00:00:12.160 Hello. I'm going to try a little experiment today. I'm going to start reading through
00:00:18.000 parts of the beginning of infinity, just the first chapter. I've sent at least one other
00:00:22.960 Aussie doing this with a different series of books. I don't know if it'll be useful. I thought I'd try
00:00:31.120 it, just as an experiment, sort of thinking out loud, I guess, and just commenting on some of my
00:00:37.680 favorite parts of the book. By doing so, it kind of clarifies my own thoughts about things. That's
00:00:44.560 one thing. The other thing that there is a sense in the critical rationalist community that
00:00:51.600 there's a lot of bad ideas out there, or false ideas, that get a lot of traction in an undeserved
00:00:59.680 kind of way, while the ideas of Popper and Deutsch, for example, don't. This might go some way
00:01:08.880 to addressing that. So in the case of Popper, we've got, of course, the open society and its enemies
00:01:16.560 and the fact that people misunderstand what democracy is about. We have the idea on the ascendancy
00:01:24.640 that Bayesianism is true, which I've commented recently, is essentially inductivism in a cheap
00:01:30.880 tuxedo. So we thought we'd dispatched with inductivism. We hadn't. It's come back stronger than ever
00:01:39.200 in a sense. And it has real practical problems. It means that science ends up being a quest for
00:01:49.120 formula, mathematical formula, that in some sense can statistically describe certain trends that we
00:01:56.560 have. So rather than being a kind of problem-solving creative exercise, people are diverted into
00:02:04.400 looking for statistical patterns. And this is very useful in some areas. There's no doubt, but when
00:02:10.720 all of science appears to be on board, we're thinking that this is the way in which science works,
00:02:16.640 it can cause delays. It can cause problems. But I digress. I'm here today to talk to about, talk about
00:02:25.200 the beginning of infinity. And I'm going to start with chapter one, the regional explanations,
00:02:32.240 and my, ooh, there's thunder outside. And my purpose here is just to read out some of my
00:02:41.360 favourite parts of each chapter. Not to serve as a summary. David, in fact, did this himself
00:02:48.240 at the conclusion of each chapter. He wrote two forms of summary. One is the meanings of the
00:02:56.080 beginning of infinity found in this chapter and an actual summary at the end of each chapter.
00:03:02.240 So if you were looking for the cribnites version of the beginning of infinity, you can simply
00:03:06.160 read the summaries at the end of each chapter. It's not a recommended way to try and understand
00:03:11.440 some of the subtleties of these ideas. But certainly you can go a long way of understanding the
00:03:17.280 philosophy by doing that rather than doing nothing. So let me begin partway through chapter one,
00:03:25.840 David says, how do we know? One of the most remarkable things about science is the contrast
00:03:31.040 between the enormous reach and power of our best theories and the precarious local means by which
00:03:36.560 we create them. No human has ever been to the surface of a star that alone visited the core where
00:03:42.320 the transmutation happens and the energy is produced. Yet we see those cold dots in our sky and
00:03:48.400 know that we are looking at the white hot surfaces of distant nuclear furnaces. Physically,
00:03:54.480 that experience consists of nothing other than our brains responding to electrical impulses from
00:03:58.720 our eyes. And eyes can detect only light that is inside them at the time. The fact that the
00:04:04.480 light was emitted very far away and long ago and that much more was happening there than just
00:04:09.920 the emission of light. Those are not things that we see. We know them only from theory.
00:04:17.280 Scientific theories are explanations, assertions about what is out there and how it behaves.
00:04:23.760 Where do these theories come from? For most of the history of science, it was mistakenly believed
00:04:29.600 that we derive them from the evidence of our senses. A philosophical doctrine known as
00:04:35.760 empiricism. Now it then provides, the beginning of infinity provides a diagram. Diagrammer has
00:04:43.520 in in form it is in the first box it says sensory experiences. There's an arrow which is labeled
00:04:50.720 derivation such as extrapolation, generalization or induction and that leads to another box that says
00:04:57.520 theories, knowledge of reality. So this is the idea of empiricism that we go out into the world,
00:05:02.960 we use our senses to read a book of nature in a sense. I'll continue. For example, the philosopher
00:05:12.320 John Locke wrote in 1689 that the mind is like white paper onto which sensory experience writes
00:05:19.360 and that that is where all of our knowledge of the physical world comes from. Another empiricist
00:05:25.280 metaphor was that one could read knowledge from the book of nature by making observations. Either
00:05:30.800 way, the discovery of knowledge is its passive recipient, not its creator. But in reality, scientific
00:05:37.280 theories are not derived from anything. We do not read them in nature nor does nature write them
00:05:42.880 into us. Our guesses, bold conjectures. Human minds create them by rearranging, combining, altering
00:05:50.480 and adding to existing ideas with the intention of improving upon them. We do not begin with
00:05:55.280 white paper at birth, but with inborn expectations and intentions and an innate ability to improve
00:06:01.920 upon them using thought and experience. Experience is indeed essential to science,
00:06:07.680 but a role is different from that supposed by empiricism. It is not the source from which
00:06:13.520 theories are derived. Its main use is to choose between theories that have already been guessed.
00:06:20.000 This is what learning from experience is. However, that was not properly understood until
00:06:26.960 the mid 20th century with the work of the philosopher Carl Popper. So historically,
00:06:31.440 it was empiricism that first provided a plausible defense for experimental science as we now know it.
00:06:38.880 Empiricist philosophers criticized and rejected traditional approaches to knowledge,
00:06:43.520 such as difference to authority of holy books and other ancient writings,
00:06:48.000 as well as human authorities such as priests and academics, and belief in traditional law,
00:06:53.760 rules of thumb and hearsay. Empiricism also contradicted the imposing and surprisingly persistent
00:07:00.640 idea that the senses are little more than sources of error to be ignored. And it was optimistic,
00:07:07.120 being all about obtaining new knowledge in contrast with the medieval fatalism that it expected
00:07:11.920 everything important to all to be known already. Thus, despite being quite wrong about where scientific
00:07:18.800 knowledge comes from, empiricism was a great step forward in both the philosophy and history
00:07:24.480 of science. Nevertheless, the question that skeptics, friendly and unfriendly,
00:07:29.680 raised from the outset always remained, how can knowledge of what has not been experienced
00:07:35.280 possibly be derived from what has? What sort of thinking could possibly constitute
00:07:41.200 a valid derivation of the one from the other? No one would expect to deduce the geography of Mars
00:07:47.520 from the map of the earth, so why should we expect to be able to learn about physics on Mars
00:07:53.280 from experiments done on earth? Evidently, logical deduction alone would not do because there
00:08:00.560 is a logical gap. No amount of deduction applied to statements describing a set of experiences
00:08:06.480 can reach a conclusion about anything other than those experiences. The conventional wisdom
00:08:12.960 was that the key is repetition. If one repeatedly has similar experiences under similar
00:08:19.200 circumstances, then one is supposed to extrapolate or generalise that pattern and predict that it
00:08:24.400 will continue. For example, why do we expect the sun to rise tomorrow morning? Because in the past,
00:08:30.080 so the argument goes, we have seen it do so whenever we have looked at the morning sky.
00:08:35.120 From this, we supposedly derive the theory that under similar circumstances, we shall always
00:08:42.240 have that experience, whether we probably shall. On each occasion, when the prediction comes
00:08:47.360 true and provided that it never fails, the probability that it will always come true is supposed
00:08:52.080 to increase. Thus, one supposedly obtains ever more reliable knowledge of the future from the past
00:08:58.400 and of the general from the particular. That alleged process was called inductive inference,
00:09:04.720 or induction. And the doctrine that scientific theories are attained in that way is called inductivism.
00:09:11.600 To bridge the logical gap, some inductive us imagine there is a principle of nature, the principle
00:09:16.000 of induction, that makes inductive inferences likely to be true. The future or resemble the past
00:09:22.480 is one popular version of this, and one could add that the distant resembles the near, the unseen
00:09:28.880 resembles the scene, and so on. But no one has ever managed to formulate a principle of induction
00:09:35.120 that is usable in practice for obtaining theories from experiences. Historically, criticism of
00:09:41.440 inductivism has focused on that failure, and on the logical gap, that cannot be bridged.
00:09:47.040 But that lets inductivism off far too lightly. For it concedes inductivism's two most serious
00:09:52.960 misconceptions. First, inductivism purports to explain how science obtains predictions about
00:09:59.200 experiences, but most of our theoretical knowledge simply does not take that form. Scientific
00:10:04.640 theories are about reality, most of which does not consist of anyone's experiences. As
00:10:10.080 Traphysics is not primarily about us, what we shall see if we look at the sky.
00:10:16.720 But rather about what stars are, their composition, and what makes them shine, and how they are
00:10:21.920 formed, and the universal laws of physics under which they happened. Most of that has never been
00:10:27.440 observed. No one has experienced a billion years or a light year, no one could have been present
00:10:32.960 at the Big Bang, no one will ever touch a law of physics except in their minds through theories.
00:10:38.880 All our prediction of how things will look at the juice from such explanations of how things are.
00:10:44.880 So inductivism fails even to address how we can know about stars and the universe as distinct
00:10:50.800 from just dots in the sky. The second fundamental misconception in inductivism
00:10:57.600 is that scientific theories predict that the future were resembled a past and that the unseen
00:11:03.520 resembles the scene and so on, or that it probably will. But in reality, the future is unlike the past,
00:11:10.160 the unseen very different from the scene. Science often predicts and brings about phenomena
00:11:16.160 spectacularly different from anything that has been experienced before. For millennia, people dreamed
00:11:22.240 about flying, but they experienced only falling. Then they discovered good explanation,
00:11:26.880 explanatory theories about flying, and then they flew in that order. So that's a part just as an
00:11:32.880 aside. This is me, this is not from the book. That David references in his TED Talk, one of
00:11:40.800 his TED Talks. These first three chapters provide much of the material for a new way to explain
00:11:51.680 explanation that TED Talk there. I'll continue. Before 1945, no human being had ever observed
00:12:03.440 a nuclear fission atomic bomb explosion. There may never have been one in the history of the universe,
00:12:08.800 yet the first such explosion and the conditions under which it would occur had been accurately
00:12:14.000 predicted. But not from the assumption that the future would be like the past. Even sunrise,
00:12:19.440 that favorite example of inductivists is not always observed every 24 hours. When viewed from
00:12:26.480 orbit, it may happen every 90 minutes or not at all. And that was known from theory long before
00:12:32.560 anyone had ever orbited the Earth. In fact, it only happens 24 hours very rarely. It's nearly a
00:12:40.240 quader or a, you know, what a mid-year solstice or something like that. If you go to Antarctica
00:12:49.120 or anywhere inside the Arctic Circle or the Antarctic Circle, you'll find that the sun's not
00:12:55.120 setting or it's not rising. I'll continue. It is no defense of inductivism to point out
00:13:04.400 that in all these cases, the future still does resemble the past in the sense that it obeys
00:13:08.400 the same underlying laws of nature. For that is an empty statement. Any purported law of nature,
00:13:14.400 true or false about the future and the past is a claim that they resemble each other by both
00:13:19.280 conforming to that law. So that version of the principle of induction could not be used to derive
00:13:25.920 any theory or prediction from experience or anything else. So this is, again, a world-class
00:13:36.480 philosopher explaining in detail why it is that induction cannot possibly work. Not only that it
00:13:48.800 isn't actually the thing that allows scientific theories to be constructed,
00:13:55.680 it's not even a thing that allows you to construct any kind of knowledge whatsoever.
00:14:03.520 It doesn't exist. Induction is irrationality. It's a way of generalizing.
00:14:10.000 What is surprising to me is that professional philosophers seem to enjoy teaching their students
00:14:20.880 logical fallacies and they understand logical fallacies quite well. One of these logical fallacies
00:14:26.480 is the fallacy of hasty generalization. This is the idea that if you see something a number of
00:14:33.360 times that you conclude that therefore it's a general rule. This is the kind of faulty reasoning
00:14:41.920 that leads to things like racism, let's say. It's where you observe a particular behaviour amongst
00:14:48.880 a certain group of people that look the same and you form the conclusion that therefore it applies
00:14:56.400 to all people that look that way. It is the assumption that if you see a bunch of white swans
00:15:06.160 that you assume that therefore all swans are white based on a limited number of observations.
00:15:15.200 But how hasty do you have to be for a generalization to be hasty?
00:15:20.800 Well I would say it's impossible to do an exhaustive search or to come up with a regularly occurring
00:15:32.960 phenomenon such that it has occurred a sufficient number of times for you to have any degree of
00:15:41.040 confidence that it will continue into the future. It's not like observing the same thing 10 times
00:15:46.160 in a row means you are 10 times less confident than if you observe the thing 100 times in a row.
00:15:55.520 It doesn't matter if you observe things every single day forever for a long, long time.
00:16:02.960 It doesn't mean that that thing is not going to be contravened the very next day. David actually
00:16:08.080 goes on to use his example of up until December 31st, 1999 he'd only ever experienced
00:16:18.400 in front of in the year he'd only ever experienced a one-nine, he'd never experienced a two-zero.
00:16:23.760 But his experience throughout his life had seen a 19 in front of the year's date in 1991, 1992,
00:16:34.880 1993, et cetera, et cetera, up to 1999. He knew that that expectation of a one-nine was about
00:16:42.080 to be refuted or, namely, on January 1st, 2000. This is always the case. We cannot know,
00:16:51.520 especially in science out there in the physical world, that the phenomena that you've seen repeating
00:16:56.080 itself isn't part of a much larger, more complex phenomena where, at any moment, it's going to reveal
00:17:03.200 itself to be utterly false. Let me go back to the beginning of infinity. So David continues later.
00:17:16.480 I'm cutting out a fair bit here. In Paris there's never did achieve its aim of liberating
00:17:21.920 science from authority. It denied the legitimacy of traditional authorities, and that was salutary.
00:17:27.840 But, unfortunately, it did this by setting up two other false authorities. Sensory experience,
00:17:33.360 and one of the fictitious process of derivation, such as induction, one imagines,
00:17:38.160 is used to extract theories from experience. The misconception that knowledge needs
00:17:43.040 authority to be genuine or reliable dates back to antiquity, and it still prevails.
00:17:47.280 To this day, most causes in the philosophy of knowledge teach that knowledge is some form of
00:17:51.200 justified true belief, where justified means designated as true, or at least probable,
00:17:56.800 by reference to some authoritative source or touchdown of knowledge. Thus, how do we know
00:18:02.800 is transformed into, by what authority do we claim? The latter question is a chimera,
00:18:08.800 that may well have wasted more philosophers' time and effort than any other idea.
00:18:14.240 It converts the quest for truth into a quest for certainty, a feeling, or for endorsement,
00:18:21.440 a social status. That, this misconception, is called justificationism. The opposing position,
00:18:31.040 namely the recognition that there are no authoritative sources of knowledge,
00:18:34.320 nor any reliable means of justifying ideas as being true or probable, is called fallibleism.
00:18:41.120 To believe is in the justified true theory of knowledge, and a justified true belief theory of
00:18:46.080 knowledge. This recognition is the occasion for despair or cynicism, because to them it means
00:18:52.480 that knowledge is unobtainable. But to those of us, for whom creating knowledge means understanding
00:18:58.000 better what is really there, and how it really behaves and why, fallibleism is part of the very
00:19:04.080 means by which this is achieved. Fallibleists expect even their best and most fundamental
00:19:10.080 explanations to contain misconceptions in addition to truth, and so they are predisposed to
00:19:15.920 try to change them for the better. In contrast, the logic of justificationism is to seek,
00:19:22.000 and typically to believe that one is found, ways of securing ideas against change. Moreover,
00:19:28.400 the logic of fallibleism is the one is, moreover, the logic of fallibleism is that one not only
00:19:37.440 seeks to correct the misconceptions of the past, but hopes in the future to find and change
00:19:42.880 mistaken ideas that no one today questions are found problematic. So it is fallibleism,
00:19:48.320 not mere rejection of authority that is essential for the initiation of unlimited knowledge growth,
00:19:55.120 the beginning of infinity. Yes, this is a wonderful defence of fallibleism. I think that if there's
00:20:05.920 a philosophy, in a sense, that unites many people who follow the work of David Deutsch,
00:20:11.760 it is this idea of fallibleism. I think this is one of the strongest defences that we find
00:20:16.320 anywhere. David's spoken about it online in various places. There's the nautilus interview,
00:20:22.400 I think there's a wonderful job of that. I think he might have mentioned it in these closer to
00:20:26.480 truth interviews as well. So this concept that because you can be wrong and you can be wrong about
00:20:33.600 the truth, that admits the truth is, that admits that there is an objective truth,
00:20:40.640 because you can be wrong about it. And so we take the idea of realism seriously,
00:20:47.200 that there is an objective reality out there, and not all claims to that reality,
00:20:53.200 not all truth claims about that reality stand on equal footing. Some of them can be shown,
00:20:58.560 or already have shown to be false. Some of them haven't yet been shown to be false, but are good
00:21:03.040 explanations. And this idea of fallibleism is that because it is people who are constructing the
00:21:08.160 knowledge and people are prone to error, that we can never be sure that we've found the perfect
00:21:15.600 theory. It's always going to be riddled with errors in ways we don't know, and therefore subject
00:21:24.080 to improvement, whereby if we can make those improvements, we've made progress, an objective
00:21:30.240 progress, because we've corrected some errors. And so the new theory that we have that doesn't
00:21:36.080 have the errors of the old theory is objectively better. So we're moving in a particular direction,
00:21:41.760 and this is all underpinned by fallibleism, the simple acknowledgement that,
00:21:46.240 for any claim that is made by human beings, or by people generally, we can be wrong about it.
00:21:54.720 There is no royal road to truth, and anyone who claims that they have possession of the absolute
00:22:04.080 final truth, have to admit that they're human. And even if an infallible source has provided them
00:22:12.720 with information that is guaranteed to be true, it is them that is now interpreting it. It is them
00:22:17.440 that is now reporting to you what this truth is. And therefore their words, their words are fallible.
00:22:24.560 Their words could be in error. Their memory could be in error. A whole bunch of things could be
00:22:30.880 going wrong. Even if the source from which they're claiming to have gained this perfect knowledge
00:22:39.600 was itself perfect, because they're reporting it to you and they're fallible. And fallibleism
00:22:45.520 is about people. It's not about supernatural entities, although we could apply it there as well.
00:22:51.280 But if a supernatural entity tries to claim perfect knowledge, then we are quite entitled to ask
00:22:59.520 by what means, by what means has this inherent knowledge come. And when you hear it from the
00:23:10.880 inherent source, how do you know you're hearing it inherently? Okay, back to the book.
00:23:19.280 David goes on. However, rebellion against authority cannot by itself be what made the difference.
00:23:25.920 Authorities have been rejected many times in history and only rarely has any lasting good
00:23:31.280 comfort. The usual sequel has merely been that new authorities replaced the old. What was needed
00:23:37.520 for the sustained rapid growth of knowledge was a tradition of criticism. Before the Enlightenment,
00:23:44.240 that was a very rare sort of tradition. Usually the whole point of her tradition was to keep
00:23:49.600 things the same. So let me pause here. There are a number of genuine discoveries that David Deutsch
00:23:59.040 has made in philosophy, in epistemology, in history that appear in the beginning of infinity.
00:24:07.200 They appear in the fabric of reality as well. Some of them. But he in the beginning of infinity,
00:24:12.800 we have some truly groundbreaking ways of attempting to understand reality, humanity,
00:24:23.360 knowledge as a whole. And these ideas, how can we say this, have not thus far reached as many
00:24:35.200 people as they deserve to have reached yet. In this one in particular, it's such a useful starting
00:24:43.200 point for any historian. Anyone who's interested in history could benefit from really trying to get
00:24:51.120 to the heart of this, this idea of tradition of criticism. This idea in particular, this idea of a
00:24:56.080 tradition of criticism is an idea that could really inform history or the study of history,
00:25:02.480 or the study of sociology. Perhaps even the study of psychology to some extent as well. This idea
00:25:07.600 that criticism is a thing that allows progress and a tradition of criticism is something that
00:25:15.200 is an explanation for the reasons why there was a stark difference between the year 1,000 and the
00:25:22.080 year 2,000. Whereas the year 1,000 was very much like the year 0 and very much like the year 1,000
00:25:28.800 BC, which was very much like the year 10,000 BC. For the majority of human history,
00:25:36.000 things were in stasis, things were the same. There wasn't much change going on. And then
00:25:41.440 something remarkable happened. And it's called the Enlightenment, and it led to the industrial
00:25:45.920 revolution. But what are the philosophical underpinnings that caused these massive transformations?
00:25:51.200 Now, we can talk about our world science arrived and there was this idea of having
00:26:00.000 science being about testable ideas and then we discovered things like steam engines very well.
00:26:09.760 These are effects of a deeper philosophical phenomenon. And the philosophical phenomenon is,
00:26:16.240 as David explains here, a tradition of criticism. Let me continue with the beginning of infinity.
00:26:23.440 Thus, the Enlightenment was a revolution in how people sought knowledge by trying not to rely
00:26:29.280 on authority. That is the context in which empiricism, purporting to rely solely on the senses for
00:26:35.200 knowledge, played such a celebratory historical role, despite being fundamentally false,
00:26:40.560 and even authoritative in its conception of how science works. One consequence of this tradition
00:26:47.120 of criticism was the emergence of a methodological rule that a scientific theory must be testable.
00:26:53.920 Though, this was not made explicit at first. That is to say, the theory must make predictions
00:26:59.920 which, if the theory were false, could be contradicted by the outcome of some possible observation.
00:27:05.840 Thus, although scientific theories are not derived from experience, they can be tested by
00:27:09.920 experience. By observation or experiment. For example, before the discovery of radioactivity,
00:27:15.120 chemists had believed and had verified in countless experiments that transmutation is impossible.
00:27:20.480 Rather food and soddy boldly conjectured that uranium spontaneously transmutes into other elements.
00:27:26.160 Then, by demonstrating the creation of the element radium in a steel container of uranium,
00:27:31.120 they refuted the prevailing theory in science progressed. They were able to do that because
00:27:36.800 that earlier theory was testable. It was possible to test for the presence of radium.
00:27:41.680 In contrast, the ancient theory that all matter is composed of combination of elements,
00:27:46.800 including earth, air, fire, and water, was untestable. Because it did not include any way of
00:27:54.320 testing for the presence of those components. So it could never be refuted by experiment.
00:27:59.360 Hence, it could never be and never was improved upon through experiment. The enlightenment was
00:28:05.440 at root, a philosophical change. The physicist, Galileo Galilei, was perhaps the first to understand
00:28:13.360 the importance of experimental tests, which he called cementing, meaning trials by all deal,
00:28:17.920 as distinct from other forms of experiment and observation, which can more easily be mistaken for
00:28:22.480 reading from the book of nature. Testability is now generally accepted as the defining characteristic
00:28:28.320 of the scientific method, proper called it the criterion of demarcation between science and non-science.
00:28:33.600 Nevertheless, testability cannot have been the decisive factor in the scientific revolution either.
00:28:40.640 Contrary to what is often said, testability predictions had always been quite common.
00:28:48.800 Every traditional rule of thumb for making a flint blade or a campfire is testable.
00:28:53.840 Every would-be prophet who claims the sun will go out next Tuesday as a testable theory.
00:28:58.800 So that is every gambler who has a hunch that this is my lucky night. I can feel it. So what
00:29:04.560 is the vital progress enabling ingredient that is present in science, but absent from the testable
00:29:11.120 theories of the prophet and the gambler? Well, pausing here. Sometimes I'm left a little bit speech
00:29:21.280 that I can read this book a hundred times and and and and still come across things where I almost
00:29:28.400 get misty-eyed because it is a remarkable, a remarkably dense work where you come across these
00:29:36.800 things that no one else previously has said. Some people may know I've been forced at university
00:29:43.280 to read all sorts of philosophy books that didn't much strike me as interesting. Some were,
00:29:49.440 but what you do find is even when you pick up a good book, let's say daycarts and meditations
00:29:54.720 I really liked it and five meditations, it's only a short bit, but there's essentially one or two
00:30:00.320 ideas in there that are kind of wow that's amazing. And the rest is fluff, whereas this work here
00:30:09.520 and here's another example makes this grand breaking claim that's true and which still hasn't
00:30:20.160 made itself reach into the zeitgeist. So let me just summarize what I heard David say there.
00:30:31.440 Papa made the claim and many people seem to agree that testability is this thing that demarcate
00:30:36.720 science from non-science and he's right. However, testable theories have always existed.
00:30:46.080 There have always been testable theories around. A caveman who's trying to make a better flint
00:30:53.680 blade has a testable theory when he makes his new flint blade if it actually turns out to be better,
00:30:59.360 then he keeps that one and he refutes the old one. If it's not better, then he refutes the new one,
00:31:03.680 he keeps the old one. So testability can't be enough to say what the purpose of science is.
00:31:16.320 There has to be more to science than this. Testability has got something to do with it,
00:31:21.040 but it's not the whole story. David is about to go on to explain his improvement.
00:31:27.840 And I don't want to rely on authority, but those of us who have
00:31:38.480 been looking into the philosophy of science for a long time have read people who have criticized
00:31:46.240 Papa or people who claim to have improved Papa. We would readily admit if we can find someone
00:31:55.200 who has done either of those things, but it is almost never the case. And so when presented with
00:32:02.000 things like Bayesianism, we shrug our shoulders because we realize that Papa already
00:32:07.360 refuted inductivism and anyone who says inductivism works or that a variation on inductivism,
00:32:12.720 like Bayesianism, might be useful for science. We've heard this argument before.
00:32:19.120 Papa heard these arguments. These arguments have been going on for decades. We've read the
00:32:23.040 surrounding material and we remain unconvinced. But then when something about David Deutsch comes along
00:32:28.640 and he's not actually saying in the book, but maybe he said it once. He's not actually saying
00:32:34.880 he's improved Papa, but he does improve Papa. He does it right here. He does it explicitly. He's
00:32:39.840 about to do it right now. So it's an amazing accomplishment. And so let's just persevere. Let's
00:32:47.040 just, I'll go into the next paragraph. The reason that testability is not enough is that
00:32:52.960 prediction is not and cannot be the purpose of science. Consider an audience watching a
00:32:58.000 conjuring trick. The problem facing them has much the same logic as a scientific problem.
00:33:04.480 Although in nature, there is no conjurer trying to deceive us intentionally. We can be mystified
00:33:09.040 in both cases for essentially the same reason. Appearances are not self-explanatory. If the
00:33:15.520 explanation of a conjuring trick were evident in its appearance, there would be no trick.
00:33:20.160 If the explanations of physical phenomena were evident in their appearance, empiricism would
00:33:25.760 be true. And there would be no need for science as we know it. The problem is not to predict
00:33:30.800 the trick's appearance. I may, for instance, predict that if a conjurer seems to place various
00:33:35.600 balls under various cups, those cups will later appear to be empty. And on I predict that if the
00:33:40.800 conjurer appears to show, appears to saw someone in half that person will later appear on stage
00:33:46.080 unharmed. Those are testable predictions. I may experience many conjuring shows and seeing my
00:33:51.840 predictions vindicated every time. But that does not even address, let alone solve the problem of
00:33:57.680 how the trick works. Solving it requires an explanation, a statement of the reality that accounts
00:34:04.880 for the appearance. Okay, so now skipping a lot of chapter one, the reach of explanation,
00:34:13.520 I encourage people to actually go to chapter one, read the whole thing. And he speaks about
00:34:19.200 science and he speaks about progress and how progress has happened in the past. And now let me go
00:34:24.720 back to the book. But even testable explanatory theories cannot be the crucial ingredient that made
00:34:29.520 the difference between no progress and progress. For they too have always been common. Consider,
00:34:34.080 for example, the ancient Greek myth for explaining the onset of winter. Long ago, Hades,
00:34:38.960 God of the underworld kidnapped and raped Persephone, goddess of spring, then Persephone's mother,
00:34:44.720 Demeter, goddess of the earth, an agriculture, negotiated a contract for a daughter's release,
00:34:48.960 which specified that Persephone would marry Hades and eat a magic seed that would compel her to
00:34:53.440 visit him once a year thereafter. Whenever Persephone was away for filling his obligation,
00:34:58.000 Demeter became sad and will command the world to become cold and bleak so that nothing could grow.
00:35:02.080 And this is the wonderful example that David uses in his TED Talk and talks about how,
00:35:09.520 because it didn't need to be Demeter, it could have indeed been some other God,
00:35:15.840 there could have been some indifferent to Persephone, it didn't need to be a marriage contract,
00:35:20.800 it could have been any other contract. The specific parts of the theory that explain the seasons
00:35:26.880 on this Greek myth explanation are easy to vary. So let me read a part where he speaks
00:35:40.880 about this in particular. So with myths, the reason why those myths are so easily variable
00:35:46.800 is that their details are barely connected to the details of the phenomena.
00:35:50.480 Nothing in the problem of why winter happens is addressed by postulating specifically
00:35:54.480 a marriage contract or a magic seed or the God's Persephone, Hades and Demeter or Freya.
00:35:59.520 Whenever a wide range of variant theories can account equally well for the phenomena
00:36:05.440 they are trying to explain, there is no reason to prefer one of them over the others,
00:36:09.440 so advocating a particular one in preference to the others is irrational.
00:36:15.040 Skipping a bit. In general, when theories are easily variable in the sense I have described
00:36:20.400 experimental testing as almost useless for correcting their errors, I call such theories bad explanations.
00:36:27.280 Being proved wrong by experiment and changing the theories to other bad explanations
00:36:31.040 does not get their holders one job closer to the truth. Because explanation plays this central
00:36:36.720 role in science and because testability is of little use in the case of bad explanations,
00:36:41.520 I myself prefer to call myths, superstitions and similar theories unscientific,
00:36:47.200 even when they make testable predictions. But it does not matter what terminology is,
00:36:52.000 so long as it does not lead you to conclude that there is something worthwhile about the
00:36:54.880 Persephone myth or the prophets apocalyptic theory or the gamblers delusion just because it is
00:37:00.000 testable. Nor is a person capable of making progress merely by the virtue of being willing
00:37:06.560 to drop a theory when it is refuted. While must also be seeking a better explanation
00:37:10.960 for the relevant phenomena, that is the scientific frame of mind. Wonderful. So, here David is saying
00:37:18.320 that even if you have a testable prediction, that does not mean that you are entitled to be referred
00:37:24.640 to as a scientific theory. I think this is a well expressed way of understanding
00:37:34.720 proper in a new way, more explicitly than perhaps proper himself was able to explain. Continue.
00:37:41.760 As the physicist Richard Feynman said, scientists what we have learned that had to keep from
00:37:46.320 fooling ourselves by adopting easily variable explanations. The gamblers in the profit are
00:37:50.640 ensuring that they will be able to continue fooling themselves no matter what happens.
00:37:55.200 Just as thoroughly as if they had adopted untestable theories,
00:37:58.080 they are insulating themselves from facing evidence that they are mistaken about what is really
00:38:02.560 there in the physical world. The quest for good explanations is I believe the basic regulating
00:38:08.080 principle not only of science but of the enlightenment generally. When he says enlightenment generally,
00:38:12.240 he means through to today. He is not just talking about that shortish period that led to the
00:38:18.320 industrial revolution in the past. He is talking about the beginning of the enlightenment continuing
00:38:23.680 through to today. So, he would say the enlightenment is still happening now and I would agree.
00:38:28.000 It is the feature, this quest for good explanations. It is the feature that distinguishes
00:38:35.920 those approaches to knowledge from all others and it implies all those other conditions for
00:38:40.800 scientific progress I have discussed. It trivially implies that prediction alone is insufficient.
00:38:47.120 Somewhat less trivially, it leads to the rejection of authority because if we adopt a theory
00:38:52.640 on authority, that means that we would also have accepted a range of different theories on authority.
00:38:59.200 And hence it also implies the need for a tradition of criticism. It also implies a methodological
00:39:03.760 rule, a criterion for reality, namely that we should conclude that a particular thing is real,
00:39:10.240 if and only if it figures in our best explanation of something.
00:39:14.240 This is also mentioned in the fabric of reality. I think it talks about it as being something
00:39:20.400 like Dr. Johnson's, Johnston's criterion. Hold on.
00:39:29.760 Yeah, so there is a section in the fabric of reality of course about precisely this chapter
00:39:37.280 4 there is called criteria for reality and the discussion there is along the lines of Dr. Johnson's
00:39:44.320 understanding of this phenomena of things kicking back in reality. So if they kick back,
00:39:52.000 that means they act in autonomous, unpredictable ways in ways that you can't predict
00:39:59.680 how beforehand. So you have to go out and test things and this doesn't mean only science.
00:40:06.720 If you're investigating anything and it reacts in a way that you didn't predict,
00:40:10.880 then you know you've got something real. Okay, now David goes on to write about the actual
00:40:16.960 explanation of seasons and how the earth is on a tilt and he says, that is a good explanation.
00:40:24.720 Hard to vary because all its details play a functional role and that is important as well.
00:40:31.760 So a good explanation is hard to vary because all of its details play a functional role.
00:40:40.000 In the axis tilt theory of the earth, you can't tilt the axis by more than what it is actually
00:40:45.440 tilted because if you did, the seasons would be different than what they are.
00:40:51.280 Good explanations are often strikingly simple or elegant as I shall discuss in chapter 14.
00:40:57.280 Also, a common way in which an explanation can be bad is by containing superfluous features
00:41:02.160 or arbitrariness and sometimes removing those yields a good explanation.
00:41:06.160 This is given rise to the misconception known as Occam's Razor, named after the 14th century
00:41:11.120 philosopher William of Occam, but dating back to antiquity. Namely, the one should always seek
00:41:18.160 the simplest explanation. One statement of it is, do not multiply assumptions beyond necessity.
00:41:23.760 I think this is what David used in the fabric of reality. However, there are plenty of very
00:41:28.560 simple explanations that are nonetheless easy to vary, such as Demet did it, and while explanations
00:41:34.960 and while assumptions beyond necessity make a theory bad by definition, there have been many
00:41:40.800 mistaken ideas of what is necessary in a theory. Instrumentalism, for instance, can sit as
00:41:46.320 explanation itself unnecessary and so did many other bad philosophies of science, as I shall
00:41:51.280 discuss in chapter 12. When a formally good explanation has been falsified by new observations,
00:41:57.040 it is no longer a good explanation because the problem has expanded to include those observations.
00:42:02.480 Thus, the standard scientific methodology of dropping theories when refuted by experiment is
00:42:07.360 implied by the requirement for good explanations. The best explanations are the ones that are most
00:42:12.640 constrained by existing knowledge, including other good explanations as well as other knowledge
00:42:17.760 of the phenomena to be explained. That is why testable explanations that have passed stringent
00:42:23.600 tests become extremely good explanations, which is in turn why the maximum of testability
00:42:29.600 promotes the growth of knowledge and science. Conjectures are the products of creative imagination,
00:42:36.320 but the problem with imagination is that it can create fiction much more easily than truth.
00:42:41.760 As I have suggested, historically, virtually all human attempts to explain experience and terms
00:42:46.400 of a wider reality have indeed been fiction in the form of myths, dogma, and mistaken common sense,
00:42:52.320 and the rule of testability is an insufficient check on such mistakes. But the quest for good
00:42:57.280 explanations does the job, inventing falsehoods is easy, and therefore they are easy to vary once
00:43:04.080 found, discovering good explanations is hard, but the harder they are to find, the harder they are
00:43:08.800 to vary once found. The ideal that explanatory science strives for is nicely described by the
00:43:14.880 quotation from Wheeler with which I began in this chapter. Behind it all is surely an idea so
00:43:21.280 simple, so beautiful that when we grasp it in a decade, a century, or a millennium, we will all say
00:43:27.520 to each other how could it have been otherwise. Now we shall see how this explanation-based
00:43:33.440 conception of science answers the question that I asked above. How do we know about such? How do we
00:43:40.160 know so much about unfamiliar aspects of reality? Again, here we have it. Here we have the claim
00:43:49.760 that testability is not a sufficient bulwark against irrationality, but the quest for good explanations
00:43:58.800 is, the quest for good explanations that good explanations being hard to vary means that you are
00:44:06.880 seeking all the time to explain phenomena that exist, things that already feature in other good
00:44:12.480 explanations of reality, in terms that aren't arbitrary, in terms of things that can't easily be
00:44:21.440 swapped out for other things, you're looking for this hard to vary quality in your explanations,
00:44:27.360 and when you manage to get that and testability is one of those things, then you're on the right
00:44:36.880 track, then you know you're pointed in the direction of progress. Okay, so there's a little more
00:44:41.200 here, let me read to the remaining parts that I've highlighted. Suppose for the sake of argument
00:44:47.040 that you thought of the axis tilt theory yourself, it is your conjecture, it is your own original
00:44:53.120 creation, yet because it is a good explanation, hard to vary, it is not yours to modify.
00:44:58.800 It has an autonomous meaning and an autonomous domain of applicability. You cannot confine its
00:45:04.560 predictions to a region of your choosing, whether you like it or not, it makes predictions about
00:45:09.520 places both known to you and unknown to you, predictions that you have thought of and ones that
00:45:14.720 you have not thought of, tilt the planets in similar orbits in other solar systems must have
00:45:19.600 seasonal heating and cooling, planets in the most distant galaxies and planets that we shall never
00:45:24.080 see because I would have destroyed aeon's ago, and also planets that have yet to form,
00:45:28.320 the theory reaches out as it were from its finite origins inside one brain that has been
00:45:34.000 infected, that has been affected only by scraps of patchy evidence from a small part of one hemisphere
00:45:39.600 of one planet, to infinity. This reach of explanations is another meaning of the beginning of
00:45:45.280 infinity. It is the ability of some of them to solve problems beyond those that they were created
00:45:50.880 to solve. Again, here we have another wonderful piece of philosophy, another, I would call it a
00:45:55.920 discovery. This idea that some explanations have reached, infinite reach, is profound. It turns
00:46:06.160 knowledge into a force of nature. It is the thing that can point itself anywhere in reality and
00:46:13.840 transform that if it's so, if there are people there that want to do that and have the knowledge
00:46:20.320 of how to do so. So I'll say that again. Knowledge is an entity in the universe, but if it points
00:46:28.160 itself at somewhere in reality, it can transform that reality into something completely different,
00:46:35.200 as long as there are people there who choose to do so and have that knowledge.
00:46:40.400 So the knowledge of how to completely transform the Andromeda galaxy,
00:46:44.640 if it's discovered here on Earth, it is the thing that will transform the Andromeda galaxy,
00:46:52.240 which means that knowledge is kind of like a supermassive black hole. If there's a supermassive
00:46:58.560 black hole wandering through the universe that we haven't yet seen yet, although we probably
00:47:03.120 would have because of such things as gravitational lensing, but let's say it's wandering through
00:47:06.960 the universe and no one's noticed it yet, and it's heading towards the Andromeda galaxy,
00:47:11.040 and if it passes through the Andromeda galaxy, it could really upset the Andromeda galaxy.
00:47:15.360 It could perturb the orbits of the stars, it could change the shape of the Andromeda galaxy.
00:47:20.080 It can do that. No one has a problem understanding that. This idea that black holes hugely
00:47:27.280 powerful bodies can eat stars, they could transform the galaxy.
00:47:33.520 But so can knowledge. Knowledge can do what huge structures in physics can do.
00:47:40.320 They can transform things. We've already done it on a very small scale.
00:47:45.120 If you want to explain why it is that a city looks the way it does, there's no point consulting
00:47:51.120 the geology. Well, okay, there is. The geology has something to do with the way a city looks.
00:47:59.440 But what also has something to do with the way a city looks is the knowledge of the people that
00:48:04.880 are there. So knowledge is like a force of nature. It's like erosion if you like. It's this thing
00:48:11.600 that can pass over physical structures and change them. Let me continue.
00:48:20.480 The axis tilt theory is an example. It was originally proposed to explain the changes in the
00:48:25.040 Sun's angle of elevation during each year. Combined with a little knowledge of heat and spinning
00:48:29.760 bodies, then explained seasons. And without any further modification, it also explained why seasons
00:48:35.200 are out of phase in the two hemispheres and why tropical regions do not have them. And while the
00:48:39.280 summer sun shines at midnight in polar regions, three phenomena of which its creators may well have
00:48:45.280 been unaware. The reach of explanations is not a principle of induction. It is not something that
00:48:51.040 the creator of the explanation can use to obtain or justify it. It is not just part of the
00:48:56.320 it is not part of the creative process at all. We find out about it only after we have the
00:49:02.880 explanation. Sometimes long after. So it has nothing to do with extrapolation or induction or with
00:49:07.920 deriving a theory in any other alleged way. It is exactly the other way around. The reason that
00:49:14.080 the explanation of seasons reaches far outside the experience of its creators is precisely because
00:49:19.600 it does not have to be extrapolated by its nature as an explanation. When its creators first thought
00:49:26.080 of it, it already applied in our planet's other hemisphere and throughout the solar system
00:49:30.800 and in other solar systems and at other times. Thus, the reach of an explanation is neither an
00:49:35.840 additional assumption nor a detachable one. It is determined by the content of the explanation
00:49:40.080 itself. The better an explanation is, the more rigidly its reach is determined because the harder
00:49:46.160 it is to vary an explanation, the harder it is in particular to construct a variant with a different
00:49:50.640 reach with a larger or smaller that is still an explanation. I am going to skip over a little more
00:49:57.840 now. It also makes sense to speak of the reach of non-explanatory forms of knowledge. Rules of thumb
00:50:03.840 and also knowledge that is implicit in the genes for biological adaptations. So as I said,
00:50:09.920 my rule of thumb about cups and ball streaks has reached to a certain class of tricks,
00:50:14.240 but I could not know what that class is without further explanation for why the rule works.
00:50:21.360 And I'll stop down and chapter one. So this idea about the reach of explanations is another
00:50:27.760 phenomenal piece of philosophy, I think. It's up there with this idea that what science is about
00:50:34.560 is not merely testable theories because they've been common forever. Instead, it's about
00:50:39.520 hard to vary explanations of the physical world, I would say that's my interpretation, but
00:50:46.080 David makes the broader point that what we're actually after in the production of knowledge
00:50:52.240 generally, so whether it's philosophy or history or mathematics, science, any particular domain,
00:50:59.680 what we're after is hard to vary explanations. And the reason that we have been making progress
00:51:07.920 since the enlightenment through today's enlightenment is because of a culture of criticism,
00:51:14.000 the culture of criticism where we criticize our best theories that exist at the moment
00:51:19.040 and we improve them and that enables progress. We're only in chapter one and already there are
00:51:25.520 these phenomenal advances in philosophy, phenomenal advances in epistemology,
00:51:31.360 excellent explanations of what our best explanations of epistemology and philosophy are
00:51:39.040 all illustrated with some excellent science. That'll do me for now, that's quite a bit of reading
00:51:46.240 and maybe tomorrow I'll try and get into chapter two. We'll see how we go. See you.
