Return to History Preamble

A Brief History of Evolutionary Genetics.
Part 1:  Early Days

ŠPaul Handford, 1998

Genetics, the science of inheritance, grew from attempts to understand evolutionary processes

The earliest efforts in the development of a science of inheritance came from attempts to understand the evolutionary process.  Darwin, in effect, simply assumed heritable variation to exist;  he was not much concerned with its origin, determination or transmission.  Most continuous variation (which he usually had in mind) seemed to blend when mixed in offspring, and after a while, blending came to be recognised as a real problem for natural selection:  wouldn't variation quickly disappear, as, each generation, offspring tended towards an intermediate form?  The struggle to reconcile inheritance with natural selection eventually led, via the rediscovery of Mendel's conclusions about inheritance being by irreducible particles, to the theory and practice of Population Genetics that we know today.

Darwin saw evolution as proceeding, probably quite slowly, through the differing success of individuals according to small natural differences ("fluctuations" in the terminology of the day) among them:  this is gradualism.  Unfortunately, as mentioned above, he (and almost everyone else) understood inheritance to be mainly of the blending type:  offspring always being intermediate between their parents.  He acknowledged the existence of large, discontinuous, heritable variants (what were called "sports"), but he felt that they were largely unimportant in evolution in the long term.  Thus there seemed to be a serious problem:  the steady generational loss of variation among individuals.  How could evolutionary change (by 'selection' among genetic variants) continue for long if the variation disappeared in a few generations of blending?  To salvage selection as the driving force of evolution, Darwin proposed, in later editions of "The Origin," and in other writings, an inheritance system which he called pangenesis, by which "messages" were passed from (environmentally-influenced) somatic cells, by hypothetical things he called "gemmules," to the germinal cells, thereby generating new variation in each individual during its lifetime, which could be passed on.  This was clearly a Lamarckian inheritance system.  What irony!  See Richard Dawkins (note 1)  for a fine discussion of the general problems with Lamarckian evolution (inheritance of acquired characters.)

Opinion quicky divided on the smoothness vs. discontinuity of evolutionary change;  this theme remains as a point of contention to this day

Everyone was quickly persuaded of evolutionary change and descent by Darwin's arguments in "On the Origin of Species," published in 1859, but Darwin & Wallace's mechanism of natural selection impressed few, even among Darwin's most ardent supporters, e.g. Huxley and Galton (Darwin's second cousin.)  Many believed that the stuff of significant evolutionary change must rather be large, discontinuous, variations.  This opposition of views — on the one hand, natural selection of small individual differences giving continuous (albeit variable-speed), gradual evolution, and on the other, sudden evolutionary jumps, or saltation, by the selection of discontinuous "sports — was to be the pedal-note of the development of a theory of inheritance until the late 1920s.  Until then, the two sorts of variation were seen as being wholly distinct, rather than representing two ends of a continuum, as we see them today.  At the time one had to choose between them, and theorize accordingly.

The opposition of smooth gradualist vs. sudden saltatory evolutionary change has been re-visited in the relatively recent past in the debate about "punctuated equilibrium," which has greatly animated many evolutionary biologists, and remains a prominent item in modern discussions.  Here is an appropriate point to reiterate the value of some historical understanding of the development of one's discipline, such as we are sketching here, and this can be done effectively by quoting a prominent, historically minded, evolutionary biologist, Ernst Mayr (he was a primary architect of the evolutionary synthesis, and is still active today):

Returning to our narrative, Galton was fascinated by the possibility of quantifying evolutionary change, and he attempted some crude analysis of the fate of population variation.  He was impressed by what he called the "regression" (the original source of the term in statistical analysis) of offspring character values towards the mean of the population (it being one of the main reasons for his eventual rejection of gradualist evolution based on small variations), and he was instrumental in founding (in 1893) a Royal Society "Committee for Conducting Statistical Inquiries into the Measurable Characteristics of Plants and Animals" to study such matters.  This committee came to include Weldon and Pearson (of correlation fame).  These two were fierce supporters of strict Darwinian (gradualist) evolution (as Galton himself was initially, though, as mentioned above, he eventually became a saltationist) and they all saw the study of evolution as a statistical problem, and essentially invented the science of biometry from scratch to permit the study of it. They tried to do things like show correlations between death rates and morphological variations, and show that they varied in turn with environmental differences.

The gradualist : saltationist debate - generating more heat than light

The calm of the late Victorian age was shattered by the sudden flaring of a controversy about the source of evolutionarily-relevant variation as Bateson, disillusioned by his earlier failed efforts with research on gradualism, and having no faith (or skill) in statistical methods, became a fervent convert to saltationism, and vigorously began to criticize "the biometricians."  There followed several heated exchanges at the Royal Society.  Galton sought to defuse the argument by appointing Bateson and other saltationists to the Evolution Committee.  This happened in due course, and blood quickly began to flow (mostly over the pages of Nature) as hostility became open and personal — much of it having little to do with the intellectual issues.  In 1900, Weldon and Pearson resigned from the committee, and Bateson and his fellow saltationists took it over.  One can imagine the bitter feelings as Weldon and Pearson retreated to their lairs, to lick the wounds of the loss of "their" committee, and to found a new journal, called Biometrika (the "k" for Karl Pearson!), wherein to publish biometrical research. 

At exactly this juncture, the works of Mendel were rediscovered and brought to light, mainly by de Vries.  Bateson accepted this new theory of particulate inheritance with alacrity, as it clearly fit well with his discontinuous, saltationist, outlook.  Thus the saltationists became known as Mendelians, and at virtually the same time, Pearson called his science Biometry.  Because the hated saltationists had embraced "Mendelism" with such fervour, the biometricians blindly denounced Mendel's discoveries, even though in a certain passage of his writings Mendel had explicitly suggested, as others were to later (including Punnet, one of the eventual arch-saltationists!), that  some of his particles could probably have even very small, and interacting, effects, thereby providing an explanation for the continuous variations which the biometricians regarded as the stuff of evolution.  How sad.  The reconciliation of particulate inheritance and continuously-varying characters had to wait until 1918 when, as we shall see below, Fisher published an amazing paper (Note 3) synthesizing the two, among many other things.  Even then, the message took a while to filter throughout the whole community of biologists interested in evolution.

This debate became very bitter, absurdly polarized, and confusing, assuming more the style of religious disputation, and it primarily succeeded in delaying the general acceptance of the synthesis that was eventually to become population genetics by 15-20 years.  On one side: the Mendelians— Bateson, de Vries, Johanssen, Punnett et al.;  on the other:  the Biometricians— Weldon, Pearson, Nicholson, Poulton et al.  On both sides there was almost willful misinterpretation of work from "the other side" (we shall see this kind of phenomenon again later in the debates between Fisher and Wright and their followers.)  There were also grand generalizations of firmly-held opinions which extended well beyond empirical results.  For example, Johanssen's work on pure lines of selfing beans nicely demonstrated the characteristics of nearly-homozygous strains:  restricted variation, and lack of response to selection.  But, in his enthusiasm, he rashly generalized this ineffectiveness of selection to hybrid populations, and also to the then almost totally unstudied natural populations, well in advance of the completion of his experiments (he had crossed pairs of homozygous strains and found the F1 populations to be even less variable;  we now know that all plants would have been identical heterozygotes).  He erroneously interpreted the variation which he did see in his pure lines as the continuous characters which the biometricians favoured as the material for selection.  He found that this variation did not respond to selection.  We would say:  no surprise there!  because we now know that he got this result because the variation he was trying to select was, necessarily, wholly environmental in origin.

Slowly, light began creeping in.  Selection experiments carried out on suitably outcrossed stocks finally showed that significant changes could be achieved by selection on continuous characters:  in effect, it eventually became clear that there was no qualitative distinction between discontinuous "sports" and the slight variants ("fluctuations") selectable by Darwinian processes.  The multifactorial nature of many characters became recognised and, following this, the significance of recombination (and its association with sexual reproduction).  It became clear that many of the characters previously interpreted by the Mendelians as discontinuous new mutants were, in reality, simply recombinant genotypes:  their true nature had not been recognised because the number of recombinant types possible was hugely greater than the number of individuals then used in most of the experiments, and so, appeared only rarely - so they seemed like mutants.  Thus became recognised the enormous store of variability potentially latent in sexual organisms:  new mutants became much less necessary.   Better to say that they became much less necessary for developing theory, because, as mentioned, there was precious little information on levels of variability in natural populations.  Altogether, however, it became clear that there was no necessary inconsistency between Darwinism and Mendelism, and that crude saltationism both was not needed, and indeed had some real problems.

This whole gradualism - saltationism debate provides a paradigm of much intellectual conflict in theorizing about evolutionary processes.  It is clear that much of the hotter parts of the argument took place in an empirical near-vacuum.  It is discussed in detail in an excellent and edifying article by John Turner (Note 4), which teaches much about evolutionary thinking and science's content and conduct.

Relationships between notions on among-individual & among-population variation

So far we have been implicitly considering developing notions of the meaning of among-individual variation, but it is a propos to make some remarks about early attitudes to among-population variation, including geographic variation, since it bears on later arguments.  We find that the gradualist - saltationist debate inserts itself here too:  the gradualists, who, as we have seen, were largely naturalists, and therefore experienced in the variations shown by wild organisms, sometimes across their very wide ranges, saw geographic variation as representative of adaptation to local conditions, and therefore necessarily genetically determined, at least in some fashion (directly, or via adaptively plastic response to the environment.)  But the saltationists, with their view of a macro-mutational origin of species, regarded geographic variation rather almost as a nuisance - if species could show gradual variations which sometimes approached the magnitude that commonly separate species, where did that leave them?  Their response was to deny that such variations were either adaptive or heritable - they were asserted to be mere "fluctuations" too, just like the variations shown among individuals within populations.

It is worth noting in this connection that these radically different attitudes to the genetic structure and composition of species also influenced attitudes to the role of geographic isolation in speciation.  For saltationists, there was no need for isolation:  species would spring into separate existence directly upon the arrival of the necessary macromutants, and so this could happen under any circumstance, even sympatrically.  But for the gradualists, with their view of extensive adaptive variability, gene exchange needed reduction so that differentiation may proceed unhindered.  This is the intellectual origin of the idea of allopatric speciation.  With the eventual demise of saltationist ideas, and the triumph of a largely gradualist picture, allopatric speciation took the place of favour, and sympatric speciation models were decidedly put in the back room, but the strength of opposition to them was probably much on account of their historical association with the hated saltationist notions.  The issue of an adaptive vs. a neutral nature of geographic variation has continued to exercise the minds of evolutionists and systematists down to the present day.  It is still an open empirical issue whether, in general, spatial differentiation is directly due to allelic differentiation, or due to adaptive plasticity, or of no adaptive significance.  Certainly, many specific cases have been investigated and allocated to one causation of another, but statements that are globally-true of all characters elude us yet.


Note 1     Dawkins, R.  1986   The Blind Watchmaker.  Longman  (also a paperback by Norton. and by Penguin)  Return to text position
Note 2.     Mayr, E. 1976  Evolution and the Diversity of Life.  Harvard Univ. Press.  Return.
Note 3     Fisher, R.A.  1918  The correlation among relatives on the supposition of Mendelian inheritance.  Trans. Roy. Soc. Edinburgh  52:  399-433  Return.
Note 4     Turner, J.R.G. 1983  "The hypothesis that explains mimetic resemblance explains evolution":  the gradualist - saltationist schism.  pp. 129-172 in Grene, M. (ed.)  Dimensions of Darwinism.  Cambridge Univ. Press.  Return.

Go to next section:  Reconciliation of Mendelian inheritance and Darwinism