[The concluding summary of my honours thesis. Links to the earlier chapters are interspersed with the text...]
It’s natural to expect that what can be known without needing to look at the world is closely tied to how the world metaphysically could or must have been. If we can only learn a fact a posteriori, through empirical investigation, we may expect that this is because there are other possible worlds in which the fact in question fails to hold. Assuming that possible worlds are wholly self-contained, we would not expect that examining the actual world could tell us anything informative about other, non-actual possibilities. Modal rationalism draws on these intuitive ideas by positing an intimate link between apriority and necessity, according to which an ideally rational agent could in principle grasp modal space – or apprehend what is possible and what is not – through the exercise of reason alone.
Kripke’s discovery of the necessary a posteriori casts doubt on this picture. There are some necessary truths – e.g. ‘water is H2O’ – which can only be known after empirical investigation. But the modal rationalist suggests that the problem here is merely semantic. We can know a priori how all the various possible worlds are in themselves; what we don’t always know is how to apply our words to them. Some terms, like ‘water’, are not semantically neutral – their application to counterfactual worlds is contingent on how the actual world turns out. That’s why empirical inquiry may be required before we can accurately assess various modal claims. The extra work is required to grant us semantic, not metaphysical, knowledge. We may avoid this need by restating a claim in neutral terms, for which the semantic values are unaffected by whether we consider a world “as actual” or “as counterfactual”. Chapter One thus established that the Kripkean challenge to modal rationalism is toothless after all; the link between apriority and necessity may be restored by restriction to semantically neutral vocabulary.
What’s needed to refute modal rationalism are “strong necessities”, i.e. claims that are true in all worlds considered as actual, despite being conceivably false. This requires that there be coherent scenarios that would not be verified by any possible world. Chapter Two explored this idea further, and assessed Yablo’s arguments for the claim that modal rationalists must recognize such strong necessities. Arguments from meta-modal conceivability provide the greatest challenge here, but I proposed that modal rationalists should respond by treating scenarios as epistemically fundamental, so that meta-modal conceivability is then uniquely determined by the sum of individually conceivable scenarios. Other arguments assume that there are unknowable necessities – an assumption we have no reason to grant, but that at least suggests the intuitive need for a non-epistemic foundation to modality.
Chapter Three set about exploring this idea further. I presented a metaphysically ‘realist’ understanding of metaphysical modality, and defended it against the conceptualist’s skepticism by highlighting its connection to our intuitive ideas about physical indeterminism, objective chance, and the open future. The realist’s primitive conception of modality forces us to take seriously the idea of strong necessities, but they need not give up on modal rationalism altogether. I suggest two principles of modal expansion – the presumption of possibility, and the consistency principle – which together serve to ground modal rationalism on a realist foundation. The end result is, I think, an attractive and defensible view, which preserves many of the intuitive claims we would wish to make about modality. And although it is arguably the conceptualist’s epistemic space that matters for key theoretical purposes, many would dispute this claim – which cannot be fully defended here – so it is worth establishing the viability of realist modal rationalism for those who would place greater weight on this metaphysical modal space.