I’ve been thinking about libertarianism a lot lately, thanks to some who have pressed me on the issue. I don’t know if much progress can be made quickly, because every time I discuss the issue with them I feel as though they are simply misusing language, but it’s terribly hard to put my finger on precisely what is going on. The following is a hypothesis which may or may not be right. First, I register two distinctions between two kinds of ‘randomness’ and two kinds of ‘determinism’, and then I move to the syntax of the sentence ‘Suzy chose A’.
One of the problems with the free will debate has been that people use slippery language. For instance, in particular, when people talk about something’s being random, there are at least two different and distinct ideas which often get conflated. First, something’s being random might just mean that given a set C of circumstances, A follows in such a way that C⊃Av~A. In other words, there is at least one logically possible world in which C and subsequently A, and at least one logically possible world in which C and subsequently not-A. If this is all one means by saying that ‘A’ is random, clearly it doesn’t follow that ‘A’ is random in the following sense: that the process leading to A was causally indeterministic. A cause fails to be indeterministic here just in case it is determined, and clearly in the case of libertarian free will a libertarian-free agent like Suzy does ‘determine’ her choice. In other words C determines A if and only if C causes A. Suzy’s choice isn’t determined by anything other than Suzy herself, acting as a first-mover or first-cause.
So, the first meaning of random is one to which all libertarians will, it seems, be very happy to agree. The second meaning of random is one with which no libertarian will agree, and indeed the libertarian will feel as though something must be being lost in translation, hidden under the misuse of language. The reverse happens too. As we’ve just seen, C determines A if and only if C causes A. Clearly, though, there are different meanings of the term ‘determines’ which have to be distinguished. If by determined somebody means that C determines A if and only if in all logically possible worlds C⊃A, then clearly the libertarian will reject this meaning. Notice that even determinists do not typically construe determinism that way (or else that, per ad hominem, would entail modal collapse). If ‘determine’ simply means ’cause’, though, then clearly Suzy is the cause of A when she libertarian-freely chooses A.
If one accepts the principle of disjunctive causation, then it seems that one will be able to accept that Suzy causes A, even if in the same circumstances Suzy may have failed to do so. There seems to be no room left to argue for the incoherence of libertarian free will at that point. Perhaps one will advance that this libertarianism remains either clearly incompatible, or at least not clearly compatible, with the principle of sufficient reason, but who cares? The debate over whether the PSR and libertarian free will are compatible is a different debate than the one over whether libertarian free will is coherent at all (unless one believes that the PSR is necessary for coherence, but that leads quickly to a reductio for libertarian free will, since the PSR, if true, doesn’t lead to modal collapse, and the only way to break the entailment from the PSR to modal collapse is to employ the very same move the libertarian is employing in her account of free will). Bracketing out the issue of how libertarian free will fares with the PSR, are there other grounds for claiming that libertarian free will is incoherent? In other words, that it is incoherent to say that Suzy caused A when she wasn’t ‘determined’ to do so in the stronger rejected sense, and that her action wasn’t random in the stronger rejected sense of being a ‘random-process’.
Somebody will argue, I anticipate, that the process Suzy employed was either random or determined. This seems like a slide right back into the misuse of language to me. Let’s quickly review some facts about what kinds of particles a sentence (and thus a proposition) has or may have. First, nouns can be in the nominative or accusative (among others), and verbs have tense (present, imperfect, future, etc), voice (active, passive and middle) and mood (indicative, imperative, subjunctive).
Certainly what Suzy did was not ‘out of her control’. She was not ‘passive’. It is not as though she underwent a procedure in which she was passive, whether that procedure be called ‘random’ or ‘determined’. Rather, she was ‘active’. The voice of the sentence expressing the proposition that ‘Suzy chose A’ is active, Suzy is in the nominative, A is the direct object of the verb which makes it accusative. When somebody asks how it has come about that Suzy chose A, I think they must imagine that the voice of the sentence is passive or middle. It’s not though – nothing has happened to Suzy according to the sentence. Suzy acted. If somebody asks what process led from Suzy to A, they cannot be imagining that the sentence has an active voice. After all, they want to know ‘what happened to Suzy’.
It is commonplace for scientific explanations to reduce things to passive objects in this way, because that language works so well. However, what the libertarian is proposing is something which isn’t strictly open to scientific investigation. To view Suzy as a passive object in this way is just to misunderstand libertarianism. So, maybe this is what’s going on here; the person who thinks that libertarianism is incoherent is reinterpreting the sentence to mean that Suzy underwent a process by which was brought about ‘A’, to which she is related as some kind of ’cause’. That misunderstands 1) that Suzy was a first-cause, and 2) that Suzy did not undergo a process which brought A about.
I invite thoughts and criticisms.