This forum gives conflicting info on the HUP

[Demystifier]

Yes, what definition are you using? If I understand correctly, these papers define an accurate measurement of A as one that produces the same distribution of outcomes as a projective measurement of A.

If so, then their definition is fine, but different from the one I used. I used a definition natural to a naive experimentalist, for whom the measurement is a procedure that would correspond to a standard notion of measurement of classical quantities.

Clearly, there are well defined procedures for simultaneous measurement of classical quantities such as position and momentum. Whatever such a procedure is, a naive experimentalist may apply the same procedure to ANY system (classical or not) and see what he will get.

 

[Ken G]

Does the definition in that paper that atyy quotes seem a bit troublesome to anyone else? First of all, the word "measurement" is used twice, which is a bit like defining something by that something. I believe what this must really mean, if one removes that circularity, is an accurate measurement is one that agrees with quantum mechanics theory. Now that is a fine way to define measurements if one has already used some other meaning of measurement to establish the value of quantum theory, but my question is, what was the definition of that other kind of measurement that was used to verify quantum theory in the first place? I believe one will then arrive at a meaning close to what Demystifier is talking about.

 

[Demystifier]

Now that is a fine way to define measurements if one has already used some other meaning of measurement to establish the value of quantum theory, but my question is, what was the definition of that other kind of measurement that was used to verify quantum theory in the first place? I believe one will then arrive at a meaning close to what Demystifier is talking about.

Again, I agree with you.

 

[atyy]

If so, then their definition is fine, but different from the one I used. I used a definition natural to a naive experimentalist, for whom the measurement is a procedure that would correspond to a standard notion of measurement of classical quantities.

Clearly, there are well defined procedures for simultaneous measurement of classical quantities such as position and momentum. Whatever such a procedure is, a naive experimentalist may apply the same procedure to ANY system (classical or not) and see what he will get.

Does the definition in that paper that atyy quotes seem a bit troublesome to anyone else? First of all, the word "measurement" is used twice, which is a bit like defining something by that something. I believe what this must really mean, if one removes that circularity, is an accurate measurement is one that agrees with quantum mechanics theory. Now that is a fine way to define measurements if one has already used some other meaning of measurement to establish the value of quantum theory, but my question is, what was the definition of that other kind of measurement that was used to verify quantum theory in the first place? I believe one will then arrive at a meaning close to what Demystifier is talking about.

Naively, I don't see how the classical definitions can be "right", since one can prove that in some sense the classical definitions don't exist. For example, there is in general no joint distribution of the values of conjugate observables.

But I do agree that there may be different definitions that are reasonable, so perhaps simultaneous accurate measurements of conjugate observables could be possible with a different definition of "accurate". Could you provide an explicit example?

 

[atyy]

In a "classical" measurement scheme, I would further say again - the theory determines what you can measure. So if one would like a classical measurement theory, one should use Bohmian mechanics.

However, do "Hamiltonian conjugate observables" exist in Bohmian theory?

Given that there isn't (yet due to the incompetence of experimentalists) a unique Bohmian dynamics, wouldn't the notion of accurate measurement depend on which Bohmian dynamics one chose?

 

[Ken G]

Naively, I don't see how the classical definitions can be "right", since one can prove that in some sense the classical definitions don't exist. For example, there is in general no joint distribution of the values of conjugate observables.

The classical definitions I'm talking about (I won't speak for Demystifier) are not definitions of the obervables, but rather definitions of the observations, and then the observations provide operational definitions of the observables (all quantities in science are just proxies for acts of observation, after all). In other words, if I say I have this apparatus, and I will claim that it "observes x", then I have a classical definition of what I mean by x based on that apparatus. If I have an apparatus that I say "observes p", same story. Now if I have an apparatus that I say "observes p and x simultaneously", then we will need to look at the ramifications of my claim.

Above I said that these claims can make good on several different levels. The weakest level is that they can agree with the predictions I used some theory to make, but they destroy the system for the purposes of further replication of my result. Since they destroy the system, they cannot be said to "confer knowledge" about the state of the system after the measurement, only about the state of the system prior to measurement (by "state" I mean "everything I can possibly need to know about the preparation of that system in order to predict subsequent behavior"). The next strongest level is an apparatus that conveys knowledge about the state going forward, but only if I already know certain things about the preparation prior to the measurement, so the apparatus is not a "complete" or self-contained measurement. Finally, the strongest level is an apparatus that conveys knowledge of the state going forward even if I know nothing about the history of that system.

Given that we can recognize three separate types of measurement, it seems natural that each might obey a different set of constraints, in particular, different versions of a HUP. If we look at EPR systems, we see that we do need past knowledge of the system, it's not a completely WYSIWYG kind of measurement. It also destroys the entanglement, so the results cannot be used to make predictions going forward that involve knowledge of the x and p of both particles. Can we say that it is the weakest type though, the type that confers knowledge about the past preparation of that system, given knowledge that we have a momentum-conserving entanglement? Yes, we can use our x and p results on an ensemble to recover completely the preparation of that system, given that prior knowledge.

But here's my point there: the information in the preparation of that system, given its entanglement, involves only the same amount of information as is in a single particle, but it is in some sense "spread out" over two particles. Hence, the past knowledge can only be the same as a wave packet of a single particle, and that knowledge is governed by the HUP of a single particle. To treat the x and p measurements on the two particles as simultaneous knowledge of x and p, one must then treat the particles as two separate systems, which involves breaking the entanglement, which means we are talking about the post-measurement preparation, not the pre-measurement preparation. It is not surprising that classical observables can impart knowledge of the post-measurement x of one particle, and the post-measurement p of another particle.

 

[atyy]

@Ken G, I think I agree to the extent I understand what you are saying. Just one question: are you agreeing with Demystifier that simultaneous or successive measurements on the same state are possible, for some definition of "accurate"? Edit: And for any state - the definition I was using means it is possible for some special states, but not for any state, at least not by the procedures considered in those papers.

Regarding EPR, if one accepts that as an accurate measurement, then it clearly is not a measurement procedure that can be defined classically. And if one needs the measurement to give accurate information about the post-measurement state, then clearly one can't do those simultaneous or joint measurements in quantum mechanics, because that is using the measurement to prepare a state, and so the preparation HUP must hold, which is just the textbook HUP.

 

[Ken G]

@Ken G, I think I agree to the extent I understand what you are saying. Just one question: are you agreeing with Demystifier that simultaneous or successive measurements on the same state are possible, for some definition of "accurate"? Edit: And for any state - the definition I was using means it is possible for some special states, but not for any state, at least not by the procedures considered in those papers.

I think your interest is in the possibility of simultaneous or successive measurements of complementary observables? If that's what you are asking, I do not think such measurements convey "observed properties" of that system, no. I think that a system is defined in part by the apparatus that it is encountering. I feel it is not even enough to say that the apparatus changes the system, I feel one must say that the apparatus is part of how we define the system. To me, the lesson of the HUP is we must get away from the classical model that systems "carry around with them" certain observables, things that in some sense "the universe already knows", and our job is to use an apparatus to figure those things out. Instead, the system is defined by both the apparatus that prepared it, and the apparatus that is measuring it, because we want to think of a system as something "real", and that means we need to be able to give it real attributes, and that means we need to be able to observe it, so that must involve both a preparation that we can know things about (by observing them or inferring them from other things we can observe), and an outcome we can know things about.

The preparation is thus just one part of that system, and determines what we call the "state" of the system, but the apparatus that is measuring that system is also part of the system, because it determines what aspects of that system are actualized. A state, or a preparation, only produces tendencies for actualizations, often expressed in terms of an "ensemble" to make it more concrete. To be considered something actual, and not just a set of tendencies, one must include the measuring device in the meaning of "the real system." The thing that Demystifier is saying that I do agree with is that the apparatuses are always classical, somewhere along the way (since ultimately, our brains are), so no quantum system is "real" until one can associate it with a set of classical pointers. The realness is in some sense the "closure" or "actualization" of that quantum system. I believe this is also what Bohr meant when he said "there is no quantum world"-- the realness comes from a system that is complete, all the way from preparation to measurement. What I don't like about the Bohmian picture is the desire to add extraneous elements to the preparation+actualization such that the classical pointers can refer to attributes that the system has all the time, and not just at the end of the closure process.

Regarding EPR, if one accepts that as an accurate measurement, then it clearly is not a measurement procedure that can be defined classically.

You can do measurements that you can define classically on the two particles, like you can measure x and p. The issue is, if when you measure the p of one particle, and know by momentum conservation it is the p of the other particle, does that then allow you to know x and p of the other particle? I say no, because the instant you think you know both the x and p of the particle, you have broken the entanglement that let you know p in the first place. I'm sure Demystifier agrees that you don't know x and p any more, after the measurement, but he holds that the particle had an x and p instantly before the measurement, and that's how you can know it. I hold that you cannot know anything without specifying the apparatus that let's you know it, and no apparatus let's you know x and p before the apparatus let's you know x and p!

And if one needs the measurement to give accurate information about the post-measurement state, then clearly one can't do those simultaneous or joint measurements in quantum mechanics, because that is using the measurement to prepare a state, and so the preparation HUP must hold, which is just the textbook HUP.

Yes, I think a key point is recognizing the difference between measurement as knowledge of a system that still exists, versus measurement as knowledge of a system just before you measured it but no longer exists in that state. I don't think of measurement as how we get knowledge of the properties of systems, I see measurement as part of the meaning of the properties of a system. So I reject the whole concept of using measurements to know the properties of a system prior to the measurement, but I do think measurements can be used to characterize the state of a system, i.e., everything we need to know about the preparation of that system to able to predict what properties it may have when those properties are actualized by measurements.

 

[DevilsAvocado]

I'm sure Demystifier agrees that you don't know x and p any more, after the measurement, but he holds that the particle had an x and p instantly before the measurement, and that's how you can know it.

(I’ve only skimmed the posts since my last post, so I trust on you that this is what he is saying)

Gentlemen, if I may: Einstein and Bohr debated x and p for 20 years (albeit without Bell) and the 'measurements' I have done all point in the direction that they both were smarter than any of us in this forum...

Therefore: My humble 'proposal', to get a better grip on EPR-Bell, is to use photon polarization instead of momentum and position, as it's easier to handle IMHO.

In the case of entangled photons, it gets instantly clear – there is no possibility of preexisting (real) values kept unchanged all the way from the source to measurement.

Why?

It's a mathematical impossibility, having 3 hidden variables A, B and C for the settings 0°, 120° and 240°. I'll use binary representation to make it even clearer (excludes 'impossible' 0 and 7):

The Yellow and Purple group are XOR mirrored (i.e. 001 XOR 111 = 110, or decimal 1 XOR 7 = 6), and since the actual values are trivial, 001 and 110 are the same when it comes to correlated hits. A hit is defined as the same values for Alice and Bob (i.e. [1, 1] or [0, 0]).

The correlation is cos²(120°) = 25% for all (relative) settings, thus we must have a minimum of 4 runs to get 25%, and the 3 hidden variables must be able to handle all three combinations of AB, AC and BC, that Alice and Bob could get jointly, and 25% are equal to one hit and three misses.

We start by picking the first three in order (i.e. decimal 1 to 3) and there are no problems in the Yellow group, it's safe regarding all possible combinations (i.e. one hit and two misses for all three AB, AC and BC settings).

But the forth pick – that must be a miss in all three combinations – is a mathematical dead end! There is no viable number left to pick, since the Purple group is a mirror of the Yellow. Fait accompli...

I would love to hear DM explain how there could be any useful (local) information there for us to know?? ()

 

[Demystifier]

I would love to hear DM explain how there could be any useful (local) information there for us to know?? ()

Your picture is a bunch of 0's and 1's, each being written at a definite position in space. And this picture, I think, is useful. Therefore, the picture itself presents a useful local information.

 

[Ken G]

Isn't it true that Bohmian mechanics allow Bell-type correlations to be true via the pilot wave? That was my impression, that if you are committed to letting systems have classical attributes all the time, not just as outcomes of measurements but also as "properties" of the system prior to measurement, you can get it with the pilot wave. You basically just take all the outcomes of the measurements you need to agree with, and reverse-engineer a pilot wave that does not itself have observable consequences, but makes the classical "properties" act quantum mechanically. You need a HUP? The pilot wave does it. You need Bell correlations? Pilot wave.

Note there is nothing wrong with this-- it agrees with the observations. It's all just a question of how badly do you want the system to maintain classical properties all the time, and if you really want that, you can get it with an invisible scaffolding that does not obey the properties assumed in von Neumann's no-go theorem. Does it seem artificial? To me, yes, but to someone committed to those classical properties, it is a requirement for getting the behavior we see-- but it does get the behavior we see (if Demystifier is right about how to make it work relativistically).

 

[Demystifier]

Naively, I don't see how the classical definitions can be "right", since one can prove that in some sense the classical definitions don't exist. For example, there is in general no joint distribution of the values of conjugate observables.

Ah, now I think I see the source of confusion. One should distinguish two things:
1) SINGLE measurement, from which no information about probability distribution can be extracted (except that the obtained value has probability larger than one).
2) Statistical ENSEMBLE of similar measurements, from which the probability distribution can be extraced.

I was talking about the former, while it seems that you are talking about the latter. If I simultaneously measure position and momentum ONLY ONCE, I cannot extract any information about the joint probability distribution.

But then again, even in classical mechanics I can repeat many times the simultaneous measurement of position and momentum. From such a measurement I CAN extract the joint distribution. Moreover, by using the theory called classical STATISTICAL mechanics I can even predict or explain the joint distribution I measured. So your claim that "classical definitions for joint distribution don't exist" is certainly wrong.

 

[Demystifier]

... but it does get the behavior we see (if Demystifier is right about how to make it work relativistically).

Even if I am wrong about that, there are also other ways to make Bohmian mechanics compatible with predictions of relativistic quantum theory. For example, one can always use a Bohmian theory with a preferred Lorentz frame at the level of hidden variables, which leads to Lorentz invariant measurable predictions. (It's only that I don't particularly like such a variant of Bohmian mechanics, because it looks somehow too cheap for me.)

 

[Demystifier]

What I don't like about the Bohmian picture is the desire to add extraneous elements to the preparation+actualization such that the classical pointers can refer to attributes that the system has all the time, and not just at the end of the closure process.

I'm sure Demystifier agrees that you don't know x and p any more, after the measurement, but he holds that the particle had an x and p instantly before the measurement, and that's how you can know it.

Note that I don't consider the Bohmian interpretation to be the only viable interpretation. What I like even more is to view the things from the point of view of DIFFERENT interpretations. In particular, I have constructed a hybrid between Copenhagen and Bohmian interpretation
http://lanl.arxiv.org/abs/1112.2034 [Int. J. Quantum Inf. 10 (2012) 1241016]
according to which only the particles of the observer are real in a Bohmian-like sense, while the observed particles are never real in that sense, not even when they are observed.

 

[Ken G]

Yes, I agree there is value in being able to see the elephant from all possible angles. There is little point in debating which is the perspective that gives us a truer look!

 

[DevilsAvocado]

Note that I don't consider the Bohmian interpretation to be the only viable interpretation. What I like even more is to view the things from the point of view of DIFFERENT interpretations. In particular, I have constructed a hybrid between Copenhagen and Bohmian interpretation

This is just great!

I'm just a bum layman, but sometimes (correct me if I'm wrong) I get a slight feeling that "interpretational fundamentalism" stands over everything else, i.e. some are willing to "look the other way", in cases which is not 'favorable' to their personal interpretation... but I could be wrong.

I'm not a scientist, but from what I know, your open-minded stance must be the correct way forward – to bend and twist this question from all angles possible.

DM, I sincerely hope that you will be successful in this work, good luck! :thumbs:

 

[DevilsAvocado]

You basically just take all the outcomes of the measurements you need to agree with,

Just to avoid any 'misunderstanding' – there is no possibility to cover all possible outcomes in EPR-Bell in preexisting hidden variables. From my picture above, it may look like we have 6 unique binary values (excluding 0 and 7), but the truth is; there are only 2 x 3 'mirrored' values, which reduces to the indisputable fact that 1/3 ≠ 1/4.

We need one more binary value to get 1/4, which the axioms of mathematics will never let us have, no matter what...

 

[DevilsAvocado]

Your picture is a bunch of 0's and 1's, each being written at a definite position in space. And this picture, I think, is useful. Therefore, the picture itself presents a useful local information.

Thanks DM, I don't understand Bohmian mechanics. What happens in an EPR-Bell experiment? You have (real) hidden variables going out from the source, and then what?

Does the pilot wave 'scan' the (space-like separated) settings and 'calculate' the correlation needed, and then 'send' this info to the hidden variables so that they can 'adjust' their values for the actual measurement?

Or, did I get this completely wrong...

 

[DevilsAvocado]

Even if I am wrong about that, there are also other ways to make Bohmian mechanics compatible with predictions of relativistic quantum theory. For example, one can always use a Bohmian theory with a preferred Lorentz frame at the level of hidden variables, which leads to Lorentz invariant measurable predictions. (It's only that I don't particularly like such a variant of Bohmian mechanics, because it looks somehow too cheap for me.)

Have you seen Lee Smolin's latest book Time Reborn? Our choice, according to Smolin, is epistemic/statistical QM or Aristotle was right!

Could there be a hidden-variables theory compatible with the principles of relativity theory? We know that the answer is no. If there were such a theory, it would violate the free-will theorem—a theorem implying that there’s no way to determine what a quantum system will do (hence no hidden-variables theory) as long as the theorem’s assumptions are satisfied. One of those assumptions is the relativity of simultaneity.

The aforementioned theorem of John Bell also rules out local hidden-variable theories—local in the sense that they involve only communication at less than the speed of light. But a hidden-variables theory is possible, if it violates relativity.

As long as we’re just checking the predictions of quantum mechanics at the level of statistics, we don’t have to ask how the correlations were actually established. It is only when we seek to describe how information is transmitted within each entangled pair that we need a notion of instantaneous communication. It’s only when we seek to go beyond the statistical predictions of quantum theory to a hidden-variables theory that we come into conflict with the relativity of simultaneity.

To describe how the correlations are established, a hidden-variables theory must embrace one observer’s definition of simultaneity. This means, in turn, that there is a preferred notion of rest. And that, in turn, implies that motion is absolute. Motion is absolutely meaningful, because you can talk absolutely about who is moving with respect to that one observer—call him Aristotle. Aristotle is at rest. Anything he sees as moving is really moving. End of story.

In other words, Einstein was wrong. Newton was wrong. Galileo was wrong. There is no relativity of motion.

This is our choice. Either quantum mechanics is the final theory and there is no penetrating its statistical veil to reach a deeper level of description, or Aristotle was right and there is a preferred version of motion and rest.

I like this book anyway...

 

[Ken G]

Just to avoid any 'misunderstanding' – there is no possibility to cover all possible outcomes in EPR-Bell in preexisting hidden variables. From my picture above, it may look like we have 6 unique binary values (excluding 0 and 7), but the truth is; there are only 2 x 3 'mirrored' values, which reduces to the indisputable fact that 1/3 ≠ 1/4.

I can't say I am following that logic, but I'm pretty confident that Bohmians are not such fools that they can't see their interpretation can be refuted by well known EPR-type observations! All the interpretations yield all the same experimental outcomes at this point, and people are trying very hard to try and find observations that can distinguish them, without a lot of success so far it seems to me.

 

[Ken G]

Have you seen Lee Smolin's latest book Time Reborn? Our choice, according to Smolin, is epistemic/statistical QM or Aristotle was right!

I find Smolin to be very thought-provoking, but I wish people would stop portraying science as a kind of "guessing game" about the "truth" such that you could either be "wrong" or "right." That's just not what science has ever been. The fact is Aristotle was right, Galileo was right, and Newton was right-- they were right in the only things they were ever saying, which was "here is a constructive way of looking at the situation that advances the goals of science." And they were right, it was. None of them ever said "here's the absolute truth that will stand for all ages", because no one who ever says that is going to be right-- if that is our standard of rightness, then none of them are it.

 

[DevilsAvocado]

I can't say I am following that logic, but I'm pretty confident that Bohmians are not such fools that they can't see their interpretation can be refuted by well known EPR-type observations!

I can only pass on the well known fact that local hidden variables are as dead as the Norwegian Blue Parrot – you just can't "take all the outcomes of the measurements you need to agree with" in advance, since this is proven impossible beyond all reasonable doubt and mathematical possibilities.

 

[DevilsAvocado]

I find Smolin to be very thought-provoking, but I wish people would stop portraying science as a kind of "guessing game" about the "truth" such that you could either be "wrong" or "right." That's just not what science has ever been. The fact is Aristotle was right, Galileo was right, and Newton was right-- they were right in the only things they were ever saying, which was "here is a constructive way of looking at the situation that advances the goals of science." And they were right, it was. None of them ever said "here's the absolute truth that will stand for all ages", because no one who ever says that is going to be right-- if that is our standard of rightness, then none of them are it.

Well, we all have our different preferences. Personally, I find it provoking to reduce the work of a faculty member at the Perimeter Institute for Theoretical Physics, an adjunct professor of physics at the University of Waterloo and a member of the graduate faculty of the philosophy department at the University of Toronto, awarded the Majorana Prize (2007) and the Klopsteg Memorial Award (2009) – as a "guessing game".

Maybe you read too much into the phrases right/wrong. Lee Smolin is of course intelligent enough to know the real premises of science, and the same thing naturally goes for Einstein:

"Newton, forgive me, you found the only way which, in your age, was just about possible for a man of highest thought and creative power." -- Albert Einstein​

 

[Ken G]

I can only pass on the well known fact that local hidden variables are as dead as the Norwegian Blue Parrot – you just can't "take all the outcomes of the measurements you need to agree with" in advance, since this is proven impossible beyond all reasonable doubt and mathematical possibilities.

Do you realize that by slipping in the word "local" in your sentence, you have disqualified your remarks from the perspective of Bohmian mechanics, which by their pilot-wave approach, reflect inherently non-local hidden variables?

 

[Ken G]

Well, we all have our different preferences. Personally, I find it provoking to reduce the work of a faculty member at the Perimeter Institute for Theoretical Physics, an adjunct professor of physics at the University of Waterloo and a member of the graduate faculty of the philosophy department at the University of Toronto, awarded the Majorana Prize (2007) and the Klopsteg Memorial Award (2009) – as a "guessing game".

You misunderstand me. I was not saying Smolin was playing a guessing game, I'm saying that Smolin, by saying "either Aristotle was right and Galileo/Newton/Einstein were wrong, or the other way around", is in effect framing what those great minds were doing as playing a guessing game, a kind of intellectual musical chairs: who will be in the chair of "rightness" when the music stops and all truth is revealed? But science never works like that, and framing it that way feeds misconceptions about what science really is. Whatever you think about his intellect, and his understanding of what science is, describing the situation as "who was right, who was wrong" is a faulty way to frame scientific progress. It seems a harmless flaw in his exposition, but actually, I think it is one of the few aspects of what he is saying that people can really understand, and glom onto. So it is important to get that part right, maybe even more important than a relativistic treatment of the Bohmian interpretation.

 

[DevilsAvocado]

Do you realize that by slipping in the word "local" in your sentence, you have disqualified your remarks from the perspective of Bohmian mechanics, which by their pilot-wave approach, reflect inherently non-local hidden variables?

Do you realize that the only logical interpretation of "take all the outcomes of the measurements you need" is local?

If you meant something else, why not explain it, instead of current unfruitful dispute?

 

[DevilsAvocado]

You misunderstand me. I was not saying Smolin was playing a guessing game, I'm saying that Smolin, by saying "either Aristotle was right and Galileo/Newton/Einstein were wrong, or the other way around", is in effect framing what those great minds were doing as playing a guessing game, a kind of intellectual musical chairs: who will be in the chair of "rightness" when the music stops and all truth is revealed?

This is your interpretation. I interpret Smolin as just basically saying that we have to choose between – based on empirical evidence and current knowledge – a world where a preferred notion of rest is possible, or there is no deeper level of description than statistical quantum mechanics.

Note that this is Smolin's 'hypothesis' on current knowledge. He is of course smart enough to realize that this could be changed, tomorrow, in case of some bright genius presenting a new idea.

I don't think Smolin would ever talk about 'metaphysical eternal truths'... everything in his book points towards the opposite direction...

Read it before judgment!

 

[Ken G]

Do you realize that the only logical interpretation of "take all the outcomes of the measurements you need" is local?

The measurements are local of course, but the hidden variables that Bohmian mechanics uses to account for them, and maintain classical properties, are nonlocal.

 

[Ken G]

This is your interpretation. I interpret Smolin as just basically saying that we have to choose between – based on empirical evidence and current knowledge – a world where a preferred notion of rest is possible, or there is no deeper level of description than statistical quantum mechanics.

Well you can choose to interpret those words he said that way, but others might not. The point is, had he said what you said, I'd have no objection, but he said what he said, and I voiced my objection. Clearly you have noticed the difference, as you needed to change his words when you inserted your interpretations. The issue here is the difference between us making "choices," which of course we must do to do science, versus scientists being "right" or "wrong." The "rightness" of science is advancing the progress of science, period, and they all did that. Was Ptolemy right or wrong? Some of both, of course, and the partition will always be a moving target. Was Copernicus right or wrong? Some of both, of course, and again that partition will always be a moving target. There's not a scorecard, there's the progress of science.

Note that this is Smolin's 'hypothesis' on current knowledge. He is of course smart enough to realize that this could be changed, tomorrow, in case of some bright genius presenting a new idea.

I am well aware that he probably knows that, the issue is what he said, and how people who do not know that can hear what he said. Ergo my point.

I don't think Smolin would ever talk about 'metaphysical eternal truths'... everything in his book points towards the opposite direction...

That doesn't surprise me, he is a deep thinker. All the same, his words were unfortunately chosen, and nothing you are saying speaks to that, indeed what you are saying here essentially support that.

Read it before judgment!

Who said I was judging Smolin? I spoke about a particular set of his words, and judged them. I did read them first, and then I pointed out the unfortunate aspect of them. What Smolin says somewhere else is of zero relevance to my point.

 

[DevilsAvocado]

Seriously Ken, this is on the brink to become hilarious...

You could at least have checked the provided Wikipedia link, before going baloney over something that obviously is completely wrong:

[my bolding]

Smolin argues for what he calls a revolutionary view that time is real, in contrast to existing scientific orthodoxy which holds that time is merely a "stubbornly persistent illusion" (Einstein's words).[1] Smolin reasons that physicists have improperly rejected the reality of time because they confuse their mathematical models—which are timeless but deal in abstractions that do not exist—with reality.[1] Smolin hypothesizes instead that the very laws of physics are not fixed, but that they actually evolve over time.

Satisfied?

 

[atyy]

Ah, now I think I see the source of confusion. One should distinguish two things:
1) SINGLE measurement, from which no information about probability distribution can be extracted (except that the obtained value has probability larger than one).
2) Statistical ENSEMBLE of similar measurements, from which the probability distribution can be extraced.

I was talking about the former, while it seems that you are talking about the latter. If I simultaneously measure position and momentum ONLY ONCE, I cannot extract any information about the joint probability distribution.

But then again, even in classical mechanics I can repeat many times the simultaneous measurement of position and momentum. From such a measurement I CAN extract the joint distribution. Moreover, by using the theory called classical STATISTICAL mechanics I can even predict or explain the joint distribution I measured. So your claim that "classical definitions for joint distribution don't exist" is certainly wrong.

My claim is correct, because it was for quantum mechanics.

What I don't understand is: what do mean by an "accurate" measurement? To define an accurate measurement in some sense, one needs a "correct" answer. In classical mechanics, one way to define a "correct" answer is that one correctly infers the value of the property that the particle had at a certain time. However, for joint measurements this definition of "correct" can't carry over to quantum mechanics, because the joint distribution of position and momentum doesn't exist in general.

 

[Ken G]

Satisfied?

I'm sorry, I don't see why you think that quote has the slightest relevance to anything that was said in our exchange. I know quite a bit about Smolin's ideas, you have not told me anything I didn't already know. I was pointing out a problem in his rhetorical device of saying that modern physics can determine whether it was Aristotle or Einstein that was right or wrong in regard to the relativity of space. Again, I can only tell you, that's just not how science works, and it is harmful to science to frame it that way. What actually happens is, scientists find insights that advance science, no one is ever right or wrong in any absolute sense. Truth in science is highly provisional, that is perhaps the main beauty of science-- it is constantly questioning and seeking knowledge. Science is not about what you know, it is about what you don't know. It seems my perspective is lost on you, but it doesn't matter, I was wrong to bring it up at all because it's not relevant to the thread and should be dropped anyway.

 

[atyy]

Bumping this just in case Demystifier did not see my response in #101.

 

[Ken G]

If Demystifier doesn't take up that cause, I would offer that measurements in science are always axiomatic. There is nothing more basic than a measurement in empirical science, nothing that we use to check that we are doing measurements "accurately"-- other than a body of other measurements we already regard as accurate by experience. We do check precision, and if ten people get ten badly different answers, we label that measurement "imprecise" and drop it from our set of approved techniques. But it is problematic to define a measurement as accurate by saying it agrees with some theory (other than the most basic theories that we already regard as axiomatic).

If quantum mechanics were ever regarded as axiomatic, then the definition of an accurate measurement as one that mimics a projection would be appropriate. I believe that Demystifier's core stance is that all axiomatic approaches to measurement must be classical, so you will always need a better definition of a measurement than that the result agrees with quantum mechanics theory. After all, if you are looking for chinks in the armor of QM as it is currently postulated, you certainly can't have someone scratching their head and saying "what did I do wrong in my measurement, my answer did not come out like QM."

But there is a case where the Ozawa definition could be appropriate, which is when we are not regarding measurements as a test of QM, but rather, as a proxy for understanding what QM is predicting, a lens on the theory if you will. In the form of a gedankenexperiment, which is used to describe a theory not reality, it is fine to use Ozawa's approach, to see in effect what QM thinks a measurement is, rather than what we have axiomatized it to be.

 

[Demystifier]

What I don't understand is: what do mean by an "accurate" measurement?

To help me answer that question, can you quote where exactly did I say that a measurement is "accurate"?