For many people, the idea of free will is bound up with the notion of “could have done otherwise”. By their lights, if only one future is possible for a person — that is, if the person cannot do otherwise — then free will is an illusion.
Philosopher Christian List — author of the recent book Why Free Will is Real — proposes an interesting species of free will based on the claim that while physics may be deterministic, behaviors at the agent level are not. Agents can do otherwise, according to List, and this is enough to ground free will even if physics is deterministic.
I think List is mistaken, but I’ll save my criticisms for the comment thread.
Readers can find List’s argument in this paper:
Free Will, Determinism, and the Possibility of Doing Otherwise
See you in the comment thread.
keiths:
Bruce:
Not exactly. He really thinks that “should” means “does” if determinism is true.
I think “should do X” means “would be better off doing X”, whether doing X is possible or not. Sentences like “Ursula should do X, but she can’t because she’s afraid” make perfect sense, including under determinism.
Bruce,
That doesn’t actually show that the constraint is needed. It just shows that if you’re going to add a qualifier, it can’t be done indiscriminately.
Leaving the qualifier off doesn’t change the conclusion of Huemer’s argument. It just clarifies the implications of Huemer’s reasoning.
Specifically, it means that if Huemer’s reasoning is correct, then the truth of determinism would imply that no one ever believes falsehoods.
Not only am I content with thin gruel, I actually find it more satisfying than fattier foods. I think being free means no more than doing what one wants because one wants to. Look for more than that and I start not having a clear sense what is being sought or why.
walto,
As Dennett says, it’s the only kind of free will worth wanting.
Yes, i think that’s right, but I believe bruce added an additional motivation.
In any case, I think #1 begs the question, whatever other problems the proof may contain. I think I could demonstrate that if i had the inclination and energy.
Right on. And it’s thinner to boot!
walto,
I addressed that additional motivation here.
keiths:
Bruce:
My counterargument to List works even without the notion of “control”. All that matters is that the microstate sequence is fixed, regardless of the reasons for that fixation. If the microstate sequence is fixed, then the macrostate sequence is also fixed.
I think List goes wrong because 1) he confuses epistemic with metaphysical indeterminism, as explained above, and 2) he forgets that although we’re employing two levels of description, the underlying reality is unified, which means that the different levels of description cannot contradict each other.
By positing metaphysical indeterminism at the psychological level, he is contradicting the determinism at the level of physics.
keiths,
I haven’t read far enough (and probably won’t) to know whether i agree with you about this stuff. My own feeling again, is that i wouldn’t let him get this far. I’d stop him at 3. We’d need to come to some agreement on ‘can’ before it’d seem worthwhile to me to get into mind-body relations and hard determinism issues.
walto,
Neither does List. He writes:
That’s why he adopts the modal interpretation.
keiths,
Right, I know that, but it’s backwards. 3* is BETTER than 3. I don’t care about 3, and 3 is what he’s interested in.
ETA: I’m now home and will look at that Huemer point you made and I left hanging. But, again, I think the main culprit there is 1.
ETA: I don’t think 3* is perfect either, because it needs that “because” phrase IMO, but it’s better than 3. As I said above, thin gruel is all I want or need.
That’s intentional. It’s a reductio.
He says it does, getting this, presumably, from Kant.
I’ve read Bruce’s comment regarding motivation for adding “with respect to the free-will issue” more carefully, and I think it’s confused. Huemer just wants a 1 that people will buy, and, as I said, he’s probably worried about some weird moral case. I think he’d accept your revision to handle that matter and would not care about raised eyebrows. From his point of view, the weirder the result the better he likes it. Again, it’s a reductio.
OK, I’ve now the argument more carefully and am pretty convinced that the only problem is with 1. It’s valid, and the other premises don’t seem to me objectionable. It’s sole problem is that 1 begs the question, and his definitions of that fallacy, inserted in an attempt to deny this, are bad.
walto,
#1 seems pretty unobjectionable to me:
I agree with that premise. We really should refrain from believing falsehoods concerning free will (modulo your concern about moral exceptions).
In my view, the problem lies with #2 and its interpretation:
The problem is clear when you combine it with #3:
Together, those give
…but nothing about determinism requires that people can’t fail to do what they should.
keiths,
2 is just “ought implies can”–a widely accepted principle of moral (and legal) theory. If Jones wasn’t able to stop the train, he can’t be blamed for not doing so. (Maybe he can be blamed for not trying, but trying is something he COULD have done.) I don’t say everybody buys into this principle, but most people have since Kant, so Huemer would likely just shrug if somebody objected to 2. (Kinda like I’ve done with your view about verifiability.) It’s an accepted dogma, even if it’s wrong.
I think 1 is question-begging, and I believe most compatibalists would say the same thing. As indicated I think I could demonstrate this with my own reductio, but life’s too short.
If you take “ought implies can” as a given, I think you’ll see why we can’t accept 1. The ‘should’ in there means we can do stuff that determinism might say we can’t do.
In sum, as Bruce intuited, it’s a dumb argument.
walto,
The truth of “ought implies can” depends on the interpretation of “can”.
Suppose Ursula lives in San Diego. Her father is near death in Manchester, NH, and she wants to rush to his bedside. Clearly, she ought to take a plane. It’s the fastest way of getting there. Flights are available, so in that sense she can fly to Manchester.
Now suppose that Ursula is so terrified of flying that she can’t will herself onto the plane — not even when her father is dying. In that sense, she can’t fly to Manchester. Her fear prevents her from doing so. (And note that this doesn’t depend on the truth of determinism.)
I think it’s still true that Ursula ought to fly to Manchester, and I think the same reasoning applies to moral and legal “oughts”. When we say that citizens shouldn’t murder each other, we don’t mean that everyone has the character and temperament to refrain from doing so. We just mean that a morally upright, law-abiding person will not murder people, and that we want everyone to be morally upright and law-abiding.
That’s why I don’t buy #2:
Ursula should fly to New Hampshire, but she simply can’t do it. Her fear is too strong.
keiths,
Many would say that because of her condition Ursula can’t be blamed for not visiting her dad and, thus, it’s not the case that she ought to have visited him.
Such folks would say, If you ought to do something and you don’t, you’re blameworthy. As Ursula is not blameworthy, it can’t be the case that she ought to have flown to NH.
Anyhow, as I said, this is all neither here nor there. There’s over a hundred years of literature on it, and it’s pretty widely (though not universally) accepted. You deny it, Huemer shrugs. The controversial premise is 1.
walto,
“Ought implies can” is widely accepted, but not with your meaning of “can”.
By your reasoning, it’s false for a determinist to say that a serial killer should stop killing, unless that’s what the killer is going to do anyway.
That’s not exactly a mainstream position, you know.
walto,
If it’s not true that Ursula should fly to NH, on account of her “condition”, then you have to accept that it’s not true that citizens should refrain from murdering each other. They have antecedent conditions too.
Are you willing to go that far?
The “going to do anyway” doesn’t mean anything there. The saying could be part of the antecedent condition.
walto,
Regarding List, you write:
List wants more, and so you’ve reached an impasse.
My deductive counterargument avoids that impasse, as it doesn’t depend on the resolution of disputes over (3), (3*), and (3**).
That’s just the common complaint about determinism: if it’s true there’s no moral responsibility. Every determinist has to deal with it–compatibalist or no. I accept that only those citizens who can should be expected to refrain from murdering each other. Like all determinists, I think punishments are administered for protection, deterrence and rehabilitation purposes only. That’s the cost of the game, bro.
You’re complaints may well be right: I’ll doubt I’ll ever know. As I said, I’m with Dennett on what we’re looking for here. Go beyond that and I don’t really know why or what is being sought.
walto,
Sure it does.
You’re arguing that Ursula can’t fly because of her condition, and that it’s therefore false for someone to say that she should fly. She isn’t going to do that.
The same reasoning applies in the case of a murderer:
Because of his “condition”, the (imminent) murderer can’t refrain from murdering, and so it’s false to say that he shouldn’t murder. He isn’t going to refrain.
That’s a pretty extreme position, and definitely not widely accepted.
The saying could be part of the antecedent condition.
Sure. Why does that matter? It’s still false, by your lights, for someone to say “you shouldn’t kill me” to someone who is going to do so anyway, because of determinism.
walto,
That’s not the price of compatibilism and determinism. One can still accept “shoulds” and “oughts” despite being a compatibilist.
I think people shouldn’t murder each other, or mistreat animals, or vote for Trump.
I’m also a compatibilist, and I see nothing inconsistent in my positions.
You are still assuming that “ought implies can”, with “can” limited to your meaning.
I accept “ought implies can”, but not with that same meaning.
walto,
In a nutshell, List is seeking to show that there are alternative possibilities at the psychological level even if there are none at the physical level.
My counterargument shows that this is impossible, given that every physical state corresponds to one and only one psychological state (it’s a surjection, in other words).
petrushka,
No one’s debating whether we have the experience. The dispute is over the interpretation of the experience. And note that even determinists think we make choices. It’s just that the outcomes are determined.
The philosophical question is highly relevant to society, because many people think that retributive punishment is justified if the perp truly could have done otherwise (in the metaphysical sense).
(Some) people are curious. They pursue knowledge for its own sake. So even if the question of free will didn’t have societal implications, it would still be worth pondering.
Keiths,
Your ‘can’ is mysterious to me. I take it that in the case you described it’s something like “some people could have gone to NH if they’d wanted.” So if ought imples can in your sense, and she ought to have gone, the world would be better if she’d gone and some people could have gone if they’d wanted to (because, e.g., planes were available).
Or something like that, no?
ETA: maybe ‘some people in her shoes…’
walto,
Yes, but even Ursula could have flown to NH if she’d wanted. It’s just that she didn’t want to ’cause she’s terrified of flying.
keiths,
There is no “can” at the heart of determinism, that’s the whole point. Choices are all illusions.
I know, crazy isn’t it. Who would believe that.
phoodoo,
As walto and I are discussing, it depends on how you construe “can”.
One last thought (I swear!) on Heumler’s proof and List thoughts of free will.
In Heumler, think the two paragraphs before the Notes are important. They discuss how to understand epistemic shoulds when determinism is true. Reasoning seems to presuppose freedom, but if someone believes determinism, then why does reasoning to make the better choice matter? What will be will be (RIP Doris).
Those concluding remarks point to an issue, I think. The proof mixes explanatory domains incorrectly. It mixes Sellar’s Space of Reasons with the Space of Causes. Or you could say it mixes List’s psychological level with the microstate physical level.
From the Sellars angle, the epistemic shoulds are in the space of reasons, the cans are in the space of causes. The proof has hidden premises about naturalizing reason.
From the List angle, the proof is mixing determinism at the microstate level with possibility of indeterminism at the psychological level. The proof has hidden premises about the metaphysical status of reduction.
I think this hidden mixing is the source of the unease regarding possible question begging in 1 and regarding the natures of the shoulds and cans in 2.
Keith, it is possible that some of your comments on can and should were about the same issue.
Bedtime reading for the philosophically obsessed (or those determined to be so)
Defining Determinism
Right, you are mudding up the water by trying to throw in more caveats and multiple definitions that are invalid, when if you just boil it down to its essence, there is no “can” in determinism- no matter how you stretch it into a belief that one can do something.
Believing one can do something is not the same as being able to do something. In determinism you aren’t even able to drink your Sprite, you either drink it or you don’t, there is no can. You either did or you didn’t.
Thus there exists no justification for you to say Ursula “can” do something, because you don’t know that. She either does it or doesn’t do it, based on what must happen.
Let’s suppose you announce to someone, I can throw a ball. In the world of determinism, you have no idea if that is true. You only know you threw one before. You can say you can do it again, but that could be totally false and you would have no way to know it. So “can” doesn’t exist.
Again, a crazy worldview, but it is what it is.
Thanks. I’ll be reading it.
As best I can tell, from browsing the first few paragraphs, it fails to define determinism. That is to say, it appears to be discussing determinism of theories, rather than determinism of the cosmos.
Could she have wanted to if she’d wanted to want to? I think not.
phoodoo,
I don’t entirely disagree with you about this, phoodoo. That is, I think keiths is largely waiving away the problem of moral responsibility, which is real. He says she ought to do it, and that in some weird way she can, but she’s not blameworthy if she doesn’t and, well, she actually can’t do it at all.
Determinists have an actual bullet to bite here, which is one of the main reasons why the Lists are so hell bent on figuring SOMETHING out. I’m content to bite the bullet, myself. Keiths says it’s an extreme view, and as you hammer down, it actually IS. But I can’t get myself to care about blameworthiness, and I’m OK with ought implying can, just as you say, and almost everybody else agrees.
You may say I’m a goddam rebel—but I’m not the only one. There are a lot of compatibalists. Shit, even FMM is one.
ETA: I kinda wish KN would weigh in here. I’m guessing he’s thought about this a lot.
phoodoo:
So if determinism is true, you don’t think you can post comments at TSZ?
That’s….. odd.
walto,
I’m not treating it as a moral issue. Ursula wants to rush to her father’s bedside. The best way to do so is to fly to NH. In order to satisfy her desire, Ursula should fly to NH.
As for whether I’m “largely waiving away the problem of moral responsibility”, I would say the answer is “no”. As I’ve stated before, I think we are proximately responsible, but not ultimately responsible, for the things we choose to do. And because we are not ultimately responsible, I am opposed to retributive punishment.
So am I. Compatibilism is not what makes your position extreme. I’ve already explained what does:
walto:
keiths:
walto:
So? Nothing about my position implies that we can do everything we want.
I’d love to be able to fly without the aid of an aircraft, but it ain’t gonna happen.
phoodoo,
That’s not an argument against determinism. It “works” just as well if determinism is false:
So according to phoodoo logic, no one should ever say that they “can” do something. Not even if determinism is false.
walto,
Earlier, you wrote:
It seems odd to me that you would accept a compatibilist view like the above, yet balk at something like this:
keiths,
I don’t balk at that at all. It doesn’t get one to moral responsibility imo though. That’s the problem for determinists. I’m content with freedom being doing what one wants because one wants to. In your story it’s quite likely the daughter couldn’t go to nh even if she wanted to.