Sunday, January 10, 2016

Free will is dead, let’s bury it.

I wish people would stop insisting they have free will. It’s terribly annoying. Insisting that free will exists is bad science, like insisting that horoscopes tell you something about the future – it’s not compatible with our knowledge about nature.

According to our best present understanding of the fundamental laws of nature, everything that happens in our universe is due to only four different forces: gravity, electromagnetism, and the strong and weak nuclear force. These forces have been extremely well studied, and they don’t leave any room for free will.

There are only two types of fundamental laws that appear in contemporary theories. One type is deterministic, which means that the past entirely predicts the future. There is no free will in such a fundamental law because there is no freedom. The other type of law we know appears in quantum mechanics and has an indeterministic component which is random. This randomness cannot be influenced by anything, and in particular it cannot be influenced by you, whatever you think “you” are. There is no free will in such a fundamental law because there is no “will” – there is just some randomness sprinkled over the determinism.

In neither case do you have free will in any meaningful way.

These are the only two options, and all other elaborations on the matter are just verbose distractions. It doesn’t matter if you start talking about chaos (which is deterministic), top-down causation (which doesn’t exist), or insist that we don’t know how consciousness really works (true but irrelevant). It doesn’t change a thing about this very basic observation: there isn’t any known law of nature that lets you meaningfully speak of “free will”.

If you don’t want to believe that, I challenge you to write down any equation for any system that allows for something one could reasonably call free will. You will almost certainly fail. The only thing really you can do to hold on to free will is to wave hands, yell “magic”, and insist that there are systems which are exempt from the laws of nature. And these systems somehow have something to do with human brains.

The only known example for a law that is neither deterministic nor random comes from myself. But it’s a baroque construct meant as proof in principle, not a realistic model that I would know how to combine with the four fundamental interactions. As an aside: The paper was rejected by several journals. Not because anyone found anything wrong with it. No, the philosophy journals complained that it was too much physics, and the physics journals complained that it was too much philosophy. And you wonder why there isn’t much interaction between the two fields.

After plain denial, the somewhat more enlightened way to insist on free will is to redefine what it means. You might settle for example on speaking of free will as long as your actions cannot be predicted by anybody, possibly not even by yourself. Clearly, it is presently impossible to make such a prediction. It remains to be seen whether it will remain impossible, but right now it’s a reasonable hope. If that’s what you want to call free will, go ahead, but better not ask yourself what determined your actions.

A popular justification for this type of free will is insisting that on comparably large scales, like those between molecules responsible for chemical interactions in your brain, there are smaller components which may have a remaining influence. If you don’t keep track of these smaller components, the behavior of the larger components might not be predictable. You can then say “free will is emergent” because of “higher level indeterminism”. It’s like saying if I give you a robot and I don’t tell you what’s in the robot, then you can’t predict what the robot will do, consequently it must have free will. I haven’t managed to bring up sufficient amounts of intellectual dishonesty to buy this argument.

But really you don’t have to bother with the details of these arguments, you just have to keep in mind that “indeterminism” doesn’t mean “free will”. Indeterminism just means there’s some element of randomness, either because that’s fundamental or because you have willfully ignored information on short distances. But there is still either no “freedom” or no “will”. Just try it. Try to write down one equation that does it. Just try it.

I have written about this a few times before and according to the statistics these are some of the most-read pieces on my blog. Following these posts, I have also received a lot of emails by readers who seem seriously troubled by the claim that our best present knowledge about the laws of nature doesn’t allow for the existence of free will. To ease your existential worries, let me therefore spell out clearly what this means and doesn’t mean.

It doesn’t mean that you are not making decisions or are not making choices. Free will or not, you have to do the thinking to arrive at a conclusion, the answer to which you previously didn’t know. Absence of free will doesn’t mean either that you are somehow forced to do something you didn’t want to do. There isn’t anything external imposing on you. You are whatever makes the decisions. Besides this, if you don’t have free will you’ve never had it, and if this hasn’t bothered you before, why start worrying now?

This conclusion that free will doesn’t exist is so obvious that I can’t help but wonder why it isn’t widely accepted. The reason, I am afraid, is not scientific but political. Denying free will is considered politically incorrect because of a wide-spread myth that free will skepticism erodes the foundation of human civilization.

For example, a 2014 article in Scientific American addressed the question “What Happens To A Society That Does not Believe in Free Will?” The piece is written by Azim F. Shariff, a Professor for Psychology, and Kathleen D. Vohs, a Professor of Excellence in Marketing (whatever that might mean).

In their essay, the authors argue that free will skepticism is dangerous: “[W]e see signs that a lack of belief in free will may end up tearing social organization apart,” they write. “[S]kepticism about free will erodes ethical behavior,” and “diminished belief in free will also seems to release urges to harm others.” And if that wasn’t scary enough already, they conclude that only the “belief in free will restrains people from engaging in the kind of wrongdoing that could unravel an ordered society.”

To begin with I find it highly problematic to suggest that the answers to some scientific questions should be taboo because they might be upsetting. They don’t explicitly say this, but the message the article send is pretty clear: If you do as much as suggest that free will doesn’t exist you are encouraging people to harm others. So please read on before you grab the axe.

The conclusion that the authors draw is highly flawed. These psychology studies always work the same. The study participants are engaged in some activity in which they receive information, either verbally or in writing, that free will doesn’t exist or is at least limited. After this, their likeliness to conduct “wrongdoing” is tested and compared to a control group. But the information the participants receive is highly misleading. It does not prime them to think they don’t have free will, it instead primes them to think that they are not responsible for their actions. Which is an entirely different thing.

Even if you don’t have free will, you are of course responsible for your actions because “you” – that mass of neurons – are making, possibly bad, decisions. If the outcome of your thinking is socially undesirable because it puts other people at risk, those other people will try to prevent you from more wrongdoing. They will either try to fix you or lock you up. In other words, you will be held responsible. Nothing of this has anything to do with free will. It’s merely a matter of finding a solution to a problem.

The only thing I conclude from these studies is that neither the scientists who conducted the research nor the study participants spent much time thinking about what the absence of free will really means. Yes, I’ve spent far too much time thinking about this.

The reason I am hitting on the free will issue is not that I want to collapse civilization, but that I am afraid the politically correct belief in free will hinders progress on the foundations of physics. Free will of the experimentalist is a relevant ingredient in the interpretation of quantum mechanics. Without free will, Bell’s theorem doesn’t hold, and all we have learned from it goes out the window.

This option of giving up free will in quantum mechanics goes under the name “superdeterminism” and is exceedingly unpopular. There seem to be but three people on the planet who work on this, ‘t Hooft, me, and a third person of whom I only learned from George Musser’s recent book (and whose name I’ve since forgotten). Chances are the three of us wouldn’t even agree on what we mean. It is highly probable we are missing something really important here, something that could very well be the basis of future technologies.

Who cares, you might think, buying into the collapse of the wave-function seems a small price to pay compared to the collapse of civilization. On that matter though, I side with Socrates “The unexamined life is not worth living.”


  1. I agree that there is nothing in physics that allows free will. Worse I don't even see how you can coherently define free will.

    But what about experience? The color red, the sound of a musical note, the pain of stubbing your toe. We don't just register these things as data as a robot would. We experience them. I would even say that our belief in free will is caused by the fact that we experience our thought process creating the illusion of control. But this ability to experience does not seem to be explainable by the laws of physics either.

    A computer program should be able to do everything the brain does. Yet if you step back and look at it all it is is a bunch of capacitors that are charging and discharging in a pattern. There is no room here for experience. You cannot look at the pattern and deduce that it has experiences nor can you assume that it has experiences and deduce something about the pattern. So the fact that we have experiences seems to have no effect on the universe. It just makes us helpless witnesses to events that we have no control over.

    Except... how do you explain people talking about experiences in the abstract? If we didn't have experiences then why would we evolve the ability to talk about experiences that we don't have? If we do have experiences but they can have no physical effect then again how can we have conversations about experiences... isn't that a physically observable effect? If we do have experiences and they do have effects then how do you explain this with physical law?

  2. "insist that we don’t know how consciousness really works (true but irrelevant)."


    What do you mean? You can't separate free will and consciousness. Neither do we really understand the collapse of the wave function, you know as well as I that research in fundamental aspects of QM are still carried out.

    You're starting to sound like Sean Carroll, a bit like an evangelical scientist preaching the 'we know everything' mantra :)

  3. Thank you! I've read a lot of free will debate essays from both sides and they rarely make this much sense!

    I think the political opposition comes from defending the American punitive system (as opposed to correctional, like in many European countries). The main argument behind treating prisoners badly is that bad guys have free will and therefore they deserve whatever punishment you throw at them, right? Whereas without the free will component you could identify the thing that caused bad behavior and fix it.

    Now, I'm not convinced that those are logical arguments, but that's the thinking that causes the political opposition, I think...

  4. Here is a defense of free will that meets your arguments:

  5. Hi, one comment on semantics:

    First of all, "will" is motivated - motivation is not free but dependent (on goals or on the past, which basically are the same thing since goal is semantically defined .. by one's past).

    Secondly, "free" is meaningless; ask "free from what?", "free to what?". Making a choice.. could you really have chosen the other alternative?

    Thirdly, an interesting exercise: Try, just for one day, to open doors using only your left hand (assuming you are right-handed). Are you free to do that? It requires to be conscious of your "robotic brain" - permanently - and monitor it. Good luck with that guy!

    I do not think it is even needed to look into physics. Even if the "spirit" does not come out of physics, the result is the same.

    So, eventually, I think all this debate is just confusion; I have to do "as if" I had free will (make my choices and decisions); and make sure the illusion of free will does not has me.


  6. I do not find "emergent" notions of free will at all intellectually dishonest. Though I don't usually think in terms of emergence, rather I am materialist enough to beleive that I am a confection living meats, and so what matters is whether this meat can make choices.

    Let's agree that physical reality is made entirely from some mix of the deterministic and the random (for short, lets call the combination "mechanistic"). Then we will not find non-mechanistic will in this universe.

    We also agree that a common mistake is to think that "mechanistic" implies "not making decisions or ... choices." Since that implication is false, we are free (ha!) to define "free will" to mean "choice-making will" or to mean "non-mechanistic will". I choose the former since choice-making seems much more relevant to freedom than does "non-mechanism".

    Now of course we are merely arguing about the definition of a phrase here, so the important thing is not what definitions we prefer, but that we remain constent when using them in arguments. So I have to agree that (my style of) free will is completely compatible with superdeterminism in quantum mechanics.

  7. "I wish people would stop insisting they have free will."

    How could they, if they have no free will! :)

  8. ppnl,

    Experience is a somewhat different issue than consciousness. I've been thinking about this recently, but it's a longer story, so maybe just let me say this. How do you know that the human "experience" of something is different from the experience a robot would have. What do you even mean by "robot". Do you distinguish between robots and humans by the material they are made of?

    Any "experience" is merely one particular reaction to some stimulus. A lot of the confusion about subjective experience (all this talk about qualia and so on) comes from mixing up the reaction to the stimulus (in person number A) with the reaction to the stimulus of watching the reaction (in person number B), which arguably isn't the same thing.

    Sure, if you could record some "experience" you could look at the pattern and then detect this experience in your test subject if it occurs again. I don't know why you think this can't be done. In fact this is pretty much what a lot of experiments do these days if they record people's brain activity in response to some stimulus. Or if they just ask people what they "experience" if you poke some part of the brain (as is often done in brain surgery).



  9. Unknown,

    Let me put it this way. "You" are some subsystem of the universe. If disabling this subsystem from executing a function it is programmed to perform removes a problem, then I would call this subsystem "responsible" for having caused the problem.

    It doesn't really matter though whether you agree on this use of the word "responsibility." The only thing that matters is that you will be facing consequences, regardless of whether you had free will. That, really, is what the test subjects should be told.

    As we come to know more about how the human brain works, it might well be that the consequences people face for wrongdoing will differ from the ones they face right now. For some crimes, for example, it seems unlikely that putting people in jail makes it less likely they will repeat the offence. Instead, it might make it more likely because it reinforces a social stigma. I tend to think that overemphasizing people's "free will" in their decisions does not improve how we deal with the situation. Best,


  10. Lauri,

    That might possibly be. I don't know much about the difference between the US American and European punitive system. It is true that most of the opposition I have gotten in reaction to my writing about (the absence of) free will comes from Americans, but then most of this blog's readers are from the US/Canada (based on stats from Google analytics), so this doesn't really mean much.

  11. If people don't have free will, how can they choose to bury the idea?

  12. Quentin,

    No, sorry, it doesn't meet my demands. It just construes up an ill-defined notion of "a-posteriori coherence" that seems to be, essentially, the judgment by some other person on whether an agent's action was "intentional". I don't see what one learns from this.

    1. The point is that intentionality and fundamental unpredictability are the two essential aspects of what we mean by (incompatibilist) free will, so in order to know if free will exists we should look for intentional and unpredictable behaviour. I see no evidence from physics that it does not exist.

  13. Sorry, could you please explain the last part a little more? "Without free will, Bell’s theorem doesn’t hold, and all we have learned from it goes out the window." But you argue against free will, therefore against Bell's theorem? And are determinism and superdeterminism equivalent?

    Your statement that there are only two options, determinism and randomness, seems close to what Bob Doyle (an astrophysicist, at least by training) calls the standard argument against free will. He makes an interesting case against it, which may be worth a look even if you're right that the best present physics doesn't leave room for free will.

  14. My free will is not described and is not obliged to be described with your sorry equations. It is not about laws of nature which are repeatable.

  15. P aom,

    The question what consciousness is doesn't play any role for this argument. If you bring in consciousness, you only get more ways to lose free will (because you might make decisions "unconsciously" without being aware of it), but you can never get back something that wasn't there to begin with.

    If you intended to insult me with your reference to Sean, you failed. Sean is a smart guy who has clearly spent some time thinking about these matters. I am looking forward to reading his new book.

  16. Jim,

    Thanks for the reference. I think I looked at this before, but I seem to have forgotten about it. Will look at it again.

    Regarding Bell's theorem. In the derivation you have to assume that the detector settings are not correlated with the state you want to measure, which means that the experimenters must be able to make "free" decisions. Take away free will and this assumption is no longer reasonable. Superdeterminism is either non-local or has a backwards causation (in which case it allows for local hidden variables, which would violate Bell's theorem if it was applicable). I don't actually know why it's called that way, sorry.

    1. The free will theorem is misnamed. As you rightly say, what is needed is not free will, but causal independence (no correlation). Pure randomness would do the job. Now you argue that randomness is not a suitable basis for free will, and then you argue that we need to assume free will to derive Bell's theorem (where what is actually needed is randomness): isn't it contradictory?

  17. I completely agree with you about free will. However I do not share your concerns about the relation with quantum physics. I have not discussed free will with any colleagues, because it did not seem relevant, so I do not have any data on what most physicist thing about this. However, I have not met anyone who have confessed in beveling in the traditional Copenhagen interpretation, where the collapse is triggered by free will or consciousnesses. I have only seen mentions about free will and/or consciousnesses in popular science descriptions, and those gets so much other things wrong so it did not bother me. And mostly it is consciousness rather than free will that is mentioned. Consciousness obviously exists, but since it is an emergent phenomena, it is not going to cause quantum wave collapses ether. If it did it would really mess upp the chemical processes of our brain.

    I have noticed, by my very limited observation, that interpretation of quantum mechanics is largely divided by generation. Most older professors believe in some kind of collapse, usually triggered by some vague notion of interaction between micro and macro system. I also heard ideas of how this would happen gradually. To me the notion that micro-macro interaction would trigger a collapse is weird since is all one physics, but at least there are no free will involved.

    Most Phd-students and younger post-docs believe in a collapse free interpretation, e.g. Everett's many world or Bohmian Mechanics. These non-collapse models does not need anything similar to the collapse trigger.
    (Many world interpretation does _not_ just replace the collapse with a split. This is an unfortunate miss understanding due to bad popular science explanations.)

    This generational divide is of course not exact and could even be just a statistical accident of whom I have been talking to. But if it is at least semi-correct, it might explain why you have a very different experience.

    You write:
    "This option of giving up free will in quantum mechanics goes under the name “superdeterminism” and is exceedingly unpopular."
    I have never heard of superdeterminism, but I will look it up. However, I know that it is not the only interpretation of quantum physics that does not need free will.

  18. Let’s be clear: the „fundamental laws of nature“, mentioned in the article, sometimes with the apposition „that appear in contemporary theories“, or „to our best present understanding“, are not irrevocable laws (as the word fundamental suggests).

    They are parts of theories, of very well proven theories in certain set-ups, but not of closed and unalterable theories.

  19. Hi Sabine. Very good and well thought out article. I have been wrestling with this as well for the past 60 years or so. A couple of points, if I may: one is that you have not clearly defined free will in a way that is self consistent, non-circular and does not use the words "is like". If we are to discuss something and use either rational thought or mathematical derivation, we must first define it. Then, if we are able to determine its properties, we can do experiments on it; another is to question why you wrote this article or why it should concern you -- did you have any choice?

    Apart from that, I agree with the conclusion you make regarding indeterminism or unpredictability; the underlying mathematics can easily be deterministic without allowing us to be able to measure or detect it. Nevertheless, if the field is not conserved, then there may be a bit of a problem with your argument. Schrodinger's equation and Maxwell's equations describe conservative systems, but the Einstein field equations contain non-conservative systems, particularly in the third ordered time differential of mass, or m-tripple-dot. I believe you would be familiar with this. If there is some non-conservatism within gravity, does that impact free will?

  20. Gerhard,

    This is entirely correct, which is why I am formulating this so carefully. I am not saying free will doesn't exist. I am just saying that according to our best present knowledge it doesn't exist. This also means though that if you want to insist it does exist, the proof is on you to show how it can be compatible with the laws of nature that we know already.

  21. Bruce,

    This is incorrect. I don't have to define what free will is to demonstrate it doesn't exist. I merely have to demonstrate that at least one necessary assumption for free will cannot be fulfilled. And that's exactly what I have done. I have been very careful spelling out all definitions in this paper.

    I don't really know what you mean with fields being not conserved or higher time derivatives of what I assume is a mass. Let me put it this way. All that matters for the issue of determinism is whether there is a well-defined initial value problem and a Hamiltonian evolution. For all we know this is the case for all existing theories. Best,


  22. "Without free will, Bell's theorem doesn't hold ..." IMO: without free will, Bell's argument for Bell's theorem doesn't hold. However, even if Bell is wrong, I doubt that empirical evidence can show that Bell is wrong. I also doubt that the Copenhagen Interpretation can be refuted. My argument would be: formulate string theory within the Copenhagen Interpretation and then use clever D-brane adjustments to provide models of any plausible (or implausible) physics. It seems to me that Bell's theorem and the Copenhagen Interpretation are sufficiently ambiguous so as to allow sophistry in the form of clever mathematics to predict anything whatsoever.

  23. Sabine Hossenfelder,

    "Experience is a somewhat different issue than consciousness."

    No I don't think it is a different issue. Consciousness is composed of experiences. Without the "qualia" that compose the fabric of consciousness there is nothing to be conscious of. We experience the very process of our thoughts and that is what leads to (The illusion of?) free will. Without the consciousness of the redness of red, the beauty of music or the calculations and planning of our next chess move the idea of free will just dissolves.

    For example our brains do an automatic calculation to tell out hearts how fast to beat. We are not aware of that calculation and so we do not attribute free will to the result. But if I am playing chess I experience that calculation and it feels like planning. Thus it feels like I have a choice on what move to make. One calculation may not be more complex or involve more "planning" than the other and in a deterministic universe they are indistinguishable.

    "How do you know that the human "experience" of something is different from the experience a robot would have. What do you even mean by "robot". Do you distinguish between robots and humans by the material they are made of?"

    Yeah the problem is that I cannot distinguish robots from humans. Worse I cannot distinguish between humans and a natural process like the weather. Does the weather have experiences? I don't know. The best I can say is that it seems to make no difference if it has experiences or not. I cannot say if a robot has experiences or not. The best I can say is that it's experiences is no part of understanding how the robot works. It cannot matter if it has experiences.

    But I can say that I do have experiences. Does it matter in understanding how I work? Maybe not but if the experience of qualia make no observable difference then how can I talk about them?

    "Any "experience" is merely one particular reaction to some stimulus."

    What do you mean "reaction"? If I drop a shoe it falls. It does not as far as I know "want" to fall. That is just how physics causes it to react. If I stub my toe I react to limit the damage. It is reasonable for a brain evolved to survive to react that way. But I also feel pain. That fact is neither derivable from physical theory nor necessary for the physical reaction.

    "A lot of the confusion about subjective experience (all this talk about qualia and so on) comes from mixing up the reaction to the stimulus (in person number A) with the reaction to the stimulus of watching the reaction (in person number B), which arguably isn't the same thing."

    No the problem is subjectively I experience things even if you don't. That fact seems neither necessary nor explainable. The mystery exists in my head and does not involve external people "A" and "B".

    "Sure, if you could record some "experience"... "

    You can only record objective events. You cannot record the subjective. I can record the weather but I cannot say if it experiences qualia. I can record you with the same result.

  24. A very interesting article Sabine addressing a question you have obviously thought a lot about. However, I do have problems with this passage:

    "It doesn’t matter ... insist that we don’t know how consciousness really works (true but irrelevant). It doesn’t change a thing about this very basic observation: there isn’t any known law of nature that lets you meaningfully speak of “free will”.

    The fact is not just that we dont know "how consciousness really works", the problem is much deeper than that: we dont have a flipping clue how electrical impulses in neurones can give rise to the "person in my head" that is consciousness. Now don't get me wrong - I'm not saying that there isnt an answer to this question, but simply that it has so far eluded us so dramatically that it should give us pause for thought when considering the issue of consciousness - and free will, which is obviously intimately conected to it (It's the person in my head who I think has this "free will").

    So, what I am saying is this: We have a phenomenon that we all know exists - consciousness (indeed strictly speaking it is the ONLY thing that each of really knows exists), but we have no idea at all what the basis for its existence is. Should we not therefore be a little cautious about decalring that the child of consciousness (free will) also does not exist just because we also cant think of a basis for its existence?

    All your statements about how the physical laws we know at present leave no room for free will could all be equally applied to consciousness. And yet it exists.

    I think this calls for "fascinated humility" (and of course some serious research). Maybe one day we will crack consciousness. Maybe our work on AI will help show us what is going on, but right now we would be wise to admit that both cosnsciousness and free will are beyond us - but as we know one of them exists we should be careful about denying the possible existence of the other.

  25. Linda linsefors,

    I don't think the Copenhagen interpretation suggests consciousness causes collapse. And I would be very surprised if Everett's many world or Bohmian Mechanics was very popular among any age group. I think Wigner started out believing in consciousness caused collapse but he dropped it he understood how fragile quantum states were with any kind of environmental interaction.

    Also the "...vague notion of interaction between micro and macro system." sounds like it is referring to decoherence. If so there is nothing vague about it. Once I understood decoherence the quantum measurement problem just didn't trouble me anymore.

  26. Our best present understanding, does not understand everything. I think the known unknowns and the unknown unknowns, when it comes to the entire nature of reality, are significant enough to not justify discounting something that many intuit in their day to day experience of waking life.

    As a layman, I respect a scientists desire to avoid the subject, but for the sake of progress and humanity, I hope we don't generally stop talking and thinking about free will.

  27. But without a definition or or at least clearly specifying which assumption cannot be fulfilled this article become completely irrelevant to a lay person reading it. By imploring people not to make bad choices based on the knowledge that free will doesn't exist, you seem to be conceding that people have the ability to make choices based information. If we have the ability to make choices then what part of free will are we missing?

    I realize that to some degree I am putting words in your mouth and possibly setting up a straw man argument, but really I've only put down my understanding of inconsistencies in your argument based on what you've written because I feel like you haven't given enough information to define more precisely what you are referring to when you use the term "free will".

  28. Sabine,

    If you are not willing to give us a definition of free will, could you at least please explain what are the necessary assumptions for free will according to your argumentation?

    Thank you,

  29. ppnl,

    "You can only record objective events. You cannot record the subjective."

    Sorry, but you just aren't making any sense to me. Experience is some state of the human brain. With a suitable device, you can measure this state. If you measure this state, of course you (the one doing the measurement) do not have yourself, subjectively, the same experience. But you can very well say that it is a state of the brain, consequently the brain is, subjectively, having this experience.

    I don't know what it is even supposed to mean that something 'experiences qualia'. There isn't any such thing as 'qualia'. There are states of your brain. Some of these you call 'experience'. If we'd know more about consciousness, we might be able to tell which these states are. We don't know yet, but one day we probably will.

  30. Quentin,

    If you still think that there are processes which are "intentional but unpredictable" and compatible with the currently known laws of nature, then you didn't understand what I explained. The only unpredictable processes we know are random. And these aren't intentional qua definition. If you could influence them they wouldn't be random.

    In the essay that you refer to the author is trying to get around this by letting a person judge on whether something was "intentional", which (as I said above) is as ill-defined as it gets.

    Try writing it down. I really encourage you. Take a pen, take a paper, try writing down anything.

    1. "The only unpredictable processes we know are random"

      I'm not sure I get the difference between "unpredictable in principle" and "random". To me it's the same for all practical purpose. The difference is metaphysical, not empirical. (Your free will function is supposed to be a counterexample: unpredictable yet not random, but as I explained in our brief twitter discussion I'm not at all convinced it works since the function would be empirically indiscernible from a random process).

      So basically you're saying that there are no "intentional but unpredictable" processes by definition (or because of your metaphysical views on unpredictability in contemporary physics), not as a matter of pure physics.

      But I think that something can be unpredictable (in principle=~"random") for A and intentional for B and that there is no contradiction in terms.

  31. Brett,

    I have stated which assumptions cannot be fulfilled in the third paragraph: If a process is deterministic, the outcome is already fixed by the initial conditions at an earlier time, consequently it isn't "free". If it is random, there is no agency to it, consequently there is no "will". In neither case do you have something like "free will" unless you want to use this word in a meaningless manner. These are the only two options we know of.

    I have explained that in more detail here. I hope you will understand that it doesn't make much sense putting too much detail into a blogpost, the only consequence would be that nobody reads it.

  32. Martin,

    I have stated this explicitly here, and it is implicitly stated in the above blogpost, third paragraph. To have free will your actions should not already have been determined by a state (of the universe, if you wish) at an earlier time because then your action is not "free". Neither should your actions be random because then you don't have agency and no "will".

    To have free will in a meaningful way you need to come up with a law that avoids both types of constraints. As I said, try it. Seriously: I would love to see an example, preferably a better one than I have been able to come up with.

  33. "chaos (which is deterministic)" Planck scale is deterministic? Ocean surfaces, noise, average overall flat given land. The Southern Ocean only has the Drake Passage, then 30 meter rogue waves elsewhere. The Dirac sea has cosmic void and black holes in which to fluctuate.

    "free will skepticism is dangerous " Absent free will, nothing can be willfully altered. Free will evolves "hope." Societies that Korporate Kulture sit on their asses get what they've earned.

    "Professor of Excellence in Marketing " Diversity hire. No economist ever grew wealthy from investment. Professors of Excellence in Marketing clearly cannot move themselves as product.

  34. I believe you're wrong. Study Conway and Kochen's 'free will theorem' and 'strong free will theorem'. Even though they don't prove that free will exists, they have proven that - in essence - "if human’s have free will, then so do particles." In other words: if particles do not have free will (and this is not the same as randomness), then as a consequence human's don't have free will, not even a tiny bit.

  35. Martin,

    I have read the paper and I commented on it my paper. The sensible consequence to draw from this "free will theorem" is of course that neither particles nor humans have free will. I don't know why you believe their argument implies I am wrong. The very opposite is the case, it supports my argument. Do you really want to argue that particles have free will? Seriously?

  36. Dear Bee,

    Prediction via mathematical laws goes from physics into the theory of computation; one has to make sure there are no loopholes introduced there.

    E.g., it would be interesting if it were possible to prove that (barring a multiverse) you are the unique physical device capable of embodying the computations necessary to predict your behavior.

  37. Sabine, the concept of free will is metaphysical, at this time. Asserting free will is restricted by gravity, the strong and weak force, and electromagnetism is far too restricting. If free will exists, it involves the quantum mechanics of spacetime itself. The observer, in a timeless setting, can measure tachyonic states. Such tachyonic states will undergo tachyon condensation, and thus give rise to bosons for the familiar four forces, as well as matter we experience temporally. But a priori, there is no arrow of time, and no geometry, as tachyon condensation has not yet occurred.

  38. The choice of observable to be measured isn't dictated by any deterministic or probabilistic law in quantum mechanics. And two different such choices mean two different states of the universe after the measurement, regardless the outcome of this measurement, even if the initial state of the universe was the same before the choice was made.

    Doesn't this simple fact leave a door open for free will, at least at our current level of understanding of physics? Of course, new insights could close this door, e.g. if a widely accepted way is found to derive the projection postulate from the other postulates of quantum mechancis, or if some deeper theory is found. Then we might be able to forget about measurements, and choices of observables to measure, as fundamental ingredients in the scientific world view.

  39. Lol.

    We don't what consciousness is, have no detailed clue as to how the brain operates, and how the two are related.

    In this regard, claims about free will are nothing more than opinions and prejudices.

    An analogy I recall is writing down Dirac's equation and then claiming that the rest of chemistry and biology are just details. Emergent notions of free will are no more "intellectually dishonest" than are emergent notions of superconductivity.
    Especially given that the have no idea what consciousness is and how it operates.

    Never mind a higher order process such as consciousness and free will, we are not even close to having a physical theory of life: dynamic self-organizing and self-reproducing non-linear systems far from thermodynamic equilibrium.

    Anyways, there is a long tradition of physicists opining about things about which they known less than zero. If nothing else, you don't lack for illustrious company.

  40. Free will is a simply a belief or axiom that among other axioms, you simply must have to anchor your consciousness. Whether illusory or not, free will simply means that you have personal responsibility for the choices that you make. Just because circumstance determines many choices that you make does not change your personal responsibility for those choices and therefore your free will. You seem to suggest that random choices based on uncertainty are not examples of your free will, but since you are also personally responsible for the random choices that you make, random choices are also examples of free will.

  41. Thanks for the response Sabine. I understand the limits of the blog post format but I still have a question about the seeming contradiction of trying to convince people to respond in a certain way if our responses are necessarily either deterministic or random.

  42. It's a bit strange to be talking about "free will" without at least mentioning the words "libertarian" or "compatibilist." Most philosophers don't think libertarian free will is tenable (which is what you seem to be talking about). As for superdeterminism, my understanding is that superdietermism isn't popular because of a belief in free will but rather because even if you reject free will the argument still requires a sort of improbable conspiracy rather than the mere existence of behaviors that are correlated in some complicated way.

  43. Sabine - I think you are mistaking ppnl's point. Yes, experience is a state of a brain (or probably also the state of a robot or program or thermostat). There are only particles and fields and charges and momenta, etc. That's the outside view. But we all know that there's also the inside view - ppnl's musical note and pain of a stubbed toe. We don't know (and many people think we'll never know) how the phenomena we observe from the outside entail what we experience from the inside. But as a factual matter, the inside is incontrovertible, as Descartes and perhaps Thales knew (even if they got everything else wrong). Regarding free will, then, we are in the same outside/inside situation with respect to causation of our actions. From the outside, there is the physics/chemistry/biology of neurons, etc., or with a robot of electronically stored 1's, 0's, currents and state transitions. From the inside, there is our experience of doing X in response to Y - you say "lift your arm" and I either do so or I don't. The outside view is a bunch of neural firings, all more or less deterministic, with, as you say, a side of quantum randomness. From the inside there is what I choose to do or not do. Just as from the outside, the plucking of a guitar string causes vibration of my ear and firing of auditory nerve cells, and from the inside I experience the sound of "D".

    We use a different language to talk about the phenomenon from the outside and from the inside. To say there is no free will because everything is determined by physical laws is a category error.

  44. You indicate that you’re familiar with Compatibilism, but reject it for reasons I don’t see. You agree with Compatibilists on the core point that free acts, far from being uncaused or random, are acts that have the right kinds of causes, arranged in the right sequences. As you say, the fact that acts are caused “doesn’t mean either that you are somehow forced to do something you didn’t want to do. There isn’t anything external imposing on you. You are whatever makes the decisions.” This is exactly right!
    Yet you insist on equating an act’s being caused with “absence of free will”, because you take it for granted that the core meaning of “free will” is a metaphysical super-causality—something that might enter into equations of physics. You probably think that the Compatiblist account of freedom and responsibility is a matter of redefining words.
    What I want to tell you is that the Compatibilist account isn’t a redefinition, it’s the correct and comprehensive linguistic analysis. Metaphysical super-causality is not the meaning that enters into actual concepts of freedom and responsibility. Yes, the specific term “free will” has a history tied to the metaphysical confusion. But this doesn’t go for the everyday uses of “free” and “responsible” and related language. To put it another way, the “free will” that may be relevant to Bell’s Theorem is not the “free will” that is relevant in courts or in moral judgments.
    So you are wrong to wage a battle against “free will”, even if some people mistakenly defend it for political reasons. It makes much more sense just to acknowledge that free acts are caused, and thus that those who resort to political arguments are confused by the metaphysics.

  45. I am looking forward to reading J.T. Ismael's upcoming book, HOW PHYSICS MAKES US FREE, which ponders these issues in more detail.

  46. What percentage of scientists or physicists believe in reductionism? I know that most particle physicists believe in reductionism but I thought that the majority of scientists (and possibly even the majority of physicists) do not.

  47. Interesting lay out Bee.

    Just to recap then......

    Determinism erases freewill, and if you are not current with science, you do not have a complete grasp of the science today?


  48. Conway and Kochen say that Super-determinism is incompatible..... Why is not it realized that observation is not found in only one location of the particle localization , rather the particles are found in the probability of distribution ....... So , is not it sufficient for free-will ? What is the answer of information loss of black-hole ? Where does this information go ? How does the Universe get started ? Is not there realized free-will and the necessary of free-will ?How is it justifiable the quantum jump of electron ?
    If we have right technology to detect the Aliens life , then we can predict how they are - Metalaw ....

  49. Great article and completely agree. Also think super-determinism is under appreciated. To me it's one of the only sensible solutions to the measurement problem by saying that the state of the universe and the unitary (which includes the SM + physics we might not know) gives rise to certain outcomes (certainties) in subsystems at particular times. It's just not a psychologically appealing solution.

  50. Sabine Hossenfelder,

    "Sorry, but you just aren't making any sense to me. Experience is some state of the human brain."

    I assume so. I cannot imagine any alternative.

    "With a suitable device, you can measure this state."

    You can measure objective states of the brain and correlate them with objective claims of subjective experience. But you cannot verify those claims of experience. For example there is a neurological condition that causes blindness yet the person is totally unaware that they are blind. They are honestly making claims about of their subjective experience that is simply not true.

    I do not know that you are conscious. All that I can do is make objective observations of your brain state and note that they correlate with my brain states. But any such correlation cannot explain the causality. As far as causality goes it might as well be magic.

    If I have a computer program that deals with color (like photoshop) then objectively there is some bit pattern that is a symbolic stand in for the color red. But the photoshop program does not experience red. I know how to do symbolic stand ins but I don't have a clue how to make photoshop experience color. All programming is or ever can be is establishing complex patterns of symbolic stand ins.

    "I don't know what it is even supposed to mean that something 'experiences qualia'."

    I don't know either and that is the problem.

    "There are states of your brain."

    Yes there are objective states of the brain. Once you know those states and know how a current state causes the next state then it seems like you know all that can be known in principle. Like a computer program one state leads to the next. Knowledge of what is or isn't conscious is not needed or even helpful here. It would seem that there is nothing more that needs explaining.

    Yet I do have subjective experiences and it seems like that needs explaining.

    "Some of these you call 'experience'."

    No, consciousness and experience is not part of the understanding of the causality of the state transitions between the brain states.

  51. I can always commit suicide, quit my job, go on a hunger strike, not get dressed in the morning, or provoke some sort of permanent, debilitating injury from which I will never recover.

    And so, within a layer of social interaction, my willingness to do so, must be provoked, yes?

    Something external must awaken within me the misery or psychological spite to do so? Maybe several degrees of separation might separate me from the true origins of a deterministic provocation, beyond my mortal capacity to intuitively understand such subtle influence. These conditions must presuppose either the anger or the depression necessary to possibly harm myself, in ways I might not "normally" decide in favor of?

    So, all this to say ...that I do not do so proves my incapacity to do so? That I cannot act without motive?

    Hmmm, I might wonder about that.

  52. Thanks for the post. I will repeat my thoughts on the subject in the doubtful case that some reader might gain anything from them.

    On subjective experience: when I smell a rose, I get a specific sensation, and when I smell an orange I get a different specific sensation. If they did not cause any sensation then I could not sense them. The physical explanation involves measurable things such as which set of neurons end up being stimulated via which nervous pathways and chemistry. The subjective sensation itself does not seem important to me. If the sensations for the rose and orange were interchanged everything would work just as well. Things have to produce some sensation to be sensed, and that is how the rose and orange sensations feel in this universe to my combination of particles and fields. (My laptop receiving my keyboard inputs is also receiving some sensation but how it "feels" to my laptop I neither know nor care.)

    On consciousness: I see it as similar to the operating system of a computer, e.g. Windows. It receives external inputs (perceiving them as sensations), gets internal inputs from sets of processing neurons, and makes some high-level decisions (which could include an unpredictable random element, e.g., based on the number of photons hitting retinas and other precise sensory perceptions which could detect quantum-level events - Feynman says in QED that the human eye was the most sensitive photon-detector available in early quantum experiments, capable of detecting a single photon). Most of its feeling of "magic" is I think due to the fact that the results of neuron processes arrive without conscious knowledge of what was done to obtain them, since there are no nerves which monitor neurons. This is like Windows getting a result from Excel and displaying it on a screen, without the Windows program itself being able to do spreadsheet calculations.The human brain's (roughly) 73 billion neurons is more processing power by orders of magnitude than any current super-computer, so it is not surprising that we are a long way from computers which we would regard as conscious, although specific computer programs can beat human experts at chess and Jeopardy.

    On free will: I agree it is a nebulous concept. I see some use for it under the following meaning: I have free will for a particular decision means that I am making the decision without constraint by external agents (no one is holding a gun to my head), and I will be responsible for the results of that decision. E.g.: did you do that of your own free will? This does not mean I disagree with anything in the excellent post, with the possible exception that if as a previous commenter mentioned, Bell's Theorem only requires uncorrelated, random choices, then as I mentioned above I think this is possible under my hypothesis of human consciousness.

  53. I may do a full critique of this if i get time. But summary is "scientific arrogance". You see me (my self) as simply the sum total of all its scientifically analytical components - expressable only in the equations and logic of science.

  54. From any of the hierarchy of transcendental subjectivity, the subject will have a consciousness in relation to its parts while at the same time, is completely mechanistic terms of maternal world, part of which is. Thus, we introduce a new fundamental relativity - relativity of consciousness.

    Indeed, the machines produced by us, are fully determined, then, as we ourselves another opinion. Clearly, there is located a step higher would be on our spirituality is not a better opinion than we have on computers. We know that the puppeteer has consciousness, but whether we have enough reason to deny that the quality of the doll?

  55. I agree randomness does not imply free will (fw), but in case there is wavefunction collapse, one can't reject fw by merely saying that one can't influence collapse.

    Determinism is compatible with fw, as explained in the references mentioned in the abstract of

    I agree that one should not support free will just because some say its absence would make people criminals. Criminal activities are sanctioned not to punish, but to make people avoid them, and the latter works even if there is no fw.

    I disagree people who accept fw have existential worries or hidden political reasons. People don't reject it easily because there is no irrefutable refutation yet.
    "I am afraid the politically correct believe in free will hinders progress on the foundations of physics". I doubt that physicists reject their peers' papers or throw away their own results for not being consistent with free will :)

    "buying into the collapse of the wave-function seems a small price to pay compared to the collapse of civilization". Quantum theorists who believe in collapse do so because they can't make sense of the outcomes of measurements without collapse, not because they want to support fw. Many of them don't even believe in fw. But the reason they don't accept easily alternatives to QM, in particular hidden-variable superdeterministic theories, is simply that standard QM works much better, not because they want to save their free will.

    From my experience, the number of those not accepting wavefunction collapse is quite large, even if we don't count those rejecting it for the wrong reasons (for instance, because they think MWI is unitary, or because they believe that decoherence somehow solves all problems). I agree that it may be possible to make sense of quantum measurements without collapse, and I explain how this is possible with Schrodinger evolution in the next issue of (sorry for the shameless self-advertising, but it is on topic).

  56. Copy/pasting my comment from Lobsters:

    If you're going to use a title this strong, you should have an equally strong argument to go along with it, but there's none to be found here.

    > But there is still either no “freedom” or no “will”. Just try it. Try to write down one equation that does it. Just try it.

    1. Um, [Math is incomplete]('s_incompleteness_theorems) (says Math), so an inability to express something in Mathematical form does not mean that thing doesn't exist.
    2. There are lots of everyday, kindof important things that cannot be expressed in mathematical form (consciousness, [qualia](, to name a few).

    So... really inadequate post on the subject of free will. Only reason it's at the top of Lobsters is because the algorithm gives [undue weight]( to controversy. */me clicks "hide"*

  57. I'm in this camp. Most of my papers have the word "Deterministic", and I've cited 't Hooft on similar quantum structures. My interest is in the "dice rolling" of QM, and the structure of vacuum that enables 'randomness' and 'spontaneity' without resorting to mystery or non-physical agency.

    It might be disheartening to realize that our own freewill is an illusion of a privileged and limited perspective, but it is the only sensible conclusion on a scientific basis. I've attended lectures where I've mentally dismissed any philosophies (especially in QM) that treat the observer as a special case, that require the measurement to be outside the system, or that define a degree of freedom that is not available to the system itself.

    We're all in the system.

  58. "It is highly probable we are missing something really important here, something that could very well be the basis of future technologies."

    You're joking, right?

  59. Thank you Sabine for replying. Yes, you are correct and did define free will as:

    1) An agent in possession of free will is able to perform an action that was
    possible to predict by nobody but the agent itself.

    At this level, there is no argument as to the validity of the definition since the definition is necessary to determine validity itself. The definition is either useful or not and this looks like a very useful definition. It requires two agents, one predicting and the other acting. Both agents are distinct, ie. they are not the same agent. An agent can be a computer, lab rat, grad student or particle. The issue is predictability, or possible predictability given the bounds of the physical sciences. So far, so good.

    Let us say that the physical sciences are described mathematically by EM, gravity and QM. What is left is to determine if these three sets of differential equations give us unique solutions in all cases to the actions of an agent and that the predicting agent has complete knowledge of the initial and boundary conditions as well as complete and unique solutions to these differential equations. In this case, you argue that the universe, or all agents within it, are completely determined. Hopefully, this defines the "problem." At least, that is how I see it. This is, of course, Laplace rising from the dead.

    If I may, I would like to stay away from determinism and stick with absolute predictability. I think that avoids all the muck. The reason for this is that of the differing properties of conservative and non-conservative fields. If a field is path independent, then it is a conservative field. Since QM relies on the Hamiltonian, then it is a conservative field and is within the realm of your argument of being deterministic. Same with EM. But, and this is a big but, gravity is not a conservative field. In other words, there are elements of the Einstein field equations which are path dependent. In other words, two end-points may have some values the same, but also have values that differ depending on the path taken. I think this applies here. The initial and boundary conditions along with the solution only yield predictability if the result is path independent. If path dependency is involved, you're hooped.

    I neither argue for nor against your argument. But if there is a hole to be found, it would emerge in studying gravity waves and information loss re: black holes. That is what I meant by the reference to third ordered differentials. It may be that a resultant position has to deal with something like the quality of the end-point rather than just the value. Almost all values in physics are conserved, but there are some that are not. Do you think this has an application here? This is not a rhetorical question, but an honest one. I would love to hear your opinion.

    As far as an experiment is concerned: let us take about 10,000 lab rats, or grad students as the case may be, and they are taught to push a blue button and get a reward and not to push a red button or they get an electrical shock. Then we remove the reward and punishment and test the rats one at a time to see if they all press the blue or red button. If any of those rats, or grad students, rebel and push the red button, he or she has free will. Does this meet your criteria for testing the existence of free will?

    Thank you so much for your time and thank you for having the courage to put forward your thoughts and your conclusions.



  60. Dear Mrs.B,
    I will rise to your bait. Definitions are always helpful.
    Determinism : the doctrine that all events, including human action, are ultimately determined by causes regarded as external to the will.
    Free Will :the power of acting without the constraint of necessity or fate.

    In your field the Bekenstein-Hawking Entropy links the smallest unit of Information to the Planck Area on the surface of a Black Hole. For me, a potential candidate for experimental evidence for a geometric reconciliation of quantum effects and gravity, and my reason for reading your blog. The “Bekenstein Bound” is a general statement of this relation and is at the heart of my Discipline where the work of Claude E. Shannon on Information and Entropy was seminal. None of this would be possible without the work of Boltzmann(1844-1906). I would like to remind you, gently, of the Fate of this underrated and tragic physicist. He was constantly in conflict with the “Traditional Authorities” in the Institutions responsible for accessing his work. He was a cultural darwinist. It is said that, unable to reconcile his insights with the recalcitrance and Philosophies of the “Establishment”, he killed himself while in a period of Depression. This is the ultimate expression of Free Will by an animate. Biology may be late to the table as a science, but its findings must be respected. As a “cultural darwinist” I too believe that Mathematics and knowledge of the “Natural Sciences” are cultural adjuncts to the “awareness” and “communication” of an evolved being and only have “meaning” in this context. Whereas the “Natural World” may have existence without us, there is no evidence that is aware of us or our tools. Boltzmann's “statistical mechanics” have lead to the equivalence of information and entropy and the ability to express the thermodynamics of inanimate processes. Is this what you mean by “Determinism” in the “Natural World” or are you misappropriating someone else’s terminology?
    Using Huxley’s terms, thanks to Darwin and the Biologists we have have a fair understanding of “Biogenesis”. The chemistry of “Abiogenesis” still eludes us. In this context may I, not so gently, remind you that Evolution is not “aware” or “rational”. It is a blind, uncaring exploration and exploitation of the resources within an environment by animates and the end product is Homeostasis. An equilibrium is state within a process . Are you asserting that this process is determinist? If so, and to steal from one of your fellow bloggers, “you are not even wrong”. Even Asimov’s science fiction hero Hari Seldon wasn’t prepared to let “Psychohistory” predict the future of an individual. Aggregate behaviour considerations always have to postulate the effects of information propagation. In my field of interest, Random Autonomous Boolean Networks, the dynamics within the systems are totally dependant on initial conditions and the nature of the event propagation. Here it is becoming more and more apparent that we do not even have the basic mathematical concepts to express process correctly. When we add perception, i.e how and when the individual animate becomes aware of choices , then I have to ask you where is there a common term for your contraposition?

  61. Brett,

    I don't understand your comment, sorry. Who is "trying to convince people to respond in a certain way if our responses are necessarily either deterministic or random" and what does this have to do with my blogpost?

  62. Jonathan,

    I don't know, but let me put it this way. I know only one physicist who does not believe in reductionism. The reason being that there doesn't presently exist any consistent theory that gets around reductionism.

    This really is the point that I don't understand. Look at this comment thread. The vast majority of people seem to believe there *must* be something wrong with this argument. Maybe there is. But given all that we presently know, the opinion I am representing is the conservative one. There isn't any other coherent explanation. I am just drawing conclusions from what is accepted knowledge. The amount of denial in this comment thread is astonishing.

    (And no, btw, a failure of effective field theory doesn't change anything about this. It merely means that you can't cleanly separate small scales from large scales as is normally assumed.)

  63. George,

    Political correctness is the only explanation I have found for this utterly strange phenomenon. The amount of wishful thinking in the literature on the topic is stunning. People literally start with the assumption that free will must exist, consequently we must somehow find an argument according to which it does exist, basically discarding what is the obvious conclusion. Tell me any other explanation for why the literature focuses so one-sidedly on what I would sum up as "desperately trying to find free will".

    I didn't say that they are "intellectually dishonest" - I think they're plainly cognitively biased. I said that *for me* denying that I see what I see would amount to intellectual dishonesty. (Read the relevant sentence in my blogpost again.)

  64. JimV,

    >On subjective experience: when I smell a rose, I get a specific sensation, and when I smell an orange I get a different specific sensation. If they did not cause any sensation then I could not sense them. The physical explanation involves measurable things such as which set of neurons end up being stimulated via which nervous pathways and chemistry.<

    Yes of course. I could build a robot that could smell roses and be attracted to roses. The physical explanation involves measurable things such as which set of electrical pathways are end up being activated by which control circuits. But does it experience the smell? Does it need to in order to function? Once you understand the physical deterministic process there is no need for consciousness nor room for it.

    Yet I do have experiences. I could be bounded in a robot and count myself a king of absolute determinism, were it not that I have experiences.

  65. Rastus,

    No, it's not a joke. You can't build a quantum computer without understanding how quantum mechanics works. I have expressed previously that I think quantum mechanics isn't fundamental. Consequently, I think that studying the matter is relevant for future technologies.

  66. itistoday,

    Your comments are just vague distractions.

    That a sufficiently complex set of axioms allows you to formulate undecidable questions doesn't change anything about the fact that we have extremely well confirmed laws of nature with the properties I listed above. It is very difficult to come up with undecidable questions that can be operationally constructed within these theories. Here is a recent example, but the interpretation is controversial. And in any case, it's irrelevant to the point because you can always remove this ambiguity within a given set of axioms by extending it.

    As to your second point... There isn't any such thing as "qualia". It's a modern version of the "élan vital". Give me a definition "qualia" that isn't mystical humbug (and see above commenters fail).

  67. Dear Sabine,

    Regarding superdeterminism. How it's supposed to work at all?

    I mean one need not only to get the violations of the Bell inequalities (that's easy to imagine how that can happen) One also need to explain why the experiments also didn't disprove quantum mechanics itself! Why their results are not just random mess?

    Different people in different locations at different times perfomed a lot of the experiments with different equipment. It's pretty obvious that in different cases the decisions depended on huge amount of the variables differently. So take experiments with EPR states the result I would expect is some arbitrary correlation violating Bell inequalities but not the specific one of QM. Or in GHZ experiment you should obtain -1 only if you are lucky but some arbitrary number in general.

    However for some magical reason in all of the experiments up to this point this dependence worked in such a way that the result agree with the predictions of the quantum mechanics. Is there any superdeterminist model (even if very simplified and unviable in the real world) that tackles this issue or all that is just talking?

  68. ppnl,

    Well, we seem to agree at least that there isn't any definition of "qualia". My conclusion from this is simply that it doesn't exist. If nobody can tell me what it is, what else am I supposed to conclude?

    You write "You can measure objective states of the brain and correlate them with objective claims of subjective experience. But you cannot verify those claims of experience. For example there is a neurological condition that causes blindness yet the person is totally unaware that they are blind. They are honestly making claims about of their subjective experience that is simply not true."

    This is a circular argument. In saying that I cannot "verify those claims" you are assuming that there is something about experience that is not measurable.

    I don't know what your example with the blind person believing that they see is supposed to tell. It seems to be saying that there are ways to have "experiences" that do not give the person a good representation of external reality, which is correct, but I don't understand how it's relevant. Best,


  69. Hi,
    Perhaps I missed it, but I did not see in the comments so far discussion or mention of how 'free will' can be accommodated within retro-casual models of quantum mechanics. In your 'free will function' paper you did mention 'backward deterministic evolution', but I suppose the addition of this has no impact on the question in your view. However, (and I guess you are aware), there is school based around the work of Aharonov et. al., which has been discussing this and related problems for some years. E.g., a recent popular article on Retro-casuality and 'free will' is here:
    I would be interested to hear your comments on their way of posing the problem and the possibility of a solution.

  70. CGT,

    Aha, how very observant :o) Indeed, I think that the two issues (retro-causality/backward determinism) and free will might be related... The superdeterministic model that I'd been working on is actually a type of retro-causality. I haven't looked at the recent work that you mention, but it's on my reading list. The last time I looked at it it seemed a reformulation of quantum mechanics, so naturally that leaves me skeptical one can actually learn something new from it that isn't merely a reinterpretation. Best,


  71. Linda,

    I think this is a misunderstanding. When I wrote that quantum mechanics needs free will, I was not referring to the wavefunction collapse or decoherence but to the detector settings. I think we all agree that free will or consciousness doesn't play a role for the act of measurement.

    It is only if you believe that the experimentalists have the freedom to chose the detector settings that you can conclude the underlying law must have been non-classical, ie local hidden variables theories are ruled out. Superdeterminism merely means that they don't have this freedom. It doesn't a priori tell you anything more than that. This could be non-local influence or it could be a backward causation (indeed the two aren't easily to distinguish). The main point is though that it is deterministic. Ie, giving up free will brings you back the option of having a deterministic time evolution.

    Now if you follow this line of thought a little further you might see why I think it's interesting. First, for what I am concerned local hidden variables are the most obvious explanation for quantum effects, and once you give up on free will this is a possible explanation. Maybe more interestingly, it means that measurement outcomes are in principle predictable (though that might be in practice difficult). If they are predictable, you can use it for an effectively superluminal messaging. I say "effective" because there isn't actually anything traveling superluminally, you're just exploiting a correlation that was already present that you didn't know of.

    I know it sounds wild, but give it a try, maybe you'll come to like it ;) Best,


  72. Orin,

    It is correct that superdeterminism is unfortunately often discarded as a "conspiracy theory". The reason is that superdeterminism is often assumed to belong to a classical, realist, formulation of quantum mechanics. In principle, since Bell's theorem doesn't apply, you could try to go for this. But this then brings up the question why, if nature is fundamentally classical, does it appear as if it is quantum. I don't know. But I don't subscribe to realism. Complex wavefunctions and Hilbert-spaces are all fine by me. I just want a deterministic 'collapse' process. Best,


  73. Sabine Hossenfelder,

    > Well, we seem to agree at least that there isn't any definition of "qualia". My conclusion from this is simply that it doesn't exist. If nobody can tell me what it is, what else am I supposed to conclude? <

    That seems like a reasonable conclusion. The problem is I could argue that consciousness does not exist on the same basis. Once you understand the deterministic mechanism of the brain there is no need for conscious experience nor any room for it.

    Except subjectively I do have conscious experiences.

    A question from before that you never answered - does the weather have conscious experiences? How can you know? You can make objective measurements of weather phenomena but how can you use them to decide if the weather is or is not having experiences?

    Ok consider my version of Searle's Chinese room - and I really hate Searle. I can barely stand to read what the man wrote.

    Say you take a dog and torture it with an electric prod. It experiences pain. It experiences so much pain that you would be breaking the law in any civilized part of the world.

    Say you recorded a movie of the event. Would replaying the movie recreate the conscious experience of pain? Most people would say no.

    But lets say your recording was so detailed that it preserved the state of every neuron and nerve cell on a picosecond level. Observing that recording can tell you all that can be objectively known about the brain states of the dog. But does replaying the movie recreate the conscious experience of pain? Again most people would say no.

    Ok the detailed recording uses an inconvenient amount of storage capacity. You note that the Kolmogrove complexity of the information is low. You should be able to produce a relatively short program that produces the long movie as output. You note that there are transition rules that govern how the next state of the neurons in the dogs brain result from the previous state. You use the initial state of the dogs brain and the transition rules to produce a program that calculates the next state. Does running the program recreate the pain? It is still just decompressing and displaying a recorded movie isn't it? Yet viewed another way you have recreated the dogs brain in software.

    You can reduce it another level. Conway's game of life is computationally universal. You can implement a universal Turing machine in it and run the data compression program on it. Looking at the output all you see is a vast pattern of black and white cells blinking in a complex pattern. Does it experience pain?

    You can record objective information about a system be it a brain, the weather or a computer program. But you cannot extract knowledge of subjective experience from objective data. The easiest thing to do is reject consciousness itself as nonsense that has no explanatory power nor theoretical mechanism.

    Except I am conscious of things. I have experiences.

  74. Sabine, today you are a beautiful depiction in art of hubris & Purity:

    "We know she's talking bullshit.....what we don't know is whether it's the right sort of bullshit"

  75. Sabine Hossenfelder,

    If a superdeterministic hidden variable theory is true wouldn't that make quantum computers impossible? After all if you are limited to a finite number of hidden variables then the quantum computer cannot be put into a superposition of an effectively endless number of states.

    Or put another way if quantum computers are built would that invalidate superdeterministic theories? Then they become the best way to test superdeterminism.

    I think Gerard t'Hooft has argued that superdeterminism means quantum computers will not work. But I don't know why or what his version of superdeterminism is.

  76. Since this comment section is suffering from an extraordinary influx of mostly ill-informed, impolite, and entirely superfluous submissions that clog my inbox, I am closing this comment section.