This will probably be a new conception to most people. Accordingly I shall have to spend a good deal of my limited space in justifying my case with the aid of evidence and argument. But evidence and argument are too frequently tedious. The average man prefers statements to arguments, conclusions to evidence. There is something in this attitude. We can often make up our minds more readily about a man after we have heard him proclaim his case briefly than we could if we had listened to him make a completely step-by-step logical justification of it.

At the risk of repetition, therefore, I propose at the outset to state my beliefs briefly in their main outlines, without the attempt at full justification by reasoned argument. This will give my readers a preliminary view; if they do not like it, there is after all no need for them to go on with the book. Later, after a personal digression, I shall come to impersonal exposition and argument; and finally shall present what seem to me the right conclusions.

What then do I believe? I believe, in the first instance, that it is necessary to believe something. Complete scepticism does not work. On the other hand, I believe equally strongly that it is always undesirable and often harmful to believe without proper evidence. Everything which we believe, except the local necessities of mathematics and formal logic, is believed on external evidence of one sort or another, although the evidence may have been assimilated long ago, or so completely or so intuitively, that we are not conscious of it. To take a simple and trivial example: when we say that a ball which we see in the distance is sperical we are basing this statement on the frequently repeated evidence of our past experience that objects which appear to the eye of a particular shape and with a particular kind of pattern of light and shade are, when explored by touch, found to possess a particular shape we call spherical; and we had to learn all this very thoroughly (although we have by now forgotten all about the learning proces) when we were babies. But even when there is not this necessary interpretation of the evidence of one sense in terms of the experience drived from another, but a direct utilisation of the materials provided by one sense only, we are still believing on evidence. When I feel a marble with my fingers, my eyes being blindfolded, I can judge directly by touch that it is a single spherical object. But everybody knows (or if he does not, let him immediately try the very simple but fundamental experiment) that if I cross two adjacent fingers and feel the marble between their crossed tips, it will be obstinately judged to be double, in spite of all knowledge to the contrary. Thus even the simple judgements of sense may be illusions, and when I say that I believe something because I saw or heard it, I am backing the view that I did not happen to be deluded. When we come to more complicated beliefs, such as the beleief that so-and-so is really angry with us, although he is doing his best to appear friendly, or that someonelse is an honest man and will never take an unfair advantage, it is still more clear weighing evidence at a conclusion (however intuitively) on the balance. And we are often wrong. How frequently it turns out that A's apparent anger was only dyspepsia, or that we were sadly mistaken as to B's honesty!

Apart from intellectual mistakes or sensory delusions, however, there is a still more potent source of error in emotional distortion. Even serious investigators have not always escaped having their conclusions colored by desires, seeing what they want to see, and even more, not seeing what they would prefer to overlook. An angry man is notoriously bad witness; and the judgements of first love about the beloved object are quite generally discounted, and that not only by cynics.

There is thus a certain practical difficulty. We must believe something, for otherwise we should never act. On the other hand, we must believe everything, or believe too readily, or we shall act wrongly. Most people would say that they are completely justified in the certain belief that the sun would rise next morning; on the other hand, there is for this no inherent necessity of the same nature as the inherent necessity for two and two to make four; something might pefectly well happen to prevent its rising; and we might believe in the existence of this something. As a matter of fact, a great many people at one time or another have believed that the world would end on a particular date, and therefore the sun would not rise one fine morning, and this belief (although always, so far, it has proved erroneous) and have often very radically affected their lives. The closing months of the year A.D.999 were accompanied by the most improbable scenes of orgy, terror and prayer, owing to the belief that the world would end at the millennium; and even in our time the members of an American sect sold all their possessions very cheap and went to await the end of the world and the translation to heaven on a convenient hill-top.

Experience has quite definitely shown (if only humanity could be persuaded to profit by her!) that some reasons for holding a belief are much more likely to be justified by the events than others. It might be naturally supposed, for instance, that the best of all reasons for belief was a strong conviction of certainty accompanying the belief. Experience however shows that this is not so, and that, as a matter of fact, conviction by itself is more likely to mislead than it is to guarantee truth. On the other hand, lack of assurance and persistent hesitation to come to any belief whatever are equally poor guarantees that the few beliefs which are arrived at are sound. Experience also shows, that assertion, however long continud, although it is unfortunately with many people an effectibe enough means of inducing belief, it is not in any way a ground for holding it.

Neither is a claim to be the receipient of a revelation the least guarantee that belief in the subject of the revelation is justified; for both madmen and false prophets have made the claim.

The method which has proved effective, as matter of actual fact, in providing a firm foundation for belief wherever it has been capable of application is what is usually called the scientific method. I believe firmly that the scientific method, although slow and never claiming to lead to complete truth, is the only method which in the long run will give satisfactory foundations for belief. The scientific method is the method, which in the intellectual sphere, is the counterpart of that method recommended by the apostle in the moral sphere - test all things; hold fast to that which is good. It consists in demanding facts as the only basis for conclusions; and of consistently and continuously testing any conclusions which may have been reached by ne facts and, wherever possible, by the crucial test of experiment, it consists also (and this is not sufficiently recognized by the generality of people) in full publication of the evidence on which conclusions are based so that other workers may have the advantage of the fatcs, to assist in new researches or, as frequently occurs, to make it possible for them to put a quite different interpretation on the facts.

There are, however, all sorts of occasions on which the scientific method is not applicable. That method involves slow testing, frequent suspension of judgement, restricted conclusions. The exigencies of everyday life, on the other hand, oten make it necessary on a hasty balancing of admittedly incomplete evidence, to take immediate action, and to draw conclusions in advance of the evidence. It is also true that such action will be necessary, and necessary in respect of ever larger issues; and this in spite of the fact that one of the most important trends of civilization is to remove sphere after sphere of life out of the domain of such intuitive judgement into the domain of rigid calculation based on science. It is here that belief plays its most important role. When we cannot be certain, we must proceed in part by faith - faith not only in the validity of our own capacity of making judgements, but also in the existence of certain factual realities, pre-eminently moral and spiritual realities. It has been said that faith consists in acting always on the nobler hypothesis; and though this definition is a trifle rhetorical, it embodies a seed of real truth.

Finally, however (and this is a truth which has been often wholly unrecognized, and never popular), and there are other occasions on which belief is not only not demanded, but is, in the phraseology of medicine, contra-indicated. When there exists no evidence or next to no evidence, and when the conclusion to which we may come can have no influence on the facts, then it is our duty to suspend judgement and hold no belief, just as definitely as it is our duty, when practical issues hang on our decision, not to suspend judgement, but to take our courage in both hands and act on the best belief at which we can arrive. This duty of refraining from belief is often imposed on men of science in their work, in order that they may in the long run arrive at greater certitude; it is also imposed upon them in other cases in order that they may not encourage false hopes of certitude. When applied to whole problems, this attitude of mind generally goes by the name (first coined by Thomas Huxley) of agnosticism. I hold it to be an important duty to know when to be agnostic. I believe that one should be agnostic when belief one way or the other is mere idle speculation, incapable of verification; when belief is held merely to gratify desires, however deep seated, and not because it is forced on us by evidnce; and when belief nay be taken by others to be more firmly grounded than it really is, and so come to encourage false hopes or wrong attitudes of mind.

That is a long exordium, I fear. It must be justified by the fact that our beliefs about Belief are among the most important that we may possess, and this all the more since we rarely stop to give their existence a thought.

I hold, then, that all our life long we are oscillating between conviction and caution, faith and agnosticism, belief and suspension of belief. That neither faith nor agnosticism is in itself the better way, but that each has its right occasions. That beliefs which are well enough for individual occasions of practical necessity may be wholly unjustified and unjustifiable when made general or when taken to dispense from further enquiry. In fact, I hold that beliefs are essential tools of the human mind - no more than tools, but no less than essential. That there is therefore no more sense in using the same sort of belief to help in solving both problems of ultimate and universal values and the practical problems of daily necessity that there would be in using a kitchen scales to determine atomic weights, or, vice versa, a string galvanometer for the purposes of the job electrician. And hat there is no more justification for wasting time and energy and hope in the drawing conclusions about subjects on which inadequate evidence exists in founding a Department of State for the breeding of hippogriffs, or inventing a method for crossing bridges before one comes to them.

SORRY! THIS ARTICLE IS STILL UNDER CONSTRUCTION