Why does nothing happen in MWI?

[Derek Potter]

Could someone explain Jan-Markus Schwindt's idea, published under the provocative title

Nothing happens in the Universe of the Everett Interpretation​

please? Frankly I can't even understand the opening paragraph of the abstract.
http://arxiv.org/abs/1210.8447

Since the scalar product is the only internal structure of a Hilbert space, all
vectors of norm 1 are equivalent, in the sense that they form a perfect sphere in
the Hilbert space, on which every vector looks the same. The state vector of the
universe contains no information that distinguishes it from other state vectors of
the same Hilbert space.

I don't get that. The state vector has direction. In fact that's all it's got. But it's a direction relative to any arbitrary basis. I don't think I'm misinterpreting Schwindt's meaning as he goes on to identify a way of factorizing the space, where each branch "peacefully rotates" and nothing else happens. He calls these Nirvana frames.

But I'm out of my depth with tensors and I find it hard to extract the logic from Schwindt's exposition so can someone explain what it all means?

 

[bahamagreen]

Maybe you can start with a look at factorization.

It seems similar to Fourier decomposition... and that any waveform may not just be decomposed into sine waves of various amplitudes, phases, and wavelengths, but that it may be decomposed into other waves... triangle, saw tooth, etc... any arbitrary shape.

The attributes that correspond to the simple waves of decomposition correspond to simple things like position and momentum, the more peculiar ones correspond to attributes requiring a much more complex experimental condition, typically quite impossible to present.

Conceptually, a waveform may be viewed as having either no attributes or all of them, subject to the choice of wave shape comprising those into which it is decomposed.

The mention of the sphere is also suggestive of this, as Bohr's idea of complementary attributes can be represented as antipodal points in the sphere, mapping all possible attributes (all possible choices of wave form of which to decompose any arbitrary wave).

I know this is all very hand wavy, but maybe it helps find where to grab onto it?

 

[Derek Potter]

Maybe you can start with a look at factorization.

It seems similar to Fourier decomposition... and that any waveform may not just be decomposed into sine waves of various amplitudes, phases, and wavelengths, but that it may be decomposed into other waves... triangle, saw tooth, etc... any arbitrary shape.

The attributes that correspond to the simple waves of decomposition correspond to simple things like position and momentum, the more peculiar ones correspond to attributes requiring a much more complex experimental condition, typically quite impossible to present.

Conceptually, a waveform may be viewed as having either no attributes or all of them, subject to the choice of wave shape comprising those into which it is decomposed.

The mention of the sphere is also suggestive of this, as Bohr's idea of complementary attributes can be represented as antipodal points in the sphere, mapping all possible attributes (all possible choices of wave form of which to decompose any arbitrary wave).

I know this is all very hand wavy, but maybe it helps find where to grab onto it?

Thank you bahamagreen but that really isn't what I was asking about. Are you familiar with Schwindt's paper that I referred to? The puzzling thing is that he doesn't just show there is a frame in which nothing happens but, if I have understood him, he claims that nothing happens in any frame. Hence the title.

 

[PeterDonis]

The state vector has direction. In fact that's all it's got. But it's a direction relative to any arbitrary basis.

Considering a single state vector, yes, I suppose that's true. But the inner product between any two given state vectors is not arbitrary; it's an invariant number, independent of your choice of basis. In fact, the "direction relative to a given basis" of any state vector is just the inner product of that state vector with a basis vector, which is itself just a state vector.

 

[Derek Potter]

Considering a single state vector, yes, I suppose that's true. But the inner product between any two given state vectors is not arbitrary; it's an invariant number, independent of your choice of basis. In fact, the "direction relative to a given basis" of any state vector is just the inner product of that state vector with a basis vector, which is itself just a state vector.

Yes I understand that - but how does that equate to "The state vector of the universe contains no information that distinguishes it from other state vectors of the same Hilbert space."? I don't think I know what Schwindt means by that.

 

[PeterDonis]

how does that equate to "The state vector of the universe contains no information that distinguishes it from other state vectors of the same Hilbert space."?

I can't say for sure, but the author's thought process may go something like this: in order to distinguish different state vectors from one another, we need to pick one particular vector and then distinguish other vectors from it using the inner product. But nothing picks out any particular vector as being "the" vector we should be using in this way.

I don't think this actually justifies the statement the author makes; but I haven't read his paper in detail, only skimmed it, so I may be missing something.

(Also, it looks to me like the author's purpose in making these statements about the state vector is to argue against the claim that the state vector captures everything we need to know about "reality". In other words, to the author, the passage you quoted is intended to make you skeptical of the idea that the state vector is all we need to do quantum physics. He's not saying he actually believes nothing ever happens: he's saying that, since the state vector alone doesn't tell you that anything happens, there must be more to QM than just the state vector.)

 

[kith]

Schwindt's argues against the notion that the only fundamental things are the state vector of the universe and its global time evolution. His position is that because we perceive the world as a collection of interacting systems, we should be able to derive the existence of these systems from the fundamental things.

So if you talk about the MWI and write down the Hamiltonian for two interacting systems, Schwindt would object "but you are not using the MWI; the MWI only has the universal state and the universal Hamiltonian and decomposing them a certain way adds something to the minimal set of assumptions which constitute the MWI".

 

[Derek Potter]

Schwindt's argues against the notion that the only fundamental things are the state vector of the universe and its global time evolution. His position is that because we perceive the world as a collection of interacting systems, we should be able to derive the existence of these systems from the fundamental things.

So if you talk about the MWI and write down the Hamiltonian for two interacting systems, Schwindt would object "but you are not using the MWI; the MWI only has the universal state and the universal Hamiltonian and decomposing them a certain way adds something to the minimal set of assumptions which constitute the MWI".

Does Schwindt say this? Last time I looked, MWI allows you to decompose the system any way you like. In fact, skimming over the words that accompany his maths, I get the impression that he first establishes that there is a decomposition in which precisely nothing happens - there is no interaction. He then finds that all decompositions share this characteristic. Remember, this thread is not about the assumptions of MWI, but about Schwindt's claim that MWI cannot have any kind of state evolution in any division of the state space.

That's an enormous claim - and since it is clear that things do happen it means that MWI cannot be viable. But I am hoping to understand how Schwindt comes to that conclusion. It wouldn't be particulary pathological if the universe as a whole doesn't evolve, that it's only within certain decompositions that time evolution is apparent. It would be odd but perhaps not too disruptive. But Schwindt seems to be claiming that what is true for the state of the universe is true for its decompositions - every single one of them. And that's what I would like to understand.

 

[Derek Potter]

I can't say for sure, but the author's thought process may go something like this: in order to distinguish different state vectors from one another, we need to pick one particular vector and then distinguish other vectors from it using the inner product. But nothing picks out any particular vector as being "the" vector we should be using in this way.

I don't think this actually justifies the statement the author makes; but I haven't read his paper in detail, only skimmed it, so I may be missing something.

(Also, it looks to me like the author's purpose in making these statements about the state vector is to argue against the claim that the state vector captures everything we need to know about "reality". In other words, to the author, the passage you quoted is intended to make you skeptical of the idea that the state vector is all we need to do quantum physics. He's not saying he actually believes nothing ever happens: he's saying that, since the state vector alone doesn't tell you that anything happens, there must be more to QM than just the state vector.)

I don't understand that reasoning. I don't see the need for MWI to identify the decompositions in which something happens, it only needs to establish that there are some in which it does. Could it be that Schwindt is falling into the trap of assuming there is a preferred decomposition? And if so, does he give a proof that it is automatically a peaceful Nirvana in which the state rotates gently but nothing else happens?

 

[craigi]

Could someone explain Jan-Markus Schwindt's idea, published under the provocative title

Nothing happens in the Universe of the Everett Interpretation​

please? Frankly I can't even understand the opening paragraph of the abstract.
http://arxiv.org/abs/1210.8447

Since the scalar product is the only internal structure of a Hilbert space, all
vectors of norm 1 are equivalent, in the sense that they form a perfect sphere in
the Hilbert space, on which every vector looks the same. The state vector of the
universe contains no information that distinguishes it from other state vectors of
the same Hilbert space.

I don't get that. The state vector has direction. In fact that's all it's got. But it's a direction relative to any arbitrary basis. I don't think I'm misinterpreting Schwindt's meaning as he goes on to identify a way of factorizing the space, where each branch "peacefully rotates" and nothing else happens. He calls these Nirvana frames.

But I'm out of my depth with tensors and I find it hard to extract the logic from Schwindt's exposition so can someone explain what it all means?

I've just skimmed it, but he's pointing out that the preferred basis problem is not solved under the WMI since there is no information available in the state vector to determine how the system should be separated into the quantum-like system and the observing system.

His contribution, seems to be a demonstratrion that the simplest separation into subsystems is of no use, implying that we need to augment the MWI and QM with some extra theory of measurement.

 

[bhobba]

Could someone explain Jan-Markus Schwindt's idea, published under the provocative title

Its the factorisation issue that decoherence only works because of factoring a system into what's observed and what's doing the observing - don't do that and nothing happens.

It an issue - but a fringe one - most don't really worry that much about it. We need some theorems to clarify the issue.

It has been discussed many many times on this forum - way beyond the interest it garners in the professional literature.

Thanks
Bill

 

[Demystifier]

We have already discussed this paper in
https://www.physicsforums.com/threads/many-worlds-proved-inconsistent.767809/
Let me just copy/paste the summary I gave there:

To define separate worlds of MWI, one needs a preferred basis, which is an old well-known problem of MWI. In modern literature, one often finds the claim that the basis problem is solved by decoherence. What J-M Schwindt points out is that decoherence is not enough. Namely, decoherence solves the basis problem only if it is already known how to split the system into subsystems (typically, the measured system and the environment). But if the state in the Hilbert space is all what exists, then such a split is not unique. Therefore, MWI claiming that state in the Hilbert space is all what exists cannot resolve the basis problem, and thus cannot define separate worlds. Period! One needs some additional structure not present in the states of the Hilbert space themselves.

As reasonable possibilities for the additional structure, he mentions observers of the Copenhagen interpretation, particles of the Bohmian interpretation, and the possibility that quantum mechanics is not fundamental at all.

 

[Derek Potter]

We have already discussed this paper in
https://www.physicsforums.com/threads/many-worlds-proved-inconsistent.767809/
Let me just copy/paste the summary I gave there:

To define separate worlds of MWI, one needs a preferred basis, which is an old well-known problem of MWI. In modern literature, one often finds the claim that the basis problem is solved by decoherence. What J-M Schwindt points out is that decoherence is not enough. Namely, decoherence solves the basis problem only if it is already known how to split the system into subsystems (typically, the measured system and the environment). But if the state in the Hilbert space is all what exists, then such a split is not unique. Therefore, MWI claiming that state in the Hilbert space is all what exists cannot resolve the basis problem, and thus cannot define separate worlds. Period! One needs some additional structure not present in the states of the Hilbert space themselves.

As reasonable possibilities for the additional structure, he mentions observers of the Copenhagen interpretation, particles of the Bohmian interpretation, and the possibility that quantum mechanics is not fundamental at all.

Why does the split need to be unique? http://plato.stanford.edu/entries/qm-manyworlds/ points out that MWI is open to various sub-interpretations and can differ from the popular “actual splitting worlds” approach in De Witt 1970 I have never seen the need for the universe to split and never seen why the metaphorical split need be unique.

After all, decoherence does provide a preferred basis in many situations but there are some in which there is no decoherence and no preferred basis! Polarization of light, for example, where the very same photon is simultaneously in a superposition of +45 and -45 degree linear polarizations; a superpostion of left and right circular polarizations and a superposition of horizontal and vertical polarizations: each component of each superposition is a world. I do not see why this is even remotely interesting - MW is Many Worlds, not Many Universes, and if the "Many" can be unpacked to mean "Many sets of Many Worlds", "Many Many" is still "Many". Some might say "Too Many Worlds" but that, I would respectfully suggest, is their problem.

Schwindt is talking about factorizing the Hilbert space, not factorizing a given wavefunction. I do not understand Schwindt's argument that, for the universe as a whole, there exists at least one factorization in which the subsystems do not interact and therefore nothing happens. I mean I literally don't understand this - it seems like a difficult argument but maybe I'm just being intimidated by the maths? It would be nice to know what exactly the assumptions are as it appears to apply only to the whole universe.

I also do not understand why, if the above is correct, it also follows that nothing happens in any factorization at all. The illusion, from within the universe, of activity whilst the real universe sits quietly doing nothing seems very satisfactory though distinctly odd.

But what does it mean that there is a factorization in which nothing happens? What are the non-interacting subsystems like?

 

[bhobba]

Why does the split need to be unique?

We need theorems showing one way or the other whether it matters or not and/or in what cases its valid.

Its only one of a number of theorems that are lacking - see for example the following where this issue (ie theorems that are lacking) is examined:
https://www.amazon.com/dp/0691004358/?tag=pfamazon01-20

Most think its only dotting the i's and crossing the t's stuff - still one never knows.

Thanks
Bill

 

[Demystifier]

I have never seen the need for the universe to split and never seen why the metaphorical split need be unique.

If the split is only metaphorical, not real and not unique, then why the actual outcomes look so real to us?

The basis in which the actual outcomes are seen are either determined by decoherence or not determined by decoherence.
If they are determined by decoherence, then, even before decoherence, we need a split into system and environment, which pure MWI does not provide.
If they are not determined by decoherence, then why do experiments show that they are?

The goal of MWI should be to explain the definite outcomes, but with such metaphorical non-unique splits I don't see how MWI can explain the definite outcomes.

 

[Derek Potter]

If the split is only metaphorical, not real and not unique, then why the actual outcomes look so real to us?

The "split" is metaphorical because the universe does not split - it is merely depicted as doing so. Metaphors always break down at some point, otherwise they would be identities not metaphors. There is no need to explain why something happens in a metaphor which has no counterpart in reality. Ther are no "actual" outcomes, there is only a superposition of observers each of whose state corresponds to a particular observation. Why do they look so real? Because in your history you have never seen a cat which is dead and alive at the same time. Thus next time you do a Schrodinger cat experiment you will enter a superposition of states, one of having seen a dead cat and one of having seen a living one. The latest experience just confirms your idea of normal reality. In one state it confirms it by adding a dead cat, in the other by adding a live cat.

The basis in which the actual outcomes are seen are either determined by decoherence or not determined by decoherence.
If they are determined by decoherence, then, even before decoherence, we need a split into system and environment, which pure MWI does not provide.

Why not? Any factorization is permissible. Where there is decoherence, a preferred basis emerges. It emerges as a result of decoherence not as a result of any structure in MWI.

If they are not determined by decoherence, then why do experiments show that they are?

Where there is no decoherence, no preferred basis emerges. Is there a preferred polarization basis? Of course not.

The goal of MWI should be to explain the definite outcomes, but with such metaphorical non-unique splits I don't see how MWI can explain the definite outcomes.

There are no definite unique outcomes in MWI. Why would you expect modern chemistry to explain the difficulty in isolating phlogiston?

 

[bhobba]

The "split" is metaphorical because the universe does not split - it is merely depicted as doing so.

Demystifer nailed it.

The issue is do all splits lead to the same result. If not decoherence as an explanation is in deep do do.

That said we don't have a proof it does depend on the factorisation

Thanks
Bill

 

[Derek Potter]

We need theorems showing one way or the other whether it matters or not and/or in what cases its valid.

Its only one of a number of theorems that are lacking - see for example the following where this issue (ie theorems that are lacking) is examined:
https://www.amazon.com/Understanding-Quantum-Mechanics-Roland-Omnès/dp/0691004358

Most think its only dotting the i's and crossing the t's stuff - still one never knows.
Thanks
Bill

Well I would agree with all that. But then the Nirvana paper that I asked about does seem to imply that "Nothing happens in the Universe of the Everett Interpretation" is a theorem (i.e. could be tidied up to be one).

 

[bhobba]

Well I would agree with all that. But then the Nirvana paper that I asked about does seem to imply that "Nothing happens in the Universe of the Everett Interpretation" is a theorem.

Even though he doesn't say it outright, he hasn't PROVEN his concern is an actual issue - merely its something that needs looking into. More research is required. But as I have mentioned its one of a number of areas that need more research with key theorems lacking.

Thanks
Bill

 

[Demystifier]

Because in your history you have never seen a cat which is dead and alive at the same time.

Can MWI explain why in my history I have never seen a cat which is dead and alive at the same time?

 

[Derek Potter]

Demystifer nailed it.
The issue is do all splits lead to the same result. If not decoherence as an explanation is in deep do do
That said we don't have a proof it does depend on the factorisation
Thanks
Bill

I thought this was elementary. Splits that support decoherence describe worlds such as the world opf Schrodinger Cats in which decoherence operates and a preferred basis emerges. Splits which do not support decoherence describe worlds in which decoherence does not operate such as polarization experiments. Clearly it would be silly to insist that decoherence is fundamental, when, in fact it depends on the factorization. This is why I say that decoherence is not fundamental to MWI, it is a welcome refinement.

 

[Derek Potter]

Can MWI explain why in my history I have never seen a cat which is dead and alive at the same time?

Yes.

 

[Demystifier]

Yes.

How can MWI explain that?

 

[Demystifier]

There are no definite unique outcomes in MWI.

But there are definite outcomes in my everyday experiences. How can MWI explain that?

 

[Derek Potter]

How can MWI explain that?

Cats decohere rapidly
Decoherence therefore applies.
Ah, I think you must have edited while I was answering - unless I simply missed half of your answer.

But there are definite outcomes in my everyday experiences. How can MWI explain that?

Because the entity you refer to as "me" and which I refer to as "you" is a member of one world and the entity you refer to as "you" and which I refer to as "me" is a member of the same world and we are both entangled with the same state of the system under observation. The weird "cat states" will not be observed very often. That's all.

 

[bhobba]

I thought this was elementary. Splits that support decoherence describe worlds such as the world opf Schrodinger Cats in which decoherence operates and a preferred basis emerges. Splits which do not support decoherence describe worlds in which decoherence does not operate such as polarization experiments. Clearly it would be silly to insist that decoherence is fundamental, when, in fact it depends on the factorization. This is why I say that decoherence is not fundamental to MWI, it is a welcome refinement.

Its not elementary - its in fact quite deep.

Conceptually its simple. Consider the particle detector in Schrodinger Cat. I have posted a very elementary argument showing if you just observe the detector then because its entangled with the emitted particle its in a mixed state and MW, or any other decoherence based explanation for that matter, handles it easy peasy. But now let's decompose it into half the particle detector and the other half + the particle; do you still get the same result? Its difficult to analyse it that way and you need rocks in your head to do it, but we are speaking of matters of principle. We do not have theorems that show you will get the same result ie the same mixed state.

Thanks
Bill

 

[bhobba]

Cats decohere rapidly Decoherence therefore applies.

You are missing the point. What if you factor it differently?

Thanks
Bill

 

[Demystifier]

You are missing the point. What if you factor it differently?

Thanks
Bill

Yes, that's exactly my point (or my question).

 

[Derek Potter]

Its not elementary - its in fact quite deep.
We do not have theorems that show you will get the same result ie the same mixed state.

You are missing the point. What if you factor it differently?
Thanks
Bill

I probably am missing the point. As I see it, MWI doesn't have proper mixed states. Factorization doesn't alter the pure state nor the improper mixed state. I can't make sense of "the same mixed state" when we are considering different factorizations which by definition means different bases and different observables.

 

[bhobba]

I probably am missing the point. As I see it, MWI doesn't have proper mixed states. Factorization doesn't alter the pure state nor the improper mixed state.

It's simple. If you factor it differently do you still get the same result ie if you do the half detector thing and the bit that flashes or clicks is in one of these halves does it still have the same state?

Thanks
Bill

 

[tzimie]

We have already discussed this paper in
https://www.physicsforums.com/threads/many-worlds-proved-inconsistent.767809/
To define separate worlds of MWI, one needs a preferred basis, which is an old well-known problem of MWI. In modern literature, one often finds the claim that the basis problem is solved by decoherence. What J-M Schwindt points out is that decoherence is not enough. Namely, decoherence solves the basis problem only if it is already known how to split the system into subsystems (typically, the measured system and the environment). But if the state in the Hilbert space is all what exists, then such a split is not unique. Therefore, MWI claiming that state in the Hilbert space is all what exists cannot resolve the basis problem, and thus cannot define separate worlds. Period! One needs some additional structure not present in the states of the Hilbert space themselves.

This problem had been solved by Max Tegmark
http://arxiv.org/abs/1401.1219

 

[bhobba]

This problem had been solved by Max Tegmark

That's his view.

I suspect most are like me and think its a load of the proverbial like a lot of Tegmarks stuff.

Thanks
Bill

 

[PeterDonis]

As I see it, MWI doesn't have proper mixed states.

The problem, according to Schwindt, is that the MWI doesn't have any states other than "the pure state of the entire universe". For example, when you say "cats decohere", you are assuming that there are "cats" picked out somewhere as identifiable quantum states. But if all we have is the pure state of the universe, there are no "cats"--or humans, or anything else. So you don't even have the structure needed to talk about "decoherence" at all.

 

[tzimie]

The problem, according to Schwindt, is that the MWI doesn't have any states other than "the pure state of the entire universe". For example, when you say "cats decohere", you are assuming that there are "cats" picked out somewhere as identifiable quantum states. But if all we have is the pure state of the universe, there are no "cats"--or humans, or anything else. So you don't even have the structure needed to talk about "decoherence" at all.

Again, please check the Max Tegmark's solution.
Please ignore the provocative title of his article.

 

[Demystifier]

This problem had been solved by Max Tegmark
http://arxiv.org/abs/1401.1219

Essentially, Tegmark proposes that the preferred basis is determined by consciousness. It could be true, but he certainly does not derive consciousness from MWI. Instead, one has to postulate consciousness as an independent assumption. I am fine with it, but this only confirms what I already said that pure MWI (i.e. MWI without any additional assumptions) cannot solve the problem. One has to assume something additional, be it Copehangen observers, Tegmark cosciousness, Bohmian trajectories, or something else.