Edinburgh, 1711–1776
David Hume
Hume asked a question so simple that nobody had thought to ask it: what entitles us to expect that the future will resemble the past? His answer was that nothing does. The question is still called Hume's problem.
The view
Hume published A Treatise of Human Nature before he was thirty and complained that it fell dead-born from the press. He restated its central arguments more briefly in An Enquiry Concerning Human Understanding (1748), and those are the pages that changed philosophy.
He begins as a follower of Locke. Everything in the mind is either an impression, a vivid experience, or an idea, a faint copy of one. Everything we can reason about is either a relation of ideas, such as the truths of arithmetic, or a matter of fact. Relations of ideas can be proved, because denying them is a contradiction. Matters of fact cannot.
That the sun will not rise to-morrow is no less intelligible a proposition, and implies no more contradiction, than the affirmation, that it will rise. Enquiry IV
The argument
All our beliefs about what we have not observed rest on cause and effect. And we know causes only from experience: nobody could work out by pure thought that bread nourishes or that fire burns. But experience tells us only what has happened. To go from "bread has always nourished me" to "this bread will nourish me" we need a further premise, that nature is uniform and the future will be like the past.
What supports that premise? It is not a relation of ideas, since a change in the course of nature is perfectly conceivable. So it must be a matter of fact, known from experience. But every argument from experience already assumes it. The reasoning runs in a circle.
Hume does not conclude that we should stop expecting the sun to rise. We cannot stop. After seeing two things go together often enough, the mind passes from one to the other by custom or habit, which he calls the great guide of human life. Belief in causes is a fact about our psychology, like hunger. It is not a conclusion of reason.
The chicken is Bertrand Russell's illustration of Hume, from 1912. Its evidence could not be better, and its confidence is highest on the last morning.
Where he meets Deutsch
Hume's logical point is exact, and Deutsch accepts every word of it. No number of observed cases entails anything about an unobserved one. There is no valid form of inference called induction, and no principle of the uniformity of nature that could be established without assuming itself. The philosophers who spent two centuries trying to answer Hume were trying to do something that cannot be done.
Hume also saw that the conclusion threatens science as well as metaphysics, and he did not look away.
Where they part
Hume destroyed induction as logic and kept it as psychology. He still assumes that we get our expectations by repetition: the bucket fills, and habit does what reason cannot. So knowledge survives only as a kind of animal instinct, and reason is left with nothing to do.
Popper's reply, which Deutsch takes over, is that Hume's psychology is as wrong as the logic he refuted. We do not acquire theories from repeated instances at all. We guess first and look afterwards. Nothing is being inferred from the past, so there is no step for Hume's argument to block, and nothing irrational in holding a theory that has survived our best attempts to refute it.
Deutsch adds the chicken's lesson. The bird went wrong by extrapolating without explaining. Had it guessed why the farmer was feeding it, the same observations would have pointed the other way. We expect the sun to rise tomorrow because we have a good explanation of what the Sun is and how the Earth turns, and the same explanation tells us that one day, billions of years from now, it will not.
| Hume | Deutsch | |
|---|---|---|
| Where does knowledge come from? | Experience, extended beyond itself by custom. | Conjecture: guesses, improved by criticism and experiment. |
| What does justification do? | It is required and cannot be had. Induction rests on habit, not reason. | Nothing. No theory is ever justified, and none needs to be. |
| Can knowledge grow without limit? | No. Reason cannot carry us beyond what experience has shown. | Yes. Problems are inevitable, and problems are soluble. |
| What is knowledge, physically? | Impressions, and the fainter ideas copied from them, linked by association. | Information that, once embodied, tends to keep itself in existence. |
Further reading
- David Hume, An Enquiry Concerning Human Understanding (1748), sections IV and V.
- Bertrand Russell, The Problems of Philosophy (1912), chapter 6, for the chicken.
- Karl Popper, "Conjectural Knowledge: My Solution of the Problem of Induction", chapter 1 of Objective Knowledge (1972).
- David Deutsch, The Fabric of Reality (1997), chapters 3 and 7.
- The Problem of Induction, Stanford Encyclopedia of Philosophy.