It is sometimes alleged that arguments that probability functions should be countably additive show toomuch, and that theymotivate uncountable additivity as well. I show this is false by giving two naturally motivated arguments for countable additivity that do not motivate uncountable additivity.

Contemporary metaphysicians have been drawn to a certain attractive picture of the structure of the world. This picture consists in classical mereology, the priority of parts over wholes, and the well-foundedness of metaphysical priority. In this short note, I show that this combination of theses entails superatomism, which is a significant strengthening of mereological atomism. This commitment has been missed in the literature due to certain sorts of models of mereology being overlooked. But the entailment is an important one: we must either accept superatomism or reject one (or other) of the most widespread theses of contemporary metaphysics.

Probabilistic theories of ‘‘should” and ‘‘ought” face a predicament. At first blush, it seems that such theories must provide different lexical entries for the epistemic and the deontic interpretations of these modals. I show that there is a new style of premise semantics that can avoid this consequence in an attractively conservative way.