Molecular Evolution Activities

Neutral Theory of Molecular Evolution

The Neutral Theory of Molecular Evolution

Panel Discussion on the Origins of Neutral Molecular Evolution

This is an edited transcript of a panel discussion held on May 20, 2004 as part of the Dibner MBL Seminar on Perspectives on Molecular Evolution.

Michael Dietrich
Richard Lewontin
Will Provine
Edna Suarez

Dietrich: Today is going to be devoted to the neutral theory and the neutralist/selectionist controversy. This morning, for the first an hour and half or so, we are going to have a panel discussion on the origins of neutral molecular evolution. … This morning’s panel is Will Provine, Edna Suarez, myself and Richard Lewontin. We will start with a presentation by Richard Lewontin ….

Lewontin: What I want to do is to start in an odd place with a schema, which, I want to claim, perpetuates itself, through a number of different aspects of what interests us.
I want to start with Darwin, believe it or not. What I understand as a tension in Darwin which isn’t talked about much because of the large amount of propaganda made about what Darwin really had to say. You all understand that everybody has his or her own Darwin, and you use Darwin to do whatever work you want to do. The Ernst Mayr Darwin is the Darwin who brought us the populational view, the non-typological populational view of organisms. Before Darwin, people thought of organisms as a type, and the problem of evolution was how you go from one type to another, but Darwin broke that by emphasizing that you should not have the typological view but should see organisms as varying. But it is not as simple as that. Because if you read the Origin, as you have, you will remember that the question arises in the Origin of the origin of variation. But to speak of the origin of variation implies that there exists a stage before the variation. Distinguish between the origin of a particular new variant and the origin of variation. Darwin has a lot to say about the circumstances under which variation appears: environmental stresses and so on, which may lead to a lot of variation. So the total Darwinian scheme is that you begin here, and you then go to that, and then somehow that cluster of variants is transformed in the space and finally, presumably, you go to that. It is a tension in Darwin, because a lot of talk about the transformation variation. In fact, I would describe Darwin’s theory, generally, as a theory that variation within the species and time and space is transformed into differences between species and time and space. So it’s a transformation of variation. Nevertheless, there is, in Darwin, this implicit tension between describing species as variable - that’s the underlying, necessary material for evolutionary change - and somehow a kind of typological view of the species before and after the process.

I want to present that because it’s a model, not historically a model, but a kind of formal model, which appears at other stages of our discussion, in particular, the actual history of the Neutral Theory, as it develops, and also the historiography of the Neutral Theory. I’m going to talk about this as the origin problem. I’m going to claim we’re not here just to talk about the origin. This is the evolutionary level. We then have the historical level. Then we have historigraphical level.
The origin problem for the neutral theory is what was the introduction or stage that the neutral theory comes into play, from previous theories. Second, we then have the “developmental stage,” not organismic development but the development system. And finally reached a new stage again. We moved from an old theory to a new one, or an old species to a new one.

I want to differentiate between the origin of the neutral theory and its early development. They are two different things. They give us a slightly different view than I had originally and, I believe, a slightly different view than Michael developed.

In the historical reconstruction of Neutral Theory, we have as you remember in the book that I wrote I emphasized exclusively the work of Motoo Kimura and his school, claiming the Neutral Theory as originated by Motoo. Originated out of a conflict about the nature of standing variation in populations, namely the classical/balance theory. And my claim was the origin of this was, in fact, on the part of Motoo, a need to cope with the observation that there was a huge amount of genetic variation within populations, without having to accept the Dobzhansky school of view that that variation has been a evidence of natural selection.

Dobzhansky had a very odd view of two aspects of the evolutionary process, which didn’t quite work. Dobzhansky’s view was that this variation, which one could observe in a natural population of organisms - he never observed it, but he just imagined it was there, from experiments, which were rather ambiguous, but his claimed that there was a lot of variation. His claim was, in fact, itself a product of natural selection, i.e., that is balanced by selective forces with superiority of heterozygotes and so on, that’s why it was there. What he didn’t seem to notice, or at least never made explicit, is that the apparent contradiction between the claim that the standing variation of population is balanced and the claim that it is the basis for this process, because if it is really balanced, it isn’t going anywhere. The whole point about balanced polymorphism is that it is balanced and it is equilibrium. But the Darwinian Theory requires a transformation of that variation in frequency space until you get down here. If you were a balance theorist, you would have to separate the process of the maintenance of the variation on which you claim evolution was going to be built and a different process which would take that variation at some moment and transform it in time and frequency. You would have to suppose one of two things: either you could say that, well, the balance theory proves, if I believe in it, that this variation is of some significance to the physiology of metabolisms in the organisms, because it is balanced and it means different fitnesses. If there are different fitnesses, then it matters to the organisms. But then some change occurs in the environment or the world, one of the homozygotes, instead of being less fit than the heterozygote, becomes more fit and sweeps through the population. That’s one thing you could say. I’m not saying its right.

Provine: Ernst Mayr claim his primary contribution to evolutionary biology need to be a concept of a genetic revolution, because he believed all the balance, all the homeostasis, everything working on the gene pool and you need a genetic revolution, which he never explained.

Lewontin: OK, it’s a genetic revolution. But it does not have to be a revolution. It could be as small as simply one homozygote now creeping up above a heterozygote and taking off. But Dobzhanksy never dealt with that. He didn’t appreciate the problem. The alternative is that you could say that environment changed in such a way that a lot of the formerly selected variation is no long selected, became neutral, and the you could have a evolutionary process by random drift. But at least you had to assert that something, not necessarily revolutionary, but a change in the process between the origin - the origin was mutational, nobody cared about that – that this developmental process was different. I want to emphasize that. That’s the balance school. That’s an inconsistency or at least a missing part in the balance school. You can’t simultaneously say that it’s balance and that it’s a phase in evolution.
The Neutralist School, however, is the reverse. The neutralist school says that the very same process, which is productive of the variation, inevitably results in some movement through the space to a change at the end, that is to say, in Motoo Kimura‘s papers, in one paper, he talked about polymorphism as a stage in the evolution of (molecular evolution). But that’s got upside-down, in a sense, for self-inflationary reasons. The theory of neutral polymorphism due to Sewall Wright, from the very beginning, intrinsically was a theory of both variation and evolution, because the dynamic of a random process includes the fact that the gene frequency is undergoing a random walk, and eventually falls off the edge and becomes fixed, as all random walks in one dimension do. You don’t require another process in addition to the process that molds the variation in that particular instant, if you were a neutralist, because neutrality is a theory simply of random walks, which includes both variation and eventual fixation. So in that sense, the neutral view of the matter, which we owe all of that to Wright. To say that I regard variation as a stage in the process of evolution. Well, you can if you like, if you’re trying to explain evolution in the first place. But I can turn that upside down and say I regard evolution is the inevitable consequence of variation. I could take the neutral variation as the fundamental point about the world, and say: “Oh, incidentally if you believe that, then organisms are going to evolve.” It’s a very important change in point of view, because the adaptive theory has this notion built into it somehow from Darwin that evolution is an adaptive process and that variation is an intermediate stage of it, whereas the random theory turns that upside down. So I’m not going to think anything about adaptation, or changes in the world. Any random process will go to fixation at some place and will almost certainly be different from what you started with. It is very important to distinguish those two.

In the historical situation of the development of neutral theory, I’m concerned with this problem, told the story of neutral theory from this standpoint, that is to say-- By the way, Edna there I disagree with you. The experiments done by Hubby and me were not part of the program of the balance theory. I was never a balance theorist. Being rebellious, anything that Dobzhansky said I disagreed with. I can honestly say - it’s not post hoc revisionism - that the purpose of the experiment was to ask the question in a completely neutral way - how much variation is there? There was meant to be a solution to that problem without a commitment to what the answer would be. In fact, I have to tell you that I was very surprised by the answer. Underneath, I thought probably there wasn’t much variation. I just want to make that historiographically clear. I don’t think anywhere in our papers did we then say that this is a manifestation of the balance theory. On the contrary, we indicated that there were real problems, because you have the load problem. Then especially in my book I said you can solve the load problem by not being so naïve about it, and back and forth and back and forth but I think its true.

Dietrich: At the end of Lewontin-Hubby, you got through three alternatives and the heterotic alleles turn out smelling sweeter than the other two.

Lewontin: Do I? OK

Dietrich: None of the come out smelling sweet.

Lewontin: OK. Fair enough. Let me say that in retrospect I still think, for that set of the data, it’s sweeter, because of the observation of the similarity in gene frequency distributions in different, quite distant populations. That would be almost impossible to explain on neutralism. But again, that’s after the fact of the observations. It was not part of the program. For example, had we found there was tremendous diversity from population to population in the gene frequencies after we had come to the conclusion. So I want to make that clear.

Historically, my book emphasizes one of the elements, namely, the response of the classical theory to the balance theory, and emphasize that the neutral theory was a solution. You could still hold on to the claim that it was not balanced and still be in agreement with the observations What I failed to do, as pointed out originally by Michael to me outside the hall where we were living and I immediately thought that he was right. It was your thesis, right? Was that there was another origin of neutral theory quite independent of the Kimura origin - Kimura just took over Wrightianism. That was from purely biochemical standpoint contained in Jukes and King’s paper. If you read the Jukes and King paper, I hadn’t realized that it was actually added in proof, but it’s clear from the structure of the paper that the numerical population genetic style argument in Jukes and King is a minor issue in the paper. What Jukes and King really are doing are saying as biochemists, we have a priori reasons to thinking that lots of amino acids substitutions will not make any difference to the organism, and certainly lots of nucleotide substitutions don’t make any difference. There is a section in their paper about nucleotide variation, which does no other work in the paper that I can find, except to increase the credibility of the biochemical kind of argument that molecular changes can be neutral, they can be neutral at the nucleotide level, because they are third positions. Of course, I would like to go back and say: “Now, we know, although they can be neutral. It is not the same thing as saying that all third positions are neutral. What I like about the Jukes and King’s paper is, in disagreement with Michael, that they are more even-handed than you said. They don’t say that they are necessarily neutral. They say that don’t forget that as chemists we know that they can be neutral. That must be a significant feature of evolution but that does not make the leitmotif of molecular evolution. It just adds to the repertoire. In fact, I have a quote here from the paper, just to emphasize that point. They were trying to be cagey, and on page 795, in a paragraph at the bottom, they say -- they first go through a load argument which would say it can’t be selected because the replacement load would be too great, and they say that the rate you have to suppose from an acceptable load, this is not in the range of observed evolutionary rates. But the expected rate becomes acceptable ten to minus six per generation within a species number of five million or favorable mutational ten to minus nine per generation or an average selective advantage 0.005. That is to say, if I play a little bit with unknown parameters, make it perfectly acceptable hypothetical values, for nobody knows what parameters really are then the observer rates of evolution at the molecular level are concerned either with predominantly non-Darwinian fixation of neutral change or with predominantly Darwinian positive selection. So they are backing off, and saying: “Look, I don’t know what these parameters really are. If I make them up, I can argue either way. I don’t think evenhandedness is a characteristic of Kimura. Kimura is selling you a bill of goods, namely, it’s neutral. They (Jukes and King) are saying it could be neutral but it doesn’t have to be neutral.

Provine: I totally disagree with your view of Kimura.

Lewontin: I knew you would

Provine: Because in 1967, he published a paper entitled “Genetic variability maintained in a finite population due to mutational production neutral and nearly neutral alleles,” he goes through and gives you physiological argument all the sorts of things that King and Jukes used at a later time. And King and Jukes got all the credit for that. Motoo had that clearly in mind. He just didn’t put it in the 68 paper.

Lewontin: More than that, the structure of the argument is very important. Just as King and Jukes used population genetics arguments to further bolster what is fundamentally a biochemical argument. So Kimura, even much more later, used biological arguments to bolster what is fundamentally a population genetics argument. One of the arguments that Kimura later used was that all the variation observed by us from the electrophoretic stuff must have been in the amino acids on the outside surface of the molecule, where they interacted with water molecules and gave rise to charge differences, and any sensible person knows that a real action in an enzyme molecule is not on those outside amino acids but is buried deep inside. So we were not finding the physiological important variation. We were finding the variation, which, on biochemical grounds alone, we wouldn’t expect to be very important What Will says is true is that both parties used both kinds of evidence. It is perfectly clear that one evidence is subsidiary, supporting for a fundamentally, originally main view, and they are opposite in that sense. We could argue that more, but that’s the point I’m trying to make, that fundamentally King and Jukes’ argument starts with biochemistry and adds a little bit, for credibility, of population genetics. Fundamentally, Kimura is a population geneticist, who adds ad hoc arguments about the protein structure as an independent support.

So that’s history. Then we have the question of the development of the theory. There, of course, we have the big “K”, that it to say, it’s Kimura who takes off. Therefore, the neutral theory, opposing its origin from its development, has the indelible stamp of population genetics, because the more sophisticated people became about the way, in which codons enter into the stability of RNA, the way in which the codon usage is important in determining the rate of turnover of RNA, its stability, its lifetime, its rate of transcription tends to mount, in more modern times, an argument against the a priori neutrality of anything. The shoe is on the other foot, and has become so increasingly with time. If you want to maintain that some nucleotide change is neutral, you can’t take that as given, from first principles. You’ve got to show it, because we had too many examples of third positions which are constrained. In fact, one of the last papers I wrote on experimental, when our group did experiments, which we found a hundred base pairs at the end of a gene, off the end of transcription and translation, in a junk DNA, which was absolutely conserved across species lines over wide species differences, despite the fact that on either side of that, there was a tremendous amount of variation. Everybody agrees now that there are constraints on nucleotides and amino acids, which it is harder to make specific arguments about from biochemistry unless you know a tremendous amount about what’s going on. So the Kimura development is largely responsible, in my view, for the picture, which we see in all textbooks, of the differential rate of accumulation of variation between forms of evolution with a high rate of accumulation – that’s talked about already in King and Jukes and other papers. But this picture, where if you plot time, where time is now independently derived from fossil evidence. We don’t infer time, but you get it from fossil evidence. And here, divergence, that the divergence for very important small protein, where every amino acid matters, it’s very low, whereas for fibrinopeptide A which is cut off, nothing but a safety catch, it’s a system where you don’t really care where your amino acids are, it’s very fast. That kind of argument is a further demonstration of the Neutral Theory. It’s not a population genetic argument. It’s a pure historical Jukes-and-King-style argument.

Dietrich: Zuckerkandl Pauling style argument.

Lewontin: Zuckerkandl Pauling style argument. What I’m trying to emphasize is that there are two kinds of arguments. I did not give them any credit except to say that Kimura used it. Nevertheless, I want to say that the subsequent development is Kimurian and Jukes and King is only brought in as supporting evidence as the theory matures. I want to make that difference.

Now in historiography itself, there is a kind of a parallel difference between origin and… By the way, I want to say that Edna’s paper, if you haven’t read it carefully, you need to read it carefully, because it distinguishes the two problematics as requiring and rolling along together with two kinds of argument and experiment and so on, namely, is your interest that or is your interest this? If your interest in that, that’s what you do. And if your interest is this, that’s what you do. I think that’s a very important contribution that paper makes on the understanding of that: are you interested in the transformation of variation? Are you interested in the change in the end product? Where you start from? And this is all you need. If you’re only interested in the end product. You do not make any population genetics argument at all. You don’t need the observation of variation within populations.
So I want to sum up that literature by saying that in the historiography of the subject itself, there is a differentiation between the explanations of the origin of change and the explanation for the process, which are two different histories, and they don’t have to be the same history. So that’s why I said I’m being a little revisionist by saying that there are two histories, and each one is itself a different historical problematic: one is the problematic of the origin and the other is the problematic of the development.

Dietrich: I have a couple of comments. They are mostly going to set some questions that I hope we can have a chance to discuss. … I want to make one point about the argument that I made in “The Origins of the Neutral Theory” paper. In Dick’s book, which is the thing that got me started on this.

Lewontin: That’s not my book.

Dietrich: That’s Motoo’s book, Yes --I differentiated two basic kinds of claims that he made: one was what I called the Historical Thesis, which was the thesis about the neutral theory being basically neo-classical; the other one was what I called the Ideological Thesis, and the Ideological Thesis never got published. I think it’s worth coming back to, because in the 74 book, in addition to the argument about the actual historical trajectory, you propose a mechanism for why this controversy didn’t resolve. And you said, it didn’t resolve because it was ideologically driven, in other words, it couldn’t stop, and it had to be transformed form classical/balance to neutralist/selectionist, because they are fundamentally opposed ideologies of stability and change that characterize the classical and the balance schools. I think that was a very interesting comment and one that people like Crow profoundly disagreed with.

Lewontin: Because he was the object of that.

Dietrich: In part, because he was the object, but in part because he also had a critique of how you could say that a group as diverse as the neutralists, nationally and ethnically, had a shared ideology about change. Right? So I set for myself a task to trying to track out the individual ideology and it showed up that they are a very diverse lot of people. It is very difficult to find that subtle evidence…
Lewontin: Could I make a remark about that? There is another case where you have to distinguish the origin and development. My claim is that the origin of Kimura-ism is Jim Crow.
Motoo wrote a glowing encomium of my years with Jim Crow. I think that Crow had a very powerful influence on Motoo. Although Motoo was a mature scientist in many ways before he got his PhD, but he did go to Crow to get a PhD. It had a lot of influence. I would claim that Crow’s ideology was very influential.

Lots of people believe things which they got from places, and they don’t know why they originated in those places, and they don’t correspond to their own ideologies. I would be guilt of the most vulgar kind of historical determinism to say that every idea which is espoused by everybody reflects their own ideology. That would be foolish.

But nevertheless, I would claim, for example, I’ve never developed this, that there are a few seminal persons whose ideologies do have a profound influence on their work and who then spread that work to others who then take it up for other reasons, but it stems originally from ideological commitments of the originators. Dobzhanksy, for example, had a very clear political ideology. Some day, maybe I will write about the relationship between that and his view of variation. Quite aside from political ideology, Dobzhansky started out life collecting beetles in Central Asia before he was ever a geneticist. He was immensely impressed with the diversity of beetles. But it isn’t just that. It is that in his whole life he was pushing the super-democratic and vaguely-socialist ideology in every sphere which he entered. The belief that genetic diversity was the characteristic not only of populations but that it was good. Genetic diversity for Dobzhansky was a virtue, a moral virtue not just an observation. All the good people in the world were heterozygotes. Muller said: “The smartest people are homozygotes, for the geniuses.” And Dobzhansky said: “No, no, the geniuses are precisely those who are heterozygotes.” So it was a deep ideological difference. And the fact that many people in his laboratory may not share that ideology but shared Dobzhansky’s view was simply due to the power and charisma of Dobzhansky. I would say, by the way at this point, that an unusually large faction of the students of Dobzhansky were members of the Communist Party. I don’t have to name names, but they were. Or if not members of the Communist Party, the members of he party, damn close to it the members of the CYM and so on, despite the fact that Dobzhansky was a declared enemy of the Soviet state - he would never be classified as an October revolutionist, not a February revolutionist.

John Beatty: Among those lines, I agree and think about the ideology of variation I think that’s why there is Marxist tension in Dobzhansky that you outlined. The tension you outlined is that he is so concerned with maintaining variation that he has no explanation for how you get substitution. But I think evolution, for Dobzhansky, was not substitution after substitution, molecule evolution, but substitutions of polymorphisms, so you evolve from one polymorphism to another poly to another poly to another one. Of course, you don’t see polymorphisms when you’re looking at species comparisons. But underneath that, I think what Dobzhansky thought was going on. …I think it’s important to look at interesting variation rather than substitution, but I wouldn’t distinguish a static vs. dynamic, because it’s dynamic variation.

Lewontin: Let’s get back to a basic model of organisms in the world. This is a kind of states-based organisms text on space that was the notion introduced by Evan Hutchinson. And our observations are that the organisms are distributed as clusters in lots of space, and there are clusters, clusters… The problematic for me, for the evolutionists, is that how you spend the empty places. Now you are quite right that for Dobzhansky evolution meant a movement of that cluster, sometimes it meant splitting the clusters, sometimes it meant moving across the space. But to describe it as a cluster in some perspective misses the point. A cluster is clustered at some place. And the problem of evolution is why you started out here, and got over here. Yes, it’s true that it’s the movement of the variation. But you still have to have a direction element to it. You still have to have to move the location of the cluster somehow in space a long way as compared to the variation, then across the resonant species. So it’s got both of those… I agree that Dobzhansky wouldn’t say that you move from that point to that point but he said you can move from that central location to that central location. And balance theory doesn’t do much for that, especially there are new genotypes…

Jan Sapp: You said that Crow’s ideology is so important in this case. So what was that?

Lewontin: Jim Crow is a political conservative.

Jan Sapp: So that’s what you mean by…?

Lewontin: In that very vulgar sense, but I think that’s a manifestation of a more general…

Dietrich: The way he says the ideology in the 1974 book is that because it is an ideology about maintaining the status quo, he’s conservative in that sense of maintaining the status quo.

Lewontin: That’s what I mean by saying that he was a conservative. I don’t mean that he wants to undertake a fascist revolution. He thinks the status quo is about as good as you can get.

Audience: How conscious is a scientist or historian in that matter ideology that they In that case, would you see youself have any particular ideology…It would seem difficult for me that only certain scientists have ideologies…How some scientists can be free…

Provine: I was his student, and he was devoid of any kind of preconceptions or anything of that sort

Lewontin: There is a tremendous amount of individual variation degree of self-consciousness about that. Dobzhansky was very self-conscious about his ideology. He spent a lot of time. I want ot give you an exmpale. He was concerned that we as geneticists should really oppose Lecyncov. Why? Of course we should, but he said I recognize there is a contradiction. After all, we oppose the Soviet state, we’d like to see them reduce in world power and perhaps disappear. Lecyncov because he believes a wrong thing about something important, namely, the source of food is destructive of the interest of the soviet state. So why don’t we encourage Lecyncov? That will further weaken this terrible regime and lead to the victory of democracy. He said we can’t do that we have a prior and higher obligation to the truth and we have to live with the fact that there is a contradiction. That’s an example of a very explicit political ideology and ethical one, which came into play in his publications and his writing about an issue in genetics, namely the Aaron Parker Hypothesis. And over and over again, as I said before, he did experiments, for example, on selectional behavior in Drosophila, because he wanted to show that population that had genetic variation adapted more quickly to behavioral challenges. He was one of those people from his very conscious, which he turned that into experimental work, propaganda work, and so on, at the scientific level in a very conscious way. Others, of course, are not so conscious of that. Most people I know say they don’t have any political ideology. Political ideology is something that only fascists, communists, and nut cases have. So you are quite right that one could not make the claim that conscious ideology informs the work of most scientists, that ideology is about what means to be a scientist, and what science means. That is not what I was talking about in the particular case.

Provine: Do you know any scientist who is devoid of ideology?

Audience: In the sense that you are talking about it, you have to imagine what scientists could be. But at the same time, I don’t know how interactive ideology and science are, and that’s what I’m trying to understand now.

Lewontin: I agree. It differs a lot. Most people are not conscious of them and would say they don’t affect their everyday decisions. And there are charismatic figures like Dobzhansky, who was very important, and Jim Crow was a very powerful person in population genetics evolution for a very long time, and he put that into practice. Dietrich: I raised this question about ideology. It’s a specific version of a much more general and much more pervasive problem, which is, controversies are by definition a dispute extended in time. So in order for a controversy to occur, it has to persist. The question of why this controversy persists is an important historical problem. The ideology thesis is one possible reason why the controversy could have persisted, but it’s not necessarily our last answer to that question.

Lewontin: Does it persist to this day?

Dietrich: That’s one of the questions maybe we’ll get to. And hopefully, in this afternoon we will get a better idea of exactly what kinds of reasons there are for this persistence. Even up in the early 80s, at least to the point we can say that controversy persisted, long enough to be very noticeable.

Audience: Mike, when you’re talking about ideology, are you talking about the very issues discussed by Dick’s …

Dietrich: Dick’s ideology thesis is one possible explanation for the persistence of this controversy, and one that would be recognized as applicable widely in a number of different controversies. But it’s not necessarily the only explanation. That’s my point.

Audience: Are there any other general explanations that you just want to toss out and …

Dietrich: And poison the well, as it were. No, I’d rather let the conversation evolve and see what explanations a group of mostly philosophers comes to of how what explanations they think there are for the persistence of the controversy by the end of the day.

Audience: … I still don’t understand what relations between Crow being conservative and pushing the classical position…

Dietrich: The classical view is a view that says that you’re mostly homozygous on average; the balance view says that you’re mostly heterozygous or variable. There is a relative frequency controversy over the relative proportion for this kind of things. If you’re maintaining the status quo, then you think that evolution through a long process of adaptation has sculpted those homozygotes to be the best they could be. It is optimized. So why would you want to change from that?

Lewontin: That’s where eugenics came out. We’ve already had the optimal genotype and all genetic variation is bad. And moreover, if you believe that individuals are mostly homozygotes, as Muller and Crow believed, then it changes your attitude about the manifested differences between human groups. Now the differences between groups are large, as compared to the variation within groups. That has tremendous political implications.

Provine: But it doesn’t mean that Crow was a big advocate of the Neutral Theory in the early days. When I first went to Michigan in 1988, Kimura was wondering whether Crow was really supporting the neutral theory. Crow was very much selectionist and was very skeptical of the Neutral Theory. He changed Motoo’s books and papers to read “junk DNA.” Later, Motoo was very upset by that. He said: “Nonono, I’m not the king of junk DNA. I’m the king of DNA!”

Lewontin: At what time was Crow a balance theorist?

Provine: Never.

Lewontin: There were a series of experiments done by the Dobzhansky School, measuring genetic variation on the average dominance of deleterious genes. Then there was the Mullerite school, which claimed that in fact the amount of variation was due to rare deleterious mutations, which were on the average semi-dominant, that is to say, these mutations had a slightly dominant effect on heterozygotes. They weren’t conditionally bad. Whereas the Dobzhansky school kept saying: “No, no, what appears deleterious in a homozygous condition is superiors in a heterozygous condition.” That struggle went on on the experimental level. The same experiments were done in Columbia, Indiana, and Madison with exactly opposite results. That struggle was a very old one, which was not the neutralist struggle. Dobzhansky and his “children” were the balance people, whereas Muller and his “children” were the purifying selection people.

Dobzhansky often constructed experiments, got the protocols, listened to the person who did the experiment, had the experimental work analyzed statistically by Howard Levine, who was his statistician, and wrote the paper, but didn’t put his name on the paper. Dobzhansky would often say: “A good professor has written many more papers than he has his name on.” That sounds like an extraordinary reversal of the Navier effect. Instead of trying to take credit for it, he’s giving it away. But this was a very clever ploy on the part of the men who understood schools of thought, namely, if he can get lots of papers in the literature, which not have his name on it, done by other people, which had results that agreed with him, that would increase the credibility of his view. So what he did was that he created a literature without his own name as a way of showing that… I don’t know if Muller ever did that, but there were certainly more papers… So as historians, don’t take the names on papers, not just the order, but whose even appears on the paper, as too serious, unless you look behind and see what was done, what the motivations were, who decided and so on.

Audience: I think you can see evidence of ideology, but this idea that organisms are already optimized in Ohta’s paper is a later development of the Neutral Theory, because she argues it’s more likely that they are slightly deleterious than they are slightly advantageous. The only argument one conceives of is that we already had been optimized.

Dietrich: So my view is that you can’t call the Neutral Theory just neo-classical. Dick makes a very strong claim in his paper. He says in his paper that to call it a Neutral Theory obscures the origins of it. And my claim was to call it a Neutral Theory obscures the origins of it, but the neo-classical tradition continues through it. That is definitely there, but you have also to take into account that there is this biochemical tradition that is feeding into it as well. So they are not mutually exclusive categories at all.

Lewontin: Nobody disagrees with that.

Audience: But it’s discontinued Dobzhansky and neutral debate, you see the same comparative debate take place, are you assuming that there is continuity between these debates.

Lewontin: No, he’s saying there is another element coming in from Jukes and King’s side, in which biochemical arguments are primary. My claim is incomplete. I only looked at the population genetic argument not the biochemical argument.

Dietrich: That explains that there is continuity but there is also something producing the transformation, and it has to be the biochemical data.

One thing that I wanted to make a small point is that when we were talking about the origins of neutral mutation, neutral evolution, I think one of the important things to realize is that people have been talking about neutral molecular evolution for many years, even in 1964 Kimura and Crow produced the infinite alleles model, and that model is basically applied to different systems of alleles: it’s applied to neutral alleles to different kinds of selection regimes. They do the neutral case because you can solve the map easily when s=0. So you can make the model work. But they say in that paper that they don’t believe that a system of neutral alleles exist in nature. I think the important move in late 1960s is that Kimura and others become convinced that the models they had produced in the past actually apply in the world.

Lewontin: What caused them to make that change in their attitude?

Dietrich: The biochemical evidence.

Lewontin: Not the revelation of huge amounts of variation…

Dietrich: Well, Kimura said it is consistent with that large amount of variability. In the 67 paper that Will has opened right here says that it’s synonymous mutations. It’s incontrovertible in 1967 that the synonymous mutations must be neutral. They have to be neutral.

Provine: Besides the paper, he’s clearly got that in mind.

Dietrich: But I think that is an important shift in the way he’s thinking about even these models, what the models are, so neutrality goes from being a mathematically elegant way of doing modeling to suddenly something that is real.

Lewontin: Are you now proposing that in fact knowing interesting in an important sense, was the population genetic side of the Neutral Theory an attempt to rationalize the observations of huge amounts of variation, that not only my suggestion is incomplete, but it’s wrong. Is that your position?

Dietrich: I’m saying that we have insufficient evidence to support that claim.

Lewontin: You begin to sound like a man of the Bush’s administration. (Laughter)

Dietrich: History should be based on evidence. Will may have more evidence than I have access to. I just had access to Kimura’s published record and some letters. But Will talked a lot, and he may have a better understanding of what Kimura was thinking at that time. But from the published record, he was thinking about it, but it is not obvious that was as decisive.

Lewontin: As a historigraphical issue, the kind of influence which I made is really not legitimate

Dietrich: It’s your influence.

Lewontin: And ought not to be taken seriously as evidence.

Dietrich: You may have had evidence more than you actually …

Lewontin: No, no. All the evidence I have is what I say, namely, what the claims had been about the nature of the genetic variation and how much there was, and the beginning wild-type, and all mutations being deleterious, and deviations from that.

Dietrich: My view is that wow, what an interesting historical correlation you pointed out.

Lewontin: I’m just trying to know where we are.

Audience: We do know that Kimura had an interesting load for a long time, leading up to the publication of that paper. He said in the paper that circulated around our group. I have read previously about Kimura’s contact with ABCC geneticists. They were all very concerned with load. Load was …localized. Load was a big issue in Japan. It certainly wouldn’t be surprising that Kimura would be very sensitive to load issues, based on standing variation. That job’s out, nobody argued for him, but it would be additional evidence that it was important for him the time and he was conditioned to be thinking that way.

Lewontin: So that’s the reason for my distinction between origin and development.

Dietrich: The key piece of evidence that led me to believe that your account of 74 wasn’t right is that he doesn’t use a load argument in 68, he used a cost-selection argument that is cast and based on a 65 evolving genes and proteins data, which are rate of evolution, is this data. He said it’s consistent with the standing variation observed by Lewontin and Hubby.

Suarez: And I think … how he changed his mental state. He was loaded with very genetic load argument, from all his immigration. I think it is consistent with the 68 paper. The evidence that made him think that this hypothetical theoretical case of the neutral mutations was the real case in the world, because of the supposed high rate of substitution.

Dietrich: Which he would have gotten from ABCC too. They are doing cost selection. This is dynamic load. It’s not different from that. What we need is the evidence of how he was thinking about the electrophoretic variability.

Cassette II

Provine: While they are setting that up, I have only a couple of minor things to say. A historical note.

First of all, Kimura had his eye on the prize in the very beginning. From 1967 on, he wanted DNA. That’s what he wanted to do by and by. He’s concerned with writing a paper about the DNA level. He had no data from it. In 1967, not one piece of DNA had been sequenced. It was not until 1968 did Rayvue even start on that problem and then it takes him a while to sequence eight phases on one side and nine on the other of lander size. So it takes a looooong time before we get any information about DNA. Yet, his eye is on the prize the whole time. So the 1986 paper that you have in your hand is the key to Kimura having arrived finally with the DNA data. He says: “Look! It’s all mine. I’m there. I’m in hold. I have no need any longer of any nearly neutral theory, whereas in his book, he put back and forth between completely neutral and his nearly neutral theory as if I will just use whichever one of them that fits the data I have better. But the problem was that he had nothing but protein data to begin with. Ohta, however, is focused on the protein data. She doesn’t care as much about the DNA level. She figures that’s Motoo’s thing. I want to work on the protein level that he is working on because he hasn’t done anything else to work on. But in time, I’ll have the DNA, and she can have the protein. And I’ll take as much from protein as I can get. So when I was there in 1988, they were fighting about the protein level. Kimura figured: “I’ve already got the DNA data.” He handed me his 1986 paper. I compared that with the book and I could see: “OK, he’s going back to his original 1968 way of looking at it. Kimura has his eyes on the prize and Otto has her eyes on the prize as well. That is something very important to keep in mind as you think about the history of the neutral theory, because there is more than one of them and the nearly neutral theory is much more important in the late 70s than is the Kimura theory. His book helps a lot, but what really helps the theory, what really puts him on the bandwagon is the arrival of DNA sequence data.

Lewontin: Why did he care about the…?

Provine: He cared about that from the beginning because he thought that the level was really the contents of DNA sequence thought.

Lewontin: According to the synonymous…?

Provine: That’s exactly what he talked about in his 1967 paper. A lot of these things should be selectively synonymous.

Lewontin: …

Provine: It doesn’t’ really, because he never was really into the protein level. It was just a way the use of data that he can get his hands on until he can get the DNA data

Lewontin: Are you telling me that Kimura was perfectly content that everybody agrees that synonymous substitution is neutral that all of the rest are synonymous. That being sufficient for him?

Provine: No, no, no, he thought a lot more of the protein stuff was selectively neutral. But there is no doubt in my mind that what he wanted from the very beginning, and this is from Otto as well as from Kimura that his eyes were on the prize of DNA, and were not focused on the protein data, which just plain messed up his model. And bygone, by 1986, the 86 paper that you have, you notice there is no mention of the early neutral model any more, nothing there that says “Alright. Now, I’m home. I’m there. It’s great and let me tell how great it is.”

Suarez: I’m going to take a more philosophical approach to this problem. I think that the origins of the neutral theory constitute a very interesting and different question because of the obvious importance of the theory and all lots in inquiring into the fascinating lives. I love to hear about Motoo Kimura, Thomas… etc. It illustrates how molecular biology has permeated all biological fields.
This is a very important historiographical problem in the history of biology. For instance, this is the view that Michael has adopted in his 94 paper. It asks very large questions of historians of biology like Lily Kay. I’d love to do history of recent science, and I’m going to tell you why. It is a very interesting subject for more philosophical and general about the nature of science. For instance, it is at the center of one of the major controversies in contemporary biology. Of course, it’s fading away. It’s an excellent example of the role of theorists in discipline building, and it was the main objective of my paper in 96. It is a wonderful case in arguing for the “epistemic diversity” of science, which is also a point I want to make in my view of the neutral theory. And also I want to say something at the end of the talk about the role historians play in the reconstruction of science. We’re talking about the history of recent science. This is a very important for historians and philosophers of science.

We have seen, by the talks of the previous speakers, the history of neutral theory has been told from basically two perspectives. One, which is illustrated by Richard Lewontin’s book, emphasizes the relations of the neutral theory with the Bellarmine’s in-population genetics. I saw these problems to be linked with the problem of intra-population variation. In this version, the role played by Kimura is emphasized. This is the only picture I could find on the internet. We need a more recent one to be circulated. In this version, King and Jukes version practically disappears. The other version puts its emphasis on different versions of the neutral theory and on its diverse origins. This is the historical approach taken by Michael Dietrich and me.

You see here, Linus Pauling, and in the middle, it’s ??. In the discussion we emphasized the role played by technical and conceptual developments in molecular biology. So Dietrich wrote a paper, in which he presented what he called the Lewontin Historical Thesis. When I wrote my paper two years after that, I wrote about the Dietrich Historical Thesis. Lewontin’s Historical Thesis sustains that the primary evidence for the Neutral Theory were the high levels of genetic polymorphisms revealed by electorphoretic methods in populations of Drosophila pseudoobscura, and that the neutral theory is but a continuation of the classical position, as we have seen before. We tried to extend Dietrich’s arguments. I wanted to make some more general claims about what these differences in interpretations meant for broader questions. I recognized that certainly there was an old involvement of Kimura with the classical views. This is all views with all Kimura integration with Crow all the problem that he has been attacking the 64 paper. There is a lot of evidence, the 64 paper with Crow, for example, in which he recognized the theoretical case of neutral mutation. But Kimura’s original paper (1968) was concerned with issues of evolutionary change. As Will Provine has told us, the commitment of Kimura original paper with this problem of protein evolution put Kimura into a number of problems. He had to change, after 68, many of views and the evidence that sustained his paper. In his 68 paper, the main evidence for neutral evolution was the high levels of substitution or the high rate of evolutionary change in proteins, but King and Jukes made Kimura see that the problem was not the high levels of substitution but the constancy rate of protein evolution. The other one was that in 68 Kimura thought that neutral mutations were the vast majority of all variation found in populations, but King and Jukes made him think that that was not the majority of the population but only between 10 and 16 % and was not enough to explain neutral evolution. Kimura made this mistake and he had to change, later, the evidence in favor of the theory.


The main evidences for neutral evolution came from molecular biology from the papers included in this conference. There are studies about sequence of globins and cytochrome C and some unexpected phenomenon recently found in molecular biology, like satellite DNA, for instance, which was the first of the repetitive sequences, which was discovered in the late 60s. These were very new evidences coming from molecular biology. They have nothing to do with population genetics. The argument that the primary structure of proteins is largely determined by the structure of the genetic code. The genetic code being an experession of functional needs. There are a lot of arguments about the structure of the genetic code and the … of the genetic code, etc. The last argument about the structure of the genetic code is very important because it illustrates one of the main points in King and Jukes paper, the role assigned to the fixation of mutants on DNA as the driving force of molecular evolution, because they are sustaining the split in the mechanisms of evolution between the organismic and the molecular level. What they are trying to do is to break the extrapolationism of the synthesis.

I think in King and Jukes’ view, random drift is a physical force. They are trying to make this point that this is a physical force. DNA is a molecule and is driven by random forces. This is the main evolutionary force at the molecular level. This force is very different than the constant mechanisms that we see other mechanisms led. They are trying to push this very philosophical and explanatory point in their paper. So the protein molecule is continually challenged by mutational changes resulting from base substitution and other mutational events in DNA. Why is it important in this historical discussion? What is the meaning of the differences in the historical interpretation of the origins of the Neutral Theory? What is the point of discussing if Dick Lewontin’s thesis is incomplete or incorrect? The answer has to do with two questions. They may be important for assessing the role of theories on discipline building. If we want to argue for something that historians and sociologists have been arguing in the last two decades, which is the epistemic diversity of science.

Despite Lewontin’s commitments, Lewontin’s Historical Thesis is related to a view of science as a mostly theoretical activity in which experiments are subordinated to the different hypotheses scientists want to test. I’m not saying that you’re sustaining, for instance, some positivistic extreme version of science like Nagel. But in this version, which is mostly eccentrified by Nagel’s theory, scientific disciplines are mostly structured around theories. For authors such as Nagel, the role of disciplines is to get unified through the reduction of their characteristic theories. It is very impressive, when you read the literature on disciplines of very broad set of philosophers. They were talking about disciplines when talking about the unification of science or the relationship between one and another. They just talk about disciplines in this way.

The big issue in the LHT as we have seen is the theoretical controversy between the balance and classical school. In his thesis, the historical controversy is emphasized. And we saw experiments are subordinated to theoretical needs such as the election between competing hypotheses. This type of experimentation can even be thought or designed before the techniques or the technology exists. So it is very clear that Hubby and Lewontin had it very clearly in their minds which methodological conditions their experiments should meet in order to deliver data to choose between the classical or balance hypothesis. I love this quotation from Lewontin’s book, because I think it’s a wonderful statement of what experiments meant to theoretical scientists: “For many years populatin genetics was an immensely rich and powerful theory with virtually no suitable facts on which to operate. It was like a complex and exquisite machine, designed to process a raw material that no one has succeeded in mining. Occasionally some unusually clever or lucky prospector would come upon a natural outcrop of high-grade ore, and part of the machinery would be started up to prove to backers that it really would work. But of the most part the machine was left to the engineers, forever tinkering, forever making improvements, an anticipation of the day when it would be called upon to carry out full production.” I think it really illustrates the state in which population genetics was in the 50s and 60s. But it also shows us about what theoretical scientists think about our experiments. Of course, they would be very happy to find these facts to fit the mechanism. They could tinker with the theoretical machine without having the effects.

Dietrich Historical Thesis, on the contrary, emphasizes the epistemic diversity of knowledge and scientific disciplines. Here we have Walter Fitch, a picture taken about ten years ago in Erwin. In his view, scientific disciplines are not always or necessarily structured around theories. Disciplines arise from the integration of very different kinds of scientific traditions. In my PhD research, I found at least in molecular evolution we could find theoretical traditions which are illustrated by population genetics, Lewontin’s and Kimura’s works, for example. These traditions which have taxonomic and theoretical inquiries, for instance, those exemplified by Fitch and Wallace’s work. In this paper, they have taxonomic concerns, so the kind of questions they asked were of the taxonomic tradition. And we have experimental traditions, which are very different from theoretical traditions.

The role of experiment in this tradition is very different. They are very concerned with the development of techniques and stabilization of phenomena, which are unexpected, not expected as experiments by Hubby and Lewontin. This interpretations by Dietrich, me, and Balawana tries to sustain the epistemic diversity of science. The Neutral Theory is an excellent example of a theory which made an important contribution to the consolidation of a discipline, not because molecular evolution is logically structured around the neutral theory, but because it brought together different kinds of reasoning, because it helped to exert a “socio-professional” break between molecular evolutionists and organismic evolutionists. The Neutral Theory played a very important socio-political role. It helped to divide an organismic evolutionist from a molecular evolutionist. It helped in the creation of this new discipline in the beginning of the 70s. You have to build disciplines on institutions, for instance, journals, societies and so on. It gave a very strong theoretical meaning to empirical concept, such as the molecular clock. The molecular clock acquired a very different meaning in the context of the Neutral Theory, which has taxonomic traditions.

And finally, I’d like to say a few words about the involvement of scientists in the writing of history. Taking into account that we have one of the protagonists here… Certainly, historians of science need to be careful and ever pram facie skeptical about scientists’ own reconstruction of events where they participated. As a case in point I recall you of the recent literature on the history of molecular biology deconstructing James Watson’s own reconstruction of the discovery of the double helix. There is a huge amount of literature in history of science, which deconstructs this position. However, this position needs a humble and symmetrical recognition that one of the main attractions in doing history of recent science is the possibility to check the historians’ own historical bias with the protagonists. For instance, Richard Lewontin has recently corrected me about my own rabid and mistaken inferences about how he and Hubby, in the mid-1960s, first began to work with electrophoresis. If you see Footnote 3 in my paper, since I knew Hubby had been at Texas before moving to Chicago. Since there was a team Lab led by Wilson Stone at Texas, which also reported the use of electrophoresis for the detection of polymorphisms, I incorrectly concluded in my 1964 paper that Hubby had taken the experiment with him to Chicago. This sequence of events wasn’t true, however. Hubby developed his skills of electrophoresis after his Texas involvement in a very independent manner, as recalled by Lewontin. So I think it’s one of the dangerous and one of the advantages of doing history of recent science. Second, of course we are here to discuss the historical thesis of one participant in the events, it is because, fortunately for us, there have always been scientists with interests broader than what may be called “science itself.” I will finish by saying that we know about Lewontin’s very broad interests. He was not just interested in the political, philosophical, and conceptual aspects of biology, and he was also interested in its historical reconstructions. That’s all.

Lewontin: Can I make a comment? One is that the role of theory in population genetics is ambiguous, stronger than any other discipline. That is just a historical fact. But we must not interpret theory in a strict sense. Dobzhansky, for example, in my view, (I’ve written a paper about it) was not essentially an experimentalist but a theoretician. He was, however, a theoretician, as I said, without tools. He was completely anumeral, could not even look up the Table of Normal Distribution, did not know how to do any statistics. Nevertheless, every set of observations that she’s ever made was intended by him to support a prior existing theory that he had. So he was a theoretician not in a practical sense, but in an ordinary, everyday sense. The entire science is structured around the pre-existing theory, so part of my narrowness in my story of the arguments of the Neutral Theory comes from the objective state of population genetics as a theoretical science, and partly, the consequence of my experience, of people like Dobzhansky, who, even though they pretend to be experimentalist, are actually theoreticians. That’s explaining my own narrowness of vision. Finally, I have to say that one has to be wary of scientists obviously telling the story of their own theory; sociologists don’t get into that list. But historians find themselves in a position … namely, historians broaden their interests and go interview other people. They don’t just look at the papers. That means they are getting the scientists’ self-delusion first-hand, rather than second-hand. But the papers are themselves part of the theoretical self-delusion of scientists, who are deeply conscious of creating a written structure, which mimics their own told story. So what you may have pointed out to us is a dilemma from which we have no obvious escape.

Suarez: We just need to check things…

Lewontin: You need to ask yourself the contrary self-delusion of different scientists. (Laughter)

Dietrich: In 1972, you wrote this something could be different

Suarez: It was Lewontin who said that … is like an intersection of many lies.

Provine: I think it’s important to understand that everyone has their own history. It’s told in the form of stories in every meeting they go to. They have stories about all the major familiars, and they are trading back and forth. I’m very skeptical of this kind of history in the sense of historical accuracy. To have all the people come to the meeting and write… when they are on different continents. It’s alright to have and wiping the blackboard. None of the stories are accurate in many ways. His jacket has no pockets, I checked it myself. But what historians have is an element of real and deep and careful and sensitive understanding of people. To this what scientists understand and what tell you about the field is a great mistake, in my opinion. should be carefully attend to, especially since we know to begin with they are never going to pay much attention to what we historians have to say and they will continue with theories. Secondly, I’m profoundly suspicious of historians writing about things, as I am, and scientists writing about the same things. I’m terribly suspicious of everything I’ve ever written, too.

Audience: I think you had a very important clarification of the issues. In your presentation, where you talked about Kimura being evolution I was just wondering if it can be clarified even further. Would you be willing to say that Kimura was interested in explaining evolution, that’s what I mean by “interested in.” Clearly he’s interested in variation, but he was interested in explaining evolution, not just evolution, but evolution in the sense of gene substitution. He was interested in explaining evolution in the sense of gene substitution, not evolution in the sense of changing gene frequencies, the old notion of evolution of Dobzhansky version. That’s not what he was trying to explain. He brought changing gene frequencies into his explanans of gene substitution, but that’s not…
Lewontin: With that particular explanans, they are just dual aspects of the same process.

Suarez: Yes, he would say later in his paper. I think he’s obviously interested in genetic variation. Muller’s influence on Kimura on this issue… Muller took Kimura seriously and they began to talk about the important of the new molecular developments. I think we don’t need to make a break between Kimura thinking inter-population variation and then he being aware of molecular biology and beginning to think in terms of molecular evolution. He had begun to think that this molecular… had to be incorporated into this problem of inter-population genetic variation. I think this book by …, putting together many papers of the Conference in 64, had a very important role in making scientists think that there are enough developments in the field to begin to take this data seriously. I think that’s why he put … to calculate these straits of change. He was thinking about the problem of evolution, but I don’t think this was a break, because as Dick Lewontin said, for the classical view, it was more natural to think that these processes are connected. I don’t think we need to put a break in Kimura’s career.

Dietrich: Generally, we’re having this conversation about polymorphisms vs. substitutions. I read that 71 paper and at the end of it, it says it very clear that he thinks polymorphisms are as important as substitutions. I don’t think you can say it’s a theory of substitution.

Audience: Didn’t he later talk about adaptation?

Dietrich: No.

Audience: If it isn’t a theory of variation war upon polymorphisms, then surely the load argument is …

Lewontin: Not if you believe it’s neutral.

Audience: No, no, but surely having too much…

Dietrich: But he was convinced by other people that the load argument can be undercut…symposium comes out and other people

Suarez: I think that for Kimura’s it’s more natural to go on from one place to the other. They decided that it’s two faces of the same problem.

Lewontin: Will emphasized over and over again, eye on the prize. All the important development of Kimura’s work is impetus to do something that makes an immense and original splash, to be really unique, not just to be just another population geneticist, working out to the second desk place of Wright’s theory of random drift.