now, perhaps it's not clear whether hawking (or if he's getting feynman right, feynman) is actually asserting that every possible universe is actual, that is, asserting, for example, that all possible incompatible physical theories are simultaneously true, or that the universe is exhaustively described by newton's physics, and is not, or that stephen hawking both does and does not exist, and so on. (these would be entailments of the view.) going to need a long run at 'possible,' i think. but what hawking asserts is that each possible universe exists 'each with its own probability.' now i would really, really like to see him or someone pay off on that.
i know what might be meant by, say, 'santa claus probably exists.' this is epistemic probability: we have more reason to think he exists than not (as of course we do). but then, to say that he exists, but exists with a certain probability, or that he or indeed the universe, exists 72.3%: i don't know what that can mean. i wonder whether the sort of existence that this universe has is probable existence rather than actual existence? or is probable existence actual existence or actually not? whatever probable existence is, could something probably exist and not exist at all?
anyway, 'existence' seems to me bivalent: one either exists or one does not. probably existing just isn't existing, probably? no idea what could be meant. but the formula is this: all possible alternative histories do exist, but each exists with a certain probability. good heavens! that they do exist entails that the probability that they exist is 1, doesn't it? for example, before the race, we might assign a probability of 5-2 to secretariat's winning. but after secretariat does in fact win the race, the probability that he won the race is 1. so if each possible universe does exist, then the probability that it exists is 1 (on a 0-1 scale). anyway...
