On 9/05/2016 10:45 pm, Bruno Marchal wrote:
On 09 May 2016, at 04:12, Bruce Kellett wrote:
On 9/05/2016 1:39 am, Bruno Marchal wrote:
Thanks Scerir. Very interesting.
On 08 May 2016, at 09:58, 'scerir' via Everything List wrote:
https://arxiv.org/abs/1501.03521
'Bell on Bell's theorem: The changing face of nonlocality'
Authors: Harvey R. Brown, Christopher G. Timpson
there are several interesting points here
ch. 9 - Locality in the Everett picture
ch. 9.1 EPR and Bell correlations in the Everettian setting
Nice.
I think that what we are trying to explain to Bruce is well summed
up in their section 9.1.2 (the Everett description of the singlet
state, case of non-align polarizer).
I have already discussed this in my reply to Saibal. The basic point
I would make again is that the splitting of the universal wave
function into separate "worlds" is an interpretive gloss that does
not actually alter anything in the theory.
As long as you don't separate them too much, as the wave describe a
pure state remaining pure all the time, the "world" are the
phenomenological views as seen by each doppelgangers involved. If not,
you lose the possible interferences in principle possible by quantum
memory erasure.
Quantum erasure is possible only in limited, tightly controlled
circumstances. Generally, decoherence into many environmental degrees of
freedom is irreversible FAPP.
Furthermore, 'who knows what about whatever' is also an irrelevance
as far as the universal wave function is concerned. If you are going
to work in the many worlds paradigm, then everything ultimately stems
from the unitary evolution of the universal wave function -- all else
is just interpretive gloss, of no fundamental significance.
This is the case for the discussion in section 9.1.2 of the paper by
Brown and Timpson. Their equation (9) contains all the relevant
results that set the universal wave function -- the additional third
measurement (or measurement-like interaction) leading to equation
(10) is, therefore, irrelevant. All that happens in eq. (10) is an
exchange of information -- but it is an exchange of information that
is already present in the universal wave function, no new information
is created at this point. Just like opening the box on Schrödinger's
cat, which is either alive or dead long before, looking changes
nothing. Eq. (10) is, similarly, just an interpretive gloss of no
fundamental significance. The important point here is that everything
is set in the universal wave function /before/ Alice and Bob meet.
The relative angle of the respective polarizers is set in the wave
function long before the light cones of Alice and Bob overlap, so
that relative angle is determined non-locally.
The universal wave function is not a local object --
I am not sure what does this mean. The SWE is linear which is a case
of extreme locality I would say.
Linearity does not entail locality. Where did you get that notion from?
the unitary evolution does not have any implicit notion of locality.
?
Locality is a human convention, and the universal wave function is
under no compulsion to take any notice of human conventions or
preferences.
The question is only: does Alice's measurement change something
instantaneously and physically at a distance? Obviously, this is not a
question of convention.
No, it is not a convention, and the violation of Bell and similar
inequalities shows that such non-local action is present -- the
measurement at particle 2 is not independent of what happened to particle 1.
There is a recent review of Bell non-locality by Brunner et al. (RMP 86
(2014) pp. 419-478) which takes non-locality as an established physical
result. This would be the position of most working physicists.
I see clearly that such action at a distance has to occur in all QM
with a physical collapse assumption, as Einstein saw already in 1927
at the Solvay Congress, and EPR-BELL-Bohm made testable. But if the
collapse is a first person view entangled with the particle in the
singlet state, I don't see any action at a distance occurring, even if
it looks like that for the person involved. I don't get your critic
of Brown and Timpson (9.1.2 in https://arxiv.org/abs/1501.03521 ).
The move from eq. (9) of section 9.1.2 to eq. (10) is unnecessary. In eq
(9) all the local measurements are complete; both Alice and Bob have
split into A(+) and A(-), and B(+) and B(-), for seeing + or - results
respectively. And these measurements were done with particular polarizer
orientations, so by the time we can write eq. (9), the relative
polarizer orientation, and the set of all possible results, are fixed.
Alice and Bob might still be spacelike separated, and Alice may not have
been split according to Bob's results, but that does not matter.
Everything that is needed for that splitting (if it occurs only after
the light cones intersect) is in place -- nothing new is added when the
light comes overlap and Alice and Bob exchange information about their
results.
Brown and Timpson state: "Following this third measurement-interaction
[leading to eq. (10)], which can only take place in the overlap of the
future light cones of the measurements at A and B, a definite outcome
for the spin measurement in one region finally obtains, relative to a
definite outcome for the measurement in the other. That is, we can only
think of the /correlations/ between measurement outcomes on the two
sides of the experiment actually obtaining in the overlap of the future
light-cones of the measurement events -- they do not obtain before then
and -- /a fortiori/ -- they do not obtain instantaneously."
But the universal wave function contains all the information about
outcomes and correlations long before the light cones overlap -- that
overlapping does not create any new information. The information might
not be available to Alice and Bob before overlap, but learning about
something does not create that thing -- the information gained generally
pre-exists (Schrödinger's cat is either dead or alive, long before we
open the box!). So the correlations are implicit in the universal wave
function as soon as Alice and Bob's measurements are complete -- the
wave function does not have to wait till Alice and Bob separately know
each other's results. So /contra/ Brown and Timpson, the correlations do
exist before the future light cones of the measurement events overlap.
They might not be known before then, but they certainly exist before
then because nothing happens at the exchange of information between
Alice and Bob that can cause the particular relative polarizer
orientation to suddenly spring into existence, and the consequent
probabilities for each of the four possible worlds to suddenly
materialize; on the contrary, the relative orientation and the
probabilities are built into the universal wave function by the
non-local interaction between the two separated measurement events.
In the MWI literature, too much is made of the fact that Bob's results
are indeterminate for Alice until she hears from him by a classical
channel. As before, merely learning about something does not cause it to
spring into existence -- Bob has split into disjoint branches for his
possible measurement results, and each of Bob's branches is duplicated
in all of Alice's branches. That is what the linear evolution of the
wave function tells us -- that is the result of eq. (9) in the paper.
Alice and Bob may only self-locate on one of these branches after
information exchange, but that does not create these potential worlds.
Brown and Simpson are close to my feeling (say), which is that Bell's
inequality violation testing does not test locality, but the MWI itself.
I think that is why you are so resistant to seeing that there is a
better account than that given by Brown and Timpson, and many others.
The reality of non-local effects does not necessarily spell the end of
the MWI -- since MWI is only an interpretation of quantum mechanics, and
gives exactly the same results as any other interpretation. Non-locality
no more kills off MWI than non-locality kills off any other
interpretation, collapse or non-collapse. I have given a perfectly
coherent account of non-locality within the Everettian paradigm. In
fact, I have shown that such non-locality is indeed necessary even in
that paradigm. The conventional MWI argument does not actually account
for the correlations at all -- because it does not explain how the
relative orientation of the polarizers enters the evolution of the
universal wave function.
Bruce
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
