Timothy Williamson (ANU): 'Causation, Determinism, and Decision' (co-authored with Alex Sandgren)

Many philosophers believe that Causal Decision Theory (CDT) is the correct normative theory for agents whose choices provide them with evidence about the causal structure of the world. There has been growing pressure to revise this belief. One of the most serious causes of concern is that CDT is supposed to go badly wrong in deterministic cases (that is, cases in which certain states in the decision situation are incompatible with certain acts being performed). Indeed, it looks as if the very reasoning that yields the verdict that you should Two-Box in Newcomb’s Problem commits the causalist to absurd courses of action in deterministic cases. For example, CDT appears to say that you should bet on the proposition that God has foreordained that you will never bet. Arif Ahmed has recently argued that there is no way for the defender of CDT to handle deterministic cases, and so CDT should be rejected.

We think that CDT can deliver the correct recommendations in deterministic cases. Or, more carefully, a modest generalisation of David Lewis’ CDT can deliver the correct recommendations in deterministic cases. We diagnose the fault in CDT as arising from a misuse of counterfactuals; standard CDT wrongly requires agents to deliberate counterfactually about outcomes that are certainly non-actual. Rather than reject counterfactual deliberation in general, however, we claim that deterministic cases show that agents must be careful about which counterfactuals they take as relevant for the purposes of practical deliberation. In particular, agents should give no deliberative weight to outcomes that could only be brought about by breaking the laws of nature (or logic, or metaphysics). This policy 1) is compatible with an independently plausible form of compatibilism, 2) respects the key motivation behind CDT that difference-making matters to rational decision, 3) can readily be formalised within the Lewisian framework for CDT, and 4) yields plausible verdicts in a range of decision problems, including those raised by Ahmed. CDT remains the correct normative theory, even for determinists.

Date & time

Tue 09 Apr 2019, 3:30pm to 5:30pm

Location

Coombs Ext Rm 1.04

Speakers

Tim Williamson (ANU)

Event series

Contacts

School of Philosophy

SHARE

Updated:  8 April 2019/Responsible Officer:  Head of School/Page Contact:  CASS Marketing & Communications