Bryan Caplan  

Free Intentions

PRINT
Why no Kansas miracle?... Boudreaux on McCloskey...
I quite enjoyed Alfred Mele's Free: Why Science Hasn't Disproven Free Will (available for pre-order now).  It's a great exercise in the debunking of debunking. 

My favorite case: Many psychologists (and laymen) argue that consciousness is epiphenomenal.  In layman's terms, unconscious urges, not conscious intentions, are the real cause of your behavior.  The evidence: Some experiments where researchers manage to predict some trivial behaviors with greater than chance probability slightly before you experience a conscious intention to act.

Mele's strongest counter-evidence:
There's an important body of research on implementation intentions - intentions to do a thing at a certain place and time or in a certain situation.  I'll give you some examples.  In one experiment, the participants were women who wanted to do a breast self-examination during the next month.  The women were divided into two groups.  There was only one difference in what they were instructed to do.  One group was asked to decide during the experiment on a place and time to do the examination the next month, and the other group wasn't.  The first group wrote down what they decided before the experiment ended and turned the note in.  Obviously, they were conscious of what they were writing down.  They had conscious implementation intentions.

The results were impressive.  All of the women given the implementation intention instruction did complete a breast exam the next month, and all but one of them did it at basically the time and place they decided on in advance.  But only 53 percent of the women from the other group performed a breast exam the following month.

In another experiment, participants were informed of the benefits of vigorous exercise.  Again, there were two groups.  One group was asked to decide during the experiment on a place and time for twenty minutes of exercise the next week, and the other group wasn't given this instruction.  The vast majority - 91 percent - of those in the implementation intention group exercised the following week, compared to only 39 percent of the other group.

In a third experiment, the participants were recovering drug addicts who would be looking for jobs soon.  All of them were supposed to write resumes by the end of the day.  One group was asked in the morning to decide on a place and time later that day to carry out that task.  The other group was asked to decide on a place and time to eat lunch.  None of the people in the second group wrote a resume by the end of the day, but 80 percent of the first group did.
Neat stuff.  My main complaint: Mele never clearly says what I want him to say, namely:
Though their methods vary, all of these free will researchers are playing a corrupt game of "Heads I win, tails I break even."  They never tell us in advance what would count as evidence in favor of free will.  The reason, to be blunt, is that no matter how their experiments turn out, they will never announce, "The results confirm free will."  Yet Bayes' Law then automatically implies that no matter how their experiments turn out, the scientists should also never say, "The results disconfirm free will."  After all, if P(hypothesis | A)>P(hypothesis), then P(hypothesis | not-A)<P(hypothesis).

Upshot: The trendy "science" of free will is question-begging, probabilistically illiterate philosophy.
But I suppose that's asking too much.  Despite his failure to express my own position, Mele's written a fine book on a subject I considered all played out.  Well done.

P.S. Just noticed Mele has a whole book on effective intentions.



COMMENTS (48 to date)
Ryan Murphy writes:

Can't you say that your Bayesian priors are such that you shouldn't be able to frequently observe "Heads I Win"? I.e., each of these unique experiments that is replicable on the margin rejects free will. Each unique experiment that exists (or what we believe exists given all the scientists digging around in this area) that shows "Tails I Break Even" actually on the margin confirms free will.

Bryan, wouldn't you be willing to bet on the number of novel experiments appearing in top XXX science journals in the next XX years rejecting free will? If they number greater than that, it is evidence against your position, etc.

Dan S writes:

Every debate about free will suffers from this same basic problem: everybody just defines the term "free will" so as to be obviously true or obviously false, and then asserts that of course we (have/don't have) free will! And speaking frankly this post is no exception. Those experiments say nothing about free will. They just describe different psychological circumstances under which people are more willing to carry out plans.

As far as I'm concerned, when discussing free will, the question to ask is, is an agent capable of taking an action other than what that agent's brain state at that time would suggest? If you say no, then I don't care what cute way you've chosen to try to define it, you haven't actually made an interesting point about free will. And if you say yes, then that pretty much goes against everything we know about how the brain works right now. So I conclude that we don't have it, and we only have it to the extent that you choose to trivialize it away by definition.

In fact I would even take it a step further and say that the whole philosophical notion of free will is a barbarous relic of a time when we just didn't understand the brain, or causation, or physicalism, or determinism vs. quantum uncertainty. Heck I would go even one step further and say that a lot of philosophy is like that.

Barry "The Economy" Soetoro writes:

We can not possibly have free will unless you believe in the supernatural (aka having a soul). You can only have free will if you believe you have the ability outside of the laws of physics to make decisions. Otherwise it is mere chemical reactions of which you can not control.

acarraro writes:

I think the belief in free will is somewhat inconsistent with my other economic beliefs the author holds.

I would argue the rational agents in most economic models have no free will at all. If that's a useful assumption to make in a model, surely it would argue free will is not a useful concept.

It also seems slightly inconsistent with the believe in free markets. The invisible hand doesn't exist and yet it looks like there is some kind of guiding agent in its results. The parallel with people is that you don't need a metaphysical entity (the free mind) to generate order, but it just happens by itself...

If you believe economies don't need some overlord to control and guide them, why does your person need it? Why do you need something outside physical reality to control your mind?

Michael Stack writes:

Dan S and acarraro nailed it. Our brains are giant pinball machines - yes, we do what we want, but what we want is out of our control.

The oddest determinist argument: Any behavior that is caused is incompatible with free will. In other words, if you do something for a reason, you are not free. In other words, actions by people who have an real reason for acting, e.g., private-sector employment, are not truly free but the actions of performance artists who live on NEA grants to come up with pointless art are free.

Tracy W writes:

Dan S: but what if your brain state suggests taking two different, mutually incompatible actions?

acarraro: what makes you think that free will must be outside physical reality?

KevinDC writes:

I'll second what Dan S said.

I'm in general a very big fan of what Bryan Caplan writes, but his writings in favor of free will are probably the weakest arguments in his intellectual quiver. This is of course not to slight Caplan - we can't all be geniuses on every topic!

Most puzzling to me is Caplan's frequent invocation of introspection in support of free will. In fact, it was an argument from introspection which convinced me that free will was false. (Not that I had any choice in the matter :-P) In debating the topic with the people who have the colossal patience required to be my friends, I've also used the argument from introspection as a way to convince many of them that free will is an illusion. This argument proves far more persuasive than hashing out the finer points of determinism or quantum mechanics. I have no real idea how Caplan thinks the argument from introspection supports free will - because in the article I linked to above he doesn't provide any actual argument. He just says "I observe that I choose freely, at least sometimes; and if you introspect, you will see it too." That's not an argument, it's just an assertion. What would Caplan say to me and many others like me who, through introspection, came to believe free will was an illusion? I am left with little to go on with the case he presents.

I'm inclined to agree with Sam Harris, who said* in one of his lectures on the topic "It's not just that free will doesn't make sense given the physics of our universe. The case I am making does not depend on determinism or materialism being true. Free will is so incoherent as an idea, it's impossible to describe any possible universe where it could be true."

*(I'm taking this comment from memory, so treat it more like a paraphrase than an exact quote.)

J writes:

I still have no idea what "free will" is. If people could agree on a reasonable definition, the topic would probably be more worthwhile to debate.

In any case, I think it is good policy to treat people like they can control their actions. (Not even sure of the relevance to free will)

Since I don't know what "free will" means, and I am not even sure there is a consensus definition, maybe some other posters could answer this: If a person were a rational utility maximizing agent (based on some personal utility function), would the person have free will?

Follow up, why does it even matter if people have free will? I do not think (hypothetically assuming I have "free will") I would live my life any differently or treat people differently if they did not have free will.

I have written programs which determine the behavior of artificial organisms, of simple "creatures" living in a computer model. This program-writing experience has brought me face-to-face with necessary assumptions. It has affected my philosophy about free will, which seems to be the subject of this post.

Start with some simple assumptions about a living thing (it can sense, act, and must find food regularly) and the universe (it has resources for life). Further assume that you are the person writing this living thing's program. It will follow your program exactly. The living thing must decide, through the program you write, what to do in each cycle (in each moment) of its existence.

Your program has to deliver a decision about how to act even in circumstances where nothing suggests how to decide, in circumstances where neither sense nor experience offers any clue. Your organism will surely die if it sits idle for too long. So you will discover that you simply have to choose an act at random sometimes. The time when you must decide often comes before you have comforting knowledge and confidence. This necessity comes close to "free will" I suppose, though I admit that I have not studied definitions of "free will".

Switching subjects a bit, to the knowledge problem, I wonder: Do deniers of free will imply that there must be a plan, somehow encoded in genes or in the environment, determining all present and future actions? If such a complete plan exists, does that tell that the Soviets were right in what they were trying to do, and that Mises got it wrong in Socialism (1922)?

I have written more about this way of modeling life, economics, and philosophy, in a paper and a blog.

Tracy W writes:

KevinDC, I was thinking about what you said about free will being incompatible with the laws of the universe, and I started wondering "which ones?" I'm no physicist, but I do have an engineering degree and I can't think of a single law that's obviously incompatible with free will. Newton's Laws of Motion, Ohmns Law, Maxwell's Equations, Heisenberg's uncertainty principle: none of them say anything overtly about free will or determinism. I've forgotten even more chemistry but I can't recall ever learning something and thinking: "well that proves determinism. "
And while I've read a bunch of things by determinists they've never said which universal rules are incompatible with free will. They just assert that we live in a deterministic universe, which is assuming the conclusion.

Dan S writes:

Tracy W,

I'm not sure what you mean by that. Do you mean, for example, what if your brain says "go left" and "go right" at the same time? Is there evidence that that is possible?

Tracy W writes:

DanS: well my brain saying "go left" and "go right" at the same time would explain some of my more impressive outbursts of clumsiness.
More generally, feeling torn between two or more choices is a pretty common figure of speech.

AS writes:

"If you believe economies don't need some overlord to control and guide them, why does your person need it? Why do you need something outside physical reality to control your mind?"

Nailed it!

Psmith_in_the_city writes:

Those of you with worries about the definition of free will might find this page useful, particularly the introductory paragraph: http://plato.stanford.edu/entries/freewill/

Bedarz Iliaci writes:
Man has free-will: otherwise counsels, exhortations, commands, prohibitions, rewards, and punishments would be in vain.
Aquinas

It is incorrect to think of free will as a randomness introduced into an otherwise deterministic evolution. We need to distinguish between a physical cause of a physical event and a logical ground of a chain of reasoning. Man knows and reasons by the faculty of intellect and intellect presents to the will various options. The will is free to select a particular option. Again Aquinas

we must observe that some things act without judgment; as a stone moves downwards; and in like manner all things which lack knowledge. And some act from judgment, but not a free judgment; as brute animals. For the sheep, seeing the wolf, judges it a thing to be shunned, from a natural and not a free judgment, because it judges, not from reason, but from natural instinct. And the same thing is to be said of any judgment of brute animals. But man acts from judgment, because by his apprehensive power he judges that something should be avoided or sought. But because this judgment, in the case of some particular act, is not from a natural instinct, but from some act of comparison in the reason, therefore he acts from free judgment and retains the power of being inclined to various things.

acarraro writes:

I don't really see how you could have free will without believing in something outside physical reality.

I assume that believers in free will are stating something stronger than an indeterministic universe. I don't think the science is settled on that issue. But it would be more interesting to discuss that directly rather than the more nebulous concept of free will...

I think introspection is a terrible experimental practice. A good experiment should isolate external influences and be repeatable. That requires the system conditions to be observable and reproducible. That's clearly not true for the human mind, which is way too complicated and impossible to interfere with.

To me stating that you believe in free will is similar to believe you believe in a soul. I don't see the difference.

Also note that even a deterministic universe can be unpredictable if you believe it is chaotic and if you believe starting conditions are unobservable. At that point the distinction chaotic and indeterministic is very small.

Another point I'd like to make is that the absence of free will doesn't really invalidate morality or ethic at all... They simply become technologies to better improve the (subjective) quality of life of each individual.

Tracy W writes:

Acarraro: I don't really see how you could have free will without believing in something outside physical reality.

On reading this, it occurs to me to ask: how are you defining physical reality?

And it also occurs to me that I do believe in things outside physical reality. Such as mathematics. That there is no largest prime number is something that I (a) know, and (b) is independent of physical reality.

Though, even though there are clearly some things outside physical reality, I can think of no reason why free will, if it exists, would need to be one of them.

To me stating that you believe in free will is similar to believe you believe in a soul. I don't see the difference.

Soul: The spiritual or immaterial part of a human being or animal, regarded as immortal.
Free will: The power of acting without the constraint of necessity or fate; the ability to act at one’s own discretion.

Implication, there is at least one difference between souls and free will; namely that a soul is immortal, while free will doesn't need to be. One can believe in free will without believing that this free will goes on beyond when you die.

Dan S writes:

Tracy W,

Ha, I too would benefit from less clumsiness in my life. But I don't see why indecision, delay, taking more time to think, etc. are not themselves perfectly respectable brain states, so I don't think they are evidence that free willing is taking place outside of the normal physics of the brain.

acarraro writes:

Tracy W: I don't really believe maths influences physical reality. It's mostly a way to describe it.

I agree there are differences between the two concepts (soul and free will), but I think it's easy to see how one can evolve from the other. If you believe in something beyond the physical, it's easy to imagine it being untouched by death and decay (which affect all physical things).

To be honest I struggle to understand what believing in free will means.

Do you think you need free will to punish bad behaviour? If my car breaks down, I fix it or destroy it... I don't just say it's not its fault since it has no freedom of not breaking down.

Do you think you need free will to give value to human life? I think self-preservation is enough.

I could accept the concept that since the human brain is such a complex machine, it's almost impossible to predict its behaviour. So free will is a reasonable heuristic. But it would then follow that as my ability to observe and understand the brain itself improves we might develop better models.

I also feel that free will implies it would be impossible to build an artificial brain. I find that very difficult to believe. How could you have free will and simulated or artificial brains at the same time?

Tracy W writes:

Dan S: I agree with you that indecision, delay, etc are perfectly respectable brain states. As for free will and the physics of brain: as far as I know the two are totally compatible. I see no reason to assume that free will takes place outside the normal physics and respectable brain states.

Acarraco: if ease of imagination was a good reason to believe in something then scientific studies would be a lot cheaper.

Brian writes:

I find many of the comments against free will to be rather odd. Most of the comments seem to be based on not knowing how to define free will. For the record, it need not have anything to do ethics, nor with an immaterial soul, nor with violating laws of physics, nor with whether we have well-defined brain states.

Let's state a couple things that should be obvious to everyone.

First, pure determinism is impossible as the sole mechanism in mental decision making. At the very least, the existence of mixed equilibria in games require us to choose one thing over another without having any objective criteria for having a preference. As we know, indeterminism is inherent to the laws of physics also. The only question, then, is whether mental indeterminism is purely random (i.e., out of our control) or affected by something else inside us.

Second, free will as a concept seems somehow bound up with consciousness. We all observe that we each seem to be conscious of our own mental states. What IS consciousness? We don't know exactly. But we observe it going on in our own heads. If free will exists, it is somehow linked to this mysterious state we call consciousness. It is this state that gives meaning to "intention" and other related free-will type words.

So, what is required for free will? I think the simplest distinction is the following. If the mental states we refer to as consciousness are reflective only of mental decisions that have already been made by our brain, i.e., they only give the illusion of having participated in the decision-making process, then we would say that we don't have free will. On the other hand, if conscious introspection plays a role in creating the decision, then we would say we have free will. Note that this criterion does not force us to accept or reject free will by definition. It becomes a well-defined empirical question, appropriate to be addressed by scientific inquiry.

So what does science say? Well, many of our conscious thoughts do indeed appear to be mere illusions. That is, scientists often find that people self-report conscious decision making AFTER the brain has already made the decision. Score one for the no free will crowd. But, not all experiments support this model. Bryan's examples above are among the experiments showing that conscious introspection does precede and influence the final decisions made by the brain. That is exactly what we mean by free will. So, although we might not use free will to the extent we think we do, it is still a well-defined and important mechanism for human decision making.

I would note that it could hardly be otherwise. From an evolutionary point of view, it's hard to see how the mental adaptation known as conscious introspection could arise without providing a survival advantage to the human organism. And that advantage is only possible if conscious introspection causes the brain to decide differently than it otherwise would. In other words, the existence of conscious introspection as a mental tool already tells us that free will must exist.

Dan S writes:

Brian,

I agree that the phenomenon of consciousness is very strange and a philosophical puzzle. It seems strange that we're not "zombies." But with respect to free will, I think you're defining it into existence like I said above. If having free will means using introspection to make decisions, then...who cares about it? That is such a low hurdle to overcome that it ceases to become important or interesting. In any case, introspection is not something that takes place outside of the physical functioning of the brain. The brain is capable of thinking about itself, pondering its own existential angst, etc. etc.

I am reminded of Austrian economists who say inflation is high by just choosing to define inflation as growth in the money supply. You can do that if you like, but it goes against common usage and mostly serves to lump all the negative associations that people have with "inflation" with something that has a totally different meaning.

acarraro writes:

Brian,

I would argue that conscious introspection is a physical act in the end. It's not much different from walking. It's the result of changes in the physical state of your brain. I would say this is a reasonably well established scientific fact. It's like saying a computer program could run without modifying the physical state of the hardware.

There are hundreds of substances which interfere with your behaviour. We know that they do that by modifying brain chemistry. How is this consistent with free will? Since we can prove your behaviour is determined by chemistry in many circumstances, it seems to follow that it should be so at all times. Whenever the behaviour is predictable it's never here nor there. We cannot perfectly predict the weather but very few people would argue that it's determined by something outside physical reality.

I don't think physical indeterminism is a settled question. We just don't really know. It might be true or not. But free will requires much more than that. So I don't think you can actually dismiss the deterministic nature of consciousness. It would also be a matter of degree since we know that macro phenomena are reasonably deterministic and it's not clear where the boundary is.

The argument about mixed equilibria doesn't really convince me. We have reasonable evidence that a lot of human behaviour is determined by heuristics. So the lack of an optimal course of action doesn't stop us from following some arbitrary rule which is certain to give us a result.

Tracy W writes:

Brian: if we have free will, but wrongly believe that we don't, then we're likely to make worst decisions than we would otherwise make. See for example belief in free will predicts better moral behaviour
and
belief in free will predicts better job performance

And I don't know what you mean by "introspection" taking place outside the physical functioning of the brain. This seems very similar to your statements in response to me about free will and brain states, you seem to think that for some reason free will and introspection can't happen in the physical brain. But you never give any reason for this belief of yours. Nor does Acarrro, despite me asking him directly.

Tracy W writes:

Acarroro: why do you think free will would have o exist outside physical reality? You keep saying this, as does Dan S, and KevinDC said he thought that free will was inconsistent with the laws of physics, but none of you ever explain why you believe these things.

You apparently take these assumptions on faith. And expect everyone else to as well.

KevinDC writes:

Tracy W,

Sorry for the late response - I've been busy. I remember a time when summers were more carefree than they are now, but such is life.

You are incorrect to say that "KevinDC said he thought that free will was inconsistent with the laws of physics." I did not say this. The only mention of physics in my post was when I was quoting someone else's point that regardless of physics or determinism, free will is so incoherent as an idea that there is no universe where it could ever be true, regardless of what the ultimate laws of this universe turn out to be. You may notice, this statement from Sam Harris kind of means something totally different from the different statement you mistakenly attributed to me.

I am totally agnostic on the matter of determinism or indeterminism in physics, because I'm not a physicist. I know that many (perhaps most?) leading physicists believe that the laws of physics mean free will must be an illusion - Einstein believed so, and Stephen Hawking said as much in his most recent book. But there are other top physicists who would disagree. Put two of them in a room and let them debate it until they are blue in the face - I'd never be able to tell you which has the better argument, because I don't know enough about physics to adjudicate. I'd feel wildly disingenuous pretending I have any grounds to say which side of that debate was better supported.

I'm also in slight disagreement with Brian when he writes "Well, many of our conscious thoughts do indeed appear to be mere illusions. That is, scientists often find that people self-report conscious decision making AFTER the brain has already made the decision. Score one for the no free will crowd." Now, I'm not disagreeing on the science. It's been very well established that conscious awareness of decision making comes after the brain activity leading to the decision making. I just don't think this is a significant point. To slightly beat the dead horse I rode in on, the case against free will I found convincing was that the very idea of free will is incoherent. Even if brain activity associated with decision making was truly simultaneous with awareness, even if determinism was false, even if minds were non physical and separate from brains, and even if we all harbored an immortal soul which was the true source of our awareness and choices, free will still would have no leg to stand on.

This is also why I am unmoved by Caplan's objection that critics of free will don't give a clear picture of what would constitute evidence of free will. To my ear, this sounds very much like asking what would constitute evidence of a square circle, or a five sided triangle. It's a nonsense question.

acarraro writes:

As I said before, we don't know if reality is deterministic or not. Quantum mechanics doesn't completely invalidate determinism and even if it did, it not obvious the brain is macroscopic enough to make quantum uncertainty somewhat irrelevant.

Even if the universe is not deterministic, it can still be described in probabilistic sense.

In such a universe, freedom seems a meaningless concept to me. Your behaviour might be only predictable in a probabilistic sense as arising from unpredictable quantum events, but that's like saying the weather is free.

I mean where is the freedom? Your brain changes at all times, based on internal and external stimuli. These changes determine your behaviour. Even if such changes are random and not deterministic, there is not much space for freedom unless you think there is a difference between you and your brain.

I mean, look at Brian statement:

And that advantage is only possible if conscious introspection causes the brain to decide differently than it otherwise would.

This statement implies that introspection doesn't happen in the brain. Or that it's separate in some way. So either introspection happens outside the brain and physical reality or that statement is just illogical as it basically states that the brain is causing the brain to behave differently from how the brain would behave.

Tracy W writes:

KevinDC: my apologies for misreading you. I am feeling very embarrassed.

acarraro: I have no idea what you think quantum mechanics has to do with whether we can have free will or not. I can't think of a single law in classical physics that is inconsistent with free will. That doesn't mean there is one, but you still have given me no reason to believe that free will is inconsistent with physical reality, even within classical physics.

As I said before, you apparently take these assumptions on faith. And expect everyone else to as well.

Most of the rest of your comment strikes me as a bunch of non-sequiturs, given your failure to establish how free will is incompatible with classical physics (or any other sort of physics).

As for your interpretation of Brian's statement about conscious introspection, which you interpret as implying that Brian believes conscious introspection takes place outside the brain, let us imagine that Brian was talking about interpretation of visual information ("vision" for brevity), and had said

From an evolutionary point of view, it's hard to see how the mental adaptation known as vision could arise without providing a survival advantage to the human organism. And that advantage is only possible if vision causes the brain to decide differently than it otherwise would.

Would you say, therefore, that Brian was implying that vision, in the sense of interpretation of the light falling on our retinas, must take place outside the brain? Do you think that vision can take place inside the brain, and yet lead our brains to decide in a different way than they would if we were blind?

acarraro writes:

Classical physics is deterministic. So every event is caused by previous events. Classical physics assumes that a system state is knowable and since it's knowable it's predictable. We might not have the instruments to currently measure a system, but those instruments are possible. Each state has only one possible evolution.

There cannot be free will in such universe as everything is pre-determined. You can re-run reality as many time as you want, and you'll get the same result. I don't think this is very controversial to be honest. Which part do you disagree with?

I don't really see your point: vision is the obvious example of a stimulus coming from outside the brain. Obviously our brain reacts to external stimuli. That's its function.

KevinDC writes:

Tracy W,

Don't sweat it - we're all guilty of misreading from time to time. I've tripped myself up that way innumerable times.

Classical physics does indeed seem deterministic to me. In particular, Newtons Laws of Motion are strictly deterministic. A thought experiment along those lines which originated with (I think) Laplace illustrates that pretty well. Suppose that we were to freeze everything in the universe at this given instant. (Yeah, I know we can't actually do that, but it's a thought experiment, so play along!) Now, further suppose we had a computer or mind vast enough to know the position of every single molecule and atom in the universe; their exact position and momentum were fully known. If you know this, you can know with absolute certainty what the universe will look like in the next moment, and the moment after that, and the moment after that, all the way to the end of the universe.

At any given moment, according to classical physics, there is exactly one and only one physically possible future which can result from that moment. This is true all the way down to every molecule and atom in your body, including the ones which compose the chemical and electrical activity in your brain, which in turn form your thoughts and actions. To say that things could have happened differently than they did only makes sense if we are arguing that the antecedents of what happened had been different from what they in fact were, so if we say "it could have been otherwise," it reduces down to saying "we would have been in a different universe had this been a different universe."

As to the concern about why free will must come from outside physical reality - I don't speak for acarraro on this matter, but I think this syllogism addresses what he was getting at:

Premise 1: Everything which occurs in physical reality is consistent with the laws of physics.

Premise 2: Free will occurs within physical reality.

Conclusion: Free will is consistent with, and therefore subject to, the laws of physics.

So, if we are to argue free will is not subject to the laws of physics, then we must change premise 2 to say it must be outside physical reality. Otherwise, we have a contradiction on our hands, or we have to come up with some ad hoc explanation for why this is the one area of physical reality not governed by the laws of physics.

Searle and others have tried to argue for free will and/or dualism in various ways, but their arguments are extremely flabby. Searle in particular - I'm reminded of something David Friedman once said about Rawls, that he had no idea why people took Rawls' arguments seriously, other than the fact that they argued for a conclusion people wanted to reach. That's my assessment of writers like Searle, Plantiga, Swindburne, and Eccles.

Eh, I've rambled enough for now. Thanks to the blogosphere for indulging my musings.

Dan S writes:

Glad to see this debate is still alive and kicking.

Let me try to explain in a slightly different way, and I apologize in advance for being vague with terminology and my liberal use of scare quotes.

We have always, even in pre-modern times, been vaguely aware that there is this computer-like thing in our heads called a brain that takes in information from its surroundings, does some calculations, and then spits out some instructions to the rest of the body, and that is what determines our behavior. But that can be very troubling. People naturally asked, doesn't this mean that we are mere slaves to this wet grey blob in our skull? Do "I," as an agent, have the ability to break out of this brain-prison and "make my own choice" in the matter? Considering the ethical and metaphysical implications involved, this is a very important question.

And the answer that modern science has given us (perhaps it will one day be shown to be wrong, but for now this is what the evidence suggests) is that no, you cannot break out of your brain prison. In fact (and here's why I say free will is a barbarous relic), the very notion that there is a "you" separate from your brain is mistaken. The two are the same, and everything that you call "you" is tied up inextricably with the neuron firings in your brain. So sadly, you don't have free will in the sense that people really care about.

Now, to my great frustration, you always get people who come along and say that free will means you weren't coerced into your actions, or that free will means you used introspection to reach your decision, or free will means you're conscious, or any number of similar such things. And in all of those cases, I feel that people are simply redefining free will so that it loses all of its teeth. If your definition of free will doesn't involve breaking out of your brain choice prison in some way, then honestly why are we even talking about it? It really does not take a brilliant mind to figure out that people sometimes use introspection when making choices or that the brain is capable of meta-decisions or what have you. I don't believe that that's what's been inspiring people to think about it for thousands of years.

Brian writes:

Dan S.,

You say "But with respect to free will, I think you're defining it into existence like I said above. If having free will means using introspection to make decisions, then...who cares about it?"

Well, you should show me where I'm defining into existence. I think it's clear that I'm not. Let me be specific. We can create definitions of free will so weak that it obviously exists. We can also create definitions, such as you are doing by saying it has to be outside the brain, that make it obviously impossible. The third possibility is to develop a definition that is not certain to exist but is testable and needs empirical input to complete the test. This latter possibility is the only one that doesn't define into or out of existence, and that's what I've done. If the conscious introspection works exclusively ex post facto, then free will is an illusion. If conscious introspection plays a role in creating the decision, then we have free will. This is an empirical issue.

Now you may not like my definition. OK. Propose another that has the same property, namely that it's open to empirical tests. It should be clear that such a definition can't violate or be separate from the laws of physics, because then it wouldn't be empirical. I do disagree strongly with you that free will has to be some nonphysical entity. I don't see where anyone in human history has required that. Please note that although the notion of a soul is often linked with free will, the two concepts are actually distinct and one need not assume one to get the other.

Tracy W writes:

Acarraco, Dan S. I am afraid that you have misunderstood classical physics. Classical physics, like quantum physics, is empirical. The laws of classical physics come from observation, and often break down under certain conditions: eg Hooke's Law breaks down when the weight is large enough, Newton's Second Law of Force is inaccurate close to the speed of light, etc.

Laplace's thought experiment is circular, not empirical. It assumes its answer. It's not good physics - at least not until someone builds said sufficiently knowledgeable computer. It's entirely possible that an empirical test would find that Laplace's experiment was false.

To put it another way: whatever your opinion of free will, you presumably agree that on some occasions a brain takes in information from the outside world, combines it somehow with information already inside the brain and produces muscle movements as a result: eg moving an arm and hand to catch a ball. Process was information> decision> external movement. That process, as far as we know, is consistent with Newton's Laws of Motion and the laws of thermodynamics and Boyle' s law and every other classical law. I can think of no reason why adding that the brain freely chose to try to catch the ball would suddenly turn the information> decision> external movement into a violation of any classical law of physics.

Tracy W writes:

acarraro: actually I defined vision in this case as the brain processing information sent from the retina. Do you agree that some processing takes place within the brain, and this processing sometimes results in actions being taken that wouldn't've been taken otherwise?

Tracy W writes:

Sorry that should've been that "it's entirely possible that an empirical test would find Laplace's predicted outcome of his thought experiment false." Truth/falsity doesn't apply to the experiment itself.

For example if there is free will then obviously the computer would not be able to predict the universe precisely like that.

Using a thought experiment that assumes the result you're arguing for is a circular argument and thus can't prove anything.

Brian writes:

acarraro,

You say "I would argue that conscious introspection is a physical act in the end."

I agree with this. The experience we have of consciousness involves some specific slosh of chemicals and firing of neurons. I'm not sure exactly how this comment responds to my comment or how it argues against free will. Why can't free will exist and be physically based in the brain?

You also say "There are hundreds of substances which interfere with your behaviour. We know that they do that by modifying brain chemistry. How is this consistent with free will?"

How is it not consistent? Does it violate free will that something can hit my arm and break it, thereby making it unusable? Free will, whether physical or not, is something that acts on the decision-making part of the brain. If we alter the ability of the brain to operate or coax it to act in a certain way, how does that show that free will cannot or does not exist? I don't think that's any different from breaking someone's arm or locking them in prison. You are merely limiting the ability of free will to act, not showing that it doesn't exist.

Your comment that "Since we can prove your behaviour is determined by chemistry in many circumstances, it seems to follow that it should be so at all times" is just plain wrong. Just because something happens often, it has to happen in all cases? This is a logical fallacy.

Likewise, you say "I don't think physical indeterminism is a settled question."

Yes, it is. Quantum mechanics is inherently indeterministic. The only possible escape is to propose a "hidden-variable" theory, but such theories can't satisfy experimental observations unless they are nonlocal, which means that two events separated in space can influence each other instantaneously. Nonlocality would mean that events could never be entirely associated with a region of space, which would prevent you from claiming that cognition occurs only in the brain. I don't think you want to go there.

Finally, you say "The argument about mixed equilibria doesn't really convince me. We have reasonable evidence that a lot of human behaviour is determined by heuristics."

Well, you need to think about it more carefully. Much of what we do is heuristic driven, which means we're boundedly rational, but we still have the capacity for perfect rationality. There's no reason to think any finite set of heuristics could ever handle decision-making in all cases. Heuristics are helpful because they allow for quick decision-making, but they can't be the whole story.

Dan S writes:

Tracy W, for the record I do not agree with those classical physics thought experiments and I think you may have accidentally put my name on there instead of KevinDC.

Brian, I believe that my view was also empirically derived. That is, the question was whether the agent was capable of doing other than what his brain state would suggest, and we now know (or think we know) that the answer is no, and that there isn't really such a thing as the agent separate from the brain. I have no issue with people studying human introspection, I just don't see how using introspection means I acted "freely." In both cases I'm using my brain to make decisions, and I have no internal agent capable of overriding the decisions that the chemicals in my brain reach, because my brain is the agent.

I find it interesting that none of us is really arguing about brain function, we're all just arguing semantics. This is exactly why I think free will debates inevitably lead nowhere and just get us talking in circles. It's because it's an obsolete concept. We don't need it anymore; let's just get rid of it.

Brian writes:

"That is, the question was whether the agent was capable of doing other than what his brain state would suggest, and we now know (or think we know) that the answer is no, and that there isn't really such a thing as the agent separate from the brain."

Dan S.,

No, that's not the question at all. The state of the brain at the time of decision-making is, by assumption, a direct reflection of what the person/brain has decided to do. The question of free will is how the brain got into that state. Is the state produced by the action of a machine-like algorithm applied to the various inputs, or is it produced in some other way? We all have an internal feeling of self that seems somewhat distinct from the body, and we refer to the actions of this self as free will. Is this feeling just an illusion independent of the decision-making process, or is it fundamentally involved in the process? Again, free will is about how we achieve that final brain state and not whether the body is a puppet of the brain.

With regard to your view, you say it is empirically derived. I don't doubt that, but that's not what I asked. I asked whether your model can make any testable empirical predictions. Tbat is, how do we know whether your model is the correct one? Your model, as far as I can see, provides no testable way for us to determine whether free will is real or not, and that makes it merely semantic (as you said). My definition of free will IS empirically testable, which makes it more than just semantics.

Tracy W writes:

Dan S: you were right, sorry about that. I have an excuse this time, I was writing on my smart phone making reviewing hard. I did think of it as a response to KevinDC.

Anyway, no one is claiming that free will is a matter of overriding our brains, or somehow outside our brains. There's no reason to think that free will has to be separate from our brains.

Let's take a simpler case - you presumably have the sensation of seeing blue, or of hearing music, or of smelling a rose or of touching sandpaper. I'll talk about blueness, meaning the sensation of seeing blue, but the same logic applies to all sense experiences.

Now scientists can do a study whereby they put a bunch of people in MRIs, show them blue things, and non-blue things, record the brain's response, compare what areas in the brain light up and what don't, and say "Okay, that area of the brain is the one that sees blue". Those neurons deal with blueness. But the result of that experiment is not the same as blueness, it doesn't even look blue, unless the analysts colour-code the results.

Even if you could open someone's brain up and observe directly what was going on when they experience blueness, this would still not be the same as experiencing blueness yourself.

The most likely explanation for the disconnect between neurons firing in our brain and experiencing blueness is that blueness, and all other sensory experiences, are emergent properties from the chemicals and neurons of the brain.

Given that blueness exists despite not being obviously inherent in neurons and chemical reactions, is it really that implausible to think that free will could also be an emergent property of the brain?

acarraro writes:
If the conscious introspection works exclusively ex post facto, then free will is an illusion. If conscious introspection plays a role in creating the decision, then we have free will. This is an empirical issue.

This makes no sense to me. If conscious introspection happens in the brain and it's a function of changes in the brain how could it not be ex post facto? Where\when does it happen? There are obviously different parts of the brain which interact with each other. But it's all driven by chemical and electrical reactions.

Given that blueness exists despite not being obviously inherent in neurons and chemical reactions, is it really that implausible to think that free will could also be an emergent property of the brain?

What do you mean by blueness? The idea? I think it is implausible to believe that a complex system can suddenly stop being ruled by the law of physics. Unpredictability is not freedom.

Yes, it is. Quantum mechanics is inherently indeterministic.

Fine. It doesn't really matter for this argument. A random reality is not much better for free will. Even if we can only describe the future in probabilistic terms, there is still not space for free will. The universe evolves from one instant to another based on zillions of random dice throws. How can "you" impact any of that? Even the concept of "you" doesn't make much sense. The energy/matter which forms your body is not different from the rest of the universe. Is a rock free? It seems meaningless to me...

Introspection/consciousness seems like a reasonable evolution to me. Once you build the infrastructure to run simulations to determine your behaviour, it seems logical to run such simulation continuously in order to update your calculations... It seems quite similar to the monte-carlo engines I use at work every day. I don't think my programs are free in a way...

Tracy W writes:

acarraro:

If conscious introspection happens in the brain and it's a function of changes in the brain how could it not be ex post facto? Where\when does it happen? There are obviously different parts of the brain which interact with each other. But it's all driven by chemical and electrical reactions.

If deterministic decision-making happens in the brain, and it's a function of changes in the brain, how could it not be ex post facto? Where\when does deterministic decision-making happen? There are obviously different parts of the brain which interact with each other. But it's all driven by chemical and electrical reactions.

Your problem is a problem regardless of whether we have free will or we don't.

What do you mean by blueness?
I defined it in the comment you are quoting from. "blueness, meaning the sensation of seeing blue."
I think it is implausible to believe that a complex system can suddenly stop being ruled by the law of physics.
Irrelevant as we have no good reason to think that free will is incompatible with any law or any combination of the laws of physics. Be those laws classical, quantum, or whatever.
acarraro writes:

Deterministic decision making implies there is no free will. It might be true or false, but it has clear implications. I am not really sure how you can argue that. Does a computer program have free will? Deterministic decision making is the definition of an algorithm.

Whatever the laws of physics are is very much material to determine if free will exists or not. How could it be otherwise if free will is part of reality?

You keep saying that no set of laws can invalidate free will. That sounds close to assuming your conclusion to me.

I don't think there is any difference between the sensation of seeing blue and the process that goes in the brain. They are the same thing. Why do you think there is a disconnect? Watching neurons firing in a brain is like watching code execute in a debugger. It's not supposed to look or feel like it does to the subject, it's supposed to describe what is really happening underneath...

Dan S writes:

Tracy W and Brian,

This has been a great convo but I think it's time for me to duck out.

Bryan Caplan says lots of things I disagree with so I'm sure I'll be there for the next one.

Tracy W writes:

Acarroro:

Deterministic decision making implies there is no free will. It might be true or false, but it has clear implications. I am not really sure how you can argue that.

Good thing I've never argued that, then, isn't it?
Whatever the laws of physics are is very much material to determine if free will exists or not. How could it be otherwise if free will is part of reality?

You misunderstand the laws of physics. They are not handed down to us on stone tablets from a perfect authority. They are empirical - physicists observe the world, formulate a hypothesis, and try to test it. Hypotheses that survive a lot of testing get called laws. Often we later discover a point at which the laws break down.

Furthermore, you have not explained how any known law of physics, or any combination of these laws, does actually violate free will. You appear to take this on faith, and to expect me to take it on faith.

You keep saying that no set of laws can invalidate free will. That sounds close to assuming your conclusion to me.
On the contrary, I have never said this.

What I have kept saying is variations on:
" I can't think of a single law in classical physics that is inconsistent with free will. That doesn't mean there is one, but you still have given me no reason to believe that free will is inconsistent with physical reality, even within classical physics."

That's quite a different argument. Newton's Laws of Motion, Newton's Law of Gravitation, Hooke's Law, Boyle's Law, none of them strike me as being incompatible with free will. Perhaps I am missing some way in which they are, in which case, please explain it to me. (I request, for the umpteenth time).

I don't think there is any difference between the sensation of seeing blue and the process that goes in the brain.
Nor do I. What am I saying is that the process that goes on in the brain does not merely consist of chemical reactions and neurons that change in response to stimulus. If you look at the neurons firing in the brain when someone sees blue, that does not look like blueness.

Acarroro, you've not managed to accurately paraphrase here a single one of the arguments I've made, instead you've mistaken my argument every single time in this comment. Please read more carefully.

KevinDC writes:

Back again, to sound off on a few more musings.

Tracy W,

For the record, Laplace's argument is not circular - it does not assume the conclusion. It's a deduction from an assumption. Specifically, assuming Newton's Laws of Motion are true, then it follows from this that at any given state of the universe there is one and only one physically possible future which can result in the next instant. If there is only one physically possible future for each instant, there is never two more possible futures. Ergo, determinism. The soundness of the deduction, like all deductions, depends on the soundness of the starting premise. And in this case, we now know that Newtonian physics is a good approximation of reality, but not the "full story," so to speak. As such, the initial premise is false, so the deduced conclusion is unsupported. The idea that Newtonian physics is deterministic is not controversial. We just now know that Newtonian physics isn't a perfect description of reality.

Modern physics has a dose of indeterminism thrown into the mix. That's all well and good, but I don't see why people who think of free will as being a matter of physics would find any comfort in that. There is no more freedom to be found in randomness that there is in determinism, and there is no mix of determinism and randomness which makes "freedom" meaningful in any worthwhile sense. Stephen Hawking said as much in his book The Grand Design:

Quantum physics might seem to undermine the idea that nature is governed by laws, but that is not the case. Instead it leads us to accept a new form of determinism: Given the state of a system at some time, the laws of nature determine the probabilities of various futures and pasts rather than determining the future and past with certainty.

So even with quantum indeterminacy, Hawking says, "free will is just an illusion." Now, I'm far from an expert on physics, or the physics of free will, but I'm comfortable deferring to Hawking. Or, I'd be very hesitant to suggest he doesn't know what he's talking about, or doesn't understand physics.

But, as I mentioned above, none of this matters to my case anyways. We can all argue physics until the heat death of the universe, and my objection would remain untouched. My objection to the idea of free will has nothing to do with determinism, materialism, or physics. Free will, the idea that the "you" you take yourself to be, the conscious witness and narrator of your inner life, is the true source of your thoughts, desires, motives, and actions etc, is so incoherent an idea it's impossible to describe it in any way which makes sense or could be possibly be true regardless of the laws of the physics or the universe.

I get the sense that this discussion is losing steam or winding down, so I probably won't be back unless someone says something really interesting I can't help but weigh in on. But I do want to throw a shout out to Dr. Caplan. (I have no idea if he reads the comments on his posts or not, so this may be an exercise in futility. But hey, I'm a sucker for lost causes.) Bryan, you've said elsewhere that "Just as the solution to sloppy statistics is better statistics, the solution to sloppy introspection is better introspection." I fully agree with you on this notion, and I agree with the value of introspection generally. My suggestion to you is that your invocation of introspection in support of free will has been a case of sloppy introspection, and that better introspection provides far stronger evidence against free will. Even when I was (reluctantly) convinced that free will as an abstract idea was incoherent, I still continued to believe in it, because it just seemed like we all have this overwhelming subjective sense of deliberating and choosing. But with introspection, I eventually changed my mind. If nothing else, be aware that many people on the opposite side of this discussion were convinced by introspection that free will was illusory, so invoking introspection in defense of free will is wildly unpersuasive, even counterproductive.

I think that's enough of this for now. Ta ta.

Brian writes:

"This makes no sense to me. If conscious introspection happens in the brain and it's a function of changes in the brain how could it not be ex post facto? Where\when does it happen? There are obviously different parts of the brain which interact with each other."

acarraro,

I think what I've written is easily understandable, but since you don't understand it, let me try to figure out where you're losing it.

The issue here is decision-making by the brain--should I have vanilla or chocolate ice cream, for example. We can map out people's brains while they're making such decisions and see the moment the decision is made. People also have a subjective, conscious impression of when they've made the decision. When I refer to conscious introspection as ex post facto, I mean that the conscious state is created in the brain only AFTER the brain has already decided. That is, we would see the decision center light up, and then attendant state of consciousness would form. In this case, the brain would merely be communicating its decision to our conscious faculty, operating in a different part of the brain, and letting us know what the result was. In this case, consciousness would play no role in the decision itself.

An example where ex post facto wouldn't hold is where a conscious state forms regarding the problem at hand, we are aware of various options and of how our brain is processing the options, and then later on the brain lights up with a final decision. In this case, consciousness is a ab initio, not ex post facto, but even at this point free will is not certain. After all, consciousness could be just a spectator, watching the brain do its stuff but having no real effect.

What is required for free will is for conscious introspection to be both ab initio and to have an effect on the outcome. That is, if the decision-making part of the brain, working on its own, makes different choices than when it's being overseen by the conscious faculty, this would be evidence for the reality of free will. Or at least we would say that it's evidence for the "will" part of free will. The "free" part is guaranteed to be possible by the indeterminism of the laws of physics.

You are correct, of course, to point out that indeterminism can mean "randomness," and that randomness is not what we mean by free will. But it's worth noting that so-called randomness is a reflection of what we can know and not something inherent in the process. The flip of a coin is completely deterministic, but because we don't know many of the relevant variables, it seems random, and indeed is our primary model for randomness. When we see a process that is fundamentally unpredictable, like the decay of a nucleus or whether we choose vanilla or chocolate ice cream on a given day, we might call it "random" but it's better to call it indeterminate and leave it at that. Such indeterminism may be an example of freedom, or it may not be, but we have no way of knowing either way. The honest thing is to say that indeterminism allows for freedom but does not require it. Indeterminism means we can't dismiss freedom out of hand.

I think most of your difficulty in understanding how free will could exist comes from your model of the brain acting like a computer running an algorithm. You're thinking the brain is like some grand Turing machine. It's not. To be sure, the brain runs and implements many algorithms in its daily activities, but that's not all it does. It has a self-referential ability that Turing machines do not, and its decision-making is the result of many different parts cooperating and competing against each other. The brain really can be of many minds, and for any given decision those conflicts have to be resolved. My suspicion, and it's only a suspicion, is that the role of conscious introspection is to regulate and resolve the conflicts among various parts of the brain, and it does so in a way that is not reducible to an algorithm.

In any case, this is why Bryan's evidence supports the notion of free will. It gives an example where conscious introspection occurring far in advance of the decision-making process has a clear effect on the final decision. This is exactly what we expect in any physical model of free will.

Tracy W writes:

KevinDC: thanks for explaining about Newton's Laws of Motion. I was wrong on that point, I did not see the full implications, assuming Newton's Laws of Motion to be wrong.

Free will, the idea that the "you" you take yourself to be, the conscious witness and narrator of your inner life, is the true source of your thoughts, desires, motives, and actions etc, is so incoherent an idea it's impossible to describe it in any way which makes sense or could be possibly be true regardless of the laws of the physics or the universe.

That idea of free will also seems deeply implausible on an empirical basis - eg where does evolution fit in? The feeling of hunger seems much more adequately explained by evolution than by a "you" sui generis.

Comments for this entry have been closed
Return to top