n the previous sections we argue that: (1) in a semantic setting, the only possible interpretation of a Ramsey conditional is as a dynamic-doxastic operator (encod- ing the agent’s belief-revision dispositions); (2) classical AGM postulates cannot be applied to an object-language containing doxastic operators (and hence, in partic- ular, to one containing Ramsey conditionals), unless the (Non-)Vacuity postulate is restricted (obtaining “Epistemic AGM”); (3) the only n atural semantics for the- ory revision operator T ∗ A for such (theories formulated in) languages containing doxastic operators is a “stat ic” one (encoded in our “correct definition” ab ove): interpreting T s ∗ A as the agent’s revised beliefs (after revision with A) about the original state of the world s (as it was before revision)