Closing the Superdeterminism Loophole?

[Mgt3]

TL;DR

Does this paper rule out superdeterminism?

This paper claims that the major superdeterminism loophole in Bell's theroem is closed, because "a local hidden variable theory consistent with relativity requires that relativistically non-invariant relations such as the time order of spacelike separated events have no physical significance, this result means that a local hidden variable theory cannot explain the correlation and reproduce all predictions of quantum mechanics even when assuming superdeterminism." I have seen Sabine Hossenfelder promoting superdeterminism on her blog, papers, and YouTube, but she never addresses the challenges posed in this paper.

In your opinions, is this paper successful?

 

[PeroK]

Summary:: Does this paper rule out superdeterminism?

This paper claims that the major superdeterminism loophole in Bell's theroem is closed, because "a local hidden variable theory consistent with relativity requires that relativistically non-invariant relations such as the time order of spacelike separated events have no physical significance, this result means that a local hidden variable theory cannot explain the correlation and reproduce all predictions of quantum mechanics even when assuming superdeterminism." I have seen Sabine Hossenfelder promoting superdeterminism on her blog, papers, and YouTube, but she never addresses the challenges posed in this paper.

In your opinions, is this paper successful?

My first thought is that the paper looks a little naive. Superdeterminism replaces QM with a set of unknown (and essentially unknowable) laws of physics that mimic QM and quantum entanglement.

Moreover, I'm not convinced that the philosophers who wrote this paper entirely understand the details of quantum entanglement in terms of spacelike separated measurements.

Hossenfelder's claim is simply that all physical data is correlated. What the authors suggest does not address this claim in any way that I can see.

 

[PeterDonis]

Moderator's note: Thread moved to QM interpretations subforum.

 

[Mgt3]

Thank you! Are there any other reasons most physicists reject superdeterminism?

 

[PeroK]

Thank you! Are there any other reasons most physicists reject superdeterminism?

Here's a brief but sharp criticism from Scott Aaronson:

https://scottaaronson.blog/?p=6215

The following sums it up for me:

It strikes me that no one who saw quantum mechanics as a profound clue about the nature of reality could ever, in a trillion years, think that superdeterminism looked like a promising route forward given our current knowledge. The only way you could think that, it seems to me, is if you saw quantum mechanics as an anti-clue: a red herring, actively misleading us about how the world really is. To be a superdeterminist is to say:

OK, fine, there’s the Bell experiment, which looks like Nature screaming the reality of ‘genuine indeterminism, as predicted by QM,’ louder than you might’ve thought it even logically possible for that to be screamed. But don’t listen to Nature, listen to us! If you just drop what you thought were foundational assumptions of science, we can explain this away! Not explain it, of course, but explain it away. What more could you ask from us?

 

[PeroK]

Thank you! Are there any other reasons most physicists reject superdeterminism?

To add to what Scott Aaronson says, I would draw a parallel with the following.

When Charles Darwin found the fossils of sea creatures high in the Andes, he took it as evidence that the Andes had somehow been pushed up by natural forces and the rocks had once been below sea level. Fitzroy, who was captain of the Beagle and a religious man, suggested that they might have been put there by God to test his faith!

The superdeterminists have just as strong an a priori faith in determinism. Darwin didn't wake up one morning and decide that it would be interesting if the Andes had been pushed up from the sea: he found scientific evidence for it. Similarly, physicists since the 1920s haven't just decided that QM is correct, they have found scientific evidence for it. The superdeterminists, on the other hand, say that this evidence is just nature playing tricks and you shouldn't let it shake your faith in determinism. They ask you to disregard the scientific evidence in support of quantum mechanics and trust their faith in determinism instead.

 

[Jarvis323]

Meh, Aaronson's view in this case is dissapointing; e.g. in the way he frames outliers in theoretical physics (Penrose, Sabine, etc.) as though there is something wrong with them and they are to be ridiculed in a debate over what people should believe. Just because 99% of people agree about something doesn't mean they're right. Science isn't a war about belief. In this level of science in particular, it's better to be exhaustive in my opinion. Having outliers focusing on unlikely possibilities is a bonus.

I also don't like how, in the same wind that he deliberately parodies SD, he remarks he deliberately tried not to. It's a little childish to be honest. Sabine was right in her comment on the blog, if someone with as much clout as Aaronson wants to blog about her theory, they should at least read it and respond to it as it's written rather than through preconcieved opinions and ridicule. And it does seem to me he's misrepresented her pretty egregiously.

Sabine is similarly arrogant and offensive in her youtube video though. So ... I don't know.

 

[martinbn]

Meh, Aaronson's view in this case is dissapointing; e.g. in the way he frames outliers in theoretical physics (Penrose, Sabine, etc.) as though there is something wrong with them and they are to be ridiculed in a debate over what people should believe.

Did he really talk about Penrose!? As far as I am aware Penrose is not a proponent of superdeterminism.

 

[Nullstein]

If you believe that the lesson of QM is that there are no hidden variables, then the issue of superdeterminism vs. non-locality needn't bother you. Only if you insist on the existence of hidden variables, you are faced with the choice between these undesireable consequences. In that case, one should notice that there is evidence that superdeterminism might require an even smaller amount of fine-tuning than non-locality.

 

[gentzen]

Did he really talk about Penrose!? As far as I am aware Penrose is not a proponent of superdeterminism.

Yes, this is what he said:

For me, Sabine, Gerard ‘t Hooft, and Roger Penrose all fall into the category of “wilful contrarians,” which is well-known to be compatible with being arbitrarily smart. Listen to them, sure, but like you’d listen to a trial lawyer. When they confidently declare things that 99% of their colleagues disagree with, make sure to listen as well to someone from the other 99%.

 

[martinbn]

Yes, this is what he said:

I see, in those comments they are discussing in general, people and their hypothetical proposals, not specifically superdeterminism. I don't see it as @Jarvis323 described it.

 

[PeroK]

For me, Sabine, Gerard ‘t Hooft, and Roger Penrose all fall into the category of “wilful contrarians,” which is well-known to be compatible with being arbitrarily smart. Listen to them, sure, but like you’d listen to a trial lawyer. When they confidently declare things that 99% of their colleagues disagree with, make sure to listen as well to someone from the other 99%.

That's perfect! This is why I feel I'm on the same wavelength as Aaronson. If he were British, he would have used the phrase too clever by half!

 

[PeterDonis]

Yes, this is what he said:

He also said this in the same comment:

This is different from (say) Sean Carroll, John Preskill, or the late Steven Weinberg, who certainly in matters of science, I never once heard to say anything unless there was an excellent reason for supposing it true.

Carroll is a proponent of the MWI, and not everyone agrees that there is "an excellent reason" for supposing that to be true. But many more physicists do believe that than believe that there is an excellent reason for supposing superdeterminism to be true. And Hossenfelder does believe that about superdeterminism, so in her view, she is perfectly justified in arguing for it.

So the standard Aaronson is appealing to here is not actually objective: it's just popularity. If enough physicists believe that "there is an excellent reason" for supposing something is true, Aaronson has no problem with physicists saying it; but if only a few "wilful contrarians" do, then he thinks you should listen to them as if they were a trial lawyer. But that isn't science.

In science, you listen to everyone as if they were a trial lawyer. Which means, in simple terms, much the same thing as it means for a jury in a trial: you don't accept what the lawyer says just because it's a popular opinion, you look for the evidence. And you can't test the MWI by evidence any more than you can test superdeterminism by evidence, so if Aaronson is going to criticize Hossenfelder for having blind faith in superdeterminism, he should be equally ready to criticize Carroll for having blind faith in the MWI.

 

[PeroK]

So the standard Aaronson is appealing to here is not actually objective: it's just popularity. If enough physicists believe that "there is an excellent reason" for supposing something is true, Aaronson has no problem with physicists saying it; but if only a few "wilful contrarians" do, then he thinks you should listen to them as if they were a trial lawyer. But that isn't science.

Absolute objectivity is not possible. On PF we do not allow discussion of personal theories unless they are published in mainstream scientific journals. But, the decision to publish a paper (or not) is ultimately an subjective decision taken by the editors and referees.

In science, you listen to everyone as if they were a trial lawyer.

This is not the case. From a practical point of view, you cannot give every crackpot a fair hearing - we certainly do not do that on PF!

if Aaronson is going to criticize Hossenfelder for having blind faith in superdeterminism, he should be equally ready to criticize Carroll for having blind faith in the MWI.

Every person is free to disagree publicly or privately with the theories or opinions with which they most disagree. They are not obliged to obey anyone else's principles. Aaronson (and I) are perfectly within our rights to object to SD and not to MWI.

 

[PeterDonis]

Absolute objectivity is not possible.

Granted, but "objectivity" in the context of science means "can be tested by evidence". In other words, it's not just based on popularity among scientists.

Also, Aaronson himself is claiming to apply an objective standard. He didn't say Carroll and others never said anything unless they believed there was an excellent reason for supposing it true. He said they never said anything unless there was an excellent reason for supposing it true. In other words, he is claiming that Carroll is meeting an objective standard that Hossenfelder is not.

But that claim doesn't hold water: it can't be a claim to absolute objectivity, which, as you point out, is not possible, and it can't be claim of objectivity in the sense of testability by evidence, which as I noted above is the proper standard for objectivity in science, because the MWI, which Carroll has argued for (and IIRC Aaronson has cited Carroll with approval on this point in previous posts on his blog), can't be tested by evidence. So if Hossenfelder doesn't meet the standard of never saying anything unless there is, objectively, an excellent reason for supposing it true, neither does Carroll. Yet Aaronson applauds the latter and criticizes the former.

Of course Carroll believes there are excellent reasons for supposing the MWI to be true, but so also does Hossenfelder believe there are excellent reasons for supposing superdeterminism to be true; so subjectively they are both doing the same thing and there is no ground for distinguishing between them. The only ground on which one can actually draw a distinction between them is popularity: Carroll's belief in the MWI is much more popular among physicists than Hossenfelder's belief in superdeterminism.

This is not the case. From a practical point of view, you cannot give every crackpot a fair hearing - we certainly do not do that on PF!

I was not intending to say one has to listen to everyone whatsoever. Perhaps a better rephrasing of my statement would be: "In science, if you choose to listen to anyone making a scientific claim, you should listen to them as if they were a trial lawyer."

Every person is free to disagree publicly or privately with the theories or opinions with which they most disagree. They are not obliged to obey anyone else's principles. Aaronson (and I) are perfectly within our rights to object to SD and not to MWI.

As a matter of personal opinion, sure. And I am free to point out that the personal opinion Aaronson is expressing seems to me to be inconsistent, for the reasons I have given. He's claiming to apply an objective standard, but the standard he is actually applying boils down to popularity.

 

[stevendaryl]

If you believe that the lesson of QM is that there are no hidden variables, then the issue of superdeterminism vs. non-locality needn't bother you. Only if you insist on the existence of hidden variables, you are faced with the choice between these undesireable consequences. In that case, one should notice that there is evidence that superdeterminism might require an even smaller amount of fine-tuning than non-locality.

It seems to me that superdeterministic requires arbitrarily much fine-tuning.

Basically, the superdeterminism loophole for EPR is that if you assume that the choices made by the experimenters (Alice and Bob) are correlated with the initial setting of the hidden variable in the twin pair, then you can easily reproduce results that violate Bell’s inequality. (I actually think you can do more than that: superdeterminism would allow you to appear to send messages FTL — if Alice’s messages are predetermined, then you don’t have to wait for them to arrive; you can just compute what they will be.)

But Alice can consult arbitrarily many other sources in making her decision. For example, she could say: “If I see a shooting star, I will perform measurement A, otherwise, if my basketball team wins, I will perform measurement B, otherwise if my geiger counter measures a uranium decay, I will perform measurement C, otherwise if there is a supernova explosion visible in this section of the sky, I will perform measurement D…”

So for the hidden variable to be correlated with Alice’s choice of measurement would require arbitrarily much of the universe to be correlated in a seemingly arbitrary way.

 

[stevendaryl]

So the standard Aaronson is appealing to here is not actually objective: it's just popularity. If enough physicists believe that "there is an excellent reason" for supposing something is true, Aaronson has no problem with physicists saying it; but if only a few "wilful contrarians" do, then he thinks you should listen to them as if they were a trial lawyer. But that isn't science.

I don’t think it’s fair to say it’s just popularity.

MWI has been around for a long time, and even though the arguments for it haven’t convinced everyone, the arguments pro and con have been offered and rebutted. I don’t think that superdeterminism is in that position.

Of course, things can change, and superdeterminism may begin to be taken seriously and the pros and cons may eventually be all aired. Eventually, it might be in the same boat as MWI. But it isn’t now.

 

[stevendaryl]

I don’t think it’s fair to say it’s just popularity.

MWI has been around for a long time, and even though the arguments for it haven’t convinced everyone, the arguments pro and con have been offered and rebutted. I don’t think that superdeterminism is in that position.

Of course, things can change, and superdeterminism may begin to be taken seriously and the pros and cons may eventually be all aired. Eventually, it might be in the same boat as MWI. But it isn’t now.

I would also point out that superdeterminism is not a theory. It is a class of theories. I would say that there isn’t actually a superdeterministic theory on the table for consideration. Just the idea that maybe it’s worth exploring.

 

[PeroK]

Actually, there's another critcism of Hossenfelder and Palmer's paper here, where Mateus Araujo has tackled the specific claims made in the paper:

https://mateusaraujo.info/2019/12/17/superdeterminism-is-unscientific/

It's funny that Hossenfelder and Palmer's are supported in arguing against Araujo by "Andrei", who is almost certainly the same Andrei who posted on here for a while and wouldn't accept that there was any experimental evidence for QM over classical electromagnetism.

It's the same on Hossenfelders blog. A sample comment from the crackpots who follow her is:

"I'm really intrigued by this explanation and I've been thinking for a long time that quantum mechanics is an anti-scientific and ad hoc theory. Finally, knowing about these technicalities of Bell's theorem, the way out of all that wrongness has become clearer, and confirms that quantum mechanics fundamentally rests on a circular reasoning:"

She and Palmer cannot claim to be naive about the damage they are doing by peddling this material that inspires the crackpots against QM.

 

[PeterDonis]

MWI has been around for a long time, and even though the arguments for it haven’t convinced everyone, the arguments pro and con have been offered and rebutted. I don’t think that superdeterminism is in that position.

While this is true, it's also true that if people who try to present arguments for superdeterminism are dismissed as "wilful contrarians", as Aaronson does, superdeterminism has no chance to ever get to be in that position. The MWI itself took quite some time to become a "recognized" interpretation because nobody bothered to read Everett's thesis since Bohr considered it too outlandish. (I think someone mentioned more details about this in another recent thread on the MWI.)

So the facts about which viewpoints on QM interpretations, and on "metaphysical" views like superdeterminism, have been around for a long time and have had a chance to have arguments pro and con offered and rebutted, are themselves a matter of popularity. After all, we are talking about things that can't be tested by experiment; if we could test the various QM interpretations against each other by experiment, the QM interpretation debate wouldn't still be going on after a century or so.

I would also point out that superdeterminism is not a theory. It is a class of theories. I would say that there isn’t actually a superdeterministic theory on the table for consideration. Just the idea that maybe it’s worth exploring.

And the MWI is not a theory either. It's an interpretation of a theory. It's just as untestable by experiment as superdeterminism is. I don't see that the MWI has any advantage here.

 

[PeterDonis]

It seems to me that superdeterministic requires arbitrarily much fine-tuning.

This post, btw, is a cogent criticism of superdeterminism (the same one I would make--in fact I think I have posted something similar in past threads), stated simply and briefly, without any unnecessary hyperbole.

 

[PeroK]

This post, btw, is a cogent criticism of superdeterminism (the same one I would make--in fact I think I have posted something similar in past threads), stated simply and briefly, without any unnecessary hyperbole.

Hossenfelder denies this. From the paper in question:

"The more polemic version of this is that in a superdeterministic theory, the universe must have been “just so” in order that the decisions of experimenters happen to reproduce the predictions of quantum mechanics every single time. Here, the term “just so” is invoked to emphasize that this seems intuitively extremely unlikely and therefore Superdeterminism relies on an implausible “conspiracy” of initial conditions that does not actually explain anything.

To address this objection, let us first define “scientific explanation” concretely to mean that the theory allows one to calculate measurement outcomes in a way that is computationally simpler than just collecting the data. This notion of “scientific explanation” may be too maths-centric to carry over to other disciplines, but will serve well for physics. The criticism leveled at Superdeterminism is, then, that if one were to accept explaining an observation merely by pointing out that an initial state and a deterministic law exists, then one would have to put all the information about the observation already in the initial state, meaning the theory is not capable of providing a scientific explanation in the above defined sense.

One problem with this argument is that just by knowing a theory violates Statistical Independence one cannot tell anything about its explanatory power. For this one needs to study a concrete model. One needs to know how much information one has to put into the initial state and the evolution law to find out whether a theory is or is not predictive."

And, if you can make any sense of that, you're doing better than I can. But, SD is not fined tuned is the conclusion.

 

[PeterDonis]

Hossenfelder denies this.

Yes, I know. I don't agree with her denial. Nothing I have said in this thread should be taken to imply that I think superdeterminism is correct. I don't. But even if one thinks a claim is wrong, one should still object to the claim being rejected for bad reasons. And conversely, even if one thinks a claim is correct, one should still object to the claim being supported with bad arguments.

 

[PeroK]

PS a more lucid example:

"Such conspiracy arguments are also often phrased as worries about the need to “fine-tune”—i.e., choose very precisely—the initial conditions (see Wood and Spekkens [15] for a quantifiable definition). The reference to fine-tuning, however, is misleading. There need be nothing a priori unscientific about a fine-tuned theory [16]. A fine-tuned theory may be unscientific if one needs to put a lot of information into the initial condition thereby losing explanatory power."

What I think she's saying is that if only we knew the laws of physics behind SD, we might find we need very little fine tuning in order to achieve the ubiquitous correlations. Hmm?

 

[PeterDonis]

What I think she's saying is that if only we knew the laws of physics behind SD, we might find we need very little fine tuning in order to achieve the ubiquitous correlations.

That appears to be what she's trying to say, but she contradicts it in the last sentence you quote:

A fine-tuned theory may be unscientific if one needs to put a lot of information into the initial condition thereby losing explanatory power.

The post by @stevendaryl that I said was a cogent criticism is basically pointing out that, to explain the kinds of correlations you would have to explain, which could involve Alice and Bob choosing to rely on incoming light signals from events arbitrarily far away to determine their measurement settings in experiments that show Bell inequality violations, yes, you would have to "put a lot of information into the initial condition". Where else could you put it? You can't put it in the laws, because the whole point of superdeterminism is to be able to have laws that are completely local.

 

[PeroK]

This is my last post on this subject, but the most popular comment on Hossenfelder YouTube Video, with 300 likes, is this:

"I've been saying for decades that EVERY observation in QM can be explained deterministically. I was shot down every time, primarily because of Bell. Now, finally, a few decades late, somebody has noticed that Bell's theorem doesn't say what everybody insisted it said. And so now we have superdeterminism, which is just plain old determinism, but with a 'super' added presumably to help someone somewhere save some face. I don't mind, in fact if people were to attach a 'super' to everything I've been saying for the last few decades and not just in the field of quantum mechanics that would be fine by me."

AFAIK, what's she done is unforgiveable.

 

[stevendaryl]

And the MWI is not a theory either. It's an interpretation of a theory.

Yes, but that’s an important contribution.

It's just as untestable by experiment as superdeterminism is. I don't see that the MWI has any advantage here.

I’m just saying that there is no substance (yet) to superdeterminism.

 

[Nullstein]

I think the claim that superdeterminism requires fine-tuning is not controversial. What should however be controversial, is the claim that non-local hidden variable theories don't require fine-tuning. In fact in probabilistic causality, both are considered to require fine-tuning (see e.g. Wood, Spekkens 2012). And then one can ask the question: Which one requires more fine-tuning? According to the paper I cited earlier, it seems that superdeterminism might get away with less fine-tuning than non-local hidden variables, which means one should be even less comfortable with non-locality than with superdeterminism. That doesn't imply that one should be comfortable with superdeterminism. It just means that one should be at least equally unsatisfied with non-locality.

Just to give some intuition: Consider that in a fully deterministic theory, a non-local action in the present can always be evolved backwards in time into a region in the past, where all relevant backward light cones once intersected. This intersection can be taken as a common cause that violates the no-superdeterminism assumption.

 

[PeterDonis]

that’s an important contribution.

Is it? As I've already said, the QM interpretation debate has been going on for a century now. Resolving it, by coming up with some way of making the difference between interpretations testable (i.e., by turning them into actual successor theories to the QM we have now), would be an important contribution, but the MWI hasn't done that, any more than any other proposed interpretation has.

I’m just saying that there is no substance (yet) to superdeterminism.

And I'm saying that on the only really relevant grounds as far as science is concerned, testability by experiment, there is no substance (yet) to the MWI, or indeed to any QM interpretation. Any grounds for saying that the MWI has more "substance" come down to the fact that more physicists believe it. Which, in the absence of any experimental testability, is a matter of popularity.

I understand that the standard I am applying here is much stricter than the standard most physicists apply. I think that is a problem with the standard most physicists apply: it does not do a good enough job of distinguishing things that can actually be tested from things that can't, and focusing attention on the former instead of the latter.

 

[PeroK]

I understand that the standard I am applying here is much stricter than the standard most physicists apply. I think that is a problem with the standard most physicists apply: it does not do a good enough job of distinguishing things that can actually be tested from things that can't, and focusing attention on the former instead of the latter.

What if I propose a SD theory of electrical circuits? Classical EM says that when I flick my light switch a current is set up in the electrical wiring and my lightbulb goes on. But, I claim, that the switch being flicked and the filament illuminating are superdeterministically correlated. No matter what one does to try to demonstrate that the switch being flicked causes the bulb to illuminate, I claim that it's just SD correlations.

And, if you put an ammeter on the wire, then the reading on the meter is likewise just SD correlated with the switch. And doesn't prove there is a current at all.

I can think of no reason to take that theory seriously! I cannot see that MWI is even remotely in the same category as that.