This interview with Daniel Dennett was conducted soon after the publication of his study of evolution, Darwin's Dangerous Idea [Simon & Schuster 1995].
HK: Why is a philosopher of mind interested in evolution?
DD: I felt I had to write this book on evolution because I had been relying on evolutionary reasoning in my work on the philosophy of mind for many years, and I kept running into what I considered to be ill-informed negative opinion about evolution from other philosophers and scientists. If I was going to continue using Darwinian reasoning I was going to have to show them it was not in as bad shape as they thought.
HK: take me through the notion of NeoDarwinism - algorithmic processes and so on.
DD: No different from orthodox Darwinism - it's the result of the synthesis of the Mendelian concept of the gene with the Darwinian theory of natural selection. Ever since Darwin wrote the Origin there have been sophistications of the theory, but it's still recognisably Darwinian.
The sense in which it's algorithmic is really very straightforward. As George Williams, the leading evolutionary biologist always stresses - a gene is a unit of information. What genetic transmission is all about is the transmission of information from one generation to the next and DNA may be the vehicle, but it's important to distinguish between the message and the medium. And what Darwin showed was that there had to be an informational process which gradually accumulated more and more design information and better and better design information, and by this process of design accumulation were built all the designs you see in nature today. He didn't have the concepts of algorithms or information processing but he was bold enough to theorise about it - he knew he didn't know about the physical substrate, he didn't know the medium, he didn't have the concept of a gene let alone the concept of DNA but he was able to think in a purely informational way about how it had to work. And he was right.
HK: So, a strong theory. Could you outline the two senses in which the word 'dangerous' comes to be applied to Darwin?
DD: The first and obvious sense in which Darwin's idea is dangerous is that there's just no denying that it attracts fans who don't understand it and misuse it in all sorts of bad ways. This has always been the case since the early social Darwinists - which means that those of us who are lovers and admirers of this theory have to protect it from some of its friends.
The more interesting reason why I call Darwin's idea dangerous is that it really completely reverses a very traditional way of thinking. Before Darwin we had a top-down or mind-first theory of design. Excellence came from on high. In the beginning there must have been a cogitative being, as John Locke said. Darwin turned that round and said all the mind, all the creativity in the world can itself be the effect not the cause - it can be created with mindless processes. Once one comes to grips with Darwin everything is still in place, it's just that the order of explanation has been inverted. A lot of people are very unhappy with that and they try to protect some part of the world, the part they are most interested in and say let's keep Darwin out of this and prevent his thinking from coming in. Misinformed and misguided thinkers have tried to rule Darwin out of their preserves and in fact there simply isn't any preserve where Darwin can be ruled out. It's not always of great use but it's always applicable.
HK: people see this inversion of cause and effect as a necessarily pessimistic thing. You seem very optimistic.
HK: What are the reasons for your optimism?
DD: For the first time we can explain some of the things which are simply left hanging in the air as mysteries in the old tradition. Any theory of learning, any theory of mind, any theory of intelligence has to break it down in the end into things which are not in themselves intelligent, which are mindless mechanical processes. Any other sort of explanation simply begs the question. As many people (Skinner, Darwin, AI people) have realised, there's no way to ground a psychology, a theory of intelligence that isn't ultimately grounded in biology.
HK: So it's positive from the point of view of the sheer power of its explanation?
DD: it's not just powerful, it's the only really powerful and general explanation.
HK: But can't you understand the fear it instils in some people - that you collapse the distinction between living and non-living for example. First, could you take me through the transition from non-life to life? And second, could you tell me why the collapse of the boundary should not be a cause for unease?
DD: Back before the mid-nineteenth century vitalism was a very popular doctrine. It said that living things had some special property which set them apart. That property was sometimes called élan vital. There was an unbridgeable gap between living and non-living. that doctrine began to unravel pretty badly with the birth of biochemistry. Once it was shown that you could synthesise urea and other bodily compounds in the lab the distinction between life-substance and non-living substance had diminished. Today I think it is uncontroversial. The processes of life get explained without residue in terms of the fundamental biological and physical processes which make them up. There's no need for any special life force - what that means, and this is a very Darwinian point - if you look at history there was a time when there wasn't anything alive and now there are definitely things alive then there's a long transitional period. It's a fundamental misunderstanding of Darwin if you think there's got to be some place where you draw a line and say "Aha! Here's the magic moment where life begins!" Life evolves out of non-life by a long series of gradual processes. There's no deep theoretical reason for marking any one point as the turning point. People are on the whole quite comfortable with that. But surprisingly they want to play the same card the vitalists play when it comes to mind or consciousness. The idea that mind or consciousness has got to arise gradually via an imperceptible process and there's no sharp threshold - to people who are worried, who think they have to have these boundaries of essentialism - then the Darwinian picture looks pretty threatening. If you consider those macromolecules which have some of the powers of the living - that is they are capable of self-replication and some degree of self-repair - the question as to whether they are alive or some kind of pseudo-organism is not an interesting question.
HK: You say that with these macromolecules we witness the birth of agency - hence meaning - hence consciousness. Is consciousness nothing more than the accretion of these processes?
DD: That's exactly what my theory of consciousness says. Consciousness Explained shows how consciousness can arise out of parts that are not themselves conscious. You can get consciousness by the way they are organised and the powers that accrue to them by their interaction. This is a model of how human consciousness, the most sophisticated and complex model of consciousness can be composed out of the same bits and pieces from which digestion and reproduction and self-repair are composed - namely little robotic scraps of tissue, or at a lower level, from molecules which work to repair themselves.
HK: You claim that one of the problems with the way the Darwinian revolution in thinking is happening is that it's taking place piecemeal. Why is the piecemeal change particularly dangerous?
DD: The Darwinian revolution is so huge that it should really be done all at once by everybody. It reminds me of the time in Sweden in September 1967 when overnight they changed from driving on the left to driving on the right. This was a potentially dangerous shift in public habits, but they planned and executed it very well, and very little harm came out of it. All the places you wanted to go were still there - nothing had changed and everything had changed. It would have been very dangerous to have listened to those who might have said let's do this gradually.
W; These metaphorical collisions are - of what? Of value systems?
DD: It's not so much collisions of the values themselves as of the ideology that justifies the values. If you think that the only thing that could justify an ethical principle was that it was handed down by God, well you might agree with someone else about the very same ethical principles, but you might be distressed that they didn't think of them as divine, but as human, as products of human rationality and intelligence. That would be a difference of ideology, not one of the values themselves.
HK: So we agree it's uncool to murder people. It's how you justify that decision.
DD: We all can agree about another area of value - we can all agree about the sublimity of Shakespeare's sonnets, and the art of Michelangelo and the music of Bach but if somebody thinks it is somehow destructive or demeaning of the beauty of this great art to have an algorithmic explanation of the processes going on in the artists' brains then that's not a disagreement of value, it's a disagreement of ideology.
HK: Which leads us into the question of ethics. You produce a Darwinian ethics. How does it work?
DD: The main difference between my Darwinian ethics and earlier versions is that I don't claim to deduce ethical views from Darwinism - I don't mean that neo-Darwinism implies an ethical theory - I claim something much more modest - that any ethical theory that is sound has to be consistent with evolution. That's a tough demand to meet because it means you can't just help yourself to Good Will or a vision of ethical truth and you can't say I don't care how it got into this universe because it's there and now I'm going to build my theory on top of it. If you're going to appeal to important ethical concepts like co-operation and selflessness you have to show how those phenomena could persist in our world. How they could be stabilised. That's a hard challenge to meet. That gives a different look to a moral theory.
HK: So we give up our foundationalism and the question of the ethicist is no longer what do I found my system on but rather how to make my system agree with what we know from Darwin?
DD: The problem is that no ethical system has ever achieved consensus. Ethical systems are completely unlike mathematics or science. This is a source of concern.
HK: Could we talk about our real world moral decisions? You say they take place 'online'.
DD: Here's a specific idea that is central to evolutionary thinking which has a direct application to the shape of moral theories. Just as Darwinian theorists have always had to confront the issue of how did all this work get done in the time available, the research and development and design improvement has to happen under time pressure. One of the fundamental insights of evolutionary theory is that the processes which work are not foolproof. They have to take risks. They cannot take all possibilities into account. That's true of evolution and it's true of ethical decision making. It doesn't mean that a process of deliberation - it's not that there are two choices - one: considers all things and is divine - two: doesn't consider all things and is evil - there are in fact processes which are good processes of discovery ,and ethical theory should consider that as a fundamental thing.
HK: So what we're doing is evolving our ethical systems towards stable and useful systems?
DD: And this is not a new idea. There have been many ethical theories which have looked at the derivation of an ethical theory out of a Darwinian predicament - John Rawls' theory of justice is basically a theory about how to justify a transition from one state where there isn't any justice to another state where there is - and it is basically an evolutionary process - before him Hobbes had such a theory.
HK: would it be fair to call it a heuristic ethics?
DD: Yes it would.
HK: How else do we use heuristic processes in our battle to survive? I'm thinking I suppose of generate-and-test.
DD: As many thinkers have noticed there's a fundamental pattern to all intelligent processes. They are at bottom what computer scientists would call generate and test. They make a lot of candidates via a relatively stupid process, you test those candidates through some sort of filter, some sort of evaluation process, you take the results and you do it again and again. That's obviously what the Darwinian algorithm is - it's generation and test at the genetic level. What Skinner proposed was that intelligence was the generation of behaviours which were then either reinforced or not. I show that in fact you can play that card many many times - there are generate and test processes everywhere we look. Heuristic search comes out of that.
HK: The 'deliberate foresightful generate and test known as science'. . .
DD: My use of science there is Germanic - all sorts of rational enquiry, all of human learning, not just academic science. What it does for the first time - of all the generate and test processes which have so far existed on this planet, this is the first one which is deliberate and foresightful - where it is done in order to winnow out the truth and is a much more powerful and swift process as a result, because it is guided by foresight - imperfect foresight of course.
HK: So science itself is a structure which changes as it itself is tested out.
DD: Absolutely - it follows that science, since it puts everything under the microscope, and engages in a continual redesign of itself, is constantly re-evaluating its own methods , its methods of evaluation and its own methods of evaluating its methods of evaluation and so forth. It evolves just the same as all the processes it studies.
HK: How close is the meme to the gene? And in what sense can we say that our ideas evolve?
DD: There are remarkable parallels - not to the concept of DNA, to the concept of the gene - information. The difference between genes and memes is that genes are for the most part dependent on a single medium - that is molecules of DNA. Even that's not strictly true - if you wrote down the codons, C P A and G on a computer, then even though a particular molecule might pass out of existence through the extinction of an animal, you might be able to splice it into another later, and it would be the same gene, the same information, the same recipe. Both genes and memes are informational units, but memes are much more promiscuous in the media in which they can be carried - in languages, in pictures, in practices, in concrete artefacts of various sorts - so it's much less likely that a strongly mathematical and predictive theory of memetics can be achieved - but it's not out of the question - if we look at the areas of human culture which most closely resemble in these features the genetic systems then memetics has a good chance - for instance it is often noted that human minds are not mindless slavish copiers of ideas but creative evaluating critical combiners of ideas. Human minds engage in a great deal of invention - they don't just take the ideas they get and take copies of them, that's true by and large, it's particularly true in science - but there are areas of human culture where that's not true, where within those subareas there is a rule - do not criticise, do not change, do not innovate, preserve exactly the traditional ideas - for instance in religions, in rituals. The formulae in most religious rituals are protected by a meme which declares do not criticise this meme, do not combine it with alien memes, do not corrupt the purity of this meme - thus in areas such as religion, memetics has a much better chance of yielding interesting results.
HK: Is this to do with the fact that memes are carried through language?
DD: The study of language and its evolution is a well developed area of what might be called memetics, since words are among the primary vehicles of ideas - we can study both the transmission and mutation of words and ideas and also their, as it were, phenotypic effects - we can see what happens to minds which get infested with particular memes.
HK: I'm interested in this question of memetic evolution - presumably you're not saying that our ideas optimise in any sense that is useful to humans?
DD: the fact of cultural evolution is just as uncontroversial as genetic evolution - everyone has known for a long time that civilisations evolve, cultures evolve, practices, agriculture - the only question is whether the theoretical perspective for examining this evolution is Darwinian, which is what Dawkins proposes with the concept of the meme. What you get when you look at cultural evolution from the memetic perspective is an appreciation that in the end the persistence of an item in our culture over time may not be dependent on our valuing it, because it may have its own independent way of encouraging its replication - the memes that are good from 'their' point of view are the ones that make lots of copies. They may not necessarily be good for 'us'.
HK: Such as a virus hoax on the net.
DD: Such as a computer virus or a virus warning on the net a chain letter. There are more than a few examples of memes which are of dubious or definitely negative value to the human minds they inhabit, but it doesn't keep them from replicating.
HK: You said to me once that on the whole those memes which have a negative effect on their carriers - you consider them probably less fit than others because they will tend to destroy their hosts.
DD: That's right - there's an imperfect but strong relationship between the fitness from a meme's point of view and its value from our point of view - by and large the memes we replicate are the ones we want to replicate because we think that they are good and we are right, however that is an imperfect relationship so there are still memes which do not replicate in spite of being good, and vice versa.
HK: When you say it is uncontroversial that civilisation and culture evolves, are we just talking about change, or are we talking about optimisation?
DD: In the sense that it's uncontroversial there is not just change but change in a descendent - that's - descent with modification. Optimisation is part of the second idea - Darwin had two ideas - one that species arrived with descent with modification, the other was that this process of descent with modification is governed by the process of natural selection. That cultures change in descent with modification is now uncontroversial.
HK: You have been quoted as appearing hostile to religious memes on the grounds that they impede a true and valid scientific understanding.
DD: That's not true - the virulence of the misrepresentation of some of my views on religion which I went to some pains to make clear in the last chapter of my book - did the Sunday Times publish my letter? I sent a letter regarding John Cornwell - I'm not hostile to religion at all. I am alarmed by fanaticisms which take their own vision of the law into their own hands and promote violence. I find that perhaps the most dangerous social force on the planet today. Aside from that I'm not hostile to religion at all, I think that religion has been a wonderful thing. Throughout the world we have already recognised that there are very clear limits to what we will consider civilised in a religion - I simply draw attention to that fact - we already recognise that religious freedom has its limits.
HK: You said something about how the technology of the infosphere has recently made it conceivable for everyone on the planet to know what you know - is the cure for fanaticism information or access to information?
DD: In the long run I certainly hope information is the cure for fanaticism, but I am afraid information is more the cause than the cure. My own unprofessional diagnosis is that when people living traditional lives confront the very alien image of the very high-tech, economically powerful lives that they see on television or through other media this is a deeply shocking and threatening discovery. In the face of those other lives which otherwise would not have been drawn particularly to their attention, they are more desperate for meaning in their own lives than they otherwise would be, and for people like that fanaticism is a trap.
HK: So the proliferation of information technology - in the short term it is promoting dissatisfaction?
DD: In some of course it's promoting disatisfaction, in others it's promoting revulsion at the complexities of the high-tech Western scientific life, and that revulsion leads people to turn to what they perceive as a strong and simple truth, a clarity which they can devote their lives to and all too often that truth is fanatical and destructive.
HK: Tell me about the question of ethical relationship towards artificial intelligences - at what point do I have an ethical relationship to a machine - is it a complexity threshold?
DD: It's possible in theory if difficult in practice to make a robot which is a sentient pursuer of its own projects, a self-protective and comprehending agent, and hence in ways that are very important a living thing - and a living thing moreover that has not just needs and desires but values that need to be respected. I think that's possible in principle, and as soon as one has created such an entity, whether by the old-fashioned way - procreation, or by modern technology, one has the responsibility to protect its rights and to treat it as more than just another artefact.
A short version of this interview appeared in Wired.