ABSTRACT
That the counterfactual analysis of causation is flawed is well known. Forty years after David Lewis lamented ‘It remains to be seen whether any regularity analysis can succeed … without piling on the epicycles’ (Lewis [1973] 1993: 194) and developed a counterfactual analysis to replace the then-standard account, there is plentiful evidence that the counterfactual approach is sharing the fate of its rival: it is piling on the epicycles. Every aspect of the original analysis has been shown to be subject to counterexamples, modifications proposed to fix it, but new counterexamples have emerged and new modifications proposed. 2
