My critique of a defense of free will in Quillette

On July 6, philosopher William Edwards wrote a defense of free will in Quillette called “The Academic quarrel over determinism.” Besides a vigorous defense of free will, included two paragraphs that I saw as deeply misguided and misleading (see my preliminary note here):

Needless to say, there is substantial evidence that people who believe in free will, or at least believe that they are in control of their own lives, are more prone to exhibit good mental health and productive, ethical behaviour. There is a not inconsiderable moral dilemma here. As we illuminate the role of DNA and other fixed factors, we will acquire knowledge that should allow us to improve and save countless lives. On the other hand, if more and more people come to accept the idea that they’re not choosing their thoughts and actions, their subsequent behaviour may guarantee that a lot more lives are in need of saving.

Thinkers like Harris and Weinstein are preoccupied with how we build a less risky world, which may be partly why their thinking appeals to conservatives. However, it is worth remembering the well-established relationship between risk and reward, because whether or not we believe in free will may turn out to be the Pascal’s Wager of the twenty-first century. With that in mind, any professional gambler worth their salt should bet on free will. There is just too much about the universe that we don’t understand, and the potential pay-off from agency is staggering.

The first paragraph involves a lot of selective citation, since the data on whether belief in free will has salubrious personal effects is contradictory and complex. The second paragraph involves one forcing oneself to believe a proposition for which there’s no evidence, simply because the unevidenced belief is beneficial.

I’ve just published a rebuttal, also in Quillette, which you can read by clicking on the screenshot below.  Comment here or (preferably) on Quillette.


  1. Evan Plommer
    Posted July 19, 2019 at 12:37 pm | Permalink

    Terrific piece. Will be interesting to see if there’s a reply from Edwards.

  2. Posted July 19, 2019 at 12:47 pm | Permalink

    Suggestion Jerry:

    “Contracausal free will is the modern equivalent of black plague, magnetism and lightning—enigmatic phenomena that were once thought to defy natural explanation but don’t.”

    That would be better phrased as: “Decision-making by the brain is the modern equivalent …”, since the comparators exist but contracausal free will does not.

  3. GBJames
    Posted July 19, 2019 at 1:03 pm | Permalink

    I left a comment there, or though I did. But It isn’t showing. So…


    • Posted July 20, 2019 at 7:59 am | Permalink

      Same happened to me, so I started over. (My Typio add-on to Chrome didn’t recover it for some reason.) When I was done, there were two. Maybe it just takes a while?

      • GBJames
        Posted July 20, 2019 at 9:00 am | Permalink

        Yeah. Mine appeared shortly after I whined about it here.

  4. Randall Schenck
    Posted July 19, 2019 at 1:22 pm | Permalink

    Left a thought over there. Always a lot of head-scratchers over there.

    • Saul Sorrell-Till
      Posted July 19, 2019 at 4:38 pm | Permalink

      It’s pretty depressing over there. Absolutely rammed with god botherers and Nietzsche-nuts. I think I find the latter more creepily unpleasant but the former tend to be the more comically obnoxious.

  5. Gasper
    Posted July 19, 2019 at 3:24 pm | Permalink

    No free will doesn’t actually bother me. To my mind, it explains a lot of things.

    What bothers me is a deterministic universe, although to some these are the same thing.

    For example: do you really believe that the guy who steps off the curb of a busy street, is distracted by something trivial, and is killed instantly by a passing car was predetermined at the moment of the Big Bang?

    • Posted July 19, 2019 at 3:41 pm | Permalink

      Probably not, given quantum indeterminacy combined with deterministic chaos (though as Jerry explains, that’s pretty irrelevant to the issue of free will).

    • Posted July 19, 2019 at 3:49 pm | Permalink

      The possibility was determined at the big bang.

    • Steve Gerrard
      Posted July 19, 2019 at 4:24 pm | Permalink

      The guy being born wasn’t pre-determined at the moment of the Big Bang. There are far too many spots along the way where some very small statistical variation could lead to an entirely different result. (e.g. Which one of those 50 million sperm cells is going to be the one?)

    • A C Harper
      Posted July 20, 2019 at 4:04 am | Permalink

      There’s a big difference between deterministic and knowable.

      The human brain is very limited in its ability to process information into top level activity. It cannot hold all the information needed to see the causal chains accurately, or very far in the past. There’s a lot of ‘beliefs’ and ‘rules of thumb’ to boil down the flood of perception into whatever is salient, and much of the history of belief and rule building is not available to introspection. Plus there’s the time lag between perception and reaction. We live by a prediction for few hundred milliseconds in the future by ‘analysing’ the perceptions a few hundred milliseconds in the past. That’s pretty spooky, but it works well enough to survive (until it fails).

      So stepping off a curb is completely deterministic as regards causes, the effect (getting mown down) is completely deterministic, but not knowable by the poor persons involved.

  6. Nicolaas Stempels
    Posted July 19, 2019 at 3:57 pm | Permalink

    Sensei Coyne demolished Edwards-kun ‘s conjecture.
    Sadly, for some reason or other, only about one in three comments I make on Quillette actually appear. Luckily Saul and Coel (and some others) do a good job there.

    • Saul Sorrell-Till
      Posted July 19, 2019 at 4:41 pm | Permalink

      Do you fill in the email address and username slots each time? I don’t know if it reminds you to or not, but it’s a requirement.

  7. Nicolaas Stempels
    Posted July 19, 2019 at 4:03 pm | Permalink

    I loved Thad’s comment on Quillette: “I believe in free will, I can’t help it” .
    I guess -and hope- that was intentional.

    • Posted July 19, 2019 at 7:49 pm | Permalink

      That’s the stock answer that both Dawkins and Hitchens gave when asked if they believed in free will. I attribute the witticism to Hitch, but it may have been someone earlier.

  8. Posted July 19, 2019 at 4:10 pm | Permalink

    You do realize that writing for Quillette automatically makes you Alt-Right?

    • Posted July 19, 2019 at 4:20 pm | Permalink

      Only if I write alt-right stuff. If anybody accuses me of that nonsense, I’ll tell them to look at what I wrote. If they say I’m alt-right because I wrote for Quillette, I’ll tell them to get stuffed.

      • Posted July 19, 2019 at 4:46 pm | Permalink


        It’s gotten insane. I just saw where Vice, in its Schadenfreude over her Twitter ban, described Lindsay Shepherd as “alt-right”.

  9. Posted July 19, 2019 at 4:13 pm | Permalink

    Excellent rebuttal to Edwards, and thanks for citing recent literature that casts doubt on the “salubrious personal and social effects of belief in free will.” We really would be much better off if folks took on board naturalism as you define it: “the view that the cosmos is completely governed by natural laws, including probabilistic ones like quantum mechanics.” This would spike the idea that, as you put it, “we could have done something other than what we did” (randomness aside). Drop that contra-causal assumption and it’s a lot harder to demonize those who end up on the wrong side of the law, morality, or our preferred politics. Have to get you on The Late Show to make all this go mainstream.

  10. Steve Gerrard
    Posted July 19, 2019 at 4:56 pm | Permalink

    Edwards says this: “Needless to say, there is substantial evidence that people who believe in free will… are more prone to exhibit good mental health…”

    That “needless to say” is a tell here. He is attempting to pass off his claim as obvious, when it is nothing of the sort.

    • Posted July 19, 2019 at 9:23 pm | Permalink

      The extremely religious are prone to believe in free will, and they are delusional.

  11. Liz
    Posted July 19, 2019 at 9:29 pm | Permalink

    “And you needn’t believe in pure physical determinism to reject free will. Much of the physical world, and what we deal with in everyday life, does follow the deterministic laws of classical mechanics, but there’s also true indeterminism in quantum mechanics. Yet even if there were quantum effects affecting our actions—and we have no evidence this is the case—that still doesn’t give us the kind of agency we want for free will. We can’t use our will to move electrons.”

    I’m a determinist and incompatiblist.

    I came across the following a few years ago from Sean Carroll and am still curious. I posted on here last year and still have the same question/interest in the correct answer.

    “The Stuff of Which We Are Made: Lecture 2 Prof. Sean Carroll,The Gifford Lectures in Natural Theology”

    Here is a shortened link to the above mentioned video which can also be found by Googling some of the words in the title.

    or (same link)

    20:46 in is about where the whole explanation of the slide begins.

    25:35 in for one of the images of this slide.

    “1. Many-Worlds. Every option becomes real in a different world. Overall evolution is deterministic.

    2. Hidden variables. Unknown quantities fix future measurement outcomes. Evolution is deterministic.

    3. Dynamical collapse. The universe observes itself, regularly but unpredictably. Truly stochastic, indeterministic.”

    25:46 in for the following referring to the dynamical collapse interpretation of quantum mechanics:

    “It’s experimentally distinguishable. The experiments are going on to test this idea right now. But in this theory the world would truly be indeterministic rather than deterministic.”

    Is it correct that the indeterminism here does not leave room for a choice somewhere in our electrons or neurons? It seems to me like it does. If there are experiments going on right now, eventually, this could be researched in our brains also.

    • Steve Gerrard
      Posted July 20, 2019 at 1:03 am | Permalink

      It leaves room for things to go one of several possible ways, with the range of possibilities being determined by quantum mechanics.

      It doesn’t provide any means for some outside influence to affect the result, though.

    • Kosmos
      Posted July 20, 2019 at 6:38 am | Permalink

      It’s hard to see how QM indeterminacy could leave room for libertarian free will when, as far as I’ve understood it, the measurement outcomes follow absolutely strict and predictable probability distributions. The outcomes sure seem forced to follow these distributions, how could we otherwise predict the shape of the probability distributions?

      And the very concept of a libertarian choice seems entirely incoherent to human reasoning. It’s hard (impossible more likely) to see how it wouldn’t entail some form of incoherent self-creation. See Galen Strawson’s argument:

      So even if libertarian free will does exist, I don’t see how we have the cognitive faculties to discover/identify it. It would just look like genuine randomness to us.

      • Liz
        Posted July 22, 2019 at 9:02 am | Permalink

        “…the measurement outcomes follow absolutely strict and predictable probability distributions. The outcomes sure seem forced to follow these distributions, how could we otherwise predict the shape of the probability distributions?”

        Thank you. This is very helpful.

        Just to be clear, when people refer to quantum indeterminacy, they are referring to the indeterminism in the dynamical collapse interpretation of quantum mechanics?

        • Kosmos
          Posted July 23, 2019 at 10:22 am | Permalink

          No, I think that most often refers to our inability to predict individual outcomes, regardless of underlying interpretation of quantum mechanics.

          • Liz
            Posted July 23, 2019 at 11:31 am | Permalink

            Is this correct? There are two different definitions of indeterminism when discussing free will (or otherwise).

            1. One refers to the “inability to predict individual outcomes”.
            (To clarify, we would be able to predict individual outcomes if we could explain quantum gravity?)

            And the other:

            2. The second refers to the progression of events as the “universe observing itself regularly but unpredictably” (from Sean Carroll at 25:46 in the above slide: according to the dynamical collapse interpretation of quantum mechanics.

            Indeterminism in the first instance is present in all three of these of these interpretations of quantum mechanics as it refers to the inability to predict outcomes of individual subatomic particles:

            a. Many-Worlds.
            b. Hidden variables.
            c. Dynamical collapse.

            Is that correct?

            Indeterminism in the second instance is only present in the third interpretation of quantum mechanics (c. dynamical collapse) and it refers to the unpredictability of the universe observing itself.

            Is that correct?

            Because, if the third interpretation of quantum mechanics is correct, the universe’s observations of itself occur regularly, even though unpredictably, the progression of events is *still* “deterministic” because “…the measurement outcomes follow absolutely strict and predictable probability distributions.”
            Is this correct? Is it because the events occur regularly in this case that they follow predictable probability distributions?

            Any clarification would be wonderful.

            • Posted July 23, 2019 at 11:41 am | Permalink

              It is a bit more complicated than that because of the “random at one level, not at the level below” possibility. (This does not seem to be the case here, but let the Bohmians try. :))

              And there’s also a debate on whether retrocausal views count as hidden variable views.

            • Kosmos
              Posted July 25, 2019 at 2:01 pm | Permalink

              I don’t think a theory of quantum gravity has anything to do with us being able to predict individual outcomes. That would imply that it would reveal a hidden variable.

              Dynamical collapse as described by Sean would mean fundamental indeterminism; genuine probabilistic randomness would be a part of the universe and the progression of events would thus not be deterministic. But the evolution of the universe would still be completely based on laws/rules, it’s just that these laws would have a probabilistic component when collapse of the wave function occurs. It is these discoverable laws that makes it possible for us to know the probability for different outcomes and therefore makes us able to predict the probability distributions of specific phenomena, such as the measured locations of electrons in hydrogen atoms.

    • Posted July 20, 2019 at 8:11 am | Permalink

      If your neurons and their constituent particles do something indeterministic, then, since those cells and particles are what you are, you also do something indeterministic. You’ve indicated that you believe indeterminism is necessary for “choice”. Indeterminism isn’t necessary in my view, but if it were, sure, that would be a choice that you made.

      The key point is not to alienate yourself from yourself and your constituent parts. As Daniel Dennett says, “if you make yourself really small, you can externalize virtually everything.” Smart people don’t.

      • Posted July 23, 2019 at 11:42 am | Permalink

        Except of course then you have no control – no more than if you decided to always use a geiger counter (or a die) to make decisions.

        • Posted July 27, 2019 at 7:42 am | Permalink

          I disagree. If your choice directly causes an action, you have control over that action. If it indirectly causes that action with the help of the randomizer, then you have control over the probabilities with which you do the actions.

  12. Evan Plommer
    Posted July 20, 2019 at 11:44 am | Permalink

    What a lot of wildly depressing comments on Quillette in response to this essay. Worst of all: many give the impression that the commenter didn’t even read the piece. That seems immoral to me.

  13. Posted July 21, 2019 at 5:43 am | Permalink

    I can’t seem to remember reading his article, which is surprising because anyone who uses Gödel’s theorems to argue for free will is immediately etched into my mind as a crackpot.

    And singularities! The argument is hilariously bad. Singularities tell us our models are wrong. They don’t exist. He seems to conflate singularities, which exist in math and the theory of general relativity, with what occurs in the center of black hole and in the Planck epoch of the Big Bang, and afterwards apply the terrible pop-sci trope of “all rules (read: our theory of gravity) break down at a singularity/the center of a black hole” to mathematical singularities.

    Also, this is somewhat minor and doesn’t impact the argument, but I’m under the impression that most people working in quantum foundations believe quantum mechanics is actually deterministic. Many-worlds can explain quantum mechanics without having to introduce indeterminism, which is entirely unheard of in physics.


    • Posted July 23, 2019 at 11:43 am | Permalink

      Singularity, effectively: x=0 in f(x) = 1/x. Nothing more.

      If that’s a *factual* equation, it simply tells our idealization is wrong at least around x=0.

      (Laws vs. law statements. Thank you Hawking.)

  14. Matthew North
    Posted July 21, 2019 at 5:54 am | Permalink

    I love it when a rational, carefully reasoned argument from a scientific mind like Jerry’s, or Dawkins or Darwin, looks at yet another half-reasoned argument for, “that mysterious something that humans can never quite fathom so we’re all better off believing in it”, and says, “Hold on now, let’s look at your argument based on the, you know,-EVIDENCE!”.

    I decided long ago that “we’re all better off”, looking at and trying to understand The Cosmos how it is, not how we hope or wish it to be.

  15. JB
    Posted July 21, 2019 at 8:36 am | Permalink

    Thought to comment at Quillette, as our host suggested, but reading through the numerous other comments there… holy ceiling cat. Seems like the loons have taken over the place.

  16. George McKee
    Posted July 22, 2019 at 11:36 pm | Permalink

    I generally don’t respond to threads with hundreds of replies, due to signal-to-noise considerations. But it may be worthwhile to bring out these points somewhere nearby, like here, in case some reader wants to take these ideas and run with them…

    And you completed your reply without mentioning Benjamin Libet’s experiments on the Bereitschaftspotential that suggested that the brain makes decisions before the mind is aware of what’s going down!

    Beyond the Hitchens/Dawkins reply, it’s valuable to explain why most of us can’t help feeling that we have free will. Briefly, in a self-aware decision-making system without free will, there occurs a long sequence of events associated with the process of coming to and executing a decision, and that sequence branches off from an equally long sequence of events leading to the reporting of the awareness of whatever that decision was.

    “Free will” requires that the decision itself and the awareness of the decision and its outcome take place exactly simultaneously, at a single quantum event in spacetime. That is, three or four different streams of causality all converge and diverge around the same point of decision. If you take our best theory of reality, quantum field theory, and try to make this convergence happen, you can integrate over all the amplitudes given by all the Feynman diagrams, and the convergence still doesn’t occur. Quantum gravity has recurrent dependencies that might work, but only Roger Penrose believes that there is gravity-related new physics happening in the conditions of human brains that aren’t visible in results from particle physics detectors.

    If there’s no free will, those streams can run in parallel without ever converging to a complex point, and no physics problems occur. You should be able to map out the evolution of the decision-making system within its state space with conventional nonlinear dynamics and chaos theory, and have the macroscopic choice outcome be determined by good old biochemical and thermal noise, which enters the reportable and actionable “mind” part of the landscape only long after the choice point of no return is long past.

    The result is a “feeling” of free will, which although it may be ontologically deterministic, is still autonomous within the person’s brain, and can’t be affected by outside forces until it’s too late. This is pragmatically free, which ought to be good enough for anyone.

%d bloggers like this: