Is Relational Quantum Mechanics the Key to Understanding Quantum Interactions?

[Demystifier]

Philosophy is off topic in this forum. Please keep discussion focused on RQM considered as a QM interpretation.

Where exactly is the borderline between interpretations of QM and philosophy of QM? And it's not just a rhetorical question (even though partly it is). Even though there is obviously no sharp borderline, it would be great to have at least some vague guide that can help to distinguish the two.

 

[Demystifier]

Today appeared a paper that claims to make no-go theorems against the relational interpretation. I haven't read it yet, but it could be interesting. https://arxiv.org/abs/2107.00670

 

[martinbn]

Such relativization is cheap and gives nothing new. Whenever you have some reasonable realistic interpretation, you can relativize it without having to add something new at all. All you have to do is to relativize all the absolute things. So all what is reached is that we have, after relativization, less valuable information than before.

If you change relativization with interpretation the same thing can be siad about any interpretation.

 

[PeterDonis]

Where exactly is the borderline between interpretations of QM and philosophy of QM?

The post I responded to isn't even about "philosophy of QM". It's just philosophy, period--not even specific to QM.

 

[*now*]

Today appeared a paper that claims to make no-go theorems against the relational interpretation. I haven't read it yet, but it could be interesting. https://arxiv.org/abs/2107.00670

?
Is there anything aside from the abstract?
I get, “Our automated source to PDF conversion system has failed to produce PDF for the paper: 2107.00670.”

 

[Demystifier]

?
Is there anything aside from the abstract?
I get, “Our automated source to PDF conversion system has failed to produce PDF for the paper: 2107.00670.”

PDF format does not work, but POSTCRIPT (.ps) format does. So click postscript! If on your computer you don't have a tool for reading postscript files and you don't want to install one, you can convert ps file to pdf file, e.g. here https://online2pdf.com/convert-ps-to-pdf .

 

[AlexCaledin]

... It's just philosophy, period--not even specific to QM.

- well, of course, it's just that the very fact of the Process philosophy being a valid philosophical variant seems reconciling all the QM interpretations as useful thinking tools, the Copenhagen pragmatic interpretation being the most useful one. (edit: because the quest for substance becomes unscientific)

 

[*now*]

PDF format does not work, but POSTCRIPT (.ps) format does. So click postscript! If on your computer you don't have a tool for reading postscript files and you don't want to install one, you can convert ps file to pdf file, e.g. here https://online2pdf.com/convert-ps-to-pdf .

thanks, I think that article is unhelpful and admits their argument does not agree with RQM primary literature and acknowledges succumbing to the temptation of entertaining such position anyway. The article, however, does refer to a recent paper that does seem helpful for a thread about RQM criticism. This recent paper referred to addresses a different article’s invalid criticism, suggesting improvement-

“As mentioned in the introduction, every interpretation of QM includes conceptual steps hard to digest. The exercise of criticising the details of an interpretation without accepting these conceptual steps is futile. I am sure we can do better, in articulating a fruitful conversation between different ways of making sense of quantum theory”. This quote is from the conclusion of that paper-

[2106.03205] A response to the Mucino-Okon-Sudarsky's Assessment of Relational Quantum Mechanics (arxiv.org)

A response to the Muciño-Okon-Sudarsky’s Assessment of Relational Quantum mechanics Carlo Rovelli (Dated: June 8, 2021)

 

[Fra]

Rovellis writes in that paper

"Even if The problem of quantum physics is not that we have no way of making sense of it. The problem is that we have many ways of making sense of it. But each of these comes with a high conceptual price. Each interpretation of quantum mechanics demand us to accept conceptual steps that for many are hard to digest"
-- https://arxiv.org/pdf/2106.03205.pdf

This is true I think, and it's the reason why i am not overly interested in "pure interpretations". Instead, if a research program "implied by one interpretation" is able to come up with new progress on open problems, then there is something to discuss. But even until we get there it's interesting to learn how other reason about the matter but not more than that.

I am also aware of the "hard to digest" points, on my own interpretation as well. I always felt that each interpretation pushes the trouble into different corners, which reveals how you think you are best equipped to finding a solution and what starting points you find minimally disturbing.

There first reason I originally started to look into Rovelli, was that. As he has many original ideas on gravity (LQG) I assumed that his interpretation of QM ought to be designed for progress in that direction.

/Fredrik

 

[EPR]

Rovellis writes in that paper

"Even if The problem of quantum physics is not that we have no way of making sense of it. The problem is that we have many ways of making sense of it. But each of these comes with a high conceptual price. Each interpretation of quantum mechanics demand us to accept conceptual steps that for many are hard to digest"
-- https://arxiv.org/pdf/2106.03205.pdf
/Fredrik

Rovelli is right, but only if one insists that the reality must be a Newtonian world. It appears to be a deeply rooted assumption that is just barely true in a very specific case. It's true, but is not the fundamental truth. Nor is it how the world works in its entirety.
There's big potential to answer some age old questions about how the world works. And maybe how we function. The parting with the Newtonian paradigm will not be painless and obviously not to everyone's taste but it's not up to us to decide what is true and what isn't. Is it? Nature doesn't care either way about what Rovelli thinks is a high conceptual price.

 

[Fra]

Rovelli is right, but only if one insists that the reality must be a Newtonian world.

I don't follow. What does the diversity of interpretations has todo with Newtona world?

It appears to be a deeply rooted assumption that is just barely true in a very specific case. It's true, but is not the fundamental truth. Nor is it how the world works in its entirety.
There's big potential to answer some age old questions about how the world works. And maybe how we function. The parting with the Newtonian paradigm will not be painless and obviously not to everyone's taste but it's not up to us to decide what is true and what isn't. Is it? Nature doesn't care either way about what Rovelli thinks is a high conceptual price.

Do you by newtomian paradigm here mean the same aa smolin as he argues for evolution of laws?

I think newtomian paradigm needs tp go out, but the alternative is not exactly smooth either. Law without metalaw? That is hard to digest for many.

/Fredrik

 

[EPR]

I meant there will necessarily be a conceptual price to pay, if one insists the reality is Newtonian. If one were to address the interpretational issue with this preconceived notion in mind.

As for the law without a metalaw - I think emergence is key here. Different organizational structures acting according to laws of particular scales. There are obviously laws and they do not necessarily amount to or refer to laws of lower scales. It could be that the fundamental laws are inaccessible and incomprehensible.

 

[EPR]

It is not even a scientific question what exists between measurements, as all our evidence comes from measurements and observations. We make necessary inferences and build models like we always do and sometimes they fail. Like the current Newtonian model.
When the model fails, we get better opportunities to learn new things about the reality.

 

[Fra]

I meant there will necessarily be a conceptual price to pay, if one insists the reality is Newtonian. If one were to address the interpretational issue with this preconceived notion in mind.

I think we arent talking about the same thing?

I do not insist on Newtonian reality, but I still see the headache.

My point was rather that, for example abandoning the Newtonian paradigm (ie. smolins definition) does causes trouble, because there are no meta laws, all there likely is, is principles or organisation or evolution, that are more fundamental than the laws of physics. Sticking to Newtonian paradigm seems easier, but instead leaves us with a unreasonable fine tuning problem, that may not even have a final solution in the first place if the paradigm is wrong.

Im not saying there is no solution to either way, I just say that with our without Newtonian paradigm we have things to mentally accept.

/Fredrik

 

[Demystifier]

[2106.03205] A response to the Mucino-Okon-Sudarsky's Assessment of Relational Quantum Mechanics (arxiv.org)

A response to the Muciño-Okon-Sudarsky’s Assessment of Relational Quantum mechanics Carlo Rovelli (Dated: June 8, 2021)

Today Muciño-Okon-Sudarsky replied:
https://arxiv.org/abs/2107.05817

 

[*now*]

Today Muciño-Okon-Sudarsky replied:
https://arxiv.org/abs/2107.05817

I don’t see any improvement on the article that has already been addressed.

 

[CarlB]

Today Muciño-Okon-Sudarsky replied:
https://arxiv.org/abs/2107.05817

Sadly the Muciño-Okon-Sudarsky paper only has two equations (1) and (2) which differ in defining the particle in the z-basis or x-basis of spin-1/2. Aren't these identical when converted into density matrices?

[edit] Oh I see, they're already identical as state vectors. Very late at night here.[/edit]Meanwhile, my paper on density matrices, RQM, quantum symmetry and the Standard Model is still under review at Foundations of Physics, most recently with "Reviewers Assigned" with a date of July 6. I'm guessing that they're getting wildly divergent reviews on it and are sending it out for more reviews in an attempt to find two that say the same thing.

 

[*now*]

I'll add the strong no-go theorem here that was mentioned in the earlier paper linked (Di Biagio, A., Rovelli, C. Stable Facts, Relative Facts. Found Phys 51, 30 (2021). https://doi.org/10.1007/s10701-021-00429-w, “Relative facts, stable facts”), and is in the following paper-

https://www.nature.com/articles/s41567-020-0990-x
Nat. phys 2020

Testing the reality of Wigner's friend's observations​

Kok-Wei Bong, Aníbal Utreras-Alarcón, Farzad Ghafari, Yeong-Cherng Liang, Nora Tischler, Eric G. Cavalcanti, Geoff J. Pryde, Howard M. Wiseman

Abstract

Does quantum theory apply at all scales, including that of observers? A resurgence of interest in the long-standing Wigner's friend paradox has shed new light on this fundamental question. Here---building on a scenario with two separated but entangled "friends" introduced by Brukner---we rigorously prove that if quantum evolution is controllable on the scale of an observer, then one of the following three assumptions must be false: "No-Superdeterminism", "Locality", or "Absoluteness of Observed Events" (i.e. that every observed event exists absolutely, not relatively). We show that although the violation of Bell-type inequalities in such scenarios is not in general sufficient to demonstrate the contradiction between those assumptions, new inequalities can be derived, in a theory-independent manner, which are violated by quantum correlations. We demonstrate this in a proof-of-principle experiment where a photon's path is deemed an observer. We discuss how this new theorem places strictly stronger constraints on quantum reality than Bell's theorem.
https://arxiv.org/abs/1907.05607
https://www.quantamagazine.org/a-new-theorem-maps-out-the-limits-of-quantum-physics-20201203/

 

[AndreiB]

Does quantum theory apply at all scales, including that of observers?

It does, but not as presented in Wigner's friend thought experiment. QM gives a simple formula for the uncertainty regarding the measured properties of various systems. For macroscopic systems, such as Wigner's friend, the uncertainty is practically 0. This means that macroscopic objects cannot be in any kind of relevant superposition. Their states can always be known with certainty by any observer.

Presumably, the uncertainty is preserved by isolating the system in a lab or a box. In order for such a lab/box to do its job it needs to stop its contents from interacting with the exterior. That means that the mass or charge of an "isolated" object is "deleted" from the universe once the lab/box is closed. This violates mass-energy or charge conservation so we can conclude that such a lab/box cannot exist.

 

[PeterDonis]

It does

More precisely, according to some physicists, it does. But according to other physicists, it does not. We do not have any conclusive experimental evidence either way. The question is open.

QM gives a simple formula for the uncertainty regarding the measured properties of various systems. For macroscopic systems, such as Wigner's friend, the uncertainty is practically 0. This means that macroscopic objects cannot be in any kind of relevant superposition.

Not in a "Schrodinger's cat" type scenario, i.e., where a macroscopic property of a macroscopic system is set up to depend on a single microscopic quantum event. For such a scenario, QM predicts a superposition of macroscopically distinct states (such as "live cat" and "dead cat"). One reason some physicists do not think QM applies at all scales is that such states do not seem physically reasonable.

 

[AlexCaledin]

QM predicts the probabilities! The superposition is a thinking/calculating tool.

 

[PeterDonis]

QM predicts the probabilities! The superposition is a thinking/calculating tool.

Under certain interpretations, yes. But not others.

 

[AndreiB]

Not in a "Schrodinger's cat" type scenario, i.e., where a macroscopic property of a macroscopic system is set up to depend on a single microscopic quantum event. For such a scenario, QM predicts a superposition of macroscopically distinct states (such as "live cat" and "dead cat"). One reason some physicists do not think QM applies at all scales is that such states do not seem physically reasonable.

The underlying assumption is that it is possible, at least in principle, to isolate the cat. It's easy to see that it cannot be done.

Let's say that a live cat moves inside the box. I can detect this motion, from outside the box, by using a torsion balance. The only way to counter that would be to build a box that can stop gravity. Such a box would, from the point of view of the outside, delete the mass of the cat from the universe, violating the mass/conservation principle. So, such a box cannot exist.

We can also notice that a live cat would have electric currents flowing in its brain. I can also detect the (static) electric field associated with those currents from outside the box, otherwise the box would violate charge conservation.

So, a cat, or any other object that has charge and mass (which includes pretty much everything) cannot be isolated from the exterior. It always interacts with the exterior. The state of such an object is knowable by any observer within the limits of the uncertainty principle.

 

[PeterDonis]

The underlying assumption is that it is possible, at least in principle, to isolate the cat. It's easy to see that it cannot be done.

The cat doesn't have to be isolated indefinitely, only long enough for the experiment to run. That might only be a few seconds if the quantum system is chosen appropriately.

Let's say that a live cat moves inside the box.

And if it doesn't, for the duration of the experiment, you can't detect the cat inside the box by this means. Or you could isolate the box from anything like a torsion balance that would allow detection of motion inside it. For example, you could have the box in free fall in deep space, not connected to anything else, for the duration of the experiment.

We can also notice that a live cat would have electric currents flowing in its brain.

Which can be isolated by making the box a Faraday cage.

a cat, or any other object that has charge and mass (which includes pretty much everything) cannot be isolated from the exterior.

If this argument were true, it would apply to electrons, since they have both charge and mass. But we know we can run experiments on electrons that show superpositions. So this argument cannot be true.

You could try to argue that a macroscopic object cannot be isolated from its environment due to its mass, or electric currents inside it, or something like that. But, as I noted above, the object would not need to be isolated forever, just long enough to run a Schrodinger's cat-type experiment. Your arguments do not show that that is impossible.

 

[AndreiB]

The cat doesn't have to be isolated indefinitely, only long enough for the experiment to run. That might only be a few seconds if the quantum system is chosen appropriately.

Sure, but the cannot be isolated for any amount of time.

Or you could isolate the box from anything like a torsion balance that would allow detection of motion inside it. For example, you could have the box in free fall in deep space, not connected to anything else, for the duration of the experiment.

This does not solve the problem. As long as there is some motion (like a beating heart) you have a change of the mass distribution. Such a change is measurable by its gravitational effects. One or more torsion balances could resolve the mass distribution inside the box. By placing the box in space you only keep its center of mass in the same place.

Which can be isolated by making the box a Faraday cage.

Such a cage cannot shield static electric or magnetic fields, only EM waves. If one could shield a charge, a violation of charge conservation would occur. So it has to be impossible.

If this argument were true, it would apply to electrons, since they have both charge and mass. But we know we can run experiments on electrons that show superpositions. So this argument cannot be true.

The electrons can be in superpositions because of the uncertainty principle, not because you isolate them in a box. The uncertainty principle applies for all observers equally, inside or outside the box. The uncertainty principle is not relevant for macroscopic objects, this is why they cannot be superposed. Sure, uncertainty applies to all objects, but for macroscopic ones it's not observable. The cat could be in a position superposition with a separation of a Planck unit or so, but this is not what is claimed here.

You could try to argue that a macroscopic object cannot be isolated from its environment due to its mass, or electric currents inside it, or something like that.

Exactly. The only limitation is the uncertainty principle. In fact microscopic objects, like electrons cannot be isolated either, for the same reason - mass and charge conservation. The mass and charge of the electron is measurable from outside the box. It's just that the measurements show the expected deviations so you gain nothing from placing the electron in a box.

But, as I noted above, the object would not need to be isolated forever, just long enough to run a Schrodinger's cat-type experiment. Your arguments do not show that that is impossible.

Time is not an issue here. The object cannot be isolated for any time, no matter how small.

 

[PeterDonis]

As long as there is some motion (like a beating heart) you have a change of the mass distribution. Such a change is measurable by its gravitational effects.

"Measurable" in principle is not the same as "measured" in the actual experiment. As long as there is nothing in the experiment that couples to whatever effects the motion has, the motion is isolated, which is all that is required.

The object cannot be isolated for any time, no matter how small.

This claim is not correct; at least, not according to standard QM. Standard QM does not forbid even macroscopic objects from being isolated for some finite amount of time. Perhaps you have some different speculative theory in mind that includes such a limitation?

 

[PeterDonis]

Such a cage cannot shield static electric or magnetic fields

You didn't say static electric or magnetic fields, you said electric currents flowing in the cat's brain. Overall the cat is electrically neutral, so putting it inside a Faraday cage will work just fine; the internal currents will not affect the (zero) external field outside the cage.

 

[PeterDonis]

The electrons can be in superpositions because of the uncertainty principle

Wrong. They are in superpositions if you prepare their states appropriately. It has nothing to do with the uncertainty principle.

The uncertainty principle is not relevant for macroscopic objects

This is not what standard QM says; standard QM says the uncertainty principle is relevant for all measurements, although the exact limitations it places on particular measurements will vary with the size of the system and the specific measurement. Again, perhaps you have some different speculative theory in mind?

 

[AndreiB]

Wrong. They are in superpositions if you prepare their states appropriately. It has nothing to do with the uncertainty principle.

And what is wrong? Yes, you prepare them "appropriately", by a measurement of a non-commuting property. You prepare a UP/DOWN spin superposition on X by passing the electrons through a Z-oriented Stern-Gerlach device. You prepare a momentum superposition by passing the electron through a narrow slit, so that its position is accurately known and so on.

This is not what standard QM says; standard QM says the uncertainty principle is relevant for all measurements, although the exact limitations it places on particular measurements will vary with the size of the system and the specific measurement. Again, perhaps you have some different speculative theory in mind?

Sure, uncertainty applies to everything. However, for a macroscopic object such an instrument pointer or a cat the uncertainty is simply irrelevant because of the high mass of the object. I do not make any speculation here.

The relevant point is that the uncertainty cannot be increased by placing the system in a box. Box or no box, it's the same.

 

[PeterDonis]

you prepare them "appropriately", by a measurement of a non-commuting property.

No, not measurement. Just passing a beam of electrons through a Stern-Gerlach magnet is not a measurement. The measurement comes when the beam hits a detector screen and you record the intensity on different parts of the screen. But if you're going to use the prepared beam for something else (such as passing through a second Stern-Gerlach magnet), you don't do that.

uncertainty applies to everything

It applies to measurements. That's not the same as "everything".

for a macroscopic object such an instrument pointer or a cat the uncertainty is simply irrelevant because of the high mass of the object

The uncertainty for some measurements, such as position vs. momentum, yes. But not all measurements.

The relevant point is that the uncertainty cannot be increased by placing the system in a box.

An experiment like Schrodinger's Cat that prepares a superposition of macroscopically distinct states has nothing to do with "increasing the uncertainty". If any basic principle of QM is involved in such a thought experiment, it's unitarity: if we have a microscopic system in a superposition, and we couple it in the right way to a macroscopic system, the macroscopic system gets included in the superposition, by unitarity.

 

[AndreiB]

No, not measurement. Just passing a beam of electrons through a Stern-Gerlach magnet is not a measurement. The measurement comes when the beam hits a detector screen and you record the intensity on different parts of the screen. But if you're going to use the prepared beam for something else (such as passing through a second Stern-Gerlach magnet), you don't do that.

OK, I stand corrected.

It applies to measurements. That's not the same as "everything".

By "everything" I meant every object, regardless of its size or nature.

The uncertainty for some measurements, such as position vs. momentum, yes. But not all measurements.

If a measurement is performed inside Schrodinger's magic box you need to make sure that an external observer is uncertain in regards to the position of that pointer. If the pointer has a mass of, say 1g there is no way you can use a preparation like in the case of an electron, so we can ignore the uncertainty principle here.

An experiment like Schrodinger's Cat that prepares a superposition of macroscopically distinct states has nothing to do with "increasing the uncertainty".

It does. According to proper QM when a measurement is performed, and the detection of a particle by a Geiger counter is such a measurement, the superposition is gone and the presence of the particle is certain. As far as I can tell there is no version of QM that has a special postulate about magical boxes that transform a measurement in a non-measurement.

If any basic principle of QM is involved in such a thought experiment, it's unitarity: if we have a microscopic system in a superposition, and we couple it in the right way to a macroscopic system, the macroscopic system gets included in the superposition, by unitarity.

As far as I can tell the unitary evolution ends when a measurement (such as a detection by a Geiger counter) occurs. The collapse is supposed to take place. You simply postulate that by using a box you can avoid that. I have provided strong arguments why this does not work. It's because, regardless of the material of the box, an outside observer can in principle observe the position of the pointer inside the box by a measurement of its gravitational field, or by a measurement of electric/magnetic fields produced by the instrument. So, the pointer cannot ever be be in a left/right superposition.

A slightly different way to think about this is to reflect on the well known question "Is the Moon There When Nobody Looks?". The answer is obviously yes, since the Moon has observable consequences everywhere, that are incompatible with a non-existence of the Moon. Without the Moon there should be no tides, Earth's orbital parameters would be different and so on. In other words it is not possible not to observe the Moon. So, if want to place the Moon in a superposition of position states by "coupling it in the right way" with a microscopic system you need to make sure that the tides are not going to "measure" the position of the Moon inside the box. How are you going to do that? What type of box are you going to use?

​

 

[PeterDonis]

If a measurement is performed inside Schrodinger's magic box

It isn't. That's the whole point of the thought experiment. The only measurement is performed when the box is opened and the cat is observed.

As far as I can tell the unitary evolution ends when a measurement (such as a detection by a Geiger counter) occurs.

That depends on which interpretation of QM you adopt. In some interpretations, like the MWI, unitary evolution is all that ever occurs; a "measurement" is just another kind of unitary evolution.

 

[PeterDonis]

A slightly different way to think about this is to reflect on the well known question "Is the Moon There When Nobody Looks?". The answer is obviously yes, since the Moon has observable consequences everywhere, that are incompatible with a non-existence of the Moon. Without the Moon there should be no tides, Earth's orbital parameters would be different and so on. In other words it is not possible not to observe the Moon. So, if want to place the Moon in a superposition of position states by "coupling it in the right way" with a microscopic system you need to make sure that the tides are not going to "measure" the position of the Moon inside the box. How are you going to do that? What type of box are you going to use?

These sorts of ideas were actually the kinds of things that Schrodinger had in mind when he proposed his cat thought experiment; as I understand it, he actually intended it to illustrate the kinds of things you are saying here--that, even though standard QM, taken literally, says it should be possible to put a cat in a superposition of being dead and alive, and Schrodinger's thought experiment describes how that could be done based on standard QM, that implication doesn't really make sense.

Schrodinger couldn't really articulate in 1935 why it doesn't really make sense, but decoherence theory, over the past few decades, has developed a viewpoint much like the one you describe. Basically, for an object with as many degrees of freedom as a cat, some of those degrees of freedom are always interacting with the environment, and those interactions are "observations". In fact, we don't even need the external environment: the degrees of freedom in the cat are always interacting with each other, and those interactions count as "observations" too.

The remaining issue, though, is that standard QM does not tell us, from first principles, how to figure out what counts as an "observation" (or a "measurement"). Those terms are simply undefined terms, and in practice physicists doing QM say that an observation or measurement occurs wherever it has to in order to make the predictions of the theory match what they see in experiments. That works in a practical sense, but it's not really satisfactory as a theory.

Different QM interpretations take different approaches to trying to fix this problem: interpretations like the MWI say that unitary evolution is all that ever happens, so a "measurement" is just more unitary evolution and doesn't require any separate theoretical machinery--but then you have to accept that the picture of "reality" that this interpretation gives you looks nothing like what we actually observe, and it's not clear how to reconcile that discrepancy. Other interpretations treat "measurement" as a distinct physical process, which is not unitary, but then you have the problem of reconciling such a process with relativity, since such a process appears to require that information about measurement results can propagate faster than light.

 

[AndreiB]

The remaining issue, though, is that standard QM does not tell us, from first principles, how to figure out what counts as an "observation" (or a "measurement"). Those terms are simply undefined terms, and in practice physicists doing QM say that an observation or measurement occurs wherever it has to in order to make the predictions of the theory match what they see in experiments. That works in a practical sense, but it's not really satisfactory as a theory.

I think we all agree that a detection of a particle by a Geiger counter is a proper measurement. True, it's not defined from "first principles" but we know from experience that it works, the predictions of QM under the assumption that such a detection is a measurement work.

standard QM, taken literally, says it should be possible to put a cat in a superposition of being dead and alive

No. If you agree that a detection of a particle by a Geiger counter is a measurement, the cat cannot be put in a live/dead superposition. QM says that the state collapses when the particle is detected.

Schrodinger couldn't really articulate in 1935 why it doesn't really make sense, but decoherence theory, over the past few decades, has developed a viewpoint much like the one you describe.

No, decoherence is a different story.

Basically, for an object with as many degrees of freedom as a cat, some of those degrees of freedom are always interacting with the environment, and those interactions are "observations". In fact, we don't even need the external environment: the degrees of freedom in the cat are always interacting with each other, and those interactions count as "observations" too.

This misses the point that even single electrons interact continuously with the environment. In a 2-slit experiment the electrons are still subject to gravitational and EM interactions. The crucial point is that those interactions cannot give you enough information about the electron so that you can determine the slit the electron passed through. This is why the electron is in superposition. In the case of the cat, the interactions DO give you enough information, so the cat is not in a superposition.

The assumption you need to make in order to speak about live-dead superpositions is that you can create a box that stops the cat from interacting with the environment, so that the information you can extract from there are again insufficient to determine the cat's state. This assumption is wrong IMHO for the reasons explained earlier.

Again, an electron does interact gravitationally. The electron would make a torsion balance move, but the uncertainty principle tells us that the accuracy in electron's position is not enough to get which-slit information. A cat interacts gravitationally exactly like the electron does, but in this case the measurement accuracy is enough to tell you if it's heart beats or not. The uncertainty principle still applies, but the measurement accuracy is good enough to distinguish between a heart that beats and one that does not. This is all.

Different QM interpretations take different approaches to trying to fix this problem: interpretations like the MWI say that unitary evolution is all that ever happens, so a "measurement" is just more unitary evolution and doesn't require any separate theoretical machinery--but then you have to accept that the picture of "reality" that this interpretation gives you looks nothing like what we actually observe, and it's not clear how to reconcile that discrepancy.

MWI fails to predict the correct probabilities, so it's not a valid interpretation.

Other interpretations treat "measurement" as a distinct physical process, which is not unitary, but then you have the problem of reconciling such a process with relativity, since such a process appears to require that information about measurement results can propagate faster than light.

Standard QM is fine enough here. Everybody agrees that a detection of a particle by a Geiger counter is a measurement. The question "what if the counter is placed in a box?" is irrelevant, since QM does not present us with a different set of postulates that apply only in the presence of a box.

What I think happens here is that a hidden assumption is made, that a measurement is only a measurement if some observer looks at the instrument. Such an assumption is not part of the standard QM, has no evidence in its favor, so it should be dismissed.

 

[PeterDonis]

I think we all agree that a detection of a particle by a Geiger counter is a proper measurement.

If by "a proper measurement" you mean "something that collapses the wave function, in a real sense instead of just as a mathematical convenience", then no, we don't all agree on this, because not all intepretations agree on it.

QM says that the state collapses when the particle is detected.

"QM" (meaning basic QM without adopting any particular interpretation) only says this as a mathematical convenience, not as a claim about what "really happens". Whether or not a collapse "really happens" is interpretation dependent.

decoherence is a different story.

Not from what you were talking about in your previous post that I quoted from. Your description of interaction with the environment and how that counts as "observation" is decoherence.

MWI fails to predict the correct probabilities

Wrong. MWI makes the same predictions as all other interpretations of QM, since all interpretations use the same or equivalent mathematical machinery to make predictions.

What I think happens here is that a hidden assumption is made, that a measurement is only a measurement if some observer looks at the instrument.

No, what is happening here is that you are confusing "basic" QM--the math that makes predictions about probabilities but makes no claims about what "really happens"--with interpretation-dependent claims.