00:00:00.000 Welcome to Topcast, and ostensibly this is part two of my discussion of chapter seven of the
00:00:08.000 fabric of reality, a conversation about justification. But I'm not actually going to get to any
00:00:13.440 readings from that chapter today. This is just an indulgence for me. It's going to be
00:00:19.760 readings from Karl Popper's book, Realism in the Aim of Science, which was a postscript to the
00:00:27.520 Logic of Scientific Discovery. And the reason I'm going through that before I get to, which should
00:00:32.720 be released soon after this, the actual discussion of chapter seven, part two of chapter seven, a
00:00:40.160 conversation about justification from the fabric of reality. The reason I'm doing this is just
00:00:45.200 because there is a lot of content in that chapter there that echoes what Popper says in this book,
00:00:53.280 Realism in the Aim of Science, because there is a section there, section four of part one,
00:00:58.640 called corroboration. So I thought it might just be illuminating and fun to go through some of the
00:01:04.960 places where I agree with what Popper is saying, and then where I disagree with what Popper is saying,
00:01:10.960 and then we can see how David Deutsch has improved on Popper and has explained what Popper was trying
00:01:18.160 to get at in books like this one. This is a 464 page term, all about realism in the aim of science,
00:01:26.880 because David Deutsch has removed entirely from epistemology, I would suggest. This whole concern
00:01:33.680 about how probable or likely to be true, how theories and explanations are. Now, Popper was
00:01:40.960 existing at a time where people really strongly believed in this problem of induction, how we could
00:01:46.240 justify us true or probably true particular theories and reducing science to a large extent to a
00:01:53.360 process of trying to predict the future rather than explain reality. The emphasis of Popper's
00:02:00.880 contemporaries and competitors even through to today is to not really understand what explanation is
00:02:07.920 in the role of explanation, let alone the role of prediction, let alone the role of science.
00:02:12.320 They get it all muddled up, and so Popper was trying to escape from the prevailing worldview that
00:02:19.520 he existed within to give us something new. It must have been hard, and I think that this chapter,
00:02:26.480 which is why I want to read just through sections of it, sections of his whole section here,
00:02:32.240 called corroboration, you can see him struggling to try and escape from false epistemology.
00:02:38.880 These unhelpful ways of thinking about the project of science, and so what I'm going to do here is
00:02:45.120 just to read through sections of this and then we'll come back in the next episode to see how
00:02:50.720 David Deutsch has actually improved on things. This chapter being titled a conversation
00:02:56.160 about justification seems to be to do with words. A conversation is a dialogue that involves
00:03:03.360 words and the word justification is a term that is used in philosophy. It's a term that I've tried
00:03:10.560 to remove in so far as possible from any discussion about epistemology, and we're going to see that
00:03:16.960 I think we can do away with it. I explain this in part one as well, and David Deutsch himself said
00:03:23.920 as much in the recording he made for the introduction to the audiobook version of the fabric
00:03:31.520 of reality. Let's just consider a basic question before I get into discussing the chapter
00:03:36.640 property. And the question is, what are we doing here? I don't just mean right now here in this
00:03:41.840 podcast I mean in life, in creating knowledge and in science. That is the subject of this episode,
00:03:49.440 and this chapter of the book after all, what's really going on with science? What are we doing
00:03:55.840 when we are doing science? Well the thing is, we are trying to understand something about the world,
00:04:03.200 but why are we doing that? It always comes back to one central issue, first stated and emphasized
00:04:09.280 by Popper and Underlined and promoted by Deutsch. And that issue is, we have a problem. Houston,
00:04:16.720 we have a problem. It is, we have a problem. It's always, we have a problem. We always have a problem,
00:04:22.800 and we want to find a solution to that problem. We want to solve a problem so we can move on to
00:04:30.560 something better. This in fact is life, and a small sliver of life is science. For most of us anyway.
00:04:39.680 For some of us it's considerably most of life, but in science we have problems, lots of problems.
00:04:46.400 As Popper rightly said of science itself, quote, I think that there is only one way to science,
00:04:54.640 or to philosophy for that matter, to meet a problem, to see its beauty, and fall in love with it,
00:05:01.440 to get married to it, and to live with it happily till death do ye part, unless you should meet
00:05:07.840 another and even more fascinating problem, or unless indeed you should obtain a solution.
00:05:12.880 But even if you do obtain a solution, you may then discover to you delight the existence of a whole
00:05:20.560 family of enchanting, though perhaps difficult, problem children, for whose welfare you may work
00:05:26.080 with a purpose to the end of your days." End quote. Now that quote there is from a book called
00:05:33.760 Realism and the Aim of Science published in 1996 after Popper's death, but it also served as a
00:05:41.200 postscript to his first book, The Logic of Scientific Discovery, a postscript that was published,
00:05:47.280 sometime well after the publication of the first edition of The Logic of Scientific Discovery.
00:05:52.560 Notice something missing there when it comes to Popper saying what science is about or what
00:05:59.440 philosophy is about. Conspicuous perhaps by its absence. Any mention of the word truth. Science,
00:06:08.240 we do not do to find the final truth, and that's why we often don't emphasize it.
00:06:15.680 We are not even finding truth of a kind we can be confident in or certain about or
00:06:22.240 anything like that, but why am I focused on this? The reason is, and the only reason for a
00:06:28.160 chapter like this particular one in the fabric of reality that I am discussing,
00:06:32.560 is that this view of science insofar as it is known, I think it's poorly understood,
00:06:38.240 and insofar as it is understood, it seems not to be taken very seriously. As for it being taught
00:06:44.160 in some sort of formal way so that people can speak sensibly about the project of science and
00:06:50.480 life more broadly. Forget it. We are mired in superstitious thinking to a large extent.
00:06:55.760 All of our knowledge has antecedents. You know the things that came before our first attempts
00:07:01.200 first approximations. Here's a story, millennia ago after the dawn of language people struggling
00:07:07.440 on the planes or in the jungles and struggling to eke out an existence and keep the children in
00:07:12.400 the tribe safe needed heuristics rules of thumb. So they invented gods and the gods knew the truth.
00:07:20.560 The gods had power and they had possession of the final truth. They knew what would happen
00:07:27.760 and why? Probably their whims. But the humans did not yet we aspire to the power of the gods
00:07:34.480 because the gods had authority over the seasons, over the weather, over natural events, over everything.
00:07:42.560 We wanted that. So the chiefs of our ancestor tribes and our medicine men and women,
00:07:48.800 those with tribal authority needed a way to persuade others to do what they were told.
00:07:55.120 Sometimes for good reasons, they knew stuff. They often didn't know why that stuff happened,
00:08:00.720 but they knew that it did happen. Out of this circumstance comes something like the more modern
00:08:07.600 religions that we have. The idea that it is known without doubt and there is no point questioning
00:08:13.680 some revealed truth or other. Only the medicine man or the chief had direct access to the gods
00:08:19.440 and later on. Only the priests or other learned people could read and interpret the holy
00:08:26.960 inherent book perhaps. The focus then there was on the final ultimate revealed truth and that
00:08:34.480 if people only followed the final ultimate revealed truth then everything would be fine.
00:08:39.920 Then the disaster or the flood could be averted. After all, the stories tell when you depart
00:08:45.760 from the truth, it's then the disaster comes, the literal flood. Okay, so this is one very
00:08:52.080 superficial, simplistic story of the past and of how we arrive at our folk epistemology today.
00:08:58.640 But what people are up to then? Well, there is much that we have inherited in our language
00:09:06.160 and our way of understanding knowledge from them through to this day, even in the most modern
00:09:11.920 incantations of epistemology. Popular in epistemology, of course. Most especially our
00:09:18.640 folk philosophy and folk epistemology, which I would include most academic versions of these,
00:09:24.320 it is still taught that what we're aiming for is to find out the truth of the matter.
00:09:30.240 Science on this account therefore is a project of finding the once and for all solution theory
00:09:36.320 or explanation that we can carve into stone and settle the matter, have a final completed
00:09:42.480 science of something or other. It's a very religious idea. It is the hopeful state of security
00:09:49.040 and safety and certainty. And along the road to certainty, although we might not be able to quite
00:09:53.920 get there yet, we're almost there and we can be very, very confident in what we have found,
00:09:59.120 confident that it is almost true. And with just a little more tinkering, we'll fill in the gaps
00:10:04.560 and we'll have the final answer. Science will come to an end. I mean, here on topcast by now,
00:10:10.800 if you've been listening for a while, that all sound ridiculous, but it still is the prevailing view.
00:10:15.840 There still is an idea that science has almost wrapped up with almost found the unification of
00:10:21.360 general relativity and quantum theory and then we'll be done in physics. And we'll biology.
00:10:26.000 Ah, we're tinkering around the edges with evolution by natural selection. We've already found
00:10:29.760 the unit of selection to Jane. What more is there to understand? Soon we'll cure all diseases,
00:10:34.480 aging will be over. We'll have artificial general intelligence and then understanding will come to an
00:10:40.320 end. And perhaps we'll all be unemployed because there'll be no creative work left to do now.
00:10:45.680 This is completely the opposite to what the vision of knowledge that Popper and David Deutsch
00:10:51.040 especially has gifted to us. But it's important to keep in mind that this is not the prevailing
00:10:57.360 view. This is counter-cultural. And sometimes I myself forget it. And then I go head long into
00:11:03.520 encountering someone or some book or some group of people who reminds me that I'm in the tiny
00:11:10.640 minority of people who think something a little bit different. There is an open-ended series of
00:11:17.440 problems before us and each time we solve a problem, we open up a whole new family of delightful
00:11:25.280 problem children if you like. I prefer daughter problems than problem children. Whatever the case,
00:11:32.800 that quote I read earlier from realism in the aim of science, that book realism in the aim of science
00:11:39.120 contains a whole bunch of really interesting chapters. But you can tell in reading it by
00:11:44.000 Carl Popper, of course, that he is writing for his contemporaries. It's filled with language about
00:11:51.280 certainty and probability. And so I want to read a little bit of it today just to place perhaps
00:11:57.680 in historical context where the fabric of reality is coming from what its intellectual
00:12:02.960 ancestor is, so to speak. Realism in the aim of science is a 460-something page tone split into
00:12:12.080 two major parts. Chapter four of the first part is called corroboration. Now corroboration is
00:12:19.520 also a term I have a little bit of difficulty with. Again, I don't think it's useful,
00:12:25.760 but it is language that is used by other philosophers of science. Popper was not perfect,
00:12:32.000 and as we'll come to, Popper may not have understood popularing epistemology as well as people
00:12:37.280 today do. Now if that sounds bizarre to you, we'll come back to it because David addresses
00:12:42.400 precisely that point in the fabric of reality. Popper, for example, focuses very heavily on the
00:12:48.640 role of probability within philosophy, within epistemology. He was trying to understand the
00:12:55.360 importance or the significance of probability. He was wrong about it. That's so what everyone's
00:13:00.480 fallible. I am no doubt wrong about a lot that I say here on topcast. And looking back 200 years
00:13:07.040 from now, people will be saying, look how primitive Brit Hall's view of epistemology was. But as I
00:13:12.480 mentioned in the last episode, I did on chapter seven on a conversation about justification from
00:13:19.360 the fabric of reality. I said that this word justified or justification may have been superfluous to
00:13:26.800 our needs. Another word that crops up throughout the chapter is this idea of corroboration. What
00:13:33.760 is corroboration? Well, in chapter four of the aim of science, which is from pages 217 all the way
00:13:40.480 through to pages 254, Popper hashes out from all angles, they rather obscure debate at the time
00:13:47.920 about the differences between confirmation and corroboration and the role of probability and
00:13:54.720 either of these things. My personal thought is just to steal my thunder for the end is it's simply
00:14:02.160 confusing the aim of science as being about primarily prediction rather than explanation.
00:14:08.880 If you focus on explanations, then as I emphasize over and again, you are exceedingly lucky in
00:14:16.400 this world if you have an explanation for a particular phenomena. And if you do have an explanation,
00:14:22.560 you have an explanation. There are no alternatives. And if someone comes along with an alternative,
00:14:28.080 then you do a crucial test, especially in science. A crucial test is an experiment which
00:14:34.240 has an outcome that goes in a particular direction. It has a particular result. And the result will
00:14:39.200 rule out one of those theories. You're extremely lucky to have two theories. It just rarely happens
00:14:46.320 in the history of science. And this is why I come back to trope examples like the competition
00:14:51.680 between Newtonian gravity and Einstein's general relativity. This is one of the rare instances where
00:14:57.920 there happened to be two competing theories where we needed to do an experiment to figure out
00:15:03.680 which one could actually explain the result to the experiment and which one couldn't.
00:15:07.600 But for any scientific mystery, it's a mystery by definition because we do not have an explanation
00:15:14.240 of why it is the way it is. Now it's very hard for me to keep up with all the areas of science.
00:15:19.840 So my focus is on astronomy. And there are a whole bunch of open-ended questions right now.
00:15:25.600 Two of the most prominent open-ended questions are, what is dark matter? Dark matter is the name
00:15:32.240 of a problem. It began as, why is it that spiral galaxies are rotating so fast? They rotate as
00:15:39.680 fast as they do because of how massive they are. This phenomena that orbiting bodies tend to rotate
00:15:46.800 as fast as the mass inside of the orbiting body. In other words, the earth goes around the sun
00:15:52.560 as fast as it does because of the mass of the sun applies to everything throughout the universe.
00:15:57.120 It's just a consequence, well, not only of general relativity, which is the explanation of gravity,
00:16:01.920 but even to the predecessor to general relativity, which is Newtonian gravity,
00:16:06.960 both of them predicted the same thing. The more massive the central body being orbited,
00:16:11.840 the faster the things go that are orbiting it. But here's the thing in modern day cosmology
00:16:16.960 when you look at galaxies and various other structures, by the way. But let's just concentrate on
00:16:20.880 the spiral galaxies. They rotate too fast. You add up all of the stars in luminous matter,
00:16:26.000 and you can see all the luminous matter at every single wavelength that's putting out light,
00:16:30.080 light of all different wavelengths. You find that these spiral galaxies are rotating too fast.
00:16:35.520 Where's the mass? We can't see the matter that's accounting for the gravity that's causing
00:16:40.000 these things to spin so fast. We don't know what the answer is. It could be that there's actually
00:16:45.200 missing matter there. That seems to be the prevailing hypothesis, but it's not really an explanatory
00:16:50.720 theory because what is this matter? Another theory is, well, we need to have a new theory of gravity,
00:16:56.320 but no one's got a good theory of gravity so far. People have suggested things, but these
00:17:02.560 theories add hawk modifications to existing explanations. Then I won't go down the road,
00:17:09.200 but epistemology would say, what is the mechanism? What precisely is it? Is it the curvature
00:17:14.400 of spacetime and why is that changing? Why can't we detect it in laboratories here on earth and
00:17:19.840 so on and so forth? We need a solution to this and we don't have one. And the other thing of course
00:17:24.560 is dark energy. No one has any clue what this is. Why is it that not only is the universe expanding,
00:17:30.880 but the rate of the expansion is accelerating? When everything else we know says it should be
00:17:35.920 slowing down, perhaps even reversing, but that's not happening. What's driving this accelerating
00:17:42.080 expansion? This is science. This is problem solving. This is where we don't even have an explanation.
00:17:50.400 Forget about having competing explanations. Forget about needing to wait in different explanations
00:17:55.200 of the probabilities of our different explanations and being able to corroborate run or confirm one
00:18:00.160 without being able to confirm the others. No, none of that. We don't even have one. When we do
00:18:04.640 have a solution to the problem, which is to say a theory explaining what's going on,
00:18:11.360 precisely in terms of things that exist in the universe that perhaps we've never thought of before,
00:18:17.600 some new physical thing that we have to postulate that then will know actually exists.
00:18:23.920 Once we have that thing, that will be the explanation. Never mind trying to observe it repeatedly
00:18:29.280 over and over again. It'll be the solution. That's what solves dark matter and dark energy.
00:18:36.080 So anyway, that's what I say science is about. I think that's consistent with what David Deutsch
00:18:41.200 says, and I think it's consistent primarily with what cow poppers says as well, but as I say,
00:18:47.760 people exist in a historical context. And papa was debating people who believe something quite
00:18:53.840 different about how science works. They believed in the primacy of observations that observations
00:18:59.920 were absolutely the be-all and end-all of everything in science. And therefore,
00:19:04.880 predicting that a certain sequence of observations would continue off into the future,
00:19:09.200 or the observations were the thing that showed your theory was true or probably true in some way.
00:19:15.680 This is what he was trying to debate against, trying to, for the first time,
00:19:20.320 mind you, for the first time, try and stand up against the entire philosophical community of his
00:19:26.640 intellectual peers. One has to be brave to do this kind of thing. And it certainly didn't make him
00:19:31.760 popular. Never mind his Jewish heritage. He struggled to find positions at universities for a while there.
00:19:39.280 And to our great historical shame here in Australia, even we rejected him for an academic post.
00:19:45.360 He went to New Zealand instead and did a lot of good work there. So I say that because
00:19:49.840 he was telling the rest of the philosophical community and even scientists. He was saying to them all,
00:19:56.160 in very, very polite terms, you're all fundamentally wrong about how science works. But the way he
00:20:02.720 couched it was in very technical philosophical jargon. Jargon, that was the language of the time
00:20:08.640 and the language of the discipline. He had to try and speak their language. So I want to go back
00:20:13.920 to realism in the aim of science. This so-called postscript to the logic of scientific discovery,
00:20:20.240 published a long time after the logic of scientific discovery, and then republished and the
00:20:24.400 version I have was published in, as I say, 1996. But written by Popper, according to my edition,
00:20:30.960 first published in 1983. Now to some extent, of course, these are esoteric considerations. It's
00:20:37.840 inside baseball to a certain extent. If you're interested in epistemology, then yeah, absolutely,
00:20:43.440 it's useful to know Popper's broad vision about how knowledge is generated and what the purpose
00:20:49.280 of science is. But if you're really interested in epistemology, then sometimes it can be even more
00:20:54.560 illuminating to figure out exactly what, going all the way back to this, to Popper's words about
00:21:01.280 corroboration may seem even a little esoteric for me. But there's so much here that speaks across
00:21:09.840 the decades through to today, and you can see it's still cutting-edge stuff that people just
00:21:15.360 don't get. People don't understand. People don't appreciate, and the wider academic community
00:21:20.880 could do worse than taking on some of Popper's old works and trying to understand what the
00:21:27.200 great man was saying about how science actually works. So I'm going to pick out just a few
00:21:33.120 excerpts from realism in the aim of science before we get to a conversation with justification.
00:21:40.000 And I'm going to begin with page 222 where Popper wrote, quote, the inductivist philosophy
00:21:47.200 not only attributes authority to science, it also, perhaps quite unwittingly,
00:21:52.560 attributes to science are cautious and indeed timid approach which is entirely foreign to our
00:21:58.800 real procedure. This philosophy, in regarding it as the aim of science to attain
00:22:05.120 high probabilities for its theories implies that science proceeds according to the rule,
00:22:10.640 go as little as possible beyond your evidence e." End quote, and skipping a little,
00:22:17.360 and I'll pick it up where Popper says, quote, all this presents a most uninspiring picture of science,
00:22:24.720 a picture more over that does not in the least resemble the original. Indeed,
00:22:30.000 what makes the original so inspiring is its boldness, its boldly conceived hypotheses,
00:22:36.640 boldly submitted to every kind of criticism, to every reputation we can think of,
00:22:41.520 including the most severe test which our imagination may help us to design, it is this boldness,
00:22:47.360 which helps us to transcend the limits narrow at first of our imagination and of ordinary language
00:22:54.240 end quote. So that first part there where this idea from induction kind of says, don't go too far
00:23:01.440 beyond the evidence, it misses the point, it misses the point, consider the great grand theories
00:23:08.640 of cosmology, the big bang theory, this idea that the universe in the deep dark past 13.7 billion
00:23:15.680 years ago was smaller than an atom, how do we come to this view that the universe we now occupy,
00:23:24.000 the universe of stupendous complexity around us and stupendous size and 13.7 billion years in age,
00:23:31.520 how did we come to that? What are the crucial pieces of evidence? Well, it's just light,
00:23:37.840 just different kinds of photons being interpreted. Among the first bits of evidence, really,
00:23:44.720 aside from the dark night sky, but let's not worry about that, were Hubble's explanations
00:23:50.720 of spectra breaking up the light from distant galaxies and that distant light when broken up into
00:23:58.560 its spectra showed spectral lines. So we're looking at spectral lines on photographic plates
00:24:05.120 and those spectral lines red shifted just a little and that evidence explained by a big bang
00:24:12.720 event, the creation of the entire universe, that is the explanation of those red shifted galaxies
00:24:20.160 and this is why Popper is saying, we're going well beyond the evidence, we are transcending
00:24:26.560 the limits of our imagination and of ordinary language, going from spectral lines to the creation
00:24:35.120 of the entire universe, that's pretty stupendous stuff. What was Darwin doing? Look here,
00:24:40.640 tortoises and finches and what was his explanation about those locally interesting things
00:24:47.200 that all of life on the planet had evolved over hundreds of millions, even billions of years
00:24:54.320 through this process of natural selection, tiny amounts of evidence and us going well beyond the
00:25:01.280 evidence to conjecture grand explanations about how reality works. That science, this process of
00:25:10.000 prediction, now in this section that I've just read from, I do have to say, I do disagree with
00:25:15.680 what Popper says in part of this section, but this is again, inside baseball. It's important for
00:25:21.840 me to understand that I disagree here, but it doesn't make a substantial difference too,
00:25:27.680 preparing epistemology as a whole. There are often disagreements among
00:25:32.240 populations. The great Danny Frederick was a great Popper scholar and he and I would engage in
00:25:39.280 debates about what the purpose of science or knowledge creation is in general. Danny's perspective
00:25:45.360 was that it wasn't that we were after truth. We were not actually looking for the truth.
00:25:52.480 As he would argue, truth cannot be a epistemic aim. Now, I often took exception to this idea,
00:26:00.560 this way of explaining what we're doing when we create knowledge. Danny was a realist
00:26:05.280 and a Popperian, but I always get worried when people deny the fact that we are after truth.
00:26:12.960 Danny endorsed the idea that truth existed. He just didn't think that this is what we were trying
00:26:17.760 to find. I guess it depends upon the person. Some people might very well be trying to find the
00:26:23.200 truth or find at least some truth, sometimes, or perhaps by removing falsity, removing misconception
00:26:31.440 in that way they're finding truth. Anyway, I do not want to try and represent Danny's position here.
00:26:36.720 You can look up Danny Frederick and look up the papers that he wrote. But I like to say that what
00:26:41.840 we're doing is, of course, solving problems. And the solution to our problems, our explanations,
00:26:48.960 our theories must contain truth. And the reason why it must contain truth is because an actual
00:26:56.000 solution to be a solution is useful. And it's only useful because it is able to solve the problem,
00:27:03.120 which means it's got something right. And right just means true. So there's something true
00:27:09.680 about that solution. If indeed, it's a genuine solution to our problem. So this kind of really
00:27:15.920 splitting hairs about what the purpose of science is at that level is it about trying to find the
00:27:22.720 truth, trying to find some truth, trying to solve a problem, get something right, create an explanation.
00:27:28.560 All of these are kind of circling the same kind of idea. But what we disagree with the
00:27:34.880 non-popurians about is that all we're doing in science is trying to make predictions. That's the
00:27:40.320 instrumentalist claim, all that we're merely telling stories to each other that are nothing but
00:27:46.880 fictional narratives that don't actually connect with objective reality at all. Some people have
00:27:52.400 caught to deny the existence of objective reality or deny the possibility that we can explain
00:27:58.160 objective reality to some extent. To any extent, as realists in the perperian mould, we think that
00:28:04.880 we are actually explaining aspects of objective reality. And we are getting more or less close
00:28:12.880 to that objective reality as time goes on by refining our explanations, by finding perhaps where
00:28:19.760 they fail. In other words, where there's a problem with them or correcting errors in them and
00:28:25.360 improving them, making progress, sometimes by radically overturning a particular theory and sometimes
00:28:32.960 by incrementally changing some aspects of our understanding of reality. So let me go back to
00:28:38.640 pop-off for a moment and to the beginning of this chapter, because I've been speaking a lot
00:28:44.720 about probability recently. And so perhaps just to satisfy listeners who have often asked me,
00:28:51.680 where do you disagree with pop-off? Well, here's a particular place. So I'm going to read
00:28:57.040 from, well, it's called section 27 and just titled, corroboration, certainty, uncertainty,
00:29:03.760 and probability, and pop-off rights, quote, I have in the preceding two chapters explored the
00:29:08.640 logical ramifications of the problem of induction. There is another ramifications, however,
00:29:14.720 which I have not yet touched on because it is not logically connected to the problem of induction.
00:29:19.760 But it is connected with it by ties that may prove even stronger than logic by the inductive
00:29:25.360 prejudice. And by a mistaken solution of the problem of induction, which unfortunately is still
00:29:32.000 widely accepted as valid. I am alluding, of course, to the view that although induction is
00:29:38.080 unable to establish an induced hypothesis with certainty, it is able to do the next best thing.
00:29:44.320 It can attribute to the induced hypothesis some degree of probability and a probability of
00:29:49.680 one would be certainty, end quote. So here, Popper is saying that that's wrong, that an induced
00:29:56.640 hypothesis cannot even establish as true anything with a degree of probability, which is what
00:30:04.720 many, many people tried to say the solution to induction was. And with this problem of induction
00:30:10.080 was, how can we rely on the theories of science if we can't prove them as true? We need a time
00:30:16.960 lesser degree of confidence in them. This was the problem of induction. How do our past observations
00:30:24.720 confer some degree of certainty on the future alliance on a particular theory? Popper keeps going
00:30:32.960 and he writes, quote, this view is radically mistaken. Yet it can be supported by a highly persuasive
00:30:38.960 argument. This argument may be presented as follows. The whole problem of induction, the argument
00:30:44.880 runs clearly arises from the fact that inductive inferences are not valid, which is the same as saying
00:30:51.440 that inductive conclusions do not follow deductively from the inductive premises. But there is no
00:30:57.040 need to get alarmed about this somewhat trite fact, especially as there exists a large and important
00:31:02.320 class of inferences, which the conclusion does not strictly follow from the premises. In fact,
00:31:07.920 every deductive inference may be modified so as to yield, an inference which is not valid,
00:31:13.520 but only more or less valid or valid to a degree. Take the following example,
00:31:19.600 here's a valid deductive argument. All men smoke, Jack is a man, therefore Jack smokes.
00:31:25.280 Here is an argument that is valid to a degree. Ex percent of men spoke, Jack is a man,
00:31:32.800 therefore Jack smokes, end quote. Now, I would just interject here and say,
00:31:37.360 well, that's clearly ridiculous. It simply doesn't follow. Jack is a man who either smokes
00:31:44.800 or he doesn't. And we won't know until we have an explanation by means of an observation
00:31:51.600 that Jack actually smokes. There is no Jack probably smokes,
00:31:56.640 and Papa agrees with this, by the way, of course. Papa goes on to explain a whole bunch of
00:32:02.320 things about why it is that some people would endorse that particular second kind of
00:32:08.000 valid to a degree argument. And after some exposition on this point, he says,
00:32:12.480 quote, last the problem of induction is to be solved by constructing a generalized logic,
00:32:18.560 a logic of probability, end quote. He's still manning their arguments, right? And I would say,
00:32:26.160 still people today endorse this kind of thing. He continues, quote,
00:32:29.600 for according to this persuasive argument, inductive logic is nothing but probability logic.
00:32:35.680 It is the logic of uncertain inference of uncertain knowledge and the probability of
00:32:42.400 some hypothesis, h, given some evidence, is the degree to which our certain knowledge of the
00:32:49.760 evidence rationally justifies our belief in the hypothesis. As I have said before, I believe
00:32:58.000 that this argument is completely mistaken. The appeal to probability does not affect the problem
00:33:04.320 of induction at all. Formally, this may be supported by the remark that every universal hypothesis
00:33:11.760 goes so far beyond any empirical evidence that its probability will always remain zero,
00:33:19.920 because the universal hypothesis makes assertions about an infinite number of cases,
00:33:26.240 while the number of observed cases can only be finite end quote. That's marvelous. That's
00:33:33.440 their popper at his best. What we're doing in science when we come up with theories, we come up
00:33:39.200 with, especially in physics, but you know, chemistry, biology, we're coming up with universal
00:33:44.080 theories, theories that apply to everything at all times at all places. In other words,
00:33:48.720 an assertion about an infinite number of cases. But how can we possibly get to that assertion
00:33:54.400 about an infinite number of cases? If we can only ever have a finite number of observed cases.
00:34:00.560 Well, that's because we begin with the theory. We don't begin with the finite number of observed
00:34:05.680 cases. It's cut before the horse kind of stuff. This is the great mistake that popper was
00:34:11.520 addressing and solving. Everyone else was saying, well, look, you begin with the observations.
00:34:16.960 Even today, you begin with the observations. This is the whole point of science. You go out into
00:34:20.960 the world and you observe stuff. And from those observations, you then derive your knowledge.
00:34:27.920 No, that's all us around. That's all cut before the horse. Put the horse before the car.
00:34:33.760 And the horse is the theory. You come up with the theory first. You explain what's going on.
00:34:39.920 You conjecture. You use your imagination. Your creativity. That's what you are. You're a human being.
00:34:44.400 That's what you're supposed to be doing in the world. It doesn't just mean in science everywhere.
00:34:48.880 Now, having come up with a solution, then you rely upon that solution and you keep on
00:34:54.400 using that solution until one day you come across a problem, which usually means you encounter
00:35:00.240 one of your finite observations, one of your finite observations that disagrees with your
00:35:05.520 prevailing view, your existing theory, what you have thought so far, for so long,
00:35:10.480 and that contradicts what you've thought all this time. And so then you're going to have to
00:35:14.800 adapt on the run. You're going to have to create, come up with a new solution.
00:35:18.080 And in science, you're coming up with this explanation that explains everything you've seen so far.
00:35:24.720 It's not being derived from what you've seen. It explains what you've seen and then
00:35:31.040 reaches out from what you have seen and where you are sitting at your desk or in your laboratory
00:35:37.200 to everywhere else in the universe to things you will never observe, but it will apply to them.
00:35:44.880 In the same way that lets assume the old wives tale, the urban legend, the myth that Newton saw
00:35:51.600 the apple fall from the tree. Let's say he did do that. Very questionable. But let's say he did.
00:35:57.440 This is one of the observations. It's not like he took that observation and from that observation
00:36:01.680 of the apple falling, derived his universal theory of gravity. No, he came up with the universal
00:36:07.360 theory of gravity, which explained the motion of planets across the sky and apples falling
00:36:12.160 and tides going in and out and it reached out from where he was in England across the earth
00:36:19.680 throughout the solar system to the other side of the galaxy and the universe that same
00:36:25.600 law applied to fruits falling from trees on the surface of planets around stars he would never
00:36:32.480 observe and perhaps no one ever will. This is the thing about science. This is what science does.
00:36:37.760 It begins there with the theory and then moving through life with your theory, you encounter a
00:36:43.600 problem and that problem is often, not always, but often an observation of some kind, especially
00:36:49.680 when it comes to science. It's an observation of some kind that disagrees with your theory.
00:36:53.360 For reasons you don't know, maybe you've just made a mistake in your observation. You think
00:36:57.840 you've made an observation that in fact you have and it's something's gone wrong. It's an
00:37:01.280 optical illusion who knows what your instrument wasn't working. But whatever the case,
00:37:05.040 you've got a problem that you've got to solve and sometimes that might be very well,
00:37:08.560 the beginning of a whole new grand theory of science. Okay, so that's everything that
00:37:14.400 Papa got right now. Here is where I emphasize things a little differently to Papa.
00:37:20.880 And I'd love to be able to speak to him about what he had in mind. I wish there were examples here,
00:37:27.200 but let me pick it up where he writes, quote, let us turn to that idea which I believe is
00:37:34.320 defensible. It is the idea that hypothesis may be distinguished according to the results of their
00:37:40.160 tests. The idea that some hypothesis are well tested by experience and others are not so well tested,
00:37:48.320 that there are further hypotheses which so far have not been tested at all and hypotheses which
00:37:54.480 have not stood up to tests and which therefore may be regarded as falsified. If we look upon a number
00:38:01.520 of hypotheses from this point of view, there can be no objection to grading them according to
00:38:06.240 the degree to which they have passed their tests. Exactly as we may grade students who have undergone
00:38:12.240 a number of tests, some of them easy, some of them difficult. The wish to grade hypotheses
00:38:17.680 according to the tests passed by them is legitimate. I do not know of any serious objection.
00:38:24.240 For reasons to be discussed in the next section, I propose to call the grade of a hypothesis
00:38:29.680 or the degree to which it has stood up to tests, it is degree of corroboration rather than its
00:38:36.720 probability end, quote. We cannot speak to him now, what a shame, but I just would love to know
00:38:44.400 what he had in mind. Where are these situations where we have got this great spectrum of
00:38:50.640 hypotheses that we need to rank order, that we need to treat like students that we are
00:38:56.560 grading. Where we say here is hypothesis one and it has a certain amount of corroboration,
00:39:01.120 here is hypothesis two that has got slightly more corroboration, and here is hypothesis three
00:39:05.280 that has got yet more corroboration. This does not happen. This does not happen. On
00:39:10.320 pop-a-zone account, this does not happen. What happens is if you have that situation ever arise,
00:39:15.200 and as I keep on saying, please write in, email me, tweet me, whatever. A situation where you
00:39:21.360 really do have these actual different, viable, good explanations of the same phenomena and you can't
00:39:31.520 distinguish between them and so you need corroboration. What really happens is, as we say, and as David
00:39:37.200 Deutsch has pointed out on more than one occasion, including most importantly in his paper on the
00:39:42.960 logic of experimental tests, what you do is you come up with a crucial test, an experiment which
00:39:48.480 will rule out all of the different alternatives, usually on me one, if you're lucky, all of the
00:39:54.800 different alternatives, leaving only one standing the one that can explain the results of the experiment.
00:40:01.920 So why pop-a-ones this degree of corroboration? I don't know. In what specific situation,
00:40:09.840 now throughout this chapter we don't actually get an example, which is a problem. He's right
00:40:14.800 to say that if you do have competing hypotheses, there's no point in talking about the probability
00:40:19.600 of one over the other. He's got that right. But degree of corroboration, and exactly how would we
00:40:25.760 measure this degree of corroboration? I don't know. The best he can do, in fact, in this section,
00:40:32.400 he says, quote, I shall later give a definition of degree of corroboration, one that permits us to
00:40:38.240 compare rival theories, such as Newtons and Einstein's, I doubt whether a numerical evaluation
00:40:45.040 of this degree will ever be of practical significance, end quote, quite right, quite right.
00:40:52.800 And if there is no such way of providing a numerical evaluation of this degree, what point is it?
00:41:00.720 What point is there in postulating this? His own example, my trope example that I like to use
00:41:05.760 following him, of course, is Newton versus Einstein. Now, there to be generous to him, I think that
00:41:12.800 maybe what he might be saying, one way of reading him is just to say that this degree of
00:41:18.400 corroboration is just perfectly synonymous with, refuted or not, refuted or not. So either Newton's
00:41:26.800 theory is refuted or it's not refuted. Now, if it is refuted, then you can say, well, Einstein's
00:41:32.560 theory is corroborated by the evidence. I would just prefer to say, explains the evidence.
00:41:38.640 It explains the evidence. It explains the evidence better than any rival. In fact, there are no
00:41:44.960 rivals. There are no longer any rivals. How can you say there are no rivals? Because we didn't
00:41:49.680 experiment and the only rival that existed up until that point was Newton's theory of gravity.
00:41:56.400 But it couldn't explain the results of this particular experiment, called it Edington's experiment,
00:42:01.280 but any number of things these days, Newton's theory can't explain this stuff. So what can
00:42:06.880 Einstein's theory? Is it well corroborated? Doesn't matter. Doesn't matter. Whatever this word
00:42:12.960 corroborated means, it doesn't matter the degree of corroboration. You can say, it explains
00:42:18.000 absolutely everything that has been asked of it. Well, maybe not. It can't quite explain what
00:42:24.240 dark matter is. It can't explain what dark energy is. Okay, fine. But nor can anything else.
00:42:29.760 For the stuff that is explicable by general relativity, by Einstein's theory, nothing else can do
00:42:35.840 that job. Nothing else explains precisely why Mercury follows the orbit that it does around the sun.
00:42:43.760 Only one theory does that. But is that theory well corroborated? That doesn't matter.
00:42:48.560 That doesn't matter. On the one hand, you can say, yeah, every single time you observe Mercury,
00:42:53.360 that's another corroboration of Einstein's theory of general relativity. But again, degree of
00:42:59.760 corroboration, if tomorrow, someone comes up with an explanation for gravity, that not only does
00:43:07.120 absolutely everything that Einstein's general theory of relativity does, not only is able to
00:43:13.920 accomplish absolutely everything that the existing theory of gravity does, but also goes a bit
00:43:18.800 further. It is able to solve the problem of what dark matter is. Then we'll be in a situation where
00:43:25.680 we have a theory of gravity that is superior to Einstein's theory of gravity because it can do
00:43:31.680 everything Einstein's theory of gravity can do. And it can explain dark matter. And in all likelihood,
00:43:39.440 the history of science suggests that when you have a good explanation, it reaches into areas
00:43:46.480 by postulating the existence of entities hit the two you never even imagined. So in the case of
00:43:52.080 general relativity, it reaches into areas where it postulates things like black holes and big bangs
00:43:57.440 and the curvature of space time and gravitational lensing and various things that weren't even
00:44:03.600 thought of prior to general relativity. And so whatever the success or ears to general relativity
00:44:11.040 was you'd expect the same kind of thing. Now I just want to pick up in this section that I'm reading
00:44:15.920 from a few other gems here that really they stand the test of time. They just written so well
00:44:21.680 because Poppa at times because Poppa wrote a vast amount and as I keep on emphasizing he was
00:44:28.720 writing for his contemporaries. But if you can get through some of the jargon at times,
00:44:33.600 the gems in here are absolutely brilliant. And well, let me just pick up one.
00:44:38.800 Quote, every observation and to an even higher degree every observation statement
00:44:45.760 is itself already an interpretation in the light of our theories. Yet even though this fact is
00:44:52.640 most important, it raises a minor issue compared with what I wish to criticize here, a general
00:44:58.640 attitude, a general philosophy of science, a philosophy which makes its main problem that of
00:45:04.640 explaining when science derives its certainty, its rational reliability, its validity or its authority.
00:45:13.280 For I hold that science has no certainty, no rational reliability, no validity, no authority.
00:45:20.160 The best we can say about it is that although it consists of our own guesses, of our own
00:45:26.000 conjectures, we are doing our very best to test them. That is to say, to criticize them and to
00:45:33.440 refute them, but the inductivist philosophy not only attributes authority to science,
00:45:38.880 it also perhaps quite unwittingly, attributes to science, a cautious and indeed timid approach,
00:45:45.840 which is entirely foreign to our real procedure end quote. And that of course is the quote I
00:45:51.920 began this episode with, but isn't it brilliant? That Poppa there still is a revolutionary
00:45:58.400 of a kind when it comes to epistemology because so many people simply don't understand that,
00:46:03.440 and yet reject him out of hand and yet here he is getting straight to the point about the purpose
00:46:09.360 of science and what we're up to, what we're doing when it comes to science and even more broadly
00:46:14.160 than science, testing our own guesses and our conjectures, trying to criticize and refute them,
00:46:20.240 but instead so many people have fixated on this idea of the authority of something or other
00:46:25.600 and the authority of science and Poppa rejected that. You're rejected authority everywhere,
00:46:31.440 politics, moral authorities, political authorities, scientific authorities, okay, just a little more
00:46:38.480 here before we, I know this is going to be an extremely long episode, but just a little more here
00:46:43.200 about corroboration. Section 28, Poppa goes on to say, in the foregoing section I introduced
00:46:49.920 the term degree of corroboration to characterize the degree to which our hypothesis has stood up
00:46:56.080 to tests. In the present section I intended to discuss merely a terminology issue my reasons
00:47:02.640 for proposing to speak of degree of corroboration rather than of the probability of a hypothesis
00:47:09.440 in the light of the tests. My main reason is of course that the latter phrase, although in itself
00:47:15.440 perfectly legitimate, is liable to lead to confusion and quote, so where I disagree is it's not perfectly
00:47:22.320 legitimate. I don't think there is such a thing as probability of a hypothesis in the light of the
00:47:28.560 tests. I just don't think there's any point of speaking in that way, and he's attempt to improve
00:47:34.640 this by saying, well I'm not going to talk about probability, I'm instead going to talk about
00:47:38.960 corroboration is a false dichotomy. Both of these approaches, unusually for Poppa, can be rejected
00:47:48.240 because of Poppa's own best solution, which is what we're doing is ruling out all the other theories
00:47:53.760 left with only one that doesn't need, so we don't need to be concerned about corroboration.
00:47:58.640 We just need to be concerned about whether or not our theory solves our problem, explains our
00:48:04.240 phenomena, has no other rivals that are yet to be ruled out by experimental tests, which in
00:48:10.320 science, they always are. At least eventually Poppa spends many, many pages criticizing, and I
00:48:17.200 think rightly the whole language of probability to a large extent and the whole application of
00:48:23.920 probability to epistemology. He highlights errors and it's very much worth reading. I'm not going
00:48:29.920 to read through it now, as I say, this is already getting very long and perhaps I'm losing part of my
00:48:35.040 audience, but let me pick it up where Poppa does try to introduce an example, and so where he tries
00:48:42.240 to justify, for what of another word, is use of this whole concept of corroboration, a degree of
00:48:49.040 corroboration. He goes on to say, quote, no rule of content holds for the first sense of probability,
00:48:55.600 which I call degree of corroboration. On the contrary, most physicists will say that Maxwell's
00:49:02.080 theory of light is more probable in the first sense that is to say better corroborated or
00:49:07.360 better tested than Fresnel's theory of light. The reason is that Maxwell's theory has been more
00:49:13.280 widely and more severely tested, even in fields in which Fresnel's theory cannot be tested.
00:49:20.000 At the same time, Maxwell's theory has a much greater logical content than Fresnel's.
00:49:24.560 Maxwell's is a wave theory of light and a theory of electromagnetism, while Fresnel's is
00:49:30.000 merely a wave theory of light. Thus, Maxwell's theory, although more probable in the sense of being
00:49:35.600 better corroborated or better tested, is at the same time less probable in the sense
00:49:40.400 of the second usage of the word, which is also very well established, especially if we are thinking
00:49:44.240 not so much of the tests successfully passed by hypothesis, but rather the chances that an event
00:49:51.040 will occur end quote. Again, this is all useless. I don't think that we need to worry about
00:49:59.040 how many, you know, counting up, how many successfully passed tests, a particular hypothesis,
00:50:07.360 has managed to pass. If this was the case, if this was the case, then my mystery about,
00:50:14.160 why is it on this view of things being more probable, the more tests they passed,
00:50:20.000 why is it that the day before the day before Newton's theory of gravity was finally
00:50:27.520 once and for all refuted by Edington's experiment? Why was it that day? Apparently,
00:50:32.720 its probability of truth was at its highest because it was that day that it had passed the most
00:50:38.560 tests, thousands, millions of tests, perhaps, every time someone observed anything that could have
00:50:44.560 been explained by Newton's theory of gravity, it was, it was, Newton's theory of gravity had passed
00:50:50.720 all of these tests and not passed some, but apparently Einstein's theory, which was brand new,
00:50:56.480 had not passed as many tests, it hadn't, it simply hadn't. Not many people had applied it to
00:51:03.120 a two certain problems, suddenly Albert Einstein hadn't, maybe a few other physicists who understood
00:51:08.560 the theory at the time in 1919, but really, it hadn't been used to solve problems, so it
00:51:14.240 wasn't passing tests, but that's irrelevant, it doesn't matter how many tests it's passed,
00:51:19.840 it can be the best explanation, it can be closer to true, closer to describing reality,
00:51:26.720 than any other alternative, including Newton's, which had have passed more tests.
00:51:32.000 This whole idea of corroboration introduces a problem, which doesn't need to be solved.
00:51:37.600 The problem would be, well, Newton's theory has been very heavily corroborated by all of these
00:51:43.920 tests that went before, but Einstein's theory had not yet been corroborated. The
00:51:48.400 kids of our corroboration, it's about refutation and explanation. And when we have two competing
00:51:54.480 explanations, we seek a method of refutation, of trying to rule it out, that's exactly what happens.
00:52:02.240 Now, let me go towards the end of this particular section and finally,
00:52:07.440 tie up, loose ends about this, and why, again, I would love to be able to speak to
00:52:12.240 Popper about this, because Popper almost admits, he almost admits in his very chapter,
00:52:16.960 that he's making a mistake. I'll just read. He says, quote, and this is coming from page number
00:52:24.480 228 of realism in the aim of science. He writes, quote, there is unfortunately the danger of
00:52:31.440 another terminological confusion until recently, in fact, until the logic of scientific discovery
00:52:36.800 was in galley proof. I did not use the term degree of corroboration, but in its place,
00:52:43.440 the term degree of confirmation. And I made use of this term for precisely the same reason,
00:52:51.200 because of the need to avoid the term probability. Therefore, I must now make clear why I have
00:52:57.600 decided to change my terminology after using it in at least half a dozen publications.
00:53:03.600 End quote. Okay. I'm not going to go on with his explanation here about why he's chosen to
00:53:10.320 change tact to using degree of corroboration. He's saying right there that he wants to avoid
00:53:16.160 the term probability. He understands that his own epistemology entails that you can't say things
00:53:23.920 like this hypothesis is probably true, and you're not even aiming for something to be
00:53:28.720 probably true or certainly true or anything like that. But despite that, because he feels like he
00:53:34.720 needed to have some replacement for this whole idea of things being more probable, a particular
00:53:41.120 theory being more probable, but he knows that that's a mistake. He knows that that's wrong.
00:53:45.360 He knows that he's not trying to, and our project rather, in science and anywhere else,
00:53:50.880 is not to try and find the most probable theory. That's not what we're trying to do.
00:53:54.640 But he thinks he needs a replacement for that. So at first, in the logic of scientific discovery,
00:53:59.680 at first, his best guess was to say, well, rather than talk about how well, how probable a
00:54:06.800 particular theory is, he's going to talk about the degree of confirmation, how well confirmed
00:54:13.520 a theory is. And now, decades later in writing this book on reflection, he's realized, whoa,
00:54:21.200 I shouldn't have been talking about degree of confirmation. We can't confirm our theories,
00:54:26.480 or partly confirm our theories or anything like that. He realizes that's wrong.
00:54:30.080 But he still wants something else. He wants corroboration now to fill the void.
00:54:34.880 But the fact is, there is no void. His own philosophy does away with that whole thing,
00:54:40.480 that whole way of speaking and thinking, this whole idea of this particular theory is more probable.
00:54:46.720 This particular theory is more confirmed, or best confirmed, or this theory even has a higher
00:54:53.360 degree of corroboration. I think all of that is not a part of Papurian epistemology.
00:55:00.560 Papurian epistemology is about trying to create good explanations. And if you're lucky,
00:55:07.200 as I've said for about the 10th time in this episode alone, if you're lucky, you might have
00:55:11.840 competing explanations. That's extremely rare. But if you do, then you do what Papur has explained.
00:55:18.000 You come up with the experimental test that rules out all but one, never mind how probable that
00:55:24.960 theory is. Never mind if it's certain, or has some degree of certainty, and never mind if it's been
00:55:31.360 corroborated by repeated observations. That's not important. What's important is whether or not
00:55:38.320 your theory or explanation actually does the job of accounting for those observations,
00:55:43.520 that finite set of observations, and then allows you to infer what is true about the rest of
00:55:49.520 physical reality, because you're explaining the rest of physical reality. So your predictions
00:55:55.040 about the rest of physical reality are derived from that explanation, but it's not the other way
00:56:01.440 around. You're not deriving the explanation from the finite set of observations, and then predicting
00:56:08.560 that therefore that finite set of observations in some way inductively infers or entails
00:56:15.440 that a particular set of observations will continue. In fact, it might say quite the opposite.
00:56:19.520 That the explanation of this particular finite set of observations says that tomorrow they're not
00:56:24.240 going to continue for whatever reason. Okay, I did say I was going to finish there,
00:56:29.280 but I can't. There is a little bit right at the end here of this section on corroboration that
00:56:35.840 well, I have to give Carl Popper the final word, almost the final word.
00:56:42.080 Let's read his conclusion here, and he writes, quote, I conclude this section by giving a summary
00:56:48.080 of my views concerning corroboration in the form of seven points, the first of which contains
00:56:53.520 the fundamental idea. Number one, the degree of corroboration of a theory is an evaluation of the
00:56:59.840 results of the empirical tests. It has undergone end quote. Okay, so that point number one there,
00:57:06.320 I've got no problem with. I mean, he could talk about the degree of corroboration as
00:57:10.320 an evaluation of the results of empirical tests. In other words, it would be a synonym for
00:57:16.240 which theory survives the process of experimental refutation. And if you want to call that degree
00:57:22.240 of corroboration fine, I think it might be a little bit misleading because we've got this concept
00:57:27.200 of degree. This one has more survived than the other, but it's a black and white thing, isn't it?
00:57:34.000 I mean, if you want to say that on and off is a degree or true and false is a degree
00:57:40.080 of truth or falsity fine, but most people what they mean by degree is a gray scale, whereas I
00:57:46.320 would say it's black and white. And I think that Popper thinks it's black and white as well,
00:57:50.320 but he's speaking in the language of his opponents. Let's continue. Quote, number two,
00:57:56.640 there are two attitudes, two ways of looking at the relations between a theory and experience.
00:58:03.200 One may look for confirmation or for refutation. These two attitudes are obviously variants
00:58:09.760 of the apologetic or dogmatic and of the critical attitude. Scientific tests are always
00:58:17.360 attempted refutations. Number three, the difference between attempted confirmations and attempted
00:58:25.360 refutations or tests is largely though not completely amenable to logical analysis. Four,
00:58:32.640 a theory will be said to be the better corroborated the more severe the tests it has passed
00:58:38.880 and the better it has passed them end quote. Again, issue probably guess what I'm going to say is
00:58:44.880 that's kind of pointless when you've got an explanation either it passes the test or it doesn't.
00:58:50.480 It's very rare to worry about which one is passing more severe tests. You might very well say
00:58:57.680 okay, we're comparing Newton and Einstein and Eddington's experiment is a severe test and in passing
00:59:03.760 that severe test, that is the reason why we endorse that theory and not Newton's theory.
00:59:09.680 But I think it's just a wrong way around the phrasing things. Rather than talk about which one is passing
00:59:15.520 tests, just talk about which one failed the test and which one failed the test. Never mind severity,
00:59:21.360 which one failed it was Newton's theory. So for all practical purposes discard that theory
00:59:28.240 when it comes to explaining things like the results of Eddington's experiment. You might not
00:59:32.800 discard the theory for a whole bunch of other things like predicting the tides or how fast
00:59:38.240 apples are falling from trees it could be very useful for that kind of thing. But pop it goes on
00:59:42.720 to say point five quote a test will be said to be the more severe the greater the probability
00:59:48.640 of failing it the absolute or prior probability as well as the probability and light of what I
00:59:53.680 call our background knowledge that is to say knowledge which by common agreement is not questioned
00:59:58.320 while testing the theory under investigation end quote. Well there we have a whole lot of stuff
1:00:03.520 about probability. The very thing that Popper has said you know it's not of much use when it comes
1:00:10.400 to epistemology. But I think he still of course thinks well probability is this real thing.
1:00:16.400 Unlike with David Deutsch who basically has concluded it's all a scam. I mean real life as I've
1:00:22.960 explained in other podcasts doesn't obey the probability calculus. Even gaming machines and
1:00:29.280 roulette wheels and so on and so forth strictly strictly do not obey the probability calculus
1:00:34.880 only approximately so but approximately so isn't reality and isn't our best explanation.
1:00:39.840 Best explanation is quantum theory which says what actually happens anyway let's just continue
1:00:46.000 point six Popper says quote. Thus every genuine test may be described intuitively as an attempt
1:00:53.600 to catch theory. It is not only a severe examination but as an examination it is an unfair one.
1:01:01.120 It is undertaken with the aim of failing the examining rather than the aim of giving him a
1:01:07.200 chance to show what he knows. The latter attitude would be that of the man who wants to confirm
1:01:12.880 or to verify his theory. And seven assuming always that we are guided in our tests by a genuinely
1:01:21.200 critical attitude and that we exert ourselves in testing the theory and assumption which cannot
1:01:27.040 be formalized. We can say that the degree of corroboration of a theory will increase with the
1:01:32.240 improbability in the light of background knowledge of the predicted test statements provided
1:01:38.080 the predictions derived with the help of the theory are successful end quote.
1:01:42.880 Yes so again so much here is as I say, couched in terms of the language dominant at the time among
1:01:53.200 the people he was trying to explain his philosophy to and he was having a uphill battle in trying
1:01:58.160 to get these ideas across. This book written decades after the logic of scientific discovery
1:02:03.360 and he's still although you know objective terms won the debate. He wasn't winning the debate
1:02:09.840 in academia. The academics didn't accept what he was saying to a large extent. All we need is
1:02:16.800 to my mind point two that he said there. You've got two attitudes, two ways of looking at the
1:02:23.840 relationship between a theory and experience. Either you can look for confirmation or you can look
1:02:28.800 for refutation and these things are not symmetrical. The truth is scientific tests and indeed
1:02:34.320 all of our critical apparatus are attempted refutations and if they fail that means you just have
1:02:41.520 to rely upon your existing theory. Your existing theory is the best thing you've got in order to
1:02:47.520 try and solve your problem situation and if it can't solve your problem situation you've still
1:02:52.800 got a problem situation. So therefore you better get about creating a new theory but this whole
1:02:58.720 scheme of things being probably true, things being more or less certain, things being corroborated
1:03:06.960 or not. He's irrelevant and Popper doesn't need it and popular in a epistemology doesn't need it.
1:03:12.560 And as I say in the next episode, okay this is sort of an episode in two parts I suppose,
1:03:18.560 we will see what David Deutsch has to say about this idea that we can understand
1:03:24.080 Popper's epistemology better than Popper in the same way as David will explain. People today
1:03:30.720 can understand Einstein's theories better than Einstein or Darwin's theory of natural selection
1:03:36.240 better than Darwin. Just because these people begin by explaining to us for the first time the theory,
1:03:43.040 we shouldn't expect that they best explain the theory. The theory rightly deserves to be
1:03:49.680 named after them but that does not mean they have some sort of ownership of it. And in fact Popper
1:03:54.240 was one of the first to actually say that. Once the theory has been explained, it ceases to be
1:04:02.320 that person's theory, they don't have that, they cannot claim any special insight into the theory,
1:04:08.000 other people can be more insightful about that theory. I'm certainly not claiming that,
1:04:11.920 what I am claiming is that Popper's own explanation of his own theory means it's sometimes
1:04:17.440 some of what he said about that theory is kind of redundant. I think we need it. But until next time,
1:04:23.680 we should be very soon. I'll release the next one very soon. Bye bye.