ABSTRACT
1. On many logical theories the word “probable” has a sort of dissolving effect. As soon as we have uttered it we are held to have removed ourselves from the region of exact results and rigid tests, and to have retired to the land of vague expectations, for which, according to some, no reason can be given, and, according to others, no reason need be given. We expect; we believe; our opinions or convictions have a certain strength; this strength has something to do with the number of instances we have observed, or perhaps with a hypothesis we have framed, but there need be no question of a test for anything so indefinite. And this treatment is extended to inductive reasoning. When difficulties are found in treating it as rigid demonstration, it is ticketed as “merely” probable, and then handed over to vagueness and indecision. From the probable to the fallacious is for some thinkers only a step, and so induction is converted, for example, into a kind of bad deduction. A flaw is supposed in its first principles, and yet it is held to give results, even perhaps good results. It is treated as a basis for the generalisations with which deductive reasoning starts,—a treatment which would make the greater part of deduction futile. Some thinkers, again, like the followers of Hume, while insisting with Bishop Butler that probability is the guide of human life, think it sufficient to point out that inductive conclusions are in fact brought about by certain alleged psychological laws without inquiring what basis, in truth or reason, this probability may have.
