ABSTRACT
THE problem of induction was known to Aristotle; but the full extent of its difficulty first came to light in Hume. How can we make an inference from examined instances to an unexamined instance or to all instances examined and unexamined? We have found hitherto that brass has expanded when heated, but how can we infer that the next piece of brass to be examined will do so too or that all pieces of brass will? As Hume put it, how do we know that the sun will rise to-morrow, simply from our knowledge that it has done so regularly in the past? There is nothing contradictory in supposing that past regularities will suddenly not be repeated.
