(I haven’t heard of this before, which I suspect means that, like L. Ron Hubbard’s science fiction, it didn’t exist in my original birth universe; I wonder if that means I’ve switched again recently. Always hard to tell.)
Anyway, the Free Will Theorem, which is described in two papers that are both quite readable really, is not actually about Free Will to speak of, at least not if you are a sensible compatibilist like we are, and I want to write down my thoughts here as to why and how that is.
What the Theorem actually shows is that, if some generally but not universally accepted parts of quantum physics and relativity are true, then if there is some behavior of some humans that can’t be predicted, even in principle if you knew every fact about the universe up to that point, then there is also some behavior of some elementary particles (as evidenced by the behavior of some macroscopic detection apparatus) that can’t be predicted, even in principle if you knew every fact about the universe up to that point.
Which is not a big surprise really; it’s hard to imagine a universe in which all elementary particle behavior was predictable but human behavior was somehow not, humans being made of elementary particles and all. But this Theorem puts a solid example under that intuition (as well as bringing up some other issues in physics that I won’t talk about more here).
Conway admits somewhere in the coverage of this that he chose the name, the Free Will Theorem, intentionally to get attention. But he’s also responded to criticism of the name by saying things worth noting.
The most obvious criticism is that being unpredictable even in principle isn’t the same as having free will (and if you’re a compatibilist it’s not even a necessary condition). Conway has said a couple of things about this.
First, he’s said informally that humans and particles are the same in this way, and since we say that humans have free will we should say that about particles too:
“That’s why I insisted on using this evocative language,” Conway says. “Many people thought I should say the particle’s behaviour is indeterminate. But it would be really rude if I told you that you were indeterminate! It’s the same property and I don’t see why we should be required to speak of it as if it were a different property. Our theorem says that if humans have it, then so do particles.”
But that’s sort of silly. The fact that humans and particles share some property X, and humans have free will, doesn’t imply that particles therefore have free will; it’s just a non-sequitur.
He’s also responded to the “randomness isn’t enough for free will” argument by claiming that the indeterminacy they’ve proved for particles isn’t just randomness. From that same link above:
and which action the particle does is free in this sense, it is not a predetermined function of the past. And that’s not the same as randomness, oh dear me no!”
Exclamatory cuteness aside, if “not determined by anything that’s come before, and not predictable even in principle” isn’t the same as randomness, I don’t know what is.
What Conway apparently has in mind here is that the randomness is weird and quantumly nonlocal: when the behavior goes from undetermined to determined, it does it at two places at once, and the places can be sufficiently far apart that no signal can get between them in time. That doesn’t mean it’s not random, it just means that as well as being random it’s also bizarre in the usual QM way; the Free Will Theorem doesn’t tell us anything particularly new about the weirdness, it’s just one of the three assumptions that it starts from.
Conway gets all sort of Penrose-like and speculates that while all the little “free decisions” made by particles usually sort of cancel out, our brains somehow avoid this canceling out, and that through some as yet unknown process our human-level free will pops out as a result. This, he says, makes the whole Compatibilism thing a moot point; since the universe isn’t deterministic, it just doesn’t matter whether free will is compatible with determinism. Compatibilism, the second of the papers says, is just “a now unnecessary attempt to allow for human free will in a deterministic world”.
Which is not quite right. :) Compatibilism is the recognition that indeterminicity is neither necessary nor sufficient for free will. Free will has to do with the freedom to express one’s preferences and goals in the world, not necessarily the ability to escape prediction by a hypothetical omniscient seer. Free will is possible with or without random or undetermined bits of fundamental physics (and those people who told Conway he ought to just say the particle behaviors are undetermined were right).
In fact free will requires that various of our actions are in fact influenced by, reflective of, if not determined by, past facts about the universe, those facts being the preferences, plans, and goals of the person acting with free will.
Hope that clears that up. :)