Another paper claiming (but failing) to give evidence for libertarian free will

June 14, 2020 • 9:00 am

UPDATE: Reader Coel pointed out in the comments—this had escaped my notice—that the author of this piece, George Ellis, won the Templeton prize in 2004 for efforts in harmonizing science and religion. This may be relevant to the article below.

Part of his citation says this:

Beyond ethics, Ellis contends that there are many areas that cannot be accounted for by physics. “Even hard-headed physicists have to acknowledge a number of different kinds of existence” beyond the basics of atoms, molecules and chemicals, he said in his prepared remarks. Directly challenging the notion that the powers of science are limitless, Ellis noted the inability of even the most advanced physics to fully explain factors that shape the physical world, including human thoughts, emotions and social constructions such as the laws of chess.


A lot of people sent me this link to an article in Aeon by physicist George Ellis, with some of them telling me that his piece deals the death blow to determinism and pumps life into the idea of free will. It doesn’t—not by a long shot. And anyone who has such a take doesn’t understand either determinism or Ellis’s arguments. For once you understand them, you see that they’re invalidated by a big fallacy.

First, here’s Ellis’s bona fides from the site (his Wikipedia bio is here):

George Ellis is the Emeritus Distinguished Professor of Complex Systems in the Department of Mathematics and Applied Mathematics at the University of Cape Town in South Africa. He co-authored The Large Scale Structure of Space-Time (1973) with Stephen Hawking.

First, let us be clear that although Ellis doesn’t define what he means by “free will”, it’s clear that he’s talking about libertarian, contracausal, “you-could-have-chosen-otherwise” free will”, not the compatiblist free will that accepts physical determinism. No, Ellis thinks that determinism is simply wrong.

And by “determinism”, I don’t mean that “with perfect knowledge of the present, or of the moment after the Big Bang, we could predict what would happen with 100% accuracy”. I don’t believe that, as there are some fundamentally non-deterministic processes—to our best knowledge at present, quantum mechanics comprises some of these fundamental unpredictabilities—that could, at a given time, act to create different futures. (Evolution might be one of these if the fuel for the process—mutation—is affected by quantum unpredictability. In that case, “rerunning the tape of life” from a given point could yield different outcomes.)

By “determinism”, I mean that the future is determined only by the laws of physics, not by some nonphysical “will” or “agency” that we can exercise. (Throughout his piece, Ellis conflates physical determinism with predictability, an odd stance for a physicist.)

The article, at six printed pages in Word in 9-point type, is very long, and larded with descriptions of the Schrödinger equation, ion channels, gene regulation, and biochemistry, all fancy science that could bamboozle the reader into thinking that Ellis’s view is backed by science. But it isn’t. In fact, his argument can be stated very simply, and I’ll try to paraphrase it:

Free will acts by human psychology changing the constraints that act on our brain molecules.  Although physical processes are constrained by physical laws, psychology can override and change these constraints. Therefore, our minds, or our psychology, can somehow “reach down” to affect the molecular processes occurring in our brains and bodies. And these psychological processes, which are apparently themselves physically unconstrained, constitute free will.

Now that is my paraphrase, but I’ll support it with quotes from Ellis (any bolding is mine):

In the case of the biomolecules that underlie the existence of life, it’s the shape of the molecule that acts as a constraint on what happens. These molecules are quite flexible, bending around joints rather like hinges. The distances between the atomic nuclei in the molecules determine what bending is possible. Any particular such molecular ‘conformation’ (a specific state of folding) constrains the motions of ions and electrons at the underlying physical level. This can happen in a time-dependent fashion, according to biological needs. In this way, biology can reach down to shape physical outcomes. It changes constraints in the applicable Schrödinger equation.

. . . So what determines which messages are conveyed to your synapses by signalling molecules? They are signals determined by thinking processes that can’t be described at any lower level because they involve concepts, cognition and emotions in an essential way. Psychological experiences drive what happens. Your thoughts and feelings reach ‘down’ to shape lower-level processes in the brain by altering the constraints on ion and electron flows in a way that changes with time.

The phrase that our thinking “reaches down” to affect our molecules recurs often in this essay, implying a psychological “invisible hand”—”thinking processes”—that is the crux of Ellis’s argument. But wait! There’s more!

How does any of this happen? As the Austrian-American doctor Eric Kandel explained in his Nobel Prize Lecture from 2000, the process of learning at the mental level leads to changed patterns of gene expression, and so specific proteins being produced, which alter the strengths of neural connections at synapses. This changes the strength of connections between neurons, thereby storing memories.

Such learning is a psychological happening. You might remember your pleasure on eating a delicious meal, the details of a Yo-Yo Ma rendition of a Bach sonata, or the painful memory of the car crash. Once again, these are irreducible psychological events: they can’t be described at any lower level. They reach down to alter neuronal connections over time. These changes can’t be predicted on the basis of the initial state of the neural connections (your neurons did not know that the car crash was about to happen) – but, afterwards, they constrain electron flows differently, because connections have changed. Learning changes structure at the macro scale (we have a ‘plastic brain’), which reaches down to alter micro connections and the details of electron flows at the bottom.

(Note the allusion to classical music; people love to show off in this way. They never talk about the moving saxophone solo of Lester Young or the poignancy of a Frank Sinatra song; it’s always Bach or Beethoven, isn’t it?)

If it’s not too early in the morning, you’ve already spotted the flaw in Ellis’s argument. First, he’s conflating predictability with physical determinism, and he’s also arguing, falsely, that a “change in constraint” means “libertarian free will”. Most important, he’s pretending that psychological phenomena are not physical.

I’ll cite just two more bits:

What these instances show is that psychological understandings reach down to shape the motions of ions and electrons by altering constraints at the physics-level over time. That is, mental states change the shape of proteins because the brain has real logical powers. This downward causation trumps the power of initial conditions. Logical implications determine the outcomes at the macro level in our thoughts, and at the micro level in terms of flows of electrons and ions.

. . . Genuine mental functioning and the ability to make decisions in a rational way is a far more persuasive explanation of how books get written. That this is possible is due to the extraordinary hierarchical structure of our brain and its functioning. And that functioning is enabled by downward causation from the psychological to the physical levels, with outcomes at the physics level determined by constraints that change over time. No violation of physical laws need occur.

That should be enough to show that Ellis’s argument is bogus. Why? Because it argues that the “psychological level” (“mental thoughts”) is somehow different from the “physical level”: that our experiences, our cogitation, and our interactions with others and with external events, are different from physical processes that shape our actions. After all, they are “top down” phenomena.

But they’re not. We can be influenced by internal and external events to so that our behaviors and actions would differ from how they’d be if those events were different.  The example I often use is kicking a dog, though I would never do that. If a dog is friendly, but you repeatedly kick it when it approaches you, it’s not going to be nearly as friendly as it would have been had you petted it instead. Your actions have rewired the dog’s brain in such a way that it regards you as an object of fear rather than affection. It’s as simple as that. And in a similar way, our experiences reprogram our brains—our onboard meat computers—in a physical way that affects our behavior, but that we don’t yet understand. There is no psychology independent of physics that can “reach down” to affect our molecules, because, in the end, our psychology is based on molecules, even if we can’t yet (or ever) predict our future behavior with a deep knowledge of our brains.

Ellis’s Big Flaw: to claim that there is an Invisible NonPhysical Psychological Hand that “reaches down” and “changes our constraints”. Rather, there are external stimuli and inputs into our brains that alter their workings. We don’t need the palaver about “constraints” and “reaching down”, which is simply literary prestidigitation that is obscurantist.

Remember, Ellisis not talking about compatibilist free will here. He’s talking about contracausal free will. Ellis rests his case on a human psychology that is independent of physics. And that is pure dualism.

In the end, perhaps his mask slips a little, for in his ending he appears to find determinism distasteful because it absolves us of the ability to make “genuine choices” or to be “accountable for our behavior”. (I’ve argued that we are still responsible for our behavior in the sense that we are the entities who behave, but that we are not “morally responsible” because we could not have acted otherwise. But this is irrelevant to his argument.)

Ellis’s ending:

Physics has made huge strides since the days of Laplace; indeed, it would be completely unrecognisable to him. Yet there are still physicists today who confidently proclaim that we can’t have free will because physics determines everything, including brain functioning – entirely ignoring the complex context and the power of constraints.

If you seriously believe that fundamental forces leave no space for free will, then it’s impossible for us to genuinely make choices as moral beings. We wouldn’t be accountable in any meaningful way for our reactions to global climate change, child trafficking or viral pandemics. The underlying physics would in reality be governing our behaviour, and responsibility wouldn’t enter into the picture.

Well, he’s wrong about “responsibility” if you conceive of it as I do—in a way that still allows for and, indeed, asks for approbation and disapprobation,  punishment and reward. After all, those are external stimuli that can alter our brains.

But one gets the sense here that Ellis’s misguided screed in favor of free will is motivated in part by his distaste for determinism and his need for all of us to be “responsible”.  But, as a scientist, Ellis should realize that wanting something to be true has no effect on whether it is true.

101 thoughts on “Another paper claiming (but failing) to give evidence for libertarian free will

  1. This says it all and refutes his arguments:

    “ (your neurons did not know that the car crash was about to happen) – but, afterwards, they constrain electron flows differently, because connections have changed. Learning changes structure at the macro scale (we have a ‘plastic brain’), which reaches down to alter micro connections and the details of electron flows at the bottom.”

    Physical things happen that cause brain changes, which is learning, so now you make a different decision.

    Determinism confirmed.

      1. The transparently fallacious arguments made by so many otherwise intelligent defenders of libertarian free will leads me to conclude provisionally that they are operating under powerful cognitive biases…which they can’t choose not to have, I suppose, but they CAN learn to recognize and work around them.

        If only we could find a way to have all such defenders put in some equivalent of Skinner boxes and have you and Sam Harris apply negative reinforcement whenever they indulge in fallacious reasoning…nothing TOO noxious, of course. Perhaps a small electric shock. As for responding when they get it right, surely logical consistency is its own reward.

      2. It’s almost like they don’t think determinism includes (environmental) inputs and outputs.
        “You changed your behavior because of some new experienc!”
        “I can build a simple circuit that does that.”

    1. Your state of mind depends on whether you witnessed the car accident, read about it, heard it or got injured by it, and goes to show that the ‘psychology that reached down’ is but a receptacle of sensory input, like action potential firings for example.

    2. Yep. I’m not sure how he goes from ‘feedback loops are real’ to ‘therefore free will.’

      We can easily think about constructing an Ellis machine that operates at his ‘different psychological level’ and illustrates how no free will emerges from the sort of process he’s talking about. You take a scanning electron microscope and hook it up to room a temperature gauge, with some program that tells it to ramp up the power in response to a temperature change. Voila, you now have a “high level” collective phenomena analogous to his “psychology” (room temperature) feeding back to atomic level processes, at which scale it imparts energy, theoretically altering the room temperature.

  2. Adding to the bio of George Ellis is: “recipient of the Templeton Prize for his work at the intersection of science and religion”.

    Also pertinent to his approach and how he sees evidence is this quote:

    “The evidence is, firstly, in seeing that all these other religious traditions have come to the same conclusion. But the deep evidence is the same in all faith. It is by beginning to comprehend the deep nature of this transforming current. If you see the deepness and the quality of transformation that is possible through this, and if you really asked me to pursue it, then it is the life of Christ, which is the example. Giving up life in order that those that persecuted Him would have freedom. That is what it is about. And in the end, it is self-authenticating. There is actually no other way of saying it. It is just something you either see or you don’t see. There is no proof. It’s something you recognize or you don’t recognize.”

    1. I thought there must be a religious connection somewhere. Apparently it is at the intersection of science and religion, though I am not sure where that is – two blocks north of reality, maybe?

      1. Wikipedia describes Ellis as an active Qauker and, if his article and its placement in Aeon is any indication, perhaps an evangelical such.

        I ignore him, since he never says anything remotely interesting.

    2. The evidence is, firstly, in seeing that all these other religious traditions have come to the same conclusion.

      But, AIUI, they haven’t. AIUI very traditional Judaism believes in a bodily resurrection, not any sort of spiritual realm. And when you compare Hinduism’s reincarnation to Christianity’s heaven to Shintoism’s ancestor spirits, you get very different outcomes for the person after death.

  3. Yeah, I don’t see how this paper is supposed to be a big deal. The biologist in me was wanting to yell back at my computer monitor as I read the quotes. ‘no, our thoughts and feelings don’t reach down to bend molecules and re-direct ions!’ and NO, we don’t choose what we learn from experience!’ Ever hear about PTSD? Addiction?

    1. Actually, my thought was “of course, our thoughts and feelings re-direct ions, just as the rotation of a bicycle wheel brings microscopic crystals of metal close to the ground, then a few feet into the air, then down again in a rotary pattern.” But that doesn’t disprove determinism in the slightest.

  4. Indeed the give away at the end tells the whole story. Another person who fears what will happen to our morality if we don’t believe in free will, just like those who fear what will happen to our morality if we stop believing in God.

    Even though this one is not a compatibilist, he is motivated by the same fear compatibilists seem to be motivated by. “People will misbehave if they don’t believe in free will?”

    The weirdest part for me is they don’t seem to notice that they are worrying that without belief in free will people will then CHOOSE to do bad things. This guy is off his rocker but compatibilists are equally off their rocker, just slightly more in line with science.

    1. I resent your implication that us Compatibilists are mostly motivated by fear of people misbehaving. I’m certainly not. In my opinion, no amount of talking about determinism will change anything having to do with human behavior. People will continue to act as agents while they draw breath and that’s what we call “free will”.

      1. Not all compatibilists are, certainly. But enough of them have used the Little People argument (Dan Dennett is one), that I take seriously that this fear of social anarchy is behind some people’s acceptance of compatibilism.

      2. “People will continue to act as agents while they draw breath and that’s what we call “free will”.”

        Why call that “free will?” Why not just say “People will continue to act as agents while they draw breath.” I don’t know any incompatibilists who would object to that phraseology. Why do you need insist on calling that “free will.” I mean if you’re going for the Templeton prize I see your motivation to hold onto the term “free will” but otherwise, why not just say “people act as agents?”

        In other news Biden is no the betting favourite! Good news for the world, bad news for my wallet! See you in November, Paul. Still hoping I have to pay you off!

        1. I agree with what Coel said. I could go further but many readers are tired of hearing me go off on free will so I’ll leave it there.

          Do we have an election bet? I don’t remember it but perhaps we do. Do you have a link?

          1. I thought we did but I might be confusing you with another person. I made an election bet with one of the regular readers here.

            I responded to Coel’s comment as well.

    2. ” the same fear compatibilists seem to be motivated by. “People will misbehave if they don’t believe in free will?””

      People will indeed misbehave if they are not held responsible for their acts. If there were no punishment for robbing banks, more people would rob banks.

      “… people will then CHOOSE to do bad things.”

      People make “choices” all the time, where a choice is a computation about what to do, based on all the input information.

      If I offer you a choice of tea or coffee, I’m not suggesting that you do something impossible under the laws of physics, I’m simply asking you to tell me what you want.

      1. “People will indeed misbehave if they are not held responsible for their acts.”

        Who said anything about not holding people responsible for their acts? Holding them responsible will deter their actions so we should definitely do that. But people believing that they have no “free will” will not cause them to do bad things. I know this because I don’t believe I have free will and that belief does not cause me to act badly. Nor does that belief make Jerry act badly or anyone else act badly. Hard incompatibilists are probably the most peaceful people on the planet. They are practically Jains. Have you ever seen Greg Caruso speak? You think that guy is a threat to society because he doesn’t believe he has free will?

        “People make “choices” all the time”

        Yes they do but they do not make those choices with something called “free will”. Compatibilism is specifically about the term “free will” not the term “choices.”

        A sorting machine makes choices, but not with its “free will.” It was programmed to make those choices by humans who were, in turn, programmed to make their choices by natural selection, biology, and the laws of physics.

        1. But compatibility are not maintaining that we need people to believe in the sort of “free will” that we don’t have, they only maintain that we need the sort of “free will” in “Did you sign this contract of your won free will or because of a gun to your head?”.

          If you agree that there is a difference between the “responsibility for their acts” of a bank teller handing over money under gunpoint, and a criminal accomplice handing over the same sum of money, the you agree that it matters whether they “wanted” to do it and were acting “freely” or not.

          1. Both the bank teller and the accomplice acted the way their respective genes and environments caused them to act. As a society we can permit the bank teller’s actions but we can not permit the accomplices actions so we jail the accomplice and let the bank teller carry on. No need to the the term “free will.” It’s silly and pointless to argue the need for that specific term. Compatibilism is silly and pointless.

          2. I should add I agree that it matters if they “wanted” to or not. But the concept of “acting freely” is incoherent, unnecessary, and pointless. Compatibilism isn’t about “wanting” it’s about the incoherent concept of “acting freely.” That is why compatibilism is itself incoherent.

            1. “That is why compatibilism is itself incoherent.”

              My concept of free will is that it’s an attribute applied to the decision-making process that goes on in all our brains. When asked in the morning whether I want coffee or tea, my brain processes the question and makes a decision. If I make the decision with a sufficiently healthy brain and without coercion, then I can be said to have made it of my own free will. I will even acknowledge that my decision was affected by the state of my brain at the time, right down to its quantum states. The state of the universe, for that matter. I don’t see how that is incoherent.

              You are now going to tell me, “But you couldn’t have made that decision otherwise.” I maintain that I could have if I’d wanted to. By this I mean that if my brain state had been different, I might have made a different choice. No incoherence there.

              The important point is that the Determinist/Incompatibilist maintains that the “could have chosen otherwise” game can only be played by considering a single instant in time. This implies that there could have been only one possible state of my brain so only one possible decision could be made. If the single-instant constraint is removed, then the state of the universe (and my brain) would be different at the two instants and I could have made the decision differently.

              The single-instant version is not how most people consider such questions. Instead, they are thinking about an imaginary world in which their brain was in a different state or that the decisions being compared were made at different times. “If I had a rough night, I would have wanted coffee. Otherwise I usually have tea.”

              Most people would not consider the single-instant version of the “could have chosen otherwise” question as coherent. Of course I couldn’t have chosen otherwise if you require that absolutely everything be held constant, including the time at which my decision was made. To think otherwise would require the introduction of some sort of external influence on my decision which would violate the premise that everything is being held constant. It’s an incoherent version of the question.

              1. “If I make the decision with a sufficiently healthy brain and without coercion, then I can be said to have made it of my own free will.”

                You can surely phrase it that way if you like but it’s not useful in any way. The term “free will” does no work in helping us understand the situation. In fact it confuses it. You are obviously going to order the beverage that you prefer, and you did not choose to prefer it, so calling that an act of free will isn’t a useful description. And if someone points a gun at your head and forces you to drink the one you don’t prefer you will also do what you want in that situation which is probably to drink the beverage that the person with the gun is telling you to drink because you WANT to live and you did not “choose” to want to live! You just do. But others might attack the person with the gun because they have different genes and they grew up in a different environment than you. In any case, everyone does what they want to do in every situation and they don’t choose what they want they just know they want it. This is a more sensible way to describe reality than people are acting on something called “free will” unless they are “coerced” by another human. That is a confusing and superfluous term.

                As for the rest of your reply on “could not have done otherwise?” I am a free will skeptic but not a determinist so nothing you said there applies to anything I am saying. I don’t need any determinist argument to show the incoherence of the concept of “free will.” I don’t even need biology or any other scientific level explanation. I can demonstrate the incoherence of free will right here in the manifest image.

              2. It sounds like your argument here just boils down to not finding “free will” useful. I don’t see any hint of “incoherence”. Let me remind you that “free will” is often referred to by contracts and in courts of law so at least they find it useful.

                Ok, so you aren’t a determinist. I defer to others on the subject as I am no physicist. My view of choice, agency, and free will is independent of determinism. The concept of a brain making a choice works fine without reference to determinism. That’s not to say that the laws of physics don’t hold sway when it comes to the operation of our brains. It’s just that making choices is at a higher level of description.

                Perhaps the version of free will that you find incoherent is the one that refers to some sort of magic external influence on how we make decisions. Some people certainly think that way. It’s what they refer to when they claim that computers can never be conscious because they lack some undefinable “life spark”, of which “free will” is a consequence. I find that point of view incoherent. Is that what you’re talking about?

                Please do me a favor. Re-read what you wrote and tell me where you say how “free will” is incoherent, not just a term you prefer not to use. What about the use of the phrase by the legal system? Do you consider that incoherent?

              3. “tell me where you say how “free will” is incoherent, not just a term you prefer not to use.”

                I’ll do it right here. You didn’t choose to be a creature that goes after what it desires. And you didn’t choose to desire tea over coffee. You just “are” a creature that goes after what it desires, and you “are” a creature that prefers tea over coffee. You chose neither of these things nor did you choose the options the waiter is offering you. Calling that “free will” is manifestly incoherent. It’s a terrible misleading description of reality. A misnomer. I don’t just prefer not to use it I actively campaign against its use because(see below).

                “What about the use of the phrase by the legal system? Do you consider that incoherent?”

                Yes!!!! I don’t have a lot of respect for our archaic legal system. Its belief in, and use of the misnomer “free will” is its greatest failing.

              4. I’m an evidence guy. I need evidence to believe in something. There is no evidence for free will at any level so I don’t believe in it. There is some evidence for determinism but also some evidence for indeterminism so I remain agnostic on that front. But even if indeterminism is true that is not evidence for free will so until I see evidence for free will I will remain a free will skeptic and a determinism skeptic.

              5. The problem is that you have your own definition of free will that is different than mine. In my world, and that of our legal system, free will doesn’t mean what you seem to think it means. For our concept of free will, there’s really no question of “believing in it”. It just refers to people making choices without coercion or mental illness. Is there a question of belief there? I don’t see one.

                Perhaps you don’t believe people make choices. If so, I think that you are choosing a definition of “choice” that goes too far. I think of making a choice as analogous to a thermostat changing state. Sure, our brain is way more complicated but the principle is the same. Sure, the thermostat’s change of state had external causes but so what? Everything does. It’s simply a matter of terminology. Making a choice refers to the change of state and the processing within the mechanism (agent, brain, or thermostat) that led to it.

                I think the version of free will that bothers you is the woo-ish one where some yet-to-be–discovered external force is involved. Some believe in that but I don’t and, I’m guessing, neither do you. Incompatibilists seem to find it convenient to pretend that all who oppose their point of view, share this broken definition of free will and that’s just not the case. As Dennett says (paraphrasing), “We have all the free will we need.” I’m also a fan of Sean Carroll’s “Free Will Is as Real as Baseball”.

              6. The legal system also doesn’t need this confusing misnomer “free will.” You can just talk about people being coerced or not without needing to refer to a mythical state of total non-coercion called “free will.” It’s useless. It does zero work. It confuses the issue.

                Also there is no such thing as a “perfectly healthy brain.” A brain tumour can cause your “choices” and so does every other brain state. You’d not be able to draw a demarkation point between the healthy brain and the unhealthy brain, whatever that means. Whatever state your brain is in it causes your “choices” and so there’s really no distinction between coerced and not coerced. Every action every human takes is coerced by their genes and their environment. A person with a gun to your head is just part of the environment just like a cultural norm or a rule for life given to you by a parent. All coercion. There is no coherence to the idea of a non-coerced decision or a perfectly healthy brain. This is the trouble with the legal definition and your definition.

                I’m not working with any specific definition. I’ll accept any definition that makes sense but yours and the legal definition suffer from not understanding that there is no such brain state as “completely un-coerced” nor is there any such thing as a “perfectly healthy brain.” You make no more sense than if you were stating that a thermostat acts with free will. It is programmed just like you are to react to external stimuli in a specific way that neither you nor the thermostat have any control over. To call that “free will” is absurd.

                Answer me these questions. Do you choose to laugh with your free will? Do you choose what you laugh at and what you don’t laugh at? How about crying? Do you choose to cry with your free will? Do you choose to be sad with your free will? Do you choose to be happy with your free will? Do you choose to breathe or eat food with your free will? The correct answer to all of those is no you don’t. So what do you do with your free will? Choose tea and shirt colors? No not those either. To state that you have free will is absurd. You don’t choose what makes you laugh or what makes you cry or what makes you happy or sad. What makes you think you choose the other things you do which are heavily influenced by what MAKES you laugh and cry.

              7. “Also there is no such thing as a “perfectly healthy brain.” ”

                Agreed. Not sure why you put quotes around this as I never referred to a perfectly healthy brain. I’m sure none of us have them.

                It sounds like you are someone that has trouble dealing with anything that is not black and white. Brain health and coercion are both things that take judgement to decide. Sometimes it isn’t easy to determine whether someone is coerced enough, or mentally ill enough, to escape conviction but that doesn’t mean the concepts are worthless. This is why we have a legal system, juries, doctors, etc.

                Your last paragraph is gibberish and makes no argument at all. It’s like you were beaten as a child with a bat that said “free will” on it. We’re done.

      2. I agree. But that definition of free-will could never be the foundation for the type of moral guilt which is at the base of religions like the christian faith. Without that foundation the whole theology of sin, original sin, redemption, etc. goes out of the window. That explains why a theist like prof. Ellis goes to such lengths trying to establish that foundation.

  5. I’ve only ever read 2 good articles on Aeon, and one of them was by Sean Carroll. Aeon does not have a good track record when itcomes to articles. Wordy, but content-less.

    And for the record, most people working in the foundations of quantum mechanics think that quantum mechanics is deterministic, simply because any interpretation that allows for indeterminism invokes wavefunction collapse, which violates causality and unitarity, and violation of the latter would lead to probabilities that do not sum to 1, which is simply incoherent.


    1. The purpose of Aeon is to have a bad, religious, track record as far as I know, since they were founded by religionists and is funded by such:

      “Cyberculture and the Integration of Science and Religion
      June 2, 2014

      I decided to ground my reflection on the frontiers of digital work on religion in a discussion of two publications that have emerged over the past year that both seek to bring science writing to new publics.1 I chose these two publications not because I think these publications are “the most exciting and productive” examples of such work – they may or may not be – but because they appear to make interesting case studies of work being done to bring together digital media and religion. The two publications, Aeon and Nautilus, are, as I mentioned, science publications, but both are set up in a way that ensures religion is among their chief areas of interest.”

      [ ]

      “Building on the tremendous success of our earlier publishing Project funded by the Templeton Religion Trust (TRT), Aeon proposes a four-part publishing Program, maximising the overlap between Templeton’s Funding Areas, Aeon’s editorial themes, and our particular, highly developed expertise in working with researchers to communicate their findings and theories to a wide public audience.”

      [ ]

    2. On the physics, wavefunction collapse such as in “shut up and calculate” [minimal theory] works. Bell test experiments show that locality is violated in order to have no hidden variables, so collapse is no biggie in that sense.

      [Which makes sense, relativity constrains locality and light cone causality, quantum mechanics works on non-local entanglement – the quantum field theory yields to both in a compromise. Determinism and contingency of statistical outcome, not unlike evolution (adaptation on variation vs sources of variation).]

      1. Bell test experiments show that one cannot have local hidden variables. It says nothing about non-hidden-variable theories. They also implicitly assume that something more is going on than just the unitary evolution of the quantum state. Unitarity is what I was addressing, not locality.


  6. Magic in – magic out. Psychoneural dualism doesn’t solve the problem, but adds another huge wack of them.

    Does this physicist not know about conservation laws?

      1. And his interest is to break laws or at least find out where they break, see his Wikipedia page on fringe cosmologies. Not fringe in that they shouldn’t be considered – e.g. the Planck collaboration tests them – but fringe in that the prior and posterior likelihoods are low.

  7. To conclude from the unpredictability that there is a free will, – Ellis is not the first and will not be the last to make this fundamental mistake.
    A german neurobiologist conducted a fruit fly experiment, the result of which (the behaviour of fruit flies in a stimulus-free environment is unpredictable) is now proof for him that free will exists, on a rudimentary level even in insects.


    “This comparatively recent evidence indicates that one common ability of most if not all brains is to choose among different behavioural options even in the absence of differences in the environment and perform genuinely novel acts. Therefore, it seems a reasonable effort for any neurobiologist to join and support a rather illustrious list of scholars who are trying to wrestle the term ‘free will’ from its metaphysical ancestry. The goal is to arrive at a scientific concept of free will, starting from these recently discovered processes….”

    The author does not seem to take into account that if a fruit fly shows a change in its behaviour at a certain point in time without this being due to a change in external stimuli, this must then be based on an internal stimulus, even if this is a kind of algorithm that determines that after a certain period of time, a different behaviour must begin.

  8. Now we know the reason why Sarah Palin found favoritism among some middle age, conservative white males.

  9. I started reading Ellis’s paper, but I stopped after encountering lines like this:

    “Once again, these are irreducible psychological events: they can’t be described at any lower level.”

    That right there tells me this paper is not worth reading. Anything one might call a “psychological event” is at a much higher level than fundamental physics and can most certainly be described at lower levels, though we may currently lack the necessary knowledge to do so.

  10. “they are operating under powerful cognitive biases”

    Beautifully stated. It probably applies to many things. To paraphrase TS Eliot: Half the harm in the world is done by people operating under powerful cognitive biases. They don’t mean to do harm; but the harm does not interest them.

  11. Jerry sez: “There is no psychology independent of physics that can ‘reach down’ to affect our molecules, because, in the end, our psychology is based on molecules, even if we can’t yet (or ever) predict our future behavior with a deep knowledge of our brains.”

    Psychology obviously needs a physical instantiation with which it’s causally consistent, but I’m not sure it’s been formally proven (perhaps it has) that the physical laws of that instantiation at the atomic/molecular level are sufficient in principle to predict, say, the position of a molecule in my left hand sometime in the future. We of course obey the laws of physics, but is there any causal autonomy of higher-level explanations when it comes to intentional systems like us, such that we’d have to invoke them to predict behavior even if we were Laplace’s demon? I of course agree we don’t have contra-causal free will no matter what level or levels of explanation are needed.

    1. Physical laws are certainly sufficient in principle to explain the position of a molecule in your left hand. Prediction is another matter because prediction requires that you know and measure all the variables and, as Robert Sapolsky as has pointed out, with highly multifactorial processes like behavior, it is impractical to know and measure them all.

      1. “Physical laws are certainly sufficient in principle to explain the position of a molecule in your left hand.”

        I was wondering if this claim has been formally proven. By physical laws I mean those of physics, e.g., Sean Carroll’s one equation of the physics of the world of everyday experience (combining QM, spacetime, gravity, other forces, matter, and Higgs), described at

        1. Your question already uses a high-level description when talking of a molecule in your left hand. Atoms and molecules are “merely” (<– that's a dangerous word) factorizations of quantum fields that are highly useful simplifications, with lots of explanatory power, in many circumstances.

          As to whether Laplace's demon would need to know psychology to predict our behavior – the counterfactual cannot be evaluated, since Laplace's demon is physically impossible. The information in your future light-cone cannot be predicted without access to your entire past light-cone. (Because of the ubiquity of chaos.) Which means it cannot be predicted by someone at your location until the moment you act, never before. And it cannot be predicted by someone at another location until later.

          1. So I take it that your answer to my question is no: it hasn’t been proved that the laws of physics alone are sufficient to predict my behavior going forward. But that’s because of inaccessible/insufficient information you say, not because downward causation exists. So it seems the question is still open about downward causation: *in principle*, do the laws of physics need to be supplemented by higher level laws in order to predict and explain human behavior?

            1. From my impression, the philosophy of science community seems divided on the question you seem to be trying to get at. To give my point of view: calling something “human behavior” already imports high-level concepts that cannot realistically be *defined* in purely low-level terms. Which may be not so much an answer to your question, as a mooting of it.

            2. Renormalization is used in quantum field theory to – I think (not having studied it) – build effective theories at appropriate energy scales based on observations setting the parameter values. The laws on each scale are linked, and the laws flows naturally from lower to higher energies [see excerpts from a Wikipedia article below].

              If laws changes with scale, the lightcone causality of signals breaks because the signals change. E.g. a change in chemical potential do not describe how an arm moves in space. In that perspective Ellis causation model seems unsupportable.

              On the other hand, if it is the laws themselves that are linked, what do we mean with “lower and higher level laws”? The Feynman diagrams that renormalization theory applies to [see the Wikipedia article] are perturbative based on vacuum physics, so starts at zero energy density (vacuum) and end with the highest energy density (Planck energy density). Therefore chemical potentials of cell machinery is “lower level” and muscle output is “higher level” – laws “flow” from cell environments out to the phenotype of an animal (say). Even recast to mao to modern physics, Ellis seems to be wrong.

              “When describing space-time as a continuum, certain statistical and quantum mechanical constructions are not well-defined. To define them, or make them unambiguous, a continuum limit must carefully remove “construction scaffolding” of lattices at various scales.”

              “All scales are linked in a broadly systematic way, and the actual physics pertinent to each is extracted with the suitable specific computational techniques appropriate for each. Wilson clarified which variables of a system are crucial and which are redundant.”

              “Changes in renormalization scale will simply affect how much of a result comes from Feynman diagrams without loops, and how much comes from the remaining finite parts of loop diagrams. One can exploit this fact to calculate the effective variation of physical constants with changes in scale. This variation is encoded by beta-functions, and the general theory of this kind of scale-dependence is known as the renormalization group.

              Colloquially, particle physicists often speak of certain physical “constants” as varying with the energy of interaction, though in fact, it is the renormalization scale that is the independent quantity. This running does, however, provide a convenient means of describing changes in the behavior of a field theory under changes in the energies involved in an interaction. For example, since the coupling in quantum chromodynamics becomes small at large energy scales, the theory behaves more like a free theory as the energy exchanged in an interaction becomes large – a phenomenon known as asymptotic freedom. Choosing an increasing energy scale and using the renormalization group makes this clear from simple Feynman diagrams; were this not done, the prediction would be the same, but would arise from complicated high-order cancellations.”

              [ ]

  12. Ellis has now written dozens of articles, some in respected journals, on “downward causation”, as have other physicists and philosophers. All of the musings about mental states causing physical states seem to hinge on this alleged process. I would like to read the arguments that claim downward causation does not or cannot exist. Sean Carroll wrote briefly about it some time ago (it can’t exist in a physical, that is non-semantic, sense because the standard model leaves no room for it) but I would like to read more. Does anyone know a good and readable source?

    1. “Sean Carroll wrote briefly about it some time ago (it can’t exist in a physical, that is non-semantic, sense because the standard model leaves no room for it)”

      That’s the basic argument. In order to have downward causation you’d need to overturn large swathes of physics. And actually, current physics works very well indeed (and our understanding of it is demonstrated by the fact that the device you’re reading this on works).

      Now of course, in principle, one could indeed overturn large swathes of physics. But, first, you’d need a damn good reason to want to do that, and second you’d need something better to replace it with. The downward-causation people have neither of those, they have only a hankering after the religion they were taught in Sunday school.

      1. Coel, do you know if it has been formally proven that in principle one need not invoke higher-level chemical, biological, and/or psychological levels of explanation to predict human behavior, but only for example Sean Carroll’s one equation of the physics of the world of everyday experience (combining QM, spacetime, gravity, other forces, matter, and Higgs)? Of course everything at the higher levels has to be *consistent* with that equation (so no overturning of physics), but that’s different from saying that it’s all that’s in principle needed to predict behavior. The equation is described at

        1. “do you know if it has been formally proven that …”

          No, it has not been formally proven, and I think there’s no way one could formally prove it.

          The brain involves around 10^27 particles, and there’s no way we can have the computational power, nor the ability to assemble all the necessary information, to apply Schroedinger’s equation to 10^27 different particles and compute all the way from particle physics up to psychology.

          1. I agree about the practical impossibility of computing from particle physics to psychology, but I think Sean Carroll might say that in principle all you need is the laws of physics to say where any particular molecule ends up at time T. To say otherwise would be to admit that downward causation is real, and that’s the sticking point here (free will aside).

              1. And that idea is nowadays captured in normalization theory. I tried to hash out what that may mean in the context Ellis want to see (“levels” and how they depend on each other), and I ended up in a place that is new for me. Responses are thus welcome!

                [See longish comment above.]

        2. The issue is not whether it is actually possible to derive higher levels of description from just using the equations of QFT and the Standard Model (which is obviously impractical). The issue is whether the lowest level is “causally closed”, i.e. there are no “missing ingredients” required to explain, in principle, an empirical observation. Panpsychists, for example, claim that a conscious element should be added to the basic elements of reality (but I heard that they also agree that it would not be possible to actually show it experimentally, so good luck with that)

    2. Apparently, Ellis claims that downward causation can be demonstrated in digital computation. Sounds unlikely to me. Does anyone know what he means?

      1. An alien from another planet who only knew about basic physical entities like particles, charges, etc. might have some hard time at first figuring out that, of all the incredible variety of basic phenomena happening in the computer circuit, the only relevant events are the electric potentials at particular points in the circuitry (which we call gates). But eventually the alien will figure that out, and that will be a perfectly valid explanation of what happens, even if the alien does not have the concept of computer program. As usual, Ellis improperly mixes different levels of description and gets different ontological entities out of those levels

    3. Sean Carroll unfortunately uses “downward causation” to mean the woo-ish idea that people like George Ellis promote. That’s not what Roger Sperry meant when he coined the term – he simply meant that high-level concepts can contribute to useful explanations of low-level events. Which is just plain true, and not even controversial. I guess the terminological abuse by woo-ists was inevitable, but I still feel it’s worth fighting.

        1. Since I specifically denied that Carroll agrees with Ellis’s woo-ish idea, no. But in an AMA, I asked him a question based on Roger Sperry’s non-woo-ish idea, and Carroll answered that a high-level description *plus some low-level descriptions* (emphasis his) can indeed explain a low-level event.

  13. As stated in the Wikipedia article, Prof. Ellis is an active Quaker. Also in one of his aeon responses to comments he states that he is a moral realist. This seems to be a typical case of religious believes shaping someone’s philosophical assumptions, while at the same time claiming it is all derived from scientific evidence. This is particularly evident in the tone of his aeon responses to comments, where he always presents his positions as the Truth, not as a possible intellectual challenge to opposite positions

    1. Ellis’s fallacy is easy to spot: he turns different levels of description into different ontological entities

      1. Ugo,
        I love this turn of phrase, but I suppose that is because I tend to share your evident nominalist leanings. There are, however, Plato-type “realists” out there, such as Christian List, who insist that the “higher level” entities discerned in higher-level explanations are just as real as anything else. I don’t know of a slam-dunk argument against this position but, even if it’s correct, there seems to be something illicit about using inferences from higher levels of analysis to trump inferences drawn from more granular levels of analysis.

        1. Ultimately it comes down to taking a particular philosophical position based on available evidence. Nothing in physics is a slam-dunk and any theory could be later modified by new evidence.
          But I find it amusing, talking about free will, to see how *predictable* are the philosophical positions *chosen* by theists

    1. Out of curiosity, I clicked the link and I think I see the problem. It appears (from p. 21 of his 2020 article, which he cites) that when Ellis says “downward causation” exists, he is using a concept of causation known as “difference-making” or “contrafactual” causation. It is an essentially correlation-based concept of causation (rather than mechanism-based) that is also touted by writers such as Christian List to “prove” that free will is real. Because, however, it infers causation from regular covariations (correlations) and does not insist on finding operative mechanisms, it is not good at discerning the causes of effects that can be “multiply-realized” or in dealing with situations where a third factor produces both the putative case and the putative effect.

      Because psychological states are rather highly correlated with behavior, a person using the “difference-making” concept of causation can easily infer that psychology causes behavior–even though (absent some hocus-pocus) there is no known mechanism by which that would be possible.

      1. I don’t agree with Ellis, but, difference-making is the only scientifically respectable approach to causation we’ve got. It is fully capable of dealing with situations where a third factor produces both of two variables that are correlated with each other. And it does not try to reduce causation to pure statistics. Instead it highlights the role of natural and controlled experiments. Try reading Judea Pearl instead of George Ellis.

        1. I have to say that the philosophical modeling of causality makes things complicated – correlations do not need to confer light cone causality even if the converse exist – but after trying to respond to Tom Clark I grokked (or not) the current way to model “lower and higher level laws”.

          And it sidesteps the whole issue, I think, there are complications that philosophers doesn’t account for – laws change over energy scales and that means a different sort of “causal” dependence in these questions than the one(s) we usually think of.

          [See longish comment above.]

  14. I started reading this article at another site a few days ago. Blather – it didn’t explain subject properly from a scientific point of view that made sense to me. Also, the moment I pick up any tone of sentient or wellness in this type of essay it sparks my bullshit detector.

  15. In the end it all goes back to Princess Elizabeth of Bavaria’s question to Descartes – how can the proposed non-physical realm affect the physical? In absence of any such process all behaviour must be considered deterministic. And in the face of the universe being a quantum field, where does this non-physical agency reside?


  16. Jerry, I don’t think your definition of determinism is in line with prevailing usage. It explicitly allows quantum randomness (even if that is fundamental) to count as determinism. That goes against mainstream scientific and engineering usage, where “determinism” means that from the current state plus laws, the future state is logically/mathematically implied.

    But I do think you’re right to point out that Ellis’s lack of a definition of determinism hurts his argument. He runs together determinism (in something like the narrow sense above) and causality, not noticing that micro-determinism does not imply micro-causality, where causality is defined as an asymmetric relation. Thereby he misses the actual hole in classic anti-free-will arguments, and has to go digging an imaginary hole instead.

  17. Oy! I wouldn’t have the energy for either Ellis religious flimflam or Jerry’s debunking.

    As an aside, I just read a cosmological article that explicitly wanted to point out “the natural nature of nature” but, because it used a religious term, annoyingly was swamped by creationists in the comments.

    Ellis conflates physical determinism with predictability, an odd stance for a physicist.

    To cite Wikipedia, Ellis is an active Quaker and writes these Platonic philosophical know-nothings from time to time. And his work concerns anisotropic, inhomogeneous cosmologies and now lately ‘top down causation’ as demonstrated here, all of which we don’t see. I have put him on my ignore list for the time being.

    1. And I now note that Ugo Corda was before me in noting Ellis religion (from the same source).

  18. “Most important, he’s pretending that psychological phenomena are not physical.” – J. Coyne

    Here’s a smart question asked by Jack Smart:

    “How could a nonphysical property or entity suddenly arise in the course of animal evolution? A change in a gene is a change in a complex molecule which causes a change in the biochemistry of the cell. This may lead to changes in the shape or organization of the developing embryo. But what sort of chemical process could lead to the springing into existence of something nonphysical? No enzyme can catalyze the production of a spook!”

    (Smart, J. J. C. “Materialism.” Journal of Philosophy 60/22 (1963): 651-662. p. 660)

    1. That’s not a smart question at all. Many animals with a brain will have “nonphysical” thoughts, tendencies, or similar evolved products that emanate from brain structure as a result of natural selection. Likewise for subjective sensations like pain. It’s really a dumb question to posit that it’s impossible for sensations or thoughts or attractions (like that of a bird of paradise to a particular male ornament) to evolve by natural selection. Only someone who isn’t thinking hard about biology could even pose a question like this.

      1. I’m sorry, but there’s a misunderstanding! For Smart never asserted that “it’s impossible for sensations or thoughts…to evolve by natural selection.” He was a reductive (not an eliminative!) materialist about the mind and consciousness, who believed that all psychological phenomena are (reductively identifiable with complexes of) physical, neurophysiological phenomena—which as such can certainly “arise in the course of animal evolution” through natural selection. What he thought cannot naturally arise or emerge so are *nonphysical, physically irreducible* mental entities as postulated by dualists, i.e. ones which aren’t composed of or constituted by neurophysical entities. What Smart disbelieved in is the possibility of a “spooky” ontological emergence of nonphysical, physically nondecomposable and thus nonreducible mental entities from physical systems such as animal brains.

        The bottom line is that his question—*which is directed at antireductionist/emergentist dualists about the mind*—is not to be read as “How could a mental property or entity suddenly arise in the course of animal evolution?”, but as “How could a *nonphysical/physically irreducible* mental property or entity suddenly arise in the course of animal evolution?”.

        1. Well then Smart’s question remains dumb, for he has no examples of a physically irreducible or “nonphysical” mental property that is convincing.
          I think he had the wrong last name. . . .

          1. There’s still a misunderstanding on your part, since Smart is a reductive materialist about the mind, who certainly doesn’t believe in any real examples of nonphysical, physically irreducible mental entities—but his opponents, the dualists, do; and he’s asking *them* how it’s naturally possible for a nonphysical mental entity to arise or emerge from purely physical entities. That is a smart question, and there is no smart answer by the dualists!

Leave a Reply