Storming the Fortress

To write a history of the world, a man would have to be slightly crazy. He would have to be possessed by some single idea which, he believed, could illuminate the whole course of history, and in the light of which all events could be judged. To write a history of biological thought, particularly one confined to thoughts about biological diversity rather than physiological mechanism, is a less ambitious undertaking, and Ernst Mayr is certainly not crazy. But the task is sufficiently formidable, and Mayr has succeeded at it only because he does indeed have a single, unified vision, which he uses to evaluate scholars from Aristotle to Theodosius Dobzhansky and R.A. Fisher.

His vision is that the history of biology has been a struggle between two world views—essentialism and population thinking. The essentialist view holds that there are a finite number of kinds of animals and plants, each kind characterized by certain essential features which it is the business of the biologist to recognize. These “kinds” correspond to what we today call species. The members of a species share an identical essence. They may differ from one another in various ways, as people differ in height, hair color, and fingerprints, but these differences are accidental and unimportant; they do not alter the essence. The populational view holds that individuals are bound together, not by the possession of a common essence, but by the fact that they interbreed.

The supreme essentialist was Linnaeus. His attitude derived in part from certain features of the real world (which I will describe in a moment) and in part from the philosophy of Aquinas. Linnaeus’s great French contemporary, Buffon, held views that derived from the alternative medieval philosophy of nominalism. He held that individuals constitute the only reality. We classify them altogether arbitrarily in groups to which we give names (hence “nominalism”). By the act of naming species, we create them.

It is interesting that, starting from such diametrically opposed positions, Linnaeus and Buffon were brought closer together as their knowledge of actual diversity increased. Linnaeus never abandoned the concept of essences, but he did abandon his earlier claim that “we count as many species as were created in the beginning.” In its place he put the genus—oddly, because if any category is real it is the species—and suggested that species might have arisen by hybridization. He was driven to this position by the sheer impossibility of characterizing each species by fixed and unvarying essences. Buffon never accepted essences, but he was driven to recognize that the groups into which organisms fall are not arbitrary and man-made. In other words, he admitted that real divisons exist independently of our judgments.

An interesting confirmation of the reality of species, which has only recently been appreciated, is that primitive peoples usually recognize the same species as do modern taxonomists. The process that ensures the uniformity of the members of a species in any one place is that they interbreed; different species remain distinct because their members do not interbreed. The gradual recognition by the naturalists preceding Linnaeus that species are real entities was an important and necessary discovery. However, it did lend support to an essentialist view of the living world, and for that reason hindered the acceptance of evolutionary ideas. For Darwin, the origin of new species was a central problem. Mayr would say that it was the central problem, but I am less sure. I think that for Darwin the most important problem was to provide a natural explanation for the adaptation of organisms to their ways of life. However that may be, Darwin certainly understood that an essentialist view of species was incompatible with evolution.

For both Darwin and Alfred Russel Wallace, a critical step in reaching their evolutionary views was the recognition that populations of a given “species,” if geographically isolated from one another (as, for example, were the animals of the Galapagos Islands), may show varying degrees of difference, from that appropriate to “varieties” of a single species up to a difference sufficient to justify their being placed in different species. If such isolated populations were once seen as incipient species, that spelled the end of the essentialist view. As Mayr says, “The fixed, essentialistic species was the fortress to be stormed and destroyed; once this had been accomplished, evolutionary thinking rushed through the breach like a flood through a break in a dike.”

Essentialism, however, did not die with Darwin. When Mendel’s laws were rediscovered in 1900, the early Mendelians saw themselves as anti-Darwinian. The mutations they were studying usually had striking effects. Each mutation was seen as the origin, at least potentially, of a new species. The relatively minor and apparently “continuous” variations in size and shape observed within natural populations were regarded by the Mendelians as irrelevant to evolution. Only mutation could alter the essential characteristics of a species, and it could do so without need for natural selection. In contrast, the Darwinists argued that the mutations of the Mendelians were monsters doomed to early death, and that minor variations were the stuff of evolution.

The reconciliation of these different views required a recognition by naturalists that the genetic variation they saw in natural populations was caused by Mendelian genes, and by geneticists that evolution was a process brought about by selection operating in natural populations, even though the selection was operating on variations that originated by genetic mutation. The story of this reconciliation is one that Mayr is well qualified to tell, since he was one of the main architects of the “modern synthesis” that resulted. However, even the modern synthesis has not seen the end of essentialist thinking in biology. Its latest manifestation is in the “punctuationist” theory of the paleontologists Stephen Jay Gould, Niles Eldredge, and Steven Stanley, whose views are reminiscent of those of Hugo de Vries and early Mendelians, but rest on a misunderstanding of the fossil record rather than of mutation.

This, reduced from 800 pages to as many lines, is the picture Mayr paints. How far does he succeed? Magnificently, I think, but perhaps not quite in the task he set himself. In his introduction, he explains that he is attempting a history of how scientists have tackled problems. He remarks on the need to avoid writing a Whig history of science, but that is the kind of history he has written. To be fair, I cannot imagine how a man who has striven all his life to understand nature, and who has fought to persuade others of the correctness of his understanding, could write any other kind of history. Indeed, I am prepared to argue that there are worse kinds of history of science to write than Whig histories—for example histories of the kind written by Gertrude Himmelfarb, based on a misunderstanding of the science whose history is described. After all, if Victorian England really had been the highest peak of civilization yet reached, and if it really had held in itself the guarantee of continued progress, Macaulay’s method of writing history would have had much to recommend it.

Unfashionable as it may be to say so, we really do have a better grasp of biology today than any generation before us, and if further progress is to be made it will have to start from where we now stand. So the story of how we got here is surely worth telling. Of course, we may not all agree about exactly where we are; indeed, as I will explain in a moment, I do not fully agree with Mayr. More important, a Whig history will seldom give the reader the pleasure that comes from a sudden and unexpected glimpse into an unfamiliar mind from the past. Mayr’s history affords a different pleasure: an understanding of a powerful and creative mind of the present, and of its intellectual roots.

In saying that this is Whig history, I risk being misunderstood. Mayr has not peopled the past with tailor’s dummies, dressed up to espouse particular causes. He has read his sources and strives to understand them. Again and again, he told me things about the history of my subject I did not know. Some examples may help. I had always pictured Aristotle as an essentialist. After all, where else did Aquinas get his ideas? Mayr has persuaded me that Aristotle explicitly rejected the essentialist approach to animal classification. I have thought, and even told my students, that the debt Darwin owed to Sir Charles Lyell was the concept of uniformitarianism. I am now convinced that Lyell’s uniformitarianism was a hindrance rather than a help, and that his real gift to Darwin was that he asked the right questions, even if he usually gave the wrong answers.

To give a third example, I have long been puzzled by the fact that, after the publication of the Origin of Species, the major advances in biology for the next fifty years took place on the Continent, particularly in Germany, and not in England. It is a strange echo of the way in which the fruits of Newton’s Principia were reaped in France. I once earned the deep hostility of Julian Huxley by suggesting that the fault lay with T.H. Huxley, who persuaded biologists in Britain and America that, after Darwin, the main task was to elucidate phylogenies by the methods of comparative anatomy, whereas in truth it was to develop a theory of heredity. I still have a soft spot for this view, but Mayr convinced me that a more significant reason was the earlier development of professional biology in German universities.

Mayr, then, is fair to his sources. My complaint, if that is the right word, is that he never reveals to me how anyone could have been so unobservant and illogical as to hold essentialist views. He knows that essentialism is synonymous with sin, and so do I. But surely men like Linnaeus, Georges Cuvier, William Bateson, and T.H. Morgan were not born in sin. They must have had some grounds for holding the views that they did. It is striking that essentialism is the automatic, unconsidered philosophy of every physical scientist. There exists the hydrogen atom, and no doubt someone, maybe Schrödinger, wrote down the equations that describe its essence. If essentialism works for them, why doesn’t it work for us? Am I quite sure that their way of seeing things can never contribute to biology?

Perhaps I would not be so conscious of these reservations if there were not a small part of Mayr’s world view that I do not share. Part of his picture is that the naturalists were usually right, and the mathematicians always wrong or, at best, irrelevant. He has a strong case. Darwin was a naturalist, as was Wallace, and he referred mathematical questions to his wrongheaded cousin Francis Galton. I was fascinated to learn that even August Weismann—for me, the greatest evolutionist after Darwin—was a lover of butterflies. Nevertheless, I think that mathematics is crucial for further progress in evolutionary biology. Mayr is on the whole generous to the mathematical geneticists R.A. Fisher, J.B.S. Haldane, and Sewall Wright (he adds the name of S.S. Chetverikov), who showed that Mendelian inheritance and Darwinian selection are compatible. But his treatment of evolutionary biology since 1950 seems to me inadequate, because it misses the central role of mathematics in analyzing, for example, the evolution of breeding systems, of social behavior, and of molecules. Mathematics without natural history is sterile, but natural history without mathematics is muddled.

The issue is best illustrated by Mayr’s attitude to what he has called “bean-bag genetics.” Population geneticists often consider the fate of a single gene at a time—or rather, of alternative forms (“alleles”) of a single gene—and ask under what circumstances a given allele will increase in frequency in a population. Mayr objects that selection acts on individuals, not on genes, and that individuals, not on genes, and that individuals are the product not of one gene, but of complex sets of “co-adapted” genes. This seems to me to be both true and largely irrelevant. A particular gene will increase in frequency, or not, depending on the effects it has on individual fitness, against the background of all other genes present and the environments experienced. It has to be a “good mixer.” But the nature of genetic transmission is such that each gene which increases in frequency must do so either by chance or on its own merits, and that is what population geneticists assume.

There are, however, two points of view, which one could almost call the “English” view, deriving from Fisher, and the “American” view, deriving from Wright. To oversimplify matters somewhat, Fisher thought that each substitution of one gene for another in evolution occurred because it was beneficial, on its own, and that the role of chance events (other than mutation) was slight. Wright thinks that, often, several gene substitutions would be beneficial if they occurred simultaneously, but that each by itself would be harmful. If so, the only way the change can take place is by chance in a small local population. In effect, the English think that evolution is a hill-climbing process, and the Americans that it also involves jumping across valleys.

As a student of Haldane’s, I take an impartial view. However, both views are essentially reductionist, and both were first formulated mathematically. There is only bean-bag genetics.

These are relatively trivial disagreements. Yet it is characteristic of Mayr’s book that the emotion it arouses in me is a wish to argue with him, not about history, but about his scientific views. Essentially, if I dare use the word, that is what his book is about.

François Jacob’s contributions to biology have been in a field as distant from Mayr’s as they could be. With Jacques Monod, he discovered how the genes of a bacterium are regulated. He is also the author of The Logic of Life, perhaps the only book that could stand comparison with Mayr’s history of biological ideas. The Possible and the Actual is a printing of Jacob’s Danz Lectures at the University of Washington, Seattle. Although Jacob is a microbiologist and Mayr a bird taxonomist, their views are astonishingly similar. For both, Darwinism is the center of biology. Both show how grotesque it would be to teach biology without teaching evolution.

In “Evolutionary Tinkering,” Jacob discusses the consequences of the fact that evolutionary adaptations arise, not de novo, but as changes in pre-existing structure. Another essay, on “Time and the Invention of the Future,” is concerned with the evolution of aging, and with our perception of our environment. I was most intrigued by the first essay, on “Myth and Science.” I also gave the Danz Lectures—an admirable institution—and I chose “Science and Myth” as the title of one of them. The similarities between what we had to say is striking. I am sure that we are both reacting to the same curious fact, that Darwinism seems invariably to be treated as myth.

Scientific theories and myths have much in common. To quote Jacob, “Myths and scientific theories operate on the same principle. The object is always to explain visible events by invisible forces, to connect what is seen with what is assumed,” and, “Scientific investigation begins by inventing a possible world…. So also begins mythical thought.” There is, however, a crucial difference: “A myth is not just a tale from which inferences can be drawn about the world. A myth has a moral content.” Darwinism has suffered, and continues to suffer, because people, imagining it is a myth, insist on drawing from it moral consequences. As Jacob remarks, “It is not clear that the theory of evolution can also fulfill this function, in spite of various attempts.” The habit of confusing myths and scientific theories is understandable but dangerous. The hard question, of course, is whether we can get on without myths, and if not, in what sense we have to believe them.