According to Conway's FWT, it appears to me that electrons collapse into one of the possible state as a result of free will (assuming the experimenters/observers have free will, and superdeterminism is not true). Conway defines free choice decision to be one that is not entirely caused by some event in the past light cone of the electron.

At the same time, physicists know from observations in Quantum Mechanics that electron wavefunction collapse is a random process (i.e. perfectly unpredictable) for a single electron.

Is there any reason or explanation in physics that precludes the conclusion that perfect unpredictability is indistinguishable from free will (when we consider a single electron and the observer)?

  • 4
    perfect randomness accounts for the "free" in "free will", not the "will". Assuming you are speaking about libertarian free will, a free choice does not only have to be unpredictable, it also has to be your choice. I.E, quantum state collapse happened in such and such a way because you wanted it to. Some decision process in us has to precede to physical reality. Therefore even if it was demonstrated that perfectly random events do exist, it would not follow that we have free will.
    – armand
    Apr 7 at 4:24
  • There is not a single definition of free will, and the question does not clarify enough which of the many versions it refers to.
    – tkruse
    Apr 7 at 9:12

6 Answers 6


Is there any reason or explanation in physics that precludes the conclusion that perfect unpredictability is indistinguishable from free will (when we consider a single electron and the observer)?


General Relativity has "perfect unpredictability" that is NOT due to free will.

You are asking for "perfect unpredictability", which I take to mean "in-principle impossible to predict to arbitrary accuracy". The light-cone structure of GR is in-principle antithetical to perfect prediction. The light-cone structure makes predictions at any finite time in the future for any agent within the universe impossible to predict to arbitrary accuracy.

This is because the future of any observer is not given only by its past light-cone, yet that is all it has access to (plus the laws of physics) to make predictions. But those are not enough. There are always "space invaders" which can appear at any finite time in the future for any observer. E.g. "particles appearing from spatial infinity without any prior warning. To put it crudely, you can’t hope to have Laplacian determinism for open systems" https://www.ucl.ac.uk/~uctytho/Determinism.pdf

And clearly this in-principle unpredictability has nothing to do with free will. It is the light-cone structure of GR, not free will, preventing predictions.

Now that was only one of your questions and not your titular one:

[are] perfect randomness and free will are indistinguishable?

Under any test possible? To me the human (with free will) under experimentation will eventually be found out if the experiments are rigorous enough. In that sense, the human who posses free will in choosing heads or tails is distinguishable from the "perfectly random" coin, as the coin never tires. Trivially the human needs a bathroom break eventually. Is this not a fair experiment? It seems perfectly doable to me.

  • Good answer referencing randomness outside free will. My own answer focused on how a free will product is similar to a random outcome. One can also mention chaos theory as another example of randomness unrelated as far as we know from free will.
    – Nikos M.
    Sep 7 at 18:38

Free will in the sense of introducing willed uncomputable novelty, that is, something novel (a decision for example) that cannot be predetermined completely from the current state of affairs, coincides with a common definition of randomness, that is, uncomputable novelty.

So at least in this sense, a product of free will coincides with a random outcome (as far as providing a mathematical model is concerned), and this is not problematic.

Note that a product of free will being similar to a random outcome does not necessarily imply that all randomness is related to some free will.

For an approach as to what randomness means and how free will can be compatible with randomness see, for example:

Free will is compatible with randomness

It is frequently claimed that randomness conflicts with free will because, if our actions are the result of purely random events, we must lack control over them. The paper challenges this view. After arguing for a product rather than a process notion of randomness, it uses an intuitive two-stage, contextual definition of free choice to show that, relative to this definition, randomness is compatible with free will. But we also stress that the argument is relative in nature: the conclusion holds only if free will is itself metaphysically possible, a claim that is beyond the purview of the paper.

You may be interested in this answer as well.

  • Free will is the ability to make deliberate choices. In that sense free will is the very opposite of randomness. However, randomness is required for making freely willed choices. In a world without randomness there would be no options to choose from. We need some random recombination of previously existing ideas to come up with alternatives to choose among. Sep 12 at 7:46
  • @PerttiRuismäki I don't disagree, except that there is a sense in which free will can be seen as creating randomness or exploiting randomness, so at least in this sense, they are not as opposite as sometimes thought. If randomness means something always meaningless and outside of any agency then of course they are opposites, but that is not necessarily so.
    – Nikos M.
    Sep 12 at 14:45
  • 1
    Free will is statistically random. It is unpredictable based on knowledge about prior events. Free will is philosophically the opposite of random. Intentional vs. unintentional / purposeful vs. purposeless / future-oriented vs. past-causal. Sep 13 at 3:49
  • @PerttiRuismäki if randomness is always unintentional then certainly is contraposed with free will. But it is possible this is not always or necessarily so.
    – Nikos M.
    Sep 13 at 6:07
  • True randomness is always unintentional. Statistical analysis does not detect or care whether the randomness is true (=unintentional) or pseudo-randomness (=intentional, fake). Sep 13 at 8:54

To riff Pertti Ruismäki's answer from a materialist perspective: free will and randomness contradict each other.

As armand writes in a comment:

perfect randomness accounts for the "free" in "free will", not the "will". Assuming you are speaking about libertarian free will, a free choice does not only have to be unpredictable, it also has to be your choice. I.E, quantum state collapse happened in such and such a way because you wanted it to. Some decision process in us has to precede to physical reality. Therefore even if it was demonstrated that perfectly random events do exist, it would not follow that we have free will.

To have free will is for your actions to be decisions taken by you, and be the decisions you want to make. Your decisions can be predicted imperfectly by others, but the only way to truly predict your decisions is to have another copy of you in the same situation and observe what that copy does. (And it has to be the same situation – for all an external observer knows, you base your decisions on the exact pattern of shadows playing on a wall.)

Non-deterministic events in your (materialist) brain fly in the face of this. Those events, which affect your decision-making process, are not a result of you. If by coincidence, they happened to all nudge you in the same direction, they could determine your choices! Rather than your decisions solely depending on you and your environment, there's an additional node in the causal graph that affects you (in ways other than informing you about the world) and that you cannot affect; it reduces your free will.

How this reduction in free will compares to that caused by your need to breathe is an exercise for the reader.

  • 2
    Those external out-of-control events do have an effect on the decisions, but they do not limit our freedom in any way. We are still completely free to choose our responses to external events. These changes in the environment are like questions, this happens, what are you going to do about it? Decisions are answers to these questions. Jul 14 at 12:19
  • @PerttiRuismäki External out-of-control events, sure. But materialism is the position that the brain is a physical system; there are internal out-of-control events, too.
    – wizzwizz4
    Jul 14 at 17:23
  • 1
    Internal physical events are external to the mind. The mind does not deal with physical events at all. The mind deals with knowledge about physical events. Materialism does not recognize the concept of knowledge at all, thus rendering itself a useless piece of knowledge. Jul 14 at 18:41
  • @PerttiRuismäki Sure it does. Look at Maxwell's demon – a physics thought experiment involving knowledge.
    – wizzwizz4
    Jul 14 at 18:44
  • @wizzwizz4 there is nothing a priori limiting mind processes to produce events which are otherwise unpredictable. That is there is no a priori reason an agent cannot create randomness but only receive it. If the universe can create randomness then we can share this ability. Hope this is clear
    – Nikos M.
    Sep 10 at 21:17

Free will and randomness are philosophically logical opposites. Physically they affect different points of causal chains.

Unpredictable outcomes come in two flavours: choice and chance.

  • Intended outcomes (=actions with a purpose) are called free choice.
  • Unintended outcomes are called random chance.

Unintentional randomness is the inaccuracy between a cause and its effect.

Intentional, freely willed, voluntary actions are caused by the decision to act, instead of being caused by a prior event.

Randomness modifies the effect independently of the cause.

Free will generates new (=uncaused) causes expecting certain desired causal consequences.

  • Downvoting because this just asserts the belief that will, decisions, purpose and so on are magically not caused by anything and can break causality just because we assert that they can.
    – tkruse
    Apr 7 at 9:02
  • 1
    There is nothing magical about not being caused. Only physical events are caused. Causality is not broken by adding new causes in the mix. Apr 7 at 9:43
  • 1
    @tkruse, Upvoted because the assumptions this answer makes are fully consistent with empirical observation, and the only reason to doubt them is to support materialism, which is an ontological theory supported by practically no rational argument. Apr 7 at 15:49

Randomness (or whatever it is that we are talking about by using the word "random") is not necessarily "perfectly unpredictable," or at least we need to spell out that phrase "perfectly unpredictable" more closely. To the extent that randomness is (and isn't) caught up with chances/probabilities, we also have to keep in mind e.g. completely subjective interpretations of probability, which although licensing "crazy" beliefs like, "The PhilosophySE is secretly in control of the entire universe" (just so long as we consistently assign the relevant probabilities), nevertheless present us with yet another problem to solve, here.

Now the motive for believing in the indeterministic form of free will is often stated nowadays in terms of being a precondition for "moral responsibility." Between the two of them, Wilfrid Sellars and Harry Frankfurt have done a decent job of undermining such a preconditional statement, Sellars more on the internal psychological side of things (his reflection on "reactive attitudes") and Frankfurt on the external such side (the mesh issue). But when Kant argues for transcendental freedom (in the second Critique), he frames it much more simply in terms of the following line of reasoning:

  1. Suppose I am obligated to do X.
  2. If I am obligated to do X, I am able to do X to some extent.
  3. Sometimes I violate my obligations.
  4. Therefore, when I am under an obligation to X, I face a choice between Xing and not Xing.

This reference to a physical example of a "mere possibilia" is not a reference to complex attitudes, at least not on its surface. "If A, then B or not B," does not clearly have anything directly to do with randomness, chance, or even a dedicated faculty of willpower. And Kant, although a faculty psychologist par excellence, is at pains to locate transcendental freedom in transworld logical space, so that he says that we could empirically build a model of human behavior that genuinely explains and predicts all empirically visible human actions and which nevertheless comports with the modal spontaneity involved in deontic affairs.

I am sympathetic to a post-physicalism that equates electrons with the primary physical substrate of consciousness. Electrons have enough quirky aspects that I don't see why we would need to advert to more fantastical conceptions of "refined spiritual matter" such as e.g. the LDS Church tries to present as the constituents of our spirit. I recognize that quantum indeterminacy is not one-to-one with the indeterminism we are looking for with respect to transcendental free will (or choice; note e.g. Kant's distinction between Wille and Willkür), so I cannot say that electronic indeterminacy is indistinguishable from such freedom. Perhaps panpsychism is true and so electrons are conscious even just by themselves, rather than only in concert as sustained neurological activity. But until I see a reason to accept the axioms that go into controverting transcendental freedom (as deontic-modal disjunction) on grounds of "randomness" or the like, such criticisms, and the preoccupation with addressing them, won't seem compelling. Whether that reopens the door to thinking of electronic indeterminacy as sufficient in order to free choice, I can't say for now.

Addendum. I guess I would say that trying to equate free will with randomness, and defending or critiquing such will on that basis, is liable to be equivocal without a more refined sense of what randomness, chance, etc. are really supposed to be, but then even on its own terms just the same. "It was random that choice A was made instead of choice B," is not the same assertion as, "A occurred randomly." No one, not even I myself, can unfailingly predict what I will do with my freedom, but that does not mean that, once I have done something with it, that I was somehow not the author of what I did. If the human mind or spirit is largely reducible to electronic activity, even the activity of a single electron in the limit, perhaps there will be the intended overlap between quantum indeterminacy and broader modal indeterminacy, but the solution to this question is for another day.


The referenced paper is explicit: The universe, and elementary particles, as asserted to have causal will. As willing is a feature of consciousness, this is an assertion of panpsychism, along with psycho-physical interaction.

The difference between randomness and willing is the consciousness and agency of the willer. These are easily distinguished internally.

From the outside, the question might be whether applying a Theory of Mind is predictively useful or not. The thesis in the paper that free will is basically purely random in its outputs, suggests that Theory of Mind would be predictively useless for these elementary particle minds, and by inference, human minds as well. Given the utility of Theory of Mind with human minds, and animal minds, this is a major counterevidence to the proposed definition of free will in the paper.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .