14. Reflections on Legality
85. The Primacy of Law
-We have seen that, on Bennett’s account, in analyzing a subjunctive conditional A > C the
closest A-worlds must be causally legal from the time of the fork onwards.
-This condition on similarity is necessary to capture many subjunctive conditionals we intuitively
regard as being true. In particular, this condition allows us to deal with those conditionals of the
form A > Big Difference. If the similarity relation was not constrained by the demands of causal
legality but was based on all-in similarity no conditional A > Big Difference, that we intuitively
regard as being true, would come out as such on our account, for any A-world in which the
consequent was false would be overall more similar to our world than any A-world in which the
consequent was true. However, we do think that many conditionals of the form A> Big
Difference are true- for example ‘If Stauffenberg had placed the bomb a foot to the right> Hitler
would have been killed’- and the reason that we think this is that we believe that with the
antecedent in place there exist causally sufficient conditions for the consequent to be the case.
-This latter point highlights a more general reason for the requirement of legality. Our
subjunctive conditionals, for the most part, concern events, where it is assumed that had a certain
event occurred, a different event would also have occurred at a different time, and in order to get
from one event to the occurrence of another we need, as Bennett puts it, “ ‘the cement of the
universe’, causation”.(pp. 222)
- Nicholas Rescher, in “”Belief-Contravening Suppositions”, gives an account of counterfactual
conditionals (he considers only subjunctive conditionals whose antecedents are false) that tries to
explain why we take certain counterfactual statements to be true, and others false.
-Bennett claims to have little sympathy for this account, and levels three objections against it. It
will be worth our time to briefly go over Rescher’s account, for I think it may provide some
further illumination for why we require that the closest worlds be ones that conform to our causal
-On Rescher’s account counterfactual conditionals fall into two categories:
2)Purely hypothetical conditionals
Nomological conditionals are ones whose consequent follows from the antecedent as a
matter of general law. So, to take an example of Rescher’s: “If Smith had eaten an ounce of
arsenic, he would have died.” This is true, according to Rescher, in virtue of a general law to the
effect that all people who eat an ounce of arsenic die.