ABSTRACT

The semantic analysis of causal statements presented in the last chapter has

led us to postulate two relations, irreducible to one another – the relation of

causal responsibility and the relation of causation between events: the

former makes true factive causal statements whereas the latter makes true

eventive causal statements. We have seen that the relation of causation

between events is more fundamental in the sense of being an essential part

of the former, whereas factive causal statements are often more informative

and more directly relevant for causal explanations. Relations between events are what makes true eventive causal statements. An event is constituted by

the set of all properties which are entirely determined within its spatio-

temporal zone. Each event is involved in a network of singular causal pro-

cesses. But to understand these processes, and to explain and manipulate

them, it is necessary to know their nomologically relevant properties.

Therefore the immediate object of a causal explanation is a relation between

a fact about the cause and a fact about the event, where the former is said

to be causally responsible for the latter. In order to elaborate our notion of causal responsibility, it will be useful

to compare our analysis with Davidson’s (1967). Recall Davidson’s solution

to the problem discussed in the preceding chapter, of reconciling two

requirements: insofar as they express a relation between particular entities,

causal statements must be extensional. However, a certain number of causal

statements do not seem to fit in this framework because they seem to violate

one criterion of extensionality. Here is how Davidson resolves this conflict:

on the one hand, he does justice to the first requirement by postulating that ‘‘the relation of causality between events can be expressed (no matter how

‘strong’ or ‘weak’ it is) by an ordinary two-place predicate in an ordinary,

extensional first order language’’ (Davidson 1967, p. 161). On the other

hand, he takes account of the observation that some causal statements are

not extensional by saying that ‘‘such sentences tell, or suggest, a causal

story. They are, in other words, rudimentary causal explanations’’ (David-

son 1967, p. 161). Here Davidson seems to join those who reduce causation

to explainability or predictability. We have seen Carnap (1966) and Stegmu¨ller (1969) defend this thesis.1 However, the apparent similarity of their