Episode 140: Immunizing Stratagems

• Listen to this episode: Apple Podcasts

Unofficial, reviewed transcript. It may still contain mistakes; check the podcast when wording matters.

Speakers are labeled by first name where identified. Unidentified-speaker labels may refer to different people. Bracketed question marks indicate uncertain wording.

Transcript

[00:00:00]  Unidentified speaker: Hello out there.

[00:00:01]  Peter: This week on the Theory of Anything podcast, we talk about immunizing stratagems and ad hoc saves. As I understand it, both ad hoc saves and immunizing stratagems are phrases Karl Popper used to describe moves humans make when we try to protect our theories from refutation. I think this is something we all do every day in one way or another. But here's the thing, immunizing stratagems hurt theories in that it makes the theories less convincing to other people. So if someone says all swans are white and suddenly we see a black swan and that original person says, well, that's not really a swan. Well, their ideas about swans are going to be less convincing. To Karl Popper, Marxism and psychoanalysis are the classic areas where he saw believers bending over backwards to save these theories from refutation from the real world. Bruce reads a lot from the excellent book, Myth of the Closed Mind by Ray Scott Percival, who we interviewed the other week on the podcast. Basically, what I hear Ray expressing in his book is the common sense assertion, though also oddly provocative, that when theories come into contact with human minds, these theories or memes that move us closer to truth have a survival advantage. So as flawed as we are, humans, all humans cannot help but seek truth, fallibly, of course. Like Bruce, I absolutely love the thesis of Ray's book.

[00:01:55]  Peter: To be honest, Ray has helped me appreciate even more, a lot of what Bruce talks about on our podcast, especially the concept of what Bruce calls Popper's ratchet, or this idea that we should always strive to modify our theories by making them more testable rather than less. I also want to issue a warning to our viewers that Bruce reads a lot from Ray's book, more than I recall him ever reading before from any book, especially from Chapter 4, Ideologies as Shapeshifters. I hope that if nothing else, this episode will encourage our listeners to read Ray's excellent book, The Myth of the Closed Mind, available on Amazon to order. I know I sure enjoyed listening to Bruce here as always.

[00:03:01]  Bruce: Welcome to The Theory of Anything podcast.

[00:03:03]  Unidentified speaker: Hey, Peter.

[00:03:04]  Unidentified speaker: Hey, Bruce, how you doing?

[00:03:06]  Unidentified speaker: Good.

[00:03:07]  Bruce: Well, we just did our interview with Ray Percival, Scott Percival. Probably the episodes will air out of order, so this may not be just the previous episode, but that was what we did last. When I had him on, I asked if I could read a substantial portion of one of his chapters from his book, because I thought it was so dang good.

[00:03:34]  Unidentified speaker: Okay.

[00:03:35]  Bruce: He gave me permission. I don't know if he realized how much I'm going to read. I did say significant portion, and he said, absolutely. That will help get advertisements for my book. That sounds good to me. But I'm going to take, and I'm going to read, and we're going to discuss, a sizable portion of Chapter Four, Ideologies as Shapeshifters. It's not even a majority of that chapter. The whole chapter is great, and I really highly recommend reading this book. This book was great anyhow, but Chapter Four just puts this book totally over the top for me. Must read for critical rationalists. Now, you're going to probably see immediately why I liked it so much. It matches my own ideas I've been developing on this podcast almost entirely. But he puts it so well. He's a great author. He just says it in a way I kept thinking, I wish I could say it like this. I'm like, oh, I should just read it.

[00:04:40]  Unidentified speaker: Are you ready, Peter?

[00:04:42]  Unidentified speaker: I am ready.

[00:04:43]  Unidentified speaker: Let's do this.

[00:04:45]  Bruce: All right. So he says, quoting now, if we want to guarantee our belief system, so just to keep this in mind, I know we already talked about this in the interview with Ray. But his book, even though it's ostensibly about people don't have closed minds, what he really takes as his thesis is something that's not as difficult to prove. And instead, he simply shows that if you try to have a message that, a theory that is immunized, that you can't help but hurt the message and make it so it's not as memetic.

[00:05:27]  Unidentified speaker: Okay.

[00:05:28]  Bruce: And most of the book, he's really doing a great job of showing that.

[00:05:33]  Unidentified speaker: Okay.

[00:05:34]  Bruce: And in the interview, I challenged him as much as I could on, come on, people are closed minded. And he basically admits that they are. Like even in the book, he admits this, right? And his point is really is stronger than that. He's kind of taking an edge of being provocative.

[00:05:55]  Unidentified speaker: He said this in the interview.

[00:05:56]  Bruce: He's being provocative, but his real thesis is something so solid that it's really hard to imagine anyone arguing with it. And yet nobody seems to ever talk about it. So I think this is why he got excited about this as kind of his thesis and wanted to develop this as his career.

[00:06:12]  Unidentified speaker: Okay.

[00:06:13]  Bruce: So if we want to guarantee our belief system, so remember that's what he's always talking about is, is it possible to guarantee a belief system by immunizing it from criticism? Our theory or our ideology against being destroyed by criticism, we may try to formulate it in such a way that it can survive any criticism because it can be reinterpreted so that no criticism could really touch it. A crude example would go like this. I predict that the world will end on a certain date. Lots of people believe me and I attract a big following. The date comes and goes without the world ending. My critics say that I was wrong. What I said would happen did not happen. Therefore, my belief system is false. I then scornfully reply that my belief system is entirely correct. My critics are making it—making a silly and superficial blunder because the world really did end on that date, and I intended to go on reporting the fact that the world did end on that date. It's just that the end of the world is not detectable by the normal methods of observation, as my very foolish critics have carelessly supposed. It doesn't matter here exactly how I developed this idea. I may say that God destroyed the world and then instantly recreated it exactly as it had been before. Or I may say that the new world is different to the old one, which really did come to an end. Don't doubt it for a moment. Because in the new world, certain trends have begun which did not exist before. There are hundreds of other ways I might develop the idea.

[00:07:50]  Bruce: I might even be very obscure about just how it is to be developed, which would go to show how I am inordinately wise. All that matters here is that I have given an interpretation to my old belief which saves it from being rejected. I deny that I was ever wrong. By giving an interpretation to my old beliefs which squares it with the criticism it has encountered, criticism of my belief has been rendered impotent and my belief has been guaranteed against ever being demolished by criticism. I've won, haven't I?

[00:08:26]  Peter: It really seems to me people do this all the time. I mean, the strategy is, I mean, the example is a little bit extreme.

[00:08:35]  Bruce: Potentially extreme.

[00:08:37]  Peter: But you know, it really, and I assume you're going to get into some positives of this too. But you know, do you know that feeling where like, you're talking to someone and trying to defend your point of view? And then, like, it's not even so much about what they're saying, really. But suddenly, you have, it's just like this feeling comes over you and you have a completely new way to defend your position. Like, it's like you're grasping at something. And then it's like this aha moment where you're like, something clicks with you and you think of a new, I mean, I'm sure that happens to you all the time, right?

[00:09:16]  Unidentified speaker: When you're in a debate.

[00:09:17]  Unidentified speaker: So, you know, like it just, it's kind of indicates that a stra, that a conversation, the value of a conversation is not just like what they're saying, but it's kind of like, realize, making your own arguments more coherent too.

[00:09:33]  Unidentified speaker: I don't know.

[00:09:34]  Peter: I mean, that's like a positive framing of a, of a immunizing strategy, maybe.

[00:09:41]  Unidentified speaker: Yes.

[00:09:42]  Bruce: Do you know, let me, so let me talk about that just for a second. So first of all, we have an upcoming podcast where I'm actually going to show a CritRat that did almost exactly what, what he's making fun of here, right? Where he would just keep reinterpreting what he meant in such a way to immunize his theory.

[00:10:03]  Unidentified speaker: Yeah.

[00:10:04]  Bruce: So yes, this happens like all the time, even amongst Popperians, like this is very common for people to just become chameleons and reinterpret their theory on the fly, to avoid being, um, avoid a criticism, immunized from the criticism.

[00:10:19]  Unidentified speaker: Which isn't always a bad thing, right?

[00:10:20]  Peter: I mean, there's got to be some utility to that too.

[00:10:23]  Bruce: So I had a CritRat, not the same one, tell me that it didn't matter if he understood me or not, because I was just supposed to get out of the conversation. Whatever I did, that was the value.

[00:10:36]  Unidentified speaker: Okay.

[00:10:39]  Bruce: You know, he wasn't wrong. I actually did get a lot of value out of my conversations with him. It did get my thinking going better. I would understand my own arguments better, but I still quit having conversations with him, like absolutely did.

[00:10:55]  Unidentified speaker: Okay.

[00:10:56]  Bruce: Because that attitude is so nasty and so wrong and honestly unethical, that there was no way I was continuing conversations with him. That's actually not true. I actually offered to continue conversations. I said, why don't we just talk on Zoom, where you can't just misunderstand what I'm saying and take something out of context. He slightly refused.

[00:11:16]  Unidentified speaker: Okay.

[00:11:18]  Bruce: I'm like, you know what? Here's the thing. He's kind of right. Two things can be true at once. It can be true that I was getting something out of the conversations, but it can also be true that I wasn't getting enough to make it worth my time because at some level, I knew he was absolutely never going to care what I said. I would spend hours writing to him only for him to take one little line out of context, reply with one little line, and everything else was lost. You would try to explain what he had done wrong. No, you've misunderstood me in this way, and he would take one more little line out of context and respond with one line. He was spending 30 seconds on each reply. I was spending hours. It was a completely losing situation for me. I finally just said, I will continue conversations if you do it on Zoom, and he completely refused. To this day, we still don't talk because he will not have conversations on Zoom with me. I'm going to make the claim that he knows that the moment he's on Zoom, his little strategy stops working, and he can no longer immunize his theories, [his criticisms I'm offering?]. I think he avoids any situation like that. He's completely safe with text. It's so easy to just take things out of context. There's two sides to this. There's, do you get something out of it? I think you always do. Do you get enough out of it to make it worth your time? I don't think you necessarily do. I do think you ultimately have to pick who you're going to have conversations with. You have to decide what the form is.

[00:12:59]  Bruce: I think private conversations, for the most part, I've gotten away from them because I just know most people have made up their minds and we're not going to make much progress. They're not even going to, in most cases, understand what I'm saying. I'd rather do it in public where I can at least know, okay, I'm putting together my own thoughts and I can put them out there for other people, so that way I'm getting more out of it. Then I'm more willing to do it. I'll go into depths. I'm known for going into depths with people trying to point by point look at what they're saying. I'm way more likely to do that if it's in public, if I can retweet myself, if I can make a point and then think, oh, that's a good way of putting things. I know the person I'm talking to doesn't care. But somebody else is going to.

[00:13:49]  Unidentified speaker: Yeah, sure.

[00:13:50]  Bruce: So I've learned to do that and that way, I am aiming more at just how does it refine my own. I know, I don’t think— I shouldn't say I've never convinced somebody of something because I'm sure I have.

[00:14:02]  Unidentified speaker: You've convinced me of a lot of things.

[00:14:04]  Unidentified speaker: So there's one.

[00:14:07]  Unidentified speaker: I've never, at least in the moment, I've never done it ever.

[00:14:09]  Bruce: I don't think I've ever changed somebody's mind. So I do think you have to ultimately look at it as, what am I getting out of this in terms of developing my own theory, and then am I getting enough out of it that this is going to be worth my time? That will depend on the form, the situation, the individual. There's a lot of things that might go into that decision to whether to have a long-term discussion or not. But you're right. I want to emphasize, you are correct, that even in an extreme case like this, where the person was really being pretty unethical about how they were going about it, I was still getting value out of it. It really strongly developed my ideas, got me thinking. My conversations with him was one of the things that really got me thinking about the fact that we have a problem with a lot of the crit rats, the way they're understanding Popper's epistemology. The way he would misunderstand things, I would go, wait, wait, wait, what's he getting wrong here? Then I would go back, I would look up at Popper, and I would say, well, I get now what Popper was saying, right? I can't deny the huge value that came out of those conversations, but I also still ended them, especially once I felt like I had gotten what I could out of them.

[00:15:28]  Unidentified speaker: Yeah.

[00:15:28]  Peter: So even though he was interpreting you very uncharitably, it was still, you still feel like you got a lot out of the conversations.

[00:15:39]  Unidentified speaker: Oh, I did.

[00:15:40]  Bruce: I totally did.

[00:15:42]  Unidentified speaker: It's about the struggle of trying to explain yourself better or dig into Popper better, more thoroughly or whatever.

[00:15:50]  Unidentified speaker: Yeah, yeah, I think, I think most people know, know that feeling a little bit.

[00:15:56]  Bruce: This was a couple of years ago, so anyone who thinks maybe I'm talking about them, I'm probably not.

[00:16:03]  Unidentified speaker: Okay.

[00:16:06]  Unidentified speaker: So, but yeah, it’s so it, and I do kind of see it that way now.

[00:16:11]  Bruce: It's like, okay, how am I going to refine my own ideas here? I mean, I know it, I know [I'm not got a weakness?] in that I get frustrated, and I'm really trying to learn to not do that, right? Because it really should just be about, how is it helping you develop your own ideas? There are part of me that believes it should be more than that. But I think there's overwhelming evidence that it just for the most part isn't, right? There is the question of dogmatism here, which is something we will explore on future podcasts, which is actually one of the reasons why I brought Ray on. I wanted to talk with him about that because I feel like it's a super important question. Why are humans dogmatic? Is it a good or a bad thing?

[00:16:50]  Unidentified speaker: Does it have an upside?

[00:16:51]  Unidentified speaker: There's all sorts of questions I've got around dogmatism, but I felt like Ray was one of the better experts on this subject to help me work through my own feelings about it.

[00:17:00]  Unidentified speaker: Yeah. Yeah.

[00:17:01]  Unidentified speaker: Um, but like, I do feel like there's a part of me that truly believes we really ought to be able to study epistemology, study what Popper taught, and we should really be less dogmatic after that because we're just more open to try to really understand the other viewpoint, to very carefully understand the criticisms and then make sure that we've got objective responses to them and not just off the cuff, armchair philosophizing, and there's a part of me that truly believes that should be the case, but I feel like it just isn't the case.

[00:17:39]  Bruce: I think humans are just humans. Of course, I include myself in this. I'm sure that I am as dogmatic as the next guy. But you don't see it in yourself, you see it in other people, and when you see it enough, you start to realize this really just may be a human trait. It may be something that there's just no easy answer for, and maybe that's why we needed special institutions to help overcome the fact that we're so dogmatic, and maybe we needed science with specific institutions to be able to see, oh, how do you make progress anyhow, even though humans have this weakness? I'm exploring several ideas around this actually, and that's why I've picked this as a reoccurring topic on the podcast.

[00:18:22]  Unidentified speaker: Yeah.

[00:18:22]  Unidentified speaker: A weakness that can also be a strength, too, and I guess that's why it's-

[00:18:26]  Unidentified speaker: It can also be a strength, yes.

[00:18:27]  Unidentified speaker: I mean, I think if it was just a weakness, it would be quite obvious how to get over dogmatism, but maybe that's why it's so tempting to embrace dogmatism in various ways in our life is because at some level, we know that it can be beneficial to be dogmatic about our ideas.

[00:18:55]  Bruce: I think there are several reasons about why dogmatism is so common, right? And I do think part of it may boil. So I raised the question in a past podcast about, will an AGI be a super intelligence compared to us because it's better at self-criticism? And I actually think that part of the reason why dogmatism exists is a natural part of how intelligence works. And in that sense, a certain amount of dogmatism, an AGI will still be dogmatic. And it may actually be enough of a positive in that sense that we're never really going to get rid of it. Now, having said that, I think humans are far more dogmatic than they need to be. And so that is still a bit of a mystery to me, why are humans an extra dollop of dogmatic compared to what would just make sense, right? Where there's a certain amount makes sense, but we're more than that. And I think by a fair margin, right?

[00:19:56]  Unidentified speaker: Okay.

[00:19:57]  Bruce: So why is that true? And there has to be a reason for it, right? Like, there's no way we would be that way. And I'm getting ahead of myself to a podcast I'm gonna do in the future. There's no way we would be that way unless there was actual survival value in it. Evolution would not develop us to be a little extra dogmatic unless there was a good reason to be a little less extra dogmatic. Good reason from the standpoint of the genes, I should say, not necessarily from the standpoint of the individual. So this is something I think is worth exploring, maybe both sides of this.

[00:20:35]  Unidentified speaker: What's the upsides of dogmatism?

[00:20:37]  Unidentified speaker: What's the downsides?

[00:20:38]  Bruce: In the case of the downsides, what was the upside to the genes? Why did we develop that way? How will AGI be different from us? How would they be less dogmatic from us? Under what circumstances would they be less dogmatic than us? So those are all questions I want to pursue in more detail in future podcasts. Let's keep going with reading Ray here. If we frame our theory, oh, I was going to mention that example he just used about the world being destroyed, that was actually the Jehovah's Witnesses. So even though we're saying it's strained, it's not. That's a real true life story that happened.

[00:21:14]  Unidentified speaker: Oh yeah, I've heard about that.

[00:21:15]  Bruce: He's putting it in a somewhat more silly way than definitely from, differently from the way Jehovah's Witnesses would explain it.

[00:21:21]  Unidentified speaker: But it's very close to what the Jehovah's Witnesses actually believe.

[00:21:26]  Unidentified speaker: Yeah.

[00:21:27]  Bruce: So okay, if we frame our theory so that what they believed was that Jesus was going to come again by a certain date, 1914. When he didn't, they declared that he had come, but it was spiritually. Okay, if we frame our theory so that logically no possible criticism, or perhaps merely no actual or likely criticism could touch it, we have immunized it against criticism. The device we use to immunize a theory is called an immunizing strategy. We give our belief system protection against criticism by setting up a logical barrier to criticism. It may seem that if we can do this, then we will have given our ideology an advantage in the competition of ideas. My aim in this chapter is not to demonstrate that a system of ideas cannot deflect criticism by logical means, but rather to show the limitations and the costs of doing this. What I will show is that immunizing stratagems either abandon the belief system they're supposed to protect or else lower the survivability of the system. Karl Popper originally used the term conventionalist stratagem, but then adopted the term immunizing stratagem from Hans Albert to describe an aspect of the unscientific methodology of certain ideologies claiming to be scientific, Marxism and Freudianism in particular. Arthur Papp had already anticipated this usage. Popper argued that Marxism, which was originally an empirically testable theory, had been recast in the form of empirically irrefutable metaphysics. This maneuver, Popper claimed, saved Marxism from refutation and immunized it against further attacks.

[00:23:08]  Bruce: Freudianism was, Popper claimed, irrefutable from the beginning. The basic theory of Freudianism does not need any immunization to make it irrefutable. Nevertheless, it does incorporate immunizing stratagems. Popper contrasted Marxism and Freudianism with the theories of Newton and Einstein, which he said were full of testable content. Thus, Popper's employment of the term immunizing stratagem arose in connection with his attempt to solve the problem of distinguishing scientific from non-scientific, including pseudoscientific theories, the demarcation problem. Popper's solution was the methodological rule to allow into science only empirically falsifiable hypotheses. If a theory can be shown to be falsifiable by an observation, an empirical test, then it is scientific. If no observation could show to be false, it is not scientific. Okay, so one thing I need to call out here. I have made two points about the demarcation criteria throughout this podcast. One, it's way more important to Popper's epistemology than you think. In fact, in his mind, it was the whole basis for his epistemology. And two, it's wrong. And it's the wrong way to actually slice things up.

[00:24:25]  Unidentified speaker: Okay, wait, let me think about this.

[00:24:27]  Peter: Okay, so the demarcation criteria, it's kind of like one of those things he's most known for, right? I mean, but most people don't know how deep it goes into how sophisticated it is in terms of epistemology.

[00:24:51]  Unidentified speaker: Is that more kind of what you're saying?

[00:24:54]  Bruce: The important thing is, is that the historical problem he was, so this is going to be in the podcast just before this one unless I move things around. Okay, so I've already just mentioned this. You listened to it in the previous podcast, even though it was recorded like a long time ago or something. In that podcast, where I'm comparing Popperian epistemology to Deutschian epistemology. Popperian here meaning original Popper, obviously Deutschian epistemology is Popperian epistemology, but it's a new interpretation on it. One of the things I point out is that the Deutschian reinterpretation sees the demarcation criteria. They agree with Popper on it, but they see it as a mere classification. It's just, oh, you happen to be in this category or not, it has no real other importance. That is absolutely historically inaccurate to what Popper was trying to do. Although, having said that, Deutsch isn't wrong about that. The way Popper happened to slice the demarcation criteria, it did end up being something closer to just a simple category, categorical difference. It didn't actually separate good and bad explanations anymore. Popper was trying to solve the problem of good versus bad explanations, and he thought the demarcation criteria solved it, at least he initially thought the demarcation criteria solved it, okay? It doesn't, because there's such a... It does go in one direction. A testable empirical theory is going to be a good explanation, but a non-empirical theory can still be a good checkable explanation.

[00:26:33]  Bruce: And so it captures a certain set of good explanations, but it leaves out a certain set of good explanations. And it takes that certain set of what we might call philosophical good explanations that are checkable instead of empirically testable. It merges those into bad explanations and treats them all the same. And that's the mistake of the demarcation criteria, in my opinion.

[00:26:57]  Unidentified speaker: Okay.

[00:26:59]  Bruce: I agree with Popper's desire to define what's a good explanation for a bad one. And I believe the demarcation criteria almost gets it right, right? Like it does slice off a whole bunch of really good explanations and separate them from this fuzzier category, right? And so I see why he was so excited about it. I see why it led to the rest of his epistemology. It was the basis for the rest of his epistemology, right? And yet it still was missing something. And Deutsch notices it's missing something. He kind of reimagines it. Oh, it's just a category. It doesn't matter that much. A metaphysical theory can be as good an explanation as a testable empirical theory. And Deutsch is right about that. But in doing so, he's now opened a loophole where you can come up with a very, very bad explanation, declare it philosophical, and the CritRat loophole has been born, right? And so I'm trying to thread the needle between these two. Demarcation criteria was way more important than you think.

[00:28:03]  Unidentified speaker: But it had a mistake in it.

[00:28:04]  Bruce: And I've tried to suggest at least a way to correct the mistake. My idea of checkability versus testability, that's how I'm trying to correct the mistake. Then you have an actual demarcation between checkable versus non-checkable. And that is the difference between good explanations and bad explanations. Or maybe it is. This is my first cut at it. I don't actually know if I'm right or not, if I'm being honest. But it sure works in a lot of cases, right? If I were to pull out something like the one I often use as an example is Church-Turing thesis. So particularly the original Church-Turing thesis before Deutsch, it changed it to the Church-Turing-Deutsch thesis. The original Church-Turing thesis behaved a lot more like a non-empirical theory, right? Where you can't go do an empirical experiment to test it. The way you test it is by trying to build a hypothetical computer that has a power different from a Turing machine, and there's no proof you can't do that, but nobody knows prior to Deutsch, nobody knows how to do it. So by doing that, you have a theory that isn't empirically testable, but it is every bit as checkable as an empirical theory and is therefore every bit as good an explanation as an empirical theory. Even though you might be tempted to try to declare it a metaphysical theory, it's got none of the problems of what you would typically think of as a metaphysical theory. This is my way of trying to separate, I have a different demarcation criteria. I use checkability instead of testability, and then everything else I think is correct after that.

[00:29:47]  Unidentified speaker: Or again, that's my first guess.

[00:29:50]  Bruce: I could be wrong. I'm putting it forward as a conjecture. I would like people to try to find problems with it. And then I would like to see if we can improve it. Okay, that would be my point about all of this, that Ray is getting this correct, that demarcation criteria is far more important than you think. It is central to understanding Popper's epistemology. Even if I'm right that he got it a little bit wrong and you need checkability instead of testability as the demarcation criteria, it's still the center of his epistemology. And you just cannot do away with it, and without really ruining the usefulness of his epistemology. It's an example of hard-to-varyness, if you will. Okay, continuing reading. Furthermore, Popper maintained theory development ought to proceed from less to more testable. Does that sound familiar there, Peter? Meaning more informative theories. If a theory is refuted and an alternative sought, it had to be more testable, not less. And the more testable, the better.

[00:30:56]  Unidentified speaker: Popper's Ratchet here.

[00:30:58]  Bruce: Of course, I'm talking about Popper's Ratchet.

[00:31:00]  Unidentified speaker: Okay.

[00:31:03]  Bruce: For to reduce testability is to reduce knowledge, whereas in science, we desire the growth of knowledge. An immunizing strategy is always a development in the theory that reduces testability. Okay, I want to make a point here too. Ray is explaining Popper's Ratchet in one paragraph, probably better than what it took me a whole podcast to do.

[00:31:27]  Unidentified speaker: Okay.

[00:31:29]  Bruce: Ray wrote this book before I ever had the idea of Popper's Ratchet. I guess you could try to argue that great minds think alike, and it just so happens that I reinvented something that Ray invented. Of course, that's not actually true. Both of us got this from Popper.

[00:31:46]  Unidentified speaker: Because this is what Popper says.

[00:31:50]  Bruce: It's in Popper. I've read the quotes on the air for this podcast. Popper's Ratchet is my term. I did invent the term. But the idea is pure Popper. It is just straight from Popper, and it's exactly what Ray is saying here.

[00:32:05]  Unidentified speaker: Okay.

[00:32:07]  Unidentified speaker: All right.

[00:32:08]  Bruce: Popper's examples of immunizing stratagems. By the way, feel free to at any time break in, and we can discuss anything you want to as I read stuff, Peter.

[00:32:18]  Unidentified speaker: That would be just fine.

[00:32:20]  Bruce: I have no pre-thought out what I'm going to say either, so it's not like you're going to ruin my flow of thought or something. I'm mostly just reading and then we're riffing as we go here.

[00:32:32]  Unidentified speaker: Okay.

[00:32:33]  Bruce: Popper says that immunizing stratagems save theories from refutation. However, Popper's own example of immunizing stratagems undermines the claim that an ideology can maintain itself against criticism by logical means. Popper's examples are not examples of safe theories, but of repudiated theories. To immunize a theory in these cases is to abandon it. I think this is really true and I think this is completely insightful that Ray has picked up on this.

[00:33:01]  Unidentified speaker: Okay.

[00:33:03]  Unidentified speaker: And I don't think Popper picked up on this.

[00:33:05]  Unidentified speaker: So I think that this is something, I know I've mentioned this on the podcast, the idea that there's not a huge difference between changing a theory and abandoning the theory.

[00:33:13]  Unidentified speaker: I had the same idea in mind.

[00:33:15]  Bruce: This may be a case where Ray and I noticed the same thing. Again, Ray seems to like he's saying it better than I ever did. The two main effects of these immunizing stratagems are saving the theorist from embarrassment at the price of abandoning the original theory and clouding the issue while reducing information content. That quote, I'm going to quote again when we do the episode on dogmatism. The two main effects of these immunizing stratagems are saving the theorist from embarrassment at the price of abandoning the original theory and clouding the issue while reducing information content. The latter obviously interferes with the growth of knowledge. To go back to my earlier crude example, my belief that the world ended on a certain date is really a new belief. Because when I predicted the world would end on that date, I actually meant it would end in a way that everyone would be able to observe. The old belief has been replaced by a new belief, which is formulated in the same words as the old belief. The verbal formula has been saved, but the actual belief is different. I think that Popper was dimly aware that immunizing stratagems do not strictly save theories. In some cases, he put forward the word saved in scare quote marks. But he did not see the full implications of this, especially for the survival of an ideology.

[00:34:43]  Unidentified speaker: Okay.

[00:34:44]  Bruce: Does any of this sound somewhat familiar to what we've talked about on this podcast?

[00:34:50]  Unidentified speaker: I'm probably stretching.

[00:34:52]  Unidentified speaker: I think most of it does, right?

[00:34:54]  Unidentified speaker: Yes.

[00:34:55]  Bruce: What I actually have in mind here, think about our episode on words versus concepts.

[00:35:00]  Unidentified speaker: Okay. Okay.

[00:35:02]  Bruce: You can always reword things, but the underlying original concept can be the same, or you can keep the words the same and the concept can be different. Popper even went further. You can change the concepts, but the theory is the same.

[00:35:18]  Unidentified speaker: Okay.

[00:35:19]  Bruce: There's something underneath, and that's the thing we're getting at, and we have to get at it with words. One of the main things that a person who's in the thrall of dogmatism does, is they concentrate on the words instead of either the concept or the theory, preferably the theory, if not the theory than the concept. There's like different levels here, words, concepts, theories. Theory is the ultimate thing that we're going to test, so it's the one we care the most about or should care the most about. It's normative, not necessarily the way it is. Concepts, if you don't have a theory yet, are better than words. There are far more distinct concepts than there are words. Words are more like instruments. They don't matter as much. They do matter politically. I've always felt like Popper vastly misunderstood how valuable words were politically. I think his life shows him making mistakes like this where he just doesn't understand this concept sometimes. But words have political importance, but no importance in terms of the truth of the world. When Ray's Jehovah's Witness example here takes place, what they're really doing is that they're trying to keep the verbal formula the same so that they can claim the theory didn't change. But really, they killed one theory and they replaced it with a new theory. The old theory is refuted, okay? But they're in denial of that fact because they found a way to make the two theories use the same verbal formula, okay? Which you can pretty much always do, right? It's the easiest thing in the world to do.

[00:37:03]  Bruce: So this is why I always try to say, you know what, let's not argue over terms here. Let's just accept all terms. If this word means this to you and it means this to me, let's write down what it means to you, what it means to me, and then let's move on. I can never get people to do it. They always say, no, that's not what the word means. It's like, I don't know how to get past this. We've got to dig a little deeper. We have to use words, but you have to see the concept or the theory that I had in my mind, or you're never gonna understand the criticism I'm offering, right? As long as you're concentrating on just the words, you're lost. You're never gonna get it. And I actually think that that is an immunizing strategy, is to start arguing over words, over how words are used.

[00:37:45]  Unidentified speaker: Okay.

[00:37:46]  Bruce: Consider the simplest of Popper's examples. Popper asks us to consider the case of a man who makes the bold claim that all swans are white, and on being presented with a black swan, promptly denies that it is a swan. After all, this man says, whiteness is part of the definition of the word swan. Popper states that the theory that all swans are white has been quote saved, scare quotes from refutation. But has the theory really been saved? What has been an empirical theory about the world has now been turned into the application of a definition. The original theory supposedly protected by the immunizing strategy, has actually been replaced by an implication of a vacuous definition. This is like almost every discussion I have on the internet, where the moment we start getting into something that might be problematic for the other theory, there's an immediate retreat to vacuous definitions. And it always brings the whole conversation to a standstill. And I actually think that's the point, right? The reason why you immediately flee to a discussion about definitions is because you don't want to have to put your theory up against a real criticism. And you can always just start an argument over definitions. And there's no way out of an argument over definitions. It is a complete black hole to the discussion.

[00:39:09]  Unidentified speaker: So, okay.

[00:39:11]  Bruce: The original theory was empirical in Popper's sense. It was capable of clashing with observable reality. The statement, all swans are white and there is a black swan, cannot both be true. A definition or implication derived exclusively from a definition, however, cannot clash with reality, for they say nothing about the world, only about the way we choose to describe the world. The statement, all swans are white, which used to convey information about how the world is, has been transformed into a statement which tells us nothing about the world, but only about how we shall use the word swan. The original theory implied you will never come across a black swan. The new theory implies you cannot possibly come across a black swan because we have defined—redefined the word swan, so that being white is part of being a swan. Thus, the original theory has been repudiated, although the words of the original theory, all swans are white, has been preserved. The repudiation is implicit and unacknowledged, and thus saving face despite abandoning the original claim. Once this point is seen, we can derive some implications from the evolution of an ideology under criticism. In real life, people do not simply make such bold assertions out of the blue. Rather, they are made with a certain intention, background assumptions, and more or less, clearly formulated problems. It is the context of the assumptions and problems that both guides us in identifying an immunizing strategy and in refuting the original assertion.

[00:40:44]  Bruce: For example, the sentence, all swans are white, might be derived from a biological theory of the coloring of birds. Knowing this allows us to exclude a whole range of immunizing stratagems that contradict this biological theory or seem to make irrelevant the intention of maintaining the biological theory as a solution to the problem of coloring in birds. Provisionally, we may define an immunizing stratagem as an evasion of falsification by the reinterpretation of its theory or the modification of its assumptions so that the modified theory is then consistent with the critical evidence. The reinterpretation or modification must consist in a reduction of the information content, which is defined as the class of all and only those statements that are logically excluded by the theory. Scientific development, which can be described in terms of concepts, theories, problems, methods, and evidence, can be described in terms of concepts, theories, problems, methods, and evidence. We can classify immunizing strategies with respect to these categories. What he's going to do now is he's actually going to go over every type of immunization that is possible, or at least that he can think of.

[00:42:00]  Unidentified speaker: These are really good too.

[00:42:01]  Bruce: I'm going to read each of them also. This is something I really want to hold on to, and maybe we can reference back to this on the podcast from time to time and say, oh, this would be an example of a conceptual immunization or something like that. Conceptual immunization, for example, conventionalist interpretation of Newton's laws of motion, portray them as definitions, and thus, taken alone, these laws could not contradict the results of any imaginable experiment. What he's trying to say is, a type of immunization is a conceptual immunization where you just re-conceptualize things so that nothing could actually contradict the theory. Theoretical immunization, a theory that is contradicted by a true observation report e, may be weakened just so that it no longer implies not e, or it may be weakened in this way, but also strengthened by the addition of a new auxiliary hypothesis, so that e becomes a consequence of the altered theory. The theory cannot be made consistent with e simply by adding extra assumptions, something I examined below. Okay, let me talk about this one just for a second. This is the no ad hoc rule we're talking about. Okay, and what he's saying is that we can imagine an immunization strategy where you have a report e, it contradicts your theory, so you simply make your theory the original theory plus or e, and at this point, no theory can ever be refuted, okay? And so you just simply made little ad hoc adjustments to the theory. You could always do this. It's always possible to do it.

[00:43:40]  Bruce: Every theory can be ad hoc saved at any time, period, end of story, okay? Now, what he says, or it may be that you've made an auxiliary hypothesis so that e becomes a consequence, now that wouldn't be valid if the auxiliary hypothesis that gets added to the theory changes the outcome of the experiment so that e is now what we expect. And if that auxiliary hypothesis is itself independently testable, then this doesn't violate the no ad hoc rule, and this is no longer to be considered an immunizing strategy, even though it may on the surface seem very similar to an ad hoc save. Because of the fact the auxiliary hypothesis had its own testable consequences, it does not count as an immunizing strategy anymore. Now, he adds something that I'm going to cover in a little bit more detail in a second. He says, the theory cannot be made consistent with e simply by adding extra assumptions, something I examined below. Now, I've seen this said many times, and it is a strictly true statement that is misleading. And so when we get down to where he talked about this, I will explain why it's a somewhat misleading statement, even though it's strictly true. Okay, so immunization through change in a problem. A theory may escape a specific criticism by a change in the problem supposedly being solved by the theory. We will see that Freud does this with his theory of dreams in order to deal with the contrary evidence of anxiety dreams. What is the logical immunization?

[00:45:13]  Bruce: One's theory might be associated with a method which either deliberately or unwittingly excludes certain domains of potential falsifiers. For example, if a Freudian only considers evidence from the couch, then providing he sticks doggedly to this method, lots of non-analytical evidence will be made impotent. He may be alarmed to discover that he has been wrong for many years, or he may be simply ignorant of the relevance of such evidence. Immunization by reinterpreting or denying the evidence. The evidence itself may simply be denied or reinterpreted, or a port of a black swan may be put down as a hallucination. Physical immunization. One's theories may be attached to a meta theory that interprets them in a certain way. For example, one might combine catastrophic, combine catastrophe theory with the meta theory that all argument is illusory. This hybrid is purely hypothetical. If taken seriously and heeded, this would amount to an exclusion of all possible criticism of catastrophe theory, since it would exclude all possible criticism of any theory. Okay, so that's the end of his list of possible ways you could immunize a theory. So he goes on to say now, Popper's demarcation criterion is useful methodological advice if our objective is to promote the growth of knowledge. The term immunizing stratagem helps us designate those moves in theory development that [flout?] the criterion. In other words, if we're interested in gaining new theoretical knowledge, we better look out for immunizing stratagems and try to avoid them.

[00:46:59]  Bruce: Up to this point, Ray is not really disagreeing with Popper on anything. He's reasserting exactly what Popper has always said, and it's exactly what we've talked about on this show. Popper's ratchet, no ad hoc rule, don't immunize your theories. That's how you actually grow knowledge. Now, he says, but as protection against criticism, immunizing stratagems possess serious limitations. Now, this is the thing that he feels Popper has not emphasized enough, that even though people do use immunizing strategies in real life, there's a reason why you can't just immunize a theory and that theory has advantage over a non-immunized theory. But as protection against criticism, immunizing stratagems possess serious limitations and certainly do not provide an easy and thorough logical means of ensuring the survival of a theory or an ideology. In other words, immunizing strategies, while they may save you from refutation, they are serious problems that you have introduced into your theory. Many immunizing stratagems involve abandoning the ideology for whose protection they have been introduced. An unplanned, often unforeseeable process that consists of numerous successive slight modifications extending sometimes over hundreds of years. Other immunizing stratagems seriously lower the survival value of the ideology through the acquisition sometimes over a long period of a burdensome and confusing protective belt of hypotheses, each of which acted at least in the short run to deflect criticism away from a privileged sector of assumptions.

[00:48:37]  Bruce: I think anarcho-capitalism fits that, what he just said perfectly, by the way, where they have a protective belt in the set of ways of looking at things. I think it has harmed the theory. It does make it so that you cannot criticize the theory to an individual who's an anarcho-capitalist, but I think it's also why the theory just has no real chance of ever catching on. Why it's, for the most part, pretty harmless, at least inside of an open society. Now, continuing with Ray here, my really different take is that the changes may not be designed, but may be unintended consequences of an attempt to deal with criticism and retain the theory. To the extent that the maneuvers abandon the original doctrine in response to the specific falsification involved, they cannot be wholly arbitrary. This reinforces my point that falsification can act like a Darwinian-like filtering device, as a Darwinian-like filtering device on ideologies, even if evasive, intentional or unintentional moves occur. It may be that although each successive immunizing strategy is intentional and produced in the knowledge that the ideology is being altered only slightly, the whole sequence of immunizing stratagems and their accumulated effects is unplanned and unforeseeable. An analogy with the evolution of language might clarify my point. Even if every change in the language were a conscious innovation, the total effects of all the unintentional ramifications of these intentional changes cannot be foreseen.

[00:50:11]  Bruce: No one living in Medieval England, for example, could have predicted the shape of today's English language. Okay, let me explain this because I feel like this is such an important point that he's making. And I want to kind of lay it out in my own way and offer some examples of this.

[00:50:28]  Unidentified speaker: Okay.

[00:50:29]  Bruce: When I am talking to someone, often a crit rat, but maybe a Bayesian, maybe back in my religious blogging days, maybe a believing Mormon or a non-believing, but liberal, practicing Mormon. One of the things that I have noticed is that people will just tweak their theories on the fly. And this is, you know, [violates?] the no ad hoc rule, but nobody knows about the no ad hoc rule.

[00:50:57]  Unidentified speaker: Okay.

[00:50:58]  Bruce: So they just make a little tweak here, a little tweak there. And it's tempting to say that, oh, that person is just dogmatic. There's no way I can get through to them, which might be true, by the way. It's even, Ray admits, you might be dogmatic your whole life, which effectively makes you fully dogmatic, right? On some particular point. You would never be dogmatic over everything.

[00:51:19]  Unidentified speaker: You would just be on something.

[00:51:23]  Bruce: But you're still, something's still happening to your theory.

[00:51:27]  Unidentified speaker: Okay.

[00:51:27]  Bruce: Every time you make a little tweak to your theory, the theory has changed a little bit. I would even go so far as to say, the original theory has been refuted, and replacement theory has shown up. It's one with, in this case, less content. And you don't know what the implications of that change are. It's not possible for you to know that. Let's use the Brett Hall theory of IQ or of intelligence as an example, something that we've used on this podcast a number of times. They would deny any problems with their theory. They would say, here's our theory. We think that it's just a matter of time and interest. Intelligence is a matter of time and interest. Everyone's equally intelligent in that regard. That's what really matters. They would have this theory and then they would work out the moral implications of it. They'd say, that means that no one can tell you, you can't be an astronaut. If that's what your real time interest is, you can do it. There's nothing that you're going to be bad at, etc. Which, by the way, isn't really necessarily bad moral advice, right? It's not like I'm arguing with the moral here, okay? It's really the moral, I think, that they were interested in, not so much the theory, okay? So I would come along and I would try to refute it, because I'm trying to treat it like a theory.

[00:52:45]  Unidentified speaker: I'm a critical rationalist.

[00:52:46]  Bruce: And I would say, well, if it's just a matter of time and interest, then how do you explain senility? This person's been alive longer, their interest hasn't changed. They've had more time because they're older, and yet they actually become less intelligent and less capable because they're becoming senile. And every single one of the defenders of Brett's theory would say, oh, yeah, well, you know, IQ theory sucks. Like they would just do a what about-ism. They wouldn't even attempt to try to explain it, right? This is an example of where, and then they would say, this theory is philosophical.

[00:53:21]  Unidentified speaker: So it can't be tested.

[00:53:22]  Bruce: So I don't even care about your test, right? The thing that was actually happening here is that they really intended this as a metaphysical theory from the outset. And they thought it was. And so they thought it could not be tested, and they had thought they had immunized the theory fully, so they could just hold on to the moral implications. And what I was really showing them was, actually you failed to fully immunize this theory. It still has implications. And if they even tried to respond to me, it was still going to have implications. There was no way they could change this theory to never have implications. So they had to instead just not [discuss it?], right? They had to just say, you know what? IQ theory sucks, so I don't care. And do a what about-ism. And that was ultimately what they had to do to immunize the theory, because it was basically impossible for them to keep ad hoc saving the theory without me eventually finding another implication that was empirical that they hadn't thought of. And this is really what Ray is getting at here, is that you may even intentionally be, intentionally or unintentionally be making an immunizing strategy move like this. And yet it's really hard to get rid of all the implications. In fact, it's probably impossible because there's probably an infinity of them. And in the end, you're going to end up with a theory that has all sorts of implications you did not intend, and you're going to probably end up with either a super bad theory or one that's totally different than the one you intended.

[00:54:57]  Bruce: In this case, you're going to end up with a theory that maybe doesn't have the moral implications you were hoping for, you know, or you're just going to end up abandoning it yourself and finally just go, you know what, I just, I can't do anything with this theory. And because of this, this theory has a serious disadvantage compared to its competitor. In this case, IQ Theory, as bad as IQ Theory is, IQ Theory is a much better theory than Brett Hall's Theory of Intelligence. Not so much because it has to be, but because of the immunizing strategies they were using. If you actually were okay with not using immunizing strategies, I think Brett Hall's Theory of Intelligence shows real promise. But you had to go into it as a theory and say, you know what, that can't explain synulity. There must be a case where there can be brain damage that can lower your intelligence. How does that impact my theory? I'm still going to say that in general, it's based around time and interest. There's an exception case, specifically damage to the brain. Notice that this is exactly the theory that Nicholas Taleb came up with. His was very similar, had the same advantages as the Brett Hall Theory of Intelligence, but it didn't have the immunizing strategies that ruined the theory to go along with it. That just made it a much better theory. Even Nicholas Taleb’s had implications he didn't wish it did. He was trying so hard to cover the implications that he didn't like. We cover all this in past podcasts, by the way.

[00:56:30]  Bruce: I think this is an example of what Ray is really talking about, where you do not control the implications of your theory. You may think you do, but you do not. Working to make your theory have zero, to be truly metaphysical where it has zero testable implications or at least checkable implications, it's so much harder than you think. You will just keep accidentally making your theory and reimagining it and it's going to have implications and they're going to be wrong. Then you're going to try to change it again, it's going to have implications, they're going to be wrong. If you don't have the attitude of the critical attitude that we're seeking here, of, oh, my theory is wrong, let's make a new one. Instead, you just keep trying to protect the essence of that theory. Then you're going to end up with something that's total [pap?] by the end, that you can barely do anything with. All right, skipping down a bit. Popper begins with a rough characterization of bold ideas. So I've talked about bold ideas in this podcast quite a bit. A theory is bold if it is a new daring hypothesis. It is daring if it takes a big risk of being false. Popper argues that this risk can be analyzed ultimately in terms of the amount that the idea excludes, the degree to which it forbids states of affairs. Severe attempts at refutation are severe critical discussions and severe empirical tests.

[00:57:53]  Unidentified speaker: Again, does this sound familiar?

[00:57:56]  Bruce: This is the Noad Hock rule. This is the idea that we should be making a choice to formulate our theories so that they take big risks. I've explained in the podcast why this is. If you're assuming realism, Popper does start with the assumption of realism, then it should be the case that a true theory can always be formulated in such a way that it takes big risks, that it constrains reality in some way, and therefore can be tested or checked. Okay, where checked means test, except that it implies like logical implications instead of empirical implications. So this idea of a choice to formulate your theories, notice that Ray has caught this, it's coming straight from Popper. Recall that in my discussions with CritRats, I had at least one say, well, you never know which direction is going to go. You have a problem, you're going to solve the problem. You can't predict where knowledge is going to grow. So it may be that actually you solve the problem by making the theory vaguer to reduce the empirical content. This is a CritRat idea. We can assess it on its own if it's correct or not. It's definitely not what Popper had in mind. Popper had in mind that you have a choice to formulate your theories boldly such that they take risks, which is exactly the same as increasing their empirical content, which is exactly the same as falsification.

[00:59:20]  Bruce: When Popper talks about conjecture and refutation, he has this idea in mind that you're always going to choose to increase the falsifiability and refutability of your theory by making it increasingly bolder, that you solve problems by doing that, not by making it vaguer and reducing the content of the theory.

[00:59:45]  Bruce: Just an explanation here, I'm skipping over part. Popper illustrates the idea of what I just read in the development of Heliocentric Theories of Copernicus is the right way to say that. Now compare that to like say Einstein's Theory of General Relativity, okay? So Popper points out that the original two, Copernicus and Aristocratus, or how to pronounce that term, that name, he points out that neither of theirs were fully scientific. They had these theories and they were not fully scientific in Popper's way of thinking. Because neither of them was bold enough to predict new observable appearances and thereby expose their theories to new empirical tests. Note that this is the Noad-Hawk rule. This is Popper's ratchet, okay? So their theories explained the known appearances, known observations, but did not explicitly suggest the existence of unknown appearances, appearances that might decide between the heliocentric and earth-centered views. If they had made such predictions, their theories would have been much more informative and would therefore have taken a larger risk of being false. But they would have have better promoted, so they would have better promoted the growth of knowledge. Now, here's the point that I think is correct that Ray is making. There is a certain boldness in the heliocentric theories. So Copernicus comes along, he says, hey, maybe actually the sun is the center of the solar system. They didn't have a concept of solar system, but what we would today call the solar system, and the earth isn't the center.

[01:01:26]  Bruce: This is bold in the sense that it broke with Aristolian tradition. But it is not bold in the Popperian sense that we care about for Popper's epistemology. Now, this is again one of the problems I've bumped into. With crit rats, they'll often say, oh, my theory is bold. It makes bold moral implications that nobody's thought of or something along those lines. I'm not denying that the theories might be bold morally in that sense, just like any heliocentric theory is bold in that sense. What Ray is saying is, this isn't really what Popper had in mind. What Popper had in mind was that it actually made new unforeseen predictions that it was not it had independently testable consequences, that it was not ad hoc, okay? Now, this doesn't mean that the heliocentric theories are valueless just because they're ad hoc, which they were. They obviously were the start of something better. And I think you do have to see this not as a is all good or it's all bad or something along those lines. It's more a matter of what a critical attitude is, okay? And I don't mean that in some generic sense that everybody thinks that they're critical. I mean it more like, do you sort of see this as a stopping point? Or do you see it as, okay, what's next? When you come up with a theory, there's certain known problems with the observations of planets. It doesn't fit their theory of how this should work because they're trying to do it centered around the Earth. They solve those problems by moving to a heliocentric theory, but it doesn't have any testable consequences, independently testable consequences.

[01:03:07]  Bruce: Does that mean that the heliocentric theory wasn't an improvement? No, it was an improvement. It actually, the fact that it solved the problem of these additional observations that didn't fit the old theory. In some sense, you might even say this heliocentric theory is now the better theory. But it doesn't really go far enough. If realism is true, there should be a way to re-imagine the heliocentric theories so they do have independently testable consequences. Of course, that's Newton's Theory or Einstein's Theory, right? There really was a better theory out there that explained this with independently testable consequences. Realism means there will always be such a theory, so you shouldn't rest on your laurels and say, look, I've solved the problem, I'm done. You should take seriously this idea that your theory needs to have additional consequences. Now, maybe you don't know how to do that.

[01:03:56]  Unidentified speaker: That's fine.

[01:03:57]  Bruce: Maybe you go with the best theory at the moment, and that's the one that's still ad hoc. That's fine.

[01:04:02]  Unidentified speaker: This isn't an, "we're going to do this and it's normative and you have to do it."

[01:04:06]  Bruce: It's more a matter of understanding what it is we're trying to do when we grow knowledge and why there should be independently testable consequences, and it's up to you to figure out how to conjecture that theory so that it has the independently testable consequences. Why we should expect that you should be able to do that if you're right? Continuing Ray, each theory, though false, represented an attempt to increase knowledge and did so because even though each was false, it had greater truth content than its predecessor and exposed itself to more tests. Popper's answer then is that it's a theory that it is a theory's openness to empirical refutation that makes it scientific. But more generally, it is whether the theory is an attempt to expand our knowledge, whether it represents an increase of information on the theory it replaces. No ad hoc rule, Popper's Ratchet. Marxism or Freudianism would not be counted as unscientific because they have been refuted. But because of the evasive way Marxists and Freudians, Marxists and Freudians have dealt with possible refutations. Now, this gets to exactly what I've been saying on the podcast, that when it comes right down to it, hard-to-vary-ness and easy-to-vary-ness have more to do with the attitude of the defenders of the theory than the theory itself. Every theory has aspects that are somewhat vague, because no theory is nailed down all the way, and so all theories have some easy-to-vary content within them.

[01:05:39]  Bruce: What really matters when we're talking about ad hocness, which I think is the better criteria, is that it has to do with the attitude of the people who are developing the theories. What really makes Marxism and Freudianism problematic isn't the theory itself. It's actually that the defenders of the theory refuse to follow the Noad Hock rule, and instead would simply immunize their theory over and over again, continuing with Ray.

[01:06:06]  Unidentified speaker: What is important for the demarcation criterion is a critical attitude and the proposal of increasingly falsifiable theories in response to refutations. Increasingly falsifiable theories in response to refutations.

[01:06:20]  Bruce: Popper's Ratchet. Kepler's elliptical orbit hypothesis represents just this sort of increase of information content in response to empirical refutations. One of the things he points out that I did not read is that Copernicus didn't have any independently testable consequences. Kepler did. That made Kepler's theory the better theory. Even though Copernicus' theory theoretically was better than the previous theory, its lack of independently testable consequences really was a problem. That's why Kepler's was just the better theory. Popper's proposal was that science was distinguished from non-science by two things.

[01:07:00]  Unidentified speaker: One, the boldness of predicting as yet unobservable phenomena, especially phenomena.

[01:07:06]  Unidentified speaker: So notice this is him summarizing Popper.

[01:07:08]  Unidentified speaker: Popper's proposal was that science was distinguished from non-science by two things.

[01:07:13]  Unidentified speaker: One, the boldness of predicting as yet unobserved phenomena, especially phenomena which will pit the theory against its competitors and allow us to decide between them.

[01:07:23]  Unidentified speaker: No ad hoc rule, independent testability. Einstein was acutely aware of the need to compare his theory with its competitors.

[01:07:30]  Unidentified speaker: Two, the boldness of looking for tests and refuting instances.

[01:07:33]  Unidentified speaker: I would also add the boldness of accepting refuting instances, which is not implied by the boldness of looking for them.

[01:07:40]  Bruce: We may generalize the methodological conclusions of Popper's investigation as follows. So this is three points that he thinks, he's claiming summarizes Popper's view. One, propound empirically testable theories. Two, aim to refute them. Three, given any theory T, aim to replace it by another theory T star, which is more general and precise, has higher information content. One that explains the success of T, explains the refuting evidence of T, and is moreover independently testable.

[01:08:15]  Unidentified speaker: Okay.

[01:08:16]  Bruce: Again, I want to point out, Ray is saying exactly what I've been saying on the show, right? And there's just no way you can explain this, if you don't realize that we're both just getting this from Popper. This is Popper's epistemology. This has always been Popper's epistemology.

[01:08:29]  Unidentified speaker: Okay.

[01:08:30]  Bruce: Now, Ray says, in his early writings, Popper would have phrased one as propound only empirically testable theories. It says propound empirically testable theories.

[01:08:41]  Unidentified speaker: He's throwing it only in there.

[01:08:42]  Bruce: Popper later put much more emphasis on the importance of non-empirical theories while retaining empirical content as the ultimate goal of theory development.

[01:08:52]  Unidentified speaker: Okay.

[01:08:54]  Bruce: Again, one of the things I've pointed out is that when you put up a empirically testable theory, a lot of crit rats I know, will put up a philosophical theory as a competitor and then declare there's the winner because there's some problem with the empirically testable theory and no problems imaginable with their philosophical theory. That isn't what Popper had in mind. He had in mind this idea that you just do not put philosophical theories up as competitors to an empirical theory because empirical theories are in a special category. Notice that that is this whole idea that Ray is discussing about propounding, originally propounding only empirically testable theories, those were the valuable ones, to later, well, prefer them, propound empirically testable theories, dropping the only. One of the things that Ray is going to say here in a moment, that I agree with him on by the way, is that number one is not necessary. That actually points two and three here, those are the ones that matter. Aim to refute theories, given any theory T, aim to replace it by another one that is more general and more precise, more specific, and has more higher information content. If you do those two things, you don't actually need to worry about if your theory is explicitly empirical or not. This is why I have tried to declare checkability as the right demarcation criteria rather than testability as the right demarcation criteria. So Ray is going to end up agreeing with me on this indirectly. But I'll let him explain as I read forward here.

[01:10:26]  Bruce: There are two aspects to the demarcation criterion, one of attitude and one of pure logic. Notice that this is something I've been saying on the podcast. Firstly, the scientist must try to find falsifying instances of his theory. This is a matter of correct attitude.

[01:10:41]  Unidentified speaker: The critical attitude.

[01:10:42]  Bruce: Secondly, the scientist must have at his disposal refutable theories. The possibility that arises of a scientist earnestly following the first injunction without realizing the theory he is dealing with is empirically irrefutable. Equally, a body of theory may be logically capable of refutation, though its adherents have refused or neglected to look for refuting instances. This is the point I was trying to make with the Bret Hall Theory of Intelligence. They weren't even looking for refuting instances that would allow them to error correct their theory. Since Popper is interested in the growth of knowledge, he is most concerned to discourage the use of immunizing strategies that flout the demarcation criterion, effectively reducing the information content of our theories. Kepler, for instance, could have described the planets that did not fit his master's model as not really planets, using a definitional change. After all, he might have said, planets do not behave like that. Planet is essentially an object with a circular orbit. This would have been an example of what Popper called an immunizing stratagem. Scratch a maneuver, Popper would say, saves the theory but at the price of a reduction to its information content. As we have seen, Kepler's actual response greatly increased the informative content of astronomy and is rightly admired for that. Now, here's where interesting point that Ray makes. Not all evasive moves are on the wrong side of the demarcation criterion. What he's going to now describe is going to be exactly the no ad hoc rule. Okay, Popper's ratchet.

[01:12:17]  Bruce: Some auxiliary hypotheses introduced to deflect a refutation from a valuable assumption, have added greatly to our knowledge. One such auxiliary hypothesis was the predictions by Adams and Le Verrier of the existence of the planet Neptune. It has been observed that the orbit, it had been observed that the orbit of the planet Uranus was not in accord to Newton's core theory, the laws of motion and the laws of gravity, plus the then known initial conditions, assumptions about the gravitational influence of other planetary bodies. Newton's theory could have been regarded as falsified by this anomaly. However, Adams and Le Verrier proposed the existence of a previously unknown planet to account for the failure of the predictions, thus saving Newton's theory. But this particular evasion brought increased information content to the Newtonian system, as was clear from the fact that the hypothesis was empirically testable by independent means, independent in the sense of doing more than simply checking whether the hypothesis agreed with the already observed perturbations of Uranus. In other words, if you introduce an auxiliary hypothesis, and it is itself independently testable, you can actually go find the other planet and observe it. Then there is nothing wrong with saving your theory through that, that's valid. So Ray goes on to say, this next part is really important, I think, by the way.

[01:13:40]  Unidentified speaker: Okay.

[01:13:41]  Bruce: He's going to explain why rule one is unnecessary, and it's exactly the reason that I've given on this podcast. He says, in 1957, Popper became very interested in the fact that metaphysical theories could not only be inspiring, but also arguably arguable and open to criticism. He held that doctrines such as determinism, that do not admit of empirical refutations, are nevertheless open to criticism as to their effectiveness at solving the problem for which they were proposed. Now, we covered this in our episode where we went over how Popper said to go about handling metaphysical theories, and I was actually very critical of Popper on this, and I still am. His choice of determinism here was about as bad as it could have been. Number one, determinism is actually true. He's wrong that he thinks that he could criticize it, and he's eliminated it as a possibility. Secondly, it's not just some metaphysical theory that people have proposed out there to solve some problem. It's an implication at the time then known laws of physics. That's a totally different status. I feel like Popper just missed this entirely. What an important thing that is, that it actually did follow at that point in time Newton's laws of physics, or Einstein's laws of physics. It's been challenged through quantum physics. Now, challenge, you can decide how to interpret this. Quantum physics shows that the universe is actually non-deterministic, but the multiverse is deterministic.

[01:15:12]  Bruce: So we might say that, look, if I'm looking at the universe, then in fact, determinism is wrong, and you can actually refute determinism by refuting the theory that implied it, because that theory itself has implications. I feel like the way Popper went about this was almost entirely wrong, but he got to the right conclusion. In light of this discussion, this is Ray again, I would like to suggest that the discussion is that there are metaphysical theories that can be criticized in the same way that you can criticize an empirical theory. In light of this discussion, I would like to suggest that the three methodological rules discussed in the previous section may be simplified by eliminating Rule 1. Rule 3 takes into account the injunction to move from metaphysical speculations to empirically testable theories. Notice that Ray is reading Popper as saying, don't rest on your laurels with a metaphysical speculation, reformulate it as an empirically testable theory. This is the thing I've argued with this crit rat friend of mine about. He wanted to say, no, you might not just go anywhere. You can't predict the way knowledge is going. That is not what Popper meant. Popper had this idea that you were going to choose to formulate your theories to be as testable as you could. Going back, we read it again. Rule 3 takes into account the injunction to move from metaphysical speculations to empirical testable theories, as well as the injunction to move from less to more informative metaphysical theories.

[01:16:41]  Bruce: If it is accepted that what is important is to move from less to more informative theories, then interesting conclusions follow. Of course, the interesting conclusion is that checkability is what matters, not testability. For example, here we come. This is the thing I loved what Ray said here. Even if Marxism has been made into untestable metaphysics, it could be made testable again. You could choose to formulate it so it's testable. Equally, Freudianism, which was never testable from its formulation, could be made testable. A Marxist or Freudian could be shown how the theories could be interpreted empirically and promptly refuted.

[01:17:31]  Bruce: This need not be as arbitrary as it seems. Even lovers of metaphysics are constrained by their speculations, by a whole network of what they regard as background knowledge. Which may consist of both empirical and metaphysical theories, and by their problem situation.

[01:17:46]  Unidentified speaker: I think that's beautifully put.

[01:17:48]  Bruce: I think Ray has really nailed this one here completely. Moving to empirical versus metaphysical criticism. But more to the point, it does look as though the ideologies most infamous for their apparent obstinacy in the face of criticism, take on a metaphysical form. He's here talking about what I would call the Crip-Rat loophole. This idea that if you really cherish your belief in some theory, you aren't going to open it up to empirical testability. Instead, you're going to declare it philosophical or metaphysical. Then you're going to say, oh, look, my theory happens to be metaphysical. There's always seems to be this assumption that it's not, that the theory just has this, not to keep picking on the Bret Hall Theory of Intelligence. I think they really thought this theory just happens to be philosophical. And it was a shock when I showed them, no, it has empirically testable consequences. So they wanted to just blanket say, no, it's metaphysical, I don't care. I'm not even going to deal with this. And I think that that is what we call dogmatism.

[01:19:03]  Unidentified speaker: Right?

[01:19:03]  Bruce: It is when we that dogmatism is that the defenders of the theory are very obstinate at when they face criticism, they make their theories take a metaphysical form to try to avoid having to put up with the danger of being refuted by observation.

[01:19:21]  Unidentified speaker: Okay.

[01:19:23]  Bruce: One might argue that economics does not make predictions of the same empirical precision as does chemistry. Some might, because think about our podcast on roughly testable theories and Hayek's approach, where he claimed that economics was a scientifically testable theory, but he had to make special provisions for how to deal with it at a different level of abstraction because it's so difficult to test in real life. We might argue that economics does not make predictions of the same empirical precision as does chemistry.

[01:19:54]  Unidentified speaker: Fair enough.

[01:19:55]  Bruce: Some might even argue that economics is not empirical at all. This was the argument I was having with my CritRat friend. Hayek agreed with me that in fact it was empirical. But you can see where my friend was coming from, that he wanted to declare economics metaphysical because it is very difficult to test. Some might even argue that economics is not empirical at all, but a very suggestive and true metaphysical theory. The analogy with chemistry would then be weakened, but we can certainly say that the neo-classical price theory has greater informative content than the Marxist unrestricted singular prediction, and may still help to undermine the Marxist case. What I would say here is that while I do think economics can in principle be tested and should be tested empirically, he's actually right that there are checkable consequences to economic theories even when you can't test them. And the checkability is what really matters, not testability, okay? It's easy, this is Ray again, it's easy to assume that empirical observation is the strongest critic. The implication would be that if a network of ideas succeeds in shielding itself from empirical counter-evidence, it will have evaded, if not all forms of criticism, at least the most damaging by psychological, both psychologically and logically. I think that’s, by the way, completely true. I think that that is exactly why defenders of a theory, if they're being dogmatic, will immediately move to my theory is metaphysical. Having said that, Ray now says, this may not be true.

[01:21:41]  Bruce: An interesting possibility is that the opposing metaphysical theories may somehow be of greater weight than empirical observations. We covered this in the episode about Popperianism versus Deutschianism. Watkins has shown how metaphysical theories serve to filter out some possible theories before they even enter the body of science. These theories do not even get discussed because they conflict with the prevalent metaphysical background assumptions. This is a place where perhaps me and Ray disagree. I'm not actually clear. I think in terms of what he strictly says here, there's nothing I really disagree with. It's often the case that a dogmatist will declare their theory to be metaphysical, to try to evade the strongest form of criticism, empirical criticisms. It's not always the case.

[01:22:33]  Unidentified speaker: I would agree with that.

[01:22:35]  Bruce: I don't know. I agree that things that we call metaphysical theories are often more important than what we call empirical theories. In the Deutschianism versus Popperian episode, I gave the example of how I agree realism is in some sense foundational to Popper's epistemology and maybe even could be thought of as having a precedent over empirical theories. Where I did argue back though, is that I think realism has real consequences that you might even call empirical. Nobody does, but I don't understand why we don't call them empirical. Like if I go kick a rock and it kicks back, as from David Deutsch's book, to me that's kind of an empirical consequence. If you don't want to call it an empirical consequence, it is at least a checkable consequence. My guess is that while it's true that things that we call metaphysical theories are sometimes more important than empirical theories, I don't think you will ever find a case where the, maybe I'm wrong, like somebody find a case for me. But I don't know that you will find a case where checkability isn't the main filter. So a metaphysical theory can have logical consequences, maybe not empirical consequences. Those should be considered as important as empirical consequences. Again, the Church-Turing thesis is probably the best example here, most straightforward example here. I don't think it's any doubt that even if you want to declare Church-Turing thesis metaphysical because it doesn't have quote unquote empirical consequences, it's such a basic theory to everything else we do.

[01:24:14]  Bruce: And the reason why it's such a good theory, even though it's a quote unquote metaphysical theory, is because it's a good metaphysical theory because it's highly, highly, highly checkable even if it's not empirically testable. So my point of view is, is that you need to move the demarcation criteria to include many metaphysical theories like Church-Turing through the use of checkability rather than testability, and that the demarcation criteria now takes on the original importance that Popper intended it to have. I think this may explain, Ray may disagree with me on this. Maybe one of these times I'll have to ask him about this. I think the reason why he's right that sometimes metaphysical, this is a quote from him, metaphysical theories may sometimes be of greater weight than empirical observations. I think he's right about that, but I think he's only going to turn out to be right about that, if we're talking about checkable metaphysical theories.

[01:25:08]  Unidentified speaker: Maybe I'm wrong.

[01:25:09]  Bruce: I would actually like to see people offer, kindly offer counter examples. It may well be that my attempt to move the demarcation criteria improved it, but still has the same problems that Popper's did, that there are still counter examples to it, which I would find exciting. That would be an exciting possibility that maybe there is still a need to improve this demarcation criteria in some way. Damaging versus eliminating a system of ideas. This is another really interesting idea that I totally agree with Ray on. In correspondence with the late W. W. Bartley III, partly conceding my point, he argued that, quote, this is now Bartley's, this is now Ray quoting Bartley in a personal communication. In a strict sense, the introduction of an immunizing strategy may be tantamount to abandoning the position, but in practice, it is more likely to be tantamount to damaging the position. Now, why is Bartley and Ray arguing this? He actually does explain this in the passages I didn't read that came before this, because I didn't want to read the entire chapter. Here's the idea that Ray is getting at, and why Ray is right and Bartley is wrong.

[01:26:22]  Bruce: Bartley is holding onto this idea that there's a difference between refuting a theory and damaging a theory.

[01:26:30]  Unidentified speaker: So imagine like, the easiest way to explain this is actually through an example that people aren't super familiar with.

[01:26:37]  Bruce: Think of atomic theory. Atomic theory dates way back to the Greeks. Their original theory was dumb by today's standards. There's these little tiny atoms, and they're indivisible, and you can't split them. Maybe they're sharp, and so when you put them on your tongue, they taste bitter. It seems kind of foolish by today's standards. But it's a well-known fact that those original theories that started off completely untestable, that they got criticized until they became what is today modern atomic theory. There's a complete line between those original theories and modern atomic theory as they simply just kept adjusting it. You could argue that atomic theory never got refuted, that it turned out to be right. Or you could argue that atomic theory was wrong, and that what we have today isn't really atomic theory. I've seen people argue both of these. It's a misunderstanding of epistemology, that either of those arguments is a misunderstanding of epistemology. Of course, the original atomic theory got refuted and got long sense replaced by a totally different theory that was inspired by the earlier one that was wrong. That's what really happened. So what the person is trying to do is if you're taking the stance that the essence of the original theory is that you can't split the atom, which the Greeks made a huge deal about, then you could point out that atoms today in modern atomic theory can be split, and you can say, see that original theory was false and we have a totally different theory today.

[01:28:27]  Bruce: You're doing it based on the assumption that the essence of the theory is that the atom can be split. Do you see the argument here? There is an intelligence to it, but it's essentialism. It's not actually a good argument at all, because it's just an argument over definitions basically. Or you can take the opposite point of view. You can say, it's true that what we call an atom can be split, but you can't split quarks, or at least we don't know how to as of today. Really, this is still atomic theory. Again, using the same essentialist argument, you can claim that actually the Greeks have been right all along. This is the exact same theory, everything else was just footnotes to the original Greek theory. Well, both of these are stupid. They're just really total misunderstanding. They're the essentialist mistake. What's really going on is that we've had a giant series of refutations of theories and replacements with better theories until we arrived at one that was super modern and useful and scientific. That's what really happened. There's no difference between damaging a theory and refuting it. It's really just an arbitrary choice to call it damage or refuting based on what you think personally the essence of the theory was. This applies to everything.

[01:29:46]  Unidentified speaker: [Is Newton—is Newtonian?] Is Einstein's physics, what did it refute Newtonian physics?

[01:29:52]  Bruce: Or is Einstein's physics actually just a footnote to Newtonian physics? Tipler argues it's just a footnote.

[01:29:58]  Unidentified speaker: There's a good basis for doing this.

[01:30:00]  Bruce: He can show you that in retrospect, we can show the Newtonian physics implied curvature. He has this whole paper where he writes this out. Based on that, he shows that actually, Einstein's theory was not a paradigm shift that we think it is, and everybody's mistaken, and it's total boulder dash. Of course, nobody perceived Newtonian physics as that at that time. You can either think of Einstein's theory as neo-Newtonian physics or you can think of it as a complete replacement and the old theory got refuted. In other words, you can think of it as it damaged the theory. You can think of the Eddington expedition as having damaged Newtonian theory or you can think of it as refuting it. It turns out there's just no difference between these two. They are exactly the same thing. Here's Ray responding to this. He says, Jeremy [Shearmur?] has made a similar criticism of my thesis. I think this contention is true but misleading. Drastic revision of a theory through use of an immunizing strategy are rare, for they are too obvious and unconvincing. The revisions are more often of a marginal nature. Bartley and [Shearmur’s?] disagreement with me rests on an unexamined assumption of theirs that there is a difference between modifying a network of ideas and making a new set of ideas. A form of essentialism that I view as mistaken. Go Ray. I'm so glad to see a Paparian get this. One might retort that a network of ideas may evolve yet survive, in the sense that the fundamentals are retained.

[01:31:41]  Bruce: My immediate response to this would be to ask how fundamentals of a network of ideas would be defined other than as the elements of the modified network of ideas that are retained. Definitions are rarely important, but asking for a definition here could be revealing. Even if we accept for the purposes of argument that to damage a position is not to eliminate it, the distinction breaks down when we look at the history of ideas. Metaphorically speaking, a sufficient number of injuries to a theory is always equivalent to its death. Each intentional or unintentional concession made by an ideologue may be individually insignificant, but a sufficient number of insignificant differences makes a significant difference. Successive slight modifications may lead from orthodoxy to radically different interpretations to what would early have been viewed as heresy. He then says, this example promoted me. He talks about the example of the Jehovah's Witnesses in the six days of creation. We already talked about this. He says, this example promoted to me the thought that in analyzing the evolution of network of ideas, we ought to distinguish the following. One, the uninterpreted terminology of the doctrine as embodied in books and other sources. Two, the interpretation placed on the terminology. Three, the interpreter's theory about how his interpretation compares with previous interpretations, his own and others. I actually just had a recent discussion with a person on Twitter where we were discussing how does a book have knowledge, can it have knowledge that is distinct from a human?

[01:33:34]  Bruce: Since like, let's say I read the Bible, I get one thing out of it, another person gets something out of it, it's the exact same words, how can you say that that book contains knowledge when you clearly need a human there to interpret it? Of course, he was sort of arguing for the non-existence of the third world, Popper's third world, the idea of objective knowledge. He wanted to see knowledge as being ultimately depending upon an actual human brain, human mind that interprets it. I was trying to explain the three worlds to him. This idea that the book contains its own knowledge, but the knowledge that is placed in the mind, is in the mind, the interpretation, is a different set of knowledge that is just based on the book. And I don't think he was quite buying it. He was kind of buying it maybe. But notice that this is what Ray is really saying, right? That we have to make a distinction here between the different worlds. You can't say the book's knowledge is equivalent to the person's interpretation because then the book has multiple sets of knowledge. We were actually, the two of us on Twitter, we were playing with different ideas of how you could really screw this up. Imagine, for example, that an author has X in mind when he's writing something, he makes a mistake and he writes Y. It just so happens that Y is correct and X is wrong in some way. Maybe X is partially correct and partially wrong. Now, the original author had X in mind, that was his interpretation. The book has a different set of knowledge, Y. It's a mutation from what the author intended.

[01:35:05]  Bruce: Then everyone who reads the book comes away with Y. Possibly, they come away with Z, maybe they misinterpret it and they go away from the true answer. But let's say they come away with Y. It may even be that you can't even necessarily look at the intention of the original author. The book may have knowledge objectively different than the intention of the original author. I don't know how common that is, but you can imagine it happening. The book has to be kept separate in its own world, the third world, distinct from the second world, the subjective knowledge. Keep in mind the subjective knowledge just means knowledge in brains. It doesn't mean like relativistic knowledge or anything like that. Popper's language, the philosophical language here is different than the way we'd use it today.

[01:35:51]  Unidentified speaker: Okay.

[01:35:53]  Bruce: Now, I think we're getting to the end here, but he goes over types of immunizing strategies. I wanted to maybe just call each of these out quickly, denying the refuting evidence E. A couple of points he makes about this that I thought were important. Suppose the refuting evidence E is denied under all circumstances. For example, in the swan case, the person who advances the theory that all swans were white, may simply deny that any black swans present to him is black. At first glance, this may look like a case in which the original theory is abandoned. But let's look more closely at least some of the information content of an empirical statement is logically equivalent to the class of basic statements with which perhaps in the presence of other assumptions it is inconsistent. In other words, the basic statement that constitutes E, basic statement would be a logical statement, there's a swan here, would be part of the meaning of T, T being the theory. Now, if no basic statement is treated as inconsistent with a purported empirical statement, then we may infer that it is after all non-empirical, Theory T. As a corollary, it follows that the original claim was either wrongly presented as empirical or was empirical and was later abandoned for another theory with the same or parallel terminology. In other words, the original claim has been abandoned. Okay, I agree with him on that. Two, reinterpreting the theory as a definition or the implications of a definition. Reinterpreting a theory is in some cases abandoning the earlier theory.

[01:37:27]  Bruce: In some cases it is simply changing the conceptual structure without changing the theory. So he goes on to use an example of Marx's labor theory of value. I'm not going to go into that, but keep in mind this idea for Popper that a theory's truth content is its testable content, not the concepts by which it is understood. You can theoretically change entirely the conceptual framework of a theory and leave the theory untouched.

[01:37:55]  Unidentified speaker: Okay.

[01:37:58]  Unidentified speaker: Changes in the conceptual system employed by a theory may function as protection against criticism, since they may disarm the critic.

[01:38:06]  Bruce: It may appear to the critic that the theory has been abandoned under the pressure of his criticism, whereas in fact the old theory is retained under the intended or unintended camouflage of the new concepts. However, such an effect has costs for the ideology survival value. That may be overlooked, the ideology has to be relearned at transmission cost. So if you have a bunch of followers and you've now changed the concept, you may avoid the criticism while keeping the same theory because you're just using a different word strategy for it. But now you have this cost of having to re-teach the new word strategy to all your followers, making it hard for this theory to propagate. To the extent that the change of the concept is unintentional, there is a loss of understanding of the theory. After all, if the ideologue believes that the conceptual transformation theory really is just the same theory, he cannot have a good grasp of the theory if the implications have changed. So it'll be hard for him to propagate it. Three, adding assumptions. One might at first think that there are two main ways in which a theory T may be immunized through changes in the assumption of the theory. One, a move from T to T prime, where T prime is the conjunction of T, and one or more auxiliary assumptions denoted by B. Two, a move from T to X, where X is T minus some of its assumptions, perhaps with replacements. Only number two represents the abandonment of assumptions of T and its replacement by another theory.

[01:39:36]  Bruce: One might argue that number one preserves the original theory within the substitute, within the substitute and therefore immunization can preserve an ideology. Thus, Lakatos says, for instance, we may have a conjecture, have it refuted and then rescued by an auxiliary hypothesis, which is not ad hoc in the sense that we have earlier discussed. It may predict novel facts, some of which may even be corroborated. So this would be following the no ad hoc rule. It's okay to introduce an auxiliary assumption if it has its own independent and testable consequences. Ray then says this, but number one is not a logically possible immunization. The modified theory cannot be consistent with the falsifying evidence. If one simply adds extra assumptions, the increased information, that increased information content. For suppose Theory T is false with respect to Theory E, then since a conjunction is false, if and only if one of its conjuncts is false, any conjunction consisting of T and an extra assumption B will also be false with respect to E. That was really hard to follow unless you're really good at logic.

[01:40:46]  Unidentified speaker: Let me explain what he's saying.

[01:40:47]  Bruce: Let me explain why it's technically correct, and let me explain why it's ultimately wrong. It is well known that logic is monotonic. What that means is it can only go in one direction.

[01:40:59]  Unidentified speaker: So if I have, you know, if A implies B, given A, I now know B to be true.

[01:41:08]  Bruce: There is no auxiliary assumption you can add that will make B now false. It's monotonic. Once you have the conclusion logically, no set of additional assumptions can ever cause a change to your inferences. This is a known fact about logic, and it clearly doesn't match the way humans reason. We, when new starting assumptions come in, we change our inferences all the time. There's been various attempts to try to model this behavior, non-monotonic logic, for example, where they're trying to work out, okay, what is it humans are really doing, and why is it that we have some form of logic, or so it seems, where new assumption can cause you to change your inferences? For example, if I were to say, birds can fly and Tweety is a bird, so can Tweety fly? And then you go, yes, that would be a decent inference in most cases. And then I go, actually Tweety is a penguin. And suddenly you go, oh no, no, Tweety can't fly. And the reason why is because in non-monotonic logic, let's say, there are certain assumptions you can make based on what you're given, and then they can be overridden and you have to rethink your inferences.

[01:42:32]  Unidentified speaker: Because of this, I've seen this as a criticism of Popper, because Popper based his epistemology around deductive logic, they've tried to point out that there simply is no way to follow the no ad hoc rule.

[01:42:47]  Unidentified speaker: The no ad hoc rule, there is never an assumption that can change an inference.

[01:42:53]  Bruce: So you can't have a new assumption that [didn't change as your inference?]. The problem with all this is, is it reassumes that we're reasoning at the level of deductive logic and we're not. Deductive logic is the final step in the process when we're trying to work out a refutation. We almost always reason in terms of causation, because that's just the way humans reason. Causation is not super well understood, Judea Pearl has put a lot of effort into trying to formalize the concept, done some great work in that area.

[01:43:22]  Unidentified speaker: I would highly recommend his stuff.

[01:43:23]  Bruce: We'll have to do a podcast on his book of why at some point. Because of this, when you are talking about introducing an auxiliary theory and the no ad hoc rule, I think you have to start with the assumption that you're not talking about a logical auxiliary statement, you're talking about a causal one. And yes, once it's been introduced, the whole network of ideas that have been written down logically, they may have to be overwritten and changed. But it can always be done that you can translate between causal statement and a logical statement at the end. Because of this, Ray's right that there is no way to actually introduce an auxiliary theory that would change the inference. But that's only true if we're starting with the assumption that the auxiliary theory in question is itself a form of deductive logic, which it never is. So just wanted to make a point about that there. Without really disagreeing with Ray there, because he's still saying something that's technically true. Four, subtracting assumptions from T. Subtracting assumptions that reduce the information content of an ideology effectively means that the original theory is abandoned. Five, reinterpreting the theory as essentialistic. Six, introducing the idea that the theory is beyond the capacity of human reasoning to criticize or test. He points out many religions will use that approach. Seven, introducing ad hoc exclusion clauses for special cases. The introduction of an ad hoc exclusion clauses is also a mode of abandoning the original claim.

[01:44:58]  Bruce: For example, suppose someone advances the theory that bread nourishes, but then notices that a certain bread, batch of bread kills some people. If he then says that bread nourishes except that particular batch that killed those people, then he has reduced the content of his claim and therefore abandoned the original theory. By the way, I would say here that there's nothing wrong with the way doing that as long as you understand this in terms of the no ad hoc rule. That saying bread nourishes could be thought of as a broad abstract statement that wasn't intended to be taken literally. When a batch kills, you could say, well, obviously, I would clarify that assuming it's not poisoned. I feel like that's actually a fair move because we have to communicate and we can't add all caveats upfront. But you do have to understand that as a universal law, bread nourishes, you've refuted it by even a single batch that kills.

[01:45:58]  Unidentified speaker: Right.

[01:45:59]  Bruce: Okay, number eight, alternating between two or more versions of a theory. The propagandists may alternate between what are really two or more theories. This is an interesting case in which the original theory is not completely abandoned. It's quite possible that two interpretations of some canonical text are maintained, each being brought to the fore when powerful criticisms make it difficult to assert the other. Frank Cioffi has noted this phenomenon in connection with Freudianism. It's characteristic of pseudoscience that the hypotheses which comprise it stand in an asymmetrical relation to the expectations they generate, being permitted to guide them and be vindicated by the fulfillment, but not to be discredited by their disappointment. In other words, one way in which it achieves this is by contriving that these hypotheses understood in a narrow and determinate sense before the event, but a broader and hazier one after it are on these occasions on which they are not borne out. Such hypotheses thus live a double life, a subdued and restrained one in the vicinity of a counter observation and another less inhibited and more exuberant one when remote from them. This is one of my main criticisms of anarcho-capitalism. Not so much anarcho-capitalism, let's say Austrian School of Economics. Austrian School of Economics is a completely legitimate novel form of economics that made real contribution to economics. Value of it as a school of thought is not in any way in doubt.

[01:47:38]  Bruce: If someone tells me, I believe in the Austrian School of Economics, with like 80 percent certainty, this is almost assuredly someone who believes in anarcho-capitalism and a whole slew of ideological ideas that really have nothing to do with the good parts of the Austrian School of Economics. What point do we admit that the term Austrian School of Economics has included and maybe intentionally included this ideological edge that isn't actually part of the goodness of its theories? How do you actually separate these two? My point of view is is that you don't. That if somebody tells me, I'm an Austrian economics person, and 20% of the time, they're talking about the good parts of the theories, and 80% of the time, they're an ideologue, then that means that that term has become to mean the ideology and not the theory. And I didn't know how else you deal with that, over time, that may change. It may be that as the criticism of the Austrian economics, that they're too ideological. Particularly Rothbard was really hugely ideological. So it depends. Hayek wasn't. Hayek was part of the Austrian school and he was a Popperian, and he only took the good parts away and he completely abandoned the bad parts of Austrian School of Economics. Which one is the Austrian, right? One of the things I've noticed is that a person who's an Austrian economics person, they'll move between these two interpretations. Though if you say, yeah, I've got problems with Austrian School of Economics. They'll go, I don't know how you could have any problems. It's just good economics.

[01:49:26]  Bruce: Then you say, oh, you just mean the good economics? Oh, I guess I accept that. They go, oh, well, then you have to be an anarcho-capitalist because that follows from our theories and that's the best theory. No, that's total bullcrap. They'll move between this ideological version and the not so ideological version, the motte-and-bailey, if you will, depending on who they're talking to and what they're trying to accomplish. I think this is just how anarcho-capitalism and Austrian economics has come to be so strongly associated. There is good economics there. It's just somehow been captured by this ideological edge that I don't know what to do about. It's really the ideology that I got problems with, not the economics that I've got problems with. But the two are so strongly associated in the adherence minds that there's no easy way to move between them anymore. Okay, so Ray then continues, the bold version is still prized for its richness of information content. This is the good version of Austrian economics, and so brought forward in certain circumstances. Well, meanwhile, the less good version, that's the one that's brought out. Only, how am I trying to say this? You bring out the ideological version when you're trying to be moralizing and you bring out only the good economics, the moment you start to be criticized in a good way, in a way that's problematic, you move back to this narrow good economic theory that's much harder to criticize because it's probably largely true, right?

[01:50:57]  Bruce: I mean, Austrian economics has problems, so it's not entirely true, but it's got a lot of truth content in it, right? And so you simply move between those and you try to associate those two together as if they're one of the same and it's really two theories. It's really two entirely different theories, okay? This is what Ray's talking about here is an immunizing strategy. Okay, that's actually the end. This went almost two hours now. This is such a fantastic book.

[01:51:28]  Unidentified speaker: I was just going to say, Bruce, you know, I, I, when I, when I read Ray's book, I, I, I, as much as I, I loved it, I was a little bit kind of like wondering at times what the audience is.

[01:51:44]  Unidentified speaker: Yeah.

[01:51:45]  Unidentified speaker: You know, like it, it, the thesis is just so compelling, that it seems like it could be just a wonderful popular book, book made for a pop—a popular audience.

[01:52:00]  Peter: But it's so dense, certainly after the first chapter or whatever, that it almost feels like an academic book. But then, I can't imagine too many academics that I know of. Well, most academics don't really like Popper, for one thing, and I don't know how many academics would read this.

[01:52:27]  Unidentified speaker: I think now I've realized who the audience was is for this book, and it's you.

[01:52:34]  Unidentified speaker: You are the audience.

[01:52:36]  Bruce: Ray wrote this book for me.

[01:52:38]  Unidentified speaker: He anticipated me coming.

[01:52:40]  Bruce: He said, I've got to write a book that Bruce is going to love.

[01:52:42]  Unidentified speaker: I think so.

[01:52:44]  Unidentified speaker: Somehow.

[01:52:44]  Bruce: He succeeded in spades. I'm like, this is so good. One of the things I did was, so Ray didn't want the book put into Google Notebook where theoretically Google then steals your text and your copyright. So he asked me [to put—input it into any AI?], but he didn't have a problem with me putting it into a local AI and I've written AI Karl Popper, so I have a local version I can run. Well, I, okay, I took the whole thing and I threw it into my audio book generator, which runs with just a local model, so it doesn't go through ElevenLabs or anything like that. I started reading the book and then I thought, oh wow, I've got this upcoming interview with Ray. We thought it was going to happen sooner than it really did. I thought it was imminent and really we were still a ways off. So I thought, I've got to get this done. So I took the second half of the book and I threw it into my converter and I listened to it as an audio book. And I'm like exercising and I get to chapter four, I'm like shouting, yeah, it's so good, you know. I'm exercising.

[01:53:55]  Unidentified speaker: Well, good.

[01:53:56]  Peter: I hope Ray appreciates how much you like it and I enjoyed listening to your thoughts here and I'm sure this is something we'll be revisiting in the future.

[01:54:08]  Unidentified speaker: All right.

[01:54:09]  Bruce: Thank you very much.

[01:54:10]  Unidentified speaker: Okay.

[01:54:11]  Unidentified speaker: Thank you, Bruce.

[01:54:13]  Unidentified speaker: Bye-bye.

[01:54:21]  Unidentified speaker: Hello again.

[01:54:22]  Peter: 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 BNielsen01. Also, please consider joining the Facebook group, The Mini Worlds of David Deutsch, where Bruce and I first started connecting.

[01:55:05]  Unidentified speaker: Thank you.