It is no mere paradox to say that religion is most interesting to those who have ceased to believe in it. The reason for this is not far to seek. Religious beliefs play so large a part in the early history of society, and are so influential in social history generally, that it is impossible to leave religion alone without forfeiting an adequate comprehension of a large part of social evolution. Human development forms a continuous record; our institutions, whatever be their nature, have their roots in the far past, and often, even when modified in form, retain their essential characteristics. No student of social history can travel far or dig deeply without finding himself in contact with religion in some form. And the mass of mankind are not yet so far removed from "primitive" humanity as to give to the study of religion an exclusively archÆological interest.

Where so much is discord it is well, if it be possible, to start with a basis of agreement. And on one point, at least, there is substantial unity among critics. There is a general agreement among students of folk lore, comparative mythology, and anthropology, that religious ideas rest ultimately upon an interpretation of nature that is now generally discarded. Differing as they do on details, there is consent upon this point. It is the world of the savage that originates the religion of the savage, and upon that rests the religions of civilized man as surely as his physical structure goes back to the animal world for its beginning. And in giving birth to a religious explanation of his world the savage was only pursuing the normal path of human development. Mankind progresses through trial and error; doubtful and erroneous theories are framed before more reliable ones are established, and while truth may crown our endeavours it seldom meets us at the outset. Religious beliefs thus form man's earliest interpretation of nature. On this there is, as I have said, general agreement, and it is as well not to permit ourselves to lose sight of that in the discussion of the various theories that are put forward as to the exact nature of the stages of religious development.

In many directions the less accurate theories of things are replaced gradually and smoothly by more reliable explanations. But in religion this is not so. For many reasons, with which we are not now immediately concerned, religious beliefs are not outgrown without considerable "growing pains." And a long time after the point of view from which religious beliefs sprang has been given up, the conclusions that were based on that point of view are held to most tenaciously. And yet if one accepts the scientific story of the origins of religious ideas there seems no justification whatever for this. Religion cannot transcend its origin. Multiply nothing to infinity and the result is still nothing. Illusion can beget nothing but illusion, even though in its pursuit we may stumble on reality. And no amount of ingenuity can extract truth from falsehood.

One's surprise at the perpetuation of this particular delusion is diminished by the reflection that the period during which we have possessed anything like an exact knowledge of the character and operations of natural forces is, after all, but an infinitesimal portion of the time the race has been in existence. Three or four centuries at most cover the period during which such knowledge has been at our command, and small as this is in relation to the thousands of generations wherein superstition has reigned unchallenged, a knowledge of the laws of mental life belongs only to the latter portion. And even then the knowledge available has been till recently the possession of a class, while to-day, large masses of the population are under the domination of the crudest of superstitions. The belief that thirteen is an unlucky number, that a horse-shoe brings luck, the extent to which palmistry and astrology flourishes, the cases of witchcraft that crop up every now and again, all bear testimony to the vast mass of superstition that is still with us. The primitive mind is still alive and active, disguised though it may be by a veneer of civilization and a terribly superficial education. And when one reflects upon all the facts there is cause for astonishment that in the face of so great a dead weight of custom and tradition against a rational interpretation of the universe so much has been done and in so short a time.

In discussing religion very much turns upon the meaning of the word, and unfortunately "religion" is to-day used in so many differing and conflicting senses that without the most careful definition no one is quite sure what is meant by it. The curious disinclination of so many to avow themselves as being without a religion must also be noted. To be without a religion, or rather to be known as one who is without a religion, would seem to mark one off as apart from the rest of one's kind, and to infringe all the tribal taboos at one sweep. And very few seem to have the courage to stand alone. Mr. Augustine Birrell once said, in introducing to the House of Commons an Education Bill, that children would rather be wicked than singular. That is quite true, and it is almost as true of adults as it is of children. There is no great objection to having a religion different from that of other people, because the religions of the world are already of so varied a character that there is always companionship in difference. But to be without a religion altogether is a degree of isolation that few can stand. The consequence is that although vast numbers have given up everything that is really religious they still cling to the name. They have left the service, but they show a curious attachment to the uniform. Thus it happens that we have a religion of Socialism, a religion of Ethics, etc., and I should not be surprised to find one day a religion of Atheism—if that has not already appeared.

But all this is a mistake, and a very serious mistake. The Freethinker, or Socialist, who calls his theory of life a religion is not causing the religionist to think more highly of him, he is making his opponent think more highly of his own opinions. Imitation becomes in such a case not alone flattery, but confirmation. The Goddite does not think more highly of Freethought because it is labelled religion, he merely becomes the more convinced of the supreme value of his own faith, and still hopes for the Freethinker's return to the fold. If Freethinkers are to command the respect of the religious world they must show not only that they can get along without religion, but that they can dispense with the name also. If strength does not command respect weakness will certainly fail to secure it. And those of us who are genuinely anxious that the world should be done with false ideas and mischievous frames of mind ought to at least take care that our own speech and thought are as free from ambiguity as is possible. There is another and deeper aspect of the matter. As I have already said, language not alone expresses thought, it also governs and directs it. Locke expressed this truth when he said, "It is impossible that men should ever truly seek, or certainly discover, the disagreement of ideas themselves whilst their thoughts flutter about, or stick only on sounds of doubtful and uncertain significance." Quite a number of theological and metaphysical conundrums would lose their significance if it were only realized that the words used are not alone of doubtful and uncertain significance, but often of no possible significance whatever. They are like counterfeit coins, which retain their value only so long as they are not tested by a proper standard. And the evil of these counterfeits is that they deceive both those who tender and those who accept them. For even though slovenliness of speech is not always the product of slovenly thought, in the long run it tends to induce it, and those who realize this need to be specially on their guard against using language which can only further confuse an already sufficiently confused public opinion, and strengthen superstitions that are already sufficiently strong without our clandestine or unintended assistance.[15]

Unfortunately, it remains a favourite policy with many writers to use and define the word religion, not in accordance with a comprehensive survey of facts, but in a way that will harmonize with existing pre-possessions. To this class belongs Matthew Arnold's famous definition of religion as "Morality touched with emotion," Professor Seeley's statement that we are entitled to call religion "any habitual and permanent admiration," or the common description of religion as consisting in devotion to an ideal. All such definitions may be set on one side as historically worthless, and as not harmonizing with the facts. Arnold's definition is in the highest degree superficial, since there exists no morality that is not touched with emotion, and on the other hand there exist phases of religion that have not any connection with morality, however slight. Professor Leuba properly rules definitions of this class out of order in the comment that, as it is "the function of words to delimitate, one defeats the purpose of language by stretching the meaning of a word until it has lost all precision and unity of meaning."[16] A definition that includes everything may as well, for all the use it is, not cover anything.

Equally faulty are those definitions that are based upon an assumed conscious effort to explain the mysteries of existence. No stranger lapse ever overtook a great thinker than occurred to Herbert Spencer when he described religion as consisting in a worship of the unknowable, and as due to the desire to explain a mystery ever pressing for interpretation. Granting the existence of an Unknowable, the sense of its presence belongs to the later stages of mental evolution, not to the earlier ones. Metaphysical and mystical theories of religion are indications of its disintegration, not of its beginnings. Primitive man began to believe in ghosts and gods for the same reasons that he believed in other things; he worshipped his gods for very concrete considerations. Even the distinction between "spiritual" and material existence is quite foreign to his mind. Such distinctions arise gradually with the progress of knowledge and its disintegrating influence on inherited beliefs. If primitive man may be credited with a philosophy, and if one may use the word in a purely convenient sense, then one may say that he is neither a dualist, nor a pluralist, but a monist. The soul or double he believes in is similar to the body he sees; the unseen forces he credits with various activities are of the same kind as those with which he is acquainted. To read our conceptions into the mind of primitive man because we use our words to explain his thoughts is a procedure that is bound to end in confusion. Man's earliest conception of things is vague and indefinite. Later, he distinguishes differences, qualitative and quantitative, his conception of things becomes more definite, and distinctions are set up that lay the foundations of science and philosophy, and which mark their separation from religion.

So far as one can see there are only two causes why people should continue to use the word religion after giving up all for which it properly stands. One is sheer conservatism. When, for instance, Thomas Paine said, "To do good is my religion," he had at least the justification of believing in a deity, but apart from this the only cause for his calling the desire to do good a religion is that there had grown up the fashion of calling one's rule of life a religion. The other cause is the ill-repute that has been attached to those who avow themselves as being without religion. Orthodoxy saw to it that they were treated as pariahs without social status, and, in many cases, legal rights. Once upon a time it was useless unless one believed in the right religion. Nowadays, any religion will do, or anything that one cares to call a religion. But not to have any religion at all still puts one outside the pale of respectability, and there seem to be few who can stand that. And supernatural religion—the only genuine article—being impossible with many, these may still, if they care to, save their face by professing to use the name, even if they have not the thing. Orthodoxy is very accommodating nowadays.

Leaving for a time the question of how religion actually does arise, we may turn to those writers who define religion in terms of ethics. It may be admitted that so far as the later stages of religion are concerned considerable emphasis is laid upon ethics. But we can only make religion a part of ethics by expanding the term morality so as to include everything, or by contracting it so as to exclude all the lower forms of religious belief. And any definition of religion that does not embrace all its forms is obviously inaccurate. It is not at all a question of defining the higher in terms of the lower, or the lower in terms of the higher, it is simply the need of so defining religion that our definition will cover all religions, high and low, and thus deal with their essential characteristics.

The only sense in which ethics may be said to be included in religion lies in the fact that in primitive times religion includes everything. The fear of unseen intelligences is one of the most powerful factors of which early humanity is conscious, and the necessity for conciliating them is always present. The religious ceremonies connected with eating and drinking, with lying down and rising up, with sowing and reaping, with disease, hunting, and almost every circumstance of primitive life prove this. Differentiation and discrimination arise very slowly, but one after another the various departments of life do shake off the controlling influence of religion. Ethics may, therefore, be said to originate in the shadow of religion—as do most other things—but in no sense can morality be said to owe its origin to religion. Its origin is deeper and more fundamental than religion. As a matter of practice morality is independent of religious belief and moral theory, and as a matter of theory the formulation of definite moral rules is substantially independent of religion and is an assertion of its independence. Indeed, the conflict between a growing moral sense and religion is almost as large a fact in the social sphere as the conflict between religion and science is in the intellectual one.

In all its earlier stages religion is at best non-moral. It becomes otherwise later only because of the reaction of a socialized morality on religious beliefs. Early religion is never concerned with the morality of its teaching, nor are the worshippers concerned with the morality of their gods. The sole question is what the gods desire and how best to satisfy them. We cannot even conceive man ascribing ethical qualities to his gods until he has first of all conceived them in regard to his fellow men. The savage has no moral reverence for his gods; they are magnified men, but not perfect ones. He worships not because he admires, but because he fears. Fear is, indeed, one of the root causes of religious belief. Professor Leuba quite admits the origin of religion is fear, but he reserves the possibility of man being occasionally placed under such favourable conditions that fear may be absent. We admit the possibility, but at present it remains a possibility only. At present all the evidence goes to prove the words of Ribot that, "The religious sentiment is composed first of all of the emotion of fear in its different degrees, from profound terror to vague uneasiness, due to faith in an unknown mysterious and impalpable power." And if that be admitted, we can scarcely find here the origin of morality. What is here overlooked is the important fact that while religion, as such, commences in a reasoned process, morality is firmly established before mankind is even aware of its existence. A formulated religion is essentially of the nature of a theory set forth to explain or to deal with certain experiences. Morality, on the other hand, takes its rise in those feelings and instincts that are developed in animal and human societies under the pressure of natural selection. The affection of the animal for its young, of the human mother for its child, the attraction of male and female, the sympathetic feelings that bind members of the same species together, these do not depend upon theory, or even upon an intellectual perception of their value. Theory tries to account for their existence, and reason justifies their being, but they are fundamentally the product of associated life. And it is precisely because morality is the inevitable condition of associated life that it has upon religion the effect of modifying it until it is at least not too great an outrage upon the conditions of social well-being. All along we can, if we will, see how the developing moral sense forces a change in religious teaching. At one time there is nothing revolting in the Christian doctrine of election which dooms one to heaven and another to hell without the slightest regard to personal merit. At another the doctrine of eternal damnation is rejected as a matter of course. Heresy hunting and heretic burning, practised as a matter of course by one generation become highly repulsive to another. In every direction we see religious beliefs undergoing a modification under the influence of moral and social growth. It is always man who moralizes his gods; never by any chance is it the gods who moralize man.

If we are to arrive at a proper understanding of religion we can, therefore, no more assume morals to be an integral part of religion than we can assume medicine or any of the special arts, all of which may be associated with religion. It will not even do to define religion with Mr. W. H. Mallock[17] as a belief that the world "has been made and is sustained by an intelligence external to and essentially independent of it." That may pass as a definition of Theism, but Theism is only one of the phases of religion, and the idea of a creator independent of the universe is one that is quite alien to the earlier stages of religion. And to deny the name of religion to primitive beliefs is to put oneself on the level of the type of Christian who declines to call any superstition but his own religion. It is for this reason impossible to agree with Professor Leuba when he says that "the idea of a creator must take precedence of ghosts and nature beings in the making of a religion." If by precedence the order of importance, from the standpoint of later and comparatively modern forms of religion, is intended, the statement may pass. But if the precedence claimed is a time order, the reply is that, instead of the idea of a creator taking precedence of ghosts and nature beings, it is from these that the idea of a creator is evolved. It is quite true Professor Leuba holds that "belief in the existence of unseen anthropopathic beings is not religion. It is only when man enters into relation with them that religion comes into existence," but so soon as man believes in the existence of them he believes himself to be in relation with them, and a large part of his efforts is expended in making these relations of an amicable and profitable character.

A further definition of religion, first given, I think, by the late Professor Fiske, but since widely used, as a craving for "fulness of life," must be dismissed as equally faulty. For if by fulness of life is meant the desire to make it morally and intellectually richer, the answer is that this desire is plainly the product of a progressive social life, of which much that now passes for religion is the adulterated expression. Apologetically, it is an attempt so to state religion that it may evade criticism of its essential character. From one point of view this may be gratifying enough, but it is no help to an understanding of the nature of religion. And how little religion does help to a fuller life will be seen by anyone who knows the part played by organized religion in mental development and how blindly obstructive it is to new ideas in all departments of life. All these attempts to define religion in terms of ethics, of metaphysics, or as the craving after an ideal are wholly misleading. It is reading history backwards, and attributing to primitive human nature feelings and conceptions which it does not and cannot possess.

In another work[18] I have traced the origin of the belief in God to the mental state of primitive mankind, and there is no need to go over the same ground here at any length. Commencing with the indisputable fact that religion is something that is acquired, an examination of the state of mind in which primitive mankind faced, and still faces, the world, led to the conclusion that the idea of god begins in the personification of natural forces by the savage. The growth of the idea of God was there traced back to the ghost, not to the exclusion of other methods of god making, but certainly as one of its prominent causes. I must refer readers to that work who desire a more extended treatment of the god-idea. What remains to be traced here, in order to understand the other factor that is common to religions, is the belief in a continued state of existence after death, or at least of a soul.

It has been shown to the point of demonstration by writers such as Spencer, Tylor, and Frazer, that the idea of a double is suggested to man by his experience of dreams, swoons, and allied normal and abnormal experiences. Even in the absence of evidence coming to us from the beliefs of existing tribes of savages, the fact that the ghost is always depicted as identical in appearance with the living person would be enough to suggest its dream origin. But there are other considerations that carry the proof further. The savage sees in his dreams the figures of dead men and assumes that there is a double that can get out of the body during sleep. But he also dreams of dead men, and this is also proof that the dead man still exists. Death does not, then, involve the death of the ghost, but only its removal to some other sphere of existence. Further, the likeness of sleep itself to death is so obvious and so striking that it has formed one of the most insistent features of human thought and speech. With primitive man it is far more than a figure of speech. The Melanesians put this point of view when they say, "the soul goes out of the body in some dreams, and if for some reason it does not come back the man is found dead in the morning." Death and dreaming have, therefore, this in common, they are both due to the withdrawal of the double. Hence we find a whole series of ceremonies designed to avert death or to facilitate the return of the double. The lingering of this practice is well illustrated by Sir Frederick Treves in his book, The Other Side of the Lantern. He there tells how he saw a Chinese mother, with the tears streaming down her face, waving at the door of the house the clothing of a recently deceased child in order to bring back the departed spirit.

Death is thus the separation of the double from the body; but if it may return, its return is not always a matter of rejoicing, for we find customs that are plainly intended to prevent the ghost recognizing the living or to find its way back to its old haunts. Thus Frazer has shown that the wearing of black is really a form of disguise. It is a method taken to disguise the living from the attentions of the dead. It is in order to avoid recognition by spirits who wish to injure them that the Tongans change their war costume at every battle. The Chinese call their best beloved children by worthless names in order to delude evil spirits. In Egypt, too, the children who were most thought of were the worst clad. In some places the corpse is never carried out through the door, but by a hole in the side of the hut, which is afterwards closed so that the ghost may not find its way back.

The ghost being conceived as at all points identical with living beings, it demands attention after death. It needs food, weapons, servants, wives. In this way there originates a whole group of burial customs, performed partly from fear of what the ghost may do if its wants are neglected. The custom of burying food and weapons with the dead thus receives a simple explanation. These things are buried with the dead man in order that their spirit may accompany his to the next world and serve the same uses there that they did here. The modern custom of scattering flowers over a grave is unquestionably a survival of this primitive belief. The killing of a wife on the husband's grave has the same origin. Her spirit goes to attend the husband in the ghost-land. In the case of a chief we have the killing of servants for the same reason. When Leonidas says, "Bury me on my shield, I will enter even Hades as a LacedÆmonian," he was exhibiting the persistence of this belief in classical times. The Chinese offer a further example by making little paper houses, filling them with paper models of the things used by the dead person, and burning them on the grave. All over the world we have the same class of customs developing from the same beliefs, and the same beliefs projected by the human mind when brought face to face with the same class of phenomena.

As the ghost is pictured as like the physical man, so the next world is more or less a replica of this. The chief distinction is that there is a greater abundance of desirable things. Hunting tribes have elysiums where there is an abundance of game. The old Norse heaven was a place where there was unlimited fighting. The gold and diamonds and rubies of the Christian heaven represent a stage of civilization where these things had acquired a special value. Social distinctions, too, are often maintained. The Caribs believe that every time they secure an enemy's head they have gained a servant in the next world. And all know the story of the French aristocrat who, when threatened with hell, replied, "God will think twice before damning a person of my quality."

Several other consequences of this service paid to the dead may be noted. The ghost being drawn to the place where the body is buried, the desire to preserve the corpse probably led to the practice of embalming. The grave becomes a place of sanctity, of pilgrimage, and of religious observance, and it has been maintained by many writers, notably by Mr. W. Simpson in his Worship of Death, that the service round the grave gives us the beginning of all temple worship.

But from this brief view of the beginnings of religion we are able to see how completely fallacious are all those efforts to derive religion from an attempt to achieve an ideal, from a desire to solve certain philosophical problems, or from any of the other sources that are paraded by modern apologists. The origin and nature of religion is comparatively simple to understand, once we have cleared our minds of all these fallacies and carefully examine the facts. Religion is no more than the explanation which the primitive mind gives of the experiences which it has of the world. And, therefore, the only definition that covers all the facts, and which stresses the essence of all religions, high and low, savage and civilized, is that given by Tylor, namely, the belief in supernatural beings. It is the one definition that expresses the feature common to all religions, and with that definition before us we are able to use language with a precision that is impossible so long as we attempt to read into religion something that is absent from all its earlier forms, and which is only introduced when advanced thought makes the belief in the supernatural more and more difficult to retain its hold over the human mind.