Register to reply

Why does quantum entanglement not allow ftl communication

Share this thread:
JesseM
#19
Apr25-08, 04:02 PM
Sci Advisor
P: 8,470
Quote Quote by Hans de Vries View Post
No, This is the reasoning of the EPR crowd:

1) The quantum state is teleported instantaneous.
2) We can not control the collapse of the wave-function.
3) Therefor we can not use it to communicate data.
4) Therefor no information is send.
Are you sure about the claim that the "quantum state is teleported instantaneously"? Do you have a reference? It seems to me that if that were the case, one could still gain probabilistic information about the original, distant state that was teleported by looking at the outcome when the teleported state was measured.
Ken G
#20
Apr25-08, 04:06 PM
PF Gold
P: 3,136
Quote Quote by Hans de Vries View Post
No, This is the reasoning of the EPR crowd:

1) The quantum state is teleported instantaneous.
There is no meaning to "instantaneous" except for the person doing the original experiment. Nothing is "teleported" unless there is classical slower-than-light communication, so it is not instantaneous.
3) Therefor we can not use it to communicate data.
That's not the correct reason why we can't communicate instantaneously, the correct reason is that nothing is transported instantaneously in the first place. I cannot speak for whoever you mean by "the EPR crowd"-- I agree that argument would be spurious, but it's not the right argument anyway.
Claim 4) violates Shannon's information theorem. that's my point
But it's all a strawman, that's my point. If the "EPR crowd" think they require that explanation, they don't understand information theory, but since a lot of people do, I don't see that as likely. There may be a difficulty in finding people interested in philosophy who are also versed in physics.
Of coarse, the majority of the EPR crowd doesn't believe this to be fundamental.
I get the impression that many of them are really chasing their Science Fiction
dreams and that statements like: Special Relativity is not really violated are
more to appease peer reviewers than that they themself believe in it.
I don't know what they believe, but I don't think personal beliefs are terribly relevant either.
Hans de Vries
#21
Apr25-08, 04:33 PM
Sci Advisor
P: 1,135
Quote Quote by Ken G View Post
There is no meaning to "instantaneous" except for the person doing the original experiment. Nothing is "teleported" unless there is classical slower-than-light communication, so it is not instantaneous.

Of coarse the whole concept of instantaneous propagation doesn't make any sense
at all in the first place if not defined with respect to a "preferred" reference frame.


Quote Quote by Ken G View Post
That's not the correct reason why we can't communicate instantaneously, the correct reason is that nothing is transported instantaneously in the first place. I cannot speak for whoever you mean by "the EPR crowd"-- I agree that argument would be spurious, but it's not the right argument anyway.

Ok, but you are using the expression "correct reason" in the sense of 2)
while I am using it in the sense of 1)

1) The reason which correctly describes the arguments used by the EPR people.
2) Something what you think (or something what I think) is the correct physics.


Regards, Hans
Hans de Vries
#22
Apr25-08, 05:20 PM
Sci Advisor
P: 1,135
Quote Quote by JesseM View Post
Are you sure about the claim that the "quantum state is teleported instantaneously"? Do you have a reference?
For instance in Zeilinger's popular article in the Scientific American (April 2000)
he claims:

Quote Quote by Zeilinger
By "spooky action at a distance", the measurement also instantly alters the
the quantum state of the faraway counter matter.
The article must be online somewhere.


Regards, Hans
JesseM
#23
Apr25-08, 05:21 PM
Sci Advisor
P: 8,470
Quote Quote by Hans de Vries View Post
Of coarse the whole concept of instantaneous propagation doesn't make any sense
at all in the first place if not defined with respect to a "preferred" reference frame.
It does make sense in the context of conventional nonrelativistic QM, though. Are you claiming that according to the equations of this theory, the quantum state of the teleported system goes instantaneously from one location to another, before the classical signal has had time to travel between the locations? If so I would like to see a reference for this.

edit: I see you reference a popular article above, but I'd like to see something more technical as the precise meaning of statements in popular articles is often unclear--Zeilinger may have only meant that the quantum state is instantly teleported after the classical signal has gone from one location to another and the information in that signal is used to prepare the far away system.
Ken G
#24
Apr25-08, 05:57 PM
PF Gold
P: 3,136
Quote Quote by Hans de Vries View Post
Of coarse the whole concept of instantaneous propagation doesn't make any sense
at all in the first place if not defined with respect to a "preferred" reference frame.
Right, and all the evidence we have is that there is no such frame.

Ok, but you are using the expression "correct reason" in the sense of 2)
while I am using it in the sense of 1)

1) The reason which correctly describes the arguments used by the EPR people.
2) Something what you think (or something what I think) is the correct physics.
I don't dispute that because I haven't seen the arguments of the "EPR people", and I agree that would be a wrong argument. I just didn't want to leave the impression that this represented the best argument against FTL communication.
Ken G
#25
Apr25-08, 06:00 PM
PF Gold
P: 3,136
Quote Quote by Hans de Vries View Post
For instance in Zeilinger's popular article in the Scientific American (April 2000)
he claims:
"By "spooky action at a distance", the measurement also instantly alters the
the quantum state of the faraway counter matter."
Sure, but that says nothing about anything being transported, not even random data. I don't know if he means this or not, but in my view the statement is perfectly correct, insofar as the "quantum state" is interpreted as "the way the observer doing the measurement would characterize the state of the faraway system". Personally, I don't know of any other meaningful definition of that phrase, but I agree that a lot of people seem to think there is one.
JesseM
#26
Apr25-08, 07:21 PM
Sci Advisor
P: 8,470
Quote Quote by Ken G View Post
I don't know if he means this or not, but in my view the statement is perfectly correct, insofar as the "quantum state" is interpreted as "the way the observer doing the measurement would characterize the state of the faraway system". .
I would interpret the quantum state to refer to the set of probability amplitudes for different outcomes when you measure the system. If you look at the schematic diagram on this page, I think the idea is that at the moment "C" turns green, it now has the same amplitudes that "A" had when it was green, up until the moment it was disrupted by becoming entangled with "B". The diagram suggests that C's state only becomes identical to A's original state (before being disrupted) after the classical data has been transmitted from the location of A to the location of C.
Ken G
#27
Apr25-08, 07:44 PM
PF Gold
P: 3,136
Quote Quote by JesseM View Post
I would interpret the quantum state to refer to the set of probability amplitudes for different outcomes when you measure the system.
So would I, but the question is, for whom? The problem with probabilities is many people treat them like absolutes, and ask questions like "what is the probability of...". But implicit in those kinds of questions is a host of information that is assumed to be known, along with a host of information that is assumed to not be known. Without those assumptions, probabilities are meaningless-- and those assumptions are often different for different people (witness a poker game).
If you look at the schematic diagram on this page, I think the idea is that at the moment "C" turns green, it now has the same amplitudes that "A" had when it was green, up until the moment it was disrupted by becoming entangled with "B".
Absolutely, and the information to do that was transported classically. Thus there is no issue all all when C "turned green", as it is a purely local event. What the observer back at A uses for the wave function of the entangled pair is irrelevant to making C "turn green", at least until the classical information arrives.
The diagram suggests that C's state only becomes identical to A's original state (before being disrupted) after the classical data has been transmitted from the location of A to the location of C.
Exactly.
JesseM
#28
Apr25-08, 07:47 PM
Sci Advisor
P: 8,470
Quote Quote by Ken G View Post
So would I, but the question is, for whom? The problem with probabilities is many people treat them like absolutes, and ask questions like "what is the probability of...". But implicit in those kinds of questions is a host of information that is assumed to be known, along with a host of information that is assumed to not be known. Without those assumptions, probabilities are meaningless-- and those assumptions are often different for different people (witness a poker game).
I'm not sure what you mean by "treat them like absolutes", but in theoretical QM every possible outcome for a measurement on a system is given an unambiguous probability amplitude, that set of amplitudes is essentially what the wavefunction for a system is.
JesseM
#29
Apr25-08, 07:52 PM
Sci Advisor
P: 8,470
Quote Quote by Hans de Vries View Post
For instance in Zeilinger's popular article in the Scientific American (April 2000)
he claims:
By "spooky action at a distance", the measurement also instantly alters the
the quantum state of the faraway counter matter.
The article must be online somewhere.
Found the article here. If you look on p. 54, where he discusses the experiment and again uses the word "instantaneous", in this context what he means is that if you have two entangled particles A and B, and you perform a certain type of "joint measurement" on A and another particle X, this will "instantaneously" create a 3-particle entangled system which also involves B. But at this point B's state is not actually identical to X's before the measurement, it only becomes identical when you interact with B in a certain way, making use of classical information about the outcome of the joint measurement on on A and X.
Ken G
#30
Apr25-08, 07:55 PM
PF Gold
P: 3,136
Quote Quote by JesseM View Post
I'm not sure what you mean by "treat them like absolutes", but in theoretical QM every possible outcome for a measurement on a system is given an unambiguous probability amplitude, that set of amplitudes is essentially what the wavefunction for a system is.
But that's just what I'm talking about-- there is no need to treat the wave function like it is unique, and in fact it is not. It is perfectly possible to imagine an experiment where two different participating physicists arrive at two different wave functions for the same system, based on different information about that system, and have "quantum mechanics work" perfectly well for both physicists. Indeed, that is precisely what can happen with entanglement. You might say "one of them has the complete wave function, and the other has an incomplete one" but there's no prescription in quantum mechanics for identifying a "complete" wave function-- we "go with the wave function we have". Indeed, Bohmians seem to feel we never are using the complete wave function. So my point is, just as "the probability" in poker is a completely relative concept, so is the "probability amplitude" of a wave function. This is annoying for people who like to think of the wave function as something real, but personally I cannot see the least bit of evidence to support that viewpoint, and it leads to all kinds of bizarre problems like "spooky action at a distance".
JesseM
#31
Apr25-08, 08:14 PM
Sci Advisor
P: 8,470
Quote Quote by Ken G View Post
But that's just what I'm talking about-- there is no need to treat the wave function like it is unique, and in fact it is not. It is perfectly possible to imagine an experiment where two different participating physicists arrive at two different wave functions for the same system, based on different information about that system, and have "quantum mechanics work" perfectly well for both physicists.
Can you give a specific example of what you mean?
Quote Quote by Ken G
Indeed, that is precisely what can happen with entanglement. You might say "one of them has the complete wave function, and the other has an incomplete one" but there's no prescription in quantum mechanics for identifying a "complete" wave function
Sure there is--if you measure a maximal set of communting operators for the system, that determines a unique wavefunction.
Quote Quote by Ken G
we "go with the wave function we have". Indeed, Bohmians seem to feel we never are using the complete wave function.
No they don't--I think you're confusing "wave function" with "complete state of the system, including hidden variables". Just because we know the complete wave function, that doesn't mean we're ruling out the possibility that there may be other hidden variables not accounted for by the wave function.
Quote Quote by Ken G
This is annoying for people who like to think of the wave function as something real, but personally I cannot see the least bit of evidence to support that viewpoint, and it leads to all kinds of bizarre problems like "spooky action at a distance".
Well, local realistic theories can be proved incompatible with QM just based on the statistics of measured outcomes predicted in QM, saying nothing about the wave function.
Ken G
#32
Apr25-08, 11:17 PM
PF Gold
P: 3,136
Quote Quote by JesseM View Post
Can you give a specific example of what you mean?
Sure, the standard 1/2-spin entangled pair with zero total angular momentum. If I do a measurement on one and get a spin of +1/2 in the "z direction", I will instantly make the wave function of your particle -1/2 in the z direction, and use that to predict the outcome of any experiment you do. You, on the other hand, will stick with a mixed-state wavefunction for your particle, with 50% up and 50% down. You will use that to predict the outcome of any experiment you can do, and you will do just fine. We both will, even though we make different predictions on that particular trial, because on an ensemble our predictions will be indistinguishable without looking at correlations (which would require slower-than-light communication to do).
Sure there is--if you measure a maximal set of communting operators for the system, that determines a unique wavefunction.
Not so. What if you do that on an entangled particle? You have no idea what correlations exist between your measurements and some other set of measurements, so your description is incomplete. What you have done is to assume that you have a single-particle wave function, but reality doesn't hand you that-- if you think a wavefunction is real, it must include everything your particle is entangled with. That's why even the wavefunction you describe is not "the complete wavefunction" that involves that particle, it is merely the most complete description you can find within the confines of a single-particle wavefunction (note also that the universe is full of identical particles and you are simply ignoring the exchange terms in the hope that they don't matter). I would say that a wave function is just a model, and hence reflects a choice by a physicist-- not a reality.
No they don't--I think you're confusing "wave function" with "complete state of the system, including hidden variables". Just because we know the complete wave function, that doesn't mean we're ruling out the possibility that there may be other hidden variables not accounted for by the wave function.
True, but if the wave function does not include that information, then it is not "the reality". That dovetails with my claim that a wave function is simply a reflection of the information we are choosing to use. That must have something to do with the reality or it would not be so useful, but "the reality" has to include more information than we are using. (Indeed, even if the Bohmian approach is a good model, I would say it still isn't going to be "the reality" because even if you have all the information, information is still reality passed through a filter, not reality itself-- but that gets philosophical).
Well, local realistic theories can be proved incompatible with QM just based on the statistics of measured outcomes predicted in QM, saying nothing about the wave function.
Correct, but a wave function is not based on local realism, so most seem to hold that wave functions are real, that there is such a thing as "the wave function" of a particle, or more correctly, a universe. Why they believe that is pretty much a mystery to me.
JesseM
#33
Apr25-08, 11:52 PM
Sci Advisor
P: 8,470
Quote Quote by Ken G View Post
Not so. What if you do that on an entangled particle? You have no idea what correlations exist between your measurements and some other set of measurements, so your description is incomplete.
What I said was that "if you measure a maximal set of commuting operators for the system, that determines a unique wavefunction"--it may not have been sufficiently clear, but what I meant was that for any entangled multiparticle system, you would have to measure a maximal set of commuting operators for all parts of the system to construct a wavefunction, not just a single particle.
Quote Quote by Ken G
True, but if the wave function does not include that information, then it is not "the reality". That dovetails with my claim that a wave function is simply a reflection of the information we are choosing to use.
I never said the wave function was "the reality", just that all the probability amplitudes can be uniquely determined with the right kind of measurements--if you've made these measurements, you don't have any "choice" of what the wavefunction should look like, even if there could be other realities to the system that aren't specified by the wavefunction.
Ken G
#34
Apr26-08, 12:24 AM
PF Gold
P: 3,136
Quote Quote by JesseM View Post
What I said was that "if you measure a maximal set of commuting operators for the system, that determines a unique wavefunction"--it may not have been sufficiently clear, but what I meant was that for any entangled multiparticle system, you would have to measure a maximal set of commuting operators for all parts of the system to construct a wavefunction, not just a single particle.
How do you know what the entangled system is? You still have to specify the system, you have to decide what entanglements you want to track, so you are still making a choice. The only system the universe hands you is the whole universe, so the only "complete" wavefunction of a system is a maximal set of all commuting operators for the whole universe. That's impossible, because an "operator" is an observable, which implies you have to do the observation from outside the system, i.e., outside the universe (there's a self-referential problem, I mean). So in reality you will consider a subsystem, but any subsystem you specify will still suffer from the incompleteness problem, because you cannot trace the entanglements and so will still be losing information about potential correlations. Completeness is impossible, so why do we pretend it isn't? Because we can achieve effective completeness in our chosen model-- but hey, it isn't the reality, which is all I'm saying.
I never said the wave function was "the reality", just that all the probability amplitudes can be uniquely determined with the right kind of measurements--if you've made these measurements, you don't have any "choice" of what the wavefunction should look like, even if there could be other realities to the system that aren't specified by the wavefunction.
The issue I was addressing is if there was a unique wavefunction that includes everything that is real about a system, or if it is merely a way for us to encode whatever information we have about the system. In other words, when we say we know "the wavefunction" in some absolute way, can we address not just questions like "what is the probabilty I'll measure X", but also questions like, "what is the probability I'll measure X given that some other entangled system gave result Y"? The answer is no, even with what you are calling the complete wavefunction for that system, we cannot answer the latter questions.

So the price for defining what you mean by a "complete wavefunction" is to rule out that you can address everything that is real about it. That is the choice-- you have chosen a weak form of completeness that cannot encompass all that is real about your system. That suffices to establish that a wavefunction is an expression of a choice we have made about modeling systems, not a complete description of the reality. That's all I'm claiming-- every physicist with different information about a system will successfully use a different wave function to describe it, and moreover, none of them will be using a complete description of all that is real about that system, so none have a claim to knowing "the wave function" of the system.
JesseM
#35
Apr26-08, 12:30 AM
Sci Advisor
P: 8,470
Quote Quote by Ken G View Post
But the point is, you still have to specify the system, so you are still making a choice. The only system the universe hands you is the whole universe, so the only "complete" wavefunction of a system is a maximal set of all commuting operators for the whole universe.
Only in the MWI, where measurements are themselves just new entanglements, is this really true. In Copenhagen QM, the act of measuring a particle can destroy previous entanglements it may have had up until that measurement (though it won't always, it depends on what measurement you perform)--subsequent measurements on this particle won't show any correlations with other particles it was entangled with prior to the first measurement.
Ken G
#36
Apr26-08, 12:54 AM
PF Gold
P: 3,136
Quote Quote by JesseM View Post
Only in the MWI, where measurements are themselves just new entanglements, is this really true. In Copenhagen QM, the act of measuring a particle can destroy previous entanglements it may have had up until that measurement (though it won't always, it depends on what measurement you perform)--subsequent measurements on this particle won't show any correlations with other particles it was entangled with prior to the first measurement.
Note I did some editing of my last post, as we're exchanging in real time! But even in the CI, the act of measuring does not destroy previous entanglements (as usual, such entanglements only show up in correlations with other measurements, never on measurements of the same system). The CI is simply more honest that you have made a choice not to track them, so the CI makes no claims that the wave function is a complete description of the reality-- even for a maximal set of commuting observations. You are right that the MWI does try to retain that "reality" property, but it still fails unless you include the whole universe in the wave function. That's the problem with MWI in the first place, there's no evidence that such a wave function exists, and we certainly know we can never use it for anything. So with MWI, all I'd have to say is "any wavefunction that any physicist could ever actually use for anything cannot be the wavefunction of that system without losing some of the reality of the situation", it's still always going to reflect a choice of some kind in MWI or CI, or Bohm.


Register to reply

Related Discussions
Communication using quantum entanglement Quantum Physics 5
Re: Quantum communication might be possible? General Physics 3
Re: Quantum communication might be possible? General Physics 2
Re: Quantum communication might be possible? General Physics 3
Quantum Entanglement and Communication Quantum Physics 21