SCIENCE AND TRUTH
I come now to the problems that Karl Popper attacked when he started his philosophical researches. The problems were very well known. By now there is a tendency to declare them rather insignificant, but in the first quarter of the century they were very much alive and they had much to do with what is known as the crisis in physics.
The crisis in physics was once something everybody talked about. Now this isn’t so anymore; in fact mentioning the crisis in physics is quite uncommon in the current literature. The crisis in physics was the inability of physical theory to live up to its promise. The promise was that all predictions that were deduced from that theory would come true. It is indeed amazing how well physical theory, especially in Newton’s mechanics, had stood up to the promise. It is amazing how many predictions based on Newton’s theory looked false but upon further examination either were found to be true or turned out to be due to inexact calculating rather than to mistakes in the theory. There was, for instance, a calculation which showed deviations of Saturn and of Jupiter from the orbits calculated from Newton’s law, and people thought maybe this is the end of Newton’s law, but Laplace showed that the calculations weren’t good enough, and that if you used better ones you got the right results. The other deviations were found, and Adams and Leverrier suggested it was a mistake to think that there are only seven planets in the solar system, that Newton’s theory can be saved by the assumption of an additional planet. The additional planet was soon found. God obviously planted that planet up there in order to make Newton’s theory right. Finally, one of the deviations of the facts from Newtonian mechanics, the motion of Mercury’s perihelion, was just unwilling to play ball; people tried to postulate new planets, or a halo around the sun, but to no avail - until Einstein came and changed the theory itself. By now we may be used to the idea of approximating truth by modifications - by a series of approximations to it. At the time it was a great shock; the very idea that Newton’s theory may be superseded was shocking. A wonderful philosophical literature - the conventionalist-instrumentalist literature - arose; just before the Einsteinian revolution and as a result of the crisis, a literature of which perhaps I should speak a little bit. The last famous conventionalist was A. S. Eddington, whose wonderful style and clarity are more famous than his own contribution to conventionalist philosophy. So let me present his views first.Consider the case of abstract painting. One is often disturbed when one looks at an abstract painting and sees nothing recognizable there. That this disturbance is rather misstated or misplaced can easily be shown: a Mondrian abstract painting does not irritate, but is usually found uninteresting, whereas a Picasso is found quite irritating: “I cannot see three musicians there”; “this is not how a guitar looks,” etc. Indeed when one looks at Picasso one cannot fail to recognize that it is, all too often, a gross distortion of a picture of three musicians, of a guitar, etc.
So much for the most central and well-known facts about abstract painting. The word ‘abstract,’ one should note, has two very different meanings. Etymologically there is little difference between ‘extract’ and ‘abstract’; yet from antiquity it was in accord with common use that though anything may be extracted, only an essence may be distracted; thus, though any idea or paragraph may be extracted from a book, only its central idea or theme may be properly abstracted. Induction was the - proper or improper - method of abstraction, i.e. of extraction of essences from facts. In accord with this use the abstract is real, indeed even more real than the concrete from which it was abstracted (since the concrete is a mixture of essence and accident); yet in some context clearly the contrast abstract-concrete means unreal-real.
Even if one does not believe in induction from facts, as long as one accepts the idea of theories as delivering essences - or forms (for non-empiricists prefer ‘form’ to its near-synonym ‘essence’) - that is, as long as theories are regarded as realistic, as having genuine referents in the world, we may still speak of the contrast abstract-concrete but not to equate it with unreal-real. Yet the latter is the artistic contrast between abstract-representational. Traditionally, the abstract represents the form and the concrete the appearances. But once we disclaim all ontological status to our theories, axioms, or definitions, they become parallel to the contrast abstract- representational. When we contrast abstract painting with more traditional paintings, we stress its being non-representational.This, indeed, is why Mondrian does not annoy viewers as much as Picasso: Mondrian does not even seem to be representational, whereas Picasso often offers us distorted representations. Many art critics have tried to fight this bias by saying that truthful representation is best done by cameras, and that ever since the camera was invented the painter was not required to make likenesses. Therefore, the argument goes, when we look at a Picasso we shouldn’t try to find there a representation, a reproduction of an object; we should see other things there, such as the interplay of color and areas and the like; hence, when we look at Picasso’s ‘Three Musicians,’ for instance, we should forget that we see three musicians there.
Obviously, this is an exaggeration. It is, I think, impossible to forget that ‘Three Musicians’ is in effect a distorted representation of three musicians - even were the title removed; since we can see one blow a distorted wind-instrument and one play a distorted string-instrument, and a piece of music writing. When we see this picture we automatically interpret it - we read a representation into it. So the art critic’s suggestion not to read any interpretation into an abstract painting is not applicable.
To this the art critic replies by way of improving his position: never mind whether it is a representation or not, he says now, it doesn’t matter what interpretation you read into it. What matters, he still insists, is the interplay of color, area, etc. If there is an interpretation, well and good, if there isn’t an interpretation, well and good - the most important thing is how the picture affects you. If, however, the picture affects you adversely because you hold an interpretation of it which makes it affect you adversely, then blame your interpretation and not the artist’s work. In particular, when the representation you read into the picture is a falsehood or a distortion, either you should not mind it or reject it in an effort to enjoy the painting. Thus, instead of objecting to interpretation in order to stress abstractness, it is better to merely play down the value of interpretation. It is important to note that we may justly say that Picasso’s ‘Three Musicians’ is, indeed, a representation (whether we like it or not and in spite of the fact that the interpretation is largely in the eye of the beholder), and as a representation it is a very bad one; we hang it on the wall not because it is a representation, however, but because it is an excellent work of abstract art.Now here ‘abstract’ is again used as opposed to ‘representation,’ yet we apply both predicates to the same work! - the same work, that is, in two different interpretations: logic is not thereby violated.
Since the present discussion does not concern aesthetics, I am not going to examine this last theory; rather, I have introduced it here as a perfect analogy to Eddington’s theory of science. The conventionalist philosophers of science declared that those who brought about the crisis in physics were mistaken in considering scientific theory as representation of reality, a xerox-copy of laws from the pages of the book of Nature. Scientific theories, they said, are not laws of Nature, but abstract equations, implicit definitions, and other highly sophisticated pieces of intellectual machinery.
Science is a very rarified system which is not representative of the world; that is, it is not given to interpretation. Hence, all questions rooted in the quest for correct interpretation may be dismissed as naive. They are questions based on the expectation that science be representative, and the expectation is out of place - as the crisis in physics shows. The crisis in physics, in other words, is not a scientific crisis, but a philosophical crisis. It is the crisis caused by the attempt to see in science representation. It is true that reports of observations in science are representations of concrete empirical facts. It is quite true that reports of experiments in science are likewise representation. Let us say that reports of experiment and observation in science ought to be true, but anything more abstract in the sense of not-concrete and not directly observational is to remain abstract in the sense of non-representational, in the sense of not read as true-or-false in any sense. Now we all read interpretations into abstract science. We all have some representation of abstract scientific entities like atoms, electrons, curved space. When we speak of space curvature, we actually see a curved space like a curved railway track or a piece of metal curved by the strong man in the circus. Once we represent or concretize our scientific theories in such a manner, we cannot but view them as true-or-false in the sense that reports of observations are true-or-false. Being scientific, and presumably true-or- false, they are naturally considered as true rather than as false; too much is expected of them, and then the expectation is disappointed, and this is the crisis in physics. To avoid crises, say the conventionalists, you should avoid all interpretations - thus avoiding all representation and thus, further, avoiding truth-value. The purpose of science, conventionalists say, is to lump together experiences in a neat way - handy, serviceable, even aesthetically pleasii^. It is a neat package and there is no more to it - especially no additional meaning, no independent interpretation above and beyond its strictly observational content. Most conventionalists are therefore simply opposed to all interpretations. Eddington alone has a more liberal conventionalist view; he has no objection to interpretations on the understanding, or on the condition, that they are strictly personal. Indeed, he adds, you cannot avoid interpreting a scientific theory. If you enjoy your interpretation, very good; if not, you may ignore it; if you are troubled about it, this is your private trouble. The main point to stress is that interpretations are not important. So much for Eddington’s view.Now I have no strong objection to all this, even though I do not agree. My heart is not there, but so far not so bad. I now come to something which is really very bad, and it is this. You have a painter who paints the model. If the painting comes recognizably close to the image of the person who sat as a model, the painter considers it a portrait and may try to sell it either to that person or to the friends of that person. But close or not, he cannot declare his portrait to be non-representational: this is the game, heads, I win, tails, you lose - which Galileo already denounced. I think one has to make up one’s mind whether one does care about representation or not - whether one willfully distorts, or intends to represent, or is indifferent to representation.4 What matters, then, is that we declare our aim, rather than fit the aim to the achievement. We should not shoot the target by first shooting and then describing the circles of the target around the bullet wherever it happens to have hit. This would be considered dishonest by most people. I contend that most amateur philosophers of science - whether scientist or other - are doing just this.
Scientists do wish to see scientific theories as representational. To view scientific theory as a mere convenience, as presenting known experience neatly but offering no new and deeper understanding of experience, is a bit demeaning to the stature of science. The deprivation of scientific theory of any depth has already been done when conventionalism starts. Perhaps the most important conventionalist philosopher is Henri Poincare. He was attacked for demeaning science, and as a reply he wrote a very charming book called The Value of Science - the points of which in my opinion and in the opinion of many others are rather dubious. There is no escape from the fact that if scientific theories are mere convenient definitions - mere instruments to aid the memory and prediction and for aesthetic gratification - then the aspirations of scientists like Copernicus, Kepler, Galileo, Newton, and others, cannot relate to science as we know it. And therefore, most innovating scientists do like to consider their theories as representational and hopefully as true representations. It is when they get into a crisis that they declare their intentions to have been more modest, and usually only in retrospect. In other words, our theories may be true or false, on condition that they are true. If they are not true, we better say that they are not meant to be considered as either true or false, but abstract. Now to say that they are true or false on condition that they are true is, of course, unfair; and this is why I put it that way. I don’t think anybody before Popper noticed it, but it is there in the literature when scrutinized carefully. Take for instance, the article in the Encyclopedia Britannica on mechanics, which was written by a mechanical engineer and a professor in the University of London. He starts the article by saying that there are two ways of looking at Newton’s theory: either as the truth, as a theory verified by experience, or as merely a convenient instrument of calculation and prediction. I think this is unfair. There is another possibility that at least should be stated, and it is that Newton’s theory is a false theory. Most scientists and many philosophers of science get indignant when you suggest that Newton’s theory is false because the word ‘false’ arouses in them hostile feelings and they conclude that you suggest that Newton was not a respectable scientist. But you can explain - to a scientist even, though not always - that you have no such intentions. Once a person is willing to view Newton’s theory as one which may either be true or false most likely that person will agree that it is false. Nevertheless, the fact remains that scientists find it hard to consider the possibility that Newton’s theory is false. And this is a point which is noteworthy. What causes this difficulty? Why is it that calling Newton’s theory false angers scientists so?
Somehow, says Popper, it is traditional to equate falsehood with sinfulness for the simple reason that traditionally it is taken for granted that falsehood may be avoided. Consider any religion, Catholicism or Protestantism, and you can see why all falsehood, to some degree or other, is viewed by them as sin. If you want to avoid all sin all you have to do is repeat what the party organizer, perhaps the party manual or the Bible, or the priest, tells you. What you are thus told comes ex-cathedra, and therefore, cannot be erroneous. Therefore if you just tow the line you too cannot err. The ability to avoid error leads almost inevitably to the equation of sin with error. Now, in the development of modern philosophy and methodology of science there was an attempt to break away from the bonds of the Church. One of the tools by which the Church was attacked was this. Aristotle’s theory may be true and it may be false, but it is not the priests’ task to declare whether it is true or false; it is a matter for scientific investigation. This is particularly the point made by Galileo who said that the priests who say otherwise were saying so out of political interest and purely for political purposes. But one can go further. The Renaissance thinkers who had a lasting influence did go further: they said that all error, whether of Aristotle or of anyone else, can in principle be avoided. Descartes in particular says that God does not lie to us, because he is benevolent; hence, if we act properly, if we use the right method, we cannot err. If I do everything the best way and yet I err, then God is a deceiver, which is absurd. Hence there exists a way, the best way, which guarantees the avoidance of all error; hence when I err I do not use the right way even though it is available; hence when I err it is my refusal to use the right way which is the cause of it; which refusal is sheer perversity; error and sin are thus equated.
This is the doctrine that truth is manifest. It was advocated by Sir Francis Bacon, and adopted by Descartes. Bacon said truth is manifest through plain facts; Descartes said through plain reasoning. Though Descartes added to our certainty that truth is manifest, it was Bacon’s doctrine that became the official doctrine of the scientific community. And Bacon’s doctrine of error was at least as influential as his doctrine of truth. Indeed, his views on how truth is achieved were widely ridiculed in the nineteenth century; but his views of error were then considered platitudes. They still are widely accepted.
Bacon’s chief problem was, what caused the decline of antiquity and the descent of the Dark Ages? His answer was, Aristotelianism, which was a self-perpetuating error. All errors are self-perpetuating and so the first task of a true scientist is not the search for truth - this is his second task - but the avoidance of error; once one has erred one is disqualified as a seeker of truth.
Bacon’s theory explains how error comes about and perpetuates itself even though truth is manifest. It is the same as the Old Testament doctrine of bribery. The Old Testament says, thou shalt not take bribes, for bribes blind the eyes of the just. Whether one intends to pervert justice or not, once one takes bribes one is in no position to prevent perversion any more. When one offers a hypothesis, says Bacon, one does so in an attempt to avoid the labour of searching for the truth because one is impatient to achieve renown.
But one cannot achieve renown by admitting that one’s views are false, or even by admitting that one’s views may be false. Hence, in order for one’s hypothesis to deliver the goods - to earn one fame in the learned community - it has to be expounded authoritatively, ex cathedra and with spurious proofs; and all arguments against it have to be dismissed offhand and with hostility. And so, the motive which makes one advance any hypothesis is the motive which makes one defend it through thick and thin. And so hypotheses are bound to become prejudices and blinkers.
It is a fact that Newton advanced his own theory of gravity as an absolute truth. It is also a fact that Newton’s own claim^cannot be accepted today; we can say that his theory is a good approximation to Einstein’s theory of gravity, or we can say it is neither true nor false; we cannot agree that it is the absolute truth. Can we therefore call Newton prejudiced? I do not think anyone would view this proposal as reasonable. And yet, we feel, calling Newton’s theory of gravity false is calling him prejudiced. Here, then, we are plainly inconsistent in that we accept Bacon’s theory of error, but only in part.
It is doubtless the case that some errors are prejudices, and perpetuate themselves in the manner described by Bacon. But to say that all erorrs are prejudices is an error, and one which may become a prejudice.
To conclude, once we overcome Bacon’s doctrine of prejudice we may view science as a portrait of nature - true or false - or perhaps abstract. Once the three alternatives are put in this novel way, the problem of choice has radically changed. Whether Popper’s views are true or false, he radically altered the philosophy of science.
III.