A general possible worlds framework for reasoning about knowledge and belief (Q1814181): Difference between revisions

It is shown that there is a unifying semantic framework --- Rantala's non-normal worlds semantics --- for various model theories of knowledge and belief which successfully avoid particular aspects of logical omniscience. The basic intuition behind the non-normal worlds semantics, intuition of compatibility, is expressed by the author as follows: ``agent \(i\) knows at world \(s\) that \(F\) iff \(F\) is true at all \(i\)'s epistemic alternatives from \(s\) (true at all worlds compatible with what \(i\) knows at \(s\))''. Non-normal worlds semantics is compared to the model theories of logics of knowledge and belief of Levesque, Fagin and Halpern, van der Hoek and Meyer. The results of the comparison: Every nontrivial structure for implicit and explicit belief (of Levesque) induces a Rantala model validating precisely the same formulas. Analogous claims are proven for Kripke structures for awareness (general awareness, local reasoning of Fagin and Halpern) and for Kripke structures for awareness and principles (of van der Hoek and Meyer): they all induce a Rantala model validating precisely the same formulas.