So Jerry Coyne is criticizing Dan Dennett for being misguided about free will. This comes in response to a review Dennett did of a book by Alfred Mele, which takes the purported scientific evidence against free will to task, and finds it at least unconvincing. By the end, Coyne will reiterate his comments that compatiblism is like “Sophisticted Theology [tm]” in his view, and will also end up, in my opinion, committing the same sins of scientism that I’ve talked about before.
I’m going to ignore the parts about the Templeton Foundation, and focus only on the free will parts. After a number of introductory points, Coyne starts to try to justify his belief that compatiblists really think that science simply cannot say anything about free will by looking at Dennett’s and Mele’s criticisms of the scientific work, and gets one objection seriously wrong from the start:
The last point puzzles me; I don’t see why contextual factors should be ruled out a priori as “not decisive”. When an authority figure in a white coat stands over you and tells you to apply more voltage to a passive victim supposedly connected to a battery, why shouldn’t that affect your behavior? Nobody denies that environmental and social pressures can change how you behave.
Including Free Will Libertarians, which is why the examples given — generally the Zimbardo and Milgram experiments — puzzled me, as no one should think that those cases said anything meaningful about free will. But this is not because of some kind of a priori reasoning that contextual factors should be ruled out. Contextual factors certainly influence our decision making, and will always be a big part of the “reasons” why we do things. The potential challenge for free will from contextual factors is that our decisions will be at least disproportionately impact by contextual factors that have no possible logical connection to the decision that we made, and that therefore cannot be used properly as reasons for our actions. Which would then imply that our decisions are not made for reasons, and that therefore reasons and reasonings are epiphenomenal; they have no actual impact on our decisions. At that point, we clearly couldn’t have anything like real choices, real reason, real rationality. We would be creatures that go through the motions of decision-making, but never really make decisions … and never really act for reasons.
Coyne is right that an authority figure telling you to do something is not the right sort of contextual factor to raise this issue. We can indeed find reasons to trust and follow the instructions of authority figures. The right sort of contextual factors might be the sorts of things that Dan Ariely talks about, but even in those case those seem to be influences and less determining factors; it biases us towards one choice or another, but doesn’t mean that we will just make that decision. In order for it to work, it has to be a factor that is a factor that overwhelmingly influences our decisions and is exceptionally irrelevant to making a reasonable decision. We would need something like the fact that I like cheeseburgers determining whether or not I, say, favour Utilitarian views or Stoic ones … or believe that God exists or not. None of the examples given rise to that level, but it is this sort of finding that Mele and Dennett and those who think that we make meaningful choices are worried about.
But never mind. What all this shows (and Dennett admits that some of those experiments have not been discredited) is that no scientific finding can refute the compatibilists’ claim that we have free will. Even if, in the future, we could predict people’s actions and future decisions with perfect accuracy using very complex brain-monitoring and knowledge of neurology, compatibilists would continue to claim that we have free will. That’s because their notion of “free will” is a philosophical one, impervious to scientific refutation. So why bother going after the science?
Dan Dennett is a staunch philosophical naturalist. He thinks that philosophy should proceed both in light of scientific data and with a scientific methodology. There is absolutely no reason to think that Dennett thinks that no scientific finding could refute the claim that we have free will. All Dennett and Mele are saying is that these scientific experiments actually refute it. Sometimes that’s because the scientific experiments don’t even prove what they claim to prove, as they’re severely flawed experiments. Sometimes that’s because while they do demonstrate what they claim to demonstrate, what they demonstrate doesn’t actually refute free will. This is what reminds me of the scientism debate, because the main objection here — if it’s made with Coyne understanding what Dennett and Coyne are saying — seems to be that since they didn’t simply accept the purported implications that scientists said followed from their experiments blindly, and pointed out how their experiments didn’t mean what they claimed it meant, that therefore they are rejecting science or insisting that science can’t refute their view, that they are choosing philosophy over science, and not even attempting to bridge the two, and making their view impervious to scientific refutation. Just because you’ve done some science — even some really good science — doesn’t mean that you’ve also done good philosophy. Again and again we see scientists wading into philosophy and claiming to have solved philosophical problems and when the philosophers point out that they haven’t really — and pointed out why — they’ve complained that philosophers just don’t want to accept the science and want to protect their own turf.
Demonstrating that all of our actions could be predicted with perfect accuracy by mapping the brain would, rather obviously, not refute compatiblist free will, because all that would do would prove determinism and materialism about mind, and compatiblists accept determinism and are generally materialists about mind. Thus, they think that even our experiences of consciousness will be determined and mapped to brain functions, which means that even the influence of conscious reasoning on our choices will be reflected there. So, again, even if science managed that it wouldn’t actually impact their philosophical view. But as pointed out above there are things that could. Anything that makes our conscious decision-making process epiphenomenal would do it, which is why the Libet experiments are so interesting to the debate, because if they work then there is reason to think that our conscious decision-making processes aren’t having enough of or perhaps any impact on our decisions … which is critical to compatiblism. Which, then, is how you refute it scientifically … if Coyne is willing to accept that as well.
So where does Dennett find free will? As he always has, he finds it in the notion that we are evolved, complex beings who reason: that is, we feel that we mull things over before coming to decisions about complex issues, and that this process of reasoning, which is an evolved part of our brain (supplemented with the environmental inputs of learning the consequences of actions), gives us free will. According to Dennett, it is this reasoning that makes us free, as opposed to decisions made when we’re constrained by other factors, like a person holding a gun to our head at the ATM and saying, “Take out $1000 and give it to me.” Without the gun, we would probably withdraw less money. The decision made at gunpoint, according to Dennett, is not “free.”
In other words, for Dennett free will lies in the ability to make reasoned as opposed to coerced choices.
While I think that Coyne is more muddled over free will than Dennett is, here is one case where Dennett might be. A person holding a gun to our head does not, in fact, impact our free will, in the philosophical sense. We are indeed still “free” to choose to not withdraw that money, and some people do. What this impacts is specifically our moral responsibility, as many people argue that that coercion is so strong that no rational person could possibly choose otherwise. But there being only one rational choice does not mean that our choices are so constrained as to be no longer free. All it means is that, well, there’s only one rational choice, and so all people choosing rationally will choose it. Note that because humans do not reason perfectly, even if we reason out our decisions we may still choose the decision that is not the most rational. Free will has to be encompass that, and if Dennett really is saying that that is not a free choice just because it is so rationally constrained then he doesn’t even maintain a free will worth wanting, and certainly not the one we have.
Also note that even the “moral responsibility” argument doesn’t hold. The Stoics would argue that one should never do anything immoral, even under threat of death. This is because they hold virtue to be the highest good, and treat death as an indifferent, something that you can have and enjoy but that you can’t put ahead of virtue. In the case of money, for the Stoics it would be pragmatic reasoning that would take over: you are choosing between indifferents, and so should choose the indifferent with the most personal gain. In the case of being asked to even steal for someone or else be killed, the Stoics would insist that you take the virtuous over the indifferent and let them kill you.
That was a bit of a digression, but I hope it clarifies what the debate is here: over morality, over whether or not that action is moral or not, and what moral consequences it has. It is still a choice you make, and still a choice that you can be held responsible for, if in no other sense than the one Coyne uses. It has nothing to do with free will per se; you are still expected to make the choice you make for reasons. It’s just that people say that the right, rational, and moral reasons have to all point to one and only one answer, objectively. But objectivity does not trump free will.
In other words, some people can make “responsible” choices, and those are the folks with free will. The others, well, they’ve been coerced.
I don’t take this as a fair summary of what Dennett is saying in the quote Coyne gives. I see it as more that we can be influenced by factors that result in us making sub-optimal decisions, but the fact that people can overcome them and that we can learn how to resist those factors proves that we do have free will, that our decision-making capacities really are, at the end of the day, the things that count, not these irrelevant factors. So it isn’t that people who don’t make those responsible choices don’t have free will, but that they have decision-making capacities that are flawed and influenced by things they shouldn’t be, and so should be corrected.
Of course, it’s possible that Dennett really does mean that … at which point he’d just be wrong.
I think this line of argument is bogus. There is no difference, I think, in being coerced by threats or social pressure, and being coerced by our neurons, which are in effect billions of tiny guns pointed at our head. You don’t have the ability to decide to “take steps to protect your autonomy”, for some people can reason in a way that makes them do that while others can’t. It’s not a free decision.
External influences short-circuit our normal decision-making processes to influence into making unreasoned and irrational decisions. Neurons implement our normal decision-making processes. That’s a major distinction … and the heart of the compatiblist view. Talking about some not having full responsibility in a free will discussion only muddles the matter, which I do think is something you can fault Dennett for. But Coyne has read and talked enough about free will to know better than to simply accept that.
Further, I think that members of some other species, like crows, elephants, and nonhuman primates, can reason and make “decisions” after some cogitation, even if their reasoning isn’t as complex as ours. Does that, then, make them “morally responsible”? If a dog attacks a human, mistakenly thinking that the human is a threat to the dog’s owner, do we hold that dog morally responsible? If not, why not?
Because, in general, they aren’t capable of moral reasoning, not because they don’t have free will. They are considered, generally, to be unable to choose actions because of their moral value and consequences as opposed to their pragmatic value and consequences. Some of them don’t even get above instinctive decisions, which are not free will decisions. Those that can understand moral reasoning — and some, particularly those who think that morality boils down to empathy, think that some animals can — could indeed be held morally responsible for their actions … but so far we haven’t found any whose ability to understand morality even approaches that of a small child, who are generally not considered at least fully morally responsible as well. So animals are not, in fact, any evidence against compatiblism.
In most cases people will indeed behave “responsibly,” for, after all, responsible behavior is behavior that endears you to society and enhances your well being. That’s precisely what our brains have evolved to do, as well as to process environmental information that is part of the evolved program. All we are doing when we make a decision is run a fixed computer program in our brain that has lots of different inputs, all of which yield a single output: the “choice.”
Which, oddly enough, is actually completely compatible with most compatiblist claims; most of them accept that a computer program with access to the right sort of information and the right decision-making processes would also have free will. There’s no way for them not to and remain consistent.
Some people’s decisions are better than others, and we say that those people are acting more “morally.” Others are “immoral”, perhaps because the reasoning process is faulty or because the reasoning process is sound but neglects important information. But in every case we’re running computer programs that have only a single possible output. How does that make us “morally” responsible? And where is the “freedom” in that?
Now, here is where things get a bit bizarre. Coyne thinks that we can have a meaningful idea of responsibility. Here, he also reveals that he thinks that we can have a meaningful idea of morality. Yet, somehow, he can’t conceive of there being a meaningful idea of “moral responsibility”? Moral responsibility would be nothing more than responsibility for actions that have a moral character, which in Coyne’s view seems to be decisions that, as he says above “… endears you to society and enhances your well being.” In short, exceptionally selfish reasons, and so if we tried to do moral reasoning on that basis it would all be about advantage. That being said, I don’t think this is what Coyne really means. But if he’s willing to accept that the term “moral” and “responsibility” have meaning, then he has to accept that the cojoined term at least potentially has meaning. Moral responsibility is not a technical term, like free will is — or has become — and really does mean responsibility for your moral actions. No more … and no less.
Coyne then goes on to do the direct comparison between Sophisticated Theology and compatiblism … but since he doesn’t really seem to understand either, it’s a bad comparison. Let’s go line-by-line:
Both redefine old notions (Biblical literalism or contracausal free will) and claim nobody believes in them any more. Like scripture is for Sophisticated Theologians™, so is free will for compatibilists: both have become metaphors for more recent notions.
Certainly false for compatiblism. They are indeed trying to capture what really is important in the free will debate, all of the things that we seem to have absolute evidence for but that people think determinism conflicts with … which is essentially meaningful choice. And they don’t claim it, but argue for it. Too many do rely on saying that “No one really believes in dualistic free will”, but the philosophical position is that dualistic free will is not the only way and is not the best way to capture choice, and the sort of choice that we seem to have.
The definitions of free will, like that of Sophisticated Gods, are concocted post facto, after compatibilists have decided in advance that their task is not to find the truth, but to buttress a conclusion they want to reach (i.e., we have free will)
No, the definitions of free will are adjusted to capture what we actually have, in light of determinism. Compatiblists are no more post facto reasoners than determinists, who often keep expanding what determinism allows when they run into cases that they can’t explain away. This is similar to the broadening of the definition of science to support scientism. However, compatiblists as philosophers are bound to philosophy and philosophical reasoning … and as such their definitions have to be justified. Simply deriding their definitions is not a philosophical refutation.
Both set humans aside as special—different from other animals (souls or free will)
Compatiblism need not set humans aside a priori, and often doesn’t. If it does, it is as a consequence of its view, not as an a priori requirement.
In both cases academic doyens (theologians or philosophers) feel that it’s dangerous for the public to know the truth (about God or about determinism).
Or, rather, because they think that it isn’t the truth, and that it is dangerous to promote a potentially harmful statement that causes radical behaviour changes for the worst without having a justified claim that it is true.
Both groups need some sense of free will to “sustain our sense of moral responsibility”
Yes, it is indeed a concern that we will no longer be justified in acting morally if the views are true. This is indeed a consequence that should concern us.
There are as many versions of compatibilism as there are conceptions of God (and no general agreement on them), so advocates can always say to critics, “you’re not attacking the best argument.”
Since this is a philosophical discussion that is still in progress, yes, this is the case. For compatiblism, most people will not be saying “You’re not attacking the best” but instead “You aren’t attacking mine while claiming to have refuted it.” That being said, Coyne in general has demonstrated a penchant for attacking views without understanding what the claims are, so take that with a grain of salt.
Both dismiss science as either irrelevant or inferior to philosophy for solving the Big Question at hand (free will or the existence of God).
Since most compatiblists are philosophical naturalists, they generally only do that when the science is irrelevant, and argue for it. Those who disagree can then argue against that conclusion. Coyne doesn’t do that. He simply declares that they are dismissing science because they don’t blindly accept its findings.
The puzzling thing to me is that I always have a hard time figuring out if Coyne is actually a hard determinist or a compatiblist. He continually tries to save traditional notions like responsibility and punishment, and proposes deterministic alternatives and consequences that compatiblists are perfectly fine with and follow just as easily from their view. Yet he continually rails against their definitions and accuses them of playing a semantic game, but it is very hard to see how Coyne’s objections to them are more than simply semantics, that he accepts their conclusions but just doesn’t like their use of the term “free will”. There needs to be a meaningful difference between the sides, and traditionally there has been. It’s only redefinition of the hard determinist side that has blurred the lines between the views … which then reminds me of the naturalism and scientism debates, which does the same thing. It’s what happens when people try to preserve their own views in light of competing evidence, and don’t have philosophical oversight on their definitions. Compatiblists have that, but where is that for populist hard determinist positions?
Tags: free will