Friday, October 21, 2011

Free Will

"Free will is probably located in the pre-frontal cortex, and we may even be able to narrow it down to the ventromedial pre-frontal cortex." --Stephen Pinker, How the Mind Works

I recently held a seminar on the subject of Free Will. The main point I tried to get across was that if the mind is produced by the brain (or that the mind somehow is the functioning of the brain), then free will is impossible. Others have of course come to the same conclusion, for example Sam Harris in his blog. But since free will seems to be necessary to make sense of almost anything that humans do  - hold each other accountable for our acts, for example - maybe we have to accept that there is something irreducible and nonmaterial to consciousness. This to me seem rather obvious. If the brain, which is a material thing ruled by material laws of cause and effect produces consciousness, then how can this consciousness be in any meaning free? Daniel Dennet, for example says: "There is only one sort of stuff, namely matter — the physical stuff of physics, chemistry, and physiology — and the mind is somehow nothing but a physical phenomenon. In short, the mind is the brain ... we can (in principle!) account for every mental phenomenon using the same physical principles, laws, and raw materials that suffice to explain radioactivity, contin­ental drift, photosynthesis, reproduction, nutrition, and growth."
And again Steven Pinker: "The mind is a system of organs of computation designed by natural selection to solve the problems faced by our evolu­tionary ancestors.” 

One of the participants said that he didn't see any contradiction between the notion that the mind is the brain, (or that the brain "produces" the mind), and the notion of free will. He suggested that even though material processes are determined by strict cause and effect, there is also randomness at work - random fluctuations on a quantum scale as well as macroscopic randomness of complex systems (as theorized in chaos theory). 
But this, to me, seems more to confuse than clarify the matter. Will is purposeful and intentional and therefore the very antithesis to randomness. And Free (in "free will") means unrestrained and is therefore an opposite of determined. All physical processes are seemingly a mix of random and determined, but how such a mix can give both freedom and will is beyond at least my ability to reason. Wether it is a predetermined chain of physical events in my brain or if it's a random release of chemicals in the brain, that forces me to choose tomatoe soup rather than mushroom soup in the supermarket, doesn't seem to matter. Both alternatives leaves me and my intentions out of the picture. 
And it also seems to me that there is something suspiscious to locate intentionality in the physical realm. If somenone tells me that this house or this forest (or this car or this computer for that matter) has ill intentions, I would believe the person saying this to be somewhat deluded. Because mental states do not have locations. Places don't have intentions. But if someone says that free will is located in the prefrontal cortex it's suddenly ok? It sounds like animism to me. 


  1. Sensei – Did you happen to read Eddy Nahmias's piece on free will, determinism, and neuroscience in the NY Times? He argues that the main problem is really not even a philosophical one: it's about semantics and definitions; how the question is framed. I won't repeat the argument here, but I found it thought-provoking.

    What I took home from it was that "is there free will?" is a bit of a dead-end question, whereas "how does free will work?" is a far more interesting one, and one that neuroscience can, perhaps, start to open up. I think the prefrontal cortex may have something to do with those answers.

    As far as I'm concerned, my answer to the dilemma is the same one I give to solipsists: I don't care if free will exists or not, since I will still continue to act as if it did.

  2. Yes, but I disagree that the question isn't philosophical. It is, and even semantics and their definitions are surely "philosophical" questions. "How it works" is to me far less interesting - being waht Chalmers have coined part of the "easy" problem with consciousness.

    I warmly recommend Raymond Tallis book "Aping mankind" on the subject.

  3. Sensei -- I expressed myself poorly. Certainly questions of definitions are philosophical question; however, an argument where the disputants are using different definitions isn't a philosophical debate, since they're talking past one another. Nothing meaningful ever comes out of that.

    Nahmias's contention is that this is what's happening between the neuroscientist/determinists and the people in Nahmias's camp. I.e., the determinists are using a definition of free will that is so narrow that it becomes useless. They effectively define free will out of existence.

    Nahmias, on the other hand, contends that there is a broader but still meaningful definition of free will that is not incompatible with determinism. That it's possible to accept that our decisions are both parts of a chain of causality, *and* acts of free will. I find that an intriguing idea.

    As to the "how" question, I find easier questions that are tractable to be more interesting than more difficult questions that may be intractable. I have a hunch the question of free will may be intractable by the methods and definitions used in science. Specifically, I think they're creating an artificial dichotomy between 'mind' and 'brain' here, and then denying that the 'mind' part is 'real.' That strikes me as a bit silly, indeed much like claiming that a house wills you ill.

    Personally, I find it more useful to treat metaphysical claims as axioms in phenomenological philosophies, than to attempt to resolve them once and for all. "Something exists, and we're part of it" is pretty good for lots of things, I find, and not too cumbersome. Whereas sometimes it's useful to think of some things as if they really existed.

    So I guess that just means I'm happy to file metaphysical questions, such as free will or consciousness, under "intractable by philosophical or scientific means."

    I have heard there are other approaches, though... ;-)

  4. Yes, I agree about how difficult it is to discuss something like "free will" without having a "clear" definition. I like Sir Arthur Eddingtons comment on the Uncertainty Principle in quantum physics: Something unknown is doing we don't know what. :)
    But I still cannot even begin to understand how there could be any meaningful kind of "free will" in a deterministic world. It simply doesn't make sense to me. To say for example that "free" means unrestrained by outside factors (like Gazzinaga in his "The Ethical Brain") would mean that a mechanical clock is free - free to do what it does, in other words - as long as no outside force interferes with it. This kind of "freedom" or "intention" seem unsatisfying to me. Or to say as Nahmias that we exercise free will "to the extent that we have the opportunity to exercise these capacities, without unreasonable external or internal pressure" again in my mind misses (or covers) the most imoportant point.
    And what is that?
    That in a physically closed and determined universe, every physical effect or state have a physical cause, and only a physical cause, leaving no room for anything nonphysical to interfere. Since physical states appear without intention (according to this kind of thinking) but rather happen because they have to and by necessity, it leaves little elbowroom for anything remotely like intention, will, belief, knowledge, and so on. And I don't believe that a physical structure (like the prefrontal cortex) can be ABOUT something else. This is why I prefer the "hard" problems that neuroscience/philosophy are facing. They point out the places where rational thougt becomes mute and the real adventure can begin.

  5. Sensei -- Let me try another tack. This is based on the point you brought up—the contention that neither randomness nor determinism are compatible with free will. I'll pick randomness, since I can make a thought experiment around it that illustrates how I see it. In my view, determinism has the same problem as randomness, with the difference that while you could prove that the universe is random (as in the thought experiment), you can never prove it's deterministic (at least not this way).

    This could get a bit long, so apologies in advance for that.

    Quantum physics has one rather scary possible implication that I'm sure you're familiar with—multiverse theory. The idea that every potentiality is an actuality in some universe; i.e., that universes diverge into an infinity of real universes at every instant in time. The proverbial cat is alive in some universes, and dead in others. As are you, or I, forever. (That's the scary implication.)

    Let's assume that this is true. Further, let's assume that some scientist—call him Tim—invents a device he calls the Higgs Resonator. The Higgs Resonator creates a connection between two universes that start to diverge at the instant the connection is created. It allows Tim to talk to his counterpart in the other universe through the resonator. So, to start with, there is only one Tim; a nanosecond after the connection is created, there are two Tims talking to each other, but they're only different in unmeasurable ways at the quantum level, whereas when the Tims return to the resonator the next day, their universes and their states will have diverged materially.

    Now, suppose Tim decides to solve this pesky free will problem once and for all. "I'll create 100 Higgs Resonators. Then tomorrow I'll phone my friend Sante Sensei and ask what he had for lunch, and post that information through the Resonators to all of the other Tims." He—and, naturally, most of the other Tims who started doing the same experiment for the same reasons—do exactly that. Then they publish the results in Nature.

    It turns out that Sante Sensei had mushroom soup in 61 of the universes, tomato soup in 33 of them, decided to skip lunch in four of of the universes, wanted to have tomato soup but spilled it and had to have Vasabröd with cheese instead in one, and said he went to McDonald's and had a Chicken McNuggets with barbecue sauce in the last one, but it turned out he was just pulling Tim's leg and really had mushroom soup.

    A random distribution. More or less what Tim would have seen had he decided to look at any semi-random physical process. (continued -->)

  6. Now, you just argued that neither randomness nor determinism is compatible with free will. Tim agrees, and considers the question settled—so-called "free will" is simply a matter of quantum probabilities resolving themselves and eventually showing up as macroscopic differences.

    But what if we asked each of the Sante Senseis in the 100 universes why they chose tomato soup or mushroom soup? Each one of them would believe he had expressed his free will in the choice. "I had tomato soup last week, so I decided to change and have mushroom soup." "I had tomato soup last week and it was really tasty, so I decided to have it again." And so on.

    Okay. Now suppose that we're not dealing with Sante Sensei's lunch, but with Saif al-Islam Qaddafi's choices following February 17 in Libya. Imagine Tim had created the connection that day, and had asked Saif about his doings. This time, it turns out that in 82 of the universes, Saif sided with Muammar and helped him crack down on the revolutionaries; in 9 of the universes he scarpered for Saudi Arabia, in 6 he mounted a coup against his father and entered into negotiations with the revolutionaries, and in 3 he successfully convinced his father to step down and go into exile in Venezuela, and in 1 of those universes he subsequently stepped down himself and helped the revolutionaries set up a national unity government.

    Random again. However, what should we think of Saif's actions in each of these universes? I contend that every instance of Saif is responsible for his choices. At least 82 Saifs should be put on trial for war crimes, whereas in at least 1 of them he should be hailed as a hero with uncommon courage and wisdom, and one of the founders of the new Libya.

    I won't go into writing up the reasoning if Tim had discovered that the only constant in all the universes he's talking with is Sante Sensei's lunch or Saif ul-Islam's political choice. In my view, it makes no difference: both Sante Sensei and Saif ul-Islam will have been responsible for their choices, even if it appears to all intents and purposes that their actions were 'determined' -- i.e., constant across all known cases. (Which is the situation we're in now, naturally, since we're only able to contact one universe.)

    "This is why I prefer the "hard" problems that neuroscience/philosophy are facing. They point out the places where rational thougt becomes mute and the real adventure can begin." -- Yes, I agree completely! The problems are intractable with those particular toolkits. The real adventure is, indeed, elsewhere.

    (Hat tip to Alastair Reynolds -- he wrote a sci-fi story where I lifted the Higgs resonator. It's from the collection Zima Blue, and it's named Signal to Noise.)

  7. Hi again Petteri! Nice that you have found this blog.
    My thoughts in it is surely random musings and not meant as "teachings".
    Still I want to put the proverbial finger on what I see as important/interesting in all this talk about determinism/randomness and free will/intentionality.

    So here I go again:
    First of all, I have my doubts about the multiverse theory of Everett. Henry Stapp gives a good critique of it in for example "Mindful Universe", and others have critizized it too.

    Secondly, I do believe that my choice of soup for lunch as well as most other choices that people intentionally perform during the course of a day, is made by PEOPLE that has the capacity to choose (or could have chosen otherwise). I believe that without this freedom any kind of "spiritual" practice is impossible.

    What I don't believe is that this (to me) obvious fact can be harboured inside a materialistic paradigm/worldview, as currently understood. Physical systems don't enjoy this kind of freedom, they do not doubt, and they do not have ideas, thoughts, fears etc. The current trend in neuroscience/phiosophy is not far from phrenology in it's sloppy use of fMri to show "where" love, hope, free will, belief in God almighty and other mental states "resides". To me this is akin to animism in the sense that it "animates" material things and structures. Free will, whatever that is, does not reside in the brain or any other structure, object, heaps of objects, assemblies of objects and so forth.

    Still it is.

  8. Sensei -- Good, then I divined your intent correctly. I didn't figure you intended these as teachings; your tone is rather different when you do. That's why I dared to comment in the first place! (And, truth be told, I am a bit intimidated as it is.)

    Re the multiverse theory: in this case, please treat it as a feature in a thought experiment rather than a posited 'reality.' I have no idea if it's true or not. I have only the most superficial understanding of QM and am totally unqualified to have an informed opinion about that. Whether we really live in a quantum multiverse or not isn't really critical to the argument. It just makes it easier to think about.

    The point I'm attempting to make is that as long as the universe is not completely mechanistic—i.e., not only deterministic but predetermined, so that you could predict any of its states if you had full knowledge of any of its other states—there is room for free will.

    I'm working on a developing this argument a bit further on my blog, actually, since it gets a bit long. I'll post it tomorrow, in case you're interested.

    In a nutshell, I believe that as long as you have a situation where there are potentialities, only some of which become actualities—there are "paths not taken"—there is room for free will. Whether these potentialities are actualities in other universes is immaterial; however, a thought experiment that lets us pretend to statistically sample some of these "paths not taken" makes it easier to think about it.

    I entirely agree with your second point, and about 90% with your third one. And I don't really disagree about that last 10%, as much as not really understand what's being claimed, or asked. Seems to me a lot depends on what definition of 'physical system' you're using.

    I do not know where free will resides, whatever it is; I do not know if it's an emergent property, an irreducible quality of the universe, or something else. In fact, I believe that any such metaphysical claims are fundamentally meaningless—like Christian fundamentalists going "Goddidit" at everything.

    I also believe that the question of free will and consciousness is fundamentally unknowable in terms expressible as conceptual structures. That's why I called these questions "uninteresting" -- not because I wouldn't like to know, but because I believe you can't know, in the conventional sense.

    You could dub this position "strong agnosticism" (cf. strong atheism—the belief that there is nothing other than 'atoms and the void.')

    Whether it is knowable in some other way I don't know either. However, I believe it is. Why do I believe that? Because of you, and people like you, and what I've read in Vasubandhu, and a few little tantalizing flashes I've had myself.

    So I return to face the wall.

  9. Maybe I'm stupid, but there are several things that are implied in this discussion without explanation:

    1. How does some non-material, non-physical realm solve the question of free will? Why it is assumed that this non-material, non-physical realm is nondeterministic? Don't reasons, motives, and desires that lead to intention and purpose have their own calculus? As far I can see, all those problems Sante-sensei sees in the material world are just lifted into non-material world. If we could pinpoint the non-physical, it would become part of physics and scientists would like to determine it's laws and free will would be removed.

    2. How does free will work with Pratītyasamutpāda (dependent arising). Free will implies agents that are able to make independent choices. Dependent co-arising seems to imply that every decision is made from circumstances.

  10. Vitali: "Maybe I'm stupid, but there are several things that are implied in this discussion without explanation"

    Ha ha. Don't pretend to be stupid! But anyway:
    1. Nothing like a "non-physical" realm is implied ere, at least not by me! I don't see what "non-physical" realm you talk about? I do not believe it is meaningful to even talk about such a "realm" except in the obvious way (intentionality, beliefs, knowledge, "aboutness" and so forth). The notion of a non-physical realm "solves" no problems except in the sense of accepting that there are basic questions the notion of "physicality" cannot solve or even get into. I do not assume that any realm (physical and non-physical) is non-deterministic, nor deterministic. (As my teacher used to say: "Don't assume anything!") Free will at least SEEMS to be free and at the same time intentional. Listen to Searle on Free Will. He makes the problem quite clear.

    2: Free will, in my mind, means that we can choose NOT to do something or other, by choosing otherwise. Pratityasamutpada puts this ability into context. Maybe our "freedom" isn't as free as we usually believe.

  11. Sante: "I do not believe it is meaningful to even talk about such a "realm" except in the obvious way (intentionality, beliefs, knowledge, "aboutness" and so forth)."

    This was not obvious to me at all. In fact I find it surprising that it's possible to believe in free will while not believing in the soul.

    The concept of free will is much easier to understand in the context of the soul essence. You just draw conceptual line and say that mental decision processes happening one side happen within the soul. Actions that originate from the soul are free will and that's it. Simple identification problem.

    How you make this work in Buddhist context is complete mystery to me. I naively assumed was that you draw the line between first skandha (rupa, material world) and the rest and call say that free will is the functioning of the remaining four skandhas. While these skandhas might be nonphysical, they are still mental processes and claiming that their functioning is free will does not seem very satisfying to me. It's just labeling. Buddhism for me has always seemed like mechanistic materialism without materialism (just with completely different laws to follow).

    My current understanding is that making choices requires evaluating the options available and that implies some kind of (possibly nonmaterial) cognitive decision mechanism behind it. Without soul that identifies the self, question of free will is just drawing the line that separates that particular function from the rest. If you don't care to draw that line, the question of free will becomes irrelevant.

    I have no idea where Buddhism draws that line or if it does it at all.

  12. "My current understanding is that making choices requires evaluating the options available and that implies some kind of (possibly nonmaterial) cognitive decision mechanism behind it." Good thinking, Vitali. But I am unsure that there is a mechanism there at all! I agree that the concept of free will is easier to understand if you posit a "soul". But why make it easy?! :)
    Complicaed is more fun, and possibly more accurate. And what exactly do we mean with a "soul"? To me free will is that which the Buddha called the "uncaused". It is in it's most fundamental form pure intentionality, or aboutness. It is undefinable.

    "How you make this work in Buddhist context is complete mystery to me. " try zen :)