Scientific American has an article by David Albert and Rivka Galchen with the New Scientist-ish headline Was Einstein Wrong?: A Quantum Threat to Special Relativity and the sub-head "Entanglement, like many quantum effects, violates some of our deepest intuitions about the world. It may also undermine Einstein's special theory of relativity."
An alternate title for this post might be "Son Of Why I Won't Make It as a Philosopher," because I really don't know what to make of it. The authors make authoritative-sounding references to a bunch of papers I haven't read, but then they also drop in passages like this:
Entanglement lies behind the new and exceedingly promising fields of quantum computation and quantum cryptography, which could provide the ability to solve certain problems that are beyond the practical range of an ordinary computer and the ability to communicate with guaranteed security from eavesdropping [see "Quantum Computing with Ions," by Christopher R. Monroe and David J. Wineland; Scientific American, August 2008].
But entanglement also appears to entail the deeply spooky and radically counterintuitive phenomenon called nonlocality—the possibility of physically affecting something without touching it or touching any series of entities reaching from here to there. Nonlocality implies that a fist in Des Moines can break a nose in Dallas without affecting any other physical thing (not a molecule of air, not an electron in a wire, not a twinkle of light) anywhere in the heartland.
Even leaving aside their disturbing choice of an example system, this just isn't true.
Or, to be a little more charitable, it's so badly phrased that it creates a deeply misleading impression. It's true that entanglement allows measurements made at one point in space to affect measurements made at a different point in space, but it's not magic. The non-local correlations between states that are the essence of entanglement have to be established through local interactions.
It is not possible to produce entanglement in a truly non-local manner. Every physical system in which entanglement has been demonstrated has had that entanglement generated through local interactions. The correlated photons used to test Bell's theorem originate from the same atom, or the same non-linear crystal. You don't get more local than that.
While it is possible to establish entanglement between two objects that have never been in contact, the crucial step involves local interactions between particles that have previously been entangled. "Quantum teleportation" uses entangled states to transmit photon information from one place to another, but it does so using an entangled pair of photons that started in the same place. Even the recent teleportation experiment from Chris Monroe's group, which entangled the states of two physically separated ytterbium ions, did so using photons that were entangled with each other by passing through the same beamsplitter.
So, the best possible interpretation of Albert and Galchen's non-local punching example is that it's incredibly badly written. For a fist in Des Moines to break a nose in Dallas, both fist and nose would need to have been in the same place-- Oklahoma City, say-- at some point in time, and interacted with each other there. Then, they would need to be separated without measuring their state, at which point the measurement of bruised knuckles in Iowa would produce a broken nose in Texas. That's still odd, but not nearly as magical as what they seem to be claiming.
The worst possible interpretation of their non-local punching example is that they've fallen into the sort of woo-woo mystical crap promoted by people like Deepak Chopra and Lionel Milgrom, in which "entanglement" is code for "magic." I doubt this is really the case, but it's hard to tell from what they've written.
So, as I said, I really don't know what to make of this article. They talk as if they've hit on something new and deep in the phenomenon of entanglement, but their examples are either solved problems or deeply confused. And the descriptions are so vague I can't tell which.
- Log in to post comments
I'm sure this will feed the various anti-relativity blogs and sites for a while, too....
I find the term quantum non-locality incredibly misleading. There are deep issues to do with reconciling causality and locality of interactions and quantum mechanics, and their resolution necessitates the introduction of quantum fields and the infinitely intricate structure of quantum field theory. This would be the interesting story to tell, except that it's old news: this development started in the 1930s, and by now is very well-understood.
Incredibly, this is often not part of the discussion when people, even serious and famous people, wax poetic about QM and special relativity. Here, in a particularly distasteful example, we are explicitly told that there is some unspecified contradiction between SR and QM, ignoring nearly a century of good solid science.
As you point out, non-local correlations are easy to generate even classically. The QM states can be correlated in stronger and more subtle ways than classical ones, but that has nothing to do with spatial relations between objects, and the locality in space. This is of course even an older subject than the story of QFT, but maybe there is a fresh way to retell it yet one more time.
http://www.csmonitor.com/2009/0317/p02s04-usgn.html
I've been told on occasion that philosophy of science and sociology of science types have useful things to say that scientists should not dismiss. It's hard to take seriously when a philosopher of physics can pontificate in Scientific American about how QM and SR are irreconcilable while being ignorant of their 80 years of happy (and exquisitely experimentally verified!) coexistence.
On the other hand, maybe this does bring up a flaw in most physics curricula, where people study QM and learn that it's weird and nonlocal, and then study QFT which is local, and are never quite confronted with how to reconcile these things. Somehow it should be stressed more that interactions in QFT are local, but states can be entangled in nonlocal ways. (And indeed, even the vacuum in QFT is a very weird, entangled, nonlocal quantum state, a fact that gets obscured by the way field theory texts build up states from Fock spaces in free theories.)
As somewhat of an insider I can say that the article definitely reads as if the author only spoke to the hardcore realist group of philosophers of QM (I suspect the list includes Albert). I guess they deserve their voice in the mainstream science media as much as anyone else. However, although some of the work described is quite worthy, I didn't find the article very newsworthy.
The Tim Maudlin book is definitely a classic and anyone interested in understanding Bell inequalities from both the physical and conceptual side should definitely read it. However, the article makes it sound like nobody else did any deep work on the subject before Maudlin. It is also the case that although the book should be very influential, I suspect that very few physicists and mathematicians who work on quantum foundations have actually read it. It is only really required reading on the philosophy side of the subject, in which it is regarded as the nonlocality bible. In my opinion, it should not be held in such awe, but equally people should read it if they want to discuss the subject seriously.
Tumulka's work is interesting and worthwhile (disclaimer: he is a former collaborator). Anything that gets GRW or Bohm type models closer to being Lorentz invariant is definitely a good thing. However, I'm not sure that I'd call it groundbreaking, since it doesn't exactly get us closer to any novel predictions. It's just one approach amongst many plausible possibilities in foundations.
As for Albert's non-narratability stuff, I personally think it is not very interesting. To find it surprising, you have to take the view that quantum wave vectors are "states of reality" (or ontological states in the jargon). If you think they are "states of knowledge" (epistemic states) then it is not too surprising because classical probability distributions are also non-narratable in exactly the same way. The hardcore realists generally just refuse to see epistemic quantum states as a possibility, so anything that can be easily explained by analogy with classical probability is still considered weird by them. In any case, the paper is very accessible for anyone with a basic knowledge of QM, so you can read it and decide for yourself. Alternatively, Wayne Myrvold gave a talk on the subject at Perimeter, which you can watch here: http://www.pirsa.org/08090079/
I've been told on occasion that philosophy of science and sociology of science types have useful things to say that scientists should not dismiss. It's hard to take seriously when a philosopher of physics can pontificate in Scientific American about how QM and SR are irreconcilable while being ignorant of their 80 years of happy (and exquisitely experimentally verified!) coexistence.
This is probably a little harsh, as the article does mention some current research-- there's a reference to this paper (I think-- there are no citations in the article), for example. I'm not really plugged into that community, but the author of that article is certainly active on the arxiv, and several of those papers are in well-regarded journals. I don't think the authors of the SciAm piece are totally uninformed-- at the very least, they're aware of papers that I'm not.
There seems to be an active community of research on this stuff, and I'm not prepared to write the whole thing off as worthless speculation. I don't quite understand what it's about, but then I don't really understand lots of theoretical physics that is unquestionably valid research.
Matt snuck in while I was typing, with: As for Albert's non-narratability stuff, I personally think it is not very interesting. To find it surprising, you have to take the view that quantum wave vectors are "states of reality" (or ontological states in the jargon). If you think they are "states of knowledge" (epistemic states) then it is not too surprising because classical probability distributions are also non-narratable in exactly the same way. The hardcore realists generally just refuse to see epistemic quantum states as a possibility, so anything that can be easily explained by analogy with classical probability is still considered weird by them. In any case, the paper is very accessible for anyone with a basic knowledge of QM, so you can read it and decide for yourself.
Do you have a cite for the article? Is it available online somewhere?
Entanglement cannot transfer information faster than lightspeed. One fails to see any problem with Relativity vis-a-vis causality. Entangled data transfer is instantaneous across arbitrary distances and through aribitrary volumes. Interpretation's components cannot superluminally propagate. Both barn doors can and cannot close while the relativistic pole proceeds through according to the needs of the observers. The universe does not tolerate contradiction.
Not "at which point the measurement of bruised knuckles in Iowa would correlate with a broken nose in Texas"?Is there research that shows it is cause and effect from the measurement, and couldn't be explained by 'the nose met the fist in OK City, but neither realized it til they got home'?
"For a fist in Des Moines to break a nose in Dallas, both fist and nose would need to have been in the same place-- Oklahoma City, say-- at some point in time, and interacted with each other there."
Or, fist is in Des Moines, nose is in Dallas, someone in Oklahoma city sends entangled quantum "hit men" (nice ones, who use their fists) either to hit the fist and the nose at the same time, to hit one or the other, or not to hit either, with probabilities that depend on the angles at which the fist and the nose are held (ie. depending on what measurement the fist and the nose make of the entangled traveling fists). How do the traveling fists know to hit or not to hit, without getting out their phones, which they are not allowed to do?
This uses the QM property of your "While it is possible to establish entanglement between two objects that have never been in contact, the crucial step involves local interactions between particles that have previously been entangled." The brotherhood of the traveling fists is established in Oklahoma.
To determine empirically how entangled the traveling fists are, we have to send many traveling fists (typically tens of thousands, as Gregor Weihs did so beautifully) and check that the statistics of measurements by the nose and the fist that happen at nearly the same time violate Bell inequalities, preferably by several standard deviations.
Looking at that, the traveling fists have to hit the nose and the fist on the left or right consistently, with different probabilities depending on orientation of the nose and the fist. Non-detection of fists leads to the detection loophole, which makes a hidden-variable model relatively straightforward, so it can't be to hit or not to hit.
This is probably a little harsh, as the article does mention some current research-- there's a reference to this paper (I think-- there are no citations in the article), for example. I'm not really plugged into that community, but the author of that article is certainly active on the arxiv, and several of those papers are in well-regarded journals. I don't think the authors of the SciAm piece are totally uninformed-- at the very least, they're aware of papers that I'm not.
Sure, they're aware of papers that I'm not, also, and -- who knows? -- maybe some of those papers are even interesting. But to write an article about special relativity and quantum mechanics and never even use the phrase "quantum field theory"? If they're well-informed, that's deeply misleading; if they're not well-informed, then they're really poorly informed. As Moshe said, this is "nearly a century of good solid science." The constraints imposed by causality and locality of interactions in a relativistic theory are well-known, and the way they are manifested in quantum field theory is also well-known. It's just an unfortunate, and admittedly confusing, linguistic accident that QFT is "local" in this well-defined sense, but "nonlocal" just as QM is in the sense that there are entangled states. I don't see any evidence from the Scientific American article that its authors understand this.
There might be interesting work involved in figuring out how decoherence works in a relativistic setting, and what this means for macroscopic emergent phenomena like "wavefunction collapse". I've never looked into it; this might be a lively and interesting area of research that these authors are trying to tell us about. But they're not going to convince me it's worth looking into when they ignore or misunderstand basic things known since the 30s.
Of course, I'm not disagreeing with you on anything other than how charitably I can read them. But it's a shame, because I would like to read them charitably; I've heard enough times that philosophers of science and people thinking about foundations of quantum mechanics are doing interesting things that I would like to encounter something that explains to me what is interesting and why I should care. In this article I just see a lot of reasons to think its authors are not credible, which tarnishes everyone else in their field by association.
It is not possible to produce entanglement in a truly non-local manner.
Interesting question -- is this sentence true in a strict sense of laws of physics as we currently understand them? I guess it is : the information that entanglement exists is information, and physics as we understand it forbids the transmission of information in violation of locality.
I worried for a minute whether it could be possible to have entanglement without knowing it -- whether entanglement could exist independently of the information that entanglement exists. But this is inconsistent language, I guess. Saying there is entanglement is really a way of saying that information exists about the possible joint states of the system; you can't separate one from the other.
Thanks for blogging about this article. I cringed at the headline. I'm glad at least the blogging community is calling SciAm out on this misleading bs.
@ #5: "maybe this does bring up a flaw in most physics curricula, where people study QM and learn that it's weird and nonlocal"
If they learn that Bohr's "quantum jumps" are necessary for an electron to transit from one orbit to another without appearing anywhere in between, they are learning a highly flawed version of something that is not even quantum mechanics. In classical QM, a transition only occurs if there is a non-zero result for the overlap integral (notice that key word, overlap) between the two states.
Entanglement is an interesting variant of what happens when two people sitting across the table from one another view a die that has been tossed between them. What makes it interesting is that even the person who tossed the die doesn't know which way it landed until the two people tell him the result, but the correlation itself is no more surprising than conservation of momentum.
so I should stop writing my PRL on teleportation schemes without a phone line betwixt Bob and Alice? Dang........
We are a group that is challenging the the current paradigm in physics which is Quantum Mechanics and String Theory. There is a new Theory of Everything Breakthrough. It exposes the flaws in both Quantum Theory and String Theory. Please Help us set the physics community back on the right course and prove that Einstein was right! Visit our site The Theory of Super Relativity: Super Relativity
To several other commenters I would like to point out that nobody involved in these discussions thinks that QFT does not provide an adequate relativistic version of quantum theory. However, there are still conceptual issues to deal with, such as Bell inequality violations and measurement collapses, which at least appear to violate the spirit of special relativity. If you are trying to develop a realist interpretation of quantum theory, then these are real problems that you have to deal with somehow. Most of the solutions, except Everett, violate Lorentz invariance in doing so.
Now, if you do not agree with the necessity for a realist interpretation, e.g. you are happy with Copenhagen, then there is no real conflict with SR to deal with. I don't think anyone involved in the debate would disagree with that. However, for someone like Albert, any reasonable interpretation of QM must be a realist one, so I am not surprised that he would call this a direct conflict between QM and SR rather than qualifying by calling it conflict between SR and realist interpretations of QM. The latter does not make such a good SciAm headline.
Anyway, regarding Chad's request for references on non-narratability, I am having a bit of trouble remembering. I definitely recall a simple account of the idea written by Albert, but I can't find it anywhere. The ideas originate from a series of papers by Aharonov and Albert from the 1980's, the most relevant of which is probably:
Aharonov, Y., & Albert, D. (1984). Is the usual notion of time evolution adequate for quantum-mechanical systems?: II. Relativistic considerations. Physical Review D, 29, 228â234.
It is also worth looking at:
Myrvold, W. "On Peaceful Coexistence: Is the Collapse Postulate Incompatible with Relativity?" Studies in History and Philosophy of Modern Physics 33 (September 2002), pp. 435-466.
because it contains the simplest example that I know of.
However, rather than wading through all the discussion in those papers, I'd advise making yourself a nice cup of tea and spending an hour or so watching the talk that I linked to in my last comment. It gets to the punchline much more quickly than the papers.
Matt: unlike reasonable, or unreasonable interpretations of QM, QFT and its underlying Lorentz invariance and quantum mechanics are experimentally established parts of reality, which are therefore in no conflict as far as results of measurements are concerned. Any statement to the contrary is plain wrong, even if it makes a good SciAm headline.
OK, I found a version of the Albert paper, although it is not exactly the same as the one I originally read. It appears to come from a conference to do with Hilary Putnam. The final version of the paper does not seem to have been posted anywhere officially yet.
http://philosophy.ucsd.edu/faculty/wuthrich/PhilPhys/AlbertDavid2008Man…
Moshe,
Nobody is arguing about the experimental correctness of QFT, or that it incorporates Lorentz invariance.
I don't want to get into an argument about whether or not one can say that there is a direct conflict between QM and SR, mainly because I do not think there is one myself. That is Albert's position and he is more than capable of defending it himself. I was just trying to qualify what he means by that, and point out that there is a line of argument that makes the statement more defensible than it initially seems. SciAm is probably not the best place for such foundational debates though.
Matt: fair enough, I suspect we both agree. Since this article is meant for non-experts, explaining precisely what is meant by this particular "was Einstein wrong?" claim (and what it doesn't) would have been more intellectually honest to my mind. I am also surprised by your statement that most interpretations violate LI, not sure what is the point then. In any event, as you say, this is a discussion for some other time, and probably some other place.
Thanks for the references, Matt. I'll try to read them in my infinite free time.
SciAm is probably not the best place for such foundational debates though.
I think this is the crux of the whole issue. The issues involved are subtle and complicated, and the style and format of a SciAm article don't really allow them to do justice to the topic.
Entanglement is not really needed to refute special relativity. A camshaft in an automobile engine does the same job: it synchronizes distant valves simultaneously
without using light. Marinov has done that already in 1975 in his coupled mirrors experiment. He measured the one-way speed of light and found that due to the high speed of our solar system (370km/s in the direction of constellation LEO)
light speed on earth in the direction of LEO is c-370km/s and it is c+370km/s in the opposite direction. He thus has
invalidated Einstein's light speed isotropy postulate and, hence, refuted special relativity which is built on light speed isotropy in all inertial frames of reference.
best regards, Hartwig Thim (www.ime.jku.at)
Entanglement is a phenomenon allowing simultaneity at two
distant points. Since Einstein postulated that due to the finite velocity of light simultaneity is impossible, Albert and Galchen conclude that special relativity is wrong. Besides entanglement simultaneity can also be achieved by employing a camshaft as used in automobile enginees.
Michelson/Morley suggested this (mechanical!) synchronisation, and Marinov had used it and found that light speed is anisotropic on earth due to the earth's motion at 370km/s towards constellation LEO. Marinov used a rotating axle with two coupled mirrors.
It is correct that entanglement contradicts special relativity as 9 has said.
But that is no problem for entanglement as special relativity is wrong anyway and has been refuted both theoretically and experimentally (see Sciam March 2009).
By the way, information can be propagated at much faster than light speed not only by entanglement according to quantum mechanics, even classical mechanics allow faster than light speed: car engines are working with cam shafts having cams which are controlling valves instantaneously,
says
Hartwig Thim, Johannes Kepler University of Linz, Austria