Elliot Temple: This is an interview with the physicist and philosopher David Deutsch about his new book The Beginning of Infinity. You can read its introduction here. David, can you briefly describe what the book is about?
David Deutsch: The "beginning of infinity" in the title refers to several things but especially to human progress. The book discusses the rapid progress we've seen since the Enlightenment, and its cause: the rational quest for good explanations. It unifies many themes of reason and unbounded progress.
Elliot: I don't want to die; I'd much prefer to be a part of your vision of infinite progress. Do you think immortality research is the most important thing that the world should be focused on?
David: I can see the argument that it should be, which is basically a sort of betting argument: death has a utility of minus infinity, so you should subordinate all other values to the value of increasing the probability of staying alive, even by the tiniest amount. But I don't think that's actually a valid argument. I do think that too little research is being done into technologies that might lead to immortality, and part of the reason for that deficiency is irrational: it's the deathism that is currently widespread in intellectual life — but that doesn't mean I think the right amount is 99% or even 10% of the whole human economy, because science is the art of the possible, as Peter Medawar said, and we have to pursue things that can be pursued. We don't even know that immortality technology when it is eventually discovered will be reached by the paths that we would today put vast amounts of money into, if we did think it merited them. For example, say we were having this conversation in the year 1900. We might divert all research efforts into medical research, and it might be that that would be a mistake, because the real immortality is going to come from computer science, or nanotechnology, which hadn't even been thought of in 1900. So it might be that even the field that's going to solve this problem doesn't exist yet. Or that it does exist but we don't know the connection to immortality yet. And so on.
Elliot: That makes a lot of sense. One thing about Aubrey de Grey's research is: each of the things he wants to do is preventing something that currently kills people, so even if it doesn't lead to immortality, it will be very useful. However, it could still be that, while this research is useful, some other field ends up solving the immortality problem instead.
David: Yes. That is the right approach because it's piecemeal and each intermediate step has value in its own right.
Elliot: The Beginning of Infinity discusses a lot of philosophy, including a section on bad philosophy. A widespread philosophical mistake is foundationalism. What are your comments on it?
David: Foundationalism is the false idea that the most important thing in any system of knowledge is the foundation from which it is either derived or justified, or both. But in reality there are never any absolute foundations, and the value of a theory is in its ability to solve problems. Foundations are put there for a purpose: to solve some kind of problem. For instance, foundational theories in physics are foundational because they're universal. They're universal because the explanations that those theories contain achieve their hard to vary status by being universal. So the reason that there are such theories in physics is they are functional in solving the problems that physics tries to solve. But people who try to look beneath those foundations, without having a problem, basically because they're looking for an ultimate foundation, aren't going to find one. You can only improve knowledge in the context of having a problem. You can't just wish to know something deeper than you already know, just for the sake of it. And you never reach the absolute foundation, or anywhere near an absolute foundation. We're always at the beginning of infinity.
David: That's the negative reason why foundationalism is really bad. The positive reason is that actual theories, which actually solve problems, are mostly not at the foundational level. This is why we didn't have to institute a crash program of redesigning all bridges when it was discovered Newton's laws were false. It really wasn't important that the foundation of the theory of bridges was false. It doesn't rest on foundations in that sense.
Elliot: What do you think of the compatibility, or not, of your ideas with the world's religions?
David: At the foundations, they are extremely incompatible. But, for the reasons that we've been discussing, that really doesn't matter all that much, and I think that for almost all practical purposes, these ideas are compatible with any major religion — with one proviso, which is quite important. Some interpretations of religions explicitly stigmatize reason — criticism and error correction and so on. And those interpretations play the same role as — in the Socrates dialogue in the book — what Socrates there calls the ultimate immoral thing, which is to destroy the means of correcting errors. So those interpretations of religions are also incompatible with the ideas of the book at the practical everyday level. But all major religions also have interpretations that aren't like that.
Elliot: There's also religious interpretations that advocate violence. Can you comment on the connections between reason or unreason and violence?
David: Reason is a property of how we manage conflicting ideas, including conflicting projects, conflicting values, and so on. Historically, most ways of dealing with this have involved violence or threats of violence. A tradition of criticism always involves, among other things, abstaining from violence in the relevant area. Ideally we would like a tradition of criticism about everything, but in real life there are specific traditions, in science, in politics, and so on, which some societies or cultures have and some don't. Those rational traditions which make our society work abhor violence, and that's the connection. Reason and violence are the antithesis of each other. Popper gives — I think it's in his autobiography 'Unended Quest' — a nice practical example from his own life, which I think he says was the thing that first made him think about the directness of this connection. When Popper was arguing with a Nazi, he said something like "what is your argument?", and the man took out his gun and said, "This is my argument."
Elliot: Are there any connections between reason and freedom?
David: A tradition of criticism entails tolerance of dissent, which in turn entails, as Bronowski said, certain values. So freedom in the intellectual sense — which means you are free to dissent, to have ideas people disagree with and dislike — is directly entailed by reason. And along with that goes values like respecting dissenting thinkers and appreciating their existence even if only with gritted teeth.
David: Freedom in the political sense is a property of political systems that obey Popper's criterion that it should be as easy as possible to remove bad rulers and bad policies without violence. And the antithesis of that, the way that society would violate that — either by having violence in its institutions or by making it difficult to remove bad rulers and bad policies — would involve impairing freedom. Either to say things, which is directly criticism, or to do things, which is to do with testing ideas. You can maintain the status quo either by preventing people from suggesting certain things or by preventing people from acting on the suggestions, both of which are matters of freedom. So suppressing freedom in both ways thwarts rational processes.
Elliot: I understand that you met Karl Popper and there was an interesting discussion about unsolved intellectual problems. Would you describe that?
David: I witnessed the following exchange between two titans: DeWitt asked Popper what Popper thought was the most important unsolved problem at the foundations of physics, and Popper said "Why are all electrons the same mass?" DeWitt said, "That's interesting; this might reflect the difference between a philosopher and a theoretical physicist. Because to me, it isn't a mystery why all electrons are the same mass. If not all electrons were the same mass, the difference between their masses would be a new field in addition to the electron field, and I would expect quantum theory to describe that field and that would be a sign of a new, unknown law of physics. That would be a mystery that we would have to solve. If there isn't such a field, and they're all the same mass, then there isn't a mystery."
Elliot: I'm curious about moral or epistemological unsolved problems. Do you think there's big unsolved problems in pure epistemology?
David: This problem may be all part of the consciousness problem, or it may be metaphysics or epistemology or it may be something else: I'm not entirely satisfied that it's okay to use a Star Trek Transporter (which disintegrates you, sends information about you to another place, and rebuilds you there). I am satisfied that the best existing theories say that it is all right. But, it does contradict some common sense, and those common sense theories reach into areas like consciousness and so on that we don't have good theories about. So the mere fact that our best theories in our best-understood fields say it's OK isn't enough to be able to conclude that this is unproblematic. It seems very problematic to me.
Elliot: This is a good example of an argument which doesn't fit foundationalism. You're starting with less derivative ideas, and using them to criticize more foundational ideas. Some people would see that as illegitimate. They would say if the more fundamental theories don't have a problem, just because common sense says there is one doesn't mean anything.
David: Yeah, well, we don't judge theories by their source. Common sense, like everything else, has to be criticized in order to be dropped, and the criticisms have to work, and there has to be a better theory, and so on.
Elliot: That's a nice way of putting it, that whether it's a foundation or not is an issue of the source, which Popperians disregard.
David: Yeah. It's funny, my professional work is all about foundational issues, both in physics and in all the other areas it reaches into. Foundations are what I study. And yet, I am vehemently opposed to what most people think is the main use of foundational theories: to justify subsidiary theories. The right use of foundational theories is as a source of criticism. If you have foundational theories, the foundations of one thing are going to be the foundations of many other things, and that links those other things. And that means that there are more constraints on ideas in all those areas because they have to satisfy the criticism that comes through to them via the foundational theories.
Elliot: You expressed respect for common sense with the transporter issue. But contrary to common sense, The Beginning of Infinity says aesthetics is objective. Please explain your stance.
David: To have an objective theory of something means that you have a theory of it that does not classify the distinction between true and false in that domain as being a matter of arbitrary cultural convention, or individual whim, or any other kind of arbitrary or subjective choice. When there's an objective truth of the matter, that means different people pursuing the objective truth, from different angles, independently of each other, starting with different problem situations and different initial theories, can expect to converge with each other. That happens because they are converging on an objective truth.
David: Some people disagree with this. They might call it "inter-subjective". But their objections are generic and apply to physics just as much as to aesthetics. The crucial thing is that there be real explanations linking ideas, not just mere aspirations for what explanations we'd like to have. We need knowledge not mere hope.
David: Aesthetics being objective leaves the question: what is it? Is it just a theorem of physics, or is it prior to physics, or is it an emergent property of physics, or what? And I think the answer to that isn't known, and isn't all that important at the moment, like many foundations are unimportant.We do ultimately want a unified worldview. Having one will help to criticize different aspects of that worldview and thereby to improve it. That's a core function of foundational studies, as I said. But it's by no means sufficient, and also it's not urgent to do that because, as Popper says, the starting point doesn't matter. It will eventually be helpful.
Elliot: Many people deny that morality is objective. What do you think?
David: In the book I stress that although you can't derive an ought from an is, nevertheless factual truths and moral truths are intimately related to each other. So intimately that you can detect people's moral views in their factual views. Just like you can detect ancient life in a fossil, which is chemically purely a non-living stone. In both cases we learn by explanation. Factual and moral explanations are not independent, even though you can't derive one from another. The deriving was unimportant in the first place because that's a relic of empiricism.
Elliot: The book discusses political systems and anti-rational memes. I've noticed a lot of political debate is unreasonable and was wondering: how could it be done differently in order to be more friendly to arguments using rational rather than anti-rational memes, so that they can discover more moral truths?
David: This is a complicated issue. Our political traditions are more rational than typical people. So the collective behaviour of people in our society is much better at creating knowledge than would be the case if you just asked those people how things should be done and then did that. What makes politics work is broader than the debates.
Elliot: What does it mean for traditions to be wiser than people?
David: There's a bit about this in the Socrates dialogue in the book. You can have traditions in a culture which have the effect that if it is seriously proposed to do something which would destroy the means of correcting errors, then the various incentives that people both inside and outside the system have, add up to produce intense, truth-seeking criticism, followed by howls of protest. Or they make the process run into the sand and never get anywhere. Or they even make it get enacted in some form, but still never go anywhere.
Elliot: You wrote a broad book, so here's a broad question: what is the human condition?
David: In one sense, the idea that there is a human condition or a human nature is the epitome of the wrong way of explaining humans. But there is another completely harmless sense of the term 'human condition', namely the condition of being universal explainers who are in a universe that is amenable to explanation. And that we are in error all the time; we're fallible. But these things — all senses of 'the human condition' in which it is true that there is such a thing — do not let you predict anything about humans. They are explanatory.
Elliot: And they apply to persons in general, not to humans as a species.
David: Yes. And the fact that we are all alike in our infinite ignorance doesn't say anything specific about any particular theory being true or false. Whereas the idea that we're conditioned by our life on the Savannah, which is supposed to justify particular theories about what humans will do, like whether they will be faithful or unfaithful to their spouses, is the height of silliness.
Elliot: Some people think it requires high innate intelligence to be able to understand advanced ideas like yours. Is that correct?
David: The limiting factor in anyone understanding anything is how interested they are. There are various qualifications to be made to that, such as that there shouldn't be anti-rational memes at work, but interest is the basic factor.
David: As to innate intelligence: I don't think that can possibly exist because of the universality of computation. Basically, intelligence or any kind of measure of quality of thinking is a measure of quality of software, not hardware. People might say, "Well, what hardware you have might affect how well your software can address problems." But because of universality, that isn't so: we know that hardware can at most affect the speed of computation. The thing that people call intelligence in everyday life — like the ability of some people like Einstein or Feynman to see their way through to a solution to a problem while other people can't — simply doesn't take the form that the person you regard as 'unintelligent' would take a year to do something that Einstein could do in a week; it's not a matter of speed. What we really mean is the person can't understand at all what Einstein can understand. And that cannot be a matter of (inborn) hardware, it is a matter of (learned) software.
Elliot: Suppose someone reads your books and they're inspired. They want to help with human progress. What's most important for them to work on? Or should they just follow their own interests, even if it's just playing the piano?
David: I don't think that playing the piano deserves the qualifier 'even'. Human knowledge is human knowledge, and it is all connected, and it's all improved by the same kind of process. And it's all impeded by the same kind of impediments. If somebody really liked either of my books, they're bound to have liked some chapters more than others. They should pursue things they find inspiring.
Elliot: If someone wanted to know as much stuff as you know, how should they go about it?
David: I don't think it's a good idea to have a theory about how much one wants to know. How can one possibly know in advance how much that will be? And suppose one did know how much, how could there possibly be a knowable path from here to there? It's much better to just pursue things one is interested in and curious about. But one piece of advice I could give (to 'know what I know', for some reason) is: learn quantum mechanics.
Elliot: Why quantum mechanics in particular?
David: In some ways quantum mechanics is to theoretical science as Popperian epistemology is to philosophy as a whole. Not only does it have a similarly central place and gets into everything, quantum mechanics is actually one of the simplest areas of physics to learn, the way I think of it. It's the place in science where the leading edge of thinking is the closest to what a layman already knows. However, unfortunately, I don't have any particular book on the topic to recommend -- until I write my quantum physics textbook.
Elliot: Speaking of quantum mechanics, what was your main point in the multiverse chapter?
David: That parallel universes are really the least remarkable thing about quantum physics, because universes are precisely the feature that is classical about quantum theory. It's all the other features that are remarkable. And all the other remarkable features are linked by fungibility. You can see them all as consequences of fungibility.
Elliot: What ideas were left out of The Beginning of Infinity and why?
David: I had been planning a chapter on scientism, to discuss all the different ways in which the perceived authority of science and the methods of science — both actual and imagined — are being misused both in science itself and purportedly to resolve issues that are in fact philosophical. Part of that found its way into the chapter A Physicist's History of Bad Philosophy (hence the subtitle "With Some Comments on Bad Science"). That is the ghost of what was to have been another chapter. There isn't any single reason it was left out. It was just too hard to write and to integrate with the rest of the book.
Elliot: Can you describe some of the history of how you learned the ideas in the book. Which ones did you invent yourself, and what are some people you learned ideas from?
David: I'll comment on some particular chapters. The Reach of Explanations: That's, as far as I know, my original idea. The idea that knowledge consists of explanations is of course due to Popper, though I think he under-stressed it and over-stressed his criterion of demarcation between science and philosophy. Between science and metaphysics, as he called it.
Elliot: In The Fabric of Reality, and also in The Beginning of Infinity, you talk about how ideas need to be good explanations in the first place before testing them matters, which Popper didn't emphasize, if he knew it.
David: Yes. I'm pretty sure he did know it, but for various reasons he didn't emphasize it. I think partly because he was sort of dazzled by his own criterion of demarcation, which was so powerful in criticizing Freud and Marx and all sorts of irrationality. He wanted that, and the time when he was successful at doing that was his early more formalistic period. I think that Popper later was fully aware of the centrality of explanation, but somehow never brought it out sufficiently.
Closer to Reality: The small idea there which I think is original to me is that scientific instruments are like conjuring tricks in reverse: that they fool us into seeing what's really there. It's slightly paradoxical that the more accurate they are, the more layers of interpretation they put between ourselves and reality. That sounds crazy from the point of view of common sense and empiricism, but if you think of it the right way it's obvious it has to be like that. What else could correct errors, apart from good theories?
Creation: I'm strongly influenced by Richard Dawkins.
Optimism: Putting it that way is original, I think. In my own thinking, it came from constructor theory. But I didn't put it that way because as a general rule one shouldn't try to persuade people of things via a route that is itself something that one has to persuade them of. So, although constructor theory looms large in my mind as the source of the optimism idea, that's not how I presented it.
The Multiverse: That's Everett and I've been one of the people who's tried to push Everett's theory forward, so in that chapter I present my take on Everett's quantum mechanics. Which is the quantum mechanics. I think the fungibility aspect is original.
A Physicist's History of Bad Philosophy: I think a lot of people think exactly this, but it's considered impolite to say that most philosophy recently has been rubbish. I don't think that's at all original to me. All I've done here, I guess, is put it into the broader context of rational and anti-rational memes, Popperian epistemology, and the reception of Everett's theory in physics.
Why are Flowers Beautiful?: I think that's an original idea of mine.
The Evolution of Culture: This builds on meme theory which was founded by Richard Dawkins, and was pursued by many people, but the idea of distinction between rational and anti-rational memes and the sort of natural history of memes that I present in that chapter is original.
The Evolution of Creativity: Similarly, the idea that humans evolved to have memes is not original. I think it's due to Susan Blackmore. But I think the particular mechanism that I suggest, with its surprising and paradoxical driving force, is original.
Elliot: Any comments on what your next book might be about, or what you're currently researching?
David: I don't know whether this will ever come to anything, but I'm considering a book called "Irrationality". My first two books are about reason and its consequences and how it's built into the physical world, and I only mention some ways in which things can go wrong. It would be nice to have a compendium of all the important ways in which reason does go wrong, and how they are all linked to each other. But on the other hand, I don't like the idea of writing a book against something. I like to be for something. So I may not write it.
David: I'm currently working on several issues at the foundations of quantum information theory, and intensively working on constructor theory. I gave a talk at the Clarendon Laboratory a few weeks ago to set out what I'm hoping constructor theory will be like.
Elliot: Final question: do you think Popper actually, genuinely solved the problem of induction and justificationism?
David: Yes, Popper solved the problem of induction. Why is that still a question?