¹ Some philosophers, of course, hold that we have a direct knowledge of the external world, which does not involve any inductive reasoning between our sense-impression and our knowledge of the reality of the external object. But various lines of evidence convince most of us that man has no such vast power of direct knowledge as this would involve.

² Of course the inductions involved in “seeing” an external object have long since become subconscious and instinctive, like the volitions which guide the fingers of a Paderewski. But this should not obscure the fact that our instinctive association of the sense-impression with the external object was acquired by inference from sense-data, and needs frequently to be checked in order to avoid error.