Induction is the process of reasoning from one set of observations to conclusions about things that haven’t been observed. We use induction when we make predictions about the future or offer explanations about the unobserved events that led up to the present. Hume’s problem of induction seems to show that we have no reason to draw these inferences.
Hume observes that we move from observations to conclusions about things that are not observed. But why are observations about how things worked in the past relevant to how they will behave in the future? Hume argues that we take for granted what he variously calls a “connecting proposition” or “medium” that I have called the uniformity principle.
You have plenty of reasons to believe what you have observed: you saw it with your own eyes. So far so good. But you will only have reasons to believe the conclusion if you have reasons to believe the uniformity principle. That’s where the trouble lies.
We cannot tell that it is true on the basis of the relation among ideas. This is because it is conceivable that nature might not be uniform. For instance, you could imagine the the laws of gravitation change so that we would be like astronauts in space even down here on earth. There is nothing wrong with the combination of ideas you would have in imagining that. By contrast, you cannot imagine that 3 ≠ 3.
But, you say, it isn’t really possible. After all, we have good reason to believe that the laws of nature never change. ‘What is that good reason?’, Hume might ask. If you say ‘because the laws of nature have never changed in the past,’ you’re in trouble.
What you would be saying would be:
And Hume will ask: ‘Why are your past observations relevant to how things will be in the future?’
You had better not answer: ‘Because the course of nature will continue to be the same.’ That’s what you’re trying to prove. But what else can you say?
That’s the problem of induction. In order for observations of the past to give you any reason to believe something about the future, you have to have reason to believe in the uniformity principle. But you do not have any reason to believe in the uniformity principle.