In this letter David Lewis argues against Reinhardt Grossmann’s claim that logic is about states of affairs (or facts), and not about sentences or propositions. It is an interesting example of Lewis’s opposition to factualist ontologies and his steadfast commitment to Goodman’s principle of nominalism: no two entities can be composed of the same parts.

The crux of the disagreement turns on the constituency relation that binds (thin) particulars and universals together to form states of affairs. As Lewis had argued, in print against D.M. Armstrong, the constituency relation is not the parthood relation. It is not even somewhat mereological, as if a constituent of a state of affairs were a sort of part. Lewis’s decree raises an interesting question: what really makes a relation mereological? what really makes a relation a parthood relation?