Causal Relevance and heterogeneity of Program Explanations in the face of Explanatory Exclusion

Abstract

n everyday causal explanations of human behaviour, known generally as ‘folk psychology,’ the causal powers of the mental seem to be taken for granted. Mental properties such as perceptions, beliefs, and desires, are all called upon in causal explanations of events that are deemed intentional. Jaegwon Kim’s exclusion principle has led him to deny mental properties causal efficacy unless they are metaphysically reduced to physical properties, but what of their causal relevance? By giving up the assumption of causally efficacious mental properties, has Kim put into question the explanatory value of explanations with mental descriptions? In other words, if a lower order neurological causal explanation involving a causally efficacious property is at hand, does it make the higher order mental explanation irrelevant and therefore redundant? If we are to save the explanatory importance of higher order predicates, and thus the causal explanations of the special sciences and folk psychology, we need an account of how such properties can be relevant as opposed to irrelevant in causal explanations, even though they may not be causally efficacious. Frank Jackson’s and Philip Pettit’s notion of program explanation tries to do just this. There are two ways to read Kim’s causal exclusion principle, metaphysically as causal exclusion, or epistemologically as causal explanatory exclusion. The explanatory exclusion principle hinders the acceptance of two causal explanations for a single effect unless an acceptable relation exists between the two purported causes. In this paper I will concentrate on the epistemological version of the principle, including Kim’s acceptable relations, and contrast it with a survey of Jackson’s and Pettit’s program explanations. My aim is to discover whether program explanations expound an acceptable relation already covered by Kim or present an additional way of understanding how two causal explanations of the same phenomenon can be accepted. Jackson and Pettit say: “The notion of a programming property does not just explain how an inefficacious property can be relevant to the causation of an event. It also shows how a program explanation can have a significance that remains in the presence of an explanation invoking the corresponding I