Brand Blanshard "Current Strictures on Reason"

I

The present writer was installed in his Association office in the month of Pearl Harbor. More than three years later, as he sends this address to press, the war in Europe is just drawing to its close. His tenure of office has thus coincided almost exactly with the chief period of the war. It is hardly conceivable that an address written in this period, even when prepared for persons as detached and dispassionate as philosophers are supposed to be, should ignore the events of the time. Nor do I intend to ignore them. The question I propose to discuss is one that has been forced by these events upon all reflective minds. Over and over again in recent years we have had to contemplate acts that seemed to belong, not to civilization and the twentieth century, but to the days of a Neanderthal past, before law or justice or pity had begun to lift up its voice against the fist and the club. What has dismayed us is not merely the cruelty and the brutality, incredible as these have been; it is also the readiness with which great numbers of kindly and sensible people have embraced absurdities that were scarcely sane. As we have seen these things and meditated on them, we have asked ourselves whether there is sense any longer in calling man the rational animal. Have we been flattering ourselves unduly? Is that life of reason that figures so largely in the discourse of philosophers—the following of an argument where it leads, the government of feeling and action by the insights of reflective thought—a practical possibility, or is it an idle hope? In the light of the last few years, is not reason best conceived as a film stretched across the mouth of a volcano?

When I first sat down to write this paper, it looked as if this might be the answer that history would record. That was shortly before Stalingrad and El Alamein. The reasonable temper, the habit of appealing to reason, the way of life marked out by reason, was threatened by a rising tide that sought to drown out everywhere the loyalty to impartial justice and objective truth. I need not remind you of what men in high places said about the scientific spirit and the intellectual life; they have had their reward; and when one has been delivered from a nightmare, one dislikes to sully the morning with the memory of it. Suffice it to say that we did barely make our escape, and that there is probably no one of us who, after living through these years, does not place a higher value upon those fragile and precious goods of the rational life that were then so nearly lost.

The outward battle will soon be over. But is the political threat the only threat by which the rational life is endangered? I do not think so. The power to think and act reasonably has inward as well as outward conditions, and though these conditions are hidden away from the general view, philosophers know that they are more important than the outward ones. If we are to follow an argument where it leads, there must be present conditions, physical, psychological, and logical, whose absence will black out reasonableness far more effectually than any withholding of civil right. And the fact is that the existence of these conditions is now widely denied. It is an illusion, we are told, that reason can play free of the human nature in which it is enmeshed; there are strings attached to its every move which, if closely observed, will be seen to govern its course completely. Is this true? If it is, then our power to follow reason has won its great outward battle only to find itself bound inwardly hand and foot.

It is this challenge to our power of being reasonable that I wish to discuss in this paper. But first what is it exactly that is questioned? What do we mean when we call a man reasonable? We mean at least this, that in his thinking and acting he shows objectivity of mind. And what is that? It means being realistic, impartial, just; seeing things as they are rather than as fears or desires would tempt one to see them. The reasonable person will suit what he thinks and what he claims to the facts. He will be ready to give up an opinion if the facts do not warrant it, and stick to the opinion in the face of inner and outer pressure if the facts require it. His claims against others and their claims against him he will view impersonally and with detachment; he will not ask more for himself than is just merely because he is he; nor will he allow himself to be put upon merely for the like reason; he bases his self-respect upon respect for the sort of justice that is itself no respecter of persons.

If such reasonableness is to be possible, two further things must be true. In the first place, there must be a set of independent facts to be grasped. It would be senseless to try to suit our opinions to the facts of a case if there were no such facts to suit them to; and if justice consisted in following our own interest or our own desire, then, as Socrates and a hundred other philosophers have shown, there is no such thing as justice at all. To be reasonable either in thought or in act requires bowing to an authority beyond ourselves, conceding that there is a truth and a right that we cannot make or unmake, to which our caprices must defer. If I have a pet theory in science and am to be reasonable about it, I must be ready to trim it, recast it, or give it up, as an impersonal logic demands; nonconformity here is not heroism but suicide. As McTaggart said, no one ever tried to break logic but logic broke him. It is the same, of course, with morals. Reasonableness in conduct implies wearing a yoke and walking a line; it implies that if you and I differ about our rights, there is an answer to our question waiting there to be found, and that we are doing what we can to find it and conform to it. To say that there is nothing right or wrong but thinking makes it so is to say that there is nothing for thinking to discover; and to say that is to deny all point in trying to be reasonable. If all our beliefs are reasonable, then none of them are.

Thus the first condition of being reasonable is that there be an independent common rule. The second condition is that this common rule should at times control the course of our thought. We must sometimes be able to say: If I thought as I did, it was because my mind was under the influence of art independent pattern, the pattern of an objective truth. This is only to say that thought, if it is to be reasonable, must be like perception when it is accurate. Suppose we look at a checker board. If there is to be any such thing as accurate perception at all, there must be, in a sense “out there,” a certain number of squares related to each other in a certain way. That corresponds to our first condition. Secondly, we must be able to say, If I see them in this way, that must be because they are this way, because that independent order acts upon my mind and makes me see it so. If this arrangement presents itself, not because it is there, but because my mind is being pulled about by wires from within, then there is no reason to believe that we ever do or can see accurately; if we did, it would be sheer luck. I am happily not concerned with the mechanism of seeing, but with a principle. If, when we perceive things, we never perceive them so because they are so, then perception is a cheat. Similarly in thinking, unless at times we think as we do because the real relations of things are acting upon our thought, laying it under constraint, governing its movement, then knowledge must be an illusion from first to last.

Let us proceed with these two conditions in mind. To be reasonable implies at the least that there is an objective truth and right which we can at times apprehend, and that if our thought follows a certain course, it is because it is laid under constraint by the objective pattern of things. If these conditions are granted, reasonableness is so far possible. If either is denied, it is not possible. To show either that the pattern we seem to find in things is not there, or that, although it is there, thought can never surrender itself to the control of that pattern, is to put reasonableness beyond reach.

Now it is by denying these conditions that the case against our power to be reasonable proceeds. At the present day there are three positions, each held by thinkers of eminence, and each formidably strong, anyone of which would, if made out, severely limit the area of reasonableness, if not destroy it altogether. The first of these positions is that the movement of thought is explicable in terms of processes in the cortex. This view is widely held among those who describe themselves as naturalists. The second is that the movement of thought is controlled by nonrational processes within the thinker’s own mind. This is an ancient theory which has been revivified in recent years by the psychoanalysts. The third is that the very ideal of rationality, conceived as the following of an objective and necessary truth and right, is an illegitimate one. This is the view of the logical positivists. It is of course impossible to discuss these positions generally within the compass of one paper. But I think it will be found in each case that the limitations imposed on reason rest upon distinct and special grounds which can be isolated without difficulty. Let us look at the three positions in order.

III

The first or naturalist theory rests on facts that physical science has led us to accept as commonplaces. We are asked if we do not concede these to be facts; we admit readily that we do; and then, as we follow out the inferences from what we have conceded, it begins to appear that we have conceded also our rational birthright. How naturally we are led on from what seems to be the most innocent facts to a conclusion that is far from innocent will perhaps be clearer if we construct a little dialogue. The physiologist interrogates us:

“When you step on a tack and feel pain, you would agree, would you not, that stepping on it is the cause of the pain?”

“Yes, of course.”

“The immediate cause?”

“No, a remote cause only. The change in the nerve ends, so I’ve been taught, induces an impulse which is carried to the cortex and induces a further change there. It is this change in the cortex that is the immediate cause of the pain.”

“Correct. And you would take the same view, would you, about other sensations, and about affections and emotions?—that is, that their immediate cause or condition is a cortical change?”

“Yes, there seems to be no doubt about that. It is true, isn’t it, that one can produce sensation artificially by stimulating the cortex?”

“Yes, and we are even learning what precisely to do to produce different kinds of experience; we can put the brain through its paces. We can turn your world yellow by giving you santonin; we can increase or diminish your anger by adrenal injections; we can lift cretinism into normality by small doses of thyroxin; and if we reduce your body’s secretion of this by about a hundredth of a grain a day, you will slide down into imbecility. It is true we haven’t found out much about the cortical correlates of ideas, but I don’t suppose you would doubt that they are there too?”

“No, there seems to be no escaping that. If sensations and affections are brain-conditioned, so must ideas be. One could hardly chop a mental state in two and say that half of it—sensation and feeling—is brain-conditioned, and the other half, involving the use of ideas, is a sort of will-o’-the-wisp, with no roots in the brain at all. If some forms of consciousness are brain-conditioned, presumably all of them are.”

“Good, I’m glad you see that so clearly; we can’t make an exception for ideas. Now suppose that one idea is followed by another; each of course is brain-conditioned?”

“Yes.”

“And the thought sequence is conditioned by the sequence in the brain?”

“Well, since we have agreed that each thought is brain-conditioned, the explanation why one follows the other must lie, I suppose, in the explanation why one brain state follows the other.”

“Obviously. And the reason why one brain state follows another is to be found, I suppose, in a physical law?”

“Since both are physical, that must, of course, be true.”

“Then the reason why one thought follows another is also given in physical law?”

“Yes, that seems right enough.”

“Thought, then, is under the control of physical law?”

“Yes, that does clearly follow.”

“Well, we seem to agree perfectly. If you are a philosopher, you are at least an unusually sensible one.”

I wonder if others have, as I do, a sense of doom closing in as this dialogue unfolds. The concessions do not seem extraordinary; nine out of ten natural scientists would grant them without hesitation, and, unless in a mood of unwonted suspicion, probably most philosophers too. That is just what makes this first argument so effective. You seem to be doing nothing more than conceding obvious facts and drawing obvious inferences. And yet I believe one can show, also by obvious reasoning, that this account cannot be correct, and that if it were, it would mean nothing less than disaster for our rational life.

Let us look at the matter more closely. I said a moment ago that if we are to be reasonable, we must be able to follow the argument where it leads, which means that thought must at times be governed, not by secret strings within, but by the pattern of what it knows. When we say that our thought is objective, we mean just that, that it is moving under the control of the object. Of course there are processes often called thinking that are not so controlled; I may sit down to a geometry problem and think first of the weather, then of my dinner, and then of my headache; but that is not thinking. Thinking proper means reasoning; and reasoning means surrendering one’s attention to the logic of the case, moving to one’s conclusion because the evidence is seen to imply it. Success here, as the experienced know, demands a wise passiveness; the best thinking is the least free, in the sense that it is most completely laid under compulsion by the course of objective necessity. If my inference moves from step 1 to step 2, and from step 2 to step 3, that is because, when I am really thinking, the facts that 1 implies 2, and that 2 implies 3, make a difference to the course of my thought; the inference takes the line it does because it is following, and is influenced by, a line of necessity that is there before it. This is what it does, for example, when, starting from the postulates of a logical or geometric system, it spins out the theorems that follow; and the account holds equally whether the necessity linking the steps is conceived as synthetic or analytic. Indeed this is what always happens when our thinking is at its best; its course is then governed and guided by the requirements of the evidence. Our conclusions are not arrived at by leaps in the dark, then checked against the evidence, and found to hold by miracle; it is rather that, starting from the evidence, our thought moves to the conclusion it reaches because the evidence requires this, in both senses of the word; the objective entailment controls the movement of inference. If this never happens, then strictly speaking, we never reason. For if, when we pass from premise to conclusion, the premise’s entailing the conclusion has nothing to do with our reaching it, then our reaching it as often as we do, indeed our reaching it at all, becomes incredible luck.

It will now be a little clearer why to explain thinking by cortical change is not to explain it, but to explain it away. The subjective process of deduction is, when really deduction, governed by an objective implication, but when one distribution of particles follows another in the brain, what we have, so far as can be seen, is not implication, but cause and effect. The sequence of brain state B upon brain state A is as little governed by any visible implication as the sequence in motion of Hume’s billiard balls. I should not deny that between the brain states correlated with the steps of inference there is more than mere conjunction; but how far this is, as we know it, from anything like implication is shown by the facts, first, that if for one of these states there had been substituted anyone of a hundred others, we should have accepted the causal relation no less readily; and secondly, that between the sequence of states in the brain that serves as the correlate of a demonstrative process and that which serves as the correlate of the loosest association there is no detectable difference. Physical causality is one thing, logical necessitation another. If therefore you say that what controls the passage from A to B in inference is physical causality, you are saying that even in reasoning at its best and clearest, where we seem to see most plainly what we are doing, we are being grossly deluded. We suppose we think as we do because the evidence requires it; we now learn that this never happens. What really happens is that a sequence of distributions of material particles, or, if you prefer, of stresses and strains, or levels of energy, one connected with its successor by nothing nearer to logical necessity than the succession of waves on a beach, produces a series of mental efflorescences which turn out by some incredible chance to bear the relation, each to its follower, of ground to consequent. That this nexus among the objects of thought exercised the slightest constraint upon the course of our thinking must be set down as illusion. The fact that A is evidence for B had no influence at all in making us think of B, or in making us accept it. The purer reasoning seems to be, the deeper is the illusion, since, speaking strictly, we never reason at all.

Must we accept this view? I do not think so, and for two reasons. First, when our thinking is at its best and clearest, our certainty that it is controlled by necessity is greater than that of any physiological speculations that can be set on the other side. Take a simple train of reasoning and observe what goes on when you follow it. Two is to four as four is to what? Four is to eight as eight is to what? Eight is to sixteen as sixteen is to what? How do you manage to hit upon the answers as you move along this series? The natural reply is, Because the rule of the series logically requires that each successive proportion should be completed in just this manner. I believe that this, which is the natural account, is also the true account. There are dozens of directions in which thought could wander off at any step in the series, and I believe that if it declines these wanderings and remains in the groove, it is because there is a groove, because thought is laid under constraint by the logic of the process. We not only see when we reach the end that this constraint did operate; we may be aware of the constraint as we proceed. And to my mind there is something fantastic in brushing aside such empirical evidence for the sake of a flight of physiological speculation. Some persons, to be sure, are so much in the habit of prostrating themselves before physical science that they are ready to snub their clearest insights if such science has shown itself cool to them. Let us recall, therefore, that what we are offered here is conjecture, not established fact. No competent physiologist professes to know exactly what happens in the cortex when any conscious state occurs, nor exactly how any cortical event leads on to another, nor exactly what is meant by parallelism between the two series—still less to have verified in detail any hypothesis about their relation. To set a theory at once so vague and so tentative against the clear, immediate assurance of the reasoning mind is not properly science at all, but the sort of philosophy bred by an uncritical idolatry of science.

But there remains a more cogent reason for denying that physical causation will account for the sequence of thought. The view is self-refuting. How is it arrived at? It is an inference from observed sequences of mental and bodily change. Now the inference to this conclusion has either been constrained by the evidence or not. If it has, the conclusion is refuted by the mode of its own attainment; for something more than physical causality was at work in attaining it. On the other hand, if the inference is not under such constraint, why should we respect its result? For then nothing more is at work in it than in the equally good causal processes of wool-gathering or derangement. It may be replied that though rational and irrational processes are equally matters of physical causation, we can see by later reflection which are necessary and which are not. But this is again self-refuting. For even if I do, in a flash of later insight, see that the conclusion was required by the evidence, I do not have this insight because the necessity is objectively there, but solely because some change in my cortex has made it appear to be there. Given the physical change, I should have “seen” it whether it was there to see or not; and hence it is the physical change, not the presence of the necessity, that makes me think I see it. This is to make all apprehension of necessity illusory, and all attempts to prove anything vain, including this one.

It is curious that the disaster implicit in the physiological account of reasoning has been so seldom noticed. But there is one school of psychologists that has seen it and explicitly sought to deal with it, the school of Gestalt. They have said boldly that there are mental processes that cannot be explained in terms of traditional natural science; that it is futile, for example, to explain a course of reasoning in terms of habit, or conditioned reflexes, or even association and that if we complete a syllogism as we do, it is for the same reason that we complete an imperfect circle as we do, because the law of structure of what is before us makes its specific demand upon us. For this insistence, at a time when psychology is threatened with ruin by technicians without vision and without philosophy, we can only be grateful.

But their theory is now being developed in what seems to me a dubious direction. Having broken with a strong tradition of natural science by finding necessity in mental sequences, they make it up to such science by putting this necessity back into the physical realm. When we reason syllogistically, we are under the control of necessity, but this necessity is literally in the brain. They have argued with some cogency that when we perceive a square or a circle there is actually a field of similar structure in the cortex. They hold that when our thought is carried along the line of necessity there is a gradient of force in the cortex, a physical tension and its resolution, and that between the physical and the conscious necessity we can detect, if we look sharply, an identical “requiredness.”

My chief difficulties with this are two: First, try as I will, I cannot see that the necessity which moves us in reasoning is the same as physical compulsion, however abstract and schematic we make their allegedly common element. What the necessity is that links premise with conclusion I do seem to see; and I also seem to see that it is something different in kind from what the physicist means when he talks about a flow of energy from higher to lower potential. To say, then, that what moves me is really the latter is to say once more that when my thought is at its clearest I am under an illusion as to what is directing it. And I do not see how you can say that without discrediting reason generally.

Secondly, the Gestaltists would agree that between the conscious and the cortical state the parallelism is not concrete and detailed, but isomorphic merely, that is, identical only to the extent of a highly abstract and formal pattern. But is this the necessity that works in consciousness? The Gestaltists themselves have taught us that it is not. They would hold, for example, and I believe with sound and important insight, that there is a necessity in music which constrains a composer to continue a melody in one way rather than in others. This necessity is one that holds among the sounds as heard; it takes its character from the terms it relates, namely these phenomenal sounds in this concrete phenomenal field. But these sounds, as the Gestaltists agree, are not themselves cortical events. Any pattern, then, that is common to brain and consciousness would have to leave them out. But a pattern in which phenomenal sound plays no part is not the pattern that works within experience. Everything depends on which pattern is to control. To say that it is the first, the abstract isomorphic schema, is to say that what really governs the musician, the painter, the moralist, is not what he believes to govern him, but something extremely different; and this seems to me in effect to discredit our actual thought in the field of value. To say that what governs is the second pattern, the pattern that takes its character from the phenomenal sounds, is to concede control by what will never be found in the cortex.

It is time to turn to the second of the contemporary theories that imperil the life of reason, a theory that to most men is more familiar and more persuasive than the first. Even if our thinking is not in servitude to non-rational forces in the body, it is still, we are told, in servitude to such forces within the mind. Man is not primarily a thinker; he is an actor, for the reason that he is still an animal, with far more animal ancestors than human clamoring in his blood. His business, and that of his forebears, has been to fight for a foothold on the earth, first by instinct, then by cunning, then by intelligence; and of these, intelligence, the latest to arrive and not yet fully mastered, is as truly as the others a tool to ends that are selected for it and not by it. Man thinks to live; if he sometimes lives to think, that only shows that his mind, like his body, is subject to distortion. Thought sprang originally, and still springs, from practical need; it is maintained by a feeling –interest–and tested by another–satisfaction; its goal is not knowledge, for knowledge itself is only a means to survival and success. Little by little the beliefs that seemed to be the products of pure reason are being shown by subtle analysis to be the daydreams of frightened men who need to be comforted, or compensations for defects that cannot well be admitted, or rationalizations of the plainly irrational bribes paid to the forces of unreason for letting us hug self-respect a little longer. Man likes to boast that he is a rational animal. How better disprove the claim than by pointing to the pitiful fact that at times like this he still makes it?

There are people who believe all this to have begun with Freud. It would be less formidable if it had. The truth is that it is the undercurrent of all philosophic history, a strain in minor key that can always be heard if you listen attentively, even when the trumpets of reason are sounding most confidently. At the very moment when Plato was heralding a reason that was the impartial spectator of all time and all existence, Protagoras in the same city was declaring πάντον χρημάτον μετρον άνθροπος and Callicles was teaching that the doctrine of justice was convention only. While Plotinus was saying at Alexandria that reason was the highest emanation of Deity, Tertullian, farther along the coast, was saying: Certum est quia absurdum est, quia impossibile est. No sooner had St. Thomas completed the edifice of his rationalism than Duns Scotus was undermining it with the doctrine that even in God the will is primary and that it manufactures truth and right in accordance with inexplicable impulse. While one great Frenchman was building rationalism into the temper of France, another was protesting: La coeur a ses raisons que la raison ne connait pas. Spinoza wrote a great book to show that the good life lay in progress in reasonableness; and before it was published Mandeville appeared in England to preach that goodness is the offspring that flattery begets upon pride, and to hear an echo from Scotland proclaiming that reason is and must be the slave of the passions. When Hegel announced at Berlin a series of five-o’clock lectures on reason in man and the world, a young gentleman named Schopenhauer set another series at precisely the same hour to show that in both man and the world the primacy belonged, not to reason, but to blind will. While Bradley in Merton was thinking out the dialectic of the Appearance, Schiller just over the wall in Corpus was teaching that “our knowing is driven and guided at every step by our subjective interests and preferences, our desires, our needs, and our ends.” So it goes; so apparently it has always gone. And thus if Freud and McDougall and Westermarck have been teaching, each in his own way, that belief is the puppet of feeling it is not as if their doctrine were something new under the sun; it is only the newest form of one of the oldest protests against reason.

Before commenting on its claim to respect, I may be permitted a remark on its political relevance. No doubt the tidal wave that has been threatening in these last years to wash us and our studies into the discard is inspired by no one philosophy, if indeed it has been tinctured by philosophy at all. But there are those who, to the amazement of some of us, have sought to link this movement in spirit to those who have made most of reason. The thinkers of the great tradition have held that our thought, if it was to be reasonable, must bow to a logic the same for all of us, absolute in its requirements, and independent of desire; some of them have gone on and said that in such a logic we had the key to a world which, if we knew it fully, would be found intelligible through and through. This view is called at times absolutism. Perhaps for that reason some persons have professed to find in it the seeds of political absolutism. To set up logic as a final authority; what is that but authoritarianism? To bow to a truth that exacts recognition regardless of our desires–is not that surrendering liberty to a metaphysical Moloch? A member of this Association appealed some years ago for a view in which, to use his own words, “logic ceases to be a bully, and makes an appeal to our better instincts.” The argument seems to be that rationalism is a kind of authoritarianism, that Fascism is also authoritarianism, and that the two are therefore somehow the same thing.

On the virtuosity of this performance as an argument I shall not comment. What is important is that its conclusion is worse than untrue; it is the very opposite of the truth. The authoritarianism of reason is about as congenial to Nazism as an eleven-ton bomb; if it is brought home to it at all, its proves shattering. Once allegiance to a common reason is admitted, the whole sombre structure of Nazi notions, the notion of “thinking with one’s blood,” of an Aryan or a Semitic truth or duty or privilege, of right as made out by might, and all the jerry-built adjuncts of other irrational rights–the right to destroy individual freedom, to claim for oneself or one’s people what is not conceded to others, to enslave and gag and exterminate–all this is recognized as the hideous nightmare that it was. It is no wonder that Fascism in all its homes and forms found it expedient to stop the mouths of the philosophers. For Fascism stands for self-assertion, and reason for self-criticism. Fascism loves force, and to reason the appeal to force must always remain irrelevant and stupid. Fascism insists on what divides men; reason is cosmopolitan. Fascism hates the intellect, and with a sound instinct fears it; for in reason it recognizes, and knows that the world recognizes, the deadliest of its enemies, an authority in which there is no authoritarianism, an absolutism that does not tyrannize, a master, and indeed the only master, in whose service there is freedom.

But to return to the argument: thought, we are told, is under constraint from within. It reflects, not the outward pattern of things, but our hidden loves and hates, desires and fears. In The Future of an Illusion Freud explained religious belief as due to the persistence of the infantile need for a father. According to Westermarck, what is expressed by our moral judgments is no character in the act, but our emotional attractions and repulsions. In a recent book a distinguished psychologist, Professor Holt, has written: “The entire history of philosophy is little else than a tiresome and futile series of pictures in which each philosopher has imagined what he most yearned to have in his own ‘best of all possible worlds’. This,” he adds, “is levity.” Such skepticism about reason, though anything but new, has perhaps never been more popular and more formidably supported than in recent years. What are we to say of it?

The first thing that we must say of it is a commonplace. It is that if the argument is pushed through and made general, nothing further is called for; like so many other attacks on reason, it disposes of itself. If it is true that we are always governed by non-rational pulls; then of course our conclusion that we are so governed is also produced by non-rational pulls. But if it is, why should it have more respect than any of the other illusions produced by such pulls? Surely the attempt to prove by rational processes that rational processes are irrational is the last irrationality.

Perhaps the reply will be made: “I admit the inference; and hence I offer my theory only as one that expresses and satisfies my own feeling, and may turn out to have the advantage of rival theories in better expressing the feeling of others also.” But the reply will not do. First, to say, “I admit the inference” is to say, “I accept it because I see that it follows,” and to say that is already to have abandoned the view that beliefs need be governed irrationally, since this one is not. Secondly, the theory is plainly not offered merely as something that pleases its maker; it is offered as true, as conforming to fact, and because it does so conform, as sounder than rival theories. If it is not so offered, why offer it? If it is, then the offer is inconsistent with the theory offered, for it offers as governed by fact the theory that, owing to subjective pulls, our theories are never governed by fact. And thirdly, when anyone says he is content to have his theory take its chances with other theories, it is hard to believe that he is really proposing to test it by its appeal to popular feeling. He is saying that as people come to know the facts better, they will see that these facts exclude the other theories and require his own. That implies that the minds to whom he takes his appeal are not puppets of feeling, but are to this extent reflectors of fact.

The truth is that in this generalized form the theory does not make sense. It says that our thought is inevitably distorted by feeling, and it is ready to say pretty precisely, as Freud does in discussing religion, where thought goes off the rails. Now you cannot recognize that another has gone off the rails unless you know what it means to stay on them. If Freud can point to the mote in other people’s religious vision, it is because he is confident he has cast out the beam from his own. He is sure that in the main he is thinking straight when he thinks about religion and about the crookedness of most people’s thought about it. What he has proved, then, is not that thinking straight is impossible–a proof that could not get under way without assuming the falsity of its conclusion–but only that thinking straight is hard, which we knew before. To say that we can never think straight is to expose oneself to that charge of fatuity which has now stood for some thousands of years against the sort of person who rises to remark that he knows he knows nothing.

I am of course not offering these few comments as an appraisal of the work that has been done by the students of man’s irrationality. We owe them a great debt. McDougall has said that Freud threw more new light on the workings of the mind than any other psychologist since Aristotle, and I should not care to deny that he is right. All I am concerned to deny is the conclusion often drawn from these researches, that the mind is so controlled by pulls from within that it is never under the control of the objective pattern of things, or follows the thread of an impersonal logic. The remarks I have offered, slender as they admittedly are, do seem to me to settle that point in principle.

We now come to the third of the current criticisms of reason. It is a peculiarly formidable criticism, because it comes not only from within the camp of the philosophers, but from a part of that camp in which clearness and accuracy are cultivated with laudable care. The attack is formidable, again, because it calls in question the very end and goal of reason as we have described it. That end is to understand, and to understand is always to follow an objective pattern or order. What kind of order is this? If it is to satisfy reason, it must be an intelligible order, and what is that? It is an order that never meets our question Why? with a final rebuff, one in which there is always an answer to be found, whether in fact we find it or not. And what sort of answer would satisfy that question?

Only an answer in terms of necessity, and ultimately of logical necessity, since of any answer that falls short of this the question Why? can be raised again. When we reach an answer that is necessary, we see that to repeat the question is idle. Of any statement of merely causal necessity, such as the law of gravitation, or Ohm’s law, or Boyle’s law, we can intelligibly ask why things should behave in this manner. But when we see that things equal to the same thing are equal to each other, we cannot sensibly ask why, because we are at the end of the line to which such questioning can take us. We have already reached the logically necessary.

Now if the world is to be the sort of world in which reason could even in theory reach its end, it must be one in which intelligence finds an answering intelligibility. I see no way in which it can assure itself beforehand that this is what it will find; I only wish I did. It may be that when we ask such questions as Why does the sun attract the earth in accordance with the law of inverse squares?, we are asking a question to which no answer that satisfies reason will ever be forthcoming, and this not because the answer is beyond our reach, but because there is no answer, because the connections of things and events are non-necessary, and therefore in one sense non-rational and unintelligible. If this is true, the attempt to understand is doomed to defeat from the outset. But I see no way of proving this either.

Here is where logical positivism comes in. It claims to have evidence that in entering upon such a program reason is bound to fail. The argument is as follows. Thought must live and move among propositions, for it is intent upon grasping what is true, and only propositions are capable of truth. Since the material with which it directly deals is thus always propositions, a review of the kinds of proposition open to it will throw light on what we may expect of it.

Now when we review the possible kinds of proposition, we find that they are all reducible to two. On the one hand are necessary propositions, such as those of logic and mathematics. Because of their necessity, they have always given delight to the rationalistically inclined. But unfortunately they are all tautologies; they unfold our own meanings only and give no knowledge of the actual world. On the other hand there are empirical propositions: this is a table; American robins have red breasts. These do assert of the actual world and, if they are true, tell us something about it. But then they are never necessary; they never report that S must be P but only that SP is the case. And if the positivists are right that these two are the only kinds of proposition that ever present themselves to thought, then the program of reason as we have conceived it is clearly impracticable. That program was to penetrate through into the intelligible structure of things. This we now see that we can never do. For though we can indeed know necessities, these necessities are never links that join actual facts; and though we can know facts, these are never necessary. The world of existence is unintelligible.

The positivist case against our program thus rests on two contentions; that all necessary propositions are tautologous, and that all factual propositions are contingent. It is important to see more precisely what these mean.

It may be supposed that the first contention, all necessary propositions are tautologous, means what Kant meant when he said that analytic propositions were tautologous. These, he said, merely set out in the predicate what is already contained in the subject. Positivists reject this account of tautology as resting on psychological grounds; it places the test, they say, in subjective intension, in the accident of how one happens to conceive of the subject named. The test they offer instead is whether the proposition in question can be denied without self-contradiction; it is necessary if it cannot. Now they admit that there are large numbers of propositions which are in this sense necessary; and if so, why should we take offense or alarm at their theory? Do not all these necessities stand for just so many intelligibilities in the nature of things, and are not these precisely what we are seeking?

Unhappily, the positivists will not let us read them in this way. They insist that the necessity here exhibited has nothing to do with the nature of things, that the contradiction involved in its denial means incoherence, not in nature, but in our own linguistic usage. Necessary propositions, writes Mr. Ayer “simply record our determination to use words in a certain fashion. We cannot deny them without infringing the conventions which are presupposed by our very denial, and so falling into self-contradiction. And this is the sole ground of their necessity.”¹ A necessary proposition of the form ”S is P” tells how we propose to use S. A necessary proposition of the form “P implies Q” illustrates a definition of implication which has been adopted arbitrarily, and which stands, not for a nexus in nature, but for a convention of our own. Let us look at these two types.

A necessary proposition of the form S is P, which in former days would have been said to state a necessary relation between concepts, is now said to state how we use, or propose to use, S. I think that what this amounts to, after all, is that such propositions are analytic in Kant’s sense; the predicate sets forth, in part or in whole, how one conceives of the subject; the addition to the older theory is that this predicate is arbitrary. Regarding this doctrine I should hold as follows: (1) the view that all propositions of this form are analytic is untrue, and (2) the addendum that the predicate is arbitrary is equally untrue.

(1) “Whatever is red is extended.” This seems to me a necessary proposition, and most positivists would, I think, agree. By saying this they mean that its contradictory would be self-contradictory. Why would this be true? Because in our first proposition we merely set forth in our predicate part of what was meant by our subject. This analysis seems to me incorrect. What I mean by extension is not what I mean by redness, nor is it part of this; the two are quite distinct. If when I think of a billiard ball as red, the extension of that red is part of what I mean by red, then when I think of another billiard ball as white, the extension of the white will be part of what I mean by calling it white; and I shall then have to say that the balls are similarly colored, which is absurd. Being extended is, to be sure, so intimately connected with being red that if a thing is red it must be extended also; the one entails the other. But surely that is the way to put it. It is quite incorrect to say that when I call a thing extended I am defining the meaning of red. Though I am asserting a relation of entailment or necessity, it is evident from inspection that that relation is not one of identity, either in whole or in part. And if so, necessities are not always tautologies. I should myself maintain that in actual thought they never are, but that is another point.

(2) To the contention that such propositions are analytic, the positivists add, as we have seen, that they are arbitrary, in the sense that they state or illustrate a convention which might have been different. Mr. Ayer writes as follows: “if I say, ‘Nothing can be coloured in different ways at the same time with respect to the same part of itself,’ I am not saying anything about the properties of any actual thing. . . I am expressing an analytic proposition, which records our determination to call a colour expanse which differs in quality from a neighbouring colour expanse a different part of a given thing. In other words, I am simply calling attention to the implications of a certain linguistic usage.”² Now I suggest that when we call two differently colored patches of a rug different it is because we see that they are and must be different, and that this, which we mean to assert, is wholly independent of linguistic usage. If it were really a matter of usage, the adoption of a different usage would make a difference to what I assert. Would it in fact? Suppose we decided that when we saw two differently colored patches we should henceforth call them the same patch; would that which we meant to assert be different from what we meant to assert before? I think not. We should still be asserting the parts to be different, because we see that they must be, and if we used the word “same,” it would now mean what we meant by “different.” The fact is—to repeat—that we call two differently colored parts different because we see that they are so, and must be; they are not so, nor are they seen to be so, because we have adopted the convention of calling them so. Language adjusts itself to the observed nature of things; the nature of things does not wait on our language. These are truisms that I am almost ashamed to set down deliberately. And yet when we are offered statements of the kind I have quoted as the final result of exact linguistic researches, a few truisms may come as a relief.

I have been dealing with necessary propositions of the S-P form, that is, propositions which assert a connection between subject and predicate. I come now to assertions of the P-implies-Q type, which assert a necessary linkage between propositions themselves. The positivists treat these in essentially the same way as the others. They would argue as follows: when we assert that a proposition P implies another, Q, we are, in the first place, asserting what we have asserted already, and in the second place, asserting a relation to hold that belongs, not to the nature of things, but to our own set of conventions. As for the first point, when we say that P implies Q, we find that we always know, or think we know, certain things about the truth of P and Q. Of the four possibilities—both true, both false, P false and Q true, P true and Q false—we know that one or other of the first three holds. But in knowing that, we know already that P implies Q, for that is what the statement means. At least that is what it means to us. For, secondly, say the positivists, you are at perfect liberty to mean by it something else if you wish. You may mean by it what, following the Principia, we have just offered, i.e., either P is false or Q is true, or what Mr. Lewis means by it, that P’s truth is inconsistent with Q’s falsity, or anyone of a large number of other things. Which of these you choose is not determined for you but by you; it is a matter of convention. All that is required is that once you choose your conventions, you stick to them, that once you have defined implication in a given way, you mean this by it consistently; otherwise you stultify yourself.

Now the first of these points, that implication is tautologous, depends on the second, that it is a matter of convention; for, in the position we are examining, what implication shall be is conventionally determined. The question before us, then, is whether it is so determined.

It seems to me that there is one very simple argument which shows that it is not. This argument is that of all the various definitions which are offered of implication, we can sensibly ask, Does this give what I really mean or not? We can not only sensibly ask that question; we can see that the various answers miss or approach what we mean in various degrees. Thus we can see that the Russell-Whitehead formula of material implication misses what we mean by a very wide margin, and that Lewis’s strict implication approximates it much more closely. This shows that we have something in mind to which all the conventions must come for testing, a relation conceived as holding independently of our usages and conventions. When we say that the premises of a syllogism imply its conclusion, or that being extended implies being divisible, we do mean something definite, however difficult to hit with words; and this is what gives the target at which our definitions aim. If there were no target there at all, how could we tell, as in fact we can, that some definitions strike close to the mark and others go wide of it? Of course our definitions are arbitrary in the sense that to the word “implication” we can attach any sense we want. But to argue from this that any sense we attach to the word will equally fit what in common use we mean by it is surely confusion. When we dispute over the nature of “justice” or “number” or “truth,” are we really free to define the term as we please? Do we not assume on both sides that we are trying to run down and capture the same thing? When we argue with each other as to whether an inference is to be admitted, is there no bar, in the form of a common understanding of what “follows” really means, to which both of us must take our appeal? If there is not, argument is futile. If there is, positivism is wrong.

This consideration is to my mind decisive, and those who hold logic to be conventional have not, I think, wholly escaped it. It is true that from differing definitions of “P implies Q” there follow “alternative logics,” in the sense of differing sets of basic logical propositions. If, for example, one defines this, not as meaning “material implication” (either “P and Q,” or “not-P and Q,” or “not-P and not-Q”) but as meaning “either ‘P and not-Q’ or ‘not-P and Q’ or ‘not-P and not-Q’” a sort of logic would follow in which a true proposition implies and is implied only by a false one. But so far as I can see, when one says that this follows, one means by “follows” what all the rest of us mean by it. The concept of following is common to all the alternative logics; to that there is apparently no alternative. Again, in a two or three valued logic, implication is commonly defined by the matrix method; for example, if P and Q may have either of the values “true” and “false,” and no others, then there are only four combinations possible from which to compound the definition. Now when it is said that these are the only four possible, is this too a convention to which an alternative is possible? I cannot think this is meant. Once more, if logic is wholly conventional, there should be logics in which the principle of contradiction is replaced by an alternative. So far as I know, there are none such; without this principle the sort of distinction required by all logics in common would be impossible. But a convention that is necessary to make all other conventions possible is not in the same sense a convention itself.

I have been dealing so far with the first position of the positivists, which would make all necessary assertions mere statements about usage. It may be asked, If not about this what else? You would not hold, would you, that they are statements about the actual world? I answer, Of course I would. “That apple yonder cannot, in the same part and under the same conditions, be colored in different ways.” I believe that when we say that, we are saying something about the apple. “X cannot at once have Y and not have it.” The positivists take this as meaning, “I do not propose to call both that which has Y and that which hasn’t by the name of X.” Bradley takes it as meaning that nothing that is real is self-contradictory. Which is right? Of course if one says, as positivists do, that all assertions except those about usage are assertions about sense experiences, Bradley is talking nonsense. There is no space here to discuss this curious and interesting revival of sensationalism. All I can say is that after an inspection of my own meaning, I wish to make it clear that I am talking Bradley’s kind of nonsense.

We turn now to the second position of the positivists, which must be dealt with in the briefest way: All factual propositions are contingent. What are we to say of it? I think that even if factual propositions are defined in the straitest positivist fashion, the statement must be set down as untrue. Before us, for example, is a series of colors arranged in order of their affinities. We perceive that in this series, orange falls, and must fall, between red and yellow. Is this an assertion about elements given in sense? Yes, and it is therefore a factual assertion. Is it a contingent assertion? No. Things are related contingently when they might be related otherwise than they are. But the relation I am here asserting could not be other than it is; if orange were not related as it is to red and yellow, it would not be orange. The Gestaltists tell us that when we “see,” as we often do, that to continue a melody in the right key we must proceed thus and not thus, we are laying hold of a genuine requiredness; and I think they are right. Here again the must holds among the given sensory elements; the insight is at once factual and necessary. And if one breaks with the narrowly sensory interpretation of “factual,” as one should, many other types of factual necessity are admitted. When I say that my present toothache is bad, am I saying that the badness is accidentally conjoined to it, so that the pain could be what it is without the badness? Clearly not; I am asserting a predicate that belongs to its subject necessarily, though that subject is an existent. When I say “I cannot doubt that I am now conscious,” I am reporting that a present fact excludes, and necessarily excludes, a predicate suggested of it. Personally I should be ready to maintain, in respect to each of the positivist positions, not only that it is false, but that the truth lies in its contrary. I think that in the end all necessary propositions must be taken to assert of existence and that no factual propositions are altogether contingent.

But it is no part of my design to argue for these positions. My present interest has been sharply limited; it has been merely to help clear the ground. But that in itself is important. For if anyone of the theories I have discussed is true, philosophy has no future, except perhaps “the future of an illusion.” If our reasoning is in truth the shadow cast by the irrational displacements of matter, if it is only the bobbing of corks on the surface, pulled about from irrational depths, if it is really a play with syntax, signifying nothing, then we should face the truth and, as Cromwell said to the cleric, we should “cease our fooling.” On the other hand, if these things are not true, as I have sought to maintain, then we should clear them out of our way and get on with our work. For that work, as the greatest philosophers have conceived and practised it, namely as an attempt to understand the world, is far too significant and exhilarating an adventure to miss. That we shall ever carry it through to the end, that we shall actually succeed in following the track of necessity across the wastes that now seem trackless, I find it hard to believe, nor do I think they believed it. But they were animated by a faith which made the adventure momentous; they believed that the track was there and that they were free to follow where it led. I think that faith should be ours.

¹ Language, Truth and Logic, 114.

² Op. cit. 104.

Posted April 10, 2007

Back to Blanshard page