From: SMITRA <[email protected]>
On 16-07-2018 23:04, Brent Meeker wrote:
On 7/16/2018 8:18 AM, Bruno Marchal wrote:
I would like to think that this were the case, but you keep coming
up with irrelevancies that contradict the straightforward account of
these phenomena. If you forget about the metaphysics and just
concentrate on Alice and Bob making real measurements and recording
them in their lab books, then all these superfluities vanish. There
are no counterfactuals, no worries with other unobserved worlds, and
Bell's theorem goes through exactly as he intended. Many-worlds does
not invalidate Bell's argument. In fact, deflecting Bell's theorem
would do no more than allow for the possibility of a local hidden
variable account. That alone does not prove that many-worlds is
local -- that would still have to be established by developing such
a local hidden variable theory. No one has to date developed such a
theory. But since Bell's theorem has not been deflected, we do not
have to worry about such contingencies.
So we really agree. You have been probably misguided when trying to
defend John Clark who claimed that there are still FTL influence in
Everett, when the Bell’s inequality relations implies FTL only
when we assume unique outcomes of the experiences (i.e. some
collapse, or Bohm’s type of hidden variable).
No need of patronizing remark either, especially when rephrasing
what I was just saying. If you agree that there is no FTL in the
many-worlds, we do agree, that was the point I was making to J.
Clark. Not sure why you defended it, especially that you have shown
implicitly that you have no problem with the step 3 of the Universal
Dovetailer Paradox. You might eventually understand that with
mechanism, Everett’s task is still incomplete, as we need to
justify the wave from all computations, as seen from some
self-referential modes (fortunately and constantly implied by
incompleteness).
Not to reignite the argument, but it originated because Bruno claimed
that MWI does away with non-locality in QM.
Brent
It reduces the non-locality to trivial common cause effects. Bruce
has been trying to prove that it doesn't by invoking the argument that
you can pick a single branch where Alice and Bob wrote their
measurement results in their lab books, and that one should therefore
be allowed to apply Bell's theorem by pretending that the other
branches do not exist and reach the same conclusion as in collapse
theories. However, one has to ask here what the violation of Bell's
inequalities implies. It only constrains extensions of "standard
instrumental QM".
It has become clear that the real argument by advocates of MWI is
that many-worlds deflects Bell's theorem, so that its implications do
not apply in MWI. I have, as Saibal points out, been arguing against
this, and I still consider my proof that selecting one branch out of
the MWI superposition is sufficient to apply Bell's theorem in its
full rigour. Since the argument applies to any branch, it applies to
the superposition as a whole, and MWI does not avoid the implications
of Bell's theorem. The implication is that no local hidden variable
theory can account for the observed EPR-type correlations. In
particular, any common cause, or Bertlmann's socks type argument,
fails in MWI for the same reasons that it fails in a single world
account.
I have no idea what Saibal means when he claims that Bell's theorem
only constrains extensions of "standard instrumental QM". Saibal has
not offered any convincing counter argument to my proof that Bell's
theorem applies in MWI.
If we assume that, in general (and not just in case of Bell-type
experiments) measurement results are deterministic, that they are
specified by hidden variables, then the violation of Bell's
inequality implies constraints on such theories. Such theories must
necessarily be non-local. But then there is no evidence for a hidden
variable theory, so there is no need to invoke non-locality on these
grounds.
That does not follow. Just because you think you have shown that
Bell's theorem does not apply, it does not follow that MWI is thereby
local, or that a local account of the correlations is available in
MWI. Similarly, the claim that there are no hidden variables says
nothing at all about whether reality is local or non-local.
Now, what is true is that if Alice and Bob perform measurements on
entangled spins such that their results are perfectly correlated and
they are space-like separated, that the non-existence of local
hidden variables has a non-local aspect to it because Bob has the
information about what Alice will find and the non-existence of
local hidden variables rules out that this piece of information is
not somehow present locally at Alice's location.
??
But this non-local effect is entirely due to a correlation mediated
by the entangled spins, in the MWI this is a common cause effect,
Bell's theorem rules out Bertlmann's socks as an explanation. You
seem to accept that my argument establishes that Bell's theorem is
valid in any branch of the MWI superposition. It follows that
Bertlmann's socks is invalid in all branches, so the possibility of a
common cause effect is ruled out in MWI, just as it is ruled out in
any single-world interpretation.
while in the Copenhagen interpretation it cannot be explained in
that way.
Why not? Many-worlds and single-worlds are entirely equivalent in
this respect -- they both have the same wave function, after all!
Bruce's elaborate argument about verifying the violation of Bell's
inequality in single branches doesn't change that conclusion. Yes,
you can verify that Bell's inequality is violated in single
branches, but as pointed out above, that violation is part of the
argument why in the MWI the non-local aspects of entangled states
are completely trivial.
That does not follow. If you want to advocate a common cause
explanation in MWI you have to actually develop the corresponding
dynamical theory-- i.e., what is the common cause and how is it
implemented? If you simply appeal to the nature of the singlet wave
function for entangled states, then you have not given a local account
since, as Maudlin correctly points out, the wave function itself is
intrinsically non-local. It seems that you are talking out of both
sides of your mouth here -- QM is non-local, but the non-locality is
entirely trivial in MWI?????
Bruce
