Questions About Quantum Theory: What's Wrong?

[juvenal]

I retract the "modified", it should have been "evolved". I never meant to imply that QM yielded incorrect results, but to imply that you make approximations in the form of perturbations etc for most problems at hand. In E&M you do not have to approximate to get an answer, maybe the solutions are not exactly tractable, but the basic equations have not been modified since Maxwell.

I spend about half my time working classical problems (Optics etc...) and the other half working in Optical Properties of Semiconductors and am truly amazed after all the years I have been doing this type of thing that QM gives reasonably correct answers that are verified experimentally. Devices I design work and the basic principles behind them are derived from the quantum mechanics of solids how do we do better?

Right - but classical E&M does not work at the atomic level. It is wrong.

There is the modern notion of "effective theory" which explains why our theories work as well as they do.

http://arxiv.org/PS_cache/cond-mat/pdf/9502/9502052.pdf

http://arxiv.org/PS_cache/gr-qc/pdf/0311/0311082.pdf

provide some background on this notion. There are other similar sorts of papers out there.

 

[quantumdude]

Right - but classical E&M does not work at the atomic level. It is wrong.

That's what I was driving at. QM may not yield exact solutions for atoms more complex than hydrogen, but the classical electrodynamics of charged particles doesn't even buy you that much.

 

[Andrew Mason]

I spend about half my time working classical problems (Optics etc...) and the other half working in Optical Properties of Semiconductors and am truly amazed after all the years I have been doing this type of thing that QM gives reasonably correct answers that are verified experimentally. Devices I design work and the basic principles behind them are derived from the quantum mechanics of solids how do we do better?

Models don't have to be correct to be useful. The model of magnetic field consisting of little lines of force works really well. The model of electricity being a fluid passing through wires under pressure has some limited usefulness as well. Treating current flow in semiconductors as electrons and holes moving through a junction works too. Kepler's nested spheres could be used to work out orbits of the planets.

Usefulness in predicting results is important. But if one wants to understand what the 'reality' is, one needs a theory that does more than usefully predict results.

AM

 

[ZapperZ]

Models don't have to be correct to be useful. The model of magnetic field consisting of little lines of force works really well. The model of electricity being a fluid passing through wires under pressure has some limited usefulness as well. Treating current flow in semiconductors as electrons and holes moving through a junction works too. Kepler's nested spheres could be used to work out orbits of the planets.

Usefulness in predicting results is important. But if one wants to understand what the 'reality' is, one needs a theory that does more than usefully predict results.

AM

Whoa!

There is a difference between a "physics model" and "visualization"! You are confusing those two! Magnetic field lines are NOT a "model", it is a visualization! I challenge you to show me there there are "magnetic lines" in the Maxwell equations.

Usefully predicting result IS a part of reality! In fact, it is ALL that we have! Again, show me a "theory" that does MORE than this. Till then, what you have said is all hypothetical.

Zz.

 

[Dr Transport]

Right - but classical E&M does not work at the atomic level. It is wrong.

Didn't I mention that exact fact in my original post, E&M works down to the point where QED takes over. There isn't a theory to my knowledge that takes into account everything. I took a Relativity course from a guy (Mendel Sachs) who claimed that after rewriting Einsteins equations in terms of quaterians and taking the linear limit he reproduced Quantum Mechanics and for a time had the best calculation of the Lamb shift. One of his students calculated the mass of the tau lepton to within an order of magnitude without using QM at all using Reletivity theory only.

Now Reletivity is about as close as I have ever seen a theory accounting for a the very large to the very small.

I still maintain that E&M is a very good, if not excellent field theory and QM is a close second.

 

[jtbell]

Usefulness in predicting results is important. But if one wants to understand what the 'reality' is, one needs a theory that does more than usefully predict results.

But how does one judge how well a theory corresponds to "reality," if not by its ability to usefully predict results?

 

[binarybob0001]

The problem we are discussing is written in Aristotles meditations. Essentially, the only thing I know for certain is that I exist. I may be writing this letter to nothing or something non-human or to a human or for nothing, but I do not know answer. The problem is we cannot reach out past our own experience. Some people argue that we can learn without experience(a priori), but I am trying to make a simple explanation, so I will only talk about learning from experience(a posteriori).
The complaint is that QM is taken from observation with no axioms to back it up. Newton was wrong. Einstein may be really close to the right answer or right on. QM may be right or close.
Newton based his theory off of experiences with the world. Newton could not see the whole structure of reality(If he could, would he know he was observing all reality?); therefore, he could not know if his theory was right or wrong. An old textbook will list Newtons theory as fact(a law), but we know that to be wrong. The point is, there cannot be axions about reality because we cannot go beyond our experience. Therefore, we cannot know what theories are right and what theories are wrong. We know that a theory explains our experiences with the world, and that is all. Whether the world really works that way, we can never know. Unless as I said before, learnig fully about the universe is possible a priori. I will let you guys decide whether or not that is possible.
QM, like or not, is a theory that explains much of the observable phenomenon in the universe. People may say that it is counter intuitive, and I understand. People should understand though that there will never be a truly provable TOE(theory of everything), only a TOE that explains all we observe. We can always ask the question is there something else? And we will never be able to answer it.

 

[ZapperZ]

People should understand though that there will never be a truly provable TOE(theory of everything), only a TOE that explains all we observe. We can always ask the question is there something else? And we will never be able to answer it.

Even that part is debatable. Condensed matter physicists like Anderson, Laughlin, and Pines would argue that the so-called TOE that is the result of a unified theory is simply a TOE for REDUCTIONISM, not a TOE of physics. I have described this at length with appropriate references elsewhere and won't bore long-time readers with this point again.

Zz.

 

[DrChinese]

But if one wants to understand what the 'reality' is, one needs a theory that does more than usefully predict results.

AM

It is very debatable that any good theory does more than provide utility, or that there is an objective reality that can be understood in any meaningful sense.

(For example, tell me ANY physical theory that explains reality in a way that folks will say, hmmm, now I completely understand that. GR gives us a nice pretty picture, for example, but how does it explain reality any more than QM does?)

 

[Stingray]

The problem with QM is that it is basically two theories stuck together in an incomplete way. You have the usual unitary evolution of states, which is nice and elegant. But you also have a completely separate process of measurement, which cannot be applied in an unambiguous way. It is certainly true that for a large class of experiments, people have learned what a measurement is (i.e. how to apply the formalism in a way that matches experiments). This has obviously been extremely useful, but is also incomplete.

The measurement postulates cannot be applied uniquely in any imaginable situation. If QM is a fundamental theory of physics, it should describe everything. In principle, it should be possible to remove any external "classical experimenter." Does the whole system then evolve unitarily? How would that be reconciled with the fact that we seem to need the projection postulate experimentally? Is that just an approximation when a large number of degrees of freedom are involved?

The resolutions of these problems will (probably) not have any profound influence on "practical" physics, but that doesn't remove their importance.

 

[ZapperZ]

The problem with QM is that it is basically two theories stuck together in an incomplete way. You have the usual unitary evolution of states, which is nice and elegant. But you also have a completely separate process of measurement, which cannot be applied in an unambiguous way. It is certainly true that for a large class of experiments, people have learned what a measurement is (i.e. how to apply the formalism in a way that matches experiments). This has obviously been extremely useful, but is also incomplete.

The measurement postulates cannot be applied uniquely in any imaginable situation. If QM is a fundamental theory of physics, it should describe everything. In principle, it should be possible to remove any external "classical experimenter." Does the whole system then evolve unitarily? How would that be reconciled with the fact that we seem to need the projection postulate experimentally? Is that just an approximation when a large number of degrees of freedom are involved?

I disagree. This is because the quantities we measure are classical! Position, momentum, energy, etc. are all "classical" quantites that we inherited out of classical mechanics. QM is simply telling us what they are if we insist on using these quantities. (Again, square objects being forced through round holes).

Unless we want to invent a new set of quantites and concepts, we're stuck with these classical ideas. If you have followed this thread, you'll know how DIFFICULT it is to make some people part with their beloved classical quantities.

Tony Leggett recently wrote a terrific article summarizing the so-called "measurement problem" of QM, and in particular, the Schrodinger Cat-type phenomenon (it is no longer a "paradox").[1] I strongly suggest people who insist that there is a "measurement problem" to read this, and his other paper in J. Phys. Cond. Matt. to look at the wealth of experimental observations and how they compare to what we know about what QM is saying.

Zz.

[1] A.J. Leggett, Science v.307, p.871 (2005).

 

[Andrew Mason]

It is very debatable that any good theory does more than provide utility, or that there is an objective reality that can be understood in any meaningful sense.

(For example, tell me ANY physical theory that explains reality in a way that folks will say, hmmm, now I completely understand that. GR gives us a nice pretty picture, for example, but how does it explain reality any more than QM does?)

Good point. A theory simply reduces a lot of little things that we didn't understand to fewer bigger things that we still don't understand: like Newton's laws which explained why things move the way they do in terms of forces and inertia.

And no doubt about it: QM does reduce a lot of little things to fewer rules and principles. The problem is that, unlike force and inertia, people lack an intuitive grasp of what these rules and principles mean physically. I long ago concluded that one should not expect to have that intuitive grasp of QM as the world at that level is quite beyond our experience.

But I don't think it is wrong for people to find QM's explanations unsatisfying. I should think that any good physicist would always want to improve understanding of the physical world. One can't do that if one is completely happy with all the current explanations of things. So that is why I complain about people who complain about people complaining about the 'unsatisfactory' explanations that QM provides.

AM

 

[ZapperZ]

Good point. A theory simply reduces a lot of little things that we didn't understand to fewer bigger things that we still don't understand: like Newton's laws which explained why things move the way they do in terms of forces and inertia.

And no doubt about it: QM does reduce a lot of little things to fewer rules and principles. The problem is that, unlike force and inertia, people lack an intuitive grasp of what these rules and principles mean physically. I long ago concluded that one should not expect to have that intuitive grasp of QM as the world at that level is quite beyond our experience.

But I don't think it is wrong for people to find QM's explanations unsatisfying. I should think that any good physicist would always want to improve understanding of the physical world. One can't do that if one is completely happy with all the current explanations of things. So that is why I complain about people who complain about people complaining about the 'unsatisfactory' explanations that QM provides.

AM

But then you miss the point of my complaint about people who complain about QM. There are ways to "complain", and just saying it based on a matter of tastes is NOT the way to complain about ANYTHING in physics, and not just QM. If you complain about E&M based simply because you dislike the idea of "fields", I would complain about your complaints too. This has nothing to do with QM. It has everything to do with how one challenges ideas in physics.

You cannot base it on "intuition", or "intuitively easy". Your intuition changes all the time as you gain more knowledge. What was intuitively difficult when you were 18 can be intuitively obvious when you're 30. So just saying QM is intuitively difficult (especially for the masses) should be completely irrelevant. Nature owes none of us any explanation on why she behaves that way. To force her to fit into our "intuition", as limited as it is, is plain arrogant. As scientists, that's the worst mistake we can ever do.

Zz.

 

[reilly]

Classical theory leaves a lot of open questions, which, typically, we don't consider often:

Whence electric charge? Why is it quantized? Why do protons and electrons have charge that is equal and opposite?


Why is the speed of light what it is?

Why is gravity, under normal human conditions, weaker than the other forces in nature?

Why does F = dP/dt?

Why are so many repeated measurements the same, relative to standard measurement error?

Why does the sun come up every day?

What is mass?

Does the elecromagnetic spectrum include arbitrarily large frequencies?

Relative to human scales, why are stars and galaxies generally so widely spread?

How, albeit under extreme circumstances, can a single person lift a car?

Why does probability work?

Why is there regression to the mean? (That is, why does statistics work?)

Why do some heavy smokers not get lung cancer?

Why is there so much nonlinearity in the world?

Why don't our children follow our superb advice?

And, on and on.

Life is full of mysteries. To suggest QM is inadequate because we do not understand the measurement process is to miss the point. Classical probability and statistics, as used in practice, suffer from the same problem, that is collapse. Prior to a coin toss the probability of a heads is 1/2. If the toss comes up tails, then the probability of tails is 1, and he probability of heads is 0. Looks like collapse of the probability to me -- not much different from measuring whether spin up or spin down.

Classical or quantum, science and physics are imperfect, but they do give us a beacon of light into the void -- and, from my own experience, that is a great gift as one gets older. Life is full of mysteries. You do the best you can with what you've got.

Regards,
Reilly Atkinson

 

[Locrian]

but the fact is that QM was guided more in its development by experiment than by actual physics.

What?? Physics by definition develops by experiment. Saying Einstein didn't use experiment is rediculous - he just wasn't the one doing the experiments. The first page of his paper on special relativity are the Lorentz transformation equations, which are based on maxwell's equations, which are 100% experimentally derived.

There is no part of physics that does not rely on experimentation, whether before or after, and without experimentation we'd still be in the dark ages. Assumptions and hypothesis are all well and good, but they require experiment to verify them. Einstein had the habit of guessing first and testing later (with exception maybe of his experiments in photon absorption), and this ended up making him wrong very, very often, and right very, very rarely. I guess if we remember his hits and forget his misses we are to be impressed?

Saying that zapperz is impeding physics is amusing, given his employment and the hard work he does on these forums. I would love to see who had contributed more to physics, Crosson or Zapperz.

 

[Eye_in_the_Sky]

As far as I can figure out, classical probability is as subject to collapse as QM. That is, measurement simply tells us at that moment what is, whether an electron in a scattering experiment, or the price of IBM stock, a sales forecast, or what you will have for dinner in two weeks.

Yes, the notion of "collapse" can be applied to classical scenarios. However, in order to put quantum "collapse" on equivalent 'footing', one must be prepared to accept as true the physical existence of "hidden variables". If "hidden variables" do not physically exist, then any 'induced' change in the quantum state-vector implies a corresponding physical change in the status of the system in question. ... I see no way around it (... except perhaps to 'deny reality', whatever that is supposed to mean).
______________

There's one thing I've never understood about the Schrodinger cat problem. It has nothing to do with QM, and everything to do with standard probability matters.

Again, this standpoint is consistent with a "hidden-variables" perspective. But from the alternative perspective (i.e. "hidden variables" do not physically exist), the Schrödinger-cat scenario has everything to do with Quantum Mechanics, and nothing to do with standard probability matters.

To see that this is so, consider – from the "no-hidden-variables" perspective – the following:

Suppose that a quantum system is in the state

|ψ> = (1/√2) [ |φ₁> + |φ₂> ] ,

where the states |φ₁> and |φ₂> are eigenstates of an observable which we can physically measure.

(Remember, we are assuming here that there are no "hidden variables". The state vector gives a "complete" characterization of the physical state of the system.)

Now, what do we want to say about this situation? Do we want to say that the quantum system is not at all actually in the said (physical) state of superposition, but that it is, in fact, in one or the other of the (physical) states |φ₁> or |φ₂> with probability equal to ½ ?

... Certainly NOT!

From this perspective, then, the Schrödinger-cat scenario is a challenge to the following contention:

The quantum-mechanical state-vector description can be meaningfully applied to systems of arbitrary "size" and "character".

But the challenge is raised only in the context of no "hidden variables".
______________

Tony Leggett recently wrote a terrific article summarizing the so-called "measurement problem" of QM, and in particular, the Schrodinger Cat-type phenomenon (it is no longer a "paradox").[1] I strongly suggest people who insist that there is a "measurement problem" to read this, and his other paper in J. Phys. Cond. Matt. to look at the wealth of experimental observations and how they compare to what we know about what QM is saying.

[1] A.J. Leggett, Science v.307, p.871 (2005).

I recently read an essay of Leggett's written some time around the mid to late 80's in which he discussed the "Measurement Problem" with a great deal of care. In that essay, he suggested "the possibility that the complexity of a physical system may itself be a relevant variable which may introduce new physical principles." I am excited to find out what conclusions he has now reached some two decades later. Thank you for posting the above reference, ZapperZ (and also for the many others which you have posted).

... Can you be more specific about the J. Phys. Cond. Matt. paper?

 

[ZapperZ]

I recently read an essay of Leggett's written some time around the mid to late 80's in which he discussed the "Measurement Problem" with a great deal of care. In that essay, he suggested "the possibility that the complexity of a physical system may itself be a relevant variable which may introduce new physical principles." I am excited to find out what conclusions he has now reached some two decades later. Thank you for posting the above reference, ZapperZ (and also for the many others which you have posted).

... Can you be more specific about the J. Phys. Cond. Matt. paper?

The reference to the J. Phys. paper is one of the papers Leggett cited in his Science article. It is a more in-depth look at the Schrodinger Cat-type scenario, especially in light of the Delft and Stony Brook's recent experiments using SQUIDs.

Zz.

 

[Eye_in_the_Sky]

The reference to the J. Phys. paper is one of the papers Leggett cited in his Science article. It is a more in-depth look at the Schrodinger Cat-type scenario, especially in light of the Delft and Stony Brook's recent experiments using SQUIDs.

Zz.

Thanks!

 

[Stingray]

I disagree. This is because the quantities we measure are classical! Position, momentum, energy, etc. are all "classical" quantites that we inherited out of classical mechanics. QM is simply telling us what they are if we insist on using these quantities. (Again, square objects being forced through round holes).

Then what should be measurable? Any physical theory must describe what its variables mean in some (operational) sense. If those variables can only be described by mixing QM and classical mechanics, then this is itself a type of incompleteness. Or are there measurement types I'm not aware of?

Also, thanks for the reference. I'll take a look at it.

 

[dextercioby]

We can measure typical QM quantities:probabilities with which certain eigenvalues occur and (integral) cross sections...What else is there to QM...?

Daniel.

 

[Haelfix]

"The Schrodinger Cat-type phenomenon (it is no longer a "paradox")."

Forgive me if I am a little bit skeptical about these sorts of matters. The paradox of Schrodingers cat ultimately is the problem of resolving when a quantum state becomes a classical one, and last I checked there is still wars going on in the measurement camp about these thorny issues. Taking to the extreme you end up with one of two scenarios.

1) Everything is quantum, classical behavior is just an emergent illusion. Ergo notions like the wave function of the universe become acceptable, despite their known enormous failure in various field theories and quantum gravity.

2) Something weird happens and there is some sort of phase change between quantum to classical mechanics. Perhaps its something not accounted for, or some tiny effect that only becomes important in the huge complexity of interactions. Eg people can't, under any circumstance, walk (tunnel) through walls.. The probability isn't just 1e-50502, but identically zero.

 

[ZapperZ]

Then what should be measurable? Any physical theory must describe what its variables mean in some (operational) sense. If those variables can only be described by mixing QM and classical mechanics, then this is itself a type of incompleteness. Or are there measurement types I'm not aware of?

The problem here is that ALL our measureables or observables are classical quantities. This is what I have been trying to stress all along. We have no other alternatives (so far). QM is our description of a world beyond classical physics USING classical physics concepts. When we do that, OF COURSE some things will simply make no sense based on our classical measurement. Our squares and cubes came through the round holes with the edges chopped off, and we struggle to still want to call them squares and cubes when they do not look quite like squares and cubes. They look funny to us. So we blame the round hole and forget all about the fact that we were forcing incompatible shapes through the round hole.

With this in mind, I do not worry about the "measurement problem", the schrodinger cat, the EPR-type measurement, the "collapsing" wave function, etc.. etc. It is what it is, and that's what Nature has decided to reveal herself so far. I have accepted and am aware of my prejudice of trying to describe nature using concepts that may not be accurate. I use the formalism, but pay more attention to the experiments. As far as I'm concerned, in the end, that is the only thing that matters.

Zz.

 

[Crosson]

ZapperZ, I like what you say here:

Unless we want to invent a new set of quantites and concepts, we're stuck with these classical ideas.

In my opinion this is the progressive approach we need to take towards QM. Take for example the particle-wave duality; what we need is a new concept, not a bunch of people saying "Sometimes its a particle and sometimes its a wave and that's just the way it is".

Energy and momentum observables are already predicted very well by QM, there is no reason to change this. We know energy and momentum are conserved quantities, and that they obey the quantized hamiltonian relationship. But unlike the state of affairs in classical mechanics, all we know in terms of experiment is that energies are proportional to the frequency of photons.

In terms of the thing we are describing, energy is not defined (without resorting to a circular definition relating momentum and energy, or the relation of "energy" to the frequency of emitted photons rather than the system we are talking about).

A separate criticism of QM which I have not brought up (because physicist seem to attack philosophers) is determinism. In my mind QM is a theory of observations, and so in cannot make claims concerning determinism. But many people embrace the idea that QM makes the universe indeterminate, and this is:

1) Not supported by experiment or quantum theory (as I read it).

2) Physically appaling, in that an indeterminate event is necessarily uncaused.

Because it is a theory of observations, QM involves indeterminate events and fails in not describing explicit causes.

Keep in mind that I know QM proves that the universe is indeterminate for an observer, but that does not rule out determinism. (indeed, it was clear from classical chaos that it would never be possible to actually predict the future this way)

I define determinism as: The present corresponds to only one future.

 

[ZapperZ]

"The Schrodinger Cat-type phenomenon (it is no longer a "paradox")."

Forgive me if I am a little bit skeptical about these sorts of matters. The paradox of Schrodingers cat ultimately is the problem of resolving when a quantum state becomes a classical one, and last I checked there is still wars going on in the measurement camp about these thorny issues. Taking to the extreme you end up with one of two scenarios.

1) Everything is quantum, classical behavior is just an emergent illusion. Ergo notions like the wave function of the universe become acceptable, despite their known enormous failure in various field theories and quantum gravity.

2) Something weird happens and there is some sort of phase change between quantum to classical mechanics. Perhaps its something not accounted for, or some tiny effect that only becomes important in the huge complexity of interactions. Eg people can't, under any circumstance, walk (tunnel) through walls.. The probability isn't just 1e-50502, but identically zero.

You have mentioned two separate aspects of the so-called Schrodinger Cat paradox. Again, most of the stuff that I have mention (and will mention here) are contained in Tony Leggett's paper, including the full treatment in his J. Phys paper.[1] I said that the Schrodinger Cat scenario is no longer a paradox because (i) it does truly occur at the QM scale.[2,3] These have been unambiguously verified.

The issue that's left, which is the 2nd part of your point, is why don't we see it at the macroscopic scale. I do not see this as another "paradox" because while it is still an active research area, we have plenty of indications of possible explanations for such dichotomy. We have seen how classical system emerges into the classical observation via careful and controlled decoherence.[4,5] Not only that, there are every indications that the emergence of "objective" properties come out of a "selective" destruction of quantum coherence, resulting in what is known as preferred pointer states.[6] This is what eventually results in what we perceive macroscopically as an "objective" and deterministic universe.

I am far from claiming this is a done issue. I just do not see this as being a "paradox" anymore as if these things simply have no explanation whatsoever. In fact, there exists many plausible explanations that connects the evolution from quantum states into classical observations. We just need more experiments to verify them convincingly.

Zz.

[1] A.J. Leggett, J. Phys. Cond. Matt. v.14, p.415 (2002).
[2] J.R. Friedman et al. Nature v.43, p.406 (2000).
[3] C.H. van der Wal et al. Science v.290, p.773 (2000).
[4] K. Hornberger et al. PRL v.90, p.160401 (2003).
[5]C.J. Myatt et al. Nature v.403, p.269 (2000).
[6] H. Ollivier et al. PRL v.93, p.220401 (2004).

 

[reilly]

Yes, the notion of "collapse" can be applied to classical scenarios. However, in order to put quantum "collapse" on equivalent 'footing', one must be prepared to accept as true the physical existence of "hidden variables". If "hidden variables" do not physically exist, then any 'induced' change in the quantum state-vector implies a corresponding physical change in the status of the system in question. ... I see no way around it (... except perhaps to 'deny reality', whatever that is supposed to mean).
______________Again, this standpoint is consistent with a "hidden-variables" perspective. But from the alternative perspective (i.e. "hidden variables" do not physically exist), the Schrödinger-cat scenario has everything to do with Quantum Mechanics, and nothing to do with standard probability matters.

To see that this is so, consider – from the "no-hidden-variables" perspective – the following:

Suppose that a quantum system is in the state

|ψ> = (1/√2) [ |φ₁> + |φ₂> ] ,

where the states |φ₁> and |φ₂> are eigenstates of an observable which we can physically measure.

(Remember, we are assuming here that there are no "hidden variables". The state vector gives a "complete" characterization of the physical state of the system.)

Now, what do we want to say about this situation? Do we want to say that the quantum system is not at all actually in the said (physical) state of superposition, but that it is, in fact, in one or the other of the (physical) states |φ₁> or |φ₂> with probability equal to ½ ?

... Certainly NOT!

From this perspective, then, the Schrödinger-cat scenario is a challenge to the following contention:

The quantum-mechanical state-vector description can be meaningfully applied to systems of arbitrary "size" and "character".

I have no clue why I would need hidden variables to suggest a correspondence between classical and quantum notions of probability. So, I would be most grateful to find out what I'm missing.

Because I've worked in the consulting business for years with probability and statistics, and in my younger years for some time as a particle theorist and teacher of QM, I've concluded from practical experience that the two probabilities are, generically the same. In fact, what made and makes sense to me is that the probabilities are statements about our knowledge, as, more or less, suggested by von Neuman and Wigner. The troubling collapse is a reflection of changes in our knowledge.

Subsequent to the long time it took me to come to this conlusion, I discovered that the great physicist, Sir Rudolph Peierls, agrees -- or, really I agree with Sir Rudolph -- with the knowledge interpretation. In Andrew Whitaker's book, Einstein, Bohr and the Quantum Dilemma(1996), Peierls " a theoretical physicist of massive achievements" In response to Bell's "Against Measurement" (1990), Peierls writes:

In my view the most fundamental statement of quantum mechanics is that the wavefunction, or more generally the density matrix, represents our knowledge of the system we are trying to describe.

Whitaker quotes more, but I shan't bore you, Whitaker's book is a review of the history of QM, particularly of QM interpretation -- Bohr to Bohm to Everett and more. The book is a very impressive piece of work.

Peierl's ideas make great sense to me. Again, quantum or classical, you don't know until you measure. Why, if we, say, see a boat sail out to sea, to vanish over the horizon, do we say we know the boat, in all probability, continues to sail after we lose sight? After all, we cannot see it. (Not a bad question for a PhD candidate's oral exam.)

With a spin doublet, superposition, according to Peierls, just says there are two possibilities. To work backwards seems to me to be an exercise in futility. Why should nature follow our conceits? I always thought science and physics were about nature, not about man's preconceived notions, whether ego driven or not.


If you don't like QM as it is, find a better way. Explain the electron microscope, or semiconductors some other way. So far, nobody has come close, and I find that very telling -- even though I'm not so foolish as to maintain there can't be a better way.

I posed a set of questions a post or two above. Why has no one dealt with them? I find that rather odd, given the lofty thoughts about the inadequacies of QM. QM ain't the only problem.

I remain a servant of Nature.

Reilly Atkinson

 

[Locrian]

1) Not supported by experiment or quantum theory (as I read it).

2) Physically appaling, in that an indeterminate event is necessarily uncaused.

I see no supporting evidence for the first, and the second is simply wrong. There is no reason that an indeterminate event is uncaused.

 

[ZapperZ]

ZapperZ, I like what you say here:



In my opinion this is the progressive approach we need to take towards QM.

Puhleeze. I'm the one hindering progress in physics, remember?

Take for example the particle-wave duality; what we need is a new concept, not a bunch of people saying "Sometimes its a particle and sometimes its a wave and that's just the way it is".

This is just plain wrong. Where exactly in QM is there this "duality"? Really now!

Such things are only used when it is being described to the general public! This duality is NOT part of QM! There is only ONE description of light and matter, not two, not three. Every single so-called "wave" behavior can be fully and consistently described via the SAME description that describe the "particle" behavior. There are no dichotomy. The dichotomy only comes in because CLASSICALLY, those two are different beasts and are incompatible with each other.

Again, this is another clear example where we are forcing the classical picture onto the QM domain. Our insistence that "wave" and "particle" are separate descriptions is rearing its ugly head, while we ignore the fact that QM has no such separation!

Energy and momentum observables are already predicted very well by QM, there is no reason to change this. We know energy and momentum are conserved quantities, and that they obey the quantized hamiltonian relationship. But unlike the state of affairs in classical mechanics, all we know in terms of experiment is that energies are proportional to the frequency of photons.

You are not describing QM in general here. You are describing only "special cases" here. Energy and momentum are not always conserved, especially when you deal with virtual particles. "Quantized" hamiltonian isn't automatic! Quantization is a direct result of the boundary condition. Energy, momentum, position, etc are NOT quantized for a free particle, for example. And it is certainly continuous in the conduction band of solids.

A separate criticism of QM which I have not brought up (because physicist seem to attack philosophers) is determinism. In my mind QM is a theory of observations, and so in cannot make claims concerning determinism. But many people embrace the idea that QM makes the universe indeterminate, and this is:

1) Not supported by experiment or quantum theory (as I read it).

2) Physically appaling, in that an indeterminate event is necessarily uncaused.

Because it is a theory of observations, QM involves indeterminate events and fails in not describing explicit causes.

Keep in mind that I know QM proves that the universe is indeterminate for an observer, but that does not rule out determinism. (indeed, it was clear from classical chaos that it would never be possible to actually predict the future this way)

I define determinism as: The present corresponds to only one future.

Again, you haven't cited any experimental observations. All you have done here is cite your distaste. I have posted a reference to a paper relating a decoherence process with an emerging pointer states. I suggest you read that.

Zz.

 

[reilly]

There are a few things worth remembering about Quantum Physics, to return to my mantra, which are evident upon reading history, and getting the facts straight.

Classical theory in the face of
Blackbody radiation
Photoelectric effect
Atomic Spectra,
Electron diffraction

was rendered totally impotent. Classical ideas failed to do the job. Nature failed to agree with our usual modes of perception and description. That was fact 100-75 years ago. In spite of revisionist attempts, classical theory and concepts still fail to describe many phenomena. ZapperZ hit this right on the head.

Nature has posed us with phenomena that appear to be logically impossible, beyond the scope of our language. JJ Thomson discovered the electron as a particle. Some years later, Davisson and Germer discovered electron diffraction. How can this be? We still grapple with this seeming contradiction, there appears to be no way around it. So, indeed the formulation of QM is grounded in experimental evidence. How, in their right mind would could anyone invent QM out of nothing?

Nature has wired us to have certain perpetual mechanisms, which give us direct experience of very little of Nature. That our logic, intuition and knowledge do not extend well into parts of Nature that are beyond our perceptual boundaries, is not surprising, and has been discussed by many. So to our normal way of thinking, Nature can be very weird. It's the job of science to accept this empirical weirdness, and try to deal with it, and if some of our concepts are found wanting, well that's that.

Determinism? I think David Hume pretty much destroyed that concept a few hundred years ago. (Read and learn)

Yes, we could use new concepts. The question is: how do we formulate them?

I guess that what's wrong with QM is that it does not easily fit into the usual modes of thinking. It is weird because what it describes is weird.

Regards,
Reilly Atkinson

 

[caribou]

I've been reading a bit about interpretation and I've got to say that I agree with a lot of what Crosson is saying.

Perhaps the most interesting attempt at interpretation is the "decoherent histories" approach (closely related to the "consistent histories" approach). What that does is takes quantum mechanics seriously and see what the theory actually says and where it leads in a modern way.

One of the things you end up with is the modern equivalent of complementarity -- like for example wave/particle duality -- of the most extreme kind when you try to describe a series of events. You are limited to talking about what is consistent or inconsistent.

A good example of this is in the two slit experiment:

A) If you choose to describe where the particle hits the screen, you can't describe which slit the particle goes through.

or

B) If you choose to describe which slit the particle goes through, you can't describe where it hits the screen.

And that is that. Trying to combine both the A and B descriptions makes no sense. Another good example is particle spin version of EPR in which not realizing that there are lots of other ways of describing the events makes many people think there is some influence traveling from one particle to another.

(Roland Omnes is someone whose books I think I'd recommend here as he has some expertise in this area.)

That there are different incompatible descriptions of the same events is mind-boggling. And these, of course, in quantum mechanics are events without causes. And these could be events without causes that split the universe into different universes as well. But that is apparently where quantum mechanics leads us.

It's like some kind of nightmare... but the theory will give the right answers if asked the right questions.

I think you can see why I might agree we need a better theory. I'm not giving up hope that we can think of something a bit less insane.

You know, suddenly determinism doesn't sound so bad anymore.

 

[ZapperZ]

One criticism that I will level off regarding your comments here is that you should leave room for the possibility that you understood QM wrongly. If you start off with the wrong premise, then of course you will not get this right. It is unfair to criticize or comment about anything, not just QM, when you start off with the wrong understanding of what it is.

I will illustrate:

I've been reading a bit about interpretation and I've got to say that I agree with a lot of what Crosson is saying.

Perhaps the most interesting attempt at interpretation is the "decoherent histories" approach (closely related to the "consistent histories" approach). What that does is takes quantum mechanics seriously and see what the theory actually says and where it leads in a modern way.

One of the things you end up with is the modern equivalent of complementarity -- like for example wave/particle duality -- of the most extreme kind when you try to describe a series of events. You are limited to talking about what is consistent or inconsistent.

A good example of this is in the two slit experiment:

A) If you choose to describe where the particle hits the screen, you can't describe which slit the particle goes through.

or

B) If you choose to describe which slit the particle goes through, you can't describe where it hits the screen.

This is incorrect. First of all, in (B), I can describe where it hits the screen equally well as where it went through the slit (your usage of the phrase "choose to describe" is something I haven't seen in physics). It is the interference pattern that goes missing when I unambiguously can determine which slit the object passes through. You get two gaussian peaks, rather than the regular interference pattern. There's nothing to prevent me from making a full description of anything.

And that is that. Trying to combine both the A and B descriptions makes no sense. Another good example is particle spin version of EPR in which not realizing that there are lots of other ways of describing the events makes many people think there is some influence traveling from one particle to another.

Again, your "sense" isn't perfect, and it EVOLVES. These things may not make sense to you, but they make perfect sense to me because I've admitted my prejudice in forcing nature of confine herself into my classical concepts of "position" and "wave" and "particle", etc. (square objects through round holes).

We cannot, and should not do physics simply via a matter of tastes. Go through this thread carefully and you will see that in every single instance of people making a so-called challenge to QM, it is done due to some personal preferences. I have said this way in the beginning of this thread. I do not understand why people continue think this is a valid way to challenge anything in physics, not just QM.

And things are made worse when people make the wrong statement about QM in the first places, such as "wave-particle duality". Everytime someone says that that is what QM is describing, I tend to think that person has never studied QM formally. I mean, where in QM is there such a thing as two separate descriptions of wave-particle? You cannot expect to be able to make a coherent comment about any subject when you only have a superficial understanding of it. QM cannot be learned and understood via its interpretation! You cannot get the full taste of a food via its description by someone else!

Zz.

 

[caribou]

ZapperZ,

I think you should also leave room for the possibility that I've understood something correctly but that I could be doing a much better job of describing it. I'm sorry if that's the case, as I think it probably is.

Yes, I agree that descriptions A and B in the two slit experiment work together just fine when measurement or interaction is allowed but this isn't allowed in the situation I mean.

What I am referring to involves only the usual set-up of the two slit experiment leading solely to the interference effect and not any other set-up which allows for measurement or interaction around the slits. This set-up I mean features no destruction of the interference effect for particles.

Then looking at this set-up with the interference effect in the theory, we can get from it basically two equally valid but incompatible descriptions of the experiment, A and B.

A is useful for describing where on the screen the particle could be detected but at the cost of not being able to describe which slit the particle could go through.

B is useful for describing which slit the particle could go through but at the cost of not being able to describe where on the screen the particle could be detected.

Obviously, we choose to use A. And that's the standard description in that we just say that if we didn't measure it in some way, then assigning a slit to the particle's journey makes no sense.

B describes the particle at the slits with no detection and it also makes the screen appear as a macroscopic quantum superposition state! It's fairly useless in any practical sense but B exists and is a valid alternative description.

Some alternative descriptions are useful, however, as in EPR they reveal a human prejudice for realism of particle properties which quantum mechanics doesn't have. Knowing that there are other ways of describing the particle properties of EPR other than just the standard one of wave function collapse is very interesting indeed.

The decoherent/consistent histories of Gell-Mann, Griffiths, Hartle and Omnes is the theory I'm describing. Their books and papers will obviously give much better details. Griffiths has part of his book and some questions and answers at his website as well:

http://quantum.phys.cmu.edu/CQT/

 

[ZapperZ]

What I am referring to involves only the usual set-up of the two slit experiment leading solely to the interference effect and not any other set-up which allows for measurement or interaction around the slits. This set-up I mean features no destruction of the interference effect for particles.

Then looking at this set-up with the interference effect in the theory, we can get from it basically two equally valid but incompatible descriptions of the experiment, A and B.

A is useful for describing where on the screen the particle could be detected but at the cost of not being able to describe which slit the particle could go through.

B is useful for describing which slit the particle could go through but at the cost of not being able to describe where on the screen the particle could be detected.

But this is, I'm sorry to say, bogus! You are claiming that by KNOWING which slit the particle goes through, you still did not destroy the interference effects. Can you cite an experiment that has shown this? Because if you can, I would nominate you for the Nobel Prize.

Obviously, we choose to use A. And that's the standard description in that we just say that if we didn't measure it in some way, then assigning a slit to the particle's journey makes no sense.

B describes the particle at the slits with no detection and it also makes the screen appear as a macroscopic quantum superposition state! It's fairly useless in any practical sense but B exists and is a valid alternative description.

Valid by whose standard? Where has this been proven to be valid? Can you show me published experiments that has shown this to be valid?

When I solve an electrostatic problem using an image charge, the solution inside a conductor exists, but it doesn't mean this is a physical solution! Thus, just because an "alternative" exist, doesn't mean it has any connection to reality, especially when there are no experimental verification that shows such a thing exist.

Some alternative descriptions are useful, however, as in EPR they reveal a human prejudice for realism of particle properties which quantum mechanics doesn't have. Knowing that there are other ways of describing the particle properties of EPR other than just the standard one of wave function collapse is very interesting indeed.

As far as ALL the EPR-type experiments have shown, ALL the results have been compatible with what QM has predicted. So how can you point out that these experiments contain things that "quantum mechanics doesn't have"? Point out exactly where in these experiments are indications that there are things that QM doesn't have.

The decoherent/consistent histories of Gell-Mann, Griffiths, Hartle and Omnes is the theory I'm describing. Their books and papers will obviously give much better details. Griffiths has part of his book and some questions and answers at his website as well:

http://quantum.phys.cmu.edu/CQT/

I know about consistent histories. What you described is NOT consistent histories especially with your 2-slit scenario, nor are consistent with what consistent histories are claiming, especially regarding the EPR-type experiments.

Zz.

 

[caribou]

But this is, I'm sorry to say, bogus! You are claiming that by KNOWING which slit the particle goes through, you still did not destroy the interference effects. Can you cite an experiment that has shown this? Because if you can, I would nominate you for the Nobel Prize.

Did I say knowing even once? No. I said describing. A lot.

To make what I mean as clear as I can, if we set up the two-slit experiment in a box completely isolated from the rest of the universe and arranged the experiment to build up an interference pattern over time then the descriptions A and B are alternative but incompatible ways to describe a single run of the experiment with a single particle.

We would not know but we could describe events in ways like A and B.

Valid by whose standard? Where has this been proven to be valid? Can you show me published experiments that has shown this to be valid?

A valid alternative description extracted from the theory. In practice, retaining macroscopic interference effects is obviously not easy due to decoherence but the description B is every bit as valid as Schrodinger's cat being described as being in a superposition.

When I solve an electrostatic problem using an image charge, the solution inside a conductor exists, but it doesn't mean this is a physical solution! Thus, just because an "alternative" exist, doesn't mean it has any connection to reality, especially when there are no experimental verification that shows such a thing exist.

As I understand it, the descriptions like A and B are coarse-grained and are collections of fine-grained descriptions which almost always interfere and don't say anything useful. Descriptions such as A and B are coarse-grained versions used to make sense of the experiment in human terms and to allow the prediction of the probabilities that fine-grained descriptions don't allow.

When we describe a particle as going through a slit or arriving at a point on the screen, we are using human concepts to coursen and make sense of something that at a fine-grained level is hard to make sense of.

As far as ALL the EPR-type experiments have shown, ALL the results have been compatible with what QM has predicted. So how can you point out that these experiments contain things that "quantum mechanics doesn't have"? Point out exactly where in these experiments are indications that there are things that QM doesn't have.

I was talking about the human preference for realism in particle properties as being the flaw and as quantum mechanics not following this human preference, not anything in quantum mechanics being flawed.

I know about consistent histories. What you described is NOT consistent histories especially with your 2-slit scenario, nor are consistent with what consistent histories are claiming, especially regarding the EPR-type experiments.

I've spent some time working to understand the fundamentals of decoherent/consistent histories. There are a lot of technical issues about the theory that I've still to resolve in my mind but I'm fairly sure I understand the fundamentals well enough.

The existence of a vast array of incompatible descriptions of the same events in quantum mechanics is what Gell-Mann has called the "Protean" nature of the theory, after the mythological prophet Proteus who resisted being asked to prophesize and would change shape and had to be held still to be made to predict the future, this being a reference to choosing a single description to make logical predictions in QM.

 

[ZapperZ]

Did I say knowing even once? No. I said describing. A lot.

To make what I mean as clear as I can, if we set up the two-slit experiment in a box completely isolated from the rest of the universe and arranged the experiment to build up an interference pattern over time then the descriptions A and B are alternative but incompatible ways to describe a single run of the experiment with a single particle.

There are already such experiments that collect low intensity photons going through a 2-slit system. What is different or new with what you are asking for?

Secondly, where exactly is the decoherence effects acting on light passing through 2 slits?

Thirdly, how do you expect to detect, or "build up" an "interference pattern over time" without having the intefering agent interact with anything since it is "isolated" from the rest of the universe?

I do not see any "incompatibility" at all. All I see is misinterpretation of what QM is describing. At best, it is the insistence that our classical concept should work in such cases. This is an a priori criteria that has no ab initio proofs that it should be valid.

A valid alternative description extracted from the theory.

What theory? You are "extracting" a description from QM? If QM is all that is needed to describe a phenomena, and you are "extracting" something MORE from it, doesn't this mean (i) you are producing something that has no experimental observation (after all, if there is, QM can be proven to be incomplete), and (ii) that you are doing nothing more than speculating?

In practice, retaining macroscopic interference effects is obviously not easy due to decoherence but the description B is every bit as valid as Schrodinger's cat being described as being in a superposition.

It is EASY. Superconductivity is the clearest manifestation of coherence effects. I've mentioned Carver Mead's paper from PNAS in another thread that clearly indicated such sentiments. So now, apply your "B" to it such as in the Stony Brook and Delft's SQUID experiments.

As I understand it, the descriptions like A and B are coarse-grained and are collections of fine-grained descriptions which almost always interfere and don't say anything useful. Descriptions such as A and B are coarse-grained versions used to make sense of the experiment in human terms and to allow the prediction of the probabilities that fine-grained descriptions don't allow.

When we describe a particle as going through a slit or arriving at a point on the screen, we are using human concepts to coursen and make sense of something that at a fine-grained level is hard to make sense of.

I will not even pretend that I understand what you are saying here. Coarse-grained? Fine-grained?

I was talking about the human preference for realism in particle properties as being the flaw and as quantum mechanics not following this human preference, not anything in quantum mechanics being flawed.

Then what's the problem? If the whole purpose of this is to "pacify" and make QM palatable for human consumption, then this is more appropriate to be in the Philosophy section. Physics has no such demands.

The existence of a vast array of incompatible descriptions of the same events in quantum mechanics is what Gell-Mann has called the "Protean" nature of the theory, after the mythological prophet Proteus who resisted being asked to prophesize and would change shape and had to be held still to be made to predict the future, this being a reference to choosing a single description to make logical predictions in QM.

What "incompatible descriptions"? You just said that QM isn't wrong. In physics, if I have incompatible descriptions of anything, that theory is suspect. Wave and particle descriptions are incompatible descrptions in classical mechanics of the same entity. That's why classical mechanics is suspect when we get into a scale where an entity can exhibit such properties. QM, instead, has NO such distinction, and has only ONE description of BOTH classical wave and particle behavior.

Again, even after reading your "A" and "B" scenario, where are these "incompatible descriptions"?

Zz.

 

[caribou]

I'm going to skip some of your questions as, although I believe I can answer them, I suspect there will be more questions following my answers. Then more questions, then more answers. And so on. If you really want to know, you should go to the source, and that source is the work of Gell-Mann, Griffiths, Hartle and Omnes.

I will not even pretend that I understand what you are saying here. Coarse-grained? Fine-grained?

That's the decoherent histories of Gell-Mann and Hartle which is closely related to the consistent histories of Griffiths and Omnes. As detailed a description as possible (a "fine-grained" description) in quantum mechanics doesn't usually allow for the assignment of probabilities but if we ignore the details, times, etc. that are irrelevant (a "coarse-grained" description), quantum mechanics can give the usual probabilities.

Gell-Mann gives a nice little introduction to the idea in his popular science book, The Quark and the Jaguar. Coarse-graining is just an explanation of what we do to extract useful probabilities from the fundamental physics.

Then what's the problem? If the whole purpose of this is to "pacify" and make QM palatable for human consumption, then this is more appropriate to be in the Philosophy section. Physics has no such demands.

The problem here is that with the EPR paradox, people are thinking the argument is over locality -- when it's actually about realism -- and this makes them think quantum mechanics and special relativity are arguing with each other and that we have faster-than-light effects when we don't.

It's important that people know there is nothing "wrong" with special relativity and there's nothing "wrong" about quantum mechanics, there's just some concepts in quantum mechanics that are not classical and are difficult and abstract to the human mind. It's either this or let people believe in physical effects and theoretical problems which don't exist.

What "incompatible descriptions"? You just said that QM isn't wrong. In physics, if I have incompatible descriptions of anything, that theory is suspect. Wave and particle descriptions are incompatible descrptions in classical mechanics of the same entity. That's why classical mechanics is suspect when we get into a scale where an entity can exhibit such properties. QM, instead, has NO such distinction, and has only ONE description of BOTH classical wave and particle behavior.

Again, even after reading your "A" and "B" scenario, where are these "incompatible descriptions"?

This central idea of "incompatible frameworks" is very difficult and even specialists in interpretation often misunderstand it. Both Gell-Mann and Griffiths have lamented this. Gell-Mann has even accused them of deliberate misunderstanding. I myself even know one person who is convinced Gell-Mann doesn't understand EPR and is talking nonsense.

It's strange, then, that when Gell-Mann and Hartle described their work at a lecture at Caltech, Richard Feynman stood up and said he "agreed with every word" they said. Then again, that's perhaps no surprise as the decoherent histories approach I'm referring to has some of its basis in on-off discussions over decades between Feynman and Gell-Mann. But I guess this person I know who says it's all nonsense must be far smarter than Feynman, Gell-Mann and Hartle, eh?

I certainly hope you are more open-minded than he was, and not for my sake either.

It's conceptually tough. If it was easy then it wouldn't have taken so long to sort out quantum mechanics and need the insight of people like Gell-Mann and Feynman to do it.

Anyhow, on the issue of incompatibility, Griffiths has a paper called "Choice of Consistent Family, and Quantum Incompatibility" that may help:

http://xxx.lanl.gov/abs/quant-ph/9708028

Of course, if you are not interested then fair enough. I'm not keen to continue this discussion as I'm not gaining anything from it, I just thought people might be interested.