I have been thinking a lot lately about a particular domain in the theory of knowledge called modal epistemology (or M.E.) the area that deals with knowledge claims about what is possible or necessary. My motivations for doing so relate to various arguments for and against the existence of God which contain premises what presuppose various modal claims. I won’t go into these arguments in this post, but I thought it pertinent to cover some basics first.
First, the rejection of Platonism (motivated by empiricism and a causal theory of knowledge) seems to be behind most skepticism about modal claims regarding possible worlds. If possible worlds are abstract (causally inert) entities, then it becomes difficult to see how they can interact with our cognitive faculties in such a way as to support beliefs about them. However, there are other non-platonic ways of understanding possible worlds along nominalist or conceptualist lines, which relieve these skeptical worries.
Second, the justification for modal claims needs to be evaluated contextually, on a case by case basis instead of being assessed according to necessary and sufficient conditions. This is because modal claims exist on a spectrum from being “everyday” (e.g. I could have purchased a different vehicle) to being “other worldly” (e.g. an exact physical replica of my body might have existed without being phenomenally conscious), and the conditions for knowing in one context need not be the same in another.
Third, just as there are multiple sources of justification for knowledge claims based on perception, memory, intuition, logic, etc, so philosophers have identified multiple sources that inform modal claims. The standard sources include (1) conceivability (whether personal relative or “ideal”), (2) imagination, (4) conceptual understanding, and (4) counterfactual reasoning. Most of the thought experiments employed by philosophers today appeal to some or all of these sources to justify their modal claims. Philosophers still debate about which sources should be privileged (and in what circumstances), and they still argue about how these sources afford justification (e.g. internally and/or externally), but these debates are similar to those about non-modal claims. They are nothing new and should not be used as a reason to throw up our hands in defeat.
Fourth, some person X can have adequate justification for a modal claim without knowing how or why it is justified. Joe has good reasons to believe that he could have chosen the career of a salesman instead of a painter, even if he is unaware of the grounds for his knowledge or of the epistemological theory that best explains it.
Conclusions? Modal claims can be warranted, but they are contextual, fallible, defeasible, involve multiple sources which the knower need not be aware of, and do not require a Platonist metaphysics.
 Notice that the ‘person relativity’ of modal intuitions, i.e. their being shaped by background beliefs, is not automatically a reason to reject them, since most knowledge of actual objects is relative to background beliefs in this way. The knowledge of actuality, possibility, and necessity all involve some kind of interpretive process in which subjectivity plays a role. This interpretive process operates in a circular (though not viciously circular way): firstly, we start with an initial understanding what is possible, then secondly, bring that understanding to the tribunal of thought experiments, counterfactual reasoning, analogies, and conceptual analysis, and finally we come to a sharpened, or refined, modal understanding. This interpretive process indicates that modal reasoning can be reliable, even if fallible and defeasible.