I'm not saying that Bell inequality violations prove that orthodox quantum mechanics is correct.
Bell's theorem sauys (and I'm not being precise here): take a certain class of theories which we can call local and real in the sense that there can be no superluminal influences and a particle locally carries all the information necessary to determine the outcome of any measurement. (Notice that this doesn't say anything about which theories are "reasonable" or not). It can be shown that certain inequalities must be satisfied by any physical theories of this type.
I have not said anything about quantum mechanics or state vector collapse or anything. Nor do I have to. It is true that the predictions of quantum mechanics violate Bell's inequality (that's why Bell's inequality is interesting).
It is a question open to experiment whether nature violates Bell's inequalities. We seem unable to agree on whether Bell violations have been measured. If they have been then locally real theories are wrong (not that quantum mechanics is right). If they weren't violated when they should have been, then quantum mechanics is wrong. There aren't any other choices.
While the quantum mechanical explanation of Bell violations relies of the collapse postulate, of course, this has nothing to do with the empirical question of whether Bell violations occur in nature. You seem unconvinced that they have been measured, and I can respect this position to the extent that there are genuine flaws in the experiments. My point is that an empirical measurement of Bell violations has nothing to do the collapse postualte or any other theoretical construction. It is simply a fact about the behavior of nature.
Secondly, while I have no doubt that the physics orthodoxy has a lot at stake in protecting certain theories, I think you misrepresent the physics community. John Bell himself hated the Copenhagen interpretation and non-determinism in general. He was a strong supporter of Bohm's theory. It seems weird to me that you prevent Bohm's theory as a realist challenge to the more "mystical" aspects of quantum theory since Bohm's theory is explicitly non-local (the pilot wave can change instantaneously throughout space), and Bohm himself was a pretty "mystical" character with his belief in the "implicate order" and whatnot.
I used to think that Bohm's theory made a lot more sense than orthodox QM, and I thought that the only reason that people held to the orthodox view was inertia or a kind of philosophical malaise left over from Bohr's ideas. Bohm's theory has its own disturbing anti-commonsensical features, however. Look up articles on "surreal trajectories".
The bottom line is that nature is non-classical, and we have not done very well at understanding that non-classical behavior. Every theory that I know of is somehow deeply unsatisfying, and yet quantum mechanics is able to predict things with incredible accuracy that classical physics completely failed to deal with.
I completely agree with your remarks concerning the projection postulate, yet I somehow fail to see the relevance for tests of Bell's inequalities. It's true that measurement is ill-defined in quantum mechanics (as the standard interpretation goes) and I'm sure that it will eventually have to be explained in a genuine, physical way.
Having said that, I don't think that the meaning of measurement and how it occurs is central to the Bell tests at all. Of course, it is central to the quantum mechanical interpretation of the tests, but (if you believe the results I mentioned earlier) Bell's inequality is violated experimentally. This does not prove quantum mechanics correct, it just proves that classical-like theories cannot be correct.
Wavefunction collapse is an ad-hoc postulate, and it is right, in my opinion, to be very critical of it. Nevertheless, the collpase postulate is central to quantum mechanics (or something very like the collapse postulate). There are countless experiments that do not involve entanglement at all that the collapse postulate correctly predicts the results of (Stern-Gerlach experiment with multiple magnets in series, quantum "watched pot" effect, etc...). While the collapse postulate is unpalatable, quantum mechanics in its current form cannot survive without it. The solution is not just to toss out the collapse postulate, but to understand the physical nature of measurement and reformulate quantum mechanics so that measurement is no longer a "deus ex machina" kind of thing.
But I honestly believe the universe is non-classical. Maybe someone will actually be able to factor a small number in 10 years using Shor's algorithm. Would you believe then?
Sackett's ion experiments are not analogous to polarizer/photon experiments. They do not use Stern-Gerlach magnets, but rather produce entagled states of ions in a trap by using lasers to drive certain transitions.
Of course, they cannot produce entangled states with 100% efficiency. But, according to Sackett's data (which I have no reason to doubt), they see Bell violations when averaging over all the data, not just those they have deemed to be entangled states. Thus (barring explanations based on hidden, subluminal communication between the ions) they see violations in the true Bell sense (i.e. taking unbiased expectation values over all data points).
I agree that we have not yet seen a perfect Bell inequality test, but the position you seem to be taking is that quantum mechanical non-locality cannot exist and that there must be a flaw in any experiment that purports to detect it. This is no more reasonable than insisting that quantum non-locality is real in the absence of any evidence.
The experiments I was referring to did not use entangled photon states, they used entangled ion states.
And I did not make up the numbers (though I did get the reference wrong). See: C.A. Sackett, et. al, "Experimental Entanglement of Four Particles" Nature 404 (2000).
It is true that these entanglement measurements are not a perfect test of the Bell inequalities since, as Sackett pointed out when I heard him speak recently, they do not close the "locality" loophole. They do, however, close the important "uniform sampling" loophole that you are (rightly) critical of.
Many other experiments (with photon entanglement) have closed the locality loophole. It is true, as far as I know, that no single experiment has gotten sufficient sensitivity using spacelike separated measurements... yet. It seems somehow perverse though to hang a defense of local realism on this fact. When the definitive test of Bell's inequalities arrives, I have no doubt that they will be violated. Again.
Mike
It is only the correlations that are non-local in measurements made on entangled states.
As long as you cannot control what result you get when you make measurement (e.g. photon polarization), which of course you can't according to quantum mechanics, there is no way to send a message using effects like this.
Which is not to say they are useless. You *can* do things like establish a random list of measurements on one of an entangled pair of particles that you know the your partner with the other particle will also measure. This is the basis of quantum cryptography and has been well-demonstrated experimentally.
Bell's inequality has been tested with entangled ions with nearly 99% detection efficiency (from the same group at Rice who demonstarted 4 particle entaglemnet in Science).
Bell violations are real. Deal with it.
I completely agree with you about the danger of science popularizations, but it's not impossible to imagine particle exchange giving rise to attraction. Imagine two ice-skaters standing back-to-back on a frozen lake. One throws a boomerang away from her in such a way that it circles around and is caught by the other skater. As they exchange boomerangs, they will feel an attractive force. Of course, this isn't meaningful or rigorous at all and it requires other factors (the air) to get that weird behavior from the boomerang-particles.
For some reason, I always thought that "cracker" was kind of a latter day invention and that "hacker" originally could refer both to someone with a skill and love for elegant coding and to someone who broke into secure systems (at one time, these traits were more closely linked than today). So, it's not wrong to say that your box was hacked, right? This was one of the original meanings of "hacked", "cracker" is the (less popular) mutation.
Re:forget factoring - quantum cyphers is where its
on
RSA slightly broken
·
· Score: 1
True, it's not going to be household stuff for a while, but the physics behind it is very solid, and it has been done.
Over pretty large distances too...
Shor's Algorithm (Re:Security through openness)
on
RSA slightly broken
·
· Score: 2
Actually, if you stretch your definitions a little, somebody has found an efficient factoring algorithm. Peter Shor's algorithm will factor a number in polynomial time, trouble is it only works on a quantum computer.
It's been proven that P QP (that is, the set of all classical polynomial-time problems is a subset of all quantum polynomial-time problems) Also, there are certain things that can be done on quantum computers more efficiently than a classical computer could (not just speculation, there are problems like this)
Slashdot Mirror
They have a name for true computer scientists: "mathematicians"
Bell's theorem sauys (and I'm not being precise here): take a certain class of theories which we can call local and real in the sense that there can be no superluminal influences and a particle locally carries all the information necessary to determine the outcome of any measurement. (Notice that this doesn't say anything about which theories are "reasonable" or not). It can be shown that certain inequalities must be satisfied by any physical theories of this type.
I have not said anything about quantum mechanics or state vector collapse or anything. Nor do I have to. It is true that the predictions of quantum mechanics violate Bell's inequality (that's why Bell's inequality is interesting).
It is a question open to experiment whether nature violates Bell's inequalities. We seem unable to agree on whether Bell violations have been measured. If they have been then locally real theories are wrong (not that quantum mechanics is right). If they weren't violated when they should have been, then quantum mechanics is wrong. There aren't any other choices.
While the quantum mechanical explanation of Bell violations relies of the collapse postulate, of course, this has nothing to do with the empirical question of whether Bell violations occur in nature. You seem unconvinced that they have been measured, and I can respect this position to the extent that there are genuine flaws in the experiments. My point is that an empirical measurement of Bell violations has nothing to do the collapse postualte or any other theoretical construction. It is simply a fact about the behavior of nature.
Secondly, while I have no doubt that the physics orthodoxy has a lot at stake in protecting certain theories, I think you misrepresent the physics community. John Bell himself hated the Copenhagen interpretation and non-determinism in general. He was a strong supporter of Bohm's theory. It seems weird to me that you prevent Bohm's theory as a realist challenge to the more "mystical" aspects of quantum theory since Bohm's theory is explicitly non-local (the pilot wave can change instantaneously throughout space), and Bohm himself was a pretty "mystical" character with his belief in the "implicate order" and whatnot.
I used to think that Bohm's theory made a lot more sense than orthodox QM, and I thought that the only reason that people held to the orthodox view was inertia or a kind of philosophical malaise left over from Bohr's ideas. Bohm's theory has its own disturbing anti-commonsensical features, however. Look up articles on "surreal trajectories".
The bottom line is that nature is non-classical, and we have not done very well at understanding that non-classical behavior. Every theory that I know of is somehow deeply unsatisfying, and yet quantum mechanics is able to predict things with incredible accuracy that classical physics completely failed to deal with.
Having said that, I don't think that the meaning of measurement and how it occurs is central to the Bell tests at all. Of course, it is central to the quantum mechanical interpretation of the tests, but (if you believe the results I mentioned earlier) Bell's inequality is violated experimentally. This does not prove quantum mechanics correct, it just proves that classical-like theories cannot be correct.
Wavefunction collapse is an ad-hoc postulate, and it is right, in my opinion, to be very critical of it. Nevertheless, the collpase postulate is central to quantum mechanics (or something very like the collapse postulate). There are countless experiments that do not involve entanglement at all that the collapse postulate correctly predicts the results of (Stern-Gerlach experiment with multiple magnets in series, quantum "watched pot" effect, etc...). While the collapse postulate is unpalatable, quantum mechanics in its current form cannot survive without it. The solution is not just to toss out the collapse postulate, but to understand the physical nature of measurement and reformulate quantum mechanics so that measurement is no longer a "deus ex machina" kind of thing.
But I honestly believe the universe is non-classical. Maybe someone will actually be able to factor a small number in 10 years using Shor's algorithm. Would you believe then?
Of course, they cannot produce entangled states with 100% efficiency. But, according to Sackett's data (which I have no reason to doubt), they see Bell violations when averaging over all the data, not just those they have deemed to be entangled states. Thus (barring explanations based on hidden, subluminal communication between the ions) they see violations in the true Bell sense (i.e. taking unbiased expectation values over all data points).
I agree that we have not yet seen a perfect Bell inequality test, but the position you seem to be taking is that quantum mechanical non-locality cannot exist and that there must be a flaw in any experiment that purports to detect it. This is no more reasonable than insisting that quantum non-locality is real in the absence of any evidence.
Mike
And I did not make up the numbers (though I did get the reference wrong). See: C.A. Sackett, et. al, "Experimental Entanglement of Four Particles" Nature 404 (2000).
It is true that these entanglement measurements are not a perfect test of the Bell inequalities since, as Sackett pointed out when I heard him speak recently, they do not close the "locality" loophole. They do, however, close the important "uniform sampling" loophole that you are (rightly) critical of.
Many other experiments (with photon entanglement) have closed the locality loophole. It is true, as far as I know, that no single experiment has gotten sufficient sensitivity using spacelike separated measurements... yet. It seems somehow perverse though to hang a defense of local realism on this fact. When the definitive test of Bell's inequalities arrives, I have no doubt that they will be violated. Again. Mike
It is only the correlations that are non-local in measurements made on entangled states. As long as you cannot control what result you get when you make measurement (e.g. photon polarization), which of course you can't according to quantum mechanics, there is no way to send a message using effects like this. Which is not to say they are useless. You *can* do things like establish a random list of measurements on one of an entangled pair of particles that you know the your partner with the other particle will also measure. This is the basis of quantum cryptography and has been well-demonstrated experimentally.
Bell's inequality has been tested with entangled ions with nearly 99% detection efficiency (from the same group at Rice who demonstarted 4 particle entaglemnet in Science). Bell violations are real. Deal with it.
I completely agree with you about the danger of science popularizations, but it's not impossible to imagine particle exchange giving rise to attraction. Imagine two ice-skaters standing back-to-back on a frozen lake. One throws a boomerang away from her in such a way that it circles around and is caught by the other skater. As they exchange boomerangs, they will feel an attractive force. Of course, this isn't meaningful or rigorous at all and it requires other factors (the air) to get that weird behavior from the boomerang-particles.
For some reason, I always thought that "cracker" was kind of a latter day invention and that "hacker" originally could refer both to someone with a skill and love for elegant coding and to someone who broke into secure systems (at one time, these traits were more closely linked than today). So, it's not wrong to say that your box was hacked, right? This was one of the original meanings of "hacked", "cracker" is the (less popular) mutation.
Over pretty large distances too...
It's been proven that P QP (that is, the set of all classical polynomial-time problems is a subset of all quantum polynomial-time problems) Also, there are certain things that can be done on quantum computers more efficiently than a classical computer could (not just speculation, there are problems like this)
I wish I knew more about this stuff.