JBS Haldane Archive
Source: Science & Society, Volume II, Number 2, Spring, 1938;
Transcribed: for marxists.org in June, 2002.
Professor J. B. S Haldane's article "A Dialectical Account of Evolution" in the Summer number of SCIENCE & SOCIETY is not only by far the most interesting paper purporting to deal with dialectics that I have ever read, but has the unusual quality of showing a scientist using dialectic in his own field of work instead of recommending it to scientists in other branches. Professor Haldane claims that dialectical materialism is "a very powerful weapon for the interpretation of biological facts" and that "in biology, at any rate, it makes for clear thinking".
A closer examination of Professor Haldane's fascinating account of evolution shows, however, that in his scheme dialectics is not used as a weapon or method of scientific research or discovery — as is usually claimed for it — but merely as a kind of pigeon-hole into which, with much ingenuity (and a little pressure) he fits the biological principles discovered by scientists who were in the main quite innocent of dialectics.
Some time ago I performed a similar feat (although I cannot claim to have accomplished it as elegantly as Professor Haldane) in a discussion with Mr. M. H. Dobb on the development of the theory of pricing in a socialist state. I found it convenient to arrange the successive points of view that had been put forward in a system of triads of theses, antitheses and syntheses. I did not use the dialectic as a method of discovering the truth about the world; I did not apply it to the phenomena of the world, but only to theories about the world; I did not suggest that it superseded or rendered invalid the normal principles of "bourgeois" logic; and I did not suggest that the arrangement of three points of view which might plausibly be considered to fit into the triad gave any cogency to the derivation of the synthesis from the thesis and the antithesis in the sort of way in which one definite conclusion follows from a major and minor premise in a "bourgeois" syllogism.
Mr. Dobb immediately exposed my use of the dialectic "merely as a pigeon-hole," and I was forced to admit that that was all for which I found it useful. Professor Haldane has done exactly the same, with the difference that in his case it does not serve very conveniently even as a pigeon-hole. Before examining Professor Haldane's use of the triad, I would like to examine very briefly one or two other dialectical phrases that are used in his article.
In discussing the difference between a small population which, by selection, may be turned into a pure line in a relatively small number of generations, and a large population in which mutations are more likely to arise before the line becomes pure, and so prevent the pure line from being reached, Professor Haldane tells us that quantity has changed into quality. This is supposed to be important because the large population, having as a result of the uneliminated mutations a reserve of variability, is able to adapt itself quickly to changes in environment. It is not clear to me, perhaps because my biology is limited to what I have learned from Professor Haldane's article, why the small population, although its line is pure, should not have an equal chance of the same favorable mutations. But in any case it is important to note that the changing of quantity into quality, as used by Professor Haldane, denotes that the change in quality is one that can be explained in terms of the change in quantity. Such explanation of changes in quality in terms of quantity seem to me to form a large part, and possibly the whole object, of scientific research. Most propagators of the dialectic, however, give that expression an almost exactly opposite meaning. They insist that only "mechanists" or "mechanical materialists" try to reduce quality changes to quantity changes, and that dialectical materialism, being free from that fetish, recognizes changes in quality which can never be reduced to quantity. In this way they repeat, in other words, the superstitions of the Holists or the Emergent Evolutionists, and are enabled to bring to the most ridiculous conclusions the apparent support of the acceptance of dialectical materialism by Marx, Engels and Lenin. It would be a pity if Professor Haldane, by using this phrase, should lend his reputation to such darkening of counsel; especially as his use of the phrase does not seem to have any real purpose.
Another example of the use of a phrase for no reason other than that it is connected traditionally with the dialectic is Professor Haldane's assertion that the probability that the gene or genes, if any, that enable their possessors to become richer under capitalism are not the same as the genes that make for biological fitness constitutes "one of the many contradictions of capitalism." The same situation is likely to exist in a socialist society. Would Professor Haldane then call this one of the contradictions of socialism?
The most important point of the article, however, is not the use of the phrases already discussed but the fitting of the development of the theory of evolution into the dialectical pigeon-hole of thesis, antithesis and synthesis. This tripartite scheme is often useful for classifying a series of consecutive arguments of a special type, namely, when one of the concepts has been inadequately defined. The first proposition uses the inadequately defined concept in one sense, and in that sense the proposition is true; the second proposition uses the same word or phrase in a different sense, and while if that is done the second proposition is also true, if the phrase is interpreted in the first sense the second proposition contradicts the first, and is therefore untrue. If in both propositions, the ambiguous phrase is interpreted in the second sense, the first proposition will be untrue. Both. propositions are true if the phrase is given a different and appropriate sense in each case. In the third proposition, the truth that is contained in each of the two previous propositions is re-stated with the ambiguity removed by a re-definition of the previously inadequately-defined term. This procedure may appropriately be called dialectical, since this is the way in which all useful arguments come to a conclusion.
It is possible by means of a simple trick to fit into the dialectical framework arguments between three or more people. One simply takes any two of them and gets the synthesis of these two, displaying one dialectical triad. One then takes this synthesis and one of the other arguments and gets the synthesis of these two, and so one can proceed and fit any number of different but reconcilable arguments into a system of dialectical triads. The procedure is exactly like the exercises in formal logic in which a long and complicated argument is arranged in a series of three jointed syllogisms. In both cases one can save a lot of time and patience if one does not believe that three is a magic number.
It should be noted further that when two apparently contradictory propositions are synthesized in this way, it does not matter in the slightest which is considered to be the thesis and which the antithesis. (Though there is usually a difference between a major and a minor premise in the syllogistic triad.) It is historical priority or majority opinion or some such irrelevant and often unverifiable distinction that confers the honour of the title Thesis to one of the propositions, and it does riot really matter. In the use of the dialectical triad that I am describing, thesis and antithesis are perfectly symmetrical and interchangeable terms.
The synthesis of apparently contradictory propositions by redefinition of terms is the only procedure I have ever come across which can plausibly be described as dialectical and which is neither scientifically illegitimate (as are the uses referred to in the last footnote) nor scientifically meaningless (as I shall suggest are Professor Haldane's examples). In a sense, the removal of an apparent contradiction by redefinition of terms is arbitrary. It is sometimes possible to come to the correct solution without the formal procedure being made explicit. The redefinition of terms is nothing but a concession to a convention of language (not to use the same symbols for contradictory propositions), that some idealist philosophers like to parade as an example of that fabulous animal, an "a priori synthetic proposition." If he does depart from this convention, he is depriving himself of the means, generally considered very useful, of distinguishing the affirmation of a statement from its denial, and consequently also of the means of knowing whether any proposition follows from another or not.
Professor Haldane's use of the dialectical triad is much looser even than the synthesis of apparent contradictions by redefinition. There is little uniformity of logical or even historical relationship between the different items. He concludes by tabling "three Hegelian triads":
|3. Selection of
|Survival of |
The relationship between tile different items in these three series can be roughly indicated by the signs or comments that I have added below in brackets.
(1) (Not) Heredity (but) Mutation (i.e.) Variation.
This means: It is not true that the qualities of descendants of a pure species are completely determined by the genes inherited from their ancestors. Sometimes there are mutations of the genes arising from accidental or outside causes. In other words, sometimes the descendants vary in their qualities from their ancestors.
(2) Variation (+) selection (=) evolution.
This means: Since individuals in a species keep on sporting variations, and variations favourable to fitness tend to be selected a species tends to evolve in the direction of the favourable variations.
|(3) Selection of
|(.'.) Loss of fitness||(.'.) Survival of |
This means: Within a (prosperous) species, variations that increase the relative fitness of individuals (in their struggle with other individuals in the same species) tend to select such individuals at the expense of the other members of the species. Since a variation which increases relative fitness probably diminishes absolute fitness (in the struggle of the species as a whole with its external environment), the success of the relatively fit leads to a diminution in the absolute fitness of the species. From this it follows that species which refrain from decreasing their (absolute) fitness in this manner are more likely to survive in their struggle against the external environment (which includes the other species).
We can now see how heterogeneous in shape are the pegs that Professor Haldane has pushed into the dialectical holes. In (1) there is a contradiction between the thesis and the antithesis, but that is simply because the thesis is wrong and the antithesis is right. The synthesis here is nothing but the repetition of the antithesis in other words.
In (2) there is not even the appearance of a contradiction, between the thesis and the antithesis. They are both true and represent two influences that are at work. The result of these two influences is called the synthesis.
In (3) there is an apparent contradiction between the thesis and the antithesis of the type discussed by me above. This apparent contradiction was dispelled by Professor Haldane by the appropriate device of what I have called "synthesis by redefinition of terms." In place of the ambiguous symbol "fitness" he substituted the two different symbols "relative fitness" and "absolute fitness." Professor Haldane does not, however, put as his synthesis the statement that directly emerges from and contains nothing but the truth that resides in the thesis and antithesis. When his redefinition of terms has dispelled the illusion of a contradiction between the thesis and the antithesis, there is found to exist quite another logical relationship between these. The antithesis is a result of the thesis (or is implied in it). Professor Haldane continues the chain of causation and presents us, under the name of synthesis, with a result (or implication) of the antithesis.
Nor are these the only sets of propositions that are declared in the course of the article under consideration to fit the triad of the dialectic. On page 481, we are given a triad closely related to triad (2). "Just as mutation negates the fixity of a species, natural selection, to a first approximation, negates the negation, and we are apparently back where we were at a uniform population"; i.e., as a first approximation we have:
(2a) Variation (+) selection (=) uniformity.
In the actual world, however, Professor Haldane points out the negation is not completely negated (or it is slightly more than negated) so that instead of the everlasting perpetuation of uniform populations we get evolution. One can hear the dialectical holes cracking as Professor Haldane inserts his scientific pegs.
On page 477, Professor Haldane writes, "Biologists to whom dialectical materialism means nothing, or means a weapon of the abominable Marx, cannot understand how harmful mutations can be in conditions of evolutionary progress." This again turns out to be a case of apparent contradiction by inadequate definition. The harmful mutations that are a condition of evolutionary progress turn out to be mutations that are harmful in one environment but beneficial when there is a change in the environment, and it is only when there is this change in the environment that the harmful mutation — now beneficial — begins to play its part in evolutionary progress. Is it really necessary to go in for dialectical contortionism to be able to understand this? What is gained by declaring that mutation is a dialectical process on the ground that "the self-reproductive organism is negated by accidents of a certain type. It can no longer reproduce itself unchanged. But since it does reproduce itself in a changed form ... the negation is "egated." (p. 477.) Is there any process that cannot be shown to be dialectical in such a way?
Professor Haldane attempts to save the face of this particularly empty piece of rigmarole by contrasting it with a much more serious possible deviation from the scientific spirit. "If we do not look at it [mutation] dialectically, we are apt to label it either as pathological or progressive. In fact it constitutes a union of these opposites."
If "dialectically" is used as synonymous with "scientifically" or "sensibly" (as it very often is), the first part of this quotation is true. The use of words implying that something is good or bad, without specifying what it is good or bad for, is inadmissible in scientific work, dialectic or no dialectic. If the environment is specified, and pathology and progress are defined, say in terms of fitness for survival in that environment, the use of these words ceases to be unscientific and the procedure is indeed described by Professor Haldane as being his own method — though it is not clear in what way, if any, it then becomes dialectical (again unless that word merely means "sensible"). What is wrong with the labelling of mutations as pathological or progressive is not that it is undialectical but simply inadequate as definition.
The second part of my last quotation from Professor Haldane's article is even more surprising. It is another dialectical triad, modestly suppressed in his summary of results, applied this time not to propositions but to actual phenomena:
It is not very clear what is the logical relationship of these concepts. It may be
|(but as far
as we are
|(it all depends
on the environment.
Anyway it's a)
|(which at the
same time is a)
None of these interpretations give the same logical relationship as any of the other triads we have previously examined. The first two make sense. The last is in conformity with dialectical rules if these are taken to mean what they say.
Professor Haldane's article clearly does not improve the interpretation of biological facts for students who are not suffering from an overpowering emotional urge to embrace the dialectic.
A. P. LERNER
I have to thank Dr. Lerner for giving me an opportunity to expand a number of points which I left undeveloped in my article "A Dialectical Account of Evolution," and for the opportunity to rectify some omissions in that article. I shall deal one by one with what I take to be his main points.
(1) "Haldane uses dialectics not as a method of research but as a scheme of exposition."
The article in question was devoted to exposition, not to an account of my researches. Nevertheless, I claim that a good deal of my recently published research has been inspired by my gradually increasing knowledge of dialectical materialism. For example, I recently published the first serious estimate of the rate of mutation of a human gene. I did this after I had been compelled, when writing an essay on "The Present State of the Theory of Natural Selection" for Gosisdat (one of the Moscow publishing houses), to try to state the Mendelian theory of evolution in dialectical terms. Some five years earlier, I had discussed the equilibrium (or near equilibrium) between mutation and selection in mathematical terms. The intellectual effort of doing so had exhausted me, and it was not until (thanks to Engels) I could state the situation verbally, that I saw how this approximate equilibrium gave the key to many surprising facts, including the frequency of hemophilia in human populations.
I do not think it is a mere coincidence that my laboratory is probably the only place outside the Soviet Union where this equilibrium and its disturbances are being studied. In the same way, a study of Engels has enabled me, after some years, to crystallize a number of heterogeneous data regarding selection of domestic animals into what I believe to be a new principle, first enunciated in The Modern Quarterly (London), I, no. 2.
I do not claim that these results could not have been obtained without a study of Engels. I merely state that they were not reached without such a study, and that so long as I find dialectical materialism a valuable tool in research, I propose to state the fact.
(2) The "passage of quantity into quality" is used to cover both changes which are readily explicable and others which are not. Thus it is perfectly clear to any reader of "Capital" why $10,000 can function as capital, but $10 cannot, at least by itself. On the other hand, it is even now not in the least clear why ice melts at 0 C., and it was utterly mysterious in Engels' time. I do not see why the phrase should not be used to cover explicable changes as well as changes which are not yet explained.
I am quite prepared to believe that the support of dialectical materialism has been claimed for a lot of nonsense. That does not disprove it. If Marshall's economic views were shown to be incorrect, that would not invalidate the differential calculus.
(3) I still think the difference between the small and large population to which Dr. Lerner refers can be regarded as qualitative. Let us consider a recessive gene which is at once eliminated by the breeder when detected in the homozygous condition, but appears by mutation with a frequency of once per million generations. The chance that such a gene will be present in a population of twenty is 0.008 per cent. The chance that it will be absent in a population of a million is very much less. In other words, it is a very safe bet (though not an absolute certainty) that the two populations will differ qualitatively. As this fact is still not realized by some critics of Darwinism, I think it should be stated in a definite way. Hence I cannot agree with Dr. Lerner that my use of the phrase "does not seem to have any real purpose."
(4) The main reason why the genes making for the accumulation of wealth under capitalism are not the same as those making for biological fitness is that childless people can accumulate wealth more easily than those with children. In so far as it remains true under socialism, it is a contradiction, though not a very important one, as the acccumulation of savings does not lead to power. In the communistic stage of socialism (to each according to his needs) it would obviously cease to be true.
(5) In so far as I have used the Hegelian triad of thesis, antithesis and synthesis, in a mechanical manner, I must plead guilty to one of Dr. Lerner's charges. If I did so, it was for the sake of clarity. A complete exposition of my point of view would have required a whole book, and the article in question was written in rather difficult circumstances as a basis for discussion rather than a complete theory. Nevertheless, I totally disagree with Dr. Lerner's statement that "it does not matter in the slightest degree which (of the two apparently contradictory propositions) is considered to be the thesis and which to be the antithesis."
Let me make this clear. I say that heredity (by which I mean that a gene in the parent gives rise to like genes in the offspring) is negated by mutation. As the result of mutation, unlike genes appear. If the "thesis and antithesis are perfectly symmetrical and interchangeable terms," I might equally well have said, "A gene in a homozygous parent gives rise to an unlike gene in the offspring, except about 99,999 times in 100,000." Similarly an average man dies every day of his life, except 19,999 out of 20,000; and all banknotes (except the genuine variety, of which several examples are known) are forgeries. Such statements are logically true, but are hardly aids to clear thinking.
Again, variation and selection cannot be interchanged, because you can have variation without selection, but not selection without variation. Nor can a species lose fitness from internal changes, unless there is first a process of some kind to cause the internal changes. I point out that even Darwinian selection may cause such changes.
Dr. Lerner says that in the case of Heredity vs. Mutation, the thesis is wrong and the antithesis right. Yet it is not uncommon for a geneticist to work for many years before he encounters a mutation. There is, for example, no sure case of the mutation of the genes for human blood group membership. Doubtless one will be discovered some day. But meanwhile, why is it wrong to use the fact that they are almost always inherited to test paternity?
(6) I quite agree with Dr. Lerner that the negation of a negation may mean several rather different things. Why not? If a formula is to be applicable to natural events in the spheres of physics, chemistry, biology, psychology and economics, and also to our thought about them, then it must be elastic. It is just because they cover so wide a field that the basic dialectical principles are so extremely useful on the one hand, and on the other, so hard to apply in detail without a considerable knowledge of detail.
(7) Harmful mutations can, in my opinion, be a condition of evolutionary progress not only when the environment is changed so that they cease to be harmful, but when it is not changed. As I hoped I had made clear, two or more mutations, each of which by itself is harmful, can cooperate to give a favorable result. I confess that I cannot follow Dr. Lerner's last paragraph but one, which seems to me to be based on a rather complete misapprehension.
Nevertheless I must thank him for his stimulating criticism. It is only with the help of criticism that those of us who are trying to apply the Marxist philosophy to science will succeed in clarifying, not only our exposition, but our actual thought. But the most valuable criticism would be from workers who were engaged in the same branch of science as myself and had accepted Marxist principles. One of my main objects in publishing the article under discussion was to increase the number of such potential critics in English-speaking countries. I do not flatter myself that it will cause "an overpowering emotional urge to embrace the dialectic." The process took me some six years, so it was hardly love at first sight. And I hope that no student of biology will become a user of the dialectic unless he or she is persuaded that it is (as I believe and Dr. Lerner does not) an aid both to the understanding of known biological facts and to the discovery of new ones.
A. P. Lerner, "Economic Theory and Socialist Economy," Review of Economic Studies (Oct., 1934)
M. H. Dobb. "A Reply." Review of Economic Studies (Feb., 1935).
A. P. Lerner. "A Rejoinder to Mr. Dobb," Review of Economic Studies (Feb., 1935).
 This is not so for those who would apply the classification not as Professor Haldane generally does to theories and propositions made by men but to historical processes; i.e., it is not so for those who would claim that the dialectic is not a classificatory device that may (as Professor Haldane claims) or may not (as I believe) clarify the exposition of known facts but a rule about how the world works. On this interpretation Thesis must come before Antithesis and Antithesis before Synthesis. But in scientific discussion such alleged pre-knowledge of the real world patently merits disqualification as simple superstition.
 p. 436
 It is possible to give a slightly different interpretation to the thesis and antithesis in this triad which would render it of the same logical structure as (2).
 It should be observed that in order to preserve the excellent sense of the biology that he is expounding, Professor Haldane has to depart a little from dialectical procedure and make his antithesis "negate the negation" — a task that is usually the prerogative of the synthesis.
 p. 477-478.