Re: photoelectric effect : hypothetical experiment

> QO people adjust for well understood physics (lossy
> detectors) in order to see how well the rest of their
> experiment matches the predictions of QM. Since the
> agreement from following this procedure is good,
> this is taken as evidence that QM is a reliable theory.

Thanks for offering one more illustration of a typical 'QO sleight of
hand' -- pretend that the fundamental QO subtractions (which are built
into the very definition of Glauber's filtered correlation functions
Gn()) are due to some kind of minor and temporary technological
imperfection, to be overcome soon. As if the actual Gn()'s are meant to
predict the actual counts & their correlations (the kinds that Bell
inequalities talk about). They are not. The QO "correlations" (the
Gn()'s) are by definition "filtered signal" functions, with
subtractions built into their definition (which in turn makes them
irrelevant, however useful they may otherwise be in optical
engineering, for the B.I. violations). They don't predict (nor are they
meant or derived as a prediction of) what the _full set of detection
and non-detection events_ (the stuff that matters for the B.I.) is
supposed to be. That limitation of the coincidence correlations Gn() to
subtracted counts is not some transient technological artifact but it
is their very definition. I just posted another longer comment on this
in sci.physics.research:

1.
http://groups.google.com/group/sci.physics.research/msg/7ee770990a31bd3f

which explains it in more detail and points out exactly how this 'QO
sleight of hand' works, with a concrete illustration from the well
known Ou & Mandel's 1988 "observation" of Bell inequality violations
(the mother of all modern PDC based B.I. "violations" experiments).

> So maybe some SED theory is better. So, to convince people
> we need a way to test cases where QM and SED differ, and
> show that SED gives better predictions. Most SED proponents,
> in my experience, prefer instead to go on about loopholes in
> the Bells test experiments, in order to show that SED has not
> been completely ruled out.

The SED, in its practical Stochastic Optics (SO) form, is a limited
scope effective theory (it is an approximation to A. Barut's Self-Field
ED, useful mostly for the optical photons phenomena and unable to
model, among others, the QED radiative corrections), which in various
cases is more or less convenient than Glauber's QO formalism (which is
a subset of the 1920s Old QED of Dirac, Heisenberg & Jordan, minus the
2nd quantization of the Dirac fields, plus the Einstein's lightquantum
heuristics and imagery). That issue (which is better, the SED/SO or the
QO) wasn't the point at all, though.

Neither the Glauber's QO nor the Marshall-Santos SO are fundamental
theories. But the SO at least doesn't pretend that it is and it doesn't
put out the pretentious (and fake) QM magic shows, as a particular
group of Quantum Opticians has been doing for the last half a century.
The SO and QO are applied physics/engineering computational techniques
covering roughly equal domains of phenomena. There are no fundamental
effects there, be it in their computational algorithms or the
observations within their domain, such as genuine observations or
predictions of Bell inequality violations or photon anticorrelations,
that would constrain the more fundamental theories (QFTs, Self-Field
ED, gravity or string theories,... etc). As explained in more detail in
[1], these "magic" effects don't exist at any level -- there are
neither observations nor QO/QED predictions of such effects.

Note also that the typical QM "predictions" of Bell inequality
violations that one finds in pedagogical & popular "proofs" are merely
suggestions/heuristics on how one might deduce genuine predictions
within the theories closer and more appropriate to a given domain (such
as QED/QO or SO for optical photons). Namely, the general QM postulates
about the existence and formal properties of observables cannot, due to
their generality, specify the concrete operational rules which map the
readouts of particular instruments to the values of the corresponding
formal observables.

Particularly critical missing piece of these operational mappings
(which the general QM postulates could at best only assert to exist)
for the B.I. violations is the question on what constitutes, in terms
of the readouts on instruments, a realization of valid measurement of a
value of some observable S (such as spin or polarization, or their
products for composite systems) in a given setup, i.e. what are the
"valid tries" (in terms of readouts) yielding eigenvalues of S and what
are the rejections (failed measurements). The general QM postulates
merely say that the operational rules _exist_ which perform such
mapping between the eigenvalues of S and the readouts of the
instruments, and make the decisions on what is a valid and what is a
failed measurement, but they cannot say how exactly these mappings
work, and most importantly for the B.I. violations, what kind and what
proportion of the readouts are rejected as invalid realizations of the
"measurement" of the formal observable S.

To predict the B.I. violations it is critical (due to the enumerative
nature of the inequalities) to predict that the rejections of the
readouts are below approximately 17% of all actual Bell EPR pairs
considered for violation. And to make that kind of prediction, which is
the very essence of the B.I. constraints, one needs much more specific
theory for the particular experimental domain. In the case of optical
photon experiments, that more specific theory, providing the concrete
operational rules for mapping between the formal observables and the
instrument readouts, is the Glauber's photo-detection theory (model),
specifically his model for the n-detector correlations for optical
photons (cf. pp 84-88 in ref [5] of post [1]). As explained in [1]
regarding the Ou & Mandel QO "prediction" based on the Glauber's
n-detector theory, the QO doesn't predict any B.I. violation for
optical photons (neither does the SO, but the SO doesn't pretend to be
doing it). The corresponding experiments agree with the QO (and SO)
predictions and don't show any violations either.

One has to take the hat off to Clauser, Aspect, Grangier, Mandel,
Zeilinger, Chiao, ... and other masters of the QO magic shows, for
their chutzpah and a kind of twisted genius to misrepresent to the
physicists this kind of double failure to violate Bell inequalities,
the complete theoretical and experimental debacle, as the double
success, based merely on the fact that the theory and the experiment
indeed agree (that there is no violations, cf. Ou & Mandel paper
discussion in [1]). The deepest bow, though, goes to the founder of
modern QO, the great (in a twisted genius kind of way) Roy Glauber, for
the masterful design and construction of the verbal and formal props,
the gear behind the curtains of the QO magic shows (in his two 1963
Phys Rev papers & his 1964 Les Houches Lectures), which had
single-handedly turned around the greatest embarrassment and defeat for
the "magic show" branch of Quantum Optics, the Hanbury Brown & Twiss
"controversy" of 1956, into the greatest stretch of PR victories and
glory (however transient and hollow it all will eventually appear in
the eyes of the history) for the QO magicians, going on to this day.
The show has probably few more years to go, provided some cocky
youngsters (such as the guys of ref [4] in [1], with their ridiculous
377 standard deviations) don't get overly greedy and careless and yank
the curtain down before its time.

> Adding ZPF to the classical EM theory might give a model capable
> of replicating predictions made by QM, but it doesn't guarantee
> a better or more convenient way of making predictions. Further,
> the Yariv chapter is all about how the SED-ish theory can play
> catch-up to a QM model, which is hardly the most convincing
> case for SED.

The existence of the SED/SO models for the alleged experimental B.I.
violations and for the photon anticorrelations is simply another, more
constructive way, to show that the QO non-classicality claims are
bogus. In that role, as a direct counter-example, SED/SO doesn't need
to be "better" or more practical than QO (just as Bohm's QM, although
not any better than regular QM, was a counter-example for the von
Neumann's HV impossibility "proof", which only later Bell dismantled
directly). It just needs to exist as a "natural" (non-contrived; it's
the theory Planck & Nernst proposed way back in 1911-1916) local theory
capable of, among others, modelling the allegedly "non-classical" QO
experimental data classically. Careful reading of the Glauber's
founding papers of QO and of their operational interpretations by the
QO experimenters (such as Ou & Mandel's 1988 paper, as sketched in [1])
leads more directly to the exactly same conclusion -- the complete
failure of QO, on the theoretical and the experimental levels, to
demonstrate existence of any such non-classical effects in the optical
photon domain.

> Make an SED prediction that differs from QM, and propose
> a loophole-free experiment to test it.

The SED/SO prediction (see for example the papers of T. Marshall & E.
Santos on quant-ph) is that there is no B.I. violation with optical
photons. That is precisely what the experiments have shown so far.
There are no loopholes in those experiments. They show exactly what the
applicable theories, be it SED/SO or QED/QO, predict for the setup (cf.
[1] on the distinction between the facts from the 'QO sleight of hand'
kind of depiction of the experimental & theoretical facts). As
explained above the general QM doesn't predict anything specific enough
to say whether B.I. will or will not be violated in these experiments.
The general QM postulates simply lack the sufficient quantitative
specificity of the operational rules, the level of specificity required
by the enumerative nature of Bell inequalities (such as the 83% or more
retained pairs), to say anything definitive about the B.I. violations.

The best that the general QM postulates can do (which is what Bell did
in his papers) is to point in the general direction of some types of
phenomena, such as the Bell EPR pairs, which are candidates for
sufficiently specific quantitative predictions. But for any specific
realization of such Bell EPR pairs you need domain specific theory of
the phenomena, especially of the instruments, which has sufficient
quantitative precision to address the critical questions essential for
the B.I. constraints. With the optical photons and the detectors with
the Poissonian photo-electron counts (the best type of detection you
can get for the optical photons), you cannot ever violate Bell
inequalities, not even using the photo-detectors with 100% "Quantum
Efficiency" (note the Q.E. number doesn't count the 'dark counts' which
are relevant for the B.I. rejection limits of max 17% pairs). The most
Bell EPR pairs retained you can get with the Poissonian p-e detectors
is 1-1/e = 63.21% which is just below the limit of the natural SED/SO
models for the Bell EPR setup, which work provided there are at most
2/Pi = 63.66 retained pairs. See the related PhysicsForum post for more
details & references:

2. http://www.physicsforums.com/showpost.php?p=540771&postcount=99

{ Submitted to sci.phys.research on Sept 23, 2005, 22:08 EST }

.

Quantcast