§ 1

Let us agree, then, to apply the instrument of “methodological doubt” to what Mr. Russell calls the “mass of common knowledge”; and let us agree to apply it without fear and without favour. But if this is to be our procedure, it will certainly prove convenient to boil down this general “mass” for the purposes of our experiment into a sort of informal creed, compounded out of the beliefs or classes of belief ordinarily entertained by ordinary people in their ordinary moments. There is no question here of attempting anything in the nature of an all-inclusive survey, however bare and abstract, of what men suppose themselves to know. This may, or may not, be worth attempting; in any case, it is not what I attempt. It will be enough for my purpose if the beliefs, or kinds of belief, I informally enumerate, are in themselves of decisive importance, are generally held, and are easily described. Convenience suggests that we should restrict our list to beliefs which, in some shape or other, are generally entertained by science and worldly good sense. Religion, metaphysics, morality, and æsthetics may, for the moment, be left on one side. It is true that these subjects have always lent themselves with peculiar facility to experiments in doubting—but, perhaps for that very reason, the doubts they suggest are much less likely to be instructive. Areas assigned by custom to military man oeuvres are rarely the theatre of decisive battles.

I propose, then, as a preliminary stage in the application of methodological doubt, to sketch a creed which shall embody certain of our familiar beliefs. But before making the attempt let me apologise for my use of the word. It is in many respects very inappropriate. A creed—be it religious, political, artistic, ethical, what you will—is usually intended to be more than a list of true propositions; it has about it some element of challenge; it is designed to separate sharply those who accept it from those who do not; it proposes to fix or stabilise beliefs on some particular subject; it is at once a device for producing conformity and for putting conformity to the test. My unpretentious list of familiar beliefs aims at none of these things. It is in intention quite uncontroversial; and could I have described it by a better word than “creed” I should certainly have done so. No better word has, however, occurred to me. Mr. Russell's “mass of common knowledge” expresses the idea, and I employ it or its equivalent when I can. But he did not devise the phrase for technical purposes; and for these it is too heavy and too long. So, bearing this explanation and apology in mind, the reader will, I hope, permit me to use the word “creed” in the loose and informal manner I have endeavoured to indicate.

§ 2

My description, then, of “the mass of common knowledge,” my creed of familiar beliefs, runs somewhat as follows:

Every man regards himself as a living being among other living beings. Especially does he think of himself as a man among men, one of a multitude of similar beings who know, feel, desire, will, and act; beings between whom intercommunication of some kind not only occurs but seems necessary to normal existence.

Every man thinks of himself and his fellows as dwellers in a world of material things, a world in space and time, a world extended and enduring. Every man believes himself to possess, under normal conditions, the power of intentionally moving some of these material things, and, among other material things, his own body. He attributes the same power to other men and to many animals. He does not, however, suppose that its exercise—all-important though it be to those concerned—seriously modifies the vast material bulk of the surrounding world. This runs its independent course, scarcely touched by living activities, and quite unmoved by the thoughts, perceptions, and desires of the human race.

Every man attributes to this natural world a large measure of regularity, and most of the persons who are likely to read these words conceive that, apart from life and will (about which there are disputes), its regularity is unbroken.1 With a complete knowledge, therefore, of the present constitution of Nature and adequate powers of calculation, we could perfectly anticipate the future of the physical universe to the end of time—unless, indeed, its course were modified by influences which are not physical. Though still infinitely remote from such a consummation, most educated moderns cherish the hope that we are moving slowly towards it; and, in the meanwhile, beliefs in the laws, or at least the habits, of Nature constitute a most important part of our common-sense equipment.

The beliefs so far enumerated are of wide scope and great generality. But to complete our sketch four other classes of belief require to be mentioned.

1. Certain abstract principles—e.g. those called logical and mathematical—which we regard as true not merely of the world in which we live, but of any world in which we find it possible to believe.

2. A large stock of minor generalisations.

3. Innumerable beliefs about particular matters of fact—some dealing with objects and events of enduring interest, others apparently referring only to the transitory and the trivial.

4. A belief that some access to truths—particular and general—is provided through such channels as perception, memory, reasoning, intuition, and human testimony.

It is important to observe that whenever the creed deals with the classes of belief numbered (2) and (3), the unanimity claimed for it is of a special type. All the world is agreed to accept some members of these classes, but there is no agreement as to what those members should be. Everybody, for example, owns a private collection of beliefs about particular matters of fact, perceived, remembered, inferred, or taken on trust. But no two people own precisely the same collection, or have gathered it together precisely in the same way. Their respective creeds, therefore, could we imagine them set out in full, would have many common characteristics, but they would never be identical; nor would they be due to the same memories, the same perceptions, the same reasoning, or the same hearsay.

If any critic cares to observe that this list of platitudes is wholly deficient in inner logic or external completeness, I shall entirely agree. It aims, after all, at no more ambitious object than to summarise, in a very general fashion, some important portions of our “mass of common knowledge,” and this I think it does. So far as I at least am concerned, it indicates in language sufficiently intelligible the beliefs and kinds of belief which I ordinarily entertain about the world of men and things in which I suppose myself to live. If it does the same for any of my readers, with them I can talk, and in their company I can proceed to make such inquiries about it as “methodological doubt” may properly suggest.

§ 3

In this investigation our point of departure is clear, and may be expressed in a simple question—“We have a creed; why ought we to believe it?” Let it be noted that this question must be sharply distinguished from another with which it is sometimes confounded—namely, “We have a creed; how came we to believe it?” The two problems may be closely related, but they are quite different, and even when treated together should never be confounded. The first, the one with which we are immediately concerned, is a problem for reflective reason. It is in the strictest sense philosophical. The second, on the other hand, is a problem for psychology, heredity, social history, and personal biography. It may conveniently be called scientific. During the course of our investigations we shall have to travel to and for between the two; but we should never do so unwittingly, nor should we, on one level of thought, discourse in a manner appropriate only to the other. For the moment we are on the philosophic level; and the question which I put a moment ago immediately suggests another. If we want a rational justification for our creed, should we not first consider what kind of reasons would satisfy us were we happy enough to find them? Since we are in general agreement about the sort of things we believe, can we not reach some general agreement about the sort of grounds on which we can reasonably believe them? This may seem a desire impossible of fulfillment; for why should philosophers, who differ about everything else, come to terms about this? Yet the problem is a narrow one; its solution does not involve the construction of a philosophy, but only the clarifying of our ideas as to the sort of philosophy we should like to have if we could get it. And even this statement goes too far, since the philosophy of which, for the moment, we are in search does not aim at satisfying all our aspirations. It has no celestial ambition, it never soars into the infinite; it modestly limits itself to the familiar commonplaces of daily life which some of the greatest masters of speculation have deemed far beneath their notice.

To what pattern, then, should a satisfactory “philosophy of the familiar” conform? Let us make a trial of some which naturally present themselves. Can we accept, for instance, a scheme which would treat all the various elements of our creed as truths which are independent of proof and severally proclaim themselves as self-sufficient? Evidently not. That “Julius Caesar is dead” (for example) or that “energy is conserved”—characteristic examples of historic and scientific beliefs—are plainly not self-evident propositions. They may be true, but their truth must be established. If this be admitted the idea that our creed is a collection of beliefs, all axiomatic and each independent of the others, need not be further pursued.

Let us, then, consider a second alternative. Let us suppose that not one of our beliefs considered separately has any validity of its own, but that all of them considered together form a rational and self-sufficient whole. This theory has at first sight a strong resemblance to the views on truth and reality entertained by some absolute idealists. For them also the parts considered separately are mere appearance. Only the Whole is real; only the Whole is self-consistent. But it must be observed that their Whole, if it exists at all, is a true Whole—individual, complete, all-inclusive; not to be increased by addition nor diminished by subtraction. Judged by this standard the mass of common knowledge never reaches the dignity of a Whole. So far from being all-inclusive, it is admittedly no more than an oasis of knowledge in an unmapped desert of ignorance. So far from being insusceptible of change, it changes every day. Moreover, if our belief in the parts ought properly to depend on our belief in the whole, until we know the whole our belief in the parts should remain in suspense. No half-way house seems possible between perfect omniscience and total ignorance. We know all, or we know naught; and since neither of these alternatives is tolerable we are, on this hypothesis, straightway reduced to speculative impotence.

It is then to some third pattern that the beliefs or classes of belief which constitute common knowledge should be capable of conforming, and surely the third pattern must require that the validity of the system be rooted in the validity of its fundamental portions; and that from these, by some rational process, the derivative portions must somehow be inferred. Under this scheme the elements of our creed would differ widely in their logical quality or status. Some might fitly serve as foundations; others must belong to the superstructure. Some would be ultimate, others derivative. In short, if the body of our common knowledge is to stand secure without transcendental support, its ultimate elements must be certain in their own right, and the connection between them and its derivative elements must conform to the canons of a sound logic.

§ 4

Here, then, in barest outline, is the “idea of a philosophy of familiar knowledge.” But before considering how far the idea has been, or can be, made actual, let us endeavour to develop it somewhat more in detail; let us, in particular, enquire what sort of connection between the derivative and the ultimate elements of our creed we should regard as reasonable, and what characteristics the ultimate elements must themselves possess if they are to give this reasonable support to the derivative. In my endeavour to deal with both these points I am again fortunate enough to be able to quote the authority of Mr. Bertrand Russell; for in the Harvard Lectures already referred to he comments on various kinds of belief and certain problems of inference in a manner which I must needs approve—so closely does it conform to the line I have myself on various occasions endeavoured to follow. In this connection, therefore, I shall draw without scruple upon his writings, and, sometimes, even upon my own.

If, then, we reflect on the first of the two questions I have just asked, if, that is to say, we enquire how those of our beliefs which, on reflection, we acknowledge to be inferred can be legitimately derived from those which, on reflection, we acknowledge to be ultimate, we find ourselves obliged to recognise four principles in the theory of knowledge on which Mr. Bertrand Russell and I seem to be in complete agreement. He holds, in the first place, that “in all inference, form alone is essential,”2 or, as I put it,3 “that every kind of logic, if it is to be philosophical, must be formal,” and that if “of two inferences which have the same form one is false and the other is true, the classification which connects them is philosophically worthless.”4 In the second place, we agree in thinking that no one has ever discovered any method by which formal inference is possible from particular judgments alone. In the third place, we both hold that all experience is experience of particulars, and can only be embodied in particular judgments; from which, in the fourth place, we conclude that particular experiences must be associated with general judgments before they can become fertile; so that in the logical pedigree of every derivative belief we ought to find some general proposition, which is not itself an inference. From these four principles taken together it evidently follows that particular experiences alone can supply no knowledge beyond themselves, and that pure empiricism is a dream.

It is true that Mr. Russell somewhat narrows the broad conclusion at which, by the same route, we both seem to have arrived, since he would limit this unfavourable verdict to what he calls the older empiricism—by which he means the empiricism of John Mill. But if empiricism means, as it surely ought to mean, philosophy based in the last resort solely on experiences, no such limitation is permissible. Every form which this philosophy can assume, call it old or call it new, stands equally condemned. For in strictness an empiricism which requires the support of principles which cannot be established by experience is empiricism no longer, and should appeal to us or approval under some other name. I admit, however, that philosophic practice has made it very inconvenient to apply this canon with too rigid a precision; nor do I propose to do so.

§ 5

These conclusions depend on abstract considerations, which to some may seem difficult and to others paradoxical. They are certainly inconsistent with any respect for the pure theory of induction; and dispose completely of the notion that by mere experiment and observation it is possible to establish all, or indeed any, general knowledge on a basis satisfactory to reason. Yet it is not this speculative conclusion which will most repel the ordinary reader. He will rather feel that all this talk about beliefs which are ultimate and beliefs which are derivative, about general principles and particular observations, is not only extremely tiresome but has no real relevance to the subject we are nominally discussing. “What” (he will ask) “have these logical subtleties to do with what you are pleased to call my ‘creed’? Many of its most important clauses describe beliefs which I began to hold in the nursery. I learnt that I was a ‘dweller in a world of material things’ when I tried to swallow my grandfather's watch. I was convinced that I was ‘a living being among other living beings’ when my elder brother knocked me down. I never doubted the regularity of a universe in which day and night, waking, eating, and sleeping followed each other with unbroken uniformity. As for the ‘innumerable beliefs about particular matters of fact,’ with some of which you truly say that all men are provided, I freely admit that my personal stock has been picked up from the people with whom I work, the newspapers I happen to read, and the gossip of my neighbours, supplemented of course by reminiscences of what I learned at school, or have observed in my passage through the world. If you are tactless enough to ask me why in the last resort I believe all these things or any of them, I may perhaps be embarrassed for a reply; but, under no circumstances, shall I go hunting after what you call the ‘fundamental elements’ in my ‘creed,’ or split hairs over the methods by which the ‘derivative elements’ in it may be legitimately inferred. Being only one of those ordinary men whose ordinary beliefs you condescend to discuss. I may be permitted to think that I have something better to do.”

Much of this is very sensible, and though petulantly expressed, is in substance well worth saying. At least it has the merit of bringing into high relief the fact that while our creed may quite possibly be reasonable, it is only in very small measure the product of explicit reasoning; that we occupy only as squatters our little clearing in the Unknown; and that if a legal title be desirable it still remains to be acquired. These are points we may ponder over with advantage, and properly understood they afford an additional incentive to the further pursuit of that “idea of a philosophy of the familiar” on which we are engaged.

§ 6

What has been already said should suffice, I think, to convince us that if this philosophy is to satisfy our critical ideal it must not only proceed by valid inferences from assured premises, but that among these premises certain types of belief must be represented. They must, in the first place, include particular beliefs of experience—for otherwise they would be out of all relation to a creed which, though not purely empirical, yet leans on experience at every step; and, in the second place, they must include one or more general beliefs, or otherwise the creed could never be developed into a rational system.

Do our premises answer to this description? Have we got self-evident general beliefs? Have we got indubitable particular experiences? Or, in default of these, have we got beliefs possessing a sufficient measure of intuitive probability?5

I share the common opinion that we have. But have we them in sufficient number to meet the requirements of our creed? This is another question, and one which cannot, I think, be solved without a survey of some speculative problems which have perhaps been too little considered by the great schools of philosophy.

Let me begin with one of the most important. I have just mentioned “self-evident beliefs and indubitable experiences.” To whom must they be self-evident? who in the last resort must regard them as indubitable? For the essential purposes of speculative reflection it must evidently be the individual who is speculating, and he alone. It must be you, or me, or some other person engaged in justifying to himself or others his own stock of “common knowledge.” For such a person the fact that some one else entertains a belief can never supply a final reason for entertaining it himself. I am not, of course, questioning the leading part which human testimony and social influence play in the formation of opinions. It can hardly be exaggerated. My contention is that even when testimony and influence are most powerful as causes, as reasons they are never self-sufficing. Always they require further support, always they lean on something which is not merely testimony and not merely influence, always that something must be a belief of one's own. If, therefore, we press our enquiries to their limit they land us at last in unqualified individualism.

And this, in itself, need cause us no uneasiness. There is no better security for truth than self-evidence, and where it exists and when it survives reflective criticism it can hardly be strengthened by authority. I, for example, am pretty sure that you hold, as I hold, that two things which, under like conditions, are equal to a third thing, are equal to each other. But though you may share my assurance, you cannot add to it. I am more confident that the axiom is true than that you believe it to be true. Indeed, I am more confident that the axiom is true than that you exist.

Even were it otherwise, your concurrence (itself, be it noted, known only by inference) provides but a weak and doubtful argument in favour of that which, so far as I am concerned, requires no argument at all. It already possesses, for me, the kind of inherent certainty for which I look. I know not how to ask for more; and you, with the best will in the world, are not in a position to give me as much.

So far, then, each one of us may put to his credit a certain number of axiomatic general judgments. Does each one of us also possess a number of indubitable particular experiences? With qualifications to be dealt with later, I think he does; but I do not think he does it to the extent which, at first sight, we might feel inclined to suppose. We must remember that what I have just said about individualism in the philosophy of knowledge applies as much to single experiences as to general truths. Indeed, it applies much more, or at least much more obviously. An axiom may not be better qualified to serve as one of the foundation-stones of my creed because apparently it is believed in by all the world. But at least it derives a certain dignity from its universality. My experiences, on the other hand, are essentially particular and essentially my own. They are never exactly similar in two individuals; and even if they were, they would never be exactly identical. Mine are always mine, as yours are always yours. And what is yet more important for the present purpose, your experiences can never do for my philosophy what they do for yours—they can never provide me with ultimate beliefs. They reach me only through your testimony; and since (as I have already pointed out) the trustworthiness of testimony is always a matter of inference, no experiences believed on testimony alone can be regarded as fundamental.

§ 7

It must be owned that these considerations limit in the most embarrassing way the basis of our beliefs. The statement that “thou” and “I” can only build on the narrow platform provided for “thee” and “me” by the experiences and principles which we severally regard as inherently self-sufficient is one at which we instinctively “boggle.” We habitually assume that, in a certain measure, we are the heirs of all mankind; that human experience has been largely pooled; that every person born into the world may draw with equal right from the common stock. And so (with all due qualifications) he may, if our familiar creed be true. Concede this, and our difficulties disappear. But, then, this is the very thing we want to prove; and as I have with perhaps undue insistence endeavoured to show, we cannot prove it without (among other things) a sufficient supply of premises, which to the individual thinker seem self-sufficing. Is this supply forthcoming for you and for me? If not, the “philosophy of the familiar,” as I have conceived it, seems, at least for us, to be in a somewhat perilous position.

The point will perhaps be clearer if I illustrate the difficulty by a concrete example: and my concrete example shall be the law of causation, on which depends, according to most philosophers, the scientific interpretation of (at least) the physical world. This great generalisation cannot rank with mathematical or other axioms. It is not self-evident, as self-evidence is ordinarily conceived; nor does the Kantian attempt6 to treat it as transcendentally involved in the “experience of succession” receive, or, in my opinion, deserve, much modern support.

Moreover, there are, as we have already seen, overwhelming logical difficulties involved in the endeavour to establish this or any other general principle by the mere accumulation of favourable instances, however overwhelming that accumulation may be.7 What I have now to insist on is that the accumulation of favourable instances at my disposal or at yours is far indeed from being overwhelming. It is true that empirical philosophers most confidently remind us that the law of causation is based on observations extending through all recorded time. But they have omitted to notice that unless we first believe in the law, we can have little reason for believing in the observations. The fact that mankind have been observing, or for the matter of that doing anything else, for hundreds or thousands of years, can never be known by direct acquaintance. Inference is always involved. And turn the problem round as we will, we shall always find that one root of this inference is the belief in the regularity of Nature–the very conclusion we desire to establish.

§ 8

The precise character of the individualist difficulty dealt with in the last section may, perhaps, be most clearly indicated if it is put in the form of such a question as the following:—How can I, on the narrow platform of my purely personal experiences, howsoever supplemented by logical and other trustworthy principles, erect the complex fabric of my creed? How can I, from premises so meagre, reasonably infer all that I insist on believing? How can I, aided only by these poor instruments of investigation, make any effective survey of a world with which my points of contact would seem to be so few, so transitory, and so uncertain?

It will be observed that this question throws no doubts upon the verities of reason or the trustworthiness of experience. It only assumes that for me the reason must be my reason, and the experience my experience; it only asks whether, if this be all I have to go upon, it is sufficient; and if it be not sufficient, how it is to be supplemented.

This is methodological doubt; and it has little relation to the labours of that “barren rascal” the professional sceptic, who is supposed to destroy his own philosophy by doubting his own doubts. Nevertheless there are philosophic circles where it is not always well looked upon. Some transcendental idealists, for instance, have nothing but hard words for those who suppose that in philosophy it is either necessary or legitimate for any man to depend in the last resort upon his own reason or his own experience. This, in their view, is mere “subjectivism”—useless and worse than useless for those who aspire to “objective” knowledge. Even if they hold that experience involves a “conscious centre” which experiences, they are prepared to anathematise anyone hardy enough to identify this “experiencing centre” with any historic individual who feels and thinks, who believes and doubts, who learns and who forgets. To encourage individualism like this is, in their opinion, to ruin philosophy; and it is British philosophers who, in this connection, have unwittingly proved themselves the worst and most obstinate offenders.

Now it is fortunately unnecessary for me to deal elaborately with this formidable controversy. I content myself with pointing out that I, and those (if any) who are of my way of thinking, dwell in regions so remote from these speculative heights as to be safe from attack by those who occupy them. Our ambitions are modest. We only want to know, each for himself, why certain beliefs, which (it seems) are common to us all, should be treated as true. Such a desire does not, at first sight, seem either foolish or perverse. But, in any case, it is ours. Let those who do not share it trouble us no further about the matter. But do not let them cherish the illusion that they can solve the sort of difficulty we feel by requiring us to assume as our point of departure “Thought” in general—thought which is not my thought, or their thought, or the thought of any individual thinker. This may turn out to be good ontology; it may provide a satisfactory account of Reality. I am far from denying that the universe may be essentially spiritual; that the essence of spirit may be Thought, and that in the unity of this transcendental Thought all minor distinctions, such as those which divide you from me, and both of us from the Absolute, may somehow be dissolved without being destroyed. But unless these high doctrines be inherently self-supporting they must be established dialectically; and the dialectic must be rooted in beliefs which are logically antecedent to them. These beliefs must in my case be my beliefs, in your case be yours, in every case the beliefs of particular individuals, of “empirical egos,” of mere men and women, who possess no more claim to constitute the universe than you do, or than I!

This, then, is the pattern to which, as I think, a satisfactory philosophy of the commonplace must conform. There is no use telling me that it is “subjectivism.” After all, I am a “subject,” and, like the rest of the world, must suffer from my limitations. From this situation there is no escape. But let me not be misunderstood. The individualism I defend is not designed to isolate the individual even in matters which are purely intellectual, still less in matters which concern the will and the affections. I myself hold, as I shall later explain, that the ordinary theory of human intercourse is narrow and defective; that it unduly ignores much of what is actual and still more of what is possible in the relation of spirit to spirit in social and religious life. But this is not our present business. What we are now concerned with is the rationalization for each one of us of his or her familiar beliefs. This is an individualistic problem which can best be expressed in individualistic terms, and cannot receive any but an individualistic solution. Whatever, then, may be said for or against the wisdom of discussing it, the form in which it has been stated is not, I trust, open to serious objection.

• 1.

  On this subject cf. “Theism and Humanism,” ch. viii.

• 2.

  “External World,” p. 44.

• 3.

  “Philosophic Doubt,” p. 11.

• 4.

  “Philosophic Doubt,” p. 10.

• 5.

  On intuitive probability, see Appendix.

• 6.

  Cf. “Philosophic Doubt,” p. 124.

• 7.

  Cf. “Philosophic Doubt,” p. 71, from which much of what follows is a quotation.