Event Causation

Abstract

Most of those who speak of causation as event causation, including latter-day Humeans, accept that it must include more than observed regularities in sequences of events. Observed regularities might support generalisations based on induction by simple enumeration, but this gives a precarious basis for predictions as to future sequences. A causal account claims to be an explanation; though an explanation is not only a prediction, it looks to prediction for its confirmation. If a sequence shows a regularity in how things happen now, it should also obtain in how they will be found to happen. Hume saw that we need inductive generalisation from a number of cases; if there is predictive confidence the latter may be a matter of expectation based on custom — we have found A’s to be followed by B’s, so we have a customary expectation that they will continue to be so followed. But they may not. Russell called the reliance on habit ‘animal induction’,¹ and gave the example of the chicken whose master comes every morning to feed it, encouraging this agreeable expectation, until the day comes when he wrings its neck, thereby disappointing its simple faith in the order of nature. Yet we do have a simple faith (‘Aber Glaube’ as Waismann said) that there are orders of causal sequence which are not just resemblances; and we need more than repetition for causal explanation.² Moreover, we need to be able to deal with counterfactual suppositions: something X which did not happen might have happened if something else, Y had happened. We cannot observe sequences of things which did not happen, but we can speculate about them and this gives us play.