idea of a shallow notion of explanation, which he believes is part of the iron rule of explanation. So to Strevin’s a shallow notion of explanatory power is a fundamental innovation of modern science. The concept underpins the iron rule of explanation, which he sees as a strategic irrationality, as I mentioned, and he says channels immense energy into empirical observations and empirical experimentation. So Strevin’s believes that this convention leads to universal agreement on what counts as a good explanation. So unlike pre -modern natural philosophy, where explanations were subjective and varied with philosophical and metaphysical commitments, modern science, according to Strevin’s, establishes a notion of explanation agreed upon by all scientists, regardless of their backgrounds or their personal inclinations. And that notion is to focus on causal derivations from observable phenomena, i.e. empirical theories with empirical tests. The phenomena is explained simply by deriving it from a theory’s causal principles and auxiliary assumptions, or what Popper would call auxiliary theories, along with background conditions.

[00:30:48]  Red: The theory must be able to entail what is observed and not entail things not observed, i.e. must make both correct predictions and not predict false things. Note that this is consistent with Popper’s own epistemology, at least so far, and matches Popper’s true understanding of falsificationism, as we discussed in the previous podcast, as a logical idea, rather than an understanding of how to prove a theory false. So we discussed that was why we did the last podcast first before this one, was so that I had talked through Popper’s concept of falsification and shown how there are various false versions of it that have been falsified, but that the true understanding of it has not been falsified. To Strevin’s, there is a disregard for philosophical or metaphysical nature of things. These shallow explanations do not require scientists to grasp the ultimate nature of the causal principles or their deeper philosophical justifications. Quoting from page 140, he says, the modern scientific standard for explanation is as empirically demanding as it is philosophically lax. This leads to, according to Strevin’s, this leads to freedom for theoretical exploration. It allows the scientists to propose just about anything, it’s a quote from Strevin’s, as a causal principle, quote, however ideological abhorrent that’s also from Strevin’s, as long as it has definitive observable consequences. So Strevin’s argues this leads to a foundation for procedural consensus. By fixing a universal standard for what constitutes an explanation, this shallow notion enables a consensus on how to legitimately test hypotheses freeing scientists from endless debates about explanatory rules and directing their efforts towards observation and experimentation. So Strevin’s gives several examples of what he means by science having shallow explanations.

[00:32:54]  Red: So for example, in Newton’s time, there was a fierce debate about how gravity could act across empty space without direct contact, action at a distance, if you will. Philosophers like Descartes insisted that true explanations require understanding the direct pushes and shoves that caused phenomena. So Descartes was looking for a deep explanation for why things are the way they are. So Descartes’ view is commonly, by the way, held by Deutsche and Critrats today. They tend to emphasize explanation over prediction and have weaponized the idea of an explanationless explanation or explanation of science as a means of dismissing unwanted theories. I will discuss that a little bit in a second. So on page 138, Strevin says, to put the Cartesian view another way, to explain something means grasping its cause. And to grasp its cause means to comprehend the sequence of pushes or collisions that bring it about. Without that comprehension, there is no understanding or no real explanation. Or that’s what Descartes was arguing. Not what Strevin’s is arguing. So I hope Critrats will recognize how their view of explanationless science sounds very much like the Cartesian view that Strevin’s is here summarizing. So Strevin’s continues, page 138. Newton, by contrast, pioneers a philosophically far shallower notion of explanation on which a phenomena is explained simply by deriving it from the causal principles of gravitational theory that is from the mathematical principles laid out in the Principia, his work, the Principia. In other words, Newton argued that it didn’t matter what the explanation was for how or why gravity worked. Only you should only look at what you could predict with the laws of gravity. So Newton famously stated, this is a somewhat famous Newtonian quote. I’m quoting it from Strevin’s book on page 150, though.

[00:34:58]  Red: He says, Newton says, I have not as yet been able to deduce from phenomena the reasons for these properties of gravity, the fact that it attracts things. And I do not feign hypotheses. For whatever is not deduced from the phenomena must be called a hypothesis, and hypotheses, whether metaphysical or physical, or based on occult qualities, are mechanical, have no place in experimental philosophy. It is enough that gravity really exists and acts according to the laws that we have set forth and is sufficient to explain all the motions of the heavenly bodies and of our sea. Here is how Strevin summarizes this view. This is from page 146 -147 of his book. The natural philosophers such as Aristotle and Descartes cared about the consequences of their theory’s causal principles, i.e. what you can predict with the laws, of course. But they also cared a great deal about their nature, about what the principles were saying about the metaphysical foundations of the world’s causal structure. Modern science, by contrast, is oblivious to the principle’s nature. Let it be as opaque as actually at a distance was to Newton’s adversary, Leibniz, who criticized Newton on that ground, or as Aristotle’s pusque, which would translate to soul, is to us. It is immaterial. All that concerns the iron rule is what kinds of things the causal principles are capable of bringing about. That’s from page 146 -147. This answer is likely to bother Deutsche and Critrats because Deutsche explicitly attacks the idea of explanationless science, which Deutsche claims is, quote, just bad philosophy disguised as science on page 321 of beginning of infinity. And he also claims, also on the same page, that explanationless science inhibits progress.

[00:37:02]  Red: What is the difference then between a shallow explanation from Stravins and an explanationless science from Deutsche? Interestingly, Deutsche claims there is no such thing as a true explanationless theory. So from page 15 of beginning of infinity, Deutsche says, for there is no such thing as a purely predictive explanationless theory. One cannot make even the simplest prediction without invoking quite a sophisticated explanatory framework. So it seems to me that Deutsche’s explanation of science might be the same as Stravins’ shallow explanation. So my first impulse is to assume that Deutsche and Stravins are disagreeing and are in contradiction to each other. I think this is an important question that goes too far off the subject. So I’ll earmark it at this time and then do a shortened reader’s digest abbreviation of it. For now, let me just point this out. At least when it comes to explaining what gravity is, it was, in fact, explanationless science. Yet there is no way Deutsche or any Deutsche and Critrat will ever claim Newton’s theory of gravity is, quote, explanationless science, that, quote, undermine science or is, quote, bad philosophy masquerading as science. And yet there’s no doubt that Newton’s theory made no attempt to explain what gravity was. It was, at least as far as that explanation goes, explanationless science. Okay. Why is it okay to be explanationless science, quote, unquote, explanationless science when it comes to explaining gravity versus some of the examples that Deutsche uses in beginning of infinity, such as generic heritability of happiness or of animal pain and stress and whether animals actually feel things? I’m not sure I see a clear cut answer to this question that was ever explained well by David Deutsche.

[00:39:07]  Red: He never addresses the problem that one could easily argue that Newton and Newtonian gravity was also explanationless science. It’s unclear when you should or shouldn’t be able to invoke explanationless science as a valid criticism from just what Deutsche says in beginning of infinity. So it isn’t surprising that in practice, Deutsche and Critrats, and sometimes I would argue even Deutsche himself, happened to invoke explanationless science as a criticism of theories that they happened to already dislike, such as genetic influence on happiness or animal qualia, but never invoke it against theories they like, such as Newtonian gravity or several other famous Deutsche and Critrat theories that, honestly, I would think would probably qualify as explanationless science. I suppose this is Streben’s point and that’s why I’m bringing it up. If you have an overly strong commitment to deep explanations about the underlying nature of things, rather than simply trying to use the theory to make predictions, you open the door for your biases to take over and for you to just simply exclude completely legitimate theories. That is, by the way, the people who were against Newtonian gravity, like Leibniz and like others, like the Aristotleians at the time, that was their main criticism of Newtonian physics is, look, we, using Aristotle’s physics, we are explaining what gravity is. Newton’s not even trying to explain what Newton’s gravity is, or what gravity is, so it’s explanationless science. They invoked that same idea. They eventually lost that war and over time, people actually got to the point where they didn’t even consider the question, what is gravity and why does gravity exist? They didn’t even consider it to be a valid question anymore as Newtonian physics continued to show amazing progress compared to Aristotle’s physics.

[00:41:06]  Blue: So was Newton an instrumentalist?

[00:41:09]  Red: No, no, that’s a very good question. Okay. Give me a second to finish this passage, and then I’m going to answer that question. Okay. So, in fact, that is exactly the right question to ask, because there is a substantial misunderstanding of instrumentalism that exists within the crit -wrack community. And so many would think he was an instrumentalist. Okay, but he’s not. Note that, oh, that was what I was going to say next on my script. Note that this is not the same as instrumentalism. So, there is no claim here on Newton’s part that the best way to understand science is as a predictive instrument. The key here is that there is, and that’s what instrumentalism is. Instrumentalism is the idea that the best way to understand science is as an instrument and that it doesn’t actually explain reality. Okay, that’s the kind of instrumentalism that Popper was against and that Popper put down. Popper would admit, though, that science is instruments. And we’re going to dig into a second what Popper actually said about this. Okay. So, the idea that we might use science that might have shallow explanations initially, and that we might use it primarily based on the predictions that it made, I don’t know that Popper would disagree with that, right? But that’s not the same as the claim that that is the best way to understand a theory. Popper definitely understood every theory as a serious attempt to understand reality and to find the truth. Okay. And that’s what made him a non -instrumentalist. So, the key here is that there is no minimum level of explanation beyond even just a simple causal claim.

[00:42:53]  Red: The explanation can have significant gaps in the explanation, but that does not bar it from being considered a legitimate scientific theory. That’s really what Strevins is arguing when he talks about science having this notion of shallow explanations. That’s really what I think Strevins means. Now, to make this debate even more interesting, which we could maybe pose this now as a debate between Deutsch, who believes in that really we should avoid shallow explanation, and we could see Strevins is arguing for shallow explanation. Okay. And maybe it by based on what I’ve read so far, maybe that’s exactly what it comes across as. Here’s what makes this so interesting. Popper agreed with Strevins, not Deutsch, because Popper also agrees that science is about shallow explanations. From Open Society, page 32, let me just read what Popper actually says. You tell me if this is agreeing with Strevins or Deutsch. The methodological nominalist will never think that a question like what is energy or what is movement or what is atom, an atom, is an important question for physics, but he will attach importance to a question like, how can the energy of the sun be made useful? Or how does a planet move? Or under what condition does an atom radiate light? And to those philosophers who tell him that before having answered the question, what the what is question, he cannot hope to give exact answers to any of the how questions. He will reply, if at all, by pointing out that he much prefers the modest degree of exactness which he can achieve by his methods to the pretentious muddle which has been achieved by theirs. So note that Popper is saying exactly what Strevins is saying.

[00:44:53]  Red: Science cares only about the kinds of things, the causal principles, the laws, what they’re capable of, bringing about, not by what the underlying nature is. So who is right, Deutsch or Popper in this case? Okay, so this raises the question of how far we can take the idea of science preferring shallow explanations, as argued by Popper and Strevins, and here I’m argued against by Deutsch, I guess, depending on how you interpret Deutsch. Can we, for example, claim that Deutsch’s view of explanation of science is falsified by this notion of shallow explanations? I actually think it’s more complicated than that. Let me lay down my own slightly different view here. As a general heuristic, I can agree with Popper that what is energy as a question is a less productive question than how can we use energy. But the idea that science doesn’t care about the question, what is energy, that is false. And Popper is wrong about that. Science today can give a pretty dang good answer to the question, what is energy? Okay, likewise, no one doubts today that Newton was incorrect that it quote unquote didn’t matter what caused gravity. For general relativity answers the question, what is gravity? And treats it as a valid scientific question. So now in the past, I’ve asked on this podcast, and in the context of Deutsch’s constructor theory of knowledge, isn’t he trying to answer the question, what is knowledge? And then I asked, does this mean Popper was wrong? Or does it mean that Deutsch’s theory is wrong? Or are they both wrong? I was trying to get people to think about this. Okay, because I’ve given a lot of thought myself on this subject. So here’s my view that I’ve come up with.

[00:46:46]  Red: And I think it requires some careful nuance. I think Popper’s point is fair if framed like this, we do not need to start with what is energy as Descartes thought, and as the Aristotelians thought, okay. In fact, how can we use energy was a much better question precisely because it allowed us to start error correcting right away, creating rapid progress, whereas the question, what is energy was at least initially an epistemological sink. And you’re done. You just cannot make progress on that question initially. Eventually working on the question, how can we use energy deepened our explanatory theories about energy until the question, what is energy became a productive question because we understood energy so much better at that point. By the same token, Aristotle tried to start with what is gravity and ended up with an epistemological sink. When I say epistemological sink, what I mean is that it’s just something that you cannot make any progress on. You can think you’re making progress. You can bring up philosophical theories. You can criticize them. You’re just stuck because it’s the wrong place to start. So he ended up with an epistemological sink that couldn’t make progress, whereas Newton simply asked, how can we use energy? And that jump started rapid progress. But eventually that led to so much progress that Einstein came along and he was able to answer, what is gravity as well? Okay, my feeling is is that Popper goes too far the way he words it. It just isn’t the case that science always asks how questions and never asks what questions. There are simply too many counter examples to Popper’s claim for it to be strictly true. But I think Deutsche’s view is also a bit problematic.

[00:48:38]  Red: As I mentioned back in particularly episodes 53, 54, 55, and then again in episode 102, where we talked about Brett Hall’s theory of intelligence, my experience with the Crittrack community was that they tend to invoke explanation of science as a criticism of theories they do not like while entirely ignoring it in equally viable ways that it could have been used against their own pet theories and that their own pet theories might also qualify as explanation of science. So it does not seem to me to create that creating a minimum standard of explanation is a viable criteria for a good explanation. It’s more of a recipe for arbitrariness. What science requires in practice is just enough explanation to start no ad hoc empirical testing. For Newtonian gravity, that didn’t even require an explanation of what gravity actually was. The moment he had a theory of how it worked, that was all we needed to get started. This is what Streven’s I think really means or at least should have meant by science being about shallow explanations. However, I have no problem with Deutsche’s view if it’s understood in a slightly different way. I would understand it like this maybe. Don’t be satisfied with shallow explanations. Always be ready to deepen and corroborate the deeper version of explanations via severe testing. Or maybe we could even formulate Deutsche’s view in a stronger formulation like this. If you want to increase the empirical content of an explanation which you should want, the way to do that is to deepen the explanation and to not let it be a mere shallow explanation. But if we formulate Deutsche’s view in that way, I think he’s totally right.

[00:50:27]  Red: But I think Deutsche goes too far when he claims that explanationless science, which by his own admission doesn’t really exist, is counterproductive. The moment you have a non ad hoc explanation, no matter how shallow, you have enough to get started. With the caveat being that you must continually follow the no ad hoc rule, testability is meaningless outside the context of the no ad hoc rule.

[00:50:52]  Blue: Can we back up just once that quick question? Was Popper saying that science doesn’t care about explanations or more that just logical positivists that he was reacting against don’t? I mean, it seems like you can read him as more saying that science does and should care about explanations.

[00:51:19]  Red: Let me re -read the quote, because he’s trying to compare a methodological nominalist. I forget what the other term was, but it was a non nominalist, an essentialist, a methodological essentialist, maybe. He says the methodological nominalist, which that’s what Popper is. He’s talking about himself and he’s talking about how science really works, because this is his opinion of how science really works. Science is methodological nominalism. He says the methodological nominalist, so a true scientist or a Popperian, will never ask a question like what is energy or what is movement or what is an atom is an important question for physics. He’ll never think that a question like, but he will attach importance to a question like how can the energy of the sun be made useful or how does a planet move or under what condition does an atom radiate light and to those other philosophers than the non nominalists who tell him, the nominalist, that because having answered the what is question, that without having answered the what is question, he cannot hope to give exact answers to any of the how to questions. He, the nominalist, the scientist, the Popperian, will reply, if at all, because he finds it a kind of a dumb question, by pointing out that he much prefers, the nominalist much prefers the modest degree of exactness that he, the nominalist, can achieve by his nominalist methods to the pretentious model which they, the non nominalists, have achieved by theirs.

[00:53:07]  Blue: And the non nominalists are the positivists or those.

[00:53:11]  Red: No, it’s no. No, no. Positives are a totally different school of philosophy that wasn’t had anything to do with essentialism.

[00:53:21]  Blue: Okay.

[00:53:22]  Red: So it, so Popper really is responding to the same, it’s the Aristotleians that he’s responding to here, right, the essentialist, the platonic view, the Aristotleian physics, this idea that Descartes, the idea that you have to answer the question, what is, before you have any hope of precisely understanding a how to question. Okay. That’s who he’s responding to.

[00:53:48]  Blue: Okay.

[00:53:49]  Red: This response is exactly the same as what Strevins is saying. And it’s over the exact same philosophical disagreement.

[00:53:57]  Blue: Okay.

[00:53:57]  Red: So Popper is agreeing with Strevins here. Okay. So here’s a possible best of both worlds approach that takes the strength of both Strevins and Popper as well as Deutsch. I think we should stop worrying about if you have a full explanatory theory or not, you don’t, you never have a full explanatory theory, period end of story. I think the correct answer is that any amount of minimal explanation is enough to get started, even just a simple causal claim. So long as it has empirical consequences and it’s, and it isn’t ad hoc, it has empirical consequences independent of the problem being solved. Start working on what you can with that shallow initial explanation and let that teach you what to do next. As you go, you can continually strengthen your theory by deepening the explanation. As the theory deepens, it explains more. It makes the theory have more implications, which is the same as saying it, it makes it more testable. So to use an analogy on the surface, this might seem a little bit like the old joke of searching for your keys under the lamp post at night, because that’s where the light happened to be, even though you know your keys were lost elsewhere. But this would make a lot more sense if you realize that searching under the lamp post makes the lamp post brighter. That’s what science is like. We start searching under the lamp post because that’s where the light is. But the reason why we do it is because it makes the lamp post brighter so that there’s more you can search in the future. That search might eventually become exactly a question of what is energy or what is gravity. And traditionally, that’s exactly what has happened. Okay.

[00:55:37]  Red: When understood in this way, I think that this is a way of understanding Deutsch’s point about explanation of science as more valid, right? Like if you can get around not using it as a weapon against theories you don’t like and accept that a shallow explanation theory is valid. And instead, understand it as you don’t want to stay with just shallow theories. You absolutely want to deepen your theories and make them more empirical over time. I think if understood in that way, I can agree with what Deutsch was getting at. Let me put it this way. It is not that science requires or even prefers shallow explanations. It’s that science cares more about formulating the theory as non -ad hoc, as independently testable, rather than about how deep its current explanation currently is. Science will gladly welcome shallow but testable explanations, just as Popper and Strevin’s claim. But it will also accept deeper questions of the nature of things if you can also make those testable too, just as Deutsch claims. But Popper is wrong to insist on only how questions and disallowing what is questions. And Deutsch is wrong to dismiss all shallow explanations as explanation of science, because those are exactly the right starting point for probably all theories. Deutsch gives several examples of explanation of science in the beginning of infinity, particularly around genetic heritability of happiness and animal qualia. As bad as some of those examples are, I sometimes wonder if maybe the studies he attacks weren’t nearly as counterproductive as he, to science as he claims. You do have to start somewhere. So if you, if you come up with a shallow explanation that’s quote unquote explanation of science, I would prefer the idea of a shallow explanation.

[00:57:23]  Red: I don’t think anything, there’s really ever true explanation of science. I think it can be very shallow explanations though. A lot of the theories that he attacks as counterproductive, almost all of them were the starting point for deeper theories that came later, right? So yeah, the individual studies he points to, I will do a separate podcast while I go into the details of this, where I actually read what Deutsch said in detail. I don’t want to get sidetracked at this point. But I think some of those examples, they are cringy bad. And so he’s right to call them out. But they were, they were the opening shot. And a lot of these theories, including about the studies of happiness, you know, whether it’s genetically heritable, including whether animals feel things, there are way stronger, more testable theories that exist today, precisely because somebody took a shot at the beginning, right? And they allowed that to start the error correcting correction process. So I’m uncomfortable with some of Deutsch’s examples. I’m comfortable that the specific examples he chose were in fact bad, like he says, I’m uncomfortable with the idea that they were counterproductive. I think that all of them turned out to be productive in the end. I think it makes little sense to disallow a scientific theory that makes testable predictions simply because the explanation doesn’t explain every single possible thing yet from top to bottom, because that’s a standard no one can reach. If that were really true, I doubt science would ever make any progress at all. So let’s move on to the next point that Stravins makes. Next Stravins argues that Kuhn got at least one thing right, that scientists are human beings. Now, consider this quote from Popper.

[00:59:00]  Red: He says, a rationalist is simply someone whom is it’s more important to learn than to be proved right. Someone who is willing to learn from others, not by simply taking over another’s opinion, but by gladly allowing others to criticize his ideas. That’s from all life’s problem solving. So Stravins points out, I think correctly, that this is an impossible standard for human beings. And I think it’s maybe even a dangerous standard. You may sincerely think that you were willing to learn more, that you’re more willing to learn than to be proved right, that you care more about learning than being proved right. And you may even think that you gladly allow others to criticize your ideas, when in fact, you may be bad at both of those. You may be even counterproductive at both of those. I’m always a bit surprised at the level to which humans can lack self -awareness of their dual standards of criticality towards their own pet theories versus other people’s theories.

[00:59:58]  Blue: Wait, now you’re dissing my favorite Popper quote? Yeah,

[01:00:04]  Red: you know, if you take it as kind of a high level quote, like it’s not a bad idea at a high level of abstraction, right? Of course, some people are better at being rationalists and are better at wanting to learn instead of having their ideas proved right. So as far as an ideal goes, I got no problem with striving for this as an ideal.

[01:00:29]  Blue: It’s an attitude that we should strive towards. Yes,

[01:00:33]  Red: but the key point I want to make here is that almost everybody thinks that it’s an attitude they already hold. And so if you’re trying to define critical rationalism only in terms of these platitudes, it’s not going to get you very far, right? It’s everyone’s going to say, oh, that’s me, right? And you’re going to be done. And really, you have to then dig into, okay, what do we really mean?

[01:01:00]  Blue: I get it.

[01:01:02]  Red: So by the way, the fact that I’ve seen this so much, this dual standard that exists between the criticality towards your own theories versus other people’s theories, it’s made me realize that that must be also true of me and that I just can’t detect it, right? I just think it’s really hard to detect the fact that you are less critical towards your own theories. Because of that, as I pointed out in the previous podcast, in the previous podcast, I mean, everyone thinks that they meet this standard, so it’s a useless standard. Similarly, even the famous saying, I may be wrong and you may be right and by an effort, we may be getting nearer to the truth, I do think it’s a nice sediment. And I think at a certain level of abstraction, it’s true. But I think nearly everybody thinks that they’re doing that no matter how much they aren’t doing that. So we really do need to look at this harder. You can’t look at critical rationalism only in terms of the platitudes, you must dig deeper as to what’s going on. So Streben’s argues that science addresses this problem by making a distinction between how scientists behave individually and particularly in private versus the nature of their public arguments. So in private, scientists, according to Streben’s, are free to utilize whatever source of inspiration they desire. It could be philosophy, it could be aesthetics, it could be theology, it could be revelation from Scripture, it could be gut feelings, it really could be anything. As far as their private thoughts go, all of it’s open game. Peter, you

[01:02:42]  Red: and I have talked about, I don’t know if we did it on the podcast or not, but we talked about that Michael Levin, it seems rather obvious, even though he’s never actually publicly said this, that his whole set of theories came out of his belief in Buddhism, that he really is behind the scenes. And he’s done amazing work. Like, I don’t think anybody doubts that Levin has upended the way we think of evolution, and that it’s going to be a drastic revision of the way we think of evolution, because now he’s shown that learning algorithms exist all the way down, probably into non -living things. Like it’s a drastic, radical revision of evolution. And it probably was inspired by Buddhism and by panpsychism. And that, I mean, I don’t believe in those things. It doesn’t matter that his theories came out of his theology, because he does not include it in his public scientific communications. And ultimately, his science is about what programs are productive that make predictions that you wouldn’t have expected and that he can then test and he can verify the tests and corroborate theories. And that turn into a productive research program, right? That’s the way he sees it. So really, that’s okay. Like, theology is a perfectly acceptable source of inspiration in science, but only in private. So Straubens argues that scientists have their own individual subjective plausibility rankings of various theories, at any given point in time. And they use these subjective plausibility rankings to decide what they believe and what to work on. So quote, subjectivity need not mean anarchy.

[01:04:37]  Red: There are rules for interpreting evidence, but they are rules that allocate a role to subject, subjectively formed estimates of a hypothesis likelihood, or as I will call them, plausibility rankings. So you may be wondering if plausibility rankings are the same as Bayesian probability rankings. In the footnote, Straubens references Bayesian reasoning as one possible way to implement this. Okay. So Straubens is here at least kind of including Bayesian reasoning into his whole epistemology. But Straubens is arguing, and this is the key, is arguing that it is explicitly the subjective plausibility rankings that, let’s be clear, science disallows in all public communications, which frankly isn’t likely to sit well with Bayesians. I’m not sure if Straubens even realized that his theory was in some very real sense anti -Bayesian. Okay. He certainly talks positively about Bayesian ideas, but if Straubens theory is to be taken seriously, we can decide if that’s if it needs to be or not. Then he’s saying that Bayesian reasoning is real, but that it’s part of the private thinking of scientists, and that that’s what we are cleansing out of all scientific communications is all that Bayesian reasoning. It’s gone. We remove it because it’s just subjective. It has no place in the public communications of science. So by the way, I happen to know that Ivan Phillips, our resident Bayesian, that he thinks highly of this book. He’s quoted it positively in the past. I don’t know. I mean, like the way the book, like I’ve mentioned that when people read things, they tend to read vibes instead of content. The vibes of this book are very positive towards things like Bayesian reasoning and induction.

[01:06:34]  Red: But if you look very carefully at the content, I’m not sure it’s so positive towards Bayesian epistemology. It’s very literally relegating Bayesian epistemology to the same status as theology. Think about that for a moment. That is what Straubens theory is saying. So for Straubens, he’s arguing that part of what caused science to make rapid progress is that it forced scientists to only utilize their subjective plausibility rankings in private as a matter of conscience, if you will, and to cleanse on sterilize those subjective subjective plausibility rankings in all their public communications as much as possible. It’s a separation highly reminiscent of private religious versus public secular divide that empowered open societies today. So Straubens uses Arthur Eddington’s paper, A Determination of the Deflection of Light by the Sun’s Gravitational Field, which that is, by the way, the famous general relativity paper that everyone knows about. He uses this as an example of how this works in real life. So we know today that Eddington was overtly biased towards general relativity because he was quote, enchanted by the elegance of Einstein’s theory. That’s from page 156, by the way. In retrospect, this bias seems does seem to unfairly impact how Eddington reports his findings. Now this may surprise many. If you’re not really familiar with the history of science and you’re used to just using kind of Popper’s example of Eddington as an example of an attempt to falsify Einstein’s theory, Popper really didn’t fully understand this. And there’s actually a lot of interesting things around Eddington’s expedition that really do need to come to life. You really want to fully understand how science works. Now this. So Eddington’s results, this is the thing that most people may not know today.

[01:08:38]  Red: Eddington’s results with the Brazilian astrographic telescope matched Newton’s theory, not Einstein’s. Okay, I’m going to say that again. The Brazilian astrographic telescope, the result matched Newton’s theory, not Einstein’s. Okay, that was a part of Eddington’s results. But because Eddington was enchanted by Einstein’s theory, he wrote things up to favor Einstein’s theory more so than the evidence at the time really suggested. From page 48 now from Stravins book, all the photographs taken by the Brazilian astrographic telescope have to say that gravity’s effect on light, that astrographic bending number is 0.86, almost exactly equal to Newton’s Newtonian prediction of 0.87 and less than half of the Einstein prediction of 1.74. And Eddington’s response to doing so, reasons for sorry, Eddington’s reasons for doing so, that he felt the Brazilian astrographic instrument has had malfunctioned, we know was based really on a little more than he thought Einstein’s theory was correct because he was enchanted by it. So here we haven’t have everything Popper claimed makes you irrational. Eddington is biased, he’s quick to ad hoc explain away negative evidence. By pure luck, he happened to pick the winning horse, meaning that future experiments did corroborate Einstein’s theory, thus making Eddington’s paper foundational. Einstein had I think it was three different telescopes, at least from what I’m reading from the quotes I have here handy from Strevin’s book. So yeah, so one of the three telescopes matched Newton’s numbers, not Einstein’s. And Arthur Eddington dismissed that telescope is not working. But the basis for him dismissing that we basically know at this point that the basis for him dismissing those results was that it didn’t match Einstein’s numbers, which makes the whole thing kind of circular, right?

[01:11:03]  Red: So now that might surprise a lot of people. The simple truth is, is that anytime you hear Eddington taught today, what happened on the editing expedition taught today, it’s always taught as if Eddington definitively falsified Newton and showed that Einstein’s theory was more correct. That just, it just isn’t the case, right? Like that’s a, that’s a revisionist history that just, it was not at all what actually happened in real life at that point. Here’s a quote from Strevin’s where he tries to explain this, page 49. If the systemat massity and objectivity of science can be seen in the pains taking measurement and calculations and the transparent presentation of the astrographic bending number, the subjectivity and unrulyness of science can be seen in what happens next. The number, with its Newtonian implications, was brushed off in a few sentences by Eddington, declared unimportant by his allies in the British scientific establishment, and ultimately dropped from the textbooks altogether, which is why you’ve never heard of this. Leaving the more Einsteinian numbers supplied by the other Brazilian telescope and the Principus telescope to decide the issue conclusively against Newton and in favor of Einstein’s theory of relativity. So yes, that’s right. The Eddington expedition did not really falsify Newton, you’ve been lied to. So Strevin’s imagines a more paparian idealized form of science where data is the sole impartial determiner. On page 50 he says, if the Eddington story is any guide, this more idealized form of science is pure mythology. There is no tribunal, no method to sort the good photographic plates from the bad. The matter was settled the old fashioned way by a mixture of partisan argument, political maneuvers and propaganda.

[01:13:01]  Red: Now, so does this mean that Kuhn is correct that science is primarily determined by sociological explanations? And does Eddington’s paper therefore deserve the foundational spot it’s been given given its bias? Strevin’s argues no Kuhn is ultimately wrong and yes, Eddington’s paper deserves its foundational spot. Why? Because ultimately Eddington follows the iron rule of explanation and honestly reports the data and minimalizes his own feelings in the write -up. This is from page 161 now. So it is, in general, when a scientific paper is written, the grounds of many of the experimenter’s crucial assumptions being partial or wholly subjective are cut away. What is left are only observation reports, statements of theories, and other assumptions and derivations that connect the two. Consequently, arguments appearing in official scientific venues, such as Eddington’s published argument for ignoring the Brazilian astrographic telescope are characteristically incomplete, perfunctory, or oddly blunted. Strange though it may seem, this is what the iron rule’s insistence on objectivity entails. This is from now page 159 to 160. The subjectivity in the paper must manifest itself then in Eddington’s argument that the Brazilian astrographic was functioning so badly that its measurements should, in effect, be ignored. But why? Comb the paper looking for that argument, however, and you’ll come up empty -handed. It is nowhere to be found. Eddington does tell his readers that the astrographic images were diffused and apparently out of focus. He then speculates after the cause, quote, the unequal expansion of the telescope’s external mirror through the sun’s heat. And he is very clear that the images are to be given much less weight. Indeed, by the end of the paper, apparently no weight whatsoever. But there is something essential missing from his chain of reasoning.

[01:15:05]  Red: Eddington would have to share his reasons for believing that there was, in fact, a scale change. He does nothing of the sort. The eclipse paper is, in a consequence, peculiarly deficient as a piece of rhetoric, although Eddington made up for that behind the scenes. At the same time, and for the same reason, it maintains not only a semblance, but a genuine claim to objectivity. There is little in the paper that can be contested. Not the measurements. He actually reported the measurements that went in favor of Newton. Okay. Not the calculations, not the observations, but the clarity about the clarity or blurriness of various sets of plates. Not the speculations about the cause of blurriness and its possible consequences. You can hardly argue with a mere speculation. And not the logic that says that data from a broken instrument ought to be given little or no weight. The objectivity is not quite complete. The bare claim that the astrographic telescope yielded no useful information can certainly be disputed. But it is close, certainly. The subjective considerations that we suppose pushed Eddington to draw that conclusion, running from the beauty of relativity to the ugliness of post -war politics, are entirely absent. It is the iron rule that demanded objectivity of Eddington’s paper as it does of all scientific communications. Everything subjective in a scientific argument, the rule says, must be eliminated. So the iron rule is that all scientific arguments must be sterilized, arguments must be sterilized of subjectivity. So page 93, the iron rule of explanation directs scientists to resolve their differences of opinions by conducting empirical tests, rather than by shouting or fighting or philosophizing or marrying or calling on a higher power.

[01:16:52]  Red: That is an impossibility and surely Eddington himself was not at all sterilized of subjectivity and his report is biased. And behind the scenes, Eddington was significantly more biased and tried to rewrite the story of the Eddington expedition as a success for Einstein more so than it really was. But the report itself was itself almost entirely sterilized of subjectivity. He honestly reports the missed prediction, because this is what it means to do science, or if you prefer, it’s what it means to do critical rationalism. Strevins argues that science can be quite tedious because of the iron rule that all subjectivity must be cleansed. So from page 256, he says, Why do scientists put up with this rule? Scientists embrace the tedium of empirical work. I’ve written because the iron rule says that’s what it is to do science. Every generation of future scientists, however, has to be convinced to subject itself to the rule. Page 192, he says, Likewise, scientists abide by the iron rule, not because it’s impossible for them to conceive of an alternative, but because they know that the rule characterizes what it is to do science. And just as chess players want to play chess, they, the scientists, want to do science. Strevins argues that it was largely, but not entirely, Newton that really perhaps accidentally invented the iron rule and therefore invented modern science, though with a considerable assist from some 17th century cultural movements. He says pages 188 to 189, the work that made Newton famous, his physics of gravity and light was pursued entirely in accordance with the iron rule’s edict that only empirical testing counts. Alchemy, theology, and

[01:18:45]  Red: the sipa sepentia, a supposed ancient wisdom about the true nature of things that Newton found was admired a lot and was a part of a lot of his inspiration, played no role in these investigations. Newton judged physical theories solely with respect to observable phenomena that they were able to explain. Yet he was not following the iron rule because it didn’t exist yet. He was not following any methodological doctrine at all, because none had been invented yet. What drove him was pure instinct, a quirk of his psychology that made him quite unlike his 17th century peers, a natural intellectual compartmentalizer. When he entered the alchemy lab, he not only put on the alchemist’s robe, he also assumed the alchemist’s persona, taking on their elusive language, their allegorical style of thought, and their conception of matter and the principles of chemical interactions. In his beliefs, his behavior and his words, he became an alchemist. And of course, this Newtonian approach to science was successful beyond imagination, leading to everyone wanting to mimic his approach. In doing this, Stravins argues that Newton invented modern science, and its attempts to sterilize and cleanse subjective criticism in favor of objective empirical criticisms only. Now, Stravins contrasts his view to Coons and Poppers. Stravins perceives Coon as arguing that scientists are limited by their paradigmatic theories. While he perceives Popper as arguing for humans to cleanse their minds of bias and open themselves up to criticism. And we’ve seen that there are some quotes for that we’ve just quoted from Popper where he effectively does say that. So, Stravins argues that instead, this is from page 238, what is severely limited is not the thinking of scientists, but the scientists’ official speech.

[01:20:44]  Red: Deprived of philosophy, religion, and beauty, and those limits, irrationally narrow in their own way, have the same ultimate effect that Coon attributed to the paradigm, to launch a fiercer, more tenacious, and more exacting search of the space of possible explanations than would have been attainable given the foibles of our psychology by any other means. Stravins sees science as constraining or perhaps prohibiting most forms of criticism to only empirical tests. Stravins sees this as a deliberate prohibition. The iron rule of explanation enforces a strict prohibition on non -empirical considerations within official scientific arguments. This means philosophy, religion, and aesthetic arguments, no matter how persuasive or well -founded, are deemed out of bounds in scientific journals and conference presentations. This is a key innovation, distinguishing modern science from natural philosophy, which readily mingled various forms of inquiry. While I don’t agree with Stravins in all particulars here, let me tell you a real -life story that did drive this home to me. I was once debating a Deutchian crit -rat about, of all things, if hunger was a cause of, or in other words an influence of, influencing people to want to eat. So I was in shock that anyone would claim hunger didn’t influence people to eat. I mean, like, we all know that’s what hunger exists for. It makes you uncomfortable, makes you so that you can’t even start to have a hard time concentrating on other things because you have this really strong uncomfortable feeling, and then it tries to get you through that uncomfortable feeling to get you to go feed yourself and to take the time to go do that and stop whatever else you want to work on.

[01:22:28]  Red: So it’s one of the many ways that our genes align our interests with theirs. Yet we all know that because human beings are universal explainers, there are exception cases like Gandhi, or for that matter a girl with a body image problem. They can learn to ignore their hunger in favor of some other desired goal. But I wasn’t arguing that hunger forces you to eat, only that it existed to create discomfort and unpleasant feelings as a cause of influence, admittedly only one of many influences, but a causal influence nonetheless. We have all experienced hunger, so we understand exactly how influential it can be to most of us most of the time. So we have all firsthand experienced the truth of the idea that hunger does coercively influence your behavior. But it’s a view that is not popular with the Deutsche and Crittratt community because they often deny any strong influence of genes upon our behavior. So this Crittratt was debating this with what to me seemed like kind of a word player, word trick. He would interpret my claim that hunger had causal power to mean hunger can force you to eat. Then he would refute that claim, the supposed claim by pointing to someone like Gandhi. I noticed that he was misreading me because of course that wasn’t what I meant. I definitely believe that a universal explainer may have other ways that they can overcome the influence of hunger, but we know that it does influence you still.

[01:24:00]  Red: And I would do my best to clarify myself that I wasn’t claiming hunger could force you to eat, rather than it would make you uncomfortable to influence you to eat, and it influence that statistically over a population would be visible and would be consistently visible, that by creating hunger you could predict that people would go eat on a regular basis and that that would align our psychology with what the genes wanted. But he was having none of it. The problem was, so for instance, he might argue against me and he might say, oh, well, we don’t really know what causes human behavior. We would need a theory of qualia to understand that. So nobody knows the answer to that question. Sure, there’s an observable fact that people do eat, but like you can’t claim that’s because hunger causes you to eat, that would to be able to make that claim, you would have to have a deep explanation, including a whole theory of qualia and you would have to be able to explain how eating, you know, was connected to these feelings. And that was the way he would kind of argue against me on this. Okay, the problem here is that the phrase to cause sometimes means to influence and sometimes means direct cause. And so does the word to influence by the way, right? So no one doubts that holding a gun to a person may well quote cause them to do something they didn’t want to do, even though we all know that they can theoretically it may even decline with a gun on them, right? Hunger and pain are I was arguing the same sort of thing.

[01:25:42]  Red: If I hold the gun on you and I say, you’re going to do this now, you may choose to ignore me, right? Like it doesn’t really force you to do anything. And yet no one would doubt it’s a coercive influence. Okay, hunger is the same way is what I was trying to argue. So some may recognize this story from the past podcast. This story did have a big influence on me. Okay, that’s why I brought it up a few times. So the question I was asking myself was, how do I reach this fellow crit rat and help him see that critical rationalism is honestly, it’s going to look foolish. If we can’t openly emit something as well corroborated as hunger influences you to eat, right? That’s very straightforward. We all know it’s true, right? So I tried making an appeal that I thought would make sense to a fellow critical rationalist. I pointed to how a theory like hunger causes or influences people to eat across the population is a wholly testable theory and that his theory by comparison that he was advancing wasn’t testable. Okay, now the reason why I did this is because I was under the impression that it was a given that all critical rationalists understood that there was this demarcation that scientific theories were given priority that if I’m presenting a scientific theory, you don’t get to present a non scientific theory, a metaphysical theory as a competitor to it. Okay, he was unimpressed. He frankly admitted his theory was metaphysical, but so what? This is an experience that I’ve had many times where I would think all critical rationalists agree with popper on certain points, only to then realize that critical rationalism actually comes in different forms.

[01:27:26]  Red: And some of them don’t agree with even some of the paradigmatic portions of popper. Okay, so for example, I was really surprised when I discovered that Danny Frederick, a very famous, at the time living, he’s since passed away, very famous, Popperian scholar, that he has created an instrumentalist version of critical rationalism and he believes in instrumentalism. I would have thought critical rationalism was wholly at odds with instrumentalism. But Danny Frederick, who really was a really good scholar of Popperian epistemology, was an instrumentalist. So I was in shock when I realized, oh, wow, there’s, Frederick’s created an instrumentalist version of critical rationalism. So maybe it’s not too surprised I was shocked. I’m just trying to say, I haven’t always realized that different critical rationalists have taken critical rationalism in different directions and have their own versions of it, if that makes sense. So in any case, this crit rat sincerely felt that his philosophical theories were better theories than testable empirical scientific theories, alternative that I was raising, and he saw no reason why he should prefer an empirical testable theory over his preferred metaphysical theory. So I made a comment about how Popper did not allow for non -empirical theories to be competitors to empirical theories, which is true. That is the whole point of the demarcation criteria. To my surprise, a whole bunch of crit rats joined the debate. This is on Twitter, so they don’t necessarily know what the context is, because it happened over the course of many weeks, and a lot of them were just joining into the middle of different conversations that Twitter happened to feed up to them. So they didn’t necessarily understand that they were defending the idea that we

[01:29:20]  Red: were actually talking about if hunger could influence you to eat or not, right? If hunger did influence you to eat. So all of them were arguing that it was just not true that you could not have metaphysical competitors to an empirical theory. Now, it’s not that they necessarily argued that Popper hadn’t said that. If they had been arguing, Popper didn’t say that. It would just be a trivial matter for me to just simply show that Popper did say that, and it would have been done. So this is explicitly what Popper’s demarcation criteria was actually about. But instead, they were arguing that it was just wrong, that even if Popper did say it, it didn’t matter. In fact, one of them explicitly said, I do not care what Popper thought. Now, I was again in shock, not realizing that there was a form of critical rationalism that felt this way. How could so many critical rationalists disagree with Popper over something that I would have thought was so basic to Popper’s whole epistemology? Now, I recall, I’ll go ahead and use the person’s real name in this case, it was Aaron Stuple, the guy who wrote the book recently, The Sovereign Child. Keep in mind that he wouldn’t have understood the context as being about hunger influencing behavior. That is kind of a strange thing that crit rats will defend it all. Of course, hunger influences you to try to eat. But he was really trying to argue strongly that we really shouldn’t put any sort of constraint. If someone puts forward an empirical theory, there’s no reason why they shouldn’t consider philosophical alternatives. So he said something like this that got me thinking.

[01:31:01]  Red: He argued that we should not disallow any theories or criticisms because all were valuable, that it made no sense to only look at empirical theories and empirical criticisms as somehow higher value than other kinds of theories and criticisms. Sure, we should use empirical criticisms when available, but why would you exclude anything? Isn’t it all valuable? This is kind of the argument that he was making to me. Why not just allow all theories and all criticisms? That’s I think pretty close to a quote that he actually used. Why constrain the conversation to only empirical theories, even if they were as I was arguing more valuable, why not consider every theory and every criticism? Okay, now let’s be honest. This argument, it makes a certain intuitive sense. I can totally see where he’s coming from. How could it make sense to constrain which theories and criticisms you allow? Why not, as he was saying, allow all kinds of theories? Just allow them all, right? Let’s just talk about them all. Let’s allow all theories, all criticisms. Let’s just put them all on the table. I believe it was a comment that got a lot of likes, a lot of people were replying and saying, yeah, I totally agree with that. There was a lot of crit rats that agreed with this idea. And doesn’t it just make sense? I mean, if I were to put it this way, how can A plus B ever be less than A, right? So on the surface, I have to admit to the intuitive appeal of this argument. In fact, let me even go say this. Those of you listening to this podcast today, right now, as I’m talking, I’ll bet you 99 % of the listeners agree with Aaron on this.

[01:32:41]  Red: Be honest with yourself and admit it. You really do not see why we should constrain which theories and which criticisms are allowed in our communication. How could that ever be a positive thing? Surely you believe that you’re perfectly capable of telling a good theory from a bad one based on your own rational ability to decide that. So why not bring all theories and all criticisms to the table and just consider all of them? Okay. This is just a natural thing for everyone to think, and I think it’s kind of the starting point for how we think of reason. So I think the crit rats in this discussion were doing a good job of vocalizing Strevin’s key point here. What Strevins is calling the iron rule of science that we cleanse all criticism except empirical ones from public scientific communication truly does look entirely perverse on the surface. It goes against our intuitive understandings of how best to reason. Here is Strevins now in his own words, trying to explain this from page 238, I think. An apparently impenetrable logical barrier once stood between those people of the past before Newton and modern science. The iron rule, science’s first commandment looked perverse. Perceiving its perversity, if they considered it at all, knowledge seekers had no reason to put it into practice. Not seeing it at work, they could have no notion of its power. If the human race was going to get its vaccines, its electric motors, its wireless communicators, the wellsprings of health, the armatures of industry, the filaments of human connectivity, something out of the ordinary had to shatter that barrier.

[01:34:31]  Red: Strevins argues Newton played a big role, along with other cultural shifts he mentions, in shattering this barrier of intuition that that was holding the human race back from reaching their full creative potential. So again, page 238, Newton’s unusual psyche, his instinctive urge to compartmentalize was the right kind of eccentricity. This is Strevins’s answer to Aaron Stuple’s argument. It isn’t that we’re doing a way with other kinds of criticism. You are free to use them privately as much as you want. But a good theory, a correct theory, should be expected to ultimately cash out in terms of actual consequences that can be tested empirically or checked. And channeling your efforts into practices and institutions that drive towards this goal of resolving all disagreements empirically are part of the explanation for why rapid progress finally exploded. Consider if we can maybe look at Strevins’s claims here, consider that Newton published the Principia in 1687. And most would consider the Enlightenment as starting around that time, often cited as being 18 or 1685. Newton represents a transition from the end of the scientific revolution and the start of the Enlightenment. And in the view that’s being expressed here, Enlightenment did not beget science, science begat the Enlightenment. It was an attempt to take this idea of cashing out all criticisms via objective testing and applying it to other walks of life, such as political governments. And on this view, it created the Enlightenment. If this is true, this is Strevins theory, not saying it’s necessarily true, I hope you can see how this has an immediate impact on Deutsch’s theory of static societies, and how it represents a fairly radical revision of his theory without nullifying the overall spirit of his theory, I would say.

[01:36:28]  Red: Put simply, Strevins argues, you can’t have rapid progress without first discovering the power of constraining one’s criticisms to just the highest quality empirical ones, at least in public communication, so that all disagreements are resolved via the higher quality criticisms that anyone can check. This is the big idea that Strevins is raising that I think has impact on the way we would view critical rationalism. There is the open question of whether this is what Popper always intended, which is what I read Popper as meaning, or if this is not what Popper always intended, which is how many other critical rationalists read Popper. I do interpret this all slightly differently than Strevins does, admittedly, but I do think the basic idea here is correct. Let’s put this to the crit -wrack community. Why hunger did influence one to eat? They would simply take the phrase influence to eat to mean forced to eat, and then they would argue that that was absurd, and it was very, very difficult to get them to understand what I really meant. In fact, what I finally did is I even said, are you claiming that hunger in no way causes or influences a person to eat? I would try to get them to take the other stance, and they would just slip away. They would say, no, I couldn’t pin them down either way, that it was or wasn’t a cause, or it was or wasn’t an influence. It would always come back to just misunderstanding what I was saying. I then offered how we might do an experiment, and this didn’t help either.

[01:38:22]  Red: The crit -wrack I was arguing with simply said that we have no idea why people that are hungry eat, that would require a theory of qualia, blah, blah, blah, okay. Therefore, my experiment, because it was explanationless science, it was just a correlation, it was just an unexplained correlation, and so it didn’t mean anything. He could ignore it. Now, suppose instead that this crit -wrack followed Popper’s no ad hoc rule, or Strevins’ iron rule of explanation, suddenly the whole argument exists in a very, very different world. Now, this person, if he’s buying this rule, if he’s holding himself to this rule, as the game he is playing, okay, my theory makes a testable prediction. Hunger over a population will cause people to eat. We can show that via experiment, and we all know it will get corroborated by experiment, right? Because it’s true. Because if he’s actually following the no ad hoc rule, he knows it’s not enough to armchair philosophize this away, or to vaguely claim that we need a theory of qualia to answer this question so there is no answer. He knows that he must put up or shut up and offer a non ad hoc testable counter theory. At this point, he’s constrained. He doesn’t have all the different escapes that were available to him anymore. The misunderstandings are almost guaranteed to melt away. Not that scientists don’t misunderstand each other, but because they’re constraining themselves to only empirical tests in terms of their public communication with each other. Because of that, they’re very strongly working to not talk at cross purposes with each other.

[01:40:07]  Red: And a lot of the kind of model that exists in almost every conversation I’ve had philosophically with the critrack community, they vanish the moment you kind of accept a constraining rule like this. And the reason why that’s true is because he doesn’t have the incentive to misunderstand me anymore. He knows that I’ve just made a move in chess and that the only counter move is he must argue by making his own theory that makes its own empirical predictions. So the incentive to misunderstand me is just gone. He’s going to get what I’m saying and all the philosophizing is just bom, bom. It’s removed in a flash. We’re now left with the actual argument I was making. Hunger does cause people to eat. Period, end of story. That’s the theory. He understands that it doesn’t mean Gandhi’s an exception case and I’ve refuted you. He knows none of that matters anymore, because now he must put his response, not in terms of philosophy, not in terms of these vague ideas. He must put his response in terms of proposing a counter theory at least as empirical as mine. My guess is that if he was buying into the iron rule or the no ad hoc rule, that he will find that it’s very difficult to dream up a good non ad hoc testable counter theory. That might be because my theory is correct that hunger does influence the behavior of eating or it might be because he hasn’t yet gotten creative enough. But in any case, if he wants to pursue his own theory counter to mine, if he’s accepting this constraint, he must now put in considerable work to get very specific with his own theory until it is capable of independent tests.

[01:41:59]  Red: This is why the no ad hoc rule or the iron rule, if you’re following Stravins, causes rapid progress while philosophizing tends to get stuck in a mire. The moment you’re forced to be very to very precisely explain your theory in such a way that it has real testable content, most of the misunderstandings that bring progress to a crawl vanish in a flash. Counterintuitive though this is, constraining criticism to the highest quality kind of criticism, to ones you can check that anyone can check, is what unleashes rapid progress. And it does so because it forces you into increasingly higher quality theories that are increasingly testable poppers ratchet, thus deepening your explanations. For this is the only way to increase testability is to deepen your explanations. So why did Stravins then consider this irrational? I do want to maybe explain his point of view, even though I don’t agree with it. Okay, Stravins considers this, this idea of only accepting empirical testing to be irrational for the same reason that Aaron Stuple considered it irrational. This ban appears illogical, unreasonable and irrational on the surface. Page 182 for an intellect open to every respectable reason, to every relevant consideration to every good argument, the now familiar doctrine that only empirical testing counts is in its illiberality quite alien. Stravins agrees, though I do not, that this is irrational. Yet Stravins argues that this is what actually did lead to modern science, the enlightenment in rapid progress. Premodern thinkers, and apparently many modern thinkers as well, viewed it as deliberate intellectual impoverishment. That’s from page 206. Stravins describes it as violating the principles of rationality by imposing a wholesale prohibition on all forms of non empirical thinking.

[01:43:58]  Red: Page 237, regardless of their track record for track record or judiciousness in the past. For instance, argues Stravins aesthetic guidance can be valuable, even crucial, but the iron rule throws it away, at least in public communication. Some might be shocked here. Doesn’t science search for, for, doesn’t science impart search for elegance or beauty in equations? Keep in mind that Stravins isn’t arguing that science scientists do not use such criteria personally. He argues that it is kept out of public official publications. For example, many leading scientists, including Paul Dirac, Albert Einstein, and Murray Gelman, privately acknowledged that beauty in one’s equations or elegance can be a chief criterion for the selection of correct hypotheses and a guide to truth. However, in their official publication, Stravins argues that quote, no invocations of beauty, no arguments grounded on the theory’s grace or charm, are found even in the writings of those that claimed aesthetics influenced them. This quote, war on beauty, Stravins’ words, not mine, exemplifies how the iron rule compels scientists to suppress even privately insightful non -empirical considerations in public discourse. That’s why he sees it as irrational. Another example that atheists might relate to better, William Hewell from the 19th century was a prominent scientist and theologian. Hewell believed that theological knowledge could genuinely inform geological research, especially regarding life’s history. Yet in his own scientific work, he adamantly insisted that theology must never be allowed to quote, must never be allowed to influence our physics or our geology, adhering strictly to quote, the ordinary evidence of science.

[01:45:53]  Red: Hewell understood that while he personally found non -empirical reasons very persuasive, that just wasn’t part of doing science and he wanted to do science, so he knew he had to leave such considerations out of his public arguments. No matter how persuasive or well -founded he personally found them. Stravins points out that Hewell struggled to explain why he both thought theology was a real source of knowledge and also that he should leave all theology out of his public scientific communications. He believed both of those. It was well -established that doing science meant living out theology, he accepted that, he just didn’t know how to explain why. So here is Hewell’s attempt to explain why he agreed he should leave out theology out of his public science, even though he also thinks it’s a source of research knowledge. So page 182 Stravins quotes him as saying, that reason whether infinite, whether finite or infinite, must be consistent with itself and that therefore the finite must be able to comprehend the infinite, to travel from one providence to the moral and material universe to another, to trace their being and to connect their boundaries. Stravins at this point quips that maybe that sounds profound to someone, but to him it sounds like pure gobbledygook. So Stravins feels Hewell was struggling to rationally explain why you’d want to leave out a reasonable source of knowledge, in this case theology, which maybe an atheist wouldn’t agree that’s a reasonable source of knowledge, but we could ask the same of philosophy and aesthetics. The fact is, is that in science’s communications, we do leave those all out, we just cut them out and science always tries to situates itself around only the highest quality criticism empirical tests.

[01:47:41]  Red: So Stravins argues that this is a strategic irrationality, despite its apparent illogic, this rule is quote, strategically brilliant irrationality, that channels immense energy into empirical observations and experimentation. It forces scientists to become quote, observational and experimental prodigies, page 182, by redirecting their competitive drive towards producing quote, exquisitely detailed empirical fact, page 172. He sees science is then producing an empirical record that all future theories must contend with and that reduce what theories are considered valid. I don’t find much here to disagree with Stravins over. When I read Popper, I perceive Popper as outright saying that empirical theories are special compared to philosophical theories precisely because of their intersubjective nature and that a theory without empirical implications was to be thought of as having a deficiency. As we’ve seen, not all critical rationalists today agree with Popper on this point anymore and they’ve gone back to seeing this as an irrational constraint on criticism. Now I disagree with Stravins on one important thing. I don’t see it as irrational personally. On the contrary, this is about as rational as you can imagine. In fact, it’s so rational, I think it’s pretty much defines what it means to be rational. We’ll consider Stravins’ arguments on this more in a separate podcast. Let me try to get through, since we’re going a little bit long here, let me try to get through the rest of his big ideas here. Let’s talk about Baconian convergence, probably the most offensive term he’s going to use for a critical rationalist. So the concept is named after Sir Francis Bacon, 17th century philosopher of course, and the chief promoter of induction.

[01:49:32]  Red: As Popper put this, this is from Myth of the Framework, page 84, Bacon’s new method, which he recommends as the true way to knowledge and also as the way to power is this. We must purge our minds of all prejudices, of all the preconceived ideas, of all theories, of all superstitions or idols, which religion, philosophy, education, or tradition may have imparted on us. When we have thus purged our minds of prejudice and impurities, we may approach nature, and nature will not mislead us. We have only to open our eyes, to observe things patiently and to write down our observations carefully without misrepresenting or distorting them, and nature or the essence of things observed will be revealed to us. I would note that a student of Popper on the preparing podcast, as I mentioned, claims this was a misunderstanding of Bacon that he felt most of Popper’s era accepted a strawman version of Bacon instead of what Bacon actually said. I say this, but I don’t really know this. I’ve never studied Bacon personally, but I’m about to read Strevins’ account of Bacon, and let’s just assess that. So it’s interesting that Strevins summarizes Bacon’s ideas like this. This is in relation to how to theorize about heat. So this is from Strevins’ book, page 107 -108, quote, Bacon’s first step, assemble all positive instances of heat, that is, all types of circumstances in which heat is present, in fire in the bodies of animals when the rays of the sun are concentrated using a magnifying glass, when two solid bodies are rubbed together, flaming meteors, natural hot baths, spicy herbs, and more. His second step, assemble all negative instances of heat, that is, all types of circumstances in which heat is not present.

[01:51:14]  Red: Of course, this list must be endless, so Bacon recommends the following alternative. For each of the positive instances, find similar circumstances in which heat is absent. For example, heat is absent in the body of dead animals, or when two solid bodies are held together but not rubbed, and apparently in moonbeams, comets, and St. Elmo’s fire. Although, as Bacon remarks, further investigation is needed. The third and final step is to assemble all the ways in which heat varies with other quantities. For example, the ways that objects get hotter and the closer they are to fire, or the way that metal takes longer to get hotter than air but retains their heat longer. With the third step completed, Bacon has the evidence he needs before him. What he asks himself can possibly explain all the assembled facts. What can explain why heat is present in live animals but not dead animals? Why is it generated by friction? Why metals soak up more heat than air? He considers all suggestions concerning the essential character of heat. Each explains at least a few of the positive and negative instances but fails to explain others, and so is rejected. This is an actual quote from Bacon now. Every contradictory instance destroys a conjecture about a form. Heat cannot be light and brightness because substances such as boiling water can be hot without being light or bright. Heat cannot be a substance because when hot iron warms another object, it loses heat but does not lose any weight and so on. I would note, this is me now, that this is essentially saying, given what we know about heat already, come to the conjecture about what heat is.

[01:53:00]  Red: At the very end of the process, there is only one hypothesis standing, only one conjecture about the nature of heat that can explain every circumstance upon which Bacon has trained his gaze. I would note that this is clearly not too far off now from conjecture and refutation. Interestingly, Bacon, in his own writings using this technique, came up with the right conjecture about his supposed inductive method. It is the hypothesis that explains how friction generates heat. The quiddity of heat is motion and nothing else, or more exactly, heat is the disordered motion, the vibration of the small particles of which all things are made. This idea, an early version of the kinetic theory, is, we now know, quite correct, a compelling advertisement for the Baconian method. The thing that I find interesting here, and that it might surprise a lot of critical rationalists, is that Bacon’s method was actually very similar to Popper’s conjecture and refutation, but in a muddled sort of way. So, Strevin says that Bacon believed that once scientists followed his method, they would quickly rule out all explanations but the correct one, while this turned out to be rather optimistic in practice, as multiple plausible explanations often compete initially. But, quote, from page 116, Bacon’s proposal is that if we take explanatory power to be our guide, then empirical testing will ultimately single out the truth. This is Strevin’s Baconian convergence. While it’s inductive in its language, and is based on Bacon’s ideas, it maybe isn’t that far off from Popper’s conjecture and refutation. The idea is that since science emphasizes empirical evidence to the near exclusion of all else, that there is always a sterilized record of observations that all future theories must explain. So, what is Baconian convergence then? So,

[01:55:06]  Red: here’s what Strevin says about it. An empirical, it’s an accumulation of evidence. An empirical evidence accumulates, as empirical evidence accumulates, even desperate plausibility rankings among scientists begin to converge. This means that differences in opinion become less extreme and consensus emerges regarding which are the leading theoretical contenders. Baconian converges, neutralizes bad data, not by correcting it, but by swamping it with overwhelming amounts of good data. For instance, Eddington’s Newtonian -leaning Eclipse data simply ceased to matter as many other Einsteinian measurements accumulated. The meticulous preservation of data in its sterilized form, stripped of authorial opinion and bias, smooths and speeds this process. Sterilized data acts as, quote, presupposition -free suspended animation ready for use by future generations. The iron rule of explanation through its procedural consensus continually channels scientists’ ambitions and competitive drive into producing immense amounts of rich intricate and revealing data. This ever -growing reservoir of evidence is the essential fuel for Baconian convergence. And while scientific reasoning is inherently subjective due to plausibility rankings, Strevin argues that Baconian convergence shows that as evidence accumulates, quote, the weight of observable fact came to overpower the differences in plausibility rankings, unquote, eventually making contra -opinion seem, quote, quirky then quirky, then eccentric, then laughable. Page 111 is where I got those quotes from. This leads to what he calls the Tychonic principle. So the Tychonic principle states that the most promising rival theories regarding structures of the world reveal their disagreement in small observable details and only in such details. It’s named after Tycho Bar Brahe, I don’t know how to pronounce that, 1546 to 1601, who went to great lengths to pinpoint the position of stars.

[01:57:14]  Red: This means that the fundamental secrets of the universe and the differences between competing theories about them lie in minute structures, nearly indiscernible details and patterns that can be detected only by the most sensitive, fragile and expensive instruments. So Strevin argues that Baconian convergence, the process by which accumulated evidence eventually singles out the truer theories and makes that possible, is only possible by way of great amounts of complex, highly accurate observations. In a quote Tychonic world where critical differences manifest in subtle details, this meticulous data collection is indispensable for scientific progress and for overpowering subjective plausibility rankings. Now, all of this sounds actually fairly close to something that Deutsch claimed. This is David Deutsch from a new way to explain explanation from a TED talk. He says, you know what the clenching evidence was for the space time that space time is curved? It was a photograph, not of space time, but of an eclipse with a dot there rather than there. The evidence for evolution, some rocks and some finches and parallel universes, again, dots there rather than on a screen. It implies that making measurements to, for example, the sixth decimal place is not a waste of time and money, but crucial for gaining knowledge. So I would note that the Tychonic principle is not at odds with what David Deutsch just said there and that there’s actually a great deal of alignment there. So the knowledge machine, Strevins’ term, has in just a few hundred years achieved Baconian convergence on many truths. For example, the nature of heat, DNA, black holes, because the weight of observable fact overpowered differences in plausibility rankings. This demonstrates that the Baconian convergence is, quote, real and well within scientists’ reach, page 111. So a few examples.

[01:59:21]  Red: The Mickelson -Morley experiment, 1887, this experiment sought to find evidence for ether by detecting an almost infinitesimally small displacement on the order of a few hundred nanometers or.0001 inches in displacement on light beams. Physics biology and beyond the Tychonic principle applies across scientific disciplines. In fundamental physics, it means that the secrets of the universe lie in minute structures. In biology, it means understanding hereditary and evolution and evolution requires pains taking attention to complex causal structures at the microscopic level. Similarly, neuroscientists, geologists, archaeologists, climate scientists, and economists require attention to exact amounts and elaborate connections. So from page 116, Strevins says the Tychonic principle says that much of the testing will be extraordinarily difficult to conduct. It is for this reason that the iron rule is essential to modern science’s success. The effort required to build a store of observable fact sufficient for Baconian convergence in a Tychonic world is so great that humans can be persuaded, induced, or impelled to take on the project only under exceptional circumstances. The iron rule engineers those circumstances and forces scientists to investigate some part of nature in a detailed depth that would otherwise be unimaginable. Then Strevins says of that quote, those are Thomas Kuhn’s words. He attributes the compulsion to paradigms not to universal procedural consensus. Otherwise, this is the best sentence he ever wrote. So as we’ve seen, despite the Baconian inductive language used throughout, there is a whole lot of potential agreement between Strevins and Popper. I would argue the following similarities. The iron rule of explanation is Popper’s demarcation criteria and the specialness of empirical tests as intersubjective and thus objective criticisms, mixed with the Doad -Hawk rule, of course.

[02:01:45]  Red: And the Tychonic principle is the same as Deutch’s claim that we use smaller and smaller evidence to explain radically strange conceptions of the world. And Baconian convergence is really just Popper’s falsification, where we require scientific theories to be vulnerable to empirical tests and therefore eliminate or falsify which theories are actually possible and converge on a small and smaller set that are more and more empirical. Strevins’ shallow explanations are just Popper’s how -over -what -if -what -is questions. And when I put it this way, it seems like Popper and Strevins are talking about very similar things but under different terms. But Strevins would not agree with me. He sees himself as wholly disagreeing with Popper. Strevins sees strong disagreement between Popper and between himself and Popper, and he sees Popper as mistaken in considerable ways. Is this merely a case of Strevins’ maybe misreading Popper as tends to happen? Or is there maybe more to it than that? So that’s an important subject. We’re going to cover that in a future podcast in detail. Let me just say, though, I do love an extended example of how science works that Strevins offers. I’m going to offer a reader’s digest version of it. This is from 93 to 96. It’s so good, I’m really going to try to just go over the whole thing, and that will be kind of where we end here. The unending script, the code of contact for scientific argument, according by which Montague and Kaplan may continue their debate indefinitely, is provided by the methodological precept that I call the iron rule of explanation. What the rule says is simple enough.

[02:03:27]  Red: It directs scientists to resolve their differences of opinion by conducting empirical tests rather than by shouting or fighting or philosophizing or moralizing or marrying or calling on a higher power. That is all. It makes no attempt to interpret the evidence to decide winners and losers. Indeed, its function is not so much to resolve the dispute as to prolong it. The first act has ended in an impasse. Montague said that heat travels across empty space takes the form of radiation. Kaplan said that it was a spume of caloric fluid jetting across the void. The iron rule shows them how to go on arguing by improvising a series of experimental scenes, by conducting a succession of empirical tests. To decide the question of how heat moves to the void, Montague and Kaplan might agree to measure the difference that a pane of glass makes to the velocity with which heat travels across a laboratory made vacuum. If Montague is right and heat is transmitted by radiation, then heat rays should speed through the glass. If Kaplan is right and heat is an invisible flowing liquid, then the glass should stop the heat or at least slow it down. Suppose they perform the experiment. The glass pane makes no difference, so the test looks to count in favor of Montague’s theory and against Kaplan’s. Kaplan might surrender, but she might fight on. After all, her caloric hypothesis predicts a significant slowdown only with certain auxiliary assumptions that the glass pane acts as an effective barrier to caloric fluid. Perhaps that’s not the way it works. Perhaps firing heat at a plain pane of glass is like directing a fire hose at a fishing net.

[02:05:16]  Red: The stuff courses through holes invisible to us, but huge by comparison to the particles that make up the fluid. To test Kaplan’s fishnet response, Montague concedes more tests will be needed. Perhaps stack up many panes of glass at different angles if the fishnet analogy holds sufficiently many layers should slow caloric fluid right down. Or suppose that the glass pane experiment comes out the other way. The glass slows heat to a crawl. Exit Montague, he’s now been falsified, not necessarily. Glass is transparent to the radiation that we call light, but perhaps it is opaque to heat radiation acting like a thick black curtain. To test the blackout response, Kaplan concedes more tests will be needed. Perhaps try various other potential barriers. If heat rays are like light rays, some ought to be transparent to heat. Whatever the outcome of the test, two things are certain. First, the loser will have the chance to save their theory by rejecting one of the auxiliary theories in favor of another. Second, both scientists will agree on the kinds of further observations which will put the new auxiliary assumptions to the empirical test. Whichever way the glass of pane experiment goes, they might agree to try different materials to find out what lets heat through and what does not. Because they follow the iron rule, Kaplan and Montague are in a certain kind of agreement. It’s not agreement as to the best theory of heat. It is not agreement as to what the evidence says. It is rather procedural agreement, an accord on how to go about arguing, by observation and experiment, and not in any other way. This is the very best account I have seen of non -ad hoc theory exhaustion version of falsification.

[02:07:10]  Red: And it came from someone who claims he’s an adductivist, and it’s not a critical rationalist. In a book where he claims he’s disproving Popper, Stravins’ key point here is interesting. He’s saying there is no such thing as falsification of a theory, per se, because you never know what creative non -ad hoc auxiliary theories the other side might come up with. So there is never a point of falsification of either theory. You always have to wait and see what the other side devises. But you always know, so long as both sides are following the rules of the game, that they will come up with a non -ad hoc alternative. So there will always be a way forward with your disputes. It’s hard to see why this does not feel like falsification to Stravins and why he rejects this account as a form of falsification. Yet it’s pretty much exactly what I claim is Popper’s falsification in the last podcast. I came up with this idea before I ever read Stravins’ book and I sincerely thought I was getting it from Popper. To put this very bluntly, I can see why this account of science, which I think is correct by the way, does not involve any actual falsification of theories, and therefore I can see why someone saying, well this isn’t falsification, why they would say that and why they would think Popper was wrong. But when I personally read Popper, this is what I thought he meant by falsification.

[02:08:48]  Red: I did not read him as seeing falsification as being about showing a theory as false, but being about a logical connection between observations and theories that allow you, through the no ad hoc rule and through Popper’s ratchet, to keep that conversation going and to keep improving your theories, even though you may never really falsify either theory all the way, right?

[02:09:13]  Blue: Just briefly, just to clarify, okay, so the Montague and Capulet thing, you know that’s from Romeo and Juliet, right?

[02:09:23]  Red: No, I didn’t. It’s a fictional example, but he said it was based on a true account of how the theory of heat progressed. Okay,

[02:09:33]  Blue: so yeah, Montague and Capulet are the families from Romeo and Juliet.

[02:09:37]  Red: So

[02:09:38]  Blue: it’s basically, as I’m reading it, an ego -driven kind of irrational feud, right? Like a family feud. But then the idea that is that through this conversation, an irrational kinds of emotional conversation, deeper truths emerge. That’s right. There’s no agreement as

[02:10:05]  Red: to the key point that Strevins is making is that there is no standard by which you falsify a theory and it plays no role that you falsify theories in science. What it is instead is it’s a rule that you get to save your theory as many times as you want, any way you care to, any way you can creatively come up with it, but you must do so through an empirical test and your theory must be non ad hoc. It must have independent tests, not just explain the one defect that you had. And that’s what we just saw with this example, right? That one believes in caloric heat, one believes in radiation heat, and I mean, we know radiation heat’s correct and caloric heat isn’t, right? But they come up with an experiment and it falsifies one of them. And then the other side goes, you know what? I don’t think I’m still falsified. I think instead, what’s going on is this. And as long as they can put it in terms of a non ad hoc theory, they have every right to not consider their theory falsified. And so you never really falsify any theory, right? What really happens is, is you exhaust non ad hoc options. Since you don’t know if you’ve exhausted them or not, you may, maybe, maybe you just haven’t thought of the right one yet, right? Because you never know if they’re gone, like if there are no more, or if you just haven’t thought of a better one. So because of this, and here’s the key thing, this is the way science works in real life. And this is the history of science.

[02:11:39]  Red: You’ll be incredibly hard -pressed to find very many definitive falsifications of any particular theory until long after the fact and only in retrospect. And that’s the truth. Now that makes sense. If falsification is actually a process of exhausting non ad hoc theories, and you are and scientists do not accept falsifications of their theories. Instead, they accept rules that they have to come up with an auxiliary theory that’s testable. And we have to test that auxiliary theory, okay? If that’s really what science is, and I’m saying, yeah, that is what science is, okay? Then it makes sense. It makes sense that you wouldn’t just never falsify theories. What you would really do is one side would just kind of stop being productive. They would have a struggle to come up with alternatives at some point. And they may never concede. They may die never conceding. But no one will ever pick up the caloric theory of heat ever again. Because the weight, the intellectual weight it must now lift is just too much. Because there’s just overwhelming observations that it has to come up with testable auxiliary theories too. And nobody knows how to do it. So at some point, the theory just sort of dies out. And years after the fact, maybe decades after the fact, we look back and we say, oh, the caloric theory of heat was falsified. But there was never a point of falsification. It just never happened as an event. And I think most people read Popper’s falsification as a sort of standard by which you can refute a theory and not have to worry about it anymore. And that’s not what it is. Because Popper’s falsification has nothing to do with proving theory’s faults. And

[02:13:32]  Red: I don’t know what to, I mean, that’s how I read Popper. I know that not everybody agrees with me. But that is how I read Popper. And I honestly think that if you read Popper in some other way, you’ve got a false form of science. And I don’t know what to tell you, right? Because any other read of Popper that doesn’t involve this actual idea of how it actually works, where you actually just exhaust the non -ad hoc theories, and eventually the theory dies out because nobody can think of a way to come up with a good ad hoc theory to save it. I mean, sorry, non -ad hoc theory to save it. That’s just really how it actually works. OK. In the past, in the previous podcast, I asked if falsification was falsifiable. Here we have another way to potentially falsify falsification itself. Use the history of science and check if Popper’s epistemology matches what science actually does in real life. So my question is of particular importance. If Popper’s falsification is how you can logically say this theory is false, then the history of science refutes Popper’s epistemology, period end of story, because that does not happen in the real history of science. But if Popper’s falsification is understood as exhausting non -ad hoc theories, then the history of science is exactly Popper’s falsification. OK. Hopefully you’re getting my point here. It’s also interesting that Strevins, like Popper before him, puts this in terms of methodological rules that scientists adopt. That is to say, science works by creating ground rules or constraints that all scientists agree to follow that are institutionally enforced. So this is the key idea that I’m going to raise, that we’re going to revisit a few times.

[02:15:18]  Red: It’s this idea of when you constrain the conversation, when you don’t have the wild open, every theory is possible, every criticism is valid, and instead you significantly constrain that conversation to just the highest quality theories that are vulnerable to the highest quality of criticisms, particularly empirical testing. OK. When you do that, that is what creates rapid progress. And yes, it’s totally counterintuitive. It is at odds with everything you think you know about what it means to be rational. And that is the point that I think Popper was making and that clearly Strevins is making. And that would be my summary now of Strevins’ book here.

[02:16:08]  Blue: OK. Well, Bruce, sometimes it takes me re -listening while I edit to really understand everything you’re saying, but you are a unique person, also a rational person, and I would be surprised if there’s a few people in the world thinking as deeply about epistemology as you are, and you have profoundly influenced how I think. So thank you. Thank you. Hello again. If you’ve made it this far, please consider giving us a nice rating on whatever platform you use, or even making a financial contribution through the link provided in the show notes. As you probably know, we are a podcast loosely tied together by the Popper -Deutsch theory of knowledge. We believe David Deutsch’s four strands tie everything together, so we discuss science, knowledge, computation, politics, art, and especially the search for artificial general intelligence. Also, please consider connecting with Bruce on X at BNILSON01. Also, please consider joining the Facebook group, The Many Worlds of David Deutsch, where Bruce and I first started connecting. Thank you.

---

Links to this episode: Spotify / Apple Podcasts

Generated with AI using PodcastTranscriptor. Unofficial AI-generated transcripts. These may contain mistakes; please verify against the actual podcast.