Interpretations of Quantum Mechanics

[Fredrik]

I must say it feels good to hear reasonable voices. So I wanted to thank you.

You're welcome.

 

[sokrates]

Yes I did, but thank you for proving that it's pointless to continue this discussion.

You are welcome :). I happily debunk any pointless claim using inaccurate terminology and weak reasoning. Any time...

 

[humanino]

Does an algorithm make predictions?

All physical laws are algorithms to cook predictions.

 

[RUTA]

By the way (this is unrelated to the content of this post), I don't see why people like the "relational interpretation" so much. I know that the basic idea, that systems don't have properties in an objective sense, but rather with respect to some other system, is pretty appealing. But to me this sounds a lot more like an ingredient of a many-worlds interpretation than like a standalone interpretation. Maybe it's possible to take Everett's MWI, which really only defines the "bird's view" properly, and complete it by adding something like the "relational interpretation" to define the "frog's view".

Speaking for myself, I use a relational approach because I'm going with non-separability rather than non-locality in dealing with quantum non-locality. Mine certainly has nothing to do with MWI, so I can't speak to your point there.

 

[Fra]

By the way (this is unrelated to the content of this post), I don't see why people like the "relational interpretation" so much. I know that the basic idea, that systems don't have properties in an objective sense, but rather with respect to some other system, is pretty appealing. But to me this sounds a lot more like an ingredient of a many-worlds interpretation than like a standalone interpretation. Maybe it's possible to take Everett's MWI, which really only defines the "bird's view" properly, and complete it by adding something like the "relational interpretation" to define the "frog's view".

IMO the notion of many worlds or multiverse is simply weird. IMO, there are many-views within one world. And the differing views implies interactions between the observers. That's my "relational view".

The idea that most people have that does "relational stuff", including rovelli, DOES take place at a non-physical bird view, where the frog views are "derived" from the birds view. But that entire construction is missing the point or an intrinsic measurement theory. There is no justification for the birds view.

The birds view is nothing but a more sophisiticated structural realist fantasy.

They way I think of relational views, replaces the birds view with evolving frog views. This is also closely related to Smolins ideas of evolving law.

The type of "modern realism" I don't like comes in different disguises. "Structural realism" is also a closely related term and it's closely related to the notion of timeless universal law. Universal objective symmetries are also modern realism in disguise.

The problem, from the point of view of instrinsic inquiry is that all these elements are added to the reasoing without beeing intrinsically inferred from observations.

I'll bite on this

All physical laws are algorithms to cook predictions.

But if we look at this in the proper context, an algoritm must have a "computer" and physical resources for implementation. This is a physical scientist, and for example a receipe that is too complex to fit the scientists brain is simply worthless. So, there are constrains on the algortitms - they must be encodable and computable relative to the observer in question. This is the essence of the intrinsic view I hold.

Edit: Now ponder, what happens as the complexity of the observer scales! That's what is totally ignored in current QM.

Alot of people (it seems to me at least) think of these algorithms as existing in some external reality, or "mathematical reality". Now again, that a kind of modern realism that is not helping me.

My point is that, if you see it this way, you also see that the algoritm we call "standard QM" has the wrong form! It is not cast in the right way to be a proper intrinsic algorithm (to use your words).

/Fredrik

 

[div78football]

Before i start, i nothing of physics other than what I've seen in sci-fi.

Is it believed that universes are created through the collision with other universes? and that this created universe assumingly expands to ultimately contract till it implodes on itself unless of course it itself collides with another universe.

Further does the rate of expansion effect the outcome of a collision...ie does a rapidy expanding universe in collision with a slowly extracting universe create something different from a slowly expanding universe with a slowly contarcting universe or any permintation of the two.

Also if there are multiverses does the outcome of a collision of many of them create something totally different.

Cheers

 

[Count Iblis]

Are you calling me the Devil?

But seriously, do you have a citation for this interpretation -- proponents, opponents, discussion, etc?

When 't Hooft first proposed the sort of deterministic models he is working on, he was thinking along these lines, see e.g. here:

http://arxiv.org/abs/gr-qc/9903084

But his models are completely deterministic and local, so what he is proposing amounts to a local deterministic hidden variable models and a necessary ingredient in these models has to be that they are superdeterministic (i.e. the observer has no freedom to choose how to set up his experiments; the fact that he/she is deterministic too cannot be ignored). In later papers 't Hooft has argued why superdeterminism is a natural feature of any deterministic model:

http://arxiv.org/abs/quant-ph/0701097

 

[Demystifier]

When 't Hooft first proposed the sort of deterministic models he is working on, he was thinking along these lines, see e.g. here:

http://arxiv.org/abs/gr-qc/9903084

But his models are completely deterministic and local, so what he is proposing amounts to a local deterministic hidden variable models and a necessary ingredient in these models has to be that they are superdeterministic (i.e. the observer has no freedom to choose how to set up his experiments; the fact that he/she is deterministic too cannot be ignored). In later papers 't Hooft has argued why superdeterminism is a natural feature of any deterministic model:

http://arxiv.org/abs/quant-ph/0701097

That is an interesting idea that superdeterminism could save locality. However, even though local superdeterminism can explain existing experiments, I think that local superdeterminism cannot reproduce ALL conceivable predictions of quantum mechanics. For example, the principles of quantum mechanics allow entangled states of systems that have NEVER been in local interaction, in which case there is no local mechanism that could provide instructions for non-free experimenters how to set up their experiments in order to get quantum-like correlations.

Anyway, Bohmian mechanics assumes that everything, including the human braines and arms, is described by deterministic equations of motion. Hence, it is not only deterministic, but also superdeterministic. Yet, it is a nonlocal theory.

 

[RUTA]

When 't Hooft first proposed the sort of deterministic models he is working on, he was thinking along these lines, see e.g. here:

http://arxiv.org/abs/gr-qc/9903084

But his models are completely deterministic and local, so what he is proposing amounts to a local deterministic hidden variable models and a necessary ingredient in these models has to be that they are superdeterministic (i.e. the observer has no freedom to choose how to set up his experiments; the fact that he/she is deterministic too cannot be ignored). In later papers 't Hooft has argued why superdeterminism is a natural feature of any deterministic model:

http://arxiv.org/abs/quant-ph/0701097

Thanks very much. I do find the claim that particles are not real but I don't find any mention of fields as mere computational devices in the path integral approach in these papers. Do I need to look at his earlier work to find this and the fact that only external lines are "real" in the computation of amplitudes?

I ask because I'm writing a paper wherein I use discrete path integrals over graphs as a model fundamental to quantum physics and the mathematical approach I'm using captures essentially these same claims, so I need to cite him if he's already made these claims in his approach. Thanks again.

 

[fleem]

If I develop a theory to describe the average behavior of a hive of bees (that is, a theory to describe how the swarm acts), and move on to study a single bee, I might see some similarities in the behavior of a single bee and the behavior of the swarm. But it would be illogical of me to insist that my swarm theory must be used as the founding theory in understanding the behavior of a single bee.

But that is exactly what we've done. "local" means a zero-length space-time interval, and "non-local" means a non-zero-length space-time interval. But "space-time" was learned and developed solely to describe the (classical) large scale, average behavior of many particles. By insisting particle-particle interactions occur in something called "space-time", we've found apparent paradoxes (entangled particles disobey the tenants of space-time). Yet even after seeing those paradoxes, we still insist space-time is the right foundation to understanding individual particle behavior! The fact that entangled particles clearly and blatantly disobey the tenants of space-time should make us realize we have no business insisting a theory learned solely from classical observation must unequivocally be the founding theory describing the behavior of individual particles. Yes, if we watch several particles interacting, we see things similar to the tenants of space-time, like the concepts of dimension (certain conserved quantities are orthogonal and can be transformed by processes similar to euclidean rotation), intervals, etc. But then several particles is a bit of a swarm, and thus a bit classical.

Entangled particles act like they are local because they are just as local as particles we classically consider local. That is, they are in the same situation as classically local particles--whatever that might be. The clincher seems to be that the particles are entangled only while there is no interaction with the rest of the universe. (Classically) local particles find it easier to behave that way because they need not remain closed as long, because they are so close. So its really a matter of closed and open, not local and non-local.

 

[alexepascual]

I have one problem with MWI, though I accept I may be looking at it too literally. The original MWI was devised to explain the double slit experiment conundrum. So the one photon goes through each slit but in different worlds. Not to upset causality or the conservation of mass or energy, each world only sees one photon – no problem. However now each photon is in a parallel world (universe), why does interference happen?

The answer to this might depend on what you choose to think of as a separate world. Usually those who adheer to the many-worlds interpretation would consider the worlds "splitting" only after wave function collapse. If we choose this definition, then when the photon goes through the slits, we could say that it has split into several copies of itself, each being "less than real" (we could call them ghost particles). But each of these copies is still within the same "world" and that's the reason you can see interference.
The other approach would be to consider each "copy" of the photon to be in a different world. But in that case we would have to consider these worlds as being part of "our world". So you could say that it is the superposition of these different worlds that make up your world. But as I said before, I don't think this is the standard way of thinking about it (which nevertheless does not mean that it has less merit or potential than the more standard MWI thinking. It seems that this is only a different use of words, but some times trivial things like the use of words can have some influence in our thinking process. So I wouldn't dismiss it as irrelevant.

 

[Adrian59]

Usually those who adheer to the many-worlds interpretation would consider the worlds "splitting" only after wave function collapse. If we choose this definition, then when the photon goes through the slits, we could say that it has split into several copies of itself, each being "less than real" (we could call them ghost particles). But each of these copies is still within the same "world" and that's the reason you can see interference.
The other approach would be to consider each "copy" of the photon to be in a different world. But in that case we would have to consider these worlds as being part of "our world". So you could say that it is the superposition of these different worlds that make up your world... It seems that this is only a different use of words, but some times trivial things like the use of words can have some influence in our thinking process.

Thanks for the reply alexepascual since I thought my comment had been lost in the general discussion, hence my delay in responding. However I still think that the MWI doesn't do what it set out do do since in your first suggestion one has to insert collapsing waves which obviously gets us back to the Copenhagen interpretation. Also, I can see no difference between ghost waves & probability waves which I thought the MWI was trying to avoid. The second suggestion is that the parallel worlds are not truly separate but different states that result in a superpostion but this also seems like back to square one & probability waves.

 

[Dmitry67]

No, modern MWI does not include any collapse at all.
'Splitting' is explained via Quantum Decoherence , so no additional postulates are needed.
http://en.wikipedia.org/wiki/Quantum_decoherence

This is a beauty of MWI - no mysterious collapse.

 

[Albert V]

I would argue for "shut up and calculate!"


Two recent papers show the relationship between mathematical undecidable propositions and randomness.

View attachment Brukner01.pdf

View attachment Zeilinger01.pdf

 

[Dmitry67]

There are many misconceptions about the undecidable statements in physics

For example, well known Conway Game of Life
Rules are simple and deterministic
And yet there are undecidable statements about the outcomes!

 

[Adrian59]

No, modern MWI does not include any collapse at all.
'Splitting' is explained via Quantum Decoherence , so no additional postulates are needed.
http://en.wikipedia.org/wiki/Quantum_decoherence

This is a beauty of MWI - no mysterious collapse.

Thanks for the reply. I would agree with your dislike of collapsing waves though I still need convincing that MWI is the best interpretation. I followed your link to wiki. The paragraph using phase space explained everything to me & the following maths only confirmed this since the first equation seemed to encapsulate the whole argument because the system ket was expanded in basis states of the environment <i |ψ>. Though on a more general point, I understood that decoherence can be considered an interpretation of QM in its own right so what MWI adds is questionable to me. Or put another way, I see it as a minimizing the adversities: does MWI give enough to the interpretation of QM in return for having to put up with infinite parallel worlds.

 

[RUTA]

Thanks for the reply. I would agree with your dislike of collapsing waves though I still need convincing that MWI is the best interpretation. I followed your link to wiki. The paragraph using phase space explained everything to me & the following maths only confirmed this since the first equation seemed to encapsulate the whole argument because the system ket was expanded in basis states of the environment <i |ψ>. Though on a more general point, I understood that decoherence can be considered an interpretation of QM in its own right so what MWI adds is questionable to me. Or put another way, I see it as a minimizing the adversities: does MWI give enough to the interpretation of QM in return for having to put up with infinite parallel worlds.

Adrian Kent gave an interesting talk at Perimeter last Sep, "Theory Confirmation in One World and its Failure in Many." In this talk he pointed out that, if MWI is true, one cannot be sure what the true probablility is for any given experiment because one doesn't know where they reside in the multiverse. Consider, for example, a simple binary outcome (call it + or -) with a "true" probability of 50% either way. If MWI is correct, there is a world (call it A) in which the experiment always comes out + and a world (call it B) in which the outcome is always -, with all other combinations realized between these extremes. Only those worlds "in the middle" would be able to deduce something close to the "true" probability, but they wouldn't "know" it. Consider two poor theorists in worlds A and B who deduce the correct theory of QM and predict the "true" 50% probability distribution. Their "correct" theories are refuted and they die in poverty and disgrace To what extent can we do science in a multiverse?

 

[Dmitry67]

RUTA, yes, exactly, it is a problem.
We discuss it here: https://www.physicsforums.com/showthread.php?t=329807
How does MWI resolve low probability multiverse events?

 

[Dmitry67]

Thanks for the reply. I would agree with your dislike of collapsing waves though I still need convincing that MWI is the best interpretation. I followed your link to wiki. The paragraph using phase space explained everything to me & the following maths only confirmed this since the first equation seemed to encapsulate the whole argument because the system ket was expanded in basis states of the environment <i |ψ>. Though on a more general point, I understood that decoherence can be considered an interpretation of QM in its own right so what MWI adds is questionable to me. Or put another way, I see it as a minimizing the adversities: does MWI give enough to the interpretation of QM in return for having to put up with infinite parallel worlds.

What is interesting is that Quantum Decoherence (QD) is mathematical fact and can not be denied, no matter if one likes or hates MWI. So Copenhagen Int. now have not one, but 2 different 'agents' for collapse: old CI collapse and New QD. Which makes CI a total mess. Well, it was quite a mess even before the discovery of QD.

Regarding your last statement, if sounds like you see the existence of another branches as something 'bad'. People tend to critisize MWI for having extra worlds using occams razor argument. But on the contrary, occams razor must be used to justify the elimination of some worlds based on the random or unknown rule!

Finally, there is still one problem with MWI: Born rule. Based on the discussions here I see that Born rule must be reformulated to be used in MWI, because there is no such thing in MWI as 'probability', because it is deterministic. But I can not answer the question what is a physical meaning of the what we used to call a probability in MWI. Or, using RUTA's example, why are we observing a world 'in the middle'

 

[kote]

I would argue for "shut up and calculate!"Two recent papers show the relationship between mathematical undecidable propositions and randomness.

View attachment 20068

View attachment 20069

I just looked at these papers... and I'm sorry, but they are just completely wrong. Analytic propositions can't be proven a posteriori. Experiments can't serve as mathematical proofs.

Or maybe I missed the point of the papers.

 

[Fra]

What is interesting is that Quantum Decoherence (QD) is mathematical fact and can not be denied, no matter if one likes or hates MWI. So Copenhagen Int. now have not one, but 2 different 'agents' for collapse: old CI collapse and New QD. Which makes CI a total mess. Well, it was quite a mess even before the discovery of QD.

I don't agree with this argument for CI beeing a mess.

The collapse vs decoherence are simply two different views. The inside view, vs the external view. So there is no surprise they are different. Note that in the decoherence view, there is an imagine massive observer. Sometimes you can say that the entire environment is an observer, or that there is a birds view.

Now if I have to choose between a birds view and a collapse, then it's easy. The birds view is IMHO a non-physical view. The collapse OTOH is a physical inside view where there is a rational revision of the observer state as a response to feedback from the environment (=new information).

SO IMO, docoherence and collapse are not competing against each others, and I see no contradiction at this level of discusssion.

/Fredrik

 

[Adrian59]

Regarding your last statement, if sounds like you see the existence of another branches as something 'bad'. People tend to critisize MWI for having extra worlds using occams razor argument. But on the contrary, occams razor must be used to justify the elimination of some worlds based on the random or unknown rule!

I wouldn't say that MWI was "bad" for postulating other universes just that, as there is no experimental evidence of any universe other than the one we're in, so MWI remains pure conjecture. Of course this may all change if the LHC shows evidence for hidden dimensions / universes (if there is a way to know the difference). However, one has to be compatable with the current evidence base & as such MWI is just a theoretical device. So for that reason alone, I am not in favour any QM interpretation that adds unobserved elements to nature, eg MWI, backwards transmitted waves (if this still is a contender).

 

[Demystifier]

I don't agree with this argument for CI beeing a mess.

The collapse vs decoherence are simply two different views. The inside view, vs the external view. So there is no surprise they are different. Note that in the decoherence view, there is an imagine massive observer. Sometimes you can say that the entire environment is an observer, or that there is a birds view.

Now if I have to choose between a birds view and a collapse, then it's easy. The birds view is IMHO a non-physical view. The collapse OTOH is a physical inside view where there is a rational revision of the observer state as a response to feedback from the environment (=new information).

SO IMO, docoherence and collapse are not competing against each others, and I see no contradiction at this level of discusssion.

First, decoherence is an experimental fact. Thus, it is not only an external view, but an inside view as well.

Second, decoherence is not a substitute for the collapse. Decoherence alone (which is both theoretically well understood and experimentally confirmed) cannot explain the appearance of collapse. Decoherence can only determine the preferred basis in which the collapse will appear, but cannot explain the collapse itself.

 

[Dmitry67]

Decoherence alone (which is both theoretically well understood and experimentally confirmed) cannot explain the appearance of collapse. Decoherence can only determine the preferred basis in which the collapse will appear, but cannot explain the collapse itself.

On the contrary, given the basis (arbitrary chosen) QD explains what is 'observed' by the system (the division of the Universe into systems is also an input). I agree that it is still recursive, but at the moment this is the best we can do. Why?

1. MWI+QD gets rids of any mysterious 'collapse agents'
2. From bird's view, world is Objective and Deterministic.

The problems starts when we try to go to the frog's view, to explain, "what is observed". But "What is observed" is not a complete question, the question have a free parameter: "What is observed by system(X)"? So you can not ask that question without specifying the X.

But by specifying X you divide the whole Universe into X and not-X and define that basis. I agree that there are sill some problems, because the state of system X is different in each sub-branch, so non only the notion of branch is basis-dependent, but also we need to redefine the basis after each act of decoherence. So ti is recursive.

But the *fundamental laws* of QM (in birds view) are not recursive

 

[karatemonkey]

I'm partial to the Quantum Bayesians. (See work by Christopher Fuchs et al on arxiv.org).
I'm not sure I understand it all but, the basic tenets make things like Bell's Theorem much less bizarre. The tenets I am speaking of are use of prior and post probabilities. In general, I like the idea that it's information and one's use of that information that is being transferred in Quantum Mechanical interactions. In the present case, information is a fundamental physical property.

 

[Fra]

To demystifier: I don't deny decoherence as such. My comments was addressed to Dmitry's reasoning with bird views. It's also true that there are decoherence like processes also in inside views, but I think that is somehow beside the point I tried to make in relation to Dmitry's reasoning. Also to me the collapse is simply a information update, there is no need for further explanation beyond that IMO. Except of course to elaborate what information is, and how it relates to matter and how information processes in nature relate to the standard model interactions. I don't have those answers though :)

But the *fundamental laws* of QM (in birds view) are not recursive

But wouldn't you agree that the process of inferring these laws from experience / interaction history, by means of a scientific method is recursive?

This is one of my objection to the birds views, and other forms of realism. There are modern forms of realism that are no like the old classical realism, but are often called structural realism, implicit often in the "reality of timeless laws".

IMO, I don't find these kind of birds views of realism consistent with beeing inferred from a scientific process. And IMO, a scientific process has a lot in common with any physical process (ie. measurement process).

/Fredrik

 

[Fra]

I'm partial to the Quantum Bayesians. (See work by Christopher Fuchs et al on arxiv.org).
I'm not sure I understand it all but, the basic tenets make things like Bell's Theorem much less bizarre. The tenets I am speaking of are use of prior and post probabilities. In general, I like the idea that it's information and one's use of that information that is being transferred in Quantum Mechanical interactions. In the present case, information is a fundamental physical property.

I'm reasonably close the bayesian view, but I think I'm even more radical, becuase a central thing in interence is the choice of rule for information the update. There are several ways here, bayes rule, but also various maxent rules with various choices of entropy. I think the rule for the information update is evolving, and it's exactly this that's subject to evolution.

In my view the question becomes, why are certain inference rules preferred over others? I think an inference rule can only be justfied ina context. Different contexts favours different rules of inference.

This also means I do not think there exists an objective, universal information measure.

/Fredrik

 

[Dmitry67]

But wouldn't you agree that the process of inferring these laws from experience / interaction history, by means of a scientific method is recursive?
/Fredrik

Yes, there is a minimum level of complexity of a system to be able to ask 'how system X observes the Universe'. You can not apply this questions to atoms, molecules, of very small systems. So while BIRD is Universum we don't need to define (and there is only one bird), FROG must be defined, and FROG is fuzzy, complicated system built of billions parts.

 

[Fra]

Yes, there is a minimum level of complexity of a system to be able to ask 'how system X observes the Universe'. You can not apply this questions to atoms, molecules, of very small systems.

In my view you can, but the question is exactly what happens to this inference process and the abstractions as the host system complexity scales down.

My conjecture is that as you go from incredibly complex systems, to simple ones, the inference processes formulated relative to the small systems, becomes the elementary physical interactions.

And unification of interactions, can be thought of in this picture as the indistinguishability between the interaction types, as the host systems scales to the most simple.

So while BIRD is Universum we don't need to define (and there is only one bird), FROG must be defined, and FROG is fuzzy, complicated system built of billions parts.

A given frog, do not need to define exactly all other frogs. From the poitn of view of a frog the prime decomposition is the self, and the environment. In the environment there are several possible decompositions of "other frogs", the preferred picture here is part of the game, and no a priori sharp definition is needed to play this game.

If you just look at nature, biology, human science, physical equilibrations, cosmology etc is it not in fact quite plasuble that the best description of nature and how it works IS in fact "recursive" as you put it? I call it evolving but that's just another word for the same thing.

/Fredrik

 

[kote]

If you just look at nature, biology, human science, physical equilibrations, cosmology etc is it not in fact quite plasuble that the best description of nature and how it works IS in fact "recursive" as you put it? I call it evolving but that's just another word for the same thing. /Fredrik

Fra,

Your post reminds me very much of David Bohm's view of QM. Check out: http://evans-experientialism.freewebspace.com/bohmphysics.htm ( Bohm, D. (1990) `A New Theory of the Relationship of Mind and Matter', Philosophical Psychology 3: 271--86. )

Highlights:

We have thus been led to an extension of the notion of implicate order, in which we have a series of inter-related levels in which the more subtle-I.e. 'the more finely woven' levels including thought, feeling and physical reactions-both unfold and enfold those that are less subtle (i.e. 'more coarsely woven'). In this series, the mental side corresponds, of course, to what is more subtle and the physical side to what is less subtle. And each mental side in turn becomes a physical side as we move in the direction of greater subtlety.
...
Finally, we may ask how we can understand this theory if the subtle levels are carried to infinity. Does the goal of comprehension constantly recede as we try to do this? I suggest that the appearance of such a recession is in essence just a feature of our language, which tends to give too much emphasis to the analytic side of our thought processes.
...
But the deeper reality is something beyond either mind or matter, both of which are only aspects that serve as terms for analysis [1].​