Insights How I Stopped Worrying and Learned to Love Orthodox Quantum Mechanics - Comments

[vanhees71]

Not "collapse" - state reduction is fine - in fact state reduction is often synonymous with "collapse". Only some people misunderstand Copenhagen and believe that "collapse" is necessarily physical.

The flaw of Peres is that he fails to state the classical-quantum cut clearly. I believe he also does not include state reduction in his axioms.

In the last chapter he very clearly discusses measurements on the example of the Stern-Gerlach experiment (the "Drosophila" of quantum physicists ;-)), and it becomes very clear that his view on the "classicality of measurement apparati" to ensure an irreversible storage of the measurement result is seen in the sense of an emergent phenomenon through the usual coarse-graining argument of quantum statistics (he calls it "blurring").

As I said, for me this book has been a relief be cause it cleans up the QT-foundational discussion of all unnecessary philosophical complications.

 

[atyy]

In the last chapter he very clearly discusses measurements on the example of the Stern-Gerlach experiment (the "Drosophila" of quantum physicists ;-)), and it becomes very clear that his view on the "classicality of measurement apparati" to ensure an irreversible storage of the measurement result is seen in the sense of an emergent phenomenon through the usual coarse-graining argument of quantum statistics (he calls it "blurring").

As I said, for me this book has been a relief be cause it cleans up the QT-foundational discussion of all unnecessary philosophical complications.

Yes, that blurring is exactly where Peres reveals his mistaken sympathies with Ballentine. It is simply wrong.

 

[Demystifier]

The flaw of Peres is that he fails to state the classical-quantum cut clearly. I believe he also does not include state reduction in his axioms.

Why would he talk about classical-quantum cut clearly if he doesn't think that there is such a cut?

Concerning the axioms, he is developing a practical instrumental approach, not an axiomatic approach. This is like complaining that a handbook of civil engineering does not state axioms of stable building construction.

 

[atyy]

I wouldn't say that Peres accepts a classical-quantum cut. What he accepts is something more like (abstract formalism)-(laboratory phenomena) cut. He says that quantum phenomena do not occur in Hilbert space, but in a laboratory. It would be akin to a statement that classical phenomena do not occur in phase space, but in a laboratory.

The book is very good, because Peres for the most part accepts the classical quantum cut. But he cannot bring himself to articulate it explicitly. He accepts it implicitly when he talks about the need for a second classical measuring apparatus, if we treat one measuring apparatus as quantum. A book about foundations should be as explicit about axioms as possible, not hide the ones that he doesn't like in implicit statements. And because he dislikes the axiom, eventually he does make a misleading statement about blurring of the Wigner function. There may be a way to read it without being misled, but as you can see from vanhees71's post #31, Peres has managed to mislead an expert.

 

[Demystifier]

The book is very good, because Peres for the most part accepts the classical quantum cut. But he cannot bring himself to articulate it explicitly. He accepts it implicitly when he talks about the need for a second classical measuring apparatus, if we treat one measuring apparatus as quantum. A book about foundations should be as explicit about axioms as possible, not hide the ones that he doesn't like in implicit statements. And because he dislikes the axiom, eventually he does make a misleading statement about blurring of the Wigner function. There may be a way to read it without being misled, but as you can see from vanhees71's post, Peres has managed to mislead an expert.

Well, Peres's and Ballentine's books have something in common. If you read the fine details of the books, you can find erroneous statements. But if you make an appropriate blurring of the books, the fine errors become invisible and the books as a whole become great.

 

[atyy]

Well, Peres's and Ballentine's books have something in common. If you read the fine details of the books, you can find erroneous statements. But if you make an appropriate blurring of the books, the fine errors cancel out and the books as a whole become great.

The problem is that the errors are fundamental, not incidental.

However, I do love Peres's book, although I dislike Ballentine's. In other words, I forgive Peres because he is so charming, at least in writing :)

 

[Demystifier]

The problem is that the errors are fundamental, not incidental.

Yes, but they are not visible after the blurring.

However, I do love Peres's book, although I dislike Ballentine's. In other words, I forgive Peres because he is so charming, at least in writing :)

I find Ballentine charming too.

 

[atyy]

I find Ballentine charming too.

But in a way, Ballentine is anti-Bohmian. If Ballentine were correct, there is no measurement problem, and Bohmian mechanics is pointless.

 

[martinbn]

The problem is that the errors are fundamental, not incidental.

You say errors, but I have the feeling that you mean opinions different than yours. Can you give examples?

However, I do love Peres's book, although I dislike Ballentine's. In other words, I forgive Peres because he is so charming, at least in writing :)

Have you read the whole book of Ballentine? May be you only dislike some parts.

 

[Demystifier]

But in a way, Ballentine is anti-Bohmian. If Ballentine were correct, there is no measurement problem, and Bohmian mechanics is pointless.

I strongly disagree, see Secs. 14.2 and 14.3. of his book. I would rather say that he is agnostic about Bohmian QM. He does not say that there is no measurement problem, but demonstrates that a lot can be understood without talking about it explicitly.

 

[atyy]

I strongly disagree, see Secs. 14.2 and 14.3. of his book. I would rather say that he is agnostic about Bohmian QM. He does not say that there is no measurement problem, but demonstrates that a lot can be understood without talking about it explicitly.

Yes, Ballentine himself is slightly sympathetic to Bohmian mechanics. But if the main point of Ballentine's book is correct, then Bohmian mechanics is pointless - vanhees71 has drawn the logical conclusion from Ballentine.

Thus in fact, it is Copenhagen - which Ballentine hates - that promotes Bohmian mechanics. As your Insights article explains, there is no reason for a Bohmian not to love Copenhagen.

 

[Demystifier]

Yes, Ballentine himself is slightly sympathetic to Bohmian mechanics. But if the main point of Ballentine's book is correct, then Bohmian mechanics is pointless - vanhees71 has drawn the logical conclusion from Ballentine.

Thus in fact, it is Copenhagen - which Ballentine hates - that promotes Bohmian mechanics. As your Insights article explains, there is no reason for a Bohmian not to love Copenhagen.

You really like to push your conclusions to the extreme.
Your style reminds me of the great philosopher of science, Feyerabend.

 

[bahamagreen]

I still don't understand what do you mean by "one particle". That the whole universe contains only one particle? That's excluded.
Concerning the number of particle types, I cannot exclude any possibility.

If one type of fundamental particle were manifest as all various types of particles/quasi-particles,
what prevents associating one particle only of only one type to all the world lines in space time?

 

[Greg Bernhardt]

Interesting thoughts Demystifier, thanks!

 

[PeterDonis]

The default hypothesis is that it persists, but a hypothesis that it doesn't is also legitimate

Isn't it known to be possible for Lorentz invariance to emerge at large distance scales from a quantum field theory that is non-relativistic on small distance scales? IIRC (I think I first came across this in Zee's QFT textbook), the Lorentz invariant speed is something like a "sound speed" in an underlying medium that emerges from the non-relativistic QFT.

 

[bluecap]

vanhees71 does not accept the classical-quantum cut, which even Peres does.

Peres is an excellent book, but it is not quantum orthodoxy. Although he hides it very well, ultimately his flawed sympathy with Ballentine shows itself in his lack of a clear statement of the measurement problem, and statements about fuzzy Wigner functions that try to avoid the measurement problem. Basically, unless a book about foundations talks about the measurement problem, it is useless as a book about foundations. The measurement problem is the most important problem in the foundations of quantum mechanics.

When we treat the wave function as objective.. we have problems with the consequence of real collapse (a varient of Copenhagen) or all the terms existing (MWI) or all terms existing but they are not real and only one particle being pushed around (Bohmian).. all of these are greatly disturbing. So why don't we just treat the wave function as just for calculation purposes and all of us happy.. I mean.. does this eliminate the classical-quantum cut by making all classical.. how does treating the wave function as calculational tool only affect the need or requirement of the classical-quantum cut?

 

[atyy]

When we treat the wave function as objective.. we have problems with the consequence of real collapse (a varient of Copenhagen) or all the terms existing (MWI) or all terms existing but they are not real and only one particle being pushed around (Bohmian).. all of these are greatly disturbing. So why don't we just treat the wave function as just for calculation purposes and all of us happy.. I mean.. does this eliminate the classical-quantum cut by making all classical.. how does treating the wave function as calculational tool only affect the need or requirement of the classical-quantum cut?

In all versions of Copenhagen, the classical/quantum cut does exists whether the wave function is real or calculational. If the wave function is not real, but the measurement apparatus and the results are real, quantum mechanics needs a reality/non-reality cut. The unitary evolution of the non-real wave function does not say what is measured, nor when a measurement occurs - it does not say when reality pops out of non-reality. There is a cut somewhere - different people call it by different names.

 

[bluecap]

In all versions of Copenhagen, the classical/quantum cut does exists whether the wave function is real or calculational. If the wave function is not real, but the measurement apparatus and the results are real, quantum mechanics needs a reality/non-reality cut. The unitary evolution of the non-real wave function does not say what is measured, nor when a measurement occurs - it does not say when reality pops out of non-reality. There is a cut somewhere - different people call it by different names.

If the wave function is not real, but the measurement apparatus and the results are real, why do quantum mechanics still need a reality/non-reality cut.. (by reality you mean classical and nonreality quantum? why didn't you use the word classica/quantum and instead use reality/non-reality?) Can you please give an example why the cut is still needed. We can treat all as classical, and the quantum result only a tool to produce probability...

 

[bluecap]

If the wave function is not real, but the measurement apparatus and the results are real, why do quantum mechanics still need a reality/non-reality cut.. (by reality you mean classical and nonreality quantum? why didn't you use the word classica/quantum and instead use reality/non-reality?) Can you please give an example why the cut is still needed. We can treat all as classical, and the quantum result only a tool to produce probability...

To elaborate my concern. I was hoping that making the wave function as simply calculation tool would make all classical.. without needing any quantum cut.. but in the double slit experiment, treating the wave function as simply calculation tool means we only deal with the output or the detector screen.. we ignore what happens between emission and detection.. is this what you mean or a good example.. and the reason we still need the cut is because we need to know what happens inbetween.. or how reality comes from non-reality?

It's really a headache to treat the wave function as objective.. I have headaches for many days lol.. so treating it as calculational tool can give us relief.. but the cost is unable to determine the reality/non-reality cut in the sense we don't know what happens between emission and detection? Is this the exact reason so I know.. thanks..

 

[Nugatory]

but the cost is being unable to determine the reality/non-reality cut in the sense we don't know what happens between emission and detection?

Yes, you're pretty much there. Now, if you could just manage to be happy about paying that cost...

 

[atyy]

To elaborate my concern. I was hoping that making the wave function as simply calculation tool would make all classical.. without needing any quantum cut.. but in the double slit experiment, treating the wave function as simply calculation tool means we only deal with the output or the detector screen.. we ignore what happens between emission and detection.. is this what you mean or a good example.. and the reason we still need the cut is because we need to know what happens inbetween.. or how reality comes from non-reality?

It's really a headache to treat the wave function as objective.. I have headaches for many days lol.. so treating it as calculational tool can give us relief.. but the cost is unable to determine the reality/non-reality cut in the sense we don't know what happens between emission and detection? Is this the exact reason so I know.. thanks..

Yes, you can try to make everything classical, eg. Bohmian mechanics or MWI are famous approaches to making quantum mechanics classical. But if you only make everything notionally classical, without equations (ie. without Bohmian Mechanics or MWI), you have no way to interact with quantum mechanics and pull quantitative predictsion about reality (or the "classical" or "macroscopic" world). Basically, quantum mechanics does not tell us when a measurement occurs, and we need that input to pull a quantitative prediction out.

 

[martinbn]

What is the problem with having a classical/quantum cut?

 

[atyy]

What is the problem with having a classical/quantum cut?

It has been stated in various equivalent ways.

- It means that classical mechanics cannot be derived from quantum mechanics without assuming classical mechanics. Thus quantum mechanics is not the most fundamental theory.

- If there is a classical/quantum cut, that means that means that quantum mechanics cannot describe the observer himself. If we extend the quantum boundary to include the observer, we need yet a second observer to observe the first observer, so there is always some bit of the universe that is excluded from quantum mechanics, again suggesting that quantum mechanics is incomplete.

It is not a problem of internal coherence of the theory itself. Rather, from how we understand theories of physics to describe reality, it appears that quantum mechanics is incomplete, even if no experimental evidence contradicts it. Thus there is a theoretical opportunity to devise more complete theories (eg. BM) or alter our understanding of reality (eg. MWI).

 

[vanhees71]

Yes, that blurring is exactly where Peres reveals his mistaken sympathies with Ballentine. It is simply wrong.

It's only your lonely opinion that the minimal interpretation is wrong. Instead, it's the interpretation nearly all physicists use in the labs and theory institutes around the world.

 

[martinbn]

- It means that classical mechanics cannot be derived from quantum mechanics without assuming classical mechanics. Thus quantum mechanics is not the most fundamental theory.

That by itself is not a problem. No matter what people prefer, this may be how nature is.

- If there is a classical/quantum cut, that means that means that quantum mechanics cannot describe the observer himself. If we extend the quantum boundary to include the observer, we need yet a second observer to observe the first observer, so there is always some bit of the universe that is excluded from quantum mechanics, again suggesting that quantum mechanics is incomplete.

It only means that in any description there will be classical mechanics involved. Why is that a problem?

It is not a problem of internal coherence of the theory itself. Rather, from how we understand theories of physics to describe reality, it appears that quantum mechanics is incomplete, even if no experimental evidence contradicts it. Thus there is a theoretical opportunity to devise more complete theories (eg. BM) or alter our understanding of reality (eg. MWI).

I don't see why it is incomplete? Any experiment can in principle be described. It may not be the way some people want it but that is irrelevant.

 

[vanhees71]

First of all QT is incomplete, because there's no satisfactory quantum description of gravity, but that's not the issue.

The state of affairs, however, nowadays is that where QT is applicable, it is more comprehensive than classical physics, whose validity is limited by empirical facts where QT effects become important. In the standard way of thinking in the physics community thus quantum theory is the more fundamental theory, and classical behavior of macroscopic systems is considered as an "emergent phenomenon", i.e., it can be understood as an effective description of macroscopic observables, which are defined by coarse-graining microscopic observables in the sense of quantum statistical mechanics. As far as I know, there's no evidence that this point of view is wrong.

 

[Demystifier]

If one type of fundamental particle were manifest as all various types of particles/quasi-particles,
what prevents associating one particle only of only one type to all the world lines in space time?

I still have no idea what are you talking about. How can one particle be associated with many world lines?

To other readers, does somebody else understand the question?

 

[Demystifier]

Isn't it known to be possible for Lorentz invariance to emerge at large distance scales from a quantum field theory that is non-relativistic on small distance scales? IIRC (I think I first came across this in Zee's QFT textbook), the Lorentz invariant speed is something like a "sound speed" in an underlying medium that emerges from the non-relativistic QFT.

Yes, exactly.

 

[Demystifier]

What is the problem with having a classical/quantum cut?

Let me use an analogy from pure mathematics. Consider naive set theory, which is known to suffer from certain paradoxes. There is a small group of mathematicians called finitists who resolve the problems by postulating that infinity doesn't exist. At first sight that looks reasonable, for no one ever seen infinity, nor ever will. However, such an approach requires a cut, namely the existence of the largest natural number. What that number is? Nobody has a clue. Whatever specific number you choose (e.g. Graham number or whatever), a smart combinatorist may find a good practical reason for existence of an even larger number.

The situation with classical/quantum cut is similar. If there is a cut, you must say where exactly the cut is. But wherever you place the cut, a smart physicist may find compelling experimental evidence that the cut cannot be there. For instance, if you say that the cut is 100 atoms, there are situations in which molecules with more than 100 atoms behave quantum mechanically. If you say that cut is 1000 atoms, there are situations in which molecules with less than 1000 atoms behave classically. Perhaps the cut is not defined by the number of atoms but by something else, but then what that else is? The amount of complexity? The existence of consciousness (something similar was proposed by mathematician von Neumann)? There are various ideas, but neither looks very convincing.

 

[martinbn]

But that by itself is not a problem, at least not in my eyes. If we go back to the analogy and we decide to use only finite sets, that doesn't mean that there will be a universal bound for the number of elements. It only means that for every problem you want to consider you will use only finite sets. For the next question you may need a bigger set, no problem as long as it is finite.