Tuesday, October 8, 2013

Quantum collapse and free choices

When I say that something is metaphysically impossible to do, I will mean: metaphysically impossible for creaturely causation. For this post I leave open the question of what God might be able to do through primary causation. The following seems quite plausible to me:

  1. If a particle is in a mixed |A>+|B> quantum state, then it is metaphysically impossible to determine the particle to collapse into the |A> state.
It is surely metaphysically possible to determine it about that the particle should have a transition from an |A>+|B> state to an |A> state. But not every transition from an |A>+|B> state to an |A> is a collapse. A collapse seems to be a natural kind of transition under the influence of the wavefunction. One can presumably take a particle in a mixed state, and then determine it to have a particular pure state. But that isn't collapse. That is our change of the particle's state. This seems very plausible to me.

The following seems to me to be just as plausible as (1):

  1. If an agent is deciding between A for reasons R and B for reasons S, then it is metaphysically impossible to determine the agent to choose A for R over B for S.
Of course, compatibilists can't say (2). But I find it surprising that in the Frankfurt literature incompatibilists typically grant the denial of (2), allowing that neural manipulators or blockers can induce particular choices. But I see very little reason for an incompatibilist to think (2) true. Of course, it may well be possible to deterministically induce a transition from the state of the agent deciding between A and B to the state of the agent attempting to execute A. But such a transition would seem to me to be very unlikely to be a choice.

Simply doing A after deciding between A and B does not constitute having chosen A. Nor is it sufficient for having chosen A that one does A because of deciding between A and B. For one to have chosen, one's doing of A must be caused in the right way by one's process of decision between A for R and B for S. But it just seems very implausible that an externally determined transition, even if it somehow causally incorporated the process of decision, would be a case of causing in the right way.

Could there perhaps be overdetermination, so that one's transition from deciding between A and B one's doing of A be both an exercise of freedom and externally determined? Quite possibly. But that wouldn't be a case where the choice is overdetermined. Rather, it would be a case where choice and external determination overdetermine the action A. The choice, however, is still un-determined.

But couldn't one make the agent choose A for R over B for S by strengthening the motive force of R or weakening that of S? I don't think so. For as long as each set of reasons has some motive force over and against the other set of reasons, it might yet win, just as a particular in a |A>+0.000001|B> state might yet collapse into the |B> state.

The above doesn't settle one question. While it is not possible to determine that one choose A over B, maybe it is possible to determine that one not choose B, by preventing a choice into a decided-for-B state, while allowing a choice in favor of the decided-for-A state? I see little reason to allow such a possibility.


Heath White said...

I don’t know enough QM to follow the analogy, but my first thought is that if (2) is true, it is just as impossible for the agent to determine this transition as for anybody else. I suspect that is not the result you intended. :-)

My second thought is that you are right that the transition from choice to behavior must take place “in the right way.” (Famous philosophical weasel words.) I became convinced a while ago (I think by Hursthouse, “Intention”) that there is no way to specify this “right way” in efficient-causal terms. You have to use teleological terms. That is, the behavior has to be undertaken *in order to* execute the decision, and there is no specifiable causal mechanism that would *necessarily* fill that role. The compatibilist can’t specify one and the agent-causalist can’t either.

My third thought is that I suspect that one issue between you (and agent-causation fans generally) and many compatibilists is that you think of choices as metaphysical simples, unanalyzable in some sense, while compatibilists tend not to: choices are the products of some brain-process or other (or some psychological interaction of motives and desires) that can in principle be analyzed into parts or sequences. The issue may come down to that.

Alexander R Pruss said...


As always, very helpful!

1. Yeah, I did mean: it's impossible for anything other than agent. It's like when people say "God created everything." They do mean to have one exception!

2. I think I agree that something like teleology is needed here. Maybe this this can be specified using Aristotelian causal powers? The behavior (or its essential kernel, the trying or intending) needs to be an activation of the causal power of the motive (I think this makes me finally see the point of the Aristotelian dictum that the effect is the actuality of the cause, though we want to qualify it for the divine case). And there is a crucial difference between something being the activation of a causal power and its being an effect of the causal power. This is teleology, of course: the causal power is a power for an effect.

Actually, I think there are two things here. The action needs to come from the motive in the right way. And the Aristotelian metaphysics may give a complete story about that.

But that's not sufficient for the action to be the result of a choice, because that the action comes from the motive in the right way is quite compatible with the agent having no other motives and hence not having a choice to make. The action needs to come from a choice for A for R *over* B for S. It needs to come from the R motives and go against the causal power of the S motives. I don't know how to give an Aristotelian specification of the latter.

Alexander R Pruss said...

3. By the way, I don't really make much use of agent causation. I believe in agent causation mainly because I think that fundamentally all causation is substance causation and the agent is the only relevant substance.

I wonder if thinking about choices as some sorts of metaphysical simples isn't required if the notion of mortal sin is to have the sharpness it seems to need to have?

Alexander R Pruss said...

Actually, on further thought about your first point, maybe what I wrote stands as is. For the agent doesn't deterministically cause herself to choose, either. She just chooses.