P. B. Medawar
P. B. Medawar; drawing by David Levine

In one of the essays in this highly readable collection, P.B. Medawar says of his hero, D’Arcy Thompson, that he “had not merely the makings but the actual accomplishments of three scholars.” Thompson was not only an eminent naturalist, but held the presidency of two classical associations, and was the author of a published paper in mathematics. He was a famous conversationalist and lecturer, and Medawar declares that his book “Growth and Form is beyond comparison the finest work of literature in all the annals of science that have been recorded in the English tongue.” Medawar’s versatility bears some comparison with that of his hero. He is Director of the National Institute for Medical Research (in Great Britain) and winner of the Nobel Prize for Medicine in 1960; he has a serious interest in the methodology of natural science, he has read more in the history of that methodology than most of its present-day practitioners, and his prose is as sharp and as witty as that of any scientist writing in English today. As one reads this collection of occasional pieces or his books, The Uniqueness of the Individual and Th Future of Man, one may also be tempted to compare him with T. H. Huxley, since Medawar has a great gift for explaining biology to laymen, he has taken pains to study philosophy and its history, and is winningly savage in dealing with theological and literary silliness about science.

Medawar’s most deserving victim is poor Teilhard de Chardin, who is followed closely by Arthur Koestler. Koestler is justly accorded more respect by Medawar, but he is also given a pretty rough time when he ventures to generalize about scientific activity in The Act of Creation. In the sharp exchange of letters that followed Medawar’s review of that book in the New Statesman it seems to me that Medawar clearly comes out on top. He also manages, in another piece which appeared originally in the New Statesman, to expose a lot of Freudian nonsense about Charles Darwin’s famous illness.

I do not wish, however, to give the impression that Medawar is merely a pundit-pummeler, a sort of scientific cop who preys on literary vagrants, and who keeps flashing his badge at scientifically wayward theologians, journalists, and psychoanalysts. It is obvious that he likes to tangle with them and is rather good at it, and I should think that untutored sages who insist on pontificating about biology would do well to be careful when Medawar is around. Yet a policeman’s lot is not a happy one, and Medawar rightly aspires to be more than that. He is much more interested in trying to answer the question: “What kind of act of reasoning leads to scientific discovery and the enlargement of the understanding?” and in analyzing certain concepts and theories of his own science, biology.

I am bound to report, however, that in dealing with the more general question Medawar is much less original and certainly less impressive than he is when discoursing about biology itself. This is particularly true when he treats such philosophical staples as induction, law, and hypothesis. On these matters he is an acknowledged disciple of Karl Popper, but it seems to me that either he misunderstands Popper’s view of the role of law in science, or he has deliberately chosen to depart from it in the direction of darkness and confusion.

POPPER has for many years conducted a spirited campaign against what he calls “inductivism,” and for the view that science proceeds by the method of “conjecture and refutation.” Inductivism is the theory that scientists offer positive support for their laws and theories by inductive inference, which in its simplest form is supposed to have the following pattern: “This A is B, and that A is B (and so on up to some finite number of A’s that are B); and no examined A is not B; therefore, it is a law of nature that all A’s are B.” Popper’s attack on inductivism takes its point of departure from Hume’s point that the “therefore” in the above pattern or argument is not deductive; it is not the “therefore” which is properly placed before the conclusion of a valid syllogism or before a mathematical conclusion when it is proven.

It follows, according to Popper, that scientists can never, on the basis of true observation statements—“This A is B” and “That A is B,” and so on—establish the truth of a supposed law of nature. On the other hand, Popper points out, we can certainly establish the falsity of an alleged law of nature, “All A is B,” by producing just one A that is not B. From this and allied considerations Popper concludes that there is no such thing as induction but that science proceeds by carrying on a relentless effort to falsify “freely invented” conjectures which are suggested but not logically dictated by the data. This critical activity, he holds, may show that some universal hypotheses which are held to be laws are false, and it may also leave the field to an unrefuted universal hypothesis. If that hypothesis survives all efforts to puncture it, and if it accounts for all of the data, it gains, according to Popper, the only sort of support that scientists can produce for their efforts at generalization. I have, of course, oversimplified Popper’s doctrine and have left out important details, such as his view of what he calls “corroboration,” but I have said enough to show, as I shall presently, that Medawar has either misunderstood or abandoned Popper’s conception of the role of law in science.


Whatever may be the philosophical merit of Popper’s attack on inductive inference as a method of establishing laws, there is no doubt that Popper rightly thinks that the formulation and testing of laws are the most important aims of science. Popper is well known for holding that every scientific explanation is a deductive argument which contains a law as one of its premises, and therefore it is surprising to find Medawar, a Popperian, saying that “scientists do not profess to be trying to discover laws” (p. 133). It may be thought that the operative word here is “discover,” in which case it may be supposed that Medawar is merely trying to say that the scientist does not passively record laws, but rather that he actively puts them forward in conjectural anticipations of nature. But a little later in the same essay Medawar says that a study of scientific language does not suggest that scientists “are busy formulating ‘laws’,” and adds that on the contrary “scientists are building explanatory structures, telling stories [Medawar’s italics] which are scrupulously tested to see if they are stories about real life” (p. 152). Surely this contrast between the task of formulating laws and building explanatory structures is not something that Medawar learned from Popper, except under the misapprehension that, because Popper rejects induction as a method of establishing scientific laws, he also thinks that the search for law is not an integral part of science. Furthermore, the concept of telling stories is left in utter darkness by Medawar.

ON THE SAME page as that on which Medawar rejects the search for law as a task of science he sends us to the writings of Claude Bernard, who, he says, constructed a philosophy of science that really “worked.” We are told that Bernard is at his best when he describes the virtue and necessity of scientific doubt, but are not told that the ultimate purpose of scientific doubt for Bernard was to clear the way for arriving at the very laws which scientists, according to Medawar, do not profess to be trying to discover. Imagine my surprise when I reopened my Bernard and found him saying that “all natural philosophy is summarized in knowing the law of phenomena” (Introduction to the Study of Experimental Medicine, p. 57, Bernard’s italics), that “a scientific law gives us the numerical relation of an effect to its cause, and that is the goal at which science stops” (Ibid., p. 67), and that “real science exists, then, only from the moment when a phenomenon is accurately defined as to its nature and rigorously determined in relation to its material conditions, that is, when its law is known” (Ibid., p. 74).

WHEN ONE TURNS from the more methodological to the more biological essays of Medawar one enters a far more interesting and fresher atmosphere. The best example of Medawar at his best may be found in his lecture, “Herbert Spencer and the Law of General Evolution.” Once again he is on the trail of those who misinterpret natural science, but this time he is stalking game much bigger than Teilhard de Chardin. Spencer, he shows convincingly, believed that there is a kind of evolution more general than organic evolution, partly because he confused the process of development from an individual seed to a tree, or from an egg to an infant, with organic evolution, which is the change of one species into another. In the same vein Medawar criticizes the use of the term “social evolution” because organic evolution takes place through genetic pathways, whereas what is called social evolution does not. All of this, Medawar thinks, is an example of what happens when we take scientific terms out of their native environment without realizing that by doing so we have changed their meanings and have lapsed into double-talk. One immediately thinks of a similar misuse of physical terminology, in particular the use of “relativity” and “indeterminacy” for intellectually nefarious purposes in ethics and metaphysics.


This is not the only important matter dealt with in the lecture on Spencer. Medawar also tries to show how misunderstanding the relationship between evolution and entropy, those two great concepts of nineteenth-century science, unfortunately led Spencer and others to suppose that belief in evolution clashed with the second law of thermodynamics. Their point was that evolution led to a long-term increase of order and coherence, whereas the second law of thermodynamics spoke gloomily of the wasting of available energy, the decay of order, and the increase of entropy. To this predicament Medawar devotes some of the most instructive pages of his book. He says that if we confine ourselves to energetics, any logical inconsistency disappears once we realize that the second law of thermodynamics tells us what happens in isolated systems, whereas living organisms are not isolated systems. An isolated system is one in which there is no external trade of matter or energy, but the second law of thermodynamics does not exclude the existence of non-isolated systems in which entropy or, as Willard Gibbs called it, “mixedupness” may be decreasing and order increasing. So, for example when we keep a room cool by refrigeration heat is transferred from that room to a place of higher temperature outside, but work must be done to effect the transfer and therefore, because the system is not isolated, we cannot say that the behavior of a refrigerator constitutes a counter-example to the second law.

Along similar lines Schrödinger has argued that living organisms maintain and add to their state of order by “drinking orderliness from the environment,” or, as Medawar puts it, “by breaking down molecules from outside the system to pay the thermodynamic bill for synthesizing molecules within the system.” But although Medawar agrees that the second law is not contradicted by the increase of thermodynamic orderliness in non-isolated systems, he resists Schrödinger’s idea that biological order may be defined as a form of thermodynamic order. Medawar holds that “order or organization as the biologist understands it means complex regularity, with the extra connotation of stability,” and that increase of complex regularity may accompany an increase of entropy or disorder even in the inanimate world. And so, when gaseous hydrogen and oxygen combine to form molecules of water, the latter are more highly “organized” in the biological sense than the parent molecules, even though this is a process in which an increase of entropy has occurred.

In a manner that I cannot usefully summarize Medawar also tries to show that certain concepts of order developed in information theory and probability theory are not equivalent to the concept of biological order. But here again his treatment, though brief and tantalizing, is illuminating and illustrative of his powers as a philosophical analyst of biological notions. Medawar has few equals among scientists when it comes to speaking out candidly and eloquently for rationality and against obscurantism masked as profundity. True, he is sometimes unclear and misleading when he deals with the problems and history of general methodology. But his writing even on those subjects is lively and provocative, and we must always remember that, however things may have been in the days of D’Arcy Thompson (1860-1948), it is given to few men in 1967 to have the accomplishments as well as the makings of three scholars.

This Issue

October 26, 1967