[Sidenote: Savages of the present day are primitive only in a relative sense, namely by comparison with civilized peoples; their customs and beliefs are in fact the product of a long course of evolution as to which we can know little or nothing.] But here it is necessary to guard against a common misapprehension. The savages of to-day are primitive only in a relative, not in an absolute sense. They are primitive by comparison with us; but they are not primitive by comparison with truly primæval man, that is, with man as he was when he first emerged from the purely bestial stage of existence. Indeed, compared with man in his absolutely pristine state even the lowest savage of to-day is doubtless a highly developed and cultured being, since all evidence and all probability are in favour of the view that every existing race of men, the rudest as well as the most civilized, has reached its present level of culture, whether it be high or low, only after a slow and painful progress upwards, which must have extended over many thousands, perhaps millions, of years. Therefore when we speak of any known savages as primitive, which the usage of the English language permits us to do, it should always be remembered that we apply the term primitive to them in a relative, not in an absolute sense. What we mean is that their culture is rudimentary compared with that of the civilized nations, but not by any means that it is identical with that of primæval man. It is necessary to emphasize this relative use of the term primitive in its application to all known savages without exception, because the ambiguity arising from the double meaning of the word has been the source of much confusion and misunderstanding. Careless or unscrupulous writers have made great play with it for purposes of controversy, using the word now in the one sense and now in the other as it suited their argument at the moment, without perceiving, or at all events without indicating, the equivocation. In order to avoid these verbal fallacies it is only necessary to bear steadily in mind that while Social Anthropology has much to say of primitive man in the relative sense, it has nothing whatever to say about primitive man in the absolute sense, {164} and that for the very simple reason that it knows nothing whatever about him, and, so far as we can see at present, is never likely to know anything. To construct a history of human society by starting from absolutely primordial man and working down through thousands or millions of years to the institutions of existing savages might possibly have merits as a flight of imagination, but it could have none as a work of science. To do this would be exactly to reverse the proper mode of scientific procedure. It would be to work _a priori_ from the unknown to the known instead of _a posteriori_ from the known to the unknown. For we do know a good deal about the social state of the savages of to-day and yesterday,