Causation and Probability-Raising
ICREA / Universitat Autònoma de Barcelona
Since the 1970s, one popular approach to analyzing causation has proposed that the cause-effect relation is essentially one of objective probability-raising. A generic cause C of effect E is something whose presence or occurrence raises the probability of E-type events; a singular cause c of effect event e is an event whose occurrence made e's occurrence more probable than it would have been had c not occurred. Though specific philosophical theories of causation based on these ideas fall prey to important counterexamples (as all theories of causation do), nevertheless the idea that causes raise the probabilities of their effects remains popular and influential.
I will discuss two broad reasons for thinking that objective probability-raising is at most a part of the story of the relationships between causation and probability: 1) There may not be enough objective probabilities to ground all the causal relations we wish to capture. 2) If we ensure that there are enough objective probabilities to cover all causal phenomena—by postulating probabilities that arise out of physical laws and cover all phenomena—then (a) we commit ourselves to a quite strong form of reductionism, and (b) we can see reasons to expect that sometimes the physically grounded probabilities will entail the (intuitively) wrong causal conclusions.
Nevertheless, it seems to me that there is something deeply right about supposing that causation and probability-raising are connected. But I propose that we locate the connection on the epistemic side: knowing about a cause for an effect raises one's subjective probability of the effect's occurrence, ceteris paribus. This connection can be captured in the form of a Cause-Probability Principle (CPP) that bears some resemblance to the so-called Principal Principle connecting subjective probability with objective chance. I will briefly discuss possible justifications for CPP, and possible objections.
Thursday, April 11th