The Limits of Science 
by Peter Medawar.
Oxford, 108 pp., £7.50, February 1985, 0 19 217744 3
Show More
Show More

Medawar presents an erudite and entertaining account of the limits of science, or mostly the lack of them, as perceived by great thinkers from Francis Bacon to Karl Popper and himself. His arguments are couched in largely epistemological terms which do not arouse my passions, but they stimulated me to think about those limits that affect laymen’s attitudes to science, about the practical limits scientists face in their everyday research, and laymen in their daily lives, and about the limits that affect industrial and public policy.

Medawar argues that science reaches its limits only when we ask ultimate questions of our existence, such as ‘How did everything begin?’ or ‘What are we all here for?’, but many other people’s attitude to science is more affected by its inability to answer the question ‘How should we behave?’ In their view, science has undermined the religious basis of morality and the belief in Heaven and Hell without putting anything in their places. It has given man immense powers over nature without suggesting any ways of improving the nature of man. Many regard these as science’s greatest failings. In his book Chance and Necessity Jacques Monod suggested that scientific truth might become the basis of a new ethics, but he did not spell out how this was to be achieved. Philosophers have shown that science can tell us only what is, and that an ‘ought’ cannot be derived from an ‘is’.

Medawar asks if there is some intrinsic limit to our understanding of the natural world, either cognitive, having to do with our powers of apprehension, or logical, arising out of the very nature of thought. His denial of both kinds of limit may be valid on philosophical grounds, because science has been defined as embracing all those problems that are in principle open to empirical observation and solution, but it is contrary to practical experience. Whatever I did discover I could have discovered years earlier, and many things that I failed to discover I could have discovered if it had not been for my limited powers of apprehension and logical thought, my obtuse blindness to the answers that lay at hand. History shows that even the greatest scientists usually advance in small steps, because the development of new concepts causes them enormous difficulties. For example, in retrospect an experiment on the scattering of alpha-particles from a gold leaf, performed by Geiger and Marsden in Rutherford’s laboratory in Manchester, ‘obviously’ suggested that the mass of the gold atom was concentrated in a tiny nucleus, but so novel and revolutionary was the idea that it took Rutherford more than a year to formulate it. I feel the lack of my own powers of apprehension and logical thought most acutely when I try to think about the evolution of the large biological molecules whose complex structures I have helped to determine, but since an explanation of this would take me deep into chemistry and physics, let me discuss instead some similar riddles that arise when we try to think about the evolution of birds.

I am a firm believer in Darwinian evolution by random mutation and natural selection; my belief has been strengthened by molecular biology, which has given us a detailed picture of the chemical basis of inheritance and of the many different ways by which chance mutations can alter the genetic apparatus. Yet I am baffled when it comes to a simple question like the evolution of birds’ wings. Darwinian evolution is driven by selective pressure, but the problem is that such pressure can act only after a rudimentary function has developed to a stage when it is of survival value. Fossils indicate that birds have evolved from crocodile-like reptiles that started to climb trees. Formation, even of rudimentary wings, would have required many separate mutations in successive generations, but until the wings had become large enough at least to allow these animals to glide from branch to branch, none of them would have made the animals fitter, and consequently there would have been no selective pressure to stabilise these mutations. This kind of enigma arises in the evolution of many other organs.

Even harder to explain is the evolution of bird migration. Warblers spend the summer months in Northern Europe. In the autumn they take off to Africa, either via the Balkans to the Nile Valley, or via Gibraltar to West Africa. They don’t fly in flocks, but singly by night, and birds hatched in Europe and flying south for the first time make the journey alone. How do they find their way? Experiments show that they navigate by the constellations of the stars. Other birds navigate by the sun as well as the stars. Some, perhaps many, also carry a compass in their brain, in the form of tiny crystals of magnetite, and navigate by the magnetic field of the earth. How could this behaviour have evolved? What use was selective pressure for the evolution of an instinct to fly south in the autumn and north in the spring before the evolution of navigation? What use would the evolution of an instinct for navigation by the stars have been to African birds before the evolution of an instinct to seek better feeding grounds in Northern Europe in the summer? How could genetic mutations have led to the crystallisation of magnetite in the brain to serve as a compass? When we think about possible ways of gradual evolution of these instincts, we come up against similar difficulties as in the evolution of wings. Recent observations brought us a small step nearer to an understanding by showing that a very few genetic mutations have been sufficient to produce striking changes in the appearance of certain varieties of maize and fruitflies, which suggests that the same might have been true for the evolution of wings. Even so, I am not sure that our imagination will ever completely fathom processes that have taken nature millions of years to accomplish.

Powers of apprehension and logical thought fail us when we try to explain these wonders of nature, but even if these powers were unlimited, science itself defines intrinsic limits to our perception of the world. In physics, these arise from Heisenberg’s uncertainty principle, which does not allow us to pinpoint the position of an atomic particle without blurring the measurement of its momentum, and vice versa. The principle also states that the shorter the lifetime of an atomic event the more blurred becomes the measurement of the energy associated with it. These restrictions affect many experiments in physics and chemistry. Uncertainties that are less fundamental, but similar in their effects, beset other fields of enquiry such as the study of behaviour in man and higher animals. For example, it has proved impossible to measure children’s innate intelligence, because the contributions of nature and nurture to mental development are inextricably mixed. Behaviour also tends to be affected by the very process of observation: tests of medical treatments can give ambiguous results, because many clinical conditions improve with the patient’s knowledge that he is being treated.

Medawar seems to regard science itself as having no frontier, but this view is not shared by all scientists. Leading physicists have sadly confessed to me that in pure physics there are now no fundamental problems left outside elementary particles. Many ambitious physicists have therefore turned to apply physics to either astronomy or biology. Fifty years ago already the great Danish physicist Niels Bohr advised his pupils that new laws of physics were most likely to be discovered in biology. Some of the young men who followed his advice are among the founders of the new science of molecular biology, but they had to make do with the known laws of physics, and rather elementary ones at that. Of course, Bohr was also wrong in believing that after relativity and quantum mechanics there was nothing fundamental left to be discovered in physics. Amongst other things, he failed to foresee the discovery of phenomena in solids which, though not fundamentally new, were quite unexpected and opened the way to the invention of transistors and high-speed computers, but despite these successes many physicists feel that there is not another wholly new set of phenomena like radioactivity, semi-or superconductivity, still waiting to be discovered.

Until Rutherford succeeded in splitting them, atoms were regarded as the ultimate indivisible units of matter. Since then, physicists have discovered many subatomic particles, suggesting, as some pre-Socratic philosophers thought, that matter might be infinitely divisible, but in fact there is likely to come a day when all subatomic particles will have been discovered. The latest particles, the W and the Z, have been detected at the European High Energy Generator in Geneva, which cost about £500m to build. A much larger accelerator now planned by American physicists is estimated to cost about five times as much. The American Government has not yet approved its construction, but supposing it does, no government or combination of governments is likely to pay for another accelerator costing five times as much again, or £12,500m, especially since none of the subatomic particles discovered during the last forty years have so far found practical applications. Those that are of practical use, the neutron, the positron and the muon, were discovered in the Thirties and Forties without the help of accelerators. On the other hand, high-energy physicists argue rightly that their technology has produced important spin-off for industry, science and medicine, such as the production of more powerful new radiations for the diagnosis and treatment of cancer. This brings us to the applications of science and their possible limits.

In medicine, the past forty years have seen the discovery of drugs against many hitherto incurable diseases. Is this stream inexhaustible? There is no absolute limit to the new drugs that might be discovered as there is to subatomic particles, but their discovery is becoming harder. In 1906, when Paul Ehrlich discovered his magic bullet against syphilis, he called it compound 606, because he and his collaborator had synthesised 605 compounds of arsenic before they hit upon the one that killed the pathogen and left the patient alive. By now drug firms have to synthesise an average of 8000 organic compounds before finding a marketable new drug. The average research and development costs of a new drug rose fivefold in real terms between 1960 and 1975 and now stand at over £50m. This means that only the largest firms can afford them, and then only if they are confident of mass sales. The recently introduced techniques of genetic manipulation may lead to a surge of new types of therapeutic agents, but not necessarily at lower cost.

In agriculture, yields of food grains have increased at an annual rate of 2 to 3 per cent, thanks in part to better pesticides. Since their period of application is limited by the growth of pesticide-resistant strains, the chemical industry continues to search for new ones, but this is becoming progressively harder, just like the discovery of new drugs. The number of compounds chemists had to synthesise to develop one marketable pesticide rose from 1800 in 1956 to about 10,000 in 1976, with a concurrent rise in development costs to the order of tens of millions of pounds. There are no absolute limits to the discovery of new drugs and pesticides, but financial limits are set by the law of diminishing returns.

The huge numbers of compounds that chemists have to synthesise before they find a marketable product accounts for only part of the large rise in development costs. The other part is due to the many additional safety tests made compulsory for new drugs after the thalidomide tragedy, and for new pesticides after the poisoning of birds and animals by DDT. In the USA, the execution of these tests and the official examination of their results has now become so cumbersome that up to ten years may elapse between the patenting and marketing of a new drug, but even after exhaustive tests absolute safety of drugs is beyond the limits of science, because individual responses vary so widely that seemingly harmless drugs like aspirin may kill people who are excessively susceptible to them.

In this field as in others, ignorance of science often goes hand in hand with unlimited expectation of its benefits. It is hard to convince the Pope that science will not be able to continue increasing food production indefinitely for an exponentially growing world population. President Nixon was confident that given enough money, science could find a cure for cancer just as fast as it had put men on the Moon. President Reagan seems to believe that science can provide continued economic growth regardless of the limits of our planet’s resources and environmental tolerance. At this moment, I believe he is wrong in trusting scientists to be capable of developing a system of computer-operated weapons in space that will automatically and reliably destroy hundreds of Soviet missiles within minutes of their launch. Even if it were to prove possible to make the required weapons, and that is still doubtful, the electronic circuits and computer programmes needed to operate them faultlessly and promptly would have to be devised and constructed not by infallible robots, but by men and women whose errors are liable to multiply with the complexity of the system. How often have we read of a space launch failing despite meticulous servicing because of a trivial fault: how much truer will this be of an entire weapons system left unattended in space? To my mind it is important to make the public and political leaders aware of science’s limits as well as its potentialities.

In summary, science cannot improve human nature nor provide a moral code. Individual powers of apprehension and logical thought are limited, and these limits can restrict our understanding of the natural world even when in principle all the facts necessary for our understanding are known. Further, the laws of physics, and the effects of the observer on the observed, place limits on attainable knowledge. Science is also limited in extent: no field of science has an unlimited frontier, even though in biology especially that frontier is still distant. In applied science, limits arise from the law of diminishing returns, and absolute safety can be bought only at infinite expense.

Should adventurous young people then turn away from science because there are no more Everests to be climbed? Certainly not. Some of the climbing may have become harder but there are many challenging mountain ranges still on the horizon. In my own field of molecular biology, the stream of fundamental discoveries shows no signs of drying up. What is more, it is beginning to yield results of practical benefit to medicine. While there may remain few fundamental laws of physics to be discovered, physics is being applied to the invention of undreamt – of advances in medical diagnosis: for example, sophisticated physical instruments now allow us to view a patient’s innards in sections without actually having to cut him into thin slices. The most recent method, an offshoot of high-energy physics called Positron Emission Tomography, lets you watch a person’s brain as he thinks. Thinking needs chemical energy in the form of glucose which the blood supplies and which the brain breaks down. The machine locates the centres of thought by finding the exact positions where this breakdown occurs. If I were writing with my head in one of these machines, it would light up the areas where glucose is being burnt to formulate my concluding sentence, but, fortunately perhaps, a machine that could also read my thoughts looks like being beyond the limits of science.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences