Gregor Mendel
Gregor Mendel; drawing by David Levine


The catalog of Harvard’s Widener Library lists 184 books about Charles Darwin, his life and work (not counting 172 volumes of self-produced letters, autobiography, and scientific opera). On the subject of Gregor Mendel, there are only seventeen. The same disproportion is reflected in the books I have before me. Darwin is represented by a 702-page collection of letters all written before the age of twenty-seven, and a 449-page biography and subsequent history of the idea of evolution written by a professional biographer with no special expertise in the subject. When I contemplate yet another book about Darwin and Darwinism, I feel a bond of sympathy with the philistine Duke of Gloucester, whose reaction to a second volume of The Decline and Fall was, “Another damned, thick, square book! Always scribble, scribble, scribble, eh, Mr. Gibbon?” For Mendel on the other hand, the services of Vitezslav Orel, a great authority who has spent more than twenty-five years in historical research on the subject, have been obtained to produce a mere one hundred pages as part of a series of lives of the intellectual saints running from Aquinas to Wyclif.

As a population geneticist professionally concerned with Mendel’s mechanism for the inheritance of variation and with Darwin’s theory of evolution by selection of that variation, I have long found the vast disproportion in interest between the two to be paradoxical. While several explanations come to mind, none seems sufficient.

First, it might be argued that Darwin’s popularity on the intellectual market is a classic case of consumer sovereignty. People are greatly concerned with the place of human beings in the universe, so the materialist theory of evolution continues to agitate and fascinate all concerned. After all, the first printing of On the Origin of Species was immediately sold out, and interest has hardly died out since, as evidenced by the legal and journalistic trials still in progress in America. But the preoccupation of the literate middle classes and the fundamentalist masses with human uniqueness cannot explain the behavior of biologists, historians, and philosophers. While the hundredth anniversary, in 1959, of the publication of Origin of Species and the centenary, three years ago, of Darwin’s death were the occasions for large numbers of international symposiums and their attendant publications, the 1965 centennial of Mendel’s paper was lightly commemorated except in Czechoslovakia, and the centenary last year of his death went completely unnoticed by the institutions of science. The Journal of the History of Biology would have to close its editorial offices if it were not constantly supplied with more and yet more about Darwin, but the Folia Mendeliana, almost a one-man industry of Dr. Orel’s, appears only annually and is hard to find. In recent years, philosophers of science have abandoned physics for the richer and more complex domain of biology which, God knows, needs their help, but they have almost all taken Darwinism as their focus of interest. The deep epistemological problems in heredity and development have been left largely to the philosophical naïfs who practice the science.

Second, it might be claimed that Mendel’s discovery was intrinsically less interesting, especially from a philosophical point of view, than Darwin’s. The uncovering of the actual mechanism of heredity might be terribly important, but it is only a question of the mechanics, of particular gears and levers. But precisely the same can be said of Darwin. Although Origin of Species made evolution popular, Darwin certainly did not invent the idea. Indeed, a good case is made by L.J. Jordanova that, if any biologist should be considered the father of evolutionary theory, it is Lamarck. Most French intellectuals have regarded the Anglo-Saxon infatuation with Darwin as a typical piece of chauvinism. Darwinism is, if anything, a particular mechanism for evolution. That mechanism is the differential rate of reproduction, under pressure from the environment, of different sorts of individuals within a population. Moreover, the success of Darwin’s mechanical explanation of evolution depends critically on Mendel. Had heredity turned out to have a fundamentally different basis, Darwin’s idea, ingenious though it was, would have been wrong.

The problem is that natural selection among variant types causes the population to lose variation as the superior type comes to characterize the species. That is, selection destroys the very population variation that is the basis for its operation. Evolution would then soon come to a stop if there were not some continued source of variation among individual organisms. If heredity takes place by a blending mechanism, either by the mixing of blood or other fluids, then any new variation that arises would be immediately diluted out by the process of mating and the production of intermediate hybrids. Darwin was acutely conscious of this problem of the loss of variation from blending inheritance and the constant need for new sources of variants. In later editions of the Origin, he allowed for the possibility that heritable variation could be directly induced by environmental action. That is, he took in Lamarck’s view that acquired traits could be inherited, which is fatal to the whole Darwinian project of explaining evolution by a variational rather than a transformational mechanism. Mendelism saved the day.


The central core of Mendelism is the distinction between the appearance of an organism (the phenotype, in modern jargon), which may indeed be a blend of the characteristics of its parents, and the physical state of the factors inherited from each parent (the genotype), which remain physically discrete and unmixed. Just as Seurat’s Grande-Jatte gives the appearance of blended pigments from a close juxtaposition of small dots of pure color which are then visually fused by the physiology of the observer, so the physiology of development fuses, at the level of the whole organism, the pointillism of heredity.

Mendel’s realization of this distinction came from his experimental crosses with garden peas. When he crossed two truebreeding varieties that differed markedly in some characteristic, say, flower color, the offspring were uniform in appearance, which is precisely what one would expect from a mixture of two varieties. In Mendel’s case, there was the minor complication that the offspring all resembled one of the two parents rather than being intermediate between them, but this is the exception rather than the rule in most organisms. Thus, when Mendel crossed red-flowered and white-flowered garden peas, the offspring were all red-flowered. Had he worked with the sweet pea, Lathyrus odoratus, rather than the edible pea, Pisum sativum, the offspring would all have been pink. Whatever the color of the offspring flowers, the uniformity among individuals is precisely what one would predict from simple notions of the mixing of heredity. One would also predict that, if the uniform hybrids were crossed with each other, they would once again produce uniform offspring, and so on, without end.

But that is not what happened. When Mendel crossed these uniform hybrids with each other, he recovered in the next generation some plants with white flowers, like one of the two grandparents. From the reappearance of grandparental characteristics, apparently uncontaminated by their passage through the hybrids, and from the exact and repeatable ratios of types appearing among the offspring, Mendel constructed the two principles of heredity—principles that Darwinism needed to make it a workable theory. First, the factors that are passed from parent to offspring in heredity—what we now call genes—are particulate and maintain their individuality despite their interaction with other genes in the development of an organism. That is, the physical basis of heredity is discrete, like the elementary quantum of physics, rather than continuous.

Second, in the process of the formation of sperm and eggs in a hybrid organism, the genes that have been mixed together in that hybrid detach from each other and are parceled out to separate sperm and egg cells. That is the principle of segregation. Those two principles guarantee that if different variants in a population mate, even though their immediate offspring may be uniform and intermediate between the parents, in later generations the variation will reappear as a consequence of segregation. Thus new variation will not be submerged and diluted by the process of mating but will always be available for selection. Mendel’s principle of segregation is the rock on which the theory of evolution by natural selection is built.

The real epistemological revolution wrought by Darwin was, in fact, identical with that created by Mendel. That identity can best be seen in the contrast with Lamarck, who was concerned with the problems both of evolution and of heredity. Jean-Baptiste Lamarck was in many ways characteristic of the intellectual movement of the French Revolution. He was a deist, he accepted La Mettrie’s homme machine, rejecting the soul, and assumed that material principles underlay all natural and human phenomena. He combined with his materialism, however, the eighteenth-century commitment to natural philosophy, to the principle that all of nature reflects a few general organizing principles. Jordanova discusses these principles and illustrates how they bore on the biological problems of evolution, taxonomy, and heredity, but, in her treatment, their origin remains mysterious. They seem a priori, or at least broad generalizations from a small base of observations. Some of the organizing principles that Lamarck espoused, such as the effect of use and disuse of organs, certainly hold across a reasonable domain of phenomena. Muscles do atrophy if they are not exercised and bones do grow larger and thicker at points where muscles are attached and produce tension. But brains do not atrophy with disuse and an important current theory of neurobiology (see below) actually maintains the opposite.


Other of Lamarck’s principles, like the inheritance of acquired characteristics, were simply a priori or based on unexamined tradition. So giraffes’ necks may indeed grow a bit longer if they stretch them to reach the tops of trees, but that change is not passed on to future generations. It took Darwin to see that giraffes that happened to be born with long necks were better able to survive than those without them. What Lamarck had in common with all natural philosophy was a typological view of phenomena, what Ernst Mayr has called “essentialism.” In general, this meant that the ontological sources of similarity between things were seen as different from the ontological sources of differences. In particular, all members of a species were held to share unalterable properties that were intrinsic to the organisms, while differences between individual members were accidental consequences of environmental modification and were subordinate to the constant features. The problem of understanding the similarities was seen as fundamentally separate from the problem of the origin of superficial differences. It was the abolition of this distinction between the ontological sources of similarity and of difference that marked the epistemological break of Darwin and Mendel.

As I have argued in some detail in a previous review in these pages, 1 Darwin changed the object of study in evolution from the type of a species to the actual variation among individual organisms within the species. The motive power for a change in the average properties of the species was in the differences from the average displayed by the organisms themselves. Thus typical differences between species in space and in time arise by the accumulation of differences that were already present as variation within a species at any one place and time. But precisely the same contrast of similarities and differences permeated the study of heredity. Before Mendel, all studies of inheritance took heredity, that is, the passage of similarity between parents and offspring, to be different from, and antithetical to, the phenomenon of variation between individuals. The object of study was not the individual organism and its variant properties but the average or collective description of groups of progeny. For the predecessors of Mendel, the appearance of white-flowered plants among the progeny from the cross of the red-flowered plants implied a source of differences between organisms that was itself different from and obscured the action of the forces of heredity. For some, it was evidence of environmental malfunction. For others, it was the consequence of a poorly specified “force of atavism.” It remained for Mendel to use the very occurrence of different types among the offspring of a single cross as the key to the laws of inheritance.

It is sometimes said that Mendel’s unique contribution came because he was an experimentalist, or that he worked with favorable material, or that he had the sense to count the numbers of different types that came from his experimental crosses. But none of these is the critical element. Alexander Seton and John Goss in 1822 and Thomas Knight in 1823 had already observed segregation of green and pale seeds in second generations of pea crosses, Mendel’s very material. Louis Vilmorin in 1856 counted the results from individual crosses and even reported three-to-one ratios of the two original parental types in the progeny of hybrids, the ratios upon which Mendel built his theory of the segregation of particulate factors in the formation of sex cells. Darwin observed “Mendelian” ratios in snapdragons but drew no conclusions. Even he did not realize that the variation among sister plants from the same parents was the proper object of study in heredity.

What Mendel understood, and what was not realized again for thirty-five years after his paper was made public, was that heredity and variation are two aspects of the same phenomenon and that only by a study of the actual variation among members of the same generation can we understand the passage of similarity across generations. This synthesis of the antithetical properties of heredity and variation is a Hegelian’s dream and represents as difficult and subtle an insight into nature as any in the history of science.

A third reason for the vastly greater concentration on Darwin is simply that there is just a lot more information about him than there is about Mendel, so, of course, there is that much more grist for the academic mill. Darwin gave us an autobiography composed after he was already a very famous person, while Mendel produced only a curriculum vitae meant to support his application for a secondary teacher’s certificate. There are hundreds of Darwin’s letters, both personal and scientific, to scores of different recipients, including leading scientific figures, and he spent a couple of hours every day on his correspondence. Mendel is represented only by ten letters to the botanist Karl Nägeli, and a handful to his mother, sister, brother-in-law, and nephew, all of which were already available to Mendel’s biographer, Hugo Iltis, in 1924.2 While Darwin’s sketches and notebooks remain as a rich source for historians, nearly all of Mendel’s notes and papers were burned at his death in 1884, leaving us only a few reports to scientific and administrative committees on which he served, and the manuscript of two metaphysical poems written by the future priest when he was a schoolboy. In the absence of materials, what is there for a historian to do? A good deal, in fact, once the great-man theory of history has been firmly put aside.

Although it has been a long while since political historians have abandoned the Suetonian ideal of history as biography, intellectual history has continued to concentrate on the individual genius as its proper focus. Did Darwin come to the idea of differential survival and reproductive success of units of different adaptive efficiency out of his head, in a true epistemological break? Or did he come to it for reasons external to scientific reasoning—for example that his income was largely derived from stocks (largely railroad shares) which he actively traded and whose rise and fall he followed daily, and with considerable care, in the newspaper?3

This struggle between internalist and externalist schools of historiography has not really changed the emphasis on individual genius because the argument has been about the sources of influence on the individual mind rather than on the central structure of causation. Even if external factors were predominant, we still call the modern theory of evolution “Darwinism” not “share marketism.” Whatever the source of influences, historians do not see the great mind of intellectual history as a mere passive nexus of external forces, but as the critical and central element in invention. Even the selfconsciously historicist Marxist J.D. Bernal made individual scientists the leading actors in his recounting of the history of science.4 Had Darwin not recovered from his attack of scarlet fever at the age of nine, would we be deprived of our understanding of natural selection? Well, not quite, since there was Alfred Wallace, there was Edward Blyth, and perhaps others that we know not of.

The standard view of Mendel makes him even more remarkable than Darwin. The son of a Moravian peasant freeholder, Mendel spent his entire life, except for a brief period of study in Vienna, in the provinces. Olomouc and Brno were not London and Cambridge. The Augustinian monastery where he lived and worked, and the secondary modern school in which he taught physics, housed a number of intellectuals including a former professor of mathematics and physics at Lemberg (now Lvov) who had been the victim of the McCarthyism of the 1850s. But none was such an eminent scientist as Lyell or Hooker, nor was there the great mass of active scientific workers in the mainstream of natural history like those with whom Darwin was in constant communication. The only scientist of note with whom Mendel corresponded was Nägeli, a contact he made only after he had completed his famous research on peas. Indeed, the lateness of his exchange with Nägeli was Mendel’s great luck. The eminent professor induced the amateur to take up a line of work with the hawkweed that cost Mendel five years of frustration, strained his weakening eyesight, and was bound to fail because of a peculiarity of the sex life of that species of which Nägeli and Mendel were unaware.

Mendel entered the monastery at Brno in 1843 at the age of twenty-one because he was hard up and saw no other way in which to complete his education and to become a teacher. The standard picture of the Königinkloster in Mendel’s time is that of a sheltered congregation of amateur intellectuals where the monk, “alone, or in converse with the sage and tranquil fathers…roamed through the monastery garden.”5 Fortunately for Mendel, the monastery was presided over by the Prelate Napp, “a man of large views, [who] was delighted that the institution under his care should become an intellectual center” where “almost all the inmates…were engaged in independent activities, either scientific or artistic.”6 On reading, as a young student, Iltis’s description I was tempted to enter an order myself. A lack of space and the prelate’s benign multiplicity of interests, it seems, forced Mendel to carry out his experiments on peas in a cramped garden plot assigned him by the abbot next to the monastery wall. Only when he himself became prelate in 1868 could Mendel enlarge his experimental domain. Thus we are led to a picture of a genteel intellectual atmosphere, encouraged, but within moderate bounds, by a “large-minded” abbot, the perfect bed for the germination and growth to maturity of Mendel’s intellectual embryo.

In large part through the studies by Orel and others associated with the Mendelianum in Brno, this picture has changed drastically in the last twenty-five years. Moravia at the beginning of the nineteenth century was a rich agricultural region, especially important in sheep, fruit trees, and wine production. The revolutionary belief in the power of science to transform nature and promote economic and social progress was put into early practice in Moravia, largely at the instigation of Christian Andreé, economic adviser to Count Salem. Together they founded the Agricultural Society, which pursued research in sheep breeding, pomiculture, and viniculture. A Pomological Association was begun in Brno as a branch of the Agricultural Society and, by the time Napp became prelate of the Königinkloster in 1825, breeding research in the region was an important activity.

Almost immediately, Napp established a fruit tree nursery in the monastery grounds and he wrote a manual on growing improved varieties. By 1830, the monastery was described as a “research establishment” for vine breeding. Napp was president of the Pomological Association. He successfully fought off an attempt to close down the monastery and replace its population with a contemplative order that would pay proper attention to the kingdom of God rather than the domain of Ceres. Napp recruited Mendel into the monastery by asking the professor of physics at the Olmütz (now Olomouc) Philosophical Institute to recommend one of his students to be a novice, and later sent his recruit to Vienna for further study in physics.

Far from grudgingly restricting Mendel to a narrow garden plot for the pursuit of his intellectual hobby, Napp built a greenhouse in 1855 in which Mendel and two research assistants from among the monks also worked. Plans for a much grander glass conservatory, never built, have survived. The monastery had a library of 20,000 volumes and an immense herbarium. Mendel was a founder of the Brünn (Brno) Society of Natural Science, to which he gave his famous paper on the laws of inheritance in 1865. When he succeeded Napp as prelate in 1868, Mendel continued the research work of the monastery. He had extensive and carefully designed experimental beehives built and was a founder of the Apiculture Society. He was a founding member of the Austrian Meteorological Society and spent many years in active weather observations which he regularly communicated. We recognize in Mendel the nineteenth-century version of the professional research scientist who, at the same time, as department chairman is in constant conflict with higher authorities on questions of budget and recruitment.

Thus the proper domain of historiography of Mendelism includes all of the scientific Enlightenment in Moravia, the structure and politics of local and regional scientific societies, an understanding of the role of monasteries and of the Church as the controller of such institutions as the Brünn Philosophical Institute. As biography, historiography must be as concerned with the life and works of André, of Napp, of Franz Klácel, Mendel’s utopian socialist fellow monk, a Hegelian who ended his days agitating in the Czech community of Illinois, as it is with the life and mind of Mendel himself. The difference between this community of institutions and personalities and the one in which Darwin was embedded was one of scale, wealth, and world influence. Britain, as the leading industrial and imperial power of the nineteenth century, also had a large positive balance of intellectual trade. Moravia was a net importer of ideas and, although the Brünn Society of Natural Science exchanged its publication with 130 other learned societies, it was to the Proceedings of the Linnean Society of London that one looked for the latest breakthrough. Modern historians of science seem still to be dazzled by Victorian values.


Despite the growing dominance of externalist views of intellectual history, we remain preoccupied with the biographies of individual intellectuals. This attention to the lives and loves of great creators is, in part, rational. It is, after all, people and not societies who think. Mental images, concepts, trial solutions to problems, abstract orderings of the world are the proximate result of physiological processes that go on inside particular human beings. At the same time, of course, the social and natural world in which those beings are embedded are conditioned by and condition those individual thoughts. The formation of an idea is the individual transformation of a social condition by a material organism that is itself the product of individual and social conditions. The material basis of that transformation is the most difficult and most seductive problem in biology.

The problem for biologists trying to understand the human central nervous function is that they cannot, like the cardiac physiologist, proceed to the problems at hand without bothering their heads about abstruse philosophical problems. Any sensible study of integrated nervous function must come immediately to the mind-body problem that has been the perpetual agony of epistemology.

There is no unique way to describe or study a natural object. We begin always with a problem that sets the conditions of our description. A description of the heart and blood vessels that ignores their function in the circulation of the blood would be possible but quite beside the point. Unlike Aristotle, we do not believe that the brain’s function is to cool the blood. Thoughts, feelings, perceptions are some kind of reflection of integrated central nervous activity, so whatever our theory of the brain, it must be so constructed that it helps us to think about thinking. Thus we cannot avoid, from the very beginning, the problem of the relationship between three pounds of flesh and the quality of mercy.

The reaction of neurobiologists to a confrontation with the mind-body problem has been largely to find one or another escape route. The molecular neurophysiologists can avoid the issue by concerning themselves entirely with the microscopic anatomy, chemistry, and physics of nerve conduction. What is a nerve impulse and how does it get from here to there? The elegant solution to this problem, which now seems more or less entirely in hand, depends only on acceptingDescartes’s machine model of the body without giving the slightest consideration to his dualism of body and mind. That is, one restricts one’s questions to the domain where materialism is unchallenged. At the other extreme, we can, like Sir John Eccles,7 give up materialism, accepting Descartes’s bête machine, but rejecting La Mettrie’s homme machine, thus giving the mind an independent existence deriving from some nonmaterial source. It is a rare modern biologist, however, who is willing to make his or her peace with the soul. Many, if not most, subscribe to epiphenomenalism, a kind of backdoor dualism that gives primacy to the physical state of the brain as the cause of the mental which is, in itself, not causally efficacious. The steam engine and its whistle are the favored metaphor of this school.

The problem with epiphenomenalism is that it cannot cope with the evident effect of mental states on physical objects. Lauren Bacall’s invitation to Humphrey Bogart—“If you want anything, all you have to do is whistle”—would not work on a steam engine. Finally, one can deny the existence of the problem altogether by denying mind and claiming with J.B. Watson and B.F. Skinner that the concept of mind is hopelessly metaphysical and that only behavior exists. Precisely how the truth tables of Whitehead and Russell are to be described as behavior is not clear, unless one allows that they represent mental behavior, in which case we are back in the soup.

The only coherent materialist position seems to be that the mental and the neural are simply two aspects of the same material physical state. Mind neither causes a physical state of the brain nor is caused by it, since cause and effect do not apply to two aspects of the same state. This view conceives the program of neural research to be one of establishing the mapping of physical and mental onto each other. What configuration of neuronal connections and circulating currents corresponds to my mental image of Gregor Mendel? How did it get that way? What features in common does it have with your image of Mendel or, for that matter, with my image of Mendel last week?

While such a dual-aspect view of the mental and the physical has been expounded before,8 it has not been so coherently and lucidly related to the newest facts of neuroanatomy and physiology as in the remarkable book of Jean-Pierre Changeux, Neuronal Man. Reading Changeux’s book I had the sensation of watching him come toward me across a boggy moor, sure that at any moment he would step off the firm path and be swallowed, like his predecessors, in one or another quaking mire. Suddenly, almost to my envious disappointment, there he was beside me, smiling, never having made a false step. He begins with a description of the facts of neuroanatomy, the parts of the brain, and what is connected to what. As I read his description of the columns of brain tissue within which all the cells respond to stimulation of the same tissue, say skin, or bone joints, or to the left or right eye, I thought, “Now he will try to tell me that the brain is divided neatly into slabs, one for each function.”

But instead, Changeux points out that the slabs, or modules, differentiating skin from joint sensitivity run at right angles to, and are crossed by, those that correspond to right and left eyes. That is, the very same bit of tissue may then correspond to two quite different sensory phenomena, just as the same stretch of DNA can code for two quite different proteins, depending upon where the reading frame begins. “Well,” I thought, “he’s got out of that one, but, reductionist that he is, he will misstep next time.” But he never does. The details of the physics and chemistry of nerve conduction are given only as a way of providing a firm material basis for further argument.

The most dangerous terrain is in the description of specific molecules that can be used to induce rage, pleasure, orgasm, pain when acting on particular regions of the brain. The temptation is to claim that there is a specific chemical label for each kind of behavior and sensation. But Changeux shows that such is not the case, that “there is no transmitter for anger or pleasure, any more than there is an isolated center for them.” At every point, Changeux counterposes the evidence for specific localization of particular function with evidence for diffuse control and interaction of parts. His message is that organization is everything, but that the organization of the brain is not simple.

The heart of Changeux’s argument about body and mind concerns mental objects. Using empirical evidence, he argues that perceptions that come from external stimuli, say the eye, are the flows of nerve impulses among well-defined groups of cells. A flow structure among nerve cells is formed when I perceive, say, a tree. If at some later time I form a mental image of that tree, what is happening is a similarly structured flow of nerve impulses in some other group of cells with a similar topology of connection. It is the similarity of that topology which allows me to identify the two mental images and to “see” the tree in my “mind’s eye.”

It is here that Changeux comes closest to a simple (although perhaps correct) physical reductionism. To say that two things are topologically similar, we may mean very directly that there are two physical structures whose three-dimensional pattern of nodes and connecting links is geometrically the same, even though they may look superficially different, rather like two alternative sets of house wires that will allow one to turn on the same set of lights with the same combination of switches. Changeux’s topological similarity is of this literal physical kind.

We might, however, mean by topological similarity, logical similarity based on totally different wiring diagrams that are physically quite incompatible, in the sense that two very different computer programs designed for completely different kinds of computers can both carry out the same computational task. It is entirely possible that two mental images may be computationally the same without being topologically similar physically. Indeed, the computer metaphor for the brain has had a powerful influence on studies of the central nervous system, and the science of artificial intelligence that exploits this metaphor is a major mode of investigation. Yet the relation between Changeux’s wiring-diagram picture and the computer metaphor is not that of simple alternatives between which a decision must be made.

There are no “computer models” of the brain, at least none that anyone would take seriously. That is, there are no hardware models of the brain in which there are biological equivalents of random access memory, central processing units, error-checking circuits, and cathode-ray tubes. There are, however, “computational” or “computer program” models of mental processes which are then set working in actual physical computers, but everyone agrees that these are so-called simulations of actual thought processes. The problem is that we cannot test or reject the hypothesis that thought processes are like computer programs because there is no hypothesis to test. Any set of propositions that can be verbalized or symbolized, any picture that we can draw, any process of choosing among alternatives, of classifying and recognizing, can necessarily be represented by a computer algorithm after the fact. Computer programs are a form of abstract thought and have no independent existence. They can always be made congruent with our mental processes as they come into our conscious minds. Thus the purpose of computer simulation of mental processes cannot be to find out how brain and mind are related, but to systematize our understanding of mind, and so provide conditions that must be met by our picture of the brain.

How the system of mind is then mapped onto physical structures is a quite different question, the question that Changeux claims to illuminate. If we are lucky, Changeux will be right, and logical topologies will be reflected in physical congruences. Unfortunately, past experience suggests that things will be more complicated. When the organization of genes of related function was first studied, it appeared that there was a one-to-one correspondence between their physical order along the chromosome and the order in which the enzymes specified by them worked in metabolism. It appeared also that, at the front end of each lineup of genes, there was a special controlling region that turned genes on and off in response to environmental demands. Changeux himself made a key contribution to the building up of this picture.

We now know that this simple physical linearity is not the general rule but that there are many different physical arrangements of related genes and many different forms of control of their function, including the simplest arrangement first discovered. The biologist is constantly confronted with a multiplicity of detailed mechanisms for particular functions, some of which are unbelievably simple, but others of which resemble the baroque creations of Rube Goldberg. As François Jacob pointed out, evolution is less a sophisticated engineer than a homeworkshop tinkerer. It seems likely that the physical organization of mental processes will bear the stamp of this natural-historical bricolage.

A good deal of Neuronal Man is taken up with the problems of the development of the set of nerve connections that underlie our memories and perceptions. Changeux makes a convincing, if not absolutely clinching, case for the selectional theory now fashionable. Turning Lamarck’s principle of use and disuse on its head, Changeux argues that an unstimulated nervous system makes very large numbers of random multiple connections which remain labile for various periods of time. Experience, i.e., stimulation of the nervous system, in this view, causes differential elimination of various of these multiple connections, leaving in place only those that form a coherent structure. In the absence of stimulation, random death of cells and rupture of connections will leave the pathways in a permanent state of disarray. That is why we can only learn to speak if we hear others speak, and why after a certain age such learning is impossible, as in the case of “wild children” who never acquire language. The analogy is to a photographic image in which certain silver grains are fixed by exposure to light, the remainder being washed out in the developing process. In fact, however, the development of the brain cannot be purely selectional, since new neuronal connections are being made all through life, and their formation is, in part, stimulated by sensory experience.

With the selectional theory of mental development, neurobiology recapitulates a historical tension between the two modes of explanation for the fit of organisms to their environments. For nineteenth-century biology, the question was how species in the process of their evolution come to fit the demands of the environments in which they find themselves. The transformational theory that we associate with Lamarck, the theory that environment leaves its direct impress on each organism, gave way to Darwin’s variational theory of the differential survival and reproduction of already existing variation. In the twentieth century, this struggle of explanations was first replayed in the field of immunology.

When our bodies are assaulted by foreign substances—bacteria, ragweed pollen, flu viruses, dog hair, bee sting toxin, or a transplanted heart—we produce specially adapted protein molecules, antibodies, that attach themselves to the invaders and make them liable to destruction. By and large, this ability to form antibodies is good thing, for without it we would soon die from infection, as do the sufferers from Acquired Immune Deficiency Syndrome. The puzzle for biologists has been how we can produce on demand the repertoire of thousands of differently shaped antibody molecules, each neatly fitted to the molecular shape of a different invading substance. One theory, the “instructional,” held that the invading substance itself served as a kind of die stamp,impressing a complementary shape on a malleable generalized antibody molecule. The alternative, “selectional” theory, was that, before being challenged, our bodies already possess an immense array of differently shaped antibody molecules, all in low numbers, and that invasion by the foreign substance results in a rapid increase in the rate of production of just that antibody that matches the intruder.

During the last twenty years, the selectionist view has won out, in large part because of discoveries in molecular biology that have clarified how such an immense variety of antibody molecules can be produced spontaneously. One of the major contributors to the molecular biological vindication of the selectional theory was Gerald Edelman of Rockefeller University, who turns up, together with Changeux, as a principal proponent of the selectional theory for mental development. If he is right, he will wind up shaking the hand of King Carl Gustaf for the second time. Neurobiology has long held to the Lamarckian instructional principle without great success. Darwinism may now serve it better.

Changeux’s successful traverse of a tricky terrain leaves much of it, perhaps the most interesting part, unvisited. If mental images are ordered structures of electrical oscillations among nerve cells, how do we account for the passage of our attention from one mental shape to another? Changeux treats consciousness as if it were simply the opposite of unconsciousness. But the heart of the problem of mind and brain is the shift of consciousness by what appears to us to be a willful act. As I tire of writing, I think first of the impending visit of a friend, then I strain to hear which Scarlatti sonata my wife is practicing, and then I return again to think about the relation of ego and mental images. I have passed among three very different mental states all under the control of the willful “I.” Some kind of information about all these states must all the while have been resident in my brain, but only one at a time was in my mind. What chooses among them? “I.” The central problem remains for neurobiology: What is “I”?

This Issue

October 10, 1985