Does Carnap’s treatment of philosophical questions about existence, such as “Are there numbers?” and “Are there physical objects?”, depend on his analytic–synthetic distinction? If so, in what way? I answer these questions by clarifying, defending, and developing the reading of Carnap’s paper “Empiricism, Semantics, and Ontology” that W. V. Quine proposes, with little justification or explanation, in his paper “On Carnap’s Views on Ontology”. The primary methodological value of studying Quine’s reading of “Empiricism, Semantics, and Ontology” is that it prompts (...) us to look for, and helps us to see the significance of, passages by Carnap that reveal the logical foundations of his views on ontology. Guided in this way by Quine’s reading, I show that (1) in “Empiricism, Semantics, and Ontology” Carnap’s preferred treatment of philosophical questions relies on paraphrasing them so that their answers are immediately obvious elementary logical truths, and are therefore, by his standards, trivially analytic; and (2) in its most general form, Carnap’s treatment of philosophical questions about existence depends on his controversial view that the analytic truths of a language L may include sentences that are not elementary logical truths, but that are nevertheless, by Carnap’s standards, analytic-in-L simply because we have stipulated that they are to be among the “meaning postulates” of L. (shrink)