Friday, January 01, 2010

Free Will Being Challenged

I have thought for a long time that "free will" is an incoherent philosophical concept. I'm not sure one can define it in any reasonable way. It is not simply the capacity for choice, because a machine flipping a coin would achieve the same result. So what is it? For the present, I will assume it refers to our feeling of being "in control".

We all have the sensation of being "in control", but how do we decide whether a biological organism other than us possesses free will? Does a bonobo have it? A dolphin? A cockroach? A bacterium? Can philosophy alone offer any guidance? I don't think so. Samuel Johnson once remarked, "All theory is against the freedom of the will; all experience for it." But we know that our common-sense experience doesn't always match up to the physical world, as in our strange system for perceiving color and how it can be fooled. So simply feeling that we have free will doesn't mean we actually do. Maybe we don't.

I think it quite possible that we lack free will in any reasonable sense - that, in fact, our actions are essentially deterministic. Despite this, I also think that our feeling that we are "in control" has a plausible basis -- I guess this makes me a "compatibilist", like Daniel Dennett. But I have a slightly different take on why, which is probably not original, but which I've never seen discussed in philosophy texts, although someone has probably done so. Namely, I'd guess that our computational hardware and software is so complex that it is not easy to predict the outcome of any situation with high probability - and in particular, we cannot even know how we ourselves will react in any given situation. We probably can do a simulation in principle, but in practice such a simulation would take too much time. So although we don't have free will in actual fact, the unpredictability of our actions makes it appear we do to beings with limited computational resources, such as ourselves. I'm hopeful that the theory of computational complexity may eventually play a role in a generally-accepted solution to the conundrum that has baffled philosophers for centuries.

The experiments of Benjamin Libet and co-authors cast doubt on our perception of being "in control". Libet found that subjects had activity in their brains about 300 milliseconds before they were aware of their volition to press a button. A more recent study found brain activity as much as 10 seconds before subjects were aware of their own conscious decisions. This popular article in Wired addresses it; for more technical details, see the article in Nature Neuroscience.

I was motivated to mention this by a recent solicitation to give money by my alma mater that mentions a freshman seminar devoted to these topics. I think it's great that cutting-edge research (the Nature Neuroscience article is from 2008) makes it so quickly to undergrad classes. And as we understand the science of decision-making better, more philosophers will be able to base their age-old speculations on some actual data instead of armchair thoughts.

28 comments:

Eamon Knight said...

W.r.t. the experiments about the time order of decision-making relative to concious awareness: I think it was Jonathan Haidt who recently remarked that, whereas we believe that our concious mind is the CEO assessing information and making rational decisions, actually it's only the spin doctor, rationalizing decisions made unconciously for reasons that are frequently anything but rational.

Paul C. Anagnostopoulos said...

Another reason for the feeling of free will might be evolutionary. It might be useful for an organism to have the feeling of control over its actions, so that it is prompted to plan ahead for future actions. Our feeling of will might feed forward into our future thinking.

And take a look at this paper:

http://portal.acm.org/citation.cfm?id=1245848

Anonymous said...

Oh, philosophers are fast to discount such reasonings with teleology, that having capability of teleological causation equals free will. I don't fathom the argument though.

ADHR said...

Odd you mention Dennett: he addresses how inconclusive those experiments are in Freedom Evolves. There's a lot of logical space between those results and the non-existence of even a robust sense of free will (let alone Dennett's modest compatibilism).

Whether free will exists is usually taken as distinct from how to account for the conscious experience of control. "Free will" usually means the sense in which our decisions make a difference to what we do. The conscious experience of control leads into the hard problem of consciousness (and the meta-hard problem, which I think you're averting to). Nastily-tangled areas of phil of mind and of action await.

Akhil Mathew said...

Interesting- this is the first time I've ever heard someone make that statement about "free will"--once which I've long wholly agreed with.

Filipe Calvario (from Brazil) said...

Some time ago, it seemed clear that the Deterministic vision of the Universe was right: if there are atoms, and everything is composed of them, and atoms obviously do not think, nor have free will, but would rather move in space/time according to certain laws, every event in the Universe, "including human cognition, behavior, decision, and action, is causally determined by an unbroken chain of prior occurrences." (Wikipedia - determinism).

However, with Quantum Mechanics, some doubt has been arisen, as the nature is starting to be described in terms of probabilities. When coming to bigger scales, the unpredictable nature of Nature would be manifested: no one would ever be able to determinate if some event just after one cause would happen ou not: one would have some probability only. Imagine, then, when one tried to analyze many events in a row: total fail!

Jeffrey Shallit said...

I agree entirely that it appears the universe is not deterministic, because of quantum events such as radioactive decay. What I am not so sure about, however, is whether these small events add up to genuine unpredictability of large systems, such as biological organisms.

Bayesian Bouffant, FCD said...

However, with Quantum Mechanics, some doubt has been arisen, as the nature is starting to be described in terms of probabilities.

Yes. And...? Quantum arguments for free will fail, because when people say they have "free will" they most certainly do not mean random chance and probability.

The same goes fro quantum arguments fro theism. A God indistinguishable from random chance is not a God worthy of belief, let alone worship.

Filipe Calvario (from BR) said...

Oh, I see!
I think your opinons, shown in the post, may match (I mean it in the sense of being the same) the ones in the following citation, from Wikipedia's article about determinism: "Some proponents of emergentist or generative philosophy, cognitive sciences and evolutionary psychology, argue that free will does not exist. They suggest instead that an illusion of free will is experienced due to the generation of infinite behaviour from the interaction of finite-deterministic set of rules and parameters. Thus the unpredictability of the emerging behaviour from deterministic processes leads to a perception of free will, even though free will as an ontological entity does not exist" Isn't it?
I think that this idea is very appealing to many people who like to think in a more "Mathematical way", including myself. For me, it makes a lot of sense, and may be true. However, there is some controversy about it.
This article, published in the Journal of Integrative Neuroscience, http://www.eecs.berkeley.edu/~lewis/LewisMacGregor.pdf , challenges this idea.
It says, in some part: "This paper argues, on rational and empirical grounds, that determinism
cannot be presently taken as a law of material nature EVEN FOR MACROSCOPIC SYSTEMS,
and therefore should not be taken, by monists, as a constraint on hypotheses concerning
free will."
It has 26 pages, and I haven't read it all, nor have I understood its arguments yet. But I think the authors share the interest in Math that we have, because they're from Dept.'s of Engineering of their University.

Filipe Calvario (from BR) said...

"Yes. And...?
And nothing, Bayesian Bouffant: I have never defended the Free Will hypotesis. I was just saying some stuff about determinism/indeterminism because I considered it to be pertinent to the matter.

-----------------------------

And my last comment ("Oh, I see! I think your opinions...") was directed to Shallit.

Veronica said...

William Provine argued that without free will there would be no means of blaming people for their actions. “Blame is useless,” he said. “It just creates a horrible system of criminal justice.”

He added that if society recognized the absence of free will, society would ultimately be much kinder to its less fortunate.


-- from http://cornellsun.com/node/22672

Not trying to prove anything; just sharing.

Miranda said...

I was thinking: When your wife or husband says to you, "I love you," doesn't it make a difference to you whether you believe he/she has free will or not?

Jeffrey Shallit said...

Miranda:

I'm not sure you understand the point of my post. It is that while it may be that at a fundamental level, we lack free will, nevertheless we appear to have it because of the computational complexity argument. So no, I am not troubled by the fact that my wife appears to have free will in exactly the same sense that I appear to.

Jeff Orchard said...

I agree with the complex-systems explanation. However, some people simply don't believe that consciousness and the feeling of free will could emerge from a complex system. It reminds me of Arthur C. Clarke's adage, "Any sufficiently advanced technology is indistinguishable from magic." In this case, any sufficiently complex system is indistinguishable from magic. Come to think of it, that's exactly the approach ID takes.

Filipe Calvario (from BR) said...

I'm starting to understand (lack of expertise coupled with non-native skills to a language kind of delays it) the article I gave the link - http://www.eecs.berkeley.edu/~lewis/LewisMacGregor.pdf

At a first moment, it tries to show that total determinism is in fact incoherent with the actual Universe, even in macroscopical scale - but recognizes predictablility in many systems, which it calls "pseudo-determinism" (not in a derogatory sense, I think). The Math envolved in the argument is interesting, worth appreciation.
At a second moment, it touches some stuff you (Dr. Shallit) have said.
You - "What I am not so sure about, however, is whether these small events add up to genuine unpredictability of large systems, such as biological organisms."
I have this doubt too.
The article, however, defends that whilst in most cases the macroscopic functions will be predictable, when it comes to the neuronal processes (from here on, NP), which is what is really important to that matter of free will, it (NP)is in fact most related to unpredictable events.
A quote:
"When one thinks of neuronal processes, one is immediately confronted with populations
of discrete elements, i.e., discrete g-protein molecules, discrete ion channels,
discrete ions, discrete synapses, and discrete packets of synaptic transmitter.
Physiological evidence suggests that these elements are frequently engaged in small
numbers during normal neuronal operation."

I think it's really worth reading. Please read it, if you can, and say what you think about it.

Miranda said...

"I am not troubled by the fact that my wife appears to have free will in exactly the same sense that I appear to."

I'm trying to understand this comment. It doesn't appear to address my comment.

Maybe I'll try it again, in a slightly different way. Let's say you hear "I love you" from your wife. You will think either:

A. She said that with free will.
B. She said that with no free will.
C. She said that with such a computationally complex system in her brain giving the /appearance/ of free will, but, at the end of the day, it was with no free will.
D. You don't analyze her thought process at all.

I maintain that "B" and "C" are essentially the same (with regard to your reaction to her statement -- "too bad she was compelled to say that" -- and that only with "A" and "D" will you come away with a real smile.


(Sorry if this post is a repeat; my computer started to act up.)

Garkbit said...

"I think it quite possible that we lack free will in any reasonable sense - that, in fact, our actions are essentially deterministic."

If our actions were essentially stochastic then that wouldn't be free will either.

Jeffrey Shallit said...

Miranda:

If I don't stop to think about it, then the answer is D. If I do think about it, then the answer is C. Either way, I get a smile on my face - so don't assume that something that fails to make you smile has the same effect on everyone else.

Miranda said...

Fair enough, Jeffrey.
Still, I don't see the difference (in the area we're talking about) between a computer program that has a one-line code that spits out a statement, and a billion-line code that spits out the same statement.

Anebo said...

I notice that there is an uptick on Atheist blogs questioning or even denying free will. But I find myself at a loss to understand it.

The experiments mentioned, suggesting that decisions are sometimes made on an unconscious level before consciousness becomes aware of it, does not suggest any limit to free will to me, because the unconscious doing the deciding is mine. Obviously we feel all sort of unconscious impulses also (hunger, desire, etc), but we decide (partly consciously and partly unconsciously) how to act on them. We also have to react to circumstances. But I don't see how any of this would impinge on free will.

Long before addressing your idea presented in this post that free will may be an illusion consequent to not being able to accurately understand ('simulate' in your terms) the process of decision making and its outcomes, I feel the need to address what seems to be a far more fundamental question in this regard.

If we don't have free will, what else could possibly be determining our actions?

The answer that seems to be on offer these days is one taken from Lucretius (though I don't see him sighted for it). A long time ago two atoms hit each other (the cline or swerve), and that everything else after that is inevitable. In fact Lucretius insisted that if the experiment were run again, and a whole new universe created (which he believed would happen), then everything else would happen just as it did. There would be another Homer who would write precisely the same words, there would be another Julius Caesar who would start a civil war and say again 'alia iact'est,' etc., just like running a film again. But that always just seemed silly to me on the grounds of personal incredulity, I suppose, but from what little I know about chaos theory, etc. (and I am sure you know far more), it just doesn't seem that kind of determinism could be possible, that you cold kick your foot in the water off the coast of West Africa and know beforehand that the wave you make would become a hurricane striking the Yucatan. I don't think it's a matter of not being able to do the calculations before hand either. Even if the initial conditions were known exactly, the way the parts in those initial conditions interact must be unpredictable on some macroscopic level.

Not to go on, but to try to reduce that rambling to coherence, I don't think that if we do not have free will, that god could be behind the determinism, or that random chance could be a deterministic force, so what else could be determining our choices except ourselves?

Filipe Calvario (from Brazil) said...

To Anebo:
"I notice that there is an uptick on Atheist blogs questioning or even denying free will."
I think that there is a (sometimes slight, but fundamental) difference between denying and questioning. I think that is very proper of the skeptical people to question every concept presented in any situation.
However, I agree with one thing: sometimes religious people will tend to ACCEPT free will, and non-believers will tend to DENY it, with no enough discussion.
Consider my case: I have, some comments before (please check it: "I'm starting to understand") proposed people to read an article which tried to demonstrate indeterminism in macroscopic systems, which may have a lot to do with the existence of free will. And I'm an atheist myself. Moreover, I did it with no scoff, some rigour to a comment in a blog and posted it a link to an scientific article.
Nevertheless, has anyone answered me?
As it is much easier to argue against dumbfounded or trolls comments, unfortunately many people would rather answer them, than answering comments made by intellectually honest people, who are interested in having a sane reasonable dialogue.

Daniel Abraham said...

I love this debate, but sometimes it feels like doing playing table tennis with marshmallows. I appreciate your post, and I think I understand what you're saying, but I don't agree with your definition of free will. I don't believe it's the *feeling* of being in control, but actually being able to choose whether or not to have eggs for breakfast.

If I understand the argument against free will as laid out by Libbet et al, if a decision is made before we're aware of having made that decision, there's no free will and we're all zombie-like automata who are too dim (or computationally simple) to notice. I feel a better interpretation might be that the choices we make aren't driven by our conscious awareness of them, and the process may spread across considerably more time than our experience of them suggests. I don't think anyone after Freud would be particularly blown away by that insight, but the confirmation is interesting and useful. But it doesn't address our ability to will, intend, guide our own futures, or direct our attention to something.

To say on that basis we have no free will seems to be either speculating past the data or using a straw man definition of "free will."

That consciousness (read: qualia) is a property of matter seems unquestionable. That will and intention should also be a property of matter doesn't seem any weirder.

Samantha said...

Does apriori brain activity make our choices no longer "free"? I think that our biggest problem in assessing "free will" both philosophically and biologically is a lack of cohesiveness in the definition. Some people think we can never have free will because all decisions are made based on our physiological selves and our past experiences interacting. Others argue that all of our choices are made out of free will because they are all self contained.

I think this research is interesting and promising, but I think it needs to be carried to further extremes before it can be considered evidence either way. We need to determine what the activity is actually doing, because being unaware of sorting out useful information for the making of a decision has far less impact than being unaware of making a decision in its entirety.

I would also love to see a study mapping brain activity when a sudden unexpected decision is asked for from a totally unprepared subject. Additionally, more complex decisions (i.e. moral decisions) should also be studied to see what kind time frame there is between first brain activity and awareness of decision making.

Still, an interesting addition to our knowledge of brain function that could definitely be expanded on.

Filipe Calvario (from Brazil) said...

This idea that free will may be a non-sense concept, independent of the truth of determinism/indeterminism is not new. In Locke's An Essay Concerning Human Understanding, Chapter XXI -"Of Power", we can read:

"12. Liberty, what.

As it is in the motions of the body, so it is in the thoughts of our minds: where any one is such, that we have power to take it up, or lay it by, according to the preference of the mind, there we are at liberty. [...] as soon as the mind regains the power to stop or continue, begin or forbear, any of these motions of the body without, or thoughts within,
according as it thinks fit to prefer either to the other, we then
consider the man as a FREE AGENT again.
[...]
14. If this be so, (as I imagine it is,) I leave it to be considered, whether it may not help to put an end to that long agitated, and, I think, unreasonable, because unintelligible question, viz. WHETHER MAN'S WILL BE FREE OR NO? For if I mistake not, it follows from what I have said, that the question itself is altogether improper; and it is as insignificant to ask whether man's WILL be free, as to ask whether his
sleep be swift, or his virtue square: liberty being as little applicable to the will, as swiftness of motion is to sleep, or squareness to virtue. Every one would laugh at the absurdity of such a question as either of these: because it is obvious that the modifications of motion
belong not to sleep, nor the difference Of figure to virtue; and when any one well considers it, I think he will as plainly perceive that liberty, which is but a power, belongs only to AGENTS, and cannot be an
attribute or modification of the will, which is also but a power."

I'm starting to agree with him.
-----------------------------
The whole work is accessible at http://www.gutenberg.org/files/10615/10615.txt

Paul P. Mealing said...

This is a very good dicussion.

There a couple of things I'd like to throw in that may interest people.

Erwin Schrodinger wrote a book called What is Life? which is a transcription of a set of lectures he delivered. In it, he puts forward an argument that quantum mechanics is more 'deterministic' than people think. He coined the term, 'statistico-deterministic'. The most recent edition of the book (2008) contains an epilogue called Determinism and Free Will and another set of lectures on Mind and Matter, where he gets unapologetically metaphysical. It's a very good book on philosophy and science and even religion.

My second point is that Roger Penrose challenges the conventional wisdom amongst AI experts that all our thoughts are computationally derived (refer Shadows of the Mind, 1994) with specific references to Turing's 'halting proof' and its close cousin, Godel's Incompleteness Theorem.

Regards, Paul.

Pseudonym said...

This reminds me of the JREF challenge. The overwhelming majority of the submitted proposals were completely incoherent about precisely what they were claiming.

That's part of the point of philosophy: We have an intuitive understanding of things like "justice", "knowledge", "consciousness" and so on, but it's hard to pin them down precisely. The world is not designed for neat human-created ontologies. It is what it is, and it's our job to try to grasp it.

So I think that Jeffrey is 100% spot on that "free will" is an inherently philosophical concept, not a scientific one.

But, whether or not it's scientific, it's still important. "Free will" has moral, ethical and legal implications, tied to the problem of responsibility. If "my brain is deterministic" is ever shown to be a scientifically accurate picture, it's only a matter of time before it's tried as a legal defence.

One thing I didn't quite get: What's the difference between a system based on true randomness, and a deterministic, self-modifying, complex, chaotic system which is constantly fed random inputs? And in what sense could you really call a machine "deterministic" if it's impossible, even in principle, to perform certain types of experiment on it twice and get the same result?

Thought experiment: If my wife's brain is deterministic, how could I reliably determine under what circumstances she would predictably ask me for a divorce?

Filipe Calvario said...

I think that, with Locke's argument I've cited above, it is possible to bring another answer to Miranda's "thought experiment" (!):
Jeffrey(if I may) would not say that his wife said "I love you" with free will, but he would rather say that she said that with freedom to will, and because of her will. Deterministic or not, he knows she wanted to do it and was not compelled by him or by other visible force.
If there is in some Universe a Laplace's Demon (http://bit.ly/rg29b) that could predict that she would say that, and say that there was no other possibility but that she would say it, it does not change the fact that it might be a sincere declaration of love.

Paul P. Mealing said...

There was something about this in Scientific American Mind. Scientists believe they've identified different parts of the brain where we 'will' the act and 'execute' it.

Regards, Paul.