Superdeterminism states that no experiment we perform has complete freedom, so no matter what you are trying to learn about how things work out in nature, nature will tell you what it wants to tell and may or may not tell the truth as per its own decision as it will decide what experiments you perform, what apparatus you use, and what results you would wound up on.

If such a theory were to be proven, would it end the realm of science, as we know? What would be the consequences of such a theory being proven?

  • Proven on which basement? You can argue that even experimental thoughts are conducted in nature. A modern scientific proof ask for falsification experiment, that is, a description of an experimental result which should be interpreted as an invalidation of the theory. Now I suppose that would the theory be considered right, one may still impishly say he can't do otherwise than believing that he is free and do as if the theory was wrong. – psychoslave Jan 18 '14 at 22:39

Anton Zeilinger argues (well: suggests) that superdeterminism would undercut science itself.

[W]e always implicitly assume the freedom of the experimentalist... This fundamental assumption is essential to doing science. If this were not true, then, I suggest, it would make no sense at all to ask nature questions in an experiment, since then nature could determine what our questions are, and that could guide our questions such that we arrive at a false picture of nature.

Source (not verified): A. Zeilinger, Dance of the Photons, Farrar, Straus and Giroux, New York, 2010, p. 266. I copied the quote from Wikipedia, but I think it is not 100% accurately quoted there, at least not according to quoting conventions used by me.

That looks a lot like your question, right? :)

You are in luck, Gerard 't Hooft (not a fan of no-go theorems in general; apparently one of the few notable advocates of investigating this loophole to Bell's theorem more seriously; and a Nobel laureate in the relevant field) responds in a comment on Physics SE (so this possibly doesn't quite count as a published statement).

I just don't agree with the Zeilinger quote. Determinism indeed implies that the experimenter's decisions, and questions, are generated by physical forces themselves, so his attitude would dismiss determinism categorically, and I am not ready to go that far. And my bottom line remains to be a simple one: I now have models telling me what might happen, and what they say does not disturb me. Important: I still keep causality intact.

Source: Physics SE. I suggest also reading the related answer by 't Hooft.

Here is a more recent interview with 't Hooft on superdeterminism.

GM: Most people can accept that our experimental decisions are determined, but the degree of freedom that determine them are usually taken as independent from the degrees of freedom of the system we’re studying.

GtH: Then you’re stuck not only with Bell’s inequalities, but more generally with the whole quantum picture of reality. So, I think you have to assume that Bob has made a decision not out of free will, but by some predetermined correlation.

In quantum physics, there’s a notion of counterfactual measurement. You measure what happens if I put the polarizer this way, and then you ask, what if I had it that way? In my opinion, that is basically illegal. There’s only one thing you can measure.

And this is his latest paper related to the subject.

Let us emphasize one thing clearly, since ‘super determinism’ raises much suspicion in general: there is no spooky acausality, or ‘retro-causality’, of any sort in the classical description of our models.

My own take on this. 't Hooft's position certainly looks fair to me, although I don't immidiately see the need to improve on quantum mechanics. But I have this nagging feeling that, if "we" ever get there, ultimately, there will turn out to be alternative and equally true ToEs. Some will be superdeterministic and some (perhaps exactly one) won't. That would mean that (non)superdeterminismness can't be considered (for non-philosophical purposes) an intrinsic property of reality. And, as such, it could never be proven that i) superdeterminism is true and ii) nonsuperdeterminism is false, thus answering your question. But that's just me, so you might choose to ignore this last part of my answer.


Such a theory can't be proven since it is not experimentally distinguishable from nonlocal hidden variables.

Such a theory is extraordinarily unlikely because it completely fails to account for the apparent compactness of description (except, apparently, when trying to do quantum physics, whereupon complex rules unfold that prevent us from noticing that we're not operating in another regime). In this sense it is not that unlike Descartes' demon (and worthy of similar treatment).

If such a theory were somehow proved to be true, however, it would still not invalidate science because science is mostly about finding what relationships allow predictive power about things that are to happen in the future; and those predictions have already worked quite well and will continue to work as well as they will work. We assume they will continue to work equally well, but there's no real reason why superdeteriminism would be more likely to make that break than any other theory. If everything suddenly stops working, it would be not because superdeterminism is true, but that superdeterminism is true and there's some bizarre change in causality at some point (assuming causality is still a sensible notion). We could also have fundamental stochasticity as indicated by conventional interpretations of QM and have some weird rule change at some point.

Science isn't invalidated just because we can't control variables as well as we would like (e.g. we can't control what experiments we run). It already corrects for that: if you don't control your variables, your predictions aren't very good. Unless Descartes' demon gets involved, that's enough to allow progress even if the progress is less expansive than one might hope.

  • 1) About your "compactness of description". I think that 't Hooft is aiming for superdeterministic descriptions that are compact. I think. Any thoughts/knowledge to share? (Perhaps to be placed under my answer, for debunking purposes.) 2) I don't see how the hypothetical "everything stops working" is an especially good part of your argument. (What do you mean?) 3) "[E]xtraordinarily unlikely" somehow seems quite meaningless under superdeterminism, but never mind that. I guess that is covered by your Descartes' demon argument. – user3164 Jan 19 '14 at 20:12
  • @GlenTheUdderboat - What is an example of an experimental outcome that you cannot explain with nonlocal hidden variables that you can explain with superdeterminism? – Rex Kerr Jan 19 '14 at 20:27
  • I've read about (but I can't find now, I apologize) the search for actual limits on computational complexity by nature that might show that quantum mechanics is not precisely true in reality. (This is mostly concerning the simulation argument.) Also, I don't understand, how are nonlocal hidden variables better than superdeterminism? – user3164 Jan 19 '14 at 20:32
  • "[I]t is not experimentally distinguishable from nonlocal hidden variables." Logically, this is true under the assumption that quantum mechanics is true, but otherwise it isn't necessarily true. – user3164 Jan 19 '14 at 20:34
  • It appears to me that the argument brought against superdeterminism is that it isn't scientific, not that it is plain illogical. In that sense it appears to be based on the historical success of the scientific method. But since we're now concerned with the "end point" (i.e. "truth"), one might wonder if that's a good argument in this context. (By the way, I'm not deliberately trying to be combative, I'm just hoping to learn.) – user3164 Jan 19 '14 at 20:50

Some modern scientists argue that the concept of determinism should be retired from philosophy of science altogether and replaced with the more accurate concept of predictability. There is a substantial difference: for example, Newtonian Mechanics is deterministic, but it has been proven relatively recently that it's not predictable: in order to compute a prediction of a sufficiently large system one would need to employ a system that is larger in an essential way.


Jeez. All this was pretty much sorted in the Critique of Pure Reason. Both "free will" and "determinism" are mental gesticulations signifying human cluelessness. 't Hooft is not -- how to put this -- intellectually cultivated to the degree that, for instance, Zeilinger most certainly is. (Also you may feel free to wonder how conditioned he has been by the emotional detritus of Dutch Reformed Calvinism ... not to laugh. When he was a kid Amsterdam was nothing like the fun place we know today.)

As AZ says, doing science has meaning only if the scientist can assume the reasonable possibility that he or she has autonomous power to interrogate the natural world employing his or her purely subjective perceptions and ideas. This is a psychological necessity. Also science is based on the assumption that discovery is an interactive process. You cannot have interaction between entities one of which determines the actions of the other. That's incoherent. Incoherent, I tell you.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.