Why is it assumed communication through entanglement would be FTL?

[vanhees71]

What do you mean by "the concept of locality to apply to correlations"? For me it doesn't make sense to say "correlations are local or nonlocal". If you refer to entanglement, it's very clear that it describes correlations between far-distant parts of quantum systems, but this has nothing to do with locality or nonlocality of interactions, i.e., microcausality. There's no constradiction between microcausality and inseparability. Entangled states are no problem within standard relativistic QFT. It's even the rule rather than the exception since already the Bose or Fermi nature of indistingushbable particles implies usually entangled states: Product states are rare since you need to symmetrize or antisymmetrize them, and that's all done automatically using the field operators to describe the states.

 

[gill1109]

Quantum mechanics itself does not predict "action at a distance". On the contrary, it predicts that there is *no* action at a distance. I am talking about action at a distance with respect to things in the manifest, external world. Things most people would certainly admit to being "real". The outcomes of coin tosses, fed into some apparatus by switching a switch. Whether or not, shortly later, detectors click and the clicks are registered indelibly, more or less, on a computer hard-disk. Bell, and the experiments, show that if the randomness of those clicks is to be explained by a "classical-like" explanation behind the scenes, in a somewhat microscopic and perhaps hidden (i.e., not directly accessible) layer, then that explanation has to be non-local. ie there is action at a distance *behind the scenes*, somewhat mysteriously not expressed in any observed action at a distance in our real world of clicks and coin tosses, tables and chairs. (Or worse still: there is predetermination, sometimes known as "conspiracy" or "no freedom").

Now, quantum mechanics does predict that there are things one can do with violation of Bell inequalities. For instance, Alice and Bob can construct shared random keys which are certainly unknown to any other agent in the universe. More prosaically, Alice and Bob can coordinate their re-actions to a joint enemy who is attacking them simultaneously on their two fronts, so separated that neither Alice nor Bob, the local commanders, knows which threat the other is experiencing. They can do this coordination through the sharing of quantum entanglement in ways which classically could only be done by having extremely rapid carrier pigeons or fax machines. On the other hand, if they did have those means, they could have done even better. 2 sqrt 2 is bigger than 2, but smaller than 4. There is now some explanation of why even without QM, Hilbert spaces and all that, one should not have expected anything better than 2 sqrt 2. This is due to a completely independent and intuitively reasonable principle of information causality, put forward by Marcin Pawlowski and others some 10 or so years ago.

I think that quantum mechanics forces us to reconsider notions like "real" and "local". Many researchers have managed to argue that QM is both local and real through sophisticated new metaphysical definitions of local and real. I think that this is just escaping the quandary through sophisticated (Jesuitical) word games. [Is God three or one? Is Jesus human or God? Is God all-powerful and caring of us poor human beings, and "therefore" he gave us free will, and let us do stupid things, consigning us to hell unless we are one of the few chosen ones?]. The academics of centuries have figured out intellectual ways to believe in several contradictory things at the same time, but this never changed the facts of life.

Alternatively, quantum mechanics forces us to retreat somewhat from the idea that we can understand the physical world by (in our minds) splitting it into pieces and understanding each piece separately.

Back to quantum teleportation. Quantum teleportation is something which only makes any sense *inside quantum mechanics*. It can work because of some simple linear algebra in C^8, the Hilbert space of three qubits. What it is, is the transportation of an unknown quantum state from location A to location B by the performance of a measurement at A which results in one of four completely random outcomes and the complete destruction of the state to be teleported. At this point, Bob can't see that anything has changed. Alice then informs Bob by carrier pigeon which of the 4 outcomes she saw. Bob carries out one of four deterministic operations on the qubit which he already had. And in the end, it's in the same state as the one which Alice was given. Since a mixed state of one qubit needs three real numbers to describe it perfectly, three completely unknown real numbers have instantaneously jumped from A to B by the transmission or two completely uninformative bits. But you couldn't have extracted those three numbers from the original qubit, anyway. So: so what?

Punchline: no-one *understands* quantum mechanics so not surprisingly, no-one understands quantum teleportation either. We can shut up and calculate, if that makes us happy. Or we can look on in wonder, get skilled with the calculations, and figure out more amazing things which QM allows to be possible in the real world, which might even be useful.

 

[Lord Jestocost]

I think that quantum mechanics forces us to reconsider notions like "real" and "local"

Exactly. But that will call for courage in case one has to abandon the notion of "reality".

 

[vanhees71]

It should first of all trigger us as physicists to define what we mean with "reality". This notion has been blurred up so much by zillions of different philosophical schools that it is pretty useless in a scientific context. Often what's meant is a deterministic description of nature but not always. In the context of foundational debates of QT it's better to avoid it completely and formulate in clear (mathematical!) terms what's meant in each case under debate.

 

[Lord Jestocost]

It should first of all trigger us as physicists to define what we mean with "reality"

The first step would be to accept that we can never be certain whether all of our putative outer experience is not mere imagining.

 

[gill1109]

Obviously, we never can be certain. But we could take it as some kind of working assumption. In fact, we could even say that our experience leads us to imagine it as very likely. After all, we can quite successfully predict what experiences we will get in the future, by "pretending" that there is an external reality, common to other "agents". So whether or not it is true, it certainly seems to be useful.

 

[Lord Jestocost]

I agree. The "external reality, common to other 'agents'", which I would call "empirical reality", is - so to speak - merely in our observations/perceptions which - when allowing intersubjective agreement - can be the objects of science. Eddington: "In science we study the linkage of pointer readings with pointer readings." Or, as pointed out by James Jeans in “PHYSICS & PHILOSOPY” (1948):

“Complete objectivity can only be regained by treating observer and observed as parts of a single system; these must now be supposed to constitute an indivisible whole, which we must now identify with nature, the object of our studies. It now appears that this [the object of our studies] does not consist of something we perceive, but of our perceptions, it [the object of our studies] is not the object of the subject-object relation, but the relation itself."

 

[zonde]

“Complete objectivity can only be regained by treating observer and observed as parts of a single system; these must now be supposed to constitute an indivisible whole, which we must now identify with nature, the object of our studies. It now appears that this [the object of our studies] does not consist of something we perceive, but of our perceptions, it [the object of our studies] is not the object of the subject-object relation, but the relation itself."

Well, this does not seem right. We gain objectivity by getting as far as possible from any specific subject-object relations. We use different subjects, different contexts while keeping the same object and look for common part from all these different relations.
But this statement seems to propose to go the opposite way to gain "complete" objectivity.
And anyways what is "complete objectivity"? Science approaches reality asymptotically. Considering this, "complete objectivity" should be unreachable anyways.

 

[Lord Jestocost]

As Ulrich Morhoff puts it in "“B” is for Bohr":

"The hallmark of empirical knowledge is objectivity. To Kant, objectivity meant two things: intersubjective agreement, and the possibility of thinking of appearances as experiences of objects." [bold by LJ]