My definition of free will

June 13, 2011 • 4:38 pm

Because everyone is demanding, rightly, that one define one’s concept of free will before discussing it, I will offer one here, subject, of course, to modification if it proves incoherent or useless:

A person makes a decision.  Now rewind the tape leading up to the moment of that decision.  Everything remains exactly the same as before:  the history of the universe, the person’s entire life, experiences, and body, and so on.  If on the rewind the person makes a decision different from the one before, then that person had free will.

Because I’m a physical determinist, I don’t think that making a different decision is possible, and so free will is impossible. But at least the definition has the merit of being specific. And it’s sort of intuitive: put a guy at the ice cream counter, and if he decides to order butter pecan instead of chocolate, well, that’s a free choice.

154 thoughts on “My definition of free will

  1. At first blush, quantum randomness would provide a source for your definition of free will. Perhaps that one C14 atom just right of the thalamus decays at that point in the decision-making process and the resulting electron flips a different neural circuit this time.

    I doubt that’s what you have in mind….



    1. I don’t think we’re in a position to even say that randomness could have happened any other way; what we observe in quantum fluctuations as random could just as well be determined (ie. couldn’t have happened any other way). I feel there’s a good chance (hah!) that the universe and all the laws of physics are all pre-determined. Lucky for our sanity, the future is still unknown (and hence possibly *look* random to us), even if it can’t be anything but what it will become.

      1. “I don’t think we’re in a position to even say that randomness could have happened any other way;”

        I agree, but likewise I think there is no way we can say that the “classical mechanics”/”complete determinist”/”Newtonian” position — the idea that the position of all particles since the Big Bang until now was perfectly predetermined, like perfectly interacting billiard balls would be, even though the necessary particle detection and calculation capabilities are forever beyond us — is scientifically proven. I don’t think we can even say that it is scientifically preferred position. My sense of it is that many more physicists would support the “true quantum randomness” position or express agnosticism.

        Especially if you add in some chaos theory, it appears that the possibility of true randomness would make it totally hopeless to attempt to prove classical determinism, which appears to be essential for disproving free will on the above definition.

        I think many physicists would say that they don’t care which interpretation of quantum mechanics is correct, because they can’t tell the difference experimentally, and it doesn’t make any difference in terms of making mathematical predictions.

        I tend to think this way about “free will” — the metaphysical “truth” of the matter may be fundamentally unreachable, due to quantum randomness and chaos and whatever else — and it doesn’t matter operationally, where we all operate as if we and others have free will, and it seems to be a decent approximation of what is going on, whatever the metaphysical truth of it.

        1. What would be interesting to discuss is the concept of “free will” as an emergent behavior. Certainly a lot of science, and parts of evolutionary biology, have been moving away from strict reductionism and towards the view that certain phenomenon are emergent at particular scales of analysis and have their own behaviors that cannot be predicted from the rules that the “particles” follow in the system in question.

          E.g. if the mind is basically the very complex interaction of electrical pulses, and takes in diverse inputs, runs some simulations of future results of various actions, and then makes decisions which produce action, all given limited knowledge and processing power — why would anyone need more, or less, “free will” than that?

        2. Just to raise a wrinkle. It’s often pointed out that random won’t do the job we’re looking for when it comes to free will. If it’s completely random whether or not I do A or B then I didn’t choose one way or the other, it was equally determined, just random. The problem is that as BG points out that’s all it takes to meet Jerry’s criterea for free will.

      2. What’s the difference between a universe in which “true randomness” exists and a universe in which the Parcae have predetermined all measurement outcomes, but have made them look random? That sounds like saying that the universe was created last week, but God was a sick jocker who wanted to give us the impression that it was billions of years old.

        1. This is a memory in your mind that I implanted when I created the universe at the start of the video.

          Yes, it is backdated.

          And yes, you have memories of reading this message before you watched the video.

          But trust me – all of those memories were implanted in your mind.

          By me.

          When I created the universe.

          I even put someone else’s name at the top of the message, just to try and trick you more.



        2. “What’s the difference between a universe in which “true randomness” exists and a universe in which the Parcae have predetermined all measurement outcomes, but have made them look random? That sounds like saying that the universe was created last week, but God was a sick jocker who wanted to give us the impression that it was billions of years old.”

          But it is also the position you get to if you think that deep down, absolutely everything is the product of exact mathematically predictable particle interactions. Whether or not we can ever tell the difference, given physical constraints on human measurement capabilities, is a different question…

      3. Yes we are in that position. It’s random, each calculation is random, if you rewind the tape and play it again you would get a random result.
        a physicist

        I think Jerry’s point is that given the same results for all particles (that would be part of the history of the universe) the brain is incapable of making a different decision. I agree.

        PS. If anyone tries to argue that quantum mechanical calculations are washed out on the macro scale I will find a way to slap you through the internet.

      4. “I don’t think we’re in a position to even say that randomness could have happened any other way”

        In our mathematical models, the randomness can happen any other way. In a thought experiment one can argue that we don’t really know, but in fact that doesn’t matter at all because this “invariable outcome of a single random event” remains hypothetical and with no evidence to suggest it is in any way a superior view to the completely random case. To say that the apparent randomness may have been predetermined is merely begging the question.

    2. Actually, I don’t think you can say that. Currently, we know that there is no means to precisely predict at which point in the future a given atom will decay, and therefore we call it “random”. However, if it were possible to rewind the “tape” of the universe and replay it perfectly, that would mean that an entity which could do so could predict everything that would happen up until the point of rewinding – down to nondeterministic effects like nuclear decay, making them deterministic. Essentially, in the view of things given by Prof Coyne’s definition, free will would not be the ability to make a different decision in the present, but to make a different decision in something else’s past.

      This definition is flawed in that it intrinsically begs the, uh, anti-question* I guess – by including in the definition of free will the possibility of rewinding and replaying history, it assumes that the universe is fundamentally deterministic and therefore that free will does not exist. It’s self contradictory; a universe in which it is possible to satisfy the condition “If on the rewind”, the consequent “the person makes a decision different from the one before” will always be false.

      Personally, I think asking whether or not we have free will is kind of a waste of time. There are no observable differences between a universe in which we have free will and one in which we do not, so why does it matter which one we live in?

      *Or maybe it anti-begs the question? Reverse begs the question? Anti-begs the reverse-question?

    3. I don’t like the definition. I think it needs to be more fixed in something that person himself is capable of doing. Even if the universe is random, and the person makes a different choice, because there’s some randomness in the universe itself, just means the person is at the will of the universe.

      There’s no free will either way, though.

  2. “If on the rewind the person makes a decision different from the one before, then that person had free will.”

    Or the person is hopelessly inconsistent.

  3. But physical determinism is false.

    Or at least it appears to be false. Maybe the many worlds interpretation or some sort of extra physics (Bohm or whatever) can save it, but that’s the usual story.

    This kind of thing is exactly what qm predicts – you set up the universe in exactly the same way, you make a measurement and you might get a different result.

    1. I agree and I think this is why JAC’s definition fails. In fact, even without non-determinism, chaos easily thrown in another wrench.

      I prefer the term “contra-causal free will” which sort of defines itself. Free Will is then any choice which is determined (or influenced) by something other than the physical and chemical constituents of our body.

      That gets around the problem of rewinding the clock (which can’t be done) and of any non-deterministic and chaotic effects.

      1. I said above that I think Jerry is saying that given the same results for those particles the person is capable of making a different decision. Even with QM you don’t get free will you just get randomness, you don’t control the randomness.

        1. Yes, but if you rewind the tape you have a very high probability of getting a different result. Also in the Many-Worlds interpretation of QM the this world and any individual world is irreducibly stochastic. It is only the ensemble that is deterministic. There is a wave-function of the universe which follows a unitary evolution that is deterministic. By the way Leneord Susskind has a new paper out linking string and the cosmic landscape and the MWI (the arXiv URL is over in a recent thread at Cosmic Variance) the MWI be be testable with future technologies.

          Our observations merely tell us which of the many worlds we reside in. Essentially stochastic free will comes from the contrafactuals arising from us splitting the histories. Roger Penrose may well be wrong about about coherent microtubules (Tegmark’s decoherence calculations) however his anaesthesiologist colleague Hameroff points only 50% off neurons with an action potential go on to release neurotransmitters. There is something stochastic going on here probably bubbling up through deterministic chaos from quantum indeterminacy at the molecular level (this by the way is is also the ultimate source of genetic variation – remember Schrödinger’s comments on X-rays, mutation and QM in “What is Life”). When a lot of neurons are involved the situation is deterministic and we get causal free will – Dennett’s form of free will. When only a few neurons are involved in switching a decision one way or the other we potentially get a QM derived stochastic free will. These are the difficult decision were you yourself are likely to end up tossing a coin or consulting the I Ching. The tape is not likely to run the same if rerun again.

        2. I agree with your intent and think we need to update the definition. As it stands, merely having a different outcome when we roll back the clock is something which happens through randomness. In suspect that this is almost the only explanation.

          If the clock is rewound perfectly and the person isn’t told that there’s anything special about the decision, you’d expect them to make the exact same choice unless there was some random factor tossed in. I think “Free Will” isn’t merely about making unpredictable choices but about making decisions which aren’t forced upon us by our brain and biochemicals.


          1. Who or what are we that you think is separate from our “brain and biochemicals”? I don’t see how our “brain and biochemicals” can force a decision upon us because I see our brain and it’s biochemistry as being us.

            1. Well yeah, exactly. If our choices are driven by our bodies (because our mind arises from it), then how can we have free will unless you believe we have something like a soul? (Even then I think the problem remains.)

        3. Sure. My point is that Jerry’s reasoning and definition can’t work, not that qm (or anything else) provides a mechanism for free will.

          The larger point is this: There is a good reason to think that free will exists; our immediate experience with it. We seem to be making decisions. This may be mistaken, certainly, and it’s interesting to look at the experiments that might shed some light on it. But, while suggestive, none are exactly a nail in the coffin for free will, partly because we understand so little about what is meant by the term, or the obviously related idea of consciousness.

          So we call for a definition. But what the attempts seem to say, to me anyway, is that our understanding of our experience with free will is too lacking to properly formulate it. There’s no shame in that. After all, there are many problems that fall in this category, and many more that used to. Maybe later we’ll know more.

          Appeals here to determinism are wrong. The world isn’t deterministic. (Probably.) In fact, I don’t see how any philosophical reasoning like this is helpful here. When people know what they’re talking about, they site evidence, not principles.

          This is also not like arguing the Existence of God. Here the phenomenon seems to exist. It’s the onus of the person arguing against the possibility of free will to show it’s an illusion. The same holds for an argument that free will exists. But until we know more, I’m just thinking that there’s no good reason to preclude the idea. We’re not talking about ghosts in the machine here, just an apparent process that defies easy explanation at the moment.

  4. Because I’m a physical determinist …

    Even ignoring QM, which we can’t, it’s impossible to measure all the variables required to test “physical determinism”. Statistical mechanics alone vitiates your definition of free will. Given that it’s physically impossible to test your definition of free will, what does it really mean?

  5. Jerry

    I take it for granted there is this thing called ‘will’ that may-or-may-not be free, depending on our understanding of what ‘free’ means in this context (in your case, ‘free’ means ‘deterministic’ – which is fine with me).

    Out of interest, it would be interesting to have you suggest a definition of ‘will’ in a way that would apply regardless of whether it is ‘free’ or not.

    I’ve attempted this myself in previous threads, but I’m interested in what you think.

    I’ve found that, once meaningfully defined, ‘will’ itself does a lot of the heavy lifting people desire from free will, but without the the pesky metaphysical argumentation that comes with it.

  6. If on the rewind the person makes a decision different from the one before, then that person had free will.

    The basic impossibility – apparently – of doing that rewinding would seem to make the determination of that ability equally impossible – sort of a formally undecidable proposition. Maybe the problem is sort of like the dichotomy, the basic duality, associated with the particle-wave duality, a question of perspective: we simply lack the technology, the knowledge or the capability to resolve the necessary details – something we have to take on faith. But reminds me of a passage in a book, Plato and a Platypus Walk Into A Bar: Understanding Philosophy Through Jokes (Cathcart & Klein):

    When asked whether he believed in free will, twentieth-century novelist Issac Bashevis Singer replied, tongue-in-cheek, “[Yes, but] I have no choice”. [pg 21]

  7. what if you rewound the tape 100 times and 99 times the person made the same decision, but 1 time was different? would we say there’s a spectrum of free will?

    of course its untestable, but it would be fascinating to discover a continuum of free will whether by individual or by unique situation.

    1. Yes… it strikes me that that’s a rather poor definition of free will for a scientist to come up with! 😉

      And this is one sense in which the title of the Social Psychology Eye article – “Free will is the new god” – is accurate: It is difficult to arrive at a coherent hypothesis that is testable!


  8. Free will occurs in the present, inarguably – and futility exists as an illusion SIMPLY BECAUSE of awareness of the past. But the past will never be accessible – so it will always be FUTILE when viewed (in the present – as there is no alternative)

    I think this is what all determinists DO NOT REALIZE they agree with.

  9. The philosophical problem of free will makes the most sense in the context of religion and saving face. If antecedent states of the universe directly caused present states, then at the very beginning we were all destined to do what we did and we never got to choose who we would become and how that person who is us will act, then we can’t be held responsible for our sins, since it was god’s fault. So the apologists had to come up with the idea of free will, the notion that our will causes our actions without anything causing the will. It makes little sense but then again what does in religion? Humans naturally assume this so the myth has been propagated, we evolved to see ourselves as the puppet-master and not a bundle of cells, some of them neurons and those neurons influenced by our external environment and their very nature, but as the commanders of the body, since that makes us act in a desirable manner that gets things done.

    The question now is how free is the conscious mind, even if it was caused by antecedent causes, to influence future behavior? Depends on the degree to which we deliberate and examine what we should do in the future.

  10. I like the definition. That’s pretty much how I define ‘free will’. But, unlike you, I don’t have faith that we live in a deterministic universe.

    As others have already mentioned, if qm theory is accurate, it indicates that we do NOT live in a deterministic world. Hence, I do not assume that if ‘rewinding’ were possible, the outcome is inevitable.

    But I don’t see anything particularly ‘magical’ about that. We effect our future through our choices in the present. We can consciously attempt to do so. I think of it as our brains modifying the algorithms that it/we use to make our decisions.

    Since that fits my definition of ‘free will’ – which is identical to yours above – I conclude that we do, in fact, have free will.

    However, I concede that if there is no ‘self’, then there is no ‘free will’ possible. I do, however, tend to believe in the idea that we are each a ‘self’ which can possess such qualities as ‘free will’.

  11. We have to realize; FREE WILL – like PROBABILITY – can only be thought about in the context of RIGID REALITY.

    The instant that we enter in hypothetical factors, like “rewinding a tape”, reality becomes extremely distorded, because these ideas allude to concepts of TIME TRAVEL, which are a deal breaker when contemplating reality. Time travel/ the past / the future are NOT THIS REALITY. there is no use contemplating them.

    To contemplate free will, we have to realize: THERE IS ONLY THE PRESENT IN THIS REALITY

  12. If on the rewind the person makes a decision different from the one before, then that person had free will.

    Even if the first decision was the best choice?

  13. I have to agree with others, that isn’t a very useful definition. I mentioned this a long time ago, but “free will” is a supernatural claim, by default, because as soon as dualism is rejected, there is nothing coherent to say about free will. Well, heck, as with all supernatural claims, it never was coherent to begin with.

  14. the ILLUSION of the past as something that exists now is what tricks us into thinking that FUTILITY IS NOW because of our seeing that there HAS BEEN FUTILITY. our knowledge of the past, and it’s clear status as FUTILE, makes us think it is in competition with the present. But does the past really exist, or do we just remember it? Yes, the past had to be what it is – but the present is INARGUABLY subject to change – and that trumps futility because we cannot say that the past existS (present tense of exist)

    1. Come on–I said a person making a decision, not a nucleus decaying! My assumption here is, of course, that quantum indeterminacy at the moment of a decision cannot influence that decision.

      1. But this raises a very serious question:

        What, precisely, is the “moment of a decision”?

        In our brains, cognition is a slow and messy process. I don’t think it’s possible to pin down a particular pair of units of Planck time, one of which there was no decision and the immediately following one where the was a decision.



        1. Exactly – this is the major weakness of Jerry’s definition, and another one Dennett spends a long time on clarifying. What is a person and how does it make a decision? And what if the person says “I’ll have the pecan” but grabs the chocolate?

          The other weakness of course being that quantum randomness could result in a different outcome, which no one would equate with free will.

          Can we get Jerry’s definition round 2?

      2. This appears to presume ALL non-linear effects tend to damp down, and never propagate to macroscopic scale. Which, given mathematics, would take a miracle.

      3. I was trying to show in a simple and maybe funny way that rewinding the clock and having someone (or something) go a different way could be explained by free will, but it could also be explained by quantum randomness.

        Perhaps this isn’t a factor but I don’t think that’s been established. I think a simple modification to the definition can avoid this debate. Maybe you remember the Beyond Belief talks a couple years ago, Stuart Hameroff had a long, bizarre and heavily criticized talk where he argued that QM explained free will and a bunch of other weirdness. He now runs a site called “Quantum Consciousness” which is really pumping this idea. I noticed people were commenting on this and I thought we could skirt the whole issue.

        Sorry if my attempt at humour came across as infuriating willful blindness. No disrespect was intended.

      4. I’ve asked before in free will discussions on this blog about whether it’d be worth getting a brain implant that uses quantum mechanical dice rolls to artificially endow someone with free will. I don’t think such a prosthetic would in any way improve quality of life, but it would seem to grant free will under Jerry’s definition of free will (even I’m pretty sure his newest revision of the definition).

  15. “Because I’m a physical determinist, I don’t think that making a different decision is possible, and so free will is impossible.”

    That gets rid of something called libertarian free will, which is what most people in the world think of as free will.

    It’s not what everyone thinks of as free will. It depends what you want your will free from. Having you will free from the kind of tape rewinding you mention doesn’t seem very useful. Why would you want to make a different decision faced with exactly the same information? You can get rid of that and you are left with something very useful: being something in the universe that can have a think about things and then decide what to do. That’s an amazing power. It’s also quite easy to come up with situations in which your will is not free: if you are imprisoned, or paralysed. It might be that you suffer from OCD or addictions and can’t resist compulsions.

    The issue of determinism and free will isn’t really very interesting. It’s not dealing with what people experience as freedom all the time (as against what they think of as their freedom). The kind of freedom people have as a result of their evolved self-aware minds actually requires determinism at some level. Random brains aren’t that useful!

    1. But Steve, is our ‘thinking’ actually different from a set of scales ‘deciding’ which side is heavier?

      I know there are loops of thought and goal setting and achieving, but is it not, fundamentally, the same as a set of scales when our brain decides which flavour of ice cream to go for? (Swapping gravity and mass for whatever set of metrics a brain uses to decide ice cream flavours.)

  16. A person makes a decision. Now rewind the tape leading up to the moment of that decision. Everything remains exactly the same as before: the history of the universe, the person’s entire life, experiences, and body, and so on. If on the rewind the person makes a decision different from the one before, then that person had free will.

    Ah, but here’s the snare: the total history of the universe contains within it the rational and emotional reasons behind that choice for that individual at that moment. Given the option of chocolate or butter pecan, the person who hates nuts in ice cream gladly chooses chocolate.

    If you rewind the tape and this time the person chooses butter pecan instead, they could only do so against their natural inclination and desire. Ironically, “free will” has produced a mindless robot.

    So I will be contrary and say that Jerry’s version of physical determinism supports a positive version of free will, a friendly and familiar sort of choice that comes from me, embedded fully in my history, my environment, and my responsive nature.

    The other kind is creepy. I don’t want it. I don’t want to look down and wonder why the hell I picked those nuts when I didn’t really want them. No. Make the nightmare stop… no contra-causal free will. It’s not really me anymore.

      1. If free will is only important for deciding the outcomes of our whims, then it’s not something that’s important to have.

  17. One of the podcasts I listen to during my commute is the brainscience podcast.
    They once described free will as “the conscious having veto power over the unconscious”.
    I guess I’m not grokking what’s being said here, since it would seem that the only choices available to us are those offered by the unconscious. We’re left to choose from a limited menu devised by processes unknown to us.
    I can choose to return a lost purse/wallet or I can choose to take any money out of it and toss it aside. Is my choice not an exercise of free will?

  18. If you define free will in that way it makes the answer irrelevant.

    To have will is to make choices. It requires desires, reasoning, and logic for choosing between different possibilities. To rewind the tape and change nothing means that you haven’t changed your reasoning or logic for making the decision you did. If, because there may be some small amount of indeterminacy in the universe, you *could* have acted differently has nothing to do with your “will”.

    Any indeterminate or random component of your choice makes that choice less “free” because it removes some of your ability to influence it.

    Free will is not only compatible with determinism, it absolutely requires it because without it you couldn’t have the types of thought processes that people point to when talking about free will.

  19. I would put it like this:

    If an event is the inevitable consequence of a prior set of conditions, we can say the event is “caused” or “determined” by those conditions.

    If an event is not the inevitable consequence of any set of prior conditions, if it “just happens,” we can say the event is “random” or “indeterminate.”

    But a freely-willed event is supposed to be in some third category. It is neither caused nor random. As Pinker puts it, it is “somehow both, somehow neither.” But there is no third category. “Caused” and “random” exhaust the logical possibilities. That is why I think free will, in this sense, is not simply an illusion, but just incoherent. It doesn’t MEAN anything.

  20. Just lurking here…

    ..but since this is only a philosophical discussion and one higly subjective in nature, I am not going to make any such declarations either way, let alone definitions. Other than to say, it’s likely the combination of the two, determinism and free will that what makes us what we are. It’s like the need to take a crap (determinism) and a decision of where and when to go to take one (free will). Beyond that, this whole conversation on the subject is nothing but wild speculation and naval gazing without the acid trip, IMO. Just saying.

  21. I agree that we don’t have free will under Jerry’s definition, but I don’t think that definition is particularly interesting for much of anything. In a lot of cases, I for one would hope to be able to make the same decision under slightly different circumstances. Also, since it’s actually never, ever going to happen that anyone finds themselves in identically the same circumstances, that’s not a sort of free will worth valuing, since no-one will ever get to partake of it’s benefits.

    So I think Jerry is missing something in his definition, something that’s probably really important. He’s mentioned before that he thinks the fact that we don’t have free will is existentially troubling, and is enough to make someone crazy if they really face up to it, hasn’t he? Well, that definition offers no clue as to why anyone should feel that way.

    I will also accuse Jerry of pulling that dirty argumentative trick that theists often pull: he’s offering a definition of free will that everyone (in this crowd) agrees describes something we don’t have (by analogy, the theists’ ground of being, or whatever). But in other arguments, he’s going to be using that other definition that he hasn’t yet shared, the one he thinks is so valuable (analogous to that Xian god that promises immortality in Heaven for those with the discipline to object to everything that the believer is personally biased against).

  22. I will “freely” admit that this discussion is over my head, and am currently trying to digest Daniel Dennett’s “freedom evolves”, where he tries to reconcile determinism and free will, which is also over my head. That said, whether or not we have free will, and even if an individual believes, in theory, that he or she doesn’t have free will, it would seem to be impossible to behave, consistently on a day-by-day basis, as if one didn’t believe in one’s own free will, and as if one didn’t believe that others have free will. My question then, isn’t whether we have free will; it’s, is “free will” a necessary illusion?

  23. I think JJonas8 (#25) is correct.

    But further, I think free will is an illusion because our actions are determined based on physical processes, internal states and feed-back loops, and external inputs. But free-will is also not predictable, for reasons such as quantum indetermincy and chaos theory making prediction incalculable in principle.

    A thought experiment:

    An astronaut (of her own free will) goes for a near light-speed spin around Alpha-Centauri, and by the time she returns several centuries have passed on Earth. After a while she begins to suspect that while she was away every human has been replaced with a robot. She feels that rather than having free-wills, these “humans” are 100% deterministic. Although they would completely pass a Turing test even in face-to-face encounters, she has a nagging doubt.

    Question: How can she test her hypothesis that these are in fact deterministic pseudo-humans with no free will at all?

    I think that if there is no possible test of free-will then the concept is meaningless.

    1. How do we know that we don’t live in the matrix and that the real world exists really and it’s not just an illusion?

      Real existence can’t be possibly tested, then existence is meaningless.

      Or maybe we should approach free will the same way we approach reality: scientifically.

  24. Well, if this were the way I saw free will, then I too would reject it. This type of definition, though it may be a common intuitive leap, appears to be internally contradictory unless some special circumstances were present.

    If “rewinding the tape” means noting the state of a system S1 (which describes my universe including me) immediately before a decision, taking note of the decision D1, and then resetting the system back to S2 = S1, and then taking note of decision D2, etc., then I think comparing D1 to D2 etc doesn’t tell you anything about free will unless S1 != S2. If S1 == S2, then if D1 != D2 wouldn’t imply free will because nothing underlying my motivations have changed (because S1 == S2) even though my decision did (because D1 != D2). This would imply that my decisions are influenced by something entirely outside of myself, which is definitely not free will. If S1 != S2, then of course it is possible that D1 != D2, as the difference may influence my motivations.

    One obvious kludge to solve this is dualism. Under dualism, the “soul” exists apart from the universe, and therefore S1 == S2, but some independent factor that is still somehow me can vary on each replay of the tape, and thus my decisions would be influenced by the “supernatural” portion of me, i.e. the soul or whatever. Obvious, this is isn’t a viable explanation.

    I guess what I’m saying is that, given the way you’ve defined free will, I can see no way out except appeals to the supernatural. However, I don’t see that your definition is the only reasonable one to be had. If this is indeed the definition you demand, I’d concede your point. However, I think a more satisfying route would be to explore the phenomenon of how people experience “free will” and modify its material description to match that rather than to start with a description of “free will” and accept or reject it based on whether it matches reality. I think the analogy to caloric by Dan L. in the other thread was quite nice. Heat is still worth talking about even if caloric was a very poor description of how it actually works. In the same way, free will is a useful concept, even if contracausal free will is bogus.

  25. Point well made. My disagreement is in the premise of single event definition. For me, the issue of free will has to do with whether I have an opportunity to make one of countless choices, most unknown.

    As for the science, I feel certain that everyone who posts here is probably more aware of the neuroscience of the day than I. Still, Libet’s conclusion that we have ‘free won’t’ is one piece of evidence supporting my pedestrian claim.

    the distinction for me is whether free will is viewed from a scientific or philosophical perspective.

  26. If you define free will as you did, Jerry – as the negation of determinism – then I agree that there is no free will. With Dennett and others, though, I don’t think that that sort of freedom is even coherent or desirable.

    When I talk about free will, I basically mean that the agent’s values and motives (character) played a role in determining his/her actions. Hume argued that it didn’t make sense to hold someone morally responsible for something unless it could be traced to a permanent *cause* (in the deterministic sense) in his character. In fact, it is arguably harder to reconcile moral responsibility with indeterminism than with determinism.

  27. Hume argued that it didn’t make sense to hold someone morally responsible for something unless it could be traced to a permanent *cause* (in the deterministic sense) in his character.

    I don’t see how this formulation resolves the paradox of moral responsibility in a deterministic universe. Why is someone responsible for an act that is “caused” by his character? Does he “choose” his character? How does that work?

    1. Well, then one could credit or blame his character in that case. But then, he is his character, so it comes out just the same.

      It seems you have the choice of attributing this to, and holding accountable, either the sum total of his habits, ambitions, goals, values, and standards – or an arbitrary, unpredictable, inscrutable something that renders all of those moot this time and may go completely differently in the future. How does that latter work?

  28. Ah, but Adam & Eve had free will of the only important kind in Eden:

    How blest was the created state
    Of man and woman. ere they fell,
    Compared to our unhappy state:
    We need not fear another hell.

    Naked beneath cool shades they lay;
    Enjoyment waited on desire;
    Each member did their wills obey,
    Nor could a wish set pleasure higher…

    (John Wilmot, Earl of Rochester)

    I seem to recall some theological discussion (Augustine perhaps?) of how in particular the behaviour of the unruly male member was a consequence of our fallen state and an affront to what should be our free will.

    1. How blest was the created state
      Of man and woman. ere they fell,
      Compared to our unhappy state:
      We need not fear another hell.

      Nice poetry, I think. But it is also probably somewhat overly romantic – seems akin to “ignorance is bliss” and ignores the “nasty, brutish and short” nature of nature. Personally, I’ll tend to the “better to rule in hell than to serve in heaven”, particularly if the cost of the latter is the loss of our consciousness.

      Although that does raise quite a number of problematic questions, the least of which might be the apprehension that it would still be somewhat worse to serve in hell than in heaven. But a more cogent one might be the question as to what, having once tasted the “Pierian Spring”, humanity’s next course might or should be. Reminds me of T.H. Huxley’s assertion that “… the ethical progress of society depends, not on imitating the cosmic process, still less in running away from it, but in combating it.” Even if his definition of the “cosmic process” was apparently a little foreshortened …

      1. ‘Better to reign in hell, than serve in Heaven’! Rhythm, among other things!

        I don’t think Rochester’s poem is romantic at all. Those two stanzas are suffused with irony and humour.It continues:

        But we, poor slaves to hope and fear,
        Are never of our joys secure;
        They lessen still as they draw near,
        And none but dull delights endure.

        Then, Chloris, while I duly pay
        The nobler tribute of my heart,
        Be not you so severe to say
        You love me for the frailler part.

        The ‘part’ is ‘frailler’ because, among other reasons, it stands up when you don’t want it to, and doesn’t stand up when you do want it to – as when you are in bed with the frustrated Chloris.

        1. ‘Better to reign in hell, than serve in Heaven’! Rhythm, among other things!

          Rhythm is among the “wages of sin”? 🙂 Unless you mean to suggest that it is an exclusive concomitant of the alternative state. But, not having much information on the comparative sexual performances of the other animals that continue to enjoy that “blest” condition, I’ll have to concede that that may be entirely the case …

          Though I do now see some of the humor and irony in the additional quoted stanzas – not having searched for the entire poem before. However, as to the “frailer” part and the poor “frustrated Chloris”, I still think that a small price to pay for having attained higher peaks – though I suppose that is a question of perspective. Particularly as the Earl of Rochester, in comparison with his modern cousins, probably did not have the same ready access to the fruits of “living better through chemistry” …

          1. You were quoting Milton (Paradise Lost, Book I, one of Satan’s speeches), perhaps unknowingly, but certainly wrongly. ‘Rhythm’ doesn’t refer to Rochester’s contraceptive methods, but to the rhythm of Milton’s pentameter line. Whether or not Rochester had access to modern methods of ‘living better through chemistry’ (he didn’t, of course), he made better use of what he had than most moderns:

            I’ll tell of whores attacked, their lords at home;
            Bawds’ quarters beaten up, and fortress won;
            Windows demolished, watches overcome;
            And handsome ills by my contrivance done.

            Nor shall our love-fits, Chloris, be forgot,
            When each the well-lookt linkboy strove t’enjoy,
            And the best kiss was the deciding lot
            Whether the boy fucked you, or I the boy.

            I must say that reading Rochester makes the question of free will seem desperately boring.

            1. You were quoting Milton (Paradise Lost, Book I, one of Satan’s speeches), perhaps unknowingly, but certainly wrongly.

              I don’t know that it was wrongly quoted as I thought to suggest – knowingly – that that phrase of Satan’s expressing his rebellion against the “natural order of God” was equally relevant to Adam and Eve’s similar rebellion – which I took as implied by that first stanza of Rochester’s that you quoted – against the “ignorance is bliss / noble savage” state imposed on them by God. But maybe I fell short in that attempt …

              I must say that reading Rochester makes the question of free will seem desperately boring.

              I wouldn’t call that question boring – desperately or otherwise – although I’ll concede that Rochester probably adds some welcome flesh – literal and figurative – to the bones of it – notably, the concept of choice. As Alice put it:

              Will you, won’t you,
              Will you, won’t you,
              Will you, won’t you, join the dance?

  29. Is there any context outside the realm of Christian theology where the question of free will is anything other than trivial?

    I know that Judaism and Islam toy with the concept of free will a bit, but in neither of those traditions is it central. AFAIK, it’s not a part of any major non-Abrahamic ecumenical tradition (Brahminism, Buddhism, Taoism, etc.)

    You only find “free will” as a subject for vigorous debate in philosophical traditions that are mired in Christianity.

    It’s rather obvious to me that free will is a) trivially true when viewed at an anthropic scale (e.g., if it looks exactly like free will, then it’s indistinguishable from free will), and b) it’s trivially false when viewed at any other scale, whether quantum or relativistic, or for that matter, karmic, dharmic, taoist, or moiraic [?].

    It seems to me that the correct answer to the question, “Is there free will?,” is, “I’m bored. Ask me something else.”

  30. Well, I almost agree. Except even if you rewind and the person makes a different choice.. that doesn’t imply free will necessarily. There could have been some other irreducibly complex random event in nature which built up to that event (some side effect from quantum mechanics?). I think it would be more accurate to define free will as the agent itself being immune to all a priori events when making a decision… which science does not support even slightly. But its your definition, not mine.

  31. “I’m a physical determinist”

    I thought randomness had plenty of empirical evidence supporting its existence.

  32. I find Jerry’s definition problematic for several reasons. As several have noted, it’s untestable, counterfactual, and at odds with Bell’s Theorem. As Steve Zara points out, it leaves nothing useful for free will to do, so why would anyone want free will of that sort? Moreover, I argue, it’s not what most people care about when they worry about free will (or the lack of it).

    Leaving aside the question of how people think free will works (and my guess is that they mostly don’t think about it), the reason people care about having it is because they want their deliberations to count for something. And they do count for something; there’s ample evidence that deliberate decisions tend to be better than impulsive ones, and that people who don’t believe in free will act more impulsively than those who do.

    Note that this is not an argument for believing in free will because it’s good for us. It’s a claim that it’s possible to define free will in such a way that it’s consistent both with physical causality and with our intuition that deliberation matters. Free will is the ability to deliberate and to have those deliberations affect our subsequent behavior. In my opinion, this is the only kind of free worth wanting or discussing.

    1. Free will is the ability to deliberate and to have those deliberations affect our subsequent behavior.

      But what do you mean, exactly, by “deliberate?” If this is not a deterministic or random process, how does it work? How does it differ from an algorithm or heuristic? If “deliberation,” as you are using the word, is deterministic, then does a computer deliberate? Does a computer have free will?

      1. I can’t give you a detailed account of how exactly deliberation works; I don’t think anybody can. But to say that it’s some combination of algorithmic and/or heuristic would not, I suspect, be too far off the mark. Certainly I’m not saying there’s anything magic about it.

        As for whether computers have free will, I’d say that no existing computer is anywhere near complex enough to exhibit the degree of self-reflection that I’d be willing to call free will. But this is a difference merely of degree, not of kind. I’m certainly willing to grant that Deep Blue plays chess and that its algorithm is responsible for choosing its moves. To claim that it’s all just physics, it seems to me, is to argue that Deep Blue’s algorithm is an illusion and the moves happen simply because the universe is structured that way. This does not strike me as a productive way to explain anything.

        1. I don’t see that your definition is any better than mine, because in ANY species an individual’s decisions “count for something.” If a fruit fly “decides” to lay its eggs on a peach that is too far gone rather than in the initial stages of rotting, that “counts for something”: the offspring may not survive. Does that animal, then, have free will? ANY deliberation matters in animals that can make decisions.

          I’m struck in these discussions that those who criticize my own definition, flawed though it may be, face equally severe problems with their own definitions.

          1. Does your fruit fly mentally model the outcome of various egg-laying strategies, consciously weighing the pros and cons of each, in order to arrive at its decision? Or does it just follow its instincts and lay the eggs in a place that smells right? I strongly suspect the latter, which is not what I would call “deliberation”.

            What makes our will freer than the fruit fly’s is precisely the fact that we have the ability to imagine alternative futures and that such imaginings help steer events toward the future we prefer. In the case of the fruit fly, natural selection does all the steering.

            1. Actually, I do think that may be very perceptive, that the thing which sets us apart from other animals – well, except those that can also do this (? — other apes, cetaceans, corvids, octopodes, psittacines, kittehs, etc., etc.); but certainly what sets us a part from fruit flies –  is to imagine the consequences of possible decisions and thus to select the one that offers the best outcome, rather than always to act based on instinct or conditioned behaviour.

              Perhaps it’s just the ability to imagine making decisions other than the one we do make that gives rise to the illusion (?) of free will… 


    2. Good point. I can’t think of what Jerry means when he says elsewhere that facing the fact that free will is an illusion could provoke an existential crisis, or whatever it is he’s claiming. But I infer from other things he writes that he means something like the idea that our deliberations don’t matter (for example, I read him as claiming that neurology that shows people confabulating implies that all claims of deliberation are just confabulation).

      1. I’m not sure that that’s a valid inference.

        Those kind of confabulations seem to me to be akin to dreams – our consciousness imposing some kind of “narrative” structure on certain mental activities (or external stimuli).

        “Deliberation” is just the name we give to the natural processes in the brain that iterate through various mental states representing, e.g., possible futures, before we come to a decision. So “deliberation” is a macro construct, just as heat is wrt molecular kinetics (h/t Dan L.).


  33. This implicitly assumes that the entire information state of the universe is determined, which may be a stronger claim than modern physics justifies.

    More certainly, the phrase “rewind the tape” treats time as a simple linear dimension, which is an inaccurate oversimplification (“wrong”) from the current QM framework. The weirdness of modern physics not only includes multiple futures, but multiple pasts – in particular, via the Feynman sum over histories. There’s a Dick-and-Jane discussion of some of this in Hawking and Mlodinow’s The Grand Design.

    One can maintain a degree of physical determinism with this; however, what is “determined” is no longer the particles or the present, but the probabilities of particles and probabilities of a present.

  34. The universe is a documentary film, not a feature film. That means, the people aren’t actors and there’s no script. They did whatever they wanted to do on filming, but it’s not changeable later. Thus I believe the deterministic block universe demanded by relativity can be reconciled with existence of free will.

  35. I am not sure why you would want to make your own definition when there is a perfectly good dictionary on the shelf that already contains it (to quote Richard Dawkins when asked to define ‘wisdom’)

    From the Collins English Dictionary:
    free will n 1.a. the apparent human ability to make choices that are not externally determined. b. the doctrine that such human freedom of choice is not illusory. Compare determinism. c. (as modifier): a free-will decision. 2. the ability to make a choice without coercion: he left of his own free will: I did not influence him

    So, what is wrong with these definitions, and working from them for testability and suitability? So long as the parties agree which version they are disputing when discussing free will these definitions are quite serviceable. It seems to me that your definition fits mostly under 1.a. as a purely deterministic question. Most of the arguments for free will come from definition 1.b. as to whether or not free will is an illusion. If, as in your definition, the universe up to the point of buying the ice cream is exactly the same right down to the quantum level, there is no reasonable objection to having a theory that the buyer will choose the same ice cream. By the definition, all of the previous experiences, biases and preferences will be exactly the same, so the outcome will also be the same.

    In any case I have a feeling that free will is untestable. If a person makes a conscious choice for something, the choice is based on ethical considerations (eg I will choose only if the choices are vegan, otherwise I will not make either choice 😉 ) taste, memory, cost, peer-pressure and other biases that are weighed up before coming to a conclusion. If the person wishes to demonstrate that they have free will by choosing to purchase the item they do not prefer, that is also not free will as the choice is based on the bias of showing that they have free will. Is there another option besides random choices?

  36. In that case, you can’t help saying that! As a physical determinist, I know that our inner computer can decide many things – thankfully to re-evaluate many of our beliefs (software).

  37. According to Jerry’s definition, free will happens in this circumstance:

    Radioactive decay is random; in the rewound universe, an atom decays, releases an energetic particle, which strikes his DNA, causes a thymine dimer, which causes a mutation, which causes cancer, which undeniably changes his decisions as compared to those in the universe where he doesn’t get cancer.

    According to other definitions of free will, this could still be determinism; but it does show that mental processes are theoretically unpredictable.

  38. As I recall, that is very close to the way Dennett thinks about it, too. Theists believe in something magical that eliminates the predictive power of rewinding the tape. In their eyes, anything goes when you rewind the tape because they believe in a kind of willpower, freewill, that is not bound to the natural world.

  39. I find the definition very bad. It amounts to randomness, not free will. I’d suggest picking up the definition of Sampolinski (an Orthodox Jew and physicist – who rejects free will):

    Free will is a doctrine with three parts.
    1) Human deliberative decisions are caused by our deliberations.
    2) These decisions could have been otherwise (this is Coyne’s defintion).
    3) This variability is, to the extent that it is “free will”, not due to randomness – to the extent that the decisions are “free”, they are caused by our “selves” in some way, not by random processes.

    His position is that the third condition is impossible to maintain so there is no free will. I think he’s right. But the key point here is that he is careful to clarify that randomness is not free-will; it is important to ascertain the causal mechanisms of the choice, not just its variability.

  40. To the dictionary definition in 42; the definition guarantees free will exists – since it defines free will as “the apparent human ability to make choices that are not externally determined.” and it appears to most people, without real evidence, that we can do so, therefore free will exists.

    Coyne’s definition is a good one: if we could go back to precisely the same state of the universe and do something different, then we would have free will. Provided that the difference is not the result of quantum randomness, being amplified by chaotic, non-linear dynamics. Although some commenters earlier indicated doubt of the probabilistic nature of nature, no (or at least very few) physicists would – hidden variables have been excluded by Bell’s Theorem and its experimental verification, and probability and uncertainty are central to quantum theory.

    Putting the universe back into precisely the same state over an over again, would not likely produce the same “futures” but this would not be due to some free will component on the part of living beings, but due to the fundamental quantum probabalistic nature of nature.

    True “free will” would imply the ability to cause constituents of nature to do things they would not otherwise do – it atoms and molecules to realign, to change bonding, to interact somehow despite the normal rules governing interactions we know in science. This would have measurable effects that could not otherwise be predicted, and so should be amenable to study and refutation or confirmation at least in principle, although not in practice now.

    1. Doesn’t Bell’s Theorem hit a rock/pebble precicely if there is no free will in what measurements are made? I think Bell said something about that.

      Found it, I think:
      “There is a way to escape the inference of superluminal speeds and spooky action at a distance. But it involves absolute determinism in the universe, the complete absence of free will. Suppose the world is super-deterministic, with not just inanimate nature running on behind-the-scenes clockwork, but with our behavior, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined, including the “decision” by the experimenter to carry out one set of measurements rather than another, the difficulty disappears. There is no need for a faster than light signal to tell particle A what measurement has been carried out on particle B, because the universe, including particle A, already “knows” what that measurement, and its outcome, will be.”

  41. I share the same view as Coyne; I am not aware of any evidence against physical determinism (such as a soul; and am not sufficiently versed in quantum mechanics to evaluate anything operative on that level). I will hedge my view with the fact that I am not a scientist or philosopher; I’m merely a layperson with sufficient interest in knowledge to learn things.

    I have been wondering, though, something that’s probably been discussed before (if it’s on this thread, show me the post number; it’s late and I haven’t read all comments).

    If things are physically determined, then if one were to know all states at a given point, and all processes that affect them, they would effectively have knowledge of the entire fate of the universe. This would, of course, include one’s own fate. The problem with this is apparent; if he knows everything the the states will produce in the future, then when it comes time to make a “decision”, could he not decide to go against what he predicted? But wouldn’t this itself be predicted? And then, wouldn’t this prediction be predicted? And so on.

    As far as I’ve been able to think, there’s only a few solutions to this problem. 1) free will does in fact exist, and even if all states and processes are known, a decision cannot be predicted with 100% accuracy. 2) It is physically impossible to store all the data and metadata (data doesn’t exist in a vacuum after all; it has to be physically represented somehow) and so one could not know all states and processes. 3) Another logical or physical barrier exists to knowing all states, such as the act of determining them changes them.

    Are there any other solutions that I haven’t thought of?

    1. I would think first that the universe is the only possible “memory” for its own states; that is the memory has to be of sufficient size to store the information. The human brain is clearly the merest fragment of the universe and could not do the job.

      Second, if you knew everything about the universe, its precise state at a moment then you would not be in the same state as you would be if you were ignorant (and neither would the universe) so the comparison is not valid. Clearly having the universe in a different state would likely result in a different future.

      Third, the universe according to our best theories is not Newtonian and predictably deterministic. Quantum indeterminacy rules – knowing the state of the universe “now” would not help you a second from now – it might look much the same on the surface, but particles would not be where you might have predicted since Newtonian dynamics does not work on the scale of particles, Quantum mechanics does.

  42. Randomness doesn’t get you free will, merely unpredictability. Your definition doesn’t work.

    Posit some uncaused random influence that changes things but which is external to our supposed free-will agent, and all you get is unpredictability, not free will.

    1. To clarify: It doesn’t matter whether the universe is deterministic, or whether it is not deterministic because there are things which happen in the universe without any cause, for no reason, randomly.

      In *either case* free will is not possible.

      Determinism precludes free will. Non-determinism does not, by itself, allow for free will.

  43. To add a bit more…

    Free will is incoherent. If free will exists, it must work *somehow*. HOwever, as soon as you nail down that *somehow*, you will have destroyed that free will, as the words are commonly understood.

    Talking about “free will”, as those words are commonly understood, is talking cross eyed badger spit.

    1. The contra-causal variety of the idea does indeed seem more common than the “combatibilist”, or pragmatical, variety.

      1. Wow, my spelling isn’t good today. I’m sort of in the process of losing my home, so mind wanders.

  44. An earlier commenter, suggested that perhaps free will is an emergent property. The idea of emergent properties is a hot one today but on the edges of science; it is a form of mysticism really. People have a magical view of the world and want to sound scientific and so they talk of emergent properties beyond current science.

    This is not to suggest that complex properties do not exist, but they are in fact based on the simple physical laws that govern interactions of particles ultimately. Of course there are at least 2 problems here: we don’t know the simplest physical laws yet and we do not have the computing power to apply those to complex systems in any reasonable timeframe for the most part.

    For example, the resistance in a circuit is an emergent property although Ohm’s law seems simple enough. It could be explained by considering each of the atoms and electrons in the wire and their interactions, but this is beyond us technically. The mass of a proton was recently calculated over several months of supercomputer time based on the quark-quark interactions – a proton seems simple but it is actually an emergent phenomenon. And likewise free will if it exists must be based either on known laws of nature or new laws to be found, in very complex large scale interactions, likely, thus appearing “emergent”.

    1. Emergent isn’t mystical. It is a shorthand for the sorts of things you just mentioned. Natural selection is a relatively simple emergent property for temporal process comprised of populations of replicating systems with heritability, mutation, and heritable fitness variation.

      1. How about a multilevel perspective?

        Suppose we were conditioned in early life to want some things and not others. Conditioning being a deterministic process, our will will not be free at that physiological level.

        Now, suppose we got in a situation that coerces us to do something that we do not want to do because of our conditioning. Even if we do the thing we are coerced to do, we will experience that we do not, therefore, also want it. We do it against our own will, and we experience our will as being free against the external forces coercing us. Our will is free at a sociological level.

        The external (social) forces can only coerce so much of us, but not that hardwired core of drives, will, or whatever you may call them.

        However, this freedom of will is only being experienced in situations of external coercion. If your environment is entirely benevolent, you may feel decadently unfree because of your internal drives.

        P.S.: In the former thread o free will, someone with the surname “King” said something similar, though I recast it as a multilevel persopective for the more scientifically bent discutants, here.

        1. I am not saying emergent means mystical, I am saying it seems the term is used that way, to give a “gee-whiz, this is new and different” feel to an idea, although in principal all events should be explicable in terms of the simplest of laws, if we find them, as per the examples I mentioned.

  45. Jerry, I would say that your notion of determinism is not quite right. On the quantum level, things can be predicted on average. For example, in the double-slit light diffraction experiment we can’t tell which slit a photon had gone through but we know that on average it will go through 1 slit 50% of the time and the other slit 50% of the time. If we were to rewind the clock to the moment that photon left its source, physics would still predict a 50-50 chance on either slit. Surely the photon doesn’t have free will, but as far as we know it could take a different path if it were given a second chance. So from a quantum point of view there is no hindrance to ‘free will’ in humans because outcomes of simple events are only predictable in aggregate and on a case-by-case basis appear somewhat random. Try predicting the roll of a die – you simply can’t do it, but you know beforehand what the outcomes should be after many thousands of rolls and in practice you’ll still have a small margin of error. Rolling a die demonstrates determinism, and the chemical reactions in a human which may cause one decision or another are governed by similar deterministic rules. Worse still, humans are very complex and there may be numerous competing deterministic events leading to a single decision, nor do we know of any fundamental reason why a human may not be able to impose some order and make one outcome more likely than another.

    1. There’s also the matter of the separation of predictability and determinism. The latter is to some extent the basis for the former, and doesn’t strictly require the former.

  46. As far as nobody argues that this free will can be readily distinguished from randomness, and nobody objects to their free will being indistinguishable from randomness.

  47. Yesterday I was driving along the motorway at 60mph, the car in front was going about 55mph. I decided to change lane and overtake.

    Then I wondered what led to my decision.

    1: I like to save fuel.
    2: My car cruises nicely at 60mph in sixth gear.
    3: Five years ago I had a chunk of money to spend and bought a more expensive car, one with six gears.
    5: I got that money from doing some contracting as a software developer.
    6: I got that contract because I had a very unique skill.
    7: I got that skill from researching a specific tool.
    8: I researched that tool after someone on an Internet chat channel said “What? You write code that talks directly to the database? hahaha”

    Ultimately if that person had not mocked me I would not have had the money to buy my more expensive car years later. I’d probably still have owned a five gear car which cruises nicely at 55mph, and I wouldn’t have changed lane.

    Now imagine if changing lane had saved my life 🙂

  48. I shall step into the cunning trap. If all the prior conditions are identical what mechanism exists for the ‘free will’ to interact differently? Does the ‘free will’ concept require dualism? If so then the debate can never be determined as some of the components are immaterial (yet the immaterial must interact with the material).

    On the other hand if you define ‘free will’ as the sensory illusion that immaterial choices exist in a determined world, then that leads to various areas of scientific research.

    Mind experiment: If a man goes to the same burger shop every day at the same time and has the same burger meal, but one day he decides to have the chicken meal instead, does this show free will? Or does this show that the unconscious priming by adverts, or the unconscious body signals following a heavy prior meal (etc.) cause an observable change in behaviour which needs his conscious justification?

      1. Indeed. I don’t believe in ‘free will’ as an ultimate ability, only a proximate explanation.

        I cannot decide ‘through the power of my mind’ to do something that is physically impossible for me (like levitate or swim the Atlantic without suppoort) so my ‘choices’ are already constrained. I just go a few further steps and argue that our unconscious constraints are just as confining – but we kid ourselves that we are obliged to act as if we were (mostly) rational agents.

  49. Drat! I’ve come to the debate late and all the clever things have already been said!
    As mentioned by several people already there is a problem with the definition of free will that Jerry suggested. It seems to assume that decision making is a momentary process – for instance you are standing before two doors and, with a click of the fingers take one or the other.
    I tend to see decision processes as a more convoluted and time consuming process. An analogy would be entering a city via one road and then leaving the other side by one of several alternative routes. It is not an instant process. There are many things that can happen during the act of crossing the city that might make you take one exit route or the other. It is what is going on during this journey that is akin to the decision making process and is, I think, more similar to how the public at large see the issue of ‘free will’. There is no soul involved, its still all the actions of a meat computer in your head, but it is also reliant on ongoing inputs to this meat computer and the knowledge and experience already stored within it.
    Replay all the same factors and yes, you will get the same outcome but the process itself, with the time involved and capacity for alternative outcomes dependent on changed inputs and memories, subjectively FEELS like free will.

  50. The OP definition isn’t for free will, it’s for determinism versus indeterminism, and a deterministic universe and an indeterministic one are completely indistinguishable, since you can’t ever do that rewind-the-tape experiment. The indistinguishable and the identical look very much alike 🙂 I conclude that there is no difference between a deterministic and an indeterministic universe.

    Forget free will and carry on making choices.

  51. ‘…well, that’s a free choice…’

    Free will Free choice? Explain please because this sounds kind of like ‘it’s turtles all the way down’ kind of thinking and we’re just playing with the words.

  52. Ah, and you said that belief in free will is like believing in god! This definition of free will won’t do, since it is not falsifiable, as a number of commenters have already suggested.

    But even if a person would choose the same thing, under the stated conditions, it doesn’t mean that it wasn’t freely chosen. Perhaps he only likes one flavour of ice-cream!

  53. If your definition of “free will” is identical with indeterminism, and you believe that determinism is true, then obviously you will believe that free will doesn’t exist. I, like most Compatibilists, think that this is just a terrible way of defining “free will” or as I think it’s better to say “agency”. The Incompatibilist wants to see “free will” as action outside the causal order. I don’t understand why anyone would want that or think it’s a good account of agency. I think you can give a much better account of deliberative rational agency within the causal order, rather than looking for the ghost in the machine opting for this or that.

    1. In fact this is the only context I can think of in which “free” is taken to mean “uncaused” ot “unphysical”. In all other usages, its meaning is more along the lines of “uncoerced” or “unconstrained”. Even physicists have no difficulty talking about the degrees of freedom of a mechanistic system, while recognizing that the future states of the system may be fully determined by its present state. There is no inherent conflict between “free” and “physical”, but quite a few people on these threads seem to think there is.

      1. Mm, “free from what?” might be an appropriate question when discussing free things. Free from gravity? Free from blackmail? Free from causation? Free from glue?

  54. If we view the world from the point of view of physics, then everything is meaningless. Humans don’t exist, only collections of atoms, forces and energy.

    Clearly, when we begin to utter a string of words and form an argument, we’re now within the realm of the human. If you are going to make an argument, then some basic assumptions have already taken place. Assumptions like meaning, like consciousness, like self, like freedom, and so on.

    Freedom simply has no meaning outside the human world, and that is why it is a mistake to try and talk about human concepts in only a reductionist and materialist sense.

    1. Egbert

      If you define “free will” as something beyond science, which you are doing, then there is no discussion, now you are in the realm of metaphysics, or religion. It shouldn’t be much of a surprise if most of the readers here do not see it your way – all phenomena in this universe are either real, and amenable to study or “magic”, which is where you are putting “free will”.

      1. Science works on the same assumptions. It uses language to communicate its ideas. It uses consciousness as a criteria for justified beliefs.

        So the word “freedom” definitely belongs to science, while remaining meaningless outside the human mind.

  55. Suppose that a starfish could make different decisions, randomly, after each “rewind”.
    But person make exactly the same decision, perhaps because he can think of differente possibilities and choose the better one.
    Would you say the starfish has free will but not the person? Does your definition depends on being a person? Why?

  56. I fear that the world has moved on already but in case anyone’s listening…

    I’ve seen “free will” used to explain why bad things happen and why bad people should be punished. You see, they had the freedom to accept or reject God, to do good or do bad.

    I ask myself: why don’t I steal anything when I go to a store. I don’t feel any urge to steal, when I was a kid and I stole a chocolate bar I felt terrible and never repeated it. Yet some people describe feeling a powerful compulsion to shoplift even though they have the money to buy it and sometimes even though they don’t even need or what what they’re taking. Which of us has free will in this case? I don’t think I could make any other decision and would only steal under duress so maybe I don’t have free will. But perhaps the people that feel compelled to steal don’t have free will because of some neurological disorder which drives them to these acts.

    If the capacity to decide to do good or bad is a necessary part of free will, wouldn’t the only humans which actually have free will are the sociopaths and psychopaths whose decisions are not unduly influenced by emotions?

    I think this ties back to points several others have raised. Eric MacDonald used the example of someone that only liked one flavour of ice cream. We may say they have the physical ability to order strawberry but if their preference for chocolate is so strong that they never make any other choice (unless forced) then are we really saying they don’t have free will?

  57. I’ll offer this definition of the type of free will worth valuing, and the type of thing that I think most people have in mind when they talk about free will:

    Will implies that an agent desires to bring about some change in the world (at least the agent’s condition in the world), and that the agent can enact some changes. Freedom of will means that the agent can change the world in the desired way by making informed decisions to take actions that will bring about that change, and to carry out actions sufficient to realize that change.

    That definition may need some refinement, but I think it’s pretty close to necessary and sufficient to call some behavior the result of free will. It’s basically opposed to the idea of fatalism, which I think is the real opposite of free will (I disagree that determinism is a good opposite to free will), and I think it also covers the gun-to-the-head scenario pretty well, too.

    I also think it doesn’t require any hypothesis about what might happen if we rewound the tape, or any amount of indeterminism. And I think it’s pretty clear about what type of freedom the agent is claiming to have.

    Also, I prefer to claim that people actually do have free will, but it’s not completely free. If I desired to be a billionaire CEO, I don’t believe I have the freedom to realize that desire. So my will is limited in that way. But I have meaningful freedom to achieve other goals.

      1. Huh? I don’t see that that’s an interesting observation (whether it’s true or false) for this conversation.

  58. Ah, I think Sastra #22 & medlefsen #24 have got it.

    Invoking QM and chaos theory is beside the point (odd for me to say so, since I’ve been so keen to throw them into the mix in previous threads!). So is arguing over determinism v. indeterminism.

    If the situation is exactly the same, you would come to exactly the same decision, because that decision would be based on exactly the same situation and information, exactly the same predilections and preferences, exactly the same attitudes and beliefs, exactly the same principles of action, exactly the same goals and objectives — in short, exactly the same you.

    So, where, then, is “free will”?

    It strikes me that it is somewhat akin to Stan’s – um, Loretta’s – right to have babies, even though he can’t (because he has nowhere for the foetus to gestate).

    “Free will” is the “right” to make a different decision, even though, in exactly the same situation, you can’t.



  59. Exactly. And if an acausal event (if we postulate the possibility of one)creates a differentiation within causal chains, such event could never be a willed event (as an acausal event cannot have a willer), yet the person could theoretically make a different choice due to that acausal event. This would also not be free will (as the chain of events have changed and not in accord to the will of the person).

    This is why I hold to the hard incompatibilist position rather than the hard determinist position. Free will is incompatible in both a deterministic universe as well as an indeterministic universe.

    Take care,
    ‘Trick Slattery

    1. Also, this is how I would define free will, and how I do so in the book I am writing (after much deliberation):

      ** Free will: The ability to choose between more than one viable option or action in which that choice was up to the chooser.

      This, I believe, is the ability most people think they have intuitively.

      1. Also, this is how I would define free will, and how I do so in the book I am writing (after much deliberation):

        I’ll keep an eye out for the book. BTW, on your book website I think “writtin” should be “written”.

        But I like the idea of choice and think that having one between two or more alternatives is central to the concept of free will. FWIW, Stuart Hameroff has several interesting papers on the concept of choice as central to the theory of quantum consciousness – not that I follow much of it very well. One of them on the role of consciousness in the Cambrian explosion has this observation:

        One possible advantage of consciousness for natural selection is the ability to make choices. As Margulis and Sagan (1995) observe (echoing similar, earlier thoughts by Erwin Schrödinger), “If we grant our ancestors even a tiny fraction of the free will, consciousness, and culture we humans experience, the increase in [life’s] complexity on Earth over the last several thousand million years becomes easier to explain: life is the product not only of blind physical forces, but also of selection in the sense that organisms choose. . .” (Scott, 1996).

        1. Thanks! I did not catch that I did that. I’ll have to adjust it.

          I think the key is the word “viable”. In other words, that all of the options are “viable” options.

          I’m not a big Hammeroff fan. 🙂

          Take care,

  60. The proposed definition is incoherent since already determinism (deterministic chaos) prevents “exactly” the same point in phase space in physics due to exponential divergence coupled to finite precision. Same for environmental contingency (say, exact same evolution of the spread of energy out of the system).

    And then quantum stochasticity compounds the problem by making sure the definition is useless – the outcome will never be the same even when close, regardless of the object considered.

    This goes back to my comment on the previous thread; it is only the empirical definition that makes sense AFAIK.

    I.e. the folk psychology model of observable situations where populations of organisms can and will take different pathways in a sufficiently complex manner that it doesn’t look algorithmic (automated).

    It is testable for its existence, despite not being a natural (factual) model as evidenced by the same observations that proves it is a coherent emergent model (say, by decisions-before-awareness observations of the brain).

  61. Two points, both well trammelled in the compatibilist literature;

    1: ‘Quantum indeterminacy’ would meet your definition of free will, conceivably, according to some accounts of quantum mechanics, but would not – I assume – match with what ‘the folk’ intend to be talking about when they talk about free will, so clearly your conception cannot be the folk one. (Why are you blaming that man? Well he could have chosen to do otherwise! No he couldn’t. Yes he could; had the wave-function of certain particles in his brain collapsed differently he would have acted otherwise! But what does that have to do with him? etc)

    2: There appear to be much more mundane notions of something like free will (based on things like models of decision making as the result of competing action-pathways manifested as differently weighted neuronal pathways) whereby the conditions you set for free will /doesn’t/ hold but on the other hand it seems manifestly sensible to hold people accountable for actions which are the result of this. (Why are you blaming that man? Because he could have chosen to do otherwise. No he couldn’t; his actions were the result of causal processes in his brain, the upshot of which were set in motion long before he was born. Sure, but the consequence of my actions is that he will adjust the weighting in the decision making areas of his frontal lobe such that he is less likely to engage in similar types of action in the future – what did you imagine was the point of my remonstrating with him, if not that.)

    The former mistake is something found quite often in the work of first year undergraduate philosophers and tenured staff at Notre Dame (in my experience). The latter is effectively what Dennett refers to as ‘the only free will worth wanting’. I don’t see why, given ‘free will’ has a clearly defined role within wider conceptual frameworks (tracking when ascriptions of responsibility are warranted) and there is a long history of mechanistic, compatibilist accounts of free will going back long before Dennett and Strawson back to at least Hobbes, why would would feel obliged to insist that ‘free will’ must mean some sort of mysterious aphysical causal process which, as you point out, doesn’t exist.

    My definition of free will; an action is free if it is the result of the decision making processes within an agents brain (regardless of the antecedents of those processes or quantum mechanics or if they could act otherwise subject to universal recurrence thought experiments etc), when it is you can hold the agent responsible and where necessary administer ‘corrective’ action to try to adjust their decision making processes so they are more likely to produce the sorts of action you want. Obviously, as with most folk psychology terms, we’ll probably find ‘will’ is pretty useless when you get into the cognitive neuroscience of how decision making actually takes place in the brain with a disconnect between actions resulting from more frontal, planning processes and actions resulting from more limbic, affective processes and this would (or should) influence what we say about the ‘responsibility’ or the appropriate response we should take to peoples actions. But don’t get hepped up on how to reconcile the way medieval theologians thought about free will with modern physics; the one thing we know about medieval theologians is that they were wrong about practically everything.

    1. A nice theory Duncan, unfortunately your description would also include computers:

      an action is free if it is the result of the decision making processes within an agents brain
      I have no interest in comparing a person or the brain to a computer, but a CPU would fulfill this function of a brain in this part.

      when it is you can hold the agent responsible and where necessary administer ‘corrective’ action to try to adjust their decision making processes so they are more likely to produce the sorts of action you want
      Computers can be trained, either by altering their programming or if the program accepts dynamic updates.

      So, do current computers have free will? A car’s fuel injection system can dynamically learn your driving style and alter its behavior accordingly, is that free will? If not, which part of your definition does it not follow?

  62. I don’t see why a person should be able to take decisions at random (i.e., to behave irrationally) in order to demonstrate the existence of free will. Free will is usually predicated for choices that remain the same across a range of similar situations, especially when they go against conventions, obligations, best interest, contrary advice.

    The idea of free will posits that the ultimate source of a decision is the subject making the decision. This is compatible with a deterministic world where the outcome of a decision is pre-determined, since we are different individuals with different life histories.

    So I don’t think you can make a metaphysical case against free will. A better way to undermine the idea of free will is to show that you can get different decisions and justifications from the same individuals by altering something at the physical level. So, when you rewind the tape you should be able to make a controlled change in the brain or the environment, i.e. a change that reliably produces an alteration in the decision pattern of similar subjects. If the person does make a different decision but cannot justify it (“I don’t know why I did it, I was in an altered state of mind”) the experiment fails. If however the person justifies the decision as a deliberation of their own you have a stronger case against free will.

  63. What confounds me is why people want to believe in free will. There certainly is no evidence nor logical reasoning for it; it always appears as a deus ex machina. It’s enough to have the functional illusion of free will—without it we could not act—but when it comes to casting blame and looking for responsibility, we have to be very careful. There is nothing that says that thinking isn’t instinctual. Just because you think you thought something, doesn’t mean you actually thought it.

    A note about quantum mechanics: quantum mechanics doesn’t allow for free will, it only allows for unpredictability.

Leave a Reply