Slashdot Mirror
'Ingenious' Experiment Closes Loopholes In Quantum Theory
Annanag writes: A Bell experiment in the Netherlands has plugged loopholes in the theory of quantum mechanics using a technique called entanglement swapping to combine the benefits of using both light and matter. It's Nobel-Prize winning stuff. Quoting: "Experiments that use entangled photons are prone to the ‘detection loophole’: not all photons produced in the experiment are detected, and sometimes as many as 80% are lost. Experimenters therefore have to assume that the properties of the photons they capture are representative of the entire set. ...
[In the new work], researchers started with two unentangled electrons sitting in diamond crystals held in different labs on the Delft campus, 1.3 kilometers apart. Each electron was individually entangled with a photon, and both of those photons were then zipped to a third location. There, the two photons were entangled with each other — and this caused both their partner electrons to become entangled, too.
This did not work every time. In total, the team managed to generate 245 entangled pairs of electrons over the course of nine days. The team's measurements exceeded Bell’s bound, once again supporting the standard quantum view. Moreover, the experiment closed both loopholes at once: because the electrons were easy to monitor, the detection loophole was not an issue, and they were separated far enough apart to close the communication loophole, too."
[In the new work], researchers started with two unentangled electrons sitting in diamond crystals held in different labs on the Delft campus, 1.3 kilometers apart. Each electron was individually entangled with a photon, and both of those photons were then zipped to a third location. There, the two photons were entangled with each other — and this caused both their partner electrons to become entangled, too.
This did not work every time. In total, the team managed to generate 245 entangled pairs of electrons over the course of nine days. The team's measurements exceeded Bell’s bound, once again supporting the standard quantum view. Moreover, the experiment closed both loopholes at once: because the electrons were easy to monitor, the detection loophole was not an issue, and they were separated far enough apart to close the communication loophole, too."
Even most physicists don't understand Feynman's point that QM is called "mechanics" for a reason: it's a set of mechanical rules for getting the right answer. It tells you nothing about how the universe operates behind the scenes so as to produce the same answer as QM. Feynman's little easy-to-understand book on QED should be read by everyone who thinks that QM is more than a tool for performing calculations. (And read Tegmark's book for an example of what happens when an intelligent person reads meaning into QM.)
Regarding the actual article: at first sight, this looks like a great experimental verification of something that no one (as far as I know) doubted; but it's always good to confirm another prediction of QM that appears bizarre to us.
I take a photon, I split it into two identical photons. I filter Photon 1 for a property X, Photon 2 goes into an experiment. I only consider results from the experiment when Photon 1 had property X (and thus so did Photon 2 and thus the experiment shows results only for photons carrying property X).
This is the issue here, the proof of entanglement is false, it is simply a filtering effect, you are not setting the photon to have X, you are detecting WHICH photons have X, and choosing the corresponding result, which... bingo.... is a result for X!
This experiment does not fix that, because they do the exact same filtering to determine the photons are 'entangled'. In other words its 245 entangled pairs out of N set where N is very large. The only time they would fail is if two photons were emitted so close together that there was not enough time between the two photons being emitted for the electronic circuit (usually called a coincidence circuit) to separate them. Bells hypothesis (a statistical claim that there is no hidden variable shared by both photons) is moot because the *TIME* they were emitted is the variable they are both tagged with. They do not carry a property, the experiment is designed to SLOWLY pump in photons so they can be split based on time. The property is time.
--------
But there is a second issue here, Protons (the +ve particle) are in the QM model, you can predict the behavior of the proton, and QM believers claim that detecting it position SETS it position, and before it was detected its position was undefined, fuzzy.
However protons are not fundamental particles anymore, deep inelastic scattering showed they are made of smaller particles. So you never detected the proton AT ALL, you simply detected the net result of the effects of these sub-proton particles. That net result jumped around, not the proton. Likewise you could not have 'set' the position of the proton, because it does not exist! It was just an effect of multiple smaller particles on the detection mechanism.
I usually describe this as the "flock of starlings" effect. If all you can see is a flock and not the individual bird, the flock appears to jump and leap and disappear and reappear. But it is simply the effect of a detector that can only see flocks and not birds.
Electrons are also considered to be fundamental in physics *currently*, yet we have a dipole experiment that shows they are not fundamental particles.
You can induce a dipole in an electron (a +ve -ve axis), showing electrons are made of both +ve and -ve stuff. In other words they're not fundamental either.
So QM model is broken here too, the electrons position appears to jump around, but its the net effect of these +ve and -ve things its made of.
I have a BA in philosophy and I took as many courses as I could on science and epistemology. The general concensus in these fields (of course with some disenters) is that you will always be able to ask this question about anything once you reach the scale boundries of our knowledge. When we say "gravity", what we really mean is a collection of rules which we are able to consistently produce accurate predictions from when applied to our observations. We can describe how a waterfall works in terms of gravity, but then when we ask how gravity works we must defer to some other system which then itself we will need to explain in terms of something else etcetera. I once grilled a chemist friend on what he meant when he said "electrons will always try to such and such" and he was stumped. It wasn't fair, because really the questions I was asking were based on a false appreciation of what the human study of natural law is able to be. Entanglement is a set of circumstances which we observe under certain conditions and believe are related to the point that we can give them a name. So are an apple, rugby, paint thinner and pornography. It is our own need for certainty that makes it difficult for us to accept this limitation of language and meaning.
"in terms of what other system could we try to explain the observed phenomena that we call entanglement?" Specifically (while I realize it cannot be used to transmit information), how is it faster than light? Is the concept of locality a defensible one?
Interestingly enough like most effects of quantum mechanics, entanglement does not have an easy macroscopic analog to compare. One way to think about it is that it is a type of emergent behavior because of the rules that QM appear to follow.
More specifically, entanglement is kind of emergent behavior that is a logical consequence of conservation rules and quantum superposition states. If you believe in the QM rules regarding conservation (e.g, conservation of say spin), and QM rules involved with superposition wave function collapse (e.g., so called "observation"), the emergent consequence of these rules is a behavior we call entanglement.
The macroscopic analog is sort of as follows. Suppose you have 1 balls and 2 boxes. By some method hidden from you, the ball into one of the two boxes and it is sealed. If you believe in conservation of balls, The two boxes are now entangled. You can move them arbitrarily far apart and then open one box, if it has a ball, you instantly know the other doesn't have a ball.
Where this breaks down is how you put the ball in the box. In the QM version of this, the method of which you put the ball into the box doesn't really put the ball into the box, it simply puts a type of probability of a ball into a box. Interestingly enough, the box can act sort of like a 1/2 ball in the box until you open it and then it "collapses" and is either a ball or not ball. The strange part is how can it act as if there is a 1/2 ball in the box before you open it? If you think of the decision being made when you seal the box, there is some sort of locality, but if you think of the decision being made when you open the 1/2 filled box, there is non-locality and you need to use a concept like entanglement as an emergent behavior.
That is 1/2 ball in the box (part particle, part wave) is QM and nobody really understands that part, so there's really not an analogous macroscopic system on which to understand it, because the systems we are familiar with don't follow those rules.
On the concept of locality, it's really unknown. We generally think of distance and time (warped by general relativity) as the way we measure locality (e.g., light cones, etc), but there isn't a clear idea if there isn't a macro-dimension or holographic way that alters our understanding what is local or non-local. Using current theories, we already speculate that there are singular violations of locality (e..g, EPR's or worm-holes, etc), and we don't understand the fabric of space-time (e.g quantum gravity) well enough to say if our current theories about this are descriptive enough to yield our current intuitions about space-time locality or if it will be as weird as QM.
"Entanglement" is a philosophically difficult arena. According to quantum theory, there is just one wavefunction for the entire Universe. However, we as observers are part of that wavefunction observing another part of that wavefunction with a really, really, big chunk of the whole wavefunction effectively unobservable but still coupled to the observer (part of the wavefunction), the measuring apparatus (part of the wavefunction), and the "experiment" (yep, part of the wavefunction. Everything is "entangled", but quantum mechanics also predicts that large systems approximated with a random phase condition will behave like a classical system, and the usual rule is that we treat a measurement apparatus as a classical system that breaks the entanglement of a measured systems and forces it "unpredictably" into a separable state. But even this is words, not equations although random phase approximations are indeed equations and are used frequently in field theory.
The only coherent explanation of this that I am aware of is the process of:
a) Starting with a density matrix (or other representation) for "the Universe".
b) Use the Nakajima-Zwanzig approach of splitting the (fully entangled) density matrix up into two parts -- a "system" that you will continue to treat as a quantum system, and a "bath" -- everything else -- which would also include the measuring apparatus if you were trying to describe an experiment. One then accepts the fact that one cannot know or prepare the state of the bath (which is really, really big being the rest of the Universe and everything) and so one makes a statistical approximation of the bath (taking the trace) which essentially eliminates the pesky entanglement but also breaks useful things like unitarity and in a sense, conservation laws. One them creates projection-valued operators and transforms the equations for the system into stochastic or semiclassical equations of motion.
c) IIRC your final result is quantum mechanics for the system expressed as a non-Markovian integrodifferential equation that is almost impossible to solve. However, if one makes a Markov approximation (forces it to be time-local, delta-correlates time) you end up with a decent explanation for things like spontaneous decay as an "exponential" process rather than a punctuated unitary process. You go one way, you can make it into a Langevin equation, go another you can make it more like Fokker-Planck.
The lovely thing about this approach is that it renders moot all sorts of nonsense, such as EPR paradoxes and "wavefunction collapse". It is perfectly clear that in the Universal wavefunction no such paradox or collapse can occur. They are simply expressions of our ignorance of phase and state whenever we try to isolate some part of the whole and pretend that it is a standalone "system" that can ever be decoupled from everything else. Schrodinger's cat paradoxes disappear as there is no paradox in the Universal wavefunction, only when we try to project the state of the cat against our ignorance of phase and interaction with the outside Universe. The cat, if you like, cannot be entangled separately from its preexisting entanglement with the rest of the Universe, and we only get into trouble when we have to force it by partitioning the system in order to get a chunk small enough to work with.
Hope this helps. I doubt it will. Very few people seem to be in touch with Nakajima-Zwanzig and the Generalized Master Equation these days, and don't treat problems like this as OPEN quantum systems as opposed to closed systems with a classical measurement apparatus, which is a place you only get to on the far side of the N-Z GME ritual.
rgb
Even when the experts all agree, they may well be mistaken. --- Bertrand Russell.
This experiment has a big problem, as an applied optics (polarization specific) expert, they use polarization entanglement, but then run the light through fiber optics.
The problem is that fiber optics (even polarization preserving designs) have a terrible issue with preservation of polarization states.
I haven't read the paper in detail yet, but I don't know how they can mitigate this issue...
The real question is - exactly wtf is entanglement anyway? I can find lots to read about what it looks like and how it behaves... but what's the underlying mechanism? Is there even the most speculative explanation of it?
Here's the best answer I can give you - I think it's true, and not so over-simplified as to be wrong.
The universe has some underlying state. We don't have direct access to that state - not only is it not directly observable, it's not directly related in any intuitive way to the state we can observe. There's this arbirtary-seeming transform between underlying state and what we observe (it only seems odd or arbitrary because all our intuitions are based on human-scale observables, and are not at all directly informed by this underlying state). This underlying state seems to be well-defined and deterministic, forwards and backwards in time. The observable universe is not.
Entanglement is a feature of how observations relate to underlying state - a feature of the transform. In very simple experiments we can measure specific properties of, say, an electron. We can't measure all of them, for a given electron, because the transform just doesn't work that way, but we can measure some. However, that's deceptive, because you can't really track that property of that electron over time, in non-trivial cases. If e.g. two electrons interact, become entangled, your observations are now a function of both electrons' underlying state, and that's a different transform from 2 non-entangled electrons.
There are two key concepts here. The first is that the whole notion of "particle" is a handy but false oversimplification. It can lead you to all sorts of false intuitions about how particles behave. Fundamentally, individual e.g. electrons don't have unique identities. The underlying state is a single electron field, which other fields can interact with, in a way that can sometimes be simplified as "particle interactions", for a simpler mental model, but you can't go too deep with that model. An example: "two electrons collide in an accelerator, and two electrons leave, which is which?" That question is "not even wrong", it's just nonsense. Thinking of electrons as billiard balls colliding is simply not a helpful model, as it just misses the point of the interaction.
"Entanglement" happens just when the "particle" mental model fails: you can no longer pick two disjoint areas in the electron field and consider them as independent "electrons", but instead you have to reason about two areas which may be quite disconnected in space and time. E.g., you might know for sure that one electron is spin-up, and one spin-down, but have 0 information about which is which. None of that matters to the underlying state: there's just one electron field, and the only truly correct way to reason about it it to reason about the whole field all the time, and so this is only half of "WTF is entanglement".
The second concept gets too much into the math to explain well, but in a hand-wavy way it's this: "what is measurement?". There are older interpretations about measurement causing wavestate collapse and so on, but they're wrong because of that word "cause". Measurement is simply the observer becoming entangled with the observed. Measuring one entangled electron doesn't "cause" the other electron to do or become anything. The underlying state is unchanged, which is why there's no faster-than-light effect. In some cases, this is an overly pedantic distinction, but it matters when the difference between QM and intuition matters. In a two-slit experiment where you see an interference pattern at your detector, if you add a measuring device to one slit suddenly you don't see that interference pattern. Informally we might say the second observer "caused" this change, but formally that's wrong, it's just that a system with 2 slits and 2 detectors behaves differently from a system with 2 slits and one detector, and it doesn't matter which detector the electron passes first, because (see above) an "electron" as a discrete particle is fiction anyway, and both detectors are entangled with the electron field already, or they couldn't measure an electron anyhow.
Socialism: a lie told by totalitarians and believed by fools.