Bell Correlations At 6 Miles
Kevin Brown
12:41 PM ET 06/26/97: Light particles communicate over six miles
LONDON (Reuter) - What Albert Einstein described as a spooky
quality of light just gets spookier, according to Swiss scientists.
They have found evidence that light particles known as photons are
somehow linked over a distance of six miles, New Scientist magazine
reported Thursday.
It said Nicolas Gisin and colleagues at the University of Geneva
generated pairs of photons and sent them along two separate optic
fibers, ending up six miles apart. ...measurements of one photon
affected the results of measurements of the other. "There was no
way the two photons could have communicated over such a distance,
meaning they must have been "entangled" Gisin said.
Alain Aspect and colleagues at the University of Paris first
showed that this does indeed happen, but in particles no more
than a few yards apart. "Gisin's experiment is really significant
because he shows that he can maintain these amazing quantum
correlations over a very long distance" Aspect told New Scientist.
John Baez
In article <33b492a7....@news.seanet.com>,
Kevin Brown <ksb...@seanet.com> wrote:
>12:41 PM ET 06/26/97: Light particles communicate over six miles
>LONDON (Reuter) - What Albert Einstein described as a spooky
>quality of light just gets spookier, according to Swiss scientists.
>They have found evidence that light particles known as photons are
>somehow linked over a distance of six miles, New Scientist magazine
>reported Thursday.
I'm still waiting for Budnik to pay me off on that bet.
e...@panix.com
ksb...@seanet.com (Kevin Brown) wrote:
> 12:41 PM ET 06/26/97: Light particles communicate over six miles
Is there any fuller reference? Or is this 'publication in the press'?
-------------------==== Posted via Deja News ====-----------------------
http://www.dejanews.com/ Search, Read, Post to Usenet
Phil Gibbs
In article <8675368...@dejanews.com>, e...@panix.com writes
> ksb...@seanet.com (Kevin Brown) wrote:
>
>> 12:41 PM ET 06/26/97: Light particles communicate over six miles
>
>Is there any fuller reference? Or is this 'publication in the press'?
The article just says that the results are "yet to be published". I
suppose we will just have to wait to see if they checked the
timing well enough to satisfy PB.
Phil Gibbs
http://www.weburbia.com/ http://www.weburbia.demon.co.uk/
"When all you've got is a hammer, everything looks like a nail"
Paul Budnik
Kevin Brown (ksb...@seanet.com) wrote:
: 12:41 PM ET 06/26/97: Light particles communicate over six miles
: LONDON (Reuter) - What Albert Einstein described as a spooky
: They have found evidence that light particles known as photons are
: somehow linked over a distance of six miles, New Scientist magazine
: reported Thursday.
: It said Nicolas Gisin and colleagues at the University of Geneva
: fibers, ending up six miles apart. ...measurements of one photon
: affected the results of measurements of the other. "There was no
: way the two photons could have communicated over such a distance,
: meaning they must have been "entangled" Gisin said.
Distance itself is not an obstacle to communication. Depending on
the details of the experiment this may be a strong argument against
a local explanation but the devil is in the details and making such
absolute statements is rather dangerous.
One must always keep in mind that if QM is wrong about quantum entanglement
then there is something happening that is completely outside of our
existing understanding of physics. Making absolute statements about
something one is completely ignorant of is particularly dangerous.
--
Paul Budnik
pa...@mtnmath.com, http://www.mtnmath.com
Ray Tomes
ksb...@seanet.com (Kevin Brown) in article
<33b492a7....@news.seanet.com> wrote:
>... according to Swiss scientists.
>They have found evidence that light particles known as photons are
>somehow linked over a distance of six miles, New Scientist magazine
>reported Thursday.
No they haven't. They have shown that they are correlated.
Any other conclusion is based on faulty logic and bad science.
-- Ray Tomes -- rto...@kcbbs.gen.nz -- Harmonics Theory --
http://www.kcbbs.gen.nz/users/rtomes/rt-home.htm
-- David A. Elm
Paul Budnik <pa...@mtnmath.com> wrote in article <5p5bqk$6...@mtnmath.com>...
>
> One must always keep in mind that if QM is wrong about quantum
entanglement
> then there is something happening that is completely outside of our
> existing understanding of physics.
QM is wrong about quantum entanglement. And if we assume that then
it is rather trival to show that no test of Bell's Inequality has any merit
since they tend to show non-local effects.
The universe is LOCAL and all physical effects occur at or below the
speed of light just like Einstein said.
-- David Elm davi...@tiac.net
http://www.tiac.net/users/davidelm/epr.htm
Jacques M. Mallah
John Baez (ba...@math.mit.edu) wrote:
: Kevin Brown <ksb...@seanet.com> wrote:
: >12:41 PM ET 06/26/97: Light particles communicate over six miles
: >LONDON (Reuter) - What Albert Einstein described as a spooky
: >somehow linked over a distance of six miles, New Scientist magazine
: >reported Thursday.
: I'm still waiting for Budnik to pay me off on that bet.
It's much harder to get the experimental conditions needed to rule
out local hidden variables than it is to just observe correlations, so I'm
sure he won't yet. However, the new data does at least place yet more
constraints on Budnick-style (BS) interpretations.
- - - - - - -
Jacques Mallah (jqm...@is2.nyu.edu)
Graduate Student / Many Worlder / Devil's Advocate
"I know what no one else knows" - 'Runaway Train', Soul Asylum
My URL: http://pages.nyu.edu/~jqm1584/
Michael J. Ramsey
On 30 Jun 1997 20:42:58 GMT, " -- David A. Elm"
<davi...@tiac.net> wrote:
[snip]
>
>QM is wrong about quantum entanglement. And if we assume that then
>it is rather trival to show that no test of Bell's Inequality has any merit
>since they tend to show non-local effects.
>
>The universe is LOCAL and all physical effects occur at or below the
>speed of light just like Einstein said.
>
David,
Is this an axiom in your system of reasoning, or do you know of
experiments which support your conclusion?
--Mike
-- David A. Elm
Mike,
It seems to me that all the experiments performed so far
support my conclusion. From the LHV view no results
measured in these tests actually violate Bell's inequality
because the plots are wrong when you scale them up to
the wrong point, due to your QM assumptions.
It is true that if you assume there is a connection between
the two particles then you can derive a chart that proves
there is a connection between these two particles.
This is all that the EPR tests prove, if you assume QM
then you can proove QM. I could say the same thing to
prove my bible is correct, It must be true, because it says so.
My reasoning says that a test is only valid if you follow good
logic also. This an axiom in my system of reasoning.
-- David Elm
Jim Carr
ksb...@seanet.com (Kevin Brown) writes:
>
>12:41 PM ET 06/26/97: Light particles communicate over six miles
>
>LONDON (Reuter) - What Albert Einstein described as a spooky
>quality of light just gets spookier, according to Swiss scientists.
>They have found evidence that light particles known as photons are
>somehow linked over a distance of six miles, New Scientist magazine
The on-line version at newscientist.com has this article.
They said it was 10 kilometers. ;-)
I have not looked to see what might be on the web at Geneva or if
there is a preprint at LANL (Question: do the net monitors freak
out about xxx.lanl.gov?). Any info on that, or scheduled publication
(I assume in Physics Letters) would be appreciated -- by more than
just me I would imagine.
--
James A. Carr <j...@scri.fsu.edu> | Commercial e-mail is _NOT_
http://www.scri.fsu.edu/~jac/ | desired to this or any address
Supercomputer Computations Res. Inst. | that resolves to my account
Florida State, Tallahassee FL 32306 | for any reason at any time.
Paul Budnik
Phil Gibbs (philip...@pobox.com) wrote:
[...]
: The article just says that the results are "yet to be published". I
: suppose we will just have to wait to see if they checked the
: timing well enough to satisfy PB.
It is certainly not a matter of satisfying me. It is a matter of
showing a direct macroscopic violation of Bell's inequality. Bell's
result is important because it went beyond all the strange and
unexpected predictions of QM to come up with a way to test for
an experimental violation of classical locality at the macroscopic
level. That you have no idea how to explain the existing experimental
results with a classical local model says as much about you as it
does about nature. Confirmed macroscopic violations of Bell's inequality
on the other hand say something quite definite about the nature of
reality as would the failure to demonstrate such violations when
QM predicts they should occur.
Bonehead
In article <33b492a7....@news.seanet.com>, ksb...@seanet.com
says...
> 12:41 PM ET 06/26/97: Light particles communicate over six miles
>
> LONDON (Reuter) - What Albert Einstein described as a spooky
> quality of light just gets spookier, according to Swiss scientists.
> They have found evidence that light particles known as photons are
> somehow linked over a distance of six miles, New Scientist magazine
> reported Thursday.
>
> generated pairs of photons and sent them along two separate optic
> fibers, ending up six miles apart. ...measurements of one photon
> affected the results of measurements of the other. "There was no
> way the two photons could have communicated over such a distance,
> meaning they must have been "entangled" Gisin said.
>
has nothing whatever to do with photon pairs *communicating* between each
other in the usual sense and a great deal to do with the nature of
*entanglement*. I have a hunch that the nature of that entanglement has
some profound implications on what we mean by separate or isolated
photons as existing as independent entities. It seems to always be the
case that in these EPR type situations we envision that photon pairs are
behaving as macro-entities, speeding off in isolation through space and
time (like the familiar billiard ball scenarios), and that they somehow
MUST be able to communicate the results of some measurement inflicted
upon them to their twin. We invoke our usual time asymmetrical cause-
effect predispositions upon these experiments and come away with such
dubious explanations as the *many worlds* and *many minds*
interpretations in order to deal with the paradoxical results. Photons
are obviously NOT independent isolated entities and their entanglement
can be explained without non-locality or hidden variables through a
detailed application of advanced action.
Bonehead
Matt McIrvin
Bonehead <dh...@cvn.net> wrote:
> In article <33b492a7....@news.seanet.com>, ksb...@seanet.com
> says...
>
> > It said Nicolas Gisin and colleagues at the University of Geneva
> > generated pairs of photons and sent them along two separate optic
> > fibers, ending up six miles apart. ...measurements of one photon
> > affected the results of measurements of the other. "There was no
> > way the two photons could have communicated over such a distance,
> > meaning they must have been "entangled" Gisin said.
> >
> I'm not surprised!
Neither am I. As I've said before about other such "shocking" results,
it's important to stress that the preservation of "entanglement" over
long distances is just what you'd expect from taking quantum mechanics
seriously at face value. It may not make these results any less *weird*,
but they are verifications of what is already thought by most physicists
to be true. Pop media reports often seem to imply that these results are
somehow at variance with known theories.
--
Font-o-Meter! Proportional Monospaced
^
http://world.std.com/~mmcirvin/
John Sidles
Matt McIrvin <mmci...@world.std.com> wrote:
>Bonehead <dh...@cvn.net> wrote:
>> > It said Nicolas Gisin and colleagues at the University of Geneva
>> > generated pairs of photons and sent them along two separate optic
>> > fibers, ending up six miles apart.
>long distances is just what you'd expect from taking quantum mechanics
>seriously at face value.
Anyone interested enough to read this thread ought to
consider reading Oxford physicist David Deutsch's new book
"The Fabric of Reality". McIrvin's phrase "taking quantum
mechanics seriously at face value" is precisely the theme of
Deutsch's book. And Deutsch's professional credentials are
impeccable.
Excerpts can be found on Deutsch's home page in the Oxford
Physics Department:
http://eve.physics.ox.ac.uk/Personal/deutsch/FabricOfReality/FoR.html
For the physics student, the downside of the book is ...
no equations! Because there are three styles of book
about quantum mechanics:
(1) No equations ... these are really books about philosophy.
They can be interesting, and even beautiful, but are of
little practical utility in establishing a physics career
or designing experiments that work.
(2) A minimal set of equations ... e.g. Gottfried or Messiah.
These books teach a parsimonious Copenhagen-style of
quantum calculation, in which details of the measurement
process are glossed over, and the role of symmetries and
conservation laws is emphasized. At the end, the student can
calculate QM, but usually has residual philosophical puzzlement
about how the classical world emerges from quantum reality.
(3) Books that begin with a fully quantum mechanical description
of the measurement process. As far as I know, there are no
such textbooks! Not that such textbooks couldn't be
written. In such a textbook, randomness would appear not as
an axiom describing Copenhagen-style measurement, but instead
would emerge naturally and inevitably, as algorithmic
randomness in the style of Chaitin (see reference below).
From this fully QM point of view, it is mathematically
*guaranteed* that typical quantum experiments will yield
random-appearing data records, for the readily understood yet
completely rigorous reason there are so many possible data
records that no concise algorithm can select any particular data
record from among the set of all possible data records. This is
the essence of Chaitin's definition of randomness.
Accession No.: 2018101.
Author: Chaitin-G-J.
Author Affil.: IBM Thomas J. Watson Res. Center, Yorktown Heights, NY,
USA.
Title: Godel's theorem and information.
Source: International Journal of Theoretical Physics. vol.21, no.12.
pp. 941-54. Dec. 1982.
Abstract: Godel's theorem may be demonstrated using arguments
having an information-theoretic flavor. In such an approach it
is possible to argue that if a theorem contains more
information than a given set of axioms, then it is impossible
for the theorem to be derived from the axioms. In contrast
with the traditional proof based on the paradox of the liar,
this new viewpoint suggests that the incompleteness phenomenon
discovered by Godel is natural and widespread rather
than pathological and unusual.
(I believe Chaitin wrote a Scientific American article that explains these
ideas in a more accessible form, but I don't have the reference handy).
Anyway, this is one area in physics where learning more mathematics
provides a very effective remedy for philosophical puzzlement.
Michael J. Ramsey
On 1 Jul 1997 14:11:42 GMT, " -- David A. Elm" <davi...@tiac.net>
wrote:
>
>
>Michael J. Ramsey <mjra...@bellatlantic.net> wrote:
>> On 30 Jun 1997 20:42:58 GMT, " -- David A. Elm"
>> <davi...@tiac.net> wrote:
>>
>> [snip]
>> >
>> >QM is wrong about quantum entanglement. And if we assume that then
>> >it is rather trival to show that no test of Bell's Inequality has any
>merit
>> >since they tend to show non-local effects.
>> >
>> >The universe is LOCAL and all physical effects occur at or below the
>> >speed of light just like Einstein said.
>> >
>>
>> David,
>> Is this an axiom in your system of reasoning, or do you know of
>> experiments which support your conclusion?
>>
>
>Mike,
>
>It seems to me that all the experiments performed so far
>support my conclusion. From the LHV view no results
>measured in these tests actually violate Bell's inequality
>because the plots are wrong when you scale them up to
>the wrong point, due to your QM assumptions.
>
I am sorry, is your concern over "counter inefficiencies, or analyzer
inefficiencies, or geometrical imperfections, and so on."?
"It is only with added assumptions, or conventional allowances for
inefficiencies and extrapolation from the real to the ideal, that one
can say the inequality is violated."
Why do you think that QM, which works so well for "inefficient
practical set-ups is [] going to fail badly when sufficient
refinements are made"? I am interested in your answer to this last.
The quotes are J.S. Bell writing on this topic in "Bertlmann's socks
and the nature of reality".
Abner Shimony also runs through the cons but concludes that the Aspect
experiments hold more water than not.
>
>My reasoning says that a test is only valid if you follow good
>logic also. This an axiom in my system of reasoning.
You are *assuming* that the universe is local and hold to that
position because you are unconvinced by the latest Bell experiments.
Locality is, therefore, an axiom in your reasoning system, isn't it?
I don't mean to give offensive. Most physicist view the world through
field theory colored glasses, and locality is part and parcel of that
world view.
--Best regards,
--Mike
Paul Budnik
Jim Carr (j...@ibms48.scri.fsu.edu) wrote:
: ksb...@seanet.com (Kevin Brown) writes:
: >
: >12:41 PM ET 06/26/97: Light particles communicate over six miles
: >
: >LONDON (Reuter) - What Albert Einstein described as a spooky
: >quality of light just gets spookier, according to Swiss scientists.
: >They have found evidence that light particles known as photons are
: >somehow linked over a distance of six miles, New Scientist magazine
Here are two choice quotes from this brief report.
Nicolas Gisin and his colleagues from the University of
Geneva have shown that the measurement of one particle will
instantaneously determine the state of the other.
[...]
but the theory holds that this collapse will immediately cause a
similar collapse in the other particle, no matter how far away it is
None of this reflects on Gisin's research. It does reflect
on the reliability and care of the New Scientist.
Paul Budnik
Bonehead (dh...@cvn.net) wrote:
: In article <33b492a7....@news.seanet.com>, ksb...@seanet.com
: says...
[...]
: I'm not surprised! I have suspected all along that what is happening here
: has nothing whatever to do with photon pairs *communicating* between each
: other in the usual sense and a great deal to do with the nature of
: *entanglement*. I have a hunch that the nature of that entanglement has
: some profound implications on what we mean by separate or isolated
: photons as existing as independent entities.
So did Bohm nearly 50 years ago. If you are not aware of his notion of
`wholeness and the implicate order' you might want to take a look at
his work.
Everyone should keep in mind that a report like this proves nothing.
As I recall my wager with Baez has a minimum two year wait between
the publication of a definitive experiment and payment of the wager. This
is to allow enough time for the report to be thoroughly reviewed. At
this stage this report has yet to be published. The New Scientist blurb
is written by someone who appears to understand next to nothing
about this aspect of physics.
Dan M.
-- David A. Elm wrote:
>
> Michael J. Ramsey <mjra...@bellatlantic.net> wrote:
> > On 30 Jun 1997 20:42:58 GMT, " -- David A. Elm"
> > <davi...@tiac.net> wrote:
> >
> > [snip]
> > >
> > >QM is wrong about quantum entanglement. And if we assume that then
> > >it is rather trival to show that no test of Bell's Inequality has any
> merit
> > >since they tend to show non-local effects.
> > >
> > >The universe is LOCAL and all physical effects occur at or below the
> > >speed of light just like Einstein said.
> > >
> >
> > David,
> > Is this an axiom in your system of reasoning, or do you know of
> > experiments which support your conclusion?
> >
>
> Mike,
>
> It seems to me that all the experiments performed so far
> support my conclusion. From the LHV view no results
> measured in these tests actually violate Bell's inequality
> because the plots are wrong when you scale them up to
> the wrong point, due to your QM assumptions.
>
> the two particles then you can derive a chart that proves
> there is a connection between these two particles.
> This is all that the EPR tests prove, if you assume QM
> then you can proove QM. I could say the same thing to
> prove my bible is correct, It must be true, because it says so.
>
> logic also. This an axiom in my system of reasoning.
>
David,
Unfortunately, you tacitly include as part of your logic the assumption
that you wish to prove. The analogy you give with a card game in your
arguments on correlations does not match the Bell correlation
experiment. I tried to give a card game analogy that did match and show
that the same problem arises.
There are no assumptions of quantum mechanics that are required for the
experiments that are considered. It is true, that predictions made by
QM are matched by the experiments, but that is a plus for QM, not a
minus for the experiments. One can frame the question totally in terms
of experimental data, counting finite numbers, the assumption of a local
realistic theory, and the validity of Aristotelian logic, and get the
inconsistency between experiment and prediction that we know and love so
well.
Dan M.
Ray Tomes
"Dan M." <sh...@wt.net> in article <33BFE0...@wt.net> wrote:
[to David Elm]
>Unfortunately, you tacitly include as part of your logic the assumption
>that you wish to prove.
Funnily enough Aspect assumes that QM is true in doing his experiments.
>The analogy you give with a card game in your
>arguments on correlations does not match the Bell correlation
>experiment. ...
Please be more specific. The shape of the curve matches exactly.
the base level may not, but that is because the EPR experimental results
as reported are after removing the base level measured. The experiments
are specifically set up so as to get the expected result.
>There are no assumptions of quantum mechanics that are required for the
>experiments that are considered. ...
Have a look at http://www.aber.ac.uk/~cat for lots of evidence to the
contrary.
>One can frame the question totally in terms
>of experimental data, counting finite numbers, the assumption of a local
>realistic theory, and the validity of Aristotelian logic, and get the
>inconsistency between experiment and prediction that we know and love so
>well.
I would like to see that demonstrated. Caroline Thompson (at the WWW
site above) has shown how to get the exact Aspect results from a local
theory by computer simulation. If you can prove that is impossible then
you have made some wrong assumptions.
Jim Carr
rto...@kcbbs.gen.nz (Ray Tomes) writes:
>
>The experiments
>are specifically set up so as to get the expected result.
I thought they were specifically set up to disprove the expected
result. What would get a bigger headline and a better shot at a
Nobel Prize:
"Bell correlations at 6 miles"
or
"Quantum mechanics proved wrong"
??
Ray Tomes
In article <5puaic$80h$1...@news.fsu.edu>, j...@ibms48.scri.fsu.edu (Jim
Carr) wrote:
>rto...@kcbbs.gen.nz (Ray Tomes) writes:
>>The experiments are specifically set up so as to get the expected result.
> I thought they were specifically set up to disprove the expected
> result. What would get a bigger headline and a better shot at a
> Nobel Prize:
> "Bell correlations at 6 miles"
> or
> "Quantum mechanics proved wrong"
The question is, would the latter get a headline at all?
Caroline Thompson has reproduced all of Aspects results from a simple
local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
According to QM theorists this is not possible. They are wrong.
She describes the process of setting up the equipment used by Aspect.
At each step QM is assumed to be right and the apparatus is adjusted so
that the QM expected result is achieved.
-- Ray Tomes -- rto...@kcbbs.gen.nz -- Harmonics Theory --
http://www.kcbbs.gen.nz/users/rtomes/rt-home.htm
subscribe to cycles-...@esosoft.com interdisciplinary cycles list
ale2
In article <34622d76....@aklobs.org.nz>
rto...@kcbbs.gen.nz (Ray Tomes) writes:
> Caroline Thompson has reproduced all of Aspects results from a simple
> local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
Anyone care to post the above here so i don't have to go up on campus
%^)
Mountain Man
al...@psu.edu wrote:.
>rto...@kcbbs.gen.nz (Ray Tomes) writes:
>> Caroline Thompson has reproduced all of Aspects results from a simple
>> local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
>Anyone care to post the above here so i don't have to go up on campus
Most of it appears to be in postscript format, so perhaps
today is the day that the hiking boots get donned? :)
All the best,
Pete Brown
--------------------------------------------------------------------
BoomerangOutPost: Mountain Man Graphics, Newport Beach, {OZ}
Thematic Threading: Publications of Peace and Of Great Souls
Webulous Coordinates: http://magna.com.au/~prfbrown/welcome.html
.
QuoteForTheDay:
.
"Life evolves out of Matter, Mind out of Life,
because they are already involved there:
Matter is a form of veiled Life,
Life a form of veiled Mind.
.
May not Mind be a form and veil of a higher power,
the Spirit, which would be supramental in its nature?
Man's highest aspiration would then only indicate
the gradual unveiling of the Spirit within,
the preparation of a higher life upon earth.
.
We speak of the evolution of Life in Matter,
the evolution of Mind in Matter;
but evolution is a word which merely states
the phenomenon without explaining it.
.
For there seems to be no reason why Life
should evolve out of material elements
or Mind out of living form,
unless we accept the Vedantic solution
that Life is already involved in Matter and Mind in Life
because in essence Matter is a form of veiled Life,
Life a form of veiled Consciousness."
.
- Sri Aurobindo, "The Future Evolution of Man"
http://magna.com.au/~prfbrown/auro.html
--------------------------------------------------------------------
Jim Carr
j...@ibms48.scri.fsu.edu (Jim Carr) wrote:
|
| rto...@kcbbs.gen.nz (Ray Tomes) writes:
| >The experiments are specifically set up so as to get the expected result.
|
| I thought they were specifically set up to disprove the expected
| result. What would get a bigger headline and a better shot at a
| Nobel Prize:
| "Bell correlations at 6 miles"
| or
| "Quantum mechanics proved wrong"
rto...@kcbbs.gen.nz (Ray Tomes) writes:
>
>The question is, would the latter get a headline at all?
The former did not even appear in our paper, unlike the news story
on the claimed asymmetry of space that questions relativty, so I
think that we would see the latter.
>Caroline Thompson has reproduced all of Aspects results from a simple
>local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
>According to QM theorists this is not possible. They are wrong.
That does not prove the prediction of QM wrong, it only says that
T ==> A as well as Q ==> A. Only (not)A would disprove Q, and
that would be big news. I am sure Aspect wished that is what he
had found, since he would be in Stockholm by now.
Since she has not disproved Bell's theorem, exactly which part she uses
would be interesting to figure out. I have not grabbed all of that
yet, just the HTML part, and the description of the mechanistic model
was not evident in it. Should I look in a particular one of those
papers for the algorithm she carried out?
Ray Tomes
In article <5q3pmg$gsj$1...@news.fsu.edu>, j...@ibms48.scri.fsu.edu (Jim
Carr) wrote:
> Since she has not disproved Bell's theorem, exactly which part she uses
> would be interesting to figure out. I have not grabbed all of that
> yet, just the HTML part, and the description of the mechanistic model
> was not evident in it. Should I look in a particular one of those
> papers for the algorithm she carried out?
In postscript at http://www.aber.ac.uk/~cat/Chaotic/chaotic.ps
also published as Found. Phys. Lett. 9, 357 (1996); also available at
http://xxx.lanl.gov/archive/quant-ph, ref. no. 9611037.
Kevin Brown
On Thu, 10 Jul 1997 rto...@kcbbs.gen.nz (Ray Tomes) wrote:
>Caroline Thompson has reproduced all of Aspects results from a simple
>local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
>According to QM theorists this is not possible. They are wrong.
>At each step QM is assumed to be right and the apparatus is adjusted
>so that the QM expected result is achieved.
The model described in Thompson's article is based on the standard
"realistic" model in which each particle is considered to possess a
definite spin (or polarization) axis, and measurement along a given
axis will return a positive or negative result corresponding to the
SIGN of cos(theta) where theta is the difference between the spin axis
and the measurement axis. Assuming two "coupled" spin-1/2 particles
in an EPR-type experiment have exactly opposite spin directions,
this model agrees with the QM predictions when the difference "alpha"
between the two measurement angles is 0, pi/2, and pi (i.e. the
measurement angles are equal, perpindicular, or opposite), where
the correlations are 0, 1/2, and 1, respectively. Of course, for
intermediate values of alpha this basic model predicts that the
correlation varies linearly, thereby differing from the predictions
of QM as shown below
1 |-----------------*-*/-
| * /
| * /
correlation | * /
| * /
1/2|----------/----------
| / * ----- basic realistic
| / * model
| / *
| / * ***** QM predictions
0 |/*_*_________________
0 pi/2 pi
alpha
Thompson's proposal is to modify this basic model so that it gives
results approximating the predictions of QM for all values of alpha.
She does this by supposing that a measurement of a particle's spin
has THREE possible outcomes: positive, negative, or null. The "null"
outcome represents cases in which the measurement fails to resolve the
spin of the particle. Further, she supposes that a measurement is
effective only if the positive or negative spin axis is within an arc
subtended by a fixed angle "beta" centered on the measurement axis.
Note that the basic model assumed beta equals pi, from which it
follows that a measurement is ALWAYS effective. Thompson's argument
is that if beta is some fixed value LESS than pi, and if we neglect
all particle-pairs in an EPR-type experiment when either of the spin
measurements return "null", then the correlations of the remaining
pairs can approximate the correlations predicted by QM.
For example, suppose beta equals just pi/2. This implies that each
individual spin measurement is effective (i.e., returns a non-null
result) over only half the range of possible spin axes. Now consider
an EPR-type experiment in which we measure the spins of two coupled
particles, with measurement axes that differ by alpha. Clearly if
alpha equals pi/2 the effective ranges of the two measurements are
mutually exclusive, so we will be unable to gather any correlation
data at all.
|
M1 = null | M1 = positive
M2 = positive | M1 axis M2 = null
* | * /
* | */
* | / *
* |/ *
--------------------------------
* |\ *
* | \ *
* | *\
* | * \
M1 = negative | M2 axis M1 = null
M2 = null | M2 = negative
However, for any other value of alpha the "effective quadrants" for
measurements 1 and 2 will overlap. Also, notice that if alpha is less
than pi/2 the overlap will consist entirely of disagreements, whereas
if alpha is GREATER than pi/2 the overlap will consist entirely of
agreements. Thus, assuming beta=pi/2, the correlation is 0.0 for
alpha less than pi/2, and 1.0 for alpha greater than pi/2. The pair-
measurement effectiveness is 0.0 at alpha=pi/2 and ramps up to 0.5
at alpha=0 and alpha=pi.
On the other hand, if we expand our assumed effectiveness arcs to
beta = 3pi/4 we will record only disagreements when alpha is less
than pi/4, and only agreements when alpha is greater than 3pi/4, and
the correlation will vary linearly in between (because now we can get
some overlap "on both sides" of the spin). Plots of correlation the
and pair-measurement effectiveness for beta = pi, 3pi/4, and pi/2 are
shown below.
____3pi/4___ 1 _______________________pi
1 | | / / |
| | / / |
| pi/2| / /pi | 3pi/4
corr | | / / eff |\ /
| | / / | \ /
1/2| |// 1/2| \ ___ / pi/2
| // |\ /
| / /| | \ /
| / / | | \ /
| / / | | \ /
| / / | | \ /
0 |/_____/____| 0 | \ /
-------------------------- --------------------------
0 pi/2 pi 0 pi/2 pi
alpha alpha
The correlations based on an assumption of beta = 3pi/4 clearly
resemble the cosine curve predicted by QM, and if we refine our
"effectiveness" model so that the chances of a "null" result vary
smoothly with theta we can certainly tailor the correlation curve
to agree exactly with the QM predictions.
Of course, all of this is quite familiar to anyone who has ever
thought about tests of Bell's inequalities. It's understood that if
our measurement efficiencies are low AND there is a systematic effect
such that the recorded pairs are a biased sample, then the simple Bell
inequalities are no longer relevant. Furthermore, it's obvious that
we can imagine realistic mechanisms that yield biases such that the
simple QM correlations are reproduced within the biased sample. The
question is (and has always been) whether the hypothesis of such a
bias is consistent with our observations.
Thompson notes that if our measurements were affected by the type of
bias described above we would expect to see the measurement efficiency
reach a minimum at alpha = pi/2. She seems to acknowledge that no
such minimum appears in the data. In particular, she quotes Aspect
as saying that although he found slight variations in his measurement
efficiencies [versus alpha], the variations were not significant.
Thompson doesn't comment on the more general question of whether the
basic premise of "unresolved" spin has any foundation, i.e., when we
pass electrons through a Stern-Gerlach apparatus, do we observe any
electrons that "can't decide" whether to be deflected up or down? My
understanding is that we do not.
To address the absence of a minimum in the efficiency at alpha=pi/2
Thompson points out that if we assume the value of beta is pi for one
of the measurements but something less than pi for the other, then the
measurement efficiency is constant. This is true enough, but it shows
that in order to match a particular set of QM result she not only
needs to invoke a systematic bias in the measurement efficiencies, she
needs an *asymmetric* systematic bias, i.e., a bias that works on only
one of the two particles in each coupled pair. Thus, we are asked to
imagine that when two particles emerge from a singlet state in
opposite directions and interact with spin-measuring devices, one of
them exhibits a definite spin when measured at any angle, but the
other sometimes doesn't exhibit spin, if the measurement angle happens
to be nearly perpindicular to its spin axis.
It seems to me that the fundamental problem with this approach is the
implication that a spin-1/2 particle (for example) will sometimes fail
to exhibit spin, or will exhibit only "partial spin" (below the
threshold of the discriminator). This seems to stand in direct
contradiction to experimental results of the Stern-Gerlach type, and
to all the other well-supported consequences of the fact that quantum
spin is a quantized characteristic of particles. Furthermore, one
would think that simplistic bias models of the kind described by
Thompson would be relatively easy to rule out by altering the design
of the experiments. Indeed, Thompson's paper includes the following
paragraph
"Clauser and Horne state on more general theoretical grounds that
[the simple form of the Bell inequality] should not be used unless
the number of pairs emitted is known. They and others devised the
third main group of ``Bell's'' inequalities (Clauser, Horne,
Shimony, Holt (CHSH) tests), members of which are not so easy to
violate as they are unaffected by the type of missing value we have
been considering... They are used in some of the most important EPR
experiments, notably the Aspect, Dalibard and Roger paired-photon
experiment of 1982 [2], and will be covered in a separate paper."
This is rather disapointing. It seems to say that the model we've
been asked to consider is known to be incompatible with the results
of improved experiments. I'm not sure which "separate paper" Thompson
is referring to here, in which she "covers" those experiments, but I
suspect it's her paper on *timing* loopholes, i.e., an entirely
different principle of explanation.
So, what becomes of the "chaotic ball" model? In addition to being
explicitly ad hoc relative to the original Aspect experiments, it's
basic premise appears to conflict with a large body of experimental
results on the quantization of spin. Furthermore, the model does not
account for the results of subsequent improved tests by Aspect, et
al., which, not surprisingly, again showed good agreement with the
predictions of QM.
____________________________________________________________
| /*\ http://www.seanet.com/~ksbrown/ |
| MathPages / \ |
|_____________/"And of some have compassion..."______________|
ksb...@seanet.com
alpha
-------------------==== Posted via Deja News ====-----------------------
Jonathan Scott
(Reposted in an attempt to get the square brackets to display
correctly this time; I was using a US terminal program on a UK IBM
mainframe and the EBCDIC/ASCII translation got messed up).
In article <33c7963f...@news.seanet.com>,
on Sat, 12 Jul 1997 20:32:21 GMT,
ksb...@seanet.com (Kevin Brown) writes:
>The model described in Thompson's article is based on the standard
>"realistic" model in which each particle is considered to possess a
>definite spin (or polarization) axis, and measurement along a given
>axis will return a positive or negative result corresponding to the
>SIGN of cos(theta) where theta is the difference between the spin axis
>and the measurement axis. ...
Thanks for summarizing this. Here's a C program which I've previously
posted a couple of times over the last few years which demonstrates
a similar realistic model that relies on non-detection, but in contrast
to Caroline Thompson's model it DOESN'T give a dependence of the
average detection rate on the relative detector angle. Unfortunately,
it gives a correlation which is too high, being even stronger than the
QM result, but it demonstrates that a realistic model is possible.
/* Program to illustrate Aspect experiment non-detection loophole */
/* Author: Jonathan Scott Version: 1.0 Date: 1990-08-26 */
/* Method involves failing to detect some photons plus random shift of
polarization which avoids detection rate dependency on relative angle.
This simple model actually gives a correlation which is TOO HIGH,
0.75 instead of 0.707 at 22.5 degrees separation, but it clearly
illustrates "deterministic" violation of the Bell inequalities. */
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#include <math.h>
#define pi 3.14159265358979323846
#define rand0to1 ((double) rand()/RAND_MAX)
main(int argc, char * argv[])
{
long int count[4] = {0,0,0,0}; /* Coincidence counts -- -+ +- ++ */
double obs1deg = 0; /* Detector 1 angle in degrees */
double obs2deg = 22.5; /* Detector 2 angle in degrees */
double obs1, obs2; /* Same orientations in radians */
double theta, phi; /* Angle and shift for each pair */
double theta1, theta2; /* Relative angle at each detector */
long int runsize = 4000; /* Number of pairs to simulate */
long int emitted; /* Number of pairs emitted so far */
long int detected = 0; /* Number of pairs detected so far */
double a1, a2; /* Amplitudes at each detector */
double corr, rate; /* Correlation and detection rate */
/* Process overrides for default parameters */
if (argc > 1) obs1deg = atof(argv[1]); /* Arg 1 = detector 1 angle */
if (argc > 2) obs2deg = atof(argv[2]); /* Arg 2 = detector 2 angle */
if (argc > 3) runsize = atol(argv[3]); /* Arg 3 = number of pairs */
srand((unsigned)time(NULL)); /* Start up random number seed */
obs1 = obs1deg * pi/180; /* Convert angles to radians */
obs2 = obs2deg * pi/180;
for (emitted = 0; emitted < runsize; emitted++) /* Main run loop */
{
theta = rand0to1 * pi; /* Generate random photon angle */
phi = (rand0to1 - 0.5) * pi / 4; /* Generate random shift angle */
theta1 = theta + phi; /* Calculate angle at observer 1 */
theta2 = theta - phi; /* Calculate angle at observer 2 */
a1 = cos(2 * (obs1 - theta1)); /* Get amplitude projections */
a2 = cos(2 * (obs2 - theta2)); /* as signed quantities */
if (a1*a1>0.5 && a2*a2>0.5) /* If both probabilities > 0.5 */
{
detected++; /* then register a coincidence */
count[(a1>0) + 2 * (a2>0)]++; /* Increment appropriate count */
}
}
corr = (double)(count[0] + count[3]
- count[1] - count[2]) / detected; /* Correlation */
rate = (double) detected / emitted; /* Detection rate */
printf("Observation angles: %g, %g\n",obs1deg,obs2deg);
printf("Pairs emitted: %ld Pairs detected: %ld\n",emitted,detected);
printf(" -- -+ +- ++\n");
printf("%6ld %6ld %6ld %6ld\n",count[0],count[1],count[2],count[3]);
printf("Correlation %g Detection rate %g\n", corr, rate);
}
/* (end of program) */
Jonathan Scott
jonatha...@vnet.ibm.com or jsc...@winvmc.vnet.ibm.com
(E-mail address in header is not valid from the internet).
Jonathan Scott
jonatha...@vnet.ibm.com or jsc...@winvmc.vnet.ibm.com
(E-mail address in header is not valid from the internet).
Jonathan Scott
In article <33c7963f...@news.seanet.com>,
on Sat, 12 Jul 1997 20:32:21 GMT,
ksb...@seanet.com (Kevin Brown) writes:
>"realistic" model in which each particle is considered to possess a
>definite spin (or polarization) axis, and measurement along a given
>axis will return a positive or negative result corresponding to the
>SIGN of cos(theta) where theta is the difference between the spin axis
Thanks for summarizing this. Here's a C program which I've previously
posted a couple of times over the last few years which demonstrates
a similar realistic model that relies on non-detection, but in contrast
to Caroline Thompson's model it DOESN'T give a dependence of the
average detection rate on the relative detector angle. Unfortunately,
it gives a correlation which is too high, being even stronger than the
QM result, but it demonstrates that a realistic model is possible.
/* Program to illustrate Aspect experiment non-detection loophole */
/* Author: Jonathan Scott Version: 1.0 Date: 1990-08-26 */
/* Method involves failing to detect some photons plus random shift of
polarization which avoids detection rate dependency on relative angle.
This simple model actually gives a correlation which is TOO HIGH,
0.75 instead of 0.707 at 22.5 degrees separation, but it clearly
illustrates "deterministic" violation of the Bell inequalities. */
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#include <math.h>
#define pi 3.14159265358979323846
#define rand0to1 ((double) rand()/RAND_MAX)
main(int argc, char * argvݨ)
{
long int countİ4¨ = {0,0,0,0}; /* Coincidence counts -- -+ +- ++ */
double obs1deg = 0; /* Detector 1 angle in degrees */
double obs2deg = 22.5; /* Detector 2 angle in degrees */
double obs1, obs2; /* Same orientations in radians */
double theta, phi; /* Angle and shift for each pair */
double theta1, theta2; /* Relative angle at each detector */
long int runsize = 4000; /* Number of pairs to simulate */
long int emitted; /* Number of pairs emitted so far */
long int detected = 0; /* Number of pairs detected so far */
double a1, a2; /* Amplitudes at each detector */
double corr, rate; /* Correlation and detection rate */
/* Process overrides for default parameters */
if (argc > 1) obs1deg = atof(argvİ1¨); /* Arg 1 = detector 1 angle */
if (argc > 2) obs2deg = atof(argvİ2¨); /* Arg 2 = detector 2 angle */
if (argc > 3) runsize = atol(argvİ3¨); /* Arg 3 = number of pairs */
srand((unsigned)time(NULL)); /* Start up random number seed */
obs1 = obs1deg * pi/180; /* Convert angles to radians */
obs2 = obs2deg * pi/180;
for (emitted = 0; emitted < runsize; emitted++) /* Main run loop */
{
theta = rand0to1 * pi; /* Generate random photon angle */
phi = (rand0to1 - 0.5) * pi / 4; /* Generate random shift angle */
theta1 = theta + phi; /* Calculate angle at observer 1 */
theta2 = theta - phi; /* Calculate angle at observer 2 */
a1 = cos(2 * (obs1 - theta1)); /* Get amplitude projections */
a2 = cos(2 * (obs2 - theta2)); /* as signed quantities */
if (a1*a1>0.5 && a2*a2>0.5) /* If both probabilities > 0.5 */
{
detected++; /* then register a coincidence */
countİ(a1>0) + 2 * (a2>0)¨++; /* Increment appropriate count */
}
}
corr = (double)(countİ0¨ + countİ3¨
- countİ1¨ - countİ2¨) / detected; /* Correlation */
rate = (double) detected / emitted; /* Detection rate */
printf("Observation angles: %g, %g\n",obs1deg,obs2deg);
printf("Pairs emitted: %ld Pairs detected: %ld\n",emitted,detected);
printf(" -- -+ +- ++\n");
printf("%6ld %6ld %6ld %6ld\n",countİ0¨,countİ1¨,countİ2¨,countİ3¨);
for Caroline Thompson
I am posting this for Caroline Thompson <xk...@dial.pipex.com> and have
not proof read it. The non-quoted text below was all written by her.
She was impressed with the ASCII diagrams. I have had to trim some
quoted text because my system will not allow more quoted text than
reply.
> In sci.physics, ksb...@seanet.com wrote:
> On Thu, 10 Jul 1997 rto...@kcbbs.gen.nz (Ray Tomes) wrote:
> >Caroline Thompson has reproduced all of Aspects results from a simple
> >local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
> >According to QM theorists this is not possible. They are wrong.
...
> 1 |-----------------*-*/-
> | * /
> | * /
> correlation | * /
> | * /
> 1/2|----------/----------
> | / * ----- basic realistic
> | / * model
> | / *
> | / * ***** QM predictions
> 0 |/*_*_________________
> 0 pi/2 pi
>
> alpha
>
> Thompson's proposal is to modify this basic model so that it gives
> results approximating the predictions of QM for all values of alpha.
> She does this by supposing that a measurement of a particle's spin
> has THREE possible outcomes: positive, negative, or null. The "null"
> outcome represents cases in which the measurement fails to resolve the
> spin of the particle. Further, she supposes that a measurement is
> effective only if the positive or negative spin axis is within an arc
> subtended by a fixed angle "beta" centered on the measurement axis.
...
Tell me more! Has any actual Stern-Gerlach experiment ever been done
with ELECTRONS?
> To address the absence of a minimum in the efficiency at alpha=pi/2
> Thompson points out that if we assume the value of beta is pi for one
> of the measurements but something less than pi for the other, then the
> measurement efficiency is constant. This is true enough, but it shows
> that in order to match a particular set of QM result she not only
> needs to invoke a systematic bias in the measurement efficiencies, she
> needs an *asymmetric* systematic bias, i.e., a bias that works on only
> one of the two particles in each coupled pair. Thus, we are asked to
> imagine that when two particles emerge from a singlet state in
> opposite directions and interact with spin-measuring devices, one of
> them exhibits a definite spin when measured at any angle, but the
> other sometimes doesn't exhibit spin, if the measurement angle happens
> to be nearly perpindicular to its spin axis.
No. When it comes to asymmetry, what I'm really interested in is
Aspect's actual EPR experiments. Here there is no shortage.
> It seems to me that the fundamental problem with this approach is the
> implication that a spin-1/2 particle (for example) will sometimes fail
> to exhibit spin, or will exhibit only "partial spin" (below the
> threshold of the discriminator). This seems to stand in direct
> contradiction to experimental results of the Stern-Gerlach type, and
I should be interested to know how anyone can know whether or not
anything is missed! But see later. Stern-Gerlach experiments are
irrelevant, as no EPR Stern-Gerlach experiment has ever been done. This
is made abundantly clear in my paper. The only experiments that are
anything like valid are optical ones.
> to all the other well-supported consequences of the fact that quantum
> spin is a quantized characteristic of particles. Furthermore, one
> would think that simplistic bias models of the kind described by
> Thompson would be relatively easy to rule out by altering the design
> of the experiments. Indeed, Thompson's paper includes the following
> paragraph
>
> "Clauser and Horne state on more general theoretical grounds that
> [the simple form of the Bell inequality] should not be used unless
> the number of pairs emitted is known. They and others devised the
> third main group of ``Bell's'' inequalities (Clauser, Horne,
> Shimony, Holt (CHSH) tests), members of which are not so easy to
> violate as they are unaffected by the type of missing value we have
> been considering... They are used in some of the most important EPR
> experiments, notably the Aspect, Dalibard and Roger paired-photon
> experiment of 1982 [2], and will be covered in a separate paper."
>
> This is rather disapointing. It seems to say that the model we've
> been asked to consider is known to be incompatible with the results
> of improved experiments.
The experiments using different Bell test are not necessarily
"improved", as they involve comparisons between counts with polarisers
present and absent. The LOGIC is otherwise sound, though.
And Yes, one might call it disappointing. That's life. Yet despite
the known possibility of null results causing bias, ONE OF ASPECT'S
EXPERIMENTS DOES USE THIS TEST! This is his second one. In fact, the
test is used, or something equivalent (measures of "visibility"), in all
EPR experiments in which both + and - outcomes are used. This (as I say
in my Chaotic Ball paper) includes some recent ones by Rarity and
Tapster. It also includes some being done in Innsbruck, at Zeilinger's
lab.
> I'm not sure which "separate paper" Thompson
> is referring to here, in which she "covers" those experiments, but I
> suspect it's her paper on *timing* loopholes, i.e., an entirely
> different principle of explanation.
Does anyone know what the structure of Gisin's experiment was? Which
Bell test? I've had a little correspondence with him and sent my papers
but he hasn't told me yet.
At least some real experiments, as I've found out only this year, are
likely to be explained by something that isn't even directly to do with
the EPR logic, but just something that is done to the data. Namely,
"accidentals" are subtracted. This is not justifiable unless each
photon
pair is emitted entirely independently of every other. That they should
be independent is a purely QUANTUM THEORY ASSUMPTION.
The case I have studied is Aspect's, which uses a calcium "atomic
cascade". Two lasers illuminate a beam of ionized calcium atoms, with
the polarisation of the lasers parallel to the direction of the beam.
The interaction region is pretty small - only fractions of a millimeter
in each direction. It is possible, is it not, that the lasers interact
with the WHOLE EM field in which the ions are embedded, setting up
resonances from time to time? These osillations of the whole field
propagate out in the form of visible light, with main components at the
two frequencies that we are filtering out and calling our two photons.
If this is the case, then there can be no "accidentals" as only one
resonance can occur at a time. Because the "coincidence window" is
finite, there might be a possibility of a second resonance within it,
but my intuition says there is likely to be some minimum time interval
between events.
I should be very interested to hear from anyone who has practical
experience in this area. Or knows about plasma physics? My knowledge
comes from Iain Main's "Vibrations and Waves in Physics" (CUP, 1978).
Here are some figures from my latest paper - "Against the Quantisation
of Light" or (soon to be added) "The EPR Magic Show", various editions
of which are creeping into my Web pages. Sorry if you're not viewing in
HTML - a completely plain version of the raw data is available from said
Web pages.
[Note by Ray Tomes: I have replaced the html list by figures from
Caroline's paper as the HTML stuff was unreadable]
Run a b Code Non Ret Ret Diff
1 0.0 0.0 A 92 21 71
2 999.9 999.9 G 245 89 156
3 0.0 999.9 F 127 46 81
4 0.0 0.0 A 98 23 75
5 22.5 0.0 B 87 23 64
6 45.0 0.0 C 62 23 39
7 67.5 0.0 D 38 23 15
8 90.0 0.0 E 28 22 6
9 112.5 0.0 D 38 22 16
10 135.0 0.0 C 63 23 40
11 157.5 0.0 B 86 22 64
12 180.0 0.0 A 98 24 74
13 180.0 0.0 A 95 22 73
14 202.5 0.0 B 87 23 64
15 225.0 0.0 C 64 23 41
16 247.5 0.0 D 38 23 15
17 270.0 0.0 E 29 23 6
18 292.5 0.0 D 39 22 17
19 315.0 0.0 C 62 23 39
20 337.5 0.0 B 87 23 64
21 360.0 0.0 A 94 23 71
22 999.9 0.0 F 125 45 80
23 999.9 999.9 G 248 91 157
24 999.9 999.9 G 250 90 160
25 999.9 999.9 G 248 89 159
26 999.9 999.9 G 250 91 159
27 0.0 0.0 A 97 23 74
28 0.0 0.0 A 97 22 75
29 0.0 0.0 A 96 22 74
30 90.0 0.0 E 28 26 2
31 180.0 0.0 A 97 23 74
32 270.0 0.0 E 27 22 5
As you will see, the adjustment subtracting "accidentals" is large.
The raw data follow a nice sine curve; the adjustment shifts it down,
which increases the visibility from <B>0.55</B> to <B>0.88</B>. If you
calculate the Bell test you will find that the raw data does
<EM>not</EM>
infringe it. The visibility is in fact only slightly greater than the
prediction (0.5) of the simplest realist model - the model that assumes
the counts obey Malus'Law <EM>exactly</EM> and that you get the expected
coincidence probabilities by multiplying the two singles probabilities,
for fixed polarisation angle. The basic QT prediction for visibility is
1.0, but it is not difficult to find excuses to reduce this to the
figure
observed!<P>
> So, what becomes of the "chaotic ball" model? In addition to being
> explicitly ad hoc relative to the original Aspect experiments, it's
> basic premise appears to conflict with a large body of experimental
> results on the quantization of spin.
See above. What results? But never mind, because quantized particle
spin doesn't come into real EPR experiments. Polarisation in the real
world can be treated as a real vector. I'm not so sure how many
dimensions we need for Rarity's "phase and momentum", but this is
certainly a continuous variable also.
> Furthermore, the model does not
> account for the results of subsequent improved tests by Aspect, et
> al., which, not surprisingly, again showed good agreement with the
> predictions of QM.
The reason Aspect (and all other EPR experimenters) is able to get good
agreement with QT is that the realist prediction is not so very
different
and that there is a great deal of freedom in the choice of parameters.
If all else fails, one can find excuses for adjusting the QT prediction.
Is it not fair that the realist prediction should also be adjusted?
The
real test between the theories ought to be not a Bell test (which covers
only an isolated point in parameter space) but how well the model
predicts results when you vary the parameters. QT has no chance - it
DOES NOT HAVE THE RIGHT PARAMETERS! The detection process is sensitive,
as the experimenters well know, to temperature, stray EM effects, choice
of detector and discriminator voltage. These are some of the true
parameters, whose effect can be modelled in terms of an effective EM
noise distribution that is added to the signal. Signals are detected
when signal + noise exceeds some threshold.
See "A non-factorable model ..." as well as new work mentioned above.
Sorry this has got so long.
Yours, Caroline
-- Ray Tomes - rto...@kcbbs.gen.nz - Cycles & Harmonics Theory --
http://www.kcbbs.gen.nz/users/rtomes/rt-home.htm
* subscribe to cycles-...@esosoft.com interdisciplinary cycles list *
Alt-Physics-FAQ - http://www.kcbbs.gen.nz/users/af/alt-faq.htm
Mountain Man
Ray for Caroline Thompson wrote:
[snip detailed response - refer earlier article]
> Is it not fair that the realist prediction should also be adjusted?
> The real test between the theories ought to be not a Bell test
> (which covers only an isolated point in parameter space) but how
> well the model predicts results when you vary the parameters.
> QT has no chance - it DOES NOT HAVE THE RIGHT PARAMETERS!
> The detection process is sensitive, as the experimenters well know,
> to temperature, stray EM effects, choice of detector and
> discriminator voltage. These are some of the true
> parameters, whose effect can be modelled in terms of an effective EM
> noise distribution that is added to the signal. Signals are detected
> when signal + noise exceeds some threshold.
The whole process sounds much like attempted weather prediction
at the quantum scale - the application of macroscopic principles
to something which is not in the realm of the macroscopic.
QT is a theory of the microscopic realm constructed by
the intellectual processes of a macroscopic being. The
penetration of the intellect into the microscopic realm
has suffered from the portage of macroscopic baggage,
and especially the "bola" of the mechanistic paradigm.
[Note: "bola" (Nepalese): "large sack carried over one's back]
We are on a path, but where does it lead?
All the best,
Pete Brown
-----------------------------------------------------------------------
BoomerangOutPost: Mountain Man Graphics, Australia
Thematics: Publications of Peace and Of Great Souls
Webulous Coordinates: http://magna.com.au/~prfbrown/welcome.html
QuoteForTheDay: "All things are Connected" - Chief Seattle (1854)
-------------------------------------------------------------------------
ksb...@seanet.com
Kevin Brown wrote:
> ... when we pass electrons through a Stern-Gerlach apparatus, do we
> observe any electrons that "can't decide" whether to be deflected up
> or down? My understanding is that we do not.
rto...@kcbbs.gen.nz (for Caroline Thompson) wrote:
> Tell me more! Has any actual Stern-Gerlach experiment ever been
> done with ELECTRONS?
The original Stern/Gerlach experiment consisted of passing a
beam of silver atoms horizontally through a magnetic field whose
intensity increased in the vertical direction. Because of the
gradient in the magnetig field, the magnetic dipole moment of
the atom causes it to be deflected up or down by an amount
proportional to the vertical component of dipole moment. After
passing through the magnetic field, the atoms arrive at a metal
plate where they condense and leave a visible trace.
Classically we would expect the axis of angular momentum, and the
associated magnetic dipole moment, to have an arbitrary orientation
in space, so the vertical components of the dipole moment should
be distributed continuously between the extremes of straight up and
straight down. Thus we would expect to find the silver atoms
distributed on the collector plate in a continuous range from
the maximum upward deflection to the maximum downward deflection.
That isn't what was found. Instead, all the silver atoms were
deflected either UP or DOWN by an amount corresponding to the
FULL dipole moment of the atom. They tried rotating the magnetic
field gradient to measure the moment component in other directions,
and in all cases the beam split into two equal parts along the
direction of the field gradient, half UP and half DOWN. They
also did this for several other kinds of atoms, and always found
that the beam split into discrete parts rather than being spread
in a continuous pattern.
These results indicate that the "preferred orientations" of atoms
in space are quantized, and therefore the dipole moments and spins
in any particular direction are quantized, i.e., we always get a
discrete result, not a continuous spectrum of results. Experiments
of this type have been repeated many times since 1922, and the
results are very well established.
Obviously these experiments were not designed to test Bell
inequalities, but they are nevertheless very damaging to any model
that tries to explain violations of Bell inequalities based on the
"realistic ball" model combined with the idea that our measurements
may fail to discriminate the value of the quantized variable when
the underlying realistic variable is close to the boundary (as you
proposed). These experiments give no support to the idea that any
significant fraction of the particles exhibits anything other than
the predicted discrete behavior.
In reaction to my comments on quantum spin you've mentioned that
you're not really interested in spin, because (almost) all tests of
Bell inequalities have involved the polarization of photons rather than
the spin of massive particles. In response I have to point out that
your paper described the "ball" analogy in terms that are, strictly
speaking, only applicable to spin-1/2 particles (such as electrons),
and not spin-1 particles (such as photons). When writing my original
comments I decided to just follow along with your spin-1/2 description,
accepting that *in principle* it could be converted to a corresponding
spin-1 model.
Of course, in doing so I let slide one of the main problems facing
all "simple realistic models" of the kind you described, namely, the
fact that whatever mechanism you visualize for particles of a certain
spin is inevitably mangled when applied to particles of any other spin.
For example, it's clear than any "realism" we attach to your "line-of-
sight" analogy based on spin-1/2 particles must be abandoned when
applied to spin-1 particles, because the geometrical relations are
completely different. For reasons like this, "simple realistic models"
have a tendancy to become very complicated and utterly unrealistic.
Anyway, the situation based on polarization of light is essentially
no different than that based on spin. Of course, the true "spin
axis" of a photon is always in the direction of motion (because photons
travel at c), but light acts like a transverse wave, i.e., the electric
field oscillates transversely in a plane normal to the direction of
motion (and of course the magnetic field oscillates in a direction
perpindicular to both the electric field and the direction of motion).
So if you are looking at a photon as it comes directly towards you, it
will "look" sort of like this
|E
|
_______|______B
|
|
|
where E signifies the oscillating electric field and B the magnetic
field. (This orientation is not necessarily fixed - it's possible
for it to rotate like a windmill - but it's simplest to concentrate
on "plane-polarized" photons.) The photon is said to be polarized in
the direction of E.
A typical beam of ordinary light has photons with all different
polarizations mixed together. However, a sheet of polaroid basically
only allows photons to pass through if their electric field is
oscillating in one a particular direction, so when you pass a beam
of light through a polaroid sheet the light that passes through is
"polarized", because all the photons have their electric fields
aligned.
Notice that a polaroid sheet cuts the intensity of an ordinary
light beam about in HALF. You might think that far fewer than half
the photons would get through, because their initial polarization
angles are uniformly distributed in all directions, and the polaroid
lets through only those that are aligned in one particular direction.
However, just as in the Stern/Gerlach experiment mentioned above, the
polaroid sheet acts as a measurement for each photon and gives one of
two answers, as if the photons were all in one of just two polarizations,
PARALLEL or PERPINDICULAR to the sheet's axis. This is analagous to
the UP/DOWN results for spin-1/2 particles like electrons. As in the
case of massive particles, we do not find a significant fraction of the
photons unable to decide whether to pass through the sheet or not, so
again this argues against any notion that we're less likely to see a
clear quantum choice when the presumed underlying "realistic variable"
is close to the boundary.
Of course, the original Aspect experiment failed to provide a way
of distinguishing between the failure of a photon to pass through the
polarizer and a failure to detect a photon that HAD passed through.
In view of the relatively low detection efficiencies this was an
obvious shortcoming. Thus the original Aspect experiment (the one
that you focused on in your paper) is well known to have been poorly
designed in this respect, and other than as an exercise in statistical
analysis there's really no point in studying making it the focus of
your investigation. The experiment that people found persuasive was
the follow-up, which used the modified form of the Bell inequality
allowing them to run with one or both of the polarizers removed. This
test not only clearly violated the Bell inequalities, but matched the
QM predictions with high precision.
Kevin Brown wrote:
> ... we are asked to imagine that when two particles emerge from a
> singlet state in opposite directions and interact with spin-measuring
> devices, one of them exhibits a definite spin when measured at any
> angle, but the other sometimes doesn't exhibit spin, if the measurement
> angle happens to be nearly perpindicular to its spin axis.
Caroline Thompson wrote:
> No. When it comes to asymmetry, what I'm really interested in is
> Aspect's actual EPR experiments. Here there is no shortage.
I think we're in agreement here, i.e., the notion of asymmetric
measurement propensities as a general physical phenomenon for coupled
particles is not very plausible. You introduced the asymmetry to your
model strictly as an ad hoc strategy in support of the thesis that
Aspect's (original) experimental results do not, by themselves,
entirely rule out the "realistic ball" explanation. As mentioned
above, no one disputes that the early experiment of Aspect left wide
open the possibility of non-QM interpretations. What's remarkable,
in view of the laxness of those early results, is that they were
still strong enough to rule out not only the "simple ball" model, but
also the *modified* model with limited "detection angles". It was
necessary to further modify the model by stipulating *asymmetry* in
the limitations on detection angles. This, by itself, doesn't logically
exclude the model, but it certainly contributes to its ad hoc character.
In any case, this is all moot in view of the subsequent experiments.
Caroline Thompson wrote:
> The visibility is in fact only slightly greater than the prediction
> (0.5) of the simplest realist model - the model that assumes the counts
> obey Malus'Law *exactly* and that you get the expected coincidence
> probabilities by multiplying the two singles probabilities, for fixed
> polarisation angle. The basic QT prediction for visibility is 1.0, but
> it is not difficult to find excuses to reduce this to the figure
> observed!
Yes, it's generally easier to explain efficiencies less than 100% than
it is to explain efficiencies greater than 100%.
Kevin Brown wrote:
> So, what becomes of the "chaotic ball" model? In addition to being
> explicitly ad hoc relative to the original Aspect experiments, it's
> basic premise appears to conflict with a large body of experimental
> results on the quantization of spin.
Caroline Thompson wrote:
> See above. What results? But never mind, because quantized particle
> spin doesn't come into real EPR experiments.
It might still be worthwhile to familiarize yourself with the physics
of quantum spin and polarization, and the experimental basis of our
understanding of those phenomena.
Caroline Thompson wrote:
> Polarisation in the real world can be treated as a real vector. I'm
> not so sure how many dimensions we need for Rarity's "phase and
> momentum", but this is certainly a continuous variable also.
I read some of the material on your web site, where you suggest that
QM is fundamentally wrong-headed because nothing is really quantized.
I'm not sure how carefully you've thought through all the consequences
of this view. For example, how do you avoid the "ultra-violet
catastrophe", and how do you explain the photo-electric effect,
the Stern-Gerlach results, etc., without quantization?
Caroline Thompson wrote:
> The reason Aspect (and all other EPR experimenters) is able to get good
> agreement with QT is that the realist prediction is not so very different
> and that there is a great deal of freedom in the choice of parameters.
> If all else fails, one can find excuses for adjusting the QT prediction.
> Is it not fair that the realist prediction should also be adjusted?
I think one should be as fair as the facts allow, but no fairer.
Ilja Schmelzer
rto...@kcbbs.gen.nz (Ray Tomes) writes:
> In article <5q3pmg$gsj$1...@news.fsu.edu>, j...@ibms48.scri.fsu.edu (Jim
> Carr) wrote:
> > Since she has not disproved Bell's theorem, exactly which part she uses
> > would be interesting to figure out. I have not grabbed all of that
> > yet, just the HTML part, and the description of the mechanistic model
> > was not evident in it. Should I look in a particular one of those
> > papers for the algorithm she carried out?
>
> In postscript at http://www.aber.ac.uk/~cat/Chaotic/chaotic.ps
> also published as Found. Phys. Lett. 9, 357 (1996); also available at
> http://xxx.lanl.gov/archive/quant-ph, ref. no. 9611037.
If I understand correctly, this criticism assumes that there is
something substracted as noise which is
1. too high
2. correlated in some special way with the used data.
In other words, Aspect does not have a good enough detector, that's
why this experiment may be explained by some local hidden variable
theory, but nothing new about Bell's inequality itself and an ideal
experiment.
Ilja
--
Ilja Schmelzer, D-10178 Berlin, Keibelstr. 38, <schm...@wias-berlin.de>
http://www.cyberpass.net/~ilja
John Sidles
In article <5qipre$5bs$2...@enterprise.desy.de>,
Patrick van Esch <van...@jamaica.desy.de> wrote:
>
>Ilja Schmelzer (schm...@fermi.wias-berlin.de) wrote:
>: something substracted as noise which is
>
>: 1. too high
>: 2. correlated in some special way with the used data.
>
>: In other words, Aspect does not have a good enough detector, that's
>: why this experiment may be explained by some local hidden variable
>: theory, but nothing new about Bell's inequality itself and an ideal
>: experiment.
>
>She uses a conspiracy in a modulation of the efficiency
>of the detector. Aspect (and every other reasonable experimenter)
>supposes that he can calibrate his detector with a known beam
>first, and then use this calibration to correct for efficiency
>losses. If somehow this efficiency is correlated with the
>measured quantity of course, no measurement is ever possible.
Latest Phys. Rev. Lett. reports on entangled atom-pairs, as
opposed to the usual entangled photon-pairs. All
data are in accord with standard quantum theory.
This will pose further problems for those who
wish to rescue classical realism.
Patrick van Esch
: 1. too high
: 2. correlated in some special way with the used data.
: In other words, Aspect does not have a good enough detector, that's
: why this experiment may be explained by some local hidden variable
: theory, but nothing new about Bell's inequality itself and an ideal
: experiment.
Indeed, this is the way I understand it also.
She uses a conspiracy in a modulation of the efficiency
of the detector. Aspect (and every other reasonable experimenter)
supposes that he can calibrate his detector with a known beam
first, and then use this calibration to correct for efficiency
losses. If somehow this efficiency is correlated with the
measured quantity of course, no measurement is ever possible.
It is even not sufficient to use better and better detectors,
because if we have to assume that the efficiency measured in
the testbeam can anyway be modulated, every possible correlation
is possible.
If however, we assume that the efficiency such as measured in
a calibration beam is more or less independent of the hidden
variables that are supposed to occur, then Thompson's argument
isn't valid.
In essence, her statement is: if the detection efficiencies
can depend on the hidden variables, bell's inequalities don't
have to hold. But on second reflection, that is evident.
cheers,
Patrick.
--
Patrick Van Esch
mail: van...@dice2.desy.de
for PGP public key: finger van...@dice2.desy.de
Douglas G. Danforth
the likelihood of a particle's detection is a function of the
polarization and a hidden variable. But this is not unreasonable
at all. Infered detectors are poor gamma ray detectors. Why?
Because the state of the particle and the state of the detector
are not "commensurate", they are not well matched. In the same way
it is perfectly reasonable to assume that the state of a particle
will determined whether the particle is detected for a given
detector state (orientation, construction).
The missing ingredient in the arguments is that a single
real valued hidden variable is sufficient which is *conserved*
between the particles produced in the singlet state (sum of
hidden variable of both particles is constant). The hidden variable
plus the orientation of spin (polarization) determines what
channel of the detector will be activated: +, -, or 0 (no detection).
These comments are unpublished results first derived in 1992 and
shown to John Clauser who concurred. I leave it as an exercise to the
reader to deduce the form of the function on the surface of a cylinder
necessary to reproduce the cos k(a-b) quantum correlation (k=1
electrons,
k=2 photons). The singlet efficiency (probability of detecting one
particle) for this model is S=1/2 +1/pi = .818. The doublet efficiency
(probability of jointly detecting two particles, one at each detector)
is D=2/pi = .636. The relative efficiency (probability of detecting
two particles given that one is detected) is R=D/S=4/(2+pi)=.778 which
is .05 (5%) less than Clauser's bound of .828.
-Doug
--
Douglas G. Danforth, Ph.D. danf...@csli.stanford.edu
Senior Research Engineer (415) 723-2487
Center for the Study of Language and Information
Stanford University
Patrick Van Esch
Douglas G. Danforth wrote:
> Yes, it is possible to construct a simple local model where
> the likelihood of a particle's detection is a function of the
> polarization and a hidden variable.
[snip ]
Of course it is possible. But that is not the working assumption
of an experimentalist. If you take it as a possibility that
any calibration can be altered in a correlated way with the
thing you are measuring, *you can't measure anything at all*.
So you have to make the assumption that that is not what is
happening, until the opposite is demonstrated.
Just to give a silly example. Suppose you are going to find
out whether there is a correlation between your body temperature
and the fact whether you've got the flue or not.
So you take a device, called a thermometer, and you calibrate
it in hot water or whatever other procedure.
Then you go and measure the body temperature of patients and
you try to find a correlation with their "illness" and the
temperature you measure.
Now suppose someone comes along with a theory that "illness"
is "posessedness by an evil spirit". He also tells you that
your thermometer will be influenced by the evil spirit and
that actually, the real temperature remains the same but you
just get that idea because of the influence of the evil spirit
on the thermometer.
If one accepts that, there is no experimental way to deny the
evil-spirit-theory.
cheers,
Patrick.
Ilja Schmelzer
sid...@u.washington.edu (John Sidles) writes:
> Latest Phys. Rev. Lett. reports on entangled atom-pairs, as
> opposed to the usual entangled photon-pairs. All
> data are in accord with standard quantum theory.
> This will pose further problems for those who
> wish to rescue classical realism.
classical realism? No, classical Einstein causality. To give up the
relativity principle solves the problem as well, without rejection of
classical realism. Use Bohm's hidden variables in a preferred frame.
Ilja Schmelzer
From: Caroline Thompson <c.h.th...@dial.pipex.com>
> The QT formula for correlated particles IS wrong. It is not based on
> logic but on blind adherence to dogma.
It is based on QT. As a QT formula, it is correct. To show that QT is
wrong, a difference between observation and QT prediction has to be
found.
> The problem that EPR identified
> back in 1935 is a real one - the incompleteness is a serious fault as
> the real world does have "hidden variables".
I agree with the EPR argumentation. But it tries to show only that
there is something behind QT, not that QT is wrong.
> If people
> had been told more of the facts they would not have accepted the claims
> of the establishment that experiments all supported QT. NONE did. The
> discrepancies have not been large. The establishment has carefully
> refrained from investigating them.
Small discrepancies are usual and normal in experimental science.
Note that the first experimental test in a new domain becomes
well-known, because of historical priority. But that means not be
best, but often the worst (again, because the first) of all
experimental tests.
> The establishment seems happy to carry on chasing the end of the rainbow
> for ever and ever. What do they expect to achieve? We live in a real
> world in which every real experiment has a real explanation! They can
> only make it appear to agree with QT by withholding information and
> twisting the data.
No. The "establishment" has an interest in finding an experiment which
proves QT wrong. It would make the scientist famous, give him a Nobel
price, if it really proves QT wrong.
> Let me repeat: NO EXPERIMENT YET HAS FAILED TO BE EXPLAINED BY LOCAL
> REALIST MODELS.
To prove this, you have to consider all experiments in this domain,
not only the first, well-known of Aspect.
> EXPERIMENTS COULD PROVE QT WRONG.
If QT is wrong.
> NO AMOUNT OF
> INFRINGEMENT OF A BELL TEST IS LIKELY TO PROVE IT RIGHT!
No experiment can prove a theory is right. Theories survive
experiments or die if they become falsified. Survival of the fittest
theory.
> I believe (and, after all, this seems to be at present a matter of faith,
> due to the obstinate refusal of the establishment to investigate relevant
> facts) that this will always be so.
I'm not. In theory, you can make an experiment without the loopholes
left by insufficient particle detectors and so on. This experiment proves
QT is wrong or local hidden variables are wrong.
I'm a theoretician, and don't know the details of current experimental
tests of Bell. Thus, I don't know if there is a hidden variable
theory which survives all existing tests like your toy theory survives
Aspect's, but I doubt. But even if this is the case, these loopholes
will be closed by future progress of experimental science, because the
scientists in this domain are interested in closing all remaining
loopholes and make better, more accurate, tests.
> But my
> main point concerns the whole style of experimentation and dissemination
> of knowledge, and for this you don't need any toy model. I invented it
> just to give the reader confidence that it is always worth searching for
> realist explanations.
I also prefer realist explanations. The realist explanation I prefer
is that there exists a hidden preferred frame where real information
is distributed in a hidden but real way with superluminal speed.
Ilja
Ilja Schmelzer
I have received the following e-mail. The To: line and the first
remark suggests that it was designed as a posting to sci.physics. If
not, sorry for publication of personal e-mail. Ilja
----------------
Subject: Re: Bell Correlations At 6 Miles
From: Caroline Thompson <c.h.th...@dial.pipex.com>
To: sci.p...@fermi.wias-berlin.de, schm...@fermi.wias-berlin.de
Cc: c...@aber.ac.uk
Date: Thu, 17 Jul 1997 11:53:54 -0700
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 2.02 (Win95; I; 16bit)
Hi
I'm having most awful troubles with my technology, so I don't know if
this will work, but here goes:
The QT formula for correlated particles IS wrong. It is not based on
logic but on blind adherence to dogma. The problem that EPR identified
back in 1935 is a real one - the incompleteness is a serious fault as
the real world does have "hidden variables". QT give ALMOST the right
predictions if we use rather underhand methods of adjusting parameters
and data, and are very selective about the results we publish. If people
had been told more of the facts they would not have accepted the claims
of the establishment that experiments all supported QT. NONE did. The
discrepancies have not been large. The establishment has carefully
refrained from investigating them.
The "subtraction of accidentals" that made Aspect's data fit QT is not
exactly a matter of noise. It is not a matter of inefficient detectors.
It is just that he used a source that was unsuitable, so that he could
not tell sufficiently clearly what to count as a pair. He made an
adjustment on the basis of the assumption of independence of cascading
atoms. He knew this was suspect; knew that Freedman back in 1972 had
made sure his results did NOT depend on it. He could have adjusted his
rates to avoid the problem. Perhaps he did! Perhaps his Bell test then
didn't get violated! WHY HAVE WE NOT BEEN TOLD? Why, since the
adjustment is suspect and large, were we not given the unadjusted results
so that we can make up our own minds?
The establishment seems happy to carry on chasing the end of the rainbow
for ever and ever. What do they expect to achieve? We live in a real
world in which every real experiment has a real explanation! They can
only make it appear to agree with QT by withholding information and
twisting the data.
Let me repeat: NO EXPERIMENT YET HAS FAILED TO BE EXPLAINED BY LOCAL
REALIST MODELS. EXPERIMENTS COULD PROVE QT WRONG. NO AMOUNT OF
INFRINGEMENT OF A BELL TEST IS LIKELY TO PROVE IT RIGHT!
I believe (and, after all, this seems to be at present a matter of faith,
due to the obstinate refusal of the establishment to investigate relevant
facts) that this will always be so. There's nothing wrong with Bell's
inequalities except that their assumptions are always violated - not just
the recognised ones, but others concerned with the selection or
adjustment of data.
I'm sorry my Chaotic Ball paper is in Postscript. Kevin Brown's version
was really very good, all in ASCII, though I've now discovered that this
all depends on your viewer. I think the principle affects many
experiments, though the geometry of optical ones is different. But my
main point concerns the whole style of experimentation and dissemination
of knowledge, and for this you don't need any toy model. I invented it
just to give the reader confidence that it is always worth searching for
realist explanations.
Yours, Caroline
Klaus Kassner
Ilja Schmelzer wrote:
> I'm a theoretician, and don't know the details of current experimental
> tests of Bell. Thus, I don't know if there is a hidden variable
I do:-) Bohm's version of QM is such a hidden-variable theory.
It will survive any test that QM itself survives.
It also is a nonlocal hidden-variable theory.
Edward Green
Jim Carr <j...@ibms48.scri.fsu.edu> wrote:
>rto...@kcbbs.gen.nz (Ray Tomes) writes:
>>Caroline Thompson has reproduced all of Aspects results from a simple
>>local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
>>According to QM theorists this is not possible. They are wrong.
>
> That does not prove the prediction of QM wrong, it only says that
> T ==> A as well as Q ==> A. Only (not)A would disprove Q, and
> that would be big news. I am sure Aspect wished that is what he
> had found, since he would be in Stockholm by now.
I think if you review that again you will see you didn't really mean
that. By "(not)A would disprove Q", you surely don't mean that any
other set of results than the set actually seen by A would overturn Q?
Or maybe you do; this becomes semantics; what are 'results'?
Anyway, I am sure Ray grasps the simple logic of the situation:
Note he did not say "QM is wrong", but that "QM theorists are wrong"
if they believe a locally realistic explanation of Aspect is not
possible. Whether such a beast exists (QM theorist who believes
this), I do not know; top men in the field seem aware the alleged
theorem is false -- cf. 'Proposal for Loophole Free Aspect
Experiment', or whatever the exact citation, that Paul Budnik is
wont to frequently post. So Ray errs slightly; his 'QM theorist'
would be an uninformed example of the genre.
Much bandwidth is needlessly wasted by young worthies here intent on
asserting the meta-theorem "Any local mechanism possible within the
window of the loopholes in the Aspect experiment is unreasonable".
You can see the problematic nature of the proposed theorem; first of
all there is the difficulty of defining 'reasonable'; then there is
the necessity of demonstrating through close analysis that any
possible mechanism that fits through the loopholes is unreasonable,
presumably in a sense agreeable to most observers. I sincerely doubt
that any of the advocates of the proposed practical extension to
Bell's work have given the matter this close analysis; they merely
cry credo, and wish you to do the same!
I thought 'scientific method' was supposed to beat such wish
fulfillment thinking out of your skull? Where are the crusty logicians
of yesteryear? Are you _sure_, Mr. Van Esch, intones Herr Professor
von Grossvonfingle, that the entire set of local explanations remaining
compatible with the observed output are individually and collectively
the work of imps, which is to say, outside of all reason? Of course I
am only Ed Green, and if I rap Mr. Van Esch's logical knuckles he is
not likely to mind me. He makes the characteristic assumption of the
intensively technically trained that specialist knowledge beyond that
of their interlocutor gives them the logical upper hand in all
exchanges.
Such people err.
They conflate the observed correlation of those not trained fully in
their discipline with the logic impaired to draw the unjustifiable
conclusion those not fully trained in their discipline are logic
impaired. One sees the same thing among some computer scientists
and lawyers. Each group wants to believe that they have some special
franchise on logical thought denied the rest of the world. As
evidence they look towards (1) the superiority of their specialist
knowledge to laypeople's (2) the statistical superiority of their
logical thews to laypeople.
Naturally, most laypeople are in fact not rocket scientists. This
supports their conclusion, just as the fact that most fish are not
man-eaters supports the contention that all waters are safe to swim in.
It was indeed the beauty of Bell's work that he was able to do an end
run around all possible local mechanisms for the exact predictions of
quantum mechanics and construct a powerful and simple result that
ruled them out. This for the exact predictions. The subsidiary
theorem being asserted; that for the _observed_ results, which do
not meet the necessary conditions to throttle all locally realistic
mechanisms, the residual possibilities are unreasonable is a _more_
subtle result, and one that, as far as I know, has not been
demonstrated.
Don't they teach suspension of judgment in the face of incomplete
evidence?
Here I invoke Mati Meron. We have no requirement to 'believe
something'. If some decision hinged on an answer to the question;
say the future of a research program, then we may have the necessity
making a best guess on the face of available evidence; which guess in
situations like this is necessarily going to hinge on some subjective
elements -- which is to say, elements with significantly varying
Bayesian weights even among the informed.
> Since she has not disproved Bell's theorem, exactly which part she uses
> would be interesting to figure out.
No part, so far as I know, nor has she any necessity of doing so!
She is addressing the actual experimental results. If she produces a
local mechanism that accounts for the actual observed results, then
effectively she is showing that these results -- actually macroscopic
laboratory observations -- do _not_ violate Bell's inequalities;
without the explicit invocation of those inequalities. At worst she
is beating a dead horse, because as I mentioned, those whose opinion
counts are apparently already aware of this fact.
I have some doubt though if she actually understands the content of
Bell's work, which of course is a different issue.
I now repeat my universal put down of Dionysiacs who like to
argue that any viewpoint other than their own is silly: If you simply
want to accept something and get on with your work, fine, but if you
are going to be a plumber, then perhaps you had best confine your
observations on the foundations of fluid mechanics to purely practical
ones. Despite what Daryl implied I at no point said that everybody
must worry about foundation issues all the time; rather those not
worried about them might have the humility to realize that if they
have not thought deeply about them they may not have anything in
particular to say about them, and can not put down the rabble with
the same authority as they can perform a solid calculation within the
paradigm, or determine what size lock-washer to use.
Edward Green
Patrick van Esch <van...@jamaica.desy.de> wrote:
>Ilja Schmelzer (schm...@fermi.wias-berlin.de) wrote:
>
>: In other words, Aspect does not have a good enough detector, that's
>: why this experiment may be explained by some local hidden variable
>: theory, but nothing new about Bell's inequality itself and an ideal
>: experiment.
>Indeed, this is the way I understand it also.
>She uses a conspiracy in a modulation of the efficiency
^^^^^^^^^^
There you go again, Patrick. Your use of the word 'conspiracy' here
is pure polemics.
>of the detector. Aspect (and every other reasonable experimenter)
^^^^^^^^^^
More polemics.
>supposes that he can calibrate his detector with a known beam
>first, and then use this calibration to correct for efficiency
>losses. If somehow this efficiency is correlated with the
>measured quantity of course, no measurement is ever possible.
That last sentence is nonsense. Why would a correlation of detector
efficiency 'with the measured quantity' mean 'no measurement is ever
possible', in any experiment at all? What is a 'measurement', in
this context, that it should be impossible? Why does a confounding
correlation make a controlled laboratory measurement impossible?
I submit you have no idea, because you haven't thought this through.
You have never thought seriously about what constitutes a
'measurement', either in QM or anything else. A measurement is
always presumed to be a function of system state and of external
noise. What ever you intend by the words 'correlation of detector
efficiency with the measured quantity' I defy you to state why it
fails that definition, or why such a would be 'conspiratorial'. Not
that I think you have seriously thought about what would constitute a
'conspiracy' of nature. You just use the word for ill-effect.
You only get away with such illogic because this is utterly worthless
sci.physics, and there are no professionals working in this field
around to call your bluff, only cranks like Ray and I, whose logic
you can with impunity dismiss. It doesn't matter what anybody outside
your clique says, because you don't bother trying to understand;
rather merely repeat your view of things.
>It is even not sufficient to use better and better detectors,
>because if we have to assume that the efficiency measured in
>the testbeam can anyway be modulated, every possible correlation
>is possible.
What on earth does this mean? You may think it means something, but
I have grave doubts.
>If however, we assume that the efficiency such as measured in
>a calibration beam is more or less independent of the hidden
>variables that are supposed to occur, then Thompson's argument
>isn't valid.
>
>In essence, her statement is: if the detection efficiencies
>can depend on the hidden variables, bell's inequalities don't
>have to hold. But on second reflection, that is evident.
Your language is opaque; what do you mean 'bell's inequalities don't
have to hold'? Mathematically? Experimentally? I assume the latter.
Now what, I ask, is so strange about the possibility of
detect/no-detect depending on some variables encoding 'the state of
the photon', which if you like, you can refer to as 'hidden
variables'. A signal reaches the detector. It either blips, or
does not blip. This may depend on the microstate of the detector,
but also on the signal contents. There is nothing strange,
outlandish or conspiratorial about such a model. In any other context
it would be apple pie physical.
The point is, a violation of Bell's inequalities extrapolated from
real detector counts to 'ideal' detector counts is not good enough to
force nature's hand. A violation among macroscopically observed
variables is required; the actual raw data. Even if subsequent
improved experiments should provide such violations, this would not
retroactively justify your view now. It would just mean you got
lucky.
You have demonstrated repeatedly that your true motive in
participating in sci.physics is to exhibit your superiority to the
rabble. Of course you are free to dismiss my arguments as unimportant,
because the truth of the matter is you mentally label everybody outside
the academy a crank, whose logic does not count for anything anyway.
I have already stated that I have my doubts whether Thompson fully
understands the Bell inequalities, but you certainly amply demonstrate
your lack of deep thought on the issue.
Thanks for helping confirm my 80/20 establishment/outsider model.
Your argument that it is 'conspiratorial' to consider the possibility
that detection rates depend on some data encoded in the incoming
light is simply rhetorical. You will have to do better. However,
not on my nickel: I would advise all working cranks not to waste
their time with you, because you are as opaquely stubborn as they
when your vision fails. That fact that theirs fails 80% of the time
and yours 20% does not really mitigate this.
Label me as you will; if you think logic takes a holiday because
nobody with any standing in your grant applications is looking, then
I at least will tell you so.
Ray Tomes
In article <5qqbaj$i...@panix2.panix.com>, e...@panix.com (Edward Green)
wrote [to Jim Carr]:
>Anyway, I am sure Ray grasps the simple logic of the situation:
>Note he did not say "QM is wrong", but that "QM theorists are wrong"
>if they believe a locally realistic explanation of Aspect is not
>possible.
Precisely.
>She [Caroline Thompson] is addressing the actual experimental results.
>If she produces a
>local mechanism that accounts for the actual observed results, then
>effectively she is showing that these results -- actually macroscopic
>laboratory observations -- do _not_ violate Bell's inequalities;
>without the explicit invocation of those inequalities. At worst she
>is beating a dead horse, because as I mentioned, those whose opinion
>counts are apparently already aware of this fact.
I think the most important conclusion to be drawn from it all is that QM
theorists make an assumption that amounts to light *travelling* as
quanta or photons. I suspect that even in Bell's logic he assumes this
for the models that he is testing against. That will not do, as the
realistic mechanistic models that work assume that light travels as
fluctuations in a continuous field and only absorption and emission are
quantised.
The simple fact is that tests to resolve this issue were carried out
long ago and proved quite clearly [to anyone with a mind not firmly
shut] that there are no *travelling* photons.
I refer to the Hanbury-Brown and Twiss (HBT) experiments from mid 1950s
which demonstrated properties of multiphoton counts using half-silvered
mirror. A weak light falls on the half-silvered mirror and two detectors
A and B are on two sides of the mirror taking the counts for reflected
and transmitted photons. It turned out that no matter how weak the light
(or how sparse the average counts), the counts on A and B were
correlated as if the photons were coming in pairs. Of course they
aren't, it is simply the case that a single emission of a photon by one
atom can result in two absorptions by different atoms.
-- Ray Tomes -- rto...@kcbbs.gen.nz -- Cycles & Harmonics Theory --
http://www.kcbbs.gen.nz/users/rtomes/rt-home.htm
Edward Green
Ray Tomes <rto...@kcbbs.gen.nz> wrote:
>In article <5qqbaj$i...@panix2.panix.com>, e...@panix.com (Edward Green)
>wrote [to Jim Carr]:
>
>>Anyway, I am sure Ray grasps the simple logic of the situation:
>>Note he did not say "QM is wrong", but that "QM theorists are wrong"
>>if they believe a locally realistic explanation of Aspect is not
>>possible.
>
>Precisely.
>I think the most important conclusion to be drawn from it all is that QM
>theorists make an assumption that amounts to light *travelling* as
>quanta or photons. I suspect that even in Bell's logic he assumes this
>for the models that he is testing against. That will not do, as the
>realistic mechanistic models that work assume that light travels as
>fluctuations in a continuous field and only absorption and emission are
>quantised.
Yes, you are quite right about Bell's logic, since after all the
entire scheme is based on sets of discrete independent events (the
assumed pair emission, followed by detections or nulls). This is the
hidden baseline assumption that is invisible to some advocates of
'Accept the results and shut up, silly man'. It goes something like
this: Because in many circumstances a model with discrete traveling
photons (which of course are weird in the way all quantum particles
are, which is to say the model is obviously slightly fucked, pardon
my English, but rather than making a mental note 'must look into
this', we are supposed to glory in it and smear it all over our
bodies ecstatically!) are an adequate description of events, the
existence of discrete traveling photons are _proved_, and we may use
their existence in all circumstances as a axiom, in order to extract
logical conclusions about the universe.
Clearly the logic is slightly flawed here, what?
This is why I have been beating a dead horse that now is in a sorry
state indeed, completely flyblown and barely recognizable as equine
remains, to the effect that to force a logical conclusion on the
universe within the ordinary assumptions of macroscopic laboratory
practice (this coffee cup exists, my memory exists, etc.) we must
observe a solid violation of an appropriately related inequality
_among actual macroscopic observations_, eg. raw detector counts over
certain time windows, and not among could-have, should-have, ought-to-
be extrapolations.
It just amazes me how smart some people think they are. This is an
extremely subtle nexus of logic and experiment, we know humans'
propensity, even smart ones, of blindness to assumptions too
obvious to question, and yet some people persist, cousins of those
who pronounced 19th century physics complete, in thinking they can
assert an uber-theorem more subtle than Bell's result itself, to the
effect that "All locally realistic explanations remaining inside the
existing experimental envelope, which class is known not to be empty,
are _whack_".
Way cool, guys. I wish I were that smart.
>The simple fact is that tests to resolve this issue were carried out
>long ago and proved quite clearly [to anyone with a mind not firmly
>shut] that there are no *travelling* photons.
>
>I refer to the Hanbury-Brown and Twiss (HBT) experiments from mid 1950s
>which demonstrated properties of multiphoton counts using half-silvered
>mirror. A weak light falls on the half-silvered mirror and two detectors
>A and B are on two sides of the mirror taking the counts for reflected
>and transmitted photons. It turned out that no matter how weak the light
>(or how sparse the average counts), the counts on A and B were
>correlated as if the photons were coming in pairs. Of course they
>aren't, it is simply the case that a single emission of a photon by one
>atom can result in two absorptions by different atoms.
I'll pass on that one, Ray. Just in the spirit of argument, I note
that explanations could be imagined which were consistent with
traveling photons; at least I think they could. Hey, we have to
give the reigning model at least the same benefit of the doubt!
How am I doin' on crankishness, by the way? I just let loose a
fusillade in various directions, and I can't see the results through
the powder smoke yet.
Jim Carr
rto...@kcbbs.gen.nz (for Caroline Thompson) writes:
>
>I am posting this for Caroline Thompson <xk...@dial.pipex.com> ...
...
>The case I have studied is Aspect's, which uses a calcium "atomic
>cascade". Two lasers illuminate a beam of ionized calcium atoms, with
>the polarisation of the lasers parallel to the direction of the beam.
>The interaction region is pretty small - only fractions of a millimeter
>in each direction. It is possible, is it not, that the lasers interact
>with the WHOLE EM field in which the ions are embedded, setting up
>resonances from time to time? These osillations of the whole field
>propagate out in the form of visible light, with main components at the
>two frequencies that we are filtering out and calling our two photons.
That would be challenging a completely different physics question,
and that is whether electromagnetic energy is carried by photons.
All of these experiments and associated analysis assume that photons
are the correct description of the physics. I don't see how you
could even look at the experiment as an EPR test without that
assumption.
alanpe...@worldnet.att.net
In article <33d6aaac....@aklobs.org.nz> Ray Tomes wrote:
>I think the most important conclusion to be drawn from it all is that QM
>theorists make an assumption that amounts to light *travelling* as
>quanta or photons. I suspect that even in Bell's logic he assumes this
>for the models that he is testing against. That will not do, as the
>realistic mechanistic models that work assume that light travels as
>fluctuations in a continuous field and only absorption and emission are
>quantised.
>
>The simple fact is that tests to resolve this issue were carried out
>long ago and proved quite clearly [to anyone with a mind not firmly
>shut] that there are no *travelling* photons.
>
>I refer to the Hanbury-Brown and Twiss (HBT) experiments from mid 1950s
>which demonstrated properties of multiphoton counts using half-silvered
>mirror. A weak light falls on the half-silvered mirror and two detectors
>A and B are on two sides of the mirror taking the counts for reflected
>and transmitted photons. It turned out that no matter how weak the light
>(or how sparse the average counts), the counts on A and B were
>correlated as if the photons were coming in pairs. Of course they
>aren't, it is simply the case that a single emission of a photon by one
>atom can result in two absorptions by different atoms.
This little note of yours interested me, so I did an AltaVista search.
The HBT effect is used in radio interferometry to measure the diameter
of nearby stars. The page at
http://www.astro.virginia.edu/~eww6n/physics/IntensityInterferometer.html
includes the intriguing sentence "Intensity interferometry was so
counterintuitive that physicists initially believed that it could not
work."
Here seems to be a case where the interpretation of QM’s formalism
directly influenced (wrongly) the predictions to be gotten from it. A
real life example that gives the lie to the claim that interpretations
don’t matter; and that the formalism is the sum toto of the theory.
Would the formalism be any different if this experiment had come out
the other way?
An explanation of HBT can be had at
http://mist.npl.washington.edu/npl/int_rep/tiqm/TI_45.html
QUOTING from that page:
There is a lesson for applications of the TI [transactional
interpretation] in this kind of interference phenomenon: particles like
photons and electrons cannot consistently be described as blobs which
travel from point A to point B. In the HBT effect a whole photon is
assembled at the detector out of half-photons contributed by each of the
two sources.
...
This two photon event may be viewed as a simple case of more general
multiphoton (or multiparticle) events, which may involve many sources
and many detectors. Such transactions can be viewed as assembling
particles at a detector from contributions derived from any number of
sources, with no one-to-one correspondence between particles emitted and
particles detected except in overall number.
END QUOTE
This would be the canonical interpretation of what’s going on. This
canonical view is plausible, if not convincing, in the case of light
coming from stars. It doesn’t make much sense, to me, in the case of the
experiment with the half-silvered mirror which Ray described above. Are
we to believe that the existence of two detectors causes photons to be
emitted (in higher proportion) as doubles?
Presumably, if the beam is not split, the time between photon arrivals
at the single detector is described by a Poisson distribution. Then, if
you split the beam, the time between the combined photon arrivals is no
longer the same distribution. What gives? What is the experts’
explanation of this experiment?
Alan Pendleton
alanpe...@worldnet.att.net
http://www.geocities.com/Athens/2740/
Edward Green
Jim Carr <j...@ibms48.scri.fsu.edu> wrote:
> That would be challenging a completely different physics question,
> and that is whether electromagnetic energy is carried by photons.
>
> All of these experiments and associated analysis assume that photons
> are the correct description of the physics. I don't see how you
> could even look at the experiment as an EPR test without that
> assumption.
Ah, but Jim, you can. That is, if by "EPR test" you mean 'test of
the adequacy of classically flavored models to account for the data'.
That _was_ after all the spirit of Bell's work. He does seem to have
incorporated the assumption of discrete pair emitted photons into his
analysis. The sometimes vociferous objections leveled against the
possibility of any alternative explanation of the output usually fully
ingests this assumption, to the extent that even temporarily
suspending it, or the accompanying assumption that the 'observed
photons' are an unbiased sample of 'all the photons' is taken as
evidence of lunacy.
In fairness, I have probably had more time to think about this in a
fundamental logical way, unfettered by excessive preconceptions than
many of people who comment upon it here. And though I can be obtuse,
stubborn, obstreperous and just downright cantankerous, I am not
logic-deaf. At least not with respect to physics.
I no longer think it is extremely important to disentangle these two
assumptions in the overall analysis. As I have been beating this
unrecognizable bag of bones before me that was once a horse, even
before I understood the logic of Bell's work, the way to bypass these
questions is subject the output to the Rutherford test -- i.e.,
analysis of the macroscopic laboratory observations in a way whose
logic could be explained to the old experimentalist, without knowledge
of quantum mechanics.
Such an analysis is quite possible, using as variables for example
detector counts over time windows. If nature will _not_ cooperate
to provide an unambiguously entangled set of sample correlations over
macroscopic variables, then over time we may become suspicious of her
coyness at painting her nails quantum green and using dainty gloves
so they cannot be seen. Kind of like one of those conspiracy theories
we hear so much about, huh?
The self-assurance of some individuals who, despite their good degree
of training, are simply mostly clueless about this kind of thinking,
is mind-boggling. I think it partially comes from studying physics,
physics, and then some more physics from a tender age; and perhaps an
innate deafness to certain species of arguments. One may come to
confuse technical prowess in manipulating the moieties of various
models for sophistication in thinking about the rules of inference
between models and the phenoumenon, a branch of thought alien to many.
Caroline Thompson
Many thanks to Ed Green and Ray Tomes for recent messages saying what I
should have liked to say about Patrick van Esch's orthodox attitude.
Now what do you make of this? Is this particular academician (anonymous,
of course) of superior intellect?
I've just heard, as expected, that my "Timing and other artifacts" paper
has been turned down by PRL. The referee (and there seems to be only the
one - I did think I had a right to two?) says:
"This paper reopens the well known problem of the 'loopholes' ...
The most important example is the question of the limited
quantum efficiency of the detectors ...
All this has been well known since the famous CHSH paper of 69
... so I think there is nothing to add ...
I would like to quote John Bell ... 'It is difficult for me
to believe that quantum mechanics, working very well for
currently practical setups, will nevertheless fail badly with
improvements in counter efficiency ...' ...
So I do not think it is worth to reanalyse 15 year old
exeriments, confirmed by many more recent ones. The only
important next step it the new generation of loophole free
experiments, and/or new type of experiments with different
schemes."
Has he *READ* my paper, I wonder? Does he not realise that I am *NOT*
talking about the standard detection loophole - one might call my
loophole the *emission* one. This has *not* been discussed except in a
very limited manner back in 1983-5, when just Marshall, Santos, Selleri,
Aspect and Grangier had a few talks, resulting in a very unsatisfactory
victory for Aspect. Who else has ever heard of the problem? And it
leads naturally to the "new type of experiments" the dear man wants - all
sorts of things with different flow rates, intensities, controlled-length
laser pulses etc leading to an actual improvement of *understanding* and
the final rejection of the particle *and* the wave-particle models of
light.
I'd better give you the actual figures in my paper:
For the "Standard" Bell test (-2 LE S LE 2):
Raw data: S = 1.55
Adjusted: S = 2.42
For the "CHSH" test (the one with polarisers present or absent, full
version with upper limit of 0):
Raw data: S = -0.121
Adjusted: S = 0.096
For the "Freedman" test (comparing both present with both absent, upper
limit 0.25):
Raw data: S = 0.195
Adjusted: S = 0.309
Was John Bell a God that he could say what experiments should and should
not do? This particular experiment would have failed badly, I say, if
Aspect had not had high emission rates and hence an excuse to do his
"adjustment", subtracting what he, on the basis of QT, decided were
"accidentals"!
Do look also at my "The EPR Magic Show" paper (in HTML) at my Web site.
This gives the table that should have been in my first message - the
straitforward average coincidence rates.
Are Coincidence Patterns "Interference"?
I'm just getting involved, by the way, in a completely new approach to
EPR: can't we argue that QT should never had put forward their simple
formula? The argument is that they are saying their formula models
*interference* between the two sides. Apparently this is in
"configuration space" or something, which is why we mere mortals can't be
expected to understand it. But the point is that the algebra they use is
as if they were dealing with ordinary interference. See a recent edition
to the quant-ph archive (http://xxx.lanl.gov) ref 9707024, Werner Hofer,
who dismisses the EPR problem in one paragraph. I'm not sure I agree
with him on main paper (he's a friend of mine, but may possibly have
got carried away by pretty formulae) but he's got a point here: true
interference would be impossible to study as time-resolutions would be
impossible. The kind of thing needed would be something like the
Pfleegor and Mandel setup (Pfleegor, R L and Mandel, L "Interference of
Independent Photon Beams", Physical Review 159, 1084-89 (1967)). I can't
remember if I actually said anything in my Chaotic Ball paper in the end,
but I meant to draw attention to the fact the "coincidence patterns" and
"interference patterns" are two very different things, taking place on
very different scales.
Even in Pfleegor and Mandel we still keep to the basic essential of
interference in that the two waves have to be physically in the same
place.
In addition to the timing problem, we have the fact that, for
interference, the waves get added whilst for coincidences they are turned
into 0/1 signals then multiplied.
Yours, Caroline
Phil Gibbs
The paper is now out. Find it at
http://xxx.lanl.gov/abs/quant-ph/9707042
Here is the abstract:
Title: Experimental demonstration of quantum-correlations over more than
10 kilometers
Authors: W. Tittel, J. Brendel, B. Gisin, T. Herzog, H. Zbinden, and N.
Gisin
Comments: 4 pages, REVTeX, 3 postscript figures, submitted to Phys. Rev.
Lett
\\
A Franson-type test of the Bell-inequality is presented. Energy and
time entangled photons at a wavelength of 1310 nm are produced by
parametric downconversion in a KNbO3 crystal and are sent into all-fiber
interferometers using a telecom fiber network. The two interferometers
are located 10.9 km aside from one another. Two-photon fringe
visibilities of up to 81.6 % are obtained. These strong correlations
yield a violation of Bell's inequality by 10 standard deviations thus
confirming the nonlocal predictions of quantum mechanics.
Phil Gibbs
http://www.weburbia.com/ http://www.weburbia.demon.co.uk/
"When all you've got is a hammer, everything looks like a nail"
- Japanese proverb
Ray Tomes
>Yes, you are quite right about Bell's logic, since after all the
>entire scheme is based on sets of discrete independent events (the
>assumed pair emission, followed by detections or nulls).
I think that it is telling that Caroline Thompson is a statistician and that I
used to be one. Have you ever been a statistician Ed? I wonder if David Elm
is? (He must be a card player at least!)
Ray Tomes wrote:
>>I refer to the Hanbury-Brown and Twiss (HBT) experiments from mid 1950s
>>which demonstrated properties of multiphoton counts using half-silvered
>>mirror. A weak light falls on the half-silvered mirror and two detectors
>>A and B are on two sides of the mirror taking the counts for reflected
>>and transmitted photons. It turned out that no matter how weak the light
>>(or how sparse the average counts), the counts on A and B were
>>correlated as if the photons were coming in pairs. Of course they
>>aren't, it is simply the case that a single emission of a photon by one
>>atom can result in two absorptions by different atoms.
>I'll pass on that one, Ray. Just in the spirit of argument, I note
>that explanations could be imagined which were consistent with
>traveling photons; at least I think they could. Hey, we have to
>give the reigning model at least the same benefit of the doubt!
They have been imagined, but only after the event. They now claim that ALL
emission events happen as poisson distributions of photons (i.e an average of 1
but potentially any number)! That is ludicrous. For a start what happens if
you only have one atom? How does it emit a poisson distribution of photons when
it has only 1 unit of energy to give up?
Even if it is argued that it has only been tested for samples with multiple
atoms there is still a fatal flaw. This same flaw seems to me to exist in the
whole way that QM says a laser works. The question is, how do you get more than
one photon travelling as a packet (in phase)?
The point of the question is that if a second photon is to be in phase with the
first, it must detect it! We are told that the only way to dect a photon is to
absorb it (or at least make its details uncertain). So how do a whole packet
(potentially large number) get into phase? The little blighters must be
communicating and giving information to each other! Bohr would turn over in his
grave if he knew that!
Patrick van Esch
Caroline Thompson (xk...@dial.pipex.com) wrote:
: Many thanks to Ed Green and Ray Tomes for recent messages saying what I
Well, I haven't seen your answer to it: how is an experiment _ever_
going to give a conclusive answer if you suppose that your measurement
apparatus can change its calibration in function of the presupposed
hidden variables one is going to detect/rule out ?
Patrick van Esch
Edward Green (e...@panix.com) wrote:
: Much bandwidth is needlessly wasted by young worthies here intent on
: asserting the meta-theorem "Any local mechanism possible within the
: window of the loopholes in the Aspect experiment is unreasonable".
: You can see the problematic nature of the proposed theorem; first of
: all there is the difficulty of defining 'reasonable'; then there is
It is obvious from the above remark that Herr Green prefers
the philosophy-from-the-sofa approach to the hard-working-in-the-lab
approach. EVERY experiment is based upon "reasonableness" assumptions.
Experimental sciences are not mathematics in which you can prove something.
Reasonableness is: "we've been careful, we've been doing the experiment
using all possible affordable care against unwanted effects and unnecessary
assumptions, and we've calibrated our instruments the best we could.
And we find result X. Anyone doubting result X will have to go in the
lab by him/her self, do an experiment and be at least as careful as we."
That is "reasonable". It means: with finite resources, we've done the
best we could and that is by definition absolute truth until a better
experiment comes along.
If someone comes by and says that somehow, still some not-well-defined
other explanation of the results is still possible given these results,
and has to use very un-reasonable conclusions such as magically
altering the calibration of instruments, then this is of not much value
until that person does a better experiment and shows that indeed, in
this better experiment it is shown that his/her theory was right
and the standard explanation is clearly wrong.
I would define "reasonable" as: using the available resources,
NOT making the assumption of which the reasonableness is questioned
cripples so much the possibility of drawing conclusions that it
becomes counter-productive.
: the necessity of demonstrating through close analysis that any
: possible mechanism that fits through the loopholes is unreasonable,
: presumably in a sense agreeable to most observers. I sincerely doubt
You should read "The Character of Physical Law" by Feynman,
what he has to say about that ...
: that any of the advocates of the proposed practical extension to
: Bell's work have given the matter this close analysis; they merely
: cry credo, and wish you to do the same!
No, it is the religion of the classicists that is shocked by
quantum mechanics and so any means is good to escape the
acceptance of the violation of classical determinism.
In no other circumstances anyone would claim that one has
overlooked the possibility that the instrument calibration has
changed by an unknown mechanism.
: I thought 'scientific method' was supposed to beat such wish
: fulfillment thinking out of your skull? Where are the crusty logicians
: of yesteryear? Are you _sure_, Mr. Van Esch, intones Herr Professor
: von Grossvonfingle, that the entire set of local explanations remaining
: compatible with the observed output are individually and collectively
: the work of imps, which is to say, outside of all reason? Of course I
: am only Ed Green, and if I rap Mr. Van Esch's logical knuckles he is
: not likely to mind me. He makes the characteristic assumption of the
: intensively technically trained that specialist knowledge beyond that
: of their interlocutor gives them the logical upper hand in all
: exchanges.
This sounds like frustration to me Ed !
What I'm saying is this: Aspect and his followers have tested
a rather far-going prediction of QM, which is normally not
obtainable by classical physics as we know it, and they found
agreement with the QM predictions. What seems to happen now
is that the people who - for religious reasons - cannot accept
the non-classical (be it quantum or something else) look into
reasons of what could possibly have happened so that they can
rescue their classical ideas. They go into the photons influencing
the computer that recorded the measurements, giving up
special relativity and locality and now, Caroline Thompsons' argument,
that the calibration (the efficiency) of the detectors is influenced
by unknown quantities.
All this is _logically_ not excluded. But I call it unreasonable,
because one doesn't hold that kind of reasoning to ANY other
experimental result. If this kind of reasoning is allowed,
most experimental results are meaningless, and the experimental
method itself is completely crippled.
Nobody who is going to take a picture of the earth from space
is going to claim that some possibility why the earth appears
round is that some unknown influence works onto the optics
of the camera, so that actually the earth IS flat but simply
appears round on those pictures, and hence that there is still
a loophole that the earth - is - flat theory can still stand
with those pictures in hand.
Why do these arguments come into play in Aspect-like experiments ?
cheers,
Patrick.
: Such people err.
--
Patrick Van Esch
Tom Potter wrote:
>
> Patrick van Esch <van...@jamaica.desy.de> wrote in article
> > approach. EVERY experiment is based upon "reasonableness" assumptions.
>
>
> 1. The probability of success.
>
> 2. The "cost" of the action.
>
> 3. The "profit" if successful.
Ok, so what is the profit of "Aspect's experiments didn't
rule out all classes of hidden variable theories. Notably,
the class in which we cannot trust our measurement apparatus
is one of them."
What is the probability of success ? I would say, since this
is not a repeatable question, the only probability that
makes sense is Bayesian subjective probability --- and that
was what all the discussion was about ?
What is the cost ? What is the cost of doing an unfeasable
experiment, namely using an apparatus that is, contrary to
what is claimed by Caroline Thompson et al., an apparatus
that is robust against an unknown influence ?
>
> The problem with institutionalized scientists
> is that they are conditioned to only consider
> high success ratios reasonable, and they rarely
> factor in the potential profit ratio.
That is not true. The problem is that institutionalised
scientists, like businessmen, make the same calculation
you just explained, and if they consider the ROI too low
they don't do it.
>
> They are willing to spend billions of dollars
> with little hope of profit, if the odds of success are high,
> and are not willing to pursue projects with little hope
> of success, no matter what the potential profit would be.
> ( Because it would make them look stupid if they failed. )
I don't think so. Sending a probe to Mars is a LOWPROBABILITY
event. They won't find any life, believe me. But IF they
find it, this will revolutionise biology.
>
> This is why astrologers invented astronomy,
> and alchemists invented chemistry.
You are right. Very low profit business. Not very
interesting.
cheers,
Patrick.
Tom Potter
> Edward Green (e...@panix.com) wrote:
>
>
> : Much bandwidth is needlessly wasted by young worthies here intent on
> : asserting the meta-theorem "Any local mechanism possible within the
> : window of the loopholes in the Aspect experiment is unreasonable".
> : You can see the problematic nature of the proposed theorem; first of
> : all there is the difficulty of defining 'reasonable'; then there is
>
> It is obvious from the above remark that Herr Green prefers
> the philosophy-from-the-sofa approach to the hard-working-in-the-lab
> approach. EVERY experiment is based upon "reasonableness" assumptions.
The "reasonableness" of an action depends upon three factors.
1. The probability of success.
2. The "cost" of the action.
3. The "profit" if successful.
In other words, it is unreasonable to invest $1.00
to get a profit of $1,000,000 if the odds are 10,000,000 to 1.
On the other hand, it is reasonable to invest $1.00
to get a profit of $1,000,000 if the odds are 100,000 to 1.
The problem with institutionalized scientists
is that they are conditioned to only consider
high success ratios reasonable, and they rarely
factor in the potential profit ratio.
They are willing to spend billions of dollars
with little hope of profit, if the odds of success are high,
and are not willing to pursue projects with little hope
of success, no matter what the potential profit would be.
( Because it would make them look stupid if they failed. )
This is why astrologers invented astronomy,
and alchemists invented chemistry.
Tom Potter http://pobox.com/~tdp
Ray Tomes
In article <33D7CA...@club.innet.be>, Patrick Van Esch
<van...@club.innet.be> wrote:
>Ok, so what is the profit of "Aspect's experiments didn't
>rule out all classes of hidden variable theories. Notably,
>the class in which we cannot trust our measurement apparatus
>is one of them."
>What is the probability of success ? I would say, since this
>is not a repeatable question, the only probability that
>makes sense is Bayesian subjective probability --- and that
>was what all the discussion was about ?
>What is the cost ? What is the cost of doing an unfeasable
>experiment, namely using an apparatus that is, contrary to
>what is claimed by Caroline Thompson et al., an apparatus
>that is robust against an unknown influence ?
Detectors with a higher probability of detection are a definite area of
improvement from everyone's point of view.
Otherwise I don't think that there is very much cost associated with anything
else required to do a proper test that determines whether a local model based on
light travelling as a non-quantised field (i.e. Maxwell's equations) can explain
all the data. This has tremendous possible value, because it can lead to
theorists to going beyond QM.
What is required is simply to not assume that QM is correct in setting up the
experiment. For example, rather than adjust the detectors so that the expected
(by QM) shaped curve results with polariser angle, it would be better to
calibrate the detector by a neutral method. I suggest that can be done with a
steady source by replacing a single polariser by three polarisers with the outer
two at the same angle as the original one and the other at a variable angle to
achieve the required adjustments in intensity to do calibration.
Kevin Brown
On Wed, 23 Jul 1997 Phil Gibbs <philip...@pobox.com> wrote:
>The paper is now out. Find it at
>http://xxx.lanl.gov/abs/quant-ph/9707042
>Here is the abstract:
>Title: Experimental demonstration of quantum-correlations over more
> than 10 kilometers
>Authors: W. Tittel, J. Brendel, B. Gisin, T. Herzog, H. Zbinden,
> and N. Gisin
>time entangled photons at a wavelength of 1310 nm are produced by
>parametric downconversion in a KNbO3 crystal and are sent into all-
>photon fringe visibilities of up to 81.6 % are obtained. These strong
>correlations yield a violation of Bell's inequality by 10 standard
>deviations thus confirming the nonlocal predictions of quantum
>mechanics.
Considering some of the discussions in this thread, it might be worth
noting the concluding remarks of the paper:
"In conclusion, this experiment gives for the first time
evidence, that the spooky action between entangled photons
does not break down when separating the particles by a
physical distance of 10 km. Further investigations with
interferometers and a pumplaser of better performance
should be made. However, we need to acknowledge that
neither this nor any of the previous experiments can close
the detection loophole [17],[18]. Hence in order to deduce
nonlocality from existing data, an additional assumption
is needed, namely that the detected pairs of particles
form a fair sample of the set of all emitted pairs."
[17] J.F. Clauser, M.A. Horne, A. Shimony, and R.A. Holt,
Phys. Rev. Lett., 23, 880, (1969).
[18] P. Pearle, Phys. Rev. D, 2, 1418, (1970); E. Santos,
Phys.Rev. A, 46, 3646, (1992).
Jim Carr
Jim Carr <j...@ibms48.scri.fsu.edu> wrote:
|
| rto...@kcbbs.gen.nz (Ray Tomes) writes:
| >
| >Caroline Thompson has reproduced all of Aspects results from a simple
| >local mechanistic based computer model (see http://www.aber.ac.uk/~cat).
| >According to QM theorists this is not possible. They are wrong.
|
| That does not prove the prediction of QM wrong, it only says that
| T ==> A as well as Q ==> A. Only (not)A would disprove Q, and
| that would be big news. I am sure Aspect wished that is what he
| had found, since he would be in Stockholm by now.
e...@panix.com (Edward Green) writes:
>
>I think if you review that again you will see you didn't really mean
>that.
What I should have said is T ==> almost-A.
>By "(not)A would disprove Q", you surely don't mean that any
>other set of results than the set actually seen by A would overturn Q?
No, I mean a set of observations E by <whomever> that clearly
disagree (99% CL, say) with the A deduced from Q. That could
also contradict T, but it is possible the "almost" is not as
good as implied by discussions here so one could be excluded
but not the other.
>Or maybe you do; this becomes semantics; what are 'results'?
'results' are imperfect.
Jim Carr
Actually from: Caroline Thompson <c.h.th...@dial.pipex.com>:
>
>The QT formula for correlated particles IS wrong. It is not based on
>logic but on blind adherence to dogma.
Physicists interpret a statement that a "formula is wrong" to mean
it does not follow from the theory. You could also say that it
disagrees with experiment, but I do not think that is justified.
As I read your paper, you allege that another explanation also
describes the data, starting from different premises. I do not
see a proof that those premises provide a counter-example to the
theorem of Bell, but maybe I missed that point.
> QT give ALMOST the right
>predictions if we use rather underhand methods of adjusting parameters
>and data, and are very selective about the results we publish.
That is a serious charge.
I did not see that charge in your "chaotic" paper.
>The "subtraction of accidentals" that made Aspect's data fit QT is not
>exactly a matter of noise. It is not a matter of inefficient detectors.
> It is just that he used a source that was unsuitable, so that he could
>not tell sufficiently clearly what to count as a pair.
All sources are unsuitable, and all detectors are inefficient. You
alluded to the existence of "computer programs" in the paper, and
the normal way of dealing with such issues is to model them. That
is, include noise, pileup, and accidentals in a monte carlo simulation
of your realistic model, since they will be present even then. For
example, if the flip rate is high enough and slightly random, one
detector can get one 'event' and the other, another. You cannot
assume that the two detectors will always have their window open at
the same instant, or that they will detect perfectly. That is where
"accidentals" come from. They are real in any model of the experiment
except an unrealistically ideal one.
>He made an
>adjustment on the basis of the assumption of independence of cascading
>atoms. He knew this was suspect; knew that Freedman back in 1972 had
>made sure his results did NOT depend on it. He could have adjusted his
>rates to avoid the problem.
Not bad for a thesis experiment. The next one should be done
more carefully. I am still looking over the Tittle preprint.
That is an important consideration.
>Perhaps he did! Perhaps his Bell test then
>didn't get violated! WHY HAVE WE NOT BEEN TOLD?
This is, again, a very serious charge of fraud by Aspect.
>Let me repeat: NO EXPERIMENT YET HAS FAILED TO BE EXPLAINED BY LOCAL
>REALIST MODELS.
I am not convinced that your model is local.
After all, it has long-range correlations. They are just of a
different kind. You still have "faster than light signals" as
that mad New York Times story put it about the Geneva experiment,
since detector A knows what B saw before B can tell A about it.
In your case the source can tell A what B will see, whereas in
quantum theory that is not possible, but you still have a super-
luminal connection between A and B.
My observation is that you have _correlated_ detection probabilities,
not variable detection probabilities, and that this is significant
for part of the conclusion. Realist but not local.
>I'm sorry my Chaotic Ball paper is in Postscript.
I'm not. ;-) I have a hassle with PDF.
What I wish it had was a comparison of the two curves, particularly
the slope at the center, as a function of delta and a comparison of
the curves to data with uncertainties for one of the experiments
with tabulated Chi^2 under different assumptions about background
sources. For example, here you make a big deal of the accidentals,
so I would like to see the data without the subtraction compared
to your prediction, but the key question not answered is what
precision data need to distinguish between the two models. That
can be used to improve experiments. Of course, that may be in a
paper I have not looked at; I was told "chaotic" was the one to get.
But it does pass my first test as a referee: my marginal notes
at the start (assumes perfect detector, fuzzy edge, add randoms)
were (except for adding randoms) addressed in some detail. In
addition, the QM prediction should have noise added to it just
imperfect detection was assumed with the ball.
The question I do not see answered is if it can deal with the
experiments on the Quantum Zeno Effect that have been done.
That interests me.
Patrick Van Esch
Kevin Brown wrote:
(quoting from the paper)
> However, we need to acknowledge that
> neither this nor any of the previous experiments can close
> the detection loophole [17],[18]. Hence in order to deduce
> nonlocality from existing data, an additional assumption
> is needed, namely that the detected pairs of particles
> form a fair sample of the set of all emitted pairs."
>
> [17] J.F. Clauser, M.A. Horne, A. Shimony, and R.A. Holt,
> Phys. Rev. Lett., 23, 880, (1969).
> [18] P. Pearle, Phys. Rev. D, 2, 1418, (1970); E. Santos,
> Phys.Rev. A, 46, 3646, (1992).
Absolutely. But the fair assumption that goes into this
is that the efficiency of a non-perfect photon detector
(meaning: a photon detector that sometimes does not detect
a photon that was there) is a fixed constant, uncorrelated
with the process under consideration. That is actually exactly
the objection Caroline Thompson has about this kind of
experiments. One assumes that of the set of N photons
that would have been detected by a perfect detector, only
on average epsilon * N are really detected by the detector,
and that this average (of a Poisson distribution of course)
is independent of other phenomena.
One has to be careful if one rejects this hypothesis. Because
in that case a "perfect" detector is not possible: even if
one calibrates the efficiency and one finds 1, this does not
mean that in the special circumstances of the experiment
that number remains 1. So even with apparently perfect equipment,
the possibility always remains that in special and unknown
conditions this equipment is arbitrarily unefficient.
Saying this is equivalent to say that ANY experiment yields
unreliable results.
cheers,
Patrick.
Douglas G. Danforth
> neither this nor any of the previous experiments can close
> the detection loophole [17],[18]. Hence in order to deduce
> nonlocality from existing data, an additional assumption
> is needed, namely that the detected pairs of particles
> form a fair sample of the set of all emitted pairs."
>
> [17] J.F. Clauser, M.A. Horne, A. Shimony, and R.A. Holt,
> Phys. Rev. Lett., 23, 880, (1969).
> [18] P. Pearle, Phys. Rev. D, 2, 1418, (1970); E. Santos,
> Phys.Rev. A, 46, 3646, (1992).
Thank you Kevin for posting this note. I was about to mention the
same thing from a slightly different point of view. The authors
of the paper focus on "visibility" of the signal and ignore
(except for the mention above) detection efficiency. It is possible
to create a local model that has 100% "visibility" and violates
Bell's inequality simply because it has less than perfect efficiency.
Visibility is not the same as efficiency.
To my knowledge no experiment has been performed with relative
efficiency (probability of detecting two photons given that one
is detected) greater than 82.8% which is Clauser's bound.
Can anyone refute this? I'd like to know.
Paul Budnik
Patrick van Esch (van...@jamaica.desy.de) wrote:
: Edward Green (e...@panix.com) wrote:
: It means: with finite resources, we've done the
: experiment comes along.
That is a novel definition of `absolute truth'.
You can never prove things (outside of formal mathematics) absolutely.
Bell's theorem and related work sets up conditions for a reasonably
conclusive experiment. To my knowledge those conditions have yet to be met.
: If someone comes by and says that somehow, still some not-well-defined
: other explanation of the results is still possible given these results,
: and has to use very un-reasonable conclusions such as magically
: altering the calibration of instruments, then this is of not much value
: until that person does a better experiment and shows that indeed, in
: this better experiment it is shown that his/her theory was right
: and the standard explanation is clearly wrong.
You have it backwards. The explanation is not going to come before an
effective experiment. As often happens with science I think a more
complete theory is going to be impossible to develop in detail without
experimental input showing where QM fails.
: I would define "reasonable" as: using the available resources,
: NOT making the assumption of which the reasonableness is questioned
: cripples so much the possibility of drawing conclusions that it
: becomes counter-productive.
That is an absurd definition. First of all the best available resources
depend on the size of your grant. I suspect an effective test has
been possible for quite some time but no one has had enough interest
in the outcome to make the necessary investment.
[...]
: No, it is the religion of the classicists that is shocked by
: quantum mechanics and so any means is good to escape the
: acceptance of the violation of classical determinism.
There are a great many conceptual problems with QM. A love for
determinism is the least of them. That is why it is quite reasonable
to expect the theory to ultimately be replaced by a deeper theory as
has happened many times in the past.
[...]
: because one doesn't hold that kind of reasoning to ANY other
: experimental result. If this kind of reasoning is allowed,
: most experimental results are meaningless, and the experimental
: method itself is completely crippled. [...]
That is not the situation. To show a violation of Bell's inequality
you must do so with direct macroscopic timing measurements. You
need to focus on the requirements of the experiment and not worry
about what mechanisms might be subject to what loopholes.
--
Paul Budnik
pa...@mtnmath.com, http://www.mtnmath.com
Caroline Thompson
Patrick Van Esch <van...@club.innet.be> wrote in article
<33D911...@club.innet.be>...
> Kevin Brown wrote:
> (quoting from the paper)
> > However, we need to acknowledge that
> > neither this nor any of the previous experiments can close
> > the detection loophole [17],[18]. Hence in order to deduce
> > nonlocality from existing data, an additional assumption
> > is needed, namely that the detected pairs of particles
> > form a fair sample of the set of all emitted pairs."
Patrick van Esch:
>
> Absolutely. But the fair assumption that goes into this
> is that the efficiency of a non-perfect photon detector
> (meaning: a photon detector that sometimes does not detect
> a photon that was there) is a fixed constant, uncorrelated
> with the process under consideration. That is actually exactly
> the objection Caroline Thompson has about this kind of
> experiments. One assumes that of the set of N photons
> that would have been detected by a perfect detector, only
> on average epsilon * N are really detected by the detector,
> and that this average (of a Poisson distribution of course)
There is no "of course" about it being a Poisson distribution, as
the total count originates from signals of varying intensities. For
fixed intensity don't we in fact have a Binomial? But this is not
important ...
> is independent of other phenomena.
> One has to be careful if one rejects this hypothesis. Because
> in that case a "perfect" detector is not possible ...
> Saying this is equivalent to say that ANY experiment yields
> unreliable results.
>
That's absolutely it! No experiment that assumes you can
really count photons is likely to yield reliable results. How
about looking at my translation of part of Section 6 of
Aspect's thesis (at http://www.aber.ac.uk/~cat)?
And in his 1986 paper with Grangier and Roger on photon
anticorrelation there is a footnote that mentions that Aspect
realised that the photoelectric effect was not in itself evidence
of the particle nature of light - it may not be quantised except
at emission and detection. Personally I am still waiting for
proof of either of these, as it seems more likely that we are
just seeing statistical correlations between intensity of input
(however measured) and number of counts from a photomuliplier.
(See Grangier, Roger and Aspect, Europhysics Letters 1 (4),
173-179 (1986))
And Patrick, could you please read my "The EPR Magic Show"
paper? Perhaps even you will be convinced that Aspect's results
cannot be taken as scientific evidence in favour of QT - not when
the part of QT we are being asked to accept is so ridiculous!
I've sent this paper to be presented at the VigierII conference, and
it has already attracted some interest. I have been asked what
Aspect thought of it. The answer is that he has not replied to any
of my letters, the last one of which included a copy of my PRL
submission ("Timing and Other Artifacts ..."). Yes, I have serious
allegations, and I hope some day Aspect will answer them. So far
I have given him the benefit of the doubt, in that it is conceivable
that he never looked at some of his data. Possibly by now my
abstract will be at http://www.vigier.yorku.ca/VigierII.
But I ask you!
INSTANTANEOUS action over *INDEFINITE*
distances? Pull the other one! The part of QT that makes this
prediction surely contradicts Bohr's idea that QT should agree
with classical theory on the macroscopic scale? And do the
followers of QED agree with this particular part? They have
non-locality, but isn't it just over tiny distances?
While I've got the bit between my teeth, I do hope that I am
not included in your category of "philosophy-from-the-sofa"!
(See Patrick van Esch reply to Edward Green, 24:7:97).
I have pursued this quest without letup for 4 years now,
travelling to anyone who will listen and with no financial reward
whatsoever. I have not been near a lab, but that is only because
it is rather difficult to organise without official status. I *have*
read Aspect's thesis (400 pages of French, which took some
considerable time) and I have studied every paper I could lay
my hands on.
Yours, Caroline
http://www.aber.ac.uk/~cat
Kevin A. Scaldeferri
In article <01bc99f8$57189820$LocalHost@eozfbrwg>,
Caroline Thompson <c.h.th...@dial.pipex.com> wrote:
>
>And Patrick, could you please read my "The EPR Magic Show"
>paper? Perhaps even you will be convinced that Aspect's results
>cannot be taken as scientific evidence in favour of QT - not when
>the part of QT we are being asked to accept is so ridiculous!
This begs one to ask, would you consider Aspect's results as evidence
in favour of QM if you did not feel that the part of QM being tested
was not ridiculous?
I think it's important to ask whether the intent behind Aspect type
experiments is to disprove local realism or to support quantum
mechanics. I feel it is the second (particularly given the loopholes
than make excluding local realism practically impossible). Quantum
mechanics has proved itself spectacularly useful and you will need to
do better than provide another theory that might also explain things
in order to get most scientists to abandon quantum mechanics. You
will need to show an experimental conflict with quantum mechanics in
order to do this.
--
======================================================================
Kevin Scaldeferri Calif. Institute of Technology
"Pragmatism! Is that all you have to offer?"
Caroline Thompson
Jim Carr <j...@ibms48.scri.fsu.edu> wrote in article
<5r92t3$amt$1...@news.fsu.edu>...
> Actually from: Caroline Thompson <c.h.th...@dial.pipex.com>:
> >
> >The QT formula for correlated particles IS wrong. It is not based on
> >logic but on blind adherence to dogma.
> >and data, and are very selective about the results we publish.
>
> That is a serious charge.
I am well aware of this. I am making a charge against the whole
fraternity, who have created a system in which rationality is no longer a
valid criterion for judging how to set parameters or analyse results. QT
was declared long ago to be perfect, and although this is now no longer
thought (or is it?) experimental methods have become corrupted by faith and
group loyalty. Perhaps SOME of the predictions are near perfect.
Experimenters do not dare to admit (at least, not in this sensitive EPR
area) that these particular predictions are not perfect. This is a dying
theory fighting a last desperate battle with no holds barred. Please do
not take it that I am criticising only Aspect. He cannot be blamed for
this collective self-deception, this joint decision to turn the blind eye.
>
> I did not see that charge in your "chaotic" paper.
I was trying to be diplomatic. And I had not yet discovered the
"accidentals".
>
> >The "subtraction of accidentals" that made Aspect's data fit QT is not
> >exactly a matter of noise. It is not a matter of inefficient detectors.
> > It is just that he used a source that was unsuitable, so that he could
> >not tell sufficiently clearly what to count as a pair.
>
> All sources are unsuitable, and all detectors are inefficient. You
> alluded to the existence of "computer programs" in the paper, and
> the normal way of dealing with such issues is to model them. That
> is, include noise, pileup, and accidentals in a monte carlo simulation
> of your realistic model, since they will be present even then. For
> example, if the flip rate is high enough and slightly random, one
> detector can get one 'event' and the other, another.
What do you mean by flip rate here? Detection rate? Or are you thinking
of some mysterious flipping of polarisation direction? I hope not, as your
remarks seem otherwise quite rational.
You say there should be computer simulation. I could not agree more!
Every one of these experiments should be accompanied by simulation. I have
not done much in this way myself - I'm a very poor programmer, far too
impatient. I know what needs doing, though. I don't really know how to
model accidentals, but that does not matter as one could try several
different ways. I have friends in Australia (BarryGilbert and Sue Sulcs)
who have done quite a bit of simulation work and are 100% convinced that
there are realist explanations. They can't be expected to find THE
explanation, as they have not got enough facts.
You know, of course, why simulation is not done - well, my theory as to
why? It is because the instant you start you realise that QT is nuts! QT
does not ALLOW you to trace the progress of individual "photons". See an
article by Richard Feynman (Int. J. Theor. Phys. 21, 467-488 (1982)) which
shows why simulation is not possible. This is just another way of saying
QT is incompatible with Hidden Variables.
> You cannot
> assume that the two detectors will always have their window open at
> the same instant, or that they will detect perfectly. That is where
> "accidentals" come from. They are real in any model of the experiment
> except an unrealistically ideal one.
I don't think you've got this quite right - see my recent papers. I'm all
to happy to see accidentals built into a simulation, though, so long as
they obey MY rules! My point about them is that the accidentals that occur
well away from the spectrum peak may be due to separate events, but in my
picture only one event CAN take place at a time at the source so in the
near vicinity of the peak ALL detections must come from this one event.
There are no accidentals here. Of course what I don't know without further
experiment is how large this no-go area actually is. With modern equipment
this could be investigated, perhaps using the visibility of the coincidence
curve as an indirect way of measuring the independence of emissions.
>
> >He made an
> >adjustment on the basis of the assumption of independence of cascading
> >atoms. He knew this was suspect; knew that Freedman back in 1972 had
> >made sure his results did NOT depend on it. He could have adjusted his
> >rates to avoid the problem.
>
> Not bad for a thesis experiment. The next one should be done
> more carefully.
Bu they are not planning a "next one"! They've done a few experiments in
Stirling that were supposed to investigate Marshall and Santos' idea of
"enhancement", but nobody to my knowledge has investigated the
Marshall/Santos/Selleri idea of accidentals.
> I am still looking over the Tittle preprint.
> That is an important consideration.
I'll look at the paper next week.
>
> >Perhaps he did! Perhaps his Bell test then
> >didn't get violated! WHY HAVE WE NOT BEEN TOLD?
>
> This is, again, a very serious charge of fraud by Aspect.
As I said earlier, a tradition of bad practice has built up. I used to
work as an experimental statistician in a horticultural research station.
Our standards said that we had to publish the data in such a way that the
reader could see what significance it had. If an adjustment was suspect
one should publish unadjusted figures as well. QT has forgotten how to do
this! A myth has built up that empirical data without interpretation via a
model is worthless. The result is that all these experiments will have to
be redone as valuable data has been chucked in the bin.
>
> >Let me repeat: NO EXPERIMENT YET HAS FAILED TO BE EXPLAINED BY LOCAL
> >REALIST MODELS.
>
> I am not convinced that your model is local.
>
Of course it is!
> After all, it has long-range correlations. They are just of a
> different kind. You still have "faster than light signals" as
> that mad New York Times story put it about the Geneva experiment,
> since detector A knows what B saw before B can tell A about it.
Rubbish.
> In your case the source can tell A what B will see, whereas in
> quantum theory that is not possible, but you still have a super-
> luminal connection between A and B.
NO.
>
> My observation is that you have _correlated_ detection probabilities,
> not variable detection probabilities, and that this is significant
> for part of the conclusion. Realist but not local.
Realist AND local, Sir. But yes, I have correlated detection probabilities
arising in a completely rational local way. If they were not correlated
they would not matter, not cause bias.
>
> What I wish it had was a comparison of the two curves, particularly
> the slope at the center, as a function of delta and a comparison of
> the curves to data with uncertainties for one of the experiments
> with tabulated Chi^2 under different assumptions about background
> sources.
Sorry but there is only the one of me and my statistics, having never been
strong, is rusty. Besides, one cannot produce a definite realist
prediction without more information. Also, look at my table in my "The EPR
Magic Show" paper - what could be clearer?
> For example, here you make a big deal of the accidentals,
> so I would like to see the data without the subtraction compared
> to your prediction, but the key question not answered is what
> precision data need to distinguish between the two models. That
> can be used to improve experiments.
I don't think Aspect needed more precision. He could have got any level he
needed simply by increased replication.
> Of course, that may be in a
> paper I have not looked at; I was told "chaotic" was the one to get.
>
> But it does pass my first test as a referee: my marginal notes
> at the start (assumes perfect detector, fuzzy edge, add randoms)
> were (except for adding randoms) addressed in some detail. In
> addition, the QM prediction should have noise added to it just
> imperfect detection was assumed with the ball.
QM does not even try to model the detection process. A Quantum Mechanic is
not allowed to try and answer the question: How does this particular photon
decide whether or not to be detected? The whole ensemble may be known to
have, say, a 10% probability, but the next signal comes along and just how
does it know whether or not it is one of the 10%? My picture says it
depends on the noise (my noise, incidentally, is a very complicated thing
that should be called "effective noise". It includes phase and frequency
so that it is not actually the simple time-varying scalar I may have seemed
to imply). I don't see how noise can enter into the QM formula. In fact I
see no future for the QM formula whatsoever, as I don't see how it can be
repared so as to model TIME sensibly. This is the problem that led Saverio
Pascazio to abandon the EPR problem after writing just a few papers, and he
is one of the most intelligent people I have met.
>
> The question I do not see answered is if it can deal with the
> experiments on the Quantum Zeno Effect that have been done.
> That interests me.
Not really my subject, but I'm sure I can explain it! Just give me a
specific paper to discuss, and a few weeks.
>
> --
> James A. Carr <j...@scri.fsu.edu> | Commercial e-mail is _NOT_
> http://www.scri.fsu.edu/~jac/ | desired to this or any address
> Supercomputer Computations Res. Inst. | that resolves to my account
> Florida State, Tallahassee FL 32306 | for any reason at any time.
>
--
Caroline H Thompson
Department of Computer Science
University of Wales Aberystwyth, SY23 3DB, UK
<http://www.aber.ac.uk/~cat>
Patrick Van Esch
Caroline Thompson wrote:
>
> Jim Carr <j...@ibms48.scri.fsu.edu> wrote in article
> <5r92t3$amt$1...@news.fsu.edu>...
> > Actually from: Caroline Thompson <c.h.th...@dial.pipex.com>:
> > >
> > >The QT formula for correlated particles IS wrong. It is not based on
> > >logic but on blind adherence to dogma.
>
> > > QT gives ALMOST the right
> > >predictions if we use rather underhand methods of adjusting parameters
> > >and data, and are very selective about the results we publish.
> >
> > That is a serious charge.
>
> I am well aware of this. I am making a charge against the whole
> fraternity, who have created a system in which rationality is no longer a
> valid criterion for judging how to set parameters or analyse results. QT
> was declared long ago to be perfect, and although this is now no longer
> thought (or is it?) experimental methods have become corrupted by faith and
> group loyalty. Perhaps SOME of the predictions are near perfect.
> Experimenters do not dare to admit (at least, not in this sensitive EPR
> area) that these particular predictions are not perfect. This is a dying
> theory fighting a last desperate battle with no holds barred. Please do
> not take it that I am criticising only Aspect. He cannot be blamed for
> this collective self-deception, this joint decision to turn the blind eye.
Hum, the conspiracy argument. It would help you to learn a bit QM
before thinking that it is a religion. There is no single experimental
result
known that is not in agreement with QM. So saying that it is
"a dying
>theory fighting a last desperate battle with no holds barred."
is the same as saying that "saying that the world is round is a dying
theory fighting a last desperate battle". I think I stop discussion
here.
You're not serious.
cheers,
Patrick.
Caroline Thompson
Kevin A. Scaldeferri <ke...@cco.caltech.edu> wrote in article
<5resjq$b...@gap.cco.caltech.edu>...
> This begs one to ask, would you consider Aspect's results as evidence
> in favour of QM if you did not feel that the part of QM being tested
> was not ridiculous?
>
I can't say. I do not pretend to be unbiased in this. Surely this is the
one place in science where it is only right and proper to be thoroughly
biased? The way to progress is to investigate results that are unexplained
or explained in a ridiculous fashion.
> I think it's important to ask whether the intent behind Aspect type
> experiments is to disprove local realism or to support quantum
> mechanics. I feel it is the second (particularly given the loopholes
> than make excluding local realism practically impossible).
Agreed, but surely it ought to be NEITHER? The reasonable thing to do is
to experiments to find out which is the better model, in terms of which
model is best at predicting not just one isolated figure (the Bell test
statistic) but how the results vary as you vary the parameters. As I've
said in various papers, QM has not even GOT the necessary parameters - it
does not, for example, have enough parameters to model real detectors (it
has only the single "quantum effeciency"), or real coincidence windows
(which need two parameters, not one, as they need both start and end times
relative to a zero that represents simultaneous emissions).
> Quantum
> mechanics has proved itself spectacularly useful
One could question this. Has its "usefulness" outweighed its disastrous
effect on a few generations of physicists whose confidence in their own
judgement has been destroyed? I gather that the STM experts are drifting
away from it. How can you have just a statistical theory when you can
manipulate individual atoms? If QM gets the right answer for a few atomic
spectra, I expect it is because in this area all you need is the right
constraints and the general idea of how to model the geometry. QM could
well be right in this area.
> and you will need to
> do better than provide another theory that might also explain things
> in order to get most scientists to abandon quantum mechanics. You
> will need to show an experimental conflict with quantum mechanics > in
order to do this.
I am sure that a conflict could be demonstrated in the EPR area. Aspect's
"anomalies" (in thesis, not PRL papers), together with faith in local
realism, show that this is so. QM is wrong in this area. A closer look at
the shapes of the coincidence curves and time spectra would show
discrepancies. It has made a false claim that rules of algebra that seem
pretty good on one scale carry over to this macroscopic realm.
But I actually think it is a bad theory on all scales. It is a bad idea to
train scientists in double-think. Wave-particles are unnecessary for
LIGHT- let's have just waves, waves and more waves, probably ad infinitum
but at least to a level much lower than the quantum.
Caroline
Caroline Thompson
--
Caroline H Thompson
Department of Computer Science
University of Wales Aberystwyth, SY23 3DB, UK
<http://www.aber.ac.uk/~cat>
Patrick van Esch <van...@jamaica.desy.de> wrote in article
<5r7859$f9h$1...@enterprise.desy.de>...
> Caroline Thompson (xk...@dial.pipex.com) wrote:
> : Many thanks to Ed Green and Ray Tomes for recent messages saying what I
> : should have liked to say about Patrick van Esch's orthodox attitude.
>
> Well, I haven't seen your answer to it: how is an experiment _ever_
> going to give a conclusive answer if you suppose that your measurement
> apparatus can change its calibration in function of the presupposed
> hidden variables one is going to detect/rule out ?
If your mind is TOTALLY blocked to the possibility that the world is
comprehensible then perhaps you will never on your own accept the evidence,
though I suspect that everyone will eventually.
Now what evidence? And how to manage if our detectors don't happen to
follow the QM rules?
The latter is really no problem. All I am saying is that the relationship
between intensity of an individual light pulse and probability of an output
signal CANNOT ALWAYS BE LINEAR. Every known piece of measuring apparatus
has its limits! This is where QM fails, as it cannot predict what happens
when the intensity is below the limits for which the detector is "set".
Under a classical picture, we SHOULD be able to make such predictions. It
is here that we can show differences between the rival theories. If the
experimenters refrain from doing experiments except when the apparatus is
set in a particular way they are preventing themselves from finding out the
truth.
There is already evidence, in the form of Aspect's anomalies, which must
have been repeatable or he would not have mentioned them. It would not
have taken a great deal more effort to increase his replication so as to
make them statistically significant.
PS: see also my reply to Kevin Scaldeferri, which partially answers your
latest message: I DO challenge the validity of QM and think its claims to
accurate prediction are vastly exaggerated. Surely the whole world now
agrees that it is NOT the ultimate theory? Surely we all agree in our
hearts that the world is in fact local, causal and the rest of it? This
taken together with the very minor, possibly negligible, role now played by
QM in computer simulations in STM I think justifies me in calling it a
dying theory. If not dying, then doomed!
Caroline
Daryl McCullough
e...@panix.com (Ed Green) says...
>
>Patrick van Esch <van...@jamaica.desy.de> wrote:
>
>>Ilja Schmelzer (schm...@fermi.wias-berlin.de) wrote:
>>
>>: In other words, Aspect does not have a good enough detector, that's
>>: why this experiment may be explained by some local hidden variable
>>: theory, but nothing new about Bell's inequality itself and an ideal
>>: experiment.
>>Indeed, this is the way I understand it also.
>>She uses a conspiracy in a modulation of the efficiency
> ^^^^^^^^^^
>There you go again, Patrick. Your use of the word 'conspiracy' here
>is pure polemics.
I don't want to get into that, but I am reminded of a comment I read once
(I can't remember who wrote it): The scientific mind is a lot like the mind
of the paranoid---while reasonable people just shrug their shoulders when
strange things happen, the paranoid and the scientist assume that there
must be a deeper explanation!
Daryl McCullough
CoGenTex, Inc.
Ithaca, NY
Edward Green
Caroline Thompson <c.h.th...@dial.pipex.com> wrote:
>Kevin A. Scaldeferri <ke...@cco.caltech.edu> wrote in article
>> I think it's important to ask whether the intent behind Aspect type
>> experiments is to disprove local realism or to support quantum
>> mechanics. I feel it is the second (particularly given the loopholes
>> than make excluding local realism practically impossible).
Whatever you feel, and whatever the practicality of the experiment, I
find this a very odd point of view. The entire raisin de etc. of
Bell's work is to provide a possible tool for 'disproving local
realism', not to conjure up yet another experiment where where the
predictions of quantum mechanics are verified. If the intent of the
Aspect experiment were truly the latter, it was a peculiarly Rube
Golbergish waste of time.
>Agreed, but surely it ought to be NEITHER? The reasonable thing to do is
>to experiments to find out which is the better model, in terms of which
>model is best at predicting not just one isolated figure (the Bell test
>statistic) but how the results vary as you vary the parameters. As I've
>said in various papers, QM has not even GOT the necessary parameters - it
>does not, for example, have enough parameters to model real detectors (it
>has only the single "quantum effeciency"),
I wonder if this is an intrinsic feature of quantum mechanics, or
merely the formulation chosen for this setup? Just asking...
>or real coincidence windows
>(which need two parameters, not one, as they need both start and end times
>relative to a zero that represents simultaneous emissions).
>
>> Quantum
>> mechanics has proved itself spectacularly useful
Yes, it is experimentally predictive. Who said otherwise?
And what is this monolithic doctrine "quantum mechanics"? Is it like
Marxism? Liberalism? Have we degenerated to such a logically
dissolute state that we are reduced to arguing the relative merits of
conceptual hodgepodges gathered up in sacks and labeled once and for
all, like a pair of racists debating the relative villainy of 'white
people' vs. 'black people'? Have the media brought us to this? Bah.
Quantum mechanics good... Quantum mechanics bad...
>One could question this. Has its "usefulness" outweighed its disastrous
>effect on a few generations of physicists whose confidence in their own
>judgement has been destroyed?
Here I huzzah loudly. From my darkened cubbyhole I see the evidence
of this here in the learned pronouncements of the university
trained next generation of physicists. It is futile to argue with
them, because their psychic armor is nigh impenetrable. First they
have mastered some collation of mathematical techniques held together
by loose conceptual chewing gum. This gives them confidence and a
feeling of self worth. Then they have imbibed the foundation mythos
of twentieth century physics, that progress is made from conceptual
triumph to conceptual triumph, always the bad being thrown out, the
good being brought in, and all objection to anything embraced by the
self-styled conceptual advant guard is reactionary. Finally, secure
in their knowledge and sophistication, they put stopples in their ears
to reduce their already limited range of logical audition, so that no
matter what you say they only hear a vague buzz which they strive to
correct, by repeating their own thoughts. Bah.
Because we can do mathematics we are deluded into believing we may
have thought carefully about what a mathematical model is. Or that we
may have thought at all. But here I will not inveigh against a
monolithic quantum mechanics. This just plays into the
opposition's hand, and allows such irrelevancies as the general
predictive fidelity of the body of techniques thus styled to be
presented as counter arguments, as if we were arguing about two
political candidates, Jim Quantum and Bob No-Quantum; so that "Your
candidate cheated on his taxes" could be 'refuted' by replying "Yes,
but yours slept with his own niece".
I agree with you about people being educated not to trust their own
reason. It is not because reason is by and large untrustworthy,
though human reason is always fallible, but rather because most
people's ratiocinative powers are just not up to the task. But they
cover their nakedness by accepting the weirdness. Bah.
From my perspective here in wu-wu land, very few trained physicists
have a clue how to think about the role of scientific models vs. the
phenoumenon (damned philosophy stuff), whilst very few philosophers
of science have a clue about how to pare down their weighty tomes,
only read by their ilk-mates I suppose, to what might be really
useful in clarifying the issues.
I have ranted. :)
Michael Weiss
Ed Green writes, in part (by which I mean I have left out the "[snip]"s):
Bah.
Bah.
Bah.
Kevin A. Scaldeferri
In article <01bc9b3c$b89d7080$LocalHost@eozfbrwg>,
Caroline Thompson <c.h.th...@dial.pipex.com> wrote:
>
>Kevin A. Scaldeferri <ke...@cco.caltech.edu> wrote in article
>
>> This begs one to ask, would you consider Aspect's results as evidence
>> in favour of QM if you did not feel that the part of QM being tested
>> was not ridiculous?
>>
>
>I can't say. I do not pretend to be unbiased in this.
Good. I'm glad to see that at least one person arguing against QM is
willing to admit their biases. Most of them are far too busy
condemning the biases of "the establishment" to spend any time in
introspection.
> Surely this is the
>one place in science where it is only right and proper to be thoroughly
>biased? The way to progress is to investigate results that are unexplained
>or explained in a ridiculous fashion.
I had to disagree her. Ridiculous is too much of a value judgement to
be useful. Furthermore, this idea that QM is ridiculous seems to be
limited to those who don't understand it or authors of books for
laymen who's goal is something other than giving their readers an
understanding of QM.
>> Quantum
>> mechanics has proved itself spectacularly useful
>
>One could question this. Has its "usefulness" outweighed its disastrous
>effect on a few generations of physicists whose confidence in their own
>judgement has been destroyed?
Huh? My confidence in my own judgement is quite high. Moreover,
familiarity with QM tends to lead to intuition which agrees with QM.
QM has a couple surprising features, but once you understand them, an
intuition for the predictions of the theory follows fairly rapidly.
> I gather that the STM experts are drifting
>away from it.
I can't see how since scanning tunneling microscopy is a distictly
quantum mechanical phenomenom. Its practitioners ought to find their
faith in QM reinforced by it.
>If QM gets the right answer for a few atomic
>spectra, I expect it is because in this area all you need is the right
>constraints and the general idea of how to model the geometry.
QM gets the right answer to far more than just a few atomic spectra.
>I am sure that a conflict could be demonstrated in the EPR area. Aspect's
>"anomalies" (in thesis, not PRL papers), together with faith in local
>realism, show that this is so.
You cannot make this statement without intellectual dishonestly.
Remember that you are the one who claims that faith in QM is the sole
reason that people have found that EPR-type experiments verify the QM
predictions.
> QM is wrong in this area. A closer look at
>the shapes of the coincidence curves and time spectra would show
>discrepancies.
You may want to add the disclaimer here that this is just your opinion
until the experiment is done to produce that closer look.
>But I actually think it is a bad theory on all scales. It is a bad idea to
>train scientists in double-think. Wave-particles are unnecessary for
>LIGHT- let's have just waves, waves and more waves, probably ad infinitum
>but at least to a level much lower than the quantum.
Physicists don't possess this "double-think". Wave-particle duality is
just the way that a couple ideas in quantum mechanics are described to
laymen. Physicists realized long ago that neither paradigm was
adequate to describe nature and quantum mechanics was created to
overcome those shortcomings.
Jim Carr
alanpe...@worldnet.att.net writes:
>
>The HBT effect is used in radio interferometry to measure the diameter
>of nearby stars.
Hanbury-Brown - Twiss is also used to study source sizes in high
energy heavy ion collisions. I can't give you a reference off hand,
but I am pretty sure Gelbke at the Michigan State NSCL pulled it
off and that they use it as a tool.
Patrick Van Esch
Caroline Thompson wrote:
> Surely the whole world now
> agrees that it is NOT the ultimate theory? Surely we all agree in our
> hearts that the world is in fact local, causal and the rest of it? This
> taken together with the very minor, possibly negligible, role now played by
> QM in computer simulations in STM I think justifies me in calling it a
> dying theory. If not dying, then doomed!
Well, the most ultimate theory (or candidate theory) I know of,
superstrings, uses QM as one of its foundations.
The currently rather successfull elementary particle theories
that , btw have quite remarkable successes, are all based
on QM.
Closer to normal phenomena, physical chemistry and solid
state physics are all solidly based upon QM.
A lot of fairly new results like boson-einstein condensates
have been predicted and now realised in the lab.
But I guess your main problem is that the thing YOU call QM
is just a lump of misunderstandings in your mind. It would help
to learn a bit QM before claiming it is doomed. Did you actually
study it ? Would you be able to make some calculations in it ?
cheers,
Patrick.
Patrick Van Esch
> e...@panix.com (Ed Green) says...
> >
> >Patrick van Esch <van...@jamaica.desy.de> wrote:
> >
> >>Ilja Schmelzer (schm...@fermi.wias-berlin.de) wrote:
> >>
> >>: In other words, Aspect does not have a good enough detector, that's
> >>: why this experiment may be explained by some local hidden variable
> >>: theory, but nothing new about Bell's inequality itself and an ideal
> >>: experiment.
> >>Indeed, this is the way I understand it also.
> >>She uses a conspiracy in a modulation of the efficiency
> > ^^^^^^^^^^
> >There you go again, Patrick. Your use of the word 'conspiracy' here
> >is pure polemics.
No it isn't Ed. What I mean is that in any normal experiment, one
calibrates a detector by, say, sending a beam of a certain polarisation
and intensity (and wavelength) onto it, taking the counting rate,
calculating how many photons normally would have hit the detector
and hence finding the so-called efficiency of the detector.
If the detector has seen 30.000 hits while there should have been
100.000
in the calibration beam (E-radiated = N * h * nu)
then one normally assumes that the detector takes a random sample
of 30 % of the photons that hit it.
So to estimate the number of photons that actuallly hit it, one has
to divide by 0.3 the number of photons one has counted.
This is also what is assumed in all those experiments: that every
photon of
a certain kind (kind = polarisation, wavelength..., say, a calibration
beam)
has that given, and INDEPENDENT OF ANY OTHER THING IN THE EXPERIMENT,
probability of having a hit in the detector.
You must realise that if you do not make that assumption, it is
absolutely
impossible to infer anything from any experiment.
What is called a "conspiracy" is that she needs those probabilities
to be correlated in such a way that the outcome, which is of course
wrongly calculated assuming that the efficiency of the detectors are
an independent probability, seem to give exactly those results one would
have obtained if the efficiencies were indeed independent and quantum
predictions were right. Such a thing is usually called a conspiracy,
because
2 things together need to mimic one other effect. Although not
impossible,
such an explanation is usually not considered the most plausible.
It is a bit like saying that, "no, officer, I was NOT driving faster
than 70 MPH,
but due to a very strange space-time curvature, your radar *measured*
that speed."
cheers,
Patrick.
Douglas G. Danforth
Edward Green wrote:
>
> Patrick van Esch <van...@jamaica.desy.de> wrote:
>
> >Ilja Schmelzer (schm...@fermi.wias-berlin.de) wrote:
> >
> >: In other words, Aspect does not have a good enough detector, that's
> >: why this experiment may be explained by some local hidden variable
> >: theory, but nothing new about Bell's inequality itself and an ideal
> >: experiment.
> >Indeed, this is the way I understand it also.
> >She uses a conspiracy in a modulation of the efficiency
> ^^^^^^^^^^
> There you go again, Patrick. Your use of the word 'conspiracy' here
> is pure polemics.
Well, I believe it was N. David Mermin who first used the word
"conspiracy"
in the context of Bell's results.
There need be no conspiracy. Only correlation between variables
that are carried with each particle. The missing incredient is
that the state of a particle determines whether it will be
detected or not. This is a local process occuring at each
analyzer/detector.
With crossed polarizers why can one get partial transmission
(say at 20 degree orientation)? If light is quantized in
a packet then one would expect that whatever the state
of the particle when it hits the second polarizer it would
either be transmitted or absorbed AND this would happen
in the same way for EVERY photon. But it doesn't work this
way. Some pass through and some don't. This means that
the first polarizer does not produce a pure state. There
are more degrees of freedom than just the polarization.
What are these other degrees of freedom? Let us simply
call all the degrees of freedom the state of the particle.
All one needs is to have the state of one particle correlated
with the state of another particle. So if the "state" determines
the transmission or absoption of a particle it is glaringly
obvious that correlated states can lead to correlated transmission
between distant detectors.
In the same manner of thinking one can envisage correlated states
that determine which channel a particle will take and whether it
will take any channel at all (be absorbed, self interfer) that
can be correlated with a distant particle and its behavior.
The state of a particle can determine whether it is detected.
That's the bottom line.
Hitech
In article <5rini3$l...@panix2.panix.com>, Edward Green <e...@panix.com> wrote:
>Caroline Thompson <c.h.th...@dial.pipex.com> wrote:
>>...
>>One could question this. Has its "usefulness" outweighed its disastrous
>>effect on a few generations of physicists whose confidence in their own
>>judgement has been destroyed?
>
>of this here in the learned pronouncements of the university
>trained next generation of physicists. It is futile to argue with
>them, because their psychic armor is nigh impenetrable. First they
>have mastered some collation of mathematical techniques held together
>by loose conceptual chewing gum. This gives them confidence and a
>feeling of self worth. Then they have imbibed the foundation mythos
>of twentieth century physics, that progress is made from conceptual
>triumph to conceptual triumph, always the bad being thrown out, the
>good being brought in, and all objection to anything embraced by the
>self-styled conceptual advant guard is reactionary. Finally, secure
>in their knowledge and sophistication, they put stopples in their ears
>to reduce their already limited range of logical audition, so that no
>matter what you say they only hear a vague buzz which they strive to
>correct, by repeating their own thoughts. Bah.
This is as it always has been. It requires great talent to maintain and
enhance the present theories. But it is not from this pool that the
precious *creator* of the new physics theory is drawn.
Patrick Van Esch
Douglas G. Danforth wrote:
>
> With crossed polarizers why can one get partial transmission
> (say at 20 degree orientation)? If light is quantized in
> a packet then one would expect that whatever the state
> of the particle when it hits the second polarizer it would
> either be transmitted or absorbed AND this would happen
> in the same way for EVERY photon.
Well, this is the hypothesis of determinism of course.
But it doesn't work this
> way. Some pass through and some don't. This means that
> the first polarizer does not produce a pure state. There
> are more degrees of freedom than just the polarization.
> What are these other degrees of freedom? Let us simply
> call all the degrees of freedom the state of the particle.
> All one needs is to have the state of one particle correlated
> with the state of another particle. So if the "state" determines
> the transmission or absoption of a particle it is glaringly
> obvious that correlated states can lead to correlated transmission
> between distant detectors.
Yes, you are describing a classical local hidden variable theory.
This is what is under test. This is however not what is discussed.
What
is discussed is that non-perfect detectors have a certain efficiency,
say, they see 20 % of all photons that pass. One assumes that this
process is independent of the hidden variables under study and that
for every "kind" of hidden variable setting, if energy and polarisation
are the same, those 20 % remain 20 % in all cases. This is called the
calibration of the photon detector. One can buy another one,
and it will turn out to be 32 %, and if one drops it, it becomes maybe
just 0.1 %. One resonably assumes that this efficiency is not
determined
by the local variables.
You see, the problem with it is that every time a new kind of photon
detector comes onto the market, with a new efficiency, you have
to increase the number of variables in your photon. At least
if you claim that the detection efficiency at one place is correlated
with the detection efficiency at another place via hidden variables.
This is what for example Caroline Thompson claims although I don't
know if she realises that.
There is no other way that all these correlations could be described
because in principle one could set up detectors of different quality,
with different efficiencies and make the decision which one to use
only AFTER the photons are space-like separated.
So in all these experiments one assumes that there may be hidden
variables, but that the detection efficiency is not determined by it.
If one would have perfect detectors (with an efficiency of 100 %)
the problem wouldn't be as visible. Unfortunately there are no
perfect instruments in the world.
>
> In the same manner of thinking one can envisage correlated states
> that determine which channel a particle will take and whether it
> will take any channel at all (be absorbed, self interfer) that
> can be correlated with a distant particle and its behavior.
>
> The state of a particle can determine whether it is detected.
> That's the bottom line.
That is indeed the bottom line of a hidden variable theory.
Bell however proved that in this way there are inequalities on the
correlations one can obtain. These inequalities are
violated by the predictions of QM and a lot of experiments,
assuming the above, have shown QM right.
cheers,
Patrick.
KRamsay
In article <01bc9b45$d8bdd720$4cfa82c1@eozfbrwg>, "Caroline Thompson"
<c.h.th...@dial.pipex.com> writes:
|PS: see also my reply to Kevin Scaldeferri, which partially answers your
|latest message: I DO challenge the validity of QM and think its claims to
|accurate prediction are vastly exaggerated.
Well, pick an instance. Enourmous mounds of applications are available
for inspection, including ones which people have said are examples of
"amazingly accurate prediction".
|Surely the whole world now agrees that it is NOT the ultimate theory?
"Quantum mechanics" is a term which has been used to mean different
things in the past. It has been used to refer to at least one of the
individual theories of nature associated with it. But for the most
part, and especially nowadays, quantum mechanics means the framework
of quantum mechanics, within which an assortment of different theories
of nature can be formulated.
For instance, the so-called "standard model" is a theory of nature. So
are various types of string theory. There's the SU(5) "grand unified
theory" which has been investigated. These are different theories, not
necessarily making the same predictions, but they are all "quantum
mechanical theories".
I don't know of anyone who believes that a particular one of those is
an ultimate theory. (Certainly it would be careless to suppose that
any theory not incorporating gravitation is "ultimate".) On the other
hand, it's very far from true that "we all agree" that the ultimate
theory will not be a quantum-mechanical theory in the sense that these
theories are. I think it might well be; on the other hand, there's no
guarantee that there even can be such a thing as an "ultimate" theory
of physics.
|Surely we all agree in our hearts that the world is in fact local,
|causal and the rest of it?
No, absolutely not. Don't kid yourself.
One of the surest ways to get yourself stuck in a complete dead end
is to become firmly convinced that this kind of belief is "obvious".
It's hard to communicate with someone who is stuck in such a state.
|This
|taken together with the very minor, possibly negligible, role now played
by
|QM in computer simulations in STM I think justifies me in calling it a
|dying theory. If not dying, then doomed!
Are these computer simulations based on an alternative explanation of
the facts? If so, what?
There are plenty of computer simulations which are based on approximate
models, ones which are found to work for certain typical situations.
That doesn't necessarily mean that there is a competing explanation
for the phenomena. As far as I know, the space program has ever needed
to consider general relativistic effects. Using only Newtonian gravity
in their calculations doesn't mean general relativity is washed up
somehow.
Keith Ramsay There is nothing on this earth, and little beyond it,
kra...@aol.com that nobody ever denounces. -- Matt McIrvin
KRamsay
In article <01bc9b3c$b89d7080$LocalHost@eozfbrwg>, "Caroline Thompson"
<c.h.th...@dial.pipex.com> writes:
...
|> I think it's important to ask whether the intent behind Aspect type
|> experiments is to disprove local realism or to support quantum
|> mechanics. I feel it is the second (particularly given the loopholes
|> than make excluding local realism practically impossible).
|
|Agreed, but surely it ought to be NEITHER? The reasonable thing to do is
|to experiments to find out which is the better model, in terms of which
|model is best at predicting not just one isolated figure (the Bell test
|statistic) but how the results vary as you vary the parameters.
Well, what's the "other" model here? Is there a second one up and
running, or is it rather a matter of "maybe we could find another
model more to our liking, is it possible"?
|As I've
|said in various papers, QM has not even GOT the necessary parameters - it
|does not, for example, have enough parameters to model real detectors (it
|has only the single "quantum effeciency"), or real coincidence windows
|(which need two parameters, not one, as they need both start and end
times
|relative to a zero that represents simultaneous emissions).
I'm not sure I know what you think you mean by saying that quantum
mechanics "has" or "doesn't have" "parameters". It appears to me that
you've read someone's analysis of an experiment, in which they model
detectors, with some parameter, and have taken this to mean that this
is the only way to model a detector, according to quantum mechanics.
...
|> Quantum
|> mechanics has proved itself spectacularly useful
|
|One could question this. Has its "usefulness" outweighed its disastrous
|effect on a few generations of physicists whose confidence in their own
|judgement has been destroyed?
I don't find physicists' self-confidence to have been "destroyed". If
it serves to warn people against overconfidence in naive intuitions,
this is all to the good.
| I gather that the STM experts are drifting
|away from it. How can you have just a statistical theory when you can
|manipulate individual atoms?
In what sense do you suppose it to be "just" statistical? If a
"probabilistic" theory tells you that something will happen to one case
with probability 1, or very close to 1, don't you suppose this says
something about the individual case?
|If QM gets the right answer for a few atomic
|spectra, I expect it is because in this area all you need is the right
|constraints and the general idea of how to model the geometry. QM could
|well be right in this area.
This is remarkably glib. There is a lot of detail which a "general
idea of how to model the geometry" won't give you.
There is also the matter of molecular spectra. If the "geometry" is
even roughly correct, one is dealing with waves in a higher
dimensional space. It gets down to really basic issues: how do
atoms bond together at all?
...
|But I actually think it is a bad theory on all scales. It is a bad idea
to
|train scientists in double-think. Wave-particles are unnecessary for
|LIGHT- let's have just waves, waves and more waves, probably ad infinitum
|but at least to a level much lower than the quantum.
Be careful to know what kind of thinking scientists are doing, before
you critique it. It's not as though anyone has to think of light as
waves on Monday, Wednesday, and Friday, and as particles on Tuesday
and Thursday, in contradictory ways. There is merely the one consistent
way, which has features which one metaphorically describe as wave-like
and particle-like.
Edward Green
Michael Weiss <colu...@pleides.camb.opengroup.org> wrote:
Thank you for your learned and insightful reply, Michael.
I suppose next time I must substitute balderdash to insulate myself
against that kind of carping...
Edward Green
Douglas G. Danforth <danf...@csli.stanford.edu> wrote:
>There need be no conspiracy. Only correlation between variables
>that are carried with each particle. The missing incredient is
>that the state of a particle determines whether it will be
>detected or not. This is a local process occuring at each
>analyzer/detector.
That was my understanding. Thanks for confirming it in more measured
prose.
>With crossed polarizers why can one get partial transmission
>(say at 20 degree orientation)? If light is quantized in
>a packet then one would expect that whatever the state
>of the particle when it hits the second polarizer it would
>either be transmitted or absorbed AND this would happen
>in the same way for EVERY photon. But it doesn't work this
>way. Some pass through and some don't. This means that
>the first polarizer does not produce a pure state. There
>are more degrees of freedom than just the polarization.
>What are these other degrees of freedom? Let us simply
>call all the degrees of freedom the state of the particle.
>All one needs is to have the state of one particle correlated
>with the state of another particle. So if the "state" determines
>the transmission or absoption of a particle it is glaringly
>obvious that correlated states can lead to correlated transmission
>between distant detectors.
Right... I point out, and I don't mean to imply you don't know this,
that the beauty of Bell's theoretical work is that he is able to rule
out _all_ such mechanisms for some range of theoretical output. But of
course moving from this theoretical result to a laboratory test which
achieves the same logical precision is problematic, though I would not
say impossible (unless of course the world does not work that way, in
which case it would indeed be impossible to show that it did!)
>In the same manner of thinking one can envisage correlated states
>that determine which channel a particle will take and whether it
>will take any channel at all (be absorbed, self interfer) that
>can be correlated with a distant particle and its behavior.
>
>The state of a particle can determine whether it is detected.
>That's the bottom line.
Yes, I agree; and I don't see why such a possibility is
conspiratorial. Thanks again for you input.
Edward Green
Patrick Van Esch <van...@club.innet.be> wrote:
>Yes, you are describing a classical local hidden variable theory.
>This is what is under test. This is however not what is discussed.
>What
>is discussed is that non-perfect detectors have a certain efficiency,
>say, they see 20 % of all photons that pass. One assumes that this
>process is independent of the hidden variables under study and that
>for every "kind" of hidden variable setting, if energy and polarisation
>are the same, those 20 % remain 20 % in all cases. This is called the
>calibration of the photon detector.
Ok. You have at least explicitly identified your assumption. Now the
question is _why_? Some wave packet, energy, what have you,
emanates from the source and impinges on the detectors. We have some
number '20%' which is supposed to represent the average probability
such a wave packet will trigger the detector. Why is it unreasonable
that the internal information in this packet -- leaving aside all
other questions about whether it is required to assume there are
discrete packets or particles, and so forth -- can swing this
detection probability away from the average? This is not in line
with your assumptions, but is hardly some bizarre 'conspiracy', now
is it? I don't think so.
>One can buy another one,
>and it will turn out to be 32 %, and if one drops it, it becomes maybe
>just 0.1 %. One resonably assumes that this efficiency is not
>determined
>by the local variables.
I don't understand why this assumption is so transparently reasonable.
Do you know in detail what is happening during a detection event? Do
you know that this question is meaningless?
>You see, the problem with it is that every time a new kind of photon
>detector comes onto the market, with a new efficiency, you have
>to increase the number of variables in your photon.
The internal state is what it is. Assuming there is such an internal
state and that it has some influence on detection efficiency, then
each new kind of photon detector requires a new _function_ of this
internal state, but not a new internal variable. Apple pie.
>At least
>if you claim that the detection efficiency at one place is correlated
>with the detection efficiency at another place via hidden variables.
>This is what for example Caroline Thompson claims although I don't
>know if she realises that.
>There is no other way that all these correlations could be described
>because in principle one could set up detectors of different quality,
>with different efficiencies and make the decision which one to use
>only AFTER the photons are space-like separated.
I don't see the relevance of this. Sure. And would this require some
bizarre conspiracy to produce a consistent relation of internal
variables to detection probability? This requires detailed analysis.
I don't see any real argument here other than that you think there
would be some unspecified problems if we swap detectors.
>So in all these experiments one assumes that there may be hidden
>variables, but that the detection efficiency is not determined by it.
>If one would have perfect detectors (with an efficiency of 100 %)
>the problem wouldn't be as visible. Unfortunately there are no
>perfect instruments in the world.
If we are unable to set up a macroscopic arrangement for which _no_
locally realistic (to use the atrocious jargon) explanation is
possible, wouldn't you find _that_ kind of bizarrely conspiratorial?
We have a correct theory, you say, we have a competing class of
theories which we know in principle how to logically eliminate through
experiment, but through some pesky problems with detector
efficiencies we can never quite achieve a macroscopically clean
refutation. Is there a refutation-of-local-realism censorship
principle operating, or what?
I have no preconception about whether such a macroscopically clean
result could be achieved or not. I just note that if we can get
close but never quite cross the line, we may be suspicious.
>> The state of a particle can determine whether it is detected.
>> That's the bottom line.
>
>That is indeed the bottom line of a hidden variable theory.
>Bell however proved that in this way there are inequalities on the
>correlations one can obtain. These inequalities are
>violated by the predictions of QM and a lot of experiments,
>assuming the above, have shown QM right.
Patrick, this logic just doesn't hold water. You should know that.
Obviously if any sort of ordinary local causal explanation is possible
for the observed results then we haven't closed the logical loop. We
can't invoke Bell's theorem to say 'local causal explanations are
impossible', but, uh, we seem to have one here, but ignore it. The
slip comes in transferring the idealized Bell set up to actual
experiments. The remaining gap between our jump and the bar does
_not_ seem to be of the character "Well, what if our photon supplier
coached the photons before the experiment", or "What if we were on
opium, and just dreamt we did the experiments"; certainly not based
on the bald assertion that internal state variables may influence
detection probabilities. Apple pie physical.
Kevin A. Scaldeferri
In article <19970729102...@ladder01.news.aol.com>,
KRamsay <kra...@aol.com> wrote:
>In article <01bc9b45$d8bdd720$4cfa82c1@eozfbrwg>, "Caroline Thompson"
><c.h.th...@dial.pipex.com> writes:
>
>|Surely we all agree in our hearts that the world is in fact local,
>|causal and the rest of it?
>
>No, absolutely not. Don't kid yourself.
>
>One of the surest ways to get yourself stuck in a complete dead end
>is to become firmly convinced that this kind of belief is "obvious".
>It's hard to communicate with someone who is stuck in such a state.
Excellent point. After all, it was "obvious" that the universe was
parity invariant until the 50's and it was "obvious" that it was
time-reversal invariant for some time after that.
Also, am I correct that string theories are, in fact, non-local?
Kevin A. Scaldeferri
In article <33DD3A...@csli.stanford.edu>,
Douglas G. Danforth <danf...@csli.stanford.edu> wrote:
>
>With crossed polarizers why can one get partial transmission
>(say at 20 degree orientation)? If light is quantized in
>a packet then one would expect that whatever the state
>of the particle when it hits the second polarizer it would
>either be transmitted or absorbed AND this would happen
>in the same way for EVERY photon.
Maybe you would expect that. Many people would not.
>Some pass through and some don't. This means that
>the first polarizer does not produce a pure state. There
>are more degrees of freedom than just the polarization.
>What are these other degrees of freedom?
In this case, the degrees of freedom are the freedom to chose a basis
for a subspace of degenerate eigenstates. The first polarizer produces
a pure state in the choice of basis specified by that particular
orientation of the polarizer. Since this basis is not identical to
the basis specified by the second polarizer, it is not a pure state
with the second choice of basis.
Douglas G. Danforth
I agree with you. That is the standard, consistent, formulation of
quantum mechanics. It is mathematically elegant and very beautiful
(complete set of commuting operators acting on vectors in a Hilbert
space resolve the maximal infomation about the state of the system).
However, that's not the question. We are looking at the foundations
upon which that formalism is built. Since there seems to be a
need for a "collapse" of a wavefunction arising from that formalism
which is a nonlocal phenomena we need to examine very closely what
is it exactly within the formalisn that leads to this nonlocality.
Quantum mechanics is totally silent* on the issue of when a particle
will or will not be detected. GIVEN that a particle is detected
we can derive the expected values of for applied operators (at least
for those systems that we can mathematically calculate or approximately
simulate). When we have two particles that must be jointly detected
we introduce a new element into the situtation since the two
must be spacially separated. The standard statistical equation
for the estimation of correlation says nothing about how or where
the data were collected. The very nature of calculating the
correlation of spacially separated particles is a nonlocal process.
But that understanding is not enough. It is the nature of the
correlation that causes the problem as Bell has shown. However,
that analysis assumed perfect detection (as has been mentioned
many times). Since perfect detection has not been realized in
practice AND since it IS possible to construct a classical model
that gives the exact quantum cosine correlation based on
less than perfect detection it behoves one to see whether the
assumption of multiparticle wave function *normalization* may
be at the heart of the problem.
With n detectors there are 2^n possible outcomes of detection
only one of which (n chose n = 1) is described by a
multiparticle wavefunction. Note that summing over the states
of k particles in a wavefuntion of n particles does NOT give
you the same results as with k missing detections.
Multiparticle wave functions are massively biased. They are those
states for which all particles are detected. What one needs is
a formalize where one can predict the outcome of n-k particles
given that a specific k are not detected. Again this is not
the same as just ignoring the k detectors (covering them up)
for that case would include detection and non detection amongst
the k. For the "cover up" case we do have the quantum mechanical
preditions
for n-k detectors.
We need to probe below the current QM formalism, yes, even given all of
its elegance and beauty.
*we can predict absorption probabilities and scattering cross section
as well as electron emmission such as the photo electric effect,
however,
my point was these are stochastic (probabilistic) processes where we
still do not have a handle on a complete state specification so that
"this" particle will be detected (it is highly probable that we never
will have such a handle, however, the issue is one of principle).
Douglas G. Danforth
Patrick Van Esch wrote:
>
> Douglas G. Danforth wrote:
...
> >
> > In the same manner of thinking one can envisage correlated states
> > that determine which channel a particle will take and whether it
> > will take any channel at all (be absorbed, self interfer) that
> > can be correlated with a distant particle and its behavior.
> >
> > That's the bottom line.
>
> That is indeed the bottom line of a hidden variable theory.
> Bell however proved that in this way there are inequalities on the
> correlations one can obtain. These inequalities are
> violated by the predictions of QM and a lot of experiments,
> assuming the above, have shown QM right.
>
> Patrick.
Careful. The inequalities are violated by QM. The inequalities
are also violated by classical models with high efficiency (>50%).
Lots of experiments with low efficiency have agreed with the QM
predictions. Experiments with low efficiencies do NOT meet the
assumptions of Bell. Hence NO experiment to date has tested
Bell's inequality. Bell needs the assumption that A^2=B^2=1
to derive the inequality. Experiments do not meet this assumption.
Edward Green
Kevin A. Scaldeferri <ke...@cco.caltech.edu> wrote:
>In this case, the degrees of freedom are the freedom to chose a basis
>for a subspace of degenerate eigenstates. The first polarizer produces
>a pure state in the choice of basis specified by that particular
>orientation of the polarizer. Since this basis is not identical to
>the basis specified by the second polarizer, it is not a pure state
>with the second choice of basis.
I think sometimes we need to step back from the language of the
particular model, and try to understand things in more abstract
terms. I have no concrete suggestions here, but you and Doug seem
almost to be writing at cross purposes; your description is within a
particular formalism and tradition, whereas I think his may be at a
somewhat more abstract level. Since it is the very _inevitability_ or
correctness of the particular formalism and tradition that is at issue
here -- just as in my discussion of the inevitability of Lorentz
invariance -- I think it may help to step back a notch. Otherwise we
are in danger of showing that a thing must be described a certain way,
because that is in fact how we describe it!
Patrick Van Esch
Edward Green wrote:
>
> Patrick Van Esch <van...@club.innet.be> wrote:
>
> >You see, the problem with it is that every time a new kind of photon
> >detector comes onto the market, with a new efficiency, you have
> >to increase the number of variables in your photon.
>
> The internal state is what it is. Assuming there is such an internal
> state and that it has some influence on detection efficiency, then
> each new kind of photon detector requires a new _function_ of this
> internal state, but not a new internal variable. Apple pie.
No, you need a new degree of freedom. Remember that we
are talking about _deterministic_ models. So the photon has to
have a list (with a copy to its twin) : if it is a photon-detector
Zircon-823, then I pass, if it is a Smotron-275 I don't, if it is an
ouligon-9889
I pass, if it is ...
Don't forget that the copy of the entries in that list of photon #1
have to average out to 20.3 % for the Zircon-823, to 0.4 %
for the Smotron, to 31.9 % for the ouligon etc...
and that on top of that the correlations with the entries of the
photon #2 have to be such that if a naive experimenter believing
he can apply independent corrections, always gets back to those
same results QM predicts, no matter to what detector the photons
have been sent. It would be extremely remarkable that this totally
different mechanism ALWAYS results in the same results (namely
_as if_ QM was right, wrongly assuming the probabilities are
independent)
without there being any a priori cause for them.
cheers,
Patrick.
Edward Green
Paul Budnik <pa...@mtnmath.com> wrote:
>Patrick van Esch (van...@jamaica.desy.de) wrote:
>: Edward Green (e...@panix.com) wrote:
>: It means: with finite resources, we've done the
>: best we could and that is by definition absolute truth until a better
>: experiment comes along.
>
>That is a novel definition of `absolute truth'.
Please be careful in your attribution.
But counting quote depth we can see that Patrick van Esch wrote those
lines, though by attribution you make it appear I did.
Patrick Van Esch
Douglas G. Danforth wrote:
> predictions. Experiments with low efficiencies do NOT meet the
> assumptions of Bell. Hence NO experiment to date has tested
> Bell's inequality. Bell needs the assumption that A^2=B^2=1
> to derive the inequality. Experiments do not meet this assumption.
You're right in principle: if the efficiencies of the detectors are
being modulated in a very peculiar way by hidden variables
(see my post to Ed), then there is a loophole. If one makes the
reasonable assumption that a photon does not have a hidden
variable telling him how to interact with EVERY thinkable kind of
detector, well then one _can_ show a violation of the Bell inequalities.
cheers,
Patrick.
Patrick Van Esch
Yes Ed, *I* wrote those ironic lines, not you. I wanted to point out
the absurd in the "quest for absolute truth" by replacing "working
hypothesis"
(the only form of "truth" that has any meaning in science) by "absolute
truth".
(this should have been obvious with the "absolute truth until something
better
comes along" contradiction, but Paul missed it I'm affraid.)
cheers,
Patrick.
Edward Green
Tom Potter <t...@earthlink.net> brilliantly wrote:
>The "reasonableness" of an action depends upon three factors.
>
>1. The probability of success.
>
>2. The "cost" of the action.
>
>3. The "profit" if successful.
Hey... that's a great definition. I must have that engraved that on a
prosthetic grade steel plaque and have it implanted into the palm of my
hand, so that next time I contemplate doing something with a large
cost of failure and vanishing probability of success I may be possibly
be recalled to reason before acting.
>In other words, it is unreasonable to invest $1.00
>to get a profit of $1,000,000 if the odds are 10,000,000 to 1.
>
>On the other hand, it is reasonable to invest $1.00
>to get a profit of $1,000,000 if the odds are 100,000 to 1.
Yeah. We can extend the paradigm to cover the costs and benefits of a
range of probable outcomes, and even throw in a non-linear utility
function, but I think your precis is enough to remind us.
>The problem with institutionalized scientists
Oh yeah, back to science...
>is that they are conditioned to only consider
>high success ratios reasonable, and they rarely
>factor in the potential profit ratio.
>
>They are willing to spend billions of dollars
>with little hope of profit, if the odds of success are high,
>and are not willing to pursue projects with little hope
>of success, no matter what the potential profit would be.
>(Because it would make them look stupid if they failed. )
And probably a sound survival strategy for all sorts of
institutionalized individuals.
>This is why astrologers invented astronomy,
>and alchemists invented chemistry.
Would you care to explicate that?
Patrick Van Esch
Douglas G. Danforth wrote:
> Quantum mechanics is totally silent* on the issue of when a particle
> will or will not be detected.
Well, it is not. QM gives you the probabilities, and, moreover, tells
you
that those probabilities are fundamental and there is NO way what
so ever that they are the result of an underlying mechanism.
This is called a non-deterministic theory, and people don't feel always
happy with it because we've that mental image that we "understand"
something when we see the levers and gears in the box.
QM tells you that there aren't any such levers and gears.
"things just happen".
Now you might say that that is ridiculous, but it is the very nature
of a fundamentally non-deterministic theory.
If there are underlying levers and gears, then QM cannot be right.
This is in essence Bell's theorem. (as a theorem, not as a proposal
for an experiment)
It might of course be that nature doesn't follow QM after all,
but given the *experimental* results we have no reason to believe
so. If one day it turns out that there ARE levers and gears, well then
QM is not _explained_ by it, but is simply wrong.
> less than perfect detection it behoves one to see whether the
> assumption of multiparticle wave function *normalization* may
> be at the heart of the problem.
You are right, better experiments would be nice. But for the moment
we DON'T HAVE perfect detectors. And, btw, how would you know
that your apparently perfect detector doesn't get a hidden signal not
to be perfect anymore when dealing with correlated pairs ?
>
> With n detectors there are 2^n possible outcomes of detection
> only one of which (n chose n = 1) is described by a
> multiparticle wavefunction.
I don't follow you here. What do you mean ? A multi - particle
wave function gives you the amplitudes for all the 2^n cases...
[...]
>
> Multiparticle wave functions are massively biased. They are those
> states for which all particles are detected.
I have absolutely no idea what you're saying.
What one needs is
> a formalize where one can predict the outcome of n-k particles
> given that a specific k are not detected.
well, a multi-particle wave function does EXACTLY that ....
> We need to probe below the current QM formalism, yes, even given all of
> its elegance and beauty.
You cannot "probe below it". If you find levers and gears, it is
broken,
and if it isn't broken, there AIN'T no levers and gears.
But you cannot have QM, and a lowel - level levers and gears
explanation.
>
> *we can predict absorption probabilities and scattering cross section
> as well as electron emmission such as the photo electric effect,
> however,
> my point was these are stochastic (probabilistic) processes where we
> still do not have a handle on a complete state specification so that
> "this" particle will be detected (it is highly probable that we never
> will have such a handle, however, the issue is one of principle).
You are stating here that nature has to be deterministic.
If QM is right, it cannot be. Even if some modification of QM is
due one day, as long as you keep the superposition principle
(the cornerstone) it cannot be deterministic.
cheers,
Patrick.
Patrick Van Esch
Edward Green wrote:
> Patrick, this logic just doesn't hold water. You should know that.
>
> Obviously if any sort of ordinary local causal explanation is possible
> for the observed results then we haven't closed the logical loop. We
> can't invoke Bell's theorem to say 'local causal explanations are
> impossible', but, uh, we seem to have one here, but ignore it.
We don't ignore it. We have an extra assumption, without which
you cannot do ANY experiment. That is, until shown otherwise, we
assume that our detectors can be calibrated using the usual
procedures. I've pointed this out previously, if you don't make that
assumption, but if you start saying that your measurement apparatus
can be influenced by just any unknown effect, you cannot do any
experiment anymore. One cannot consider the class of theories
"in which my measurements are influenced by something unknown".
The
> slip comes in transferring the idealized Bell set up to actual
> experiments. The remaining gap between our jump and the bar does
> _not_ seem to be of the character "Well, what if our photon supplier
> coached the photons before the experiment", or "What if we were on
> opium, and just dreamt we did the experiments"; certainly not based
> on the bald assertion that internal state variables may influence
> detection probabilities. Apple pie physical.
The problem is that if you don't specify a CLASS of possible internal
state variables that can influence detection probabiliities, ANYTHING
can happen in the lab.
It could be, for example, that godalmighty poured so many
extra internal variables into every little particle in the universe a
loooong
time ago in which he "programmed" the whole universe. Those particles
have LONG LONG lists of hidden variables and know exactly what to do
when. When they are destroyed, they pass there lists to new particles.
Such a theory can NEVER be experimentally refuted. It can give just
any answer. The only way out is to assume that this is NOT the case.
One has never observed the efficiency of a photon detector to
depend (usually weakly) on anything else than
wavelength and polarisation. And now
suddenly it should depend on a god-known variable, and moreover,
in such a specific way that if you didn't know this, and did wrongly
a correction with a constant efficiency, you got the QM predictions ?!
And this for different classes of detectors ? Opium, you said ?
Conspiracy ?
cheers,
Patrick.
Edward Green
Ray Tomes <rto...@kcbbs.gen.nz> wrote:
> e...@panix.com (Edward Green) wrote:
>
>>Yes, you are quite right about Bell's logic, since after all the
>>entire scheme is based on sets of discrete independent events (the
>>assumed pair emission, followed by detections or nulls).
>
>I think that it is telling that Caroline Thompson is a statistician and that I
>used to be one. Have you ever been a statistician Ed? I wonder if David Elm
>is? (He must be a card player at least!)
Well, I was a TA for a statistician for several semesters once; is that
close enough? :)
Seriously, I think I imbibed several habits of thought which are very
important and illuminating, and whose inculcation seems to be missing
from a standard physics education. I'm not saying this is a
substitute for all other forms of physical understanding (despite his
amazing annoying quality, van Esch really does know a lot more
physics than I do by most measures), but their absence seems to be
absolutely fatal in carrying on some kinds of discussion, and may
account for the feeling sometimes of reasoning carefully with a brick
wall. Sometimes I think getting outside of the daily language of the
thing, even harboring a certain controlled ignorance about the
details, lets you see some details of the landscape more clearly.
Within a profession, this role is usually accorded to 'senior men',
who despite spending a career in one area almost cannot help by power
of mere accumulated experience getting a big picture. What is hard
for some people to admit is that an outsider could at any point be in
possession of any part of such a picture, or even could conceivably
know how to think -- strange notion -- even if thinking leads Patrick
(we _are_ ganging up on him now, aren't we :) with perhaps
understandable scorn 'an armchair philosopher'. Well yes, I am a
philosopher, and armchairs are most comfortable; though I prefer a
couch for really serious work. :)
I wouldn't tell him how to run an experiment; but on the other hand
his belief that dependence of detection probabilities on some inner
state of the detected is 'conspiratorial' mainly shows -- and I have
been rather nasty about this, I admit -- being wedded to some model to
such an extent it is literally impossible to see outside it! It can't
be like that because in his working model it is not like that, and he
uses his working model everyday, and what do I know about calibrating
detectors anyway?
As I mentioned, lawyers and computer scientists seem to go through a
similar career phase.
>Ray Tomes wrote:
>>>I refer to the Hanbury-Brown and Twiss (HBT) experiments from mid 1950s
[snip]
>>I'll pass on that one, Ray. Just in the spirit of argument, I note
>>that explanations could be imagined which were consistent with
>>traveling photons...
>
>They have been imagined, but only after the event. They now claim that ALL
>emission events happen as poisson distributions of photons (i.e an average of 1
>but potentially any number)! That is ludicrous. For a start what happens if
>you only have one atom? How does it emit a poisson distribution of photons when
>it has only 1 unit of energy to give up?
Again, I am unqualified to comment, reluctantly. You know I share
your skepticism at many things, even when this may label me a
non-standard thinker, but I am simply uncertain how far quantum
optics goes to ironing out the conceptual difficulties -- or at least
avoiding them -- that spring from a simple How and Why book picture of
photons. I would fight other battles; this will fall out in the
wash, at least I think so.
Jim Carr
xk...@dial.pipex.com (Caroline Thompson) writes:
>
> This particular experiment would have failed badly, I say, if
>Aspect had not had high emission rates and hence an excuse to do his
>"adjustment", subtracting what he, on the basis of QT, decided were
>"accidentals"!
The problem of correcting for accidental coincidences is a real
one, but avoiding it lowers the rate at which one can do the
experiment. However, I believe that a _real_ implementation of
Caroline's realistic model would also have an accidental rate
if the data rate gets close to the sampling window acceptance.
>I'm just getting involved, by the way, in a completely new approach to
>EPR: can't we argue that QT should never had put forward their simple
>formula? The argument is that they are saying their formula models
>*interference* between the two sides.
I would say they describe a correlation between them.
The 'realistic' model puts that correlation in there also, but makes
it deterministic from moment to moment. That is the difference. I
am not quite sure how to put the 2-body correlation function into
the statistical language you prefer, but it is not unique to
configuration space. It is there in any Hilbert space.
>I can't
>remember if I actually said anything in my Chaotic Ball paper in the end,
>but I meant to draw attention to the fact the "coincidence patterns" and
>"interference patterns" are two very different things, taking place on
>very different scales.
That is a good point, and a valid distinction IMO.
ale2
In article <33DE69...@club.innet.be>
Patrick Van Esch <van...@club.innet.be> writes:
QM gives you the probabilities, and, moreover, tells
> you
> that those probabilities are fundamental and there is NO way what
> so ever that they are the result of an underlying mechanism.
This must have been proven somewhere, any references?
Patrick Van Esch
Bell's theorem ! Ok, my statement was too wide, sorry.
I forgot to add: locally realistic underlying mechanism.
cheers,
Patrick.