David Deutsch is the author of The Fabric of Reality and The Beginning of Infinity. He works on fundamental issues in physics, particularly the quantum theory of computation and information, and constructor theory. He is known as the father of quantum computing for his contributions to the field. He is an advocate of Taking Children Seriously, a new, non-paternalistic view of children.

Topics we discuss are well captured by the timestamps below.

Watch on YouTube or X. Listen on Apple Podcasts, Spotify, or any other podcast platform. Read the full transcript below. Follow me on X for updates on future episodes.

Click here to support my work.

Timestamps

0:22 - Happiness is a state of continually solving one’s problems

4:54 - Both free-will and the self exist

12:06 - The principle of optimism

17:28 - Any ultimate explanation is a bad explanation

20:22 - The origins of Taking Children Seriously

25:33 - Why children are the most coerced members in society

31:33 - Anarcho-capitalism

Full transcript

Arjun Khemani: I thought we could talk about some of the wider implications of taking our best understanding of epistemology seriously. Since it is universal for all kinds of knowledge and knowledge creation, Popper’s epistemology has some profound and far reaching implications. It is also the only kind of philosophy that I know of that is actually very practical.

In chapter 12 of The Beginning of Infinity, A Physicist’s History of Bad Philosophy, you write a bit about happiness. And you conjecture an explanation about the cause of human happiness. You say, “Happiness is a state of continually solving one’s problems. Unhappiness is caused by being chronically balked in one’s attempts to do that. And solving problems itself depends on knowing how. So, external factors aside, unhappiness is caused by not knowing how.” Can you please expand on that? It seems that this theory is a special case of the principle of optimism. Which you define as, “All evils are caused by insufficient knowledge”.

David Deutsch: In answering that specific question, I think you have to start with suffering, human suffering, because that’s really what all evils are. I mean, some people would disagree. Some people would say that the Earth can suffer. On the other hand, I think it’s not the case that reducing or abolishing suffering is the definition of morality. It’s just that they’re connected via epistemology.

Arjun Khemani: Yes, so I just wanted you to expand on your definition or view of happiness, which you say is a state of continually solving one’s problems.

David Deutsch: Yes. Well, I’d rather not define things. I’d rather say there is an issue, there is a problem about what we should aim for and what is right and wrong and so on. There’s a whole constellation of problems that arise from the fact that we are capable of creativity and of making creative decisions. And one way of putting that is that we seek happiness, but we don’t seek happiness in any—not always anyway—in any straightforward way, because, you know, somebody might decide to join in a war because they think it’s right. And they know that there is a risk that they will suffer as a result of this. Now you might say, yes, but they would suffer even more if they refused to go because they wouldn’t be thinking of themselves in the same way and as the same kind of person and so on. Yes, but that is approaching the problem backwards.

There is a reason why they want to think of themselves as in one way rather than another. And that way of thinking about themselves is rooted in their theory of morality. So, what they would be happy or unhappy doing depends on, in part, on their morality. It also depends in part on other things like their environment and their culture and their memes that they’ve inherited from other people and so on. And all those things ranging from what they have thought of themselves to what is hardwired in their genes or whatever, all that is mutable. So they can, in principle, change all those things, and if they’re changing them successfully and aren’t thwarted or balked in doing that, then I think we can call that happy—to a good approximation—even if they are in the more superficial sense, suffering as a result.

I should have begun by saying, like Popper, I’m not too keen on defining things. The question is always, “What problem are we faced with?” And as for defining things, “What concept comes up again and again when we’re thinking about particular kinds of problems and is it worth giving those a name?” And then we can call that happiness. But if you say, yes, but it doesn’t conform to the usual way of using the word happiness. I mean, I often use the word fun. Well, in that case, I’m happy to use someone else’s terminology so long as their terminology does not define away distinctions that I want to make in discussing problems.

Arjun Khemani: Some people think that it is the transcendence of the self gives rise to true happiness or to enlightenment. They say that the self is an illusion and free will does not exist. What do you think about this claim which some people think can be proven by experience?

David Deutsch: Well, nothing can be proven by experience, so that’s an easy one. I think those two things are different—the existence of the self and the existence of free will.

I think they both exist, and I suppose connected, but, there are differences between them. I think, for example, it’s true that when we are deeply engaged in a problem, an absorbing problem, then typically one does not think of oneself while conjecturing and criticizing in regard to that problem. Notoriously, people forget to eat when they’re immersed in the problem.

I was just reading a Sherlock Holmes story yesterday, I forget which one, in which Holmes remarks that he hasn’t eaten since yesterday or something. So that’s a notorious thing and I think it’s true.

Arjun Khemani: It’s a sign that you should be doing what you’re doing.

David Deutsch: Yes. Yes. I mean, it’s a sign. It’s not an infallible sign. Nothing is.

Now, that doesn’t mean that the self doesn’t exist, though. Any more than, you know, so sometimes we think of ourselves, qua self, and sometimes we think of the early universe. And when we’re thinking of ourself and not the early universe, that doesn’t mean that the early universe didn’t exist. It existed, we’re just not thinking about it. Similarly, the self exists and sometimes we don’t think about it. It’s very hard to define because it is connected with consciousness, which we don’t have a very good theory of.

Now, free will, it seems to me that a lot of the discussion about free will is really literally a discussion about nothing because people define free will as a type of thinking that violates the laws of physics. And then they say, but you can’t violate the laws of physics. Therefore, there is no such thing as free will. Well, there’s no such thing as free will thus defined—yes. But I think the way free will is used, both in everyday life and in philosophy, it isn’t like that, it’s got nothing to do with violating laws of physics.

We need that concept in issues like when person X kills person Y, did they do that of their own free will, or was it that a gust of wind pushed them towards the other person and then the other person fell into the path of the train, and therefore that wasn’t initiated by them. And then you can go into detail, like saying, well, if they have a hateful state of mind, and that was inculcated in them by their parents, are they really exercising their free will when they commit the crime or putative crime, or when they enact their parents’ crime, or however you want to phrase it.

Well, that’s a matter of fact. It’s a matter of fact whether the process that led from their parents inculcation to their pushing the person into the path of the train, whether that ever involved a process in their mind that is worth calling free will. So in that kind of situation, we can ask, well, you know, when we say what kind of process is worth calling free will, we really mean what kind of process is worth considering to have interpolated between the parent’s inculcation and the pushing action in such a way that it shifts the moral responsibility?

And I think the answer to that, or a big part of the answer to that, is whether the person in this thinking, during this thinking, has created something new in the world—new knowledge. But, you know, it might be false, but whether they have created something. Now, again, many people would say there is no such thing as creating something new in the world. Because everything you create, not only is it caused by your parents’ inculcation or whatever, but it’s caused by the laws of physics. It’s caused by the state of the universe as it was a hundred years ago, as it was billions of years ago, as it was at the Big Bang. So, the Big Bang is what caused this act of pushing. Now, I think that’s an incoherent view of the novelty in the universe. It is clearly not true that—I say clearly advisedly—that someone’s actions today are caused by the state of the universe yesterday because all the arguments that it is apply equally well to tomorrow. So you could say he pushed him into the path of the train because he’s going to be sitting in a jail cell tomorrow. And that has the same logic. The laws of physics are time reversible. So this kind of arguing away of the existence of novelty, and therefore of free will, is incoherent. It’s literally infinitely ambiguous.

And some of the ambiguous ways of deploying that argument are mutually contradictory. And common sense says that Einstein’s theory of relativity was not contained in the Big Bang. It was not contained in the thoughts of physicists at the beginning of the 19th century, but it was created in the mind of Einstein.

So humans create things, create novelty all the time in their minds. And therefore it’s meaningful to ask whether a particular idea was created by them or by somebody else or by nobody, whether just sheer chance. And, it can be a matter of degree. I mean, you can say the idea was partly created by Einstein and partly by Lorentz and Hilbert and so on. And it might be hard to tease out exactly what idea was created by Einstein, but you can tell it was because those people did not write those papers. So, yeah, I think both free will and the self exist, and the arguments against them are no good.

Arjun Khemani: In your most recent book, you explain that a momentous dichotomy exists, which is that any physical transformation is either impossible, because it is forbidden by the laws of physics, or it is achievable given the right knowledge. The principle of optimism is that all evils are caused by insufficient knowledge. Doesn’t this principle assume that the things that are forbidden by the laws of physics are not evil?

David Deutsch: Yes. So logically they could be. It could be that the universe is evil and certainly the universe is not as friendly to thinking beings as a lot of people think it is. But it’s not as evil as a lot of other people think it is. I mean, it’s not evil in that sense in that, well, I think it’s best to think of it as indifferent rather than evil. It doesn’t care whether we’re happy or miserable. There’s nothing in the universe that’s directed towards our being happy or miserable. And, you could imagine, as I say, logically, you could imagine a universe which was like that, where it was directed towards human affairs, just like religions sometimes often say that there are supernatural. Things in existence that care about whether humans are happy or not. And sometimes they are cast as wanting humans to be happy and sometimes vice versa and so on.

In regard to the sort of the modern scientific worldview, this would be very surprising. Because the facts of the universe are arranged according to laws of physics, which are highly universal. They apply to the planet Venus, and to formation of galaxies many billions of years ago, and to events in the universe that haven’t happened yet and so on. And nothing about them is specialized to our planet or to our species. So, it would be very odd if somebody found one day that the real laws governing the electron are the Dirac equation plus an extra term that we hadn’t noticed that said that, if Arjun is happy, thwart him in such and such a way. It wouldn’t fit in with the way we now know the universe runs. But you can’t rule it out logically, but there is no motivation for assuming such a thing.

Arjun Khemani: As you know, bad outcomes can have good intentions tied to them, or perhaps they can even have no intentions tied to them. I’m trying to think how, even if the universe or the proverbial asteroid or something is a problem to us humans, or to any human in particular, or to a person, then, like, is that not an evil because the universe is indifferent to us or…

David Deutsch: Yeah. So, if an asteroid turns up in such a way, I mean, I think by now it would have to be a bit of a minor planet. If it turns up in such a way that we haven’t noticed it and it’s going to wipe us out in too short a time for us to prepare for it, or to deflect it, then that was caused by a lack of knowledge in the sense that we could have generated that knowledge. And our slowness in generating that knowledge is itself due to a lack of knowledge.

We didn’t have that knowledge for a thousand years during the dark ages. And even before that, we only had it in certain subcultures and so on. And, we have existed as a species for two or 300,000 years. Most of which were wasted from the point of view of creating knowledge. So, you know, maybe in some universes we were wiped out, or presumably in some universes we were wiped out early on before we even in principle would have had the chance to create the knowledge.

Yes, I think that would better be called indifference than evil. Because the process that brought that about was an unlikely one. And it wasn’t tuned to causing human suffering, or death or whatever. And we see that the earth has been struck and there’s nothing in that event that could be called targeting. That is, there’s nothing in that asteroid or where it came from, right back to the beginning of the universe, right back to the big bang. There’s nothing in that process that knew about dinosaurs and wanted to wipe them out. So it was an accident.

Arjun Khemani: You’ve said before that if all knowledge is conjectural and subject to improvement, then protecting the means of improving knowledge is more important than any particular piece of knowledge. This goes against so many common assumptions about what knowledge is, what morality is, and how we know things. Some people think that we have to reason up from a few axioms or a foundation that we know to be unquestionably true. But you think that any ultimate explanation or foundation is a bad explanation. Why is that?

David Deutsch: Yeah, so a bad explanation is one that is easily varied. And if you have an explanation that cannot be questioned, then that means that the question of why the unquestioned thing should be that rather than a different unquestioned thing, which somebody else advocates as being unquestionably true is unanswerable.

And so that’s the very epitome of a bad explanation, because if, you know, if somebody believes a particular version of this, and a different person believes in a different version of it, and it comes from a different culture. Then they have nothing to offer in terms of a reason for why they’re adopting their view. And the structure of the explanation of morality or whatever it is, applies equally well to the other person.

Some people say, well, what if we have an explanation that doesn’t need any assumptions, which just is self evident and nobody has a rival theory. Well, if nobody has a rival theory, then, then, then we don’t need morality, like everybody will agree. But first of all, there is no such thing. And secondly, even then, even if everybody agreed, that still wouldn’t make it true. It would still be important to criticize that thing and try to discover a deeper meaning to why it was true. Why is it that everybody agrees? Is it perhaps because of a widespread irrationality inculcated in them, or is it because it’s really true? And because it’s really true—that possibility requires some critical thought about this unquestioned thing, even if true. I think J. S. Mills said something like this. No doubt he said it much better than I could, but something like, even if you’re absolutely right, unless you,… I can’t say it properly, but unless you know why your opponent is wrong, you haven’t really understood why you are right. And that requires criticism of both your idea and the opponent’s.

Arjun Khemani: After reading The Beginning of Infinity, I started quoting you everywhere and I was writing a blog post about parenting and I quoted you about something against authority. And I extended that argument to education and to children. And, at that point, I hadn’t heard of taking children seriously, but I just naturally seemed to overlap with those ideas. And our mutual friend, Brett Hall, read my piece and he told me that it may be no accident that I was quoting you, because you founded a non paternalistic movement with Sara Fitz-Claridge called Taking Children Seriously. Then I dug up the content online and, to me, it all immediately just made sense because It was all an application of Popperian epistemology to children and to education and parenting.

So I thought we could talk a little bit about Taking Children Seriously, starting with what was kind of your and Sarah’s motive for starting it as a movement?

David Deutsch: Well, I can’t really speak for Sarah. I think our motives were different. Mine came from Popper and I noticed, well, I suppose it came in some stages, but they happened quite fast, you know, one after the other. First of all, realizing as Popper did, but you know, I didn’t really get it at first, that his epistemology is very general. He applies it at great lengths to political philosophy and to science. He didn’t ever apply it to economics, for example, and he didn’t apply it to education. At least he didn’t write—he hardly ever wrote about education, but that was because he didn’t want to write about psychology because he didn’t want to give any ammunition to people who adopt subjectivist theories of truth and knowledge. But occasionally, you know, there’s a little gem here and there in Popper where he says the right thing, or more or less the right thing.

So having, having realized the generality of this epistemological theory and having been convinced that its opponents are wrong, I said to myself, well, if this is true, meaning this epistemology I was just reading, if this is true, then everything in existing educational theory is false. It simply destroys it and it destroys it with sort of philosophical firepower that’s overwhelming because the existing education theory is very parochial. It’s centered on alleged properties of humans and it’s mixed with theories of psychology and so on. And yes, certainly one does need theories of psychology in order to understand education. But if one violates epistemological principles, then one is wrong. The arguments for why one is wrong are just too powerful.

So then I started thinking about it and that was the beginning of my journey into educational theory, and I found that a lot of this realization was not at all new. There’s a long history of what you might call non-coercive educational theory. Rousseau, Godwin, and in the 20th century, Montessori.

Arjun Khemani: But you might have some quibbles with that, right?

David Deutsch: Well, I have quibbles with all of them. I think all of them have some things in common. And what they have in common is that they view educational theory as a branch, a small branch of—they would put it in different ways, but I would say of—epistemology. Or a philosophy or of liberalism. Even John Locke realized this, even though his actual educational theory, it was awful, but John Locke was a long time ago. And by the time it came to Godwin, it was pretty far advanced. Though, even he was awful in some places and Rousseau and so on. So I think putting it all in a, in a more, I don’t know what’s the right… I used the word powerful just now. It’s not quite the right word, but anyway, in a proper philosophical framework changed all that into TCS, and really it didn’t require any more than Popperian epistemology.

I don’t know what he would have thought of it. There’s this, you know, he was just this guy, his theories were the best we have so far in epistemology and related things, but he made many mistakes and maybe he would have made a mistake or two there as well. So anyway, that, that’s how I came to it via Popper.

Arjun Khemani: What do you think… is there an explanation for such widespread, I want to say coercive control over children in society at large? Do you have something that explains that?

David Deutsch: I think basically if you have a conception of knowledge, well, one can put this in several ways, but as Popper would put it: if you have the bucket theory of the mind, which is that knowledge is like a fluid that exists in one generation and can be poured into the next generation mechanically, then you will automatically have a coercive theory of education. You might be very kind about it. You might be harsh about it. You might be anything in between. You might have idiosyncratic quirks, but you will basically be wrong because that’s just as bad as, just as error prone as trying to—my other simile—it’s like trying to fly by jumping high. It’s just the wrong thing. It’s the wrong picture of what the problem is.

Because of this, the right picture of what the problem is, is highly counterintuitive. It’s like, whenever I criticize a particular feature of existing educational practice, like, say, exams, then the natural thing is to, is to say, well, how will the knowledge get into them? OK, some of them will want to learn that, but some of them won’t, and what about those? And this, some of them won’t, and some of them will, that’s all the wrong theory. It’s all the wrong picture. Like if somebody doesn’t want to learn something on the curriculum, it’s because they are learning something else.

They might be making a mistake or you might be. And the way to solve that kind of issue is to discuss it and to bring criticism to bear and to respect not only the principles of epistemology broadly, but specifically the principles of liberalism, which already tell us,… we’re lucky for hundreds of years now, we’ve had liberalism, which has already solved the apparently intractable problem of how you get people who have different views about things to live together without violence.

Now that’s not enough for an education theory, but it’s a thing an educational theory ought to respect and conform to. And if we still have the educational theory that we had in pre-liberalism days, then that’s a sign of something that needs correcting just in itself, but it might be that, that certain practices are justifiable under liberalism as well. It’s an issue that, that needs to be addressed.

The reason, now, this is a speculation. I think what I just said is fairly straightforward, but the reason why it’s especially intractable in our society is that it is connected with the transmission of memes. So education as conceived at present is entirely the transmission of memes. Now without transmission of memes, we would be reinventing the wheel all the time. So it’s highly desirable that memes be transmitted, but it’s also highly desirable that they not be transmitted faithfully. Because otherwise there won’t be improvement. And if there’s anything worse than having to reinvent the wheel, it’s having no improvement because whereas reinventing the wheel all the time is terribly inefficient, not having improvement is fatal.

It’s certainly fatal. So we need to solve this problem and liberalism has solved it in the political sphere. Capitalism goes a long way to solving it in the economic sphere. The canons of rationality in science, especially as improved by Popper, have more or less solved this in the scientific sphere. But in education, because educational theory is conjectured in the context of the wrong problem situation, it’s especially resistant. This is all speculation.

Arjun Khemani: I think with parenting and with most other things that are worthwhile, it just requires creativity. And I see a lot of people with this zero-sum mindset that if, you know, if the child gets what they want, then the parents can’t get what they want. Or there’s a compromise and nobody gets what they wanted. But there can be a solution and it just requires creativity.

David Deutsch: Yes. And it is interesting that in regard to liberalism and capitalism and so on, a few hundred years ago, the analogous arguments were being made to say that liberalism is impossible, you know, democracy is impossible. Because, you know, what might the people vote for? What if they voted for all institutions to be overturned? What if the majority voted for the minority to be dispossessed? Well that can happen. But under the institutions that have evolved that doesn’t happen. That’s not the problem. Democracy has plenty of problems, but that’s not one of them.

Arjun Khemani: This ties to the next thing, the next topic I want to talk to you about, which is anarcho-capitalism. From my understanding of the ideas of Popper and yourself, a system of anarcho-capitalism really does seem like an ideal system for maximizing knowledge growth and progress in society.

What are your thoughts on anarcho-capitalism? Do you think it is desirable or do we need certain institutions like the state to maintain stability? Feel free to talk as much as you like about it.

David Deutsch: Well, I don’t think anarcho-capitalism is a system. It’s a state or a condition that a system might promote, it might have that property. Just like, you know, before banks were invented, you might say, well, you could set up a capitalist society, but, all sorts of things that the king used to do aren’t really possible because somebody who doesn’t have the money to start up a business to do the thing cant start it up and so it won’t be started up. Whereas the king could just order it to be started up and so on. I don’t know if that’s a real example, but institutions have non-trivial knowledge in them. And there was a time when people invented what we now think of as a bank. And there was a time when people invented what we now think of as money.

And the institutions that we have involving limited government and liberalism and elections and so on, all these details came about as solutions to problems that came up because they didn’t have them before. So freedom of speech and so on. It wasn’t instituted all in one go. It was instituted in bits and each bit was thought to solve a problem. And sometimes there were false steps. Where people thought something would solve a problem, and then it didn’t, and so on. And what we call modernity, or liberal democracy, and so on, the features of it, like the banks and so on, they are features that were installed to solve problems. And if you went back to prehistoric times and try to found a bank, you couldn’t because the problem that a bank solves hadn’t arisen. You can’t have an accounting system among people who haven’t yet invented numbers or voting for that matter. If you can’t count the votes, I suppose you could tally them, but even tallying was invented.

So anarcho-capitalism is a state where no monopoly of violence is considered legitimate. But that can only exist when the problems of not having a monopoly of violence have been solved somehow by some institutions. And so people can be in favor of those institutions rather than in that you can’t be in favor of just the outcome without, without specifying the institutions that would cause it. We have institutions that make it unthinkable that, mass violence would erupt over some political issue, at least, you know, in, in many countries, that’s the case. In other countries, it’s not the case and, and nobody can put their finger on exactly what it is about existing institutions that have that property. If you abolished existing institutions, then there would be violent conflicts among the force users, and I think in real life, what would happen is that they would want to revert to the present system. They would, decide that the present system is better than the warfare they’re having.

Anarcho-capitalists, I’ve heard them say that, no, they, they would sit down and they would come to an agreement about how to resolve such disagreements about violence. One company and their customers think that there should be copyright laws enforced violently if necessary, and then the others think there shouldn’t be, and that should be enforced violently, and then they should sit down with each other and determine who would win a war if they fought one, and then consent. The losers would consent to losing the argument, without ever having to fight the war. Now, there are no such institutions. Nobody has ever done this. Nobody has ever opted, who was about to have a war, has ever opted to not have it because they’re going to lose. The issue of whether they’re going to lose was considered long ago when they, when they declared war, in fact, long before that too.

So I do think that anarcho-capitalism points to some real problems with existing methods of, existing institutions of consent. They sometimes violate consent unnecessarily and greatly. It is not obvious what to do instead, and it can’t be done by fiat and it can’t be done instantly. What I would advocate is instead of trying to create an overriding, overarching system, which would automatically solve all those problems I would rather address the problems and let the overarching system evolve from addressing the problems.

So, there was a time when, in Britain, it was a consensus theory adopted by the majority that the government should seize control of the commanding heights of the economy, as they put it. And so that was done. And it failed in its own terms. And now nobody wants to seize control of the commanding heights of the economy. So fortunately this happened in Britain. Because it happened in Britain, it didn’t interfere with the memes controlling the use of violence. So the government was able to institute all these stupid things without facing significant violence in opposition. And then when it was undone later there was, again, not significant violence opposing undoing it. This is unusual. Usually when a state based economic system is imposed in a political culture, it’s done violently and it encounters violent opposition and then, well, either it fails or it succeeds. If it succeeds, it succeeds by wiping out the opposition militarily. Obviously, that’s an inefficient way of running things and it’s also a way that is antagonistic to the growth of knowledge. So I wouldn’t want to do that for anarcho-capitalism. I would rather see something which with the benefit of hindsight, we will say, ah, yes, that was what we have now. This system is definitely better than what we had before. And important features of it were foreseen by David Friedman, but there are these other features of it, which are the things that make it work, which weren’t always foreseen by David Friedman, except in very general terms.

Arjun Khemani: Right. So we need to take it incrementally and step by step, solving problems, relevant problems along the way, and kind of foreseeing it as a system, like playing out, I guess you could foresee some of the problems and, you know, propose solutions to them but I guess like the point about the growth of knowledge being unpredictable does apply here. That you just cannot foresee… You can’t even foresee what future humans will want or what future you will want. So, yeah, I think that’s an interesting point to keep in mind.

Well, David, I could ask you questions all day, but I think that’s a good place to end our conversation for now. Thank you so much for joining me.

David Deutsch: It’s been fun as always.