Why do we judge an event to be a cause? Holger Andreas and Mario Günther offer a fresh answer: causation is epochetic dependence along an inferential path where each step genuinely depends on the cause. The dependence is called epochetic because the effect must be inferable from its cause after suspending judgment on both. This theory matches everyday judgment in virtually all scenarios, including the classic problems of overdetermination, preemption, and omissions. The authors go on to solve the problems of spurious and simultaneous causation and thereby develop a reductive theory of causation in the spirit of Hume. The book will be useful for students and researchers in philosophy, computer science, cognitive science, and law. It is a must read for anyone who wants to keep up to date with research on one of philosophy's great problems.

‘The account developed in this deep and sophisticated book is a welcome addition to the pantheon of analyses of causation, one whose novelty and ingenuity will repay close study, as well as further development. Andreas and Günther's analysis unfolds in a thorough and systematic fashion, with detailed attention to the contemporary literature, guided throughout by an attractive vision of causation as essentially connected to inference. And a treatise that could easily have been dry is, thanks to their clean and elegant prose, anything but.’

Ned Hall - Harvard University

‘Andreas and Günther have been developing an exciting new theory of causation and present the final version here for the first time. Their presentation is clear and rigorous, presupposing a minimum of formalism. They connect their theory with belief revision and deductive inference, so the book will be of interest to logicians and epistemologists as well as those working on the metaphysics of causation.’

Christopher Hitchcock - California Institute of Technology

‘A long-standing difficult problem in philosophy has been to find a reductive definition of actual causation that does not presuppose any causal or modal concepts as primitive. Andreas and Günther propose a new reductive definition that gives an intuitively correct answer for a wide array of standard situations that serve as counterexamples to existing proposed definitions. Philosophers interested in causation will find this well worth reading.’

Peter Spirtes - Carnegie Mellon University