California Wild: Dr. Dawkins, when you talk about the "gene," what do you mean exactly by the gene?
Richard Dawkins: Oh, I thought you were going to say, "What do you mean by 'selfish'?" The gene is that which lasts for an appreciably long time in coded form. By long time I mean over many generations. So molecular biologists define genes in various alternative ways. For my purposes, the important point about the selfish gene is that it is coded information that is very, very long-lived compared to the bodies which it temporarily inhabits.
CW: You say it is "that which" persists. Is that how you define the gene? Something does persist but how do you define just what that "that" is?
RD: There are two possible confusions here. One is that the physical atoms of the DNA molecule are constantly turning over so there is no persistence there. The only thing that is persistent is the coded information. It's like reprinting a book over and over.
The other possible source of confusion is the exact size of the portion of DNA we are talking about. Molecular biologists conveniently talk about a cistron, meaning approximately that which codes for one protein chain. And that's fine. I mean that'll do for my purposes, but so also will bigger bits of DNA because the cutting of chromosomes in crossing-over at meiosis [the process whereby a cell nucleus divides and separates its genetic contents] means that large chunks of chromosomes are short-lived. Very small chunks of chromosomes are long-lived. The longer lived they are, in terms of generations, the more they qualify to play this almost mythic role of "selfish gene"-- that which persists.
CW: Are there occasions when something which is in the interest of the gene is not in the interest of the species?
RD: Oh, very frequently. We're talking about interests. That means when you look at adaptations, when you look at things like eyes and hair, you ask what they're good for. If they were for the good of the species they would look very different from how they do look, which is for the good of the gene.
CW: In what way is it in the interest of the gene for an eye to be blue instead of brown? And if there is a good explanation, then why aren't all eyes blue?
RD: Well, first of all in what way is it in the interest of the gene to have eyes at all? Clearly having eyes helps the body to survive and that means the body can preserve the DNA and pass it on in reproduction. Now you ask a much more restricted question, blue versus brown. There is very little difference there and we're talking now about something very small and statistical. Blue eyes are slightly more prone to be dazzled than dark eyes, and in general that isn't going to matter. But one of the things that Darwinians have learned by doing theoretical calculations is that very, very tiny margins in safety can be very important in natural selection. It's only necessary to say that some individuals have died as a direct result of being dazzled. Which is quite plausible. The crucial moment the saber-toothed tiger is creeping up, you're dazzled and don't have enough time to get up a tree or whatever one does. It's counterintuitive, but very slight risks, by ordinary human judgment, can accumulate when averaged over many generations.
CW: So why do blue eyes persist at all?
RD: In cases where we have polymorphisms [various forms of the same trait], like blue eyes and brown eyes, it could be that in different parts of the world different color eyes convey different advantages. I hasten to say that I don't know what the answer is in the case of eye color, but in the case of blood groups which is another well known polymorphism, it appears that different blood groups confer relative resistance to different diseases.
CW: Do you find high concentrations of one blood group in a geographical region?
RD: They certainly vary, yes. They certainly vary in different races.
CW: Is that why the predominant color of the peppered moth in England changed from white to black after the onset of the Industrial Revolution and then back again after smokeless fuels became common?
RD: Well, the colors were continually recurring by mutation, so when there was no industrial pollution the black moth would have been constantly cropping up by mutation but swiftly removed by predation. Only when the environment changed so that blackness became an advantage did selection start to favor the hitherto rare black moth.
CW: That mutation presumably would have been catalyzed by cosmic rays.
RD: Many mutations are, yes.
CW: Are species any better today than the first microorganisms were at reducing the mutation rate from cosmic rays?
RD: Yes, we are. Mutation rates have been cut out to a fantastic extent. They're very low now. Fortunately, they're not totally zero because then evolution would come to a halt. But the error-correcting mechanisms that have evolved are exceedingly efficient and they compare very favorably with the best manmade error-correcting mechanism.
CW: What about the persistence of such traits as homosexuality and traits that would seem to be counterproductive for reproduction?
RD: Yes. Well this is only a problem to the extent that homosexuality is genetically influenced. It probably is. The homosexual phenotype [a physical trait determined by genetic instructions and environmental influences] seems to exist in males with sufficient frequency that it is something to give a Darwinian a bit of a worry. There have been various theories trying to explain why in our wild ancestors it might have been favored in a minority. One suggestion is the 'worker ant hypothesis.' Homosexual males would have looked after their nephews and nieces and perhaps freed the father to go off and do other things. Another theory is the 'sneaky male theory,' which is that homosexuals are frequently bisexual, and so when you live in a species that has a dominant male harem system then the best way for a subordinate male to get females might be to pretend not to be interested in them and win the trust of dominant males.
Actually, I'm not that impressed with either of these theories. The one that I prefer is an example of a general, rather important lesson, which is that when we talk about genes for X there isn't necessarily any irrevocable linkage between genes and X. There are lots of genes where the phenotypic effect depends very heavily upon the conditions.
For example, there might be a gene (this is purely hypothetical) that makes you homosexual if you're bottle-fed but not if you're breast-fed. That's not implausible. So before bottle-feeding was invented the phenotype of that gene might have been something quite different. And so there wouldn't have been any selection against it. As soon as you change the environment (and I'm just using bottle-feeding as an illustration) then a new phenotype arises. So we're asking the wrong question if we ask, What is the Darwinian survival value of homosexual behavior? We should be asking, What was that gene doing at a time when natural selection really counted?
CW: And that kind of thing could be extremely difficult to sort out.
RD: Yes, it could. Genes don't have their phenotypic effect written on them in characters of stone. They just emerge as a consequence of complicated interactions not only with the environment but with other genes.
CW: In what ways has the computer made it easier for us to grasp evolution's mechanisms and implications?
RD: Computers are extremely complicated, very versatile machines that can be reprogrammed to, in effect, be anything that you can specify with sufficient precision. So you can reprogram a computer to be a chess player, or the City of Vancouver, or almost anything. And similarly, if you've got an idea, an hypothesis, of what's going on in any system, you can program the computer to behave like that.
Evolution is no exception, and so computers have been used to simulate mathematical models of what's going on in populations. I've used computers to perform a more didactic role of simulating how you can make forms change in strange and sometimes beautiful ways by artificial selection of randomly varying, genetically controlled forms--my so-called "biomorphs," which I described in The Blind Watchmaker. The whole science of artificial life is largely devoted to doing similar things and not necessarily simulating the way we think life on this planet actually is, but inventing new kinds of quasi-life which have some of the properties of life as we know it, but not all of them. So it's an aid to the imagination. It's an aid to the teacher. It may be an aid to working out, at a more mundane level, the consequences of various models of what would actually go on in life as we know it.
CW: You do seem to use a lot of very convenient analogies between computers and genes. For instance, if you delete a file from a disk, the reference disappears, but all the information is still there. Without computers these concepts might be harder to visualize.
RD: Yes, I agree with you. There might be something similar in paper documents but it's not obvious that there is. The computer disk is absolutely spot on. That's precisely what happens. But, I suppose the same has been true of technology for a long time. Computers offer a new range of metaphors. But no doubt other machines did before them.
CW: Recently, a computer scientist, tracing the history of analogies of the human mind and various other machines, such as subway systems, or sewer systems, or calculators, said that the computer metaphor is different. He said the mind isn't like a computer, the mind is a computer.
RD: Well, that's misleading if it leads people to think that it works like a computer. But, yes, I see what he means.
CW: Few biologists seem to be interested in artificial life. Yet, it seems like such a fruitful area of research.
RD: Is that so? Well if you're right, then that's rather boring of them. I'm not sure it is my experience.
I think that I have met a few who are very tied to the facts and the data as we actually see them and have no interest in a wider, more speculative, imaginative view. I think this hit me most strongly when I was promoting the view, which is in the last chapter of The Blind Watchmaker, that it's very hard to pin down what life on other planets might be like. It could be exceedingly alien and hard to predict. But I gave reasons spelled out in some detail why I thought that, whatever else happened on other planets, it would have evolved by Darwinian natural selection of some kind. I gave reason to suppose that there is no other way of getting life than some form of Darwinism. Nobody has ever criticized that reasoning. It doesn't seem to be flawed. But there are quite a few people who just think it's not interesting because "who cares about other planets?"
CW: Some biologists, like Tom Ray, say that evolution is going on in computers already. Do you feel like life could be leaping now from a wet, carbon-based medium to a dry, digital one?
RD: In one sense Ray is right. But it isn't going to leap. There are plenty of things that computer life does not do which in ordinary language we think of life as doing. For one thing, it's confined within the computer memory. But if we imagine an artificial life program that didn't sit in a static computer on a table but sat in a mobile robot that walked around picking things up and looking at them and turning them over and manipulating them, shaping them, cutting them, finding the paths to build new machines, then we've now got a continuum towards machines that actually make other machines similar to themselves. So if we think of this machine as having as part of its mechanism not only a computer, not only wheels or legs to walk around on, but an industrial robot arm and hand that's capable of manipulating objects and assembling the bits to form another machine like itself, then we would have something far closer to what we understand by life. And then we might really start getting alarmed about a sort of takeover.
CW: In addition to artificial life, there are many other things that humans now design including, in some sense, our own destinies. That seems to have been moved out of the realm of natural selection and into the realm of design. This may be a semantic question, but is this changing the nature of natural selection, or is this entirely set apart from natural selection?
RD: Before humans came on the scene, natural selection shaped living creatures to become very well adapted and to have the illusion of design. But natural selection never had foresight. It was never capable of doing more than take the optimal path from the present position. It could never say, "Okay, let's put up with some temporary bad times to make things better in the long run." That could never happen. When humans start manipulating evolution they can do that because with our brains we can actually calculate or imagine what the long-term consequences of various courses of action might be.
There are two different things we can do. We can practice artificial selection, which we have been doing for thousands of years. And we can actually practice artificial mutation, directed mutation which we can only just about begin to do now. But either of those things can be used with the gift of foresight. So we could decide to breed a race of domestic animals which is useful for some purpose and we can go through, say, a hundred generations where these animals are worse than useless for anything let alone the purpose that we're trying to breed them for. But we're steering them through evolutionary space towards some distant target and we'll eventually get there, and then we end up with creatures, or plants, that are useful for whatever it might be. Directed mutation is so new that it's really quite hard to know where that's going to lead us.
CW: There's another way in which humans direct and design things in a more short-term way all the time. Making decisions that may be counter to genetic benefit but are in line with some other chosen benefit. That, too, must be a relatively new innovation and must portend a profound change. Will evolution ever be the same?
RD: Yes, I think you're quite right. A simple example, I suppose, is contraception. When we use contraceptives at a time when we are economically capable of having children then we are subverting the ordinary direction of evolution. But in more subtle ways we're doing that all the time when we devote our time and our energy and our enthusiasm and our goal-seeking proclivities to following goals which have nothing to do with survival or reproduction.
It looks to me as though the way to think about it is that natural selection built into our brains, goal-seeking mechanisms for good reasons. I mean originally they were harnessed to goals of survival and reproduction. But now that we've got those goal-seeking mechanisms, we can harness them to finishing a book or writing a symphony. We go after these goals with just the same single-minded determination, but they have nothing to do with gene propagation.
CW: Is this a temporary aberration?
RD: Well it's an aberration from the gene's point of view. I'm not sure how temporary.
CW: Cultural evolution is subverting, or eclipsing, natural selection.
CW: Why do you think that religion persists so strongly?
RD: There are positive attractions of religion which must be very powerful. Even people who are perfectly well aware of the scientific explanation of the world and life remain unconvinced that God is no longer necessary in our description of the world. I suspect that religion persists partly because of a kind of imprinting mechanism in the brain. If children are told something with sufficient conviction when they're sufficiently young it's quite hard for them to shake it off. And, again, there are very good Darwinian reasons for this. It's very easy to see that a young child, to survive in its culture, needs to be taught a whole lot of very important information. Not only the language of the culture, and customs, but also useful things like which berries not to eat, and not to put your hand in the fire. And so a sort of rule of thumb of the juvenile brain, "believe everything you are told," would be a sensible rule for natural selection to have built in. But of course such a rule of thumb is vulnerable to parasitization by any information whether true of false. So if the child is told "there is a Juju in the sky who will eat you if you don't sacrifice once a week," then the child will believe that. How can the child distinguish that kind of information from: "Don't eat red berries?"
Well, if that imprinting mechanism is so powerful that a significant number of those children never shake it off, even in adulthood, they will then pass on the same message to their own children and so on. Like a computer virus, this imprinting mechanism that IÕm postulating is vulnerable to picking up all kinds of parasitic information down the generations. If this were true, of course, you'd expect it to be very difficult to stamp out. You'd also expect that in different cultures the details would be different. You would expect them to be entirely arbitrary and variable from culture to culture. But you would expect them, however arbitrary, to be very hard to eradicate.
CW: But is it feasible that a parasitic idea could become so dominant, could loom so large in so many cultures?
RD: Clearly some kinds of information in a non-arbitrary way are going to survive better than others. So there probably is some bit of cultural Darwinism going on. Some ideas survive not just for purely arbitrary reasons but because they have survival value (and remember survival value doesn't mean for the individual life or society, it means for the idea itself). It' the equivalent of the gene, the mental virus. So any mental virus which has what it takes to survive, will survive.
It' not enough that your parents told you "X", and therefore you pass it on to your children, but "X" has some particular resonance with the mind. It has what it takes to survive. Suppose 'X': "If you do not do whatever's required of you by the religion of the society you will go to hell--and hell will be elaborated in graphic and terrifying detail, which may give you nightmares--that already is a virus that has a little bit more of what it takes to survive than just the mere fact that you're told it as a child.
Another one would be, "When you go out into the world you will find people who will deny that "x" is true. They will have very plausible arguments, but don't listen to them. That's just the Devil talking. The more plausible they sound the more this is the Devil talking and the more you've got to guard yourself against them." Well that, you can see, is a kind of self-protecting mental virus. It's planting a hedge around itself.
I know that in the United States people who have been, so to speak, "educated" in creationism, are systematically forearmed, forewarned against the arguments of scientists.
CW: Religion would seem to have moved people to act in socially responsible ways. If the selfish gene theory is taken to heart, and if it does evaporate the dew of religion and superstition that remains, what kind of ethical and social ramifications might that have? What would we be left with?
RD: I would be extremely unhappy if anybody took the selfish gene theory as a guiding principle for life. I think almost the opposite. You could take it as a guiding principle as to how not to behave. So, we're not going to use selfish genes as our guiding principle. We're going to use something else. I don't know what it is, and you're putting me on the spot asking me what it is, because it would have to be some sort of moral, philosophical, ethical principle which we work out by discussion and argument and democratic politics.
Now religion, purportedly, gives us an ethical system by which to live and we only have to look in the Good Book to see how to behave. But actually you will quickly notice if you read the Good Book that people don't follow what it says. If they did they'd be stoning adulteresses to death and making human sacrifices and all sorts of horrific things. Religious people do not get their morals from the Bible: they pick and choose. You can find plenty of the verses that have moral precepts that you, emphatically, don't follow. How do they choose? Well, again they do it using the same modern, liberal, consensus, democratic principles of moral philosophy. Somehow we all think that the Good Samaritan is worth emulating but Abraham sacrificing Isaac is not. We do it on non-biblical principles.
So to sum up, I suppose what I'm saying is that I want to throw out the ancient texts of DNA and I also want to throw out the ancient texts of holy books. They're both equally bad as guiding principles for modern behavior and we've got to find something else. And I'm not saying it's easy to find something else, but religious people certainly have no better clue about where to find that something else than any of the rest of us.