What does it take to solve the measurement problem? (new paper published)

[Madeleine Birchfield]

TL;DR Summary

preprint by Jonte Hance and Sabine Hossenfelder on the measurement problem and quantum foundations

Yesterday Jonte Hance and Sabine Hossenfelder published this preprint on the arXiv: https://arxiv.org/abs/2206.10445

What does it take to solve the measurement problem?​

We summarise different aspects of the measurement problem in quantum mechanics. We argue that it is a real problem which requires a solution, and identify the properties a theory needs to solve the problem. We show that no current interpretation of quantum mechanics solves the problem, and that, being interpretations rather than extensions of quantum mechanics, they cannot solve it. Finally, we speculate what a solution of the measurement problem might be good for.

 

[Morbert]

"We show that no current interpretation of quantum mechanics solves the problem"

They don't even come close to demonstrating this imo. They would need to carry out a much longer literature review of foundations in QM to establish this.

I'm also not too impressed with the paper's novelty. It's reiterating views the authors have expressed before.

 

[Demystifier]

I've sent an e-mail to the authors and now we are having a lively discussion.

 

[Morbert]

I've sent an e-mail to the authors and now we are having a lively discussion.

I know it's a private correspondence but can you broadly say what the gist of your critiques/their responses are?

 

[martinbn]

I don't see what the paper adds! Except, of course, that it adds one to their paper count.

 

[vanhees71]

I wonder, where this has a chance to be published...

 

[gentzen]

I wonder, where this has a chance to be published...

One day later, N. David Mermin published a related preprint on the arXiv: https://arxiv.org/abs/2206.10741

A note on the quantum measurement problem​

The idea that wave-function collapse is a physical process stems from a misunderstanding of probability and the role it plays in quantum mechanics.

That one says:

A slightly revised version of what appeared as a “Quick Study” in Physics Today, June 2022, 63-64.

So if the preprint by Hance&Hossenfelder were shorter, it could be published in a similar place.

 

[Demystifier]

I wonder, where this has a chance to be published...

Foundations of physics?

 

[Demystifier]

I know it's a private correspondence but can you broadly say what the gist of your critiques/their responses are?

Well, I objected against their claims that Bohmian mechanics does not help to solve the measurement problem. Let me not reveal their responses.

 

[Fra]

"We show that no current interpretation of quantum mechanics solves the problem"

They don't even come close to demonstrating this imo.

I also don't see the argument against the psi-epistemic resolution as convincing.

One problem of many psi-epistemic advocates though is that I think they rarely mention what I see as the obvious duality in the views. Ie. the agent itself must be ontic, but the psi-epistemic wavefunction is encoded in the ontic agent, and the agents must of course be normal physical systems(what else?), that interact according to physical laws. So it's both ontic and epistemic, but the ontic part lies in the agent. But of course the agent does not "describe itself", this is why the ontics of the agent, from its own perspective "just is". So there is all along a sort of holographic duality here. The agent encodes en illusion of reality. The the illusion itself makes up part of the reality for other observers.

A good examele is Mermin in this paper which gentzen quoted. While I agree with the abstrac of Mermins paper, his arguments reveals that he is not taking the duality into account, making qbism look simplistic. I do not subscribe to that "minimal version" of qbism. I think we can do more with the epistemic view that this.

"Just what happens with space-time curvaturewhen a photon passes through a beam splitter? Saying that the wave function is epistemic doesn’t answer these questions."
-- Sabine, https://arxiv.org/abs/2206.10445

I think to have a chance to explain this, in the epistemic view, we need a reconstruction of the theory in the radical qbist direction, where "qbist-agents" can interact. I just fond it ironic that this is mentioned as a (-) for the epistemic view. I see it the oppositve way. But the epistemic view is also apparently very heterogenous.

/Fredrik

 

[Quantumental]

Their 'treatise' on Many Worlds also did not make much sense to me.

 

[Morbert]

@Fra My position is epistemic interpretations fare better when the knowledge is not "knowledge of the ontic state of the system". We can say $\rho_0(t_0)$ is knowledge of the likelihood of a sequence of measurement outcomes to happen, if the appropriate measurements are made after $t_0$. The probability for sequence $\alpha$ is $$p(\alpha) = \mathrm{Tr}\left[C_\alpha\rho_0(t_0)C_\alpha^\dagger\right]$$
and $\{\rho_0(t_0), \rho_f(t_f)\}$ is knowledge of a sequence of measurement outcomes to have happened if the appropriate measurements were already made between $t_0$ and $t_f$, with probability given by $$p(\alpha) = \frac{\mathrm{Tr}\left[C_\alpha\rho_0(t_0)C^\dagger_\alpha\rho_f(t_f)\right]}{\mathrm{Tr}\left[\rho_0(t_0)\rho_f(t_f)\right]}$$An epistemic interpretation of states as knowledge of measurement outcomes is safe. I'd also go one further and say consistent histories can generalise this to an epistemic interpretation of states as knowledge of events that occur, whether or not we want to understand them as measurement outcomes.

"Just what happens with space-time curvaturewhen a photon passes through a beam splitter? Saying that the wave function is epistemic doesn’t answer these questions."
-- Sabine, https://arxiv.org/abs/2206.10445

I'm not sure I understand her point here. If we are considering a quantum theory of spacetime, our theory would presumably be quite general. We could e.g. write down a quantum state that returns probabilities for the alternative metrics that correspond to the photon going through the alternative paths a la ordinary QM.

 

[Fra]

@Fra My position is epistemic interpretations fare better when the knowledge is not "knowledge of the ontic state of the system". We can say $\rho_0(t_0)$ is knowledge of the likelihood of a sequence of measurement outcomes to happen, if the appropriate measurements are made after $t_0$. The probability for sequence $\alpha$ is $$p(\alpha) = \mathrm{Tr}\left[C_\alpha\rho_0(t_0)C_\alpha^\dagger\right]$$
...

I share this perspective. The "expectations" of the agent in my view, is best understood as its own conditional probability of the future state, relative to the lossy retained past information. This makes the observers own dynamics a bit like reflection upon itself. This follows a some information preserving evolution - until new information arrives.

An epistemic interpretation of states as knowledge of measurement outcomes is safe. I'd also go one further and say consistent histories can generalise this to an epistemic interpretation of states as knowledge of events that occur, whether or not we want to understand them as measurement outcomes.

I have tried to look at consistent histories, and from what I remember it was not quite in line how I prefer it.

I'm not sure I understand her point here.

She seems to have the opinion the epistemic vs ontic perspective adds not real solutions to the open questions. But I disagree, I see potential here. But I concluded from other things Sabine wrote that she thinks in a different direction.

/Fredrik

 

[Demystifier]

I've sent an e-mail to the authors and now we are having a lively discussion.

In the revised version they mentioned my in Acknowledgements, but didn't change anything related to the stuff we discussed.

 

[gentzen]

Their 'treatise' on Many Worlds also did not make much sense to me.

I disagree with her conclusions, but their 'treatise' on Bohmian Mechanics feels worse to me:

Bohmian Mechanics is usually formulated in position space, but one can easily extend this definition just by requiring it to respect invariance under basis transformations.

It basically combines the Schrödinger evolution and the Collapse Postulate to one local evolution for the particle and a non-local one for the guiding field.

Maybe one could try to defend those statements, but they don't even admit that those statements need to be defended. They refuse to give any relevant references to justify them.

 

[A. Neumaier]

Summary: preprint by Jonte Hance and Sabine Hossenfelder on the measurement problem and quantum foundations

Yesterday Jonte Hance and Sabine Hossenfelder published this preprint on the arXiv: https://arxiv.org/abs/2206.10445

What does it take to solve the measurement problem?​

See also one of Sabine Hossenfelder's blog entries:
https://backreaction.blogspot.com/2022/05/chaos-real-problem-with-quantum.html

 

[Fra]

See also one of Sabine Hossenfelder's blog entries:
https://backreaction.blogspot.com/2022/05/chaos-real-problem-with-quantum.html

Hmm, Sabine writes on the blog:

"But quantum mechanics is supposed to be a theory of all matter. So we should be able to use quantum mechanics to describe large objects, right? "

I agree that the a good general theory of meaurement should be able to describe large objects. But i think quantum mechanics as it stands is not really designed todo that as it is designed for small subsystems. But yes this is a "problem".

The larger ambition i think requires a modification, not just pure reinterpretations. If that is the main point i fully agree. I just think some epistemological views suggest a reasearch direction. I would agree that a "pure epistemological interpretation" that has no ambition to modify QM also will not solve any problems.

But not sure what she really means. Perhaps i got her emphasis wrong.

/Fredrik

 

[Demystifier]

See also one of Sabine Hossenfelder's blog entries:
https://backreaction.blogspot.com/2022/05/chaos-real-problem-with-quantum.html

My problem with this argument is the following: What knew do we learn from the Hyperion problem that we didn't already know from the Schrodinger cat problem?

 

[A. Neumaier]

But i think quantum mechanics as it stands is not really designed todo that as it is designed for small subsystems

This is not true. For example, quantum mechanics is needed for the correct description of the thermal properties of bulk matte (metals, ice) at very low temperature.

 

[Fra]

This is not true. For example, quantum mechanics is needed for the correct description of the thermal properties of bulk matte (metals, ice) at very low temperature.

Yes but the mechanism at play in bulk matter (chemistry) is still "small" and "fast" in comparasion to anything that contains signifiicant gravity.

Even with your view of the tomographics process of the classical environment the missing piece to make a measuremnt a physical interaction is to understand how the tomographic information is focused in the agent making the meaurement so we can understand the causal backreaction. In QM the observer ia never an real active player, it just passively prepares observes ans records with dominance. As long as this is the case i don't see how we can unify inference rulea with physical law?

/Fredrik

 

[vanhees71]

Hmm, Sabine writes on the blog:

"But quantum mechanics is supposed to be a theory of all matter. So we should be able to use quantum mechanics to describe large objects, right? "

I agree that the a good general theory of meaurement should be able to describe large objects. But i think quantum mechanics as it stands is not really designed todo that as it is designed for small subsystems. But yes this is a "problem".

The larger ambition i think requires a modification, not just pure reinterpretations. If that is the main point i fully agree. I just think some epistemological views suggest a reasearch direction. I would agree that a "pure epistemological interpretation" that has no ambition to modify QM also will not solve any problems.

But not sure what she really means. Perhaps i got her emphasis wrong.

/Fredrik

But quantum mechanics DOES describe large objects. Condensed-matter physicists are very successful in doing so. For me the measurement problem is a pseudo-problem. The only thing that's unsolved is to find a viable quntum theory of gravitation and, likely, thus of spacetime. That may need a real "revolutionary" revision of our concepts, but it's not "the measurement problem" that has to be solved.

 

[Fra]

My problem with this argument is the following: What knew do we learn from the Hyperion problem that we didn't already know from the Schrodinger cat problem?

It adds gravity?
/Fredrik

 

[Fra]

But quantum mechanics DOES describe large objects. Condensed-matter physicists are very successful in doing so.

Yes but this is small things without gravity.

For me the measurement problem is a pseudo-problem. The only thing that's unsolved is to find a viable quntum theory of gravitation and, likely, thus of spacetime. That may need a real "revolutionary" revision of our concepts, but it's not "the measurement problem" that has to be solved.

I disagree. From my stance the two problems are related. But i agree its fair to say a revolution is needed. Both a conceptual one and probably new mathematics.

/Fredrik

 

[Demystifier]

It adds gravity?

Since it is Newtonian gravity (not general relativity), I don't think that it changes anything.

 

[martinbn]

My problem with this argument is the following: What knew do we learn from the Hyperion problem that we didn't already know from the Schrodinger cat problem?

I think the point she makes is that it is puzzling to have chaos in classical systems it they are just a limiting case of quantum system. How does the chaos arise? Say you have an interpretation/modification of QM that solves the measurement problem, does it have an explanation of classical chaos? If not, then that is an additional problem for the foundations of QM, it is not just the measurement problem.

 

[Demystifier]

I think the point she makes is that it is puzzling to have chaos in classical systems it they are just a limiting case of quantum system. How does the chaos arise? Say you have an interpretation/modification of QM that solves the measurement problem, does it have an explanation of classical chaos? If not, then that is an additional problem for the foundations of QM, it is not just the measurement problem.

But if one shows in general that the limiting case of QM is classical mechanics, then the origin of chaos is a trivial corollary. And of course, the measurement problem is closely related to the explanation of the classical limit.

 

[A. Neumaier]

if one shows in general that the limiting case of QM is classical mechanics

Her point is that these limiting arguments only hold for times that are not too large. 100 years are too large.

 

[Demystifier]

Her point is that these limiting arguments only hold for times that are not too large. 100 years are too large.

I would expect that for larger times the classical limit is even more valid.

 

[martinbn]

But if one shows in general that the limiting case of QM is classical mechanics, then the origin of chaos is a trivial corollary. And of course, the measurement problem is closely related to the explanation of the classical limit.

But that's the puzzling point. If there is chaos in classical mechanics, how can there be a classical limit!

 

[malawi_glenn]

isn't Hossenfelder on the verge of crackpot soon?

The main sources which are to be discussed in this thread is an arxiv manuscript and a blogpost. I thought only peer-review published work were allowed on this forum? I am just trying to understand

 

[A. Neumaier]

I would expect that for larger times the classical limit is even more valid.

Can you give a reasonably rigorous argument for this?

The problem is that one has to bound the errors in the classical approximation, and the bounds obtainable by the standard techniques are usually exponential in the duration.

 

[A. Neumaier]

I thought only peer-review published work were allowed on this forum?

For justifying an otherwise unsupported claim made by a contributor to the forum, only peer-review published work counts.

Sabine Hossenfelder supports her observations with references from the published literature, though the paper itself is not published.

 

[Demystifier]

But that's the puzzling point. If there is chaos in classical mechanics, how can there be a classical limit!

I don't understand what's puzzling.

 

[Demystifier]

Can you give a reasonably rigorous argument for this?

Far from being rigorous, my argument is roughly $\Delta E \Delta t \sim \hbar$, so for larger $\Delta t$, $\Delta E$ is smaller i.e. more classical.

 

[martinbn]

I don't understand what's puzzling.

All I said was that I think that this is the point she makes.