NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/sci/physics/21839

Internet Message Format | 1992-12-29 | 16.6 KB

Xref: sparky sci.physics:21839 alt.sci.physics.new-theories:2637
Newsgroups: sci.physics,alt.sci.physics.new-theories
Path: sparky!uunet!well!sarfatti
From: [e-mail] (Jack Sarfatti)
Subject: re: Budnik's thesis that Aspect's experiment does not show FTL influences.
Message-ID: <[e-mail]>
Sender: [e-mail]
Organization: Whole Earth 'Lectronic Link
Date: Tue, 29 Dec 1992 11:02:35 GMT
Lines: 366


Sarfatti comments on:
Subject: Aspect's experiment, was Re: Peer review ...
Date: 28 Dec 92 21:19:33 GMT
Followup-To: sci.physics
Organization: Mountain Math Software, P. O. Box 2124, Saratoga. CA 95070
Lines: 187


        Arguing with Jack Sarfatti is a bit like shoveling manure in a barn
full of horses. No matter how much you shovel today there will be just
as much to shovel tomorrow. It is not that Jack is unintelligent or does
not know physics. It is just that he is extremely careless. He seems to
confuse lack of scientific discipline with scientific adventurousness.
They are not the same thing. Scientific adventurousness is impossible
without scientific discipline.

        It is nonetheless sometimes useful to respond to him, because he
raises issues that others may also be confused about.

In article <[e-mail]>, [e-mail] (Jack
Sarfatti) wr
ites:
> > I do not what Aspect thinks but the two references that I am relying on
> > are:
> > J. D. Franson, Physical Review D, pgs. 2529-2532, Vol. 31, No. 10, May
> > 1985.
> > A. Leggett, Foundations of Physics, V 17, p. 875.
>
> *OK, Paul, good. Since I do not have easy access to these references, and
> since others interested in the truth here may not, please take the time
to
> type it relevant quotes from Franson and Leggett that you think explain,
> clarify, and support your point - a point that, so far, evades me since
 your idea about "delay" is like Merlin's many changing forms.*

I have previously posted some quotes and have included them at the end
of this article. I suspect it is your careless reading and not my
definition
of delay that is responsible for the `many changing forms'.

> > Both make it clear that Aspect's experiment was not conclusive.
>
> *But do they make it clear in your one sense of "dely" that I thought I
> understood, namely, the delay in the photon passing from a polarizer to a
> detector - then you changed it in midstream to the uncertainty in the
time
> of emission at the source - so anyone trying to follow you will be
totally
> confused!*

If you had read my article carefully you would have understood that I only
introduced the issue of the uncertainty in emission time because it was
relevant to how accurately you can measure the critical delay. The critical
delay is the time between when you change polarizer angles and this has
an observable effect on joint detections.

You know Paul it is not so much my "careless reading" but your carless
writing! For example your obscure sentence:
"The critical delay is the time between when you change polarizer angles
and this has
an observable effect on joint detections."

What is the fragment "The critical delay is the time between when you
change polarizer angles " supposed to mean? At first it looks like you are
going to define "critical delay" and then you never complete the idea. "the
time between when you change polarizer angles" - do you mean the switching
time between two alternative positions of one of the polarizers? Keep the
other one fixed. The important point is that this time, if that's the one
you mean, must be short compared to flight time of photon from source to
polarizer to get space-like interval - and it was in Aspect's experiment.

"and this has an observable effect on joint detections." Oh yeah? Then tell
us what it is explicitly.
"
> > I am not aware of any refutation of the arguments these two present. I
have
> >  read  Aspect's original papers and I think the arguments that Franson
and Leggett
> > make are correct.
-
> *What the hell are their arguments? And how do thir arguments connect
with
> whatever it is you are trying to say?*

Their arguments are that Aspect did not directly measure the critical delay
I just described and his experiment did not put a tight enough constraint
on
this delay to prove that locality was violated.

>
> > [...]
> > In my analysis this delay is negligible (i.e. short compared to flight
> > times from source to polarizer) and essentially irrelevant. Why do you
> > think it important?
>
> Unless you measure it directly you have no proof that locality is
violated.

No Paul - this is dead wrong.  Bell showed that the standard quantum
prediction of  (cos@)^2/2
is not consistent with Bell's locality inequality.  Aspect's experimental
data showed actual data that fit the standard QM prediction and did not fit
the Bell inequality prediction with a few standard deviations to spare if I
remember correctly. Furthermore, Aspect's acoustically driven switching
time for the polarizer orientations was short compared to flight time by a
factor of order 10 as I recall - so the interval between detections of any
two photons in the same pair is definitely spacelike. And I don't think
it's clear to anyone but you if your "critical delay" is the same as
Aspect's switching time.

> *Paul , I do not understand your point here at all. Can you draw a
picture?
> Can anyone else reading this explain what Paul means?

"I do not know what could be clearer. To prove Bell's inequality is
violated
you must show that changing the polarizer angles produces a change
in correlation values as predicted by QM and that these two events, the
change in polarizer angles and the resulting change in joint detections
have a space-like separation. The only way to do this is to measure the
times and distances between these events."

No Paul - your above paragraph is totally garbled and confused! I mean
specifically your remark:

" these two events, the change in polarizer angles and the resulting change
in joint detections
have a space-like separation."

Those are not the "two events" that must have a spacelike separation in
order "to prove Bell's inequality is violated" - what you talk about as
"events" are not "events" in the sense of special relativity. An event is a
point (i.e. a small local region) in flat spacetime. Your so-called events
are not localized. When you write "the change in polarizer angles" you are
refering to the nonlocal relationship between the far-apart polarizers one
for each photon in the backto back pair - unless you mean the change in
orientation of one polarizer in a short time - that would be an event. But
then your other event which is to be spacelike separated from this one is
"the resulting change in joint detections" - that's no event - that's a
delocalized prcess you have to compare data from both ends do a lot of
computing etc.

On the contrary Paul - the two events that must be spacelike separated are
the two irreversible detections (events 2&3) of both photons in the same
pair emitted from the same calcium atom in the same atomic cascade (even
1)!

                                                  \2              / 3
                                                    \            /
                                                      \        /
space-time diagram
                                                  a     \    /  b
                                                          \/
                                                          1

and the source must prepare an ensemble of these pairs all statistically
independent of each other so each pair has the same J= 0->1->0 wavefunction
(1/rt2)[|a+>|.b+> + |a->|b->] for any common (+,-) photon linear
polarization basis.
>[...]
> It is also central to my argument that QM is incomplete. You can measure
> the
> statistics of these delays, but QM does not predict what those statistics
> are.
>
> *Paul, I still do not understand which delays you are talking about and
how
> you would measure them in principle - you must be much more specific !*

"I do not know what could be simpler. You change polarizer angles and you
measure the time it takes for this change to have an observable effect."

This is operationally meaningless. How do you measure the "observable
effect" is it a change in coincidence rate or what?

Doing this with enough accuracy is not simple, but conceptually what
you need to do is very simple.

>[...]
> > You cannot know what any of the times are that you discuss.
>
> *Fine, Paul, I agree that in the actual experiment done by Aspect that
you
> cannot know those times. But, that is not the point, the point is you do
> not need to know them! And if you did need to know them, it would be a
> different experiment! Why do you want to know them? Aspect was trying to
> measure (cos@)^2/2 - or, rather, the deviation from it predicted by
Bell's
- locality inequality - and his curves are quite beautiful - and the
> correlation is over a spacelike interval between detections of both
photons
> in same pair - that's all that matters.*

Aspect showed that the joint detections had a space-like separation. This
is
*irrelevant* to the question of whether Bell's inequality is violated.

The events that must be space-like separated are the changes in polarizer
angles and the resulting effects on joint detections.

There you go again Paul. I'll reread Aspect's paper to try to see how you
get this insane notion. If you're right on this I'll eat my hat!  Does
anyone else out there agree with Paul on this specific point? Maybe I've
flipped universe and am in the one next door. I'm beginning to feel like
Alice in Wonderland.

Aspect estimated this delay based on flight time. This is not a legitimate
thing to do
in QM the way it is in classical mechanics.

>[...]
> *As Ronald Reagan would say "There you go again Paul!"  I thought we
agreed
> that Aspect did not measure, or need to measure (well maybe we do not
agree
> on the need part) the delays from polarizer to detector for each photon
in
> the pair - who gives a damn?

The measurements are of *macroscopic events* not of photon
positions.

The critical measurement is the delay between when a polarizer
changes angles and this has an observable effect.

You keep saying this Paul as if by repeating it we will know your meaning.

This must be measured
directly. One cannot make estimates based on assumptions about what is
happening at a microscopic level. After all the point of this experiment is
to
test the correctness of QM itself. Thus you should not assume QM in
analyzing
these experiments. Even if you assume QM you still do not when the
particles traversed the polarizers because you cannot know in QM where
a particle is at a given time unless you observe it *at that time*.

> *Yeah, Paul - but what does that mean operationally- what do you actually
> do -I mean in principle what would you try to have done? You are still
> being very vague - what does "between states" mean, for example?*

I repeat what can be simpler then changing polarizer angles and seeing
how long it takes for this to effect joint detections.

>[...]
> *But the point is, to get good measurements, you want to pulse the
emission
> of the pairs so that no photons are in the polarizer during the time the
> orientation is changing.

You do not need to this and cannot do this because you have no idea when
the photon is `in the polarizer'.

> Or you want to do it so fast that the noise
> introduced is very small compared to what you are looking for.

This is a constraint.

> Also these
- delays are not what Bell's inequality is all about - [...]

On the contrary they are precisely what Bell's inequality is about.

Paul Budnik

OK Paul I'll re-read what Aspect actually wrote before I comment on this
again - but I think you got it all wrong.
_______________

Anthony Leggett observed in a review of several books including
"Quantum Reality" by Nick Herbert:

        The one point on which I believe Herbert seriously misleads the
reader
        is his repeated and emphatic statements on pp. 230-235 that Bell's
        theorem, and the related experimental results requires superluminal
        responses at the $macroscopic$ level ...

Henry Stapp and Phillipe Eberhard agree with Herbert on this one - that
superluminal influences are necessary - I say influence not communication
so I think Legget is wrong if that's what he says.

         No it doesn't --- not if the word "macroscopic" has anything
remotely
        resembling its everyday use and we stick rigorously to experiments
        which have actually been done rather than extrapolating their
results
        according to our theoretical prejudices.
        ... if we look in detail at the geometry of the
        experiments, the lifetime of the intermediate atomic state,...,
        we see that even in the latest (Aspect) experiments ... the
        "macroscopic"(?) events ..., or at least an appreciable fraction
-       of them, $were probably never separated by a space-like interval$
...

So Paul you seem to be insinuating that Leggett is attacking Aspect as an
incompetent experimental physicist who has conned us all!

        [Foundations of Physics, V 17, p. 875]

Franson in discussing how long the delay might have been in Aspect's
experiment between when a polarizer angle changed and this had an effect
notes:

    The time interval over which the probability amplitude discussed
    above may simultaneously exist and interact in the experiment by
    Aspect, Dalibard, and Roger could conceivably be comparable to
    the 89-nsec lifetime[12] of the excited atomic state which produces
    the pair of photons. If the photon emission time remains indeterminate
    for this length of time, then it is plausible that the final outcome
    of the event may remain indeterminate for a comparable amount of time.

Franson goes on to suggest that it might even be indeterminate for the
coherence time of the lasers used to excite the photon source. These
times are long enough to allow local effects to produce the correlations
Aspect observed.

OK Paul, if you send me copies of the papers by Franson and Leggett I will
read them objectively but they seem cranky- but maybe I'm wrong. But
whatever it is you are trying to say you are still not saying it clearly.
It's curious that no one else on net has defended your position on this or
tried to explain it. If they have please e-mail it to me.

Mail the Franson&Legget papers to POB 26548 San Francisco, CA 94126  if you
will be so kind.

In the 1976 paper before the actual experiment Aspect writes that he has no
need for "monitoring the change in time of the orientations" ... "We
believe that these difficulties could be overcome by using optical
commutators. During a short time interval, the commutator CA directs
photons vA towards the polarizer I1;then its state changes and, during the
following period, it directs vA towards the polarizer I2. The commutator
works similarly with the photons vB independently of CA. The time intervals
between two commutations are taken to be stochastic, so that two states of
the commutator, separated by a time longer than the autocorrelation
time,are statistically independent. The autocorrelation time of each
commutator is taken as short as L/c; L denotes the the distance between the
commutators and c denotes the speed of light... The four joint detection
rates are monitored, and the orientations a1,a2,b1 and b2 are not changed
for the whole experiment."

So Aspect's whole idea appears to be totally different from what you have
in mind. My use of "switiching time" above corresponds to Aspects single
commutator "autocorrelation" time. I see nothing in Aspect's idea that
corresponds to your oft-repeated phrase "The critical measurement is the
delay between when a polarizer changes angles and this has an observable
effect."

Note that Abner Shimony writes:
"The polarization correlation test of Bell's inequality was first performed
in 1972 by S. Freedman and J.C. Clauser ...... the parameters of the two
polarization analysers were kept at fixed values during time intervals of
the order of a minute.... Nevertheless there was a loophole.for the
defenders of local hidden variables theories who could say that a minute is
enormously long compared to th time needed for a signal traveling at the
velocity of light to go from one polarization analyser to the other....
This loophole is blocked .. by the spectacular experiment of Aspect,
Dalibard and Roger (1982) in which the choice ... is made by electro-
acoustical devices within time intervals of 10ns, wheeas the time required
for a signal at light velocity to connect the two analysers is about 40
ns... But the data of Aspect ... conflict with Bell's inequality and
confirm the predictions of quantum mechanics. Consequently, Bell's
independence conditions must fail, in spite of the spacelike separation of
the tsts on photons 1 and 2"  p.p.386-7 NEW PHYSICS ed. by P Davies.
"