Page:Mind (New Series) Volume 6.djvu/370

 354 J. ELLIS MCTAGGAET : us to predicate any quality, which was possessed by any In- dividual, of every other Individual in the universe. It is evident that on this we can found, not only no ap- proach to certainty, but no probability or presumption, even of the lightest kind. If two Universals are never connected, except in our subjective expectations, then the presence of two Universals in one Individual can never give the least probability that one of them will be accompanied by the other in another Individual. If, on the other hand, two Uni- versals are really connected, then general laws have objective existence, and are not merely our inferences from particular cases. Our antagonist may, however, take up a more definitely sceptical position. He may admit that the inference which is made in Induction is perfectly unjustifiable from a logical point of view, that the conclusion is not made in the slightest degree probable by the premises. But he may say that he never put it forward as logically justifiable, but simply as actually existing. We have, he may say, a natural tendency to expect B to accompany A in one case, if we have seen that it accompanies A in another. We may not be able to justify this impulse, but we cannot deny the psychological fact that we have it. This position however involves a contradiction. For it denies the validity of general propositions by an argument of which general propositions are essential links. No general proposition is logically defensible this is itself a general proposition. If, therefore, we make it, by that very act we condemn it as logically indefensible. It is impossible to state this view without denying it, and the result would be, not merely the rejection of one species of knowledge, but complete and utter scepticism. And complete scepticism is in the same plight as self-contradiction. For if you assert that nothing is certain, you assert, among other things, that your denial of certainty is not certain. Besides these general considerations an argument ad homines may be addressed to those who assert the basis of inference to be an irrational impulse. They do, in point of fact, trust to it. And not only do they trust inferences when they have made them, but they take great trouble to put themselves in a position to make more. They conduct, or speak with approval of others who conduct, researches in physical science. They laboriously accumulate instances, and examine what general qualities are found in combination in them, for no other purpose, on their own showing, than that they may become the victims of an irresistible, though