Principles of World Theory say, more or less, that: - a proposition (whatever) is *necessary* iff it is true in all worlds; - a proposition (whatever) is *possible* iff there is some world in which it is true; - there is only one *actual* world; - there are propositions which are true at the *actual* world; - there are propositions which are not true at the *actual* world, but they are true at some *non-actual* *possible* world.
It is not much. But, in any case, we must start from these points :-)
Very nice. Except perhaps that it is the principle of the Old World Theory, implicit in Aristotle and Leibniz, where all the worlds are accessible from each other. It is formalised by the modal logic S5. Abbreviating "necessary p" by box p, or p, and "possible p" by <>p (which itself can be seen as an abbreviation of "not box not p", i.e. - - p), its main axiom are p -> p, p->p, and <>p -> <>p. Search S5, in the archive, I have given the precise axioms and inference rules, at some time. Kripke has relativized such sort of modal logic by saying that p is true in a world A if p is true in all world accessible from A. Then for each choice of an accessibility relation, you get a different modal logic. And modal logic is the best tool for being precise on the invariant truth (like laws) in a multiverse, or in contextual frame, etc. Also, modal logics can be used to simulate in the classical settings, non classical logics, like intutionistic logic, quantum logic. I agree with we should, well perhaps not start from that, but invoking them when we disagree about the validity of an argument in *apparently* fuzzy context. I work mainly with the modal logic G and G* (see the archive). This can help for giving axiomatic notion of self-identities.