3.2 David Hume: Concluding Remarks

What this lecture covers

This short lecture closes the discussion of Hume by looking at how philosophy and science responded to his skepticism. Millican explains Immanuel Kant's attempt to show that Hume must be wrong, then argues that later developments in science undermined Kant's confidence and vindicated Hume's more modest view of what science can achieve.

You'll come away able to state Kant's reasons for thinking some truths about the world are knowable with certainty, and explain why Millican thinks the two-slit experiment supports Hume's position over Kant's.

Key ideas

• Kant's response to Hume: Kant held that certain truths about the world are known with certainty, including universal causation, Euclidean geometry and Newtonian conservation laws.
• Kant's explanation: he argued our minds condition how the world appears to us, so the "phenomenal" world we experience must satisfy truths like Euclidean geometry a priori.
• Darwin's challenge: On the Origin of Species (1859) reinforced Hume's view that humans are part of nature rather than above it.
• Einstein's challenge: general relativity (1915) showed space is gravitationally curved, undermining the claim that Euclidean axioms are true of the world.
• Quantum mechanics' challenge: developments from around 1925 undermined the idea that the world is fully law-governed and intelligible in Kant's sense.
• The two-slit experiment: individual photons fired at two slits form an interference pattern, but the pattern disappears if detectors reveal which slit each photon passed through.
• Codifying, not explaining: mathematics can describe how the two-slit result occurs but not why, which Millican treats as a modern vindication of Hume's view that science gives systematic description rather than understanding.

Walkthrough

Kant's answer to Hume (0:10)

Millican introduces Kant as starting from the premise that Hume's skeptical conclusions must be false, since Kant believed we know with certainty that the world obeys universal causation, that Euclidean geometry is true of the world, and that principles like conservation of momentum can be known a priori. Kant's theory held that the mind itself shapes the world as it appears to us, guaranteeing these truths.

Darwin, Einstein and the collapse of certainty (2:12)

The lecture traces how later science undercut Kant's premises: Darwin placed humans firmly within nature, and Einstein's general relativity showed that space is curved, meaning Euclidean axioms are not straightforwardly true of the physical world and are not knowable a priori as Kant claimed.

The two-slit experiment (4:13)

Using a simulation, Millican demonstrates the classic two-slit experiment. Light passing through two slits produces an interference pattern, as if it behaves as a wave. Firing single photons with detectors at each slit removes the pattern, but removing the detectors—while still firing photons one at a time—brings the interference pattern back, even though no photon can be observed going through both slits at once.

Hume vindicated (8:17)

Millican argues that while quantum mechanics lets us calculate what happens mathematically, it offers no explanation of why detecting which slit a photon uses changes the outcome. He presents this as confirming Hume's claim that science can only codify regularities, not deliver ultimate understanding, in contrast to the confidence Newton's physics once seemed to justify.

Before you watch

• Watch the previous lecture on Hume's philosophy of science and the problem of induction, since this lecture directly continues that discussion.
• No prior physics knowledge is assumed for the two-slit example, but a rough sense of "wave interference" helps.

Check your understanding

1. What three kinds of truths did Kant believe we could know about the world with absolute certainty?
2. How did Kant explain why the phenomenal world must conform to Euclidean geometry?
3. Why does Millican see Darwin's and Einstein's work as undermining Kant's premises?
4. What makes the two-slit experiment's detector result so puzzling, according to the lecture?