Causation, further themes

1. Causal relata

Philosophers usually draw a distinction between type and token causation. A type causal claim such as ‘Smoking causes lung cancer’ describes causal relations between types of events, whereas a token causal claim such as ‘Smith’s smoking caused his lung cancer’ describes causal relations between particular events (of variable duration). Philosophers have focused their attention almost exclusively on token causal relations even though type causal relations are the central concern of the natural and social sciences.

One of the many controversies surrounding token causal relations concerns the nature of the events that are their relata. One proposal is that of Donald Davidson (1967), who argued that these events should be understood as unstructured particulars. Events are like particular objects, he argued, in having spatio-temporal locations and identity conditions; and they are like objects in being describable in many different ways (see Events). Famously, Davidson (1970) appealed to this last feature to explain how mental events can cause physical behaviour even though there are no true psychophysical laws: they can do so because, in addition to their mental descriptions, they have physical descriptions which allow them to fall under purely physical laws (see Anomalous monism).

However, many philosophers have objected that Davidson’s proposal fails to capture the idea that when a causal relation holds between its relata, it does so in virtue of properties that are in some way constitutive of those relata. In order to capture this idea, many philosophers have proposed property-based conceptions of causal relata. For example, Jaegwon Kim (1980) understands events that are the basic causal relata to be instantiations of properties by objects at certain times; D. H. Mellor (1995) takes causes and effects to be facts, which are the truth-makers of true contingent sentences (see Facts); Douglas Ehring (1997) thinks that the basic causal relata are tropes or the concrete instances of a property (see Particulars §4); and Judea Pearl (2000) and James Woodward (2003) take them to be values of quantitative variables. On any of these proposals, causes and effects are more fine-grained and numerous than Davidson’s unstructured particular events. There is some reason to think that commonplace causal statements are primarily concerned with such fine-grained causal relata (see Bennett 1988).