Impossible worlds are here to stay

I address objections to impossible worlds (IWs) by Timothy Williamson and Kit Fine. Two species of IWs Mark Jago and I had in our Impossible Worlds book were FDE worlds (worlds used in the semantics of the nonclassical logic of First Degree Entailment) and open worlds (worlds not closed under any non-trivial logical consequence relation). Williamson attacks the idea that propositional contents are sets of open worlds; but we explicitly disavowed that very idea. He endorses uses of IWs we developed, except he calls these ‘pseudo-worlds’. In Angellic Content (AC), Fine champions a truthmaker semantics and in Constructing the Impossible advocates its superiority over IWs. But his semantics also includes states that are IWs, as characterized in the literature: representations of absolute impossibilities. They are like Barwise and Perry’s situations — which is how FDE worlds have been interpreted for decades. A key difference between AC and FDE is that only the latter validates ‘absorption principles’ one may find unwelcome in a characterization of propositional content. But a more topic-sensitive FDE can make them fail. Williamson has also objected to the idea of giving truth(making) and falsity(making) conditions separately, as in both AC and FDE. But FDE can work with truth conditions only and the compatibility semantics for negation. AC can work the same way.