called the categories of the understanding: ideas of time, space,4 class, number, cause, substance, personality, etc. They correspond to the most universal properties of things. They are like the solid frame which encloses all thought; this does not seem to be able to liberate itself from them without destroying itself, for it seems that we cannot think of objects that are not in time and space, which have no number, etc. Other ideas are contingent and unsteady; we can conceive of their being unknown to a man, a society or an epoch; but these others appear to be nearly inseparable from the normal working of the intellect. They are like the framework of the intelligence. Now when primitive religious beliefs are systematically analysed, the principal categories are naturally found. They are born in religion and of religion; they are a product of religious thought. This is a statement that we are going to have occasion to make many times in the course of this work.
This remark has some interest of itself already; but here is what gives it its real importance.
The general conclusion of the book which the reader has before him is that religion is something eminently social. Religious representations are collective representations which express collective realities; the rites are a manner of acting which take rise in the midst of the assembled groups and which are destined to excite, maintain or recreate certain mental states in these groups. So if the categories are of religious origin, they ought to participate in this nature common to all religious facts; they too should be social affairs and the product of collective thought. At least — for in the actual condition of our knowledge of these matters, one should be careful to avoid all radical and exclusive statements — it is allowable to suppose that they are rich in social elements.
Even at present, these can be imperfectly seen in some of them. For example, try to represent what the notion of time would be without the processes by which we divide it, measure it or express it with objective signs, a time which is not a succession of years, months, weeks, days and hours! This is something nearly unthinkable. We cannot conceive of time, except on condition of distinguishing its different moments. Now what is the origin of this differentiation? Undoubtedly, the states of consciousness which we have already experienced can be reproduced in us in the same order in which they passed in the first place; thus portions of our past become present again, though being clearly distinguished from the present. But howsoever important this distinction may be for our private experience, it is far from being enough to constitute the notion or category of time. This does not consist merely in a commemoration, either partial or integral, of our past life. It is an abstract and impersonal frame which surrounds, not only our individual existence, but that of all humanity. It is like an endless chart, where all duration is spread out before the mind, and upon which all possible events can be located in relation to fixed and determined guide lines. It is not my time that is thus arranged; it is time in general, such as it is objectively thought of by everybody in a single civilization. That alone is enough to give us a hint that such an arrangement ought to be collective. And in reality, observation proves that these indispensable guide lines, in relation to which all things are temporally located, are taken from social life. The divisions into days, weeks, months, years, etc., correspond to the periodical recurrence of rites, feasts, and public ceremonies.5 A calendar expresses the rhythm of the collective activities, while at the same time its function is to assure their regularity.6
It is the same thing with space. As Hamelin has shown,7 space is not the vague and indetermined medium which Kant imagined; if purely and absolutely homogeneous, it would be of no use, and could not be grasped by the mind. Spatial representation consists essentially in a primary co-ordination of the data of sensuous experience. But this co-ordination would be impossible if the parts of space were qualitatively equivalent and if they were really interchangeable. To dispose things spatially there must be a possibility of placing them differently, of putting some at the right, others at the left, these above, those below, at the north of or at the south of, east or west of, etc., etc., just as to dispose states of consciousness temporally there must be a possibility of localizing them at determined dates. That is to say that space could not be what it is if it were not, like time, divided and differentiated. But whence come these divisions which are so essential? By themselves, there are neither right nor left, up nor down, north nor south, etc. All these distinctions evidently come from the fact that different sympathetic values have been attributed to various regions. Since all the men of a single civilization represent space in the same way, it is clearly necessary that these sympathetic values, and the distinctions which depend upon them, should be equally universal, and that almost necessarily implies that they be of social origin.8
Besides that, there are cases where this social character is made manifest. There are societies in Australia and North America where space is conceived in the form of an immense circle, because the camp has a circular form;9 and this spatial circle is divided up exactly like the tribal circle, and is in its image. There are as many regions distinguished as there are clans in the tribe, and it is the place occupied by the clans inside the encampment which has determined the orientation of these regions. Each region is defined by the totem of the clan to which it is assigned. Among the Zuñi, for example, the pueblo contains seven quarters; each of these is a group of clans which has had a unity: in all probability it was originally a single clan which was later subdivided. Now their space also contains seven quarters, and each of these seven quarters of the world is in intimate connection with a quarter of the pueblo, that is to say with a group of clans.10 "Thus," says Cushing, "one division is thought to be in relation with the north, another represents the west, another the south," etc.11 Each quarter of the pueblo has its characteristic colour, which symbolizes it; each region has its colour, which is exactly the same as that of the corresponding quarter. In the course of history the number of fundamental clans has varied; the number of the fundamental regions of space has varied with them. Thus the social organization has been the model for the spatial organization and a reproduction of it. It is thus even up to the distinction between right and left which, far from being inherent in the nature of man in general, is very probably the product of representations which are religious and therefore collective.12
Analogous proofs will be found presently in regard to the ideas of class, force, personality and efficacy. It is even possible to ask if the idea of contradiction does not also depend upon social conditions. What makes one tend to believe this is that the empire which the idea has exercised over human thought has varied with times and societies. To-day the principle of identity dominates scientific thought; but there are vast systems of representations which have played a considerable rôle in the history of ideas where it has frequently been set aside: these are the mythologies, from the grossest up to the most reasonable.13 There, we are continually coming upon beings which have the most contradictory attributes simultaneously, who are at the same time one and many, material and spiritual, who can divide themselves up indefinitely without losing anything of their constitution; in mythology it is an axiom that the part is worth the whole. These variations through which the rules which seem to govern our present logic have passed prove that, far from being engraven through all eternity upon the mental constitution of men, they depend, at least in part, upon factors that are historical and consequently social. We do not know exactly what they are, but we may presume that they exist.14
This hypothesis once admitted, the problem of knowledge is posed in new terms.
Up to the present there have been only two doctrines in the field. For some, the categories cannot be derived from experience: they
