Does wave function of an entangled particle collapse instantly?

[andrewkirk]

Dear generous and helpful physicists,

A number of threads here contemplate strategies for transmitting information faster than light by observing an entangled particle in one place, allegedly causing the wave function of its entangled twin to instantly collapse in another, far away place. Leaving aside the question of whether it is possible to devise a way of using this to transmit information faster than light, I am interested in whether the suggestion that the wave function of particle B collapses the instant its entangled twin A is observed is
* part of the generally accepted principles of quantum mechanics,
* a point of controversy, or perhaps
* just a matter of interpretation, with no practical implications.

Further, is it possible to even tell that particle B's wave function has collapsed without observing it? Of course the wave function of B collapses when you observe it, but how can you know whether it has collapsed just then because of the observation you made, or earlier, when the observation of A was made?

If it's not possible to tell does that mean that my initial question is just a moot point - a matter of metaphysics rather than physics?

 

[Ken G]

You can count my vote for "just a matter of interpretation, with no practical implications." I find it quite significant that no experiment you can ever do on just one particle of the pair will ever seem the slightest bit strange. There is no trace of the entanglement at all-- until you correlate with observations on the other particle. Hence, there is never any need to imagine that anything "instantaneous" happened to the other particle-- instead, what happened instantaneously is your information about and expectations for that other particle changed. Since those changes occurred in your head, it is not surprising that they could be instantaneous-- you did an experiment on a system, and gained instantaneous information about that system.

 

[DrChinese]

...I am interested in whether the suggestion that the wave function of particle B collapses the instant its entangled twin A is observed is
* part of the generally accepted principles of quantum mechanics,
* a point of controversy, or perhaps
* just a matter of interpretation, with no practical implications.

...

To add to Ken's nice answer: I think all three of the above are true.

While the results seem instantaneous in some views, there is no existing experiment that demonstrates this to complete satisfaction for all. For example, the *lower* bound for collapse seems to be 10,000 c according to experiment. However, other experiments give results which seem to have time order reversed (collapse before entanglement, and yes I know it sounds impossible).

So I think you can see how difficult it is to put words on it. Yet all of this follows from the formalisms.

 

[xts]

Count my vote for "pure metaphysics - interpretation, with no practical implications."
As Ken G pointed - you cannot notice entanglement before you correlate results of both measurements. The fact you measured (or not) one particle not does not change the outcome of other measurement.
Thus, if you speak about 'collapse' when you measured one particle it doesn't affect by any means the other one - it affects only your knowledge about its behaviour. Collapse is not a real process affecting the other particle. It is just a measure of your knowledge. And, of course, your knowledge changes instantly, regardless how far away the particle is (or if it exists at all - maybe it already had been destructively measured some time ago?)

 

[unusualname]

Count my vote for "pure metaphysics - interpretation, with no practical implications."
As Ken G pointed - you cannot notice entanglement before you correlate results of both measurements. The fact you measured (or not) one particle not does not change the outcome of other measurement.
Thus, if you speak about 'collapse' when you measured one particle it doesn't affect by any means the other one - it affects only your knowledge about its behaviour. Collapse is not a real process affecting the other particle. It is just a measure of your knowledge. And, of course, your knowledge changes instantly, regardless how far away the particle is (or if it exists at all - maybe it already had been destructively measured some time ago?)

NO, it does not change the probability distribution of the other measurement (prior to measurement), so it seems to the person measuring the other particle that nothing has changed (but, as the person at the other end knows, the measurement is decided)

And you must not keep suggesting that wavefunction collapse is equatable to simply a revelation of unknown facts a la "Bertlmann's socks" (paper here)

 

[xts]

I guess we may agree to:
"measurement does not influence neither the outcome of the other measurement nor even its probability distribution"

And you must not keep suggesting that wavefunction collapse is equatable to simply a revelation of unknown facts a la "Bertlmann's socks" (paper here)

Why? How the Berlemann's socks relate to 'collapse'? The paper relates to Bell's inequality, nonlocality (I never said those are not true), and unapplicability of 'hidden variables' view, but what all those have to 'collapse'?

I always admit that entanglement involves stronger correlation than Bertlemann and his socks, but it is still only (nonlocal) correlation between outcomes of the measurements, which may be found only after both measurements are performed. Measurement of one Berlemann's leg has no influence on other leg, regardless we measure them simultaneously, left first, or right one first.
The "collapse" of right-leg wavefunction is only a change of our description of the right leg (we got some knowledge, so we may substitute a variable in the equation with fixed value), which we may use, since we know what sock is on the left one.
All three models lead to the same results:
1. We measure branch A, then we use 'collapsed' wf of B, then we measure B (the result is in accordance with probability density predicted by "collapsed" wf)
2. The same, but B first, then result in "A" is in accordance with "collapsed" wavefunction for A
3. We simultaneously measure both A and B as resulting from "uncollapsed" entangled wavefunction.
All three approaches lead to the same predictions and the same results: pairs of (A,B) exhibiting some correlation.

In order to assign some real meaning to the 'collapse', even in metaphysical sense only, I would expect causality in either direction. Either A causes 'collapse' in B, or B causes 'collapse' in A. But there is nothing like that. The same experiment is equally good described as A->B or B->A. Thus - rather than calling it 'a mystery' I prefer to call it 'measure of experimenter knowledge' - as the only what differentiates between A->B and B->A is the order in which I possesed the knowledge about measurement outcomes.

 

[unusualname]

I guess we may agree to:
"measurement does not influence neither the outcome of the other measurement nor even its probability distribution"

No, I can only agree that measurement at one end does not influence the probability distribution of any observable at the other end prior to its measurement.

Why? How the Berlemann's socks relate to 'collapse'? The paper relates to Bell's inequality, nonlocality (I never said those are not true), and unapplicability of 'hidden variables' view, but what all those have to 'collapse'?

Because you seem to suggest collapse is simply a process whereby previous unknown but existing facts (elements of reality) are revealed. And this has been shown incorrect by Bell's theorem and subsequent experiments.

From your various posts, you also appear to think entanglement and non-local correlations in Quantum Mechanics are pretty trivial ideas, and require no more explanation than classical wave-mechanics or similar.

Whatever the world is, it is not the naive world of classical realism, and there is (still) a mystery about how Quantum Mechanics works. It is ridiculous to say QM is pretty straightforward apart from the mystery of entanglement. Entanglement (and superposition) are the mystery!

 

[xts]

Because you seem to suggest collapse is simply a process whereby previous unknown but existing facts (elements of reality) are revealed.

So you totally missed my point!
I suggest that 'collapse' is not a process at all - it causes no change to reality - it is only one of the mathematical operations we may perform to make our calculations easier (like reducing fractions).
I suggest there is no real difference between 'collapsed' wavefunction and 'uncollapsed' one, except of fixing some variables, which may be fixed later as well.

From your various posts, you also appear to think entanglement and non-local correlations in Quantum Mechanics are pretty trivial ideas, and require no more explanation than classical wave-mechanics or similar.

That's not true. Just contrary. They are not trivial, and there is something deeply counterintuitive in Bell's nonlocality. Nevertheless, most of the 'mysterious' experiments discussed with emphasis do not exploit that Bell's mystery. All Quantum-Eraser and similar experiments, use Berlemann: only they take from entanglement is that Dr.B randomly puts a pair of pink-green socks.
If you took my post as denying Bell - you are more than wrong. I am just telling people, who are impressed by titles like 'quantum eraser', that not everything shining is gold and the experiments they are so excited are no more mysterious than 200 years old Young's experiment

Whatever the world is, it is not the naive world of classical realism, and there is (still) a mystery about how Quantum Mechanics works. It is ridiculous to say QM is pretty straightforward apart from the mystery of entanglement. Entanglement (and superposition) are the mystery!

I fully agree.
Feynmann: "entire mystery of quantum mechanics is in the double-slit experiment".
Yes, I agree, both Young's and Aspect's experiments are mysterious!
But all the rest are either overinterpreted or equivalent to one of those two.

I can't find any place for 'collapse' in reality - even in this mysterious reality of Young and Aspect - no justification to call it 'physical process' rather than 'calculation trick'.

ADDED>
Thanks for making me to read Bell's "Bertlemann's socks" again!

I am just advocating Bohr's view, quoted by Bell as:
There is no quantum world. There is only an abstract quantum mechanical description. It is wrong to think that the task of physics is to find out how Nature is. Physics concerns what we can say about Nature.

The idea of 'collapse' does not tell us anything about Nature...

 

[PAllen]

The idea of 'collapse' does not tell us anything about Nature...

Doesn't the idea of collapse correspond to a certain number of pages of "Road to Reality"? Without it, wouldn't this book be significantly shorter?

 

[xts]

Doesn't the idea of collapse correspond to a certain number of pages of "Road to Reality"? Without it, wouldn't this book be significantly shorter?

Sure! There is a whole chapter devoted to ontology and interpretations of 'measurement paradox'.

Anyway, the very first quote from Penrose's book on this issue I found appropriate is:
The jumping of the quantum state to one of the eigenstates of Q is the process referred to as state-vector reduction or collapse of the wavefunction. It is one of quantum theory’s most puzzling features, and we shall be coming back to this issue many times in this book. I believe that most quantum physicists would not regard state-vector reduction as a real action of the physical world, but that it reflects the fact that we should not regard the state vector as describing an ‘actual’ quantum-level physical reality.

Emphasis - mine.

 

[unusualname]

Roger Penrose's ideas on Quantum Mechanics are not really respected. He's had some interesting ideas on General Relativity and Geometry, but on QM he's almost crackpot status.

 

[Ken G]

But note that Penrose in that quote is testifying not just to his own opinion on the matter, but to the general view of the physicists he knows! I'll bet that's quite a few of them. I happen to think he is dead-on right-- and so was Bohr. It is just demonstrably true that physics is about what we can say about nature, to think otherwise is to enter into a kind of pretense that has never been correct for thousands of years of doing physics.

 

[andrewkirk]

For example, the *lower* bound for collapse seems to be 10,000 c according to experiment. However, other experiments give results which seem to have time order reversed (collapse before entanglement, and yes I know it sounds impossible).

I would love to understand what those experiments are. Can you provide a link to anything describing what was done? I am intrigued by how one can tell whether a wave function has collapsed.

When you say the lower bound for collapse is 10,000c do you mean that some experiment has been performed which was able to demonstrate that, when a measurement of quantity m was performed on a particle at A at rime t, the wavefunction of its entangled twin at B had 'collapsed' to an eigenket of the operator M, no later than t+d(A,B)/(10,000c) where d(A,B) is the spatial separation from A to B?

Thank you, and the others, for the responses. I find this fascinating. I had never heard of Dr Bertelman's socks before.

 

[Cthugha]

I would love to understand what those experiments are. Can you provide a link to anything describing what was done? I am intrigued by how one can tell whether a wave function has collapsed.

The experiment can be found in: "Testing the speed of 'spooky action at a distance'" by D. Salart et al., Nature 454, 861-864 (2008).

If you do not have a sucscription to Nature, it is also available for free here: http://arxiv.org/abs/0808.3316" .

However, one should take this with a grain of salt. See for example Zeilinger's comment on that paper: http://arxiv.org/abs/0810.4452"

 

[Fra]

there is never any need to imagine that anything "instantaneous" happened to the other particle-- instead, what happened instantaneously is your information about and expectations for that other particle changed. Since those changes occurred in your head, it is not surprising that they could be instantaneous-- you did an experiment on a system, and gained instantaneous information about that system.

I think that OP is answered and I pretty much agree with Ken's view, but I'd like to make a note.

I don't quite like when we using phrases like "in your head" as it sounds like the wavefunction somehow is something that exists only in the human brain only. This way to expressing it is I think why some people react, and for good reasons. I do not think that in an way that demonstrated quantum mechanical interactions between subatomic particles, such as spectra lines etc has anything at all to do with human brain. It rather has to do with how electrons, nucleus and fields "see each other". ie. Why would the stability of atoms have anything to do with collapses in our heads?

It makes no sense. I suspect Ken would agree on this too so far?

Maybe there is a better way of putting it.

I like to think of the information observer O has about system A "living" in the observing system O, encoding information about the observed system A. Of course the converse also holds, the information A has about O is encoded in A. The action of O is local in the sense that it depends only on information at has, the same goes for A.

But in the most general case, construction actions in a rational way like you construct oddds from a set of empirical evidence is simply not yet understood. Instead what we do have, is that in hte limit of O -> classical and A is much smaller compared to A, the regular hilbert space structures works. In this way the problems where O is UNABLE to encode all historically available information of A is removed, since when I is classicla it has effectively infinite information capacity. But then we hit into another problem which is that we get hard to compare infinites, when considering two quantum interacting classical systems that are strongly coupled.

So if we replace encoding "in our head" with in the encoding "observing systems microstructures or "memory"" I feel more at ease, as it removes the human qualities which I honestly think has noting to do with it, except in the most superficial obvious way that EVERYTHING we do are "just" human activities.

/Fredrik

 

[Fra]

Roger Penrose's ideas on Quantum Mechanics are not really respected. He's had some interesting ideas on General Relativity and Geometry, but on QM he's almost crackpot status.

I suppose you're referring this Penrose idea of gravitationally induced objective collapse?

I think Penrose is touching on an interesting possibility there that connects inertia and gravitation, but I do not at all share his quest for objective collapse. I think the other way around is more interesting - to use the subjective collapse to explain gravity as emergent. Penrose lookes for the other way around.

/Fredrik

 

[Ken G]

It makes no sense. I suspect Ken would agree on this too so far?
Maybe there is a better way of putting it.

Perhaps the key is to recognize the wave function is in our heads, without saying it is only in our heads. The first part is undeniably true, but the word "only" carries unwanted connotations.

I agree with you that it is a process of building expectation and noticing surprise, but I don't think it is something that atoms literally do to other atoms, I think it is something that intelligence as we know it does as we interact with what we interpret as atoms. Indeed, I regard that as a demonstrable fact. But there must be something going on that we benefit from describing as "atoms doing to other atoms"-- it wouldn't work so well if there wasn't some connection to reality. All I'm saying is our fingerprints are all over our language about atoms, and always will be. Bohr said it best-- physics is about what we can say about nature. The three key parts are "us", "nature", and "language". Sure sounds like a conversation to me. When we are in a conversation, are we part of the conversation?

I like to think of the information observer O has about system A "living" in the observing system O, encoding information about the observed system A. Of course the converse also holds, the information A has about O is encoded in A. The action of O is local in the sense that it depends only on information at has, the same goes for A.

I see a lot of value in your prescription. I would just add that the information you refer to is always interpreted by O-- even when O is attributing that information as being something A "knows" about O. Atoms don't really know, we do, but we can imagine a "mini me" sitting on that atom, and that might be a valuable tool-- as long as we don't attribute superhuman powers to the "mini me", it must process the same information we do and in the same ways. We're always in the conversation.

In this way the problems where O is UNABLE to encode all historically available information of A is removed, since when I is classicla it has effectively infinite information capacity.

I'm not sure that making things classical affords them added information capacity-- I see it the other way around. A single electron has vastly more information capacity than the ability of a brain to understand that electron. The act of understanding is a process of reduction, where we throw out huge amounts of information to find the "beating heart" of the process, which is really just the part we care about. Thermodynamics is a classical example-- if I have N free particles, I have 6N things to know about those particles classically (and much more quantum mechanically), but in the limit as N gets huge and we are in thermodynamic equilibrium, I only have 2 things to know-- temperature and density. So we have three levels of "classicalness", and the more classical, the less information.

I believe this is why we perceive "collapses"-- our brains cannot handle the full amount of information there (which is much more than a wave function, the wave function is already a reduction because it only answers the kinds of questions we are capable of asking), so we focus on "measurements", and that very limited vocabulary for talking about nature forces upon us the concept of "mixed state." The combination of that concept, and the act of perception, forces upon us the concept of "collapse." Each is a loss of information-- bringing nature into our heads is an incredibly lossy process. But the MWI camp imagines they can bring it into their heads and unpack it into what it was before. I view that as inverting the uninvertible.

So if we replace encoding "in our head" with in the encoding "observing systems microstructures or "memory"" I feel more at ease, as it removes the human qualities which I honestly think has noting to do with it, except in the most superficial obvious way that EVERYTHING we do are "just" human activities.

I agree that the word "just" in that sentence would be unwanted, but the sentence without the "just" is still something I view as crucial. We would have to enter into a kind of pretense to ignore that sentence.

 

[xts]

but we can imagine a "mini me" sitting on that atom, and that might be a valuable tool-- as long as we don't attribute superhuman powers to the "mini me", it must process the same information we do and in the same ways.

We must be very carefull using such views. It is veeeery easy to fall in some pitfall thinking that way.
Great example of such fallacy is Lev Veidman's article "On schizophrenic experiences of neutron" (arXiv:quant-ph/9609006 v1), denying non-realism (thus leading to Many Worlds), because of absurdity of the world viewed from the perspective of quantum particle.
It is very difficult for us (impossible for L.Veidman) to imagine ourselves as quantum particles (especially: deprived of all sensory information we always have).

 

[Ken G]

It is very difficult for us (impossible for L.Veidman) to imagine ourselves as quantum particles (especially: deprived of all sensory information we always have).

I agree, the "mini me" has to follow all the rules of gedankenexperiments-- it's perceptions must not be conceptualized, they must be actualized by some possible (in principle) experiment, keeping in mind that the presence of such an experiment is probing a fundamentally different reality (one that includes that experiment). We are not imagining that "we are neutrons", that would be replacing the reality with a physically impossible one. By "mini me", I meant the opposite-- we are not bringing ourselves to the neutron, we are bringing the neutron to us. That will cause collapse and so on, and will be a different reality than the one without the "mini me." This is indeed a severe limitation, that's what I think is the challenge that Fra's envisaged prescription faces: how to discuss a reality that doesn't exist as a way of knowing something about one that does.

 

[UChr]

Dear generous and helpful physicists,

* just a matter of interpretation, with no practical implications.

Either one can show that there are differences - for example to change the sequence of experiments provide changes in measurements - or: there's maybe just talk about measurement problems when the distances are significant enough to result in A can be detected before B is measured.

See for example [3] N Gisin, B Gisin: ‘A local hidden variable model of quantum correlation exploiting the detection loophole’, Physics Letters A 260, 323--327 (1999)
http://www.gap-optique.unige.ch/wiki/_media/publications:bib:pla240799.pdf

 

[DrChinese]

See for example [3] N Gisin, B Gisin: ‘A local hidden variable model of quantum correlation exploiting the detection loophole’, Physics Letters A 260, 323--327 (1999)
http://www.gap-optique.unige.ch/wiki/_media/publications:bib:pla240799.pdf

Not that the reference is bad, but it appears to have no obvious connection to collapse. Assuming that this thread is not turning into a debate on the interpretations themselves, Bell loopholes, local realism, etc.

Please, don't take us there!

 

[Fra]

Ken, I think we almost agree, but I sense that I like to be more radical/speculative than you. I mean in the MWI vs CI discussion we agree, but withing the "CI variants" camp I think I want to go further.

My point is that I think "atoms do know", atoms do have expectations of other atoms. Except of course the big question is what mathematical formalism that may be a good abstraction of this, and how to use this to construct expectations of composite systems from the point of view of another observer. This is the area where I have some crazy ideas.

Given that we agree quite well, I honestly think that you are soundly sceptical to these more radical ideas simply because it is not at all clear exactly how to make sense out of them. This makes perfect sense.

But let's for a second assume that I can make sense out of these concerns, then I would think that it would be hard to deny that we have made some non-trivial progress also somehow in line with the original CI spirit. Because it would use the empirisim to explain the origin of interactions even, in a very deep way.

/Fredrik

 

[Ken G]

Ken, I think we almost agree, but I sense that I like to be more radical/speculative than you.

More power to you-- faint heart never won fair maid, nor fair breakthrough.

My point is that I think "atoms do know", atoms do have expectations of other atoms. Except of course the big question is what mathematical formalism that may be a good abstraction of this, and how to use this to construct expectations of composite systems from the point of view of another observer. This is the area where I have some crazy ideas.

I would probably have to be dragged kicking and screaming into the notion that atoms actually do know, but I would certainly entertain the notion that if we model atoms as if they thought like we did, perhaps we could understand their interactions better from our own point of view. If so, that would presumably not be because atoms really do know, but rather, because the way we perceive and interpret their interactions is so deeply intwined with what we know, that there is analytical power in treating them as if they "thought" like us.

But let's for a second assume that I can make sense out of these concerns, then I would think that it would be hard to deny that we have made some non-trivial progress also somehow in line with the original CI spirit. Because it would use the empirisim to explain the origin of interactions even, in a very deep way.

Yes, I think success along those lines would very much fall under the heading of having an interpretation that was useful in designing the next theory. Once we have a theory all laid out, I think there is a lot of sense in the "shut up and calculate" approach, but when it comes to contrasting with past theories, or creating the next great theory, it is essential not to shut up at all.

 

[Fra]

I would probably have to be dragged kicking and screaming into the notion that atoms actually do know, but I would certainly entertain the notion that if we model atoms as if they thought like we did, perhaps we could understand their interactions better from our own point of view. If so, that would presumably not be because atoms really do know, but rather, because the way we perceive and interpret their interactions is so deeply intwined with what we know, that there is analytical power in treating them as if they "thought" like us.

Sure, you concerns here are precisely because it's unclear what exactly it means that "atoms know", and I agree with you, that I am certainly not an atom, so when I speak of atoms know, what I mean is a model where explanatory power is generated by assuming that atoms interacting with each other AS IF they hold expecetations about each other. Wether they "actaully do" in some realism sense is quite meaningless, becauase they I see atoms act is their only way to communicate to me.

This in fact no different that how other PEOPLE communicate to you. If we were to take the objections to "atoms know" to it's extreme, then you might even say that knowledge is not only constrained just to brains, but even just to KenG's brain! :) From the point of view of KenG, Fra's brain is no more different in principle than a pile of cells or an atom. you can only make indirect inferences from how I or the atoms act anyway.

When I say "KenG knows", byt the same token I REALLY don't see inside your head, all I can tell is that your actions are consistent with a certain presumed state of knowledge together with some conjectured mechnism explaiing your actions.

/Fredrik

 

[Fra]

I would probably have to be dragged kicking and screaming into the notion that atoms actually do know
...
that there is analytical power in treating them as if they "thought" like us.

They way I think about this it's not that atoms think like humans, rather than way my working direction is means that matter systems are characterised by their encoding structure.

To take the simplest possible example, you can consider how a counter, beeing able to count and store distinguishable states, is implicitly encoding expectationa of the future. The expectation of the future is simply based on "statistics" of history. But since the counter can't count ALL evens in an infinite history, when the counter is saturated, for each new data point it must randomly loose a count. Statisticall the spectrum of the released information conincides with the prior distribution encoded by the counter. Loosely speaking you can think of this as radiation, which is how the counter communicates with the environment. The counter is doing like a "random walk" on it's own prior.

This is just the simplset possible example of how an inanimate information structure can have expectations and know ="encode information".

So my abstraction of a general matter system, is an encoding structure, whose "random action" according to the "evolving prior expectation", should encode physical interactions as we know it. In this picture concepts like inertia are identified as information capacity, a system fully adapts to new information until it's satureated, then it startes to evolve by releasing random information according to the prior profile, as well as putt in the new information.

/Fredrik

 

[Ken G]

This in fact no different that how other PEOPLE communicate to you. If we were to take the objections to "atoms know" to it's extreme, then you might even say that knowledge is not only constrained just to brains, but even just to KenG's brain!

Yes, idealism is the only fully self-contained philosophy. But I reject it on symmetry principles-- I see myself as too closely symmetric to other people not to imagine they think like I do. But with atoms, I need not invoke any such symmetry.

When I say "KenG knows", byt the same token I REALLY don't see inside your head, all I can tell is that your actions are consistent with a certain presumed state of knowledge together with some conjectured mechnism explaiing your actions.

Yes, I'm fine with that perspective, it's a kind of "mini me" at the atomic level. There is probably no other option-- we have no choice but to think like we do, so any process that we are going to treat as involving information is going to have to involve what we mean by information.

 

[Ken G]

So my abstraction of a general matter system, is an encoding structure, whose "random action" according to the "evolving prior expectation", should encode physical interactions as we know it. In this picture concepts like inertia are identified as information capacity, a system fully adapts to new information until it's satureated, then it startes to evolve by releasing random information according to the prior profile, as well as putt in the new information.

I'm certainly game to see what you can come up with using that approach.