A nonreductive physicalist libertarian free will

ABSTRACT

Libertarian free will is, roughly, the view that the same agential states can cause different possible actions. Nonreductive physicalism is, roughly, the view that mental states cause actions to occur, while these actions also have sufficient physical causes. Though libertarian free will and nonreductive physicalism have overlapping subject matter, and while libertarian free will is currently trending at the same time as nonreductive physicalism is a dominant metaphysical posture, there are few sustained expositions of a nonreductive physicalist model of libertarian free will – indeed some tell against such an admixture. This paper concocts such a blend by articulating and defending, with some caveats, a nonreductive physicalist model of libertarian free will.

KEYWORDS:

• Free will
• libertarianism
• nonreductive physicalism
• supervenience
• physicalism

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 Since indeterminism is the negation of determinism, the general definition of indeterministic causation is the view that C causes, without nomologically necessitating, E. I leave open the particular model of indeterministic causation. Most commonly, C is an indeterministic cause of E when C raises the chance of E occurring (Suppes 1984; Lewis 1986, 176– 177; Ramachandran 2004, 152). Relevant concerns with this model include the fact that causation can occur where probabilities are not raised (Glynn 2011, 349ff), and vice versa (Hitchcock 2004), which may disentangle causation from probability-raising. Others take indeterministic causation to involve actual causal processes operating within indeterministic contexts, where C is an indeterministic cause of E when there is a chance of E occurring, and C causes E via an actual causal process (Schaffer 2001). To borrow an example from Christopher Hitchcock, two gunmen are shooting at a vase, each with a fifty percent chance of hitting the vase, thus jointly there is a seventy five percent chance of the vase smashing. Gunman A strikes the vase, while gunman B misses. The firing from gunman B raises the chance the vase will break from fifty percent to seventy five percent, but does not cause the vase to break. Rather, it is gunman A that actually causes the vase to break (Hitchcock 2004, 410).

2 Agential Indeterminism may not be a sufficient condition for libertarian free will. Many advocates of libertarian free will posit additional features for libertarian free will, and I do not dispute these possibilities. Indeed, in Section 3, I argue that libertarian free will also requires agents to have some control over what actions occur. Agential Indeterminism may not be a necessary condition for libertarian free will either. So-called non-causal libertarians reject the Agent Causation component, thereby rejecting Agential Indeterminism. But a bulk of libertarians about free will endorse Agential Indeterminism, so it suffices as an approximate definition of libertarian free will here.

3 It is common to deploy a remote-to-local strategy to show that agents lack control over indeterministic physical processes in their brains (Pereboom 2001, 50ff; Shabo 2014; Haji 2000, 333–336; Van Inwagen 1983, 134–142; McCall 1985, 672–674). On this strategy, it is universally granted that agents have no control over indeterministic physical processes – say an atom that decays – in remote space. But, by slowly bringing the indeterministic processes closer to brains, it becomes clear that there is no substantial difference between indeterministic physical processes in remote space and indeterministic physical processes in brains. This being the case, agents have no control over the indeterministic physical processes in their brains either.

4 The Bottom-Up Objection raises two additional worries. First, isn’t it possible to reformulate the Bottom-Up Objection and apply it to the agential cause A instead of the effect A*? Agential cause A is itself determined by its physical base P, which was itself caused by some prior physical process that agents lack control over, so agents lack control over their agential causes A, so agents do not freely cause their actions. This reformulated Bottom-Up problem is solved by reformulating the reply as well. Namely, A itself had causal contributions from some prior agential cause A_-1 as well, where the agent has some control over A_-1, so the physical base P of A is not the sole determinant of A. Possibly, if the Downward Causation principle stands as well, agential deliberations A_-1 causally contribute to the occurrence of P as well, so Agential Indeterminism is true even if the agent cannot control the fact that P metaphysically necessitates A. Here is the second additional worry: according to the Bottom-Up Objection, P* determines A* (from Supervenience). This fact alone calls into question Physical Indeterminism for effect A*, as A* is determined with metaphysical necessity by physical base P*. This calls into question Agential Indeterminism for effect A*, as A* is determined to occur. The same replies are still available. A* is not only determined by its base P*, but A* must also have an agential cause A, so A* is not determined by metaphysical necessity from P*. And, on the plausible Downward Causation view, A can influence whether P* occurs or not, so A can influence whether A* occurs or not.

5 Some even go so far as to suggest that P fixes the probability of P* occurring (satisfying Indeterministic Physical Causal Completeness on a probability-raising model), but A settles whether or not P* actually occurs (satisfying Agential Indeterminism) (cp. Rigato 2018, 388–392). This possibility shows that, even if A fails to raise the probability of P*, it is not established that A does not influence whether P* occurs or not.