Scholarpedia article on Bell's Theorem

[Maui]

So it would be much clearer to put it this way: denying realism implies nonlocality. Or equivalently: locality implies realism.

If by 'locality' you mean just the abscence of nonlocal influences, how does it imply realism? How does "denying realism imply nonlocality"?

 

[harrylin]

If by 'locality' you mean just the abscence of nonlocal influences, how does it imply realism? How does "denying realism imply nonlocality"?

That doesn't make sense to me either: there are certainly theories possible that are "realistic but non-local, and Bell admitted that non-realism could be an alternative interpretation to his non-locality ("it might be that there is no reality below some "classical" "microscopic" level").

 

[lugita15]

If by 'locality' you mean just the abscence of nonlocal influences, how does it imply realism? How does "denying realism imply nonlocality"?

Locality alone isn't supposed to imply realism. What is required to get realism in the argument is locality plus perfect correlation, which is the quantum mechanical prediction that if you have two entangled photons, and you send them to distant polarization detectors oriented at the same angle, then they will either both go through the detectors or they will both not go through. (And many/most people will argue that what you need to imply realism is four assumptions: locality, counterfactual definiteness, perfect correlations, and the no-conspiracy condition). Then once you have realism in hand, you can derive a Bell inequality which is in contradiction to other predictions of QM. (See "quantumtantra.com/bell2.html" for a really simple and easy-to-understand explanation of all this.) So the conclusion is supposed to be that a local theory can't reproduce all the predictions of quantum mechanics (QM is often considered nonlocal because of wavefunction collapse, but this is controversial).

 

[lugita15]

That doesn't make sense to me either: there are certainly theories possible that are "realistic but non-local

Yes, but ttn's claim is not that realism implies locality, but rather that locality (plus the no-conspiracy condition and perfect correlation at identical angle settings) implies realism. I'm inclined to disagree with this claim of his, because it seems to me that counterfactual definiteness should be necessary, but for now I'm stymied by his argument quoted in my post #316.

 

[ttn]

If by 'locality' you mean just the abscence of nonlocal influences, how does it imply realism? How does "denying realism imply nonlocality"?

See our explanation of the argument here:

http://www.scholarpedia.org/article/Bell%27s_theorem#The_EPR_argument_for_pre-existing_values

The idea is just that the only way to explain the perfect correlations without recourse to nonlocal influences is to say that each particle carries pre-existing values for all the various polarization/spin components. (This last is what Dr C and others insist on calling "realism".)

 

[ttn]

Yes, but ttn's claim is not that realism implies locality, but rather that locality (plus the no-conspiracy condition and perfect correlation at identical angle settings) implies realism.

That's right.

I'm inclined to disagree with this claim of his, because it seems to me that counterfactual definiteness should be necessary, but for now I'm stymied by his argument quoted in my post #316.

Stymied by it? You just mean you can't refute it? Good luck with that. =)

 

[Demystifier]

Travis, have you seen my cheap "proof" of nonlocality in
https://www.physicsforums.com/showpost.php?p=3310924&postcount=1

It is surprising how difficult for most readers it was to identify the real error in the proof, which I eventually revealed at
https://www.physicsforums.com/showpost.php?p=3317384&postcount=57

 

[ttn]

Travis, have you seen my cheap "proof" of nonlocality in
https://www.physicsforums.com/showpost.php?p=3310924&postcount=1

It is surprising how difficult for most readers it was to identify the real error in the proof, which I eventually revealed at
https://www.physicsforums.com/showpost.php?p=3317384&postcount=57

Actually I basically agree with the "cheap proof". What you call the "real error in the proof" -- that H means two different things in the two parts of the argument -- might be true, but need not be. That is, it is entirely possible to run the EPR argument and get out exactly the H you need for Bell's theorem. (Doing so uses also the "no conspiracy" premise, but that barely counts as an extra assumption.)

It sounds as if your point with this brain teaser was to acknowledge that something like "counterfactual definiteness" actually is an extra assumption needed to get to Bell's conclusion of nonlocality. If that's what you meant, I think you are wrong.

 

[ttn]

Travis, have you seen my cheap "proof" of nonlocality in
https://www.physicsforums.com/showpost.php?p=3310924&postcount=1

It is surprising how difficult for most readers it was to identify the real error in the proof, which I eventually revealed at
https://www.physicsforums.com/showpost.php?p=3317384&postcount=57

I looked very quickly at some other posts from that old thread and found this gem from Dr C:

The EPR premise was that if you assumed H and L, then you would conclude a more complete specification of the system was possible than QM provided. They made the argument that you had to assume H if the (simultaneous, course) elements of reality existed, which they felt was a reasonable position to take.

I would be interested in how you got L -> H out of their reasoning. I think you will see H is an assumption, not a deduction.

That is a perfectly clear and terse statement of his view, I think. And it gives me a nice opportunity to summarize what I've been saying against him in this thread.

The key point is that "H" alone implies (really, just *means*) that "a more complete specification of the system was possible than QM provided". There is no need at all to bring in "L" if you are going to *assume* "H". Now the EPR paper, written by Podolsky without Einstein ever having seen it before it went to press, is definitely confusingly written. But none of these guys was stupid. If they were going to just *assume* "H", they wouldn't have bothered with any of the business about "L".

So why do they bring in "L"? Because the actual heart of the argument is that "L" --> "H". That is, "H" is *not assumed, it is inferred*.

This is what Dr C fails/refuses to grasp.

 

[Demystifier]

Actually I basically agree with the "cheap proof". What you call the "real error in the proof" -- that H means two different things in the two parts of the argument -- might be true, but need not be. That is, it is entirely possible to run the EPR argument and get out exactly the H you need for Bell's theorem. (Doing so uses also the "no conspiracy" premise, but that barely counts as an extra assumption.)

If you are right, then my "cheap proof" is much more valuable than I thought. For that purpose, can you specify more precisely how the EPR argument should be run to get exactly the H needed for Bell's theorem?

Let me explain why I find it difficult. The EPR argument runs roughly as follows. At measurement, the wave function either collapses or does not collapse. So

(i) It if collapses, then the collapse is not local, which contradicts L. That's the easy part.

(ii) The difficult part is what if it does not collapse? Then there must be something that determines the measured value which is not contained in the wave function alone. There must be something more, which we generically call H. But what that H is? What properties it must have? I don't see how EPR can answer that question. On the other hand, the H_Bell used in the Bell theorem seems to have more specified properties, which I don't see how to extract from the EPR argument alone.

In other words, it seems to me that people who try to avoid the Bell theorem (including myself with my H_solipsistic attempt) try to construct H appropriate for the EPR argument which lack some properties of the H_Bell.

 

[Demystifier]

I looked very quickly at some other posts from that old thread and found this gem from Dr C:



That is a perfectly clear and terse statement of his view, I think. And it gives me a nice opportunity to summarize what I've been saying against him in this thread.

The key point is that "H" alone implies (really, just *means*) that "a more complete specification of the system was possible than QM provided". There is no need at all to bring in "L" if you are going to *assume* "H". Now the EPR paper, written by Podolsky without Einstein ever having seen it before it went to press, is definitely confusingly written. But none of these guys was stupid. If they were going to just *assume* "H", they wouldn't have bothered with any of the business about "L".

So why do they bring in "L"? Because the actual heart of the argument is that "L" --> "H". That is, "H" is *not assumed, it is inferred*.

This is what Dr C fails/refuses to grasp.

Yes, I also tried to explain him the same thing in that old thread.

 

[DrChinese]

This is what Dr C fails/refuses to grasp.

Now you're hurting my feelings...

We've agreed to disagree before. However, I did add your Scholarpedia article to my Bell links page. And I have had a link to your paper "EPR and Bell Locality" on my page for years. See:

Bell's Theorem: An Overview with Lotsa Links

 

[Demystifier]

It sounds as if your point with this brain teaser was to acknowledge that something ... actually is an extra assumption needed to get to Bell's conclusion of nonlocality. If that's what you meant, I think you are wrong.

I would like to be wrong, but I just can't see that clearly enough. My H_solipsistic reduces nonlocality down to microscopic distances inside the observer, which suggests that perhaps nonlocality could be eliminated completely.

 

[Demystifier]

However, I did add your Scholarpedia article to my Bell links page. And I have had a link to your paper "EPR and Bell Locality" on my page for years. See:

Bell's Theorem: An Overview with Lotsa Links

You are not the only fan of his paper.
I cited it in the revised version of my recent EPR paper
http://xxx.lanl.gov/abs/1203.1139

 

[ttn]

If you are right, then my "cheap proof" is much more valuable than I thought. For that purpose, can you specify more precisely how the EPR argument should be run to get exactly the H needed for Bell's theorem?

See section 2, "The EPR argument for pre-existing values" in the article.

Really briefly, I think the issue is that you are thinking of EPR as merely a proof that (if you assume locality) then ordinary QM descriptions of physical states must be incomplete. It is of course that. But the particular way that EPR argue for that allows it to be somewhat more as well. What I mean is this: the way they argue for incompleteness is by establishing the existence of certain facts about the distant particle which are not endorsed by ordinary QM. But if you are really assuming locality (and the thing we call "no conspiracies") then you can establish by this same argument the existence of *many* such facts. Establishing merely *one* such fact is sufficient to show that QM is incomplete. But in fact the argument can be used to establish the existence of *many* -- as many as you need (it turns out) to derive a Bell inequality by standard methods.

 

[ttn]

Now you're hurting my feelings...

We've agreed to disagree before. However, I did add your Scholarpedia article to my Bell links page. And I have had a link to your paper "EPR and Bell Locality" on my page for years. See:

Bell's Theorem: An Overview with Lotsa Links

And that's why, at the end of the day, I still love you!

BTW, I appreciate the link to the scholarpedia piece, because I think that's a really important and high-quality article. On the other hand, that old paper "EPR and Bell Locality" is not really that great. Maybe you could swap it out for something newer and greater, like "JS Bell's concept of local causality"? =)

 

[ttn]

You are not the only fan of his paper.
I cited it in the revised version of my recent EPR paper
http://xxx.lanl.gov/abs/1203.1139

For that paper you might also want to check out my "Einstein's Boxes" paper that was in AmJPhys several years ago. It is also about "EPR before EPR".

 

[ttn]

I would like to be wrong, but I just can't see that clearly enough. My H_solipsistic reduces nonlocality down to microscopic distances inside the observer, which suggests that perhaps nonlocality could be eliminated completely.

Yes, as soon as you start saying that the things we normally think of as physically real events (like the outcomes of experiments) are actually not physically real, but just exist as phantasies in some mind somewhere, obviously you can "eliminate" the nonlocality. Of course, you aren't eliminating it in favor of a local theory -- you're eliminating it in favor of a "theory" in which there are no causal influences at all, because there are no physically real things/events in physical space. So ... whatever else you want to say about it ... this isn't a way of "saving locality".

Note also that this is not so much a way of challenging Bell's argument (that explaining the quantum correlations requires nonlocality) as it is denying that those correlations actually occur. So, really, your beef is with the experimentalists like Aspect who claim to have observed the correlations -- not with Bell. But basically as soon as you start going solipsistic we're off into la-la land and none of the usual ways of reasoning apply. It's similar to MWI in that respect.

 

[Demystifier]

For that paper you might also want to check out my "Einstein's Boxes" paper that was in AmJPhys several years ago. It is also about "EPR before EPR".

Thanks, I will read it.

 

[Demystifier]

It's similar to MWI in that respect.

I agree even more than you might think. My solipsistic HV model was to a large extent inspired by MWI and their claim that "dBB is MWI in a denial". While I certainly don't agree with that claim literally, some of the arguments used by MWI-ers do have a point.

In particular, I think that Bohmians should not ignore the fact that, in Bohmian theory, a single electron is not observed because it has a position, but because the measuring apparatus (the wave function of which is entangled with electron wave function) has a position.

My solipsistic HV's can be thought of as an attempt to combine some vague ideas of Copenhagen and MWI with conceptual clarity of dBB. Perhaps no many people like the resulting hybrid because at the same it is less minimalistic than MWI and Copenhagen and also less conceptually clear than dBB. On the other hand, it can be argued that it is more conceptually clear than MWI and Copenhagen, and also more minimalistic than dBB, so I still believe that solipsistic HV's do deserve one of the distinguished positions in the web of different interpretations of QM, perhaps as a link between other distinguished interpretations such as dBB, MWI and Copenhagen.

 

[Demystifier]

See section 2, "The EPR argument for pre-existing values" in the article.

Really briefly, I think the issue is that you are thinking of EPR as merely a proof that (if you assume locality) then ordinary QM descriptions of physical states must be incomplete. It is of course that. But the particular way that EPR argue for that allows it to be somewhat more as well. What I mean is this: the way they argue for incompleteness is by establishing the existence of certain facts about the distant particle which are not endorsed by ordinary QM. But if you are really assuming locality (and the thing we call "no conspiracies") then you can establish by this same argument the existence of *many* such facts. Establishing merely *one* such fact is sufficient to show that QM is incomplete. But in fact the argument can be used to establish the existence of *many* -- as many as you need (it turns out) to derive a Bell inequality by standard methods.

For me, the crucial step in the argument of Sec. 2 is the sentence
"But without any such interaction, the only way to ensure the perfect anti-correlation between the results on the two sides is to have each particle carry a pre-existing determinate value."

This is a logically correct argument, but it contains a hidden assumption that results on the two sides DO exist and ARE correlated. It is certainly a very very natural assumption, but is it strictly and absolutely a logical necessity? It is not, as long as we allow for the logical possibility that observations are kind of illusions or dreams (which indeed is the idea of solipsistic HV's). For example, if Aspect dreamed that his experiment violated Bell inequalities, would he interpret it as evidence for nonlocality?

Furthermore, the assumption that results on two sides do exist can be thought of as an ASSUMPTION of certain "H". Let us call this assumption H_outcome. So EPR argument can be summarized as follows:
Assuming QM and H_outcome, then L => H_spin
So EPR really assumes one kind of H (H_outcome) and locality to derive another kind of H (H_spin). (DrChinese may like it.) If macroscopic outcomes of experiments exist and are correlated, and if nature is local, then some microscopic reality (e.g., spin in all directions) also exists - that's what EPR seems to really say.

So, can Bell and EPR be combined as in my "cheap proof" to really derive nonlocality without assuming H? It depends on what exactly do we mean by H. My analysis above shows that if we assume H_outcome, then we can prove nonlocality without assuming H_spin. A physicist may reasonably argue that H_outcome is an observed fact so it does not need to be assumed, but strictly logically H_outcome may not exist if the only thing that exists is H_solipsistic.

Finally, I want to stress that I don't think that it diminishes the value of the Bell-EPR theorem. It just refines it by making its assumptions more explicit, so that those who want to play with possibilities of avoiding nonlocality know more precisely what of the common-sense assumptions they need to drop.

 

[ttn]

This is a logically correct argument, but it contains a hidden assumption that results on the two sides DO exist and ARE correlated. It is certainly a very very natural assumption, but is it strictly and absolutely a logical necessity? It is not, as long as we allow for the logical possibility that observations are kind of illusions or dreams (which indeed is the idea of solipsistic HV's). For example, if Aspect dreamed that his experiment violated Bell inequalities, would he interpret it as evidence for nonlocality?

Yes, it's certainly true that when we say that only a theory with pre-determined outcomes can locally explain the perfect correlations, we do indeed assume that the perfect correlations actually exist. I thought you were worrying about some kind of assumption about the reality of *counter*-factual outcomes... but you're worried about the reality of the *factual* ones! Well then I don't know what to say. Is it "strictly and absolutely a logical necessity"? I don't know. But it's an assumption that no sane person and certainly no scientist fails to make. But we can leave that aside. I'm happy to just stipulate that, yes, there is an additional assumption: Aspect wasn't merely dreaming that his experiment had the outcomes he published.

Furthermore, the assumption that results on two sides do exist can be thought of as an ASSUMPTION of certain "H". Let us call this assumption H_outcome. So EPR argument can be summarized as follows:
Assuming QM and H_outcome, then L => H_spin
So EPR really assumes one kind of H (H_outcome) and locality to derive another kind of H (H_spin). (DrChinese may like it.) If macroscopic outcomes of experiments exist and are correlated, and if nature is local, then some microscopic reality (e.g., spin in all directions) also exists - that's what EPR seems to really say.

Sure, fine. But who seriously doubts that macroscopic outcomes of experiments exist?

So, can Bell and EPR be combined as in my "cheap proof" to really derive nonlocality without assuming H? It depends on what exactly do we mean by H. My analysis above shows that if we assume H_outcome, then we can prove nonlocality without assuming H_spin. A physicist may reasonably argue that H_outcome is an observed fact so it does not need to be assumed, but strictly logically H_outcome may not exist if the only thing that exists is H_solipsistic.

It doesn't sound like we disagree about anything except how seriously solipsism should be taken. Incidentally, here is Bell on solipsism: "Solipsism cannot be refuted. But if such a theory were taken seriously it would hardly be possible to take anything else seriously." I think that's exactly right.

Finally, I want to stress that I don't think that it diminishes the value of the Bell-EPR theorem. It just refines it by making its assumptions more explicit, so that those who want to play with possibilities of avoiding nonlocality know more precisely what of the common-sense assumptions they need to drop.

I don't think it is any big news that if you adopt solipsism you will no longer accept nonlocality. You will also no longer accept that the Earth is round, that it goes around the sun, etc. Whatever this is, it is certainly not a *scientific* objection to the EPR argument.

 

[Demystifier]

Yes, it's certainly true that when we say that only a theory with pre-determined outcomes can locally explain the perfect correlations, we do indeed assume that the perfect correlations actually exist. I thought you were worrying about some kind of assumption about the reality of *counter*-factual outcomes... but you're worried about the reality of the *factual* ones! Well then I don't know what to say. Is it "strictly and absolutely a logical necessity"? I don't know. But it's an assumption that no sane person and certainly no scientist fails to make. But we can leave that aside. I'm happy to just stipulate that, yes, there is an additional assumption: Aspect wasn't merely dreaming that his experiment had the outcomes he published.

Sure, fine. But who seriously doubts that macroscopic outcomes of experiments exist?

It doesn't sound like we disagree about anything except how seriously solipsism should be taken. Incidentally, here is Bell on solipsism: "Solipsism cannot be refuted. But if such a theory were taken seriously it would hardly be possible to take anything else seriously." I think that's exactly right.

I don't think it is any big news that if you adopt solipsism you will no longer accept nonlocality. You will also no longer accept that the Earth is round, that it goes around the sun, etc. Whatever this is, it is certainly not a *scientific* objection to the EPR argument.

I agree with most of that, but let me discuss only those aspects with which I disagree.

First, how can Aspect be sure that it wasn't just a dream? Maybe he was in a Matrix from the Matrix movie?

Second, more seriously, I think there are some sane persons and scientists who in one way or another seriously deny or question factual outcomes. Well known examples include: Mermin - correlations without correlata, Zeilinger - message of the quantum, and Rovelli - relational interpretation.

So, if those guys are scientists, and if they publish such ideas in scientific journals, and if those papers receive a lot of attention by other scientists, then is it really justified to say that these ideas are not scientific? Even if they are not "scientific" in a certain narrow traditional meaning of that word, the mere fact that distinguished scientists seriously write about such ideas suggests that we should redefine what "scientific" means.

Third, if one takes solipsism seriously, it does not necessarily mean that one can't take anything else seriously. One should distinguish the general idea of solipsism from a concrete model of solipsism. In the case of general idea, perhaps all you can say is "cogito ergo sum". But in a concrete model, you can say much more. Specifically, in my solipsistic HV model, there are at least 3 things which you can take seriously: 1. space and time, 2. wave function in the configuration space, and 3. particle trajectories essential for working of the observer's consciousness.

All this does not mean that I like solipsism. Actually, I don't. But sometimes you can learn from thinking about something you don't like. In particular, I liked the mentioned papers by Mermin, Zeilinger and Rovelli even less, but it's not only that I didn't like them; these papers did not even make sense to me. Still, I respected these guys for their other scientific achievements, so I decided to think more about their ideas and to translate these ideas into something that makes more sense to me. As a result, I have constructed my solipsistic HV's. After that, I can say that even Mermin's, Zeilinger's and Rovelli's strange ideas are no longer so absurd to me as they used to be. Now whenever someone tells me something about QM in the spirit of Mermin, Zeilinger or Rovelli, I can translate it to my solipsistic-HV language and in this way understand much better what he is talking about. So if nothing else, solipsistic HV's are at least a useful mental tool (which, by the way, is also how many think of Bohmian HV's.)

 

[ttn]

I agree with most of that, but let me discuss only those aspects with which I disagree.

Sure, that's always the fun part. =)

First, how can Aspect be sure that it wasn't just a dream? Maybe he was in a Matrix from the Matrix movie?

But don't you see how, as Bell said, if you take this seriously you can't take anything seriously? I mean, how do you know you weren't dreaming just now when you came up with this argument about how maybe Aspect was dreaming? Maybe it seems to you like this is a really good argument that shows this important flaw in EPR's proof, but maybe it is actually a terrible argument and it just feels good in your dream? Or maybe all of the rules we're taking for granted here, about how you distinguish good arguments from bad ones, etc., maybe those were all just something in a dream and maybe the *real* standards for what is a good argument are completely different than what you're thinking right now. And so on. It is literally the case that, once you allow this kind of arbitrary doubt into your thinking, and take it seriously as if it were somehow logical or legitimate, *nothing* can remain. It wipes everything out. Now does that prove that it's wrong? Not exactly. But it proves you better not do this if you want to do science.

Second, more seriously, I think there are some sane persons and scientists who in one way or another seriously deny or question factual outcomes. Well known examples include: Mermin - correlations without correlata, Zeilinger - message of the quantum, and Rovelli - relational interpretation.

Rovelli maybe. But for sure Mermin and Zeilinger accept that the outcome of an actually-performed experiment is somehow real. They are only "anti-realist" in the sense that they deny the reality of unperformed experiments -- i.e., they deny that there is some fact about "how it would have come out" that is carried by some hidden variable. But here they simply miss the fact that no such *assumption* need ever be made. Anyway, as far as I know, the only people who really take seriously what I described before as "insane" are MWI people (or here equivalently "relational interpretation" people).

So, if those guys are scientists, and if they publish such ideas in scientific journals, and if those papers receive a lot of attention by other scientists, then is it really justified to say that these ideas are not scientific? Even if they are not "scientific" in a certain narrow traditional meaning of that word, the mere fact that distinguished scientists seriously write about such ideas suggests that we should redefine what "scientific" means.

I don't want to get into a big argument about this here, but for sure there is no contradiction in principle between saying "Mr. X is a scientist" and "Mr. X holds some views that are unscientific". For example there are at least a few people in the US who have PHDs in science and who hold professorships in science departments, but who believe in things like "young Earth creationism".

Third, if one takes solipsism seriously, it does not necessarily mean that one can't take anything else seriously. One should distinguish the general idea of solipsism from a concrete model of solipsism. In the case of general idea, perhaps all you can say is "cogito ergo sum". But in a concrete model, you can say much more. Specifically, in my solipsistic HV model, there are at least 3 things which you can take seriously: 1. space and time, 2. wave function in the configuration space, and 3. particle trajectories essential for working of the observer's consciousness.

Actually, as you know, I think your "concrete model" is incoherent or pointless -- or some such bad thing -- for just this reason. For example, what in the world makes you think that there exist brains made of particles and that conscious experience somehow arises or emerges from these physically real objects? Think back to the actual evidence you've encountered in your lifetime that makes you believe these things. I submit that *every single thing on that list* (of all the pieces of evidence that you take to justify your belief in the real existence of brains, etc.) is something that, according to your "concrete model", is a kind of unreal fantasy/dream/delusion/whatever. So, on the assumption that your "concrete model" is true, you should no longer believe in the existence of brains, etc., which of course in turn renders your concrete model completely pointless/incoherent. Let me put it this way: the idea that we could explain our perception of the regular world in the way that your model purports to do, actually *presupposes* a bunch of stuff that your model explicitly denies. The model is in some funny way self-refuting.

All this does not mean that I like solipsism. Actually, I don't. But sometimes you can learn from thinking about something you don't like. In particular, I liked the mentioned papers by Mermin, Zeilinger and Rovelli even less, but it's not only they I didn't like them; these papers did not even make sense to me. Still, I respected these guys for their other scientific achievements, so I decided to think more about their ideas and to translate these ideas into something that makes more sense to me.

Well, as you know, I think your model actually makes *less* sense than theirs. If you are going to say that basically conscious experience is all delusional, so there is no need to have really-existing trees, cats, planets, etc., in the ontology of the theory, then it is much better, much more elegant, much more plausible, to go "whole hog" and have conscious experience somehow emerging directly from the universal wave function (as in MWI). Putting in some brain particles so that there can be one "physically real" blob of stuff for all the delusionary experience to arise from, is just silly and pointless. But I've told you all of this before and there's no reason to get into it here.

As a result, I have constructed my solipsistic HV's. After that, I can say that even Mermin's, Zeilinger's and Rovelli's strange ideas are no longer so absurd to me as they used to be. Now whenever someone tells me something about QM in the spirit of Mermin, Zeilinger or Rovelli, I can translate it to my solipsistic-HV language and in this way understand much better what he is talking about. So if nothing else, solipsistic HV's are at least a useful mental tool (which, by the way, is also how many think of Bohmian HV's.)

Well, I find myself perfectly well able to understand these "crazy" sorts of views already, without any silly crutches. Also, despite understanding them, I have no problem concluding that, still, they are crazy, even irrational or unscientific.

 

[Demystifier]

But for sure Mermin and Zeilinger accept that the outcome of an actually-performed experiment is somehow real. They are only "anti-realist" in the sense that they deny the reality of unperformed experiments -- i.e., they deny that there is some fact about "how it would have come out" that is carried by some hidden variable.

I disagree. Unfortunately, Mermin and Zeilinger are not here to say what they mean, but see Mermin's paper "What is quantum mechanics trying to tell us".

Actually, as you know, I think your "concrete model" is incoherent or pointless ...
So, on the assumption that your "concrete model" is true, you should no longer believe in the existence of brains, etc., which of course in turn renders your concrete model completely pointless/incoherent.

Perhaps you misunderstood my model. It is not really about brain being real, but about our consciousness being real. Additional auxiliary (but not essential) assumptions are that reality of consciousness is related to reality of SOME particles, and that these particles are IN brain, but not that the brain as a whole is real.

Well, as you know, I think your model actually makes *less* sense than theirs. If you are going to say that basically conscious experience is all delusional, so there is no need to have really-existing trees, cats, planets, etc., in the ontology of the theory, then it is much better, much more elegant, much more plausible, to go "whole hog" and have conscious experience somehow emerging directly from the universal wave function (as in MWI).

Yes, you have a point, but I find such MWI-like theories more vague than mine. Of course, vagueness is subjective, so there is no point in arguing too much on that.

Also, despite understanding them, I have no problem concluding that, still, they are crazy, even irrational or unscientific.

Unlike you, I do have a problem with concluding it.

 

[ttn]

I disagree. Unfortunately, Mermin and Zeilinger are not here to say what they mean, but see Mermin's paper "What is quantum mechanics trying to tell us".

Well, who cares really. Maybe you're right about those two guys (though I don't think so). The point is just that there are more people who think the problem (with Bell's alleged proof of nonlocality) is something to do with *counter-factual definiteness* than there are people who think the problem is with *factual definiteness*. (By the former I mean the idea that unperformed experiments don't have definite outcomes; by the latter I mean the idea that even performed experiments don't have the definite outcomes we think they do.) Mermin, for example, famously thinks that the moon is not there when nobody looks. But I don't think he doubts that the moon is there when somebody *does* look. But who cares about Mermin, really. It doesn't matter. The real point here, actually, is that it is very dangerous to use words like "realism" that could mean lots of very very different things. Bohr, for example, was surely "anti-realist" (if "realism" means hidden variables) but was staunchly "realist" (if that means that the classically-macroscopically registered outcomes of experiments "really exist"). I think Mermin follows Bohr here. But I don't actually care if I'm right!

Perhaps you misunderstood my model.

Or perhaps you did!

It is not really about brain being real, but about our consciousness being real. Additional auxiliary (but not essential) assumptions are that reality of consciousness is related to reality of SOME particles, and that these particles are IN brain, but not that the brain as a whole is real.

I think I understood all that. My point was that it seems artificial, and indeed quite pointless, to introduce real physical particles (for the consciousness to arise from) when really the whole point of the model is that you could have consciousness (and in particular, consciousness "of", or at least "as if of", the usual QM predictions) without the usual naively-assumed *referents* of those beliefs actually existing physically the way we normally think they do. If that's the game you're going to play (namely, making QM be about delusional beliefs in some consciousness instead of being about physically real outcomes of physically real experiments) then you can play it better by having no particles in the picture at all. The particles in fact serve no purpose whatsoever, and introduce all kinds of embarrassing questions. (For example: what possible grounds could you have for even believing in "brains", let alone "particles", in the first place, if you're going to be solipsist about things like heads, scalpals, surgeons, etc.? And: what happened before the guy (whose consciousness your theory is about) was (as it is commonly, but according to your theory, erroneously, put) "born" and what will happen after he "dies"? And: What is so special about these particular degrees of freedom in the theory's hamiltonian, that they get particles associated with them, but other seemingly equivalent ones don't? And so on.)

Yes, you have a point, but I find such MWI-like theories more vague than mine. Of course, vagueness is subjective, so there is no point in arguing too much on that.

I agree, there's not much point in arguing about which theory is more vague. But it would actually help if you could say what you find "vague" about MWI. I really don't know what you mean by that, and I wouldn't myself say that the problem with MWI is that it is "vague". I think it's perfectly clear what it says -- namely, what's physically real is only the universal wave function (there are no "local beables"), and consciousness somehow emerges directly from that, and the structure of the wf (in particular decoherence) causes those emerging consciousnesses to be "as if of" the kind of (macroscopic) world that would ordinarily be thought of as described by a single branch of the universal wf. I personally find that to be a perfectly clear (non-vague) idea -- it's just an idea that is a little too crazy to be taken seriously, at least given that there are other (far less crazy) options on the table.

 

[ttn]

Unlike you, I do have a problem with concluding it.

OK, just because I'm curious, do you feel that way also about (e.g.) the "young Earth creationism" I mentioned before? This is the idea that the whole universe was created 6,000 years ago, on a Tuesday afternoon, just as it says in the bible. God arranged for there to be buried dinosaur bones (from dinosaurs that never lived or even existed) and other such things, evidently to try to trick us into believing the evil idea that actually the current species evolved over many hundreds of millions of years, and the Earth has been around for at least a couple billion years, etc.

There are actually people who believe this stuff. Suppose one of them has a phd in physics and is a physics prof with lots of good physics publications (about experimental condensed matter physics or some such). Would you thus refrain from saying that "young Earth creationism" is crazy and unscientific? I'm just trying to understand if your idea is "by definition, nothing espoused by a legitimate scientist can be unscientific and crazy" or is instead "MWI / relational versions of QM are unusual and weird, but not unscientific or crazy". I assume it's the latter, in which case we don't really disagree much. It's just that sometimes you give the impression it's the former, e.g., by citing things I consider totally irrelevant (like the credentials of the people who espouse MWI-ish things).

 

[Demystifier]

OK, just because I'm curious, do you feel that way also about (e.g.) the "young Earth creationism" I mentioned before?

I don't.

 

[Demystifier]

The particles in fact serve no purpose whatsoever,
...
But it would actually help if you could say what you find "vague" about MWI.

Perhaps you didn't read my solipsistic paper carefully enough. See again Sec. 2.2, especially the first and the third paragraph.

 

[Demystifier]

I'm just trying to understand if your idea is "by definition, nothing espoused by a legitimate scientist can be unscientific and crazy" or is instead "MWI / relational versions of QM are unusual and weird, but not unscientific or crazy". I assume it's the latter, in which case we don't really disagree much.

The latter, of course.

It's just that sometimes you give the impression it's the former, e.g., by citing things I consider totally irrelevant (like the credentials of the people who espouse MWI-ish things).

You will probably say that it doesn't matter WHO says something, but only WHAT he/she says. That's indeed true in theory, but not in practice. In practice, humans (which includes scientists) are not perfectly rational machines. They often use intuition and heuristics rather than logic. For example, if I see a new title on arXiv "Proof that quantum mechanics is local", will I bother to read it? If it is written by someone who I know as a crackpot, be sure that I will not. If it is written by A. Zeilinger, I will give the paper chance, but perhaps not too much. But if it is written by you or S. Goldstein for example, be sure that I will carefully read every word of it. And if it will not make sense to me at first, I will read it again and again, because it would be hard to believe that you or Goldstein could say something like that without having a really good argument.

On the other hand, if I see a paper with a title "The solution of the mind-body problem", I would read it if it was written by you, but I would pay much more attention to it if it was written by David Chalmers. I guess I don't need to explain why.