Metaphysics

Hume on Causation

We never observe causation directly — just constant conjunction

David Hume (1748) argued we never observe causation directly. We see one event followed by another (constant conjunction); never a "necessary connection." Causation is inferred, not perceived. Three components of causation Hume identified: temporal precedence (cause before effect), contiguity (close in space), constant conjunction (regular pattern). Hume: causation is just habit of mind expecting effect after cause. Reduces causation to psychological projection. Influenced: Kant (response), modern philosophy of science. Different views since: regularity theory, counterfactual theory, mechanistic.

  • SourceHume, "Enquiry Concerning Human Understanding" (1748)
  • Three componentsTemporal precedence, contiguity, constant conjunction
  • Hume's viewNo necessary connection observed
  • ConclusionCausation is psychological habit
  • InfluenceKant's response; modern philosophy of science
  • Modern theoriesRegularity, counterfactual, mechanistic

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

Why Hume on causation matters

  • Philosophy of science. What science studies.
  • Metaphysics. Nature of causal relations.
  • Free will. Causation and choice.
  • Religion. Cosmological arguments.
  • Statistics. Correlation vs causation.
  • Critical thinking. Causal inference.
  • Modern philosophy. Foundation of empiricism.

Common misconceptions

  • Hume denied causation. Reduced to regularity, not denied.
  • Causation easy. Centuries of debate.
  • Same as correlation. Scientifically distinct.
  • Solved by science. Continues philosophical debate.
  • Just metaphysics. Practical implications too.
  • Hume's argument refuted. Influences continue.

Frequently asked questions

What's Hume's argument?

Three steps. (1) We have idea of causation involving necessary connection (when A causes B, B must follow A). (2) But: where does this idea come from? Empirically: we never observe necessary connection. We see fire (cause); we see warmth (effect). Don't see "the necessity." (3) Conclusion: necessary connection isn't from sensation but from habit/expectation built from constant conjunction. Causation is psychological, not in the world.

What are the three components?

Hume's analysis of what we DO observe. (1) Temporal precedence: cause comes before effect. (2) Contiguity: cause and effect are close in space and time. (3) Constant conjunction: when A occurs, B regularly follows. Together: build our notion of causation. But: no necessary connection — just regularity.

Why is this radical?

Conventional view: causation is real feature of world; necessary connection holds. Hume's view: causation is projection of mind. We see regularity; expect it to continue (induction); call this "causation." Implication: science about causation may be psychological, not metaphysical. Affects: free will, scientific knowledge, religious arguments using causation.

What's the modern regularity theory?

Direct descendant of Hume. Causation = constant conjunction. A causes B if A's occurrences are regularly followed by B's. Problem: spurious correlations. Day causing night? Same regularity. Need additional constraints. Modern: causation usually requires more than just regularity; some sort of mechanism or law-like connection.

What's the counterfactual theory?

David Lewis (1973). A causes B if: had A not occurred, B would not have occurred. Implicit comparison with possible worlds. Captures: dependence relation. Examples: throwing rock at window caused break — without throw, no break. Problem: pre-emption (if you didn't break it, someone else would have). Modern philosophy of causation: lots of variants on this theme.

How did Kant respond?

Immanuel Kant (1781): "It was Hume who awakened me from dogmatic slumber." Kant: causation is necessary feature of how we think about world. Not empirical claim about world; categorical structure of experience. We can't experience world without using causal categories. Synthetic a priori. Modern Kantianism continues this line.

Is causation real?

Major debate. (1) Realists: yes, causation is real feature of world. (2) Hume: psychological projection. (3) Pragmatic: useful concept; ontology unclear. (4) Russell (1913): science doesn't use cause-effect; uses functional relations. (5) Modern interventionists (Woodward): cause = what manipulating changes effect. Various positions; Hume's challenge remains influential.