Evolution, 7(4): 386-387 (Dec. 1953)
THE "BALDWIN EFFECT," "GENETIC ASSIMILATION" AND "HOMEOSTASIS"
C. H. WADDINGTON
Edinburgh, Scotland

1 Evolution, Vol. VII, No. 2, p. 118.
2 Ibid., p. 110.

When preparing my article on "Genetic Assimilation of an Acquired Character,"1 I debated with myself whether to include a discussion of the old train of thought usually referred to as "organic selection," but in the interests of brevity decided not to. However, the immediately preceding article by Simpson on "The Baldwin Effect"2 dealt with just this subject, and it would now perhaps he useful to indicate the way in which the idea which I was putting forward differs from those which he recapitulated.

Simpson describes the Baldwin Effect as taking place in three stages, which put very shortly are: (1) Individual organisms interact with the environment in such a way as to produce nonhereditary adaptations; (2) Genetic factors producing similar traits occur in the population (3) These factors increase in frequency under natural selection. The gap in the argument is between steps (1) and (2). Is there supposed to be any connection between the developmental adaptations and the genes with similar effect, and if so, what? Simpson (p. 115) says that either there is no particular connection, in which case the theory signifies very little, or the connection must be by way of a neo-Lamarckian causal connection. Huxley (Evolution: The Modern Synthesis, 1942, p. 304) seems to put the point originally made by Baldwin and Lloyd Morgan more clearly when he writes that the adaptive modifications operate "by holding the strain in an environment where mutations tending in the same direction will be selected." Thus according to both Simpson and Huxley, the theory of "organic selection" still leaves the actual nature of the adaptive changes produced to the operation of random mutations or Lamarckism.

3 Schmalhausen in his "Factors of Evolution," 1947, meant, I think, either nearly or quite the same thing, but seems to confuse the "normalising" action of selection, which keeps a population uniform, with its "stabilising" action, which renders development relatively independent of environmental variations.

The theory which I put forward in 1942, in an article which Simpson seems to have overlooked3 (Nature, Vol. 150, p. 563), and which provided the basis for the experiments described in my Evolution paper, attempted to go considerably further. I argued that natural selection for the ability to produce an adaptive phenotype would change the genotypes in such a way as to encourage the appearance of genetically controlled variants mimicking the adaptive type. The initial non-hereditary response therefore does not merely allow the organism to persist in a new environment and become adapted to it; it enables natural selection to set the stage in such a way that the useful genetic effect is likely to occur.

Simpson comes to the conclusion that the Baldwin effect, in the sense he describes it, has probably played a rather small role in evolution. The genetic assimilation mechanism, however, must be a factor in all natural selection, since the properties with which that process is concerned are always phenotypic; properties, that is, which are the products of genotypes interacting with environments. By speaking of mutations as "random," which is true enough at the level of the gene as a protein-DNA complex, we obscure the fact that the effect of a mutation, as far as natural selection is concerned, is conditioned by the way it modifies the reaction with the environment of a genotype which has already been selected on the basis of its response to that environment. This is not neo-Lamarckism, but it is a point which has been unduly neglected by neo-Darwinism.

The idea of the "canalization" or "buffering" of development (Waddingtnn, 1939, 1940), which underlies the theory of genetic assimilation, is also closely related to the concept of homeostasis, which has recently been introduced in genetical theory. This word has been used in two senses, which should be distinguished. Lerner (1950) has applied it in connection with the tendency of the gene frequencies in a population, after disturbance by artificial selection, to return, under the influence of natural selection, to an equilibrium state. This may be called "homeostasis of gene ratios" or "genetic homeostasis." The aspect of natural selection concerned in maintaining it is what I have called "normalising selection" (Waddington, 1953). Dobzhansky and Wallace (1953) use the term in quite a different connection; an organism is said to be homeostatic when it "adjusts itself to recurrent environmental changes in such a way that its function continues unimpaired." This might perhaps he called "developmental homeostasis." But the word homeostasis is perhaps an unfortunate one, since it seems to imply a stationary state, whereas development essentially involves change in time. It is for this reason that I have preferred the words canalization or buffering, which refer to an equilibrium sequence of states rather than to one unchanging equilibrium state. For the aspect of natural selection concerned with setting op such buffered developmental systems, I have recently (1953) used the term "stabilising selection," which was introduced by Schmalhausen (1949) who, however, does not clearly distinguish it from normalising selection. But again this is perhaps not a very satisfactory expression, since the stabilisation of development would seem to imply that change had been brought to a standstill. Probably the best expression would be "canalizing selection," derived from the word used in the original discussion of the concept.

REFERENCES

Waddington Bibliography