Everybody sees the same elephant (says Carlo Rovelli)

[marcus]

It is clear that everybody sees the same elephant. More precisely: everybody hears everybody else stating that they see the same elephant he sees. This, after all, is the best definition of objectivity.

Page 6 of this paper:
Relational EPR
Matteo Smerlak, Carlo Rovelli
7 pages
"We argue that EPR-type correlations do not entail any form of "non-locality", when viewed in the context of a relational interpretation of quantum mechanics. The abandonment of strict Einstein realism advocated by this interpretation permits to reconcile quantum mechanics, completeness, (operationally defined) separability, and locality."
http://arxiv.org/abs/quant-ph/0604064

 

[marcus]

I think it could be beneficial to try and understand what this paper is saying. I will look for some more quotes. Here is one:

DEFINITION OF LOCALITY (a principle---a feature that physical theories can have or not have)

"2.3. Locality. We call locality the principle demanding that two spatially separated objects cannot have instantaneous mutual influence. We will argue that this is not contradicted by EPR-type correlations, if we take the relational perspective on quantum mechanics. In fact, locality is at the roots of the observation that different observers do not describe the same reality. As emphasized by Einstein, it is locality that makes possible the individuation of physical systems, including those we call observers 7 ... "

 

[marcus]

...Bohr’s epistemological position, as presented for instance in [26]:

"It is wrong to think that the task of physics is to find out how nature is. Physics concerns what we can say about nature.[/size][/color]"

[26] Petersen, A.: The philosophy of Niels Bohr, Bulletin of the Atomic Scientist 19, n 7, 8-14

I guess I would paraphrase it by saying that nature is not a collection of THINGS, but a collection of FACTS.

Do you not think this is an adequate paraphrase? Then please give your own. The article also quotes the famous saying of Wittgenstein from the Tractatus. See footnote 8 on page 3:

"8. We can take this observation as an echo in fundamental physics of the celebrated: “7. Whereof one cannot speak, thereof one must be silent”[/size][/color] [25]. "

 

[marcus]

Fascinating stuff. Here is what he has to say about Einstein's supernatural realism (realism carried to what I judge to be an impractical or unrealistic extreme):

"What is missing in Einstein’s quotation above is the distinction between “elements of physical reality” (quantum events) relative to A and “elements of physical reality” relative to B. Observer A can of course measure the state of B (or, for that matter, beta), but only when A is back into causal contact with B [14]. This is, needless to say, in the future light-cone of A, and therefore poses no challenge for locality. In other words, Einstein’s reasoning requires the existence of a hypothetical super-observer[/size][/color] that can instantaneously measure the state of A and B. It is the hypothetical existence of such a nonlocal super-being[/size][/color], and not QM, that violates locality."Actually I do not believe in such a superobserver super-being. And so, for me, there is no one official set of facts.

I think Rovelli is saying that each observer is a quantum animal like everything else (there are no classical systems, or classical clocks, or classical observers) and his state Phi in his Hilbert space of states represents all he has learned about the world so far---all the facts which are fruits of his experience.

Perhaps my natural tendencies---to be skeptical of self-appointed authorities who tear down everybody and everything that is not them----to be suspicious of Official dogma----perhaps my own nature prepares me to find Rovelli's message acceptable.

But so far I have not made up my mind about this by Rovelli et al---except that it is a very readable and charming short paper.

=======================

Another thing is, it reminds me of Padmanabhan's recent paper (the one from the Paris Einstein Centennial). Because for Padmanabhan EACH OBSERVER HAS A BOUNDARY. to describe all the observers you describe all the boundaries. there is no one superbeing superobserver who instantaneously can observe all the other ones. all observers are morally equal.

 

[marcus]

Found a beautiful article in Rovelli's citations

http://arxiv.org/abs/quant-ph/0310010
Einstein, Podolsky, Rosen, and Shannon
Asher Peres

"The EPR paradox (1935) is reexamined in the light of Shannon's information theory (1948). The EPR argument did not take into account that the observers' information was localized, like any other physical object."

 

[Kea]

Hee, hee. The Elephant is a book on Category Theory by Johnstone.

 

[marcus]

Hee, hee. The Elephant is a book on Category Theory by Johnstone.

Neat title. Is it a good book about Category Theory, and entertaining?Hmmm it seems the real title is "Sketches of an Elephant" and the Wikipedia article on Topos has this to say:

Peter T. Johnstone: Sketches of an Elephant: A Topos Theory Compendium, Oxford Science Publications, Oxford, 2002. Johnstone’s overwhelming compendium. As of early 2006, two of the scheduled three volumes were available.

http://en.wikipedia.org/wiki/Topos

 

[Kea]

Johnstone’s overwhelming compendium...

Yes, well, elephants are rather large.

And by the way, Johnstone is using the word in just the same way as Rovelli.

 

[Thomas Larsson]

Rovelli is undoubtedly one of the clearest thinkers around. I especially like this paper, presumably because it is philosophically very close to my own position. Five years ago, I tried (presumably unsuccessfully) in http://www.arxiv.org/abs/hep-th/0110166 to express the need for observer dependence as follows:

"The resolution forced upon us by representation theory is as radical as it is simple: do not consider space-like separations at all. After all, a quantum theory should only deal directly with observable quantities, and space-like distances can not be observed; no observer can be in two different places at the same time. To introduce two observers does not help, because the second observer belongs to the system being observed by the first observer. Of course, I do not propose that space-like distances do not exist, only that they are not described explicitly within the formalism. What I do propose is a very strong notion of locality. Not only should all interactions be local in spacetime, but the theory should only deal directly with quantities that are local to the observer, i.e. objects on the observer’s trajectory. A drastic example: a terrestial observer does not observe the sun itself, but only photons and other particles that reach terrestial detectors, including the naked eye."

This is apparently the same idea as "Whereof one cannot speak, thereof one must be silent."

However, whereas Rovelli arrived at this position by thinking deeply about physics, I had a completely different starting point; I needed a physical interpretation of the mathematical structures arising in the multi-dimensional Virasoro algebra, which is the mathematical statement of background independence on the quantum level.

 

[Careful]

**Rovelli is undoubtedly one of the clearest thinkers around. I especially like this paper, presumably because it is philosophically very close to my own position. **

There are many clear thinkers around with equally different points of view. It is not such a big deal to come up with a story which avoids the need for non-local collapses if you allow for distinguished elements obeying different dynamics (see MWI - Rovelli's version is a clear follow up of that story). The entire difficulty of physics is to find a unifying story which (a) gives the correct predictions (b) is maximally economic (c) truly allows for a unified dynamics (or at least for an extension in that direction). Rovelli's interpretation violates (c) and (b) and perhaps also (a) - cfr. cosmological constant, dark matter, pioneer anomaly, etc... . And again, why to think that causality (in the sense cause -> effect) is restrained within the (dynamical) lightcone ? So in that sense Rovelli's clarity is the consequence of a clear historical embedding. For example : 't Hooft is a very sharp and quick thinker but probably less clear for many people.


**
This is apparently the same idea as "Whereof one cannot speak, thereof one must be silent." **

I think you still speak about the sun, moon, etc... and I assume you still give them a definite shape even if you do not receive photons from them - and neither can you speak about different worlds since no-one experienced this (and there is no ``proof´´ that these exist). Anyway, anybody can believe what he/she wants to... my bet is that perfect Bell tests do not exist (but at least avoid the citation of claims which can be turned into any direction).

Cheers,

Careful

 

[marcus]

I guess it should be pointed out that Rovelli's relational QM has nothing to do with the "Many Worlds Interpretation" and is not derived from it.Traditional QM allows for several observers, as many as one pleases. According to my reading of the paper this thread is about----"Relational EPR"------Relational QM also allows for many observers. But:

1. THEY ARE REAL PHYSICAL OBSERVERS (part of quantum nature) and
2. THERE IS ONE FEWER OBSERVER than in the traditional QM picture.

To rephrase these differences, which I would say make RQM at once more realistic and more economical.

1. in RQM the observers are not classical.

2. in RQM there is no omniscient super-observer making measurments and statements about the circumstances of all the others observers

==============

in RQM two observers can only compare notes if they are causally connected----that is, if one of them is in the lightcone of the other.

One dispenses with the wish to have supernatural or metaphysical facts which transcend the communication between ordinary observers. Objective reality becomes what these ordinary rank-and-file observers can agree on.

RQM, I would say, is a SIMPLIFICATION achieved by what James Hartle calls throwing out excess baggage.
http://arxiv.org/abs/gr-qc/0508001
One thought that one needed a supreme classical observer to watch all the others, but then one finds one does not----one can keep track of the world without using that unnecessary baggage.

 

[Thomas Larsson]

My point was that the multi-dimensional Virasoro algebra seems to know about Rovelli's relational QM. I find that quite remarkable, and some sort of triumph for both of us.

If anything, my position is more extreme than Rovelli's. The problem with quantum gravity is that QM, or rather QFT, is not completely quantum - there is an residual classical element which causes problems. Obviously, classical physics rears its ugly head in the Copenhagen interpretation - the observer is placed on the classical side of the Heisenberg cut. More seriously, however, is that an assumption about a classical observer sneaks into the formalism of QM itself.

This is most easily seen in the Hamiltonian formulation. Here one starts by foliating spacetime into fixed-time slices. But since time is defined by the observer's clock, this step implicitly assumes the existence of a macroscopic, classical observer. One can of course pick another foliation, which corresponds to a different choice of observer, but once time has been defined, it remains the same independently of what happens.

This is an unphysical assumption. In order to observe a system, the observer must interact with it. This interaction will transfer momentum to the observer, making her undergo a Lorentz transformation, and change the definition of time. Thus, the act of observation changes the foliation. Only if the observer is macroscopic, and thus classical, can we ignore this effect.

This is not so serious if we ignore gravity, since detectors are usually much bigger than the quantum phenomena we want to observe. However, a macroscopic observer has infinite mass. Hence the assumption about an a priori foliation secretely introduces an infinitely massive observer into the universe. Since gravity interacts with this infinite mass, this assumption will most likely wreck havoc in a quantum theory of gravity, in agreement with experience.

One may expect to recover ordinary QFT from observer-dependent QFT in the limit that the observer's mass goes to infinity, in the same sense that one recovers Newtonian mechanics from QM when hbar -> zero and from SR when c -> infinity.

 

[Careful]

**I guess it should be pointed out that Rovelli's relational QM has nothing to do with the "Many Worlds Interpretation" and is not derived from it.**

Calm down Marcus (f-h referred to MWI too in the context of Rovelli's LQG). In MWI, the observers are also QUANTUM (that means, there exist at least an aleph_0 number of copies of them) and there is no superobserver (God, outside of the universe who watches it all) - no foliation and all that. What I mean by uneconomical is that observers are experienced as classical in our world - so why having an infinite number of copies of them and what makes that we observe only one possibility out of an infinity of them?

Cheers,

Careful

 

[marcus]

I guess it should be pointed out that Rovelli's relational QM has nothing to do with the "Many Worlds Interpretation" and is not derived from it.
...

didn't mean to sound uncalm! I just thought the above should be pointed out.
No one has established a connection AFAIK. Anyone who wants to establish a connection should spell it out. Give links and quotes.

I don't think it is actually possible to do this, though. One would come up with only at most some superficial resemblances that don't survive scrutiny.

If F-H actually thinks that RQM is a "follow-up" of MWI, then let him speak for himself. He can tell me this himself and explain in what sense he means it.

 

[Careful]

**
If anything, my position is more extreme than Rovelli's. The problem with quantum gravity is that QM, or rather QFT, is not completely quantum - there is an residual classical element which causes problems. **

Not in the path integral formulation (see later).


**
This is most easily seen in the Hamiltonian formulation. Here one starts by foliating spacetime into fixed-time slices. But since time is defined by the observer's clock, this step implicitly assumes the existence of a macroscopic, classical observer. **

I do not see why (the foliation is a problem for quantum covariance in the Hamiltonian formulation true), doing QFT on any background *should* (at least according to our wishes) not be dependent upon the choice of foliation (that is the theories are expected to be unitarily equivalent) - this is not true in case of the thermal state calculated in the context of the Unruh effect of course, but the latter is due to a singular coordinate transformation.

** One can of course pick another foliation, which corresponds to a different choice of observer, but once time has been defined, it remains the same independently of what happens. **

Ok, but physical measurements should be independent of choice of global foliation in background dependent QFT (they indeed depend only upon the local classical clock of the observer).

**
This is an unphysical assumption. In order to observe a system, the observer must interact with it. This interaction will transfer momentum to the observer, making her undergo a Lorentz transformation, and change the definition of time. Thus, the act of observation changes the foliation. Only if the observer is macroscopic, and thus classical, can we ignore this effect. **

You mean the LOCAL foliation I presume.
Ok, but in an entirely classical theory, it would be ``easy´´ to calculate such backreaction effects and even in the context of QFT one could calculate the higher momenta of the momentum transfer and impose - as a first order correction - accordingly a statistical motion upon the observer (this not a local procedure in the strict sense of course but the same would be in the quantum case). More general: within the framework of Hartle and Sorkin, you basically only need an initial hypersurface and wave function to ask any spacetime question concerning any field observable you want to (on a fixed spacetime background) given a certain notion of coarse graining. This does not depend upon any foliation at all, you can treat everything quantum.

**
This is not so serious if we ignore gravity, since detectors are usually much bigger than the quantum phenomena we want to observe. However, a macroscopic observer has infinite mass. Hence the assumption about an a priori foliation secretely introduces an infinitely massive observer into the universe. **

No, it does not, the foliation is entirely kinematical.

**Since gravity interacts with this infinite mass, this assumption will most likely wreck havoc in a quantum theory of gravity, in agreement with experience.**

No, something like CDT has a classical time notion and some observables come out right.

** One may expect to recover ordinary QFT from observer-dependent QFT in the limit that the observer's mass goes to infinity, in the same sense that one recovers Newtonian mechanics from QM when hbar -> zero and from SR when c -> infinity. **

Classical mechanics cannot be retrieved from quantum mechanics (for N particle systems), taking limits can a be subtle and nasty process.

Cheers,

Careful

 

[Careful]

didn't mean to sound uncalm! I just thought the above should be pointed out.
No one has established a connection AFAIK. Anyone who wants to establish a connection should spell it out. Give links and quotes.

I don't think it is actually possible to do this, though. One would come up with only at most some superficial resemblances that don't survive scrutiny.

If F-H actually thinks that RQM is a "follow-up" of MWI, then let him speak for himself. He can tell me this himself and explain in what sense he means it.

The argument in the paper is almost the same as the one of Vanesch with that difference that in the relational intepretation Vanesch's consciousness is replaced by the usual discrete measurement operation. You can find evidence for the latter claim at page 2, paragraph 2.2

Cheers,

Careful

 

[vanesch]

The argument in the paper is almost the same as the one of Vanesch with that difference that in the relational intepretation Vanesch's consciousness is replaced by the usual discrete measurement operation. You can find evidence for the latter claim at page 2, paragraph 2.2

I was hesitating to jump in (I do not often hang around here), but yes, I have to agree with Careful. I skimmed to the paper, and this sounds seriously as just another MWI variant (but from the point of view of one observer where one simply doesn't talk about the copies).

Look at the thread https://www.physicsforums.com/showthread.php?t=114207
for instance.
In MWI there is NOT this "superobserver which sees all spacelike events at once" either, and, within a branch, all observers agree on seeing the same elephant also.

 

[Careful]

I was hesitating to jump in (I do not often hang around here), but yes, I have to agree with Careful. I skimmed to the paper, and this sounds seriously as just another MWI variant (but from the point of view of one observer where one simply doesn't talk about the copies).

Look at the thread https://www.physicsforums.com/showthread.php?t=114207
for instance.
In MWI there is NOT this "superobserver which sees all spacelike events at once" either, and, within a branch, all observers agree on seeing the same elephant also.

For completion, there are two differences:
(a) in Rovelli's story each observer has his own wavefunction on which he/she alone can perform the reduction rule. So your zombies are by definition merely interactions in his framework. Hence, there is no omnium and conscious beings (those who do the reduction) are by definition all in different universes.
(b) consciousness is replaced by good old fashioned reduction.

Actually, I do not feel like speaking about all this, it is just the 100'th variation to the MWI theme (although Bach certainly proved that variations on the same theme can be beautiful )

Cheers,

Careful

 

[hossi]

nonlocal super-being

http://en.wikipedia.org/wiki/The_Little_Prince"

[...]after some work with a colored pencil I succeeded in making my first drawing. My drawing number 1. It looked like this:

I showed my masterpiece to the grown-ups, and asked them whether the drawing frightened them.
They answered me: "Why should anyone be frightened by a hat?"
My drawing was not a picture of a hat. It was a picture of a boa constrictor digesting an elephant. Then, I drew the inside of the boa constrictor, so that the grown-ups could see it clearly. They always need to have things explained. My drawing number 2 looked like this:

The grown-ups' response, this time, was to advise me to lay aside my drawings of boa constrictors, whether from the inside or the outside, and devote myself instead to geography, history, arithmetic and grammar. That is why, at the age of six, I gave up what might have been a magnificent career as a painter. I had been disheartened by the failure of my drawing number 1 and my drawing number 2. Grown-ups never understand anything by themselves, and it is tiresome for children to be always and forever explaining things to them.




I admit not having any idea what the status on the field is. So could someone explain me what is new about the Rovelli-paper? I kind of like it, it goes along with my believe that there is no paradox, but I don't see how it helps in any other regard. Is the central point that the observers need to actually interact to compare their information?

The elephant issue seems to me a rather philosophical one. As scientists, I agree, we should stick to what we can say about nature, but does that really answer the question why we see what we see? You might claim, that's not a good question to ask, but I would like to know nevertheless.

Indeed, it's more like we see the prince's hat - or is it an elephant inside a boa? That's the question we can't answer. To be pragmatic, it's a question that we most likely don't need to answer. Can't avoid hoping to finally make some sense out of the quantization.



B.

 

[josh1]

I admit not having any idea what the status on the field is.

Maybe that's because you prefer physics to, well, not physics. Of course, when you're dealing with issues that require a radical conceptual change, unless you have some specific physical intuition about it that doesn't lead to a dead end, the only recourse you'll have is to philosophy. This is the case with Rovelli.

 

[hossi]

Maybe that's because you prefer physics to, well, not physics. Of course, when you're dealing with issues that require a radical conceptual change, unless you have some specific physical intuition about it that doesn't lead to a dead end, the only recourse you'll have is to philosophy. This is the case with Rovelli.

Hi josh,

I am quite flexible with my opinion what physics should or should not be, and I don't mind philosophy. It might be useful in several regards, to discuss the foundational issues of physics, esp. when being stuck at the front of research.

I have no idea what Rovelli is aiming at, but it seems to me like an attempt to question what you call 'physical intuition'. Unless you assume that we are born with a natural connection that tells us what the truth about reality is, most of our intuition comes from the education. And so far, it seems to be a pretty dead end - at least I don't want to be stuck with the 'unfinished revolution' for the rest of my scientific life.



B.

 

[Careful]

Maybe that's because you prefer physics to, well, not physics. Of course, when you're dealing with issues that require a radical conceptual change, unless you have some specific physical intuition about it that doesn't lead to a dead end, the only recourse you'll have is to philosophy. This is the case with Rovelli.

I fully agree, either you are putting in new physics like 't Hooft, Leggett and Penrose suggest or else you simply add another slightly different interpretation to the rest of them. Personally, I do not see the benefits of the latter...

Cheers,

Careful

 

[hossi]

well, that discussion here is definately a benefit for me

 

[marcus]

Relational QM is not a novel explanation of quantum mechanics

http://plato.stanford.edu/entries/qm-relational/

but more a way of adapting one's ideas of the world and of nature to reflect the lessons learned from ordinary usual QM. Instead of changing QM to fit one's ideas, one adjusts one's ideas to fit QM.

the link is to an article in the
Stanford Encyclopedia of Philosophy

specifically about Relational Quantum Mechanics dated July 2005 that was written by Rovelli and a friend.
( I guess it can be considered authoritative )

the summary begins:
"Relational quantum mechanics is an interpretation of quantum theory which discards the notions of absolute state of a system, absolute value of its physical quantities, or absolute event. The theory describes only the way systems affect each other in the course of physical interactions..."

For more info, there is the main paper on it (Rovelli 1996)

http://arxiv.org/abs/quant-ph/9609002
Relational Quantum Mechanics
Carlo Rovelli
Int. J. of Theor. Phys. 35 (1996) 1637

 

[Thomas Larsson]

Maybe that's because you prefer physics to, well, not physics. Of course, when you're dealing with issues that require a radical conceptual change, unless you have some specific physical intuition about it that doesn't lead to a dead end, the only recourse you'll have is to philosophy. This is the case with Rovelli.

I am not Rovelli, and I disagree with him (and with everybody else) about certain things (gauge anomalies). However, the explicit introduction of an observer makes a hard mathematical difference. This is because it makes it possible to construct diff anomalies, which generalize the Virasoro algebra from one to higher (in particular four) dimensions. To explicitly introduce an observer is necessary, because in all known representations, the relevant cocycles are functionals of the observer's trajectory.

Virasoro-like extensions in N dimensions are encoded in the Lie algebra cohomology group H^2(vect(N), (Z_N-1)^), where vect(N) is the algebra of vector fields in N dimensions and (Z_N-1)^ is dual to the module of closed (N-1)-forms. In particular, when N=1, a closed 0-form is a constant function, so the Virasoro extension is central in 1D, but not otherwise. The number of independent non-trivial extensions, dim H^2 = 1 if N = 1, and dim H^2 = 2, N >= 2.

This is a mathematical fact, which no philosophy (or lack thereof) in the world can change.

 

[josh1]

I am not Rovelli

This is the part of your post that made sense to me.

 

[marcus]

http://en.wikipedia.org/wiki/The_Little_Prince"
View attachment 6688
... could someone explain me what is new about the Rovelli-paper? I kind of like it, it goes along with my believe that there is no paradox, but I don't see how it helps in any other regard. Is the central point that the observers need to actually interact to compare their information?

Yes! that is a key point. There is no absolute overseer who can instantaneously report all the observers' results. (Not even in a Gedankenexperiment! )

... the explicit introduction of an observer makes a hard mathematical difference.

By explicit I understand "labeled". Each observer's results are labeled according to who is doing the observing. I think this highlights the observer's importance and allows for different observers to query/ have information about each other as autonomous quantum systems.

"Relational quantum mechanics is an interpretation of quantum theory which discards the notions of absolute state of a system, absolute value of its physical quantities, or absolute event. The theory describes only the way systems affect each other in the course of physical interactions..."

http://plato.stanford.edu/entries/qm-relational/
Stanford Encyclopedia of Philosophy (July 2005)

 

[Thomas Larsson]

Then let me try again. It is widely appreciated that spacetime diffeomorphisms play a crucial role in GR, although there is some confusion about what the terms general covariance/diffeomorphism invariance/diffeomorphism covariance/background independence really mean. However, this is a question which belongs to philosophy, or perhaps semantics.

What is completely clear, however, is that the multi-dimensional (especially the 4D) Virasoro algebra is related to spacetime diffeomorphism invariance in exactly the same way as (twice) the ordinary (1D) Virasoro algebra is related to conformal invariance in 2D. Some of us think that it might be a good idea to know about the Virasoro algebra when one studies conformally invariant theories such as string theory. In the same sense, it is a good idea to know about the 4D generalization of the Virasoro algebra if one studies diffeomorphism invariant theories such as GR; it is simply the correct quantum form of the constraint algebra (in covariant formulations).

The outstanding lesson from the multi-dimensional Virasoro algebra is that in order to construct representations, one must first expand all fields around an operator-valued curve, which is naturally identified as the observer's trajectory in spacetime. Hence background independence on the quantum level forces us to explicitly consider the observer. It is quite remarkable that one arrives from this mathematical starting point to a need for observer dependence, which is very similar to what Rovelli finds on purely physical grounds.

 

[selfAdjoint]

But TL, if your observer (I remember your making this same argument years ago on spr) has a trajectory, then he isn't a "global observer" but a "local one", no? He may be outside the system constrained by the Virasoro algebra, but he is not like God or Laplace's demon.

 

[marcus]

BTW a young QG student at U. Nottingham has made a long blog post about the Rovelli paper

http://realityconditions.blogspot.com/2006/04/relational-quantum-mechanics.html

The student (or maybe postdoc) name is Alejandro Satz.

Nottingham is a good place, I think it has John Barrett of the Barrett-Crane spinfoam model and also Kiril Krasnov----who started the GFT (group field theory) treatment of spinfoam QG along with Laurent Freidel IIRC.

I would say it is interesting what Alejandro Satz has to say about Rovelli's paper.

=============

Alejandro also gives a link to the QG blog of Christine Dantas called "Background Independence"
where he says there is currently opportunity to discuss the Rovelli paper
http://christinedantas.blogspot.com/2006/04/unfinished-revolution.html

 

[marcus]

It turns out that Alejandro Satz is a second year PhD student at Nottingham.

His blog is called REALITY CONDITIONS
http://realityconditions.blogspot.com/2006/04/relational-quantum-mechanics.html
and his explanation of Rovelli's paper seems to me to give a description of Relational QM that more clear and accurate than some I have seen (even if by older commenter's) so I will quote a piece of it:

---from Reality Conditions blog---
...The main idea of the relational interpretation is that a quantum state is not an "absolute" description of a system, but only relative to a given observer, and that a same system may be described at the same time by many different states. For example, in the "Wigner’s Friend" version of the Schroedinger Cat paradox, one observer inside a box makes a measurement of a quantum system and sees a definite result, while for a second observer outside the box the whole system including the first observer is still in an indeterminate "superposition" state. The relational interpretation has a simple description of the situation: the state is collapsed relative to the first observer and superposed relative to the second observer. (In contrast, the better known "many worlds" interpretation would say that the "true" state is the superposed one and that the first observer’s impression is a kind of illusion produced by the "branching" of his consciousness. The relational interpretation is more "democratic"; none of the descriptions is privileged.) A key feature of the relational interpretation is that according to it any quantum system can be called an "observer"; conscious beings have no special status, and any interaction can be a "measurement"...
---endquote---

 

[vanesch]

Maybe that's because you prefer physics to, well, not physics. Of course, when you're dealing with issues that require a radical conceptual change, unless you have some specific physical intuition about it that doesn't lead to a dead end, the only recourse you'll have is to philosophy. This is the case with Rovelli.

I'd say that a little philosophy can save you sometimes a lot of useless troubles. In the same way as a little philosophy in social life can make you accept a priori unfairness in life and avoid you to lose your energy in trying to set up a revolution and try to exterminate all sources of unfairness (which is in any case a lost cause and will bring you lots of useless troubles), in the same way a little philosophy in physics can make you accept certain aspects of physics you deem "unfair" and save yourself a lot of intellectual and career trouble (I'm not thinking of anybody in particular here ). But of course you won't be Napoleon then...

 

[Careful]

**Then let me try again. It is widely appreciated that spacetime diffeomorphisms play a crucial role in GR, although there is some confusion about what the terms general covariance/diffeomorphism invariance/diffeomorphism covariance/background independence really mean. However, this is a question which belongs to philosophy, or perhaps semantics. **

I do not agree, your ideas about covariance are technically different from the LQG type of quantum covariance and certainly different from string theoretical QFT ideas. Also, they differ from Hartle and Sorkin's approach within the context of decoherence functional quantum mechanics.

I must say however that I find your paper ``manifestly covariant canonical quantization I´´ quite interesting and have spent today something like one hour studying it. I have some questions and some silly (technical) remarks - since I looked a bit in the details I shall also give some of the typos.

(a) the first remark concerns the computation of the cohomology on page 10 - there you say that each function which contains pi is not closed, that is not true, a counterexample is pi*e + (psi*)*K*(pi*), however this one is in the image of the KT derivative.
(b) in general, your idea is to quantize first and then impose the dynamics, but are you not running then in similar problems as canonical quantization in the interaction picture for non linear theories?
(c) in formula 4.8, the second psi* should be \bar{psi} and similarly in 4.10, it is correct again in 4.18
(d) On page 17, the purpose of your splitting of the Hamiltonians, that is the constraint H_0 and the ``observer´´ H is to define the time derivative relative to the quantum worldline of ``the observer´´ and associated to this, the definition of the Fock vacuum state relative to the worldline and the parameter time t. However, t by itself is just window dressing and should have not any physical significance, this calls the question for reparametrisation invariance of the measured quantities. This issue is adressed in section 8 where you mention that extra matter coupling is necessary to make sense of this (did I get that right?).
Now here I am confused in the beginning, since at page 28 you mention that every bosonic p-jet bundle contributes 2(N+p,N) to the central charge (and minus for the fermions) while in formula 8.6 I suddenly get entirely different numbers.
(e) Also, in the latter construction , one would expect the relative energy to be a measurable quantity only in case the worldline would interact with the matter fields. How would this reflect upon your relative positive energy condition? (sorry, did not really think about this )

Another comment/remark, since *t* is some unphyiscal parameter, it becomes obscure what happens to your equal time commutators, and more in particular to causality itself. Actually, I see no reason why field operators corresponding to causal separations should commute. Could you elaborate more upon the relationship between the standard Fock QFT quantisation and your framework?

That's all for now...


Cheers,

Careful

 

[Careful]

But of course you won't be Napoleon then...

Elba wasn't such a bad place, he only escaped once !

 

[marcus]

http://realityconditions.blogspot.com/2006/04/relational-quantum-mechanics.html
...

continuing what I said in post #30 and 31, Alejandro Satz (in his blog "Reality Conditions") does an outstanding job reviewing Smerlak Rovelli "Relational EPR". Here is a key quote:

-------quote from Satz blog----

I think this is a quite elegant solution to the measurement problem. It does not involve any change in the testable predictions of QM, unlike the models with a physical wavefunction collapse; it does not involve extra physical baggage like hidden variables models do, and it does not involve the extra ontological baggage of the many worlds interpretation. (From the point of view of the relational interpretation, the many worlds interpretation would seem to privilege as the only "true" state one which is not relative to any particular observer; God’s point of view, so to say. I think that the relationist should deny that there is any such state, like a "wavefunction of the universe". This should have implications for quantum cosmology.) Even more attractive, for me at least, is that the interpretation is not instrumentalistic: quantum mechanics is not merely a tool for calculating and predicting but a true description of how the world works; the description must be done from the "point of view" of some physical system, but there is no privileged choice for the reference system (much like the situation with reference frames in special relativity).
---endquote---

 

[Careful]

**
I think this is a quite elegant solution to the measurement problem. **

It is not a solution to the measurement problem (remember the born rule is still in there), neither is MWI from the strict reductionist point of view.


**It does not involve any change in the testable predictions of QM, unlike the models with a physical wavefunction collapse;**

It remains to be seen whether that is a good or a bad thing

**it does not involve extra physical baggage like hidden variables models do, and it does not involve the extra ontological baggage of the many worlds interpretation. (From the point of view of the relational interpretation, the many worlds interpretation would seem to privilege as the only "true" state one which is not relative to any particular observer; God’s point of view, so to say.**

There is no need for God in MWI, and if you do the counting MWI is actually equally economical as the relative state interpretation (remember, any observer needs his own wavefunction).

** I think that the relationist should deny that there is any such state, like a "wavefunction of the universe". This should have implications for quantum cosmology.) Even more attractive, for me at least, is that the interpretation is not instrumentalistic: quantum mechanics is not merely a tool for calculating and predicting but a true description of how the world works; the description must be done from the "point of view" of some physical system, but there is no privileged choice for the reference system (much like the situation with reference frames in special relativity).**

It is still instrumentalistic wrt to the observer though.

Cheers,

Careful

 

[Sauron]

Another comment/remark, since *t* is some unphyiscal parameter, it becomes obscure what happens to your equal time commutators, and more in particular to causality itself. Actually, I see no reason why field operators corresponding to causal separations should commute. Could you elaborate more upon the relationship between the standard Fock QFT quantisation and your framework?

I have my own idea about these (I still haven´t readed all these thrread to see how it applies to these relational viewpoint).

The key is to extend quantum mechanics probability interpretation also to the time parameter.that is, \Phi(x,t) gives the probability to find the particle in a given position and in a given time.

The source of these attemp (and how to read it) comes from the following kind of imaginary experiments.

Imagine a particle who can go from point A to point B (maay be they are the same point) by two different classical paths of different length.

Imagine the particle is a "quantum clock". Them when it arrives to point B it is in a superposition state of tow states of internal time. And only when it is measured you select a particular time.

A way to build a "auntum clock" could be modifiying the energy of the particle making, for example, the two diferent paths go throught different intensities of a gravitational field. So the indicative of the "clock time" would be the energy of the particle.

Another interesting thing is that in some enviroments you can make clasical clocks and in others, like for example the very first "time" of universe you can´t.

Of course i still didn´t develope all aspectos of the idea and i am not sure if they wil dive me tosome interesting place. But i see some relations betwen it and the problems that LQG faces about the time.

 

[Thomas Larsson]

I must say however that I find your paper ``manifestly covariant canonical quantization I´´ quite interesting and have spent today something like one hour studying it. Careful

Thank you. Coming from you I take that as a serious compliment.

Before addressing the more technical details below, I should state that there are (at least) two serious flaws in my paper:
1. There is an overcounting of states already for the classical harmonic oscillator, as stated in the introduction. At the time I thought this problem would go away if you add interactions, but I was wrong.
2. No inner product was defined.
It turns out to be possible to resolve both these problems by adding extra antifields to kill the unwanted cohomology. One needs to implement a condition which identifies momenta and velocities. Then the involution defined by making frequencies negative gives the correct inner product. With these modifications my method does work for the harmonic oscillator, and by extension for all free theories. Interacting theories give rise to additional complications, which I do not yet understand.

I have some questions and some silly (technical) remarks - since I looked a bit in the details I shall also give some of the typos.

(a) the first remark concerns the computation of the cohomology on page 10 - there you say that each function which contains pi is not closed, that is not true, a counterexample is pi*e + (psi*)*K*(pi*), however this one is in the image of the KT derivative. Careful

This is true. I had completely missed that one. Thank you.

(b) in general, your idea is to quantize first and then impose the dynamics, but are you not running then in similar problems as canonical quantization in the interaction picture for non linear theories? Careful

I am not sure what these problems are.

(c) in formula 4.8, the second psi* should be \bar{psi} and similarly in 4.10, it is correct again in 4.18 Careful

OK.

(d) On page 17, the purpose of your splitting of the Hamiltonians, that is the constraint H_0 and the ``observer´´ H is to define the time derivative relative to the quantum worldline of ``the observer´´ and associated to this, the definition of the Fock vacuum state relative to the worldline and the parameter time t. However, t by itself is just window dressing and should have not any physical significance, this calls the question for reparametrisation invariance of the measured quantities. Careful

On page 16 I specialize to Minkowski spacetime. By replacing the geodesic equation by d^2 q/dt^2 = 0 I made t physical, defined in terms of the gravitational field. This could have been empasized more, though.

This issue is adressed in section 8 where you mention that extra matter coupling is necessary to make sense of this (did I get that right?). Now here I am confused in the beginning, since at page 28 you mention that every bosonic p-jet bundle contributes 2(N+p,N) to the central charge (and minus for the fermions) while in formula 8.6 I suddenly get entirely different numbers. Careful

Apart from a factor of 2, the numbers (N+p, N-k) should be replaced by (N+p-k, N). The numbers came from Lemma 7.1 at page 15 of http://www.arxiv.org/abs/math-ph/0101007, but somewhere on the way they mysteriously mutated. This is of course the problem with doing something that nobody cares about; there is no careful proof-reader.

(e) Also, in the latter construction , one would expect the relative energy to be a measurable quantity only in case the worldline would interact with the matter fields. How would this reflect upon your relative positive energy condition? (sorry, did not really think about this ) Careful

Why blushing? I had worried about this myself. Maybe the worldline in Minkowski space should satisfy something like M d^2 q/dt^2 = F, where F is a force from the fields at q(t). The straight line condition would then correspond to M = infinity, and the backreaction of this infinite mass would cause trouble in gravity.

Another comment/remark, since *t* is some unphyiscal parameter, it becomes obscure what happens to your equal time commutators, and more in particular to causality itself. Actually, I see no reason why field operators corresponding to causal separations should commute. Careful

You mean non-causal separation, no? Since all points on the observer's trajectory are supposedly causally related, and the Taylor coefficients live on this trajectory, the notion of spacelike separated events disappear from the horizon. This is confusing, and I am not sure that I have digested it yet. However, it is the same miracle that underlies the notion of analyticity.

 

[marcus]

I see some points need amplification in my earlier post which quoted Satz. In particular we need to see what is the measurement problem and how relational QM avoids or addresses it. So I will quote Wiki. Here is my post, some non-essentials eliminated and with points highlighted that need discussion.

continuing what I said in post #30 and 31, Alejandro Satz (in his blog "Reality Conditions") does an outstanding job reviewing Smerlak Rovelli "Relational EPR". Here is a key quote:

-------quote from Satz blog----

I think this is a quite elegant solution to the measurement problem. It does not involve any change in the testable predictions of QM, unlike the models with a physical wavefunction collapse; it does not involve extra physical baggage like hidden variables models do, and it does not involve the extra ontological baggage of the many worlds interpretation. (From the point of view of the relational interpretation, the many worlds interpretation would seem to privilege as the only "true" state one which is not relative to any particular observer; God’s point of view, so to say. I think that the relationist should deny that there is any such state, ...
---endquote---

Here is an exerpt from Wiki article on MEASUREMENT PROBLEM

---quote Wiki---
The measurement problem is the key set of questions that every interpretation of quantum mechanics must answer. The problem is that the wavefunction in quantum mechanics evolves according to the Schrödinger equation into a linear superposition of different states, but the actual measurements always find the physical system in a definite state, typically a position eigenstate. Any future evolution will be based on the system having the measured value at that point in time, meaning that the measurement "did something" to the process under examination. Whatever that "something" may be does not appear to be explained by the basic theory.
The best known example is the "paradox" of the Schrödinger's cat: a cat is apparently evolving into a linear superposition of basis vectors that can be characterized as an "alive cat" and states that can be described as a "dead cat". Each of these possibilities is associated with a specific nonzero probability amplitude; the cat seems to be in a "mixed" state. However, a single particular observation of the cat does not measure the probabilities: it always finds either an alive cat, or a dead cat. After that measurement the cat stays alive or dead. The measurement problem is the question: how are the probabilities converted to an actual, sharply well-defined outcome?
Different interpretations of quantum mechanics propose different solutions of the measurement problem.
The old Copenhagen interpretation was rooted in the philosophical positivism. It claimed that the probabilities are the only quantities that should be discussed, and all other questions were considered as unscientific ones. One could either imagine that the wavefunction collapses, or one could think of the wavefunction as an auxiliary mathematical tool with no direct physical interpretation whose only role is to calculate the probabilities.
While this viewpoint was sufficient to understand the outcome of all known experiments, it did not explain why it was legitimate to imagine that the cat's wavefunction collapses once the cat is observed, but it is not possible to collapse the wavefunction of the cat or the electron before it is measured. The collapse of the wavefunction used to be linked to one of two different properties of the measurement:
The measurement is done by a conscious being. In this specific interpretation, it was the presence of a conscious being that caused the wavefunction to collapse. However, this interpretation depends on a definition of "consciousness". Because of its spiritual flavor, this interpretation was never fully accepted as a scientific explanation.
The measurement apparatus is a macroscopic object. Perhaps, it is the macroscopic character of the apparata that allows us to replace the logic of quantum mechanics with the classical intuition where the positions are well-defined quantities.
The latter approach was put on firm ground in the 1980s when the phenomenon of quantum decoherence was understood. The calculations of quantum decoherence allow the physicists to identify the fuzzy boundary between the quantum microworld and the world where the classical intuition is applicable. Quantum decoherence was proposed in the context of the many-worlds interpretation, but it has also become an important part of modern update of the Copenhagen interpretation that is based on consistent histories ("Copenhagen done right"). Quantum decoherence does not describe the actual process of the wavefunction collapse, but it explains the conversion of the quantum probabilities (that are able to interfere) to the ordinary classical probabilities.
The Hugh Everett's relative state interpretation, often inaccurately referred to as the many-worlds interpretation, attempts to avoid the problem by suggesting it is an illusion. Under this system there is only one wavefunction, the superposition of the entire universe, and it never collapses -- so there is no measurement problem. Instead the act of measurement is actually an interaction between two quantum entities, which entangle to form a single larger entity, for instance living cat/happy scientist. Unfortunately Everett was never able to "close the loop", and demonstrate the way that this system would result in real-world measurements, ones in which the probabilistic nature of quantum mechanics could appear.
The many-worlds interpretation is a development of Everett's that attempts to provide a model under which the system becomes "obvious". Everett's interpretation posits a single universal wavefunction, but with the added proviso that "reality" is defined as a single path in time through the superpositions. That is, "you" have a history that is made of the outcomes of measurements you made in the past, but there are many other "yous" with slight variations in history. Under this system our reality is one of many similar ones.
---endquote---

It is probably obvious to some readers that the measurement problem (as described here) does not arise in RQM. Indeed relational QM avoids the problem rather gracefully by simply labeling the state.
If someone objects as per Wiki
"... it did not explain why it was legitimate to imagine that the cat's wavefunction collapses once the cat is observed, but it is not possible to collapse the wavefunction of the cat or the electron before it is measured. "

Then the RQM person has a very simple answer. The wavefunction in question, which you call the "cat's wavefunction" is actually a state in a Hilbertspace labeled by the name of the OBSERVER that summarizes the information he has acquired about the world. It does not have an absolute existence apart from that observer and his experience. There is a "before-and-after" operator on the observer's hilbertspace that transforms the state from the way it is BEFORE he opens the door to the way it is AFTER he opens the door of the box and has new information. He then incorporates the new information in the subspace logic latticework of his Hilbertspace that he uses to store the fruits of his experience.

Nothing could be more natural than that the state is modified by the operator of Observing the Cat, at precisely the time that the observer opens the door and observes the cat.

Other readers may not understand and want more discussion of this, so I will make a separate post to discuss further.

BTW notice that one should not confuse RQM with "Relative State Interpretation" which is the technical name for Everett's proposal widely known as MWI ("many worlds interpretation").

"Relative State Interpretation", since it an idea of Everett's associated with MWI, should not be carelessly used as a synonym for Relatonal QM---which, as Satz comment explains, is signficantly different.

 

[marcus]

I think this is a quite elegant solution to the measurement problem.

An indication of the confusion some people are in, and the need for more explanation is post #36 in this thread

It is not a solution to the measurement problem

But indeed RQM avoids the measurement paradox as it does others because these cliché contradictions can only exist in the mind of a Superbeing who can simultaneously see everybody's state of knowledge (the cat inside the box and the scientist outside---the eeper measuring a spin on Jupiter and the other eeper measuring one on Uranus)

...it does not involve extra physical baggage like hidden variables models do, and it does not involve the extra ontological baggage of the many worlds interpretation. (From the point of view of the relational interpretation, the many worlds interpretation would seem to privilege as the only "true" state one which is not relative to any particular observer; God’s point of view, so to say.

To which post #36 replies

There is no need for God in MWI, and if you do the counting MWI is actually equally economical as the relative state interpretation (remember, any observer needs his own wavefunction).

Notice that the RELATIONAL QM of rovelli is being discussed, but the poster SAYS "relative state interpretation" as if he thinks that is the same thing or wants to blur the distinction and have readers get that impression.

Indeed MWI can be regarded as equally economical with Relative State because they are essentially the SAME THING----both products of the fertile mind of Hugh Everett. So this is literally correct.
But saying this gives readers the impression that MWI is equally economical with RELATIONAL QM.

and the detail "remember any observer needs his own..." MAKES IT SEEM as if Relational is being talked about because in RQM each observer has his own hilbertspace and state within that representing his information.

But MWI has a lot of additional ontological baggage (as Satz puts it) like millions of parallel universes or whatever, and branching. It is very Baroque. You have to believe in a lot of stuff you can't see. Plus it has, like Wiki says, A SINGLE MASTER WAVEFUNCTION, implying some kind of Super-observer. Wiki says about MWI:
"Under this system[the many-worlds interpretation] there is only one wavefunction, the superposition of the entire universe, and it never collapses..."

This is preposterous. A wavefunction is the state in some observer's hilberspace logic that represents that particular observer's state of knowledge. If you fantasize a huge omniscient wavefunction that implies a huge omniscient observer with all-embracing knowledge and because the state never collapses THE SUPERBEING NEVER EVEN LEARNS ANYTHING! It is an amusing picture, but it is hard to take seriously.

Alejandro Satz puts it forthrightly enough

I think that the relationist should deny that there is any such state, like a "wavefunction of the universe".

 

[Mike2]

Section 5.2 reads,
"The properties of the system are established by its interaction with other quantum systems, and these properties are represented by the corresponding projection operators on the Hilbert space. These projectors are elements of a Boolean sigma-algebra, determined by the physics of the interaction between the two systems. ... The family of all Boolean sigma-algebras forms a category, with the sets of the projectors of each sigma-algebra as objects."

Does this mean to say that the algebra used in the relational quantum mechanics stems uniquely from category theory? Thanks.

 

[hossi]

Yes! that is a key point. There is no absolute overseer who
can instantaneously report all the observers' results. (Not even in a
Gedankenexperiment! )

Thanks, marcus. Excuse me for asking more dumb questions. I know about nothing about the topic, but I am too interested to be embarrased

So, I have spatially separated observers, measuring the outcome of an experiment with an initially entangled state. Each observer measures some 'collapsed' state. Usually, the measurements have to be in some kind of relationship that has to be instantaneous, therefore the problem with locality.

Now I say instead, well, there is no evidence for a non-local collapse as long as the observers haven't actually compared their measurements. So, I bounce back the information from one measurement to the other, to compare both. No non-locality in this, which is good. The comparison requires an interaction. Does this interaction process then make sure that the measurements fit together as parts of the entangled state? And if so, what is the difference to saying, that the 'collapse' propagates locally? I.e. imagine a continuum of observers whose measurements get compared. Or what did I miss with Rovelli's interpretation?

Still hoping for enlightenment

B.

 

[Mike2]

And if so, what is the difference to saying, that the 'collapse' propagates locally? I.e. imagine a continuum of observers whose measurements get compared. Or what did I miss with Rovelli's interpretation?

I take it that the universe is a superposition for each observer until he measures something. Then everything collapses to a eignstate in which everything is consistent with that observation, including the information from other observers.

The question then becomes does the universe go back to a superposition until that observer makes another measurement? Or is all the past and future determined by that one observers measurement (at least for him)?

 

[marcus]

Hossi and the Seven Dwarves

So, I have spatially separated observers, measuring the outcome of an experiment with an initially entangled state. Each observer measures some 'collapsed' state. Usually, the measurements have to be in some kind of relationship that has to be instantaneous, therefore the problem with locality.

Now I say instead, well, there is no evidence for a non-local collapse as long as the observers haven't actually compared their measurements.

Still hoping for enlightenment

B.

One day there was an interplanetary adventuress named Hossi who prepared two particles in aligned spinstates. One she kept, and the other she gave to some Dwarves, who were her good friends. Then the Dwarves went off to the planet Pluto. Their plan was that when it got to be Eastertime Hossi and the Dwarves would each measure their spinstates to see if they are East and West.

Now it is Easter, and Hossi doesn't KNOW that the Dwarves actually REMEMBERED to do it! Or they might have accidentally measured in a different direction besides East/West. Or the Dwarves might have just totally screwed up.

In fact, one of the Dwarves, the one called Careful John, actually LIKES screwing up. He always wants to confuse the others and get things wrong. Nothing in this life is certain, so Careful John may have prevailed or he may not have.

As far as Hossi goes there is only one Hilbertspace, hers, and just this one state of the precious pair of particles which as a token of friendship she has divided with the Dwarves. This state represents her Knowledge...

TO BE CONTINUED

BTW I really liked the hatsnakeelephant pictures you put in. they were a beautiful illustration, thanks

 

[hossi]

I take it that the universe is a superposition for each observer until he measures something. Then everything collapses to a eignstate in which everything is consistent with that observation, including the information from other observers.

The question then becomes does the universe go back to a superposition until that observer makes another measurement? Or is all the past and future determined by that one observers measurement (at least for him)?

Now you have completely confused me.

I thought: everything collapses for each observer, but not neccessarily consistently, as long as they have not observed each other. Otherwise: everything is consistent with which observation? And how would that differ from the usual collapse?

 

[marcus]

I should interrupt the story about the Seven Dwarves to make sure everybody knows what Dirac said about the Heisenberg picture. It is the right one---more fundamental.

this is what Dirac said near the end of his life in the last public seminar he gave. It was on the Island of Sicily and he had only one slide for the the whole lecture. the slide said:

The Heisenberg picture is the right one.[/size][/color]

---quote from Smerlak Rovelli----
ψ is the coding of the information that A has about S. Because of this irreducible epistemic character, ψ is a relative state, and cannot be taken to be an objective property of the single system S, independent from A. Every state of quantum theory is a relative state. 5

On the other hand, the state ψ is a tool that can be used by A to predict future outcomes of interactions between S and A. In general these predictions depend on the time t at which the interaction will take place. In the Schrödinger picture this time dependence is coded into a time evolution of the state ψ itself. In this picture, there are therefore two distinct manners in which ψ can evolve: (i) in a discrete way, when S and A interact, in order for the information to be adjusted, and (ii) in a continuous way, to reflect the time dependence of the probabilistic relation between past and future events.

From the relational perspective the Heisenberg picture appears far more natural: ψ codes the information that can be extracted from past interactions and has no explicit dependence on time; it is adjusted only as a result of an interaction, namely as a result of a new quantum event relative to the observer. If physical reality is the set of these bipartite interactions, and nothing else, our description of dynamics by means of relative states should better mirror this fact: discrete changes of the relative state, when information is updated, and nothing else. What evolves with time are the operators, whose expectation values code the time-dependent probabilities that can be computed on the basis of the past quantum events. [Footnote 6]Footnote 6: This was also Dirac’s opinion: in the first edition of his celebrated book on quantum mechanics, Dirac uses Heisenberg states (he calls them relativistic) [29]. In later editions, he switches to Schrödinger states, explaining in the preface that it is easier to calculate with these, but it is “a pity” to give up Heisenberg states, which are more fundamental. In what was perhaps his last public seminar, in Sicily, Dirac used a single transparency, with just one sentence: “The Heisenberg picture is the right one”.
---endquote---

 

[vanesch]

BTW notice that one should not confuse RQM with "Relative State Interpretation" which is the technical name for Everett's proposal widely known as MWI ("many worlds interpretation").

"Relative State Interpretation", since it an idea of Everett's associated with MWI, should not be carelessly used as a synonym for Relatonal QM---which, as Satz comment explains, is signficantly different.

I'm really sorry, but I don't see any difference.

What happens in MWI is the following. Let's say that we have Alice who is going to "measure" using a "measuring device", the spin state of a particle. We say that *before the measurement* (which is nothing but an interaction), the state is:

|alice> (|up> + |down>)

The "measurement interaction" will evolve the state:
|alice>|up> into |alice_who_saw_up> |up>
and
|alice>|down> into |alice_who_saw_down>|down>
(this is how the interactions in her apparatus are constructed, and linked, finally to her body by her senses)

which means that the overall state evolves into:

|alice_who_saw_up>|up> + |alice_who_saw_down> |down>

But, "alice" being a subjectively generated experience by a classically-looking state, we have now that the original experience |alice> will evolve into two of these. As if alice's classical body was copied.
The quantum state of Alice's body is now "generator" of two classically-experiencable states, and can hence generate two associated, different, subjective experiences. So Alice's original subjective experience will not be able to "ride both of them" in the same way as twin brothers do not have a "single conscious experience" but have two of them. We had "one" classically generated state before the interaction (associated with a subjective experience) and now we have two. If we would only have had one, we would have made no objections that there was a "continuity" of the original Alice experience (we wouldn't wonder why qlice wasn't suddenly experiencing the next moment "Mary's body" or something, right ?). But there are now two of them.
Which one will be "the original" and which one will be the "copy" ? Answer: use the Born rule. Now, most MWI variants try to naturally deduce this Born rule from an equal-probability rule and world counting, which is impossible without introducing extra hypotheses, but you can simply STATE so. So we say that the "original" alice experience makes a random choice between the two "new" ones using the Born rule and we mark this with an asterix:

before:
|alice*> (|up> + |down>)

after:
|alice_who_saw_up*>|up> + |alice_who_saw_down> |down>

So alice's "original" experience was first connected (in the same term) with a superposition |up> + |down> and after the interaction, with |up>.

Now, if I understand well, the relational approach ONLY CONSIDERS the starred states, and then we see that the complete original "alice-experience" evolved from:

|alice>(|up> + |down>)
into
|alice_who_saw_up*> |up>

and things happen as if projection occurred from her point of view.

This is exactly what's claimed in MWI too - except that one considers that the quantum body is carrier of DIFFERENT classical states which can be subjectively experienced.

In a typical EPR setup with an Alice and a Bob, we can do the same, and we can find that from the point of alice's experience, things happen as expected, and from Bob's point of view, too. However, the difference is of course that "the original bob" can be in another branch than the "original alice" (but they won't be able to find out).

I really fail to see the difference with the relational approach *if you limit yourself to one single observer* - and of course ALL you can only find out is what one single observer has ever observed (eventually through the observation of other physical bodies).

 

[marcus]

Hi Vanesch,
I trust you have read the earlier Wikipedia quotes. In this thread I would like to use Wiki articles, which are clickable for everybody, to keep the use of terminology reasonably consistent (democracy of sources + everybody on same page = happy discussion )

What we have about MWI so far is what I quoted in post #39
----Wiki---
The Hugh Everett's relative state interpretation, often inaccurately referred to as the many-worlds interpretation, attempts to avoid the problem by suggesting it is an illusion. Under this system there is only one wavefunction, the superposition of the entire universe, and it never collapses -- so there is no measurement problem. Instead the act of measurement is actually an interaction between two quantum entities, which entangle to form a single larger entity, for instance living cat/happy scientist. Unfortunately Everett was never able to "close the loop", and demonstrate the way that this system would result in real-world measurements, ones in which the probabilistic nature of quantum mechanics could appear.
The many-worlds interpretation is a development of Everett's that attempts to provide a model under which the system becomes "obvious". Everett's interpretation posits a single universal wavefunction, but with the added proviso that "reality" is defined as a single path in time through the superpositions. That is, "you" have a history that is made of the outcomes of measurements you made in the past, but there are many other "yous" with slight variations in history. Under this system our reality is one of many similar ones.
---endquote---

How are you with this description of MWI? Would you like to provide a link to some alternative definition? I can't promise I will accept it for use in the context of this thread, but I am curious what it would be (if you were to propose another link) and would try to accommodate your proposed change.

As for the definition of Relational QM, I think we have to take Rovelli's article called "Relational Quantum Mechanics" as defining it, don't you?

I suppose the safest thing is to read what Rovelli has to say comparing RQM with various QM versions and interpretations. He goes into the similarities and differences at some length. what is your opinion about that?

Please let me know if you find Rovelli's definition of his own theory acceptable, and whether you believe his account of how it compares to other QM pictures.

If you don't like Rovelli's account of RQM and how it compares, it would be great if you would point to specific paragraphs in the article that you judge questionable.

VANESCH HERE IS THE WIKI LINK that provides the quoted description of MWI:
http://en.wikipedia.org/wiki/Measurement_problem

HERE IS ROVELLI ON RQM:
http://arxiv.org/abs/quant-ph/9609002

FOR MORE ON MWI:
http://en.wikipedia.org/wiki/Many-worlds_interpretation

 

[marcus]

...I really fail to see the difference with the relational approach *if you limit yourself to one single observer* - and of course ALL you can only find out is what one single observer has ever observed (eventually through the observation of other physical bodies).

Perhaps this is not so surprising since RQM is intended to have the look and feel of traditional QM. Especially in the case of *one single observer*
RQM is primarily an adaptation of one's philosophy in such a way that certain paradoxes (associated with two or more observers) fail to materialize, without tampering with the tried and true formalism of standard QM.

This is simply my opinion, as is what follows. To be sure, please don't take my sayso, but instead use authoritative source material, especially Rovelli's writings since it is his version of QM.

One more non-authoritative thing from me. I think that RQM does not involve postulation of additional mechanisms or consciousness or hidden variables or branchings of reality or any of that stuff. It does not give any mechanisms explaining where uncertainty comes from or where probabilities come from. As far as innovation, RQM is minimal and severely economical.

So it is not even in the same ballpark with MRI. In my humble opinion.

However *if you limit yourself to one single observer* then RQM is going to look just like ordinary QM has looked for 75 years.

and so it is going to look pretty much like other Heisenberg picture QMs including what you, Vanesch, might have in mind . I certainly wouldn't argue with that!

 

[vanesch]

What we have about MWI so far is what I quoted in post #39
----Wiki---
The Hugh Everett's relative state interpretation, often inaccurately referred to as the many-worlds interpretation, attempts to avoid the problem by suggesting it is an illusion. Under this system there is only one wavefunction, the superposition of the entire universe, and it never collapses -- so there is no measurement problem. Instead the act of measurement is actually an interaction between two quantum entities, which entangle to form a single larger entity, for instance living cat/happy scientist. Unfortunately Everett was never able to "close the loop", and demonstrate the way that this system would result in real-world measurements, ones in which the probabilistic nature of quantum mechanics could appear.
The many-worlds interpretation is a development of Everett's that attempts to provide a model under which the system becomes "obvious". Everett's interpretation posits a single universal wavefunction, but with the added proviso that "reality" is defined as a single path in time through the superpositions. That is, "you" have a history that is made of the outcomes of measurements you made in the past, but there are many other "yous" with slight variations in history. Under this system our reality is one of many similar ones.
---endquote---

How are you with this description of MWI? Would you like to provide a link to some alternative definition?

Well, in great lines I agree with what's said there, except for one point. I think there are as many flavors of MWI as there are people thinking about it, so I'll give you mine (that I've been telling about since ages on PF), but which is just a mixture of ideas which are since long around. The main idea is that "the wave function" evolves unitarily ; but even for this to make sense, one has to place oneself into a certain reference frame (a Lorentz transformation gives you *another* evolution and *another* wavefunction). And even *within* such a frame, one should consider a coarse-grained Schmidt decomposition into two systems: "observer" x "rest of universe". THESE are the branches in MWI - and clearly they are observer-dependent! They are observer-dependent for the choice of inertial frame (hence how to split up the unitary structure into "state" and "unitary evolution") AND they are observer-dependent in the Schmidt decomposition "observer/rest-of-universe".
The "number of branches" is not equal for all observers, for instance, so there's nothing "objective" about this splitting. It is only in the case when two observer bodies are in contact with the same big thermal bath that there is any hope that they will have decohered in similar branches.

Coarse-grained here means: not making distinction between microscopically different quantum states which would give rise to identical macroscopic observations (while this can include many, many different orthogonal states which may continuously evolve into one another)

As for the definition of Relational QM, I think we have to take Rovelli's article called "Relational Quantum Mechanics" as defining it, don't you?

I suppose the safest thing is to read what Rovelli has to say comparing RQM with various QM versions and interpretations. He goes into the similarities and differences at some length. what is your opinion about that?

Well, almost everything I read in Rovelli's paper made me say "yes, that's also how I see things". For instance, his "Main observation" and his "Hypothesis 1" are in complete sync with how I see things too from an MWI viewpoint.

However, his comments in "objection 7" make me think that Rovelli didn't quite understand (modern views on) MWI, and got stuck with Everett's original idea, while these have been evolving over time. He seems to think that these branches are absolute and objective, and not observer-dependent. That's of course not the case: already the choice of the split between "wavefunction" and "time evolution" (choice of inertial frame) is observer dependent; but also the "split" in branches is observer dependent because depending upon the Schmidt decomposition between "observer body" and "rest of universe".
As far as I understood, MWI starts EXACTLY from the "main observation".

As a simple example, imagine an EPR like experiment, in a frame where Alice did already her measurement, but Bob not.
We then have:

|bob-init> ( |alice+>|-> - |alice->|+>)

As long as both didn't decohere together with a common thermal bath (cannot happen if they are still spacelike separated) there is ONE branch for Bob, and there are two branches for Alice.

However, in another frame, bob made maybe already his measurement, and not alice, so there we have the opposite case. This is entirely dependent on how we "slice" the unitary structure in "state" and "unitary evolution" which is nothing else but the choice of reference frame.
And in yet another frame, both made their measurements. If the axes aren't aligned, however, each appears in a superposition to the other (until they MEET and INTERACT - exchange data) in which case they get entangled, decohere and end up in the same number of branches.

Please let me know if you find Rovelli's definition of his own theory acceptable, and whether you believe his account of how it compares to other QM pictures.

Well, I fully accept what Rovelli writes, but it seems to me that that view WAS already present in different MWI flavors.

What is the "difficulty" (which I think, is not a difficulty) in MWI, namely the "derivation" of the probability rule by world counting, is solved by Rovelli in the same way as I think it should be solved in the same way: by POSTULATING it - but I admit here to be dissonant with most MWI views, which still have the hope of _deriving_ it, which I am profoundly convinced is impossible with an extra postulate anyways.
Just say that for a specific mind to have a "bob" experience, is given by the Born rule, applied to the different branches that appear for the body state of Bob.

Further, the "coarse grained" Schmidt decomposition corresponds exactly to Rovelli's "postulate 1" (namely, a finite amount of information can be "extracted" from the universe by an observer, which comes down to saying that the observer is in one of a finite number of distinguishable states at a certain point - this is exactly the *coarse graining* needed.) His postulate 2 comes down to saying that an observer state can always entangle with some extra stuff, and hence split into two or more states. Postulate 3 is unitary quantum theory.

Nevertheless, there's one problem Rovelli runs into, and that is exactly the same problem as any other view, which is the "preferred basis" problem ; except by using the coarse-grained Schmidt decomposition + decoherence approach. Given the physical structure of a measurement apparatus, there's no way for him to find out with what hermitean measurement operator that apparatus (fully described by a unitary evolution operator of the interaction with the system) is going to correspond to, without specifying what are its pointer states. Pointer states which ONLY have a meaning when we take into account the coarse-grained decohered Schmidt decomposition between "observer" and "rest of universe".
But by even *considering* this Schmidt decomposition, one assumes the DIFFERENT TERMS in the wavefunction - which is MWI-like, no ?

I think the final comment by Rovelli on the first part of p19 makes me think that he missed the essence of "many minds": namely he gives me the impression that these minds ARE the brains, and hence are "physical" - in which case the brains would indeed be "special" things which are treated differently than other things. But minds are NOT brains: minds are "emergent states of awareness" *generated* by a physical brain state. Hence a single brain can give rise to several minds, consciousnesses, subjective experiences or whatever. And in MWI (in this = my) flavor, THIS is the ultimate "observation" (the subjective experience of a mind, as one of several, generated by a brain state). As this is not part of the physical system per se, this does NOT violate his hypothesis 1.

Rovelli leaves in the middle what is an observer ; if I fill in this "mind" stuff then I'm in agreement with all he says. However, if he understands by "observer" *any* physical system, then he has the same problem as any other. For instance, let us consider an electron as an "observer", and consider a 2-slit experiment with that "observer" ; he then has the same problems as everybody else. And given his hypothesis 1, an electron IS a valid observer !
According to the electron, which slit did it go through ? The answer is that the electron, if it ever KNEW, FORGOT through which slit it went.
And I'd like to see how Rovelli talks himself out of this "observing and forgetting" electron, without getting into decoherence, generated classically looking states, and all the stuff that finally makes up an MWI view (in which we would take the position that an electron has a mind and that to each different electron state, corresponds also an electron-mind-experience - but in what basis now ??)...