II. Developments in the Logic and the Sociology of Science


   If, at this distance in time, neither Mapother's nominalism nor Golla's vitalism-holism are at all satisfying, this is because so much has happened to the philosophy of science in the interval. The writings of Sir Karl Popper have had a revolutionary influence; and the important work he has done over thirty years or so is readily available in The Logic of Scientific Discovery (LSD, 1968) and Conjectures and Refutations (CR, 1969). Popper starts from the position that no logical basis for the accumula­tion of knowledge can be laid on inductive processes (and he also maintains that this is not the way the mind of the scientist works). From observing crows, and seeing that one after another they are all black, one cannot conclude that all crows are black. [1] In general terms, it is not possible to proceed logically from single existential ('there are ...') statements to universal statements; and an epistemological process that based itself on induction would be rocky at its very foundations. How is it then that science progresses, making mistakes from time to time, but recovering from them, providing for itself a firmer foundation for further work, anchoring it more deeply and extending it more widely? This could only be if the basic logical processes were not, in fact, inductive, as the nineteenth century scientist supposed. The search for an acceptable basic principle came down to finding a criterion of demarcation which would distinguish the em­pirical sciences from other branches of learning. Popper found it in the principle of refutability.

    Though existential statements cannot be used as a basis for asserting universal statements, they can he used to refute them. The universal statement 'all crows are black' is refuted by breeding a white crow from a pair of black parents. The way science advances, Popper says, is by making a universal statement, which without logical foundation is, by itself, in the nature of a guess or conjecture. It will then be refutable by some existential statement; indeed, it. will imply the prohibition of a class of existen­tial statements. A theory, or hypothesis, which is in the nature of a universal statement, will be the better the wider the range of existeniai statements which it prohibits. The next step will be to try to produce one of the forbidden phenomena, and if possible one that goes to the very heart of the universal statcnsent that is the scientific theory. !fan experiment can be con­ducted in which the forbidden phenomenon is shown to occur, then the theory is overthrown; if the phenomenon does not occur, then the theory is to that extent confirmed. But it is not proved; the theory will have other implications than the one tested, and a later test may still refute it. Our acceptance of scientific theories, then, is always provisional. This formulation emphasizes the way in which scientific theories, by suggesting ways in which they may be tested, provoke experimentation and therewith the growth of knowledge.

   A number of important consequences follow on the adoption of the criterion of refutability. The first is that there is an element of un­certainty even about refutation. There can always be a mistake made in the conduct of a critical experiment, so that the refutation of a theory cannot rely on a single experiment alone. Refutation experiments or observations must be repeatable. Popper writes (LSD): 'If accepted basic statements contradict a theory, then we take them as providing sufficient grounds for its falsification only if they consti­tute a falsifying hypothesis at the same time'. The scientist cannot be without his instruments. Even if a theory is beginning to break down, one has to go on using it for suggesting further work to do, until an alternative theory can be offered as an instrument in its place.Just as there is nothing quite clear‑cut about a refutation, so there is a theoretical element in every observation. For instance, the statement 'this object is made of glass' clearly contains within it a load of physical and chemical theory about glass and glass‑making. With reference to clinical medicine, Popper writes (CR):

Clinical observations, like all other observations, are interpretations in the light of theories and for this reason alone they are apt to seem to support those theories in the light of which they were interpreted. But real support can be obtained only from observa­tions undertaken as tests, by attempted refutations; and for this purpose criteria of refutation have to be laid down beforehand.

    It has been said that after a theoretical exposi­tion one should always ask, 'But, sir, what experiment could disprove your hypothesis ?', and, after every account of an experiment, 'But, sir, what hypothesis does your experiment disprove?'

   It also follows from the criterion of refutability that £10 theory is useful, or capable of carrying information about the empirical world, unless it has implications which are capable of clashing with experience; or, more precisely, unless those implications can be systematically tested. Trivial theories are those that have trivial implications; great theories have wide‑ranging implications which can be tested in a number of fields, and, perhaps, have to be tested over a very wide range to be tested at all. One might say that this was the case with such a great theory as the theory of Evolution.

    We must also take note that the content of a theory has an inverse relationship with its probability in the Keynesian sense, with its plausibility. The less the theory implies, the more probable it is, and the more useless for heuristic purposes. We can illustrate this fron), our teaching of psychiatry. We teach our students that mental disorders are the complex products of interaction involving environmental stimuli and stresses with. the complex psycho­biological entity of the individual personality, having a constitution derived from genetical make‑up, early life experiences, strains resulting from previous environmental pressures, etc. This is the 'theory' that Per Dalcn has mocked with the comment that it is like saying that bad weather is due to the complex interaction of geogenic, heliogenic and seicuogenic factors. It is only by breaking such an extremely plausible, 'satisfying' and almost self‑evident (but quite unrefutabie) theory into its constituent parts that we are able to get hypotheses with which we can work scientifically.

   Suppose we break off part of the genetical bit, and say, 'there is a genetical component in the aetiology of schizophrenia'. This is an existential statement, and accordingly irrefutable as it stands. It must be translated into a universal statement which can be refuted by existential statements, i.e. by empirical observations. The universal statement is something like 'all schizo­phrenic illnesses occur indifferently of the genetical background, i.e. without evidence of correlation with genetical factors'. This state­ment is refuted by a range of observational data, e.g. by showing that the mnonozygotic twin of a schizophrenic is at greater risk than the same­sexed dizygotic twin, and so forth. One can then advance to a new universal statement, which might run, 'there is a genetical contribution in all cases of schizophrenia', or, for practical purposes, 'it will prove impossible to collect a systematic series of schizophrenics, in which evidence of genctical predisposition if looked for will not be found'. This statement is again refuted, since in the case of the schizophrenic psychoses manifested by patients with temporal­lobe epilepsies a genetical investigation yielded a negative result. The dialectical process of assertion and refutation leads to one forward step after another, leading in fact to a systematic advance, every falsification opening the vistas of new problems.

   One of the first requisites of a scientific theory is that it should he consistent, consistent that is to say both with the theoretical structure of science on which it abuts, and also, most necessarily of all, internally consistent. A self contradictory system is uninformative, since from it we can draw any conclusion we like. It can be shown by a purely logical argument (anti Popper has done this) that a self‑contra­dictory theory can embrace all statements, with none being found incompatible. An internal contradictoriness is particularly likely to appear when a theory involves forces working in oppo­site directions. There is no human behaviour of any kind which cannot be interpreted in the light of interaction, say, between God and Devil, or Life Instinct and Death Instinct.

   There are other important virtues to which a good scientific theory can aspire, such as universality, precision, simplicity and parsi­mony. With all this we must bear in mind what is the nature of a scientific hypothesis, that is a guess, a leap in the dark, the bolder the better, risking total destruction. At the back of these guesses we see the non‑scientific forces that go to the inspiration and moulding of scientific work: unconscious drives, fantasy, imagination, dereistic thinking.

    Popper denies that the central aim of science is the accumulation of knowledge. [2] Instead, it is rather to proceed from problems to problems, going always to ones that lie deeper. The accumulation of knowledge is a by‑product; and in this knowledge there is no final certainty. 'Every scientific statement must remain tentative for ever' (LSD). It is only in a wrong view of science that one can find the craving to be right. Popper writes (LSD, 281):

… for it is not his possession of knowledge, of irrefutable truth, that makes the man of science, but his persistent and recklessly critical quest for truth… Science never pursues the illusory aim of making its answers final, or even probable. Its advance is, rather, towards an infinite yet attainable aim: that of ever discovering new, deeper, and more genera] problems, and of subjecting our ever tentative answers to ever renewed and ever more rigorous tests.

   Although Popper is concerned with the empirical sciences, the generalizations he makes are not themselves scientific. Thus, his criterion of demarcation is in the nature of a proposal for an agreement or convention (LSD, 37), a proposition about logic not based on logic but to be justified by its results. Again, he does not maintain that the concept of causality, on which science in effect relies, is anything other than metaphysical. But it can be re‑worded as a methodological rule which supplies the same driving force (LSD, 61): 'that we are not to abandon the search for universal laws and for a coherent theoretical system, nor ever give up our attempts to explain causally any kind of event we can describe'. In the same way, his statement, quoted above, about the way science progresses, is a metaphysical one, irrefutable, nonscientific; the logic of scientific discovery is not itself a part of science. What Popper gives us here is a strategy of advance. It is open to the critic to reject it, if he can persuade himself that he has something better to offer in its place.

   Nevertheless, the plan is so rationally grounded in a basic realism, while it aspires so loftily towards truth, that it makes an immense appeal to working scientists. These are the men who accept the view that there is an adamantine reality outside our minds and independent of them, which we may aim to understand, and which may provide us with an unmistakable 'no' when our aim shoots wide; who believe there is a truth which we can aspire to by way of ever increasing verisimilitude; that while there is no logical basis for a belief in causality, it provides us with the motive to search for natural laws; that there cannot be an empirical statement with methodological consequences which would make us abandon that search.


    To the working scientist, the picture that Popper presents looks rather idealized. It is descriptive more of the progress of science than of the performance of the scientist. The man who has the big idea is likely to fight for it, and to seek confirmation rather than refutation. The innovator will be happy if he can show that his hypothesis explains a range of facts, especially if some of them are as yet inadequately accoun­ted for. By his achievements in this direction, he shows that his hypothesis has the form and the content to make it interesting. So he is very willing to leave the task of refutation to others, wishing them but ill success. The force of Popper's analysis lies in the fact that other scientists always do get around in the cud to attempting refutations, with the result that some things stand and others fall. Not only the noble spirit of the search for truth, but also the less beautiful but still human impulses of ambition, envy, and the rest, are channelled willy‑nilly into an additive self‑correcting process.

   Some scientists have also found themselves dissatisfied by Popper's devaluation of inductivism to a point where it is hardly allowed any part in scientific discovery. Neither science nor a scientific guess spring from a void; and faced by an uncomprehended world, the human mind, in the path of comprehension, begins by looking at the facts it is able to collect. The first step towards astronomy was the making of star maps and inventories; and the history of spectro­scopy 1870‑1900 was one of mapping, in which the spectra of the elements were described with ever increasing precision (l'earcc Williams, 1970) 'There is precious little "refutation" going on here, yet it would be hard to deny Angstrom the title of scientist.'

   Another case in point is the theory of Evolu­tion. Darwin relied on inductive reasoning to support his arguments. He showed that evidence for the processes lie hypothesized could be found in such a range of biological systems that one might think they involved all life and extended over all time. By formulating the hypothetical processes of environmental and sexual selection, he gave evolution an intelligible machinery. In the end the theory of Evolution came to stand as an immense ünifying conception, no longer refutable as a single whole by any limited series of observations, but still a good scientific theory in the sense that it was richly productive heuristically. Such large‑scale and complex theories, of which Freudian psychoanalysis is a psychiatric example, cannot fall victim to refutation by any single experiment; they have to bleed to death from many wounds, if die they must.

   The historical and conceptual background to the Darwinian revolution has been described by Ernst Mayr (1972), most instructively for our present purpose. The age of evolutionism began even before Buffon, and there was an ever increasing ground swell of evolutionary ideas from the beginning of the eighteenth century. A great deal of evidence had been accumulated by 1830; but professional geologists and biologists remained blind to manifestations of evolution staring them in the face. Creationism dominated the thinking of the leading biologists; and if belief in a single Act of Creation had to yield to evidence from. the fossil record, then repeated Creations must be postulated, up to fifty or eighty or more. Refutation, in a Popperian sense, was blocked by faith in a Creator God and in His wisdom and constant attention.

   In the way of advance to a truer understand­ing stood roadblocks, built on a firm theological foundation and cemented by the prevailing philosophy. This was one of an essentialism deriving from Plato. The world was modelled on fixed types and unchanging essences. It was felt that if deprived of their immutability species would have no reality. As change was unimagin­able, and as species had come and gone, they must have been repeatedly exterminated and repeatedly replaced. Essentialisni and crea­tionism, fitting one another like tenon and mortise, gave a reinforced solidity to dogma. A great time span was required for its erosion and for time acceptance, not only of a new theory, but of a new conceptual world.

   Whole groups of new ideas had to make their way against these resistances: recognition of the great age of the earth; abandonment of catastrophism and creationism, but abandonment also of the alternative of a steady state; the replacement of essentialism and norninalism by population thinking. There had to be a change from one Weltanschauung to another, with far reaching shifts of emphasis, perspective and significance in religion, philosophy and human­isrn. Some of the most important of these changes are still in process of birth. We still have to be weaned from the comforting belief in an auto­matic upward evolution; and we are constantly forgetting that change and development may proceed not only upwards, but in any direction whatever, including downwards, or regressively, or along a blind alley. In human societies, it is usually held, all change or any change is, ulti­mately, for the good. We still have very far to go to abandon anthropocentrisin. The unifying of man with the evolutionary stream was an idea of extreme distastefulness to the Victorians. Even now, to regard man, not as the Lord of nature, but as part of Nature, is an idea almost confined to biologists, and to all others anathema.

   Mayr maintains that the course of this cen­turies‑long debate does not fit well with either Popperian or Kuhnian models, Those who wished to replace the direct intervention of the Deity by the operation of natural laws (e.g. Chambers in 1844) were faced by demands for conclusive proof. Manifestly no such proof was possible; there can he no disproof of Divine intervention. As Mayr points out, much of science consists merely in showing that one interpretation is more 'probable' than another, or is consistent with more facts. Those taking part in the debate attempted to 'falsify' the statements of their opponents, but were often quite uncertain what argument or what kind of evidence constituted a falsification. There was a blind spot for the possibility that there might be more than two opposed alternatives, such as physical laws vs. plan of creation. It was said that the abundant regularities in nature demonstrate the plan of a Divine Intelligence, since they could not he the result of blind physical forces. The pre‑Darwinian biologists could not see that these arguments were irrelevant to a third hypothesis: descent with modification. Those significant regularities could be gradually evolved by natural processes such as are daily tobe observed. None of them imagined that both processes, extinction and speciation, were going on simultaneously, and the turnover of the fauna was the result of a balance. Two alternatives may appear to conflict and yet both be true.


    Mayr's instructive example shows us that science does not have to advance; it may remain bogged down for generation after generation. The Popperiani dialectic of conjecture and refu­tation can happily go on, getting nowhere at all, if men blinded by their preconceptions are aiming their conjectures at a target that exists only in their imagination. Then the questions that are asked are irrelevant to the crucial problems, and the answers received are unin formative or misleading. Time psychiatrist can think of other fields than pre‑Darwinian biology where devoted research goes on for decades, and the stream of ideas circulates in an eddy, turning over and over, always renewing itself and never leaving its origins behind. We have to bear in mind, in fact, that science is conducted by scientists.

   It is with the behaviour of scientists that Thomas Kuhn (1970) has concerned himself; and he gives us a very different picture from Popper's, based not on philosophy and logic but on a historiographic investigation of the development of science. While Popper started with a search for a criterion to distinguish science from other branches of learning, Kuhn begins with a consideration of its basic tools. He sees these as paradigms, that is universally recognized scientific achievements that for a time serve a community of scientists with model problems and model solutions. Kuhn emphasizes the importance of textbooks, from which the beginner learns an idealized version of these models, and so finds his place in an ongoing tradition. The paradigm is used like a map, and teils one about what is known of the country traversed so far, where one stands now, where one can go to next, and how to go about the process of map‑making. By virtue of his paradigm, the scientist knows what a datum is, by what instruments it may be retrieved, by what concepts it may be interpreted.

   There is, of course, a stage before there is any consensus at all, before a paradigm can he constructed. Perhaps psychiatry is not yet entirely out of the pre‑paradigmatic stage. At such a time there are scientists but not yet a science. Fact‑gathering then is a much more random process than it becomes when the ‑paradigm is available. One works on the data that lie ready to hand, and some of the material is too complex to be at all suitable. As Popper says, all facts have a theoretical aspect; and in the pre‑paradigmatic stage the information gained may be uninterpretabie, because theory is still in its infancy. At this stage, then, there are ‑many competing schools, and evidence of ‑progress, except within schools, is hard to find. The same range of phenomena, though perhaps not the same particular phenomena, are seen, from these different points of view, in a very different way. The differences between schools may he resolved by their gradual convergence, by one of the competitors losing its dynamism, or sometimes quite abruptly, as when biometrical genetics and Mendelisns were suddenly seen to be complementary instead of being rivals.

   The paradigm is always faulty; it never does account for all the phenomena to which it can be applied. But once it is there the productive work of normal science can go on. Problems are selected which are amenable to treatment with the means available. The means, conceptual and instrumental, are further developed and refined. The scientist knows what he wants to do and how to go about it. Progress is steady, and seems assured; there are no conflicts. With a wide acceptance of views held in common, the body of scientists becomes a community, talking to one another, rather isolated from the rest of the world. The isolation of the scientist is to a considerable extent a consequence of his education, which, unlike the education of the student in the arts and humanities, is not at all concerned with the work of the great men who have gone before him, but is based upon text­books. The community of like‑minded and limited men then becomes one great team, in which the individual tends to merge, perhaps the most efficient organization one could devise for the solving of puzzles.

   Paradigms are faulty, and they are not corri­gible. Now, it is in the nature of scientific work to push the paradigm as far as it will go, to extend its application to the point where it no longer provides an adequate fit for the hard and knobby face of reality. We then enter a crisis situation, in which ordinary science has to give way to extra‑ordinary science. Before the crisis is reached, things have probably been going slightly wrong for some time. The basic credo of the scientist, that science knows what the world is like, necessarily inspires defensive reactions. 'Normal science', says Kulm, 'often suppresses fundamental novelties because they are necessarily subversive of its basic commit­merits.' But eventually the novelties can be suppressed no longer. There comes a time when an apparently normal problem, one that ought to be solved by the known rules, resists the attack of the ablest members of the groups within whose competence it lies. Everything is in a partial but uneasy disarray. Ad hoc adjustments to the paradigm are made, to the satisfaction of few. New ways are tried. Recourse is had to philo­sophy and reconsideration of fundamentals. Then there comes the discontinuity, the break in normal research, and eventually the building of a new paradigm.

   This is a painful process, that goes through with difficulty and against resistance. The older and now inadequate paradigm is never aban­doned, whatever its faults, until there is a new one there ready for adoption. This needs must be, since there is no such thing as research without a paradigm. The invention of the new paradigm is always the work of men either very young or very 'new to the field. With habits of thought not so rigidly fixed, capable of standing a greater emotional strain, they are able to descend deeper into the disarticulation of the existing gestalt, permitting the momentary, confusion of figure and ground, and, in a moment of insight, seeing, all at once, the new gestalt.

   The transfer of allegiance from one paradigm to another is a conversion experience that cannot be forced. It may be beyond the powers of the older men; and, indeed, they are right to resist the change since the battle of the paradigms must be won and lost on their merits. Perhaps the lapse of a generation may be needed for the process to complete itself. Max Planck, in his autobiography, made the classic comment: 'A new scientific truth does not triumph by convincing its opponents and making them sec the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.'

   With two paradigms in the field, we have the stage set for the conduct of critical experiments. This is something that has recently occurred, quite suddenly, in the field of psychiatric genetics. That the risk of schizophrenia is greater for the sib or the child of a schizophrenic than for a member of the general population is something that can he explained either genetic­ally or by postulating common factors in the early environments. The genetical paradigm has been takers more or less for granted in European psychiatry; hut when American psychiatrists became aware of it, it seemed to them a heretical newcomer, it was therefore in America that it started to inspire doubts and questionings, and at last radical attempts to settle the question whether family resemblances in morbidity risks were to be accounted for on a genetical or an environmental basis. We have now had three critical experiments: the investi­gation of the monozygotic co‑twins of schizo­phrenic probands; the follow‑up of those children of schizophrenics who have been adopted and brought up in the homes of normal foster‑parents; and the comparison of the liabilities to schizophrenia of the biological and foster‑parents (and their relatives) of adoptees who have eventually become schizo­phrenic. Nature's answer to the question put to her in these three experiments, i.e. can the gcnetical contribution to the aetiology of schizophrenia be dismissed?, is the same in each case, an unambiguous 'no'.[3]

   There are, of course, many teachers (and even a few workers) to whom the clearest of the replies that Nature can give is, simply, un­acceptable. For their conversion to a more realistic point of view we may have to wait, as Planck's principle informs us, for the kindly offices of death. But critical experiments cannot be dispensed with; without them, argument alone is insufficient. The supporters 0f the two paradigms hack their fancy by applying them each in its own defence. The circularity is unavoidable. The two sides talk through one another, and for most of the time are at cross purposes. The early versions of the new para­digni are inevitably crude; no other paradigm solves all the problems it defines; each side can see many of the weaknesses of the other, and rather less of their own. One of the qualities that may help the new paradigm is an aesthetic advantage, an aspect of simplicity and elegance. Bit by bit more scientists are seduced into trying it out; the paradigm is strengthened, stream­lined, extended, explored further and further. In the end, if it is destined to survive, it will explain all that the old paradigm explained, and more. With good fortune, it will make contact with a neighbouring part of the field and show its potentialities for unification. Then, at last, it becomes standard and part of the establish­ment. The old textbooks that have had their day are thrown out, and the new ones are written to go onto the students' shelves.

   Kuhn presents us with a picture of the evolution of science which is complex, with linear develop­ment and the accretion of knowledge again and again being interrupted by a scientific revolu­tion. The growth of science is not so simple and cumulative as the textbooks make it appear, with their re‑writing of history and their retro­spective falsifications. The nature of the pro­blems that have obsessed each successive generation have been widely different; and we have forgotten more of the problems of our predecessors than we have ever solved.

   Can we say then that we approach always nearer to an ultimate truth? Perhaps, with every Kuhnian revolution, we lose, if not as much as we gain, yet something that is an essential part of a greater totality. Perhaps that greater totality is in any case a figment of our imagination. 'Does a field make progress because it is a science, or is it a science because it makes progress?' Kuhn points to a parallel between the evolution of organisms and the evolution of scientific ideas. Darwin abolished teleology, beloved of the theologian, from his Origin of Species, and accounted for evolution from the more crude and primitive to the more highly adapted and efficient by natural selec­tion, without the help of God. Similarly the debate between paradigms, by which one is abandoned and another survives, can be accounted for as a selection based on relative efficiencies, judged in the here and now. We cannot say that along this path lies the way to a transcendental goal. Compared with Popper's aspirant optimism, this seems a strangely dis­couraging conclusion to reach.


    Adherents of Popperian and Kuhnian points of view, not all of them fully committed, met in a lively symposium in 5965 (ed. Lakatos and Musgrave, 1970). Here Kuhn developed his thesis and made it more precise. Asked to say whether his account of the way in which scientists work and science itself moves was to be regarded as descriptive or prescriptive, he replied that it was both. That was the way science was done, and the way in which it should be done. The normal phases in scientific development, in which puzzle after puzzle is solved by the application of a standard para­digm, must not be despised. These modes of behaviour have certain essential functions; lacking an alternative that would serve similar functions, scientists should behave essentially as they do. This is, of course, no very satisfying answer. The fact that a system works gives us no reason to think that another system would not work heiter.

   In Kuhn's view, normal science (=puzzle­solving) cannot be validated by any absolute criterion, but only by ones which are in the last resort subjective. The values by which one can decide, for instance, whether a research pro­gramme is progressive or degenerative are those of the group mind, wholly, and there is nothing beyond. Once one has a good working theory, the time for steady criticism and theory prolifera­tion has passed. One can then work on eliminating conflicts between different theories or applications of a theory. It is not good strategy to be constantly picking at weaknesses in a theory and trying to set up alternatives. One should not over‑react to anomalies. Of course there must be a critical level at which a tolerable amount of anomaly becomes intolerable, but that does not have to be the same for everyone. The approval of the group is what matters, and group unanimity is of paramount value. Groups try to minimize the occasions for conflict and to reunite quickly, even at a price (e.g. by excluding a formerly productive member, or by subdividing the specialty).

   The Popperian attack on Kuhn's position seems to have been motivated in part by a difference in taste. Popper himself said he recognized Kuhn's 'normal' science, and didn't much like it; it was, in effect, 'applied' science. He considered that one would never learn about the aims of science by turning to psychology or sociology. Compared with physics, these two are riddled with fashions and uncontrolled dogmas. While the Logic of Discovery has little to learn from the Psychology of Research, the latter has much to learn from the former.

   However, the spear‑point of the critique lay in the fact that Kuhn teaches us no way to distinguish science from non‑science; puzzle solving as a criterion helps us not at all. The Kuhn ian description of 'normal' science could be just as well applied to, say, astrology, or to the increasingly refined techniques of the pro­fessional criminal to attain his aims. Science is something quite different. One may indeed work successfully with a theory without testing it, but it is essential that it should be testable. Thus one may abandon a theory, without testing it, simply because another and more testable theory presents itself Lakatos (1970) comments that, in Popper's view, virtue lies not in caution in avoiding errors (as in 'normal' science), but in ruthlessness in eliminating them. 'Intellectual honesty consists rather in specifying precisely the conditions under which one is willing to give up one's position. Committed Marxists and Freudians refuse to specify such conditions: this is the hallmark of their intellectual dishonesty. Belief may be a regrettably unavoidable bio­logical weakness to be kept under the control of criticism; but commitment is for Popper an outright crime.'

   The attack by Popper's critics does not seem to have been made on his criterion, but rather on his view of the way scientists behave, the critics maintaining that in their human frailty the scientists go bumbling along without much regard for their logical foundations. Thus Harré (1970) points out that in thinking scientifically one does not think only in propositions, but also in pictures, models, diagrams. We should think both of Picturing and Sentencing, both of imagining structures and formulating proposi­tions. 'Theory construction is primarily model building, in particular in imagining para­morphs – imaginary processes among real or imagined entities.' He regards it as a reductio ad absurdum of the deductivist position that, in accordance with the canons of received logic, the only relation theory can have with fact is that a theory or hypothesis is falsified by the discovery of states implying the contradiction of conclusions drawn from the theory. Since infinitely many theories can imply the same consequences, the falsifying of one of them can give no guidance in the rational choice of its successor.

   Harré is concerned more with what scientists actually do and think than with the logical basis on which their results can achieve semi‑perma­nence. He says plainly that his aim is to preserve the persistent intuitions of scientists against the philosophers. 'If received logic runs counter to an important intuition then I take it as a prima fade hypothesis that received logic is inadequate in some way.' Thus it is not in accordance with logic that statements can be ordered by the relative strengths of belief which scientists give [hem, from those which are easily abandoned to those which are treated as practically immune to falsification. It is a feature of scientific work that theory is sometimes used as a powerful corrective to naive acceptance of fact.

   However, Harré is with Popper and against Kuhn in holding to the concept of an objective reality or truth to which we may make an asymptotic approach. He lays great stress on 'causes' (rather necessary counters in a scientist's thinking), and rejects Flume's view that 'cause' can be replaced by a statement that X 'is the invariable antecedent of' Y. Causation involves understanding of the mechanism by which cause produces effect. 'The aim of science is to try to find the structures, states and inner constitutions from which the phenomena of nature flow... to look for the causal mechanisms of which the patterns and regularities of pheno­mena are the effects ... to find the inner aspects of things of which the phenomena are the outward aspects.'

   Kuhn may perhaps be allowed the last word here. Not only does knowledge increase, but ignorance also. Theory evolves, and the points of attachment, between theory and nature multiply‑but the interstices between these points of attachment get ever wider.


   To those who harbour the vain hope that we may ultimately achieve total understanding of nature, Kuhn's vision of our achieving only an ever‑increasing ignorance may seem shattering. To everyone else it should be bracing. Whatever happens we are set on a journey whose challenge and whose interest cannot grow less. But there are those who would deny us even this, scientists who confront science with a denial of her integrity.

   The critics come in the main from the social sciences where humane considerations rank high and rigour of denmonstration is seldom sought because so difficult to attain. The criti­cism that really goes home concerns not science but its practitioners, and less the scientists than their employers, the departments of state and the great corporations that find the funds and call the tunes. There is heavy endowment of enquiries designed to provide profits for the rich and to give more power to the powerful. Industry benefits while spiritual welfare starves. But even if what is principally at fault lies in the motiva­tions of power politics, scientists cannot be entirely acquitted. Scientific enquiry, to be effective, has to be conducted in an impartial spirit. One cannot be emotionally committed to getting one result rather than another. The scientist gets trained in emotional detachment, and may train himself beyond that to the point where lie becomes dehumanized. For there are scientists prepared to make a career in any field however antisocially orientated, for instance even in the further sophistication of armaments already powerful enough to destroy mankind. But these faults are human faults and not faults of the discipline. They do not touch the episte­mological status of science, which is attacked on other grounds.

   It is said that the scientific paper is a fraud, on two grounds. First, the presentation of the work­report in standard format, with introduction, discussion of previous work, materials, methods, results, discussion and conclusions, does not disclose the thought processes of the research and is a long way from describing how the task was performed. This complaint is without substance. A detailed account of how the problem was originally envisaged, what false starts were made, how it was re‑envisaged, how a genuine start was eventually made, etc., is not germane to the interests of the readers involved with similar problems. They wish to be shown, in the shortest compass, what the territory is that has been cleared, and how clear it is, in order that they can get on with their own work. A scientific paper is a communication to fellow scientists of the information which they wish to know; the devious journey by wtdch that infor­mation was eventually reached is strictly non­rclevant‑however interesting it may be, in the case of a great achievement, to a more general public who like to read about human foibles and emotional entanglements.

   The second attack on the scientific paper is that it is invariably subjectively biased. The scientist is the creature of his age and his back­ground, full of prejudices, led by presuppositions to seek ways of confirming them, providing accounts of what he has done which are coloured by emotional bias. Science itself is in hopeless case, since its subjectivities can never be eliminated‑‑any correction will inevitably he as biased as what is corrected. Those who make this attack are very often men who are corn­rnitted to liberal‑humane ideals; and their real target is the misuse both of scientific enquiry and of technological know‑how in the service of world atrocities. However, the attack is mis­guided in its aim, and leaves its target un­touched. It does not matter for the progress of science that scientific enquiry is made by human beings who have the defects, not only of those biases of which mention is made, but of countless other frailties. The discipline of science is such that errors of all kinds, occurring whether from technical or emotional causes, once on record, can he subjected to test. Facts can be shown. to be non‑facts, conclusions can be shown to be mistaken, in any work which can be checked by others. In the end, it can only be what is true in an objective sense, true for all time and for all persons, that will stand ii to critical test and retest. The emotional biases of scientists, with which the scientific left wing is so properly concerned, should not cause us uselessly to deny the self‑correcting nature of scientific advance, but rather set us to testing more rigorously the results claimed by workers who, we believe, have been sent on a wild‑goose chase by the demands of their unconsciouses.

   However, the attack on science by those who are horrified at the way in which science is misapplied can lead only to neglect of the real enemy. To take an example from the psycho­logical field, we have the sequence: (1) tests of IQ have shown American negroes scoring lower on average than Ameiican whites; (2) this observation has been used in argument by pressure groups determined to support anti­-negro social discrimination; (3) those to whom such discrimination is an abomination seek to deprive their opponents of an argument by attacking 'intelligence' as an operational con­cept, or 'IQ’ as held distinct from intelligence, or the psychological tests on. which estimates of IQ are made.

   This is a mistaken way of thinking, and leads to a mistaken strategy for reform. As the sociologist Bressler has pointed out (1963): 'the liberal emphasis on biological equality as the basis for social ethics is misplaced because it unnecessarily rests on empirical generalizations, which are, in principle, subject to verification and therefore vulnerable'. If one maintains that negroes should not be treated so badly because the IQ tests which show them at a disadvantage are no true tests of mental ability or potential, one runs the risk of being driven further and further into a corner with any continuing demonstrations, convincing to others if not to oneself, that these tests are valid in the context under discussion. The only durable strategy for reform is to maintain that human rights and human differences in mental ability do not belong in the same universe of discourse. Negroes should not be treated so badly because nobody should he treated so badly. To quote Bertrand Russell (1954):

Ethical considerations can only legitimately appear when the truth has been ascertained: they can and should appear as determining our feelings towards the truth, and our manner of ordering our lives in view of the truth, but not as themselves dictating what the truth is to be.


   The outstanding feature of scientific advance is its capacity for self‑correction, This cannot he accounted for on the basis of inductive reasoning. It would not be enough to accumulate observa­tions and then invent theories to account for them, for such conjectures would have no power to compel conviction, and a variety of theories could be advanced and held according to taste. In science theories are advanced, are criticized, are destroyed or survive criticism, and then perhaps become part of a generally accepted scientific world‑view. It is important to be able to distinguish those argumentative processes which are capable of compelling con­viction from those which may leave differences of opinion for ever unresolved. Karl Popper has proposed a criterion of demarcation for this purpose: a scientific theory is indeed a guess or conjecture, but is one which is capable of refutation. It is a generalization, a universal statement, which can be disproved by showing a contrary instance. If the theory explains how things happen one way, it denies that they happen another way; and it is the prohibition that matters, because this is what can he tested. The test is made by attempting to produce one of the forbidden range of phenomena. If such attempts at refutation fail, the hypothesis is to that extent confirmed. But it is not proved, since there will he other implications than the ones tested, and a later test may still refute it. Our acceptance of scientific theories is always provisional. This requires an attitude in the scientist of being willing to be proved wrong; though he may believe a theory, he should not be committed to it, or give it his unreserved loyalty.

   A consequence of adopting Popper's criterion is that we are enabled to recognize, among all the theories purporting to give information about the empirical world, those which are worth scientific attention arid those which are not. It is only theories which involve conse­quences capable of clashing with experience that are worth such attention. When we adopt a theory we must be able to say fairly precisely what observations we would accept as refuting it, and under what circumstances we should feel compelled to abandon it.

   This criterion is a serviceable one, covering all that we think of as the empirical sciences. The criterion excludes, for instance, law, theology, and history, for in them no part is played by deductive reasoning from empirical observations. It equally excludes philosophy, logic and mathematics, [4] disciplines not pri­marily concerned with observational subject matter and not dependent on it. If advance occurs in any of these fields, it is not because contentions are settled by the specific self­testing process which is at the heart of science.

   Popper's work still stands, when all the criticisms are heeded to which it has been subjected. The point has been justly made, by Kuhn and others, that scientists are various and conform to no invariable pattern in the way they work. We cannot deny that they are rooted in the ideologies of their era arid their society, brainwashed by their teachers, blinded by prejudices, hogticd by obsessions, beguiled by will‑o'‑the‑‑wisps, and. at times even tempted to cheat and to lie. These lapses cause no more than eddies in a river. The curiosity of the scientist about the world, his rejection of authority and appeal to fact, lead to an in­formation flow. Conjecture and refutation are integral to the flow, polarities necessarily engendered by the curiosity and the appeal. They are, in fact, higher order effects, and their interaction appertains to science and not to scientists.


[1] Bertrand Russell takes it to a paradoxical extreme: “… the rejection of induction makes all expectation as to the future irrational, even the expectation that we shall continue to feel expectations. I do not mean merely that our expectations may be mistaken; that in any case must be admitted. I mean that, taking even our firmest expecta­tions, such as that the sun will rise tomorrow, there is not a shadow of reason for supposing them more likely to he verified than not.” (History of Western Philosophy, 1946, p. 693). This is, of course, an absurdity, to winch one might reply by saying, 'Very well. I take it that, if you were a betting roan, you would be willing to wager Io at even odds that the sun will not rise tomorrow?' Russell has misused the words 'u‑rational' and 'shadow of reason'. There are abundant rational grounds for betting that the sun will rise tomorrow, but these are not logically com­pelling. Russell's illustration shows the difference between logic and rationality. A. J. Ayer (Probability and Evidence, London, Macmillan, 1972) concedes the point that, certain knowledge may be impossible in the realm of fact but asserts that there can still be highly probable know­ledge, and this kind of knowledge can be made logically respectable if not demonstrable. We do have justification for our factual inferences, and we can show our procedures to be rational. We have to gamble. He says that where he aiffcrs from Runic is in riot seeing this as a reason for scepticism.

[2] One must protest even more strongly against the view that it is the aim of science, or any acceptable aim for the human spirit, to. 'control nature'. It is the mis­conceived attempts to control nature rather than under­stand her and live with her that have led to the present critical disharmony between man and his environment.

[3] The twin experiment also puts the corresponding question about the environmental contribution to aetiology, again with a negative answer; the other two experiments are non‑contributory in this sense.

[4] I am indebted to Dr. Michael Slater for an elegant example showing the conceptual diflerence between mathematics and empirical science, the famous result by Vinogradov which says that every odd number, from some point (X) on, is a sum of three primes. This is meaningful and sound mathematically, since the existence of X is proved by extracting a contradiction from the opposite assertion. But it is not a scientific statement, since there is no way of finding what X is, so that the statement is incapable of clashing with experience. No number of trials producing odd numbers which were not the sum of three primes would invalidate it, since they would merely show that X had not yet been reached. Similarly, a demonstration that all odd numbers from, au', onwards up to any finite number are in fact capable of expression as the suns of three primes is irrelevant, since no proof is provided that X is not still to be reached.