I recently discovered a paper on modal realism posted by Damon Woolsey, and I liked it very much. Here is a brief take on it.
Woolsey endorses modal realism, but has some problems with David Lewis’ version. He first considers the alternative of branching worlds as a way of reducing the extravagant number of non-overlapping causally distinct worlds in Lewis’ theory. This gives one a system of world-lines in a tree-like structure. A virtue of this may be that it is consistent with our intuition of indeterminism and an open future (for Lewis, the actual and possible worlds are deterministically fixed). On the other hand, the idea of a persistent individual persisting across different world-lines is problematic (vs. Lewis’ use of counterpart theory). Also, with branching worlds, we must accept that the “worlds” are constructions of a sort, rather than really unified worlds. Given this, we should bite the bullet and give upon fundamental possibilia as unified worlds and consider them instead as the set of possibilities considered from one’s particular point of view.
Without postulating all the details of a causal theory or a theory of mind, Woolsey asks us to consider that each of us comprises a causal nexus. We have direct acquaintance with the state of affairs which enter into our nexus. There could be various possibilities going on at points distant from us (patterns of sand in the Gobi Desert) which are consistent with our local facts. So, real possibilities are defined as those states of affairs which can mutually co-exist with a given causal nexus. Given this vision, a possible world is derived as a specific “maximally consistent” state of affairs. The paper has a good discussion fleshing out these notions.
To digress for a moment: one thing I like about this idea is how it coheres with the relational interpretation of quantum mechanics. Possibilia are real, but this reality is relative to a given quantum system=causal nexus (as opposed to a modal realism which features truly branching worlds, which would be consistent with the many worlds interpretation).
In Woolsey’s view, one can derive worlds also by tracing the chain of events back and looking at what other possible world-lines could have happened (he calls this the “causally constrained principle of recombination”). I like this, because it coheres with my discussion in prior posts about how we can know about metaphysical possibilities (see this post, including the comment thread, also this one on Alexander Pruss’ theory). The idea is that the real possibilia that ground our modal intuitions are those adjacent to us. Possible worlds and the unicorns and such which inhabit them are more distant entities which are extrapolated from this raw material. Our conceivability apparatus “rewinds the tape” from our current situation to infer the nature of more distant possibilities. It gets more difficult to imagine how this works when we go beyond nomologically possible worlds and individuals, but I trust our intuition that physical laws are contingent as well, and that our ability to conceive the full metaphysically possible landscape is basically sound. After all, on this account, the intuition is a form of knowledge grounded by contact with real adjacent possibilia. (One area where Woolsey differs is that he is reluctant to endorse possibilia which go beyond the ones consistent with our physical law and the initial state of our world).