One of the conspicuous handicaps is the conservatism of the scientific mind in its corporate aspect. The collective matrix of a science at a given time is determined by a kind of establishment, which includes universities, learned societies, and, more recently, the editorial offices of technical journals. Like other establishments, they are consciously or unconsciously bent on preserving the status quo -- partly because unorthodox innovations are a threat to their authority, but also because of the deeper fear that their laboriously erected intellectual edifice might collapse under the impact. Corporate orthodoxy has been the curse of genius from Aristarchus to Galileo, to Harvey, Darwin, and Freud; throughout the centuries its phalanxes have sturdily defended habit against originality. The uses of hypnotism in dental surgery, child-birth, etc., are regarded as a modern discovery. In fact, Esdaile, who lived from 1808 to 1859, carried out three hundred major operations under 'Mesmeric trance'; but since Mesmer had been declared an impostor, medical journals refused to print Esdaile's papers. In 1842 Ward amputated a leg painlessly under hypnotic trance and made a Report to the Royal Medical and Chirurgical society. The society refused to believe him. One of its most eminent members argued that the patient had merely pretended not to feel the pain, and the note of the paper having been read was struck from the minutes of the Society.
The martyrology of science mentions only a few conspicuous cases which ended in public tragedies. Robert Mayer, co-discoverer of the principle of the Conservation of Energy, went insane because of lack of recognition for his work. So did Ignaz Semmelweiss, who discovered, in 1847, that the cause of childbed fever was infection of the patient with the 'cadaveric material' which surgeons and students carried on their hands. As an assistant at the General Hospital in Vienna, Semmelweiss introduced the strict rule of washing hands in chlorinated lime water before entering the ward. Before this innovation, one out of every eight women in the ward had died of puerperal fever; immediately afterwards mortality fell to one in thirty, and the next year to one in a hundred. Semmelweiss's reward was to be hounded out of Vienna by the medical profession -- which was moved, apart from stupidity, by resentment of the suggestion that they might be carrying death on their hands. He went to Budapest, but made little headway with his doctrine, denounced his opponents as murderers, became raving mad, was put into a restraining jacket, and died in a mental hospital.
Apart from a few lurid cases of this kind we have no record of the countless lesser tragedies, no statistics on the numbers of lives wasted in frustration and despair, of discoveries which passed unnoticed. The history of science has its Pantheon of celebrated revolutionaries -- and its catacombs, where the unsuccessful rebels lie, anonymous and forgotten.
Limits of Confirmation
From the days of Greece to the present, that history echoes with the sound and fury of passionate controversies. This fact in itself is sufficient proof that the same 'bundle of data', and even the same 'crucial experiment', can be interpreted in more than one way.
To mention only a few of the more recent among these historic controversies: the cosmology of Tycho de Brahe explained the facts, as they were known at the time, just as well as the system of Copernicus. In the dispute between Galileo and the Jesuit Father Sarsi on the nature of comets, we now know that both were wrong, and that Galileo was more wrong than his forgotten opponent. Newton upheld a corpusculary, Huyghens a wave-theory of light. In certain types of experiment the evidence favoured Newton, in other types Huyghens; at present we tend to believe that both are true. Leibniz derided gravity and accused Newton of introducting 'occult qualities and miracles' into science. The theories of Kekulé and Van't Hoff on the structure of organic molecules were denounced by leading authorities of the period as a 'tissue of fancies.' [14] Liebig and Wöhler -- who had synthesized urea from anorganic materials -- were among the greatest chemists of the nineteenth century; but they poured scorn on those of their colleagues who maintained that the yeast which caused alcoholic fermentation consisted of living cellular organisms. They even went so far as to publish, in 1839, an elaborate skit in the Annalen der Chemie, in which yeast was described 'with a considerable degree of anatomical realism, as consisting of eggs which developed into minute animals shaped like distilling apparatus. These creatures took in sugar as food and digested it into carbonic acid and alcohol, which were separately excreted.' [15] The great controversy on fermentation lasted nearly forty years, and overlapped with the even more passionate dispute on 'spontaneous generation' -- the question whether living organisms could be created out of dead matter. In both Pasteur figured prominently; and in both controversies the philosophical preconceptions of 'vitalists' opposed to 'mechanists' played a decisive part in designing and interpreting the experiments -- most of which were inconclusive and could be interpreted either way.
I have compared the nineteenth century to a majestic river-delta, the great confluence of previously separate branches of knowledge. This was the reason for its optimism -- and its hubris; the general convergence of the various sciences created the conviction that within the foreseeable future the whole world, including the mind of man, would be 'reducible' to a few basic mechanical laws. Yet as we enter our present century, we find that in spite of this great process of unification, virtually every main province of science is torn by even deeper controversies than before.
Thus, for instance, the most exact of the exact sciences has been split, for the last twenty years, into two camps: those who assert (with Bohr, Heisenberg, von Neumann) that strict physical causality must be replaced by statistical probability because subatomic events are indeterminate and unpredictable; and those who assert (with Einstein, Planck, Bohm, and Vigier) that there is order hidden beneath the apparent disorder; governed by as yet undiscovered laws, because they 'cannot believe that God plays with dice'. Another controversy opposes the upholders of the 'big-bang theory', according to which the universe originated in the explosion of a single, densely packed mass some thirty thousand million years ago and has been expanding ever since -- and the upholders of the 'steady-state theory', according to which matter is continually bring created in a stable cosmos. In genetics, the neo-Darwinian orthodoxy maintains that evolution is the result of chance mutations, against the neo-Lamarckian heretics, who maintain that evolution is not a dice-game either -- that some of the improvements due to adaptive effort can be transmitted by heredity to successive generations. In neuro-physiology, one school maintains that there is rigid localization of functions in the brain, another, that the brain works in a more flexible manner. In mathematics, 'intuitionists' are aligned against 'formalists'; in the medical profession, opinions are divided regarding the psychological or somatological origin of a great number of diseases; therapeutic methods vary accordingly, and each school is subdivided into factions.
Some of these controversies were decided by cumulative evidence in favour of one of the competing theories. In other cases the contradiction between thesis and antithesis was resolved in a synthesis of a higher order. But what we call 'scientific evidence' can never confirm that a theory is true; it can only confirm that it is more true than another.
I have repeatedly emphasized this point -- not in order to run down science, but to run down the imaginary barrier which separates 'science' from 'art' in the contemporary mind. The main obstacle which prevents us from seeing that the two domains form a single continuum is the belief that the scientist, unlike the artist, is in a position to attain to 'objective truth' by submitting theories to experimental tests. In fact, as I have said before, experimental evidence can confirm certain expectations based on a theory, but it cannot confirm the theory itself. The astronomers of Babylon were able to make astonishingly precise predictions: they calculated the length of the year with a deviation of only 0.001 per cent from the correct value; their figures relating to the motions of sun and moon, which form a continuous record starting with the reign of Nabonasser 747 B.C., were the foundation on which the Ptolemaic, and later the Copernican, sy
stems were built. Theirs was certainly an exact science, and it 'worked'; but that does not prove the truth of their theories, which asserted that the planets were gods whose motions had a direct influence on the health of men and the fortunes of states. Columbus put his theories to a rather remarkable experimental test; what did the evidence prove? He and his contemporaries navigated with the aid of planetary tables, computed by astronomers who thought the planets ran on circles, knew nothing of gravity and elliptic orbits, yet the theory worked -- though they had the wrong idea why it worked. Time and again new drugs against various diseases were tried in hospital wards, and improvement in the patients' condition was considered experimental evidence for the efficacity of the drug; until the use of dummy pills indicated that other explanations were equally valid. Eysenck has questioned the value of psychotherapy in general, by suggesting that the statistical evidence for successful cures should be reinterpreted in the light of the corresponding numbers of spontaneous recoveries of untreated patients. His conclusions may be quite wrong; but his method of argument has many honourable precedents in the history of science. To quote Polànyi:
For many prehistoric centuries the theories embodied in magic and witchcraft appeared to be strikingly confirmed by events in the eyes of those who beheved in magic and witchcraft. . . . The destruction of belief in witchcraft during the sixteenth and seventeenth centuries was achieved in the face of an overwhelming, and still rapidly growing body of evidence for its reality. Those who denied that witches existed did not attempt to explain this evidence at all, but successfully urged that it be disregarded. Glanvill, who was one of the founders of the Royal Society, not unreasonably denounced this proposal as unscientific, on the ground of the professed empiricism of contemporary science. Some of the unexplained evidence for witchcraft was indeed buried for good, and only struggled painfully to light two centuries later when it was eventually recognized as the manifestation of hypnotic powers. [16]
It is generally thought that physical theories are less ambiguous than medical and psychological theories, and can be confirmed or refuted by harder and cleaner experimental tests. Speaking in relative terms, this is, of course, true; physics is much closer to the 'ultra-violet' than to the 'infra-red' end of the continuous spectrum of the sciences and arts. But a last example will show on what shaky 'empirical evidence' a generally accepted theory can rest; and in this case I am talking of the cornerstone of modern physics, Einstein's Theory of Relativity.
According to the story told in the textbooks, the initial impulse which set Einstein's mind working was a famous experiment carried out by Michelson and Morley in 1887. They measured the speed of light and found, so we are told, that it was the same whether the light travelled in the direction of the earth or in the opposite direction; although in the first case it ought to have appeared slower, in the second faster, because in the first case the earth was 'catching up' with the light-ray, in the second was racing away from it. This unexpected result, so the story goes, convinced Einstein that it was nonsense to talk of the earth moving through space which was at rest, as a body moves through a stationary liquid (the ether); the constancy of the speed of light proved that Newton's concept of an absolute frame of space, which allowed us to distinguish between 'motion' and 'rest', had to be dropped.
Now this official account of the genesis of Relativity is not fact but fiction. In the first place, on Einstein's own testimony the Michelson-Morley experiment 'had no role in the foundation of the theory'. That foundation was laid on theoretical, indeed speculative, considerations. And in the second place, the famous experiment did not in fact confirm, but contradicted Einstein's theory. The speed of light was not at all the same in all directions. Light-signals sent 'ahead' along the earth's orbit travelled slower than signals 'left behind'. It is true that the difference amounted to only about one-fourth of the magnitude to be expected on the assumption that the earth was drifting-through a stationary ether. But the 'ether-drift' still amounted to the respectable velocity of about five miles per second. The same results were obtained by D. C. Miller and his collaborators, who repeated the Michelson-Morley experiments, with more precise instruments, in a series of experiments extending over twenty-five years (1902 to 1926). The rest of the story is best told by quoting Polànyi again:
The layman, taught to revere scientists for their absolute respect for the observed facts, and for the judiciously detached and purely provisional manner in which they hold scientific theories (always ready to abandon a theory at the sight of any contradictory evidence) might well have thought that, at Miller's announcement of this overwhelming evidence of a 'positive effect' in his presidential address to the American Physical Society on December 29th, 1925, his audience would have instantly abandoned the theory of relativity. Or, at the very least, that scientists -- wont to look down from the pinnacle of their intellectual humility upon the rest of dogmatic mankind -- might suspend judgement in this matter until Miller's results could be accounted for without impairing the theory of relativity. But no: by that time they had so well closed their minds to any suggestion which threatened the new rationality achieved by Einstein's world-picture, that it was almost impossible for them to think again in different terms. Little attention was paid to the experiments, the evidence being set aside in the hope that it would one day turn out to be wrong. [17]
So it may. Or it may not. Miller devoted his life to disproving Relativity -- and on face value, so far as experimental data are concerned, he succeeded.* A whole generation later, W. Kantor of the U.S. Navy Electronics Laboratory repeated once more the 'crucial experiment'. Again his instruments were far more accurate than Miller's, and again they seemed to confirm that the speed of light was not independent from the motion of the observer -- as Einstein's theory demands. And yet the vast majority of physicists are convinced -- and I think rightly so -- that Einstein's universe is superior to Newton's. Partly this trust is based on evidence less controversial than the 'crucial' experiments that I have mentioned; but mainly on the intuitive feeling that the whole picture 'looks right', regardless of some ugly spots that will, with God's help, vanish some day. One of the most prominent among them, Max Born, who inclines to a positivistic philosophy, betrayed his true feelings, when he hailed the advent of Relativity because it made the universe of science 'more beautiful and grander'.
Paul Dirac, undoubtedly the greatest living British physicist, was even more outspoken on the subject. He and my late friend Erwin Schrödinger shared the Nobel Prize in 1933 as founding fathers of quantum mechanics. In an article [17a] on the development of modern physics, Dirac related how Schrödinger discovered his famous wave equation of the electron. 'Schrödinger got his equation by pure thought, looking for some beautiful generalization . . . and not by keeping close to the experimental developments of the subject', Dirac remarks approvingly. He then continues to describe how Schrödinger, when he tried to apply his equation, 'got results that did not agree with experiment. The disagreement arose because at that time it was not known that the electron has a spin.' This was a great disappointment to Schrödinger, and induced him to publish, instead of his original formula, an imperfect (non-relativistic) approximation. Only later on, by taking the electron's spin into account, did he revert to his original equation. Dirac concludes:
I think there is a moral to this story, namely that it is more important to have beauty in one's equations than to have them fit experiment. If Schrödinger had been more confident of his work, he could have published it some months earlier, and he could have published a more accurate equation . . . It seems that if one is working from the point of view of getting beauty in one's equations, and if one has really a sound insight, one is on a sure line of progress. If there is not complete agreement between the results of one's work and experiment, one should not allow oneself to be too discouraged, because the discrepancy may well be due to minor features that are not properly taken into account and that will get cleared up with further developments of the theory
. . .
In other words, a physicist should not allow his subjective conviction that he is on the right track to be shaken by contrary experimental data. And vice versa, its apparent confirmation by experimental data does not necessarily prove a theory to be right. There is a rather hideous trick used in modern quantum mechanics called the 'renormalization method'. Dirac's comment on it is:
I am inclined to suspect that the renormalization theory is something that will not survive in the future, and that the remarkable agreement between its results and experiment should be looked on as a fluke . . .