Jacques Loeb
Jacques Loeb; drawing by David Levine


The history of biology is the history of struggles over the difference between the animate and the inanimate. Natural philosophy, through the Renaissance, and folk wisdom for a much longer time, saw the entire natural world as a single interconnected system in which radical transformations of qualities of both living and nonliving things were entirely credible. It was not merely that one inanimate kind of substance could, by alchemical transformation, be made into another, or that a vain boy could become a flower, but that the inanimate and the animate were interchangeable. Men could be petrified and marble statues turned to warm flesh in the embrace of their admirers. Papal staves put forth leaves, while moldy cheese and rags bred forth mice.

Aristotle believed animals could come from mud and that the animate and inanimate graded imperceptibly into one another on the scala naturae. But even the ancients were ambivalent about the ease with which inanimate matter could make that imperceptible transition. Despite Lucretius’ assurance that “even today many animals spring from the earth, formed from the sun’s heat and rain,” it was not regarded as an everyday occurrence. In the Metamorphoses, Ovid gives few examples like the case of Pygmalion’s statue and the creation of man from clay by Prometheus.

The raising of the dead by Jesus was, after all, the evidence of special powers and the leafing out of the Pope’s staff a sign of special grace. Moreover, the transition from Aristotle’s view to our present belief that the living are separated from the dead by a one-way bridge was a long and problematical one. Already in the seventeenth century William Harvey had declared ex ovo omnia, but the idea of spontaneous generation—that life originated from nonliving matter—was given a boost when Leeuwenhoek, looking through his microscope, saw a multitude of tiny living particles swimming by.

The Whig history of science we learn in school tells us that by the end of the eighteenth century Spallanzani had nailed down the case against spontaneous generation by his experimental approach to what was purely metaphysical speculation by wicked Aristotelians. But, as a matter of fact, the very same experiments done by others got the opposite result, supporting spontaneous generation. The disproofs of spontaneous generation that we now regard as definitive, those of Pasteur showing that microorganisms reproduce, were carried out in response to a public solicitation by the French Academy of Sciences for someone to finally settle the issue—in 1860.

We should not imagine, like the Whig historians of science, that the struggle over spontaneous generation was a story of the triumph of materialism and empiricism over superstition and a priori natural philosophy. On the contrary, nineteenth-century materialists took sides against the biogenetic law, the rule of “all life from life.” For if there were an unbridgeable gap between the nonliving and the living, how could we explain the primal origin of life except by the infusion of a vital spirit into clay by a Promethean God? Moreover, that vital spirit, distinct from the known material forces of the universe, must be lurking in all living organisms, impalpable and unmeasurable. Nothing could be more anti-materialist than the claim for the uniqueness of life.

The struggle over spontaneous generation embodies the contradiction that plagues biology even today. On the one hand, mechanist materialist biology assumes that living beings are simply another form of the motion of matter and that the reductionist tactic of tearing matter into smaller and smaller bits will reveal all there is to know about life. On the other, biologists have never been able to create the living from the nonliving, nor do they even know where to begin. The biogenetic law seems as unbreakable as ever.

In support of the multibillion-dollar project to determine the DNA sequence of the entire set of human genes, the eminent molecular biologist Walter Gilbert has claimed that when we know the entire human genome we will know what it is to be human. But that must mean that at the very minimum we will know what it is to be living flesh. Yet neither Gilbert nor any other molecular biologist that I know of has suggested that knowing what a human being is will enable us to make one. That is, biologists, while believing that living organisms are nothing more than a form of matter and its motions, also believe that there is some principle of organization of living matter that is shared with no other natural assemblage of atoms. The question biologists keep asking themselves is “Why is this matter different from all other matter?”

The distinction between knowing what things are made of and knowing how to create or manipulate them permeates science and yet is a source of confusion for reductionist biologists. The problem is that in the very operation of determining what things are made of we take them to bits, and in ways that destroy the very relations that may be of the essence. “We murder to dissect.” Nor is this true only for whole organisms or cells. Despite a knowledge of the structure of protein molecules down to the very placement of their atoms in exact three-dimensional space, we do not have the faintest idea of what the rules are for folding them up into their natural form. That does not prevent us, however, from being able to fold them up correctly in some cases by blind empirical fiddling. In the view of most biologists the disjunction between knowing what and knowing how is only a reflection of temporary ignorance, and, anyway, the successful manipulation of the world is only secondary to the primary goal of understanding. But that view has not been universal and some have regarded the control of life to be the object of the enterprise, putting the question of understanding quite aside. The most famous of the bearers of this “engineering ideal” in biology, Jacques Loeb, is the subject of Philip Pauly’s superb book, Controlling Life. Rarely does a scientific biography so clearly illumine deep and long-lasting ideological differences in the conduct of scientific work.1


Jacques Loeb was trained as a medical doctor at the University of Strasbourg and came to the US in 1891. When he moved to the University of California at Berkeley in 1902 after ten years at the University of Chicago, a page one headline in Hearst’s San Francisco Examiner announced that “Illustrious Biologist Joins Faculty of State University,” accompanied by a four-column artist’s sketch of Professor Loeb and his magnifying glass. Of course, Hearst was a California booster and November 1902 was during a rare slack period when the United States, having completed one of those Caribbean adventures so loved by Mr. Hearst, had not yet started on its next. Nevertheless, even a Nobel Prize winner these days can hardly count on more than a one-column feature on the day of the big news, and is unlikely to be noticed by the local stringer for the Times when he decides to trade in State Street for Union Square.

The reason for Loeb’s popular and journalistic fame had been announced just three years earlier in the Chicago Tribune: “Science Nears the Secret of Life: Professor Jacques Loeb Develops Young Sea Urchin by Chemical Treatment—Discovery That Reproduction by This Means is Possible a Long Step Towards Realizing the Dream of Biologists, ‘to Create Life in a Test Tube.’ ”

Indeed for many Loeb had created “life in a test tube.” The modern anti-abortion movement did not invent the idea that life begins at the mystical moment of fertilization. The passive and comatose egg is quickened into life by the active wriggling sperm, like Sleeping Beauty recalled to life by a princely embrace. Loeb’s successful induction of embryonic development without the benefit of sperm—“artificial parthenogenesis”—seemed closely akin to spontaneous generation.

Indeed, a headline in the Boston Herald completed the connection: “Creation of Life. Startling Discovery of Prof. Loeb. Lower Animals Produced by Chemical Means. Process May Apply to Human Species. Immaculate Conception Explained.” The intoxication of the press was extraordinary. The confusion between artificial parthenogenesis and spontaneous generation might be expected, but the conflation of the doctrine of the Immaculate Conception with the doctrine of the Virgin Birth seems inexcusable in a Boston newspaper.2

As Pauly so clearly shows, Loeb’s triumph was neither the accidental consequence of a program devoted to broader ends nor a critical step in an analytical project designed to “understand” development and reproduction. It was, rather, “a natural consequence of his conviction that biology was and should be an engineering science concerned with transforming the natural order.” It was the coming together of the nineteenth-century ideological commitments to materialism, on the one hand, and an optimistic progressivism on the other. The phenomenal world was material and only material, and through the workings of human intellect that material world could be manipulated for any desired end. It was not that all things lay within the possibility of human understanding but that they lay within the sphere of human action. Indeed, we owe to Loeb this extraordinary epistemological position, an extension of Ernst Mach’s operationalist view that

the proof of the explicability of any single life phenomenon is furnished as soon as it is successfully controlled unequivocally through physical or chemical means or is repeated in all details with nonliving materials….

We cannot allow any barrier to stand in the path of our complete control and thereby understanding of the life phenomena. I believe that anyone will reach the same view who considers the control of natural phenomena is the essential problem of scientific research.3

Moreover, such control of life was the object of the entire enterprise: “I believe that it can only help science if younger investigators realize that experimental abiogenesis is the goal of biology.”4


Pauly notes the inevitable journalistic comparison of Loeb with Victor Frankenstein, but he makes nothing of the problem that the engineering ideal raises, the problem of the unintended consequences of pragmatism that is the central theme of Mary Shelley’s Frankenstein, or the Modern Prometheus. Shelley and her husband were greatly interested in and greatly disturbed by the instrumentalism of the most eminent and influential English scientist of the early nineteenth century, Sir Humphry Davy. Davy’s Discourse, Introductory to a Course of Lectures on Chemistry, which Mary Shelley read just before beginning her own Frankenstein, was the inspiration for her fictional Professor Waldman, Frankenstein’s teacher and model.

Davy’s scientific work was a series of diverse researches in chemistry, biology, and practical physics that were often instigated by practical demands. He made discoveries in agricultural chemistry and invented the Davy lamp, which allowed miners safe illumination in gas-filled mines. He investigated the electricity of the torpedo fish and the composition of ancient coloring materials. He was the very model of a modern scientist general, solving the mysteries of nature for the benefit of human life. But his philosophical writings on chemistry show that for him, as for Loeb, understanding was simply control; scholarship was secondary to artisanry.

It was this untheoretical pragmatism that was Frankenstein’s error. In a blinding moment of truth he discovered a single secret of how to create life and then on “a dreary night of November” (November was the month of Loeb’s announcement!), “I collected the instruments of life around me, that I might infuse a spark of being into the lifeless thing that lay at my feet” (my emphasis). While Shelley does not say, those “instruments of life” were surely the apparatus of galvanic electric stimulation with which Galvani’s nephew, Aldini, in a public demonstration in 1803, had made the jaw of a hanged murderer quiver, his fist clench, and his eye open.5 But the possession of this single trick is like the secret ingredient of an alchemical formula or the secret incantation of a sorcerer. It can call spirits from the vasty deep, but it cannot control them because there is no understanding of their true nature.

In the most revealing chapter for the intellectual historian, Pauly traces the pedigree of Loeb’s empiricism into later generations. Although Loeb continued to work on parthenogenesis, adjusting the chemical condition to optimize the process, he also returned to his earlier work on animal behavior in a similar mode. While others tried to analyze behavior as being the result of intrinsic physiological processes, Loeb’s instrumentalist view led him to emphasize the importance of the environment in eliciting responses from organisms. Just as in the case of reproduction, the goal of experimental science was to be the control of behavior by the appropriate external stimuli rather than a “metaphysical” program of analyzing the internal states of an organism.

John B. Watson, who founded behaviorism, was a disciple of Loeb’s at Chicago. In his behaviorist manifesto of 1912, Watson declared that psychology’s “theoretical goal is the prediction and control of behavior.” And the network grew. Another disciple of Loeb’s was W.J. Crozier, a physiologist at Harvard, and two of his students were Gregory Pincus, who erroneously claimed to have produced parthenogenesis in rabbits on his way to inventing the contraceptive pill, and B.F. Skinner, the inheritor and elaborator of Watson’s behaviorism. Yet another admirer of Loeb’s was H.J. Muller, who won the Nobel Prize for his discovery that mutations could be produced by X-rays. While Loeb wanted biology to create life, Muller’s goal was to change it, to produce directed evolution both by controlling the process of mutation and by eugenic programs of controlled breeding. By general agreement a demigod of genetics, Muller never made a single contribution to the analysis of the underlying physiological, cellular, and biochemical mechanisms of inheritance. For him, as for Loeb, the control of life was the central issue.

Loeb’s claim to be interested only in control, relegating analysis of internal mechanisms to the sphere of the “metaphysical,” ran against the entire trend of the reductionist science of his day. He was under constant pressure to rationalize his work by the articulation of an analytical program, pressure to which he finally gave way. Between 1910 and 1918 Loeb gave up his radical Machian commitment to the control of life and integrated himself into the established epistemological order. “He no longer saw scientists as leaders in the transformation of the world, but as cloistral figures, removed from society, seeking pure knowledge.” Prometheus was bound: more than bound, utterly transformed. The image of Loeb as the founder of the mechanistic concept of life comes down to us in the form of Max Gottlieb in Sinclair Lewis’s Arrowsmith, the epitome of the pure scientist.

The struggle between control and analysis as the goal of biology continues to the present, albeit in a less disinterested form. The commitment to understanding, to knowledge as an end in itself, is deeply embedded in scholarly culture. It is a first cause of our collective scholarly existence. But as the easy problems of biology are solved, the costs of solving the remaining hard ones become greater and greater, and the strain on the credulity of those who pay has not been lessened by changing the tax brackets. It is simply impossible to justify the expenditure of a billion dollars on a project to put in sequence the complete DNA of a “typical” human being or corn plant on the grounds that it would be a lovely thing to behold. So are we assured that it is really all in the interest of curing cancer, relieving schizophrenia, and making groceries cheaper. What for Jacques Loeb was an honest epistemological position has become, for the post-modern Prometheus, a piece of directmail advertising.


Living organisms are characterized by five properties: they reproduce, they evolve, they recognize themselves, they develop, and they feel. These properties have given rise to the five major problems in biological science. Three have turned out to be “easy” for mechanistic biology, and two are very hard. The problem of how organisms reproduce, how they pass on to their offspring the information that they are to be lions rather than lambs, is now largely solved thanks to Gregor Mendel and his molecular biological epigones. So, too, we know the global features of the evolutionary process, although local mysteries remain. The ways in which organisms recognize themselves as opposed to others are varied so that there is no single mechanism of self-recognition; but a major mode, the formation of antibodies in higher animals, is now well understood at the molecular and cellular level.

That leaves us with the two hard problems: What is going on inside my head as I write these words and how, starting from a single fertilized egg in my mother’s uterus, did I develop the brain, eyes, and fingers that make it all possible? The problem is not simply that we do not have single coherent stories to tell about these processes, but that we do not know how to produce well-framed questions of whose relevance we are sure. Instead we have faddish models that succeed each other at five- or ten-year intervals, driven largely by changes in available technology in other branches of science, rather than by any coherent intellectual program.

Even the “easy” problems were not, of course, all that easy, and great fame has accrued to those who like Mendel or Watson and Crick have made major contributions to the solution. Gerald Edelman became famous (at least within scientific circles) and won a Nobel Prize for his major contribution to solving the self-recognition problem at the molecular level. Like other molecular biologists, having succeeded in solving his “easy” problem at a fairly early age, he decided to try a “hard” one, and for the last dozen years he has worked on the formation of the central nervous system. As a consequence, working along lines related to those of Pierre Changeux6 and others, he has produced a really new theory of how the thinking brain develops,7 by analogy with Darwin’s theory of evolution by natural selection. But the problem of the brain is a problem in development in a larger sense; it is a problem of how so immensely complex a cognitive system is created out of “rudiments of form and sense.” And so, with an intellectual ambition unmatched since Sir Humphry Davy, Edelman slid into the other hard problem, embryonic development. Topobiology is his attempt to make a coherent story of embryonic development out of the vast research on the subject that has accumulated since the turn of the century.

The problem of embryonic development is often said to be that of the origin of heterogeneity from homogeneity. The problem arises at two levels. First, all sperm and eggs look pretty much alike, but even a child can tell a frog from a prince. That is, what is the origin of the immense difference between organisms starting with what appears to be pretty much the same stuff? Second, starting with a single rather dull looking egg cell, what is “that strange eventful history” by which we acquire, only to lose in the end, our teeth, eyes, taste, everything? That is the problem of embryonic differentiation. With the hubris born of a surfeit of Nobel Prizes, the geneticists tell us that the answer is obvious, it’s all in the genes. Frogs and princes have very different sorts of genes despite the superficial resemblance of their gametes, or reproductive cells, so we should expect the result of development to be very different. As for the development of differentiated tissues and organs within an individual organism, it is simply the expression of different members of our set of genes at different times and different places in the developing embryo.

For Edelman the difficulty of this facile non-explanation is one of dimension. How, he asks, can one-dimensional information (the string of DNA) become manifest in a three-dimensional organism? Indeed, the problem seems to be even worse since an organism is really four-dimensional, changing in a patterned way with time. The time dimension is, in fact, expressed not in time units but in a sequential order of changes. So the standard description of the development of a frog is not in terms of hours and days but in terms of developmental stages with clear morphological markers (the one-cell, two-cell, four-cell stages, the first appearance of a nerve cord, and so on). This method of marking time is a consequence of another major feature of embryogeny. Although the absolute time rate of development is sensitive to external conditions like temperature, the major features of embryonic development follow each other in an invariant order.

Edelman’s statement of the problem of development as one of dimension does not quite epitomize either the difficulty or his solution to it. In principle there is no difficulty at all in expressing instructions about three dimensions in a one-dimensional form. When I addressed the envelope in which this review was sent to “Robert Silvers, New York Review of Books, 250 West 57th Street,” I did just that and in a very precise way, given Mr. Silvers’s size relative to that of the rest of the universe. Moreover, had I added the instruction “To be opened immediately,” I would have provided both temporal and functional information at the same time.

The problem is not one of dimension, but of size. The nucleus of a cell of the fruit-fly Drosophila, the favorite organism of geneticists, has enough DNA to specify the structure of about five thousand different proteins and about thirty times that much DNA is available to provide spatial and temporal instructions about when the production of proteins by those genes should be turned on and turned off. But this is simply too little, by many orders of magnitude, to tell every cell when it should divide, exactly where it should move next, and what cellular structures it should produce, over the entire developmental history of the fly. One needs to imagine an instruction manual that will tell every New Yorker when to wake up, where to go, and what to do, hour by hour, day by day, for the next century. There is just not enough DNA to go around.

It must be, Edelman argues, that the present location of a cell and its present activity provide most of the information on what it is to do next. It is this contingency on position that makes biology into “topobiology.” It is this contingency on position that explains why, in taking an organism to pieces, we lose its organismal property. Legs and arms have exactly the same kinds of skin, bone, muscle, hair, nail, and connective tissue in about the same proportions. And they are very similar in their overall dimensions and gross structure. Yet, luckily for us, whether they developed on our front or rear ends made the critical difference in their final form and function. Victor Frankenstein’s real “instruments of life” are not electric generators and spark gaps, but microscopic compasses, rulers, and protractors.

The idea that the position of a bit of protoplasm relative to other bits provides critical developmental information is not a new one. On the contrary, most theories of embryonic development in the last seventy years have attempted to make something of the notion of positional information. Metaphors of “field” and “gradient” and spatial waves of chemical concentration have dominated embryology. The problem is that no one has been able to make these metaphors work, or to give them a material molecular basis. That is the task that Edelman sets for himself. His strategy is to push the notion of local positional information to its extreme by supposing that essentially all the action is at the level of small collectives of cells acting as a group on their immediate neighbors. There are no global gradients over the organism or large-scale fields in which cells are moving. Central planning has been replaced by local initiative in a kind of perestroika of the protoplasm.

In the spirit of the local autonomy of small collectives, Edelman divides the cellular processes of differentiation into two kinds. First there are cell population processes—the division, migration, and death of cells. Much of embryonic development, when given a bare-bones description, does indeed consist of differential rates of cell division and cell death and the movement of small clumps or sheets of cells from one place to another, accompanied by the folding and rolling of such sheets as they move. The remainder of development is cytodifferentiation, the qualitative change in the actual structure and function of individual cells. Some cells, like those in our hair follicles, begin to pump out huge amounts of a particular protein; hence the demand for barbers. Others grow tiny hair-like appendages themselves, whose beating and waving keep microscopic dust from accumulating in our lungs.

During his work on the development of the central nervous system Edelman had his attention drawn to molecules that acted like glues between cells, and it is on these molecules that he builds the entire edifice of his general theory of development. These molecules, called CAMs (cell adhesion molecules) and SAMs (surface adhesion molecules), are turned on and off in cycles. By affecting the surface properties of cells they cause cells to aggregate; this is followed by the movements of sheets of cells across each other and along noncellular matrices of materials that are secreted by other cells. This is the origin of the large movements and foldings of tissue that give rise to the general shape of an organism and account for the location of tissue. In some unspecified way the surface interactions of cells with their neighbors then turn on special regulatory genes within cells, which, in turn, switch on and off the genes responsible for cytodifferentiation. At every stage it is the local interactions of cells and tissues that determine the further movement, division, and differentiation of cells in the locality, which lead to yet further new local interactions, and so on to adulthood.

But, you may object, how does an entire, integrated, functioning organism arise from this anarchy of local control? Where is the invisible hand? By asking that question, Edelman says, you reveal that you are still mired in nineteenth-century pre-Darwinian teleology. The invisible hand is natural selection. There is nothing intrinsic to the process of development that leads to an integrated functioning organism, any more than there is anything intrinsic to the process of mutation that leads to better-adapted organisms. Those developmental processes that lead to a nonfunctioning organism have been lost in evolution because the bearers of those faulty programs left no progeny. All that is left is the collection of local processes that give the appearance of overall coordination because they work. The invisible hand of development is the very one that the Scottish economists extended to Darwin. Although the molecular details of the process of development occupy the attention of Edelman, and are likely to be the center of attention for most of his biologically trained readers, it is his sweeping away of the teleological element lurking in most accounts of development that is most insightful and most radical. Although biologists constantly decry teleology, they have not been able to free themselves of it in their explanation either of development or of cognitive functions. By constructing a theory of development that is nothing but the collection of quasi-independent local events, filtered through natural selection, Edelman has offered reductionist biology its last chance of encompassing development in its epistemological program.

Topobiology is a hard book to read, even for a professional biologist. Part of the reason is that it is a hard subject with an immense phenomenology ranging from the anatomical to the molecular and with a relevant literature extending back into the mists of antiquity. But partly the subject has been obfuscated by Edelman’s language. As I read, I thought of the famous sentence in Czech, “Put your finger down your throat,” which is said not to have a single vowel. There are whole paragraphs in Topobiology without a monosyllable (including “and” and “the”). In a euphuistic frenzy of polysyllabism Edelman refers to chewing and swallowing as “mastication and deglutition.” The author is certainly not an uncultured scientific idiot savant (as adolescents, he and I together learned to declaim Corneille, Racine, and lesser French classics at a school devoted to such manifestations of high culture). It may be that Edelman believes that large subjects demand large words. In some instances, at least, the imprecision of high-flown vocabulary seems designed to substitute for a precision of ideas. If so, it is too bad, because the importance of his project, and the opportunity it offers biologists to vindicate their faith in the analytic method where they have consistently failed before, would be more than enough justification for Edelman to use, repeatedly, those three little Anglo-Saxon words: “I don’t know.”

This Issue

April 27, 1989