En Stuart, S. Dodig-Crnkovic, G. Computation, Information, Cognition. The Nexus and the Liminal. Introduction. (2007)

The word ‘cause’ and its cognates, such as ‘bring about’, ‘produce’, ‘make’, etc., are all common in our vernacular. In many cases we have no problems to determine whether we should use the word ‘cause’ or not; our intuitions are clear. But when we try to spell out our criteria it has proven all too easy to invent counterexamples. This is a philosophical challenge and an incentive for debate. Another reason for discussing causation is that in many sciences we are interested in causal relations and we urgently need a well-defined concept.

Disagreements about the nature of causation have been a perennial trait in philosophy. In recent years at least four quite different views are proposed. One is the conserved quantity theory according to which causation is transfer of a conserved quantity. Another is the agency theory according to which a cause is something we humans can manipulate; a third popular approach is to analyse causes in terms of counterfactuals and a fourth account is to define causes in terms of INUS-conditions. Each approach can muster strong intuitions and paradigmatic examples in favour of its view, but all have difficulties which become visible when generalising beyond paradigmatic examples.

In response to this, some have suggested an ecumenical approach, saying that causation should be analysed as a disjunction of all these accounts. One could then in each different case apply the analysis that fits best. But this sounds a bit ad hoc; what is the unifying principle? Why do we call all these apparently different things for cases of causation? Something stronger than mere disjunction is needed.

In this paper I will propose a more unified approach involving three of the abovementioned approaches, viz., Mackie’s analysis in terms of INUS-conditions, the agency theory and the conserved quantity theory. But why leave out counterfactual analysis?