David Hume · Philosophy
Hume asks what justifies our reasoning from past to future, from observed to unobserved. His answer is unsettling: nothing in reason can justify it. Inductive inference rests on an assumption that cannot be proved.
We are constantly reasoning from what we have experienced to what we have not. The sun has risen every morning, so we expect it tomorrow. Fire has always burned, so we expect this fire to burn. Bread has always nourished us, so we expect this loaf to nourish too. This kind of reasoning - from observed cases to unobserved ones, from the past to the future - is called induction, and it is the basis of all our practical knowledge and all of empirical science. Every prediction, every expectation, every scientific law generalised from experiments depends on it. Hume now asks a simple but devastating question about this universal habit of mind: what justifies it? When I infer that the future will resemble the past - that the sun will rise, that fire will burn - what is the reasoning that licenses this inference? On what ground do I move from ‘it has always been so’ to ‘it will be so’? It seems the most obvious and reasonable thing in the world. But Hume, examining it closely, finds that it cannot be justified by reason at all - and this is the ‘problem of induction’, perhaps the single most famous problem in the theory of knowledge.
Hume’s argument is a dilemma built on his fork (relations of ideas vs. matters of fact). How could the uniformity principle - that the future will resemble the past - be justified? There are only two kinds of reasoning. The first is demonstrative reasoning (relations of ideas, like mathematics), which establishes necessary truths whose denial is a contradiction. But the uniformity principle cannot be established this way, because its denial is not a contradiction: there is nothing logically impossible in supposing that the course of nature changes, that the sun fails to rise, that fire ceases to burn. We can clearly conceive these things; they are not self-contradictory. So no demonstration can prove that nature is uniform. The second kind of reasoning is probable or experimental reasoning (matters of fact) - reasoning from experience. But here Hume springs the trap: any attempt to justify the uniformity principle from experience is circular. For all that experience can show is that nature has been uniform so far, in the cases we have observed. To get from ‘nature has been uniform in observed cases’ to ‘nature is uniform (including unobserved cases)’ - that is, to project past uniformity onto the future - you must already assume that the future will resemble the past. But that is the very principle you were trying to prove. So the argument from experience assumes the uniformity principle in order to establish it, and is hopelessly circular.
The problem of induction looks like a recipe for total skepticism - if none of our reasoning about the unobserved world can be justified, then all of science and all our practical expectations are groundless. But Hume’s response is characteristically subtle, and it is not despair. The skeptical argument is, he thinks, unanswerable as an argument: reason genuinely cannot justify induction. Yet this skeptical conclusion is one we cannot actually live by or even believe for long. Nature does not wait for reason’s permission; the moment we step out of the study, custom takes over and we expect, predict, and act exactly as before. “Nature, by an absolute and uncontrollable necessity, has determined us to judge as well as to breathe and feel.” The doubt is real and the argument valid, but human nature overrides it: we cannot help reasoning inductively, any more than we can help breathing. So Hume’s conclusion is not that we should stop trusting induction - we cannot, and we should not - but that we should understand its true basis. Our confidence in the uniformity of nature is not a deliverance of reason but a gift (or compulsion) of instinct. This is the mitigated skepticism again: the argument humbles reason and corrects our pretensions to rational certainty, while leaving our actual beliefs and practices, grounded in nature, exactly where they were - only now understood for what they are.
This is the opening of the lesson. The rest — the dialogue, the primary source, and the recall — is in the app.
You learned that Hume showed inductive reasoning (from past to future, observed to unobserved) cannot be justified by reason, since it rests on the unprovable assumption that nature is uniform. Explain the problem and why it is so troubling.
Leads to Bertrand Russell.
Begin this lesson →epoché — a humanities education that remembers you.