Roger James Return to Reason, Open Books, Somerset, 1980
The importance of criticism
'The central mistake is ... the quest for certainty.'
Although by no means everybody makes a study of hilosophy in the way that we all study arithmetic,
everybody has what may be called his own philosophy, picked up in much the same way as he picks up his native language. But whereas his language is out in the open and subject to correction, and other people's language is out in the open and available as good or bad example, philosophies
tend to be private and seldom made explicit. This fact, Popper believes, is the main justification for the study of philosophy - because our private, almost unconscious, philosophies are, unless they have been clearly expressed and revised, usually full of errors and mistaken assumptions.
Popper cites a very common mistaken philosophical assumption which he calls the conspiracy theory. This is the assumption that bad things like wars, slumps, unemployment, rising prices etc., are the result of well laid plans by those who stand to gain privately by the public discomfort - armaments manufacturers, employers of cheap labour, profiteers of all kinds. In fact this is not true.
Nearly all those bad things are the unintended consequences of the actions of individuals or firms or
governments with quite other intentions. When you negotiate to buy a house, the last thing you intend is to put up the price of houses in that district. Nevertheless your action has that tendency; and if a number of people enter the market and bid for the same house, a rise of price is certainly the result. The enormous rise in property prices in the early 1970s was not the result of a sinister plot by
those who owned property, but the unforeseen consequence of well-intentioned, if naive, attempts to
enable more people to own their houses by, amongst other things, making available 100 per cent mortgages, without increasing the supply of houses. (We are now preparing to do it again - for `First-time buyers'.) Similarly the high unemployment together with high inflation which we are now experiencing (an impossible combination in a laissez-faire economy) is the result of `good' attemps to
prevent a wage explosion while at the same time providing unemployment - and supplementary - benefits at a level which discourages work for very low wages. Efforts to improve the productivity of industry, to increase the output per man, were not undertaken with the object of creating unemployment, yet unemployment has certainly been one consequence. What looks like conspiracy is more often what I have called tunnel-vision or mere muddle. `T do not wish to imply', Popper writes
that conspiracies never happen. On the contrary, they are typical social phenomena. They become important, for example, whenever people who believe in the conspiracy theory get into power. And people who sincerely believe that they know how to make heaven on earth are most likely to adopt the conspiracy theory, and to get involved in a counter-conspiracy against non-existing conspirators. For the only explanation of their failure to produce their heaven is the evil intention of the Devil,
who has a vested interest in hell.
Conspiracies occur, it must be admitted. But the striking fact which, in spite of their occurrence,
disproves the conspiracy theory is that few of these conspiracies are ultimately successful. (O. S., ii, 95) This is not to deny that there are occasional successful conspiracies. According to Mick Hamer (Wheels within'' Wheels) the strategic motorway network was such a one.
But they are not often successful, certainly not more often than the overt plans of individuals and governments. They are therefore not very important in their consequences – nothing like so important as the unintended consequences of well-meant or rational actions. The analysis and, where
possible, the foreseeing of these unintended `side-effects' is, Popper believes, the main task of the social sciences, not long-range prophecy. (It is only fair to say that Marx was probably the first to state this sort of view; but for a full discussion of this I refer the reader to The Open Society, ii, p. 323, note 11 to Chapter 14.)
The fallacy of the conspiracy theory is the first among several of Popper's philosophical conclusions which have consequences for the way we think about our private lives and public affairs, and which differ, in some cases rather strikingly, from the philosophical assumptions which underlie what we read in the press, see on television, and hear from public figures both in government and
A second fallacy is a particularly pernicious one: `A man's opinions are always determined by his economic or political interests'. This, Popper says, becomes `If you do not hold the same views as I, you must be dominated by some sinister economic (or political) motives' (Magee, 1971). The evil effect of this belief is that it makes discussion impossible. It diverts attention from the important question of what is the truth, what are the facts, and what can we learn from them, to the comparatively trivial question of motives in asking such questions and interpreting the answers. It
leads to a belief that only people who already share a framework of assumptions can hope to reach agreement in a rational discussion. This sounds plausible and reasonable, but has a terrible divisive effect. It breaks mankind into mutually exclusive ideological groups who cannot discuss with one another, only fight each other. `It ignores the likelihood that our western civilization is itself the result of the confrontation between societies who had no common "framework" - the Greeks and their neighbours, then the Greeks and the Romans, Romans and Jews and Germanic peoples, and later still, Christians and Moslems.'
It rests on a false assumption - that truth can be discovered only by eliminating bias and prejudice. Popper shows that this is impossible. Everybody is biased, prejudiced, and interested (in both senses of the word). The fact that, in scientific discussion, for example, and even in courts of law, we can approximate to the truth, is due to the public nature of the discussions, not to lack of bias in the
participants. It is on the contrary an advocate's job to be biased. We all know how, though knowing ourselves wrong, we can stick to our position in a private argument and just say `well, I don't agree'. But in the presence of several onlookers, even if they play no part in the discussion, it is far more difficult to hold on to a logically losing position. We know that somebody is going to say `It's no good. Admit you are wrong'. Freedom and publicity of discussion, not lack of bias, are the pathways to the truth.
Following on from this and fundamental to Popper's philosophy is the importance of criticism. Some of
Popper's ideas have filtered, often in a somewhat garbled form, into popular thought and have had some influence; but some of the most important remain known only to the few people who have studied his works in detail. The idea of criticism as the source of the growth of knowledge is one of these.
Criticism is one of the functions of language and, in evolutionary terms, probably the most recent. The first two functions - expression of feeling, alerting calls and signals - are possessed by many animals. For these, even in man, words are not necessary. The chewing of gum or the puffing at a pipe can say: `I am not worried' or `I don't care' and in so saying often tell a lie. A third function, description, is possessed perhaps in rudimentary fashion by some animals. Bees seem to be able to tell each other
where nectar is to be found. This descriptive function introduces the standard: true or false. Criticism is the fourth function and is peculiar to man. It arises from the third. The way to the nectar is disputed or there is more of it over there. This critical function introduces the idea of validity, in argument and it does require words (see O.K. 235). Knowledge grows by criticism, by the weeding out of wrong ideas, just as the species of animals and plants develop by the elimination of maladjusted ones, and skills
by the elimination of useless movements and habits. It is a case of the survival of the fittest - in ideas, organisms, and techniques - as a result of the getting rid of the unfit.
The problem of induction
I must substantiate this view and emphasise its far-reaching consequences for human thought by going back to David Hume and the problem of induction. Until Hume demolished the idea in his Treatise on Human Nature, published in instalments from 1737 to 1740, it was generally' agreed that the rationality of science depended on the process of induction, that is of making generalisations
from a limited number of facts. We observe that the sun rises every morning and we conclude from this that it will rise again tomorrow and every other morning - the more frequent the observation, the more secure the generalisation. Hume showed that the logical process was not valid, that there is no 'reason for generalising in this way; and so scientific laws derived in this way could not be said to be founded on reason and experience.
Having established that because A follows B a hundred times there is no reason to suppose that on the 101st occasion A will again follow B, Hume fell back on a weaker, psychological, theory. It was that, although there is no reason to suppose this, it is in fact what we all do. We observe repetitions and then act on the assumption that they will go on happening. In spite of its logical invalidity, induction is indispensible in practical life. We live by relying on the continuation of repetition. Association strengthened by repetition is the main mechanism of our intellect, by which we live and act.
So we are left with a paradox - even our intellect (not just our emotions or intuition) does not work rationally. The pre-dawn human sacrifices of the Aztecs provide a chilling example of the consequences of an unquestioning belief in association. The sun invariably rose after the sacrifice; the practice therefore had to go on to ensure that it would continue to rise every morning.
I am not going to give Hume's arguments for rejecting induction because they are intertwined, as was the Aztec reasoning, with ideas about causation; and this complicates the matter. But one can easily imagine how he came to be led towards his conclusion; for, in practice, repetitions do not always occur. Rain may follow the dawn on six days of the week, but this makes it less rather than more likely
that there will be rain on the seventh day. Bertrand Russell (1946) said of Hume that he developed to its logical conclusion the empirical philosophy of Locke and Berkeley and, by making it `self-consistent, made it incredible'. British empiricist philosophers from then on, according to Russell, preferred to reject Hume's scepticism without ever refuting it, while German philosophers simply ignored it. Hume himself said that his treatise `fell dead-born from the press'. Hume's rational scepticism was ignored and smothered by Rousseau's romantic irrationalism. `Rousseau was mad but influential', was Russell's comment, `Hume was sane but had no followers.'
Nevertheless Hume's success in apparently proving that experience and reason have no necessary connection with one another, that there is no such thing as rational belief, was `an intellectual time-bomb which after sizzling away for two hundred years has only just gone off', according to ord (Kenneth) Clark. If not even science was rational the way was clear for mysticism. To quote Russell again: The growth of unreason throughout the nineteenth century and what has passed of the twentieth is a natural sequel to Hume's destruction of empiricism. It is, therefore, important to discover whether there is an answer to Hume within the framework of a philosophy that is wholly or mainly empirical ... If not, there is no intellectual difference between sanity and insanity, the lunatic who believes he is a poached egg is to be condemned solely on the grounds that he is in a minority. This is a desperate point of view and it must be hoped [my, italics] that there is some way of escaping
I shall take up in Chapter 6 Hume's psychological theory and give Popper's answer to it. The mistaken quest for certainty was the factor that misled those rationalist philosophers who genuinely sought a solution. Russell (1948) recognised the fact of uncertainty. With a characteristically homely comparison, he put it like this: `All knowledge is in some degree doubtful, and we cannot say what degree of doubtfulness makes it cease to be knowledge, any more that we can say how much loss of
hair makes a man bald.' Possibly he was diverted by the vividness of his own imagery into asking, by implication, the wrong question. For it is not a matter of defining the word knowledge but, as we shall see, of ranking approximations to the truth. The search for a solution within an empirical philosophy was also misleading. The answer is rational but not empirical.
It was when Popper considered the impact made by Einstein on Newton's theory of gravitation that he realised that the solution lay in this very uncertainty about our knowledge of the external world. Newton's theory had been - still is - so astonishingly successful that the problem always seemed to be how to explain man's ability to know about the universe so exactly. The fact that Einstein's totally different theory explained some observations better than did Newton's made Popper realise that both are theories - hypotheses, not facts. This thought then led to Popper's solution of the problem of induction. This was not the disproving of induction. That had already been done by Hume two centuries before, to his and Russell's and many others' satisfaction. It was to explain how ational action is nevertheless possible, how we can be reasonably sure that the man who believes he is a poached egg is wrong.
Popper's solution to this important problem is easily explained. He simply denied that we do act inductively on the assumption that repetitions will go on recurring. We do not act upon the assumption that the future will be the same as the past (when we are behaving ationally) but upon the best-tested theory - the theory for which we have the best reasons for believing that it is the nearest to the truth (O.K.95). It is true that we act upon the assumption that the sun will rise tomorrow; but this is not only because it rose today and yesterday, but because it is the best theory. If we were to believe that another celestial body was going to pass near the earth in such a way as to stop its rotation, then we should act on the assumption that the sun would not rise or not set, whichever the best theory might predict. In fact the sixteenth century mariner exploring uncharted seas did not assume that because he saw no land for twenty days on and that he would never see land again. On the contrary. Bertrand Russell (1946), using one of his favourite food examples, said: `I see an apple, past experience makes me expect that it will taste like an apple, and not like roast beef; but there is no rational justification for this expectation.' Popper's answer to this would be that we expect the apple to taste of apple not only because of past experience but because it is the best theory. Changing to a less fantastic example: if we are told that what looks like a ripe apple is in fact unripe or over-ripe or riddled with maggots we should expect it not to taste like a good ripe apple. Our expectations are certainly not based only on past experience.
This is all there is to it. Popper is saying that we do not act inductively - at any rate when we are using our reason - but on the best available theory whether it involves repetitions or not.
To sum up: the process that looks like induction is really one of choosing the theory that is best supported by reason. It is one of a number of similar conceptual illusions. I have already mentioned tunnel-vision or muddle looking like conspiracy. Another is natural selection, which looks like instruction. It looks as though the giraffe acquired its long neck because its ancestors wanted to
nibble higher branches and taught their young to stretch their necks and they taught their young likewise. What actually happened, we now believe, was that those pre-giraffes with longer than average necks survived better because they could nibble higher; and, where the longer neck was an hereditary variant, their progeny again survived better and so on. I mention two more of these
conceptual illusions in Chapter 6.
The demarcation of science
One of Popper's earliest philosophical innovations was to suggest a line of demarcation between science and non-science or metaphysics. It was that a scientific theory is, in principle, capable of being shown to be false. The impetus that led him to reverse what had hitherto been the generally accepted view that scientific theories could be confirmed by observations was provided by a number of friends of his youth who were admirers of either Marx, Freud, or Adler.
These theories appeared to be able to explain practically everything that happened within the fields to which they referred. The study of them seemed to have the effect of an intellectual conversion or revelation, opening your eyes to a new truth hidden from those not yet initiated. Once your eyes were opened you saw confirming instances everywhere: the world was full of verifications of the theory. Whatever happened always confirmed it. Thus the truth appeared manifest; and unbelievers were
clearly people who did not want to see the manifest truth, either because it was against their class interest, or because of their repressions which were still `unanalysed' and crying loud for treatment. (C. R.34)
Popper was impressed by his observation that `a Marxist could not open a newspaper without finding on every page confirming evidence for his interpretation of history' and that Freudian analysts found that their theories were `constantly verified by their "clinical observations"'. He realised that the great appeal of these theories was that they enabled one `to know in advance', as Bryan Magee (1973)
puts it, `that whatever happens one will be able to understand it'. At about the same time, in 1919, he heard that Einstein had said that %bservation failed to show the shift to the red of the lines of the spectrum as he had predicted, `then the general theory of relativity will be untenable'. Einstein would regard his own theory as untenable if it should fail in certain tests. This, Popper realised, was the true attitude of science. (U.Q.38)
In Popper's view all organisms are endowed with propensities, expectations, one could call them rudimentary theories. The newborn child expects to be led and cared for. But the expectations are not necessarily fulfilled and this leads to problems, a gap between theory and practice. We survive by learning to solve problems, and we do this by modifying our conjectures. `The new solution, new
behaviour, new theory, may work; or it may fail. Thus we learn by trial and error, or, more precisely, by tentative solutions and their elimination if they prove erroneous.'
This method is used by even the most primitive of animals; but its theories are its behaviour, and if they are wrong it succumbs. But there is a most important difference between what an amoeba does and what a scientist does. In a radio conversation with Bryan Magee, Popper explained it like this:
On the pre-scientific level we hate the very idea that we may be mistaken. So we cling dogmatically to our conjectures as long as possible. On the scientific level, we systematically search for our mistakes. This is the great thing; we are consciously critical in order to detect our errors. Thus on the pte-scientific level we are often ourselves destroyed, eliminated with our false theories. On the scientific level, we systematically try to eliminate our false theories, we try to let our false theories die in our stead. This is the critical method of error elimination. It is the method of science. It presupposes
that we can look at our theories critically, as something outside ourselves. They are not any longer our subjective beliefs [World 2, see page 25]. They are our objective conjectures [World 3].
(Modern British Philosophy)
Popper sums up the general picture of science as follows: We choose some interesting problem. We propose a bold theory as a tentative solution. We try our best to criticise the theory; and this means that we try to refute it. If we succeed in our refutation, then we try to produce a new theory, which we shall again criticise, and so on ... The whole procedure can be summed up by the words: bold
conjectures, controlled by severe criticism which includes severe tests. And criticism, and tests, are attempted refutations (Magee, 1971).
Thus a theory cannot be proved, it can only be disproved. But the more it stands up to attempts to refute it, the more secure it becomes, although it can never be regarded as certain. Popper encourages the formulation of theories in such a form as to say: `If such and such experiment were
performed with such and such a result then the theory would fail.' Scientific theories, as it were, forbid certain eventualities; and `the more a theory forbids the more it tells us'. While `a theory which cannot clash with any possible or conceivable event is ... outside science'. (U. Q.41)
To see straight away what is implied by this criterion of demarcation for science, one can look at the continuing controversy on intelligence (measured by I.Q.) and race. Professor Hans Eysenck, prominent among the protagonists, has recently stated, in an article in New Scientist (1979) that the basis of the psychologists' case is `the theory that all cognitive performances are to a variable degree a function of a single underlying ability (intelligence or g) ... ' Now this theory is so imprecise that one cannot conceive of any observation or experiment that could refute it. Yet endless discussion proceeds, looking like scientific argument and involving abstruse mathematics, of what is, by Popper's criterion, metaphysics.
Popper cannot identify the process whereby we get the hunches, or whatever, that form the theories that have to be tested- `New ideas have a striking similarity to genetic mutations' (S.B.54) - but he does suggest that they are not arrived at by induction. This is because the observation of repetitions is the result not the cause of the theory. We make a guess that A may be followed by B and then we
watch out and see if it is so. We do not go and observe with a blank mind. We have expectations and then we observe to see whether they are confirmed.
A good example of the contrast between the inductive approach and the Popperian view of science is provided by the attitudes of Francis Bacon" and Galileo in the seventeenth century. Bacon taught that men must rid their minds of prejudices (theories) and just go and observe and, as it were, `read the open book of Nature'. Galileo realised that all is interpretation, that we have to arrive at the correct interpretation of what we appear to see. He was full of admiration for the astronomers Copernicus and
Kepler because they refused to believe what they saw. Their eyes told them that the sun goes round the earth. Anybody can see it. You just go and look. Copernicus preferred to believe his reason which told him that this was an illusion.
The distinguished neuro-physiologist, Sir John Eccles (co-author with Popper of The Self and Its Brain), has recorded his personal experience of conversion to the Popperian view of science. At the time when he first met Popper, he was in a state of depression, desperately clinging to a theory he had proposed about the mechanism of nervous transmission which, in his heart of hearts, he had begun to suspect was wrong. Popper persuaded him that it was no disgrace to have one's hypotheses refuted. Indeed this was the means by which science advances. Encouraged by this view, he proceeded to join the `killing' of his own `brain-child' and to assist in the advancement of the theory of his rivals. `It was in this most personal manner', Eccles wrote, `that I experienced the great
liberating power of Popper's teachings on scientific method.' (The Philosophy of Karl Popper)
This story contrasts strongly with what happens in politics where the behaviour of politicians, civil servants, and political parties resembles much more the amoeba than the Popperian scientist. They can seldom admit themselves to have been wrong and they tend to bind themselves to ";particular solutions rather than to solving problems.
Richard Grossman relates in his diaries how he asked Mr Callaghan, then Home Secretary, why it was that he was now in April (1969) proposing an obviously sensible reform in respect of immigration which he had himself opposed in the previous October. What had made the difference? The answer was that the chief immigration officer had been shifted. Similarly, in his analysis of the way in which three municipal authorities dealt with their transport problems, John Grant showed how only by getting rid of the leading participants - in one case by defeat at the polls, in another by the death of the City Engineer - could the policy be changed. In the third case, where both party and leading actors survived, the road-building ambitions of the 1960s were only chipped away bit by bit.
Popper's great innovation, then, is that human knowledge, especially scientific knowledge, depends upon criticism for its advances. This idea is linked with another, which is also contrary to a popular saw - that you can never prove a negative. On the contrary, if you are talking about general propositions and theories, negatives are the only things you can prove. (The confusion arises from the fact that the opposite is true of singular events.) As Popper puts it: `No number of sightings of white swans can establish the theory that all swans are white; but the first observation of a black swan can refute it' - assuming of course that the total number of swans in the world is, not known.
I think it is fair to say that Popper's formulation of the process of growth of scientific knowledge is now generally accepted by natural scientists themselves. It is interesting to compare it with what had immediately preceded it, the theory held by the logical positivists that any theory that could not in principle be confirmed, was not just non-science but nonsense. In Popper's view what is non-science, as defined by his subtly different criterion, is not necessarily nonsense. Metaphysical ideas can, with the growth of knowledge, become scientific theories. To take a fantastic example, a century ago the theory that the moon was made of green cheese was metaphysical. There was no possible way of proving it false. Now it could be considered a scientific theory that has been conclusively disproved.
Or, as a realistic example, take the recently publicised theory that anorexia nervosa (complete loss of appetite and consequent severe loss of weight, usually in adolescent girls or young women) is caused by a fear of reaching maturity. Now this may or may not be true; but it would be difficult to put this hypothesis into a form such that it, or its logical consequences, could be refuted. I do not deny
that it may be a helpful idea to have in one's mind when dealiig with a particular case. But there is a great danger in believing it to be a fact. For such a belief will tend to make one disregard other evidence, especially any statements by the patient herself, which appear to contradict it.
Science is not necessarily connected with microscopes and electronic machinery. It is a method of acquiring knowledge. What is called scientific knowledge has been laboriously built up over the years and thoroughly tested but is not certain. It includes social as well as physical science. Theories which are not testable and, therefore, do not belong to science on Popper's criterion, may be
interesting and fruitful - as for example Marxism and psychoanalysis - but cannot command a degree of reliance in any way approaching that commanded by scientific theories that have stood up to tests.
In concluding this chapter I must emphasise the fact that the growth of scientific knowledge depends on the publication of theories and on freedom of speech in criticising them. Technology may thrive in secrecy and under tyranny but science cannot. Two groups of people are required for the growth of knowledge: those putting forward the theories and the critics, who may or may not be experts of some sort themselves, trying to find fault with the theories.