Local realism ruled out? (was: Photon entanglement and )

[yoda jedi]

Cute, but metaphysics and philosophy, and not helpful when discussing non-locality.

i am not discussing NON LOCALITY.

"real" has too many definitions otherwise, we need somewhere to begin, yes?

be real is independent of any conceptual consideration.

We need to discuss something yoda jedi, and reality is a standard we believe we experience.

Reality does not need you.

 

[Frame Dragger]

Ahh, it's good to be back. I see this thread has not changed much.

@Yoda Jedi: what kind of reality are you talking about in a local realism thread? I'm not getting what you're driving at, and I've been reading this thread for a while. The title is photon entanglment, so, I'm genuinely not getting your drift here.

 

[IcedEcliptic]

i am not discussing NON LOCALITY.

You are not discussing the topic of the thread? I'm completely confused, perhaps if you spoke in more than single sentences I could learn more from you.

 

[yoda jedi]

You are not discussing the topic of the thread? I'm completely confused, perhaps if you spoke in more than single sentences I could learn more from you.

Discussing:

"real"has too many definitions otherwise, we need somewhere to begin, yes?

be real is independent of any conceptual consideration.

reality is a standard we believe we experience.

Reality does not need you.

 

[DrChinese]

Ahh, it's good to be back. I see this thread has not changed much.

 

[Eye_in_the_Sky]

say NO to "local realism" – say YES to 'nonlocal' 'reality'

I think it is true to say that Quantum Mechanics implies 'nonlocality'.

This 'nonlocality' is either in a sense of 'causation' or in a sense of 'existence' or (say maybe) 'identity'.

 

[Eye_in_the_Sky]

Your use of the term "locality" encompasses both causal locality and separability, but otherwise it looks like the EPR argument with the same conclusion. To finish the story you've only to add QM's predicted violation of the Bell inequality with its subsequent experimental confirmation whence people believe QM is complete. Get rid of superdeterminism (keep free will) and that leaves you having to discard causal locality and/or separability, which is where the debate is centered.

______________

As a matter orientation with regard to perspective, here are three (takes on) takes on "separability" I have come across:

State Separability: The "state" 'assigned' to a "compound physical system" at any time is supervenient on the "states" then 'assigned' to its "component subsystems".

... that which we conceive as 'existing' ('real') should somehow be localized in time and space. That is, the 'real' in one part of space, A, should (in theory) somehow 'exist' independently of that which is thought of as 'real' in another part of space, B. If a physical system stretches over the parts of space A and B, then what is 'present' in B should somehow have an 'existence' independent of what is 'present' in A.

SEPARABILITY: mutually independent 'existence' of spatially distant 'things'.
______________

In connection with the scenario of Alice and Bob, I am trying to imagine a 'reality' in which the "microsystem" 'exists' in a 'manner' which is not "existentially separable", whereas, on the other hand, the "macro-instruments" (of Alice and Bob) do 'exist' in a 'manner' which is "existentially separable".
______

So, for example – as applied to the "macro-instruments" (of Alice and Bob) – by "existentially separable" I mean (something like):

The 'real' "state of Alice's instrument" and the 'real' "state of Bob's instrument" 'exist' independently of one another.

That is ... in any theory in which a notion of "state" is 'assigned' to the "instruments" of Alice and Bob, the following two conditions will hold:

1) The "state of Alice's instrument" and the "state of Bob's instrument" can be 'specified' independently of one another;

and

2) A 'specification' of the "joint state of Alice's instrument and Bob's instrument" is equivalent to a joint 'specification' of the "state of Alice's instrument" and the "state of Bob's instrument".
______

So, to repeat:

I am trying to imagine a 'reality' in which:

the "microsystem" (i.e. "singlet state") 'exists' in a 'manner' which is not "existentially separable",

whereas, on the other hand,

the "macro-instruments" (of Alice and Bob) do 'exist' in a 'manner' which is "existentially separable" [... except(,) perhaps(?) possibly(??) where/when their mutual "instruments" happen to be 'linked' via a common, "existentially nonseparable" 'onething' (such as, a "singlet state")].


... I seem to be getting stuck at this spot.

 

[Eye_in_the_Sky]

This was the EPR argument. Local causality + HUP -> (QM is incomplete) or (Reality is observer dependent - in this case Alice).

Yes. The only essential difference between the argument I have given and that of original EPR lies in the "completeness" condition.

I agree. Even if the argument I have posed can go through, its 'lesson' can be no different from that of original EPR.

So ... I see then ... as far as original EPR is concerned, you have no objection to the type of CF used. Okay. That helps clarify for me your position on CF. Good.

So, we are left with the question of which notion(s) ought to be relinquished:

... one of the following must be relinquished:

(i) 'free-choice' ,

(ii) QM is "local" ,

(iii) QM is "complete" ,

(iv) some other (implicit, currently unidentified) assumption .

You suggest:

... Bob's reality is dependent on a choice made by Alice if QM is complete. I would say this is a generally accepted conclusion: that either locality does not hold, or reality is dependent on observeration.

Okay. Let us write this as:

(QM is complete) Λ (local causality) → Bob's 'reality' depends on Alice's choice ,

where the 'reality'-dependence is "non-causal".
_________

For clarity, let us consider an example.

Suppose Alice measures S_x and gets the result "+". Then Bob's 'reality' is such that

if Bob measures S_x then he cannot obtain the result "+".

On the other hand, if Alice had measured S_y (instead of S_x), then Bob's 'reality' would have been such that

if Bob measures S_x then he can obtain the result "+".

... DrC, is this example included in what you mean by "Bob's reality is dependent on a choice made by Alice"? ... or is it not?
__________________________

Only now is it beginning to become clearer to me (although, not yet quite 'altogether') what is going on here.

First, let me explain the two motivations I had for my having posed the argument in the manner I did:

motivation 1: Somehow, vaguely, I felt that by stripping the microsystem of all 'reality', then (as a consequence) the "nonseparability" issue would – simply – disappear; [... Now, however, I see it seems that the issue has not just disappeared, but rather, it has been transferred over to the macroscopic experimental arrangement;]

and

motivation 2: Since Bell's "local causality" criterion is about 'probability' 'assignments' made on the basis of "complete" 'information', I suspected that by couching the quantum state in terms of 'information', then somehow, a previously hidden insight would emerge. [... And indeed (... I think)... I see it now.]

Bell's "local causality" criterion goes like this ["types" of emphasis added] (diagram):

A ["complete" stochastic] theory will be said to be "locally causal" if:

The 'probabilities' 'attached' to 'values' of "local beables" in a spacetime region 1 are unaltered by 'specification' of 'values' of "local beables" in a spacelike separated region 2, when what happens in the backward light cone of 1 is already sufficiently 'specified', for example by a full 'specification' of ['values' of] "local beables" in a spacetime region 3.

Now here comes the 'catch':

... what sort of 'existence' do these "local" 'beables' have?

These "local beables" belong to a 'realm' regarding which the principle of "separability" applies.
________

For example, the following four quantities are all construed (by Bell) as being "local beables":

a ≡ Alice's setting ,
b ≡ Bob's setting ,
A ≡ Alice's outcome ,
B ≡ Bob's outcome .
________

So ... "separability" as applied to (these) "local beables" (in this context of Alice and Bob) would (seem to) mean (among other things (something like this)):

The 'real' "state of Alice's instrument" and the 'real' "state of Bob's instrument" 'exist' independently of one another.

This then is (supposed) to imply that in any theory in which a notion of "state" is 'assigned' to the "instruments" of Alice and Bob, the following two conditions will hold:

1) The "state of Alice's instrument" and the "state of Bob's instrument" can be 'specified' independently of one another;

and

2) A 'specification' of the "joint state of Alice's instrument and Bob's instrument" is equivalent to a joint 'specification' of the "state of Alice's instrument" and the "state of Bob's instrument".
_____________________

I believe it is correct to see the conjunction of assumptions in Bell: locality + realism.

In connection with "stage 2" of Bell's argument, I agree with you. But in connection with "stage 1" I do not see it.

Okay, now I see it. That is, what I am now seeing regarding "stage 1" (in terms of a conjunction of assumptions) in Bell is very much along the lines of what you had put as:

locality + realism .

(After quite some thought ... I think) I would (like to) put it like this:

Bell's "local causality" criterion ↔

"causally local" 'reality' Λ "existentially separable" 'macro-apparatus-world' .


... Does this make sense to you?

 

[DrChinese]

Y

For clarity, let us consider an example.

Suppose Alice measures S_x and gets the result "+". Then Bob's 'reality' is such that

if Bob measures S_x then he cannot obtain the result "+".

On the other hand, if Alice had measured S_y (instead of S_x), then Bob's 'reality' would have been such that

if Bob measures S_x then he can obtain the result "+".

... DrC, is this example included in what you mean by "Bob's reality is dependent on a choice made by Alice"? ... or is it not?
__________________________

Yes, that pretty well sums it up. By the EPR reasoning, Bob's reality is determined by a choice of measurement by Alice. This is required by the HUP.

 

[RUTA]

I think it is true to say that Quantum Mechanics implies 'nonlocality'.

This 'nonlocality' is either in a sense of 'causation' or in a sense of 'existence' or (say maybe) 'identity'.

Or both.

 

[RUTA]

motivation 1: Somehow, vaguely, I felt that by stripping the microsystem of all 'reality', then (as a consequence) the "nonseparability" issue would – simply – disappear; [... Now, however, I see it seems that the issue has not just disappeared, but rather, it has been transferred over to the macroscopic experimental arrangement;]

Correct. This is the basis for the Relational Blockworld interpretation, i.e., no microsystem plus nonseparable experimental equipment.

 

[DevilsAvocado]

Suppose Alice measures S_x and gets the result "+". Then Bob's 'reality' is such that

if Bob measures S_x then he cannot obtain the result "+".

And if we add http://en.wikipedia.org/wiki/Relativity_of_simultaneity" to that – I say Bob can do whatever he likes = free-choice.

 

[jambaugh]

For clarity, let us consider an example.

Suppose Alice measures S_x and gets the result "+". Then Bob's 'reality' is such that

if Bob measures S_x then he cannot obtain the result "+".

On the other hand, if Alice had measured S_y (instead of S_x), then Bob's 'reality' would have been such that

if Bob measures S_x then he can obtain the result "+".

... DrC, is this example included in what you mean by "Bob's reality is dependent on a choice made by Alice"? ... or is it not?
__________________________

There is a critical distinction I would make here. When you speak of what might happen from Bob's end you are not describing an Objective Reality (what is) but rather as you've presented it the actuality of what may happen.

If you further interpret the "Why" of this actualization then you may invoke an objective reality, i.e. the state of Bob's system is in the subset of states which exclude those where property $S_x$ has value "+". In so doing you will run into the Bell type inequalities which are violated by QM because the presumption that the system and its environment is in a set of states implies the probabilities of outcomes defines a probability measure over that set.

It is this transition from "what happens" actuality to a "what is" reality that we non-realists object to. But note that you can avoid this, still speak of what happens in an objective way (which is why in CI you need the measuring devices to be classical level, you need objective reality of your measurement record) and avoid the need to invoke non-local causation.

That's I believe the heart of classic Copenhagen interpretation.

You needn't invoke CI if you prefer another interpretation (though I think you'd be incorrect) but you should be careful to distinguish when you are describing a reality=objective state of existence vs. actuality=behavior.

It is hard to recognize this distinction at first since we grow up thinking classically where all that happens can be equivalently described in terms of what is.

 

[Count Iblis]

It is not clear to me how you can get an "objective reality" at all from a theory in which it doesn't exist at the fundamental level.

 

[jambaugh]

And if we add http://en.wikipedia.org/wiki/Relativity_of_simultaneity" to that – I say Bob can do whatever he likes = free-choice.

Speaking of Relativity of Simultaneity, our heritage of thinking in terms of objective realty goes back to the fiber-bundle structure of pre-Einstein space /time. Space/time was a http://en.wikipedia.org/wiki/Fiber_bundle" with time as the base and each slice of space (and the state of all within it) is a fiber indexed by this time base. Very much the continuum analogue of a movie with each frame a reality snapshot.

Relativity of simultaneity already begins to sand away at our old concept of objective reality. To preserve it in classical SR we invoke the frozen history of all past and future in a composite space-time. This of course doesn't allow choice or change except by selecting a whole new space-time universe.

In order to still speak of possibilities and probabilities and choices in a deterministic setting we invoke typically a classical field theory, again a fiber-bundle (space-time base with fibers the possible configurations of local reality and typically also some gauge degrees of freedom).

Now any time we see fiber-bundles we can be sure there is a relativity principle which may unify base with fiber and with is a group deformation of actions where the one-way dependence of fiber action on base coordinate becomes a two-way interaction. In this case matter affects the space-time just as space-time affects the matter and we get GR.

Well I'm getting off track... the point is that we already have good reason to soften up our traditional "objective reality" mindset and begin thinking in terms closer to the epistemological basis of science, the actuality of what we observe instead of the imagined state of reality with which we modeled it in past.

 

[jambaugh]

It is not clear to me how you can get an "objective reality" at all from a theory in which it doesn't exist at the fundamental level.

This is exactly the process of transitioning to the classical scale. We look at the necessary conditions for a system of events to behave classically, e.g. commutativity of observables on the scale to which we make distinctions. When the pointer on a particle counter is set to large units so that the momentum of that pointer can be similarly refined, we don't care about the hbar's worth or error. The device has amplified the microscopic observable (of say a particle's spin) to one which is classical in scale (say the loud clicks of one of two particle detectors). (and this amplification is an irreversible thermodynamic process b.t.w.)

The critical question is "does a classical reality-model work adequately for the system?" if so then we have no problem treating the system classically. But we can also embed this classical system inside a large quantum one (the classical variables are a commuting subset of the larger class of quantum observables.) This is what we must do in order to see that classical measuring devices in a quantum universe is perfectly consistent and not dualistic.

Think of it in terms of the actuality of people living in the US and the construct of "Government" and "Law" which has a more objective behavior (ideally) than the actual people but which none-the-less is a function of and embedded within the world of people doing what people do. Note we also see in this analogy that when pushed to cases the idealized law breaks down (corruption,miscarriages of justice, civil disobedience, et al) because at the fundamental level the people are not just clockwork objects and thus their implementation of law is not perfect according to how the law itself defines "what ought to be". And yet to function as a society we must work with an objective system of government and laws, recognizing them as not the fundamental nature of us but a useful and necessary construct.

EDIT: So too I say is "reality" a useful and necessary construct but not fundamental.

 

[RUTA]

Speaking of Relativity of Simultaneity, our heritage of thinking in terms of objective realty goes back to the fiber-bundle structure of pre-Einstein space /time. Space/time was a http://en.wikipedia.org/wiki/Fiber_bundle" with time as the base and each slice of space (and the state of all within it) is a fiber indexed by this time base. Very much the continuum analogue of a movie with each frame a reality snapshot.

Relativity of simultaneity already begins to sand away at our old concept of objective reality. To preserve it in classical SR we invoke the frozen history of all past and future in a composite space-time. This of course doesn't allow choice or change except by selecting a whole new space-time universe.

In order to still speak of possibilities and probabilities and choices in a deterministic setting we invoke typically a classical field theory, again a fiber-bundle (space-time base with fibers the possible configurations of local reality and typically also some gauge degrees of freedom).

Now any time we see fiber-bundles we can be sure there is a relativity principle which may unify base with fiber and with is a group deformation of actions where the one-way dependence of fiber action on base coordinate becomes a two-way interaction. In this case matter affects the space-time just as space-time affects the matter and we get GR.

Well I'm getting off track... the point is that we already have good reason to soften up our traditional "objective reality" mindset and begin thinking in terms closer to the epistemological basis of science, the actuality of what we observe instead of the imagined state of reality with which we modeled it in past.

I don't see how you can distinguish base from fibers in the case of GR as you can with Newtonian space/time, since the local configurations are given by the stress-energy tensor (SET), the components of which require the notions of space and time (can define SET via variation of matter-energy Lagrangian with respect to the spacetime metric). The so-called "interaction" between base and fiber here is not a relation between distinguishable concepts. Therefore, I would say GR is rather a self-consistency criterion for the co-construction of the two.

 

[GeorgCantor]

It is not clear to me how you can get an "objective reality" at all from a theory in which it doesn't exist at the fundamental level.

Yes that seems to be true, at least in the perspective of quantum entities in time and space. There can be no 1 electron, if there was a universe with exactly one electron, then that electron has NO objective reality, i.e. it doesn't exist. It doesn't matter if you believe in decoherence, measurement-causes-collapse, MWI, BM or some other stuff. All these interpretations require a CONTEXT, i.e. they require relationships with other quantum entities. No interpretation that i know of can restore the objective reality of objects in time and space existing in and of themselves, without a context. In this view, local realism is not ruled out but dead as a complete picture of reality.

 

[DevilsAvocado]

Speaking of Relativity of Simultaneity, our heritage of thinking in terms of objective realty goes back to the fiber-bundle structure of pre-Einstein space /time.

All I wanted to say is that there is no way to tell if Alice or Bob does the measurement first – therefore they both have absolute non-deterministic "QM-probability-freedom" to do any measurement they like.

 

[zonde]

Relativity of simultaneity already begins to sand away at our old concept of objective reality.

It's not so much about concept of objective reality as about concept of rigidity of our measurements.
So we have good reason to soften up our traditional "rigid measurements" mindset.

 

[Geigerclick]

It's not so much about concept of objective reality as about concept of rigidity of our measurements.
So we have good reason to soften up our traditional "rigid measurements" mindset.

What do you mean?

 

[zonde]

What do you mean?

I mean that relativity of simultaneity applies to both measured object and measurement equipment. So we can't have non-contextual (rigid) measurement.
And I think that the same applies to QM measurements i.e. they are contextual.

Objective reality however means that we can fit it all together when we take into account contextuality of measurements. That is easily demonstrated is SR - all measurements fit together when using Lorentz transformations.

 

[filegraphy]

"There may be a measurement problem, but I doubt it is the problem you think it is. It is kind of like the problem of why there is more matter in the universe than anti-matter. Something it would be nice to understand, but not something that is actually in contradiction to theory." Dr. Chinese
In the book Antimatter by Frank Close, there is a process when matter and antimatter form that allows more matter to remain than antimatter.

 

[jambaugh]

I don't see how you can distinguish base from fibers in the case of GR as you can with Newtonian space/time, since the local configurations are given by the stress-energy tensor (SET), the components of which require the notions of space and time (can define SET via variation of matter-energy Lagrangian with respect to the spacetime metric). The so-called "interaction" between base and fiber here is not a relation between distinguishable concepts. Therefore, I would say GR is rather a self-consistency criterion for the co-construction of the two.

The point I inferred was that GR is less separated into fiber/base than field theory in SR. But still the definition of locality makes a distinction. One localizes with regard to the base space-time manifold and not with regard to to the fiber matter fields. (But note asymptotic freedom seems to localize in the momentum domain as well) We also --of course-- have the the fiber bundle structure of the tangent bundle on the manifold, but we may view this as our linearization of the description. Kaluza-Klein type theories show how the field/space-time may be unified into a space-time-gauge manifold. We then better see the tangent bundle structure as one of convention and not essence.

I presume then that brane theories attempt to quantize from this point. I'm inclined to think the objectification of space-time is a "wrong tract" and that a more Eulerian than Lagrangian description is "the way to go". But I haven't much in the way of example theories to suggest in that direction.

We can also view the classical probability description as a fiber-bundle with base, the state manifold and fibers of probability density. This is the heart of the Bell inequality derivation which is equivalent to the assumption that probabilities form a measure over a manifold of objective states of reality. No locality issues need apply. (And "rigidity" or its lack in measurement is not the issue.)

 

[jambaugh]

I mean that relativity of simultaneity applies to both measured object and measurement equipment. So we can't have non-contextual (rigid) measurement.
And I think that the same applies to QM measurements i.e. they are contextual.

This is true I believe but the QM case goes beyond that. With the classical SR case you get mappings from each observer's measurements, which we then see as perspectives on an objective whole. i.e. the length of an object is, in the whole, seen as one observer's cross-section of the object's world-volume (the locus of all space-time events associated directly to that object.) One still has an objective reality-history.

The contextual nature of quantum systems is in the objective reality (or reality-history) itself. We may have the objective reality of anyone set of commuting measurements. But one does not map given measurements to given measurements in the unitary transformations (in the sense of outcomes=values of observables). One rather maps certainty of measurements to probabilities of measurements, or in the more general case, map amplitudes to amplitudes some of which may take on the value of certainty. One has thereby abstracted from the objective (though relative) description of the system itself to the statistical description of our knowledge about how the system might behave.

Yes the math parallels but the "thing" upon which the relativity group acts is no longer the system state. It is the "state vector" or "mode of preparation" vector identifying a class of actual systems. One cannot narrow this class to the point of all systems acting identically under any possible measurement and thus one cannot speak of a instantiation of the class as being in an objective state of reality in that this state determines the outcome of all measurements exactly. Contextuality is an important feature of understanding the quantum description but there is more than that going on here.

 

[RUTA]

We can also view the classical probability description as a fiber-bundle with base, the state manifold and fibers of probability density. This is the heart of the Bell inequality derivation which is equivalent to the assumption that probabilities form a measure over a manifold of objective states of reality. No locality issues need apply. (And "rigidity" or its lack in measurement is not the issue.)

The forms of locality involved with violations of Bell's inequality are associated with the spacetime manifold -- causal and constitutive. The manifold of objective states of reality necessarily contain violations of one or both when the states are those of QM, but to "see" that locality is in jeopardy, one needs to go to spacetime. That's why so many physicists don't "get it," i.e., they work in Hilbert space where it all makes sense. I teach QM using both Heisenberg and Schrodinger formalisms, QM makes perfect sense. You have to move beyond playing with the formalism to appreciate the ontological implications (those highlighted in the popular literature). However, if you're only concerned with formal consequences, you have them -- if QM is right, GR can't be right because GR is both causally and constitutively local. Where do you fall on that issue?

 

[zonde]

Yes the math parallels but the "thing" upon which the relativity group acts is no longer the system state. It is the "state vector" or "mode of preparation" vector identifying a class of actual systems. One cannot narrow this class to the point of all systems acting identically under any possible measurement and thus one cannot speak of a instantiation of the class as being in an objective state of reality in that this state determines the outcome of all measurements exactly. Contextuality is an important feature of understanding the quantum description but there is more than that going on here.

Why would you require that all representations act identically under any possible measurement. Maybe you mean in exact manner?

Well in relativity contextuality means that it is quite useless to talk about preferred reference frame.
In QM contextuality makes it quite hard to talk about preferred measurement base.
But here is equivalence between representations of ensemble in different measurement bases. That's the idea of "state vector", isn't it?

But otherwise absence of preferred measurement base is of course only a small part of QM.

 

[jambaugh]

Why would you require that all representations act identically under any possible measurement. Maybe you mean in exact manner?

If many representations are representing the same physical entity then the many representations must transform isomorphically under the relativity group. But that is not what I'm talking about...rather the reverse. Two distinct categories of entities may transform isomorphically but this isomorphism does not imply they are the same type of entity.

In the case to which I refer, there is the objective reality of a classical object, the traditional observables of which are as you say "contextual" and as I would say relative. To each classical objective measurable quantity acting as coordinate the object's reality there is the act of measurement, the classes of experimental procedures which yield identical information about the state of that classical entity. These classes (of observations) necessarily transform isomorphically (or dually depending on the representation) to the objective reality they measure. They are none the less a distinct category of entities in the physics from the category of physical objective states.

Now hop into QM and you have the same (and typically a larger relativity group of) transformations acting on the classes of measurement actions/devices. You however lose the whole of the dual objective reality that in the classical case they were presumed to measure for a single physical entity. Instead you have each observation individually corresponding to an instance of actuality but only commuting subsets able to correspond to a single physical instance of the quantum entity.

Indeed in QM the term "system" refers more to the system of empirical actions on the physical entity, rather than to the entity directly. Contextuality is a prerequisite to this quantum non-objective actuality but (especially since contextuality can be invoked classically), by no means is it the sole defining characteristic.

Well in relativity contextuality means that it is quite useless to talk about preferred reference frame.
In QM contextuality makes it quite hard to talk about preferred measurement base.
But here is equivalence between representations of ensemble in different measurement bases. That's the idea of "state vector", isn't it?

(A side note, and repetition of one of my usual speeches)

The essence of a state vector is a maximal measurement of the physical system. Since this measurement is not classically total, it should no longer be referred to as a "state vector" but more properly (as you'll find in some literature) as a mode vector as in describing the mode of measurement or equivalently mode of preparation of the actual physical entity.

With that in mind, when we speak of the relativity group it again has passive and active context, i.e. we can rotate the physical entity or reverse rotate our measuring devices and achieve the same change of outcomes but in both cases we work within the same representation framework (of "state" vector i.e. measured values i.e. measurement processes) because we cease to have the "metaphysical" duality of measurement process+ objective state.

This is proper, and indeed imperative in the discipline of science since science is an epistemological discipline. Within the doctrine of modern science the observation is most fundamental component of a theory and not the objective state.

But otherwise absence of preferred measurement base is of course only a small part of QM.

Yes, that was principally my point.

Another note, in this non-object understanding of QM one can still be reductive in the sense of say reducing the behavior of the moon to the behaviors of its component elementary particles, however the contextuality gets "squared" in the treatment of composite systems. Not only do we have the sum of relativity transformations for the components, we have the product which implies there is not only a contextual aspect to how you measure the components to derive a quantity corresponding to the composite but also that there is a contextual aspect to how you actually subdivide the composite into components.

For a concrete example consider a ground state helium-4 atom. In subdividing into nucleus and 2 electrons we may speak of the spin z +1/2 electron and spin z -1/2 electron, or alternatively into spin x +1/2 and spin x -1/2 (since we know the electron pair is in a singlet [STRIKE]state[/STRIKE] mode ).

In the two cases we are subdividing the electron pair into two electrons in very distinct ways. This is something we must be conscious of when we parse e.g. EPR type experiments especially with our common language which has evolved to describe classical rather than quantum entities. This is where the conterfactuality landmine can trip us up.

 

[jambaugh]

The forms of locality involved with violations of Bell's inequality are associated with the spacetime manifold -- causal and constitutive.

To this I disagree strongly. Bell (and Einstein, Podolsky, & Rosen) invokes space-time locality only in so far as it enables him (them) to exemplify the more basic concept of independent acts of measurement. One can also derive, and then observe the violation of Bell's inequality by considering say two independent observables of a single localized particle. Assuming probabilities derive from a measure on a state manifold for the outcomes and assuming causal independence in the process of the two classes of measurements one may derive Bell's inequality. By entangling and then measuring one can demonstrate (or predict via QM) violation of the inequality.

For example one could take a spin-3/2 system (4 dimensional Hilbert space) and consider the cross commuting pair of observables constructable via complex superpositions from z-spin > 0 vs z-spin < 0, and separately |z-spin component| = 3/2 vs |z-spin component| = 1/2. The observables sets (block) commute which means they are causally isolated.

Of course as a practical matter it is terribly terribly difficult to isolate the two measurement processes. But for cross commuting sets one can in principle construct the devices to carry out actual experiments. Distance is the easiest means but not the only means.

The manifold of objective states of reality necessarily contain violations of one or both when the states are those of QM, but to "see" that locality is in jeopardy, one needs to go to spacetime.

It is a question of what assumptions one wants to make. ( I don't see locality as ever having been in jeopardy). Just as we empirically verify spatial locality to assure ourselves it is a valid assumption we can similarly verify that say gross position and spin measurements are similarly independent (commuting). And in both cases we may hypothesize that our assumption is wrong when we see Bell inequality violation and that there is some mechanism of interaction beyond the theoretical prediction that they are causally independent.

But QM predicts any pair of commuting observables may be none-the-less entangled and thus that you can both derive a Bell inequality from "reality assumptions" and that said inequality gets violated. It isn't about the locality! It's about the reality!

However, if you're only concerned with formal consequences, you have them -- if QM is right, GR can't be right because GR is both causally and constitutively local. Where do you fall on that issue?

I don't agree with that statement. GR and QM are perfectly compatible beyond GR being as yet still a classical theory. With regard to causal vs constitutive locality I fully believe in causal locality but am not sure how you mean constitutive locality. Especially I'm not sure "constitutive anything" is proper in QM if by that you are invoking an objective reality of the physical system or its constituents.

I think in the end, as long as by "constitutively local" one is referring to the ability to expand measurement processes into constituent local measurements (invoking superposition) and thus so that this becomes a further qualification on the causal locality of the measurements then your fine.

Said better one may postulate "a complete set of causally local observables."

If you mean otherwise then you may be reifing the wave-function further than I think is proper.

 

[zonde]

Indeed in QM the term "system" refers more to the system of empirical actions on the physical entity, rather than to the entity directly.

That is no different in relativity. In relativity physical entity is not described directly with length and time but rather with measurements of length (rulers) and measurements of time (clocks) and of course along with synchronization procedure for clocks instead of universal simultaneity.

But I agree that there is significant diference between limits of objective reality under relativity and QM.
Let's formulate it this way and see if you will agree that this is the key difference.
Relativity allow reductionism and that is one of the key parts of objective reality. You can split description of reality however you like and all parts will still obey the same isomorphism of different measurements.

Actually you say yourself something like that here:

Another note, in this non-object understanding of QM one can still be reductive in the sense of say reducing the behavior of the moon to the behaviors of its component elementary particles, however the contextuality gets "squared" in the treatment of composite systems. Not only do we have the sum of relativity transformations for the components, we have the product which implies there is not only a contextual aspect to how you measure the components to derive a quantity corresponding to the composite but also that there is a contextual aspect to how you actually subdivide the composite into components.

With that in mind, when we speak of the relativity group it again has passive and active context, i.e. we can rotate the physical entity or reverse rotate our measuring devices and achieve the same change of outcomes but in both cases we work within the same representation framework (of "state" vector i.e. measured values i.e. measurement processes) because we cease to have the "metaphysical" duality of measurement process+ objective state.

Actually I didn't mean that when I said that there is no preferred measurement base.
What I mean is that when we talk about non-commuting measurements ensemble is represented using two orthogonal vectors and you do not necessarily have preferred basis for representation of those two vectors. Actually you have preferred basis in case when in that basis one of the vectors becomes zero.
But then if we want to relate this back to relativity then you have preferred representation there in special case when you have reference frame where every object under consideration is at rest. In that reference frame all effects of relativity will disappear.

And there is another thing where have different viewpoints that prevents me from accepting your arguments about probabilistic measurements.
You talk about uncertainty of measurement and with that you imply that single entity of ensemble is fair representative of whole ensemble and you acquire some certainty of measurement only as statistical build-up of individual independent probabilistic measurements.
I say that there is more than statistics in QM and ensemble is not statistical ensemble (at least not always) but physical ensemble i.e. measurement of ensemble can acquire such certainties that are not possible for simple statistical ensemble. So I talk about certainty of measurement of ensemble (rate of clicks versus individual click).
That's a bit similar as we talk about length measurement of a stick instead of count of atoms along the length of a stick.

 

[RUTA]

Jambaugh, clearly you’re not familiar with the terminology of the foundations community. Let me provide the background via excerpts from “Reconciling Spacetime and the Quantum: Relational Blockworld and the Quantum Liar Paradox,” W.M. Stuckey, Michael Silberstein & Michael Cifone, Foundations of Physics 38, No. 4, 348 – 383 (2008), quant-ph/0510090 and arXiv 0908.4348 (accepted for presentation at PSA 2010, revised version under re-review at FoP).

Fm second paper:

In Healey’s language, strong nonseparability might be dubbed a kind of non-locality, not “causal non-locality” but rather “constitutive non-locality” (Healey, R.: Gauging What’s Real: The Conceptual Foundations of Gauge Theories. Oxford University Press, Oxford (2007), p 127). As he says, strong nonseparability strongly suggests physical property holism, i.e., “There is some set of physical objects from a domain D subject only to type P processes, not all of whose qualitative intrinsic physical properties and relations supervene on qualitative intrinsic physical properties and relations in the supervenience basis of their basic physical parts (relative to D and P) (Healey, 2007, p 125).”

From first paper:

In particular, the implied metric isn’t an “extreme embodiment of the separability principle” (D. Howard, in Potentiality, Entanglement and Passion-at-a-Distance, edited by R.S. Cohen et al. (Kluwer Academic, Great Britain, 1997), p 122).

As Howard notes in the following passage, one of the central debates between the founding fathers of quantum mechanics was over the conflict between the spacetime picture and the quantum picture of reality and how they may be reconciled (Howard, 1997, pp 114-115):

"The second striking feature of Pauli’s last-quoted paragraph is that it points backward to what was by 1935 an old debate over the nonseparable manner in which quantum mechanics describes interacting systems. The fact that this was the central issue in the pre-1935 debate over the adequacy of the quantum theory disappeared from the collective memory of the physics community after EPR….Einstein had been trying in every which way to convince his colleagues that this was sufficient reason to abandon the quantum path…But it was not just Einstein who worried about quantum nonseparability in the years before 1935. It was at the forefront of the thinking of Bohr and Schrödinger."

In today’s terminology we would say that the spacetime picture of relativity adheres to the following principles (Howard, 1997, pp 124-125):

Separability principle: any two systems A and B, regardless of the history of their interactions, separated by a non-null spatiotemporal interval have their own independent real states such that the joint state is completely determined by the independent states.

Locality principle: any two space-like separated systems A and B are such that the separate real state of A let us say, cannot be influenced by events in the neighborhood of B.

It is now generally believed that Einstein-Podolsky-Rosen (EPR) correlations, i.e., correlated space-like separated experimental outcomes which violate Bell’s inequality, force us to abandon either the separability or locality principle.

As Howard notes, Einstein thought that both these principles, but especially the latter, were transcendental grounds for the very possibility of science. Einstein’s spatiotemporal realism is summarized in his own words (A. Einstein, Deutsche Literaturzeitung 45, 1685-1692 (1924)):

"Is there not an experiential reality that one encounters directly and that is also, indirectly, the source of that which science designates as real? Moreover, are the realists and, with them, all natural scientists not right if they allow themselves to be led by the startling possibility of ordering all experience in a (spatio-temporal-causal) conceptual system to postulate something real that exists independently of their own thought and being?"

Minkowski spacetime (M4) is a perfect realization of Einstein’s vision but as Howard says (D. Howard, “Einstein and the Development of Twentieth-Century Philosophy of Science” to appear in Cambridge Companion to Einstein, from his website):

"Schrödinger’s introduction of entangled n-particle wave functions written not in 3-space but in 3n-dimensional configuration space offends against space-time description because it denies the mutual independence of spatially separated systems that is a fundamental feature of a space-time description."

In this sense, we agree with Howard (Howard, 1997, pp 124-129) that NRQM is best understood as violating “separability” (i.e., independence) rather than “locality” (i.e., no action at a distance, no super-luminal signaling), and we take to heart Pauli’s admonition that “in providing a systematic foundation for quantum mechanics, one should start more from the composition and separation of systems than has until now (with Dirac, e.g.) been the case” (W. Pauli, Scientific Correspondence with Bohr, Einstein, Heisenberg a.o., Vol 2, 1930-1939, edited by Karl von Meyenn (Springer-Verlag, Berlin, 1985), pp 402-404).
***************************************************

Given your postings to date, Jambaugh, I’m guessing you’ll fall into our camp, i.e., causal locality is maintained, QM is “right” and GR is “wrong” in that the separability of GR is only an approximation. So, what say you?

 

[jambaugh]

Jambaugh, clearly you’re not familiar with the terminology of the foundations community. Let me provide [...]

In Healey’s language, strong nonseparability might be dubbed a kind of non-locality, not “causal non-locality” but rather “constitutive non-locality”

Thanks for the translation. Yes I'm not familiar with "constitutive (non)locality" as a phrase. To my mind "nonseparability" is more encompassing since it reflects not just spatial issues. inseparability is clear enough in the subadditivity of entropy for quantum systems.

Separability principle: any two systems A and B, regardless of the history of their interactions, separated by a non-null spatiotemporal interval have their own independent real states such that the joint state is completely determined by the independent states.

But I am of the camp that feels even a single system A has no "independent real state" as such so this definition of separability fails from the start. (I think the issue being considered in defining separability vs nonseparability is one of trying to reconcile QM with a ontology... a futile quest IMNSHO).

"Locality principle: any two space-like separated systems A and B are such that the separate real state of A let us say, cannot be influenced by events in the neighborhood of B."

Here again I see a bias toward "statism";-) if you pardon the misuse of the term. Rather try:

Observational locality principle: An action carried out in region A spatially separated from region B can have no effect on measurements made in region B.

Probably I could word that better given time but you see the point. Avoid reference to operationally meaningless unobserved states of reality and stick to the operationally meaningful actions such as measurements.

Given your postings to date, Jambaugh, I’m guessing you’ll fall into our camp, i.e., causal locality is maintained, QM is “right” and GR is “wrong” in that the separability of GR is only an approximation. So, what say you?

Fairly accurate except I see nothing wrong with GR at its foundation, only in the categorization of the geometric model of GR as an ontological theory as opposed to being a model. The elimination of the gravitational force qua dynamic force is to my mind a "gauge condition" and the full power of the equivalence principle has yet to be invoked in attempts to quantize GR.

I bring this up because I think the separability of GR is a function of its typical geometric formulation (model) and not the theory itself when "properly" (i.e. operationally) interpreted.

 

[RUTA]

But I am of the camp that feels even a single system A has no "independent real state" as such so this definition of separability fails from the start. (I think the issue being considered in defining separability vs nonseparability is one of trying to reconcile QM with a ontology... a futile quest IMNSHO).

You have to have some ontology to do physics; the formalism is meaningless in and of itself. I teach QM using actual experiments, so all the formalism translates immediately to actual experimental configurations and measurement devices. Anyway, no ontology, no physics.

Probably I could word that better given time but you see the point. Avoid reference to operationally meaningless unobserved states of reality and stick to the operationally meaningful actions such as measurements.

And when I teach QM according to experiments, like I said before, QM is perfectly clear. It's not until you ask, "How can that be?" that you run into confusion.

Fairly accurate except I see nothing wrong with GR at its foundation, only in the categorization of the geometric model of GR as an ontological theory as opposed to being a model. The elimination of the gravitational force qua dynamic force is to my mind a "gauge condition" and the full power of the equivalence principle has yet to be invoked in attempts to quantize GR.

I bring this up because I think the separability of GR is a function of its typical geometric formulation (model) and not the theory itself when "properly" (i.e. operationally) interpreted.

And this is where we in the foundations community see a benefit to asking, "How can that be?" By understanding that QM is nonlocal and/or nonseparable while GR is local and separable, you see immediately that one or both have to be corrected in one or both respects. Our formalism has GR as a statistical limit to quantum physics when the approximation of separability holds. We developed our approach to QG via our interpretation of QM. So, while foundational issues may be irrelevant to you, they were sine qua non for us.

 

[jambaugh]

You have to have some ontology to do physics; the formalism is meaningless in and of itself. I teach QM using actual experiments, so all the formalism translates immediately to actual experimental configurations and measurement devices. Anyway, no ontology, no physics.

Not to my mind. The formalism can be meaningfully interpreted without invoking ontology of the system; e.g. "an electron" is the phenomenon of an electron detector going "click". Of course one must invoke an ontological description of the measuring devices and the records of measurements.

And when I teach QM according to experiments, like I said before, QM is perfectly clear. It's not until you ask, "How can that be?" that you run into confusion.

Right, and in teaching QM according to experiments you should be demonstrating that the formalism is operationally applied to the configuration of experimental devices. I assert that the confusion arises when one tries to push beyond that operational interpretation.

And this is where we in the foundations community see a benefit to asking, "How can that be?" By understanding that QM is nonlocal and/or nonseparable while GR is local and separable, you see immediately that one or both have to be corrected in one or both respects. Our formalism has GR as a statistical limit to quantum physics when the approximation of separability holds. We developed our approach to QG via our interpretation of QM. So, while foundational issues may be irrelevant to you, they were sine qua non for us.

I rather see QM as non-separable, causally local, while CM is separable, causally local. Classical GR is separable, causally local and a QGR should be non-separable, causally local.

 

[RUTA]

Not to my mind. The formalism can be meaningfully interpreted without invoking ontology of the system; e.g. "an electron" is the phenomenon of an electron detector going "click". Of course one must invoke an ontological description of the measuring devices and the records of measurements.

Exactly my point. In fact, in RBW the ontology is "there is no system" (other than the experimental equipment).

Right, and in teaching QM according to experiments you should be demonstrating that the formalism is operationally applied to the configuration of experimental devices. I assert that the confusion arises when one tries to push beyond that operational interpretation.

I start my QM course with the Mermin device, interaction-free measurement, delayed choice (Zeilinger and Aharonov have done some cool experiments that I show them), and the quantum liar experiment. Then we use the QM formalism to describe all these experiments. I have them read many articles, but the texts are Shankar and Albert. For example, we work through "Entangled photons, nonlocality, and Bell inequalities in the undergraduate laboratory," D. Dehlinger & M.W. Mitchell, Am. J. Phys. 58, Sep 2002, 903-910, in detail. We also reproduce all the results in Mermin's three AJP papers. So, the students see how QM works and why most physicists don't see anything "weird" about it. But, they also see that QM violates separability and/or locality, which strikes them as "weird," so they can appreciate all the "fuss" made over this fact.

I rather see QM as non-separable, causally local, while CM is separable, causally local. Classical GR is separable, causally local and a QGR should be non-separable, causally local.

Exactly what we believe. Our approach to QG can be described as non-separable Regge calculus. The manner by which this unifies physics is explained in 0908.4348. What is your approach to QG?