Why Does the Measurement Problem Terrify Physicists?

[tommyburgey]

I recently watched a program called 'Atom' on bbc4 that said something along the lines of: "If you want to strike fear into the eyes a physicist; mention the measurement problem" - What is the measurement problem and why is it such a problem?

 

[ZapperZ]

I must not be a physicist then, because the "measurement problem(?)" strikes no fear in my eyes - and I'm an experimentalist too and I measure things for a living!

Zz.

 

[Slaviks]

BBC4 is, of course, exaggerating. What is known as "the measurement problem in quantum physics" is a certain set of open interpretational and philosophical questions regarding the relation between micro- ("quantum") and macro ("clasical") levels of physical theories. It should noted that these open questions do not signal any flaw or inconsistency of modern physical theories, including quantum mechanics.

 

[Demystifier]

There are two kinds of physicists: practical ones and those who want to understand nature on the deepest possible level.
The measurement problem "strikes fear into the eyes" only for the latter.

What that problem actually is? Perhaps not everyone will agree, but I would summarize the problem by the following questions:
Do physical observables have some values even when we do not measure them?
If no, then how exactly measurement makes them?
If yes, then how exactly measurement changes them?

 

[Slaviks]

There are two kinds of physicists: practical ones and those who want to understand nature on the deepest possible level.
The measurement problem "strikes fear into the eyes" only for the latter.

But why fear? I would say "fascination and excitement". Many students of quantum mechanics (including me) get "hooked" precisely by the contraintuitive nature of QM and the sense that it never completely fits into your mind.

 

[ZapperZ]

There are two kinds of physicists: practical ones and those who want to understand nature on the deepest possible level.
The measurement problem "strikes fear into the eyes" only for the latter.

Ah, but you're missing another option here. There could be physicists who are "practical ones" but also interested in understanding nature "on the deepest possible level". The difference being that until there's truly something that can distinguish and answer those questions, everything that is being argued is simply a matter of tastes and would not lead to any kind of agreement.

Physicists like Tony Leggett make testable proposals on the measurement problem rather than simply argue based on tastes. I would consider that to be very "practical".

Zz.

 

[tommyburgey]

So is it philosophical rather than physical?

 

[Slaviks]

It depends. The problem and solutions proposed in the literature are mostly philosophical. However, if a particular mechanism of solution, like Legett's [mentioned above] or Penrose's leads, to experimentally verifiable predictions in the areas not yet tested, then it is a matter of physics.

 

[Demystifier]

The guys like Legett (BTW, can someone provide me an appropriate reference or link?) are not really practical. Instead, they propose experimental tests just because they want to be sure that their "deep understanding of nature" is correct. They probably do not intend to do anything practical with that. In fact, I am also one of such guys:
http://xxx.lanl.gov/abs/quant-ph/0406173
http://xxx.lanl.gov/abs/0705.3542

 

[ZapperZ]

I guess it depends on what you mean by "practical". Leggett is "practical" in my book because (i) he's a condensed matter theorist (ii) he always focuses on experiments and even proposes what to measure (iii) his CM background allows him to come up with CM experiments that have some of the highest degree of certainty of any experiments conducted in physics. I don't associate "practical" with something that has applications.

As for references, he wrote about the quantum measurement problem in an article in Science a few years ago (Science v. 307, p.871 (2005)). But his seminal work on this issue was a few years earlier after the Schrödinger Cat-type experiments by the groups at Stony Brook and Delft that essentially did an experiment he sugested (J. Phys.: Condens. Matt., v.14, p.R415 (2002)).

Zz.

 

[JesseM]

It depends. The problem and solutions proposed in the literature are mostly philosophical. However, if a particular mechanism of solution, like Legett's [mentioned above] or Penrose's leads, to experimentally verifiable predictions in the areas not yet tested, then it is a matter of physics.

Also, my sense is that it would be hard to really develop a theory of quantum cosmology (where the entire universe is treated using quantum rules, with no observer outside to measure it) without at least partially addressing the measurement problem, so it's not purely philosophical in this sense.

 

[Anonym]

What is the measurement problem and why is it such a problem?

I suggest reading “The Quantum Challenge” by G Greenstein and AG Zajonc. It is good and clear presentation of the problem and it current status (2005) almost without math.

Regards, Dany.

 

[lalbatros]

Simply out-dated:

If you want to strike fear into the eyes a physicist; mention the measurement problem

 

[jostpuur]

If you don't know how to drive a car, you cannot go behind a steering wheel and pretend that you know how to drive a car, because everybody can see that you cannot drive a car.

But if you don't know what the quantum mechanical measurement is, you can pretend that you don't have any problems with it, and nothing will force you to admit that you have no idea what it is.

 

[Anonym]

if you don't know what the quantum mechanical measurement is, you can pretend that you don't have any problems with it, and nothing will force you to admit that you have no idea what it is.

Please, define what the quantum mechanical measurement is (the definition that you use).

Regards, Dany.

 

[Anonym]

What that problem actually is? Perhaps not everyone will agree, but I would summarize the problem by the following questions:
Do physical observables have some values even when we do not measure them?
If no, then how exactly measurement makes them?
If yes, then how exactly measurement changes them?

You have no idea what the measurement problem is. It is clearly defined during last 80 years. You confuse OP and violate PF guidelines:

Specifically, we aren't here to defend mainstream science, we are here to explain it.

I will therefore ask people … to please refrain from posting statements of their opinions unless they can quote a peer reviewed published paper, textbook, or other source which supports this as a current subject of scientific debate in accordance with our guidelines.

Regards, Dany.

 

[jostpuur]

Please, define what the quantum mechanical measurement is (the definition that you use).

Regards, Dany.

I was talking about the process, that is mathematically described as the state being projected onto some eigenstates. This is what quantum mechanical measurement usually means.

It is mathematically clear, but what is it physically? When does the projection occur? What initiates the projection?

You have no idea what the measurement problem is. It is clearly defined during last 80 years. You confuse OP and violate PF guidelines:

We are violating only the philosophy "shut up and calculate", but that yet is not a big crime.

 

[Anonym]

I was talking about the process, that is mathematically described as the state being projected onto some eigenstates. This is what quantum mechanical measurement usually means.

No. It is not a measurement. You are talking about the universally valid phenomenon called the collapse of the wave packet.

It is mathematically clear, but what is it physically?

Let us talk physics. Your definitions still not clear to me. What is the physics the measurement apparatus obey? Is it the classical physics only (macroscopic)? If so (CI), then why you call it the quantum mechanical measurement? If your physical system under test is macroscopic (obey laws of the classical physics), you will find that every dynamical variable is measurable (observable); if your physical system under test is microscopic (obey laws of the quantum physics), then only the dynamical variables that form the mutually commuting set of self-adjoint operators will be observable. But the measurement set-ups are the same. Clearly for me that you perform the same measurements, only the results are different. Therefore, it is the demonstration that the quantum world is not the classical world. In addition, the above type of measurements is accompanied by the collapse of the wave packet.

R.Penrose calls it the “R-process” to make distinction from the unitary evolution of the QM system (“U-process”). But it is not a process, there is no dynamics that associated with it and it is not needed. The act of the measurement is instant.

If you agree with me, please redefine the quantum mechanical measurement.

When does the projection occur?

At the same instant when you perform the measurement. It is the transition from the Quantum World to the Classical World. Symbolically we call it the E. Schrödinger Cat.

What initiates the projection?

Interaction with the measurement apparatus.

Regards, Dany.

 

[jostpuur]

At the same instant when you perform the measurement.

If a Gaussian wave packet, that is badly delocalized, hits a wall, and collapses into a smaller area, when does the collapse occur? This is a measurement of the position. When does the measurement occur?

I can be more specific. Suppose we can approximate some wave packet solution in one dimension as

$$\psi(t,x) = e^{-(x-vt)^2}$$

Okey, that's not a real solution of SE, but it doesn't matter. The peak of the wave function is at location $x=vt$ at any given instant t. Now, if there is a wall at location L, can you tell me when this "wave packet hits that wall"? An answer "at time $t=L/v$" is not valid, because that is only the instant when the center of the packet hits the wall. An answer "when the packet first touches the wall" isn't valid either, because it is touching it all the time.

So, when does the wave packet hit the wall? When does the measurement of the position occur?

 

[Demystifier]

You have no idea what the measurement problem is. It is clearly defined during last 80 years. You confuse OP and violate PF guidelines:

Be careful before such statements. I have published several papers on this issue in peer reviewed physics journals. For one of these journals the editor is the Nobel-prize winner G. 't Hooft (who, by the way, also explores the measurement problem in QM). How about you?

For people like you who think that all conceptual problems of QM are already solved a long time ago, I have written a review:
http://xxx.lanl.gov/abs/quant-ph/0609163
And yes, it is accepted for publication in a peer reviewed journal (for which the editor is the mentioned Nobel-prize winner).

 

[Demystifier]

If you don't know how to drive a car, you cannot go behind a steering wheel and pretend that you know how to drive a car, because everybody can see that you cannot drive a car.

But if you don't know what the quantum mechanical measurement is, you can pretend that you don't have any problems with it, and nothing will force you to admit that you have no idea what it is.

Brilliant!
Is that written by you? Do I have a permission to cite it somewhere?

 

[tommyburgey]

but I also heard that some wackos actually think that measurement and discovery can change past events? Also I don't understand how Schrödinger's cat can be anything other than pure philosophy as it doesn't change anything physically it's just our understanding of the events.

 

[jostpuur]

Brilliant!
Is that written by you? Do I have a permission to cite it somewhere?

hmhm.. yeah it's mine. I used to think this way in programming, one older programmer emphasized this to me. In social life you can always accuse others when things don't work, and nothing forces you to see your own mistake, but in programming your own program, that doesn't work, forces you to see your mistake mercilessly. And watch how the justice works here: If you refuse to find your own mistake in the code, you cannot get the program working anymore!

My comment about wave function collapse was part a more general philosophy. There is usually two kinds of "not understanding". Kind of, where nothing forces you to see the problem, and kind of where it is impossible to keep pretending.

Citing it... hmhmh... I didn't think it's that good myself :/

 

[jostpuur]

but I also heard that some wackos actually think that measurement and discovery can change past events?

I remember hearing this too. It sounds like misunderstanding of quantum mechanics. Perhaps measurements can change past events in some very exotic interpretation, but its not part of the popular Copenhagenian interpretation at least.

Also I don't understand how Schrödinger's cat can be anything other than pure philosophy as it doesn't change anything physically it's just our understanding of the events.

I'm not sure what you are saying here, but the Shrodinger's cat paradox is a real paradox. You'll have to understand QM before it. What is paradoxical is how macroscopic and living beings could be in superposition of very different states.

 

[Anonym]

Be careful before such statements.

The OP question is: What is the measurement problem and why is it such a problem? There are several sessions here at PF related to that question but at least during last year nobody suggest the discussion of the roots and the present status of the problem which is now under intensive development.

In addition, I consider CI out-dated. First of all, I try through discussion with Jostpuur to understand how the problem is defined today. Later, what was done and what remains open. Knowledge at least of W&Z “Quantum Theory and Measurement” is required to contribute to our discussion.

Your post #4 is related to the M. Born statistical interpretation of QM and has nothing to do with the Measurement Problem. As to your warning to be careful, it may be considered by OP, that we don’t fear anything: you have no idea what the measurement problem is. Try to focus on the specific problem or just open the relevant session.

For people like you who think that all conceptual problems of QM are already solved a long time ago.

I have no idea what you are talking about.

Regards, Dany.

 

[genneth]

For a fairly modern attack on the problem, see the work of Christopher Fuchs, esp. http://arxiv.org/abs/quant-ph/0205039

If nothing else, the style is invigorating.

 

[Anonym]

For a fairly modern attack on the problem, see the work of Christopher Fuchs, esp. http://arxiv.org/abs/quant-ph/0205039

Thank you. I missed it and will study. It looks a good review of the activity. However, the paper apparently is not a physical. There are no refs on E. Schrödinger, W. Heisenberg, P.A.M.Dirac and E.P.Wigner. It is not possible to formulate the physical theory of measurements without well defined notion of the objective reality and I don’t know how it may be done without language of the theory of the continuous groups (H.Weyl, E.P.Wigner, C.N.Yang, etc.). It is why EPR paper is wrong.

Regards, Dany.

 

[Anonym]

If a Gaussian wave packet, that is badly delocalized, hits a wall, and collapses into a smaller area, when does the collapse occur? This is a measurement of the position. When does the measurement occur?

Let me don’t answer your question now. We will discuss it later. You are focused on the details of the collapse. But the collapse is only how the measurement is actually take place. I would like to continue the discussion you initially started: what one would like to know (measurement) and how it may be done.

In the classical physics I know that the third Newton law governs the measurement procedure. I claim that as far as CI is valid the procedure will be the same. The physics of the procedure should be such that: 1) the special relativity requirements must be satisfied; 2) in case when the system under test is QM, the discontinuous transition from delta(x)*delta (p)>0 (for example) to delta(x)*delta (p) =0 should take place.

However, CI statements that the measurement apparatus must be macroscopic seem to me obsolete. No doubt that it is part of the story but not all the story. For example, in AB experiments we measure the EM potential which is unobservable classically. Moreover, the quantum computation may be considered as a measurement and I may store the gathered data without direct communication with the classical world. Later using our good friend photon that seems does not know anything about the boundary between the quantum world and the classical world I will communicate the data to the end user. Don’t interpret what I said as the statement that there is no collapse in ED. I don’t forget about the retarded solutions and the statmech.

But let us return to the definitions of what the measurement is. In the classical life it is clear for me why I should know where the moon or sun is located. But in the quantum system the question where the electron is located in the hydrogen ground state is the stupid question and the Mother Nature answer: don’t ask stupid questions.

If you agree with me, please redefine the quantum mechanical measurement.

Regards, Dany.

 

[tommyburgey]

I think I'm missing something here concerning Schrödinger's cat.
At some moment in time the cat will die. If one person was watching the cat with a camera inside the box then it wouldn't be a paradox for him as he'd see the cat die. But if someone did not see inside the box - it would be a paradox for him. Doesn't this mean that Schrödinger's cat is just concerning human perception and not an actual paradox such as the train paradox ?
I do now understand the measurement problem slightly better in the way that by viewing an object photons have had to hit it (moving it slightly) or could hit an electron and change the atom's physical state but Schrödinger's cat confuses me.

 

[jostpuur]

I think I'm missing something here concerning Schrödinger's cat.
At some moment in time the cat will die. If one person was watching the cat with a camera inside the box then it wouldn't be a paradox for him as he'd see the cat die. But if someone did not see inside the box - it would be a paradox for him. Doesn't this mean that Schrödinger's cat is just concerning human perception and not an actual paradox such as the train paradox ?
I do now understand the measurement problem slightly better in the way that by viewing an object photons have had to hit it (moving it slightly) or could hit an electron and change the atom's physical state but Schrödinger's cat confuses me.

I remember, that when I first read about Shrodinger's paradox, I thought that the release of the poison into a box is a random event, but it either happens or doesn't happen in some time period, and when it happens the cat dies. The fact that a human outside doesn't know what's happening inside brings no paradox.

Is this what you are thinking?