There is a story about a wonder-rabbi, who on his deathbed whispers to his chief assistant, “Life is like a bagel.” The word spreads through the crowd waiting outside the rabbi’s house, “Life is like a bagel; the rabbi says life is like a bagel,” until finally, at the edge of the crowd, it reaches the town fool, who asks, “What does it mean: life is like a bagel?” The question spreads back through the crowd, “What does it mean, life is like a bagel?” until it reaches the bedside of the rabbi. “Rabbi,” his assistant asks, “What does it mean, life is like a bagel?” “Nu,” the rabbi says, weakly shrugging his shoulders, “so life is not like a bagel.”

By the timely intervention of a naïve mind, we are saved from a metaphor. We do not have to live our lives as if they were seamless circles without beginning or end, firm at their exteriors but soft within, substantial around the periphery but at the center empty. We are not always so fortunate, however. Every philosopher does not have a divinely appointed fool to keep nature from mimicking his art. On the contrary, our view of the world is so dominated by powerful metaphors that we often turn similes into identities. Life ceases to be like a bagel, and becomes one.

What is surely the most powerful and influential metaphor-become-real in Western civilization was provided in 1637 by René Descartes in Part V of his Discourse on Method and further elaborated in his Principles of Philosophy. It is the organism as machine,

…a machine which having been made by the hands of God is incomparably better arranged, and is more admirable in its motions, than any that could be invented by men.


if there were such machines, which had the organs and appearances of an ape or some other nonrational animal, we would have no way of realizing that they were not of exactly the same nature as the animals….1

Indeed, the whole world, animate and inanimate, is like a clock:

I have described this earth, and the whole visible world in general, as if it were a machine in the shape and movements of its parts … for example, when a clock marks the hours by means of the wheels of which it is made, it is no less natural for it to do so than it is for a tree to produce its fruits.2

What has happened since 1637 is that, in the minds of natural scientists and a large fraction of social scientists as well, the world has ceased to be like a machine, but instead is seen as if it were a machine. Cartesian reductionism, which regards the entire world of things as, in fact, a very complicated electro-mechanical device, is not simply the dominant mode of thought in natural science, but the only mode to enter the consciousness of the vast majority of modern scientists. It is no exaggeration to say that most scientists simply do not know how to think about the world except as a machine.

The demonstration that living organisms in all their aspects (with the exception of the human conscious mind) could be explained as a mechanical system was at the heart of Descartes’s argument for a clocklike world. Part V of the Discourse contains an extensive description of the circulation of the blood, partly based on William Harvey’s De motu cordis et sanguinis, although in characteristic Gallic style Descartes refers to him only as “an English physician” who “broke the ice in this area.”3 One of the two great ironies of the history of Cartesian thought is that biology did not become thoroughly mechanistic and reductionist until 250 years after Harvey. What Newton did for physics toward the end of the seventeenth century, and Lavoisier did for chemistry at the end of the eighteenth, was not carried out in biology until Darwin gave a mechanical explanation of evolution and the German school of Entwicklungsmechanik provided a physico-chemical model of embryonic development in the latter part of the nineteenth century. The other great irony is that biology is likely to be the science that destroys the hegemony of Cartesian reductionism. Last in and first out, so to speak.

There have been two major sources of challenge to Cartesianism. One has been from Marxists who see reductionism as an intellectual reflection of the individualist bourgeois political revolution of the seventeenth and eighteenth centuries, and who oppose its atomization of the world into separate bits and pieces. The other has been from biologists who study either the central nervous system or embryonic development, fields which have made remarkably little progress using the same physico-chemical models that have been so fruitful in other fields of investigation. Both sources of discontent with reductionism were brought together in the conference in Bressanone, Italy, organized by the Dialectics of Biology Group in 1980, whose proceedings have now been published as Against Biological Determinism and Towards a Liberatory Biology.


Even the titles are in dialectical opposition. The group assembled in Bressanone were biologists concerned with brain function, embryonic development, and evolution. Most were Marxists, although two, Gerry Webster and Brian Goodwin, announce themselves as “structuralists” and prove it by ending their essay with a quotation from Lévi-Strauss, albeit a rather Marxian one.4 All have in common a frustration with a metaphor that is also a world view, and that seems to obstruct both the process of understanding the physical world and of changing the social one. When Alexis de Tocqueville sat in the Chamber of Deputies on February 24, 1848, watching the dissolution of the bourgeois monarchy, he observed that “the men of the first revolution were living in every mind, their deeds and words present to every memory.” So the Bressanone conference self-consciously compared itself to another revolutionary congress when its manifesto began: “A strange fate has overcome traditional Western philosophy of mind.” Not quite a “specter haunting Europe,” but close enough.

The natural historical approach to understanding the world consists in attempting to reconstruct the causes of events from observing systems in their normal state of motion or stasis. The experimental approach, on the other hand, uses perturbation as its primary tool. The object under study is pushed, pricked, and nicked, bits and pieces are removed, foreign agents added and the normal working of the system generally disturbed in the hope that its response to these alterations will reveal its inner workings. The Cartesian reductionist view confuses the nature of the perturbation itself with the “cause” of the system’s normal functioning. A Russian story tells of the psychologist who proves that fleas hear with their legs by training them to jump on command, and then observing that they no longer respond when their legs are amputated.

As Giorgio Bignami and Lesley Rogers show in their essays on reductionism in medicine in Against Biological Determinism, ex juvantibus logic confuses the cause of a disease with its treatment. If giving a patient the drug dopamine reduces the tremors of Parkinson’s disease, then parkinsonism is held to be caused by a deficiency of dopamine. If, on the other hand, giving a drug that is an antagonist of dopamine reduces the symptoms of schizophrenia, then an excess of dopamine is said to be the cause of schizophrenia. We all learned in school that bacteria are the cause of infectious disease, yet we manage to carry tubercle bacilli without contracting tuberculosis, while our grandparents in the mills and sweatshops regularly contracted and died from it. In view of the history of infectious disease in the last 150 years, it is at least as sensible to say that unregulated capitalism causes bronchitis as it is to blame the bacilli. Clearly, the failure of psychoneurology to understand the origins of mental and neurological disease can be traced to the vulgar reductionist belief that if a behavior is modified by a pill or a scalpel, the cause of the behavior has been touched. Nor is ex juvantibus logic without its social and political impact, for psychosurgery has been used repeatedly by the state to “cure” violent prisoners of their “pathological” reaction to the penal system. The Russian trainer of fleas was well within the tradition of modern psychoneurology.

The political uses of reductionism are not confined to the treatment of recalcitrant prisoners. Theories of the physical body and the body politic come together in biological determinism, an ideology that both justifies current social arrangements and claims them to be the inevitable consequences of the facts of life. For sociobiologists and believers in natural meritocracies of class and sex, the properties of society are determined by the intrinsic properties of individual human beings, individuals are the expression of their genes, and genes are nothing but self-replicating molecules. We are selfish because we are formed by selfish DNA. As Richard Dawkins puts it in The Selfish Gene,5 we are “lumbering robots” controlled “body and mind” by our genes. So politics becomes a branch of molecular biology, and our social and political institutions are as immutable as the chemicals of which we are made. The first essays in the Bressanone conference show how Cartesian mechanism, with its emphasis on the individual elements and the clock-like relations among them, is both a mirror of, and a theoretical basis for, a society of alienation and division of labor governed by iron laws of economic behavior.6

The confusion of perturbation with cause is nowhere more powerful than in what has come to be called with unconscious self-mockery the “Central Dogma” of molecular genetics. According to this dogma, the proteins that make our bodies are coded in our genes in the structure of the “master molecule” DNA. From Science to Science Digest to the Reader’s Digest, the word has spread that DNA makes proteins and proteins make us. In an essay that provoked great controversy, even among the ranks of the dialecticians of biology, Ruth Hubbard points out that genes do not “make” proteins and that DNA is only one part of a complex molecular ensemble, all of which must be in place and functioning for the production of the stuff of which bodies are made. DNA by itself makes nothing, not even more DNA, so it is not even “self-replicating.” Nor are the genes all that we inherit, for the sperm and egg from which each of us began contained a very large number of enzymes and energy-producing compounds, all held together spatially in microscopic structures within these cells.7


It is precisely in the mechanism8 of inheritance that the dialectical process of life is most clearly manifest. Particular genes do indeed contain the information in their DNA sequences to specify that this rather than that protein can be made. The actual manufacture of proteins requires the intervention of enzymes, which are themselves proteins and so have had their structure specified by yet other DNA sequences. Moreover, enzymes make new DNA using old DNA as a model. So, genes “make” the proteins that “make” genes, while proteins “make” the genes that “make” proteins.

There is another ideological bias lurking in the Central Dogma of genetics that is not taken up in Hubbard’s essay, the belief in the dominance of intellectual labor over mere production, of design over execution. We speak of Cheops “building” the Great Pyramid and of Napoleon “conquering” Europe, although of course laborers built the pyramids and French soldiers conquered Europe. In the production metaphor that dominates molecular biology, the genes play the role of blueprints, of designs that contain the true essence of the thing to be produced, while the enzymes and cell structures are “merely” the production machinery itself, undifferentiated labor power that can be put to any use. One hardly needs to be a Marxist to see the evident penetration of the bourgeois world view into biology.

It is said that Thomas Henry Huxley, when challenged to explain why, despite generations of circumcision, Jews continued to be born with foreskins, instantly replied, “There’s a divinity that shapes our ends, rough-hew them how we will.” No field of investigation has produced more crises of faith in Cartesian reductionism than the study of embryonic development. Over and over again the great embryologists of the last century have begun as mechanists and have ended searching for the immortal hand that framed the symmetries they studied. The great German embryologist Hans Spemann thought he was on the track of the chemical organizers of development until his student Holtfreter showed that he could get frog embryos to make eyes and legs by shoving toothbrush bristles and cellophane from cigarette packs into them. Hans Driesch, one of the founders of Entwicklungsmechanik, wound up believing in vitalism and parapsychology.

The problem has been that when developing embryos are challenged by perturbations, they often respond by ignoring them. The capacity of a developing frog to end up as a normal adult frog despite insults and injuries heaped on it by the embryologist has led many biologists to believe in some mysterious force or inner will contained in living stuff. So, for example, simple Cartesian reductionism leads us to analyze a leg as being composed of toes, a foot, and a shank, and each of these, in turn, as made up of constituent muscle, bone, blood vessels, nerves, and skin. Yet if one cuts out the developing limb bud of an amphibian, shakes the cells loose from each other, reaggregates them into a random lump, and puts the lump back in place, a normal leg will develop. It is as if each cell had within it the information “limb” without knowing whether it is to be toe or shank, muscle, bone, or nerve. Unlike a machine whose totality is created by the juxtaposition of bits and pieces with different functions and properties, the bits and pieces of a developing organism seem to come into existence as a consequence of their spatial position at critical moments in the embryo’s development. Such an object is less like a machine than it is like a language whose elements, despite a superficial identity of morphology, take unique meaning from their context. The very different properties of language-like and machine-like systems became apparent when attempts were made to build translation machines. How does one build a machine that will translate correctly into, say, French, the two sentences, “Fruit flies like bananas” and “Time flies like an arrow”?

In the past, the failure of embryogenesis to yield to the Cartesian program of cut and paste has led to an alternative of an obscurantist holism or, worse, a belief in nonmaterial life forces. The members of the Dialectics of Biology Group who have concerned themselves directly or indirectly with problems of development9 are, above all, materialists, so they reject vitalism; but they reject holism as well, maintaining that the physico-chemical properties of molecules and subcellular parts are indeed relevant to an understanding of development. It does matter, after all, that organisms are made of flesh and blood and not of stone.

Their solutions, although not all the same, are versions of what we might call constrained relationalism. The properties of organisms are consequences of the particular interactions that occur between bits and pieces of matter, so it is laws of interaction that we must study. But these interactions, although unique to different interacting parts, are constrained by the nature of the parts themselves. The properties of water are the outcome of a unique interaction of hydrogen and oxygen, but the kinds of interactions that oxygen can enter into are different from those in which, say, nitrogen takes part and are, to some extent, predictably different.

Of course, it is one thing to call for a biology that is relational rather than compositional, and quite another to put the call into practice. One of the simplest examples of a force that is created by the very objects that are under its influence is a gravitational field, yet a complete description of what happens when only three bodies interact eludes us. If development is really, in an important sense, a consequence of the relations between things, how are we to reduce the incredible complexity of relationships to a manageable set of regularities? And how are we to do this using an experimental method that is itself so wedded to Cartesian analysis?

If thinking about embryos has troubled Cartesian reductionists, thinking about brains has driven everyone to distraction. Descartes’s solution to the problem of mind was, as we all know, to exclude it from the universal machine and to claim that soul was separate from body. In the biology of the twentieth century, however, Cartesian dualism is simply not acceptable. Whether reductionist, holist, or dialectical, all parties agree on a materialist view that everything in the world is some manifestation of matter and energy. Reductionists have interpreted this materialism in a rather naïve way, as they so often do, providing some amusement in a really dismal science. So, for example, it was fashionable not long ago to suppose that specific memories were encoded in specific nucleic acid molecules, and there were several reports, now discredited, that one flat worm could learn what another flat worm knew by eating it. After all, genes are made of dna, and aren’t they just a form of “information,” a “memory” of our ancestors? The price of metaphor is eternal vigilance.

A more serious attempt has been to analogize the brain to an electronic computer, a metaphor with administrative consequences, since it led to the formation of a number of university departments of Artificial Intelligence. While these groups have produced a few elementary robots and some very clever problem-solving computer programs, including one that will do high-school geometry proofs, it is now universally agreed that if the brain is like a computer, it is like a computer no one has ever designed. Among the difficulties with the computer analogy is that, unlike an electronic memory, human memory does not store specific events in specific spots, and new experiences can actually induce the formation of new circuits. Yet every materialist would have to admit that if I could build with transistors or wires or whatever hardware an exact copy of my brain, and then set the electronic currents flowing in it, as they are now flowing in my brain, that machine could have written this sentence. That is, it could have contemplated is own existence.

The reductionist view of the relation of the brain to the mind has been either to claim that the brain produces the mind, or to deny the relevance of the mental altogether, saying that it is “nothing but” a brain state. In either case, the physical brain is at the center of attention, the proper object of study for the scientist, while the mind can be left for the allegedly ineffectual ruminations of epistemologists, linguisticians, and psychologists, none of whom will ever get to the real root of things. The dialecticians of the Bressanone group, however, assert the symmetry of mind and brain as two aspects of the same reality, each embodying a different set of properties of that reality.10 Meaning and syntax, for example, are not in the brain, although there is a unique physical configuration of the brain corresponding to every thought. There is, simply, a set of correspondences between brain state and mind state, and these in turn manifest themselves in social experience. Stephan and Beatrice Chorover argue that physical states of the neocortex, the limbic system, and the core brain are represented in individual human beings as thinking, feeling, and doing, and in society as facts, values, and practices.11 Changes that arise out of activities at one level do not cause, but are manifested in and represented at, the others.

While the members of the Bressanone group are undoubtedly right that mind and brain are aspects of the same reality, it is not clear how we can use that realization in the solution to the biological problem of why we think with our brains and not with our noses. Somehow the juxtaposition of physical elements and functions that is in our heads possesses an aspect, the mental, that is denied our lower parts and lower forms of life. The answer does not lie in the physical stuff itself, which is the same in all animals, but somehow in the structure of the central nervous system as a system. Yet a study of all or any significant part of the interconnections and coupled oscillations of our millions of brain cells seems a task that would consume the three billion years left before the sun becomes a red giant and fries all of life. We must face the possibility that we will never understand the organization of the central nervous system at any but the most superficial level.

What then is the insight that dialectical thought offers? Surely it is not simply the idea that things interact to produce unique properties that do not belong to the components themselves. Even vulgar reductionists agree that hydrogen and oxygen are toxic gases but that water is wet and sustains life. And dialectics must mean more than that one thing can have two very different co-equal aspects, since that was already decided at the Council of Nicaea, 1,500 years before Hegel and Marx.

What characterizes dialectics is its rejection of the terms of the argument that places all questions somewhere on a line between explanation by properties of parts and explanation by emergent properties of wholes. It is not that a whole is more than the sum of its parts, but that the parts themselves are redefined and re-created in the process of their interaction. So the reductionist sociobiologists argue that individual human limitations place constraints on society, but, in fact, social organization is the negation of individual limitations. No human being can fly by flapping her or his arms and legs. Yet we do fly because of the existence of airplanes, airlines, pilots, fuel production, radios—all products of social organization. Most important, it is not society that flies, but individuals who have acquired a property as a consequence of socialization. The reductionist geneticist says that DNA is self-replicating, but neither DNA nor any molecule or cell component can reproduce itself in isolation. In their organization into cells, which are self-replicating systems, the individual molecules and cell particles have acquired the property of replication.

Cartesian biology objectifies organisms. They are seen as the passive consequences of internal and external forces, genes, and environment. Organisms are objects, the internal and external forces the subjects. What dialectical biology attempts to do is to break down the alienation of subject and object, to insist on the interpenetration of gene, organism, and environment. Thus, in place of the metaphor of adaptation of organisms to a preexistent environmental “niche,” dialectical biology emphasizes the way in which organisms define and alter their environment in the process of their life activities.12 Organism and environment are both in a constant state of becoming, mutually determining each other.

The problem for dialectical biology is to turn its vision into concrete results. One of the participants in the Bressanone Conference told his friends when he got home that he had just attended reductionism’s funeral. If so, the pall-bearers must have had an easy time of it, because I am pretty sure I saw the corpse going up in the elevator in my department last week. Thus far the dialecticians have only interpreted science in various ways; the point, however, is to change it.

This Issue

January 20, 1983