Episode 93: Philosophical Theories vs Bad Explanations
- 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
[00:00:07] Blue: Today, on the Theory of Anything podcast, I’m very happy to listen to Bruce consider philosophical theories versus bad explanations. Can philosophical theories be refuted? What is a bad explanation? Can all theories be made more empirical? Bruce gets deep into these questions as he takes a dive into the correct way to apply popper’s ratchet to metaphysical or philosophical theories. Along the way, Bruce puts forward a generalization of testability he calls checkability and explains why vague manning our theories is worse than dogmatism. I hope you enjoy this.
[00:00:54] Red: Welcome back to the Theory of Anything podcast. Hey, Peter. Hello, Bruce. How are you? Good. So we’re finally going to do my version of how to apply popper’s epistemology to philosophical theories. On the last episode, we talked about popper’s treatment of that subject. And I had some criticisms of popper’s treatment. But before we get into that, let me ask you a question. And this isn’t a trick question. So I have to always say that because everybody always thinks I’m making a trick question because probably most of the time I am. So popper famously created a demarcation between scientific and philosophical or sometimes we say metaphysical theories.
[00:01:39] Blue: Yes.
[00:01:40] Red: So what side of that line, the demarcation line, is popper’s own epistemology on?
[00:01:47] Blue: I mean, they’re philosophical theories, right?
[00:01:50] Red: So, you know, that’s exactly what he says. He says that his theory is a philosophical theory, a metaphysical theory. And therefore it is not a scientific theory and it exists on the philosophical side of the line. OK. So now here’s a second question for you. Which side of the line is induction on?
[00:02:13] Blue: I guess I’m going to have to go again with philosophical.
[00:02:18] Red: OK. Again, I think popper would agree with you that it is on the philosophical side. OK. Yeah. So we agree that induction and popper’s epistemology, they’re philosophical theories, right? Albeit in the case of induction, we think it’s an incorrect philosophical theory, correct?
[00:02:38] Blue: I pretty
[00:02:39] Red: much, yeah. OK. All right. In the last episode, this will become relevant later. In our last episode, I talked about how popper in Conjecture and Refutations, Chapter 8, a very popular chapter amongst crit rats, that he, in my opinion, at least, accidentally created a magnificent loophole in his epistemology. And I call this the crit rat loophole, or sometimes the crit rat backdoor.
[00:03:06] Unknown: OK.
[00:03:06] Red: All you need to do is declare your favorite theories to be philosophical and seemingly popper, well, at least if you interpret him in the way he gets popularly interpreted in this chapter, says that you get to ignore all potential refutations, and instead you get to have a discussion about much if your criteria, such as does, and this is a quote from Conjecture and Refutations, page 269, does it solve the problem? Does it solve it better than other theories? As it perhaps merely shifted the problem? Is the solution simple? Is it fruitful? He makes a big deal out of fruitfulness for his own theory. Does it perhaps contradict other philosophical theories needed for solving other problems? That was the quote from popper from 269. Those are much if your questions, though, right? They’re nowhere near as straightforward as did we this theory makes this prediction with perform an experiment and it either matches the experiment or the experiment matches the prediction or it doesn’t, right?
[00:04:07] Blue: OK. So what you have just outlined is popper’s ideas on how we judge the validity of non empirical theories.
[00:04:18] Red: That’s right.
[00:04:19] Blue: Just to I just want to be absolutely clear. OK.
[00:04:22] Unknown: Yeah.
[00:04:22] Blue: I got it.
[00:04:23] Red: Now, presumably, Popper never intended to encourage people to try to declare their theories philosophical just so they can get easier standards of criticisms. In fact, I’m quite certain that he absolutely never foresaw that possibility. OK.
[00:04:37] Blue: And that’s the case you’ve made, I would say, quite convincingly in past podcasts about popper’s ratchet.
[00:04:44] Red: Yeah. And ad hoc saves. Right.
[00:04:47] Blue: Which which I really like your ad hoc save is his language. Popper’s ratchet is your language.
[00:04:55] Red: That’s right.
[00:04:57] Blue: They kind of mean the same thing. I actually prefer. Your language makes a heck of a lot more intuitive sense to me.
[00:05:06] Unknown: For
[00:05:06] Blue: some reason, it took like what you were what you were. He was meant by ad hoc save. It took quite a bit for that to sink in. But Popper’s ratchet was like, oh, yeah, we always strive to make our theories more empirical, even our philosophical theories. That kind of seems like a no brainer, really. I mean, it just makes it just just connects.
[00:05:32] Red: So that’s I’m going to make that case today. That
[00:05:35] Blue: OK.
[00:05:36] Red: Popper, that we should always be applying Popper’s ratchet to all our theories, including philosophical theories. OK, I need to explain what I mean by that because I need to make a generalization of Popper’s ratchet for that to work. OK, because the original one that I created was empirical only. And obviously, that doesn’t apply to non -empirical theories. But I’m going to make the case that it’s not that hard to generalize. So now, as I was saying, presumably, Popper never intended to encourage people to try to declare the theories philosophical just so they can get easier standards of criticism. His dark demarcation criteria was really meant to say, look, don’t intentionally make your theories philosophical if you can avoid it. And here’s the quote from. Conjectures, refutations, page 48 from Popper, irrefutability is not a virtue of a theory as people often think, but a vice. OK, so that’s Popper saying, look, you’re not supposed to declare your theories philosophical just to get easier standards of criticism. You should be seeing that as a problem. OK, now, I’ve argued that it’s the reason why it’s I’ve had crit rats when I say this, when I point out that Popper has said this, they’ll say, oh, sure, you’re just saying that every philosophical theory is worse than every empirical theory. It’s like, like, of course, I didn’t say that, right? If we’re looking at competing theories and all of them are philosophical, then all of them have this vice that they can’t be empirically tested and therefore doesn’t really get held against them, right? But it’s still a vice. You have to still see it as a vice, right? And that’s what Popper is really trying to say.
[00:07:15] Red: OK, now, I know I harp on Brett Hall’s theory of intelligence a lot just as a quick repeat. Brett Hall in past episodes, I showed that he and the defenders of his theory, they they have this theory that all intelligence differences because we’re universal explainers that really are no intelligence differences between people and therefore all intelligence differences boil down to two things, interest and time spent studying something. There’s really nothing whatsoever stopping Brett’s theory from being a scientifically or empirically testable theory other than Brett Hall and the theory’s defenders really badly don’t want to see their favorite theory get refuted. I’ve argued throughout this podcast, what a mistake it is to protect your pet theories like this because you can then not improve them. You can no longer error correct them. Now, how do they go about doing this? How do what is the crit rap loophole that I referenced? The crit rap loophole allows bad explanations to masquerade as good philosophical explanations by using the following fool proof trick. Declare your theory to be philosophical. If someone points out that it can be refuted via testing as can be done with Brett Hall’s theory of intelligence, it’s very easy to point to Joe Biden and say he’s got tons of time in interest. Why is he cognitively declining? Right. And so clearly this is how would you explain Joe Biden’s cognitive decline in terms of time and interest, which is Brett’s theory that those are the only two things that matter in terms of intelligence? OK, you right there, you have a hard refutation of Brett’s theory. OK, but how would they deal with? I’ve actually talked with them. I know how they deal with it.
[00:09:03] Red: If someone points out that it can be refuted via testing ad hoc, save the theory. For it is always possible to ad hoc, save any empirical theory by introducing a new, wholly untestable auxiliary theory. Even Popper talks about this, that you’re supposed to make a rule that no ad hoc, save rule to not do this. If someone points out that you just violated poppers, no ad hoc, save rule. Respond that you already said your theory was philosophical. So thus, the no ad hoc rule doesn’t apply to your theory, because that was something that only applied to empirical theories. Then point out how your theory is fruitful and simple or how it solves some arbitrarily selected problem. At a minimum, whatever theory you are against can be declared itself a moral problem, because pretty much everything has moral consequences and therefore can be declared a moral problem. And your theory can can be fruitful by solving this moral problem. OK, I am arguing that literally any theory, any theory can be declared a best theory using this simple trick. OK, and that’s what I call the Cret Rat loophole.
[00:10:12] Blue: And just to review an ad hoc, save just just for our audience and for myself, I just want to make sure I have this right. An ad hoc save is when you try to save a theory by making it less empirical or just not more empirical.
[00:10:29] Red: So according to Popper,
[00:10:30] Unknown: an
[00:10:30] Red: ad hoc save, an ad hoc theory is one that’s not independently testable.
[00:10:35] Unknown: So
[00:10:35] Red: to Popper, a theory, what you’re trying to minimize according to Popper is ad hocness. He actually says that. I’ve got the quotes from the past episodes where I actually quote him saying that what you’re really trying to do is you’re trying to minimize ad hocness with your explanations. So what is ad hocness? It’s theories that have its theories that lack reach. So the theory solves whatever problem you introduced it for, but it doesn’t have any other implications that can be tested. OK, OK. So what you want is you want a theory that solves the problem, but you want the theory to solve the problem by in such a way that it has its own independently testable consequences so that you can go try to corroborate or refute that save. Right. And if you do that and it does have this if it does have the quality of that you can test it independently like that. In other words, that it has a reach that the theory that you’re introducing to save your theory has reach. Then it’s OK to save your theory. OK, now, it’s not too hard to see that if you’re only allowed to save theories by introducing new theories that themselves have testable consequences, then the total amount of empirical empirical content of your theories, the collection of your theories, must always go up. OK. And that’s what I call Popper’s Ratchet. So it’s a Popper’s Ratchet is a natural consequence of Popper’s no ad hoc saver rule. OK. I think it was just
[00:11:58] Blue: I think the concept, I think that the connotation of ad hocness in this sense is just different than what I always thought ad hoc meant. So that’s why it was it just took a while to sink in.
[00:12:11] Unknown: But
[00:12:11] Blue: yes, especially now that you’re saying out Popper’s Ratchet, I’m like, oh, OK, that that makes a lot of sense.
[00:12:18] Red: I think people do struggle with this, right? They kind of have this idea in their mind of ad hocness. It’s kind of a vague notion. And they say, oh, yeah, well, you shouldn’t ad hoc save theories. But you never feel like you’re ad hoc saving a theory. What Popper does is he makes this very straightforward. He says, no, to ad hoc save a theory is to introduce a save for your theory that isn’t independently testable, doesn’t have its own empirical consequences. OK. Yeah. And everybody seems to miss that he actually does this, that he actually says, by ad hoc save, I mean this. Or rather, he does it the other way around. I have this concept in mind that you must not save your theories unless you do it with a theory that is itself testable. I’m going to call that to failing to do that ad hoc saving. And I’m going to call that. And the theory that does that an ad hoc theory. So he’s very clear about what he means. And yet it’s just really hard to release that he’s not quite talking about what you think he’s talking about. Right. And there was no other word he could have picked. You know, ad hoc was probably the best word he could have picked. It’s even weirder when you realize that what you’re really trying to do is you’re trying to minimize ad hocness. So what are you trying to maximize? Well, he says boldness. Well, that’s even vaguer, right? Like that’s a great thing to call it, because what you’re really trying to maximize is how risky your theories are. OK. And we’re going to talk about that today. I’ll give some examples of that.
[00:13:45] Red: But to say, well, my theory is bold. Well, that could mean almost anything. So I’m sure the defenders of Brett Hall’s theory would say, this is a bold theory. I’m saying something that goes against what people think. Intelligence really is. It’s bold. Well, no, that’s not what Popper meant by bold. He meant it had independently testable consequences, which is exactly what they deny their theory has. So they’re just you’re always struggling for words to explain the concept you have in mind. And I felt like Popper’s ratchet, for me, at least, made it way more clear what we’re talking about.
[00:14:21] Blue: OK, and I would agree.
[00:14:23] Red: So Popper has at times, even just in the quotes we just considered, considered suggested fruitfulness as a criteria by which to judge philosophical theories. So on logic of scientific discovery, my only reason for proposing my criterion of demarcation is that it is fruitful that a great many points can be clarified and explained with its help. That’s logic of scientific discovery, page 33. And is that the
[00:14:45] Blue: same as far reaching fruitfulness, far reaching kind of the same?
[00:14:50] Red: Well, you know, I don’t think he ever really explains what he means by fruitfulness. And yeah, it’s it’s kind of a vague concept. And that’s what I’m kind of arguing here. OK, I don’t. It sounds
[00:15:01] Blue: a lot like far reaching, though, right? So I mean,
[00:15:04] Red: I don’t even know if that’s what he meant. I think he meant it explained things. It is a bit vague as to what it means. OK, so this is practically begging communists to declare their theory a best critical rationalist theory. Marxists can now say, oh, but our theory is so fruitful. Induction has been incredibly fruitful in both pushing science forward. Even Popper admits Bacon did a great job of pushing science forward in creating machine learning and other fields. Plus, it is just intuitively correct that humans can quote, induce things that is generalized in that we generalize from specifics all the time. Bayesian reasoning has been incredibly fruitful in machine learning and other fields. And it is also now dominates epistemology because of how fruitful it has been. It is also popular because it seems intuitively correct. I’m emphasizing the word intuitively here. We’re going to get to this. Fruitfulness is such a loosey goosey criteria that I don’t have any real hope of it being any sort of guide towards preferring one theory over another. Likely, you’re just to declare whatever theory you happen to prefer to be fruitful and the whole thing will be circular.
[00:16:16] Blue: Yeah.
[00:16:17] Red: So let’s take a step back and let’s ask the question. Why does science work so well? Why does progress in science work so well? OK. And let’s try to learn from that. And this is in my opinion, that’s this is what Popper was trying to explain. He was trying. He was looking at why science is so progressive so fast compared to other kinds of theories. And he was trying to work out what are the institutions? What are the things, the conventions, the methodologies that exist that allow them to be so good at error correcting their theories? OK. So now human beings have literally hundreds of thousands of years or maybe even arguably millions of years, depending on which hominid you consider to be the first true humans to develop knowledge using their universal explainer minds. And humans fail to progress almost that entire time. This was true in every culture, save one across time, even in the culture, Western culture, that did progress. It does so very, very slowly at first. And honestly, is only barely distinguishable technology wise in terms of cultures around them, arguably inferior to many of the cultures around them. It’s easy here to you’ve probably heard much about how Western culture was technologically inferior to China or the Islamic world. OK. And that’s probably open for to. Well, the Islamic world was probably open for debate. I don’t know anyone debates that China was more advanced in the Western world for a very long period of time. So it and so Western culture, the culture that ultimately ends up coming up with this rapid sense of kinds of progression. It’s not really immediately obvious that it’s a superior culture at knowledge creation for a very long period of time.
[00:18:07] Red: There is a sudden explosion when science is invented, or that’s what we say. But when was, quote, science invented? So I’ve heard different options as to who, quote, invented science. So we might have Ptolemy, that would be 305 B.C. Or maybe Copernicus 1473 A.D. Or a really popular one is Galileo 1564 A.D. nearly a century later or Kepler is another one, a contemporary to Galileo 1571 A.D. Another one that I’ve often heard maybe not as often as Galileo is Newton 1642 A.D. That’s a century later. Hopper claimed in our last episode that you can’t even have a critical conversation about Newton without understanding Galileo and Kepler. This was in Conjection and Refutations, Chapter 8. Understanding Galileo in turn is impossible, or so says Hopper, without understanding Copernicus. In fact, their theories are so similar that really what Galileo was doing was he was advancing Copernicus’s theories. And so Copernicus a century before Galileo was the really the first one who came up with the idea that everything revolves around the sun. Galileo did not come up with that.
[00:19:31] Blue: He just kind of used the telescope, right? That’s right. Copernicus’s theories. Yes. Is that right?
[00:19:37] Red: We have such strong misunderstanding of what happened with Galileo today. We’ve created myths around him that are kind of partially true and partially just made up to serve certain stories, narratives that we want to tell. It’s interesting that nobody considered Copernicus a heretic for both for having a theory that everything goes around the sun. So there was something more going on with Galileo. It wasn’t really just his theory that everything revolves around the sun that got him in trouble. So and, of course, then you can’t really make sense of Copernicus without understanding Ptolemy. So it’s difficult to draw a line as to who invented or when science was invented. OK, I think Deutsch would argue Galileo was an attempt to invent science, but he got put down critical discussion, wasn’t available yet. I think that’s what we would probably say to that. So where do and then there’s a completely fair question is where do the natural philosophers fit in? Socrates, Aristotle, Plato, all about a century before Ptolemy, by the way. So the answer science evolved over time is undoubtedly true, but not everyone on that list is equal when considering the question of quote, who invented science? So Deutsch did make a good case that Galileo deserves extra credit for having the right critical attitude, which is what really got him in trouble. But I would say that and maybe this isn’t going to be a popular opinion, but I’d say Newton looms large over the rest because he was the first nearly all encompassing and wholly calculable version of physics. And Newton, perhaps the first truly modern scientific theory was Newton’s. And then it stood the test of time for centuries after that unchanged up until Einstein.
[00:21:34] Red: So Newton perhaps deserves more credit than the rest. And this is an argument that Leslie Valiant makes in his recent book, The Importance of Being Educable on page two, fourteen. He says the case of Isaac Newton is noteworthy. His ability to create fundamental scientific theories is unsurpassed in human history. Though his though his theories. Through his theories, he was uniquely impactful, both in creating physical theories and in defining by example what later generations considered to be valid science. I wanted to emphasize that because I actually think that’s exactly correct, that in some sense, Newton maybe not intentionally invented science by creating the first true scientific theory that then everybody strived to emulate. They wanted all their theories to be like his, this highly empirical, highly calculable theory that makes amazing predictions using math. Right. So. There is now a shift in the historical record after Newton’s theory in the very essence of what people mean by science. Now, namely, they now often use the term synonymous with empirical theories due to the Newton, which is by the way, where where Popper gets it. Popper, I’ve argued that he talks about his demarcation being between science and non science. And I’ve argued he’s wrong about that because unless you understand the word science as meaning the set of empirical theories, which is what I think Popper did mean by it. OK, but that’s not what people generally mean by the word science. Science is a more encompassing thing that includes many non empirical theories like string theory, let’s say. So but because science started to become so strongly equated, at least the best of scientific theories, the goal of scientific theories was to be empirical. This is becomes a huge aspect of science after Newton’s theory.
[00:23:27] Red: Now, ironically, Newton, if we consider him as inventing science, he really did it as a side project to his real interests, which were alchemy and working out biblical prophecies like in the measure of David’s temple. Leslie Valin continues on page two fourteen. The science Newton is now famous for, however, occupied only a small part of his theory creating efforts. He applied much of his energies to areas such as alchemy and the occult that would now be considered bogus for from any academic standpoint. The voluminous manuscripts that he left on these subjects were little wanted, little read and generally embarrassing to his admirers down the centuries. These writers are these these writings are evidence based in that they put seven scare quotes in that they refer to experiments, historical data, the Bible and the dimensions of Egyptian pyramids. They include predictions of the end of the world, twenty sixty being one date he mentions. They include about a million words on alchemy. That’s page two fourteen. Yet Newton’s physics theories went on to be the standard of what good scientific theories must look like. How could the inventor of science literally fail so badly to understand what makes the kind of explanation that leads to progress? It wasn’t that Newton was uncritical about his alchemy and biblical theories. He still rooted them in evidence and they were explanatory theories in his mind, at least. Deutsch would say, I think, would say here that Newton’s false theories were easy to vary, so empirical evidence did not matter.
[00:25:04] Blue: Can I just make one point about that? I just read that book, The Ascent of Man by Jacob Bernowski. Which is just excellent. I think it’s one of the best science books I’ve read or the history of science books I’ve read so much that I could say about it. But they talk he talks about that this specific thing about Newton being into all this weird stuff. And but one of the points he makes about it is that even in Newton’s own mind, he saw a real clear distinction between defending his empirical theory theories using data and facts and all this and all these all these other occult kind of weird stuff he was into, too. I wish I could find the exact quote, but he. So, you know, I mean, I think at least based on that, it still seems that Newton had what you might consider a some sort of demarcation attitude and the attitude in line with critical rationalism, even even him.
[00:26:14] Red: So what was Newton doing wrong? And as I mentioned, Deutsch would say here that his theories were easy to vary. In episode 81, easy to vary versus versus ad hocness, I pointed out that it’s unclear what counts as easy to vary versus hard to vary. The criterion seems to be vague because no one, even amongst the crit rat community agrees on what it means. And I also argued that that criterion, while on the right path, tries tries to entirely root the criteria, the criterion in the nature of the theory itself, whereas in real life, it is just as much or maybe more to do with how the defenders of the theory treat the theory. Basically, if they’re willing to ad hoc, save it or not, whether they’re going to obey that convention and apply poppers ratchet or if they’re going to not obey it. Um, if you don’t agree, then take a look at the pseudoscientific theories of today. Let’s let’s go ahead and let’s talk about a couple. So if I were to say, Peter, give me examples of pseudoscience, you could probably give me quite a few, but ones that you might come up with might be psychic phenomenon or maybe let’s say near death experiences. OK, something along those lines. OK, would you disagree with either of those that those are pseudoscience? I know the near death experiences are somewhat more interesting, but I would still say say pseudoscience. OK, but these are absolutely scientific endeavors that have proved some great have produced some great legitimate science. OK, there’s a there’s a fantastic journal of near death experiences that is entirely well known, well respected scientific journal, for example.
[00:28:02] Red: OK, the thing that I would say here, like the near death experiences is kind of typical, you would find that some of the authors in the journal believe that near death experiences are views of heaven or something like that. But like many of them don’t, right? It’s they’re just trying to study the phenomenon of near death experiences in a scientific way. And the same could be said of psychics, right? Is there’s a legitimacy to just studying if. And ESP exists, that is a scientific question, right? Now, the real reason why you think of them as pseudoscience isn’t because it rightly so. By the way, I’m not actually saying you were wrong to say that they were pseudoscience is because we isn’t because we can’t do real science in those fields. Usually we do do real science in those fields and then we outright refute them using that science. It’s because each of these fields have a community built up around them that continually ad hoc saves these theories from refutation. This is who we are really thinking of when we think of these fields and refer to them as pseudoscience. OK, that is why I’ve argued that following the no ad hoc rule is actually more important than the nature of the theory itself. All easy to vary theories can quickly become scientific theories if you are required to only save them from reputation using independently testable theories. If you do that, your theory becomes hard to vary. And this is why I kind of downplayed it’s not that I actually think easy to vary or hard to vary are not useful concepts. But the reason why they’re vague is because it makes the wrong distinction.
[00:29:48] Red: It tries to root into the theory when what really matters is the way the defenders of the theory treat the theory, whether they’re willing to follow the no ad hoc save rule or not. OK, and even things that you think of as pseudoscience, all of them are actually legitimate science. There’s they’re all legitimate scientific questions. There are there is no pseudoscientific field that doesn’t have a legitimate scientific side to it, at least to the point of proving ESP doesn’t exist. Right. And refuting the existence of ESP. And you really only start to think of them as pseudoscience. Once a community of people who don’t obey the no ad hoc save rule grows up around that field.
[00:30:30] Blue: That sounds a lot like moralizing. Once you start kind of moralizing about your your favorite theories, it’s you’re almost inevitably going to start. I agree. I think trying
[00:30:41] Unknown: to try
[00:30:41] Blue: to ad hoc safe.
[00:30:42] Red: I think I think moral theories are very, very dangerous because of that. I I’ve I’ve always said they’re obviously all the good comes from moral theories, too. I’m not trying to say more or inherently bad. But I think I think we we often downplay the degree to which moral theories are just inherently dangerous.
[00:30:59] Blue: Yeah, but you can have a moral theory. To me, it seems like moralizing is is sort of a different different than just having a moral theory, maybe.
[00:31:11] Red: Obviously, we a different connotation. We wouldn’t exist without moral theory. Science wouldn’t exist without moral theories. You’re right. So I think we use the term moralizing when we see it as bad.
[00:31:22] Blue: But it’s often not clear.
[00:31:24] Red: I mean, the person who’s moralizing thinks they’re simply being moral. So I mean, this is like a totally separate podcast, how to deal with moral moral moralizing versus morals or something along those
[00:31:35] Blue: lines. That that could be good.
[00:31:38] Red: All right. So another common argument you hear is that progress comes from a culture of criticism. So David Dwight says in a tweet or somebody tweeted this, I don’t know where the quote comes from. The only admission card you need to this club of rationality is the willingness to expose your ideas to other people’s criticisms. Again, this is undoubtedly correct in some vague sort of way. So I’m not saying it’s wrong in any way. It’s it’s it is a true statement. But why does a culture of criticism lead to progress? I tend to get when I ask this question, I tend to get a vague answer here of something like, well, you criticize theories and then you error correct them so they continually improve. And OK, that is vaguely undoubtedly correct. But literally anything you count as a anything can count as a criticism. Presumably, we must somehow sort out more or less valid criticisms or this process wouldn’t work. Moreover, let’s take anyone that you can think of that you feel is not rational. Just think of somebody in your mind that’s not a rational person, OK, due to their unwillingness to expose their ideas to other people’s criticisms. And I’ve got a sincere question for you. Did this person believe they were uncritical and unwilling to expose their ideas to criticism or did they claim they were super critical and were constantly exposing their ideas to criticism? And in fact, drew conclusions only after carefully considering the strongest versions of other viewpoints. I don’t never once. I don’t think even once in my life have I come across someone, no matter how uncritical, that did not sincerely think they were willing to expose their ideas to other people’s criticisms.
[00:33:25] Red: So clearly, we need to make a culture of criticism by somewhat less somewhat less vague to figure out what it really means because everybody considers them part themselves, part of the culture of criticism.
[00:33:38] Blue: Well, to be fair, let’s not say everybody. I mean, there are there are people out there who will straight up want to kill you if you disagree with them.
[00:33:47] Red: You know, that’s probably true. But if you were to ask them, they would offer you an explanation and they would argue with you rationally. You know, why did you kill that person? Why do you want to kill that person? Right. Everybody always considers themselves rational. They always consider themselves that way.
[00:34:04] Blue: Yeah. Yeah. I mean, they always want to make an argument for their actions, I suppose, been impressed. You know, that’s why I like I like Hopper’s idea about about reason, is that really the core of reason is that we’re making argument where we’re letting our our our words do our fighting and dying rather than our actions.
[00:34:29] Red: Yeah.
[00:34:30] Blue: And in some ways that it seems to me the main the main the the theory that really is the the birth of of science and progress, you know, kind of like what we were talking about before is as soon as people enter a space where they’re willing to to to put their theories out there and defend them and and maybe maybe change them. And and, you know, that’s that seems to me the the the core of the of progress and reason and science and all that.
[00:35:04] Red: All right. Bruce Caldwell from Clarifying Popper. Excellent paper we’ve gone over several times on this podcast. There’s even a podcast specifically about his paper. He says of criticism when it comes to criticize criticizability, aren’t all theories criticizable? After all, even an uncriticizable theory can always be criticized on the grounds that it is uncriticizable. And just how is criticizability to be identified anyway? Are there any criteria by which the effectiveness of a criticism might be assessed? How does one decide when a criticism has been successful? Clearly, not all criticism should carry the same weight. Otherwise, a proof of logical inconsistency would have to be considered on equal footing with the following criticism. And then he goes on to give a cute little poem that basically says, I do not likely I can’t tell why, but I don’t. So the question we’re really asking needs a far more explicit and precise answer than merely we somehow criticize things and thereby improve the error correction. We need to know how to tell a good criticism from a bad criticism. And we also need to know how to tell what kind of theory is preferable. Moreover, we need to understand how those two statements are really one in the same. One answer that will not do is, well, we can just tell, presumably using our intuitions. Perhaps someone arguing this will point to the fact that we arguing that we do, we can just tell the point of the fact that we do make continual progress. So clearly, we must be able to tell good and bad criticisms for the most part.
[00:36:46] Red: But this answer is pretty unconvincing once you realize that before our very recent explosion of progress in open societies, we as a race had hundreds of thousands of years or even maybe millions of years of non progress. So if anything, the correct conclusion here must be that humans have generally poor ability to tell good criticism from bad criticism. And in fact, Newton himself apparently often couldn’t tell the man who maybe arguably invented science. So you often hear crit rats say that experiments are just a kind of criticism. And for the moment, let’s accept that it’s correct. And actually it is correct. It’s a little misleading, but it is correct statement. So I’m going to get into why I feel it’s a little misleading. So our questions, question must be as follows. How do we tell a criticism is valid and thereby detect an error or maybe more valid? I don’t know how to phrase that. When I was asking Brett Hall and the defenders of his theories of intelligence about how they use their theory to explain loss of cognition due to old age or to explain someone that was severely mentally challenged that couldn’t take care of themselves. They absolutely did not see that as a valid criticism of the theory worthy of error correction. When I level criticisms against crit rats that believe in praxeology, a theory rooted in ductivism and justificationism, I have yet to have even a single one of them say, hmm, that’s a valid criticism. Let me think about how to error correct my theories to deal with that criticism.
[00:38:16] Red: Instead, they always claim there is no error worth correcting or take the part of the section of the crit rat community that believes that it is a best explanation that animals don’t have feelings or qualia, not merely that they claim we don’t know. There’s one section that claims we don’t know and one section that claims that it’s a best theory. I’ve pointed out to them that there is a full body of experimental work on this that they aren’t even aware of, including experiments around animal grief that literally can’t be explained without reference to some or at least can’t be non -ad hoc explained without reference to animals actually feeling grief. And I’ve been told that there simply is no error in their theories to correct here. Those experiments are simply empiricism or justificationism or is another common claim I hear and thus wrong and can be dismissed without even looking at them. Now, this is hardly a problem unique to the crit rat community, though I’m using them as my example due to the fact that this is community community that spends considerable amount of time studying the epistemology of Karl Popper, yet I’ve never really seen them even be able to detect errors in their own theories, even what seems to me at least to be a fairly obvious case, like the ones I just listed. And undoubtedly each of each of these exchanges, if you were to go talk to the person involved, they would say, well, we offered criticism to Bruce and he claimed our criticisms are invalid and he couldn’t see the errors we were pointing out. So maybe I’m not better at detecting errors than anyone else. I mean, how would I know? That’s what I’m asking, right?
[00:39:53] Red: By that criteria, can I or anyone else tell when a criticism validly points to a real error worthy of correction? And even if we’re all bad at recognizing valid criticisms, at least we’re in good company with Sir Isaac Newton. So let’s look to science as the answer here. Or at least how science answered this question. I want to suggest that Popper sort of already answered these questions, though apparently even he didn’t quite fully get what he had discovered because of my criticisms of his chapter eight. Keep in mind that Popper’s main interest was working out why science makes such incredible progress. He barely put any effort at all into working out how to apply his epistemology to non scientific or non empirical theories. Big exception there would be politics. Obviously, he put quite a bit of effort into that. So it’s not surprising that his answer on how to generalize his theory to non empirical theories was, in my opinion, a bit of a misfire, though I would say it was at least on the right path. Go, Popper. Let’s ask the question this way. Why are scientific or rather empirical theories special? Popper’s answer answer was they are not unless you follow certain conventions. The two most important of which was one, formulate your theories to be testable. That means that they constrain reality via universal laws such that they are independently testable via predictions that can be falsified. Here’s a quote from Popper himself. Every good scientific theory is a prohibition. It forbids certain things to happen. The more a theory forbids, the better the theory is. That’s page 48, conjecture and refutation. Moreover, these experiments must be experiments that anyone can perform. This is going to become important later.
[00:41:45] Red: Only save the theory from refutations. So the second thing that Popper brings up is only save the theory from refutation by countering example in this case. If the new theory you’re introducing to save the original theory is itself independently testable. This is the no ad hoc rule. OK, if you follow those conventions, then Popper claims that your theories that there is something special about your theories now being empirical. So here, keep in mind, independently testable means the theory has consequences or reach unrelated to the problem you’re trying to solve. That we can then do an experiment on by testing those predictions. Popper claimed that you can do this, that to do this, you had to axiomatize your theories. That is his term, by the way. And I quoted that extensively in a past podcast where I showed where Popper talked quite a bit about how you’re supposed to axiomatize your theories as best you can and why that’s an important thing to do. Which means making them very precise and explicit, such that you couldn’t just accidentally or surreptitiously modify the theory on the fly to avoid refutation. That was why he made such a big deal about this. The word axiomatize is used by Popper multiple times and it covered that in episodes 81 and 82. By the way, based on these conventions, what can we learn about what makes, at least within science, a valid or at least preferred criticism? We’re here, I’m treating experiments as just a kind of criticism. So here are three things that I can see. Number one, the theory itself must be made very explicit and precise so that the theory can’t just keep slightly changing what it means to avoid refutation.
[00:43:29] Red: From Open Society Volume 2, page 218, Popper says, in order to avoid speaking at cross purposes, scientists try to express their theories in such a form that they can be tested, i.e. refuted or corroborated by such experience. Point number two, the criticisms, experiments in this case, because we’re talking about science, must be repeatable by anyone, i.e. what Popper calls intersubjective. Popper claims that this intersubjectivity is what makes science objective. Here’s some quotes from Popper on that, Logic of Scientific Discovery, page 22. I say that the objective objective of scientific statements lies in the fact that they can be intersubjectively tested or Open Society Volume 2, page 218. Everybody who takes the trouble can repeat it is what he says. Number three, even given the above two criteria, all theories are impossible to error correct without also accepting that all saves from refutation must themselves match the above two criteria. So Popper is often misunderstood as saying you must not shrink from refutations, like let your theory be refuted. You know, if somebody offers you refutation. Kuhn quickly pointed out that can’t be right because all theories have counter examples and thus are refuted. But this isn’t what Popper really meant. It’s actually a misunderstanding of Popper, in my opinion. Popper is actually quite open to even a dogmatic insistence on saving your theories from refutation. So long as you do so by following the conventions we just mentioned, particularly the no ad hoc rule and increasing the empirical content of your theories when you when you do the save. This is what I call Popper’s ratchet or solve all problems with your theories by increasing the empirical content of your theories. Which requires you to always axiomatize your saves from refutation.
[00:45:28] Red: Also work out testable predictions and then test that those theories independently that you’re using to save your theory. So what Popper’s so what was Popper’s answer to why science makes progress then? Scientific theories are not just any kind of theory. They are theories which where a choice has been made to frame the theory in a special explicit or even axiomatized way such that they make specific predictions. That we can find counter examples to and that we can completely independent of the specific problem we can test sorry independent of the problem that we were trying to solve. We can test these independent predictions. This is the same as saying the theory must not be ad hoc and experiments are not just a criticism. They are a very special case of criticism and objective or rather inter subjective criticism. They do not rely on you using your subjective intuitions to decide if the criticism landed for me or not. And a scientist is not just any old person advancing a theory. They are someone that has made the choice to participate in a community where every theory advanced even if it mean even if meant to dogmatically save their pet theory must be not ad hoc and independently testable as well. Here’s Popper page to 18 from open society volume two. Scientists try to avoid talking cross purposes. They were very they try very seriously to speak one in the same language in the natural sciences. This is achieved by recognizing experience as the impartial arbiter of their controversies.
[00:47:09] Red: When speaking of experience, I have in mind experience of a public character like observations and experiments as opposed to experience of the sense of more private or subjective, in other words, subjective experience in order to avoid speaking across purposes. Scientists try to express their theories in such a form that they can be tested, i.e. refuted or corroborated by such experience. OK, I feel like Popper is being very clear here, right? Like he’s explaining what it is that makes science special. It is that you participate in a community. In this chapter, he makes a big deal about the importance of the community of scientists. And I’m not going to quote all of that, but he basically says, I don’t think people can even do science unless they’re part of a scientific community. And he refers to science done without outside of a community as the same as revealed knowledge, right? Like revealed from God or something because it’s not really part of this community. I don’t know if I agree with him on that, but he goes that far when he’s trying to explain this. And the thing that makes the scientific community, according to Popper, so special is that they made this choice to let the arbiter be experienced, not subjective intuitions, but things that we can all intersubjectively look at, OK, and try for ourselves. Popper says, page two, 18, this is what constitutes scientific objectivity. This aspect of scientific method shows what can be achieved by institutions designed to make public control possible. It is therefore advisable to characterize science by its methods.
[00:48:49] Red: So Popper’s arguing that the the the reason why science is so special at making such great progress is because it exists in a community made up of institutions that allow public control and therefore we characterize science by the methods of that it used that are used to criticize theories. Specifically, through experiments that are intersubjectively testable. Is there any way to generalize these two points across the demarcation line between empirical and philosophical theories? Now, we’ve already made some progress here. Let’s list what we’ve learned so far. Science cares only or at least prefers a specific kind of criticism, ones that are objective, by which we really mean anyone can go intersubjectively perform the experiments for themselves. Science avoids criticism by intuition that is to say it avoids subjective criticisms, or at least it does not prefer them. But science also recognizes that a theory can only risk itself against objective, objective criticisms by formulating the theory in a special way, namely so explicitly and so precisely that we can apply outright logical refutations to it via its mis predictions from an experiment. And then, even with all the above, it is all for naught unless you also adopt the methodological convention of obeying Popper’s ratchet or obeying the no ad hoc rule by always formulating your saves via increasing empirical content. Now, here’s where things get interesting. Let’s take a closer look at Popper’s own example of his reputation of induction. In Conjecture’s Refutations, chapter eight, we covered Popper’s reputation of induction two episodes ago in detail, and we showed that it was a correct refutation. Popper criticized induction, or at least a certain explicit form of induction anyhow, on several grounds.
[00:50:46] Red: He showed that it was not credible, especially since observation statements are imprecise and error prone compared to Newton’s universal theory. He showed it was historically false that Newton derived theory from observation statements. Instead, he shows that Newton relied on errors in past theories. And Popper logically showed via a logical proof that it’s impossible to induce a theory from solely from observations. We also showed that inductive machine learning independently came to the same conclusions, basically number three there. This is called by machine learning experts, Tom Mitchell, the futility of bias free learning, and I covered that in detail in the same podcast. Mitchell showed that if that if you try to make your inductive learner bias free, i.e. that there are no assumptions about what the answer must look like, that it becomes impossible to generalize at all. But wait, induction is a philosophical theory. So by definition, it can’t be refuted, right? Yet everyone, at least in the crit rack community, knows about Popper’s refutation of induction in Conjecture to Refutations chapter eight. Critical rationalists never seem to think that this is a contradiction to Popper’s claim that philosophical theories can’t be refuted. We somehow accept both that Popper is correct that philosophical theories can’t be refuted, which he says in chapter eight. And also that Popper was able to refute induction, a philosophical philosophical theory in chapter eight. It gets worse. Popper does his refutation of induction for a specific purpose. He wants to show that it was impossible for Newton’s theory to be derived from observations, much less proven true by them. Popper isn’t content with merely saying, oh, well, my theory is more fruitful. And arguably it isn’t or that it is simpler. Again, arguably it isn’t.
[00:52:43] Red: He wants to be sure that he left induction or at least this explicit form of it, no room to exist at all by refuting it all together and showing that it was a myth. Conjecture refutations pages 70 induction, i.e. inference based on many observations, is a myth, says Popper. And logic of scientific discovery page 18. In my view, there is no such thing as induction. So Popper was able to refute induction, despite it being a philosophical theory. His refutation is not at all subjective or rooted in intuitions. It was a logical roof. Even his other refutations of induction had the character of not relying on subjective intuitions as criticism. For example, you can go check for yourself that observation statements take a different character than predictions from theories do, if nothing else, due to imprecise measurements. And you can go check for yourself that the history of how Newton of the history of how Newton came up with his theories and see if Popper was right about that or not. So is induction a philosophical theory? Popper tried to create three categories of theories that we covered in the last podcast. Category one was mathematical and logical theories. Category two was empirical or scientific theories. Category three was philosophical or metaphysical theories. And we quoted Popper in chapter eight. He claims that you can only refute or prove theories in category one. And he did leave the possibility of doing proving theories in category one. That you can only empirically refute theories in category two. And for category three, he asked, what reasonable argument can we deduce for and against a theory which we know to be need to neither demonstrate that it can be neither demonstrable nor refutable.
[00:54:31] Red: But he asked this right after logically refuting induction, I suppose, philosophical theory, this is the dead giveaway that Popper missed something important in this chapter. In fact, we just put our first crack into Popper’s demarcation criteria. Induction, or at least the version Popper refutes in chapter eight, is really a logical theory that can be refuted, at least when formulated explicitly as induction is the theory is the theory where you can induce a general theory from solely specific observations alone, which is what Popper was assuming induction meant when he refuted it. However, Popper’s refutation of induction does not work if you simply are a bit vaguer about what you mean by induction. And surely if induction just means somehow by some means you can generalize via some process that includes specific observations, but may or may not contain other specified stuff, then induction can’t be refuted anymore. And Popper’s refutation no longer applies to this far, far vaguer version of induction. Now, at the time I use this to show why there was no real inconsistency between the induction of Douglas Hofstetter and Popper’s epistemology. And at the time I said that there was at least two forms of induction and I treated this more general form as if it was legitimate. But now we’re looking at now that we’re looking at a bit more deeply, there seems to be something wrong with this more general form of induction. Namely, it is very, very, very vague. To put this another way, if one wants to defend induction, one must make a choice between offering it up as an explicit axiomatized theory that can then be refuted and thus risks refutation or offering it up as a vague theory that can’t be refuted.
[00:56:24] Red: But wait, isn’t that identical to what we said about what makes scientific epistemology special? How you choose to formulate your theories determines if you can be refuted or not, even for philosophical theories. You must either choose to formulate your theories, even philosophical theories, in a way that they are explicit and precise such that we know what an objective criticism of it would look like or your theory can’t be shown to have problems and thus can’t be error corrected. I can’t possibly overemphasize how important this is. Popper’s epistemology is about how you choose to formulate your theories from conjectures and refutations, page 48. Some theories are more testable, more exposed, reputation and others than others. They take, as it were, greater risks. This is the central aspect of Popper’s epistemology, how you choose to formulate your theories to take greater risk. All theories, including empirical ones, are constantly in danger of being formulated in such a way that the theory can no longer be refuted by counter example and thus can’t be error corrected with empirical theories. The big the big danger is ad hoc saving the theory, which Popper defined as adding an auxiliary theory to save your theory that is not itself independently testable or refutable. But now we see that philosophical theories work the same way. Induction can be explicitly formulated such that it is refutable logically or it can be vaguely formulated such that it can’t be the real reason why Popper’s reputation of in chapter eight isn’t perceived as quote refuting induction more generally is because inductivists save induction by making it vaguer, because it’s always possible for all theories to escape refutation by simply making your theory vaguer.
[00:58:22] Red: Let’s take a harder look now at Popper’s ratchet using Popper’s own words previously quoted liberally in past podcasts from Conjection of Refutations page 259 to 260. The better theory is the one that has the greater explanatory power that explains more, that explains with greater precision and that allows us to make better predictions. Let’s add a few more popper quotes to make this point clear. This is one that Brian Curd found and put up on your page. And I don’t know what the source is, but it’s a popper quote. The conjecture that it is the aim of science to find satisfactory explanations leads us further to to the idea of improving the degree of satisfactoriness of the explanations by improving their degree of testability, that is to say by proceeding to better testable theories, which means proceeding to theories of even richer content or higher degrees of universality, of higher degrees of precision. Yet criticism will be fruitful only if we state our problem. This is from logical scientific discovery in the introduction. Yet criticism will be fruitful only if we state our problem as clearly as we can and put our solution into a sufficiently definitive form, a form in which it can be critically discussed. Popper says this all over the place.
[00:59:43] Red: It’s strange that it’s it’s somehow I think we all kind of know it that Popper said these things, but we’ve missed that this is the center of his epistemology induction in the sense of you can derive general theories solely from specific observations is quite explicit and precise so it can be refuted induction in the sense of in some unspecified way we generalize and it includes specific observations somewhere in the process that’s vague, non explicit and imprecise so it cannot be refuted or really even criticized at all or discussed critically at all. But putting this another way, it appears that at least in the case of one philosophical theory induction, we can apply a form of Popper’s ratchet. In fact, as discussed in episode 82 Popper’s ratchet as well as 81 easy to vary versus ad hocness, I made the claim that this is true for many traditionally easy to vary or philosophical theories. I gave the example of spontaneous generation. Crit Ratz hold this theory out as a quintessential example of an easy to vary theory that can’t be refuted, yet science did in fact refute it by experiment, thanks to pastor and others. Let’s now discuss how to refute creationism, because we’ve brought this up a few times, I used to be a creationist and I keep mentioning that I refuted it, even though that’s supposed to be impossible. This is another classic example of a supposedly easy to vary or irrefutable philosophical theory. I was able to refute creationism back when I was a creationist like this. I kept taking the burden of explanation on myself because I intuitively understood that that was what was required of a rational person.
[01:01:22] Red: So I might start with the idea that really the earth is only 6,000 years old, just like the Bible says today, this is called young earth creationism. Young earth creationism is derided, but it’s actually a good example of a proper way to apply poppers ratchet. It’s a choice to frame creationism so precisely and explicitly that it’s possible to test it. And of course, it immediately gets refuted by taking measurements of the age of the earth.
[01:01:49] Blue: Creationism is a philosophical theory or an empirical theory. I mean, it sounds like you’re
[01:01:55] Red: I’m arguing that it entirely depends on how you choose to frame it.
[01:01:58] Blue: OK, so you’re basically applying poppers ratchet to creationism by trying to make it more sort of turning it into an empirical theory.
[01:02:09] Red: That’s right.
[01:02:10] Blue: OK.
[01:02:11] Red: So I might so now as a creationist, I might think to myself, oh, that’s a problem. But maybe the periods of creation in the Bible aren’t really are really eons of unspecified time. So notice how I’m unratcheting the theory. I’m making it less explicit, vaguer and less testable. But even this framing of a theory is still quite testable. For example, humans have actually been around for hundreds of thousands or years long of years or longer than where the Bible places Adam, the supposed first man, right? So Adam’s supposed to only be a few thousand years ago. So again, even in this unratcheted, less testable version, it still has testable consequences and I can still refute the theory. Now, sure, I could eventually keep unratcheting the theory and all move all the way to God created the Earth. And it’s a mystery how it happened. And I could, at that point, save creationism from all refutations.
[01:03:07] Blue: I feel it in my gut.
[01:03:08] Red: Yes, that is to say, I could make the theory so vague and so non explicit and so imprecise that it was impossible to refute or criticize the theory at all. But this to me just felt like an irrational thing to do. So I eventually realized that, at least for my sake, I had effectively refuted creationism scientifically and moved on. And I did refute creationism because creationism can be formulated in a refutable way through the whole. Now, I did admit that the whole incident left me with the deep sneaking suspicion that evolutionists were actually irrational to and only accepted evolution on faith due to how they consistently ad hoc saved their theories from refutation rather than taking the problems of the theory seriously by making their theories increasingly vague. And in episode seventy seventy four, the problem of open -endedness, I gave examples of evolutionists unratcheting their theory to save it from refutation. But I could see at least that evolution had one big advantage over creationism. Some of it was quite testable. But I could find no explicit framing of creationism that didn’t immediately get refuted by observation. And I was going to need something like a theory of evolution to explain life as we see it because it’s a mystery, explains everything, and thus explains nothing. So here is Popper on the importance of how we frame theories to be risky again. Irrefutability is not a virtue of a theory as people think, but advice. Page 48, Conjection of Refutations. Some theories are more testable, more exposed refutation than others. They take, as it were, greater risk. And then after saying that, speaking of astrology as a negative example, he says, by
[01:04:54] Red: making their interpretations and prophecies sufficiently vague, they were able to explain away anything that might have been a refutation of the theory, had the theory and the prophecies been more precise. In order to escape falsification, they destroyed the testability of their theory. It is the typical Seuth sayers trick to predict things so vaguely that the predictions can hardly fail that they come that they become irrefutable. That would be page 48, 49. And then in myth of the framework, he says, and we choose to set ourselves standards of explanation and methodological rules, which can help us to achieve our goal and which it is not easy for any theory or framework to satisfy. Of course, we may choose to not do this. We may decide to make our ideas self -reinforcing. We may set ourselves no task other than the one we know our present ideas can fulfill. I’m going to argue this is precisely the problem with me asking the crit rats, defending Brett’s theory, Brett Hall’s theory of intelligence. When I asked them to explain the cognitive decline of age, like Joe Biden, in terms of their theory of interest in time spent learning a subject and having them not take that as a serious problem with a theory to resolve, but instead replying fairly consistently, well, IQ theory is an explanatory as if the problem of their theory is somehow mitigated by a what about isn’t. I even showed. So what I’m trying to argue here is that you make a choice. You really do, right? Every theory can be defended in this way.
[01:06:37] Red: If you’re going to just do a what about isn’t or you’re going to just make it increasingly vague, OK, you have to actually that none of them, not a single one. I asked multiple ones when I said, how do you through your theory explain age? Cognitive decline due to age, not one took up the challenge of trying to figure out how to apply their theory to that case and instead tried to start to talk about defects of the what they saw as the competing theories. They simply did not take these problems seriously. And instead, they wanted to, as Popper is saying here, they want to make their theory self -reinforcing. They set no task other than the one they know their present theory can fulfill is to quote Popper there. I even show that David Deutsch’s own example of easy to vary of an easy to vary theory, the story of Persephone’s to explain the seasons fell immediately to empirical testing. This is back in the podcast on easy to varyness versus ad hocness that so long as you’re willing to follow Popper’s ratchet by disallowing ad hoc saves the theory that even the myth of Persephone’s immediately falls to empirical testing and that’s exactly what happened in real life. It’s these theories could fall to empirical testing by simply taking the no ad hoc rule seriously, because then you have to offer a save to finding out that Australia has the opposite seasons, not by simply modifying your theory to save it, but by modifying it in a way that’s independently testable, you take that idea seriously. The story of Persephone’s, it’s a myth, it’s completely philosophical, and it can also be empirically tested.
[01:08:21] Red: So long as you keep it explicit and you follow the no ad hoc rule. So Popper’s ratchet appears to have more general use than merely empirical theories. And this is great news. It means we should be able to come up with a generalization of Popper’s ratchet that applies to physical to philosophical theories. So let’s do that. How do we apply Popper’s ratchet to philosophical theories? So we’re going to use Popper’s refutation of induction in chapter eight as our kind of guide in our example. Popper produced a proof that it was impossible to derive general theories from from specific observations because there can never be a logical inconsistency between two observations without a theory to connect them. This is from two episodes ago. This is clearly not an experimental refutation, but it does have something like the same quality of an experimental refutation. What Popper does is he starts with a very explicit and precise formulation of induction, and then he shows that it results in a logical contradiction. An experimental refutation is also a kind of logical contradiction. So we’re on to something here. Instead of thinking of a refutation as an empirical outcome counter to the theory predicted, let’s generalize to the notion of a refutation is allowed to be any kind of logical contradiction. As Popper put this in Myths of the Framework 51 to 60, examine and challenge a theory’s consequences, its explanatory power, its consistency, and its compatibility with other theories.
[01:09:54] Red: You work out what the implications of your theory are and you try to see if it results in this case, we’re going to start with just the idea that if it results in a contradiction or not, we’re going to generalize that a bit more in a second to other kinds of problems. But let’s start with just this. OK, we now have a generalization of Popper’s ratchet that will apply to something like induction. That statement I just made, examine and challenge theories consequences of explanatory power, its consistency, and its compatibility with other theories. Popper did say that of scientific theories. What I’m arguing is that it applies equally well to many philosophical theories. And so we’re going to use it as the basis for generalizing Popper’s ratchet.
[01:10:35] Blue: And can you define? Sorry, if you already did this, but can you define philosophical theory? Because if we’re defining philosophical theory as a non -empirical theory, but then we’re trying to make these theories more empirical, it kind of screws up the definition of what a philosophical theory is, in a way.
[01:10:58] Red: Yes, it does.
[01:11:00] Blue: OK.
[01:11:01] Red: So one of the things that I am demonstrating is that there is, in fact, something wrong with Popper’s demarcation criteria. OK. The fact is, is that creationism can either be a philosophical theory or an empirical one, depending on how you want to frame it, right? Induction. I think everyone would say it’s a philosophical theory in that it’s not an empirical theory,
[01:11:24] Blue: but
[01:11:25] Red: it can be framed as a logical theory, which can be refuted, even according to Popper, OK, by showing logical consequences or contradictions.
[01:11:34] Blue: So basically, we need a better definition of philosophical theory.
[01:11:38] Red: So I
[01:11:38] Blue: think the
[01:11:39] Red: basic idea is of there’s a difference between a theory that can be empirically tested and those that can’t is maybe not such a bad distinction. So from that point of view, I can buy into Popper’s demarcation criteria. And induction clearly isn’t an empirical theory in that sense. But it can be refuted by logical contradiction. You can formulate it explicitly, which Popper did. And you can show that it results in a logical contradiction. So that’s not still not an empirical theory. So I would still call it a philosophical theory. If you want, you could say, well, it’s no longer a lot of philosophical theory. It’s now a lot. It’s a now a logical or mathematical theory. OK, that that’s fine. Popper does make a third category like that. But I think that’s what I’m saying is that take every single one of your theories, I don’t care if they’re empirical or not, and make them explicit such that it’s possible for them to have logical contradictions. Let’s just start with that basic idea that Popper demonstrated very well with induction, even though we normally think of it as a philosophical theory.
[01:12:45] Blue: OK.
[01:12:46] Red: And yes, this is putting cracks in the demarcation criteria. It’s showing that there’s the demarcation criterion needs an overhaul.
[01:12:54] Blue: Yeah. Well, I kind of like how you put that in a theory that you would refute using logical contradictions, broadly defined, I guess.
[01:13:05] Red: Yeah.
[01:13:05] Blue: Rather something that you would refute in a laboratory or something.
[01:13:12] Red: Yeah. OK, well, let’s let’s challenge this, though. OK, so I’m saying let’s start. We’re starting with this idea of a logical contradiction. But I’m going to I’ve been challenged on this by people I’ve been talking to. So let me give an example. OK, so let’s bring up this idea that I’ve mentioned past podcast number of times. I’ve given the example of me asking Sadiah to give me a refuting example to the Church -Turing -Deutsch thesis. So Sadiah feels that there’s something wrong with the Church -Turing -Deutsch thesis. She actually talked about this in several four podcasts that we did with her, and she covered that in quite a bit of detail. So I asked her at one point, give me a refuting example. And Sadiah paraphrasing Lee Cronin and David Chalmers and other respected scientists and philosophers suggested Kualya as a refuting or counter example that falsifies the Church -Turing -Deutsch thesis. Now, I don’t want to keep picking on Sadiah here, especially since this was only one of her two arguments and the other being the problem of time and not even the more important of the two. And she was just using a very common, well -known criticism of computationalism that originated, I think, with David Chalmers, who’s a pan -psychist. Sadiah is not a pan -psychist, but David Chalmers is. OK. And the argument goes something like this from David Chalmers. Church -Turing -Deutsch thesis says everything can be modeled or simulated as as computations, but this means Kualya must also be computation. But how can that be? Kualya is a feeling and every single individual step of the computation has no feeling or Kualya. So this is a logical contradiction.
[01:14:49] Red: So David Chalmers would say that Kualya is a logical contradiction to the Church -Turing -Deutsch thesis, and thus he has refuted it. OK. So can
[01:15:00] Blue: I just say one thing? I don’t think David Chalmers is a pan -psychist exactly. I think he’s just he might have some sympathy for that.
[01:15:07] Red: That’s probably more accurate than sympathy for it. He like that’s like the specific like I’ve seen him talk and he always says science will come around to this and maybe it’ll be something like this. And he always brings up pan -psychism. Yeah,
[01:15:19] Blue: but he’s not he’s not like an outright pan -psychist.
[01:15:22] Red: Yeah.
[01:15:22] Blue: Yeah.
[01:15:23] Red: He’s more open, I would say. I think I agree with that.
[01:15:26] Blue: OK. OK.
[01:15:27] Red: So but that seems to be where he thinks it needs to go is that we’re going to have to explain consciousness in terms of being something that’s an atomic, you know, at the atomic level that everything is conscious.
[01:15:38] Blue: OK.
[01:15:40] Red: So even declaring that we’re looking for logical contradictions appears to not be enough because we can apparently disagree over what counts as a logical contradiction. Or can we? So I definitely think this is where Saudi was coming from. She saw it as a logical contradiction. I’m going to challenge that idea, though, even though I just tried to frame it as a logical contradiction and tried to steelman it as a logical contradiction, I want to point out that there is something different from what we would normally think of. So let’s apply Popper’s ratchet in this in this finding. We know what we care about our logical contradictions and that logical contradictions can apply to even some philosophical theories, such as induction, if they are formulated explicitly and precisely enough. But does Chalmers’ use of qualia count as a true refuting case or logical contradiction? The problem is that it’s not obvious if this is actually a contradiction or not. So now I offered to Saudi that maybe we just don’t know yet what the algorithm for qualia is. Now, I don’t remember exactly how Saudi responded to me asking that question, but I have asked that question of other people who believe similar things and typically the answer
[01:16:54] Blue: is something like this.
[01:16:55] Red: They usually turn the question around, said something like, well, until you can offer me the algorithm of qualia, I consider this a refutation of Church -Turing -Deutsch thesis. Now, is that a fair response? If they instead said, I see the legitimate problem, that would be a fair response. Without a doubt, it is a problem that we can’t currently explain qualia in terms of computation. OK,
[01:17:20] Blue: it’s an interesting response.
[01:17:21] Red: Yeah. But I mean, the fact that the Church -Turing -Deutsch thesis can’t today explain qualia, that really is a true legitimate problem with Church -Turing -Deutsch thesis, right? Surely that’s true. But who gets to decide what our null hypothesis is? And that’s the problem with this reply. And why doesn’t the highly corroborated theory, QM and Church -Turing -Deutsch thesis, in this case, get to be the null hypothesis instead of the vaguely formulated encyclism or whatever vegan, untestable theory, alternative theories being offered as a competitor? OK. So there’s something amiss here. Herein lies the problem with Chalmers argument. Chalmers feels that it’s intuitively obvious that qualia is a contradiction to any view of reality, to any view of reality that everything can be described via computations. But Daniel Dennett has the exact opposite intuition. He feels that it’s rather intuitively obvious that Chalmers intuition is wrong and that instead of course, qualia will turn out to be computational. While these supposed refutations are being worded as logical contradictions, it turns out that it’s a subjective matter of opinion, whether or not they are actually a refuting case or a logical contradiction. If qualia is a logical contradiction to Church -Turing -Deutsch thesis, we should be able to write out and show the logical contradiction in logical syntax, but that would require an axiomatized version of qualia, which we don’t have. And in fact, their their claiming can’t exist. So instead, the claim that this is a logical contradiction is really just a gut feeling, and that’s all it really is. Ditto to Dennis Hackathol offering up animals making automatic movements or mechanical movements as being a contradiction to animals having feelings. It just comes down to a gut feel.
[01:19:17] Red: You either have the gut feel that this is a contradiction or you don’t. This is literally the opposite of what a refuting counter example is under Popper’s epistemology, which is an objective counter example to a theory that anyone can check for themselves by experiment. This is the inner subjectivity. Suppose that I have no intuition of my own here. How would I test between Chalmers and Dennett’s opposing intuitions? It would appear I can’t. If Chalmers is right, and perhaps he is, his current argument is literally inaccessible to Daniel Dennett, because Dennett happens to not have the same intuition Chalmers has. Ditto here for anyone that happens to have an intuition that automatic movements in an animal is not a contradiction to the idea that animals have feelings. How might we make a convention to resolve this problem? Let’s take a look at Popper’s epistemology for an answer here. Hopper has already solved this problem on the empirical side of the demarcation boundary. Hopper simply required that all refutations are intersubjective. That is to say, if you want to offer up a refuting observation by experiment, the experiment must be repeatable by anyone that can check it. If you can’t do that, it doesn’t count as a refuting observation until you can figure out how to make it intersubjective. In logic of scientific discovery, page 22, he says, I say that the objective objectivity of science, scientific statements, lies in the fact that they can be intersubjectively tested. I’ve quoted that previously just to emphasize them. Let’s now apply this to Chalmers theory. I offer up the following convention, which is clearly just a slight generalization of Popper’s already existing convention requiring inner subjectivity.
[01:21:02] Red: It is as follows, a criticism only counts if it is a non -subjective, non -intuition based criticism. The criticism must be checkable by anyone. In fact, I got this from Popper himself, and we’ve already done some of these quotes, but let me do them again for emphasis. I say that the objectivity of scientific statements lies in the fact that they can be intersubjectively tested. Then he goes on to say on the same page, I have since generalized this formulation for intersubjective testing is merely a very important aspect of the more general idea of intersubjective criticism, or in other words, of the idea of mutual rational control by critical discussion. That’s on page 22 of logic of scientific discovery in the footnotes. Now, compare Popper’s disproof of induction to Chalmers’ supposed disproof of Church -Turing -Deutsch thesis. I can go work out Popper’s proof for myself and see that it is correct, at least for the specific formulation of induction that Popper laid out precisely and explicitly, the same just is not true for Chalmers supposed refutation of Church -Turing -Deutsch thesis. It requires me to happen to hold a certain subjective intuition first with no possible way for me to check things for myself if I happen to hold a different intuition. I want to make an observation here about human beings. Human beings really badly seem to want to argue using their subjective intuitions to determine what counts as a valid criticism or not, or a good theory or not. You hear all the time statements like that criticism didn’t land for me. The crit -rat community has practically formalized the belief that what sounds like a good explanation is in fact a good explanation.
[01:22:50] Red: And they don’t even, in my opinion, bother any more with little things like if the theory is not ad hoc. I actually saw a thread where I saw crit -rats arguing over whether falsification was part of Popper’s epistemology or not because they had changed this idea. They instead said, well, no, it doesn’t really matter if it’s falsifiable or not. It’s really just if it’s hard to vary or not. And they almost turned this into a formalization that it’s really about maximizing explanation and therefore it’s a matter of, does this explain well? The thing that we’re trying to, the problem that we’re trying to solve and have really turned that into just subjectively, does it seem to me it’s a good explanation? And I’m not sure much else is going on in a lot of the theories I’ve seen with the crit -rat community offer up. We sincerely seem to believe that we are truth detectors or at least error detectors. We put amazing amounts of faith in our ability to detect a good explanation or a good criticism of a bad explanation. But what I want to suggest is that actually the opposite is true. Human beings are very, very, very bad at detecting truth, error, or good or bad explanations using their subjective intuitions. And Popper’s epistemology, when rightly understood, is really about really this amazing accomplishment, because Popper was the first to formally work out how to go about, how we go about in science, not having to rely on our subjective intuitions to detect error. Or to put another way, Popper’s big discovery is that he determined that some kinds of criticisms, objective criticisms matter more than other kinds of criticisms, subjective criticisms. We might even frame this convention as follows.
[01:24:42] Red: While you are never welcome to you, you’re always welcome to use your intuitions as much as you want to come up with interesting conjectures. You are never allowed to count subjective intuitions as criticisms or reputations. The only criticisms that are considered valid are the ones that result in objective criticisms that anyone can check for themselves. Popper’s refutation of induction qualifies, Chalmers does not. What shall we refer to this new convention as? I’m tempted to simply call it testability, because I think that’s the best word for it. But that term is already taken by Popper and refers specifically to empirical testing, so I’m going to call this checkability. But what I really mean, just mean, is a generalization of testability that now includes any sort of object of objective criticism that anyone can check for themselves, and they don’t have to rely on subjective gut intent intuitions to make make the criticism. This is a bold claim, so I’m going to word it as boldly as as possible. I’m claiming the following. When applying criticisms to theories distinct from when you’re choosing conjectures, objective criticisms are far more important than subjective criticisms that rely on needing a specific intuition. I’m very tempted to word this even more boldly as you can completely discount all criticisms that rely on gut intuitions and are thus subjective in nature. And in fact, I actually think that with some caveats that I would have to explain and we’re not going to in this podcast, that’s actually correct. For now, I’m not going to go that far. There’s some problems with going that far that need some explanation. And so maybe we can revisit that later.
[01:26:21] Red: But either way, this is a very harsh standard, because if you pay attention nearly every criticism in every typical internet argument relies almost entirely on intuitions and subjective feelings. But wait, didn’t we say popper’s refutation of induction didn’t apply to the more general form of induction? So what is to stop someone from saying, OK, you refuted this specific formulation of induction, hopper refuted this specific formulation of induction. But I disagree that formulation that is the correct formulation of induction. I think induction was never about creating theories solely from observations. And in fact, that is exactly how real life inductivists respond to popper’s refutation of induction. They just claim popper is straw manning them and offer a vaguer theory of induction and say popper didn’t refute that in previous podcasts. I acted as like this more fuzzy or general kind of induction was valid in some way, but now we’re in a position to see that there is something really wrong with this more general form of induction. The real reason it is not refuted is because it’s so vague that it can’t even in principle be anything meaningfully said about it. What we need is not only a generalization of testability, which I’m now calling checkability, but also a generalization of the no ad hoc rule. Recall that the no ad hoc rule is that you’re not allowed to save your theories from refutation via introducing an auxiliary theory. Unless the new auxiliary theory is itself independently testable and has been corroborated via testing, we can now apply the same idea to philosophical theories. We might. How might we take the basic ideas of the no ad hoc rule and generalize them to non empirical theories? The induction example is instructive here.
[01:28:03] Red: Those that say to popper, I disagree with you. Formulation of induction should be expected to offer their own formulation of induction that does allow creation of new generalizations. This is what I’ve called the burden of explanation. They must take that burden of explanation on themselves by offering their own non vague, equally explicit version of induction that allows for general aid, generalization from specific observations. Likewise, I see the same thing with the whole Chalmers argument. They have the burden of explanation, not me, right? Because that’s the problem is that they need to explain it themselves. They need to offer a theory of qualia before they get to use it as a refutation of the competing theory. That’s what they’re not doing. OK, it isn’t enough to say, unless you show me the algorithm for qualia, I consider your theory refuted. They have the burden of offering us a testable theory that both explain the success of the current best theory, including the Church -Turing Doge thesis and also explains qualia, what while the Church -Turing Doge thesis still can’t, that’s what offers epistemology requires of them. OK, inductive philosophers mostly just refuse the burden of explanation and try to leave induction a vague, non explicit, non -precise concept that can’t be checked at all. We need a term for this, so I’m going to dub it vague manning your theory. The beauty of vague manning a theory is that not only is your theory irrefutable, but you also get to claim all attempts to refute it, which require making it more explicit, are straw manning your theory. This is one of the reasons why a claim of straw manning my theory may or may not be valid.
[01:29:48] Red: You have to actually look at the person’s willingness to make their theory explicit. Now, let me give you an example of the right way to go about this. OK, so Popper offers an explicit form of induction saying you somehow generalize solely from observations. He shows that’s impossible. OK, he refutes it, logically refutes it. OK, if you’re an inductivist, what you have to do is you have to come up with your own equally explicit version of induction that does allow for generalization from observations. That’s what Tom Mitchell did. OK, that’s why I gave the example of Tom Mitchell in the same episode. He took that challenge seriously. And the end result was something known as computational learning theory that precisely works out under what conditions and circumstances you can generalize from specific observations, given certain starting assumptions or starting theories, in other words. Now, this new form of induction that’s being offered, we can now see it’s not at odds with Popper’s epistemology, because it’s so explicit we can now understand, oh, this actually fits Popper’s epistemology just fine. OK, this form of induction, we can still call it induction. I don’t care what we call it is not at odds with Popper’s epistemology anymore because it used Popper’s ratchet and came up with an explicit version that is not refuted by Popper’s refutation, but also now is compatible with Popper’s epistemology. So let’s think of this as a third kind of induction, which I suspect many Popperians will argue is induction in name only. OK, but let’s not get caught up with words because words don’t matter. What we call it doesn’t matter. Let’s call it computational learning theory if you don’t want to call it induction.
[01:31:34] Red: OK, it’s not the induction of philosophers anymore, which is vague. It’s not the induction that Popper refuted, which is explicit and refuted. It’s really its own thing completely compatible with Popper’s epistemology. This is the generalization of the no ad hoc rule we’re seeking. You are not allowed to save your your your refutation of your philosophical theories by making your theory vaguer. You’re only allowed to save your philosophical theory from criticism by making it more explicit and more precise. So what is the philosophical equivalent of a refutation at a minimum? Some other objective problem we can all go check for ourselves and doesn’t require subjective intuitions. Preferably, it will be a logical contradiction of some sort, but it may not be. There are other things that may qualify. So let me give let me give an example of this from the church Turing thesis itself. OK, so see episodes nine to eleven of this podcast, if you’re interested in that. The church Turing thesis states that all computing machines that you could possibly conceive will turn out to be equivalent to a Turing machine. Now, we later did have to amend that to the church Turing Deutsch thesis and say, well, it’s the quantum Turing machine. But the same principle applies. And I’m for simplification’s sakes, I’m usually going to just say Turing machine, but I mean the quantum Turing machine. Is this an empirical theory or is this in a philosophical theory? It really can’t be said to be an empirical theory per se, because there’s no experimental test you can do to refute it. So it must be a philosophical theory, right? But then
[01:33:14] Red: this is a great philosophical theory because it is so precise and explicit that we know exactly what a counter example or contradiction would look like. A counter example to the church Turing thesis would be a computing machine that exceeds the power of a Turing machine. And it just so happens that no one can find such a machine. So the church Turing Deutsch thesis may well be a philosophical, that is to say, non empirical theory, but it is just as risky as any empirical theory because the theory explains exactly what an objective counter example would look like, even though none exist. We just put another major crack in Popper’s demarcation criteria because it literally does not matter if you want to frame the church Turing Deutsch thesis as scientific or philosophical. What matters is is that it’s so precisely defined that we know exactly what a counter example would look like. So let’s make this convention a good explanation, whether or not it is philosophical or empirical is one that is so precise and explicit that we know what an objective problem would look like, such as such as a counter example to it. Even if we can’t find such a counter example, I can I know in case I’m not being clear, let me just explain this one more time. I cannot tell you an example of a machine that exceeds the power of a Turing machine. Nobody knows how to do that, right? Because it may presumably because it doesn’t exist.
[01:34:49] Red: But I I would immediately if someone showed me a machine that was more powerful than had more power than a Turing machine, it would be easy for me to go check it for myself, see that that’s correct, that everyone just missed it and then say, oh, you know what? We just disproved the church Turing thesis. That’s the power of a good explanation is that it’s risky like this. But to be risky, you have to be so explicit that it’s obvious what a counter example or a contradiction or an objective problem would look like. This is the correct understanding of checkability. A checkable theory is one that is so precise and explicit that we know what a counter example or contradiction to it looks like. We can now formulate a generalization of Popper’s ratchet that applies to both empirical and at least some philosophical theories. Here is this the revised generalized version. Formulate your theory so explicitly and precisely such that they can be objectively checked by anyone via objective problems such as logical contradictions or counter examples and then only save your theories from contradiction by making your theories more explicit and more precise and thus easier to understand what a contradiction or objective problem would look like. In other words, make it more checkable. This is the generalization of Popper’s ratchet that we’ve been seeking and it no longer applies solely to empirical theories and can also apply to at least some philosophical theories. So does this new version of Popper’s ratchet apply to all philosophical theories? Maybe it only happens to apply to weird cases like induction and the church Turing thesis, right? I think the short answer is no, it does not always apply to all philosophical theories. Here’s the thing that
[01:36:34] Red: people who ask this might be missing, though. It is up to you to figure out how to create to creatively formulate your theories so that they are checkable. If your philosophical theory isn’t checkable, that’s a problem. And you need to see it as a problem. OK, this is what Popper’s been saying. So when we try to compare Popper’s refutation of induction, which was very explicit to an adductivist who’s trying to save the theory and his theories very vague, what would a counter example to their form of induction even look like? Right? Like, what does it even mean? Like, it’s so non explicit, you have to immediately realize that’s a problem. That’s why we don’t care about that theory and why Popper’s is the better theory. OK, and you can say it’s straw manning all you like. Doesn’t matter, right? Popper straw man didn’t he were still refuted an explicit form of induction that now is impossible. And it’s really up to the inductivist to figure out, OK, how can I reformulate induction so that it is something explicit? If you do that, you end up with machine learning, not with what the philosophers thought you could do, but machine learning simply does not allow you to. You still have to have starting theories under machine learning. OK, the inductive bias as Tom Mitchell would would call it. So it’s not that this applies to all theories. It’s that you have to choose to formulate your theories so that it applies. That’s the thing that people keep missing. There was someone on the critical rationalist Facebook page that was asking a question about how to argue with irrational arguments.
[01:38:14] Red: Now, no matter how you answered him, he simply responded with increasingly vague theories that you responded and apply to underlying his approach was a tacit assumption that that if made explicit would be as follows. How do you always determine which of two competing theories is correct? The answer is you don’t. I think the answer to this question, even on the scientific side, even on the empirical side of the demarcation boundary line is most of the time, you simply do not know which theory is the better theory. It’s very difficult to come up with an explanation that is at once so explicit and with so much reach that it literally subsumes all its competitors. And anything short of that likely means that likely the competing theories have some different pros and cons compared to each other and that none is the clear cut winner. So Popper talks about this in the framework, page 161, he says, thus the result of a scientific discussion is very often inconclusive, not only in the sense that we cannot conclusively verify or even falsify any of the theories under discussion, but also in the sense that we cannot say that one of the theories seems to have definitive advantages over its competitors. That’s kind of the norm. It’s like it’s it’s rare. You have a best theory, right? Being a best theory is this objective thing. It’s got nothing to do with how well you in your gut feel this theory explains things well, right? It’s got nothing to do with that. It’s an objective thing where it literally subsumes its competitors while being increasingly checkable. OK, whether it’s checkability includes testability.
[01:39:50] Red: So it sort of doesn’t matter that this new formal poppers ratchet isn’t going to solve all your philosophical problems. And again, I emphasize the burden is always on you to choose to formulate your theory such that poppers ratchet can and will apply. But it should, if correctly applied, allow you to immediately see which theories within the crit rack community are not best theories. Don’t confuse this with not being true theories. A theory could be true, but could be currently formulated very poorly, very vaguely, OK, but ultimately, there will be some more explicit version that’s true. OK. We really kind of don’t assess the truth of theories. We assess how good the theories are. And it seems weird. And this is why instrumentalism seemed true to so many people. You’re really just assessing how checkable is this theory, how possible is this theory compared to its competitors? It hasn’t been refuted or not yet. There is nothing wrong with using your subjective intuitions to have a favored pet theory. In fact, you probably can’t help it, right? You’re just going to do it. But there is something wrong with trying to raise your pet theory to best theory status in the Popperian sense. If it doesn’t deserve that status status at this time. So all my problems with the correct community, and I do mean all of them, always boil down to arguments with them over various pet theories that they claim are best theories that really just don’t qualify as best theories. And
[01:41:13] Red: all my discussions, they always think I’m arguing about the theory they’re interested in and always place me in the in the role of believing in the compete competing theory, but I’m really just trying to say, no, that’s not a best theory yet. Like, yes, I get it that you have reasons why you have a preference for this theory, but this theory does not qualify as a best theory until it has certain objective criteria that you’re not matching at this point. OK. The fact that. Brett Hall’s theory of intelligence can’t explain cognitive decline. Whereas G factor can, OK, at least very, very loosely. That’s an example of how his theory just did not qualify as a best theory, because it doesn’t subsume its competitors yet. OK, a best theory to Popper was something quite objective. It wasn’t just in a subjective matter of opinion based on one’s intuition that this theory landed for me or this criticism lands for me. Or even this is the sole theory remaining after I personally tried criticizing the perceived competitors as best I could, which means very little because you are the easiest person you know how to fool. A theory only becomes a best theory to Popper if it subsumes its competitors to the point that we don’t really need the other theories anymore. Method of the framework page 60, the correct method of critical discussion tries to replace all our theories by better ones. It’s hard to see how Brett’s theory of intelligence subsumes its competitors when it can’t explain Joe Biden’s cognitive decline in terms of interest in time. This brings us to the to the answer to the question I keep asking. How do you tell a good philosophical explanation from a bad explanation?
[01:42:52] Red: It’s important to keep in mind that we are not asking you to tell how to tell if the explanation is true or not. This is a purely epistemological question about the nature of the theory itself and how it is formulated or framed, whether it’s true or not, a different question. Surprisingly, induction, at least the version Popper refuted, is very much a good explanation, precisely because it was framed such that it could be refuted. It’s just not a true explanation. But the more general form of induction, the one that we keep treating as if it’s OK, it’s valid, because it’s not refuted, is actually very much a bad explanation because there’s no way to check it. However, Tom Mitchell’s version of induction is both a good explanation and is also not refuted yet, but it’s simply not the kind of induction that Popper refuted either and it’s compatible with Popper’s epistemology. So let’s formulate things like this, a bad philosophical explanation is a bad explanation for the same reason any explanation is bad because it relies on your intuitions rather than on being formulated such that we can check it. A good philosophical explanation is a good explanation for the very same reason any explanation is good because it relies on an explicit and precise formulation of the theory such that we can reasonably check various independent implications of the theory. So how do we tell a good philosophical explanation from a bad explanation? The answer is exactly the same way you do with empirical theories via Popper’s ratchet. Do not vague man your theories. Do not root your theories in subjective intuitions.
[01:44:30] Red: Instead, frame your philosophical theory such that they make prohibitions such that it is obvious what a problem or counter example or contradiction would look like. Make your theories bold and risky and thus error correctable. So again, a few more quotes from Popper. A couple of these I think I’ve already done, but kind of apply in this place. Conjecture refutations 48, some theories are more testable, more exposed refutation than others. They take as or greater risks. That’s what you’re trying to do. By contrast, the correct method of critical discussion starts from the question. What are the consequences of our thesis or our theory? You’re trying to work out the implications of your theory. So your theory has to have actual real life implications that can be checked. Critical discussion tries to test the theory under discussion by finding out whether its logical consequences are acceptable. That’s all page 60 for method of the framework. Then only solve problems with your theory by making them increasingly precise and explicit, such that it is even easier to tell what a problem would look like. This is our generalization of Popper’s ratchet. And that is what I believe is the correct way to treat theories, metaphysical theories to apply Popper’s epistemology to metaphysical theories. So, Peter.
[01:45:47] Blue: Well, I love everything you’re saying. I love that checkability, vague manning. I think you might have just coined two new new terms here. Now, I assume that you’re not using this as you’re not advocating using this as a way to shut down any theory that’s not a best theory. I mean, it seems to me that whole whole range of things that are just really interesting theories to discuss, whether panpsychism or the simulation hypothesis or Joe Rogan was going on this thing about the stone date hypothesis the other day. I thought it was really interesting. I mean, these are all like they’re obviously not best theories. They might not even be testable, but they’re they’re still interesting to talk about.
[01:46:39] Unknown: Right.
[01:46:39] Blue: Would you say that?
[01:46:40] Red: Yeah, totally. So yeah, OK, let’s put this in a different slightly more formal way. Let’s say you took the stance that anything that isn’t a best theory is garbage and we’re going to ignore it. OK, yeah, how would you? How would you ever get a replacement theory off the ground? Right?
[01:46:58] Blue: Yeah, yeah, it would be impossible.
[01:47:01] Red: OK, so and this is why I’ve suggested that you have to kind of think of Popper’s epistemology as having maybe two stages, two phases. Admittedly, you’re you’re probably intermixing the phases, right? But this is just for us to make it easy in our minds. Are you coming up with an interesting conjecture or are you advancing a best theory? OK, would yeah, as something that’s ready to be criticized, right? That can survive criticism. OK,
[01:47:27] Blue: so it’s really just a matter of being more mindful about what you’re doing.
[01:47:30] Red: Yes. So my experience, this is not unique to the crit community at all, my experience with just people in general is that they always, even if they’re not critical rationalists, even if they don’t know what a quote best theory is from a Popperian standpoint, is that they see, they always see their theory as it’s the obviously best theory. And the other theories just have all these problems with them. I don’t care if you’re a creationist. You know, if you’re a young earth creationist, you see your theory as just kind of obviously true. Bible said so. And I’ve got all these problems that I’ve offered to evolution, my competitor, that really are problems and therefore your theory is refuted. And so mine’s the only theory left. And I think that is what folk epistemology is, right? Is we do very naturally try to refute the competing theory and try to leave your explanation as the soul standing explanation. Popper didn’t invent that part. What what really makes Popper unique is this realization that it’s really about formulating your theory such that they can be tested or checked. OK, and that’s a choice you have to make. That’s really the big breakthrough with Popper. OK, not refutation, because everybody thinks they’ve refuted all the competing theories all the time and in a sense they have. And in a very real sense, creationists, their arguments, some of them are terrible arguments, but some of their arguments are valid against evolution. And they really have refuted evolution. It doesn’t matter that they’ve refuted evolution because evolution just simply is going to end up with a replacement theory that still isn’t creationism. And we’ll eventually get there, right?
[01:49:07] Red: And creationism is not even being posed as a testable theory to begin with. It’s almost always vague manned to the point of there’s just nothing you can do with the theory. And you just have to kind of think about this, right? You have to think about we get so excited like, oh, no, I’ve refuted your theory. In some sense, all theories were refuted anyhow, right? It’s so it doesn’t matter that you refuted the theory. It’s it’s about making your competitor subsume the other theories. If you were to tell me. Hey, Bruce, you know, praxeology, Misi and economics, that’s a really fruitful theory. I totally agree with you, right? And there’s like tons we can learn from it. But it’s not a best theory. Like it’s nowhere near even close to being a paparian best theory. If if the critrack community, instead of trying to declare animals as a best theory, we are best theory in animals don’t feel things, which, like Dennis Hackful actually said that, right? In I’ve got tweets and articles like that. If he had just said, look, we don’t know for sure. Here’s an interesting theory that’s not yet testable. What if animals don’t feel things, right? And he actually put it out there as an interesting theory. Let’s think about the consequences of this. OK, there’s nothing wrong with that. That that really is interesting at that point. It’s worthy of discussion at that point.
[01:50:33] Blue: Yeah,
[01:50:33] Red: yeah. There’s this giant difference between if if David Chalmers or Saudi or whoever is trying to disprove Church -Turing and Deutsch thesis, if they were to advance their theory, hey, there’s this interesting idea, what if there’s going to be new physics and these new physics are going to make time, you know, a sentence is Lee Cronin almost to a T, right? I’m going to make time, you know, fundamental and space is emergent from it. And we’re working on assembly theory and it’s kind of showing us the way towards this. No, this isn’t the best theory, right? Like it’s not even close. OK, but sure, it’s an interesting theory. We absolutely should have discussions about it and see where we’re going with this. It’s really only a problem once you start to say, oh, I’ve refuted your theory, though, so we’re going to have to go this way. No, you haven’t, right? You offered to get feeling for why you personally don’t like Church -Turing Deutsch thesis, right? It’s by far the stronger theory still.
[01:51:29] Blue: Well, this kind of goes back to our dogmatism episode. In a way, I mean, it might not be such a bad thing that some people out there think they have a best theory, it motivates them to to to defend their theories and maybe maybe make them best theories.
[01:51:47] Red: So I’m going to say something very controversial.
[01:51:49] Blue: Yeah,
[01:51:50] Red: because I never do that. I never say controversial things. So I was on the increments podcast and we talked about dogmatism. And I actually said, dogmatism is not all bad. And Ben kind of defended that. And then Vedin kind of goes, we’re not encouraging you to be dogmatist. No, no, no, we’re not, you know? And here’s the honest truth. OK, and I kind of said, you know what? I agree with every position we’ve talked about here. The fact is, is dogmatism is often very, very, very good.
[01:52:14] Blue: In
[01:52:15] Red: fact, a dogmatic scientist that understands the intuition, the institutions of science that he must not big man his theories, that he must come up with testable theories. A dogmatic scientist is incredibly valuable. Imagine someone so dogmatic that, but, but they obey Popper’s ratchet. OK, they take that seriously. That’s something that they need to do. OK, they’re going to go to great lengths to try to figure out, OK, I got this refutation, this refutation of my theory. How can I potential refutation of my theory? How can I come up with an alternative explicit alternative explanation that has its own explicit independent tests? OK, and a dogmatic person may be the only person who can spend the time necessary to figure that out, right? Because they’re sure that they’re right and the other side is wrong. And so it’s just a matter of they, but they know there must therefore be a good explicit alternative explanation. So they go find it and they come up with it and they try a thousand things till they find it, OK? There’s nothing wrong with that, that kind of dogmatism, the kind that is following Popper’s ratchet, OK, just following the no ad hoc convention is all good. There are no downsides to it. And there’s nothing wrong with that kind of dogmatism, which is why I’ve been so much more positive on dogmatism on this podcast than people normally are. And I know it makes people a little uncomfortable. It’s not really dogmatism. That’s the problem. It’s vague manning your theories that is the problem. If you vague man your theories because you’re a dogmatist, which is usually why you do it, that’s a problem. OK, it’s not the dogmatism that’s the problem, though, not by itself.
[01:54:05] Red: Does that make sense? Sorry.
[01:54:07] Blue: Yes, it does. And I assume that you you would recognize that it’s just. Being who we are. That as we go through this life following our curiosity and a million different ways, we pretty much have to go with our gut. Yes, just on a day to day basis. I don’t even know what a conversation between two people would really look like if they were not going with their gut, at least at some level. I mean, we just don’t have the time or energy or ability to make all of our theories.
[01:54:49] Red: Best theories.
[01:54:50] Blue: Yeah, as theories, I guess, right? I mean, I assume you were completely
[01:54:54] Red: right. You’re completely right. It’s so hard. That was the quote I had from Popper about we make a framework that’s very hard to satisfy. That’s exactly what science is, right? It is super hard to get to best theory status. And even when you do it, like I’ve argued that animals filling things is a best theory. OK, but I always caveat it with, but it’s still a bad theory. It’s really only a best theory compared to its competitor, which is an even worse theory, right? It’s a terrible theory, in fact. We know so little about qualia and when the critrack community is at their best, what they really do is they point out we make assumptions about animals filling things and we haven’t really we don’t really have a detailed theory of qualia. So it’s really possible, is at least a possibility that it will turn out that animals don’t fill things or some animals don’t fill things or it’s not the animals we expect or it isn’t they fill things, but not in the ways we expected. You’d have to have a very open minded kind of look at it. And I think that’s beautiful, right? To really that’s a proper use of critical rationalism to kind of put out there, here’s what we don’t know at this point. Now, there’s a giant body of work. The reason why and I’ve used this as an example in the past on the podcast, it used to be that everyone believed animals didn’t fill things and all scientists at least believed animals didn’t fill things. I mean, obviously, individual pet owners have always thought their animals fill things, but it was widely accepted that animals did not feel things.
[01:56:27] Red: And there was a small group of scientists, DeWall and others that put some really excellent science together and changed the minds of the scientific community, not entirely. It’s not, you know, unanimous, but and it really caused us to rethink whether animals feel things or not based on some really clever experiments that they’ve come up with. I dislike it. I mean, it’s it’s cringey when the crit -rat community tries to ignore that body of work, tries to say, oh, it’s all justificationism, it’s all empiricism or whatever. Right.
[01:57:05] Blue: Yeah. Well, it’s just off -putting for people. Right. I mean, it just doesn’t doesn’t conform with just common sense. When you see the video of the dog and the dogs watching Star Wars and Darth Vader comes out and the dog like hides behind the couch. The dog must be feeling something there.
[01:57:23] Red: So, well, let me put it this way. So let’s take the grief experiments as an example. There’s a ton of these, like there’s a ton of these, like this giant body of literature, right? Let’s just take the grief experiments where they actually. So animals, some animals, not all will die from grief, right? They’ll actually stop taking care of themselves. OK, now we how do we explain grief evolutionarily for humans? Well, we explain it in terms of humans developed this feelings and the feelings have consequences, but that the feeling in general has reproductive value, but it can be negative in some circumstances. Specifically, you have this feeling of how terrible it would be if you lost your loved ones, let’s say, OK, and then grief is what actually happens when you actually do lose your loved ones. That overall feeling of love and attachment, that’s what evolution evolved. And the grief is this impact that takes place afterwards. And yes, it’s a negative, but it’s offset properly by the fact that you want to take care of your loved ones, which then passes your genes on. OK, so it fits evolutionary standard Darwinian evolution perfectly, right? This theory to explain why humans feel grief references a feeling to be able to explain it, and so therefore, as per Deutsche’s criteria for reality, if it’s in our best theory, then we counted as real, at least for now. Then how do you explain that animals also die from lack of self care? Now, of course, of course, you can vague man your theory. I had so I had a crit rat right to me. And he argued, well, no, that doesn’t prove animals. Feel grief just because they die from lack of self care. It
[01:59:21] Red: could just be that they have this algorithm and it’s to stay near their young and then the young die and the algorithm kind of screws up and they’re not sure what to do because the the young isn’t moving and therefore they don’t do anything. And I tried to point out, I said, look, that’s not a testable theory. And he didn’t care because, of course, most crit rats don’t actually understand Hopper’s ratchet or care about it. And I do. And I was trying to explain, look, I’m not saying you’re wrong. I’m not saying you’re wrong. It could be that they feel nothing and there’s some other explanation for why yours explanation could even be the right one. But you must make it explicit. You must explain, look, here’s how we’re going to test this theory as opposed to having it be feelings, the feeling of grief. Here’s how the animal will behave differently if they’re not feeling grief and they just have some sort of algorithm that gets screwed up when they’re young die. OK, so they end up not knowing what to do. By the way, doesn’t that violate evolutionary theory? I mean, like in the case of grief, the grief was a national consequence of attachment and it was the attachment that was evolved. In this case, shouldn’t evolution have pushed away the negative so that if the animal, the young die, then it just takes care of itself and goes, oh, well, it’s not moving. I guess I guess that’s fine. Why stop taking care of itself? You’re trying to explain this in terms of a coincidence. You’re trying to explain this in a bad explanation.
[02:00:49] Red: You’re trying to say, oh, it just so happens that they have some sort of algorithm and I’m not telling you exactly what and it just sort of looks like grief. But I can’t even tell you explicitly what it is so that I could tell you how it differs from grief, from the feeling of how an animal would behave if it feels grief. Well, it’s obvious this theory doesn’t prohibit anything. Whereas the theory that the animal feels something does. It in fact, the theory that the animal feels things has immediate consequences that we can go test. For example, they should have the exact same hormone levels as us. So they go, they actually went out and they checked the feces of the animals. And it turns out animals that recently had a lost loved one. This was an apes had the same wacky hormone changes that a human would have. OK, there’s there’s no way like it’s just a coincidence. I mean, you can see how this doesn’t prove anything I admit. But it’s exactly following Popper’s ratchet where it had these implications that we could go check. OK, this is why people started to buy these theories. And then they also had these interesting cases that just couldn’t be explained like this theory that that this rat was putting forward that maybe the animal just somehow screws up. One of the ones that died was one of the ones that died from self care. One of the apes that died from self care was because the mother died and it was an adult ape and it just felt so bad from its mother dying that it didn’t take care of itself and ended up dying. Right.
[02:02:16] Red: I mean, how do you explain that in terms of this loose theory? Now, of course, you can always come up with a vague man theory and you can say, well, there’s just some sort of other algorithm in its head. And it just so happens, it looks like grief. Now you’re not being explicit. Now you’re not prohibiting anything. Your theory basically says animals have an algorithm that looks exactly like grief and acts exactly like grief in every conceivable way, except that it’s not grief. And that’s really your theory at this point. And that is a bad explanation. That’s what we mean by bad explanation. If you want to put forward an alternative theory to animals filling grief, you must offer an equally explicit explanation that explains the same things and has its own testable consequences. OK. And that is poppers epistemology. You keep applying poppers ratchet. You keep error correcting your theories by making them increasingly specific so that you can check them and they have consequences that you can go check. You can’t check the whole theory, but you can check some of the consequences of the theory. There isn’t a good alternative that we need to do an episode where we actually talk about what the alternatives are. There are some theories out there that maybe animals don’t feel things in the same way humans do. Usually they claim animals do feel something, right? But there’s like the two circuits theory where they claim that most of what we see in animals is actually not feelings at all. And they base that on the fact that most of like if you see somebody experiencing fear,
[02:03:48] Red: a lot of times they don’t report feeling fear, but their brain knows to be in a fearful circumstance and they behave in a certain way. That’s the two circuits theory. And we know we know it’s true for humans. That humans don’t always feel things, but they do act in a certain way that we can recognize as emotion, even though they don’t feel a thing. So emotions are the way you is the mode you’re in. And feelings is an actual instantiation where you actually experience it. Right.
[02:04:14] Blue: That’s that’s interesting. So I would almost say that a real like really what we think of as feelings is almost like a self reflection kind of a thing. Whereas if you’re really in a situation where you’re scared or something like that, that’s almost like a different kind of a feeling than a self reflective feeling.
[02:04:34] Red: Right.
[02:04:34] Blue: Is that what I’m kind of? It is. And
[02:04:36] Red: I’d have to go get the actual papers. I’d have to go find them again and we could like do a podcast on it. So could you use the two circuits theory to try to come up with an alternative theory of animals not feeling things? Well, you probably could, right? Like, I would love to see the crit rap community behave like critical rationalists and offer alternative theories of I don’t care if they’re dogmatic, like if you really believe animals don’t feel things great. If you dogmatically believe animals don’t feel things great, but just obey Popper’s ratchet, right? Come up with interesting alternative theories or admit that you don’t have them. Say, you know, mine’s not a best theory yet. You know,
[02:05:10] Blue: my gut says that that’s a that’s a better theory. We’ll say that. But so
[02:05:17] Red: anyway,
[02:05:17] Blue: we should wrap this up before we get too much over two hours here.
[02:05:21] Red: Yeah. OK. I was just going to say the two circuits theory, the guy who came up with it, they do believe animals actually feel things. So they just they just believe that that it’s a
[02:05:31] Blue: different kind of feelings.
[02:05:32] Red: It’s different kinds of feelings. Yeah. So they well, the theory is that the animals still have the two circuits. It’s just that the circuits much stronger for humans.
[02:05:41] Blue: Oh, OK. OK. That’s interesting.
[02:05:44] Red: Hmm.
[02:05:45] Blue: Well, well, Bruce, this has been top top shelf Bruce Nielsen, I’d say. One of your best episodes. Checkability, vague manning. I love it. I’m going to start using it in my everyday lexicon, as if people didn’t think I’m weird enough already. But it’s very cool. And I’ll look forward to listening again and editing. And thank you.
[02:06:11] Red: All right. Thank you very much.
[02:06:20] Blue: 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 BN Nielsen 01. Also, please consider joining the Facebook group, the mini 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.