Ways to Deny the Existence of Free Will

A while back, I posted about two general strategies for trying to show that free will exists.  Here, I want to suggest that there are also two general strategies for trying to show that free will doesn't exist and get your input on them.

Before I get to that, however, I want to say that there won't be anything parallel to the indirect approach for showing the existence of free will.  In essence, the indirect approach argued (i) that free will is necessary for some other thing x, and (ii) that x exists.  But if we're focusing on the non-existence of free will, saying that free will is necessary for x and then showing that x doens't exist won't establish that free will doesn't exist.  Perhaps the closest thing to an indirect argument for the non-existence of free will would be to argue that free will is sufficient for some other thing x and then argue that x doesn't exist.  But I'm not aware of anyone who makes this kind of argument.

If this is right, then it looks like all arguments for the non-existence of free will will be 'direct'.  But there are two different ways that one could directly argue for the non-existence of free will, which I shall call contingent denials and categorical denials.  A contingent denial will be a view which holds that while it is possible that free will exists, it is a contingent truth that free will doesn't exist.  Categorical denials will be stronger: free will does not exist because it is impossible for free will to exist.  Pereboom is an example of a contigent denier, while impossibilists like Galen Strawson will be categorical deniers.  

What do you think about this way of demarcating views which deny the existence of free will?

Comments

[Neil]

This seems right. It maps onto the (typical) reasons for denying the existence of free will. Incompatibilists will be contingent denies: since (as a matter of fact) the laws of nature are deterministic, no free will. No free will either way theorists - those who accept incompatibilist arguments and couple them with the luck objection, say - will be categorical deniers. There is also logical space for the categorical incompatibilist. Such an incompatibilist would hold that the laws are necessarily deterministic.

[Kip]

Neil:

"Incompatibilists will be contingent deniers."

Incompatibilists aren't necessarily contingent deniers. G. Strawson is an incompatibilist, but he's not a contingent denier.

"There is also logical space for the categorical incompatibilist. Such an incompatibilist would hold that the laws are necessarily deterministic."

This also doesn't seem right, for the same reason as above. Someone like G. Strawson could be a categorical denier, without holding that the laws are necessarily deterministic. Such a person could hold that the laws aren't deterministic, or that they are contingently deterministic, but that free will doesn't exist in either case. In other words, that person could say "it's contingent whether the laws are deterministic; it's not contingent whether free will exists."

[Kip]

Kevin,

I agree that there is an ostensible distinction here. But I don't put that much weight on it for the following reasons:

1. I tend to agree with G. Strawson that free will is logically impossible. The idea is simple enough: free will involves self-creation, and self-creation is impossible, because X can't create itself if X doesn't exist yet.

2. To the extent that people disagree with Strawson, and think that there is some kind of free will that is possible, there is a dilemma of a fork: (i) either they are still talking about the same thing as Strawson, but are just confused about its logical possibility or (ii) they are starting to deviate from the Strawson's idea and to talk about a different kind of free will. In (ii), this is like what happens with event-causal libertarians who say that free will is possible, but their kind of free will is a bizarre, watered down idea very different than what others think "free will" is.

Personally, I think Pereboom is guilty of (i), if anything. But, then again, I hesitate to criticize Pereboom, because he's smarter than me and I learned a lot of what I know about this subject through his writings.

I think a much more interesting distinction to make between fw deniers is:

1. conservative deniers (Nichols, Smilansky, Double, Sommers-since-2006);
2. radical deniers (Pereboom, myself, Tom Clark, Joshua Greene).

I'm not sure where G. Strawson falls on that line, but I would guess 2.

The key idea here is between those who deny that free will exists, but think that would, or should, have practically no effect on our social practices, and those who deny that free will exists, and think that this leads to a very progressive view on crime and punishment.

[malmeida]

In essence, the indirect approach argued (i) that free will is necessary for some other thing x, and (ii) that x exists.

Does this look like the following?

1. O --> F

where O is the relevant observation and F is free will. We observe O and conclude F. So, why not the same against free will.

2. O --> ~F

where the observation that O entails that ~F. In that case we have an indirect argument against free will that takes the same form as the indirect argument for free will. Either both are indirect or neither is. In both cases we can reformulate in terms of sufficient conditions. van Inwagen gives something like an argument based on (2) that appeals to the (in principle) observation of chancy behavior in indeterminsitic worlds (the replay argument).

[Sofia Jeppsson]

Kip, are we back at discussing the criminal law? As I (and others) have already stated, even if it's possible to justify blame in general, it doesn't follow that it's possible to justify any certain view on criminal politics. It's really easy to find examples on people who argue both for the justification of blame in general and for a humane criminal law, without any contradictions.

Back to the subject at hand: I think that compatibilists, sceptics and libertarians more often than not talk about different things when they say "free will". The real debate is usually about what kind of free will matters, is interesting, would be needed to ground morality (but then there's no guarantee that people mean the same thing with "morality" either - compare Parfit and Korsgaard for example, where Parfit doesn't think that Korsgaard's account of morality is "real" morality or matters, and Korsgaard pretty much thinks it is the other way around).

Sure, one can add "desert-entailing" to "free will" but it's not like that specify things either, since people have different ideas of what desert means. Is it some metaphysical lump that sticks to people after they've acted, is it simply part of morality where "morality" is given a constructivist meaning without metaphysical implications, is it something else altogether?

I'm perfectly happy to admit that G Strawson's idea of self-creation is impossible, I just can't see why anyone would be interested in that kind of self-creation. Or to put it this way: If someone was to convince me that G.S-style self-creation after all IS possible and not just that, it exists in the real world - well, why would that new fact have any impact whatsoever on any normative view I hold, on the way I see myself and others, etc? Obviously G Strawson thinks this is of tremendous importance, but I can't see why.

[Kevin Timpe]

Mike,

That's a good suggestion, but I want to make sure that I understand it. Are you saying something like the following?

The chanciness of behaviour in indeterministic worlds is sufficient for there not being free will (because of worries about luck, replays, etc...). The world is chancy in this sort of way; therefore there's no free will.

[Kip]

Sofia,

I was not trying to make any controversial statement about the criminal law. I was simply noting a distinction between two kinds of anti-realists about free will:

A. deniers who say "free will doesn't exist, but that doesn't really change anything, and it shouldn't change anything, because our social practices of praise and blame are useful"; and
B. deniers who say "free will doesn't exist, and that should substantially change our social practices."

For example, Pereboom says that free will doesn't exist, and therefore we should only quarantine criminals without otherwise punishing them. That's a radical view. It's quite different than the view of Nichols or Smilansky, who argue that things should pretty much stay the same, because our current practices provide essential benefits.

I still think that's a valid and important distinction to make. I wasn't trying to make any claim about the justification or foundation of the criminal law.

[Sofia Jeppsson]

Your post made me re-read Smilansky's "centrality of illusion" and Pereboom's "living without free will". What struck me was that although their positions are different, they're actually not as far apart as I thought. (So I'm not arguing that you're distinction is useless or anything - but perhaps softer than you think?)

Smilansky does write in his "centrality of illusion" that incompatibilists have a real point when they say that ultimately any criminal is just a victim of his circumstances, and therefore he shouldn't be made to undergo misery. He doesn't elaborate that point, but it sure sounds as if he wouldn't be in favour of a harsh and retributivist criminal system.

What Smilansky does see as important to keep are such social praxises as tipping an attentive waiter more than a sloppy one (he even specifies in that example that they're both young and healthy, i e have the same possibility in a compatibilist sense of being attentive to the customers), of taking responsibility for our own actions, respecting each other as persons (and it sounds as if he thinks that a real belief in determinism would make us unable to do that, perhaps because we would regard each other as some kind of machines then...), stuff like that.

Pereboom, on the other hand, actually thinks that many social praxises can be kept in a slightly altered form - although there is a sense in which we couldn't be grateful to one another if we came to embrace hard incompatibilism, we could still express joy over good deeds in a way very similar to gratefulness. And the same goes for lots of other praxises - he just argues that we'd replace them with something very similar. He actually writes that the biggest difference would be that we wouldn't be as angry with each other as before, if we came to embrace hard incompatibilism.

But then again, Smilansky doesn't argue that anger is particularly important, it's other things he's mainly concerned with - taking responsibility for what you do and respecting one another.

I think the largest difference between them is actually that they make different empirical predictions on what effect a belief in hard incompatibilism would have on people. Smilansky think we would loose respect for each other and stop caring about our actions, while Pereboom think we could have a really nice moral community even with this belief - in fact a NICER community than the current one, since we'd be less prone to anger.
But they actually don't seem to differ that much in what kind of society they see as desirable.

[Kip]

Sofia:

The biggest difference, to my knowledge, between Smilansky and Pereboom's recommendations, is that Smilanksky endorses pretending that we have free will (and therefore that nothing, or very little, will change), whereas Pereboom endorses proclaiming the non-existence of free will from the rooftops. Pereboom claims that recognizing that nobody is morally responsible for anything will lead to softer treatment of criminals (e.g. just quarantining them).

[Sofia Jeppsson]

Yes, Smilansky's article that I just read is even called "the centrality of illusion", so of course there's a difference there. Although it's quite ironical that Smilansky talks about "respect for persons" as much as he does, since it sounds pretty disrespectful in my ears to claim that common people need to keep this illusion (while presumably Smilansky himself and his philosopher colleagues can do without it, otherwise he shouldn't have written that article to start with).

But the reason Smilansky thinks there's need of an illusion and Pereboom doesn't, seems to be mainly that they have different empirical speculations of how people react to the idea of determinism. They don't seem to differ a lot when it comes to normative views of what a society ought to look like.

[Saul Smilansky]

Sofia,

Let me try to clarify my position, and contrast it with Derk's. I think that philosophically compatibilism is a large part of the truth (50%?) on the compatibility question, and in practice, although it is in an important way morally shallow, compatibilism ought to be the basis of social (and indeed almost all personal and inter-personal) practice. The partial but also true insights of hard determinism ought to influence our thinking as philosophers, and in part our practice as well, but here only in a limited way. This is in part because of what I call in my book (Free Will and Illusion) the "Pragmatic Priority of Compatibilism", roughly, the idea that hard determinism cannot form a basis for social and personal life. We need people to take responsibility, we need to make moral distinctions, we need people to pay when they wrong others, etc, and hard determinism is bad in providing these things. In a more positive way, I believe that compatibilism ought to be central in practice because a Community of Responsibility largely following compatibilist distinctions is a condition for a life of respect for persons, personal decency, and a civilized society (as much as that is possible). So, even if we bracket the need for illusion, the insights of the hard determinist perspective can at most serve as limiting factors on a society that will largely look like (non-utilitarian) compatibilists think it should. In other words, what happens to people will broadly track their compatibilistically-free actions, and particular attention will be given to excusing them (only) when they lack free will, compatibilistically interpreted.

Pereboom will not buy into this.

(Then I have an additional story about the positive moral importance of illusion, and in particular the illusion of libertarian free will, which is crucial for supporting the basically compatibilist moral order and serves other purposes, but that's a different matter.)

Derk gives us an optimistic, progressive, fairly simple, unitary, and harmonic view. My view (both about the philosophical truth and about the ways in which we ought to try, or not to try, to implement it in practice) is pessimistic, conservative, distasteful, complex, fragmented, and messy. I wish he was right.

[Sofia Jeppsson]

Saul, but in the final paragraph you seem to agree that the difference between you, even if a pretty big one, is still largely about empirics? I e, you have different views on psychology and how people react to determinism?

[Saul Smilansky]

Sofia - well, insofar as my Illusionism goes then, yes, mostly, but in terms of the compatibility question then much of my disagreement with Derk is not reducible to different predictions about empirical matters. No one seems to believe me, but I do think that compatibilism is true quite a long part of the way, while Derk of course disagrees. I am quite happy, for example, to talk about desert. My example of the waiter that you mentioned was supposed to help us to see how desert can make sense in a deterministic world. It's going to be a rather morally shallow sort of desert, but "shallow" does not mean unimportant, in my view. It's possible to push the claim about convergence from the other end, perhaps: much of Derk's "substitutes" for commonsense notions of free will, moral responsibility, desert and the reactive attitudes arguably approach dangerously close to compatibilism. But that's for him to respond to. From my end, I disagree with hard determinism when it claims that if there is no libertarian free will then there can be no free will, moral responsibility, desert, or just praise, blame, or punishment. All these are possible, but they exist in a limited way alongside the also limited hard determinist truths. The true position on the compatibility question is a complex one combining the insights of compatibilism and hard determinism. "Monists", whether they are compatibilists (like yourself, it seems) or hard determinists (like Derk) are only seeing part of it. Or so, at least, it seems to me.