High School Collapse and unitary evolution

[PaleMoon]

In his book "the black hole war" Susskind writes that in quantum mechanics information cannot be lost because it is unitary. As collapse is not unitary does he say that collapse never occurs?

 

[Demystifier]

Not necessarily. In the usual collapse interpretation of QM, the non-unitary evolution happens only at the instant of measurement. So even if one accepts the collapse interpretation associated with measurement, one still expects that evolution should be unitary before measurement. On the other hand, black hole evaporation seems to imply a non-unitary evolution before the measurement.

 

[PaleMoon]

my question is about Susskind's point of view.
According to him
"Hawking's conclusions violated one of the most basic scientific laws of the universe, the conservation of information"
it seems that he is talking about unitarity. is unitarity a basic law that exists before the birth or a black hole during its life and during its evaporarion?

 

[PeterDonis]

As collapse is not unitary does he say that collapse never occurs?

I believe Susskind favors the MWI, which is a no collapse interpretation, so he would probably say that yes, collapse never occurs.

However, as @Demystifier says, the issue with black holes does not depend on whether collapse occurs; it is present even in "no collapse" interpretations like the MWI.

 

[PaleMoon]

susskind does not use the word collapse.
he recalls that black holes evolution can be equvalently described by two
theories GR in 3+1 dimension and an ADS quantum theory in a different space.
he concludes that Hawking lost the war because information is never lost in quantum mechanics
is there a bias in his conclusion?

 

[PaleMoon]

i think that wigner's friend provides an argument against objective collapse.
when wigner chooses something to be measured there is a quantum object in a spacetime region outside him.
wigner himself is not a quantum object.
suppose that there is a collapse in this spacetime region.
wigner's friend can do another choice and include wigner as a quantum object in a larger spacetime region
for him wigner is in a superposition and collapse could occur later when he will decide to do delayed measurement
may be collapse is observer dependent whereas unitarity is more fundamental

 

[atyy]

susskind does not use the word collapse.
he recalls that black holes evolution can be equvalently described by two
theories GR in 3+1 dimension and an ADS quantum theory in a different space.
he concludes that Hawking lost the war because information is never lost in quantum mechanics
is there a bias in his conclusion?

There is only a bit of bias. The AdS/CFT correspondence is one of the most stunning developments in theoretical physics. It may provide a complete theory of quantum gravity in some universe. The caveats are that although there is much evidence for the correspondence, it remains unproved; also it may not be able to describe qusntum gravity in our universe.

 

[PaleMoon]

i have of course no answer to the measurement problem but i think that collapse is a wrong concept.
mw interpretation is not the only possibible way to avoid it. collapse only concerns individual measurements.
Statisrical ensemble interpretation does not need collapse and unitarity is safe.

 

[atyy]

i have of course no answer to the measurement problem but i think that collapse is a wrong concept.
mw interpretation is not the only possibible way to avoid it. collapse only concerns individual measurements.
Statisrical ensemble interpretation does not need collapse and unitarity is safe.

The statistical ensemble interpretation has collapse. If you wish to preserve unitarity, then you should look at either Bohmian mechanics or the Many-worlds interpretation as apptoaches to the measurement problem.

 

[PeterDonis]

wigner himself is not a quantum object.

Why not?

 

[PaleMoon]

What do you think about Susskind's sentence: in quantum mechanics informationd can never be lost?
this is the core of his conclusion

i read this
"The first merit of Maldacena's work has been to solve the paradox of information, at least in the particular case of a black hole in AdS5. The latter is indeed equivalent to a hot plasma on the boundary, characterized by the temperature of Hawking TH and described by a gauge theory; both have the same entropy. In addition, plasma obeys the usual Rules of Quantum Mechanics, and in particular it evolves unitarily, which automatically leads the black hole to evolve also unitarily and to respect the principles of quantum mechanics [6]. This result led Hawking to revise its position and to announce in 2005 that the paradox was indeed solved by the AdS / CFT correspondence for the conservation of information."

this is written by Jean Pierre Luminet
any collapse would break the unitarity. so physicists insist on the fact that unitarity is a law of quantum field theory.
one often read that states evolve unitarily except when they do not...

 

[PeterDonis]

What do you think about Susskind's sentence: in quantum mechanics informationd can never be lost?

Please give a specific reference. We can't comment on out of context quotes.

any collapse would break the unitarity

The models you are talking about do not contain any measurements, so the question of whether collapse takes place or not is irrelevant. These models are just the same as, for example, the "internals" of a double slit experiment, where even collapse interpretations agree that the evolution of the wave function is unitary; the only "collapse" is at the end of the experiment when the pattern is observed on the detector screen. The equivalent of that in the models you refer to is the universe in the infinite future, when all of the black holes have evaporated and all that is left is an infinite expanse of radiation at extremely low temperature. What "unitary evolution" means in this context is that, for a hypothetical observer in that infinite future universe, they can't tell from any of their measurements whether the infinite expanse of radiation came from the evaporation of black holes or from some other process (like matter-antimatter annihilation leaving only radiation behind) that didn't involve black holes at all.

Note also that all of these models are set in anti-de Sitter space, which is not the universe we live in. A major unresolved issue in this field is whether similar models can be constructed in de Sitter space, which is at least approximately like the universe we live in.

 

[PaleMoon]

it is in "the black hole war" by Susskind at the end of paragraph 22

 

[nrqed]

Not necessarily. In the usual collapse interpretation of QM, the non-unitary evolution happens only at the instant of measurement. So even if one accepts the collapse interpretation associated with measurement, one still expects that evolution should be unitary before measurement. On the other hand, black hole evaporation seems to imply a non-unitary evolution before the measurement.

I have

Not necessarily. In the usual collapse interpretation of QM, the non-unitary evolution happens only at the instant of measurement. So even if one accepts the collapse interpretation associated with measurement, one still expects that evolution should be unitary before measurement. On the other hand, black hole evaporation seems to imply a non-unitary evolution before the measurement.

I have always wondered why the absorption of matter by a black hole could not be considered a type of measurement, which would then take care of the loss o information. I am sure this is stupid for some reason but I have never seen a clear explanation why so.

 

[PeterDonis]

it is in "the black hole war" by Susskind at the end of paragraph 22

Which is a pop science book, so it's not a valid source for discussion here. You need to look at actual textbooks or peer-reviewed papers.

 

[bhobba]

wigner himself is not a quantum object.l

Yes he is - we all are. There is a coarse graining argument that shows how our classical world emerges from the scale below us - but that scale is quantum and so is our scale - its just - as I mention below different physics emerges at different scales - but it's all quantum stuff. You can't choose to view and not view him as a quantum object - he is one all the time.

Just reviewing re-normalization theory at the moment and one of the big insights of Ken Wilson was this idea of different but self similar physics in different scales of nature. Its a very interesting phenomena:
https://websites.pmc.ucsc.edu/~wrs/Project/2014-summer seminar/Renorm/Wilson-many scales-Sci Am-79.pdf

You have to be very wary of things said by the early pioneers like Wigner - great mathematical physicist though he was. Things have moved on a lot since then and from today's vantage things like Wigner's friend are not the issues they were in the early days.

Thanks
Bill

 

[bhobba]

Which is a pop science book, so it's not a valid source for discussion here. You need to look at actual textbooks or peer-reviewed papers.

I just want to mention Susskind does write some excellent actual textbooks for a 'lay' audience under the Theoretical Minimum series. You just need a bit of calculus and its quite accessible if you think a bit - you are not not spoon fed - its real physics. They of course can be freely discussed here. Popular writings, while often interesting to read, I read them myself, are problematical to the aims of a forum like this.

Thanks
Bill

 

[PeroK]

Yes he is - we all are.

Suppose we say that "Wigner is a physicist". Can that be translated into quantum mechanics? I guess it's possible that you could define some properties of a bunch of atoms that represented physicist from non-physicist. But, it seems to me, that it may be impossible to disentangle that property from the atomic configuration.

Or, "Wigner is a US Citizen".

How could you, by studying the atomic configuration of a bunch of atoms tell whether they represented a US Citizen?

In that case, if citizenship is a defining property of a human being, then a human being is not a quantum object.

In general, the question is whether a complex system can assume properties not inherent in the underlying atomic configuration. The reductionist position would be that it cannot.

But, does that mean that what someone has written or achieved in life is either not a defining part of them or is inherent in their current atomic structure?

In any case, Wigner cannot be a quantum object but a continuously evolving set of quantum systems. He's effectively never the same quantum system from one moment to the next.

 

[martinbn]

The statistical ensemble interpretation has collapse.

Why is that?

 

[Fra]

this is written by Jean Pierre Luminet
any collapse would break the unitarity. so physicists insist on the fact that unitarity is a law of quantum field theory.
one often read that states evolve unitarily except when they do not...

This seems a wise statement that is as close to the truth we can come, and not as but not as silly as it may first seem.

As I see it quantum mechanics, deduces the future expectations of the state given expectations of the current state, from requring a consistent self-evoluation of the state. Its is from the self-consistent evolution (given that no perturbations to the deductive system takes place), that unitarity follows deductively.

But let's not forget that the deductive system here is effectively part of the "initial conditions". Ie. one can not really separate initial conditions from laws, except during special conditions. Conditions that happens to be met in regular particle physics (where we have a massive detector controlloing and preparing subatomic events in a small detector), but not in cosmology.

So the idea that unitarity in quantum mechanics "must hold" universally when we talk about general QG or cosmological models, is IMO a fallacy. The reaons for unitarity in quantum mechanics makes sense, but it should be equally clear why one can not wildly extrapolate this beyond the scope of the observer control.

"future evolves as per my expectations, except when they do not" actually constitutes the basis for action, in the context of evolution. Ie. the best we can do, is to act according to our best prognosis, and revise or die when wrong. I expect physical law to be no differently constructed.

I this view one will understand consistent unitarity as a kind of attractor or steady state in theory space and observer population ecology. The biggest difference between that and thinking that its some eternal mathematical constraint is when it comes to try to understand unification of forces. There need not be a conflict between them.

At the heart of the black hole info paradox is indeed the observer role in theory formulation - which is an open issue. So there is not really a "paradox".

/Fredrik

 

[PaleMoon]

Hi Bhobba
i am glad to read you.
i am also fond of things like Busch-Gleason theorem. probabilities are traces
and in your two axioms all is about operators and outputs

i wonder if the purification theorem is not the key to this no information loss
observed operators obey Lindblad equations and entropy may increase.
but it is possible to enlarge the hilbert space so that the global state follpws a Schrödinger evolution
could it be what is suggested by these physicists?

 

[stevendaryl]

I have always wondered why the absorption of matter by a black hole could not be considered a type of measurement, which would then take care of the loss o information. I am sure this is stupid for some reason but I have never seen a clear explanation why so.

Well, something that is strange about the loss of information in a black hole, compared with a normal measurement is this: In a normal measurement, you choose an observable to measure, and (presumably) after the measurement, the system is in an eigenstate of that observable. If you choose to measure the spin of an electron along the x-axis, then afterward the electron is either spin-up or spin-down along that axis (and not a superposition of those two possibilities). If you don't know the result of the measurement, but only know what was measured, then you would describe the system afterward as a mixed state of the two possibilities (which is mathematically different from a superposition of the two possibilities).

In contrast, when one particle of an entangled pair falls into a black hole which then evaporates, the other particle of the pair is in a mixed state, but it's not due to any choice of an observable to measure.

 

[Fra]

I have always wondered why the absorption of matter by a black hole could not be considered a type of measurement, which would then take care of the loss o information. I am sure this is stupid for some reason but I have never seen a clear explanation why so.

I do not see anything stupid with this. As extension to this, I have always "wondered" why not ANY interaction can not be seen as an "observation" - give the right perspective (ie. choice of observing system). Any my own understanding and journey has come to make me completely convinced that this CAN indeed be so. And that its even the KEY to understanding interactions and their unification. But there is still no one that has structured and explained this in a formal way. So except for the fact that we have still to see a fundamental paper that solves and explains this, i an totally convinced it is the best way to understand this.

Information losses are then simply related to that not all observers encode the same information And maybe more important that not all observers can even in princip have equivalent information due to their differing encoding capacities.

/Fredrik

 

[stevendaryl]

I do not see anything stupid with this. As extension to this, I have always "wondered" why not ANY interaction can not be seen as an "observation" - give the right perspective (ie. choice of observing system).

Well, there are two different roles of measurement in QM:

1. Singling out a preferred basis (namely, the observable being measured)
2. The selection of one outcome out of a set of possibilities, according to the Born rule.

Simple interactions do not do either of these.

 

[bhobba]

Suppose we say that "Wigner is a physicist". Can that be translated into quantum mechanics?.

QM is about matter and energy - not human concepts such as occupation.

Thanks
Bill

 

[bhobba]

Hi Bhobba
i am glad to read you. i am also fond of things like Busch-Gleason theorem. probabilities are traces
and in your two axioms all is about operators and outputs

i wonder if the purification theorem is not the key to this no information loss
observed operators obey Lindblad equations and entropy may increase.
but it is possible to enlarge the hilbert space so that the global state follpws a Schrödinger evolution
could it be what is suggested by these physicists?

I am not into areas like game theory the purification theorem comes from so I can't say. I am also sorry to admit I do not know much about this information thing - my knowledge of such matters is Shannon's sampling theorem and related areas in communication theory - information as used by Susskind and others I am almost certain is something different.

Yes indeed that QM comes from just two axioms (well almost but that requires another thread) is quite amazing, and an excellent reason to get and study Ballentine. That the second axiom almost follows from the first makes it mind boggling really. As Weinberg puts it QM seems to be an island in theory space. It's hard to get across the true beauty of QM to those that have not actually studied it. When you understand that the tension between the theory and what it means becomes somewhat different. It leads me to believe interpretations are not really about the meaning of QM, but shedding light on what the theory says. Its just so damn beautiful at it's formal underpinnings.

Thanks
Bill

 

[bhobba]

Well, there are two different roles of measurement in QM:

1. Singling out a preferred basis (namely, the observable being measured)
2. The selection of one outcome out of a set of possibilities, according to the Born rule.

Simple interactions do not do either of these.

I think 1. has been solved, at least Schlosshauer thinks it has (I do as well but am not as expert as he is). Its 2 that's the issue. Some don't even think its an issue - others think its critical. I am in the former camp for what its worth - but that means nothing.

Thanks
Bill

 

[PaleMoon]

. I have always "wondered" why not ANY interaction can not be seen as an "observation" - give the right perspective (ie. choice of observing system). Any my own understanding and journey has come to make me completely convinced that this CAN indeed be so. And that its even the KEY to understanding interactions and their unification. But there is still no one that has structured and explained this in a formal way. So except for the fact that we have still to see a fundamental paper that solves and explains this, i an totally convinced it is the best way to understand this.

/Fredrik

take an electron, you want to observe its spin along some direction. you send it through a stein gerlach apparatus and it inreracts with it. at this level you have measured nothing. you need a screen in front of the possible paths. you need something to encode the output.
so any interaction is not an observation. it has something to do with reversibility and unitarity.

 

[Fra]

Well, there are two different roles of measurement in QM:

1. Singling out a preferred basis (namely, the observable being measured)
2. The selection of one outcome out of a set of possibilities, according to the Born rule.

Simple interactions do not do either of these.

In my post i should have emphasized more that no one that has structured and explained this in a formal way yet. Surely the standard formalism is inadequate to described things in the spirit i envisioned. This kind of doesn't belong in this section though. Perhaps this post https://www.physicsforums.com/threads/ed-witten-on-symmetry-and-emergence.927897/#post-5860052 explains a little but more what i have in mind.

Ie. singling out the preferred basis, or "preparation", is taking care of by the interaction history of the observer in question. Each observer has and encodes their own history.

Maybe i missed nrqeds point, but i just wanted to add support for the idea here.

/Fredrik

 

[stevendaryl]

In my post i should have emphasized more that no one that has structured and explained this in a formal way yet. Surely the standard formalism is inadequate to described things in the spirit i envisioned.

Well, the standard formalism assumes that when someone performs a measurement, he knows what it is that he is measuring, and furthermore that he can recognize a distinct outcome for the measurement. It doesn't really explain these two things, but just assumes them. The founders of the Copenhagen interpretation, Bohr and Heisenberg and those guys, made the distinction between microscopic systems, which are described by quantum mechanics, and measurements/observations, which are (approximately) described by classical mechanics.