The search for a systematization of inferences depending upon ‘it is known that p’ (Kp). The usual backbone of such systems, for example in the work of Hintikka, is the idea of epistemically possible worlds, or worlds left open by a piece of information. This enables ‘known’ and ‘unknown’ to be seen as analogous to the ‘necessary’ and ‘possible’ of modal logic, and similar questions arise, such as whether we should adopt the principle that Kp → KKp. Particular problems emerge in the interaction of K with names and quantifiers, since it induces intensional contexts: it may be known that the murderer is dangerous but not known that the postman is dangerous, although the postman is the murderer.