Once again with free will: a question for readers

August 16, 2016 • 11:30 am

I swear, it’s harder to convince the average person that their behaviors and “choices” are determined solely by the laws of physics—and not by some kind of dualistic “ghost in the machine”—than it is to convince the average religionist that there’s no evidence for God. That’s because not only do goddies have a special reason to believe in dualistic free will—their attainment of paradise or hell depends on their free ability to choose—but all of us have a strong feeling of agency, as if we really could have chosen otherwise when making a decision. (The classic dilemma is a burger vs a salad at a restaurant.)

Let me define terms before I pose my question, a sort of survey of readers. And let me divide up people into three categories, A, B, and C.

A. Libertarians: Those who believe in a kind of dualism: that somehow our brains can really overcome the laws of physics and, were we to return to a previous situation of “choosing”, with every particle in the universe configured as it was before, we really could have chosen differently from how we did.

As I’ve shown before, a study by Sarkissian et al., surveying people in four countries, found that this is indeed the way most people conceive of the world: between 65% and 85% of people say that were they to return to an identical situation of choice, they could have chosen differently from how they did. (You can, if you wish, deny the Sarkissian et al. results, but they do match my anecdotal experience with people who have never discussed determinism and behavior.) Further, 60%-85% of people surveyed say that in such a deterministic world, people would not be considered “fully morally responsible” for their actions.

B. Hard determinists. (I am one of these.) Those are people who believe that our brains, being material objects operating under the laws of physics, can give only a single output from the inputs they receive (barring any quantum indeterminacy operating in our neurons). Our behaviors are solely and uniquely decided by our genes and our environments, and nothing else. There is no dualism, and if you returned to the “original situation” described above, you would always decide the same thing. We feel as if we are agents who could have chosen otherwise, but in reality we can’t. Hard determinists like me feel it’s pointless to talk about “free will.” Besides me, to other hard determinists are Alex Rosenberg and Sam Harris.

C. Determinist “compatibilists.”  Members of this class share the view of hard determinists that in a given situation, with all molecules configured identically, we can do only a single thing. As Sean Carroll argues in his new book The Big Picture (p. 295):

Under naturalism [Carroll’s a naturalist] there isn’t that much difference between a human being and a robot. We are all just complicated collections of matter moving in patterns, obeying impersonal laws of physics in an environment with an arrow of time.

The difference between members of this class and hard determinists is that the class “C” members think that determinism is compatible with some conception of free will, though of course not the version adhered to by libertarians.

How compatilists conceive of free will differs: some say our “freedom” is simply the complexity of the human brain, which allows us to weigh different inputs (“reasons”) before acting in a way no other animal can, even if those weights are simply aspects of our neurons existing in our brains. Others say our freedom resides simply in not acting under duress: a person cannot “choose freely” to go to the store if he’s locked in jail. (My response, of course, is that the bars of a jail are no different from the bars in our mind that compel us to do one thing rather than another.) Because compatibilists disagree on what constitutes “free will”; the only thing they agree on is that we can conceive of human actions so that we have something called “free will.” Examples of compatibilists are Dan Dennett, Sean Carroll, and our own reader Vaal, who has argued elegantly for compatibilism on this site.

Readers who have followed our discussions know my view: I fit into class B, and consider the difference between classes B and C to be largely semantic. If you want to call the complexity of human brain programming as “free will,” so be it, even though that’s not what most people think of it.  To me, it’s like saying to a Brit, “Okay, if you want to call a cookie a ‘biscuit,’ fine. They’re still the same thing.”  But of course others disagree.

You’ll also know that the reason I bang on about this at length—frustrating compatibilist readers—is because I believe that fully grasping determinism has a huge potential effect on human behavior, including in particular how we treat transgressors or criminals. It also has import in politics in general: e.g., many Republicans believe in their “just world” philosophy that many people are poor simply because they made the wrong choices. Such a philosophy makes no sense under determinism. Finally, we all surely agree that accepting determinism will sink the libertarian free will inherent in many religions, which I think is a good thing. You simply CANNOT freely accept whether or not to hold Christ as your savior, or Muhammad as Allah’s prophet. To punish people for eternity on the basis that they could have chosen otherwise makes no sense at all.

This is a long-winded preface to my question for readers, which I’ll put in bold. It’s this:

Philosophers squabble about the difference between classes B and C, whereas to Professor Ceiling Cat (Emeritus), a far more important argument is to be had between members of combined class (B + C)—the determinists—versus members of class A, the libertarians. To me, the latter argument, B + C vs. A, is of vital importance for making society better, while the argument between B vs. C is basically a semantic squabble that has an import on academic philosophy but not on society.

Do you agree with me or not? State you reasons. (Try to be briefer than I’ve been!)


509 thoughts on “Once again with free will: a question for readers

  1. Hi Jerry,

    To me, the latter argument, B + C vs. A, is of vital importance for making society better, …

    As a compatibilist I agree with you that A versus (B or C) is the more important argument.

    I think, however, that the harmful consequences of the libertarian view are mostly down to religiosity, and that trying to reduce religiosity would be the more effective strategy.

    1. I agree. Outside of religiosity, one absolutely needs the visceral experience of free will to operate in the day-to-day world. There’s no escape from it! If you want to commit 100% to reductionist / deterministic physics as your view of the world, you will end up a mystic, believing only things like “separateness is an illusion” and “all is one.” Best to save that for time sitting at home.

    2. Disagree, for a hard determinists to be tough on crime is totally unjustifiable.

      I don’t think we can make the same generalization about defenders of compatibilism.

      If we would live in a different world, I I maybe could agree with you.

      But the ambiguous use of the label free- will leads currently to a lot of unnecessary suffering in the world.

      1. for a hard determinists to be tough on crime is totally unjustifiable.

        How so? In C-type determinism the prosecutors, punishers etc. have no more choice than the criminal; to say we must stop thinking of the criminal as morally bad carries with it the paradoxical requirement that we must also stop thinking about the punishers of that criminal as morally bad.

        Moreover, there is nothing inherent in C-type determinism that says highly punitive responses won’t work. We have to determine empirically whether they do or not. They might not work (I agree with Jerry that they probably don’t), but the question of whether they do or don’t has zero to do with the question of whether we live in A-world, B-world, or C-world. Its solely about how human meat-machines are wired together, and we very well could be meat-machines that are wired so that we make our behavior more sociable in response to excessive punishment.

        1. A related problem is that if we must stop thinking of criminals as deserving blame or punishment, then logically we must also stop thinking of them as deserving compassion. So the claim that determinism necessarily leads to more humane treatment fails; one can just as easily cite determinism in support of capital punishment: the sensible thing to do with malfunctioning meat robots is to put them down.

          Obviously Jerry doesn’t agree with that view (and neither do I). But the point is that determinism isn’t the silver bullet he thinks it is for penal reform.

          1. Gregory,

            Yep, that’s what I bang on about here as well: consistency. The determinist has to be consistent. Everything prescriptive you say about the past applies to the present and the future, because it is all equally fixed in terms of being determined.

            IF you say that you should feel bad about a past action because “I couldn’t have done otherwise” that same logic applies to what you are doing right now, or plan to do.
            If you had choked a person because you were mad at him at the time, and you can relieve a sense of responsibility and negative feelings by saying “I couldn’t have done otherwise” then you can apply the same logic to make yourself feel less responsible, and
            less bad when your hands are around someone’s throat right now. Or in your plans to do it tomorrow.

            But…isn’t it an appropriate emotional reaction to feel bad that you did something unfair or hurtful to a person? Brought upon by understanding the reasons you really should not have done it? It would seem so. And that type of appropriate negative emotional reaction to “doing something wrong” is just what you’d want to arise to help dissuade you when you are doing something wrong, or considering doing something wrong in the future.

            1. Ugh…second paragraph should start:

              “IF you say that you SHOULDN’T feel bad about a past action because “I couldn’t have done otherwise”….

              1. Vaal, were you born a determinist or did you come to that conclusion later? 🙂

                Personally, having internalized (in)determinism have affected the way I view certain social arrangements as well as my own emotional states. I’m guessing such consequences are part of what hard determinists want to highlight.

                It could be that some people would feel less sympathy by not believing in free will but for the average person I don’t think that would be the case. I feel more sympathetic towards other when realizing we are all part of the same causal web and I still feel bad if I hurt someone. But it comes a point where further guilt is not productive, at least when you have wronged yourself by making a bad decision.

                Even if I were to accept the compatibilist position on the issue of free will I have a hard time believing I wouldn’t have less angst over past decisions, having realized we are just big collections of elementary particles interacting according to the laws of physics.

    3. I agree Coel. The most pernicious implications in dualist free will generally stay afloat due to the theist’s need to justify how God treats humans. If people can actually end up being put in hell, well then we must somehow DESERVE hell. If this world is so afflicted, well then somehow we must DESERVE all these affliction. Because the goodness of God can not be questioned.

      If you take God out of the scenario, then even if you keep dualist free will, you can still argue that, for instance, wherever possible forgiveness and rehabilitation is better than retribution.

  2. Often the distinction between (b) and (c) is whether or not one thinks a notion of moral responsibility is possible given the determinist premiss. Dennett, Fischer and Ravizza, and most “big name” philosophers have said yes. There are others, lesser known, usually, who deny this.

    In my view this is actually a false dichotomy, somewhat,as one could hold that moral responsibility comes in degrees in some relevant way. I once tried to see if one could do some sort of “self origination” a la Kane in degrees (without dubious quantum mechanics and just using neural system self assembly or spontaneous firing, a la Bunge), but it does fall foul to the regress going back to “not responsible for your character”. (It is this argument, the argument that one cannot be self-originating, that convinces me). I am not sure this sort of argument (a *correct* use of slippery slope?) is successful completely, however.

    Consequently I am (b) usually with the proviso that I sometimes use “moral responsibility” language as shorthand or as a means of being understood.

    1. To me, “moral” responsibility” just means “susceptible to the deterrent effect of social opprobrium” responsibility.

      PCC-E agrees that we can change people’s behaviour by social opprobrium. We’d only disagree on whether to use the word “moral” for that concept, or to instead grant the word “moral” to the religious.

      1. having moral responsibility simply means that sentient creatures learn.

        The useful question is not whether actions should have consequences, but what are the most effective ways of applying consequences.

      2. Take an example from the Inuit. Children are normally never address in (what look like) moral terms. For example, a European might say: it is bad for you to leave camp, you will be eaten by a bear. An Inuk would say (except to the intellectually handicapped and to visiting anthropologists) traditionally: “If you leave camp you’ll be eaten by a bear.”

        Is the latter “social opprobrium”?

        Also, it is a factual question, even if that is supposedly moral because it is social here, that this is the *best* way to change behavior, too. (And note the distinction between the Inuit and the European ways; this could been enough to make a difference consequentialistically speaking.)

        1. In case it is held that this is bad example, another one is:

          “Don’t hit your brother, it is wrong.” (or mean, or whatever.
          “If you keep hitting your brother, I (or he) will hit you.”
          (And then a follow up on the threat, if necessary.)

          1. The latter, maybe, but “rude” I suspect wouldn’t be used, given its moral overtone. But usually it is just stating the result of an action, so even ‘stop’ is unlikely. So, in your case my prediction would be “If you keep fighting with your brother, I will fight you!”

            1. Surely that last sentence implies disapproval by the father (and the boy would surely realise that) — and that is all there is to what I called “social opprobrium”. “Moral” or “immoral” acts are simply those that other people might approve or disapprove of.

  3. I’m in class C: Compatibilist. Hard determinism is a category mistake. Reducing human behavior to the deterministic interactions of elementary particles has no explanatory power, even in principle.

    1. In principle, of course it does, it has more explanatory power than anything else. In practice, you can’t process or have access to all that information so you think in simplified terms that have less explanatory power in principal but are perhaps the best one can do. That however does not make determinism untrue and it doesn’t make your actions or will “free”.

      1. Particle physics can (in principle) give a complete description of the history of, say, a particular mass of bronze. But knowing that history doesn’t enable you to generalize to an explanation of why people erect statues of war heroes in public squares.

        In answering questions of the latter sort — why people engage in certain broad categories of behavior and not others — particle physics is of no help whatever, since it doesn’t even recognize the existence of such categories. All it can do is track the trajectories of individual particles in specific cases.

        1. Knowing the history of all bronze etc. would exactly enable you to say why the collections of particles we call people build the collections we call statues, and also why they don’t when they don’t. It’s far *more* general.

          Particle physics doesn’t *need* such categories as “a statue” but that doesn’t make statues a mystery to physics. Whatever you arbitrarily decide delineates the category “statues” in terms of particle arrangements, a complete particle picture tells you what the things in that category (and every other) do. You “simply” translate back into your higher-level categories if you want to. (Although doing so loses information, so for example the best description of a cause may not reduce to your simple categories of “a guy made it ‘cuz he wanted to”.)

          This is perhaps easier to understand with a simpler example. Does a sail move because the wind hits it? Sure, but that is just the Cliff’s Notes Jr. version of it. The more complete explanation is that the collection of particles you call the sail and the one you call the air all interact with each other according to the general laws of physics and the final configuration of those particles still fits in the categories you called the sail and the air but with the gross displacement of those clumsy summarizing terms.

          In no way do I suggest trying to actually process your life this way, but it is the underlying truth to the best of our knowledge and that is the relevant point when asking if things are determined.

          1. The point is that knowing that things are determined is not the same as explaining them. Determinism, if true, is simply a brute fact about physics, not a theory of human behavior.

            I think you would be very hard-pressed to come up with a robust definition of “war memorial” purely in terms of particle configurations. Any attempt to do so would necessarily include not just the physical substance of the statue, but also the brain states of the people who commissioned and created it, and the thoughts and intentions represented by those brain states. Expressing all of that in purely physical terms would not just be massively impractical; it would impede understanding by radically obfuscating those features of a war memorial that are most relevant, namely its conceptual and symbolic attributes. So I don’t see how that sort of reductionist description can be said to have greater explanatory power.

            1. Does the process which leads to the construction of the war memorial completely reduce to physics or doesn’t it? This is the reductionist question, beyond the more abstract determinist one. Will an ideal simulation beginning at the big bang or somewhere more recent end up showing the creation of the war memorial? If so, then all of the psychology and sociology is shown to be a bit flimsy…

            2. Agreed: use the level of abstraction which is appropriate. E.g. don’t try to debug code using an oscilloscope.

            3. “The point is that knowing that things are determined is not the same as explaining them.”

              So first, do you now concede that Stephen Barnard’s claim that hard determinism is a category mistake is false, since you now opt to separate determinism and explanation?

              “Determinism, if true, is simply a brute fact about physics, not a theory of human behavior.”

              It would also be a fact, though it’s not brute, about human behavior. But if you know determinism is true that tells you that an explanation exists. In this case, physics is the theory of human behavior, determinism is a quality of that theory. (I’m tabling questions of quantum mechanics and how well we actually know our theory is deterministic.)

              “I think you would be very hard-pressed to come up with a robust definition of “war memorial” purely in terms of particle configurations.”

              That doesn’t matter, it’s not my job to do it, it’s yours to explain why it could never be done. You would be hard pressed to robustly define any abstract or concrete thing, regardless of doing it in physics terms. More importantly however, you miss the point in the following:

              “Expressing all of that in purely physical terms would not just be massively impractical; it would impede understanding by radically obfuscating those features of a war memorial that are most relevant, namely its conceptual and symbolic attributes.”

              You wish to focus on the “most relevant” attributes and details would “obfuscate” them because your brain can’t handle the full explanation. (Neither can mine.) But I am talking about the explanation in the abstract, not your ability to grasp it. You want the executive summary, which is suitable for many cases, but it is not the *best* or *full* explanation.

              The reductionist picture explains, in principle, every feature of the summarized picture, but the summarized picture can’t explain all the details. Hence the reductionist picture has more power. The “conceptual and symbolic” attributes are ultimately relations between patterns of particles in our brains. We have limited ability to say why the memorial moves some people and not others, or why it’s specific shape was chosen, or why we think of it at certain times in terms of human-level feelings and conceptual attributes. Those terms simply run out of explanatory steam. They will always be *at best* approximate, vague and incomplete.

              1. Joshua, we seem to be talking past each other, and I think it’s because we take “explanatory power” to mean different things. You seem to think it means accounting for all the details of a specific case. I think it means the ability to generalize to other cases. If we want a general theory of war memorials, understanding particle interactions won’t help us, not because we’re not smart enough, but because the necessary concepts of war, sacrifice, honor, etc. aren’t available at that level. That’s what I mean when I say that discarding high-level concepts in favor of particle descriptions obfuscates rather than enlightens.

                Now perhaps you’re saying that understanding all the particle motions necessarily implies all that other stuff, that Laplace’s demon is not just a wizard at physics; he’s a polymath who understands literally everything there is to know, simply by virtue of knowing the low-level physics in detail. I don’t think that’s a reasonable conclusion. The understanding of the higher-level stuff doesn’t come for free; you actually have to construct workable theories at each level, and doing so necessarily involves pruning away some of the lower-level details in order to gain generality and (my sense of) explanatory power. After a couple of steps up the ladder, an understanding of physics is pretty much irrelevant; our politics didn’t change much with the transition from classical to quantum physics.

                Yes, reductionism lets us account for more detail than generalization does, but it’s generalization that allows us to recognize and characterize broad categories of similar phenomena, and that’s how science advances.

              2. @ Gregory
                Whoops, didn’t mean to have my full name show up! Anyhow, apologies for the length, but…

                “You seem to think it means accounting for all the details of a specific case. I think it means the ability to generalize to other cases.”

                Not really. If there are details that can be explained but that aren’t on your model then you don’t have a full explanation, that’s true; but I also agree that an explanation is only good if it generalizes. That is, ‘A because B because C ad infinitum’ is not a good explanation if each step only “explains” the next. Rather B should explain A, and A’ and etc. so that we are reducing the number of parameters to explain all the data. However, you don’t seem to grasp that the particle picture (for example) is the *more general* picture. It applies to all war memorials, and all people, and all memorials period, and all planets and etc. It is you who want to specify some subclass of all particle configurations and say “how does everything in this particular class behave”. But the general picture has already told you in principle; not one specific case but all specific cases.

                Contrary to your claim, this reductionism is the heart of scientific advancement. Newtonian Gravity is a breakthrough not (just) because it gave us a more detailed prediction of planetary orbits, but because it recognized that the motion of all planets and all earthly objects could be explained by the same principle. Realizing that planets and monuments and people are ultimately made of the same stuff acting under the same set of rules and therefore explained by the same set of rules is probably the greatest insight humanity has ever gained.

                “…the necessary concepts of war, sacrifice, honor, etc. aren’t available at that level.” No, this is you assuming that an explanation for something must be in terms of nearby concepts in your everyday thinking. It’s like saying you can’t understand an atom in terms of nucleons and electrons because the term “atom” doesn’t exist at that level. The truth is the atom exists only as an approximation of the fuller picture. It is not needed at the deeper level, but you can translate back to it if you want your outcomes in terms of atoms.

                “The understanding of the higher-level stuff doesn’t come for free; you actually have to construct workable theories at each level, and doing so necessarily involves pruning away some of the lower-level details in order to gain generality…”

                Again, you seem to be confusing practical science with explanation in principle. For many purposes it’s practical to use atoms and the subatomic details are pruned, but that is only because you can’t follow the full details in practice, which is a statement about our cognitive faculties, not about the nature of the universe. The thing is, you aren’t generalizing the rules, you are just blurring the specific cases into approximations. You say “everything that is close enough to [this] behaves close enough to [that way]” by ignoring details. But that is not more general than the full theory, which implies the same thing *and* gives you much more. As you say, we can build many levels of this, and that’s absolutely useful given our limitations. However, you’re not gaining generality, you are losing it in order to handle a manageable complexity.

                Simple example: I have two ideal spherical particles bound by a spring. Given their initial positions & velocities I tell you where they end up and by what path at a certain future time. You say that’s too hard. You take the net momentum and tell me how the center of mass moves. Have I failed to explain anything you have? No, but I know and explain things you don’t. Is your rule about how centers of mass move more general? No, it’s merely a consequence of a specific calculation in my rules. Have my rules failed because they don’t require me to talk in terms of a center of mass? No.

                “After a couple of steps up the ladder, an understanding of physics is pretty much irrelevant; our politics didn’t change much with the transition from classical to quantum physics.”

                There is no reason the high level behavior should change just because you understand the fundamental level. And, again, understanding that the fundamental level exists and is in principle the explanation in no way implies that you can practically calculate with it and thereby use that knowledge. However, if you actually follow science you know that once the limitations of a “level” are reached the only way to improve your predictions is to go a level down. That’s why biologists now worry about molecular biology, why chemists started thinking in terms of electron orbitals, why nuclear physicists worry about gluons, and so on.

              3. Josh, I don’t have time to pursue this today, and it seems we’re not really getting anywhere anyway. To me, it’s pointless to talk about “explanation in principle”; an explanation must above all be useful and promote understanding of the phenomena it describes. Catalogs of particle trajectories fail to do that for anything above the level of particles, but if you insist they count as “explanations in principle”, I won’t argue the point further.

    2. I call straw man but I forgive you because you didn’t have any choice. In fact if we rolled back the tape and you made your post again it would be the exact same mischaracterization of the Hard Determinist position.

      1. How have I mischaracterized the Hard Determinist position? I didn’t mean to. Would you care to state the Hard Determinist position in ten words or less?

    3. It seems to me you could just as well say that it is the compatibilists that make the category mistake.

      I think that for most people the question of free will is not a descriptic-behavioural one concerning which level of description is the most explanatory. It is a philosophical question about whether or not we are the source of our own willings.

      1. Is it meaningful to think and talk about people’s thoughts, desires, motivations, etc.? And if it is, how would a complete and accurate microscopic description of the universe shed any light and these things? It wouldn’t. That’s the category error.

        1. Sean addresses this in his book.

          First, the emergent theory must be fully consistent with the fundamental one. If you had a version of the Ideal Gas Law in which the kinetic energy of the system was not conserved, you’d have a problem. As such, the “free” part of “free will” is highly suspect from the get-go, because it contradicts the fact that we know that all of physics really is deterministic.

          There’s another part of it that Sean addresses. You can’t mix and match your descriptive language. You can describe human beings either by listing all their constituent subatomic particles with their positions and momenta or as conscious agents with passions and dreams, but not both at the same time. The particles don’t have consciousness, passions, and dreams; and the consciousness, passions, and dreams don’t have submicroscopic positions and momenta.

          “Free will” fails that test, too; “freedom” would be a property of the underlying physics (which doesn’t exist), and the “will” is clearly a property of the conscious agents.




        2. “Is it meaningful to think and talk about people’s thoughts, desires, motivations, etc.?”

          Sure it is, in the same way it is useful to talk about a dune moving across the desert. But knowing the behavior of every particle of sand in the dune would tell you almost infinitely more about what is going on then lumping it all into “the dune”. There is no category error here, just the unsupportable assumption that thoughts, desires and such are best explained in terms of thoughts, desires and such.

          1. The only difference is that when humans use words like thoughts, desires, motivations, they mean something subjective, not objective. A dune or a brain can be understood in reductionist terms, but we don’t presume there is subjective experience “going on” in them. The problem of other minds is especially acute in reductionism / determinism because then you are confident of a completely different model to explain the behavior of a walking talking human.

            1. Yes. That’s why a microscopic description of the universe has no potency when it comes to thoughts, desires, and motivations.

      2. To answer your philosophical question, you first have to understand what the self actually is. And Sam Harris has done a superlative job of demonstrating that the self itself really is an illusory construct. When you try to look inward to find your “self,” what you find is not what you think you’re looking for.

        Try it right now. Just sit still and pay attention to your thoughts. All sorts of random thoughts will pop into your mind. Where do they come from? You’ll start to analyze those thoughts in some manner. What part of “you” is analyzing them?

        What are you, really? Are you the thoughts themselves? Are you that which is thinking the thoughts? Are you that which is somehow aware of the thoughts you’re thinking — and is that not just another thought in and of itself? Are you the framework that manipulates the thoughts? And what are these thoughts, anyway? And when you come to a conclusion, what part of “you” is it that’s concluding, and what’s the significance of the conclusion? Given the exact circumstances that held at the moment you reached the conclusion, could you really have reached a different conclusion What does it mean to your concept of “self” when circumstances change and you later come to a different conclusion? Are you the same “self”? If so, how could that be; if not, what happened to the old “self”?

        It doesn’t take much of this sort of introspection to come to viscerally understand that, yes, we really are Rube Goldberg contraptions unfolding like clockwork — the exact antithesis of everything that the very confused religious and philosophical attempts at defining the self-contradiction of “free will” is supposed to be.




        1. Hello, Mr. Goren.

          Philosophers usually argue that the self is identical to some psychological feature, or some physical feature, or some combination of psychological and physical features, of a person. So, if your criticisms are directed at substance dualism, that’s not a view that a philosopher is likely to actually hold.

          My view is that your self is your values. Your values are constant in you, as opposed to the constant flux of thoughts and emotions. You are currently thinking about something different than you were an hour ago, but your values are the same. If there was a radical change in your values (say, you suddenly had a religious experience and became a Christian), there would be a change in your sense of self.

          What do you think?

          1. My view is that your self is your values. Your values are constant in you, as opposed to the constant flux of thoughts and emotions. You are currently thinking about something different than you were an hour ago, but your values are the same. If there was a radical change in your values (say, you suddenly had a religious experience and became a Christian), there would be a change in your sense of self.

            Your values are, of course, a function of the physical functioning of your brain. And they’re about as constant as your body. Your values as a teenager are radically different from your values as a toddler — but so is your body. Your values as an adult are significantly but not radically different from your values as a teenager. Your values continue to change throughout your life.

            It’s the Ship of Theseus all over again.

            I increasingly find it difficult to get excessively worked up over concerns about the self. Carpe diem is much more fruitful — especially if today you’re busy building a better tomorrow. Build for the future, learn from the past, live in the moment. Who you are now is not who you were nor will be, but your future self will thank your current self just as your current self should be thankful for your past self. Pay it forward, and so on.




  4. Is this a typo? “a semantic squabble that has no import on academic philosophy but not on society”. Replace “no import” with “an import”?

  5. Is this a typo?
    … a semantic squabble that has no import on academic philosophy but not on society.
    Should it be:
    … a semantic squabble that HAS import on academic philosophy but not on society.
    … a semantic squabble that has no import on academic philosophy but HAS on society.
    Or: something else?

  6. … some [compatibilists] say our “freedom” is simply the complexity of the human brain, which allows us to weigh different inputs (“reasons”) before acting in a way no other animal can, […] Others say our freedom resides simply in not acting under duress: …

    These are two aspects of the same thing. To a compatibilist, our brains are like an chess-playing computer. The brain can weigh up all the inputs and make the decision that best reaches its goal.

    That flexibility, the possibility of different choices given different inputs is what we call “freedom” (similar to the physics or engineering concept “degrees of freedom”).

    A very simple animal only has a limited range of possible “responses” to a range of environmental inputs. A complex brain would have a greater range of possible responses to that same range of inputs. That’s why we evolved complex brains.

    If something then reduces the range of sensible responses, say an opponent putting the chess-playing computer in check, or someone holding a gun to the head of a person, then the “freedom” is reduced.

    A chess-playing computer has fewer sensible responses when in check, and a person has fewer sensible responses when coerced. That’s a reduction in “freedom” in the compatibilist sense.

    Thus I don’t agree that compatibilists are disagreeing about what “free will” is, they’re just explaining it differently.

    I also wouldn’t agree that compatibilists would say that “no other animal” has it. We’d surely say that other animals do have it, perhaps to a lesser degree. Any computational device (biological brain or otherwise) that evaluates inputs in order to select a response in order to attain a particular aim has “free will” in the compatibilist sense.

    1. If humans are complex automata operating within some complex set of ‘degrees of freedom’, then at what point do we decide to call it ‘free will’? There’s a convenient barrier at the moment between humans and all other instances of such automata. Was that always the case? Where there pre-humans that we would have said that they don’t have free will?

      From a determinist’s perspective it seems an arbitrary distinction, that coincidentally aligns with the same entities that dualist theists think have dualist free will.

      “Thus I don’t agree that compatibilists are disagreeing about what “free will” is, they’re just explaining it differently.”

      I disagree. They are not explaining the same phenomenon differently, but attributing the label of ‘free will’ to the something other than what determinist incompatibilists and duaists call ‘free will’.

      Further, when compatibilists respond to incompatibilists that ‘free will’ is an illusion they know (or should know) that the incompatibilist is referring to ‘free will’ of dualism. We cannot detect the source of our will events, our decisions, within the brain. The brain’s operations viewed by the brain itself, it’s self-observation, it appears to itself as a ‘mind’ not a brain. That’s the nature of the illusion: we feel we have a mind that is not the brain, and we feel like our decisions just appear, as if our will is free of physical cause.

      I really don’t get why compatibilists insist on claiming ‘free will’ as a label for the decision events of a physical autonomous brain.

  7. I don’t agree that defeating libertarianism is the most important or efficacious step for improving social circumstances or moral thinking. Moral and social concepts are not completely reductive to underlying neuroscience or physics. There are cases where determinism can help to get a point across, but there always seem to be more persuasive and accessible approaches that avoid descending into the neurological or metaphysical basement. Too much emphasis on free will vs determinism can, I think, undermine the broader social point because the topic is so far removed from common intuition.

      1. Sort of. I haven’t had time to process the whole discussion going on here, but this is what I think about reductionism: it’s basically a foundationalist theory of knowledge that can’t work. It’s a physicist’s dream to obtain a foundational unifying theory that accounts for all phenomena. But in principle, any such theory must depend on our theories about theories (mathematics, logic, epistemology, etc) which are built semi-independently. Furthermore there’s no guarantee that you can compute the properties of an emergent system from knowledge of all its constituent parts; emergent behavior may be indeterminate in principle, and also in practice due to any minuscule uncertainties or inaccuracies in the foundational theory. If there’s something I don’t know about the foundation, the uncertainty is amplified at higher levels of inference. Lastly as some have mentioned reductionism has no explanatory power in the sense of prediction and control. If I try and decide an engineering solution based on the composition of all underlying particles, the complexity explodes enormously. We can’t always approximate upward from a foundational theory. Somebody higher up in the thread mentioned gravity as a point in favor of reductionism; but Newton’s theory of gravity is not reductive at all. It’s based on directly accessible observations and mathematical models without any knowledge of deeper physics. For a huge variety of scientific and engineering tasks, deeper theory would only serve to obfuscate the relevant phenomena. Somebody else mentioned trying to debug software with an oscilloscope. In that same direction of thought, I can design a complex arithmetic machine without any knowledge of how it will be built; it can be mapped to mechanical or electronic or optical or hydraulic devices, all with totally different particle composition and different underlying physics. For that purpose, the underlying physics is only relevant up to the component level. Beyond that, the physics is discardable and I have a truly non reductive entity; it’s perfectly well defined and demonstrably physical, yet without being based on any specific physical system.

        1. If there’s something I don’t know about the foundation, the uncertainty is amplified at higher levels of inference.

          Those are all practical limits if you wish to build a computational model here on Earth to make the kind of brute-force predictions we’re discussing. But there’s no need to actually build such a super-duper-ultra-mega-hypercomputer to understand that the model is sound.

          Newton’s theory of gravity is not reductive at all. It’s based on directly accessible observations and mathematical models without any knowledge of deeper physics

          Read what I’ve written elsewhere here about effective theories. We don’t need to know what underlies an effective theory to be confident of the robustness of the effective theory or to be equally confident that any “mysteries” about the fundamental theory are relevant to questions that apply within the scale of the effective theory.




        2. Kepler reduces to Newton. Chemistry – as near as anyone can tell – reduces to physics. The fact that Chemistry has its own “laws” which are more practical to use than the exploding complexity of trying to use physics to do chemistry doesn’t mean otherwise. We have a standard model of physics, but don’t have or seek a standard model of chemistry with new causse and effect relationship, then a standard model of biology alongside the others.

          1. The explanatory power of a theory is entirely wrapped up in its predictive utility. If you want to predict the trajectory of a complex chemical reaction by solving Schroedinger’s equation for the whole system, you will get nowhere very slowly. To say that chemistry reduces to physics is almost completely meaningless if you can’t actually do it. It’s more than wishful thinking, but in the end it’s a grand untestable claim to say that all laws of chemistry are reducible to lower physics. The simple fact is that we DO have standard models of chemical systems which directly describe and predict the observable characteristics of those systems. “Exploding complexity” isn’t just an inconvenience, it makes most things uncomputable in principle. Reductive theories are not really theories if they can’t be used to make real predictions. They might as well be metaphysics.

            1. To say that chemistry reduces to physics is almost completely meaningless if you can’t actually do it.

              Except, of course, for the part where we can do it, only at small scales. See Folding@Home, for example, if you wish to participate in the process yourself.

              If your complaint is that we can computationally model protein folding but we can’t computationally model an entire human brain at the same level…sorry, but I’m not into chasing goalposts powered by rocket sleds.




              1. Sheesh. You can verify that very small-scale examples are reducible in very narrow ways. The protein-folding project is not reduced to the standard model of particle physics. That wouldn’t be possible or even useful. If you want to see an exquisite example of a no reductive theory, go check out the Wikipedia page on protein folding under the “principle of minimum frustration.” Read further and notice that Folding@Home uses a Markov model, which is the king of nonreductive model strategies.

                It sounds nice to label something as a theory-of-everything, to say “in principle, this theory predicts everything that happens in the universe,” but it doesn’t predict anything if you can’t actually compute the prediction. When computing your prediction requires more computational resources than exist in the universe, it isn’t really a prediction at all. Maybe it feels good but how can you call it knowledge if it can’t possibly answer specific questions about real phenomena? If other non-reductive theories do answer those questions, then they are necessary for real science and reductionism is inadequate.

              2. You don’t have to build the complete model from top to bottom. Instead, you do exactly what we’ve done: break it down into bite-sized pieces of multiple levels and confirm that each level properly fits with the one underneath.

                Frankly, what you’re proposing is that physics is so poorly understood that there still remains room for any sort of paranormal nonsense simply because we don’t have a Planck-scale entire universe simulator. And yet I’m sure you don’t wait to drink a cup of water until after you’ve performed a complete inventory of all the atoms in the glass….




              3. You’re not getting what I’m saying at all. In between debating with you about reductive theory, I’m preparing course materials that will rocket through Shroedinger’s equations, Fermi Dirac statistics, covalent bond theory and the octet rule, diffusion equations (derived from a Markov model abstracting away the complexity of lower physics), up to nonlinear circuit theory, eventually reaching up to technology-independent abstractions. This is a large bag of linked theories, but the story is not one of successive layers in a tower built on a singular foundation. There are links between all these things, but the lower levels are not sufficient to fully establish the higher layers. You can know every single thing about quantum states and it will never get you to energy band theory. Not without pulling in some other knowledge (statistical mechanics and thermodynamics) that is external to quantum theory. There is not one foundational story, there are many linked domains of science and mathematics that have mutual influences. They cohere (hopefully) but they don’t neatly stack.

              4. Again, your complaint is that science is not complete; ergo, it is broken. Such a position isn’t even remotely close to supportable.

                Will you at least agree with me that the Ideal Gas Law is derivable in its entirety from Newtonian Mechanics? That if you were to create a computer model of Newtonian spheres in an enclosed chamber that you would observe those spheres behaving as described by the Ideal Gas Law?

                If so, then I don’t see what the problem is in other cases where we as yet lack the computational power to perform such simulations but in which all observations are consistent with the same pattern.

                Now, of course, if you have evidence that cognition is is incompatible with Sean’s Big Equation, that’s another matter. But I trust you’ll forgive me if I don’t pay much attention to objections in this vein until such evidence is produced, at the very least, if not until the evidence is presented in the award presentation at a Nobel ceremony.




              5. 1. In my posts I’m not claiming that science is incomplete (although it is). 2. “Incomplete” does not mean “broken.” It’s a feature not a bug. 3. No, the ideal gas law is not derivable from Newtonian mechanics. It needs extra stuff. Your simulation will implicitly bring in extra stuff. 4. I’m not talking about computational inconvenience, I’m talking about intractability. There are physical limits to computation that render it impossible to scale up a simulation. A lot of important problems have exponential complexity. For example, many systems will conform to a minimum energy configuration, and finding it is similar to a traveling salesman problem. Example: on a trip from Hokkaido to Kagoshima visiting all Japanese prefectures the number of overland routes is 5,039,760,385,115,189,594,926,092,397,238,616,064. That’s 5×10^44 routes, and I need to discover the shortest path or paths. Any CS freshman can write an algorithm to solve this problem, but it probably won’t finish before the extinction of mankind. We could try to parallelize the calculation, but it will probably require more silicon atoms than there are on earth. It’s simply false to say we have the knowledge just because we have a method for computing it. We don’t have the knowledge until the computation is done (and even then, it’s not very much knowledge). If you want reductive theories to be meaningful, then you’ll have to prove that P=NP, and I will happily attend your Nobel ceremony when you do.

              6. Example: on a trip from Hokkaido to Kagoshima visiting all Japanese prefectures the number of overland routes is 5,039,760,385,115,189,594,926,092,397,238,616,064. That’s 5×10^44 routes, and I need to discover the shortest path or paths.

                How do you know that your 44-digit number is the actual number of possible routes? Did you sit down and count all the possibilities, or did you calculate it based on some simpler assumptions? Without actually enumerating every possible route, how can you be sure that the answer you posted here is correct? Even if you used a calculator to do the calculation, how do you know that the calculator did the calculation correctly unless you personally verified the results? And if you did personally verify the results, how do you know that you didn’t make an error yourself?

                When you understand the absurdity of my above complaint, you will understand the absurdity of your own complaints.




              7. You have not comprehended my point at all, in any amount. That makes me sad. I encourage you to read back through the discussion and try again, but I won’t be replying to further comments in this discussion.

              8. Then you’re equally not comprehending my point.

                It is no more necessary for you to manually calculate the number of possible traveling salesman routes to know what the number is than it is necessary to actually compute — by whatever means, manual or computational or whatever — the folding of a protein using a quantum mechanics level simulation to know that protein folding is, indeed, an inevitable consequence of quantum mechanics. That we haven’t actually performed such a calculation is by no means a valid reason to conclude that it’s not the case, any more than that you haven’t actually enumerated every travel route between Japanese prefectures is reason for you to conclude that the number of such routes is the one you gave.




              9. Okay, last reply: what you seem to be arguing is that the whole evolution of the universe is determined by some foundational laws; e.g. the whole universe is one big quantum state or one big particle system. That’s just a statement of determinism, and I think that’s probably correct. I’m not disagreeing with that. Reductionism is not the same idea. Reductionism is about knowledge and the structure of theories. There are lots of different versions of reductionism, but what I’m referring to is the notion that all knowledge can be deduced from a foundational theory. By “knowledge” I mean something with predictive utility that lets me make correct judgements about the world, something that helps me understand a system and control it for applied uses.

                As an extreme example, I’ve had bitter arguments with animal rights activists who argue that medical testing is unnecessary since we can simulate all the relevant biochemistry in a human body. No need to do experiments. If you have a correct core theory that explains all of chemistry and consequently all of biology, and if you could run the simulation (you can’t, but let’s suppose you could), then you should have total confidence in the result, right? So what’s the experiment for? I regularly deal with people who don’t see the point of paying for an experiment since we can just simulate things for cheap. That’s a form of reductive thinking.

                Of course there are lighter versions of reductionism, like a hypothetical grant administrator who expects to see big impacts in public health follow from investments in neuroscience. Maybe, maybe not. It remains to be seen how much practical knowledge can be deduced from neuroscience.

                If you go back to the top of the thread, what I’ve really been discussing is whether low-level physics can deliver much useful knowledge about morality or criminal justice. I’m not talking about whether our behavior is ultimately determined by basic particles, I’m talking about how much A can help me to correctly understand B.

              10. (On the phone…please excuse the short reply.)

                In that case, we’re pretty much on the same sheet of music. The only quibble would be that, though we’re a loooioooong ways from full computer modeling for medical research, every step in that direction improves medicine and reduces the need for animal testing. Someday, maybe not I our lifetimes, but someday, such animal rights complaints will be valid…but also moot because we’ll by then long since have abandoned a map testing as ineffective…in the distant future.




              11. My brain is made of elementary particles. One of the particles moves from point A to point B. What moves it? Forces described in the standard model, or is something required to be added? If something needs to be added to account for the motion of that particle, then the standard model will need to be changed. The only alternative is to suggest some kind of mind over matter or primitive animism is moving that particle.

              12. If something is predicted to happen in the brain and it doesn’t, then there may be a problem with your theory. But you won’t be looking to fix the standard model of particle physics. If the folks at CERN discover a big change in particle theory, it won’t have any impact whatsoever on neuroscience.

              13. No. If we couldn’t account for what moved a particle (which we can), we’d figure it out and update the standard model. You’re doing “God of the gaps”.

              14. Biological entities like neurons are observed to move in a certain way. If that way doesn’t resolve to physics, we may need to supplement the standard model of physics with a “standard model of biology”. I don’t think anyone at CERN would sit still for that. Any evidence that runs counter to reductionism is a red flag for physics.

              15. Every observation of every phenomenon here on Earth is fully consistent with Sean’s Big Equation. His equation is not adequate to explain certain far-distant exotic phenomena at far-removed scales, especially including black holes, dark energy, and cosmogenesis; yet we should be overwhelmingly confident of its applicability here on Earth, precisely because of the overwhelmingly vast mountain of evidence that supports it.

                If you’ve got actual evidence that neurons are inconsistent with Sean’s Big Equation, there’s a Nobel with your name written on it.




              16. @Diana MacPherson ” If we couldn’t account for what moved a particle (which we can), we’d figure it out and update the standard model. You’re doing “God of the gaps”.” First of all, we can’t. We account in general but quantum mechanics is not strictly deterministic. We have an unbridgeable gap there. We put in many worlds or God or mind-over-matter or something else. But my point was more that autonomous conscious agency could only exist in the gaps. It was part of a larger discussion about reductionism in this thread.

              17. We put in many worlds or God or mind-over-matter or something else.

                If you think gods are on an equal footing with the Schrödinger equation, you’re barking up the worng tree.




              18. Skipping over all the weird claims you’ve made about QM, you’ve just equated god, “mind over matter” and many worlds theory and that’s completely fallacious.

                Many worlds uses indirect evidence (facts) and uses scientific tools (math), to propose an answer to a question. “Mind over matter” and “god” posit an answer (one that is often defended as absolutely true and unchangeable) where facts are unknown with no evidence and no scientific evaluation. It’s an answer pulled out of nowhere. It’s faith, not fact.

                Or as Tim Minchin put it,

                Science adjusts it’s views based on what’s observed
                Faith is the denial of observation so that Belief can be preserved.

              19. @Diana MacPherson Didn’t equate. I listed a # of things which are brought in by different people at different times to bridge the gap which is observed. I could never make enough “Weird claims” about QM to do it justice. And with that, I’m done – you may have the last word.

            2. There are certainly cases in chemistry where something unexpected is happening and then more scrutiny is placed on the situation in terms of underlying physics. Yes, this is how it is done. The physics is believed to underlie the chemistry, even if it cannot be simulated out to a practical level in most cases. But the physics is there to look in cases where needed. Where it comes through, this confirms the reduction. Where it seems more problematic, physicists may seek to tweak the standard model.

              1. “There are cases” where one level of theory can link up with another. That’s not the same thing as “reducible”. Scientific theories are claims about knowledge. Knowing everything about particle physics doesn’t endow you with much knowledge about chemistry or biology, although it may give you a little, as in those cases you mention. But in nearly all cases there is no path of inference from particle physics up to biology or neuroscience or electronics. And if there’s a single tiny wrong with your beliefs about particle physics, or a single tiny thing left out, then all of your inferences become suspect unless there are independent, nonreductive means of testing those beliefs at higher levels. It’s satisfying to believe that “there is a theory” that accounts for all phenomena, but you can never really know if you have it, and even if you do have it you can’t use it for very much without help from above (not supernatural help, just help from methods that don’t directly derive from your theory).

              2. The particular cases I was meant to indicate are not simply cherry-picked – they are picked as something difficult for chemistry that makes sense when the usually-more-practical “laws” of chemistry are abandoned and then the more-reliable but harder-to-apply-en-masse underlying physics rules have to be brought in.

  8. I am in Class C: Compatibilist. Reduction of physical interactions is not important so long as it cannot be proven that tomorrow’s future is not fully predicable which makes a free will metaphysical.

  9. Even if macroscopic brain processes operated non-deterministically, this would still not support A. So, there is no need to commit oneself to ‘Hard Determinism’ in order to fight against A. If determinism is false, that still does not give us the ‘free will’ that most people want to believe in, because it simply makes our choices random.

    Thus, in my opinion, the crucial distinction should be between substance dualism vs physicalism/monism, not free will vs determinism.

    1. Condition A should be parsed into at least two groups:

      1. Dualism with Free Will
      2. Dualism without Free Will

      I contend that Libertarians want a n immortal soul, I think most of them find the issue of free will as tangential. They want heaven to be real and Condition A is pretty much the only game in town.

      Empirically, Condition A is in the perpetual quick sand of determinism. Conditions B and C do not share this plight.

      1. In the FW&D class I alluded to, my classmate Lisa Fuller remarked at one point we what we were really doing was skirting around the mind-body problem. I agreed, though I also said I regard that as solved. (I’m a combination emergentist-eliminativist-functionalist!) She replied something like not everyone is so sure, which is correct, needless to say.

  10. I am not compelled to pick any of the options from the “menu”. It looks to me that people are trying to reach a conclusion about objective reality of free will (or its absence) from purely metaphysical premises. That has never worked to resolve any issue in philosophy – the same options (possibly worded a bit differently) will be on the menu a thousand years from now. The warring fractions simply refuse to admit that they lack the instruments to objectively choose one menu option over another.

    Professor Ceiling Cat believes “that our brains, being material objects operating under the laws of physics, can give only a single output from the inputs they receive (barring any quantum indeterminacy operating in our neurons).” In other words, PCC believes that we know enough physics to pick the right answer from the menu from what what we know about physics seems to tell us. However, it is only a _belief_ – a metaphysical one – and given the intricacies of quantum mechanics and its interpretation, and anything else that we don’t know, not even quantum physics experts (especially not them) will necessarily accept that belief. We are only a century away from learning about quantum mechanics, much less so from quantum entanglement and standard model and basic principles of cosmology, and it is naive to think that our current knowledge will remain stable enough over the next several centuries in a way that won’t undermine PCC’s current belief.

    What I have consistently disagreed with PCC are the practical consequences of free will or lack thereof. What I firmly (metaphysically) believe is that, if somebody or something told us with certainty which one of the above menu options is the correct one, it would make exactly zero difference for how we (choose to) live our life. For that reason, I think that invoking philosophical free will for discussing ethical responsibility, punishment and penal systems is a category mistake.

  11. I agree that our actions are “solely and uniquely decided by our genes and our environments, and nothing else.” However, why does that therefore mean that there is one and only one outcome possible in any given situation? In math you can have problems with two or more right answers–could this not also be true here? Or what about a deterministic model that accepts outcomes within a narrow statistical range? I’m sure these questions have been answered somewhere; I’d love to know by whom.

    1. In math you can have problems with two or more right answers–could this not also be true here?

      Yes and no.

      Yes, in the sense that simplistic models of reality will produce statistical predictions.

      No, in the sense that, ever since Laplace, we have understood that physics is entirely deterministic in such a way that, given the complete state of the Universe at any given point in time plus the laws of physics, you can precisely predict the entire future and past of the Universe.

      That was Laplace’s big insight, as anthropomorphized by his eponymous Daemon, and it’s remained a feature of all subsequent physics.

      Yes, yes — constructing a machine capable of calculating the Universe would require substantially more “stuff” than there actually is in the Universe, and actually observing the Universe without interacting with it poses other problems, and so on. This isn’t meant as a recipe for how you’d actually go about doing it, but it is our best understanding of what really does happen.




      1. given the complete state of the Universe at any given point in time plus the laws of physics, you can precisely predict the entire future and past of the Universe.

        Doesn’t the existence of black holes mess up figuring out the past of the universe? How can you determine what went into the black hole?

        1. Black holes are one of the big ares of mystery in modern physics. We don’t know how they operate. It might be the case that “information” (in the sense used by physicists) must be sacrificed in our understanding of them, but it might not be the case. We just don’t know.

          We do know, of course, that, outside of black holes, the universe behaves as if information is always conserved. Even if conservation of information breaks down at some point, that breakdown is so far removed from life on Earth that it’s utterly irrelevant to cognition.

          If information is not conserved, Laplace’s Daemon wouldn’t work at universal scales. But we know with overwhelming confidence that it does work at all scales short of black holes (and similar discontinuities, such as the Big Bang).

          (And I personally suspect that the solution to the black hole problem will preserve information…not that my suspicions have any significance in the slightest.)



        2. Yes, given classical physics. But what about quantum physics? Add chaos to the mix and you get a very unpredictable universe.

          1. Quantum physics is statistically deterministic. Given the complete state of the Universe at any given point in time plus the laws of physics, you cannot predict precisely which atoms in a lump of uranium will still be there, as uranium, a minute in the future. Instead you can predict approximately how many of them will decay. (Chaos is different–a chaotic system is basically one which you can predict the future but only by perfectly modelling the whole system and letting the model run. Tiny differences at the start make big differences later, so you can’t make the usual simplifications.)

            1. Quantum physics is statistically deterministic.

              Not just statistically, but actually so. The determinism is trivial to see in the Everettian Many-Worlds interpretation. All the other interpretations are equally deterministic but have different means of explaining why particular outcomes are observed with the frequencies given by the Born Rule.




              1. Ben, as usual I’m enjoying your posts, but I disagree in this case.
                When you say deterministic, I suspect you’re talking about the evolution of a wavefunction. Whereas the observables are probabalistic.
                Given the same setup of the universe, I suppose one couldn’t tell which observations will occur in reality, only the relative probability.
                So, if you could rewind the universe, somehow, then couldn’t you get different outcomes to the same measurement…?

              2. Remember, in Everettian Many-Worlds, all outcomes do happen, and they’re all observed. Do the double-slit with single electrons, and you will observe the electron on the left and you will also observe the electron on the right!

                The confusion comes from the fact that the respective wavefunctions will diverge, in essence creating two “you”s, each of which has equal access to your current memories and equal claim of being the “real” you — but no way of interacting with each other.

                Imagine you’re given a raffle ticket upon entering a room, and there’re a number of other people in the room. You each have a number on the ticket, and no two numbers match. What significance are you going to place on the number you’ve been given? You can even have a plainly structured means of generating the tickets, but the perception of each ticket holder is one of randomness.




              3. Thanks for the reply!

                I’d suggest the relevant thing about the double-slit experiment is not so much the question of which side the electron goes through, so much as that by trying to measure which side, one disturbs the far-field distribution. Bluntly, the idea that the electron goes through one slit or the other is not compatible with experimental observation of a coherent interference pattern.

                The raffle ticket analogy sounds worryingly like a hidden variable theory…

              4. Ah, but in the context of the original post, this resurrects the exact question that Everettian Many Worlds theory can’t yet crisply answer:

                “…were we to return to a previous situation of “choosing”, with every particle in the universe configured as it was before,”
                would the branch that “I” end up in (the “I” that I consider to be “me”) be the same branch, or could it be one of the other branches, in which I made a different choice, and experienced a different outcome?

                Strictly speaking, an Everettian would have to respond, yes, with a certain probability!

              5. The question is unanswerable because it’s ill-posed. There is no unique you that distinguishes one branch from all the rest. Each branch contains an instance of your brain that believes — correctly — that it’s you. So on both the original pass and on the replay, you end up in all branches, and there’s no meaningful sense in which any one of them is more truly you than the others.

              6. There’s another aspect to it…and that’s the fact that not all conceivable outcomes are actually possible.

                There’s a popular misconception that the laws of physics permit all the air molecules in the room to randomly start moving in unison to an upper corner of the room, leaving the bulk of the room in vacuum and that one corner in an high-pressure state indefinitely. This misconception results from a failure to understand the underlying physics of gases; namely, that gas molecules are, essentially, Newtonian objects. And Newtonian objects bump into each other. And fall down in gravitational fields. No amount of “randomness” is going to overcome that sort of behavior. Furthermore, long before the bulk of the room was in vacuum, the air would be in an extraordinarily low entropy state…which is an high-level-language way of expressing the same thing.

                A more relevant example would be with the “universe splitter” app that somebody here, maybe you, mentioned. The idea is that you come up with two different actions you’re committed to taking — a choice. And an expensive piece of machinery somewhere performs some sort of measurement that results in equiprobable observations that, according to Everettian Many-Worlds, will both be observed on different branches of the universal wavefunction.

                So, if you’re at the ice cream parlor and can’t decide between vanilla and chocolate, no worries: flip the quantum coin and do what it says. The Universe will split, and one you will live a life in which you had vanilla and another you will have had chocolate.

                Let’s say the choice is now between having ice cream and licking the terminals on a 9-volt battery. Most of us have done the latter at some point…it’s decidedly unpleasant, but there’s no real risk to it. Do you still think you’ll follow the advice of the quantum coin? You might, depending on the sort of person you are and the mood of the room and the like.

                Now, the choice is between ice cream and committing some unspeakably horrific criminal act upon your most loved ones. I think we can safely conclude that all of us would have the ice cream, regardless of what the quantum coin said — and that, for any exceptions, they’d likely commit the act sooner or later whether or not the quantum coin said to do it right then.

                At this point, it should be obvious that the “all possible outcomes” really doesn’t mean that, in some other branch of the wavefunction, you’re the Emperor of the World. For the most part, you’re pretty much recognizably you.

                …and we know this to be the case, because of both the overwhelming regularity we observe (there aren’t any random people who’re Emperor of the World, horrific criminals are few and far between), and because the Born Rule dictates that classical Newtonian Mechanics is as inevitable result of Quantum Mechanics as the Ideal Gas Law is of Newtonian Mechanics.

                I should add: closely related to this discussion are the questions of whether time and / or space are infinite. And if both are infinite, then we can be confident that there’s some sort of trans-universal equivalent of the Born Rule as it works in Everettian Many-Worlds, where the number of branches of the wavefunction are proportional to the frequencies of observation.




      2. How does knowing the current state allow prediction of past states? E.g. if I find a Rubik’s Cube in a particular configuration, how can I tell which of the many possible configurations it was in before the most recent rotation was made? A fully deterministic universe may mandate only one possible successor state to any given state, but I don’t think it mandates only one possible predecessor state to any given state.

        1. The microscopic laws of physics are time-symmetric and information-conserving. Distinct predecessor states never evolve into microscopically identical successor states since that would violate time symmetry and destroy information.

          At the macroscopic level it’s a different story: distinct macrostates can evolve into pragmatically indistinguishable successor macrostates; that’s the basis of entropy and the arrow of time. But at the micro level, reversibility is the rule.

            1. “Information” as used by physicists is much, much different from the everyday definition. In a Classical perspective such as Laplace would have been familiar with and as is easy for lay people to understand today, it would be the composition, position, and momentum of every submicroscopic piece of stuff in the Universe. In the world of Quantum Field Theory, it would be the state of all the fields…

              …quick explanation of QFT. Laplace’s big idea was to take Newton’s Gravity and turn it on its head. Rather than spooky action at a distance such that you had to know everything about the entire Universe to know what local gravity is like, he posited a gravitational “field.” Imagine a grid laid over all of space, with arrows at each point in the grid. The direction of the arrow shows you what direction something would be pulled by gravity, and the length of the arrow would show you how hard it would be pulled. Now, you just have to know the state of the field at that one point to know what it’s going to do.

              In modern physics, everything is a field. That’s the answer to the big question your teachers might have posed but never answered…are electrons particles or waves? They’re waves, nothing but waves — and they’re waves in the electron field. The electron field would be best visualized much like the gravitational field, only the “density” of the field at a given point is the likelihood that you’d observe an electron at that point.

              There are also quark fields (which can be conceptually combined into proton, neutron, and nucleus fields, or even with electrons if you want to think of atomic fields), photon fields, electromagnetic fields, and so on.

              The combined state of all those fields would be what a modern physicist means by the “information” that we think is probably conserved.




              1. Thanks, Ben. I knew lots but not all of that. Still, the question remains.

                Why I don’t do information theory is all that jazz about Maxwell’s daemon (why is it spelled with an “a’?) needing a note pad to keep track of things. And you read that entropy is either information or its negative, so no one seems quite clear on the matter. I admit that I am not interested enough to read Shannon.

              2. Any notepad that Maxwell’s Demon (with whatever spelling you normally use) might possess is purely metaphorical. For all sorts of reasons, the Daemon can’t actually exist and / or interact with the Universe, save unless some sort of conspiracy theory actually holds. For example, if everything since the Big Bang is “merely” a Matrix-style computer simulation being run on a computer vastly bigger than the Universe as we perceive it, that computer could rightly be considered Maxwell’s Demon.

                Entropy is most emphatically distinct from information. Entropy is simply a statistical measurement of the number of microscopic states which are macroscopically indistinguishable. Your latte has lower entropy before you stir it because there’re more ways for milk-and-coffee to be homogeneous than there are to have a layer of coffee on bottom and a layer of cream floating on top.

                As a side note, complexity is loosely related to entropy. The initial condition is simple: cream on top and coffee on bottom. The final condition is also simple: creamy coffee everywhere. But the transition from the one to the other is (usually, but not guaranteed to be) filled with all sorts of complex whorls and vortices and gradients and what-not.

                It is probably no coincidence that we “just happen” to live in a period in the Universe’s lifetime roughly midway between minimum entropy (the Big Bang) and maximum entropy (heat death) when complexity is at a maximum.




              3. “Entropy is most emphatically distinct from information.”

                Well, information is defined in some cases in terms of the log of the probability distribution. That sounds pretty much like entropy. And I read in one place, which I forget, where the author discusses whether information should be considered negative or positive entropy. So there is some link — and a fair amount of confusion — there.

                As for the demon’s metaphorical note pad, I”ve never understood why he needed to take notes at all. Why doesn’t he just look at a molecule, pass it or reject it, and then forget about it?

                Actually, we are getting off topic, aren’t we. In any case, thanks for your … uh, information. 😉

              4. As for the demon’s metaphorical note pad, I”ve never understood why he needed to take notes at all. Why doesn’t he just look at a molecule, pass it or reject it, and then forget about it?

                Different Daemon.

                Laplace’s Daemon knows the state of the Universe at a given moment in time plus the laws of physics and can use that information to calculate the state of the Universe at any other point in time. Laplace’s Daemon has to be bigger than and external to the Universe yet able to observe it non-interactively, which can only be possible if the Universe as we observe it is really some sort of conspiracy — such as a Matrix-style computer simulation. But the thought experiment remains superbly useful even if we’re not dupes (which we almost certainly aren’t).

                Maxwell’s Daemon operates a gate within a two-chambered vessel filled with gas. It knows the average kinetic energy of the gas and it knows the actual kinetic energy of a molecule approaching his gate. He opens the gate in a sorting manner, sending all higher-energy molecules to the one side and all lower-energy molecules to the other. In so doing, he creates a state of low entropy in the system where initially there was high energy. This would appear to violate the Second Law of Thermodynamics — and, indeed, it does unless Maxwell’s Daemon either extracts energy from the gas in the system (thus cooling it) or is externally powered.




          1. The laws of physics as we know them are time symmetric (well, CPT symmetric actually.) But Richard is right that determinism doesn’t require time-reversal symmetry.

        2. The Rubik’s Cube states (where the pieces and rotations are) are large aggregates of microstates. So one can run the laws of motion backwards and forwards from *that*, not from the macrostate.

  12. Members of B and C to spending copious amounts of time distinguishing why they are distinct is similar to the far left eating their own. . .

    Even most varieties of Christians have figured out that their survival is strengthened if they ban together against non-Christians. This is why there are so many ecumenical Christian events. Similar with the interfaith community, which has figured out that supporting other faiths helps support the faiths of its diverse members.

    B and C having a semantic fight seems to be about establishing group identity. It may even reveal a thread of tribalism within the atheist community to the extent that the distinctions divide atheists into identity silos.

    It would be smarter for B and C to team up and focus on engaging members of A. But that would take actual interaction with A, and when people are stuck on proclaiming their own affiliations, engaging groups that are grossly different isn’t so compelling.

    But this is one case where B and C should take a hint from how other “minority” groups have survived. Team up.

    1. Good analysis. I would think, though, that to engage and oppose libertarian free will is a bit like tilting at windmills. Far better to attack the religious/moral segment at it’s proven vulnerability – simple poor thinking skills (irrationality) on a number of far more intelligible and urgent topics – of which there are very many.

  13. I’m in class B: We are the authors of our thoughts and decisions, but neither of those are free: it’s physics all the way up. We could speak of thresholds or scales of freedom for practical purposes only (we already hold mentally ill criminals to different standards, for example).

    Yes, disagreements between B’s and C’s are semantics or nitpicking, and are secondary to disagreements with A’s, who happen to be the most numerous by far.

  14. I’m category B and agree that the argument against A is important. The problem is that the rank and file member of group A sees group C as somewhat of an ally, so by espousing that position you’re tacitly supporting A rather than opposing it.

    1. My sense is that people in Group C are concerned with the implications of ‘volition’, which is so ethereal and little understood at this point that adding the term does little to help. If volition is largely an illusory sensation added as a piece of subjective information to an outcome that was going to happen no matter what, then we should treat a calculating criminal the way we treat someone with a brain tumor. If volition is tied to agency in a way that is far more influenceable by environmental input and genuinely a product of whatever neurological program the sense of ‘self’ is, then it’s important to distinguish acts that are volitional (if not ultimately ‘free’) from acts that are not volitional (and grey cases like brain tumors, where we intuit the volition as being there but not the person’s ‘true self’) when structuring societies. It seems to me that volition is the intermediate step of interest here.

  15. As with so many philosophical conundrums, the problem comes with really bad thinking laying a foundation of presuppositions that simply don’t make sense to begin with, and then building sky castles thereupon.

    “Free Will” is a married bachelor, no matter how you try to spin it. Just as marriage and bachelorhood are very useful terms in their own contexts that each separately make perfect sense but can’t even hypothetically be simultaneously juxtaposed, so, too, is the case with freedom and the will.

    The whole point of freedom is the lack of constraints. The whole point of the will is to impose constraints. Even pure theological “Free Will” with non-corporeal spiritual souls and the like doesn’t make sense; either the will has some sort of predictability and structure and meaning to it, in which case it is not free from constrain; or else it is free from constraint but utterly meaningless and the antithesis of willful.

    Might as well argue over whether or not “Organized Randomness” is real or not.

    Even worse, “Free Will” is inextricably intertwined with theology. Perform a survey of the general public asking for their understanding of why bad things happen, especially why people do bad things, and an overwhelming majority will answer that “Free Will” somehow excuses the incompetence and / or immorality of their favored gods, in some unspecifiable fashion.

    If you take a step backwards, and ask people to point to their “Free Will” and when they’re exercising it, you’ll see that most people equate “Free Will” with their decision-making process. That decision-making process is very much understandable and real, but it has no more to do with “Free Will” than biology does with “Elan Vital,” and we do ourselves a great disservice to confuse matters by equating the two.

    The closest I come to Sean and the compatibilists is in noting that the decision-making process is best described as an high-level phenomenon. Just as you’re much better off describing the air in the room using the Ideal Gas Law rather than listing the position and momenta of each molecule, you’re better off describing decision-making in terms of counterfactuals and the like rather than the electrochemical potentials of every neuron.

    And, when you do so, much of the popular language remains intact; you’re “free” to “choose” between chocolate and vanilla, even though, in reality, you have no more choice in the matter than a pebble does when falling down a ravine in a landslide. We might say that a given pebble could go to the left or the right of a given tree, and we can say that certain percentages of pebbles will go to the left and to the right — but we also understand that there’s no actual choice or fuzziness involved, other than in our ability to predict events in that level of detail.

    Indeed, the subjective perception of that decision-making process is a perfect fit with Jerry’s folk definition of “Free Will”: the ability to rewind the tape of history and play events out differently. That’s exactly how we imagine different scenarios in order to pick our preferred decision, as well as how we analyze our past actions. Never mind that no actual rewinding of the universe is involved and that even this exact mental activity is itself perfectly deterministic; we pretend that that’s what’s happening. And that pretense is superbly useful and an excellent way of optimizing one’s actions.

    I think, if we can get people to that level of comprehension, we’re pretty much home free.



    1. and this works how in the Many Worlds Interpretation where you make all possible choices simultaneously?

      1. The determinism of Many-Worlds should be especially obvious. All outcomes present, in frequencies given by Born’s Rule. If it’s possible, it happens; if it’s not possible, it doesn’t. The only question for observers is whether or not they observe the outcome — but the outcome does happen and is observed. Each observer only observes one outcome and no observer observes all outcomes. The outcome observed on any given “branch” is mere anthropics.




    2. “Free Will” is a married bachelor

      A glaring sign that you have probably failed to exercise charitable interpretation.

      Even worse, “Free Will” is inextricably intertwined with theology.

      Theologians love to get their hands on important but poorly understood aspects of human life, like consciousness, morality, and free will. That’s not a good reason to cede the territory.

      Indeed, the subjective perception of that decision-making process is a perfect fit with Jerry’s folk definition of “Free Will”: the ability to rewind the tape of history and play events out differently. That’s exactly how we imagine different scenarios in order to pick our preferred decision

      The second sentence here, minus the first three words, is absolutely right. The first is an elaborate misinterpretation. A better interpretation can be built on the basis of Judea Pearl’s analysis of Causality. We can trace forward the effects of various interventions in the causal network. That’s all we need. No universe-rewinding required.

      1. “Free Will” is a married bachelor

        A glaring sign that you have probably failed to exercise charitable interpretation.

        That, or I have and you have failed to follow the arguments of compatiblism fully to their conclusion in the manner summarized by my quip.

        Again, freedom is the property of being without constraint. The human will is the very essence of that which constrains our actions. I did not choose the analogy of the married bachelor lightly!

        The first is an elaborate misinterpretation.

        How so? Whenever I wish to analyze some complex situation in order to make a decision, that’s exactly what I do: I imagine I make one choice, imagine how that scenario might play out, and then start all over again imagining what might happen if I made a different choice. Sometimes I leap back and forth between the scenarios; sometimes the conclusions are obvious enough from the get-go that it all subjectively seems to be over in an instant.

        So either you’re taking issue with my description of the contemplative decision-making process — which I would find astonishing — or you think that it’s a poor fit for Jerry’s folk definition of “Free Will” — which I would find equally astonishing.

        A better interpretation can be built on the basis of Judea Pearl’s analysis of Causality. We can trace forward the effects of various interventions in the causal network. That’s all we need. No universe-rewinding required.

        I’m unfamiliar with the book and certainly not going to go read it in the next five minutes. But I would also suggest that, even if Pearl has laid out some valid theory for understanding causality, it is utterly irrelevant to the way humans actually perceive the process of making decisions — and it’s that subjective experience that I’m referring to.




        1. “Constraint” is relative to an analytical framework for studying a causal network. As Judea Pearl says of the distinction between exogenous and endogenous variables:

          The scientist carves a piece from the universe and proclaims that piece in… The rest of the universe is then considered out… This choice of ins and outs creates asymmetry in the way we look at things and it is this asymmetry that permits us to talk about ‘outside intervention’ and hence about causality (pp. xiii-xiv).

          When we ask what we are free to do, we take our decision as an exogenous variable, and what happens to us afterward as an endogenous one. A man in jail is constrained much more than one who is not.

          Similarly to how “probability” still makes sense in a deterministic universe – because it’s relative to the position of a knower – so too do degrees of freedom vs constraint make sense in a deterministic universe. The latter are relative to the position of an agent.

          There’s no “rewind” of time in decision making; instead there are parallel imaginings each with their own time line. If I “rewound” time as you describe, it would take me twice as long to make a decision (in the limit of a large number of options considered). One forward-traverse per option will do.

          1. A man in jail is constrained much more than one who is not.

            But that’s simple freedom — not freedom of the will. The man in jail has every bit as much will as anybody else, and his will is subject to the same constraints. His ability to act upon his will, yes, is more constrained. But his will itself? Does he not long to roam free in the forest with the breeze in his hair every bit as much as you do? How is his will to do so any different from your will to do so? And how is he any different to chose to wish to go for a walk in the park than you are?

            If I “rewound” time as you describe, it would take me twice as long to make a decision (in the limit of a large number of options considered).

            Only if you think I’m proposing an actual rewinding of time. Instead, I’m noting that we imagine playing potential future events multiple different times. And, of course, these imaginings are far less detailed than reality — and it can take a split second to play out.

            Here’s a demonstration. By virtue of my writing these words, you’re now aware that you have a choice between doing whatever else you were going to do today and going out and robbing a bank. Do you really need to spend the next several decades to imagine the consequences of robbing a bank? No; just that quick moment of imagining the rest of your life behind bars doing hard time for armed robbery is more than enough for you to conclude that that’s not a future you wish to experience. Right there, you’ve run out an imagined future for the rest of your life, rewound the tape to an imagined future where you go about doing whatever you were planning on doing, and chosen the latter.




            1. You can distinguish between free will and freedom simpliciter, but they’re related. If you want to ask what distinguishes free will from freedom in general, it would be the ability to consider a wide range of options, evaluate them in depth, and stick to a decision unless some good reason came up to change it. So basically reasoning ability, maturity, etc. – which explains why mentally healthy adult humans are the paradigm case of free willed beings.

              As far as rewinding time, I’m suggesting there’s a simpler metaphor. Something along the lines of Multiple Worlds in QM (with one extra copy who temporarily stays put). The agents in the multiple worlds explore the terrain until all but one meet unacceptable fates, or the one meets success. Then the extra copy, the actual decider, tries to follow the path that looked successful. This doesn’t require – please note – that all the imaginary copies start at the exact same position. Generally they don’t. They start at *a position the decider knows how to reach*.

              1. If you want to ask what distinguishes free will from freedom in general, it would be the ability to consider a wide range of options, evaluate them in depth, and stick to a decision unless some good reason came up to change it.

                Ergo, a simple thermostat (bimetallic spring and mercury tube) doesn’t have much in the way of “Free Will,” but a modern programmable thermometer has significantly more.

                …and this is a meaningful concept with significance…how, exactly? And in what way does it align with common and historical definitions of and uses for the term?



              2. A simple thermostat is like a room with one molecule of dirt in it; a modern thermostat is like one with a few dozen. The latter is many times “dirtier” than the former, but neither one would even begin to be called “a dirty room” in practical circumstances.

              3. I’ll grant you a form of consistency with such a definition / usage.

                But I still don’t see the freedom in even a simple thermostat.

                “Will” I could certainly see in the thermostat, exactly as you laid out: a very little bit, but not much.

                But “freedom”? In something so clearly and obviously deterministic?



              4. There’s a good point in what you say. The determinism isn’t a problem, but the simple thermostat doesn’t simulate multiple courses of action, so it’s lacking in freedom. Let’s suppose that a complex advanced thermostat has control over both the blinds and the furnace. The room is too cold, so the thermostat evaluates what will happen if it opens the blinds more. That will generate enough heat, and not too much glare. The furnace would work too, but is more expensive. So it goes for solar. That would be a teeny tiny step toward freedom. It’s a long way from General Artificial Intelligence, but it’s something.

              5. So, following your analogy…a single train on a single track has no freedom. The rail system in its entirety has much more freedom. The interstate highway system perhaps more.

                Again, I can see how you’re being consistent in your usage…but I remain utterly befuddled as to how the word, “free” can even begin to be suitable to the description.

                “Complex,” sure. And maybe “flexible” or “useful” or other variations on that theme. But “free”?

                Writ Orwell, “Freedom is slavery.” That’s exactly what you’re proposing — trains on tracks are “free” to go wherever they want, so long as it’s exactly where the tracks lead. A la Ford, you can have your Model T in any color you want, so long as the color you want is black. What choice, what freedom is that?



              6. I dunno – does the rail system as a whole have an advanced computer controller (or a board of human governors) that evaluates multiple scheduling options, then picks the best one? If so, then I’d say it has some degree of freedom (in proportion to its intelligence, flexibility, etc) and makes choices.

                But the “rails” are notably missing from human life. A human being is more like a hovercraft than a train: where it goes depends to an extreme degree on internal factors. And sure, those internal factors are part of a deterministic causal network, but that doesn’t diminish us, nor our influence. “The devil made me do it” could be a valid excuse, if there were such things as devils. “I made me do it” is not; nor is “my brain made me do it”, which says the very same thing, thinly disguised.

              7. As far as rewinding time, I’m suggesting there’s a simpler metaphor. Something along the lines of Multiple Worlds in QM (with one extra copy who temporarily stays put). The agents in the multiple worlds explore the terrain until all but one meet unacceptable fates, or the one meets success. Then the extra copy, the actual decider, tries to follow the path that looked successful.

                If is is meant to describe the cognitive process, it’s inaccurate. The brain does not multi-task like that. If it’s meant to describe actual physics, it has no bearing on reality whatsoever.



              8. The brain does occasionally multi-task, in that sometimes we develop scenario A a little, then develop scenario B, then investigate scenario A some more. But I’ll concede that sequential play of scenarios is more common.

                The more important point is that the scenarios don’t have to start from the same exact position. In fact “exact” positions are not mentioned at all – only the gross macroscopic level, sufficient to engage the decider’s model of how stuff works. In this respect quantum “multiple worlds” is also the wrong model. Oh well.

              9. The brain does occasionally multi-task, in that sometimes we develop scenario A a little, then develop scenario B, then investigate scenario A some more.

                That’s task switching in the parlance of modern computers. It’s sometimes considered a “poor man’s” multitasking, and is how Ye Olde Mac OS System 5’s Multifinder functioned. True multitasking requires parallel execution, such as is typical on today’s multi-core CPUs. And, yes, humans pretty much do task switching, with notable exceptions for unconscious and semi-conscious subsystems such as breathing.

                The more important point is that the scenarios don’t have to start from the same exact position. In fact “exact” positions are not mentioned at all – only the gross macroscopic level, sufficient to engage the decider’s model of how stuff works.

                Indeed, you’re really making my point for me…the imagined worlds of our decision-making process are very rough and lacking in all but the most important (perceived) details. But, then again, so is our conscious perception of the real world! Which is probably why it’s so easy to get lost in thought, to willingly suspend disbelief in a story, and so on.

                But, again…the entirety of the imagination is, itself, every bit as deterministic as a thermostat. Vastly more complex and interesting, to be sure…but no more “free.”




              10. As a compatibilist, I’m completely unbothered by the fact that imagination is a deterministic process in the brain. That’s exactly what I’ve been assuming all along – while demonstrating that this is perfectly compatible with the actual perceptual judgments we experience in decision making. The actual experience, as opposed to theological, or even science-inspired, theories about it. I think Jerry’s “rewind” scenario is a science-inspired theory about everyone’s experience, or maybe even a science-inspired theory about a theological theory about the experience.

    3. “The whole point of freedom is the lack of constraints.”

      Disagree. Typically, “freedom” means a lack of *certain* constraints. So you have built an argument starting off on the wrong premise, IMO.

      “The closest I come to Sean and the compatibilists is in noting that the decision-making process is best described as an high-level phenomenon. Just as you’re much better off describing the air in the room using the Ideal Gas Law rather than listing the position and momenta of each molecule, you’re better off describing decision-making in terms of counterfactuals and the like rather than the electrochemical potentials of every neuron.

      And, when you do so, much of the popular language remains intact; you’re “free” to “choose” between chocolate and vanilla, even though, in reality, you have no more choice in the matter than a pebble does when falling down a ravine in a landslide.”

      Ben, this is where you seem to be contradicting yourself. And it happens right when you make the mistake of saying “even though, IN REALITY, you have no more choice…”

      It’s not reality that is changing, it’s the context and level of description, so BOTH are “in reality.” Remember, you are acknowledging that it is “better” to describe actual real world human phenomena
      like decision making in terms of “counterfactuals” and the like. But if what you are describing isn’t reality…why could it ever be “better” to use such descriptions?
      We want to describe reality, right? Essentially you’ve said it is the right thing to describe having a choice…when “in reality” we don’t have a choice. Which doesn’t seem to make sense.

      I think it would be more consistent to acknowledge that both describing the physics that underlie human behavior…and human behavior at the decision-making level, are two ways of describing reality…different levels of describing the real….rather than taking only one as being “reality.”

      1. Disagree. Typically, “freedom” means a lack of *certain* constraints.

        As the old saw goes, you can will what you want, but you can’t want what you will. The will is that which constrains your actions, and the will is utterly powerless to constrain itself. There are exactly zero degrees of freedom for the will — and, again, that’s true not only in the real world, but even in fantasy theological constructs of disembodied ethereal souls.

        It’s not reality that is changing, it’s the context and level of description, so BOTH are “in reality.”

        The one context is the imagined inner virtual reality in which we perceive our decision-making process. But to claim that that is itself as deserving of the appellation, real…well, why stop there? Why not declare Nevernever Land, Oz, and Tattooine to be real as well? You can point to all sorts of actual real manifestations of those concepts — but only in the context of the imagination.

        That’s perhaps a better way of getting to the heart of the divide.

        We can both agree that we have the perception of playing fast and loose with the inviolate determinism that underlies existence. And I think we can both agree that, on occasion, such as in a dream, we have the perception of defying gravity (flying) or otherwise wreaking havoc with physics.

        The question is whether we wish to describe those phenomena as, “real.” And it should be obvious why I argue so vehemently against such as description.




        1. Ben,

          I don’t see that you answered my question.

          Again: you wrote: “you’re better off describing decision-making in terms of counterfactuals and the like”

          And I agree.

          But then you start implying that one would be refering simply to the imaginary and not to the real. And if that’s the case, as I asked, WHY would it be “better” to describe
          decision making in terms of counterfactuals and the like? If you don’t think it’s “better” to do so, because wouldn’t be describing reality, then why did you write that in the first place? That’s the contradiction I’m getting at.

          My take, as I’ve gone to in mind-numbing detail before, is that we can and often do make true statements about reality, via counterfactual and possible-world descriptions. It’s a true description about the nature of our world to say IF you put the glass of water you are holding in the freezer, the water will freeze. You don’t have to actually put the water in the freezer for that to be a true statement, don’t you agree?

          If you don’t agree, then all the If/Then and counterfactual conceptualizing by which scientists describe fundamental physics would be just as imaginary and illusory, hence this undermines any appeal you’d make to how physics work.

          It is by exactly the same form of understanding that I would say “IF I desire to walk to the store I COULD do so, and IF I desired instead to drive I COULD do so.” These are If/Then truth statements, and can be empirically demonstrated just as one would demonstrate similar claims for physics.
          And in just this way we can decide whether they are empirically justified (by testing the claim and seeing the evidence) from If/Then truth statements that would be false or imaginary – e.g. IF I jump hard enough I would remain levitating in the air is going to be fantasy.

          There’s no jump to “illusion” or untruth when moving from talk about physics to talk about things humans can do in various situations. But this becomes incoherent if you take the stance, as you seem to, that one is talking only about illusion and the imaginary, when entering the realm of If/Then thinking and decision making.

          1. The problem comes down to the fact that “Free Will” is supposed to be a fundamental property of either human nature or the Universe.

            “All else being equal,” yes, you can certainly choose between chocolate and vanilla. But we know for a fact that all else is never equal, at which point it becomes much more complicated. “If I have coffee with dessert I’ll have chocolate ice cream, but if I have tea I’ll have vanilla.” Well, now are you going to have coffee or tea? Depends on if they’ve got your favorite tea blend. But if somebody puts a gun to your head and tells you to order the coffee, even if they’ve got your favorite tea, you’re definitely going with the coffee. Unless you know that the person with the gun is an actor who’ll pretend to shoot you if you order tea, and the script calls for you to pretend to get shot.

            “All else being equal” thus becomes quite obviously a purely imaginary construct. Yes, it’s a wonderful conceit to aid the decision-making process…

            …but even the most cursory after-the-fact examination of how we actually make decisions reveals that, in the circumstances that actually presented, we didn’t have a choice and our decisions are fully constrained by circumstances.

            Indeed, “Free Will” amounts to no less than an assumption that you can influences the circumstances which determine your decision — a clearly preposterous proposition.

            That is, yes, in a counterfactual sense you could choose either chocolate or vanilla…but in what sense can you choose the circumstances that exist at the time you place your order that narrow the choice down to the only option you actually have at that moment?

            When you’re telling the server you want chocolate, at that moment where the rubber meets the road, you really don’t have any remaining possibility of choosing vanilla. And a dispassionate analysis of the events leading up to it will reveal that vanilla never was an option to begin with — even if the next night a similar set of events plays out and you have vanilla then, for on that night vanilla was, ultimately, your only option.

            So…yeah. Arguing for “Free Will,” in whatever form, must ultimately equate to an argument against determinism, because, in reality, all else is not equal — even if there’s a lot of rhyming going on.




            1. The problem comes down to the fact that “Free Will” is supposed to be a fundamental property of either human nature or the Universe.

              We are putting aside for a moment the question of definitions of “Free Will.”

              The question we are focusing on right now is: If we assume determinism, how ought we to conceive of and describe human choice-making?

              “All else being equal,” yes, you can certainly choose between chocolate and vanilla. But we know for a fact that all else is never equal, at which point it becomes much more complicated.

              Ben, whenever you follow the logic you have begun there, I would ask that you keep in mind that you need to apply it to EVERY instance of empirical knowledge, and see if you are being consistent.

              A science teacher imparts this information about water to his class. Holding up a glass of water, he says “IF you place this water in below zero degrees Celsius, the water molecules will begin to lose energy, move more slowly until it becomes what we call solid ice.”

              Now this imparts actual information – that is truth – about the nature of water and how it reacts in freezing temperature, right?
              And it’s precisely the type of language scientists use all the time to describe truths about any specific phenomena. Sean Carroll’s book, like any book on science, is full of knowledge provided in this manner.

              If you think empirical claims of this type are unjustified because “all else is never equal” then you have just thrown away scientific knowledge – at best you leave only “X happened” available to science, and have done away with knowledge we use to manipulate the world. (But then, you’d be left unable to explain our predictive powers).

              We both know that you can’t really undermine empirical description in this manner without catastrophe. And we understand that we can talk reasonably, even truthfully, about probabilities even if we don’t know an outcome for sure. One really does have a better grasp of how the world IS to predict that not putting someone under anesthetic will likely cause them pain when sawing off a limb. Or that someone pulling out a gun and yelling ‘Allahu akbar’ on a commercial American Airlines flight is going to alarm the passengers. Someone not KNOWING that probability would be deemed amazingly obtuse and ignorant. We would not think a person denying these probabilities is reasonable to say: “I don’t think those predictions are justified, because all things are never equal”

              So, again, I’m saying we allow for THE SAME
              type of empirical talk about our choice-making: IF I wanted to go outside for a walk I COULD go outside for a walk is precisely the same type of empirical claim that the scientist makes when he says “IF I put water in the freezer the water would freeze.”

              So either:

              1. You recognize this logic and acknowledge that speaking of choice in this way is fully compatible with determinism and allows us to speak truthfully and coherently about options.

              Which is actually what it seems you started out agreeing with. You deny this at pains of denying scientific knowledge.

              But then you can’t turn around and claim (on an after-the-fact examination):

              2. That we “really” DON’T have a choice in our decisions.

              Now you have contradicted #1.

              Why can’t you just say that #1, using If/Then and counterfactual reasoning can give us truth statements, not merely “illusion?” You implicitly know this already, because you start off acknowledging it seems necessary and useful…but then you are so dead set on denying “freedom” and “choice” that your follow up steps deny it again.

              It’s the way you are using “really” and “reality” that messes things up: you apply that term only to one mode of thinking, implying the previous was about falsehoods or illusion.

              If you said something along the lines of “We DO have choice in the SENSE understood within counterfactual reasoning, but we DON’T have choice within the sense of contra-causal reasoning” then you would be making sense.

              1. “IF you place this water in below zero degrees Celsius, the water molecules will begin to lose energy, move more slowly until it becomes what we call solid ice.”


                And just how free is the water to chose which phase (liquid or solid) to be in?

                If you understand why questions of freedom and choice are absurd in the context of water, you should also understand why questions of freedom and choice are equally absurd in the context of the human decision-making process.

                Put it in terms of conditionals, if-this-then-that, and all is good. You only go off the rails when you ignore the fact that there are rails — which is another analogy. Can a train “choose” to go anywhere other than where the tracks lead? Can it “choose” to switch tracks? No, clearly; if the train is on a particular track at a safe speed, it will follow that track. If a switch directs it to another track, it will follow that one instead.



  16. I’m going to have to read through that a bit slower than most to make sure I get the gist of it 🙂

    I will say here though, that I am a non-believer in free will for the same reason I am a non-believer in other things…

    If having free will means being able to make a choice differently, then the evidence needed for that would be to show you could have made a choice differently… as we do not have that evidence (unable to travel back in time, unable to set up the same situation exactly once more) I am simply not convinced of the assertion we have free will…

  17. The way religionists answer JAC’s challenge above is that they usually dont: they go on the offensive. They say that whatever flaws the Bible has, we all know there is an ‘objective morality’ and that the only way you can have an objective morality is with a God. Otherwise everyone can choose whatever morality they like. I think this is a terrible argument but I’ve never seen a good refutation of it in a stand-up debate.
    I’d say the error begins by conflating two different definitions of the word ‘objective’ It can mean 1. arising from a higher level or 2. transcendent. The first definition applies to morality because it comes from the interactions of large groups of people but the second doesnt. Morality isn’t written into the fabric of space-time or in stone by a God.

    1. They say that whatever flaws the Bible has, we all know there is an ‘objective morality’ and that the only way you can have an objective morality is with a God. Otherwise everyone can choose whatever morality they like. I think this is a terrible argument but I’ve never seen a good refutation of it in a stand-up debate.

      The trivial response is to ask how they know that the Bible isn’t a cookbook.

      Then you can segue into explaining that morality is best understood as an individual’s optimal strategy (in the sense used in Game Theory) for success within a cooperative society. The knee-jerk response to that is that that somehow must mean that cheating is your best bet, but the very first discoveries in Game Theory were that cheating is your worst bet for a long-term strategy, and that cooperation and mutual support (but warily…”trust but verify”) are far more effective.

      Once you put it into that perspective, it becomes plain that morality really must emerge as a balance between the individual and the society. If the individual wants support and safety from the society, the individual must support and not threaten the society. If the society wants support and safety from individuals, the society must support and not threaten individuals. Each individual is going to wish support in different ways, and each collection of individuals is going to have different collective wishes — meaning that the particular actions that are considered supportive or threatening are going to be relative to the particular individuals and societies. But the basic framework by which that can emerge is simple and obvious and universal.




      1. Suppose there is a society that is the only collection of conscious beings in the universe. Suppose also that the majority in this group has realized it must cause a maximum of suffering to an unwilling minority of its members in order to allow the next generation of individuals to be born. I’d argue that there are situations in which it is morally right for that society to all die childless, ending the existence of conscious life. I’d also argue that my previous statement shows an existence of non-emergent, objective morality–perhaps it would be an arbitrary choice between right and wrong, but the choice would be made via an axiomatic definition of rightness that is unfounded on game theory. (It doesn’t require any divine revelation either.)

        1. It’s trivial to construct imaginary hypothetical situations which haven’t the slightest bearing on reality, and follow those through to equally absurd conclusions.

          Frankly, I don’t care one whit about whether or not Martians should or shouldn’t torture their hatchlings.

          What I do care about is that, in this society, the real one that we all live in…in our society, we most emphatically should not be torturing anybody, that we should provide a sound safety net for all people (including access to health care, housing, education, and so on), that our criminal justice system should make its first priority protection of society from criminals and its second priority as-rapid-as-possible reintegration of criminals safely back into society, and that sort of thing.

          Let the Martians worry about what sort of society works best for them.

          (But I’ll note: you specified, “unwilling minority,” and that right there tells you that their solution is decidedly suboptimal. Reasons why should be obvious and are left as an exercise….)




  18. I am in the class B camp (my blood is B neg, so I can’t help it 😊).

    I think educating the majority who are in the class A camp is the more important task and it will have a large effect on society if accomplished.

    Most compatible lists feel very strongly about their view. I feel that the huge multitude of inputs available to make a decision is what makes us feel like there are alternative choices to be made, but only one selection will happen with identical inputs.

  19. To me, the latter argument, B + C vs. A, is of vital importance for making society better

    Why it makes sense “to make a society better” when its outcome is predetermined by natural laws? You are not able to choose a “better future”. You are just a marionette as Harris would say. You are in an illusion that you can!

    Better reverse the question based in this below:

    …all of us have a strong feeling of agency…

    On how this feeling is based in a natural reality we have to discover and not how it is an illusion whatever its strength (universally admitted).

    Here is a serious mistake in your thinking. You think like a Judge who judges according man made laws absolute at the moment of decision. Natural laws is not like that, are not absolute as they are what we know or understand about them. Knowledge is not final. Far from that! We have not met the “end of science” already.

    I cannot evaluate matters about quantum mechanics but appears that can give ideas. Like an article in Aeon with title “Life is quantum. Weird quantum effects are so delicate it seems they could only happen in a lab. How on Earth can life depend on them?” by professor Johnjoe McFadden of molecular genetics at the University of Surrey: https://aeon.co/essays/quantum-weirdness-is-everywhere-in-the-living-world

    Some theoretical physicist (Luboš Motl:) did an extensive argument about “free will” and you commented “not very good, based on quantum mechanics”. Not very good but good? What this means? But here you ignored to refer to this.

    I think that some time questions can really improve. Is not against… natural laws, I suppose! Else why to bother again and again?

  20. It seem to me that the debate between B and C is really about the societal implications of rejecting A. Hard determinists think there are huge implications for rejecting A; compatabilists think there aren’t any implications for rejecting A, because we can still have the kinds of free will worth wanting. Hard determinists therefore think that attacking A is urgent and important, whereas compatibilists don’t really care much about attacking A. The debate between B and C needs to be resolved, because that will determine whether attacking A is viewed as a priority.

    1. Very well put! There’s another advantage to compatibilism, which is that it’s a lot easier to let go of the anti-scientific ideas in A, once someone realizes how little societal implications they have.

  21. To start with, I think your definition of hard determinism (A) is somewhat muddled. You acknowledge the possibility of quantum indeterminism, but still insist there’s only one possible outcome to any choice, and make that uniqueness a cornerstone of your argument. This has always been a weakness in your case against libertarianism, which logically does not require a single unique future. All it requires is that our decisions have physical causes, whether those causes are deterministic or not. The insistence on unique outcomes is a red herring.

    Sean Carroll, a compatibilist and an Everettian, certainly does not subscribe to the notion that “in a given situation, with all molecules configured identically, we can do only a single thing.” His view would be that we do all physically possible things, on different branches of the wave function. Rewinding the universe would not yield the same unique outcome; it would yield the same ensemble of possible outcomes.

    Also, in the interests of defining your terms, can you give a clear statement of what “environments” means in “genes and environments”? If (as I suspect) it means “all physical causes that aren’t genetic”, then how is “Our behaviors are solely and uniquely decided by our genes and our environments, and nothing else” meaningfully different from “our behaviors have physical causes”? If your argument is based on physical determinism, why bring genes into it at all?

    So my answer to your question in bold is that if we want to refute libertarianism (A), then we should use coherent, consistent arguments to do so. In my view, compatibilism (C) presents a more coherent rebuttal of (A) than your version of (B) does. As I’ve argued here several times, effective rehabilitation of offenders would seem to require an essentially compatibilist perspective on the possibility of reform and self-control. So that’s why the difference between (B) and (C) matters.

    1. Greg. Do you think that a choice that I could have made, but didn’t, is real in the same sense as the electron that went through the “other” slit is real under MWI? I would think that JAC does not really need “hard” determinism to defend his claim, even though he professes it. It seems to me that all he needs is “adequate” determinism which applies to macro (i.e., non-quantum) events. Perhaps you disagree that a brain making a decision is a macro event.

      1. I’m not arguing that making a choice is at bottom a quantum event in which both outcomes are realized. What I’m saying is that in EQM, the wave function branches so frequently that as a practical matter there are almost certainly going to be branches in which you decided differently due to cumulative chaotic interactions bubbling up to macroscopic scale.

        Quantum events have consequences on all size scales. If we rewind all the way back to the Big Bang, there are branches in which the large-scale arrangement of galaxies is different and the Earth doesn’t even exist.

        So in my view the whole notion of rewinding and obtaining a unique outcome is fundamentally misguided, since we know that physics doesn’t work that way. If there’s an argument for Jerry’s position on penal reform that doesn’t depend on this notion of unique outcomes, then I would be delighted to see him make it.

        1. Okay, we agree that the all-will-repeat-if-we-rewind-to-the-big bang scenario is very remote. Let me ask about a specific case that bothers me. On August 21 2017 the sun’s shadow will pass over central oregon. If we rewind to the big bang, this may or may not occur again, but as of now the eclipse is all but certain. To me, whether I choose to travel there to observe is uncertain, and up to me. Is my choice to observe (whether yes or no) in fact as certain as the eclipse but simply harder to predict and unknown even to me at this time, or is it in some fundamental way different than the certainty of the eclipse? If my decision is as certain as the eclipse but simply not predictable, it seems JAC is right.

          1. If you consider things like cosmic ray impacts on DNA and microelectronics, it seems to me there’s ample scope for quantum events to affect the course of human affairs. Computer glitches, dropped phone calls, unplanned trips to the dermatologist, and so on all have the potential to fork history onto an alternate path.

            So no, I see no reason to think that your eclipse-watching trip next year is as certain as the eclipse itself. There are any number of microscopic causes that could snowball into macroscopic effects between now and then.

            But if you’re still in doubt, download the app and split the universe yourself!

            1. Cool app. Better than pseudo random, I hope.

              Yes, I understand that I am more likely to die than the moon is to leave orbit. I meant controlling for those differential risks and considering only my decision to observe or not to.

              1. But you don’t make your decision in isolation; you’re part of a network of social influences. Suppose a coworker takes time off to visit a sick relative, leaving your office understaffed. Might that affect your travel plans?

                It doesn’t take a catastrophe to change the outcome. The world is constantly forking, just in the nature of things. Once forked, the branches evolve independently, providing subtly different inputs to your meat computer. Eventually you’ve diverged to the point where you start making different decisions, for reasons that seem completely reasonable and natural to the version of you that makes them.

              2. But, if you accept (adequate) determinism, the relative getting sick and my co-worker taking time off is determined by antecedent events, which are in turn determined by antecedent events, etc. I just don’t know about these events yet, but the universe will unwind and I will choose to miss the eclipse as deterministically as the eclipse itself. An omniscient demon could predict it.

              3. My point is that in general you don’t have to trace chains of antecedent events back very far to find a cosmic ray or other quantum event. If your coworker’s relative’s illness is cancer, we’re probably there already.

                Divergence of histories is not a rare occurrence; it’s the norm. That’s the real truth about determinism in a quantum universe.

          2. Your choice to observe the eclipse could easily fall into a third category: “certain” (“in principle” predictable) to someone else but “uncertain” to you, precisely because you get to decide it. It’s one of the “paradoxes” of self-reference.

            Try this trick: you fill in a circle to make a pie chart, where the colors of the subregions represent the fraction of the pie chart’s area occupied by that color. Your “predictions” for the pie chart will be interpreted by this rule. Meanwhile, I’ll try to predict your pie chart’s fractions, using another pie chart.

            Note that you can put any damn colors you want, in any proportion, and your “prediction” will be correct. I cannot. I actually have to do the hard work of prediction – because I’m not the decider.

        2. Even if Everett is right, all it seems to mean is that there are many “counterparts” (to borrow a term from D. Lewis). All of *those* had no self-origination power either; they just found themselves doing one action rather than another. (See also S. McCall’s version of many worlds, which unfortunately also suffers from the Kane problem, but at least is clear as to the goal.)

  22. If one is convinced that others are “just physics” then one has no reason to regard them as conscious. Then there is no reason to imagine that they “believe” in anything, free will or otherwise. All of their behavior are phenomena caused by the big bang. At the formation of the galaxy – no such thing as consciousness or conscious agency. Same thing today. As for one’s own consciousness – this can be filed under solipsism. It cannot be examined using the scientific method any more than the consciousness of one’s neighbors.

    1. Agreed. Nothing special about consciousness, but to borrow from Nagel: What’s it like to be an electron?

      From what we know, electrons do not have beliefs and do not care about free will. Could an electron tell the difference if the universe had free will? Likewise, could we determine, as humans, if an electron was ever acted upon by an agent of free will? And how would we know, using the electron as a tool to measure such an action?

      1. Sean Carroll addresses exactly those questions in his latest book. And the short answer is that modern physics has performed a complete accounting of the degrees of freedom available to electrons (etc.) and the forces that can act upon them. They behave exactly according to the modern theory describing them, with no observation ever performed indicating otherwise and no room left over for unknowns or other sorts of mysteries. Your only option is to propose that everything we think we know about physics etc.) is worng; that some incomprehensibly vast conspiracy holds instead.

        For details, his book is most highly recommended….




        1. I can second the recommendation on Carroll’s new book. While it goes over many issues I’ve seen before, it is a great summary of the extent of human knowledge. Certainly he deals a death blow to non-materialists of all sorts, for those who thought they were still even barely alive and standing.

      2. There is *everything* special about consciousness – so much so that we cannot say empirically whether it exists. So it is especially transcendent or especially imaginary. It is certainly not an ordinary attribute of a system like temperature.

  23. I very much agree with you. I find myself leaning more and more to hard determinism, just by the fact that I treat any sort of extra-physical existence with the same skepticism that finally led me out of Christianity. Like you, I think that the drive behind compatibilism is the feeling that when I make a choice of any sort, I’m doing so with full agency, and with the full spectrum of possibilities at my disposal, but that I’m freely making that choice, and not being forced in any sense – except that the synapses that drive my choice are in fact those very natural processes that make that choice the only one I would ever make given all other conditions remain the same.

    And I actually think this is one of the KEY issues in this country – that getting over this meritocratic hurdle of “rugged individualism” is absolutely necessary to solving a number of our societal problems and breaking the hold of big money on our government.

  24. I agree with you.

    And I am a B.

    Coel, at 5, above likens the human mind to a chess-playing computer. But, unless the programmer inserts a random number generator into the program, the chess playing computer will make the same move every time, for a given configuration of the board. The fact that cogitation happens first doesn’t change to outcome.

    The outcome is the product of the boundary conditions and operator’s configuration (program).

    To me, free will is just the name we give to the feeling you get when your brain has made a decision for you.

    1. But if, as Jerry argues, the overriding goal is to refute libertarian dualism, then I suggest that drawing artificial distinctions between “you” and “your brain” is a poor way to go about it.

        1. Computer science can help, a bit.

          The brain is obviously the hardware. The wiring of the neurons can be thought of as the software, and the thoughts would be equivalent to the running state.

          But anybody familiar with actual computer systems will recognize that these distinctions are somewhat arbitrary even in digital computers, as everything is a matter of physical states, just with some states more permanent or fleeting. When you flip a bit, you’re actually changing the voltage in a memory location just as if you had a teeny tiny lever mechanism that moved a bunch of electrons from one spot to another. But that’s as useless a model of computer science as a listing of the positions and vectors of all the air molecules in the room is of fluid dynamics.

          Just as the Ideal Gas Law is a better way of describing the air in your room, your favored computer language is a better way of describing the actions of your computer…and the language of psychology is a better way of describing thoughts and desires than a listing of the electrochemical states of all the neurons of your brain.




          1. Humorous side note: deviations to the Ideal Gas Law (van der Walls and other Virial equations of state) give rise to phenomena that, to the best of our ability, defy prediction. As surface scientists jokingly concede, the Devil lives in interactions.

            1. I use the derivation and analysis of the ideal gas law as an example of an abduction when doing argumentation schemes to a sophisticated audience. Is there a good reference for explorations of the “phenomenal” constants in the van der Waals law? My (admittedly) short collection which discusses it has them just produced by tables and measurements (no theory to speak of, except to explain why one would expect the deviations from the ideal).

    2. But … the chess playing computer will make the same move every time, for a given configuration of the board.

      The compatibilist entirely agrees with that, which is not in any way an argument against compatibilism.

      The fact that cogitation happens first doesn’t change to outcome.

      The cogitation *determines* the outcome.

      1. In the chess-playing computer, the cogitation happens in a programmed way. The exact same way every time. (Given the same boundary conditions.)

        There’s no freedom in there.

        And if the programmer changed the order in which the program executes the various simulations before arriving at the final output, it would just run that tape each time.

        The length of time the program runs before the outcome is reached (or the number of program steps or the CPU run time) is not a metric of freedom.

        1. In the chess-playing computer, the cogitation happens in a programmed way. The exact same way every time. (Given the same boundary conditions.)

          Yes, the compatibilist agrees with you 100% there.

          There’s no freedom in there.

          You are using a “libertarian” conception of “freedom” in which it means action outwith the laws of physics. You’re right, there is none of *that* sort of freedom there.

          But the compatibilist has a different conception of “freedom”, one that is entirely compatible with determinism and with your first statement above.

          If that sounds peculiar, realise that every other usage of “freedom” does *not* imply something acting outwith physics. For example, if I describe an object as in “free fall”, then its motion is 100% determined by physics, exactly as the chess-playing computer is.

          1. OK, I think I understand your meaning.

            I don’t find that for me it has any meaning as a type of freedom with regard to free will. Or as a kind of will.

            We do things.

            I agree that it is a different conception of freedom: One that is very analogous to the degrees of freedom described in computer modeling of physical phenomena, for instance.

            I have a (strong) feeling that my consciousness is directing my actions. But science tells me it is not. I’m certainly more complex than a thermostat; but adding degrees of freedom of that nature doesn’t add up to the feeling I get.

            Or, rather, it does, since all are phenomena of my meat computer; but it isn’t apparent at all — it’s all below my level of perception (hence freedom and will as people normally understand them aren’t in it).

            I’m not sure redefining freedom and will is helpful.

            1. What science tells me is that my conscious experience is a physical phenomenon that is inextricably part of the causal chain of behavior. I find it baffling that hard determinists seem to think otherwise.

              1. But you are not responsible for your experience, so even if there is “top down” there, it doesn’t do you any good for the question of self-origination, and hence for responsibility. (Or such is the argument.)

              2. I guess I’m not much impressed by the alleged “problem” of self-origination. Self-authorship doesn’t have to be an all-or-nothing proposition. We can presumably agree that Jerry is the author of his own books, without also requiring that he be the inventor of the English language. (And if Jerry himself sincerely believes he’s not the author of those books, why is he touting them in the sidebar of this website?)

      2. The cogitation is itself also determined. But the thing is, no one calls a chess-playing computer “free”. (Well, maybe you do but it’s a neologism at best.) You seem to equate “freedom” with, e.g., the number of open moves in chess. But that is a discrete number of options considered based on the programmed “rules”. So an altered program where every piece moves like a queen has more freedom? So a go-playing computer has more freedom than a chess-playing one?

        It’s then not clear what this has to do with human will. If I can think of more options in a given situation am I more free? If I’m physically handicapped am I less so? It seems like you are interested rather in complexity, loosely speaking, and I don’t see the reason to drag the word “free” into it, especially in the context of a debate with libertarian free willies. What exactly is the argument *for* compatibilism? Why is it important to preserve the phrase “free will” as opposed to the “hard determinist” who can simply say “no, that’s not a thing”?

        You single out coercion as an example of reduced free will, which seems to be a desired outcome of your scheme, but on your account coercion isn’t special: my “freedom” is equally limited by a rock in my path. And of course, coercion doesn’t really limit my number of choices, it simply changes the outcomes, as again would any number of other extraneous or internal factors. Supposing though, that you have some coherent theory of how this is all supposed to work as a theory of ethical responsibility, say, there is still no obvious reason to adopt the “freedom” terminology which is clearly not the standard usage in these contexts.

        In the end I don’t see how compatibilism explains anything or resolves any problems. It just seems to be a rejection of non-compatible determinism while claiming to agree on all the salient points.

        1. So an altered program where every piece moves like a queen has more freedom?

          Yes! This is the “degrees of freedom” concept as used in physics and engineering.

          If I can think of more options in a given situation am I more free? If I’m physically handicapped am I less so?

          Yes. If you’re able-bodied you can decide to stand or sit; if you’re disabled you might be unable to stand. You have less freedom to act. Just as someone in jail has less freedom to be where they want to be.

          This is, of course, not the *libertarian* concept of freedom, it’s a very different, and also very common, conception of freedom.

          What exactly is the argument *for* compatibilism? Why is it important to preserve the phrase “free will” …

          It’s not that important to preserve the particular phrase “free will”, but we do need concepts of “choice” and “decision”, and we need to understand what they actually are.

          Incompatibilists never proceed to dealing with those concepts, though if you listen to them talk they say “choose” and “decide” and “blame” and similar things every day.

          … but on your account coercion isn’t special: my “freedom” is equally limited by a rock in my path.

          Yes, or the walls of a jail cell.

          … there is still no obvious reason to adopt the “freedom” terminology which is clearly not the standard usage in these contexts.

          Actually, in all other contexts *except* when discussing the philosophical issue of “free will”, the compatibilist account of freedom *is* the standard meaning!

          1. “Yes! This is the “degrees of freedom” concept as used in physics and engineering.”

            I’m nitpicking a bit, but no it’s not. Degrees of freedom refers to how many independent variables one has in a model. Your idea is far too vague to be equated, it seems to include the range of variables also, for example.

            “Yes…This is of course not the libertarian concept…Yes…”

            All of this just agrees that your usage of “freedom” doesn’t match up with the normal usage in the context of wills and decision making. We *don’t* normally say that my free will fluctuates wildly throughout the day as I encounter new situations, think of new courses of action, stub my toe, etc. So I still don’t see the point.

            “…we do need concepts of “choice” and “decision”, and we need to understand what they actually are.”

            But nothing in your discussion has elucidated those terms. I make a decision between two options or a thousand. And the difficulty/complexity of the decision may be harder between two choices where I have to carefully weigh many factors than between a thousand simple options with one clear winner. And really, unlike with degrees of freedom, there is no real theory here of how to compute my degree of choosiness or how that should enter into responsibility.

            “Incompatibilists never proceed to dealing with those concepts…”

            Okay, now you’re just making things up. 🙂 Jerry has talked about these things but more generally, there is nothing about being an incompatibilist that precludes discussing them. I make choices, they are determined. I am a choice-making machine. My choices are not free.

            “Actually, in all other contexts…”

            Citation needed! But anyhow, this is like saying that in all other contexts “fired” means “deliciously flame-broiled” when we are talking about jobs. We’re talking about free will, not free samples at the bakery. Which, come to think of it, is different from “free of worries” or “free jazz”. All of these can perhaps be related to a general notion of “without or in the absence of” something, but A) there is no useful way to quantify absence of, what?, limitations/options/coercive forces/imagined scenarios in your scheme and B) free will means “absent complete deterministic constraints” to most people so there is no language argument down this route.

            1. I’m nitpicking a bit, but no it’s not. Degrees of freedom refers to how many independent variables one has in a model.

              Or, for example, how many ways in which a gas molecule can translate, rotate or vibrate. Which is pretty close to the usage I’m giving.

              But nothing in your discussion has elucidated those terms.

              True, I didn’t elucidate it in that comment. But INcompatibilism doesn’t have a concept of “choice” or “decision”. If incompatibilism develops concepts of “choose” it immediately becomes exactly what compatibilism is!

              Jerry has talked about these things but more generally, there is nothing about being an incompatibilist that precludes discussing them. I make choices, they are determined.

              Yes, your choices are determined; both compatibilists and incompatibilists agree on that.

              Now, what do you, as an incompatibilist, mean by “choose”?

              1. “Or, for example, how many ways in which a gas molecule can translate, rotate or vibrate.”

                That’s just the independent variables in your model of a gas. Like I said, it has a specific, useful meaning and so far your usage is vague and unhelpful.

                “If incompatibilism develops concepts of “choose” it immediately becomes exactly what compatibilism is!”

                No, the difference between compatibilism and incompatibilism is not whether or not “choice” can be discussed, it is whether “free choice” exists. For example:

                “Now, what do you, as an incompatibilist, mean by “choose”?”

                Loosely speaking your chess-playing computer is a fine analogue of incompatibilist choice, and as mentioned above, people don’t call the computer an example of free will. In more detail: your brain simulates the world around you and particularly it tries to simulate the future. It runs various scenarios with different simulated actions on your part and acts on the one that seems to optimize achieving your desires. You consciously experience that decision making process. From the outside the outcome is fixed, but of course from the inside your brain can’t know the conclusion it will reach before it reaches it, nor does it necessarily know consciously the criteria that determine the outcome. Nothing about this is “free”. The difference between you and a rock is the complicated attempt to simulate the future and optimize your actions with it. That makes conversation with you more interesting than with the rock, but complexity doesn’t add freedom.

              2. Joshua Sayre,

                Your reply is usually what we get from incompatibilists…and as usual it misses the
                real issue Coel was asking about. Your attempt to describe the decision making process leaves out the very crux of the matter.

                When you write: “your brain simulates the world around you and particularly it tries to simulate the future. It runs various scenarios with different simulated actions on your part and acts on the one that seems to optimize achieving your desires.”

                That doesn’t answer the question: Are the thoughts our brains are thinking when making a decision TRUE? And to answer that question, you still have to answer a deeper question not simply about “making a choice” but about “HAVING a choice.”

                What does it mean to say you HAVE a choice?
                E.g. a choice between steak or seafood at the restaurant, or any other situation of choice-making we encounter? This entails explaining whether or not it makes sense to say either option is “really” possible and therefore what do you mean by “really possible.”

                If you don’t think that when we have a choice that alternate options aren’t “really” possible…then for one thing you are not using the word “choice” in the way it’s normally understood. You have to re-define it. But more important, you will have to re-define it in a way that when one is offered a “choice” one isn’t actually offering “real” alternatives. But how in the world will that work coherently given all the “choice-making” we must do each day?

                These are the aspects of “what does it mean to have a choice?” that your description of brain processes do not answer…yet.

                The compatibilists argue that once you DO start making sense of the word “choice” and “people HAVING choices” within a deterministic context…you end up speaking the language of compatibilism…which also happens to be pretty much everyday language.

              3. That doesn’t answer the question: Are the thoughts our brains are thinking when making a decision TRUE?

                That question can only make sense in one of two contexts.

                Most charitably, it’s a question about how well the mental models our brains have align with the reality of the Universe…in which case it’s clearly irrelevant save for analyzing the effectiveness of the decision-making process.

                But to me it reads much more like a Platonic reification of minds, which is something I’m more than prepared to reject outright without further consideration at this point in my life.

                E.g. a choice between steak or seafood at the restaurant, or any other situation of choice-making we encounter? This entails explaining whether or not it makes sense to say either option is “really” possible and therefore what do you mean by “really possible.”

                This supports my suspicion of superstition, for it comes with it a very heavy implication that choosing involves manipulation of the circumstances that cause choices to be decided the way they are.

                When you open the menu at the restaurant, it is certainly (often) the case that you don’t yet know what you will order; yet, it is equally certain that you will order either the steak or the seafood. (Please sweep under the rug all other possible options, such as leaving without ordering, a plane crashing into the restaurant, and so on.)

                The decision-making process is wholly deterministic and the various circumstances that present at the time you sit down will get filtered through your decision-making process to lead you to the single decision you’ll ultimately make. As part of that process, you’ll likely imagine how you expect the stake to taste and, separately, imagine how you expect the seafood to taste. Those factors will certainly influence your decision — as will the waiter’s presentation of the special of the day, your thoughts and memories before you entered the restaurant, the condition of the lobster tank you walked past in the lobby, and so on.

                Your choice is the Rube Goldberg contraption of the ball bouncing down all the different pathways until it lands in either the steak or seafood cubbyhole at the bottom. You can’t predict which cubbyhole it’ll land in until perhaps the very last moment…but that doesn’t mean that the ball itself has any say in the matter.

                It’s convenient to describe that whole process as choosing between steak and seafood…but the unavoidable determinism at every step should make clear that any sense in which both steak and seafood are somehow both “really” possible when you sit down is purely an artifact of ignorance and has no basis in reality whatsoever.




              4. Ben,

                Yet again…the answers you are giving are (I have argued) like those given by Jerry to these questions: they miss the gist of the problem.

                It’s a question about our concepts, not about the mechanistic way concepts arise, but about what our concepts MEAN.
                You can describe in an objective fashion HOW someone’s mind produces a belief or argument, but that doesn’t answer
                whether the belief is TRUE or whether the argument is SOUND or not. To do that, you have to engage in the logic of the concepts themselves.

                Your description of a mind being a Rube Goldberg contraption with bouncing balls could apply both to Ray Comfort’s “banana” argument for God, or to the Christian
                claim the bible is true because it says it is true, or to Jerry’s
                promulgation of the arguments and evidence for evolution. But that doesn’t tell us which of those arguments are poor, or uselessly circular, or sound.

                To do that, you have to get into what concepts are being employed by someone, and whether they are reasonable and consistent concepts, from which reasonable and consistent conclusions are being drawn.

                And I’m asking, granting determinism as we both do, what can it MEAN to say we have a “choice?”

                How will we navigate both conceptually and linguistically all the situations described by our current “choice-making” language and concepts?

                If you are a waiter offering the days specials “You can order
                the catch of the day, OR you can order the chef’s special lamb dish” what can you MEAN by this?

                If there is not some sense, some conception of choice within a deterministic context, in which you are pointing out options that are “real” for someone, then it seems you have rendered such talk to be illusion, nonsense. And if you tell your son he ought to be studying instead of watching TV, he can say “but Dad, I know you don’t think such choices are real, it’s just a bullshit way of talking, so I don’t need to take
                your recommendation as making any sense.”

                How do you MAKE SENSE, in terms of conceptual consistency, to your son when trying to describe other ways for him to behave and influence his behavior?

                Until you make sense, unfortunately (short of threats of physical violence) nobody needs to pay attention to you.

                Just like in other posts, when you write something like:

                “It’s convenient to describe that whole process as choosing between steak and seafood”

                Everything you are being asked to answer is being hidden in words like “convenient.” What could that mean? Does it mean talking of choices between alternatives is bullshit and fantasy…but hey it might influence someone anyway? Basically, lying. Which, first of all doesn’t seem something someone who cares about the truth would want to purvey, and second of all only works until someone notices you are lying.

                Or do you mean by “convenient” that talking of choice is an easy way to convey some real information, some truth, to people to help them decide? But if you mean that such modes of speech are “convenient” messengers for truth…then why in the world do you deny the compatibilist position I keep giving? And why would you keep saying that we don’t ‘really’ have options if our ‘convenient’ language of choice actually IS a way of talking truth?

                So you either need to answer what you would MEAN by telling someone she has a “choice,” or giving different options, or if you are going to give up such conceptual talk, explain what you’d put in it’s place. And objective descriptions of “how our brain works” isn’t going to fill that void, any more than the descriptive statement “some people are starving in Africa due to drought” can tell us whether it’s “good” that they are starving or not, or whether we “should” help them or not.

              5. If there is not some sense, some conception of choice within a deterministic context, in which you are pointing out options that are “real” for someone, then it seems you have rendered such talk to be illusion, nonsense. And if you tell your son he ought to be studying instead of watching TV, he can say “but Dad, I know you don’t think such choices are real, it’s just a bullshit way of talking, so I don’t need to take your recommendation as making any sense.”

                Now you’re arguing against fatalism, not against determinism. As I’ve noted elsewhere in this thread, as inspired by a link somebody else posted, determinists still look both ways before crossing the street.

                Everything you are being asked to answer is being hidden in words like “convenient.” What could that mean? Does it mean talking of choices between alternatives is bullshit and fantasy…but hey it might influence someone anyway?

                It’s convenient to talk of a balloon as having pressure, volume, a number of molecules, and temperature, all related by the constant R, but it’s very important to know that none of those descriptors are fundamentally real — even as they’re, in this case, superbly convenient.

                Or, it’s also convenient to teach young kids about the solar system model of the atom — but, only if, when doing so, you start with a disclaimer that this is a fundamentally flawed concept that you’re only using as a placeholder to build to a fuller understanding.

                So, it’s convenient to ponder if you’d rather have steak or seafood at the restaurant, but you do yourself a grave disservice if you fail to understand the fundamental determinism underlying your actions. In that context, “choice” is a much closer synonym for “ignorance” than it is for anything else.




              6. Temperature is sort of real. It’s the average kinetic energy of all the molecules in question. It’s a statistic that can be derived from a microscopic description, but in practice (since we never have a microscopic description) it’s usually measured indirectly.

        2. the “hard determinist” who can simply say “no, that’s not a thing”

          But it’s not that simple. Sarkissian et al. notwithstanding, it’s not the case that every use of the words “free will” in everyday speech is inextricably dependent on libertarian dualism. So it behooves us to find out what sort of phenomenon or behavioral capacity “free will” refers to in any given instance before unilaterally declaring “that’s not a thing”. Compatibilism tries to do that, and that’s why it matters.

          1. “Compatibilism tries to do that, and that’s why it matters.”

            Linguists and psychologists try to do that. Compatibilists try to preserve the usage of “free will” in philosophical discussions. I note one shouldn’t assume that usage connotes a particularly coherent concept, it doesn’t in the libertarian case.

            But if you’re confused, just ask! E.g., one signs a confession “of my own free will” to aver that it was not forced by threats. Now we might have a long discussion about what does or should constitute illegal coercion but the phrase ‘free will’ and the compatibilist attempts to analyze it add nothing. One helpful point is to note that “free” is superfluous here, any time you can remove it, I say do so. An incompatibilist can talk all day about one’s will.

            Regardless, the question “do we/robots/animals/etc. have free will” is not about legal documents.

        3. About to put myself on the spot.
          Even ‘science” does not explain anything. It just describes in finer and finer detail.and allows us to conjecture further realms of discovery. Strict determinism is very hard to argue with-and seems to provide a fertile ground for enquiry,You come up hard against such questions as-‘What is the nature of mind and conscioussnes’ or-‘ Is there a ‘self’ ?. Determinism leaves one no alternative but to dig deeper. Compatibilism describes how determinism spins out and I agree with several others here that it is human based and a matter of language. If we were to use a different word or have a deeper understanding of the question we were tackling the discrepancy would disappear. It is a bit like the use of teleological terms when talking about evolution. And I tend to think the careless use of teleological terms when discussing evolution for a ‘lay’ audience tends to muddy the waters and make a true insight more difficult.
          As complex beings with large and complex brains and sensitive organs of perception we have more ways of perceiving the world and processing our experience-but how all that spins out is determined by the prior condition of any and all of the multiple processes involved and their interaction. There is no ‘I’ or ‘me’, which can stand apart and direct the fork in the road to be taken. No homunculus taken to a mind level.

    3. The chess playing computer will only make the same move every time, for a given configuration of the board, if it is a relatively dumb program. Smarter ones learn from past mistakes; they bring some summarized knowledge of their history to each game. Humans definitely do so.

      1. Smarter ones learn from past mistakes

        …at which point they’re no longer the same program as when initially run.

        In other words, you’re missing the whole point of Jerry’s thought experiment.

        It’s not a question of whether or not, if you knew then what you know now, you’d make the same decision. We can all agree that, very often, you wouldn’t. As you point out, that’s pretty much the whole point of learning from experience.

        The question, rather, is, given the same conditions, including your ignorance and inexperience as well as all other environmental factors and all of everything, you would make the same decision. And, framed that way, the answer is a very obvious, “yes.” Any other conclusion at best represents ignorance of the fundamentally deterministic nature of physics and, at worst, presupposes some sort of time-travelling soul that can influence past choices of the self.

        Now, there’s another aspect of this to consider. When making decisions, we frequently “visualize” the results of different choices. And if we’re at all good at it, we will learn from our mental simulation of reality just as well as if we experienced the same situation in real life. As a bonus, in the theatre of the mind, we can go back and re-run the sequence of events in as many different permutations as we like, with all sorts of different decisions at any given point.

        And, subjectively, this feels exactly like what you’re naïvely misunderstanding the proposal to be, as well as being a perfect fit for Jerry’s summary of the folk definition of “Free Will” (rewinding the tape of the Universe and making a different decision). It’s a very real phenomenon, and vital to the human condition and experience.

        …but it also is an entirely deterministic one itself….




        1. I’m not missing the point of Jerry’s thought experiment; I’m just not very interested in it. I’m more interested in some of the ways humans (and some other animals) beat the pants off our best computers to date, at many tasks.

          For the record though, I’d hope that in Jerry’s thought experiment I’d make the same decision every time, if it was an important decision and I had some reasonable basis to go on. Libertarian “free will” has less than zero appeal to me. It would be a freaking curse.

  25. Jerry,

    I really appreciate your work. As well as your recent advocacy not just for good science but also your critiques against the regressive and censorious elements of the left.

    That said, I’m completely bewildered by this controversy over free will. And so much of this topic seems to involve talking past eachother.

    There is so much in this world that we don’t distill down to just physics. Though of course it’s entirely based on it. When a lion stalks a gazelle, we don’t refer to that in terms of physics. We refer to that scenario as a lion being driven by hunger and a proclivity to hunt engaging in one behavior and a gazelle being driven by fear and a disposition towards self-preservation engaging in another behavior. We don’t go down to the level of quantum mechanics to understand what is truly going on. Same as when a chimpanzee problem solved. Or a human for that matter. Again, we don’t default to understanding what is going on at a base quantum level.

    Same with “free will.” What we are really interested in is agency. Is there such a thing as agency. In the same sense as is there such a thing as any of what I mentioned above.

    I suggest there is. If you disagree, then the alternative would have to be something like epiphenomenalism. And then we should focus the conversation on that. Do mental events that come into play in conscious deliberation have absolutely no direct role in our decisions? Rather we make our decisions without agency prior to any deliberation and only experience the illusion of having deliberated after the fact. I suggest the answer to this is no. For most higher order decisions. With the caveate of perhaps yes for automatic, lower level ones.

    But if we admit to the role of agency, then there is no reason to have this debate. Yes our agency is still underlied by physics. But it is still “our” agency. Even if there is no magic outside of that which could cause our agency and its associated values and processes to be “other than it is.” If we don’t accept this, then we might as well throw everything else out as well. There is no “problem solving.” There is no “creativity.” And so on. And personally I see no need for that at all.

    1. Same with “free will.” What we are really interested in is agency.

      The heart of the debate can be distilled down to whether or not “Free Will” is to agency as “Elan Vital” is to biology.

      Essentially, the compatibilists are insisting that we mustn’t abandon the concept of “Elan Vital” at the heart of our understanding of biology, at the same time as some of them agree that “Elan Vital” and everything it is understood to be is nonexistent and / or incoherent.

      The incompatibilists are wondering why such primitive superstitions are still so prevalent, and why can’t we get past them?

      Considering that the libertarians are convinced that “Elan Vital” is the proper explanation for biology, it should be obvious why the incompatibilists consider the compatibilists much closer to the libertarians than the incompatibilists.

      After all, who amongst us would take seriously a biologist in this day and age who still explained Evolution in terms of “Elan Vital,” even if that biologist’s pet definition of “Elan Vital” was somehow twisted into something logically equivalent to the Modern Synthesis?



      1. Ben, even after all this time, you still haven’t appreciated what compatibilism actually is.

        You always, always, always interpret compatibilist FW as a hankering after libertarian FW.

        1. That would be because compatibilists keep insisting that “Elan Vital” — er…”Free Will”…is necessary to understanding biology. Cognition. Whatever.



          1. My reading of Carroll is a bit different. I think he said “free will” is part of a useful language for talking about the higher levels of reality in which we live day-to-day. If “Elan Vital” was merely useful in language needed to discuss life, it would be a supportable concept, but I think it really has no use. It is completely adequate to speak of life as a chemical process.

          2. No, it’s more like insisting that the concept “life” is useful for understanding biology, against opponents who insist that, because “living” things are merely composed of dead atoms, we should ditch the concept “life”.

            1. Coel,

              Beautifully put. I wish certain incompatibilists would frame what you just wrote and try to contemplate those words a few minutes per day, every day, over the course of a month. Perhaps then this would stop going in circles all the time.

            2. I think we can agree that “Elan Vital” is hopelessly bogged down with all sorts of nonsense that we know doesn’t actually reflect reality. And I think we can also agree that “life” is a perfectly reasonable term that has no significant baggage.

              I further think that we can agree that “cognition” and “decision-making processes” and “agents” and similar terms are as free of baggage as “life.”

              What I can’t fathom, for the life of me, is how you can fail to see that such a theologically and philosophically bogged-down and inherently self-contradictory term as “free will” is one worth attempting to salvage.

              I mean, really. Seriously, truly, honestly. Go do a survey asking people why bad things happen to good people, and observe just how integral “Free Will” is to their answers. And, while you’re at it, ask them if the legal phrase, “of one’s own free will,” is referring to the same phenomenon as the “Free Will” that explains theological incompetence and / or malevolence. Overwhelmingly, people will tell you that the two terms are at best tangentially related and really don’t have anything to do with each other.

              We both know what the results of that survey will be. And yet you want us to think that “Free Will” is the mechanism by which we make decisions, but you simultaneously want us to think that “Free Will” is not an incoherent theological construct.

              Sorry. Not buying it. Words have clear meanings, and you’re coopting these words to mean exactly what they’re explicitly defined to contrast with.

              You want to talk about the will, great — it’s a perfectly valid concept describing that which motivates human action. You want to talk about freedom, fantastic; there’s no better way to explain the notions set forth in the First Amendment. Just keep the two distinct — exactly as you would “marriage” and “bachelorhood.”




              1. To reiterate what Coel &al. have written repeatedly, in every other context free does not have a supernatural meaning either, so I fail to see why it should be different for free will. To reiterate what I have written repeatedly, out of my own free will is merely the direct, one-to-one Germanic translation of voluntarily (see freiwillig in my native language). There is thus no contradiction whatsoever between free and will, or at least not more so than between free and fall.

                I further think that we can agree that “cognition” and “decision-making processes” and “agents” and similar terms are as free of baggage as “life.”

                That is interesting given that even on this website incompatibilists have in the past argued on the lines of: we don’t make choices, our brain chemistry makes choices for us (which I, personally, cannot help but interpret as a mind-body dualist statement).

                If your whole beef is only about applying free will in academic philosophical discussions we have no serious disagreement, for the simple reason that such discussions are pretty irrelevant to me (other compatibilists may see it differently). What concerns me are the we don’t make choices style arguments, because they appear very confused about what “me” is, and if successful they would make discussing human decision making virtually impossible.

              2. “Agent” is enormously fraught. Is not an agent a source of causation – a white hole of effects? Causation proceeds from an agent, rather than cause-and-effect chains simply passing through – else we would not call it an agent.

              3. Hi Ben,

                … how you can fail to see that such a theologically and philosophically bogged-down and inherently self-contradictory term as “free will” is one worth attempting to salvage.

                You live in a highly religious society, whereas I (and other compatibilists here, such as Vaal I think) live in the much more secular UK.

                The term “free will” thus has far fewer of the theological connotations to us. Religiosity is your enemy, not compatibilism.

  26. Physics rules, all else is empty headed blather. Meat-bots we are born, as meat-bots we live our lives, as maeat-bots we die. We have as much free will as the bacteria that consume our corpse’s.

    1. Damien, they are not even conscious, at least not according to any result of the scientific method. We should be thorough – not only denying that beings have agency, but also denying that they have consciousness.

  27. Both, A and C, reject B for the same reason: A kind of belief in belief, the suspicion (openly stated by A people; convolutedly by C people) that without libertarian free will we will be moral monsters, that everything should be permitted and criminals should not be punished. The argument between B and C is important because C people provide the intellectual ammunition for A people to go on believing there is libertarian free will.

    1. I like that one. Very good. C people want to be just a little bit pregnant and we are pretty sure that doesn’t happen.

  28. I would agree that B+C vs A is presently more important in the social context, but the distinction between B and C is not merely semantics (although it is often made out to seem so). For me the big problem are the fundamentals of physics here, quantum level cannot simply be barred.

    The problems with hard determinism are that 1. it assumes we already have the Theory of Everything and 2. that even the bits we have are confusing to philosophers. E.g. Sam Harris’s recent podcast discussion with Dan Dennett makes it painfully obvious, that they do not distinguish clearly enough between classical and quantum randomness, and do not take into account the axiomatic status of probability in quantum physics (Frequentist vs Bayesian anyone?). So they tend to get their intuition from classical physics, which is not really random, forget that we don’t know which quantum interpretation is correct (if any), and speculate.

    It could very much be the case that rewinding the world will NOT lead to exactly the same evolution of the system, and we could give a precise description of “could have done otherwise” – not only in similar circumstances (which already matters for C), but in _exactly_ the same ones. It does not mean that the choices would be chaotic (in the popular sense) because the whole brain would weigh in in a consistent way, nor would they be fully determined by external influence and finally they would not always come out the same way.

    Perhaps there is no way to quantify such effects today, even theoretically, but I would be very careful about physical foundations of determinism.

    1. For me the big problem are the fundamentals of physics here, quantum level cannot simply be barred.

      Actually…it can.

      Sean Carroll loves to explain why the physics of the everyday world are completely known. There’s lots we don’t know, with quantum gravity his current favorite, along with dark matter, dark energy, cosmogenesis, high-energy particle physics, and more. But we know everything there is to know about the physics relevant here on Earth…and that includes human brains.

      None of that physics has any wiggle room for woo of any kind. Put simply, brains are far too big, hot, and messy for Quantum Mechanics to be of any significance in our understanding of cognition. You’ll need QM at some level in sub-cellular biology, but neurons are entirely classical entities — and cognition happens at the level of neurons, not near-atomic physics.




      1. “neurons are entirely classical entities — and cognition happens at the level of neurons, not near-atomic physics.”

        I do not think that either of those assertions is established fact. We currently have an understanding of neurons based entirely on classical entities but we do not understand them entirely. Further, it is not established how non-neuronal brain cells may contribute to thought and decision-making activities.

        1. To be fair, I was painting a low-resolution picture of the process. Yes, there’re cells other than neurons in the brain, and, no, we don’t know everything about how the human body works.

          But we do know far more than enough to rule out any sort of quantum weirdness as a significant factor in any macro-scale human phenomenon.




          1. I’m disagreeing with you again! One can find quite a lot of literature on both sides of this question (for example, the Aeon article someone linked above). So I’m not sure it’s as definitive as you suggest.
            I’d view most macroscopic events as the sum of many QM events, which eventually starts to look classical, but obviously that’s just a convenient approximation…

            1. Sean Carroll would describe Newtonian Mechanics as an “effective” theory. As such, it makes not one whit what “underlies” it; we can still operate using Newtonian Mechanics as absolute over its relative domain.

              A much better example of this (for explanatory purposes) is the Ideal Gas Law. We don’t care if the gas in question is air or purified chlorine…or even a tin can filled with ball bearings in orbit. When you look closely enough at any of those examples, the underlying reality is radically different. But with sufficiently large samples, we observe that an inevitable consequence of Newtonian Mechanics is that, if you’ve got a system that’s close enough to a collection of uniformly-sized spheres not being accelerated with perfect elasticity…then you’ll observe that the pressure on the walls of the container times the container’s volume is equal to the number of particles times the average kinetic energy (temperature) times a constant.

              In the case of Newtonian Mechanics…it really is absolute, set in stone. We know the scales over which it is applicable. But, at those scales, it really doesn’t matter what sorts of funky things can happen at other scales: Newton rules the day, absolutely.

              You will never, ever, ever, ever, ever see, for example, the Sun rise in the West or apples fall up to trees — outside, of course, carefully-constructed circumstances which are equally well understood.

              It is worth noting that Quantum Mechanics and Relativity are both, themselves, effective theories. We know for a fact that they’re not the complete picture…but we also know for a fact that they’re as rock solid as Newton over their respective domains. Nothing we might learn about new physics is going to change diffraction gratings or GPS.

              That’s the amazing power of effective theories….




  29. “Free will”, contrary to God, is evident at least as a strong feeling, admitted from everybody. So “hard determinists” must prove and explain how it is an illusion, how this illusion is created. Else, paraphrasing a saying of Hitchens, “claims without prove can be dismissed… without prove”.

    1. What kind of “proof” can one person offer to another of being conscious? It is easier to believe that reductionist physics may one day explain the human *behavior* of espousing free will in terms of particles and evolution than any need to bother with any sort of “subjective illusion” taking place in “minds”.

    2. When people point to their “Free Will,” they’re inevitably describing their decision-making process, which involves imagining the outcomes of different decisions. This has the subjective perception of exactly what Jerry describes: rewinding the tape of history with different outcomes. But the decision-making process is itself entirely deterministic, and we don’t actually spawn off other real universes in which we make different decisions.




      1. Nobody really rewinds. Imagining etc is part (input and feedback) of the future decision making and so on. Decision making never stops. But always someone can say that all this is “deterministic”. The whole life is sensed and is applied strongly non deterministically. Nothing really makes any sense if not. The question is, how really, determinism can make sense. “Hard determinists” appear to fail to make sense speaking in a not “free will language”. How to speak about improving society without speaking about changing a predetermined outcome? Is this possible? Prove it!

        1. How to speak about improving society without speaking about changing a predetermined outcome? Is this possible? Prove it!

          You have this ALL confused. You improve society by changing and/or adding inputs which can change the output. This is still deterministic.

          1. The output is predetermined. Cannot change it. But you can have the illusion of doing that if determinists is taken seriously. You see it in the short run. The real point is in the long run. In the long run… all changes are predetermined sequences of cause and effect starting from the big bang. I have this expressed clearly in a comment at Jerry’s previous post about a cartoon depicting what you say.

            1. But just because the outcome is “predetermined” doesn’t mean that you *know* what it will be.

              And because you don’t know the outcome, it *feels* like you have a choice.

              Similarly, because hard determinists don’t know the outcome either, they *feel* like they can influence the future (as with the topic at hand).

              This is what an algorithm feels like from the inside.

              1. But just because the outcome is “predetermined” doesn’t mean that you *know* what it will be.

                That has much more to do with the entropic arrow of time and the so-called “Past Hypothesis,” which itself explains why we can remember the past but not the future.

                Time actually plays only a relatively minor role in physics — just as direction doesn’t either.

                Here on Earth, there’s a very clear distinction between “up” and “down.” But in space, all directions are equal. The reason we have “up” and “down” here on Earth is because we’re in the proximity of a massive body that creates a gravitational field, and the direction towards that center of gravity is “down” and its opposite is “up.”

                It’s a similar case with time. We life in (relatively) close proximity to a state of extraordinarily low entropy, a state we call the “Big Bang.” There are statistically more ways for something to have high entropy than low entropy…and that basic fact of statistics accounts for all the evolution of everything that’s happened since the Big Bang.




        2. There’s a vast difference between determinism and predestination.

          For a physicist, determinism is the observation that the Universe unfolds like clockwork or a Rube Goldberg contraption. Just as apples “simply” fall because of gravity, everything else is just doing that which things in such situations do.

          Predestination, on the other hand, is the concept that the Fates have decided that you will die on a given day, or that you will marry your mother, or what-not, and nothing you can do will change your destiny. For example, if you try to cheat the Oracle by marrying a girl younger than you, a cruel twist of fate will reveal that she really is your mother and a magic spell only gave her the appearance of youth.

          Predestination makes for great literature, but has nothing whatsoever to do with reality.




  30. Dear Dr. Ceiling Cat.

    I agree with you.

    To be honest, I never really considered the problem of free will until I started reading WEIT. I must confess that initially I had a difficult time understanding some of the concepts. I don’t really know why I had such difficulty at first. I suppose it is because I had simply internalized a kind of “common knowledge” about free will; I assumed I had it. But I kept reading WEIT (and links therefrom) and I now think I grok what you “B” types are saying. I agree that the difference between you “B”s and me – a “C”, I’m afraid (more below) is essentially semantic.

    So in my own formulation of the issue I agree that free will is an illusion; the choices we make are constrained and determined by our genes, the environment and how we’ve developed – they are not in any sense a dualist’s choice. They are, essentially the result of compliance with laws of physics.

    However, I like to think of some of my choices as local violations of those laws. Just as life itself is an emergent property of complex chemical reactions, I think there is room to think of free will as an emergent property of consciousness. As we all know the process of life can appear to violate certain laws of physics (the second law of thermodynamics is an easy case), but of course those violations are only local and ultimately the laws of physics pertain. They are illusory violations.

    Just-so, I like to think at least some choices I make are local violations of the laws of physics; the choice I’m making is an emergent property of consciousness even though that state is determined by the laws of physics.

    Of course, any local violations that amount to free will are ultimately illusory and it is therefore mere semantic difference to what folks who are “B”s would say about choice. I’m ok with that.

    I’d love to hear a critique of this idea to see where I’ve gone wrong. Or a link to one (I am certain something like this has been addressed somewhere).

    No matter what, thanks Dr. Ceiling Cat for providing a website where biology is king, cats are celebrated and such things are discussed.

    1. I celebrate Ceiling Cat’s mom and dad, High on a tree branch Cat and Up on the Roof Cat. They live forever and will never be road-kill. Blessed be the Cats!—and all other Fur, Feather, and Scaly Critters.

    2. Just-so, I like to think at least some choices I make are local violations of the laws of physics;

      Hmm, I’m not sure I understand your account at all, but it does not sound like compatibilism (which is a wholehearted embrace of determinism, without asking for “local violations”, whether “illusory” or not).

    3. I agree with lots in this presentation. But I don’t understand one thing. How does the 2nd law of thermodynamics violate the laws of physics?

      These days, young people — or just people — have teachers like Jerry, Richard Dawkins and Sean Carroll to introduce them to science and its effects on our lives. When I grew up, that was lacking a bit. (I never even heard of Carl Sagan until well into this century. Wonder where I was.) The authors I read to give me a sense of what life really was were much more literary and philosophical — Kazantzakis and Camus and others like that. And that is a bit where I still am.

      So I really dig (to use a term from the period i just referred to) Sean Carroll’s poetic realism. I feel grudginly obliged to accept determinism and what it means — on the level where physics insists on it. But there are emergent concepts (water or self) and, as Sean says, it is useful to talk about these things in some cases, as long as one maintains their compatibility with the other levels. So, just as in the existentialism of my youth I imagined that Sisyphus thought he really might that ol’ rock to stay this time, I find it useful (and encouraging) now to act as though I did some sort of choice in some decisions, be they conscious or unconscious.

      Bottom line: I really do not like not being able to feel good when I have made a good choice. So even tho I know that is the case, I ignore it. Is that B or C?

      Bottom line no. 2: I think there is much to do in improving our justice system, but a lot of it can be defended without the need of invoking determinism.

      Bottom line no. 3: In any case, I am afraid hell… — no, bad metaphor — the sun will swell up into a red giant before you get most people to admit determinism in their lives and the legal system.

    4. Hard determinists make much of the notion of the “illusion” of free will. I think this positing of an “illusion” is itself an illusion, borne from accepting uncritically libertarian presuppositions. I devote a page to countering this erroneous idea at http://www.RationalRealm.com/philosophy/metaphysics/freewill-compatibilism-page9.html

      I give an example to illustrate my point:

      Think about a paradigmatic free act, such as choosing coffee over tea at your friend’s house (assuming you enjoy both kinds of beverage). Sure, you don’t feel any kind of physical impulse forcing your finger to point to the coffee. You would be the first to know if there were such a force acting on your finger. Furthermore, you don’t feel any kind of impulse forcing neurons in your brain to adopt particular action potentials. It’s the feeling of absence of such a force that the libertarian cashes out as the feeling of having free will.

      But notice the queerness of the latter kind of feeling. Feelings of your finger being forced are natural and we all know what that feels like. However, what would it feel like to have particular neurons in your brain forced into particular states—states that go on to generate forces that determine the states of the motor neurons that ultimately trigger your finger pointing? There are currently billions of neurons in your brain that at this very moment have their states determined by the action potentials of other neurons. And yet you have no feeling of these neurons being forced. If you cannot know the feeling of neurons in your brain being forced into particular states, how can you or anyone else possibly know what it feels like to have some neuronal states underdetermined by other neuronal states? The feeling of free will that we have must be something other than the feeling that some of our brain states are underdetermined by physical forces.

  31. Dualist Incompatibilist (Libertarian) – The claim that dualist ‘free will’ exists (coz religion, spirituality, souls, minds, etc.) in that the will is free of physical determination (at least enough). And that is ‘incompatible’ with determinism, and so the world is not completely deterministic.

    Determinist Incompatibilist (Hard Determinism) – No evidence of souls, minds; and all science suggests determinism. So there is no dualist free will. Dualist free will is ‘incompatible’ with determinism. The free will we feel we have is illusory – we can’t ‘feel’ the deterministic processes of the brain, so out willed decisions feel free of physical cause: that’s the illusion.

    There’s nothing in Determinist Incompatibilism that Compatibilists disagree with when pushed (do you really think our decisions are sensed; can you really feel your neurons ticking?) And yet many insist free will is not illusory. The only way they can do that is calling our determined localised autonomous behaviour ‘free will’. It’s a fudge. This is the extent of ‘talking past eachother’. This is the sense in which a Compatibilist might say HD is a category error – because they define the category within which ‘free will’ resides, ignoring the place where the big argument lies: A v B.

    A v B is where two ‘incompatible’ positions are being argued, and as such is one of the key battles with religion.

    C confuses the issue, because when Compatibilists defend free will they are defending it alongside dualists whether they want to or not.

    Compatibilism is often expressed in terms of ‘freedom’, which boils down to no more than a version of ‘degrees of freedom’, even for complex systems like humans. And ‘degrees of freedom’ is contained entirely within a deterministic framework.

  32. Partial disagreement based on the Douglas Hofstadter concept of “sphexishness” (used a lot by Dan Dennett). A way to construe this is that some sentient beings really have more “free will” than other sentient beings based on the degree that they have self-modifying behavior. An alcoholic or heroin addict has less free will than someone free of addictions (I am not saying this is MY position, simply that it is a possible one.)
    This is why the law allows some crimes to be judged ‘not guilty by reason of insanity’ and others not to be.
    Thus rehabilitative punishment is appropriate in some cases and not in others. (Of course, where one draws the line is VERY subjective. I realize there are real problems with this position.)

    I’m utterly agnostic on B vs C, but the compatibilist position seems to me to boil down to two things.
    1) A statement widely attributed to Arthur Schopenhauer “Man can do what he wills but he cannot will what he wills.” (The provenance of this quote has become dubious)
    2) Hofstadter’s concept of sphexishness. Non-sphexish beings have self-modifying behavior, in a way the sphexish ones do not. The latter behavior is what Dennett calls “simply determined” and has no sense of agency behind it.

    Finally, You can be a libertarian free-willist and still think that the notion of salvation based on belief in Jesus/Mohammed is utter moral nonsense!!

    1. I’ll second one of the ideas you mention: I’m curious as to if Prof. Ceiling Cat has read Godel Escher Bach?

      I must admit that I’ve struggled with that!

  33. Let me be of no help whatsoever. I am not sure if I fall into group B or C. I have agreements and disagreements with well spoken proponents of both groups. If I had to choose one as best fit I’d go with group C, today at least.

    Regarding the question, I just don’t know. I am sure though that I don’t think it would hurt to try as some compatibilists argue. I disagree with Dennett’s argument that (massively paraphrased) if you force the general population to accept determinism that that could have severe negative consequences for society because many people will no longer feel responsible for their actions.

    So by all means, try and convince people to be less retributive and more rehabilitative by convincing them that dualistic free will does not exist. I don’t think there would be any negative consequences and pursuing multiple methods of moving society towards a more humane future, as opposed to a single method, is certainly a good thing.

  34. Imagine really wanting chocolate ice cream and going to your favourite ice cream shop. When you get to the front, you inexplicably order strawberry. Surely you would come out of the store feeling rather puzzled and when your friend says “I thought you wanted chocolate?” you would say “I did. I have no idea why I ordered strawberry”.

    Wouldn’t libertarian free will entail rather bizarre such behaviour?

    That said, as I am writing this, it occurs to me that I occasionally do behave in such an inexplicable way so maybe libertarian free will is real!

  35. I’m not sure where I fall on the B vs. C debate. What I don’t understand about the compatibilists is why they’re trying so hard to wrestle the phrase “free will” out of the hands of history. To me, there are certain aspects of the concept of free will which are appealing when discussing the deliberative process I go through when I make decisions, the fact that I can be held responsible when I sign a contract but a chimpanzee can’t. It seems to me that it would be trivially easy to simply come up with a new phrase like “effective will” which takes on the burden of describing what it is humans do when they make a decision, while being free of the anachronistic baggage of the old phrase. I think the problem with the B+C vs. A debate is that there is a loss of language that occurs once one is disabused of the notion of free will. They can of course use the term free will in a compatibilist way, but I think doing this is a recipe for confusion, since to many people, the word “free” is not compatible with the phrase “sort of,” in the same way that “pregnant” is not compatible with the phrase “sort of.” I think that semantic debate between B’s and C’s is less important than the debate between B+C’s and A’s. But if the semantic debate were resolved, it would be much easier to win the debate between B+C vs. A.

  36. Did someone willingly write this post?

    Must human consciousness be unitary?

    If human consciousness need not unitary, can some parts will freely, and others not?

    1. What is “willingly”, what is “consciousness”; and what is “unitary” here?

      And what do they have to do with “free will”?

        1. What would “Every word is defined in terms of every other word” even mean?

          You haven’t made any definition. And a useful definition must, like hypotheses, be judged against observations.

          What you have produced so far looks like mysticism, which is meaningless.

          1. Wittgenstein had a lot to say about word-games. Every word is defined only by other sets of words, as are those words in turn, in an ever-expanding network of words. But words are our attempt to model reality, not reality itself.

            Kant, and he was certainly not the first or the last, distinguished between phenomena and noumena. Phenomena is our model of reality and is all we can ever know – we can never know noumena. People persistently forget they’re creating models.

            Vision, hearing, smell, touch, taste – all these inputs create models of reality. The colors we see, the smells, etc., exist only in our own brains to model reality. They are phenomena, not noumena.

            Even Schrodinger created his dead/not dead cat thought-experiment to illustrate that Quantum Physics was a model of reality, not reality, and as such, it was incomplete and misleading.

            I don’t think that’s mystical at all, merely an awareness of the shortcomings of humanity’s primary tool – language. I haven’t defined words in the way one might wish because words can’t do what people assume they can do – enable us to reach noumena.

  37. Put me with the B’s, but I might be swayed away with good evidence. The inexplicability of the subjective illusion of being, even after explaining the mechanism of consciousness, still muddles me.

  38. I’m now a C (after having read The Big Picture). But we can make society better whatever our view on free will. It’s irrelevant. Think about the moral progress we have made already. Before we hit children to chastise them. Now we don’t hit children. This and other improvements in our behaviour hasn’t depended on our view of free will. We have used reason. And it doesn’t work that if we all became B then this would lead to sweeping changes in how we think about crime and punishment. What improvements occur happen piecemeal. One solution to a problem at a time.

  39. I don’t agree, because I think it is all a semantic (philosophic) squabble.

    Organisms are “free” (autonomous) agents, they have goals (taxis or tropism) and if they can make choices you may call the result “free will” if you want.

    But it is still dependent on physical law (whether deterministic or probabilistic).

    If we study the 3 options as given verbatim, they describe impossible realizations of physics.

    A. Described as not obeying physics.

    B. Puts an impossible initial condition of same “particle” configuration, which deterministic chaos shows is impossible to enact. It is an idealized notion – and I think Laplace used it – but it doesn’t contribute anything new.

    C. Same as B.

    So that is my take, “free will” is obeying physics.

    If anyone wants to argue that roughly the same initial conditions will lead to roughly the same outcome for humans, I guess we know enough biology to agree. (Reflexes, illusions, et cetera.)

    1. And the ‘could have chosen otherwise’ condition is not sufficient to give us the ‘free will’ that most people want to believe in.

      Radioactive decay is inherently unpredictable (in principle, it could have decayed at any other time), but we don’t ascribe ‘free will’ to it.

      1. I am not sure you agree or disagree with me. I didn’t mention the ‘could have chosen otherwise’ idea since I find the analysis lacking in empiricism before that.

        But yes, a lot of organism traits are found in other natural systems, as we would expect.

        My opinion is that humans are not special, just a species among others. So some traits are unique of course. Being an autonomous agent and having goals aren’t among those traits.

    1. And you assume we know and understand all the laws of nature that apply in this case in sufficient depth and with sufficient completeness?

      Am I the only one who thinks that to scientifically literate people two centuries from now, our own understanding of nature will most likely look about as incomplete and occasionally way off as the understanding of early 19th century looks to us now?

      1. Sean Carroll says we do understand the laws of nature this way. The standard model+general relativity can explain everything within a large energy range around the one we inhabit, and excludes new forces and matter in said energy range. Two hundred years from now, that will still be true.

        1. 300 years ago, Newton’s mechanics was able to explain everything people knew about motion. 300 years later, it is still true – it can still explain everything people knew 300 years ago. Just like in 200 years standard model + general relativity will be able to explain everything we know now. About things they will know in 200 years, I am not so sure.

          In every case in distant enough history when people stated they now know everything important there is to know about a phenomenon, they were proven wrong, often spectacularly so. Do not fall into that trap. We are not special. The time in which we live is not special. In 25th century, 20th and 21st centuries will be viewed merely the time between 19th and 22th century,with some remarkable breakthroughs but not more remarkable than those in later centuries. Much of what we know will look like antique ignorance.

          1. That’s why Sean Carroll is very careful in his claims. His mathematical formalism has an explicit UV cutoff.

            His common English formulation is that the laws of physics underlying everyday experience are completely known. Every experiment ever performed on Earth is perfectly consistent with the Standard Model plus Relativity — with a very exciting footnote about hints of novel particles maybe about to be produced at the LHC.

            More importantly, we know more than enough about the stuff that makes up the everyday experience to know that nothing we will ever discover will ever change our understanding of it. We fully expect to discover new physics, but we also know with overwhelming confidence that that physics is irrelevant to terrestrial environments.

            Specifically, humans are made entirely of electrons and quarks, with some notable footnotes for gluons and the other stuff that holds atoms together. A claim that new physics might change our understanding of everyday physics is a claim that there’s some other way to interact with electrons and quarks. The problem with this claim is that, because Feynman diagrams can be rotated, it is a claim that smashing electrons and quarks together at the relevant energies will produce new particles associated with the proposed as-yet-unknown force.

            And physicists are very, very, very good at smashing together electrons and quarks. We’ve smashed them at every energy that could possibly be relevant, and discovered only that which is accounted for by the Standard Model.

            As such, we know that, if there are any unknown forces that can interact with electrons and quarks, they do so either at energies way beyond anything that’s ever existed on Earth (far beyond even those at the heart of the Sun) or they’re ten thousand times or more weaker than gravity. Gravity itself is extraordinarily weak — pick something up, and you’re overcoming the combined gravitational field of the entire Earth. And we already know that gravity itself is irrelevant to cognition, since astronauts in microgravity don’t experience any changes in thinking (outside, of course, the expected awe at the magnitude of the experience and that sort of thing).

            So, it’s a very specific claim, and not at all like the grandiose ones of history you’re referring to. In the 25th century, they will, indeed, look back at us in the 21st century and note that we lived at the dawn of the age when the physics of everyday life was fully accounted for, even if they’re quite comfortable with the physics of the radically exotic.




  40. I would bet that most compatibilists would point to both the come it’s of our brains and the gun-to-the-head scenario in fleshing out their conception of free will. Those are not mutually exclusive components.

    I also think it’s perfectly reasonable to make a distinction between actual prison bars and internal, biological constraints. You’d presumably see a meaningful distinction between the sugar in those cherry pies and, say, coal, even though they’re both just conglomerations of carbon, hydrogen, and oxygen.

  41. I hate to complicate matters, but I have to in this case (can not do otherwise). I am a member a fourth category, the Non-Free Willist. I can not be called a hard determinist because I am not convinced that the universe is hard in its determinism. It may be that indeterminism is also part of the mix. BUT, and this is important, interderminism does not impart freedom to the will. It merely introduces an element of randomness, which does not comport freedom. So, long ago, I rejected the label determinist (hard or otherwise), and adopted non-free willism to more properly identify my position on this matter. (I am as far as I know the coiner of that label, and first member of the non-free willist category.) Though I may have coined the label, I hardly feel like a founder, as there were many before me saying that there is no libertarian freedom to the will.

    Non-free willists believe the same as hard determinists, that human behaviors are solely a function of our genes and our environmental inputs, (Nature and Nurture, if you will) and nothing else besides some unmeasurabe random element. There is no dualism, and if at any point in time you could rewind the clock to some previous point in time, events would most assuredly happen as before (except accounting for any truly random influences).

    Also, if asked I would be happy to share my case of the non-corporeal decision maker, as an argument for non-free willism.

  42. At this point in discussion, I would like to warmly recommend an essay by philosopher Eric Dietrich from SUNY Binghamton entitled “There is no Progress in Philosophy” (google it, I cannot post links here). A couple of relevant excerpts:

    “I’m a professor in a philosophy department. Most of my philosophical colleagues study ethics of one sort or another. We have in our department several consequentialists, a couple of deontologists and moral essentialists, a couple of virtue ethicists, and a few relativists. It is a commonplace that these views, at least in certain well-known formulations, are incompatible with each other. Certainly, most of my colleagues believe this. Most also believe that he or she is _right_. Since they also believe in theory-incompatiablism, they believe that their colleagues are _wrong_. The consequentialists (a group to which I do not belong) are particularly vociferous (no doubt just by chance). They passionately and earnestly explain to the rest of us that we are wrong, and they give us arguments both old and new to get us to change our views. We never do.

    This is not to say that their arguments don’t affect us. Like philosophers everywhere and of every stripe, we nonconsequentialists are strongly affected by our colleagues’ arguments for consequentialism: they cause us to draw distinctions and find false premises and errors in our colleagues’ reasoning. Of course, when we point these out, the consequentialists, like us, don’t change their minds, they muster even more resolve and begin afresh.”

    A little later,

    “Philosophy does not even stumble forward. Philosophy does not move forward at all. It is the exactly the same today as it was 3000 years ago; indeed, as it was from the beginning. What it does do is stay current; philosophers confuse this with advancing, with making progress. Staying current is not moving forward any more than staying up on the latest fashions or music is movement toward greater social justice.”

    And finally,

    “If space aliens show up and give us the solutions to our philosophical problems, then McGinn will be proven right and Nagel wrong. But, McGinn denies that this can happen: we wouldn’t understand their solutions. Again, think of giving dogs our solution to making dog toys (factories, synthetic, harmless fibers, plastic squeakers, etc.). They wouldn’t get it, to put it mildly. And, in any case, the arguments are about what to believe _now_. Of course, we might wake up tomorrow with a sudden understanding of the freewill/determinism problem. But if we ask, _today_, what that understanding would be, we’d be doing philosophy and we’d get nowhere.”

    In summary: A, B or C: Doing philosophy. Getting nowhere.

    1. “In summary: A, B or C: Doing philosophy. Getting nowhere.”

      Do you have the same approach to the questions 1) “Am I a brain in a vat” and 2) “Are other people conscious?”

      1. There are people who believe those things, or that we cannot know answers to those questions. Go ahead and try to persuade them that they are _wrong_ and that you are _right_.

        Get it?

  43. Count me as a B. I also agree that B & C vs. A is currently a more relevant and important discussion than B vs. C. However, I am very tentative on the latter. The studies I’ve seen investigating what people actually believe about free will, how it affects their behavior short- and long-term, etc. seems to be very limited in size and very muddled in result. Without a better understanding of what people generally believe, I can’t be too confident about which argument is more important.

    Perhaps there are far fewer true libertarians in the general population than I suspect. Perhaps it is especially difficult to change their minds by arguing against libertarian free will directly, and it would be more effective to argue against religion, or to fight against poverty, or some other confounding variable that causes a belief in libertarianism. Any of these is plausible, and would reduce the importance of B & C teaming up against A.

    In the end, those are the aspects around the topic I find most interesting — getting a better characterization of what people actually believe about free will, how those beliefs change (or don’t) given different circumstances, how those beliefs do (or do not) impact their behavior, how their beliefs can (or cannot) be changed, etc.

  44. I’m not sure if (C) and (D) are really different from one another but it doesn’t matter because (A) is the correct answer, especially as you’ve now stated it more clearly as ‘could the universe turn out differently from an identical starting point’. THe answer is yes due to well-understood, thoroughly well researched quantum mechanical considerations.

    This has nothing to do with our minds ‘overcoming’ the laws of physics, it *is* the laws of physics…

    1. But how does QM imply free will?

      If the universe was rewound and replayed the exact same way (“with every particle in the universe configured as it was before”, including all QM effects), the result would *have* to be the same.

      If the universe was rewound and replayed and some QM effects changed (photons scattering in slightly different directions or whatnot), then you might see a different result. But that would, by definition, not include “every particle in the universe configured as it was before”.

      Your thoughts, created by firing neurons, are part of physics. Just because you can’t understand how, doesn’t mean that they aren’t.

      1. If the universe was rewound and replayed the exact same way (“with every particle in the universe configured as it was before”) then some uranium atoms that decayed in the first minute in the original setup might not decay in the replay. Either that or I have really misunderstood some of my classes…feel free to correct me! Of course the statistical likelihood of decay will be the same, but in a given lump of uranium there is nothing to determine which atoms decay at which moment.

        1. This may just be a(nother) issue of semantics, but I took “with every particle in the universe configured as it was before” as including QM effects (so, your uranium decayed a given way the first time around, it would have to do the same the second, or else it wouldn’t be “configured as it was before”).

          So, if what you’re saying is that QM has an effect on things, I agree, but since we live up here in the “big world”, it’s not particularly relevant to the issue of determinism vs. free-will.

        2. I don’t think I properly explained what I meant in the previous comment, so I’m going to try one more time.

          Determinism, as I see it, is all about cause and effect. Some of that is at the particle level, but by and large, we’re talking about emotions, knowledge, etc. – things are amorphous concepts, not subatomic particles.

          This is why I have a hard time seeing how QM has much of an effect on our determinism: we are affected by quantum events, but only after they propagate up to our scale. And once they have propagated to our scale, they will affect us at our scale, not at the quantum level.

          Indeed, since quantum mechanics underlies all of physics, every cause will be traceable back to the quantum scale, but we won’t feel their effects until it hits our scale.

          And, at our scale, events cause neurons to fire, which has the effect of us doing something.

      2. Note that the fact that QM is not strictly deterministic doesn’t imply some kind of dualism. It may be that nothing as purely statistical as radioactive decay can have effects on decision making process. Even if it did, it would just mean replacing the idea that you would act in precisely the same way (in the identical situation), with the idea that you would act that way 90% of the time in repeated identical situations, with no way to predict which way your decision would fall in any one instance.

        1. And with that comment, you have explained what I was trying to get at much more succinctly.


          Now I only wish I had read that before I wrote those other two….

        2. An action which has a certain chance of happening or not happening, with the outcome being impossible to predict in advance is pretty much a textbook definition of free will as I understand it. Your choice of a 90%/10% example creates the impression of less than 100% freedom but you might just as well choose a 50%/50% example.

          Disclaimer: I may not understand exactly what free will is supposed to be!

          1. Re your discliamer: no, you don’t. Predictability has nothing to do with libertarian free will, at least. A coin toss is not predictable, but it is determined, and the coin, of course, has no “free will”.

  45. I agree as I think convincing people of determinism is much more important than fighting over non-libertarian free-will. Think of the consequences of understanding determinism! For one, the stigma for mental health will completely vanish. No more “snap out of it” harshly hollered at depressed people. I often refer to the trouble of not understanding determinism as the “dangers of dualism” because if you believe you are a ghost in a machine, enjoying the view behind your eyes, you will readily dismiss the plights of the mentally ill or the behaviour of a child with an immature brain and expect these people to behave in ways they just can’t.

    1. On the other hand, to have ethical regard for the welfare of people, whereas you would not have any ethical regard for a pile of their constituent chemicals, is also a form of dualism, and not sanctioned by science per se, no? If you think that other people deserve your respect, and a rock does not, perhaps it is you yourself who are being dualistic.

      1. Huh? You keep using that word. I do not think it means what you think it means.

        Dualism and empathy are two different things.

        1. One does not have empathy for a pile of rocks or a vat of chemicals. One has empathy when one comes to feel that one is dealing with a conscious person. How does that happen? Subjectively, one regards a person differently than just chemicals. This is a form of dualism, eh?

          1. Empathy is reasonably well understood from the functioning of so-called “mirror neurons” in the brain. Those same neurons are likely to be significantly responsible for our own sense of self.




            1. Mirror neurons are so argued over as to be completely debunked as a source of empathy. Subjective “sense” of anything cannot be verified externally to even exist. There may indeed be a “turing test” sweet spot in the brain where a subject could easily be triggered to report that any inanimate object was worthy of empathy.

              1. We are therefore in disagreement of the facts…and, as far as your second sentence goes, I’m afraid that I actually am somewhat of an expert on the matter. More than enough to tell you that you couldn’t be more wornger.

                Specifically, the spreadsheet linked on this page:


                has all the information you need to precisely predict (within certain limits not relevant to this discussion) the subjective perception you or anybody else (again, with well-understood caveats) will experience when you observe light with any given spectral power distribution.

                For you to claim that such a well-understood phenomenon doesn’t even exist is as absurd as suggesting that maybe matter isn’t real, either.




              2. @1977ub: Empathy for inanimate objects has indeed been demonstrated many times. To give but one example: people watching a cartoon, in which a square seems to be chasing a triangle, root for the triangle.
                @Ben Goren: Whether or not matter is ‘real’ is a hotly discussed subject, especially since nobody knows what it is. To quote Wikipedia: “matter does not have a universal definition, nor is it a fundamental concept in physics today.”

              3. Whether or not matter is ‘real’ is a hotly discussed subject, especially since nobody knows what it is.

                Our resident sand sorter…rock examiner? Gravel inspector!

                Our very own Aidan has a perfect way of demonstrating to anybody who questions the reality of matter that matter really is really real. It’s a trivial experiment, with only two pieces of equipment: a sock and a rock.

                You put the sock in the ro — er, no; sorry. Got that backwards…you put the lime in the cocon…

                …aw, hell. Just whap the dude upside the head with a clue-by-four repeatedly until reality sinks in sufficiently to merit acknowledgement; thus, ’tis refudiated.




              4. So it’s the same as with free will: you really *feel* it’s there, so it must exist.

              5. If you’re going to so casually dismiss every physics experiment ever performed, including all the ones you yourself must have performed in order to graduate high school, including your own minute-by-minute daily existence, and dismiss freakin’ matter as actually being real…

                …then, whatever reality you’re living in, it’s so far disconnected from the one I’m living in, that no meaningful communication is even hypothetically possible.

                The illusory (or, at least, ephemeral) nature of the self is very well demonstrated, beginning with ancient contemplative and meditative practices and continuing all the way through modern anesthesia and cognitive neuroscience — not to mention nightly sleep cycles.

                The reality of matter (whatever its ultimate form) is even better demonstrated.

                If we can’t even agree on that much…then this really would be the end of the line.




              6. Is that what you call charitable interpretation? Obviously, the problem is “what the ultimate form” of matter is. You just casually dismissed the fact that no physicist has yet come up with a definition. So, if you are as smart as you think you are, why not provide the world with one?

              7. This is another example of what Sean Carroll describes in The Big Picture as an effective theory.

                An effective theory is one which remains true regardless of the foundations upon which it is built, with the Ideal Gas Law being a perfect example. Your gas can be the familiar air in a baloon, a tin can filled with ball bearings in orbit, or a computer physics simulation. The fundamental reality underlying all those examples is radically different, and yet the effective theory of the Ideal Gas Law is equally applicable and all can be explained in terms of pressure, volume, number of particles, a constant, and temperature.

                We know that classical Newtonian Mechanics is unassailable at relevant scales, because we know equally well that the Quantum and Relativistic Mechanics from which it emerges are just as unassailable at far more extreme scales. We know that Quantum and Relativistic Mechanics are themselves effective theories, because we know that they break down — and the big job of physicists today is to figure out what theory underlies them.

                And we already know that, at quantum scales, the familiar world of solidity simply doesn’t apply; instead, it’s all wavelike fluctuations in the dozens of quantum fields of Hilbert Space. At the scales at which QM is known to break down, it could be something else entirely, again radically different — such as the vibrating strings of String Theory.

                But all that’s irrelevant, as whatever’s beneath Quantum Mechanics most emphatically and unquestionably resolves to Quantum Field Theory, and Quantum Field Theory most emphatically and unquestionably resolves to Newtonian Mechanics at the scale of a rock in a sock.

                So, again: if you wish to insist that matter isn’t real because we don’t have an ultimate grand unifying theory of everything, you are again invited to bang your head against the wall, repeatedly, until the reality of the solidity of matter finally sinks in — figuratively or literally — and your objections thus effectively refuted.

                Or, you might as well object that water doesn’t exist, because the electrons, quarks, and gluons that mostly make up the hydrogen and oxygen atoms of water molecules aren’t wet…in which case I’d advise you to take a long walk on a short pier and get back to us after you’ve reached your conclusion.




              8. @Ben Goren: “every physics experiment ever performed” will not settle the issue of Descarte’s demon / Brain-in-vat.

              9. @Arno Matthias Re: “triangle” this is my broader point as well – that Empathy cannot be interpreted as a response to consciousness per se, only one’s concept of the other mind.

              10. Your spreadsheet can help one person decide that another person is having some sort of experience. But for the very idea of subjective experience to have scientific validity, another’s subjectivity would have to deliver behaviors above and beyond what reductionist physics does. But if you believe that, you’re a dualist or at least not a reductionist. Which is more empirically valid:
                Physics —> Experience —> Behavior
                Physics —> Behavior

          2. Dualism is thinking you are separate from your brain and are essentially a soul peeking out from behind your eyes. It has nothing to do with empathy.

            1. Empathy happens when you imagine there to be more to one’s neighbor than simply the sum of their particles. You don’t have empathy for particles. On some level you must be imagining a “soul peeking from behind the other’s eyes”. That is the dualism I’m talking about.

              1. But you’re redefining dualism. From Wikipedia:

                In philosophy of mind, dualism is a view about the relationship between mind and matter which claims that mind and matter are two ontologically separate categories. Mind-body dualism claims that neither the mind nor matter can be reduced to each other in any way.

                Having theory of mind is not dualism.

                Theory of mind (often abbreviated ToM) is the ability to attribute mental states—beliefs, intents, desires, pretending, knowledge, etc.—to oneself and others and to understand that others have beliefs, desires, intentions, and perspectives that are different from one’s own.

              2. “Having a theory of mind is not dualism.” Ok, when the ancients attributed the cause of storms to be anger in the mind of Zeus, was that dualism? Today we would scoff, calling it primitive animism, and we would do this particularly because we have physical reductionist explanations for weather phenomenza. If we are satisfied with physical reductionist explanations for a neighbor’s behaviors, why do we also not simply abandon the primitive animism of seeing some kind of conscious agent at work? If I think that other people are conscious agents – in the face of substantial evidence their behaviors might be explained by reductionist physical explanations, is that not some form of stubborn dualism?

              3. You are presupposing that a non-superstitious explanation of consciousness is impossible. Every cognitive neuroscientist on the planet would find your objections incoherent.



              4. No none of that is dualism. That was simply attributing the cause of something to the wrong thing, which in this case is an imagined entity.

              5. It’s called solipsism, a perfectly logical stance that’s best avoided in social encounters.

              6. No, because I have evidence my neighbour and his mind exist. No such evidence exists for gods.

              7. I think you’ll find that many of us here are no more impressed with “pure” philosophy than we are with theology. If you want to get our attention, give us evidence — and the evidence is overwhelming that there are other humans, that the other humans behave similarly, that we perfectly fit that pattern, that we have minds; ergo, other humans have minds, too.




              8. What Ben said. Your link to is to pure philosophical musings. Observational science gives me indirect evidence that people have minds. It’s nowhere in the same neighbourhood as god’s, with minds, exist.

            2. Ben Goren, you will THIS thread again by default. (I will point out that since consciousness – the word used by humsns – is not a name for any type of “phenomenon”, and so it is a non-sequitur to seek an “explanation” for it.) Now get out there and celebrate!!!!

  46. I largely agree with Gregory Kusnick above. I don’t think the argument between (B+C) and A is as important as it is sometimes made out to be. In my opinion, there are more effective and convincing ways to argue against religious thought, and it is not necessary to reject even libertarian notions of free will to argue that a restorative justice system is far more effective and beneficial for society overall.

    My biggest problems with hard determinism are (1) the brushoff of quantum indeterminacy, and (2) taking as given very nebulous ideas of causality. When we are talking about the interaction of electrical impulses in the brain being “determined” by the state of the universe at a given point in time, it is not at all clear to me how we can just ignore the composite quantum indeterminacy of the underlying system and declare the outcome to be deterministic. Moreover, I see no way such a complex system can be adequately described in practice using currently established causal concepts. I don’t reject the possibility that hard determinism may hold, but I don’t find the current argument for it convincing.

  47. I agree with PCC(E) and am a (B) in the above list.

    I agree as well that most of the time, the disagreements between B & C are semantic in nature (the exception being those compatibilists that believe, mistakenly, that quantum mechanics provides a way to have true “free will”).

    I also agree that our time and energy would be better spent fighting against the patently false concept of Libertarian free will. However, I think it will be at least as difficult as trying to convince everyone of the truth of evolution, simply because the illusion of free will is much stronger in most people than their illusion of perceiving design and ultimate agency in the world (and look how hard it’s been to get rid of *that* idea).

    Maybe we (as a species) will figure it out in another hundred years.

    1. those compatibilists that believe, mistakenly, that quantum mechanics provides a way to have true “free will”

      Which compatibilists would those be? I don’t think anyone here believes that.

      1. Maybe I misunderstood, but that seemed to be the points that Ed Kroc and Tomasz were both making – that things are generally deterministic, but that quantum indeterminacy gives a way to have a free will of sorts.

        FWIW, I think it’s also the argument swordfish was using to identify as a Libertarian.

        1. The way I read Ed and Tomasz is not that they think QM provides a mechanism for free will, but rather that physical determinists should reject position (B) as defined by Jerry, because it gets the physics wrong. Their complaint (and mine) is that if hard determinists want to teach the rest of us the truth about determinism, they should first figure out what determinism actually means in a quantum universe.

          1. [I]f hard determinists want to teach the rest of us the truth about determinism, they should first figure out what determinism actually means in a quantum universe.

            Whether or not you accept Everettian Many-Worlds, it is superb for demonstrating that you’re carping about a red herring (if you’ll forgive me).

            According to Everett, the universal wave function evolves according to Schrodinger’s equation, with all possible outcomes equally presenting in reality, in proportions aligned with the Born Rule. The electron really does go through both the left and the right sides of the slit, and it really does do so in a diffracting pattern that really does get recorded by each and every photosite (or whatever) on the detector. And that evolution of the function is purely, entirely deterministic.

            The only reason there’s any sort of mystery as to what you’ll observe is because the different branches of the waveform can’t “talk” to each other. There’ll be many different Gregories after the experiment, all of whom trace their histories back to the single Gregory reading these words…but each will have observed something different. For one of your future selves to ponder why you observed it on the left is as pointless for some other future self of yours to ponder why you observed it on the right.

            Again, whether or not you accept Everett is irrelevant…because the other formulations of Quantum Mechanics have their own means of resolving the mystery, and all of them are equally deterministic. I only use Everett here because the determinism is especially easy to grasp in this model, even if you might find other aspects of the model disagreeable.

            This, of course, also ignores the fact that there’s no soul that can manipulate the outcomes of quantum indeterminism — which is the only possible way that even naïve interpretations of quantum randomness could be relevant to questions of cognition. There isn’t even any hypothetical “Will,” “Free” or otherwise, in a coin toss…so the question of how deterministic the Universe is at its smallest scales is relevant…how, exactly?

            …and, on top of it all, once again, brains are entirely classical phenomenon, given their size, temperature, and complexity. There isn’t even any theoretical room for quantum indeterminacy to manifest in a brain….




            1. Ben, please click here to experience quantum indeterminacy manifesting in the visual cortex of your brain. The specific colors you see are derived from the output of a quantum RNG.

              Also your claim that all formulations of QM are equally deterministic is incorrect. Dynamical collapse theories are explicitly nondeterministic.

              1. That’s certainly a rather nifty demonstration…but, remember. We’re discussing cognition. In reality, what difference will it make to the way your life unfolds if you see a blue patch followed by a green one or a yellow one followed by a purple one?

                It might help to think of quantum indeterminacy in this context in terms of entropy — and that random color generator is a perfect example to use.

                Let it run for a bit and it’ll fill the screen with random blocks of color. Reset it and let it run for the same amount of time and you’ll get another screen with different random blocks of color. Repeat as often as you like.

                What you find is that, though the microstates of all the screens are different, the macrostates are indistinguishable. That’s the very definition of high entropy.

                So, while it’s certainly possible to do something meaningful with a system in an high state of entropy, it’s highly unlikely and / or expensive to do so.

                Let’s take it a step further, and use your quantum RNG to generate lottery numbers — and, indeed, let’s even set it up in such a way that everybody participating in the lottery is guaranteed to have at least one winning Many-Worlds branch. Everybody who plays will have future selves who both won and lost the lottery. By Born’s Rule, we know that the overwhelming majority of a particular individual’s future selves will have lost, but some small percentage will have won.

                We can agree that winning a lottery, especially if the payout is sufficiently large, can be a dramatically life-changing event.

                But what does this say about cognition, about choice, about “Free Will”? The future selves who won no more chose to win than the future selves who lost chose to lose. And the decisions all of them make follow inevitably from the state of the Universe at the time of the decision. A loser can’t decide to spend the money he didn’t win any more than a winer can decide to pay taxes based on the money his alternate didn’t win.

                So, again…while Quantum Mechanics is full of all sorts of amazing and counterintuitive true facts about the Universe…it really, truly, honestly is completely and totally and utterly irrelevant to human cognition and the decision-making process.




              2. Ben, it seems you’ve just argued that winning the lottery is “completely and totally and utterly irrelevant” to my decision whether or not to buy a yacht.

                All inputs are fodder for the meat computer, even those arising from macro-scale manifestations of quantum events. As I said elsewhere in this thread, the wave function forks constantly, and we fork with it, and somewhere downstream we make different decisions as a result. So QM is as relevant to our decisions as “the laws of physics” or “genes and environment”.

                None of this, of course, refutes physicalism or implies that QM confers magical free will. The point, as I said a few comments back, is that Jerry needs a more coherent formulation of determinism than he gives in (B) if he wants his case to be convincing to people who know something about physics.

  48. I think I’m a C and agree that the A’s are really the problem. And I think the way to deal with the A’s is to convince them that there are no gods. Religion is the problem that needs solving not free will. Most non religious people already “know” that free will is bogus – they know that people do what they do because of who they are and how they were raised.

  49. call me devil’s advocate if you like or call these dumb questions:

    Is the choice either determinist or libertarian, and if so, what says it’s either determinist or libertarian? Sounds like a false dilemma…. if i follow… then choice C in the scheme above would magically harmonize them?

    If libertarian-determinist then a range or a “spectrum”, what defines this range? What says there will never be a philosopher who comes along in 50 years and gives another possibility or dimension to the question? Are all the pieces before us now?

    What question does free will answer?

  50. Consciousness, free will, the subconscious, imagination are sensations produced by the brain. Can we describe the above non-anecdotally?

    Study the brain unencumbered by such metaphysical concepts and we might learn something of value.

  51. The debate between A and (B + C) only partly maps to what people mean by free will. If the term, as used by people, means a variety of things, then drawing out a philosophical boundary around that is going to be problematic. The issue is it you say “Free will doesn’t exist”, some of the things that people take it to mean are things that are perfectly compatible with hard determinism.

    So in the philosophical debate, I’d agree that the arguments for the combined class (B+C) are in opposition to A, but I’m not sure how that translates down to the everyday use of free will.

    As a side what, what I’ve never being able to fully understand is why the debate takes place at a metaphysical level, rather than at a sociobiological level. Aren’t we concerned with the human ability to make judgements, decisions, exercise will, etc.? If so, arguing over the laws of physics doesn’t make a lot of sense. What we’re after is a theory of people.

  52. I agree that B+C vs A is much more important and that the difference between B and C is just semantics. I’m a compatibilist.

  53. If we decide that justice is merely acting to maximize wellness in society by rehabilitating offenders, and sometimes deterring potential offenders, then the distinctions between A,B and C have no practical effects. However, there is surely some correlation in the real world due to human nature: if you start with B or C as your view then you maybe are more likely to choose rehabilitative justice. If you start with A you can certainly still choose rehabilitation, but you maybe are less likely to rule out extremes of harsh punishment based on the idea that the offender “deserves it.”

  54. [reads terms again]… so there’s two flavors of free will… one for A and one for B…. that aren’t equal….

  55. … I can’t come up with an answer – I don’t understand free will. It seems to me that free will is ignoring external factors like incentives, asteroids, competitors,… things or patterns made of atoms or electrical gradients in neurons that operate without the free willer’s will….

  56. One of the main reasons I’m a compatibilist is because the libertarian version is completely self contradictory. We don’t need a word for a four sided triangle and we don’t need the phrase ‘free will’ to refer to

    were we to return to a previous situation of “choosing”, with every particle in the universe configured as it was before, we really could have chosen differently from how we did

    Even if you add supernatural souls that can have thoughts, memories, or whatever sort of qualities you want, you still cannot get libertarian free will. If the physical universe AND the non-physical whatever were both in configured the same way they were for a previously made choice, either you would always make that choice (determinism holds) or you wouldn’t (determinism with some randomness holds).

  57. I have no doubt that the question whether libertarian free will exists is more important than the semantic quibbles between compatibilism and incompatibilism.

    It follows then, however, that telling compatibilists that they shouldn’t use words (decide, free will, choice…) in their well-established, common sense ways is not where incompatibilists should invest their efforts.

    Another observation one might make is that while the question of libertarian free will is important, it is also trivially answered and thus not very interesting. The idea was incoherent from the get-go because the options always reduce to predictability versus randomness, and the latter is not what libertarians mean. So after two seconds of thought it is like, “done now, what’s next?”

  58. Has no one here ever looked at a restaurant menu and selected what they wanted to eat? Next time, watch the process.

    1. The event in which you decide what to eat has a myriad contributory causes – that you are not aware of. A lot of chemicals are moving around in your brain and you are not aware of that. You can subjectively “watch the process” all day long and not have anything approaching a reductionist physical picture. Perhaps you have heard of framing? An interesting subject there, too.

      1. “Myriad” may be far too small a quantity.

        If you can watch the process, who is it that is watching? And can you then watch the watching? How far can that be carried?

        Feedback loops.
        I have read that the brain has far more feedback loops than inputs from outside the brain. Feedback loops monitor processes, with large levels of loop nesting. Billions of feedback loops, interconnecting and nesting.

        It can still be argued that the brain’s operation is deterministic despite all these loops and the myriad of “myriads” of feedback loop-caused modifications, but that seems to push the definition of determinism out to absurdity, where we must take it on faith that – in the operation of the human brain – determinism rules completely.

        1. I think the question of reductionism is more to the point. If the brain is composed of particles and forces – if it is ONLY composed of them, and the operation of a particle is the same whether that particle is part of a “feedback loop” or just sitting around in some intracellular fluid, and the particles don’t have agency, then the patterns comprised of them don’t have conscious agency. There is some faith to take reductionism as 100% since quantum has raised obstacles, but on the other hand reductionism marches on.

          1. Sean Carroll addresses the thinking error you’re demonstrating in The Big Picture.

            To use a different analogy, you can either describe the contents of a balloon by listing all the molecules and giving the position and momentum vector for each or you can describe it in terms of the Ideal Gas Law with pressure and temperature and volume and number of particles.

            If you try to mix the two languages, you run into trouble. That is, you can’t say that individual molecules have pressure or that the volume of the balloon is mere quarks and electrons.

            Your objections with respect to reductionism make exactly as much sense. When you understand the fundamental flaw in the preceding paragraph — why it doesn’t even make superficial sense — you’ll understand the fundamental flaw in your objections.




  59. No, I don’t think the question of free will is the battleground where the future of society is decided. People will argue ad hoc for or against free will, depending on what their real goal is. Listen to how Quine frames it in this television interview:

  60. I am with you in group B, and I agree that convincing A group to become a B or C adherent is of most importance. There is no Magic Sauce(TM) in our brains that allows us to escape or bend the laws of physics. One could conceive of quantum indeterminates triggering events earlier or later, were one to reset the clock and replay it, but this in no way effects my “ability” to decide what I’ll do next.

  61. I’m in category C and I think the really crucial issue here is at the end when you talk about the consequences of not having Libertarian Free Will.

    You use as an example the idea that “many people are poor simply because they made the wrong choices.”

    To me, whether this is true or not is independent of whether or not Libertarian Free Will exists. Arguments for and against make equal sense in both contexts.

    Even though we are deterministic, we still make choices through computational processes in our brains and those choices are informed by qualities that are unique to each person. We can evaluate those choices and we can also evaluate people based on their qualities and the behavioral trends they produce. We can call someone lazy, violent, generous, etc… and use those evaluations to inform our social interactions.

    It’s entirely possible for a deterministic world to exist where the sole reason some people are better off than others is because those people make better choices based on some “superior” set of qualities (in fact, that is basically a description of genetic algorithms).

    That of course isn’t true of our world, but it’s because it’s empirically wrong, not because it’s incompatible with determinism.

    I think if common ground is to be found between groups B and C, people in group B should focus on making the case for why a lack of Libertarian Free Will impacts arguments like your example and other topics like criminal justice.

  62. No matter how you slice it and re-define terms, contra-causal free will requires magic. If you don’t believe in magic, you should not believe in free will of the type beloved of religious folk but deprecated by Dan Dennett 🙂

  63. No free will is a hard pill to swallow but the good doctor’s (JAC et al) diagnosis and prescription looks to be the way to go. I feel myself healthier already.

  64. I guess I fit into B.
    I think that the greatest difference between A and (B+C) is that those fitting into A want to change people to make society different (I hesitate to say better, but they probably view it that way) while those fitting into B+C would need to change the environment if they are to expect change in actions. The consequences extend into all public policy as well as day-to-day interactions.

  65. Agree to the question asked and the body of the post.
    Given we have some astounding optical, cognitive illusions some we cannot eliminate and others only by training and being aware. I find, given all my understanding of reading and discussions of the science, free will an illusion grounded in the status of ‘I’ (reproductive and energy rewards)to wit thousands of years of selective / cultural pressure to be just another tool of the survival kit.. like short cut heuristics and acute optical illusions. For all intensive purpose, real enough in the brain but an illusion no less in the real world.
    Simply, a brain artefact born of the human exclusive phenomena of self realisation and agency to others.

  66. I agree with the statement in your question for readers, and I’m Type C.
    The main reason I agree is that I think Dualism is corrosive to our understanding of the way the universe works, and our existence as a vibrant part of it. It proposes a mystical, non-physical escape hatch to consider ourselves as “souls” that are in some way separate from (and importantly, in some way superior to) the matter, energy, and information that make up everything that is. We’re not really stardust, Dualism claims, we’re better than yucky stardust. Jesus-dust or something.

    But I’d like to add in turn something that annoys me: the often poorly-expressed claim by hard determinists in such discussions that determinism implies we are simply the hardware in our brains. In fact, we are the combination of our hardware and our software and the information we come preloaded with and the information we absorb from our environment — and this is an important distinction.
    Consider an analogy: if a human was an “Angry Birds” cellphone game, the neurophysiologists could poke and prod at the chips and transistors, even describe them to an atomic level of detail, without ever understanding what made those birds so interesting (or so angry)!

  67. In my opinion, determinism as a theory of human behaviour would work if the world itself were simple and deterministic.

    But the world is complex and probabilistic and to some extent chaotic.

    We have evolved to cope with such a world, and if 50% of our coping mechanisms are genetic and 50% learned,then our responses may appear to be free will, at least subjectively.

    A reductionist may be able to set up controlled laboratory experiments that demonstrate that certain responses have been inherited or learned.

    But would such experiments tell us much about the behaviour of the same experimental subjects in the wide wild world?

    Since 1960, I have lived and studied and worked in 18 countries engaged in over 40 team activities with people of at least a dozen religions and ethnic groups.


    I get a phone call or an email now and then and a week or so later I touch down in another country. It takes 3 to 6 months to learn enough of a new language to get along socially. Sometimes I already know the language or a language I know is widely spoken.

    In Timor Leste, for example, Indonesian and English are widely spoken. Laws and government documents are in Portuguese, which is easy to read if you know Spanish.

    Although I theorize that most behaviour I have witnessed is programmed by nature or nurture, because so much of life is complex and chaotic, much of human behaviour appears indistinguishable from free will.

    I therefore conclude that my “no-free-will” theory is not empirically verifiable, that is I cannot use the theory to make reliable predictions.

    The theory is therefore on the borderline between science and pseudoscience, better than psychoanalysis but not as scientific as psychiatry.

    So my strategy is to “play by ear” more or less like playing Dixieland jazz. I know I will get my turn to improvise, to act as facilitator of the “music” the team is playing. And once in a while, I am selected as the lead artist whether or not that is my official role.

    Anthropologists tell us that leadership by authority (in contrast to leadership based on power)arises from trust within groups up to about 30 because all members know all other members.

    When a trusted leader cuts a cake s/he uses authority to allow others in the group to choose their portions in an egalitarian way. A leader with power allocates the size of portions according to the power structure within the group.

    In my experience, trust probably works in unstructured unrelated groups up to 75. Perhaps, trust dominates more than power even in groups of greater size, but I do not personally have experience of such large groups.

    I believe social organization based on trust is characterized as “leadership by authority without power”. As groups become larger, leadership tends to rely more on power to compel followers.

    I am aware that my will to lead comes as a result of being the eldest in a single-working-parent family. Not exactly a matter of free-will.

    My lifestyle may be a direct result of reading adventure and travel books to escape a troubled childhood environment. That and the fact that I could not understand why Tarzan would want to live with apes when all he had to do was make friends with the natives, as did Louis Leakey.

  68. The never ending argument. I don’t wish to add anything to it but prefer to throw in some name calling instead. I didn’t invent these, just picked them up from a long past issue of Philosophy Now (my apologies to the true author).

    1. Clockwork Determinist (that’s Jerry of course.)
    2. Free Willie (with or without a ghost in the machine)
    3. Beneath Contemptibilist (why is compatibilism beneath contempt? Simple. These folks want to have their cake and eat it too. Is there anything more contemptible than that?)

  69. Dr Coyne, I agree with you because finding myself in category B has made me a more selfless, compassionate, and charitable person. I also notice that others I know in categories B and C are similarly selfless, while I see a normal mix from those in category A (some compassionate, some “selfish”, if that’s the right word).

  70. Take this problem from the top-down. What is human reality that as self aware beings we experience? It is described like this, every word matters: We try to choose a good future and avoid a bad future. At individual lever and at group level. In this process we fail or succeed. Individually or as a group. This describes evolution too, not surprisingly. This is the reality we are living and is equivalent to a non predetermined future. With such a future, only, this reality makes sense. Accepting this is the way our mind to seek order, as it should, in understanding the world. This is what we always do. This is what made us.

    The deterministic assumption denies that reality. Following its consequences instead of solving problems suggests their elimination. When all actions, all changes are predetermined the future is predetermined. Full stop. After that starts chaos. Nothing makes sense. A mind seeks chaos now. Why to do that?

    Between deterministic assumption and human reality, we have to choose. Like between order and chaos. Both cannot be truth. But human reality is conceived as the only way we can conceive a fact. The way we know anything about anything. So by reductio ad absurdum deterministic assumption should be wrong.

    Or… a better assumption is needed.

  71. I fully agree with your conclusion. The difference between hard and soft determinists is mainly a play of words since both have a deterministic worldview.
    One note though about the study of Sarkissian. I believe the survey was held among philosophy students. I am sure that if one would ask the average person in the street about this the percentage of ‘libertarians’ would be much higher, probably around 90-95%.

  72. To your question: Do you agree that the difference between (B + C) and A is more important than the difference between B and C?

    I lean toward “no.” Here is a first pass at some reasons. On preview, they’re not as brief as I would have liked. My apologies for that.

    I think the answer depends on (1) whether it really is the case that B and C make similar policy recommendations with respect to things like how society ought to treat criminals, and (2) whether — assuming that B and C make similar policy recommendations — it really is the case that A makes importantly *different* policy recommendations.

    Let’s take (2) first. I don’t see any reason to think that A implies importantly different policies than does B.

    You want to say that retributive punishment is nonsensical on a B-type view. But I’m not so sure. I’m not a fan of retributivism myself, but I don’t think that it is ruled out by commitment to hard determinism. If I understand you, your argument is supposed to be something like:

    [0] Hard determinism is true.

    [1] If hard determinism is true, then no one has moral responsibility.

    [2] If no one has moral responsibility, then the policy of retributive punishment is irrational.

    [3] We should not have irrational policies.

    [4] We should not have the policy of retributive punishment.

    I’m worried about premiss [2]. It seems to me that any argument for [2] would have to show that the ideas of moral responsibility and moral desert are necessarily or conceptually connected. Many people seem to think that they are so connected: the section on desert in the SEP article on retributivism is an illustration (http://plato.stanford.edu/entries/justice-retributive/#Des). But I’m not so sure. It seems to me that one may be strictly liable for some harm — and hence deserve punishment in the sense the retributivist wants — without being morally responsible for the harm (perhaps because moral responsibility requires either intention to cause the harm or negligence). If so, then a policy of retributive punishment *might* be rational even if no one is ever morally responsible for anything. Of course, it might not be rational for other reasons (as I think is the case), but our lack of moral responsibility wouldn’t be one of those reasons.

    Now, suppose you’re right, and hard determinism implies that retributive punishment is a bad approach. Does libertarianism have the implication that retributivism is a *good* approach. I don’t think it does. Rather, it seems to me that libertarianism merely leaves open the possibility that retributivism is a good approach. If so, then if you are right about the implications of hard determinism, then A and B type positions differ only in that one closes off retributivism while the other does not. But if I’m right, then a libertarian could still think that retributivism is a bad approach. Just to put some flesh on this, take a view like the one Socrates sketches in the Republic: people have souls and the point of punishment is to refine the soul and make it more just. One might be a libertarian committed to souls and still think that the purpose of punishment is not retribution but improvement. If so, then giving up libertarianism wouldn’t necessarily make any difference to the policies about punishment that one endorses.

    (The foregoing seems to assume that the reasons one gives for punishments are really important, even when the punishments themselves are the same. I’m not so sure that such an assumption is correct. If you recommend incarceration for five years on the grounds that it will rehabilitate the criminal and deter further crimes, and I recommend identical incarceration on the grounds that the criminal deserves punishment, does it really make any difference to the criminal? It seems to me that in order to establish that there are important policy differences, the positions need to do more than assert different grounds for the same punishments. They need to recommend *different punishments*.)

    Suppose you either reject my arguments up to now or think that the difference we’re left with at this point — namely, that B type views rule out retributivism, while A type views do not — is sufficiently important to warrant a “yes” answer to your main question. So far, we’ve been assuming that B and C type views have similar policy implications, but if ruling out versus not ruling out retributivism is a significant difference, then B and C type views do not appear to have similar policy implications. After all, several philosophers have appealed to compatibilism specifically to defend the idea of retributivist desert in the face of arguments similar to the one I’ve attributed to you. Maybe those philosophers are making mistakes, and C type views really do rule out retributivism. I would be interested to see an argument to that effect.

    Setting aside the abstract debate, I think the real crux is in one of your parentheticals. Is it the case, as you assert, that “the bars of a jail are no different from the bars in our mind that compel us to do one thing rather than another.” It seems to me that there are relevant differences between the two, which make talk of free will and moral responsibility perfectly reasonable.

  73. Yes and No.

    Yes, it is broadly speaking Libertarians vs Compatibilists and Hard Determinists regarding the metaphysical status of “Free Will”. B and C may disagree, but it is a somewhat semantic argument.

    But no, Compatibilists harshly disagree with Hard Determinists about the relevance this has to society. I, and many others, absolutely reject that “fully grasping determinism has a huge potential effect on human behavior, including in particular how we treat transgressors or criminals”. As other commenters have put it, that’s a category mistake; our notions of justice and responsibility can survive basically intact. Therefore, *in the social sphere*, Compatibilists are more on the side of Libertarians, who are “right for the wrong reasons”. (Needless to say, being a Compatibilist does not mean you support harsh retributive justice, only that that is a very different argument that has nothing to do with the metaphysical status of Free Will.)

    P.S. I can’t resist:
    “we all surely agree that accepting determinism will sink the libertarian free will inherent in many religions”
    Clearly, you’ve never heard of Calvinism.

  74. I’m solidly in class C. The long standing debates here at WEIT have only clarified/verified/justified my view in this matter. I am convinced that B is grossly over-reductionist and moreover, is actually quite incoherent. Although I agree the debate between C and B is mostly a philosophical one, I still feel that B is dangerously wrong – not only wrong philosophically but wrong as a representation of reality.
    As for the justification of argument from CONSEQUENCES, which seems to motivate many B debaters (eliminating blame etc.) I can only comment that this should have no place in scientific debate. In any case the consequences of B can be argued to be highly deleterious to social well being.
    So in conclusion, I also feel it is wise to stop the B vs. C squabble and concentrate on promoting an understanding of determinism – this is not only a practical way to advance social well being, it is most aligned with the evidential basis of reality which we adhere to in Science.

  75. I broadly agree with comments at 7, 10, and 40. Physics and metaphysics are of little value for building useful models of human behaviour or constructing systems of ethics. Ethical philosophy, psychology, and sociology can be useful. I also think evolutionary biology could help, and that’s mostly what I want to talk about.

    I think that self-awareness, and a ‘sense of free will’ or a ‘feeling of agency’ are adaptive, though they also involve some self-deception which may well be nonadaptive. I do not think I have a solution for the hard problem of consciousness (subjective experience).

    Self-awareness. Imagine you are an individual in a group of smart social animals, where the behaviour of others is very important to your prospects, and that behaviour cannot be assigned to roles (alpha female, my child, etc) but requires modelling the behaviour of individuals. It become important to track this behaviour (individuals’ strengths and weaknesses, a tally of favours given and received, etc). Each individual acquires a reputation, and your reputation is important to you.

    Suppose also that these animals have a language which allows them to name individuals, and talk about their behviours (as well as talk about features of the environment). This could be a primitive language, consisting of little more than juxtapositions of nouns and proper nouns. Now the animals can talk about one another, and affect reputations. Gossip is born.

    If these animals haven’t already started to be self-aware, each seeing themself as another individual, tracking their own behaviour, and talking about themselves (in the third person), the selection pressure to develop such an ability seems clear.

    Your ‘self’ is first and foremost a social construct. It exists to represent your behaviour to others. it’s like the PR department of your brain. If you believe it to be your soul, or your true self, or your inner core, you have fallen for your own publicity.

    Free will. This is similar but simpler. The capacity for imagination is adaptive. It is useful to be able to imagine alternative scenarios which might happen or might have happened, depending on choices you make or made. The more effort you put into mapping out alternatives in a way that seems plausible or ‘realistic’ to you, the more real they seem to you. If you misinterpret your imagined scenarios as actual possibilities, you will have a sense of free will.

    A, B, C. It should by now be obvious I am not in category A. I don’t like B or C either. Some days I’m a hard nondeterminist, some days I’m a neutral monist. Neither is popular.

  76. I’m in the B camp. I agree that dispelling libertarian free will is the most important issue and that your perception of free will affects your attitudes towards different social issues (and even personal ones). But I still believe that there are some differences between group B and C that are worth highlighting.

    First off, saying that we can have FREE will under determinism is just being too loose with the language for me. This perception of free will should be called something else. I could accept functional free will https://danielmiessler.com/blog/functional-free-will/.

    Secondly, I believe that most people’s first intuition about free will is of the libertarian variety and probably more so historically.

    Finally, compatibilists insistence that we have free will can easily get confused with and give support to the libertarian version: “Hey, even an atheist like Daniel Dennett says we have free will!” I don’t think the average person grasps that he, and other compatibilists like him, still believe that we are 100% determined by causes that we cannot control.

  77. I am a libertarian in the sense used here, yet I do not consider myself a person denying the laws of physics, for I do not know any law of physics “forcing” living systems to work in a rigidly deterministic way (one has to leave room for random changes, if not anything else).

    I also fear that widespread acceptance of determinism in human decisions may lead to even more widespread and dismal excesses of applied behavior analysis (e.g. “normalizing” a “sissy” boy by beating him when he “behaves like a girl” – something that has been attempted).

  78. Those who believe to be autonomous agents might find it interesting to think a bit about autonomous vehicles (Google cars and the like). Their actions depend on hardware, algorithms and inputs. We wouldn’t say they have free will, would we? What would have to be changed to make an autonomous vehicle truly autonomous?

    1. As far as I’m concerned, a self-driving vehicle has (more or less) the same capacity for self-directed navigation as a live human taxi driver. In ordinary speech, when we’re not quibbling about semantics, we say that they’re both free to choose an appropriate route, while recognizing that they both use (quasi-)algorithmic processes — judgment rather than unmotivated whimsy — to make their routing decisions.

      Is that “free will” in any meaningful sense? Now we’re back to quibbling about semantics.