Below is a post inspired by reading this paper by Georgetown’s Alexander Pruss (thanks to Tychic for the pointer to this). It has a number of interesting ideas relating to modality and causality.
Early on in this paper, Pruss makes the point that our every-day modal claims are mainly about “local” or adjacent possibilities. For example, “were I to drop this glass, it would fall”. Importantly, our intuition is to view these claims as being about possibilities local to us in this concrete world, not located in another possible world.
He then discusses the philosophical tradition of analyzing of modal statements in terms of possible worlds -- we might call this a “global” treatment of modality. Pruss concurs with the consensus on this subject that there are compelling analytic reasons to approach modality in this way. He also discusses the idea of truthmakers for modal claims, which leads to a discussion of realist accounts of possible worlds.
He spends part of the paper critiquing two main versions of “possible worlds” modal realism; the concrete version of Lewis, and the abstract versions of Plantinga and Robert M. Adams (with emphasis on Adams). One of his two main critiques of Lewis is the charge that inductive inferences about the future are invalid in this scheme (the second is an argument about ethical paradoxes). We could be in a world where gravity fails tomorrow! This point is familiar from discussions of Hume’s treatment of causation. Lewis’ overall scheme of counterfactual analysis of causation plus modal realism is self-described by him as Humean in spirit, so presumably he wouldn’t have been bothered by this criticism.
Pruss mentions a prominent criticism of abstract possible worlds, which is that what distinguishes one possible maximal state of affairs as our concrete world is unexplained – it is a primitive of the theory.
His next criticism is the important one for where he’s going in this paper: He criticizes the abstract system for the fact that all of these possible worlds already exist, prior to any seeming need for them to explain a local possibility in our concrete world. The simultaneous existence of all these abstract possibilities is a problem for Pruss, because it conflicts with the Aristotelian maxim that “actuality is prior to possibility”. And if the maxim is valid and possibility is grounded in actuality, then it means that the actuality “has some powers, capacities or dispositions capable of producing that possibility, which of course once produced would no longer be a mere possibility.”
This again ties the discussion of modality back to a discussion of causality. Abstract entities are usually viewed as categorically unable to enter into causal relations. Aristotle’s system of causation is a real causal production system which follows from the causal capabilities and dispositions of actual entities in the (concrete) world.
If what Pruss says is right and this Aristotelian view can produce possibilities, could this be the basis for a full alternative treatment of modality which obviates the need for a scheme like Lewis’ or Adams? The idea is that “a non-actual state of affairs is possible if there actually was a substance capable of initiating a causal chain…that could lead to the state of affairs we claim is possible”.
Pruss himself sees two obstacles for this approach. First, while it works for local possibilities, it’s not clear you can get global possibilities (whole possible worlds) out of it. The second problem is the initial grounding of the contingent concrete things to begin with: how did this party get started?
Finishing the paper with a theistic turn, Pruss proposes that a possible solution to these two problems would involve a necessary first cause. A possible world which is not actual would be linked to the actual by following the chain of causation backward until we found a starting point which could serve as a branching point for both worlds. This would be the first cause.
So, starting with an analysis of what makes modal claims true, we’ve taken a roundabout route to the cosmological argument for the existence of God. Whew! (In a second paper here, Pruss has offers further arguments about the Principle of Sufficient Reason, the Cosmological Argument and related topics).
There is lots of thought-provoking stuff here. The point I want to emphasize again is the linkage between solving the problems of modality and causality. David Lewis’ system for handling both topics seems rigorous and consistent, but violates our intuition that real causation and possibility are active in our concrete world. It’s unclear that abstract possible worlds can be linked to causality given the normal view of abstract objects as causally inert, so this is a weakness of this approach. Using one view of real causation, the Aristotelian one, Pruss can get to a treatment of modality, but it requires a big commitment to a necessary first cause to make it work. I'm not sure about that move, but it seems right to me that the path to a solution does does depend on working out a system of real causality.