I previously outlined three modes of possibility: logical, metaphysical, and actual. Logical possibilities are those permitted by logic, metaphysical possibilities are those permitted by logic and the laws of metaphysics, and actual possibilities are those permitted by logic, the laws of metaphysics, and actuality. I would like to describe a particularly interesting subset of metaphysical possibilities.
Mere metaphysical possibility is often too broad in scope to be relevant to a given project. Consider a group of people discussing who could have won the ’07 presidential election. A gadfly among them proposes that Mr. T. could have won, and goes on at length about how such a scenario could have obtained.
While our gadfly may have some interesting things to say, his attitude is annoying because he is bringing up a possibility which is not relevant to the project at hand.
But the next rung down the modal ladder is “actual”, and in many cases when metaphysical possibility is too broad to be relevant, actual possibility is too narrow. Perhaps among the group of people discussing who could have won the election another gadfly jumps in and argues that only Obama could have won, because there are no actual alternate possibilities. Though quite correct, he is equally annoying, for the same reason that our first gadfly was annoying: his point is not relevant to the project at hand.
What our friends are really interested in is a subset of the metaphysically possible worlds in which things like the physical constants, quantities, and laws of the actual universe obtain, and which share the same history as the actual universe up to some point prior to the election. They want to know whether in any of these such possible worlds, anyone other than Obama wins the election, and if so, what else is different about such worlds that brought about the difference.
Let’s call these “real” possibilities.