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

[Demystifier]

A very interesting read, thank you. If you don't mind sharing, what was the deep conceptual error in arXiv:1309.0400?

Eq. (182) is only valid after the measurement. On the other hand, Eq. (185) contains a tacit (but wrong) assumption that (182) is valid at all times.
http://www.thethingswesay.com/there...integrity-than-his-behavior-when-he-is-wrong/
http://www.quote-coyote.com/quotes/authors/l/bruce-lee/quote-3714.html

 

[vanhees71]

Here's what seems strange to me. You have system A, an electron, say. For simplicity, only consider the property of being spin-up in the z-direction. You have system B, some measuring device. Among other things, it has a pointer that can swing from pointing left, where there is a label "U" to pointing right, where there is a label "D". You somehow connect the two systems so that system B measures the spin of system A: If system A is spin-up in the z-direction, then system B will go into the state of pointing to "U", and if system A is spin-down in the z-direction, then system B will go into the state of pointing to "D".

So for people say that properties of system A are meaningless, or have no definite value, until they are measured by system B seems weird if they are both quantum systems. Does system B need a third system, C to make its pointer-value meaningful? That would lead to an infinite regress.

The way I feel about it is that unless one can formulate the Rules of Quantum Mechanics in a way that does not mention, at the fundamental level, any macroscopic quantities such as "measurement", "preparation procedure", "average over many, many systems", then we don't really understand quantum mechanics. That might be fine. There might be limits to what we can understand. But I object to people pretending otherwise.

I think the problematic thing is to call properties which are not prepared as "meaningless". If you have a system in a state, the observables that have no determined value are of course not meaning less but measurable, and in measuring them you usually have an influence on the state of the measured system. Which one this is, depends on the interaction between measurement apparatus and measured system. In the special case of von-Neumann-filter measurements (I'd rather call them a certain kind of preparation procedure) you have prepared a state, where the observable takes the corresponding determined value.

 

[martinbn]

Here's what seems strange to me. You have system A, an electron, say. For simplicity, only consider the property of being spin-up in the z-direction. You have system B, some measuring device. Among other things, it has a pointer that can swing from pointing left, where there is a label "U" to pointing right, where there is a label "D". You somehow connect the two systems so that system B measures the spin of system A: If system A is spin-up in the z-direction, then system B will go into the state of pointing to "U", and if system A is spin-down in the z-direction, then system B will go into the state of pointing to "D".

So for people say that properties of system A are meaningless, or have no definite value, until they are measured by system B seems weird if they are both quantum systems. Does system B need a third system, C to make its pointer-value meaningful? That would lead to an infinite regress.

The way I feel about it is that unless one can formulate the Rules of Quantum Mechanics in a way that does not mention, at the fundamental level, any macroscopic quantities such as "measurement", "preparation procedure", "average over many, many systems", then we don't really understand quantum mechanics. That might be fine. There might be limits to what we can understand. But I object to people pretending otherwise.

This is what I don't understand. Why do you insist on the theory being of certain type? Why is it not ok to mention these notions? It seems to me it is a matter of taste. Almost as saying as long as the theory uses differential equations it is not a good explanation. It is incomplete until a purely algebraic description is found.

 

[Demystifier]

This is what I don't understand. Why do you insist on the theory being of certain type? Why is it not ok to mention these notions? It seems to me it is a matter of taste. Almost as saying as long as the theory uses differential equations it is not a good explanation. It is incomplete until a purely algebraic description is found.

In physics, there is a widespread belief that fundamental laws must be fully microscopic. You can compare it with a widespread belief in pure math that all math must be based on set theory. Proposing that macro laws could be fundamental can be compared to a proposal that math should be based on category theory (rather than set theory). Yes, some people propose it, but the mainstream does not buy it.

 

[stevendaryl]

This is what I don't understand. Why do you insist on the theory being of certain type?

I'm not insisting on anything. I'm just explaining why I feel there is something not yet understood about quantum mechanics. My feeling is that macroscopic properties should be derivable from microscopic properties, so that in principle, any mention of macroscopic properties should be eliminable. That's part of the reductionist program, it seems to me.

 

[stevendaryl]

I think the problematic thing is to call properties which are not prepared as "meaningless". If you have a system in a state, the observables that have no determined value are of course not meaning less but measurable, and in measuring them you usually have an influence on the state of the measured system.

But to me, calling something "measurable" is the issue. A property is measurable if some procedure can make it's value correlated with a macroscopic property (such as a pointer position). But what makes pointer positions different than properties such as the z-component of spin? Why does the first not need to be measured to have a value? Of course, that would lead to an infinite regress, but how do you stop the regress? It seems to me by saying that there is something special about pointer positions.

Can one electron measure the spin of another electron?

 

[stevendaryl]

In physics, there is a widespread belief that fundamental laws must be fully microscopic. You can compare it with a widespread belief in pure math that all math must be based on set theory. Proposing that macro laws could be fundamental can be compared to a proposal that math should be based on category theory (rather than set theory). Yes, some people propose it, but the mainstream does not buy it.

I think it's different from that. I couldn't care less whether you base your math on set theory or category theory. But if you had a law of physics that mentions macroscopic systems---say, that cats always land on their feet--it seems to me that either the law should be derivable from particle dynamics (since a cat is made up of particles, after all) or else particle dynamics is actually violated when the particles are part of a cat.

I suppose a third possibility is some kind of superdeterminism. The cat's particles just obey ordinary particle dyanmics, but the initial conditions are such that the cat always lands on its feet.

 

[stevendaryl]

This is what I don't understand. Why do you insist on the theory being of certain type? Why is it not ok to mention these notions? It seems to me it is a matter of taste. Almost as saying as long as the theory uses differential equations it is not a good explanation. It is incomplete until a purely algebraic description is found.

The macroscopic description of a measurement situation might be something like this:

• if you prepare an electron in the spin state $\alpha |U\rangle + \beta |D\rangle$ and perform a measurement of spin in the z-direction, then the device will make a transition to the "Measured spin-up" state with probability $|\alpha|^2$ and to "Measured spin-down" with probability $|\beta|^2$.

Since "Measured spin-up" and "Measured spin-down" are presumably states of the measuring device, and the measuring device is made up of ordinary particles, then it seems that in principle, the above rule should be re-expressible in the form:

• If you prepare a collection of particles in such-and-such a state, then later they will be in such-and-such a state with probability $|\alpha|^2$. (Not "measured to be" in that state, because the system already includes the measuring device, which presumably doesn't need to be measured by a third system. Or does it?)

It seems that in principle, it should be possible to eliminate the measurement aspect of the theory and re-express it as a theory of pure particles. If that can't be done, that seems pretty weird to me. On the other hand, we're pretty sure that it can't be done, because the dynamics of pure particles is deterministic (Schrodinger's equation) regardless of how many particles are involved. So there's something weird going on.

 

[vanhees71]

But to me, calling something "measurable" is the issue. A property is measurable if some procedure can make it's value correlated with a macroscopic property (such as a pointer position). But what makes pointer positions different than properties such as the z-component of spin? Why does the first not need to be measured to have a value? Of course, that would lead to an infinite regress, but how do you stop the regress? It seems to me by saying that there is something special about pointer positions.

Can one electron measure the spin of another electron?

The difference between "macroscopic" and "microscopic" observables is that the former are coarse grained, i.e., averages over many microscopic degrees of freedom, which have the tendency to behave classical.

The infinite regress you mention simply stops by construction of a macroscopic measurement apparatus and its verification by experiments to really measure what it's supposed to measure. Physicists must be to a certain amount practitioners and must indeed stop to worry about such purely philosophical problems. Although Bell didn't like the expression "for all practical purposes" ("FAPP"), at a certain point you must get practical not to get lost in infinite regress of purposeless philosophical pondering. It's the art of the physicist to disinguish between interesting and pointless questions!

Finally, as I said before, one should say that an observable of a quantum system takes a determined value if the system is prepared in a corresponding state. Just to measure it doesn't give it a certain value. Quite often the measured system (like a photon) is destroyed in the act of measurement, and then you can only say, you've measured some value on this individual system. To check the predictions of QT you need to prepare a sufficiently large ensemble to test the probabilities predicted to a given significance level.

 

[Demystifier]

I think it's different from that. I couldn't care less whether you base your math on set theory or category theory. But if you had a law of physics that mentions macroscopic systems---say, that cats always land on their feet--it seems to me that either the law should be derivable from particle dynamics (since a cat is made up of particles, after all) or else particle dynamics is actually violated when the particles are part of a cat.

I have another mathematical analogy that can be useful. Saying that cat is made of particles is like saying that a function f(x) is made of points - at each local point x you have to specify f(x). However, you can make a Fourier transform and say that the function is not made of local points but of global functions sin and cos. In other words, you can have locality in the k-space rather than the x-space. Which space is fundamental? We don't know a priori. If forces of nature are local in the x-space, then it seems reasonable that x-space is more fundamental than the k-space. But if forces are not local in x-space (as violation of Bell inequalities suggests), then perhaps "fundamental" does not mean "micro". Perhaps we need to make some big functional transform of all our known laws of physics and obtain a truly local laws in some completely different space. To connect this (farfetched?) speculation with something more familiar, perhaps this big transform is somehow related to AdS/CFT and EPR=ER conjectures.

 

[Physics Footnotes]

The macroscopic description of a measurement situation might be something like this:

• if you prepare an electron in the spin state $\alpha |U\rangle + \beta |D\rangle$ and perform a measurement of spin in the z-direction, then the device will make a transition to the "Measured spin-up" state with probability $|\alpha|^2$ and to "Measured spin-down" with probability $|\beta|^2$.

Since "Measured spin-up" and "Measured spin-down" are presumably states of the measuring device, and the measuring device is made up of ordinary particles, then it seems that in principle, the above rule should be re-expressible in the form:

• If you prepare a collection of particles in such-and-such a state, then later they will be in such-and-such a state with probability $|\alpha|^2$. (Not "measured to be" in that state, because the system already includes the measuring device, which presumably doesn't need to be measured by a third system. Or does it?)

It seems that in principle, it should be possible to eliminate the measurement aspect of the theory and re-express it as a theory of pure particles. If that can't be done, that seems pretty weird to me. On the other hand, we're pretty sure that it can't be done, because the dynamics of pure particles is deterministic (Schrodinger's equation) regardless of how many particles are involved. So there's something weird going on.

This is a very elegant articulation of the so-called Measurement Problem, and makes very clear why it is called a 'problem', namely: the experiments used to justify quantum mechanics are, by that very theory, not dynamically possible!

Despite what you may read to the contrary (here or elsewhere), this problem has not been resolved, and so it should be no surprise that it is causing you so much head-scratching. Many physicists like to pretend it has been solved by hand-wavy arguments littered with terms like 'decoherence', 'coarse-graining', 'Ehrenfest's Theorem', and so forth, however no such techniques have succeeded in putting this problem in the solved tray. (If it was true that this problem had been solved, you would not have otherwise perfectly level-headed physicists desperately introducing unobservable multi-verses with abandon!)

One can certainly avoid the problem by resorting to interpretations that avoid one or more assumptions you have made, but they then have their own problems, resulting in no consensus whatsoever about the best way to respond to the problem.

 

[Lord Jestocost]

I'm not insisting on anything. I'm just explaining why I feel there is something not yet understood about quantum mechanics. My feeling is that macroscopic properties should be derivable from microscopic properties, so that in principle, any mention of macroscopic properties should be eliminable. That's part of the reductionist program, it seems to me.

There is – so to speak - indeed a “problem” with quantum theory (QT) and it seems to me to be an “insoluble” problem – despite opposite claims. The “problem” can be explained in a simple way: There is one equation and one quantity which define the theory – the Schroedinger equation and the associated wave function – and those don’t describe how we - as conscious observers - experience our world. That is the fundamental essence of Schroedinger’s cat fable.

If you accept QT as a fundamental physical theory, you have to apply the theory straightforward at all stages, there is no way out. However, QT allows at no level definite outcomes to be realized, whereas at the level of our human consciousness it seems a matter of direct experience that such outcomes occur. That means, suddenly, when you make a measurement (observation), there is somehow a “cut" or "collapse”, something seems to become “concrete” and “real” stuff. And QT says nothing about it: the conceptual transition from quantum to classical “knowing” had to be put in “by hand”. This is indeed something not yet understood about quantum theory when considering it as a fundamental physical theory about "reality".

 

[bhobba]

Despite what you may read to the contrary (here or elsewhere), this problem has not been resolved, and so it should be no surprise that it is causing you so much head-scratching

Nobody here says its been solved. Consistently we all say issues remain.

The question is - are they worth worrying about. A number of people here, including myself and Vanhees, believe in the Ensemble Interpretation of Einstein - updated for modern times of course. But unlike Einstein many of us accept its just the way the world is and don't get worked up over it. There is no way to tell the difference between an improper mixed state and a proper one so who cares? Yes they are different but so what?

Just our view, but we are happy with it. Einstein probably wouldn't be - but to each their own.

Bottom line - no it has not been fully solved, but its purely a matter of opinion if its worth getting upset about. Every theory, every single one has things it accepts, it simply a matter of taste if you get worked up over them or not. I for one am perfectly happy with the state affairs. Some get worked up about foundational issues in probability, but most don't care a hoot and simply use it. Its not shut-up and calculate - its more like - well yes they are their but exactly why is it a worry? Its not like any theory explains everything.

I like to study various interpretations, not because I want to get to the bottom of how it all really works or anything like that. I am perfectly happy with the Ensemble interpretation. It just helps me understand the formalism better. For example its easy to get the impression from the formalism it has collapse - it doesn't but its not really clear until you study non-collapse interpretations - or even exactly how nuanced the question of just what collapse means - it's by no means straight forward.

Thanks
Bill

 

[atyy]

One can certainly avoid the problem by resorting to interpretations that avoid one or more assumptions you have made, but they then have their own problems, resulting in no consensus whatsoever about the best way to respond to the problem.

There need not be a best way without experiment. We would like to know all ways of responding, then deciding the best way by experiment.

 

[atyy]

Its not shut-up and calculate - its more like - well yes they are their but exactly why is it a worry? Its not like any theory explains everything.

Indeed why worry about quantum gravity - the existing theory is perfectly fine.

 

[bhobba]

Indeed why worry about quantum gravity - the existing theory is perfectly fine.

Well actually it is - up to the plank scale:
https://arxiv.org/abs/1209.3511

We want to know beyond that.

QM does not have that issue - as far as we know it works everywhere. It explains all phenomena in its domain - in gravity there is a domain about which we really know nothing. That may require a revision in QM to resolve - but any theory - any theory at all is just provisional.

Thanks
Bill

 

[vanhees71]

This is a very elegant articulation of the so-called Measurement Problem, and makes very clear why it is called a 'problem', namely: the experiments used to justify quantum mechanics are, by that very theory, not dynamically possible!

It's only not dynamically possible, if you insist on a collapse assumption, but that's not even part of many flavors of the Copenhagen interpretation!

 

[bhobba]

and those don’t describe how we - as conscious observers - experience our world.

Why yes - that is the central mystery - why do we get outcomes at all - colloquially of course - not technically which requires considerable detail to explain properly.

But the question is why don't you just accept that's how nature is? Why get worked up about it? Even if you answer it , and its experimentally proven, there will be another unknown that replaces it. Its just a matter of taste if you like some assumptions and not others.

If the issue interests you - great - research away but I get this sneaky feeling those that harp on it have some sort of evangelist bent this is the Earth shattering thing about QM that needs immediate attention. Sorry - but I don't agree.

Thanks
Bill

 

[kith]

@Demystifier, I find it very interesting to read about your evolving views on QM. Thanks for writing it up!

I don't know if this is incidental but David Wallace has written an article with the same film reference in its name and about a similar topic (Decoherence and Ontology, or: How I Learned To Stop Worring And Love FAPP). He also talks about emergent structures and uses analogies about quasi-particles, albeit from a MWI perspective.

 

[kith]

I also have a question regarding arXiv:1703.08341. If all particles are actually quasi-particles emerging from the behaviour of hypothetical non-relativistic "atoms", how are these "atoms" different from the aether?

For photons, don't the same counterarguments apply to this idea as to the aether in classical electromagnetism?

 

[kith]

But the question is why don't you just accept that's how nature is? Why get worked up about it?

Personally, I'm still figuring out how much I want to get worked up about it. ;-)

I have a bit of sympathy for the view that we can't remove the observer from science, that QM is a broad hint in this direction and that there's a limit to our understanding of Nature. But this still leaves a number of things to ponder about. How much exactly can we say and where is the limit? Can this view be reconciled with everyday realism? What does QM tell us about the nature of probabilities? etc.

 

[Lord Jestocost]

The question is - are they worth worrying about. A number of people here, including myself and Vanhees, believe in the Ensemble Interpretation of Einstein - updated for modern times of course. But unlike Einstein many of us accept its just the way the world is and don't get worked up over it. There is no way to tell the difference between an improper mixed state and a proper one so who cares? Yes they are different but so what?

I care. There is a difference, because even if the "syntaxes" (when looking at the density matrices) seem to be the same, the "semantics" are fundamentally different. Physics should as a matter of principle avoid to present interpretations which might somehow be subject to "confirmation bias".

 

[fanieh]

The macroscopic description of a measurement situation might be something like this:

• if you prepare an electron in the spin state $\alpha |U\rangle + \beta |D\rangle$ and perform a measurement of spin in the z-direction, then the device will make a transition to the "Measured spin-up" state with probability $|\alpha|^2$ and to "Measured spin-down" with probability $|\beta|^2$.

Since "Measured spin-up" and "Measured spin-down" are presumably states of the measuring device, and the measuring device is made up of ordinary particles, then it seems that in principle, the above rule should be re-expressible in the form:

• If you prepare a collection of particles in such-and-such a state, then later they will be in such-and-such a state with probability $|\alpha|^2$. (Not "measured to be" in that state, because the system already includes the measuring device, which presumably doesn't need to be measured by a third system. Or does it?)

It seems that in principle, it should be possible to eliminate the measurement aspect of the theory and re-express it as a theory of pure particles. If that can't be done, that seems pretty weird to me. On the other hand, we're pretty sure that it can't be done, because the dynamics of pure particles is deterministic (Schrodinger's equation) regardless of how many particles are involved. So there's something weird going on.

You are basically asking why the detector in the double slit experiment can only detect one electron and not have multiple hits for only one electron emitted. But isn't it that according to:

1. Bohmian Mechanics.. there is a trajectory for the one electron being emitted so it hits the detector at one point...
2. Many Worlds.. there are multiple hits in the screen.. but we only viewed one of them because we are entangled with only one of them...
3. Copenhagen.. the wave function may pass through both slits but it collapses into one hit when it reached the screen...

These are the explanations why there is only one detector hit in the screen and supposed to address your "So there's something weird going on" ... are you saying you don't believe in the explanations? I can't seem to get your point... it's as if the interpretations are not related to your concern? Kindly elaborate.

 

[bhobba]

Personally, I'm still figuring out how much I want to get worked up about it. ;-)

That's the issue isn't it.

Some get almost evangelistic about it, others just shrug and accept say what Ballentine says.

Still others like me, while agreeing mostly with Ballentine gain insight by studying various interpretations to understand the formalism better.

And either one of those can get really 'into it' as many threads here show.

Thanks
Bill

 

[fanieh]

That's the issue isn't it.

Some get almost evangelistic about it, others just shrug and accept say what Ballentine says.

Still others like me, while agreeing mostly with Ballentine gain insight by studying various interpretations to understand the formalism better.

And either one of those can get really 'into it' as many threads here show.

Thanks
Bill

Something I want to know. If quantum state and even the quantum fields are just smoke and mirrors or not really there.. but statistical.. why are there forces of nature such as the electroweak force.. if objects are just smoke and mirror.. why do they seem to exist as stable object. Are you saying that symmetry and gauge symmetry is what created our universe.. so it's a battle or difference between pure mathematical symmetry and quantum state having objective properties.. but can't they occur at same time.. that is.. our universe results from mathematical symmetry and quantum state can be real? Or only one thing is true and why is that?

 

[atyy]

Well actually it is - up to the plank scale:
https://arxiv.org/abs/1209.3511

We want to know beyond that.

How do you know there is a "beyond that"?

QM does not have that issue - as far as we know it works everywhere. It explains all phenomena in its domain - in gravity there is a domain about which we really know nothing. That may require a revision in QM to resolve - but any theory - any theory at all is just provisional.

QM does not work everywhere. It requires the classical-quantum cut, so there is always somewhere that is not described by QM.

 

[Demystifier]

@Demystifier, I find it very interesting to read about your evolving views on QM. Thanks for writing it up!

I'm glad that I had the opportunity to share my evolving views with others.

I don't know if this is incidental but David Wallace has written an article with the same film reference in its name and about a similar topic (Decoherence and Ontology, or: How I Learned To Stop Worring And Love FAPP). He also talks about emergent structures and uses analogies about quasi-particles, albeit from a MWI perspective.

This is not incidental. The title of this paper was an inspiration for me, as was the title of https://arxiv.org/abs/1201.2714 .
And all those titles are inspired by the famous Kubrick's movie https://en.wikipedia.org/wiki/Dr._Strangelove .

 

[Demystifier]

I also have a question regarding arXiv:1703.08341. If all particles are actually quasi-particles emerging from the behaviour of hypothetical non-relativistic "atoms", how are these "atoms" different from the aether?

They aren't. That's aether in disguise.

For photons, don't the same counterarguments apply to this idea as to the aether in classical electromagnetism?

Of course. Nobody ever proved that aether doesn't exist. What has Einstein (and others) demonstrated is that it is merely simpler to describe the observed phenomena without the aether. But simpler doesn't always mean "more correct". For instance, it is simpler to describe a fluid as a continuum, yet today we know that it is more correct to describe it as a discrete set of atoms.

 

[vanhees71]

If there is something close to an (a)ether in contemporary physics it's the vacuum state of QFT, but it's by construction a Poincare invariant state, i.e., it doesn't introduce a preferred frame of reference as the old (a)ether ideas did, and for me that's as good as saying that there's no (a)ether in this old sense.

 

[bhobba]

Something I want to know. If quantum state and even the quantum fields are just smoke and mirrors or not really there.. but statistical.. why are there forces of nature such as the electroweak force.

Observations really do exist.

We do not know why the electro-weak force exists - it has a beautiful symmetry from which its properties follow - but why it exists we do not as yet know.

As one person said about the standard model - it has some parts of dazzling beauty - that would be the symmetry bit - other parts are an ugly kluge - that would be the constants that need to be put in by hand.

Thanks
Bill

 

[Demystifier]

If there is something close to an (a)ether in contemporary physics it's the vacuum state of QFT, but it's by construction a Poincare invariant state, i.e., it doesn't introduce a preferred frame of reference as the old (a)ether ideas did, and for me that's as good as saying that there's no (a)ether in this old sense.

Of course, you are talking about QFT in Minkowski spacetime. But in QFT in curved spacetime there is no Poincare invariant vacuum. Typically one finds several "natural" vacuums in a given spacetime background and it is not easy to decide which, if any, is "the physical one".

 

[bhobba]

How do you know there is a "beyond that"?

Well the theory we have is an EFT that is known, like all effective theories, to be untrustworthy below a certain scale. For the EFT of gravity that's the Plank scale. Its wise to keep this in mind because gravity is really in the same boat as our other theories - we don't know how well they fare either below the Plank scale eg we have things like the Landau pole. It was once thought string theory would help but that didn't quite work out.

Thanks
Bill

 

[atyy]

Well the theory we have is an EFT that is known, like all effective theories, to be untrustworthy below a certain scale. For the EFT of gravity that's the Plank scale. Its wise to keep this in mind because gravity is really in the same boat as our other theories - we don't know how well they fare either below the Plank scale eg we have things like the Landau pole. It was once thought string theory would help but that didn't quite work out.

So just cut the theory off at that scale. Is there any problem with that?

 

[atyy]

If there is something close to an (a)ether in contemporary physics it's the vacuum state of QFT, but it's by construction a Poincare invariant state, i.e., it doesn't introduce a preferred frame of reference as the old (a)ether ideas did, and for me that's as good as saying that there's no (a)ether in this old sense.

But by the Landau pole, QED and the standard model are suspected not to exist above a certain energy, ie. the standard model is usually thought to be an effective field theory. So the standard model may not come from Poincare invariant theory.

 

[vanhees71]

That may well be true, and maybe after all nature is not Poincare invariant at very large scales where we have no observations yet. All this is, of course, wild speculation, which won't be solved by theory alone but one needs some phenomenological hint to the highly desirable "physics beyond the standard model"!