The Breadth of Possibility

Okay, with the initial characterization of metaphysical possibility clarified, we can now ask: just what worlds really are possible? We can begin to answer this question by working backwards.

Firstly, the actual world is certainly possible. So our modal space is non-empty. (That's a good start!)

The next obvious opportunity for inclusion arises if our world is indeterministic. Suppose it is not yet metaphysically fixed whether some particular instance of radioactive decay will occur at time t. This then presents us with a range of genuine possibilities. (The atom might decay at t, or it might not. The world really could turn out either way.)

Let us call possibilities of this type, "dynamic possibilities". We may define a dynamically possible world as any nomologically possible world that shares our actual history up to time t. (If determinism is true, then the actual world is the only dynamically possible world, since the first moment - in conjunction with the laws of nature - determines all.) Dynamic possibility reflects our intuitive notion of what still can be. The past is fixed, but perhaps a variety of future possibilities remain open.

(This is related to our concept of objective chance. If determinism is true then every event has an objective probability of either 0 or 1. It will either definitely occur, or definitely not occur. But a world might be governed by indeterministic laws, and the chance of an event E occuring might plausibly be analyzed as the proportion of dynamically possible worlds in which E occurs. Libertarian free will might complicate things; our analysis would then see 'chance' as not purely random. Undetermined agent-causal influences, if coherent, would also be included. I take the idea of this strong sense of "free will" to be that the agent really could -- categorically, and not merely hypothetically -- have brought about a variety of different outcomes.)

Anyway, the relevant point for now is this: if a world W is or was dynamically possible at any time, then W is clearly metaphysically possible in my sense. It really could have been actualized. The only reason it wasn't, was due to random chance, or perhaps exercises of free will, which nothing prevented from going the other way.

This has gotten us some distance towards elucidating the breadth of modal space, in what is hopefully a clearer way than my initial invocation of the "rewind and replay" heuristic. But we are still stuck within the realm of nomological possibility, further bound by initial conditions. So now we must ask: was there the opportunity for our universe to have different initial conditions, or even different laws of nature?

I don't know how we could go about answering that, or even how to make clearer sense of the question. (One might invoke the God heuristic again, imagining that he had a prior decision about which world to actualize. But I don't know how helpful that is.) So I'm just going to stop here for now.

Still, we've made some progress at least. We have seen that any world that was ever dynamically possible is ipso facto metaphysically possible. It's just when we try to work all the way back to the very beginning of our universe, and ask if that could have been different, that our intuitive modal concept runs out of gas.

Categories:

• modality
• metaphysics
• probability
• freedom
• philosophy