On Thu, Jun 21, 2018 at 11:48 PM, Bruce Kellett
<[email protected] <mailto:[email protected]>> wrote:
From: *Jason Resch* <[email protected]
<mailto:[email protected]>>
On Thu, Jun 21, 2018 at 12:56 AM, Bruce Kellett
<[email protected] <mailto:[email protected]>> wrote:
There are only two photons, but each has two possible
polarizations. When you measure the polarization, you split
into two branches, one for each possible result. The
partner photon reaches the other person on each of your
branches, but if everything is purely local, the photon that
is remotely measured cannot know which result you obtained
(it cannot know which of your branches it is actually on),
so it has indeterminate polarization, and when measured,
there is necessarily equal probability for either result.
I think this is the heart of our disagreement. You are seeing the
entangled photons are still distinct objects without a correlation.
The entangled photons are a non-separable unity. But if everything
is local, the measurements by Alice and Bob, being space-like
separated, must be equivalent. If Alice splits into two branches,
so must Bob. The correlation arises because the entangled pair is
not a local object -- it has no purely local description.
Something that spreads at <= speed of light, and effects and interacts
with only the local particles/fields is not a non-local phenomenon.
Here the photon pair (under many worlds) matches both of these
qualifiers perfectly.
There are two events where some human experimenter gets entangled
with a photon (you could saw under MWI that two splits occur).
Now I think you ask why does the second measurement know to
"split the right way", but it doesn't, and doesn't need to. Both
experimenters contagiously contract the superposition of the
photon they measure (entangle themselves with).
What on earth does that mean? "contagiously contract the
superposition". It can only mean that the superposition is
non-local, and that you are actually making use of this
non-locality without being aware of it.
The superposition starts with the creation of the photon pair, and
spreads to everything that measures/interacts with it, and anything
that measures/interacts with the thing that measured it, ad infinitum.
This means that the photon that is on the branch in which
your photon passed the polarizer can either pass the remote
polarizer, or be absorbed, with 50% probability for each.
Similarly for the photon that is on the branch in which your
photon was absorbed. The outcome by considering both branches
is four possible worlds, one for each combination of 'pass'
and 'absorb' results. Two of these worlds violate angular
momentum conservation. How do you rule out these worlds with
only local interactions?
The photon pair was created at one point in space time, it
traveled only at light speed to two locations, where its
superpositional state became entangled with the local environment
at its point(s) of measurement.
To say there are 4 possible worlds here, I think is to assume
measuring the same photon twice using the same polarization
angle, can produce inconsistent results.
You are confusing measurements on the entangled pair with repeated
measurement of the same single photon. The entangled state is a
unity, but it is not the same as a single photon.
They are analogous, and by rotating the picture of space-time when
looking at the electron-positron pair in the original EPR thought
experiment, you can view the electron-positron pair as the same
particle. In the conventional view we see it as a pi meson decaying
into e- and e+, with opposite spins. But rotate things about
90-degrees and you see it as an electron interacting with a Pi meson
and changing directions.
Look at it this way. The two measurements are made at
space-like separation. If everything is local, the
measurements must be independent.
There is where I disagree. The actions are independent, but the
results are not.
Then there must be a non-local effect! If the measurements are
made independently at a space-like separation, there can be no
correlation without either a common cause or non-locality. Common
cause is ruled out by the statistics of repeated measurements of
such entangled pairs at different angles.
Common cause isn't ruled out by Bell for measurements that have more
than one outcome.
If the measurements are independent they cannot be correlated
-- that is one possible operational definition of independence.
But when measuring an entangled photon pair, they must be correlated.
Exactly. So how did this correlation arise?
From the time of the pair's creation, each element of the the pair has
already measured the other. Each element of the pair (while in this
superposition) proceeds at <= light speed to a location where it will
be measured. Since each element of the pair has already measured the
other, measuring either element is equivalent to measuring both (or
you might say the same particle, under Feynman's view of antiparticles).
Since the measurement results are known to be correlated,
they cannot be independent. Since there can be no sub-luminal
interaction between the two measurements, this correlation
can only be a non-local effect. In the case that I have been
discussing, quantum mechanics predicts 100% correlation.
There is no way this can be achieved locally because the
singlet you are measuring is rotationally symmetric and has
no intrinsic polarization state that can be carried
subluminally between the experimenters.
In other words, the structure of the singlet state rules out
a common cause explanation for the 100% correlation. Bell's
theorem then rules out any /local/ hidden variable explanation.
Look, the singlet state is:
|psi> = (|+>|-> + |->|+>).
When Alice makes her measurement she effectively splits this
state into the |+>|-> state on one branch, and the |->|+>
state on the other branch.
I would not say she splits the state, I would say she splits
herself, by now becoming part of the state. The super position
never goes away so you get two Alices: (Alice * |+>|->) +
(Alice |->|+>)
The "(Alice * |+>|->)" knows that Bob she will hear from who
performs the same measurement of the photon of the entangled pair
will be the Bob that sees the - photon, and "(Alice |->|+>)"
knows the that the Bob she will hear from who performs the same
measurement of the photon of the entangled pair will be the Bob
who sees the + photon.
How does she know this? And why can't the Alice who sees the |+>
not hear from the Bob who sees the |+>.
|+> Alice is no longer in the same branch as the one containing |+>
Bob. They're in the same superposition though.
Both must exist if everything is purely local, you know.
Both exist, but the |+> Alice is in the branch where the |-> photon
belongs to "her" Bob.
But if everything is purely local, this split does not happen
for Bob before he makes his measurement.
True Bob has not yet entangled himself. But you speak of splits
as if they are instantaneous things that create two whole
universes instantly. This is not what MWI predicts.
We are doing measurements with people and macro apparatus that is
not isolated from the environment. So the decoherence following
the measurement splits the world -- locally, of course, but that
is all that is required. Bob is at a space-like separation, so the
split occasioned by Alice's measurement has not encompassed him at
the time he makes his measurement. Is that so hard to grasp?
I agree with this.
So he, too, measures the original entangled |psi> state, and
he also must have 50% probability of either result. However,
quantum mechanics says that when Alice measures |+>, Bob
necessarily measures only the |-> component of his photon;
and when Alice measures |->, Bob necessarily measures only
the |+> component of his photon.
Correct.
This is how the correlation comes about. But this is
non-local -- the non-separable initial state is separated
non-locally by the measurements.
But they're measuring (entangling themselves with) the same
superpositional state: the entangled (mutually already measured)
photon pair.
What you say here makes no sense. Alice cannot entangle herself
with Bob's photon when she makes her measurement -- it is at a
spacelike separation. Or rather, any entanglement that does occur
must be non-local. The non-separability of the state means the
correlations are non-local in origin.
I am not sure what is meant by "non-separability of the state".
If I send a photon through a filter orientated at 0
degrees, and it passes through, and goes all the way to
Pluto where you measure the filter at 0 degrees and it also
passes, you would not say this violates locality, would you?
No, because its passage to Pluto is at the speed of light.
Same as with the traversal of the photons in the photon pair, is
it not?
If you think these cases are equivalent, then I am sorry, but
there is nothing more I can say that will ever convince you of
your error.
You deleted my quote from Feynman. Did you disagree with what he says?
One obvious problem with your attempt to equate the measurements
on the entangled pair with repeated measurements of a single
photon is that in neither case do you make any use of many-worlds
to explain the EPR-type correlations.There is no violation of
counterfactual definiteness, no use is made of measurements that
could have been performed, but were not.
The EPR-type correlations were explained in the paper I provided, but
you said this was "being opaque" and adding confusion. It was your
suggestion to keep things simple with an identical orientation for the
polarization filter.
So what, exactly, is MWI giving you in your purported explanation
of EPR that is not already available in a collapse model?
Determinism, locality, time-symmetry, reversibility, QM obeyed at all
times, no classical-QM cut off point, the absence of FTL effects or
influences, and a simpler theory of QM.
Jason
