ON THE NATURE OF THE NOTION OF EXTERNALITY. 213
- reasons ' for so regarding material bodies must be taken
rather as setting out implications of the phrase " material body ". For not all presentations (e.g. things seen in dreams and ' optical illusions ') are called material, but only such things as have to be affiliated to the external world. That the belief in the externality of an object is associated in a special manner with the sensation of touch, is only to be explained by the fact that, in practice, that sense is the best direct test of externality : the feeling does not contain the belief in the externality of that which is felt. At the same time, the test in question is one which is peculiarly ap- plicable to ' objects ' as such, rather than to changes in objects ; and this is probably the source of the prevalent errors concerning the nature of the notion of externality itself. Having once got hold of the notion of externality, it is not necessary to imagine the external world as being in the exact form of our perceptions, if by representing it in some other way we can more successfully achieve the purpose of introducing order into the chaos of sense. But in some form or other we must imagine it, if that purpose is to be fulfilled in the slightest degree. It may be whittled down to what is, by comparison with the fulness of per- ception, the veriest thread, but we cannot dispense with it altogether. To attempt to do so is as if a performer on the tight-rope, unduly exultant in his own prowess, were to essay to walk from bank to bank of the Niagara abyss on the insufficient air. This is what was meant by saying (footnote, p. 206) that " representative fictions " and in general, it may be added, the distinction between primary and secondary attributes of matter are essentially a refinement on the primitive distinction between the stream of consciousness and the external world. That " representative fictions," whatever else they may be, are not descriptions of the " routine of sense-impressions," is obvious from the fact that there is no such routine to be described. And that they do not describe our sense-impressions in any ordinary sense of the word, is obvious from the admitted fact that, e.g., light waves do not resemble our colour-sensations. To describe a thing is to describe it as it is, not as it is not. Briefly, no amount of description of the cause of any event will of itself convey an idea of the effect: and the 'facts' treated of in these theories are supposed to stand to the sensations (or the underlying neuroses) in the relation of cause to effect. Professor
