GR/StatMech/QM foundations, epistemic views only please

[DennisN]

Some 2014 Edge links:
Gefter http://www.edge.org/response-detail/25513 (retire *the* universe, i.e. a unique account of reality)
Zeilinger quantums is too about real stuff! http://www.edge.org/response-detail/25548
Rovelli retire geometry-as-the-description-of-physical-space http://www.edge.org/response-detail/25345
Giddings retire spacetime-as-fundamental http://www.edge.org/response-detail/25477

Some comments from me on the interesting essays:

Gefter on "The Universe":

In light of the universe's retirement, this all looks slightly less miraculous. After all, superpositions are really superpositions of reference frames. In any single reference frame, an animal's vitals are well defined. Cats are only alive and dead when you try to piece together multiple frames under the false assumption that they're all part of the same universe. (my bolding)

I don't get this at all. Is this the Many-Worlds Interpretation in disguise? If so, why didn't she just say MWI? And I would be very interested to hear what the frame of reference of a photon looks like... Nah, I don't think the Universe is close to retirement just yet. We need more observational/experimental data that support multi-hypotheses, I think .

Anton Zeilinger on "Reality":
I liked his essay.

But you don't have to go so far. It is enough to assume that quantum mechanics just describes probabilities of possible measurement results. Then making an observation turns potentiality into actuality and, in our case, the position of the particle becomes a quantity one can talk reasonably about. But, whether it has a well-defined position or not, the buckyball very well exists. It is real in the double-slit experiment, even when it is impossible to assign its position a well-defined value. (my bolding)

It's hard to argue against that.

Rovelli on "Geometry": I don't know enough about it to comment on it.

Giddings on "Spacetime":

The problem with fundamental spacetime is even more strongly hinted at from multiple developing perspectives. Notable among these hints is the physics of black holes, where it appears that evolution that respects quantum principles must violate the classical spacetime dictum that information does not propagate faster than the speed of light. (my bolding)

I don't understand/know anything about this. Anyone know what he is referring to?

In the debate at this year's FQXI conference (Jan 5-10, Puerto Rico) Raphael Bousso argued for Loop Quantum Gravity and skillfully attacked String Theory. His debate adversary, Carlo Rovelli, stubbornly defended String Theory and criticized LQG with zeal and gusto.

Very funny, thanks!

 

[strangerep]

In light of the universe's retirement, this all looks slightly less miraculous. After all, superpositions are really superpositions of reference frames.In any single reference frame, an animal's vitals are well defined. Cats are only alive and dead when you try to piece together multiple frames under the false assumption that they're all part of the same universe. (my bolding)

I don't get this at all. Is this the Many-Worlds Interpretation in disguise? If so, why didn't she just say MWI? And I would be very interested to hear what the frame of reference of a photon looks like... [...]

I see I'm not the only one who perceives that particular paragraph of Gefter's as poop on the carpet.

 

[atyy]

@strangerep, you have to consider that some people may read your comments while having dinner

 

[DennisN]

I see I'm not the only one who perceives that particular paragraph of Gefter's as poop on the carpet.

Our noses seem to have detected that something was not right on that carpet...

Regarding my previous post #36:

"

The problem with fundamental spacetime is even more strongly hinted at from multiple developing perspectives. Notable among these hints is the physics of black holes, where it appears that evolution that respects quantum principles must violate the classical spacetime dictum that information does not propagate faster than the speed of light. (my bolding)

I don't understand/know anything about this. Anyone know what he is referring to?"

Clarification:

I suspect it has something to do with the recent black hole research/debates considering firewalls, complementarity and unitarity, but I'm not aware of what Giddings exactly meant... nevermind, I won't hi-jack the thread with it - if anyone knows, please PM me .

 

[RUTA]

I should clarify the term epistemic by generalizing the blue highlight in post #2.

" A central issue is whether any mathematical model describes reality (the ontic view) or an agent's knowledge of reality (the epistemic view)."

The original highlighted quote from Mermin merely applied to "a quantum state" and I'm extending it to include classical states and physical models in general.

This discussion is a good idea, marcus. By exploring what we believe physics represents, we can consider changes to the way we understand the formalism. This can lead to entirely new approaches to physics.

1. The above quote can be understood as a false dichotomy. How can you argue that our models don't represent what we know (at least what we think we know, based on our experiments and observations)? So, that "any mathematical model describes ... an agent's knowledge of reality" holds regardless of whether or not you define reality ontologically by that which is represented by (or tacitly included in) your models. You may of course always assume reality contains more than is represented by your models.

2. Ontology can't be done without making epistemological assumptions and vice-versa.

Physics is a game of reconciling disparate perceptions. All I have to go on concerning your perceptions is what you tell me you perceive, and you telling me what you perceive is a subset of my perceptions, so I start playing the game by verifying what you tell me. For example, if you tell me you saw a pink elephant in my dining room, I will have to go into the dining room and see a pink elephant or I won't have a perception consistent with yours to reconcile. Another example, I see a cereal box between us and we tell each other what we see on our side of the box. I can then rotate the box 180 deg and verify what you reported. You then tell me that you see what I reported to you. I can then construct a model of you, me and a two-sided box between us. Most people reify this model calling it "reality" and assume equal ontic status of self, other, and box.

Where QBism/CBism deviates from business as usual is to acknowledge that the nature of the game is in fact the reconciliation of perceptions (knowledge). And, to question whether or not this reconciliation requires a unique model, e.g., self, other, and box, or a unique spacetime manifold. Thus, there is perhaps no unique model possible in the reconciliation even though there is a mathematical formalism that links the disparate perceptions. It sounds like some here assume that there exists a unique, underlying reality responsible for the disparate set of perceptions. What does this gain you by way of theory construction? If you succeed in constructing a model of this unique, underlying reality, you will have violated a central premise of QBism/CBism as I understand it. Am I missing something?

 

[marcus]

Excellent summary RUTA! I appreciate this both because it helps me personally to see the issues and because it contributes substantially to thread. I will excerpt:

... By exploring what we believe physics represents, we can consider changes to the way we understand the formalism. This can lead to entirely new approaches to physics.

==quote==
Physics is a game of reconciling disparate perceptions. ...
Where QBism/CBism deviates from business as usual is to acknowledge that the nature of the game is in fact the reconciliation of perceptions (knowledge). And, to question whether or not this reconciliation requires a unique model, e.g., self, other, and box, or a unique spacetime manifold. Thus, there is perhaps no unique model possible in the reconciliation even though there is a mathematical formalism that links the disparate perceptions. It sounds like some here assume that there exists a unique, underlying reality responsible for the disparate set of perceptions. What does this gain you by way of theory construction? If you succeed in constructing a model of this unique, underlying reality, you will have violated a central premise of QBism/CBism as I understand it. Am I missing something?
==endquote==

I think this misses nothing, but I would put the epistemic viewpoint a bit more softly. Yes, go right ahead and try to find a "unique underlying reality" which somehow gives rise to all these disparate perceptions. Do try to construct a math model of this "U.U.R."! Just don't be too upset if you cannot find it. We have not been promised by Nature that a single official account of what is really happening can be obtained.
Maybe it is, as you said, a game of reconciling disparate accounts without there being a unique correct one.

So it seems that someone like Mermin is saying (reinforced with all his quotes from Bohr and Schroedinger and others) that we should simply regard things like QM (and generel, specialrel, thermodynamics) and mathematical tools that an individual can use to organize his/her experience. Then all the paradoxes go away. As long as each observer has his/her own hilbert space, what's the problem?

 

[strangerep]

So it seems that someone like Mermin is saying (reinforced with all his quotes from Bohr and Schroedinger and others) that we should simply regard things like QM (and generel, specialrel, thermodynamics) and mathematical tools that an individual can use to organize his/her experience. Then all the paradoxes go away. As long as each observer has his/her own hilbert space, what's the problem?

ISTM, the tricky bit starts when one confronts the questions of how to model interactions in detail. After two (initially uncorrelated) systems have interacted, they are typically correlated (to some extent) with each other. Although one might start with (say) a tensor product Hilbert space, the interaction term must mix up the component spaces to establish the correlation, and... we're back to the notion of a single larger Hilbert space. But maybe that's ok, since Hilbert spaces are abstract.

In Gielen+Wise observer space, I see some ideas reminiscent of the decades-old attempts at many-time relativity, which now seems regarded as something of a fringe subject. Trump & Schieve wrote a book on that (which I feel lukewarm towards), but I don't detect much recent progress.

 

[marcus]

Just to clarify one point in what I said here,
==quote==
So it seems that someone like Mermin is saying (reinforced with all his quotes from Bohr and Schroedinger and others) that we should simply regard things like QM (and generel, speciarel, thermodynamics) as mathematical tools that an individual can use to organize his/her experience. Then all the paradoxes go away. As long as each observer has his/her own hilbert space, what's the problem?
==endquote==
I did not mean that each observer has a personal hilbert describing his OWN STATE, he has a hilbert to represent his knowledge and experience of the world around him including other observers (which he views as quantum entities). So when you have two observers A and B, there is no possibility to form the tensor product (then a transcendent being would be looking down at both of them). You simply have H_A = what classic A knows about quantum B and the rest of the world, and H_B = what classic B knows about quantum A and the rest of the world.

Each hilbert H_O is associated with a specified observer and is used by that observer to organize and understand that person's experience of the entire rest of the world, and calculate odds, keep records, make predictions, etc.
So it would not make sense, as I see it, to tensor two hilbert spaces together. If they are separate versions of qM that separate agents are using as conceptual/math devices to understand the world. Hence I don't grasp the import of this part of your post, or the context.

... Although one might start with (say) a tensor product Hilbert space, the interaction term must mix up the component spaces to establish the correlation, and... we're back to the notion of a single larger Hilbert space. ...

What am I missing about your intent?

 

[marcus]

Since we've turned a page, I'll bring forward a kind of thematic post as a reminder of what's involved in epistemic approaches such as QB or RQM…
====quote post#27====
... a reminder about the basic reason epistemic approaches like "QB" and "RQM" avoid various puzzles and pitfalls is that although there is a basic reality that all observers observe and although they can RECONCILE differences in the accounts of it different observers arrive at, they do have different accounts.
There is no one single official story.

To take a trivial illustration: Alice considers herself an "observer" and Bob as a quantum system, so her account of reality is obviously different from that of Bob, who considers himself the observer and Allce part of the quantum world he is trying to understand. Both use Quantum Mechanics to organize their experience and inevitably their stories differ, but yet can be reconciled.
Special rel gives us something analogous where observers can disagree about the order in which events occurred, and yet the disagreements can be explained by the observers' relative motion.

So there is a clear and serious cost. I think we all here realize this. Just for extra clarity I quote some excerpts from Mermin's recent writings and from the 2006 paper "Relational EPR".

==google "mermin now arxiv" ==
In a Physics Today Commentary, and more carefully, extensively, and convincingly with Chris Fuchs and Ruediger Schack, I argued that stubborn longstanding problems in the interpretation of quantum mechanics fade away if one takes literally Niels Bohr’s dictum that the purpose of science is not to reveal “the real essence of the phenomena” but to find “relations between the manifold aspects of our experience.” Here I note that the view of science as a tool that each of us can use to organize our own personal experience, called QBism by Fuchs and Schack, clarifies more than just quantum foundational problems. Recognizing that science is about the subject (the user of science) and not just about the object (the world external to that user) can eliminate well entrenched confusion in classical physics too.
==endquote==​

==google "relational EPR" ==
The relational approach claims that a number of confusing puzzles raised by Quantum Mechanics (QM) result from the unjustified use of the notion of objective, absolute, ‘state’ of a physical system, or from the notion of absolute, real, ‘event’.
The way out from the confusion suggested by RQM consists in acknowledging that different observers can give different accounts of the actuality of the same physical property [6]. This fact implies that the occurrence of an event is not something absolutely real or not, but it is only real in relation to a specific observer. Notice that, in this context, an observer can be any physical system.

Thus, the central idea of RQM is to apply Bohr and Heisenberg’s key intuition that “no phenomenon is a phenomenon until it is an observed phenomenon” to each observer independently. This description of physical reality, though fundamentally fragmented, is assumed in RQM to be the best possible one, i.e. to be complete [6]:

“Quantum mechanics is a theory about the physical description of physical systems relative to other systems, and this is a complete description of the world”.
==endquote==​

My comment: Note that "complete" here means best possible. The completest description we can hope for,
admittedly fragmented into versions whose differences are, however, explainable. It means giving up on the hope for one single official account as seen from transcendent perspective.
====endquote====

 

[strangerep]

What am I missing about your intent?

Evidently it was I who was missing something about your intent.

[...]
I did not mean that each observer has a personal hilbert describing his OWN STATE, he has a hilbert to represent his knowledge and experience of the world around him including other observers (which he views as quantum entities). So when you have two observers A and B, there is no possibility to form the tensor product (then a transcendent being would be looking down at both of them). You simply have H_A = what classic A knows about quantum B and the rest of the world, and H_B = what classic B knows about quantum A and the rest of the world.

Each hilbert H_O is associated with a specified observer and is used by that observer to organize and understand that person's experience of the entire rest of the world, and calculate odds, keep records, make predictions, etc.
[...]

The basic question still remains though: how to express the establishment of consistent correlations across these Hilbert space? And how to model interactions between the observers within such a framework?

Edit: Since we're discussing some of Mermin's views, one should probably (re-)read his older paper: What is quantum mechanics trying to tell us?, especially the "SSC Theorem: Subsystem correlations determine the state".

(But not you, Atty -- I'm sure this paper would just annoy you. )

 

[marcus]

...
The basic question still remains though: how to express the establishment of consistent correlations across these Hilbert space? And how to model interactions between the observers within such a framework?

Heh, heh well I'm not an official spokesperson for RQM but let me just tell you my immediate reaction. In the RQM picture AFAICS there ARE no interactions between observers, as such. No need to model.

This view is taken in the original paper "Relational EPR": A considers herself observer and B a quantum system. The interaction between A and B is modeled in her H_A version of QM.
It is not the "interaction between two observers".
It is the interaction between observer A and part of the quantum world, namely B and his instruments.

She can radio to B and ask what spin he observed and he can report, this all takes time and this ensures no instant "action at distance".

I expect you read "relational EPR" so this is familiar.

Conversely B considers himself an observer and A as a quantum system. So he sees the B interaction with A NOT as interaction between two observers but between himself B the observer, and a quantum system A.

there is no need to model the interaction because it is ALREADY MODELED by each person individually

I think I already sketched this earlier in post#27 so I am just repeating in more detail. It's how EPR is resolved in the 2006 Smerlak Rovelli paper, and how locality is confirmed.

...
There is no one single official story.

To take a trivial illustration: Alice considers herself an "observer" and Bob as a quantum system, so her account of reality is obviously different from that of Bob, who considers himself the observer and Allce part of the quantum world he is trying to understand. Both use Quantum Mechanics to organize their experience and inevitably their stories differ, but yet can be reconciled.
Special rel gives us something analogous where observers can disagree about the order in which events occurred, and yet the disagreements can be explained by the observers' relative motion.
...

 

[strangerep]

(I edited my previous post while you were typing.)

Heh, heh well I'm not an official spokesperson for RQM but let me just tell you my immediate reaction. In the RQM picture AFAICS there ARE no interactions between observers, as such. No need to model.

Ah, but I meant in the generalized sense: correlations must somehow be established between them. I wasn't thinking "action at a distance".

 

[marcus]

... correlations must somehow be established between them. I wasn't thinking "action at a distance".

I probably was editing too while YOU were writing, or editing It is a liability of lively discussions. Maybe I will let the dust settle and come back in the morning. What time zone are you in? I am in PST (pacific)

Not sure what you mean by correlations must be established… In whose quantum theory? Is there a third observer? Someone who sees both A and B as quantum systems?

Hopefully RUTA will help clarify. He's a prof at a university in PA and has a research interest in this kind of thing. But it is nearly 11 PM here so it will be approaching 2AM in PA, won't hear from him until tomorrow I guess.

 

[Paulibus]

May I ask whether Qbism is all about multiple observers who concoct multiple descriptions of
quantum systems, differing among themselves because the systems observed include other observers with their differing perceptions? Is why Qbism is called 'subjective’?

Or if instead Qbism is a description of future 'reality' that involves personal guesses, rendered
quantitative and respectable by a calculated ‘probability’, making this the reason for calling Qbism
'subjective’ --- since ‘probability’ is a word that links the known past and present to the
unknown and mysterious future?

I suspect the latter, because a ‘probable’ event is something which may be expected to happen in
the future, rather than linking multiple past or current perceptions.

 

[atyy]

I suspect it has something to do with the recent black hole research/debates considering firewalls, complementarity and unitarity, but I'm not aware of what Giddings exactly meant... nevermind, I won't hi-jack the thread with it - if anyone knows, please PM me .

I'm not sure if the inability to signal classical information faster than light is exactly the same as the notion of locality used in Hawking's derivation of thermal radiation from a black hole. However, the derivation does depend on a notion of locality, and one hypothesis is that locality breaks down in order to preserve quantum mechanics.

Giddings sketches how locality is used starting from Eq 4.1 of http://arxiv.org/abs/1105.2036.
Mathur gives a lot more detail on the use of locality in getting Hawking's result http://arxiv.org/abs/0909.1038.

May I ask whether Qbism is all about multiple observers who concoct multiple descriptions of
quantum systems, differing among themselves because the systems observed include other observers with their differing perceptions? Is why Qbism is called 'subjective’?

Or if instead Qbism is a description of future 'reality' that involves personal guesses, rendered
quantitative and respectable by a calculated ‘probability’, making this the reason for calling Qbism
'subjective’ --- since ‘probability’ is a word that links the known past and present to the
unknown and mysterious future?

I suspect the latter, because a ‘probable’ event is something which may be expected to happen in
the future, rather than linking multiple past or current perceptions.

Yes, that's part of the reason, as they say in http://arxiv.org/abs/1301.3274 . Also, there are various interpretations of probability including frequentist, objective Bayesian, and subjective Bayesian. A key theorem in classical subjective Bayesian inference is the beautiful de Finetti representation theorem. Hudson and Moody proved a quantum version, and Caves, Fuchs and Schack gave a later, different proof of it that makes the parallel to the classical de Finetti representation theorem more obvious http://arxiv.org/abs/quant-ph/0104088 .

 

[marcus]

May I ask whether Qbism is all about multiple observers …

Or if instead Qbism is a description of future 'reality' that involves personal guesses, rendered
quantitative and respectable by a calculated ‘probability’, making this the reason for calling Qbism 'subjective’ ...

I think the latter. I merely use the two-observer situation as a trivial example of how different views MUST differ. Each is subjective from that observer's standpoint.
But in applying this conception of QM it is not essential to have more than one observer. In a bizarre universe in which there were ONLY ONE, that observer could still be making subjective judgements of probabilities based on the information accumulating from his/her experiences.

What is essential to the concept is to have AT LEAST one observer. However as we all know there are lots of us so the discussion quickly moves to considering that additional complication.

Here's a nice quote from the conclusions of the 1301 paper that Atyy just linked:
==quote page 27, 28 of http://arxiv.org/abs/1301.3274 ==
Of course, as a single-user theory, quantum mechanics is available to any agent to guide and better prepare him for his own encounters with the world. And although quantum mechanics has nothing to say about another agent’s personal experiences, agents can communicate and use the information gained from each other to update their probability assignments.

In the spirit of the Paulian Idea, however, querying another agent means taking an action on him. Whenever “I” encounter a quantum system, and take an action upon it, it catalyzes a consequence in my experience that my experience could not have foreseen. Similarly, by a Copernican-style principle, I should assume the same for “you”: Whenever you encounter a quantum system, taking an action upon it, it catalyzes a consequence in your experience.

By one category of thought, we are agents, but by another category of thought we are physical systems. And when we take actions upon each other, the category distinctions are symmetrical. Like with the Rubin vase, the best the eye can do is flit back and forth between the two formulations.

The previous paragraph should have made clear that viewing quantum mechanics as a single user theory does not mean there is only one user. QBism does not lead to solipsism. Any charge of solipsism is further refuted by two points central to the Paulian Idea. (Fuchs, 2002b).
One is the conceptual split of the world into two parts—one an agent and the other an external quantum system—that gets the discussion of quantum measurement off the ground in the first place. If such a split were not needed for making sense of the question of actions (actions upon what? in what? with respect to what?), it would not have been made. Imagining a quantum mea- surement without an autonomous quantum system participating in the process would be as paradoxical as the Zen koan of the sound of a single hand clapping.
The second point is that once the agent chooses an action {E_i}, the particular consequence E_k of it is beyond his control. That is to say, the particular outcome of a quantum measurement is not a product of his desires, whims, or fancies—this is the very reason he uses the calculus of probabilities in the first place: they quantify his uncertainty (Lindley, 2006), an uncertainty that, try as he might, he cannot get around. So, implicit in this whole picture—this whole Paulian Idea—is an “external world . . . made of something,” just as Martin Gardner calls for…
==endquote==

 

[atyy]

The way I like to think about QBism is that it is "Copenhagen" in the style of the "orthodox" or "naive textbook" interpretation given in Landau and Lifshitz, except that where L&L had a classical measurement apparatus, QBism asks can we have a classical rational agent, where rational is defined analogously to that in subjective Bayesian inference. As long as one does not go on and say, and there is no deeper reality (what can such a thing mean, since reality is just something in our model, which is only our useful approximation to the more mystical notion of reality), I think one could even say QBism can be derived from Bohmian mechanics, which is certainly a proposal for a deeper reality. Why? Because Bohmian mechanics does use probability, but is agnostic as to whether that probability is frequentist, objective Bayesian or subjective Bayesian.

Here is an approach to combining Bohmian mechanics and subjective Bayesian inference:

http://arxiv.org/abs/0706.2522
Grounding Bohmian Mechanics in Weak Values and Bayesianism
H. M. Wiseman

"I believe the most fruitful viewpoint to adopt, in BM as elsewhere, is the subjective or Bayesian interpretation of probabilities. This approach is perhaps best summed up by the slogan "probability is not real" [36]."

"[36] B. de Finetti, Theory of Probability (Wiley, New York, 1974–5)."

QBism does not solve the problem of locality (because it wasn't a problem to start), and it does not solve the measurement problem, because there is still the fuzzy cut between classical and quantum. The cut between classical agent and quantum world is fundamental in QBism, but we can still ask exactly which neurons of an agent are in his agent self, and which neurons are part of the quantum world.

 

[strangerep]

QBism [...] does not solve the measurement problem, because there is still the fuzzy cut between classical and quantum. The cut between classical agent and quantum world is fundamental in QBism, but we can still ask exactly which neurons of an agent are in his agent self, and which neurons are part of the quantum world.

I don't see the need for any such cut at all. Classical systems are just quantum systems in a certain limit.

 

[atyy]

I don't see the need for any such cut at all. Classical systems are just quantum systems in a certain limit.

The cut is fundamental in Qbism. The agent is a primitive notion, just like the classical apparatus is a primitive notion in the orthodox interpretation (Landau & Lifshitz) and other interpretations with a fundamental collapse postulate (or Bayesian updating). It is true that in all interpretations, classical systems are a limit of quantum mechanics, but in the orthodox interpretation, the cut is also necessary to collapse the wave function.

Here is a QBist exposition in which the agent and the cut between the agent and the external world is a primitive notion http://arxiv.org/abs/1301.3274 (p27): "Primitive notions: a) the agent, b) things external to the agent, or, more commonly, “systems,” c) the agent’s actions on the systems, and d) the consequences of those actions for her experience."

For why I say Bayesian updating is a sort of collapse, an example is found in Bohmian mechanics with the Bayesian interpretation of probability http://arxiv.org/abs/0706.2522 (p10): "As soon as an innocent observer were to open her eyes she would collapse her state of belief about x from P_prior(x; t) to a much sharper P(x; t), by observing the location of objects (from the pointer on a meter to the stars in the sky) relative to her. Note that this “collapse” is completely classical: it is just Bayesian updating of her beliefs about the positions of macroscopic objects. The guiding function |ψ(t)> of course does not collapse."

Although both QBiism and Bohmian mechanics with subjective probability have cuts. The difference is that the observer in QBism cannot consider himself quantum, but the external world is. If the QBist agent includes himself in the wave function, experiments will not yield definite outcomes for him (in QBism, the wave function applies to single systems). In QBism, it is not clear if there is any wave function if there is no observer. In Bohmian mechanics, the observer can consider himself made of the same stuff as the external world. Furthermore, Bohmian mechanics can be conceived to run whether or not there is an observer in the universe.

Incidentally, the collapse of the wave function can also be rigourously derived from quantum mechanics as a limit. This was done by Klaus Hepp in http://dx.doi.org/10.5169/seals-114381. John Bell argued that this mathematical limit was correct, but not of physical significance for solving the measurement problem http://dx.doi.org/10.5169/seals-114661. (I found out about Bell's paper in Allahverdyan et al's http://arxiv.org/abs/1107.2138)

 

[marcus]

Mermin talks about where the dividing line is, it doesn't seem all that important to me---could be put where the individual observer says.
I suppose another possibility could be to imagine that each agent A has a Laptop programmed to do quantum mechanics, with the essential information about that agent's hilbert space H_A stored in its memory.
When the agent gets information he/she enters it through the keyboard and it takes mathematical form. The quantum state is updated. Maybe the classic observer is defined to be some portion of the Laptop.

IOW whatever file system registers the new information and gets updated, THAT we could consider the observer. To be precise, the mathematical structure represented in the database. And any other isomorphic to it.

And then the wording of some of what Mermin says would have to be changed a bit, but I'm not sure it would matter.

What makes the human image so compelling, I think, is that the observer can be imagined as a professional bettor, to whom probabilities have a clear subjective meaning. If the bookie thinks the odds of something happening are 80% what that means is that he will pay anything UP TO $0.80 for a bet that pays $1.00 if the event occurs. And he will be willing to sell such a bet (agree to pay $1.00 if it occurs) for any offer in excess of $0.80.

I suppose a successful professional bookie could be constructed as an automaton. Maybe that is what physics PhDs do these days when they go to work in Finance. But none of that matters. What I want to do here is focus on the information that is registered at the observer in some form, and that is processed according to a quantum mechanical model, leading to the state at that observer being updated.

 

[atyy]

Mermin talks about where the dividing line is, it doesn't seem all that important to me---could be put where the individual observer says.
I suppose another possibility could be to imagine that each agent A has a Laptop programmed to do quantum mechanics, with the essential information about that agent's hilbert space H_A stored in its memory.
When the agent gets information he/she enters it through the keyboard and it takes mathematical form. The quantum state is updated. Maybe the classic observer is defined to be some portion of the Laptop.

IOW whatever file system registers the new information and gets updated, THAT we could consider the observer. To be precise, the mathematical structure represented in the database. And any other isomorphic to it.

And then the wording of some of what Mermin says would have to be changed a bit, but I'm not sure it would matter.

What makes the human image so compelling, I think, is that the observer can be imagined as a professional bettor, to whom probabilities have a clear subjective meaning. If the bookie thinks the odds of something happening are 80% what that means is that he will pay anything UP TO $0.80 for a bet that pays $1.00 if the event occurs. And he will be willing to sell such a bet (agree to pay $1.00 if it occurs) for any offer in excess of $0.80.

I suppose a successful professional bookie could be constructed as an automaton. Maybe that is what physics PhDs do these days when they go to work in Finance. But none of that matters. What I want to do here is focus on the information that is registered at the observer in some form, and that is processed according to a quantum mechanical model, leading to the state at that observer being updated.

The Bayesian rational observer can certainly be a classical computer (it's a standard concept in machine learning). But there is an external world described by the wave function. And the dividing line between the classical computer and its external world is primitive in QBism, and fuzzy when we have to implement it. As with the orthodox interpretation, there hasn't been a problem in practice. But it is a problem in principle.

 

[RUTA]

Hopefully RUTA will help clarify. He's a prof at a university in PA and has a research interest in this kind of thing. But it is nearly 11 PM here so it will be approaching 2AM in PA, won't hear from him until tomorrow I guess.

I wasn't notified about posts on this thread today. I just now ck'd and was surprised to see all this activity.

Anyway, I was hoping you guys would explain QBism to me, haha. What little I know of it is consistent with what marcus posted today.

I need to understand how we can have a unique classical reality that is supposed to result from quantum events when each person's collection of quantum events (housed in their view of the unique classical spacetime) is different. Even the inconsistency of simultaneity and spatial and temporal measurements of SR is accommodated by a single structure, i.e., M4. So, we have this unique spacetime structure of classical physics that is constructed from all observers' input of classical events. Then each observer conducts QM experiments and analyses using that classical structure (that's how you get your operators, eigenvectors, and eigenvalues for spin measurement, for example). Then, when it's time to record and report your QM measurements, the single unique spacetime structure of classical physics disappears and it's every man for himself. Yet, this unique spacetime structure is supposed to be nothing but a multitude of quantum events. I haven't wrapped my head around that yet.

 

[atyy]

@RUTA: that part of QBism, especially as represented in the FMS article, makes no sense to me either.

 

[marcus]

...I need to understand how we can have a unique classical reality that is supposed to result ...

Happily enough it turns out that (as I understand epistemic approaches such as QB or RQM) we are relieved from having to understand that, because a single unique classical reality is not assumed.

There is no one unique official mathematical description of reality. Various constructs in classical physics and quantum mechanics are used by the physicist to organize his experience, give an account of it, calculate probabilities, make predictions. But they are viewed as conceptual tools, and not equated with reality.

 

[atyy]

Happily enough it turns out that (as I understand epistemic approaches such as QB or RQM) we are relieved from having to understand that, because a single unique classical reality is not assumed.

There is no one unique official mathematical description of reality. Various constructs in classical physics and quantum mechanics are used by the physicist to organize his experience, give an account of it, calculate probabilities, make predictions. But they are viewed as conceptual tools, and not equated with reality.

There is a classical spacetime which is part of one observer's classical reality (since apparently "locality" is real). Then there are also multiple classical observers. Why are they not part of anyone observer's classical reality even though there is a real classical spacetime on which the wave function describing other observers (who are classical to themselves) are not? Also, in what space is "classical observer A" identified with "quantum observer A"? If the classical and quantum versions of each observer are not identified, there isn't really a problem. But they do seem to want each observer to be either classical or quantum, depending on context.

 

[marcus]

Physics is a game of reconciling disparate perceptions...

Where QBism/CBism deviates from business as usual is to acknowledge that the nature of the game is in fact the reconciliation of perceptions (knowledge). And, to question whether or not this reconciliation requires a unique model, e.g., self, other, and box, or a unique spacetime manifold. Thus, there is perhaps no unique model possible in the reconciliation even though there is a mathematical formalism that links the disparate perceptions...

I think you put it well earlier. In fact (although I don't think it's known at least at present) I personally suspect that it is NOT possible to have a unique model, a unique mathematical description.

IOW, whatever successful math devices people come up with (whether QM, or GR, or thermodynamics, or an unnamed conceptual tool still to be invented) will turn out to have a subjective part---will turn out to depend on the perspective of the observer. Just a suspicion.

 

[atyy]

I think you put it well earlier. In fact (although I don't think it's known at least at present) I personally suspect that it is NOT possible to have a unique model, a unique mathematical description.

IOW, whatever successful math devices people come up with (whether QM, or GR, or thermodynamics, or an unnamed conceptual tool still to be invented) will turn out to have a subjective part---will turn out to depend on the perspective of the observer. Just a suspicion.

That's far too general a reply. No one would disagree with that. The problem in QBism is if I am an agent, I have my classical world and my quantum world, and I am the only classical agent. There doesn't seem to me to be any other classical agents. Yet QBism seems to have multiple classical agents, so in whose model of reality do these agents exist? They seem to exist in FMS's god-like view, even though to me FMS are just quantum things and have no classical experiences.

 

[marcus]

... in what space is "classical observer A" identified with "quantum observer A"? If the classical and quantum versions of each observer are not identified, there isn't really a problem...

Not sure what you mean by "quantum observer A". I don't think I used the phrase "quantum observer" and can't recall reading it in the QB and RQM papers.

With B as the observer, A is not an observer but is just a quantum subsystem, part of the world that B (the observer) is using various theories, models, tools to understand.

Certainly A as a quantum chunk of the quantum world that B is observing is not equated to A the observer, in the disjoint context where A is the observer and B is part of the quantum world that A is observing.

Or so it seems to me, anyway. If you find anything to the contrary either in "Relational EPR" or the recent Mermin papers please let me know!

So maybe, as you say, "there isn't really a problem."

 

[atyy]

Not sure what you mean by "quantum observer A". I don't think I used the phrase "quantum observer" and can't recall reading it in the QB and RQM papers.

With B as the observer, A is not an observer but is just a quantum subsystem, part of the world that B (the observer) is using various theories, models, tools to understand.

Certainly A as a quantum chunk of the quantum world that B is observing is not equated to A the observer, in the disjoint context where A is the observer and B is part of the quantum world that A is observing.

Or so it seems to me, anyway. If you find anything to the contrary either in "Relational EPR" or the recent Mermin papers please let me know!

So maybe, as you say, "there isn't really a problem."

Yes, but the basic question is in which space are there multiple agents?

And the problem remains A is a quantum subsystem and classical observer - how and in which space is that identification made?

 

[marcus]

...Yet QBism seems to have multiple classical agents, so in whose model of reality do these agents exist?...

"Multiple classical agents"? I did not get that impression. Why do you say it "seems" to have them?

Notice QB is not an ontology. It is an way of regarding QM in a single-user context. In every implementation of QM, there is a specific observer, a physicist who is being aided by QM to understand and organize his experience.
AFAIK there is no concrete application or implementation of QM in which there are multiple observers, that I have ever seen referred to in QB. It sounds contrary to the spirit and intent of QB.
Of course I'm hardly and expert You've read more of it than I have, I think. Some of these links are really nice.

 

[atyy]

"Multiple classical agents"? I did not get that impression. Why do you say it "seems" to have them?

Notice QB is not an ontology. It is an way of regarding QM in a single-user context. In every implementation of QM, there is a specific observer, a physicist who is being aided by QM to understand and organize his experience.
AFAIK there is no concrete application or implementation of QM in which there are multiple observers, that I have ever seen referred to in QB. It sounds contrary to the spirit and intent of QB.
Of course I'm hardly and expert You've read more of it than I have, I think. Some of these links are really nice.

FMS say "There is a vestigial remnant in QBism of the Copenhagen classical domain, but the vestige of this “classical domain” varies from one agent to another and is limited to that agent’s directly perceived personal experience." Or http://arxiv.org/abs/1301.3274 say things like "two agents looking at the same data". So they do wish to say that there is more than one agent. They do identify the classical domain with an "objective (“classical”) domain". So my question is in which space do multiple agents exist? In my mind, as a QBist agent, there is only my objective experience - I cannot buy the claim that multiple agents exist. If I am to buy the claim that mutiple agents exist, I must operate on a different level than a QBist agent, the reality of FMS.

 

[strangerep]

The [classical/quantum] cut is fundamental in Qbism. The agent is a primitive notion, just like the classical apparatus is a primitive notion in the orthodox interpretation (Landau & Lifshitz) and other interpretations with a fundamental collapse postulate (or Bayesian updating). It is true that in all interpretations, classical systems are a limit of quantum mechanics, but in the orthodox interpretation, the cut is also necessary to collapse the wave function.

Well, then perhaps I was reading more into the FMS paper than they intended. I had in mind a collection of systems (of any kind).

So... I think I just ceased to be a QBist. It all starting to seem like a great volume of emptiness, clothed in sophistry.

I retreat to shut-up-and-calculate, since that produces all the results that are actually useful, with no need for a classical/quantum cut.

 

[Paulibus]

I now join Strangerep in his/her retreat to shut-up-and-calculate. Feynman is persuasive, as usual, and many of the fine words in this thread butter no parsnips!

Uneasily, I can't agree with Marcus' statements (e.g. post # 59) along the lines of "There is no one unique official mathematical description of reality". Instead, I maintain that the very purpose of physics is to devise such storied descriptions, quantified with the invented language of mathematics, which cleverly includes the statistical concept of quantified probability. This convenient feature let's physicists tell a unique story that incorporates a single-user's uncertain personal expectations of future happenings. This is how varying outcomes can be reconciled iff they come to pass.

Such happenings are quite unpredictable, sometimes grossly (e.g. horse racing) sometimes in detail (e.g. double-slit with individual electrons). The devil is in the future detail.

The real mystery is not the story (we anthropoid primates are uniquely adept at telling adaptable stories), but in the universal scale of 'the detail', manifested by the unexplained size of h. Perhaps the scale of 'the detail' in Nature is set by the Devil Herself?

 

[Haelfix]

I must admit as to being completely mystified by what you gentlemen are talking about in this thread. There are about 5 different completely different concepts, about twenty preprints with concepts ranging from the interpretations of quantum mechanics, to speculative proposals about modifying stat mech and quantum mechanics, to black hole physics, to ads/cft and into philosophy of science. THe edge magazine interview questions are discussing very different specific physics scenarios that really don't have anything to do with one another.

 

[marcus]

Hi Haelfix,
going back to the OP, here's what I intend the thread to discuss. There's a bunch of interesting current research including some by prominent mainstream authors having to do with taking an EPISTEMIC view of mathematical models. I.e. that they don't describe reality in any simple naive sense, they are tools by which the individual scientist understands, organizes experience, predicts--they represent knowledge of reality specific to the agent, not a uniquely correct version of reality itself. In fact according to this view one does not expect a unique account of reality, or a single correct version of the model.

If you have seen interesting recent research papers on foundations/interpretation of [GR/StatMech/QM], please share your links and thoughts. Argument along ontic versus epistemic lines is not approved--it is frequently a waste of time. So if you please use a separate thread if your views are ontic.

For a simple explanation of the difference, google "Mermin pirsa". You get a 45 minute video lecture http://pirsa.org/09090077/ "Confusing Ontic and Epistemic Causes Trouble in Classical Physics Too"
And the summary which you can read immediately without watching the talk says:
" A central issue is whether quantum states describe reality (the ontic view) or an agent's knowledge of reality (the epistemic view)."

The title of that video talk is a reminder that "QM interpretation" is only part of a tangle of foundational puzzles involving GR, Statistical Mechanics, Thermodynamics as well. In other words foundational confusion can cause trouble in CLASSICAL branches as well as quantum. I would like to know what other people have learned about this nexus of problems.

If you want a paper that is easy to read and quickly covers the material of that video, google "Mermin problem of the now". http://arxiv.org/abs/1312.7825 This defines a conceptual problem common to a lot of physics, notably (but not exclusively) classical. You might be interested in how Mermin resolves it.
That is a December 2013 paper---we're especially interested in recent work in this thread.

I want to mention some foundations connections between GR, StatMech, Thermodynamics, but will make a separate post of that so this one doesn't get too long.

If you google "introduction QBism" the top hit will be this November 2013 paper by Fuchs Mermin Schack http://arxiv.org/abs/1311.5253
An Introduction to QBism with an Application to the Locality of Quantum Mechanics
We give an introduction to the QBist interpretation of quantum mechanics. We note that it removes the paradoxes, conundra, and pseudo-problems that have plagued quantum foundations for the past nine decades. As an example, we show in detail how it eliminates quantum "non locality".
11 pages.

I don't think that's an empty claim and it signals a kind of change in the weather around quantum foundations and interpretation. Basically they say "let's put the agent (the subject, the physicist) into the picture instead of pretending that there's only the objective real world, and let's acknowledge that agents can communicate about their common reality." There is a kind of common sense realism here, I find.

This paper is one of two which for me personally characterize an epistemic view of QM. Bear in mind that there is more to this than merely Quantum Mechanics. There are significant epistemic developments in GR, StatMech, Thermodynamcs and in the interconnections among these fields. But just looking at QM for the moment, the OTHER paper personally significant for me is what you get when you google "relational EPR"

If you google "relational EPR" the top hit will be this April 2006 paper by Smerlak and Rovelli:
http://arxiv.org/abs/quant-ph/0604064
Relational EPR
We study the EPR-type correlations from the perspective of the relational interpretation of quantum mechanics. We argue that these correlations do not entail any form of 'non-locality', when viewed in the context of this interpretation. The abandonment of strict Einstein realism implied by the relational stance permits to reconcile quantum mechanics, completeness, (operationally defined) separability, and locality.
10 pages
==excerpt==
... It is far from the spirit of RQM to assume that each observer has a “solipsistic” picture of reality, disconnected from the picture of all the other observers. In fact, the very reason we can do science is because of the consistency we find in nature: if I see an elephant and I ask you what you see, I expect you to tell me that you too see an elephant. If not, something is wrong.
But, as claimed above, any such conversation about elephants is ultimately an interaction between quantum systems. This fact may be irrelevant in everyday life, but disregarding it may give rise to subtle confusions, such as the one leading to the conclusion of non-local EPR influences.
In the EPR situation, A and B can be considered two distinct observers, both making measurements on α and β. The comparison of the results of their measurements, we have argued, cannot be instantaneous, that is, it requires A and B to be in causal contact. More importantly, with respect to A, B is to be considered as a normal quantum system (and, of course, with respect to B, A is a normal quantum system). So, what happens if A and B compare notes? Have they seen the same elephant?
It is one of the most remarkable features of quantum mechanics that indeed it automatically guarantees precisely the kind of consistency that we see in nature [6]…
==endquote==

Both these papers are so thematically similar that I continue to find it odd that the November 2013 one does not cite the April 2006 one as a reference! In any case both have helped to form my own views and thinking about this topic.

Since we've turned a page, I'll bring forward a kind of thematic post as a reminder of what's involved in epistemic approaches such as QB or RQM…
====quote post#27====
... a reminder about the basic reason epistemic approaches like "QB" and "RQM" avoid various puzzles and pitfalls is that although there is a basic reality that all observers observe and although they can RECONCILE differences in the accounts of it different observers arrive at, they do have different accounts.
There is no one single official story.

To take a trivial illustration: Alice considers herself an "observer" and Bob as a quantum system, so her account of reality is obviously different from that of Bob, who considers himself the observer and Allce part of the quantum world he is trying to understand. Both use Quantum Mechanics to organize their experience and inevitably their stories differ, but yet can be reconciled.
Special rel gives us something analogous where observers can disagree about the order in which events occurred, and yet the disagreements can be explained by the observers' relative motion.

So there is a clear and serious cost. I think we all here realize this. Just for extra clarity I quote some excerpts from Mermin's recent writings and from the 2006 paper "Relational EPR".

==google "mermin now arxiv" ==
In a Physics Today Commentary, and more carefully, extensively, and convincingly with Chris Fuchs and Ruediger Schack, I argued that stubborn longstanding problems in the interpretation of quantum mechanics fade away if one takes literally Niels Bohr’s dictum that the purpose of science is not to reveal “the real essence of the phenomena” but to find “relations between the manifold aspects of our experience.” Here I note that the view of science as a tool that each of us can use to organize our own personal experience, called QBism by Fuchs and Schack, clarifies more than just quantum foundational problems. Recognizing that science is about the subject (the user of science) and not just about the object (the world external to that user) can eliminate well entrenched confusion in classical physics too.
==endquote==​

==google "relational EPR" ==
The relational approach claims that a number of confusing puzzles raised by Quantum Mechanics (QM) result from the unjustified use of the notion of objective, absolute, ‘state’ of a physical system, or from the notion of absolute, real, ‘event’.
The way out from the confusion suggested by RQM consists in acknowledging that different observers can give different accounts of the actuality of the same physical property [6]. This fact implies that the occurrence of an event is not something absolutely real or not, but it is only real in relation to a specific observer. Notice that, in this context, an observer can be any physical system.

Thus, the central idea of RQM is to apply Bohr and Heisenberg’s key intuition that “no phenomenon is a phenomenon until it is an observed phenomenon” to each observer independently. This description of physical reality, though fundamentally fragmented, is assumed in RQM to be the best possible one, i.e. to be complete [6]:

“Quantum mechanics is a theory about the physical description of physical systems relative to other systems, and this is a complete description of the world”.
==endquote==​

My comment: Note that "complete" here means best possible. The completest description we can hope for,
admittedly fragmented into versions whose differences are, however, explainable. It means giving up on the hope for one single official account as seen from transcendent perspective.
====endquote====