Experience, causation, and induction › Where necessary connection enters

Section VII asks a more specific question. What supplies our idea that a cause and an effect are necessarily connected? Watching one billiard ball strike another gives a sequence of events, Hume argues, rather than an impression of an intrinsic tie that makes any other result impossible. Repeating the sequence does not insert a newly visible tie between the balls. Something changes in the observer: the first event now brings an expectation of the second. E 7.6 and 7.28–30.

That customary transition supplies the impression from which Hume explains the idea of necessary connection. His account distinguishes the observed conjunction from the felt determination of thought. It does not require us to deny that scientists can observe forces through their effects or construct successful explanatory models. The disputed issue is access to a necessity that makes the effect follow from the cause through its very nature.

In E 7.29, Hume offers two definitions of cause. One describes an object followed by another, with similar objects followed by similar results. The other describes an object whose appearance carries thought to the other. The first emphasizes regularity; the second emphasizes inference. He also introduces a formulation concerning what would have happened had the first object not existed. His compressed presentation leaves questions about how these formulations fit together. It should not be advertised as a complete modern account of causal explanation. Section VII.

A difficulty remains even before those interpretive questions. Reliable sequences can occur without the first event producing the second. Imagine a station clock chiming shortly before a train arrives because both follow a schedule. The chime alone does not propel the train. This invented example raises a challenge for any simple equation of causation with regular succession. Hume's central achievement here is an analysis of what experience and mental transition contribute to our causal ideas; that achievement does not settle every question about distinguishing causes from common effects.