This thesis compares the epistemological positions open to exponents of the three varieties of modal metaphysics: modal realists, ersatz theorists, and deflationists. We seem to have modal knowledge, so any metaphysics of modality ought to be able to explain how we have that knowledge (or explain why we think we have it).
Lewisian modal realists suffer from an objection presented by Richards, which is similar in structure to one side of Benacerraf’s dilemma for mathematical platonism. Since modal realist worlds are, by definition, spatiotemporally disconnected from us, how can we know what is in them? Modal realism is shown to be internally consistent, epistemologically: various methodological criticisms of Lewis’s reply are dismissed, including criticisms of S5 as the logic of metaphysical possibility and necessity, on which Lewis’s reply depends. However, modal realism is also shown to be undesirably epistemologically mysterious. Suggestions for bridging the gap between our knowledge and realist objects are rejected. The parallel solutions in mathematics are discussed and found to be inapplicable to modal realism because of the concrete nature of the modal realist’s possible worlds, and an alternative sceptical epistemology is presented as the modal realist’s best option.
Ersatz theorists are also attacked by Richards because of the apparently infinite nature of knowledge of necessities, but a priori knowledge of necessities is discussed and accepted as coherent, and Richards’ objection to ersatz modal knowledge is thus dismissed. McGinn’s argument that an ersatzer must be a realist is accepted, and this realism is shown to be less problematic than Lewisian modal realism, as it can adapt mathematical solutions to the problems of knowledge of realist objects and create an epistemology using plenitude principles.
Lastly, two strong deflationist epistemologies are presented: one from Chalmers’ secondary conceivability, and one based on Peacocke’s theory of possible worlds. The classic conception of conceivability as a guide to possibility is, however, rejected.
It is concluded that on an epistemological basis, modal realism ought to be ruled out as the best explanation of the metaphysics of modality, but that ersatz and deflationist theories are on an equal footing in epistemological terms.