| C. D. Broad, Perception, Physics, and Reality, 1914 |
THE present essay has as its object an attempt to discover how much natural science can actually tell us about the nature of reality, and what kind of assumptions it has to make before we can be sure that it tells us anything. By natural science, for the present purpose, I mean physics. INTRODUCTION
When a certain way of looking at the universe meets with the extraordinary success with which that of physics has met it becomes the duty of the philosopher to investigate it with care; for it is likely to offer a very much better Cosmology than his own unaided efforts can do. And, if philosophy is to take into account empirical facts -- and it is extremely difficult to see what it will be able to tell us about the existent unless it does -- it can hardly neglect the most fruitful and thorough investigation of certain large branches of empirical facts that has yet been made.
But natural science starts with certain assumptions, and, as it goes on, it developes certain general conclusions about the real world. For instance, it starts with a position not far removed from naif realism, and, in its progress, it draws a distinction between the reality of primary and secondary qualities, and developes a causal theory of perception. Now the distinction between and primary and secondary qualities as to their reality is primary a metaphysical question, and science seems historically to have taken over its answer to it from Descartes. This has led quite reasonably to an attack on science from later philosophers who have not agreed that it was possible to stop at the point at which Descartes and natural science stopped in this matter. If we look into the arguments of philosophers on this important question we shall, however, find that they have little right to cast stones, as their reasonings are confused and full of implicit assumptions which, when rendered explicit, tend to cease to be probable. It therefore becomes very important to reopen this discussion, and to consider what really is the truth about primaries and secondaries, and how far the particular traditional belief which science has inherited on this question is relevant to its real progress and is consistent with its other beliefs. I have therefore begun by discussing all the common arguments against realism which do not depend on the belief that our perceptions are caused by real things identical with or rather like their objects. I come to the conclusion that none of these arguments which are so confidently repeated by philosophers really give conclusive reasons for dropping even the crudest kind of realism. I of course assume the distinction between a perception and its object which Mr Moore, in his Refutation of Idealism, showed to have been so largely ignored. Whilst the recognition of this distinction is all-important, and whilst it is perfectly true that many of the most plausible arguments used by philosophers against naif realism depend on ignoring it, it is not true to say that all arguments for Idealism rest on this confusion. Its recognition is perfectly compatible with the belief that the objects of none of our perceptions continue to exist when we cease to be aware of them. And the arguments that I discuss in my first chapter are attempts to prove this proposition, at any rate for some objects. Whilst I do not think, as will be seen, that they do absolutely prove anything of the kind, it is not true that they rest on the confusion that Mr Moore pointed out.
In the second chapter I discuss Causality, with particular reference to the objections that have been brought against causal laws on various grounds by philosophers like Lotze and Mr Bradley. It is clear that, unless causal laws have some kind of truth, science in general cannot tell us about the nature of the real; and, in particular, we cannot investigate arguments against realism that rest on the belief that our perceptions are the products of causal laws that include our minds, our bodily organs, and events in a real world. I have been much assisted here by the views on Causality in Mr Russell's Principles of Mathematics, where they were relevant. At the same time I think that they suffer by having been stated with particular reference to the science of mechanics. Mechanics is so abstract and has been so successful that by keeping it too much in mind we are liable to forget the difficulties that beset causation in other regions where we certainly suppose that it holds. I have been led in the end to state causal laws in terms of probability.
I must say a word at this point as to the use of probability in this essay. I have constantly put my conclusions in terms of probability and not of certainty. This will perhaps seem peculiar in a work which claims to be philosophical. It seems to me that one of the most unfortunate of Kant's obiter dicta is that philosophy only deals with certainty, and not with probability. So far is this from being the case that to many philosophical questions about the nature of reality no answer except one in terms of probability can be offered; whilst to some there seems no prospect of an answer even in these terms. Few things are more pathetic than the assumption which practically every philosopher1 makes that his answer to such questions is the unique possible answer; and few things are funnier than the sight of a philosopher with a theory about the real and the nature of perception founded on numberless implicit assumptions which, when made explicit, carry no conviction whatever, telling the scientist de haut en bas that his atoms and ether are mere economical hypotheses. Into the meaning and nature of probability I have entered as little as I could. This is not because I underrate its importance, but because I do not feel competent to attempt a task which has happily fallen into better hands than mine. But, as there is little question about the actual laws of probability, this omission is less important than it would otherwise have been.
In the third chapter I discuss a form of the doctrine known as phenomenalism. This theory is commoner among physicists than among philosophers, which is rather ominous for a philosophic theory. The reason for the wide belief that it enjoys in this quarter is that the physicist who is interested in the philosophy of his science naturally turns to the works of Mach and his school. I should be the last to deny the many excellencies of those works ; but they do not form an adequate philosophic equipment for the decision of the questions which they so confidently solve in the phenomenalistic sense. I have therefore devoted a very short chapter to what seem to me to be the essential points in the discussion of phenomenalism. This is to be taken mainly as prolegomena to the fourth chapter on the causal theory of perception. Here I have discussed as fully as I could this theory, which nearly everyone assumes, and which has been so little discussed in its purity. It is curious to note how often philosophers seem to have held that, when they have offered an explanation of the causes of our perceptions of certain objects or of the origins of certain beliefs, they have proved that those objects cannot be real and that the beliefs must be false. This has been particularly the case in the matter of perceptual and conceptual space. An honourable exception to this general failing is Lotze, who very justly points out that presumably all beliefs and perceptions have causes, whether they be valid beliefs and perceptions of real objects or not. I have tried to clear up the relations between the causes of our perceptions and the reality of their objects; and have been forced to conclude that, in all probability, the objects of our perceptions do not exist when they are not perceived, although there is no absolutely conclusive proof of this. The question then arises as to whether this ought to reduce us to complete agnosticism about the nature of the real causes of our perceptions. I have tried to show that the scientific account of the causation of our perceptions can be so stated that its success will enable us to gain probable judgments about the nature of those real causes, in spite of the fact that it seems to be stated in terms of what have been seen to be, in all probability, mere appearances.
I have concluded with a discussion of Newtonian mechanics. This may seen, to be out of order in a work which is primarily philosophical, especially at a time when there is some reason to think that the classical mechanics is only a particular case of electromagnetic laws. But the philosophical problems raised are essentially the same as those raised by the laws of the motion of of electricity. These problems are that of absolute or relative motion, and that of the reality of force. And it is clearly better for philosophic purposes to discuss these questions with reference to the ordinary laws of motion which everybody knows and which are clearly applicable to a very large class of motions, if not to all, within the limits of experimental error. I have been most helped here by the relevant chapters in the Principles of Mathematics. It will be seen, however, that I do not think that Mr Russell proves his point about absolute space.
The greater part of the present work was submitted as a dissertation at the examination for Fellowships at Trinity College, Cambridge, in 1911. I am far from thinking it conclusive, and I still hanker after a more realistic view than it reaches. But, if it succeeds in the showing old arguments and old difficulties in a clearer light, and in bringing forward latent assumptions, it will not have failed in its object, and may serve as a stepping-stone to further investigations.
I have to acknowledge but little direct indebtedness to books, except in the last chapter. But I owe much to the lectures and conversation of Mr Bertrand Russell, and should have found his little book on Philosophy useful if it had appeaxed when I was writing. I also desire to express my obligations to the lectures of Mr W. E. Johnson of King's, whose rooted antipathy to publishing anything till he is sure of everything is a great misfortune to philosophy. I must also thank Prof. Stout for kindly reading the proofs, and for making many suggestions and criticisms. Sometimes I have modified my argument, or made additions, to meet points raised by him; and when I have not done so, it does not mean that I have failed to see the force of his objections.
C. D. B.
CAMBRIDGE,
September, 1913.
Notes 1 Cournot is an honourable exception.
Contents -- Chapter 1