“Counterfactual triviality: A Lewis-impossibility proof for counterfactuals” Philosophy and Phenomenological Research (2012).
Lewis’s famously argued that the obvious formulation of the “Ramsey test” connecting belief in conditionals to conditional belief, trivialized. I show analogous results hold for counterfactual conditionals and a corresponding counterfactual Ramsey test, and identify the parallels and potential disanalogies.
Counterfactual desire as belief (recent draft)
Intimately tied to Lewis’s triviality result for conditionals was his putative reductio of the “desire as belief” thesis—the idea that degree of belief about the goodness of an option were (ideally) tied to levels of desire for that option. Lewis framed the result using evidential decision theory as a fix on rational desirability, and it’s been argued that he should have used the causal decision theory he himself argued was the appropriate model of rational action; and furthermore, the triviality results do not apply to this setting. I argue that this is only so for some formulations of causal decision theory, which are independently objectionable. On some not-so-objectionable causal decision theories, triviality is regained.
Suppositions and decisions (in preparation, with Daniel Elstein)
Daniel and I look carefully at the kind of theory of rationality offered in standard (among philosophers) formulations of decision theory. We distinguish between the practical rationality that links certain suppositional beliefs to desires/actions, and the theoretical rationality that tells us what the rational constraints are between those suppositional beliefs and our wider mental state. We further identify a role for beliefs in conditionals as reporting the key suppositional states, thus forging a link between practical rationality and ordinary modes of practical reasoning. Under the assumption of perfect rationality, we may ignore these distinctions, since all three characterizations march in step. For imperfect agents, they can come apart. And we identify two families notions: indicative supposition, tied to ratios of categorical beliefs and reported by indicative conditionals; and counterfactual supposition, tied to expectations of conditional chance and reported by counterfactual conditionals. We note that in both cases, triviality results threaten the proper reporting of suppositional states of mind by conditionals.
An argument for conjunction conditionalization (recent draft, with Lee Walters).
If a conjunction A&B is true at a world, should the corresponding conditional “if A then B” be true there too? You might think not, intuitively. Lee and I measure the costs of dropping this principle.
“Defending conditional excluded middle” Nous 44(4): 650-668 (August 2010). Penultimate Draft
I present arguments that Conditional Excluded Middle is a costly principle to drop from one’s logic of conditionals. I consider Bennett’s arguments against it, based on its interaction with epistemic modality, and find them wanting.
“Vagueness, conditionals and probability” in Erkenntnis 70(2): 151-171 (March 2009). Penultimate Draft
In his defense of conditional excluded middle, Stalnaker argued that the conditionals Lewis regards as false, are in fact indeterminate. He also maintains that intermediate belief in conditionals such as “If I flipped this fair coin, it’d land heads” is appropriate. I note that these theses will clash on several understandings of the cognitive role of indeterminacy.
Tenable conditionals (recent draft)
I construct an account of indicative/counterfactual conditionals intended to systematically bypass Lewisian triviality accounts. It has the result that almost all conditionals are indeterminate; but I argue (picking up themes from the Erkenntnis paper above) that this should not pose an obstacle to substantive characterizations of the appropriate degree of belief to have in them.
“Chances, counterfactuals and similarity”; Philosophy and Phenomenological Research, vol 77(2), 2008. Penultimate Draft
Lewis’s account of the worldly conditionals that make-true a counterfactual were designed for deterministic worlds. Hawthorne has presented arguments that his attempted patch misfires. I respond to Hawthorne’s arguments, and present an improved Lewis-style treatment of chancy counterfactuals, based on identifying “fit with chancy laws” on a notion of “being typical according to those laws”.
“Counterfactuals and chance, redux” Analytic Philosophy (forthcoming).
I respond to Dylan Dodd’s interesting objections to the account developed in the PPR paper above—he points out some counterintuitive predictions that argue to be simply a reflection of a presupposed classical treamtent of the vagueness of “typicality”.
“Conversation and conditionals”; Philosophical Studies 138(2): 211-223 (March, 2008). Penultimate draft.
I look at the way that Stalnakerian conversational dynamics play with the truth-conditions of conditionals. I offer explanations in these terms of two puzzles concerning indicative conditionals: the infelicity of reverse Sobel sequences, and the “Gibbard phenomenon”—apparent violations of conditional non-contradiction.