Heim Sequences and Why Most Unqualified ‘Would’-Counterfactuals Are Not True

Abstract

The apparent consistency of Sobel sequences (example below) famously motivated David Lewis to defend a variably strict conditional semantics for counterfactuals. (a) If Sophie had gone to the parade she would have seen Pedro. (b) If Sophie had gone to the parade and been stuck behind someone tall she would not have seen Pedro. But if the order of the counterfactuals in a Sobel sequence is reversed-in the example, if (b) is asserted prior to (a)-the second counterfactual asserted no longer rings true. This is the Heim sequence problem. That the order of assertion makes this difference is surprising on the variably strict account. Some argue that this is reason to reject the Lewis-Stalnaker semantics outright. Others argue that the problem motivates a contextualist rendering of counterfactuals. Still others maintain that the explanation for the phenomenon is merely pragmatic. I argue that none of these are right, and defend a novel way to understand the phenomenon. My proposal avoids the problems faced by the alternative analyses and enjoys independent support. There is, however, a difficulty for my view: it entails that many ordinarily-accepted counterfactuals are not true. I argue that this (apparent) cost is acceptable.

Should we be skeptics or contextualists about counterfactual conditionals?

“…Once such reasons are made explicit through (2), we recognize (1)'s falsity.7 The following is similar to an argument made inLoewenstein (2021). For a different kind of retraction argument against contextualism, see Hájek (2020).8 For some different objections to antecedent-contextualism, as well as some objections aimed specifically at Sandgren and Steele's account, seeHájek (2020).9 See alsoIchikawa (2011), for an alternative connective-contextualist account.10 For additional objections to Lewis's contextualism seeHájek (2020) andLoewenstein (2021).11 Of course, the speaker who drops the 'probably' here is not thereby saying something false, whereas the skeptic thinks that the person who drops the 'probably' in the consequent of the counterfactual, is. But this difference is not a problem for the argument that speakers drop the 'probably' for the same reasons in both cases if speakers don't realize that there is this difference.…”

“…Once such reasons are made explicit through (2), we recognize (1)'s falsity. 7 The following is similar to an argument made in Loewenstein (2021). For a different kind of retraction argument against contextualism, see Hájek (2020).…”

“…10 For additional objections to Lewis's contextualism see Hájek (2020) and Loewenstein (2021). 11 Of course, the speaker who drops the 'probably' here is not thereby saying something false, whereas the skeptic thinks that the person who drops the 'probably' in the consequent of the counterfactual, is.…”

Counterfactual skepticism is (just) skepticism

“…Loewenstein (2021a) gives a different argument based on Sobel sequences. I believe her position is also targeted by the kinds of arguments I give below.4 Thanks to an anonymous referee for pushing me to say more here.…”