Some time ago, Steve Hays wrote some comments on Peter van Inwagen’s arguments showing that free will and determinism are incompatible. (http://triablogue.blogspot.com/2005_07_01_triablogue_archive.html)
I want to use Hays’ comments to illustrate some mistakes people (particularly Calvinists like Hays) make when thinking about free will and determinism.
The first general problem is in the definitions that Hays provides. He says that if libertarian free will existed then there are only three logical possibilities: Hard Determinism, Soft Determinism, and Indeterminism. As I make clear if libertarian freedom existed the first two options are logically impossible. Hays attempts to pick out the concept of Hard Determinism by saying that it is the idea that “We are not free to do otherwise even if we wanted to do otherwise.”
This isn’t the idea of Hard Determinism. Hard Determinism put forward by the likes of Derek Pereboom is in part the idea that Determinism is true and Libertarianism is false. If Libertarianism is false, then there cannot be libertarian free will. Hard Determinism is furthermore the idea that since determinism is true we have to modify our ascriptions of freedom and moral praise/blame and moral responsibility accordingly since our everyday or pre-theoretical notions of moral responsibility and freedom are not compatible with determinism. Consequently it is hardly informative to say as Hays does that Hard Determinism is compatible with the idea that freedom is an illusion since that is exactly what the position maintains. Hays is clearly confused as to what is Hard Determinism.
How does Hays fair when trying to grasp the idea of Soft Determinism? I am afraid not much better. He glosses Soft Determinism as “We are free to do otherwise if we want to do otherwise—although we are not free to want to do otherwise.” The problem is that this is not Soft Determinism as much as a gloss on what Compatibilist freedom amounts to. Soft Determinism is the idea that determinism is true, we have freedom and determinism is logically compatible with freedom where such freedom does not include being the ultimate source or terminus of one’s actions and having alternative possibilities. Soft Determinism excludes even “wanting” to do otherwise since it excludes any form of alternative possibilities simpliciter. Alternative possibilities are the exclusive domain of Libertarianism.
What Hays is trying to say is that Soft Determinism which entails the Compatibilist notion of freedom can make sense of our every language of doing otherwise by what has been dubbed conditional analysis. The main idea is that statements of the form “x could have done y” can be explanatorily reduced to statments of the form “x would have done y.” If antecedent causal states had been different than Jones would have done other than he did do. Conditional analysis has been pretty much been shown to be a philosophical flop. (See Gary Watson’s anthology on Free Will from Oxford, specifically Chisholm and Austin’s essays.) Generally it is a flop because it fails to capture the necessary and sufficient conditions of concept of the ability to do otherwise. What an agent can do is not exhausted by what does do and noting what an agent would do doesn’t imply that they could do otherwise.
Soft Determinism is built off a Compatibilist notion of freedom which can be thought of either hierarchically (Frankfurt) or historically (Fischer). Compatibilism is the idea that determinism and freedom (and perhaps moral responsibility as well) are logically compatible while being silent as to whether we have freedom or determinism is true. Soft Determinism is therefore a stronger claim than Compatibilism.
Frankfurt’s hierarchical view is roughly that an agent is free (and an agent/person) if they can have second order volitions. So if I can will what I will in a self reflexive way then I am a genuine agent and therefore a suitable candidate for praise or blame regardless of how I got to be the way I am. Fischer’s historical account is that an agent is free if their actions are formed by reasons in the appropriate way-if they are “reasons responsive.”
As to Indeterminism Hays interprets it as “We are free to want to do otherwise.” Here again he seems to miss the mark. Indeterminism isn’t a thesis about desires or willing at all because it isn’t a thesis about agency. Indeterminism like determinism is a thesis regarding causation. Logically it is possible to be an indeterminist and not believe that there are any agents at all. Libtertarians and Incompatibilists in general take Indeterminism to be a necessary condition for libertarian free will but it is not a sufficient condition and certainly not a thesis regarding agency per se. To his credit Hays correctly notes that the Westminster Confession opts for a kind of Soft Determinism but this is hardly news.
Next Hays takes up the challenges of deploying the famous Frankfurt Counter-examples. I take Hays’ discussion of Frankfurt Cases to be the “cleft pallet” version of Frankfurt Cases. It is a clear a deformity. Hays writes,
“An example of hard determinism would be Frankfurt-cases. A Frankfurt-case is a thought-experiment in which the subject, unbeknownst to himself, has a failsafe device implanted in his brain which would prevent him from making a certain choice. Frankfurt-cases are generally deployed to show that LFW is not a necessary condition of moral responsibility. But aside from their relevance to the ethical issues raised in the debate between compatibilism and incompatibilism, they are also relevant to the epistemic question of what would count as evidence for LFW, were it true.
The problem which Frankfurt-cases pose for libertarians is that the subject of the experiment believes himself to be free, even though he isn’t. There is nothing in his experience to falsify his belief that he is other than free, even though his belief is false. On this view, not only is hard determinism compatible with moral responsibility, it is also compatible with the illusion LFW. It is not my purpose to make a case for hard determinism. Rather, I’m arguing from the greater to the lesser. If the indeterminist can’t even disprove hard determinism, he can scarcely disprove soft determinism. The problem is that an agent is in no position to know, from the inside out, whether his actions are determined by an external source. As such, this question can only be resolved by revelation rather than reason.”
To be thorough I must note that off the bat that Frankfurt cases are not cases of Hard Determinism. Frankfurt cases were taken to show that alternative possibilities are not necessary for freedom and/or moral responsibility without the assumption of the truth of determinism. Hard Determinism is in part the idea that determinism is true and is incompatible with freedom thereby precluding freedom. Clearly Hard Determinism and Frankfurt Cases are not expressing the same ideas.
Frankfurt argued that his cases show that alternative possibilities aren’t necessary for freedom and/or moral responsibility because the counter-factual covert controller could flip the switch and alter the subject’s choice if the subject chooses B instead of A. In the actual course of the events, the subject chooses A and the covert controller does nothing at all except observe. Frankfurt argues that the subject appears to be free and to be morally responsible for choosing A even though he could not have done otherwise.
Others, like Stump, have argued that the lesson of Frankfurt cases is not that freedom and moral responsibility are compatible but that freedom and moral responsibility don’t require alternative possibilities and only that the agent be the source of their act. On such a view alternative possibilities wouldn’t be necessary for freedom and moral responsibility but the latter two ideas would still be incompatible with determinism.
Others, like Kane and Widerker, have argued that Frankfurt cases either presuppose determinism since the prior sign that tips the controller off as to what the subject is going to do can only indicate what the subject is going to do if it is a causally sufficient condition for the subject’s action. If there is no antecedently sufficient prior sign, then the controller can’t preempt the subject’s choice. The current discussion concerning Frankfurt cases has pretty much reached a philosophical stalemate. That being the case I am not clear on what value Hays thinks that they have contra Libertarianism.
Is the problem for the subject in Frankfurt cases that he believes himself to be free but isn’t? If he isn’t free, then this certainly not what Frankfurt aimed to show. Rather Frankfurt aimed to show that the subject was free even though he could not have done otherwise. Here Hays gets it wrong. (This is the polite way of saying that he can’t seem to accurately reproduce the ideas of others reliably.) Moreover, since Frankfurt cases were designed to help grasp a concept in metaphysics, the subject’s knowledge or lack thereof plays little or no explanatory role as to whether libertarian freedom is a coherent concept or alternative possibilities are necessary for free will or what the concept of freedom is.
As to epistemology, I am not sure as to what the problem is supposed to be. I grant that in cases of say Descartes’ evil demon, the god of Calvinism (is there a difference?) or other covert controllers of exceeding power that the subject in question either can’t or probably can’t, find out that their alternative possibilities have been counter-factually and covertly eliminated. So what? The only thing of significance that appears to follow is that the agent doesn’t know what they took themselves to know and that is hardly a big deal. What we need is an argument to show that such a situation is analogous to the situation we are fact in and this is an argument that Hays doesn’t suggest let alone give. In any case, Frankfurt cases aren’t about if you can tell if you have free will or not, but what constitutes freedom. That is to say, it isn’t a question about how we know something but about what that something is.
Furthermore, even if we were in such a case and all our actions were predetermined by God, how would say a professing Calvinist be in a position to know that they were elect or had genuine faith? To appeal to self authentication or an inner witness bakes no bread since one could be determined to think that they had the experience of self authenticating faith or an inner witness without in fact having it. How could the professing Calvinist tell the difference? How could they find out if God had determined them to have genuine or spurious faith and hence a reprobate? By Calvinist lights God does either infralapsarianly or supralapsarianly predestine people to be reprobate. How are they to know if they are elect or reprobate? What could possibly function as evidence that one was one or the other? What evidence could the Calvinist give for any of his beliefs being true? If God determines that most of the Calvinists’ theological beliefs are to be false and that all of the data that the Calvinist appeals to is determined to serve that end, how is such a person in a position to find out and know that any of their theological beliefs, including their Calvinism is true? If it is equally possible that God could determine most of their theological beliefs to be true or false, then this would presumably imply that first, none of their theological beliefs have any greater probability of being true than 50% and second, that their belief in theological determinism is just as probably false as well as true. Consequently, I don’t see how the Libertarian is in any worse position (and probably better) than the Calvinist in cases of covert controllers of exceeding power. At least the Libertarian only has to worry about the Devil and self deception, neither of which determine the agent’s actions or beliefs. And the libertarian wouldn’t be responsible or so I would argue in cases of pervasive and exceedingly powerful covert counter-factual control. He would just be a sad and helpless victim. The Calvinist has to worry about the Devil, self deception and God as potential obstacles to finding out if they have any sort of freedom (compatibilist included) or if any of their beliefs are true. They also have to worry about being responsible for acts that they are not the source of. It seems hard to differentiate cases of predetermined control from cases of manipulation and people certainly aren’t (fully) responsible in the latter cases and so not on the former either. If the Calvinist god is the source of evil acts it becomes difficult to see at worst how he is not evil, Calvinist protestations aside, or at best innocent.
And per Hays does the “Indeterminist,” by which I think he means Libertarian, have to disprove Hard Determinism? Remember that Hard Determinism is in part the thesis that determinism is true. From my reading, determinism is thought to be false by most scientists. As to who has to show what in the theological realm the burden seems born by the Calvinist just as equally as the Libertarian. I admit that if there is such a controlling deity then it is hard to see how there could be any epistemological basis for thinking that we had libertarian freedom. But is there such a deity? Has Hays shown that there is? Moreover, as to what could serve as theological evidence for Libertarian freedom I think there is a strong case to be made from the Biblical corpus for it. In a nutshell God has libertarian freedom and we are made in God’s image and therefore enjoy a measure of his kind of freedom. The same language that is employed concerning God’s choices is generally employed with respect to human and angelic choices plausibly giving us grounds for thinking that the freedom is at least of the same kind even if not of the same measure or degree. To deny that such language implies libertarian freedom to humans by the same token denies it to God contradicting every major Christian theological traditon, Protestant, Catholic and Orthodox.
Moreover, it seems hard to see how there could be any basis for any of our beliefs if there were such a deity which entails beliefs about having compatibilist type freedom as well. Certainly Calvinists want to hold to Soft Determinism which entails compatibilist freedom. The reasons that Hays gives as undercutting any possible evidence for libertarian freedom also undercuts any possible evidence for compatibilist freedom as well and therefore undercuts the Westminster Confession’s adherence to Soft Determinism. (It really sucks when you cut off the branch you are sitting on.) Calvinists and Libertarians already agree that Hard Determinism is false so I fail to see why the “Indeterminist” has to prove the falsity of Hard Determinism any more than the Soft Determinist Calvinist. Again, Hays is confused.
In short I agree that if determinism is true, then there seems to be no way for the agent to know if his actions are determined by an external source. But this is given the truth of determinism and I am not sure why a Libertarian has to grant that determinism is true. Moreover, the Libertarian seems to be able to argue that since determinism would render any knowledge claim, let alone a claim about having freedom so problematic, that it is a reason to think that determinism is false since the knowledge we have isn’t that problematic to attain. In any case, Frankfurt cases aren’t intended to presuppose determinism which is why Hays’ understanding of them is muddled and epileptic.
Hays then moves to construct something like an argument against Libertarianism by claiming that Libertarianism implies the same kinds of absurdities as “retrocausation.” Without reproducing Hays comments here I don’t think he is right. Libertarians can and do quite easily agree with the majority of philosophers today that time travel is impossible and that the past is fixed and hence accidentally necessary. (For a helpful discussion of accidental necessity see http://www.nd.edu/%7Eafreddos/papers/anld.htm ) Since the past is fixed time travel is impossible.
Second, on Libertarianism it is false that if you replayed the past you would alter the future. On Libertarianism it is possible to alter the future but it is not necessary that given the same past circumstances an agent will do otherwise but only that they could do otherwise or at least will to do otherwise. It is important to keep in mind that Libertarians are minimally committed to the idea that the agent could will to do otherwise regardless of the success of those volitional acts, that is regardless of whether I do in fact do otherwise. It doesn’t follow that because an agent is free to do otherwise given the same past circumstances prior to some volition that they will do otherwise. So it is possible that if we could replay the past a million times prior to Jones choosing A as opposed to B that for every time that Jones chooses A. This doesn’t imply that the past determines Jones choosing A but only that the past isn’t causally sufficient to explain Jones’ choice. It might be that Jones always has good reasons to choose A and never has (because say, there aren’t any) good reasons to choose B. Hays seems to be confusing “could have done otherwise” with “would have done otherewise” and conceptually these are not identical. It is possible that Jones could have chosen B some of the times but it doesn’t follow that since Jones had the power to do otherwise that as Hays writes, “If LFW were true, and you kept replicating the past, then, of necessity, the same agent would do otherwise in the same situation. If he really could do otherwise, and you keep giving him enough chances to do otherwise, he would do otherwise—sooner or later.” If Jones has libertarian freedom, he doesn’t do anything, A or otherwise “of necessity.” Hays has adeptly created a strawman.
Libertarians are committed to the idea that the future is in some sense open to free agents. Like Tolkien they believe that the smallest person can make a difference to how the world comes about which is why Libertarians have been concerned (at least should be) about public acts and social policies that eliminate large members of the human community (abortion/euthanasia). Eliminating 40 million children in the US certainly seems to have contributed to altering the way the world would have gone. I don’t see what is so problematic about such an idea and it certainly doesn’t commit us to the idea that the past can be altered but only that the past and the future can be understood counter factually. If Jones in world W had willed B at time T, the future would be X. If Jones in world W1 had willed A at time T, the future would be Y. Where is the supposed implication to retroactive causation here?
Furthermore, if the future ended up being different, it wouldn’t render the initial choice to do otherwise impossible or paradoxical since the different choice or the same choice would still be the antecedent state and contributing but not sufficient cause for whatever future came about. Libertarians don’t deny that the past influences or contributes to the future. A different choice, say for B by Jones isn’t analogous to retroactive causation since the Libertarian isn’t claiming that Jones’ choice for B changes the fixed future or the fixed past.
In sum, it seems to me that Hays needs to do some more reading in the current philosophical literature concerning freedom and determinism. His comments on divine freedom in the comments section would take too long to dissect here but they seem just as confused and contrary to his own confessional tradition.