Two EPR questions [Archive]

Blake Winter

Sep27-04, 03:31 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nFirst, I\'m curious how one can explain the EPR paradox in relativistic\nQFT, since one sets (as I understand it, anyway) [q1(x1),q2(x2)]=0\nwhen x1 and x2 are spacelike related points and q1 and q2 are\narbitrary operators. This seems to destroy the possibility of having\nEPR states. Why doesn\'t it?\n\nSecondly, in considering Bell\'s proof that hidden variable theories\nhave to be nonlocal, isn\'t there a loophole in the sense that no\nmatter how one constructs the experiment (even if one uses quantum\nprocesses to choose the polarization measurement orientations) there\nwould be no way to really get "random" orientations since the theory\nis, in fact, deterministic? That is, in some way the initial\nconditions which set up the whole experiment would determine what\nmeasurements would be made as well, thereby getting rid of the need\nfor nonlocal effects. Not that I think this is a particularly\nattractive loophole, since it raises all sorts of questions about how\nthe information would be stored about which measurement would be made,\nbut I\'m still curious if its technically a loophole.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>First, I'm curious how one can explain the EPR paradox in relativistic
QFT, since one sets (as I understand it, anyway) [q1(x1),q2(x2)]=0
when x1 and x2 are spacelike related points and q1 and q2 are
arbitrary operators. This seems to destroy the possibility of having
EPR states. Why doesn't it?

Secondly, in considering Bell's proof that hidden variable theories
have to be nonlocal, isn't there a loophole in the sense that no
matter how one constructs the experiment (even if one uses quantum
processes to choose the polarization measurement orientations) there
would be no way to really get "random" orientations since the theory
is, in fact, deterministic? That is, in some way the initial
conditions which set up the whole experiment would determine what
measurements would be made as well, thereby getting rid of the need
for nonlocal effects. Not that I think this is a particularly
attractive loophole, since it raises all sorts of questions about how
the information would be stored about which measurement would be made,
but I'm still curious if its technically a loophole.

Ralph Hartley

Sep27-04, 10:20 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\nBlake Winter wrote:\n\n> Secondly, in considering Bell\'s proof that hidden variable theories\n> have to be nonlocal, isn\'t there a loophole in the sense that no\n> matter how one constructs the experiment (even if one uses quantum\n> processes to choose the polarization measurement orientations) there\n> would be no way to really get "random" orientations since the theory\n> is, in fact, deterministic? That is, in some way the initial\n> conditions which set up the whole experiment would determine what\n> measurements would be made as well, thereby getting rid of the need\n> for nonlocal effects. Not that I think this is a particularly\n> attractive loophole, since it raises all sorts of questions about how\n> the information would be stored about which measurement would be made,\n> but I\'m still curious if its technically a loophole.\n\nYes, it is, but it is a very small loophole. I\'ve seen it used by people\nwho obviously had no conception of how *very* small it is.\n\nWhat makes it so small is that Bell\'s theorem is violated for *all*\nmechanisms for deciding which measurements to make (at least all that have\nbeen tried).\n\nTo take advantage of that loophole would require "nature" do do something\nequivalent to reverse engineering the apparatus to figure out how the\ndecision depends on quantum events (if at all). The *exact* workings, to\nthe last logic gate out of millions, of your pseudo random number generator\nwould have to be taken into account, just to decide on the state of one photon.\n\nIt would require "The Lord" to be more than subtle, more than malicious,\nbut deliberately and fiendishly deceptive.\n\nIf that were so, locality would be the *least* of our problems.\n\nIt\'s like the possibility that the world was created in seven days, and\nonly a week ago, complete with fossils and our memories. You can\'t rule it\nout, but at that level you can\'t rule *anything* out.\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Blake Winter wrote:

> Secondly, in considering Bell's proof that hidden variable theories
> have to be nonlocal, isn't there a loophole in the sense that no
> matter how one constructs the experiment (even if one uses quantum
> processes to choose the polarization measurement orientations) there
> would be no way to really get "random" orientations since the theory
> is, in fact, deterministic? That is, in some way the initial
> conditions which set up the whole experiment would determine what
> measurements would be made as well, thereby getting rid of the need
> for nonlocal effects. Not that I think this is a particularly
> attractive loophole, since it raises all sorts of questions about how
> the information would be stored about which measurement would be made,
> but I'm still curious if its technically a loophole.

Yes, it is, but it is a very small loophole. I've seen it used by people
who obviously had no conception of how *very* small it is.

What makes it so small is that Bell's theorem is violated for *all*
mechanisms for deciding which measurements to make (at least all that have
been tried).

To take advantage of that loophole would require "nature" do do something
equivalent to reverse engineering the apparatus to figure out how the
decision depends on quantum events (if at all). The *exact* workings, to
the last logic gate out of millions, of your pseudo random number generator
would have to be taken into account, just to decide on the state of one photon.

It would require "The Lord" to be more than subtle, more than malicious,
but deliberately and fiendishly deceptive.

If that were so, locality would be the *least* of our problems.

It's like the possibility that the world was created in seven days, and
only a week ago, complete with fossils and our memories. You can't rule it
out, but at that level you can't rule *anything* out.

Ralph Hartley

Thomas Trotter

Sep28-04, 11:50 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>blake.winter@houghton.edu (Blake Winter) wrote in message news:<87423d2a.0409261455.69146e11@posting.google. com>...\n> First, I\'m curious how one can explain the EPR paradox in relativistic\n> QFT, since one sets (as I understand it, anyway) [q1(x1),q2(x2)]=0\n> when x1 and x2 are spacelike related points and q1 and q2 are\n> arbitrary operators. This seems to destroy the possibility of having\n> EPR states. Why doesn\'t it?\n\nMy understanding is that paired detection attributes in, say,\noptical Bell experiments are related insofar as they correspond\nto paired photons produced via the same atomic emission(s). If so,\nthen because emission-paired photons *incident on* the analyzers\nare polarized identically (via conservation of angular momentum),\nthen you get nonlinear data graphs when you plot the rates of\ncoincidental detection (identical detection attributes) for the\nvarious angular differences (theta) of analyzer settings.\n\n>\n> Secondly, in considering Bell\'s proof that hidden variable theories\n> have to be nonlocal, isn\'t there a loophole in the sense that no\n> matter how one constructs the experiment (even if one uses quantum\n> processes to choose the polarization measurement orientations) there\n> would be no way to really get "random" orientations since the theory\n> is, in fact, deterministic?\n\nVarying the analyzer settings (randomly, pseudorandomly, or however)\nwhile the photons are in flight shouldn\'t matter anyway -- because,\nfor any given pair of detection attributes there\'s one and only one\ncorresponding theta.\n\n> That is, in some way the initial conditions which set up the whole\n> experiment would determine what measurements would be made as well,\n> thereby getting rid of the need for nonlocal effects.\n\nThere\'s no need for \'nonlocal effects\' anyway, imo. Bell showed that\nif you model the biparticle setup in a certain way, then you\'ll get\nexpectation values which are incompatible with qm for some theta.\nAnd, if you believe the subsequent experiments, then Bell\'s model is\nincompatible with reality also. You *can* tweak the model so that\nit\'s compatible with qm by allowing spacelike separated events to\ninstantaneously interact -- but this is an artificial and misleading\nway of representing the experimental/observational context and why/how\nthe measurement results are related.\n\nSo, what\'s wrong with the model? Well, for one thing, it assumes\nthat rate of coincidental detection is a linear function of theta.\nBut, qm and experiments say otherwise.\n\nIt\'s not the combination of the polarization of photon 1 wrt polarizer\nA and the polarization of photon 2 wrt polarizer B that\'s being\nmeasured in the combined context. Rather, it\'s the polarization\nof photon 1 wrt the polarization of photon 2 wrt theta that\'s being\nanalyzed. Now, if you assume, as the emission model does, that\nphoton 1 and photon 2 of any given pair of polarizer-incident\nphotons are polarized identically, then how would you formulate\nthe experimental context? Why does the probability of coincidental\ndetection then vary as a nonlinear function of theta?\n\n> Not that I think this is a particularly attractive loophole, since\n> it raises all sorts of questions about how the information would be\n> stored about which measurement would be made, but I\'m still curious\n> if its technically a loophole.\n\nImo, this particular consideration wrt Bell tests isn\'t important,\nthat is, no, it\'s not technically a loophole wrt my understanding\nof the meaning of Bell\'s theorem and violations of Bell inequalities.\nIn other words, if paired detection attributes correspond to\nemission-paired photons (and experimenters do take great care\nto facilitate this), then Bell inequalities will be experimentally\nviolated because they\'re based on a model which incorrectly\nrepresents the experimental/observational context. My interpretation\nof the qm formulation is that the spatially separated events\ndon\'t need to be instantaneously interacting in order to\nunderstand nonlinear curves of rate of coincidental detection wrt\ntheta. (see the Law of Malus)\n\nHowever, the consideration of experimental loopholes in general is\nvery valuable insofar as it produces refinements/improvements in\nexperimental techniques.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>blake.winter@houghton.edu (Blake Winter) wrote in message news:<87423d2a.0409261455.69146e11@posting.google.com>...
> First, I'm curious how one can explain the EPR paradox in relativistic
> QFT, since one sets (as I understand it, anyway) [q1(x1),q2(x2)]=0
> when x1 and x2 are spacelike related points and q1 and q2 are
> arbitrary operators. This seems to destroy the possibility of having
> EPR states. Why doesn't it?

My understanding is that paired detection attributes in, say,
optical Bell experiments are related insofar as they correspond
to paired photons produced via the same atomic emission(s). If so,
then because emission-paired photons *incident on* the analyzers
are polarized identically (via conservation of angular momentum),
then you get nonlinear data graphs when you plot the rates of
coincidental detection (identical detection attributes) for the
various angular differences (\theta) of analyzer settings.

>
> Secondly, in considering Bell's proof that hidden variable theories
> have to be nonlocal, isn't there a loophole in the sense that no
> matter how one constructs the experiment (even if one uses quantum
> processes to choose the polarization measurement orientations) there
> would be no way to really get "random" orientations since the theory
> is, in fact, deterministic?

Varying the analyzer settings (randomly, pseudorandomly, or however)
while the photons are in flight shouldn't matter anyway -- because,
for any given pair of detection attributes there's one and only one
corresponding \theta.

> That is, in some way the initial conditions which set up the whole
> experiment would determine what measurements would be made as well,
> thereby getting rid of the need for nonlocal effects.

There's no need for 'nonlocal effects' anyway, imo. Bell showed that
if you model the biparticle setup in a certain way, then you'll get
expectation values which are incompatible with qm for some \theta.
And, if you believe the subsequent experiments, then Bell's model is
incompatible with reality also. You *can* tweak the model so that
it's compatible with qm by allowing spacelike separated events to
instantaneously interact -- but this is an artificial and misleading
way of representing the experimental/observational context and why/how
the measurement results are related.

So, what's wrong with the model? Well, for one thing, it assumes
that rate of coincidental detection is a linear function of \theta.
But, qm and experiments say otherwise.

It's not the combination of the polarization of photon 1 wrt polarizer
A and the polarization of photon 2 wrt polarizer B that's being
measured in the combined context. Rather, it's the polarization
of photon 1 wrt the polarization of photon 2 wrt \theta that's being
analyzed. Now, if you assume, as the emission model does, that
photon 1 and photon 2 of any given pair of polarizer-incident
photons are polarized identically, then how would you formulate
the experimental context? Why does the probability of coincidental
detection then vary as a nonlinear function of \theta?

> Not that I think this is a particularly attractive loophole, since
> it raises all sorts of questions about how the information would be
> stored about which measurement would be made, but I'm still curious
> if its technically a loophole.

Imo, this particular consideration wrt Bell tests isn't important,
that is, no, it's not technically a loophole wrt my understanding
of the meaning of Bell's theorem and violations of Bell inequalities.
In other words, if paired detection attributes correspond to
emission-paired photons (and experimenters do take great care
to facilitate this), then Bell inequalities will be experimentally
violated because they're based on a model which incorrectly
represents the experimental/observational context. My interpretation
of the qm formulation is that the spatially separated events
don't need to be instantaneously interacting in order to
understand nonlinear curves of rate of coincidental detection wrt
\theta. (see the Law of Malus)

However, the consideration of experimental loopholes in general is
very valuable insofar as it produces refinements/improvements in
experimental techniques.

Ilja Schmelzer

Oct1-04, 04:22 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb\n> Blake Winter wrote:\n> > Secondly, in considering Bell\'s proof that hidden variable theories\n> > have to be nonlocal, isn\'t there a loophole in the sense that no\n> > matter how one constructs the experiment (even if one uses quantum\n> > processes to choose the polarization measurement orientations) there\n> > would be no way to really get "random" orientations since the theory\n> > is, in fact, deterministic? That is, in some way the initial\n> > conditions which set up the whole experiment would determine what\n> > measurements would be made as well, thereby getting rid of the need\n> > for nonlocal effects. Not that I think this is a particularly\n> > attractive loophole, since it raises all sorts of questions about how\n> > the information would be stored about which measurement would be made,\n> > but I\'m still curious if its technically a loophole.\n>\n> Yes, it is, but it is a very small loophole. I\'ve seen it used by people\n> who obviously had no conception of how *very* small it is.\n>\n> What makes it so small is that Bell\'s theorem is violated for *all*\n> mechanisms for deciding which measurements to make (at least all that have\n> been tried).\n\nI have proposed another argument which works against this loophole:\n\nImagine there is a preferred frame and we learn in some future how to\ncommunicate FTL. You have a working FTL phone - two black boxes\nand some channel between them which allows you to talk with somebody\non Mars without any time delay. Surely, this FTL phone falsifies\nrelativity.\n\nNow imagine some old crank who, despite the obvious falsification\nof relativity, wants to save relativity and presents some arguments which\nsuggest that relativity is not violated.\nYou don\'t really want to be as stupid as that crank? In this case,\nit should be clear that there is something wrong with his arguments.\nThus, any sort of argument which allows to save relativity if there\nexists an FTL phone should be wrong.\n\nNow, the point is that the loophole can be used by that crank too.\nIf there is no free will of the experimenter, you can do whatever\nyou like with your FTL phone but you will be unable to prove\nthat there exists some FTL effects inside the phone.\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb
> Blake Winter wrote:
> > Secondly, in considering Bell's proof that hidden variable theories
> > have to be nonlocal, isn't there a loophole in the sense that no
> > matter how one constructs the experiment (even if one uses quantum
> > processes to choose the polarization measurement orientations) there
> > would be no way to really get "random" orientations since the theory
> > is, in fact, deterministic? That is, in some way the initial
> > conditions which set up the whole experiment would determine what
> > measurements would be made as well, thereby getting rid of the need
> > for nonlocal effects. Not that I think this is a particularly
> > attractive loophole, since it raises all sorts of questions about how
> > the information would be stored about which measurement would be made,
> > but I'm still curious if its technically a loophole.
>
> Yes, it is, but it is a very small loophole. I've seen it used by people
> who obviously had no conception of how *very* small it is.
>
> What makes it so small is that Bell's theorem is violated for *all*
> mechanisms for deciding which measurements to make (at least all that have
> been tried).

I have proposed another argument which works against this loophole:

Imagine there is a preferred frame and we learn in some future how to
communicate FTL. You have a working FTL phone - two black boxes
and some channel between them which allows you to talk with somebody
on Mars without any time delay. Surely, this FTL phone falsifies
relativity.

Now imagine some old crank who, despite the obvious falsification
of relativity, wants to save relativity and presents some arguments which
suggest that relativity is not violated.
You don't really want to be as stupid as that crank? In this case,
it should be clear that there is something wrong with his arguments.
Thus, any sort of argument which allows to save relativity if there
exists an FTL phone should be wrong.

Now, the point is that the loophole can be used by that crank too.
If there is no free will of the experimenter, you can do whatever
you like with your FTL phone but you will be unable to prove
that there exists some FTL effects inside the phone.

Ilja

Ralph Hartley

Oct1-04, 09:43 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nIlja Schmelzer wrote:\n> I have proposed another argument which works against this loophole\n> Now, the point is that the loophole can be used by that crank too.\n\nI think I have seen your argument before, but didn\'t read it carefully\nenough to understand it. It may have been part of a thread I was starting\nto tune out.\n\n> If there is no free will of the experimenter, you can do whatever\n> you like with your FTL phone but you will be unable to prove\n> that there exists some FTL effects inside the phone.\n\nThe problem with that argument is that it uses something like a triple\nnegative and a very unlikely hypothetical, which while logically correct,\nis more than you can expect the reader to follow. Especially a reader who\nis already struggling with the concepts, as most of those who need the\nargument are.\n\nA simpler way to put it would be that the loophole could explain away *any*\nnonlocal effect whatsoever, regardless of any conceivable evidence.\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Ilja Schmelzer wrote:
> I have proposed another argument which works against this loophole
> Now, the point is that the loophole can be used by that crank too.

I think I have seen your argument before, but didn't read it carefully
enough to understand it. It may have been part of a thread I was starting
to tune out.

> If there is no free will of the experimenter, you can do whatever
> you like with your FTL phone but you will be unable to prove
> that there exists some FTL effects inside the phone.

The problem with that argument is that it uses something like a triple
negative and a very unlikely hypothetical, which while logically correct,
is more than you can expect the reader to follow. Especially a reader who
is already struggling with the concepts, as most of those who need the
argument are.

A simpler way to put it would be that the loophole could explain away *any*
nonlocal effect whatsoever, regardless of any conceivable evidence.

Ralph Hartley

Blake Winter

Oct3-04, 03:49 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nIt\'s certainly a rather ridiculous loophole, and I didn\'t mean to\nsuggest it as being a likely loophole. I was just curious if it was\nan actual loophole.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>It's certainly a rather ridiculous loophole, and I didn't mean to
suggest it as being a likely loophole. I was just curious if it was
an actual loophole.

rof@maths.tcd.ie

Oct5-04, 06:42 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n>Imagine there is a preferred frame and we learn in some future how to\n>communicate FTL. You have a working FTL phone - two black boxes\n>and some channel between them which allows you to talk with somebody\n>on Mars without any time delay. Surely, this FTL phone falsifies\n>relativity.\n\n>Now imagine some old crank who, despite the obvious falsification\n>of relativity, wants to save relativity and presents some arguments which\n>suggest that relativity is not violated.\n>You don\'t really want to be as stupid as that crank? In this case,\n>it should be clear that there is something wrong with his arguments.\n>Thus, any sort of argument which allows to save relativity if there\n>exists an FTL phone should be wrong.\n\n>Now, the point is that the loophole can be used by that crank too.\n>If there is no free will of the experimenter, you can do whatever\n>you like with your FTL phone but you will be unable to prove\n>that there exists some FTL effects inside the phone.\n\nI should mention that the usual utterance produced by those who\nargue against taking Bell\'s inequalities seriously, namely the\ndogmatic assertion that "it is not meaningful" to consider the\nresults of unperformed experiments, such as what would have happened\nhad the magnetic field of the apparatus been aligned differently,\nis also subject to your objection.\n\nThis is because anybody can always say that what person B would\nhave heard through the FTL telephone had person A said something\ndifferent is "undefined", or "not meaningful" or (if they want\nto summon up some pomposity) "not physically meaningful." The\nvalidity of the assertion that the telephone is a device which\ncommunicates what A says (whatever that might be) to B can then not\nbe established; only the fact that B heard what A said in this\ninstance can be established.\n\nThe rejection of so-called counterfactuals is a prescription\nfor never drawing any inferences from observed data, no matter\nhow consistently and reproducibly one event is found to\nfollow another. One would have thought that such a bizarre\nattitude would be restricted to skeptical philosophers, whose\ninsistence on rigour is so great that any claims to knowledge\nof an external world are denied. And yet the people who\nassert that counterfactuals are contraband (the majority\nof physicists who have managed to get this far in the debate) will\nat the same time express their disdain for philosophy.\n\n(It is, of course, useful for the factory-owner to instill in his\nworkers a certain amount of contempt for those who think compared\nto those who simply work without thinking. I wonder how many\nphysicists acquired their professed contempt of philosophy\nfrom other physicists, rather than having the idea on their own.)\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

>Imagine there is a preferred frame and we learn in some future how to
>communicate FTL. You have a working FTL phone - two black boxes
>and some channel between them which allows you to talk with somebody
>on Mars without any time delay. Surely, this FTL phone falsifies
>relativity.

>Now imagine some old crank who, despite the obvious falsification
>of relativity, wants to save relativity and presents some arguments which
>suggest that relativity is not violated.
>You don't really want to be as stupid as that crank? In this case,
>it should be clear that there is something wrong with his arguments.
>Thus, any sort of argument which allows to save relativity if there
>exists an FTL phone should be wrong.

>Now, the point is that the loophole can be used by that crank too.
>If there is no free will of the experimenter, you can do whatever
>you like with your FTL phone but you will be unable to prove
>that there exists some FTL effects inside the phone.

I should mention that the usual utterance produced by those who
argue against taking Bell's inequalities seriously, namely the
dogmatic assertion that "it is not meaningful" to consider the
results of unperformed experiments, such as what would have happened
had the magnetic field of the apparatus been aligned differently,
is also subject to your objection.

This is because anybody can always say that what person B would
have heard through the FTL telephone had person A said something
different is "undefined", or "not meaningful" or (if they want
to summon up some pomposity) "not physically meaningful." The
validity of the assertion that the telephone is a device which
communicates what A says (whatever that might be) to B can then not
be established; only the fact that B heard what A said in this
instance can be established.

The rejection of so-called counterfactuals is a prescription
for never drawing any inferences from observed data, no matter
how consistently and reproducibly one event is found to
follow another. One would have thought that such a bizarre
attitude would be restricted to skeptical philosophers, whose
insistence on rigour is so great that any claims to knowledge
of an external world are denied. And yet the people who
assert that counterfactuals are contraband (the majority
of physicists who have managed to get this far in the debate) will
at the same time express their disdain for philosophy.

(It is, of course, useful for the factory-owner to instill in his
workers a certain amount of contempt for those who think compared
to those who simply work without thinking. I wonder how many
physicists acquired their professed contempt of philosophy
from other physicists, rather than having the idea on their own.)

R.

Ilja Schmelzer

Oct7-04, 07:04 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb\n> Ilja Schmelzer wrote:\n> > If there is no free will of the experimenter, you can do whatever\n> > you like with your FTL phone but you will be unable to prove\n> > that there exists some FTL effects inside the phone.\n\n> The problem with that argument is that it uses something like a triple\n> negative and a very unlikely hypothetical, which while logically correct,\n> is more than you can expect the reader to follow. Especially a reader who\n> is already struggling with the concepts, as most of those who need the\n> argument are.\n\nHm. At the current moment, it is the majority of the scientific community\nwhich does not accept my interpretation - that the violation of Bell\'s\ninequality is an experimental falsification of Einstein causality, as good\nas\nany other falsification. IOW, whose "who need it" are the majority of\nscientists. I would not expect that this majority is unable to follow\na logical argument of this type.\n\n> A simpler way to put it would be that the loophole could explain\n> away *any* nonlocal effect whatsoever, regardless of any\n> conceivable evidence.\n\nHm. Let\'s remember why I prefer the FTL variant. The argument is\ndirected against Einstein causality - my claim is that the violation of\nBE is a falsification of Einstein causality. Now, if an EPRB loophole\nis sufficient to explain away an FTL phone, Einstein causality becomes\nobviously unfalsifiable and is, therefore, no longer part of empirical\nscience.\n\nNot sure what is simpler.\n\nIlja\n\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb
> Ilja Schmelzer wrote:
> > If there is no free will of the experimenter, you can do whatever
> > you like with your FTL phone but you will be unable to prove
> > that there exists some FTL effects inside the phone.

> The problem with that argument is that it uses something like a triple
> negative and a very unlikely hypothetical, which while logically correct,
> is more than you can expect the reader to follow. Especially a reader who
> is already struggling with the concepts, as most of those who need the
> argument are.

Hm. At the current moment, it is the majority of the scientific community
which does not accept my interpretation - that the violation of Bell's
inequality is an experimental falsification of Einstein causality, as good
as
any other falsification. IOW, whose "who need it" are the majority of
scientists. I would not expect that this majority is unable to follow
a logical argument of this type.

> A simpler way to put it would be that the loophole could explain
> away *any* nonlocal effect whatsoever, regardless of any
> conceivable evidence.

Hm. Let's remember why I prefer the FTL variant. The argument is
directed against Einstein causality - my claim is that the violation of
BE is a falsification of Einstein causality. Now, if an EPRB loophole
is sufficient to explain away an FTL phone, Einstein causality becomes
obviously unfalsifiable and is, therefore, no longer part of empirical
science.

Not sure what is simpler.

Ilja

Ilja Schmelzer

Oct7-04, 07:04 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n>\n> >Imagine there is a preferred frame and we learn in some future how to\n> >communicate FTL. You have a working FTL phone - two black boxes\n> >and some channel between them which allows you to talk with somebody\n> >on Mars without any time delay. Surely, this FTL phone falsifies\n> >relativity.\n>\n> >Now imagine some old crank who, despite the obvious falsification\n> >of relativity, wants to save relativity and presents some arguments which\n> >suggest that relativity is not violated.\n> >You don\'t really want to be as stupid as that crank? In this case,\n> >it should be clear that there is something wrong with his arguments.\n> >Thus, any sort of argument which allows to save relativity if there\n> >exists an FTL phone should be wrong.\n\n> I should mention that the usual utterance produced by those who\n> argue against taking Bell\'s inequalities seriously, namely the\n> dogmatic assertion that "it is not meaningful" to consider the\n> results of unperformed experiments, such as what would have happened\n> had the magnetic field of the apparatus been aligned differently,\n> is also subject to your objection.\n\nThank you for making that point.\n\nMy FTL phone argument is, indeed, directed against a whole\nclass of arguments around the Bell inequalities.\n\nThe only arguments I know which fail the test are arguments\nbased on the impossibility to apply the effects for information\ntransfer, and that the inequalities allow two realistic interpretations:\nAn information transfer A->B or an information transfer B->A.\n\nThese two are related. Once we have some effect which can be\nexplained in two ways - as A->B or B->A - then it cannot be\nused for information transfer: A transfer A->B would be in contra-\ndiction with the explanation B->A.\n\nWhat remains is, therefore, a rejection of indirect observation.\nObserving "A->B or B->A" is not accepted, only "A->B" pure\nwould be accepted. Such a rejection of indirect observation\nwould be obviously a very serious restriction of the scientific\nmethod if applied everywhere.\n\nIlja\n\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
>
> >Imagine there is a preferred frame and we learn in some future how to
> >communicate FTL. You have a working FTL phone - two black boxes
> >and some channel between them which allows you to talk with somebody
> >on Mars without any time delay. Surely, this FTL phone falsifies
> >relativity.
>
> >Now imagine some old crank who, despite the obvious falsification
> >of relativity, wants to save relativity and presents some arguments which
> >suggest that relativity is not violated.
> >You don't really want to be as stupid as that crank? In this case,
> >it should be clear that there is something wrong with his arguments.
> >Thus, any sort of argument which allows to save relativity if there
> >exists an FTL phone should be wrong.

> I should mention that the usual utterance produced by those who
> argue against taking Bell's inequalities seriously, namely the
> dogmatic assertion that "it is not meaningful" to consider the
> results of unperformed experiments, such as what would have happened
> had the magnetic field of the apparatus been aligned differently,
> is also subject to your objection.

Thank you for making that point.

My FTL phone argument is, indeed, directed against a whole
class of arguments around the Bell inequalities.

The only arguments I know which fail the test are arguments
based on the impossibility to apply the effects for information
transfer, and that the inequalities allow two realistic interpretations:
An information transfer A->B or an information transfer B->A.

These two are related. Once we have some effect which can be
explained in two ways - as A->B or B->A - then it cannot be
used for information transfer: A transfer A->B would be in contra-
diction with the explanation B->A.

What remains is, therefore, a rejection of indirect observation.
Observing "A->B or B->A" is not accepted, only "A->B" pure
would be accepted. Such a rejection of indirect observation
would be obviously a very serious restriction of the scientific
method if applied everywhere.

Ilja

Blake Winter

Oct8-04, 06:20 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\nScientists don\'t accept the Bell inequality violations as a proof\nagainst relativity because one can\'t communicate information using the\nwave function "collapse". All relativity says is taht information\ncan\'t be communicated faster than c. Also note that one cannot\nexperimentally tell whether the measurement at A or B was made first -\nthe results are the same either way so there\'s no meaning to ordering\nthem in time.\nFurthermore its possible to show that the Bell inequalities are also\nresult from time symmetry. That is, its possible to understand them\nin a similar way as the Wheeler-Feynman theory of radiation\nresistance. Under such time symmetry one would expect to get Bell\ninequalities even from a deterministic classical theory, which leaves\nquantum physics ok.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Scientists don't accept the Bell inequality violations as a proof
against relativity because one can't communicate information using the
wave function "collapse". All relativity says is taht information
can't be communicated faster than c. Also note that one cannot
experimentally tell whether the measurement at A or B was made first -
the results are the same either way so there's no meaning to ordering
them in time.
Furthermore its possible to show that the Bell inequalities are also
result from time symmetry. That is, its possible to understand them
in a similar way as the Wheeler-Feynman theory of radiation
resistance. Under such time symmetry one would expect to get Bell
inequalities even from a deterministic classical theory, which leaves
quantum physics ok.

rof@maths.tcd.ie

Oct8-04, 06:20 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n\n>> I should mention that the usual utterance produced by those who\n>> argue against taking Bell\'s inequalities seriously, namely the\n>> dogmatic assertion that "it is not meaningful" to consider the\n>> results of unperformed experiments, such as what would have happened\n>> had the magnetic field of the apparatus been aligned differently,\n>> is also subject to your objection.\n\n>Thank you for making that point.\n\n>My FTL phone argument is, indeed, directed against a whole\n>class of arguments around the Bell inequalities.\n\n>The only arguments I know which fail the test are arguments\n>based on the impossibility to apply the effects for information\n>transfer, and that the inequalities allow two realistic interpretations:\n>An information transfer A->B or an information transfer B->A.\n\n>These two are related. Once we have some effect which can be\n>explained in two ways - as A->B or B->A - then it cannot be\n>used for information transfer: A transfer A->B would be in contra-\n>diction with the explanation B->A.\n\nThis reaction to the violation of Bell\'s inequalities is even more\ncommon than the objection to counterfactuals. It\'s still not really\na genuine counterargument, though, since nobody ever claimed that\nthe phenomenon could be used to transmit information. It\'s more\nof an attitude than an argument - "You can\'t use this effect to send\ninformation, so it\'s not worth thinking about."\n\nIf one does think about it a little, though, it quickly becomes\napparent that something very strange is indeed going on - there is\na conflict between Einstein causality and the observed experimental\nfacts (modulo the very small probability that, as experiments improve,\nthe observed effect will disappear). It means, essentially, that\nthe view of space-time as a structure that constrains which events\ncan affect which other events is incorrect. Space-time can still\nhowever be considered as a structure which constrains information\nflows (since information, rather than influence, is what can not\nleap across unlimited distances of space to affect the observations\nof distant observers).\n\nThe natural question for your ether theory is whether it strictly\nforbids information from travelling faster than light, since it\nis evidently comfortable with superluminal influences.\n\n>What remains is, therefore, a rejection of indirect observation.\n>Observing "A->B or B->A" is not accepted, only "A->B" pure\n>would be accepted. Such a rejection of indirect observation\n>would be obviously a very serious restriction of the scientific\n>method if applied everywhere.\n\nI don\'t believe I\'ve ever heard anybody express that position\nas their own; that is, a refusal to accept inferences of statements\nin the form "X or Y" from data which suggests it. Obviously\nsomebody who rejected statements like that would be severely\ncrippled in everyday life, rejecting the inference that x or\ny must be zero from the observation that xy is zero, for example.\nOn the other hand, the requirement that one\'s position make sense,\nor not lead to ridiculous consequences, doesn\'t appear to be one\nwhich many physicists make of themselves when dealing with quantum\nmechanics.\n\nRegards,\nR.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb

>> I should mention that the usual utterance produced by those who
>> argue against taking Bell's inequalities seriously, namely the
>> dogmatic assertion that "it is not meaningful" to consider the
>> results of unperformed experiments, such as what would have happened
>> had the magnetic field of the apparatus been aligned differently,
>> is also subject to your objection.

>Thank you for making that point.

>My FTL phone argument is, indeed, directed against a whole
>class of arguments around the Bell inequalities.

>The only arguments I know which fail the test are arguments
>based on the impossibility to apply the effects for information
>transfer, and that the inequalities allow two realistic interpretations:
>An information transfer A->B or an information transfer B->A.

>These two are related. Once we have some effect which can be
>explained in two ways - as A->B or B->A - then it cannot be
>used for information transfer: A transfer A->B would be in contra-
>diction with the explanation B->A.

This reaction to the violation of Bell's inequalities is even more
common than the objection to counterfactuals. It's still not really
a genuine counterargument, though, since nobody ever claimed that
the phenomenon could be used to transmit information. It's more
of an attitude than an argument - "You can't use this effect to send
information, so it's not worth thinking about."

If one does think about it a little, though, it quickly becomes
apparent that something very strange is indeed going on - there is
a conflict between Einstein causality and the observed experimental
facts (modulo the very small probability that, as experiments improve,
the observed effect will disappear). It means, essentially, that
the view of space-time as a structure that constrains which events
can affect which other events is incorrect. Space-time can still
however be considered as a structure which constrains information
flows (since information, rather than influence, is what can not
leap across unlimited distances of space to affect the observations
of distant observers).

The natural question for your ether theory is whether it strictly
forbids information from travelling faster than light, since it
is evidently comfortable with superluminal influences.

>What remains is, therefore, a rejection of indirect observation.
>Observing "A->B or B->A" is not accepted, only "A->B" pure
>would be accepted. Such a rejection of indirect observation
>would be obviously a very serious restriction of the scientific
>method if applied everywhere.

I don't believe I've ever heard anybody express that position
as their own; that is, a refusal to accept inferences of statements
in the form "X or Y" from data which suggests it. Obviously
somebody who rejected statements like that would be severely
crippled in everyday life, rejecting the inference that x or
y must be zero from the observation that xy is zero, for example.
On the other hand, the requirement that one's position make sense,
or not lead to ridiculous consequences, doesn't appear to be one
which many physicists make of themselves when dealing with quantum
mechanics.

Regards,
R.

Thomas Trotter

Oct8-04, 06:20 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<ck2ofa\\$1i4\\$1@beech.fernuni-hagen.de>...\n> <rof@maths.tcd.ie> schrieb\n> > "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> >\n> > >Imagine there is a preferred frame and we learn in some future how to\n> > >communicate FTL. You have a working FTL phone - two black boxes\n> > >and some channel between them which allows you to talk with somebody\n> > >on Mars without any time delay. Surely, this FTL phone falsifies\n> > >relativity.\n>\n> > >Now imagine some old crank who, despite the obvious falsification\n> > >of relativity, wants to save relativity and presents some arguments which\n> > >suggest that relativity is not violated.\n> > >You don\'t really want to be as stupid as that crank? In this case,\n> > >it should be clear that there is something wrong with his arguments.\n> > >Thus, any sort of argument which allows to save relativity if there\n> > >exists an FTL phone should be wrong.\n>\n> > I should mention that the usual utterance produced by those who\n> > argue against taking Bell\'s inequalities seriously, namely the\n> > dogmatic assertion that "it is not meaningful" to consider the\n> > results of unperformed experiments, such as what would have happened\n> > had the magnetic field of the apparatus been aligned differently,\n> > is also subject to your objection.\n>\n> Thank you for making that point.\n>\n> My FTL phone argument is, indeed, directed against a whole\n> class of arguments around the Bell inequalities.\n>\n> The only arguments I know which fail the test are arguments\n> based on the impossibility to apply the effects for information\n> transfer, and that the inequalities allow two realistic interpretations:\n> An information transfer A->B or an information transfer B->A.\n>\n> These two are related. Once we have some effect which can be\n> explained in two ways - as A->B or B->A - then it cannot be\n> used for information transfer: A transfer A->B would be in contra-\n> diction with the explanation B->A.\n>\n> What remains is, therefore, a rejection of indirect observation.\n> Observing "A->B or B->A" is not accepted, only "A->B" pure\n> would be accepted. Such a rejection of indirect observation\n> would be obviously a very serious restriction of the scientific\n> method if applied everywhere.\n\nIn optical Bell tests, in a given coincidence\nwindow, the relationship of the polarizations\n(wrt each other) of two photons from the same\noscillator are being analyzed wrt some theta.\n\nThe relationship between qm predictions,the\nobserved correlation curves (coincidence rates),\nand Malus\' Law seems clear enough.\n\nThus, considerations of an instantaneous effect,\nA<->B, would seem unnecessary -- and BI has\nnothing to do with the principle of locality.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<ck2ofa$1i4$1@beech.fernuni-hagen.de>...
> <rof@maths.tcd.ie> schrieb
> > "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> >
> > >Imagine there is a preferred frame and we learn in some future how to
> > >communicate FTL. You have a working FTL phone - two black boxes
> > >and some channel between them which allows you to talk with somebody
> > >on Mars without any time delay. Surely, this FTL phone falsifies
> > >relativity.
>
> > >Now imagine some old crank who, despite the obvious falsification
> > >of relativity, wants to save relativity and presents some arguments which
> > >suggest that relativity is not violated.
> > >You don't really want to be as stupid as that crank? In this case,
> > >it should be clear that there is something wrong with his arguments.
> > >Thus, any sort of argument which allows to save relativity if there
> > >exists an FTL phone should be wrong.
>
> > I should mention that the usual utterance produced by those who
> > argue against taking Bell's inequalities seriously, namely the
> > dogmatic assertion that "it is not meaningful" to consider the
> > results of unperformed experiments, such as what would have happened
> > had the magnetic field of the apparatus been aligned differently,
> > is also subject to your objection.
>
> Thank you for making that point.
>
> My FTL phone argument is, indeed, directed against a whole
> class of arguments around the Bell inequalities.
>
> The only arguments I know which fail the test are arguments
> based on the impossibility to apply the effects for information
> transfer, and that the inequalities allow two realistic interpretations:
> An information transfer A->B or an information transfer B->A.
>
> These two are related. Once we have some effect which can be
> explained in two ways - as A->B or B->A - then it cannot be
> used for information transfer: A transfer A->B would be in contra-
> diction with the explanation B->A.
>
> What remains is, therefore, a rejection of indirect observation.
> Observing "A->B or B->A" is not accepted, only "A->B" pure
> would be accepted. Such a rejection of indirect observation
> would be obviously a very serious restriction of the scientific
> method if applied everywhere.

In optical Bell tests, in a given coincidence
window, the relationship of the polarizations
(wrt each other) of two photons from the same
oscillator are being analyzed wrt some \theta.

The relationship between qm predictions,the
observed correlation curves (coincidence rates),
and Malus' Law seems clear enough.

Thus, considerations of an instantaneous effect,
A<->B, would seem unnecessary -- and BI has
nothing to do with the principle of locality.

Patrick Van Esch

Oct11-04, 03:51 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\nblake.winter@houghton.edu (Blake Winter) wrote in message news:<87423d2a.0409261455.69146e11@posting.google. com>...\n> Secondly, in considering Bell\'s proof that hidden variable theories\n> have to be nonlocal, isn\'t there a loophole in the sense that no\n> matter how one constructs the experiment (even if one uses quantum\n> processes to choose the polarization measurement orientations) there\n> would be no way to really get "random" orientations since the theory\n> is, in fact, deterministic?\n\nI think a slightly more serious loophole for the moment is given by\nthe inefficiency of the photon detectors, in that if the hidden\nvariables are allowed to change the detection efficiency per impact\n(while keeping the average efficiency) you have a loophole until you\nhave an average quantum efficiency, I think, of about 87%.\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>blake.winter@houghton.edu (Blake Winter) wrote in message news:<87423d2a.0409261455.69146e11@posting.google.com>...
> Secondly, in considering Bell's proof that hidden variable theories
> have to be nonlocal, isn't there a loophole in the sense that no
> matter how one constructs the experiment (even if one uses quantum
> processes to choose the polarization measurement orientations) there
> would be no way to really get "random" orientations since the theory
> is, in fact, deterministic?

I think a slightly more serious loophole for the moment is given by
the inefficiency of the photon detectors, in that if the hidden
variables are allowed to change the detection efficiency per impact
(while keeping the average efficiency) you have a loophole until you
have an average quantum efficiency, I think, of about 87%.

cheers,
Patrick.

Joe Rongen

Oct11-04, 03:52 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Blake Winter" <blake.winter@houghton.edu> wrote in message\nnews:87423d2a.0410071557.6cbf3170@posting .google.com...\n>\n> Scientists don\'t accept the Bell inequality violations as a proof\n> against relativity because one can\'t communicate information using\n> the wave function "collapse".\n\nActually, that was precisely how Gisin\nperformed some of his experiments.\n\nFrom: ENtanglement, Amir D. Aczel,2002,\nISBN 1-55192-549-4 Page 238:\n\n"Gisin believed that while entanglement doesn\'t allow us to send\nreadable messages faster than light, the phenomenon still violates\nthe \'spirit\' of special relativity. He thus wanted to test the entanglement\nphenomenon within a relativistic framework. In one of his experiments,\nGisin used an absorbing black surface, placed at the ends of the optical\nfiber, to collapse the wave function.\n[.....]\nThis complex experiment using moving reference frames resulted in a\nstrong confirmation of nonlocal entanglement and the prediction of QM."\n\nRegards Joe\n\n\n"We won in Iraq. Now we have our very own West Bank."\n\n\n---\nOutgoing mail is certified Virus Free.\nChecked by AVG anti-virus system (http://www.grisoft.com).\nVersion: 6.0.774 / Virus Database: 521 - Release Date: 10/7/04\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Blake Winter" <blake.winter@houghton.edu> wrote in message
news:87423d2a.0410071557.6cbf3170@posting.google.c om...
>
> Scientists don't accept the Bell inequality violations as a proof
> against relativity because one can't communicate information using
> the wave function "collapse".

Actually, that was precisely how Gisin
performed some of his experiments.

From: ENtanglement, Amir D. Aczel,2002,
ISBN 1-55192-549-4 Page 238:

"Gisin believed that while entanglement doesn't allow us to send
readable messages faster than light, the phenomenon still violates
the 'spirit' of special relativity. He thus wanted to test the entanglement
phenomenon within a relativistic framework. In one of his experiments,
Gisin used an absorbing black surface, placed at the ends of the optical
fiber, to collapse the wave function.
[.....]
This complex experiment using moving reference frames resulted in a
strong confirmation of nonlocal entanglement and the prediction of QM."

Regards Joe

"We won in Iraq. Now we have our very own West Bank."

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6..774 / Virus Database: 521 - Release Date: 10/7/04

Ilja Schmelzer

Oct11-04, 03:53 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> >These two are related. Once we have some effect which can be\n> >explained in two ways - as A->B or B->A - then it cannot be\n> >used for information transfer: A transfer A->B would be in contra-\n> >diction with the explanation B->A.\n>\n> This reaction to the violation of Bell\'s inequalities is even more\n> common than the objection to counterfactuals. It\'s still not really\n> a genuine counterargument, though, since nobody ever claimed that\n> the phenomenon could be used to transmit information. It\'s more\n> of an attitude than an argument - "You can\'t use this effect to send\n> information, so it\'s not worth thinking about."\n\nYep.\n\n> If one does think about it a little, though, it quickly becomes\n> apparent that something very strange is indeed going on - there is\n> a conflict between Einstein causality and the observed experimental\n> facts (modulo the very small probability that, as experiments improve,\n> the observed effect will disappear).\n\nYep. Except that I don\'t see a reason to name this "strange".\nThat\'s what happens all the time in science - theories are\nfalsified by experiments. Moreover, we have alternative\ntheories which are not falsified (Bohmian mechanics).\nWhat\'s strange?\n\nThe only strange thing is the reaction of the scientific\ncommunity which prefers to reject realism instead of\naccepting the facts. Unfortunately, as you say:\n\n> On the other hand, the requirement that one\'s position make sense,\n> or not lead to ridiculous consequences, doesn\'t appear to be one\n> which many physicists make of themselves when dealing with quantum\n> mechanics.\n------------------------------------\n> It means, essentially, that\n> the view of space-time as a structure that constrains which events\n> can affect which other events is incorrect. Space-time can still\n> however be considered as a structure which constrains information\n> flows (since information, rather than influence, is what can not\n> leap across unlimited distances of space to affect the observations\n> of distant observers).\n>\n> The natural question for your ether theory is whether it strictly\n> forbids information from travelling faster than light, since it\n> is evidently comfortable with superluminal influences.\n\nIn the case of my classical ether theory I have proven the EEP.\nSee gr-qc/0205035. Once the theory is classical, causal and\ndeterministic, it follows from the EEP that the light cone restricts\ninformation transfer.\n\nIn the pure quantum case we have no longer a classical light\ncone which allows to restrict the flow of information. Therefore\nin full quantum theory we should not expect such a restriction.\n\nThe open question is, therefore, one about the semiclassical\nlimit where density, velocity and pressure of the ether are\nconsidered as classical fields but the other degrees of freedom\nas quantum fields. I don\'t think this will be a serious problem.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> >These two are related. Once we have some effect which can be
> >explained in two ways - as A->B or B->A - then it cannot be
> >used for information transfer: A transfer A->B would be in contra-
> >diction with the explanation B->A.
>
> This reaction to the violation of Bell's inequalities is even more
> common than the objection to counterfactuals. It's still not really
> a genuine counterargument, though, since nobody ever claimed that
> the phenomenon could be used to transmit information. It's more
> of an attitude than an argument - "You can't use this effect to send
> information, so it's not worth thinking about."

Yep.

> If one does think about it a little, though, it quickly becomes
> apparent that something very strange is indeed going on - there is
> a conflict between Einstein causality and the observed experimental
> facts (modulo the very small probability that, as experiments improve,
> the observed effect will disappear).

Yep. Except that I don't see a reason to name this "strange".
That's what happens all the time in science - theories are
falsified by experiments. Moreover, we have alternative
theories which are not falsified (Bohmian mechanics).
What's strange?

The only strange thing is the reaction of the scientific
community which prefers to reject realism instead of
accepting the facts. Unfortunately, as you say:

> On the other hand, the requirement that one's position make sense,
> or not lead to ridiculous consequences, doesn't appear to be one
> which many physicists make of themselves when dealing with quantum
> mechanics.
------------------------------------
> It means, essentially, that
> the view of space-time as a structure that constrains which events
> can affect which other events is incorrect. Space-time can still
> however be considered as a structure which constrains information
> flows (since information, rather than influence, is what can not
> leap across unlimited distances of space to affect the observations
> of distant observers).
>
> The natural question for your ether theory is whether it strictly
> forbids information from travelling faster than light, since it
> is evidently comfortable with superluminal influences.

In the case of my classical ether theory I have proven the EEP.
See http://www.arxiv.org/abs/gr-qc/0205035. Once the theory is classical, causal and
deterministic, it follows from the EEP that the light cone restricts
information transfer.

In the pure quantum case we have no longer a classical light
cone which allows to restrict the flow of information. Therefore
in full quantum theory we should not expect such a restriction.

The open question is, therefore, one about the semiclassical
limit where density, velocity and pressure of the ether are
considered as classical fields but the other degrees of freedom
as quantum fields. I don't think this will be a serious problem.

Ilja

Ilja Schmelzer

Oct11-04, 03:53 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>NNTP-Posting-Host: lfa222122.richmond.edu\nX-Trace: rumor.richmond.edu 1097485444 8872 141.166.222.122 (11 Oct 2004 09:04:04 GMT)\nX-Complaints-To: usenet@rumor.richmond.edu\nNNTP-Posting-Date: Mon, 11 Oct 2004 09:04:04 +0000 (UTC)\nXref: core-easynews sci.physics.research:59806\n\n\n\n"Blake Winter" <blake.winter@houghton.edu> schrieb\n> Scientists don\'t accept the Bell inequality violations as a proof\n> against relativity because one can\'t communicate information using the\n> wave function "collapse".\n\nThis is something I have a problem with. I provide arguments against\nthis position, but these arguments are simply ignored.\n\n> All relativity says is that information\n> can\'t be communicated faster than c.\n\nThat is a positivistic reduction of relativity.\nIt should be distinguished from a fundamental claim about reality.\n\nThe problem with such reductions is that people like to switch\nbetween different versions. If confronted with the violation of\nBell\'s inequality, they move to the positivistic version.\n\nThe positivistic version is compatible with Bohmian mechanics,\nwhich has a hidden preferred frame. But relativists like to reject\nBohmian mechanics because it has a preferred frame, using the\nother interpretation of relativity.\n\n> Also note that one cannot\n> experimentally tell whether the measurement at A or B was made first -\n> the results are the same either way so there\'s no meaning to ordering\n> them in time.\n\nI have already spend some postings in this thread "noting" this difference.\nWe observe "A->B or B->A". Above explanations violate Einstein\ncausality. If you would like to reject indirect observation as a way of\nfalsification, you should throw away a lot of science. Interesting what\nremains at all.\n\n> Furthermore its possible to show that the Bell inequalities are also\n> result from time symmetry. That is, its possible to understand them\n> in a similar way as the Wheeler-Feynman theory of radiation\n> resistance. Under such time symmetry one would expect to get Bell\n> inequalities even from a deterministic classical theory, which leaves\n> quantum physics ok.\n\nOf course you can give up causality at all, allow causal influences\nbackward in time in some time-symmetric way. That\'s another nice\nexample of an "explanation" which allows to explain away even\na working FLT phone (see my previous postings in this thread).\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>NNTP-Posting-Host: lfa222122.richmond.edu
X-Trace: rumor.richmond.edu 1097485444 8872 141.166.222.122 (11 Oct 2004 09:04:04 GMT)
X-Complaints-To: usenet@rumor.richmond.edu
NNTP-Posting-Date: Mon, 11 Oct 2004 09:04:04 +0000 (UTC)
Xref: core-easynews sci.physics.research:59806

"Blake Winter" <blake.winter@houghton.edu> schrieb
> Scientists don't accept the Bell inequality violations as a proof
> against relativity because one can't communicate information using the
> wave function "collapse".

This is something I have a problem with. I provide arguments against
this position, but these arguments are simply ignored.

> All relativity says is that information
> can't be communicated faster than c.

That is a positivistic reduction of relativity.
It should be distinguished from a fundamental claim about reality.

The problem with such reductions is that people like to switch
between different versions. If confronted with the violation of
Bell's inequality, they move to the positivistic version.

The positivistic version is compatible with Bohmian mechanics,
which has a hidden preferred frame. But relativists like to reject
Bohmian mechanics because it has a preferred frame, using the
other interpretation of relativity.

> Also note that one cannot
> experimentally tell whether the measurement at A or B was made first -
> the results are the same either way so there's no meaning to ordering
> them in time.

I have already spend some postings in this thread "noting" this difference.
We observe "A->B or B->A". Above explanations violate Einstein
causality. If you would like to reject indirect observation as a way of
falsification, you should throw away a lot of science. Interesting what
remains at all.

> Furthermore its possible to show that the Bell inequalities are also
> result from time symmetry. That is, its possible to understand them
> in a similar way as the Wheeler-Feynman theory of radiation
> resistance. Under such time symmetry one would expect to get Bell
> inequalities even from a deterministic classical theory, which leaves
> quantum physics ok.

Of course you can give up causality at all, allow causal influences
backward in time in some time-symmetric way. That's another nice
example of an "explanation" which allows to explain away even
a working FLT phone (see my previous postings in this thread).

Ilja

Ilja Schmelzer

Oct11-04, 04:02 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Thomas Trotter" <thomastrotter2005@juno.com> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> > My FTL phone argument is, indeed, directed against a whole\n> > class of arguments around the Bell inequalities.\n> >\n> > The only arguments I know which fail the test are arguments\n> > based on the impossibility to apply the effects for information\n> > transfer, and that the inequalities allow two realistic interpretations:\n> > An information transfer A->B or an information transfer B->A.\n> >\n> > These two are related. Once we have some effect which can be\n> > explained in two ways - as A->B or B->A - then it cannot be\n> > used for information transfer: A transfer A->B would be in contra-\n> > diction with the explanation B->A.\n> >\n> > What remains is, therefore, a rejection of indirect observation.\n> > Observing "A->B or B->A" is not accepted, only "A->B" pure\n> > would be accepted. Such a rejection of indirect observation\n> > would be obviously a very serious restriction of the scientific\n> > method if applied everywhere.\n>\n> In optical Bell tests, in a given coincidence\n> window, the relationship of the polarizations\n> (wrt each other) of two photons from the same\n> oscillator are being analyzed wrt some theta.\n>\n> The relationship between qm predictions,the\n> observed correlation curves (coincidence rates),\n> and Malus\' Law seems clear enough.\n>\n> Thus, considerations of an instantaneous effect,\n> A<->B, would seem unnecessary -- and BI has\n> nothing to do with the principle of locality.\n\nI don\'t understand your point.\n\nIn case you refer to known loopholes of actual tests\n(like detector efficiency): That\'s nothing I care about.\nBut that does not seem to be your point. It sounds more\nlike "we can compute the observed probabilities with QM,\nso what?"\n\nThe point is that QM is not a realistic theory, it does not\ndescribe what really happens, it allows to compute only\nsome probabilities without explaining them. But there is\nno necessity to reject classical realism, realistic theories\nwhich are compatible with observation exist\n(Bohmian mechanics).\n\nOf course, as long as BM has not been known, it was\na reasonable decision to "shut up and calculate" without\nthinking too much about the foundations: There were a\nlot of things to compute with QM, a lot of things which\ncould be done in this way. But today?\n\nIt is also possible to understand that Einstein did not like\nBM. Last not least it requires a preferred frame. But\nEinstein has not known Bell\'s inequality and Aspect\'s\nexperiment. But today?\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Thomas Trotter" <thomastrotter2005@juno.com> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> > My FTL phone argument is, indeed, directed against a whole
> > class of arguments around the Bell inequalities.
> >
> > The only arguments I know which fail the test are arguments
> > based on the impossibility to apply the effects for information
> > transfer, and that the inequalities allow two realistic interpretations:
> > An information transfer A->B or an information transfer B->A.
> >
> > These two are related. Once we have some effect which can be
> > explained in two ways - as A->B or B->A - then it cannot be
> > used for information transfer: A transfer A->B would be in contra-
> > diction with the explanation B->A.
> >
> > What remains is, therefore, a rejection of indirect observation.
> > Observing "A->B or B->A" is not accepted, only "A->B" pure
> > would be accepted. Such a rejection of indirect observation
> > would be obviously a very serious restriction of the scientific
> > method if applied everywhere.
>
> In optical Bell tests, in a given coincidence
> window, the relationship of the polarizations
> (wrt each other) of two photons from the same
> oscillator are being analyzed wrt some \theta.
>
> The relationship between qm predictions,the
> observed correlation curves (coincidence rates),
> and Malus' Law seems clear enough.
>
> Thus, considerations of an instantaneous effect,
> A<->B, would seem unnecessary -- and BI has
> nothing to do with the principle of locality.

I don't understand your point.

In case you refer to known loopholes of actual tests
(like detector efficiency): That's nothing I care about.
But that does not seem to be your point. It sounds more
like "we can compute the observed probabilities with QM,
so what?"

The point is that QM is not a realistic theory, it does not
describe what really happens, it allows to compute only
some probabilities without explaining them. But there is
no necessity to reject classical realism, realistic theories
which are compatible with observation exist
(Bohmian mechanics).

Of course, as long as BM has not been known, it was
a reasonable decision to "shut up and calculate" without
thinking too much about the foundations: There were a
lot of things to compute with QM, a lot of things which
could be done in this way. But today?

It is also possible to understand that Einstein did not like
BM. Last not least it requires a preferred frame. But
Einstein has not known Bell's inequality and Aspect's
experiment. But today?

Ilja

scerir

Oct11-04, 01:07 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\nBlake Winter:\n> ... in considering Bell\'s proof that hidden variable theories\n> have to be nonlocal, isn\'t there a loophole in the sense that no\n> matter how one constructs the experiment (even if one uses quantum\n> processes to choose the polarization measurement orientations) there\n> would be no way to really get "random" orientations since the theory\n> is, in fact, deterministic?\n\nNot sure to understand which theory is "deterministic".\nSince orthodox QM is not "deterministic" and Bell\ndid not assume "determinism" but only "local causality".\nAnyway I (perhaps) understood your point. And Bell\n(and then Zurek and Peres) made a similar point.\nIt is called super-determinism.\n\n"It has been argued that quantum mechanics\nis not locally causal and cannot be embedded\nin a locally causal theory. That conclusion\ndepends on treating certain experimental parameters,\ntypically the orientations of polarization filters,\nas free variables. But it might be that this apparent\nfreedom is illusory. Perhaps experimental parameters\nand experimental results are both consequences,\nor partially so, of some common hidden mechanism.\nThen the apparent non-locality could be simulated."\n- John Bell, "Free Variables and Local Causality",\n\'Epistemological Letters\', 15, (1977)\n\nRegards,\ns.\n\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Blake Winter:
> ... in considering Bell's proof that hidden variable theories
> have to be nonlocal, isn't there a loophole in the sense that no
> matter how one constructs the experiment (even if one uses quantum
> processes to choose the polarization measurement orientations) there
> would be no way to really get "random" orientations since the theory
> is, in fact, deterministic?

Not sure to understand which theory is "deterministic".
Since orthodox QM is not "deterministic" and Bell
did not assume "determinism" but only "local causality".
Anyway I (perhaps) understood your point. And Bell
(and then Zurek and Peres) made a similar point.
It is called super-determinism.

"It has been argued that quantum mechanics
is not locally causal and cannot be embedded
in a locally causal theory. That conclusion
depends on treating certain experimental parameters,
typically the orientations of polarization filters,
as free variables. But it might be that this apparent
freedom is illusory. Perhaps experimental parameters
and experimental results are both consequences,
or partially so, of some common hidden mechanism.
Then the apparent non-locality could be simulated."
- John Bell, "Free Variables and Local Causality",
'Epistemological Letters', 15, (1977)

Regards,
s.

rof@maths.tcd.ie

Oct12-04, 10:49 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n\n>> It\'s more\n>> of an attitude than an argument - "You can\'t use this effect to send\n>> information, so it\'s not worth thinking about."\n\n>Yep.\n\n>> If one does think about it a little, though, it quickly becomes\n>> apparent that something very strange is indeed going on - there is\n>> a conflict between Einstein causality and the observed experimental\n>> facts (modulo the very small probability that, as experiments improve,\n>> the observed effect will disappear).\n\n>Yep. Except that I don\'t see a reason to name this "strange".\n>That\'s what happens all the time in science - theories are\n>falsified by experiments. Moreover, we have alternative\n>theories which are not falsified (Bohmian mechanics).\n>What\'s strange?\n\nFor me, nonlocality itself is rather odd. If the world is nonlocal,\nthen how does locality emerge? It\'s not satisfying to merely say\nthat we can\'t use it to send signals, where what is really meant\nis that it can\'t be used to *deliberately* send signals; signals\ncan be sent, but only accidentally. If one wants a realistic theory,\nthe appearance of such a human notion as deliberate intention in\nsuch an important role is distateful, to say the least.\n\nOn the other hand, you\'re right - the realist model of special\nrelativity, with space-time as the stage in which local things\nplay out their local acts, has been falsified (notwithstanding\nloopholes and arguments which fail your FTL telephone test).\n\nSo what do we do? The choices are to abandon realism, or abandon\nspecial relativity. But the arguments in favour of relativity\nare very strong. Specifically, the notion that if two observers\nget the same results for all experiments, then the description\nof nature that they use should be the same, seems self-evident.\nCombine this with the fact that no known or conceivable experiment\ncan distinguish between rest and "uniform motion", and there\'s\na powerful argument for special relativity, based on simplicity.\n\nThere may also be independent reasons to abandon realism, but\ndiscussions of these quickly degenerate into nonsense.\n\nIt is usually the case that a given physicist is merely repeating\nwhat he was told as a student of quantum mechanics, without having\nthought about it himself. I know of very very few physicists who\nwould think deeply and then reject realism were they not told to\ndo so. In fact, the proliferation of interpretations of quantum\nmechanics, Bohmian, many-worlds, Penrose\'s gravitational collapse,\nand so on are indications that when physicists do put a little\nthought into their understanding of quantum mechanics, they try\nvery hard to rescue realism. I was just at the KITP\'s miniprogram\non the future of physics, which had an afternoon of quantum\nmechanics:\nhttp://online.itp.ucsb.edu/online/kitp25/\n\nNobody even brought up the possibility that the wavefunction\nmight describe knowledge (\'t Hooft was closest - he considers\nquantum states as equivalence classes of underlying\ndeterministic states which lose information over time; he was\nalso the only one concerned about nonlocality). Once upon a\ntime, most physicists believed that it did describe knowledge.\n\nHow do we view the change from the dominance of an epistemological\ninterpretation to the prevalence of an ontological one? Is it\nlike the waning influence of a once-dictatorial church? Or is it\nmore like a group of people collectively forgetting something\nthat their predecessors paid attention to? What arguments have been\npresented against the wavefunction-represents-knowledge view?\nI can\'t remember any, but they must be pretty convincing if\nthey\'ve managed to change the viewpoint of everyone who has thought\nabout the question.\n\n>> The natural question for your ether theory is whether it strictly\n>> forbids information from travelling faster than light, since it\n>> is evidently comfortable with superluminal influences.\n\n>In the case of my classical ether theory I have proven the EEP.\n>See gr-qc/0205035. Once the theory is classical, causal and\n>deterministic, it follows from the EEP that the light cone restricts\n>information transfer.\n\n>In the pure quantum case we have no longer a classical light\n>cone which allows to restrict the flow of information. Therefore\n>in full quantum theory we should not expect such a restriction.\n\nLooking at your reference, it looks like the theory says that\nthe universe has no beginning in time, although this is put\nin by hand in the axioms. Do you know this a priori?\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb

>> It's more
>> of an attitude than an argument - "You can't use this effect to send
>> information, so it's not worth thinking about."

>Yep.

>> If one does think about it a little, though, it quickly becomes
>> apparent that something very strange is indeed going on - there is
>> a conflict between Einstein causality and the observed experimental
>> facts (modulo the very small probability that, as experiments improve,
>> the observed effect will disappear).

>Yep. Except that I don't see a reason to name this "strange".
>That's what happens all the time in science - theories are
>falsified by experiments. Moreover, we have alternative
>theories which are not falsified (Bohmian mechanics).
>What's strange?

For me, nonlocality itself is rather odd. If the world is nonlocal,
then how does locality emerge? It's not satisfying to merely say
that we can't use it to send signals, where what is really meant
is that it can't be used to *deliberately* send signals; signals
can be sent, but only accidentally. If one wants a realistic theory,
the appearance of such a human notion as deliberate intention in
such an important role is distateful, to say the least.

On the other hand, you're right - the realist model of special
relativity, with space-time as the stage in which local things
play out their local acts, has been falsified (notwithstanding
loopholes and arguments which fail your FTL telephone test).

So what do we do? The choices are to abandon realism, or abandon
special relativity. But the arguments in favour of relativity
are very strong. Specifically, the notion that if two observers
get the same results for all experiments, then the description
of nature that they use should be the same, seems self-evident.
Combine this with the fact that no known or conceivable experiment
can distinguish between rest and "uniform motion", and there's
a powerful argument for special relativity, based on simplicity.

There may also be independent reasons to abandon realism, but
discussions of these quickly degenerate into nonsense.

It is usually the case that a given physicist is merely repeating
what he was told as a student of quantum mechanics, without having
thought about it himself. I know of very very few physicists who
would think deeply and then reject realism were they not told to
do so. In fact, the proliferation of interpretations of quantum
mechanics, Bohmian, many-worlds, Penrose's gravitational collapse,
and so on are indications that when physicists do put a little
thought into their understanding of quantum mechanics, they try
very hard to rescue realism. I was just at the KITP's miniprogram
on the future of physics, which had an afternoon of quantum
mechanics:
http://online.itp.ucsb.edu/online/kitp25/

Nobody even brought up the possibility that the wavefunction
might describe knowledge ('t Hooft was closest - he considers
quantum states as equivalence classes of underlying
deterministic states which lose information over time; he was
also the only one concerned about nonlocality). Once upon a
time, most physicists believed that it did describe knowledge.

How do we view the change from the dominance of an epistemological
interpretation to the prevalence of an ontological one? Is it
like the waning influence of a once-dictatorial church? Or is it
more like a group of people collectively forgetting something
that their predecessors paid attention to? What arguments have been
presented against the wavefunction-represents-knowledge view?
I can't remember any, but they must be pretty convincing if
they've managed to change the viewpoint of everyone who has thought
about the question.

>> The natural question for your ether theory is whether it strictly
>> forbids information from travelling faster than light, since it
>> is evidently comfortable with superluminal influences.

>In the case of my classical ether theory I have proven the EEP.
>See http://www.arxiv.org/abs/gr-qc/0205035. Once the theory is classical, causal and
>deterministic, it follows from the EEP that the light cone restricts
>information transfer.

>In the pure quantum case we have no longer a classical light
>cone which allows to restrict the flow of information. Therefore
>in full quantum theory we should not expect such a restriction.

Looking at your reference, it looks like the theory says that
the universe has no beginning in time, although this is put
in by hand in the axioms. Do you know this a priori?

R.

Ilja Schmelzer

Oct13-04, 03:45 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb\n> Ilja Schmelzer wrote:\n> > "Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb\n> >> Ilja Schmelzer wrote:\n> >> > If there is no free will of the experimenter, you can do whatever\n> >> > you like with your FTL phone but you will be unable to prove\n> >> > that there exists some FTL effects inside the phone.\n> >> The problem with that argument is that it uses something like a triple\n> >> negative and a very unlikely hypothetical, which while logically\ncorrect,\n> >> is more than you can expect the reader to follow. Especially a reader\nwho\n> >> is already struggling with the concepts, as most of those who need the\n> >> argument are.\n> >\n> > Hm. At the current moment, it is the majority of the scientific\ncommunity\n> > which does not accept my interpretation - that the violation of Bell\'s\n> > inequality is an experimental falsification of Einstein causality, as\ngood\n> > as any other falsification. IOW, whose "who need it" are the majority\nof\n> > scientists. I would not expect that this majority is unable to follow\n> > a logical argument of this type.\n\n> In that case, I don\'t understand after all.\n> Quantum Mechanics does not permit an "FTL phone", and neither do *any* of\n> its interpretations.\n> So how is it relevant?\n\nIt is an argument about the scientific method.\n\nLet\'s formulate it as a critical question:\n\nX defends Einstein causality. He claims: "You are unable\nto build an FTL phone."\n\nI ask him: "Ok, but would you accept that relativity is\nfalse if I\'m able to build an FTL phone?"\n\nThe expected answer is "Of course".\n\nAfter this, he can no longer use arguments which fail\nmy FTL phone test.\n\n> > Now, if an EPRB loophole\n> > is sufficient to explain away an FTL phone, Einstein causality becomes\n> > obviously unfalsifiable and is, therefore, no longer part of empirical\n> > science.\n\n> I don\'t understand. The "no free will" loophole exists (and is not worth\n> taking seriously) for *any* experiment. How does that prove anything?\n\nThe point of the argument is, of course, that a loophole which exists for\nany\nexperiment is not worth to be taken seriously. Once you accept for other\nreasons that it is not worth to be taken seriously, fine, the argument is\nnot\nnecessary.\n\nBut it is not that obvious. At least Bell has taken it seriously, that means\nmentioned it and rejected it based on (different) arguments. Thus, its\nworth to have arguments against it.\n\nMoreover, my FTL phone argument may be used against other arguments\nas well. Even against yet unknown arguments. Once somebody presents\na new argument in defense of Einstein causality, all you have to do is to\ncheck: Does it work if confronted with an FTL phone? If yes, its trash.\n\n> Quantum mechanics (together with special relativity) *does* have a\n> particular kind of locality. That kind of locality *is* falsifiable (an\nFTL\n> phone would do quite nicely), so what\'s the problem?\n\nThe let\'s apply the FTL phone test to the following argument:\n\n"We have a mathematical theory which allows to compute\nthe probabilities we observe. It is not a realistic theory, it does\nnot explain how these probabilities appear, but that does not matter."\n\nCan this argument be applied to an FTL phone? I think yes.\nWe use some realistic theory with preferred frame which allows\nto compute the observed probabilities (the analogon of BM).\nThen we remove the realistic, interpretational part from the theory,\nthat means everything except the directly observable predictions.\nAs part of this scheme, we can interpret the preferred time T as a\nphysical field T(x) which fulfills a relativistic wave equation.\n\nUsing this quite artificial "theory" a relativistic dogmatic is able\nto defend relativity in a similar way even if confronted with\nan FTL phone. Therefore, the argument should be rejected\nat least in the way I have formulated it.\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb
> Ilja Schmelzer wrote:
> > "Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb
> >> Ilja Schmelzer wrote:
> >> > If there is no free will of the experimenter, you can do whatever
> >> > you like with your FTL phone but you will be unable to prove
> >> > that there exists some FTL effects inside the phone.
> >> The problem with that argument is that it uses something like a triple
> >> negative and a very unlikely hypothetical, which while logically
correct,
> >> is more than you can expect the reader to follow. Especially a reader
who
> >> is already struggling with the concepts, as most of those who need the
> >> argument are.
> >
> > Hm. At the current moment, it is the majority of the scientific
community
> > which does not accept my interpretation - that the violation of Bell's
> > inequality is an experimental falsification of Einstein causality, as
good
> > as any other falsification. IOW, whose "who need it" are the majority
of
> > scientists. I would not expect that this majority is unable to follow
> > a logical argument of this type.

> In that case, I don't understand after all.
> Quantum Mechanics does not permit an "FTL phone", and neither do *any* of
> its interpretations.
> So how is it relevant?

It is an argument about the scientific method.

Let's formulate it as a critical question:

X defends Einstein causality. He claims: "You are unable
to build an FTL phone."

I ask him: "Ok, but would you accept that relativity is
false if I'm able to build an FTL phone?"

The expected answer is "Of course".

After this, he can no longer use arguments which fail
my FTL phone test.

> > Now, if an EPRB loophole
> > is sufficient to explain away an FTL phone, Einstein causality becomes
> > obviously unfalsifiable and is, therefore, no longer part of empirical
> > science.

> I don't understand. The "no free will" loophole exists (and is not worth
> taking seriously) for *any* experiment. How does that prove anything?

The point of the argument is, of course, that a loophole which exists for
any
experiment is not worth to be taken seriously. Once you accept for other
reasons that it is not worth to be taken seriously, fine, the argument is
not
necessary.

But it is not that obvious. At least Bell has taken it seriously, that means
mentioned it and rejected it based on (different) arguments. Thus, its
worth to have arguments against it.

Moreover, my FTL phone argument may be used against other arguments
as well. Even against yet unknown arguments. Once somebody presents
a new argument in defense of Einstein causality, all you have to do is to
check: Does it work if confronted with an FTL phone? If yes, its trash.

> Quantum mechanics (together with special relativity) *does* have a
> particular kind of locality. That kind of locality *is* falsifiable (an
FTL
> phone would do quite nicely), so what's the problem?

The let's apply the FTL phone test to the following argument:

"We have a mathematical theory which allows to compute
the probabilities we observe. It is not a realistic theory, it does
not explain how these probabilities appear, but that does not matter."

Can this argument be applied to an FTL phone? I think yes.
We use some realistic theory with preferred frame which allows
to compute the observed probabilities (the analogon of BM).
Then we remove the realistic, interpretational part from the theory,
that means everything except the directly observable predictions.
As part of this scheme, we can interpret the preferred time T as a
physical field T(x) which fulfills a relativistic wave equation.

Using this quite artificial "theory" a relativistic dogmatic is able
to defend relativity in a similar way even if confronted with
an FTL phone. Therefore, the argument should be rejected
at least in the way I have formulated it.

Ilja

Ilja Schmelzer

Oct13-04, 03:45 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> >Except that I don\'t see a reason to name this "strange".\n> >That\'s what happens all the time in science - theories are\n> >falsified by experiments. Moreover, we have alternative\n> >theories which are not falsified (Bohmian mechanics).\n> >What\'s strange?\n\n> For me, nonlocality itself is rather odd.\n\nNot that odd. Last not least, Newtonian mechanics is\nnonlocal too.\n\n> If the world is nonlocal, then how does locality emerge?\n\nAs the rather small speed of sound emerges in a world\nwhere the speed of light is the limiting velocity.\n\n> It\'s not satisfying to merely say\n> that we can\'t use it to send signals, where what is really meant\n> is that it can\'t be used to *deliberately* send signals; signals\n> can be sent, but only accidentally. If one wants a realistic theory,\n> the appearance of such a human notion as deliberate intention in\n> such an important role is distateful, to say the least.\n\nI don\'t think so. You don\'t have to introduce deliberate action\nor intention. It is sufficient to prove that some part of the really\nexisting information is not observable for internal observers.\nFor an example of such a proof see (in the context of the\npreferred frame in a classical ether) my paper gr-qc/0205035.\n\n> On the other hand, you\'re right - the realist model of special\n> relativity, with space-time as the stage in which local things\n> play out their local acts, has been falsified (notwithstanding\n> loopholes and arguments which fail your FTL telephone test).\n\n> So what do we do? The choices are to abandon realism,\n> or abandon special relativity. But the arguments in favour\n> of relativity are very strong.\n> Specifically, the notion that if two observers\n> get the same results for all experiments, then the description\n> of nature that they use should be the same, seems self-evident.\n\nBut it isn\'t.\n\nLook at it from another point. Imagine some world. Imagine\nsome internal creatures of this world. Is it obvious that these\ninternal observers are able to distinguish by observation all\nsituations which are really different? Not at all. I would even\nsay that in some sense it is impossible at all.\n\nThus, there will be different situations which cannot be\ndistinguished by observation. Internal relativists will\ncertainly fail to describe their world correctly.\n\n> Combine this with the fact that no known or conceivable experiment\n> can distinguish between rest and "uniform motion", and there\'s\n> a powerful argument for special relativity, based on simplicity.\n\nA symmetry group for observables is, indeed, a powerful argument\nfor proposing the same symmetry group as fundamental. But if\nwe have the following two things:\n\n1.) There is no fundamental theory with this symmetry group\n(the failure of quantum GR)\n\n2.) The appearence of this symmetry group as an approximate\nsymmetry of some (large distance) limit is quite natural, that means,\nthere exists theories with different fundamental symmetry which\nhave the observable symmetry in some limit\n\nthen I would no longer say that the argument is strong.\n\n> There may also be independent reasons to abandon realism, but\n> discussions of these quickly degenerate into nonsense.\n\nThere may be but at least I don\'t know any\nindependent reasons.\n\nInstead, I believe that realism (that means all what is necessary to\nprove Bell\'s inequality except Einstein causality) should be\npostulated, considered as part of the scientific method, as some\nextension of logic.\n\nIf we reject logic, we give up science.\nIf we reject realism, we also give up some part of\nscience, namely the search for realistic explanations\nof observable correlations.\n\n> Nobody even brought up the possibility that the wavefunction\n> might describe knowledge (\'t Hooft was closest - he considers\n> quantum states as equivalence classes of underlying\n> deterministic states which lose information over time; he was\n> also the only one concerned about nonlocality). Once upon a\n> time, most physicists believed that it did describe knowledge.\n\nI also prefer ontological interpretations. I don\'t believe that\nthe wave function itself is what really exists. I think it is a function\nof what really exists.\n\n> How do we view the change from the dominance of an epistemological\n> interpretation to the prevalence of an ontological one? Is it\n> like the waning influence of a once-dictatorial church? Or is it\n> more like a group of people collectively forgetting something\n> that their predecessors paid attention to? What arguments have been\n> presented against the wavefunction-represents-knowledge view?\n> I can\'t remember any, but they must be pretty convincing if\n> they\'ve managed to change the viewpoint of everyone who has thought\n> about the question.\n\nI would not say that I have much arguments against the\nwavefunction-represents-knowledge view. I have to think\nabout it but at the current moment I think that this view may\nbe compatible with a realistic interpretation.\n\n> Looking at your reference, it looks like the theory says that\n> the universe has no beginning in time, although this is put\n> in by hand in the axioms. Do you know this a priori?\n\nNo. As in every theory I have to postulate something.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> >Except that I don't see a reason to name this "strange".
> >That's what happens all the time in science - theories are
> >falsified by experiments. Moreover, we have alternative
> >theories which are not falsified (Bohmian mechanics).
> >What's strange?

> For me, nonlocality itself is rather odd.

Not that odd. Last not least, Newtonian mechanics is
nonlocal too.

> If the world is nonlocal, then how does locality emerge?

As the rather small speed of sound emerges in a world
where the speed of light is the limiting velocity.

> It's not satisfying to merely say
> that we can't use it to send signals, where what is really meant
> is that it can't be used to *deliberately* send signals; signals
> can be sent, but only accidentally. If one wants a realistic theory,
> the appearance of such a human notion as deliberate intention in
> such an important role is distateful, to say the least.

I don't think so. You don't have to introduce deliberate action
or intention. It is sufficient to prove that some part of the really
existing information is not observable for internal observers.
For an example of such a proof see (in the context of the
preferred frame in a classical ether) my paper http://www.arxiv.org/abs/gr-qc/0205035.

> On the other hand, you're right - the realist model of special
> relativity, with space-time as the stage in which local things
> play out their local acts, has been falsified (notwithstanding
> loopholes and arguments which fail your FTL telephone test).

> So what do we do? The choices are to abandon realism,
> or abandon special relativity. But the arguments in favour
> of relativity are very strong.
> Specifically, the notion that if two observers
> get the same results for all experiments, then the description
> of nature that they use should be the same, seems self-evident.

But it isn't.

Look at it from another point. Imagine some world. Imagine
some internal creatures of this world. Is it obvious that these
internal observers are able to distinguish by observation all
situations which are really different? Not at all. I would even
say that in some sense it is impossible at all.

Thus, there will be different situations which cannot be
distinguished by observation. Internal relativists will
certainly fail to describe their world correctly.

> Combine this with the fact that no known or conceivable experiment
> can distinguish between rest and "uniform motion", and there's
> a powerful argument for special relativity, based on simplicity.

A symmetry group for observables is, indeed, a powerful argument
for proposing the same symmetry group as fundamental. But if
we have the following two things:

1.) There is no fundamental theory with this symmetry group
(the failure of quantum GR)

2.) The appearence of this symmetry group as an approximate
symmetry of some (large distance) limit is quite natural, that means,
there exists theories with different fundamental symmetry which
have the observable symmetry in some limit

then I would no longer say that the argument is strong.

> There may also be independent reasons to abandon realism, but
> discussions of these quickly degenerate into nonsense.

There may be but at least I don't know any
independent reasons.

Instead, I believe that realism (that means all what is necessary to
prove Bell's inequality except Einstein causality) should be
postulated, considered as part of the scientific method, as some
extension of logic.

If we reject logic, we give up science.
If we reject realism, we also give up some part of
science, namely the search for realistic explanations
of observable correlations.

> Nobody even brought up the possibility that the wavefunction
> might describe knowledge ('t Hooft was closest - he considers
> quantum states as equivalence classes of underlying
> deterministic states which lose information over time; he was
> also the only one concerned about nonlocality). Once upon a
> time, most physicists believed that it did describe knowledge.

I also prefer ontological interpretations. I don't believe that
the wave function itself is what really exists. I think it is a function
of what really exists.

> How do we view the change from the dominance of an epistemological
> interpretation to the prevalence of an ontological one? Is it
> like the waning influence of a once-dictatorial church? Or is it
> more like a group of people collectively forgetting something
> that their predecessors paid attention to? What arguments have been
> presented against the wavefunction-represents-knowledge view?
> I can't remember any, but they must be pretty convincing if
> they've managed to change the viewpoint of everyone who has thought
> about the question.

I would not say that I have much arguments against the
wavefunction-represents-knowledge view. I have to think
about it but at the current moment I think that this view may
be compatible with a realistic interpretation.

> Looking at your reference, it looks like the theory says that
> the universe has no beginning in time, although this is put
> in by hand in the axioms. Do you know this a priori?

No. As in every theory I have to postulate something.

Ilja

Ralph Hartley

Oct13-04, 02:13 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nrof@maths.tcd.ie wrote:\n> the possibility that the wavefunction\n> might describe knowledge\n\nThere is a sense in which that is a reasonable view.\n\nBut "knowledge" isn\'t *exactly* the right word. What it really describes is\nwhat knowledge you *could* have.\n\nThe formulation of QM in terms of density matrices expresses things a\nlittle bit better, since it has a representation for ignorance (mixed\nstates), but it still doesn\'t quite capture the sense in which the state\ndescribes *objective* knowledge.\n\nThe distinction exists (and is essentially the same) in classical\nprobability theory.\n\nProbabilities can be be used to describe knowledge (or belief), but they\nhave an objective existence as well.\n\nConsider a gambler about to make a bet on a roll in a Craps game. He may\nbelieve that the probability of winning is 50%, but there is a *real*\nprobability as well (quite likely much lower).\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>rof@maths.tcd.ie wrote:
> the possibility that the wavefunction
> might describe knowledge

There is a sense in which that is a reasonable view.

But "knowledge" isn't *exactly* the right word. What it really describes is
what knowledge you *could* have.

The formulation of QM in terms of density matrices expresses things a
little bit better, since it has a representation for ignorance (mixed
states), but it still doesn't quite capture the sense in which the state
describes *objective* knowledge.

The distinction exists (and is essentially the same) in classical
probability theory.

Probabilities can be be used to describe knowledge (or belief), but they
have an objective existence as well.

Consider a gambler about to make a bet on a roll in a Craps game. He may
believe that the probability of winning is 50%, but there is a *real*
probability as well (quite likely much lower).

Ralph Hartley

Patrick Van Esch

Oct14-04, 11:07 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nrof@maths.tcd.ie wrote in message news:<ckf91p\\$79j\\$1@lanczos.maths.tcd.ie>...\n> I know of very very few physicists who\n> would think deeply and then reject realism were they not told to\n> do so.\n\nWell, I think I found a way (just for my peace of mind) :-) It comes\nclose to Wigner\'s idea in the 60-ies, and keeps as well a certain kind\nof realism, as well as locality. But the price to pay can seem to be\nhigh: it is a form of solipsism.\n\nThe trick is that there is ONE true measurement apparatus in this\nuniverse, namely my personal consciousness, and only mine, and it is\nmy own observation which makes the collapse. In an EPR-like\nexperiment, when Bob and Alice make their respective measurements, and\nsend me their notes by mail, these notes (and Bob and Alice) remain in\nsuperposed states which are entangled by the EPR photon pair, until *I\npersonally* look at them. This collapses locally the states of these\nnotes into readable, classical notes, and it is during this collapse\nthat the necessary correlations appear. Before I was here, the\nuniverse evolved unitarily, and after I\'ll die, it will also (until I\nreincarnate :-))\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>rof@maths.tcd.ie wrote in message news:<ckf91p$79j$1@lanczos.maths.tcd.ie>...
> I know of very very few physicists who
> would think deeply and then reject realism were they not told to
> do so.

Well, I think I found a way (just for my peace of mind) :-) It comes
close to Wigner's idea in the 60-ies, and keeps as well a certain kind
of realism, as well as locality. But the price to pay can seem to be
high: it is a form of solipsism.

The trick is that there is ONE true measurement apparatus in this
universe, namely my personal consciousness, and only mine, and it is
my own observation which makes the collapse. In an EPR-like
experiment, when Bob and Alice make their respective measurements, and
send me their notes by mail, these notes (and Bob and Alice) remain in
superposed states which are entangled by the EPR photon pair, until *I
personally* look at them. This collapses locally the states of these
notes into readable, classical notes, and it is during this collapse
that the necessary correlations appear. Before I was here, the
universe evolved unitarily, and after I'll die, it will also (until I
reincarnate :-))

cheers,
Patrick.

Ralph Hartley

Oct14-04, 11:07 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nIlja Schmelzer wrote:\n> "Ralph Hartley" :\n\n>>Quantum mechanics (together with special relativity) *does* have a\n>>particular kind of locality. That kind of locality *is* falsifiable (an\n>>FTL phone would do quite nicely), so what\'s the problem?\n>\n> The let\'s apply the FTL phone test to the following argument:\n>\n> "We have a mathematical theory which allows to compute\n> the probabilities we observe. It is not a realistic theory,\n> it does\n> not explain how these probabilities appear, but that does not matter."\n\nBut it *is* a local theory. It absolutely forbids any FTL phone.\n\n> Can this argument be applied to an FTL phone? I think yes.\n> We use some realistic theory with preferred frame which allows\n> to compute the observed probabilities\n\nBut to give the same probabilities, that theory has to be local (in the\nsense of forbidding FTL information transfer) as well. It may not *look*\nlocal the way it is formulated, but its nonlocal features can never have\nany observable effect. It can\'t permit an FTL phone, since by definition it\nmakes the same predictions as a local theory.\n\nBM is only OK because it has that property.\n\n> Then we remove the realistic, interpretational part from the theory,\n> that means everything except the directly observable predictions.\n\nSo it\'s back where it started, isn\'t it?\n\n> As part of this scheme, we can interpret the preferred time T as a\n> physical field T(x) which fulfills a relativistic wave equation.\n\nYou can introduce any (unobservable) fields you like. An FTL phone is still\nforbidden by the theory.\n\n> Using this quite artificial "theory" a relativistic dogmatic is able\n> to defend relativity in a similar way even if confronted with\n> an FTL phone.\n\nI don\'t see how. I don\'t even see how that is *supposed* to follow.\n\n> Therefore, the argument should be rejected\n> at least in the way I have formulated it.\n\nYou can always formulate things in a way that must be rejected. What does\nthat prove?\n\nBut I still don\'t understand which argument you *claim* must be rejected.\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Ilja Schmelzer wrote:
> "Ralph Hartley" :

>>Quantum mechanics (together with special relativity) *does* have a
>>particular kind of locality. That kind of locality *is* falsifiable (an
>>FTL phone would do quite nicely), so what's the problem?
>
> The let's apply the FTL phone test to the following argument:
>
> "We have a mathematical theory which allows to compute
> the probabilities we observe. It is not a realistic theory,
> it does
> not explain how these probabilities appear, but that does not matter."

But it *is* a local theory. It absolutely forbids any FTL phone.

> Can this argument be applied to an FTL phone? I think yes.
> We use some realistic theory with preferred frame which allows
> to compute the observed probabilities

But to give the same probabilities, that theory has to be local (in the
sense of forbidding FTL information transfer) as well. It may not *look*
local the way it is formulated, but its nonlocal features can never have
any observable effect. It can't permit an FTL phone, since by definition it
makes the same predictions as a local theory.

BM is only OK because it has that property.

> Then we remove the realistic, interpretational part from the theory,
> that means everything except the directly observable predictions.

So it's back where it started, isn't it?

> As part of this scheme, we can interpret the preferred time T as a
> physical field T(x) which fulfills a relativistic wave equation.

You can introduce any (unobservable) fields you like. An FTL phone is still
forbidden by the theory.

> Using this quite artificial "theory" a relativistic dogmatic is able
> to defend relativity in a similar way even if confronted with
> an FTL phone.

I don't see how. I don't even see how that is *supposed* to follow.

> Therefore, the argument should be rejected
> at least in the way I have formulated it.

You can always formulate things in a way that must be rejected. What does
that prove?

But I still don't understand which argument you *claim* must be rejected.

Ralph Hartley

Ilja Schmelzer

Oct15-04, 01:54 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb im Newsbeitrag\nnews:416DA0AD.4000808@aic.nrl.navy.mi l...\n> Ilja Schmelzer wrote:\n> > "Ralph Hartley" :\n> >>Quantum mechanics (together with special relativity) *does* have a\n> >>particular kind of locality. That kind of locality *is* falsifiable (an\n> >>FTL phone would do quite nicely), so what\'s the problem?\n> >\n> > The let\'s apply the FTL phone test to the following argument:\n> >\n> > "We have a mathematical theory which allows to compute\n> > the probabilities we observe. It is not a realistic theory,\n> > it does not explain how these probabilities appear, but\n> > that does not matter."\n\n> But it *is* a local theory. It absolutely forbids any FTL phone.\n\nThis is not a criterion of being a local theory. BM also absolutely\nforbids any FTL phone but is obviously nonlocal.\n\nOnce QM is not a realistic theory (in the sense of EPRB realism)\nI don\'t think it is reasonable to name it "local" or "nonlocal".\n\n> > Can this argument be applied to an FTL phone? I think yes.\n> > We use some realistic theory with preferred frame which allows\n> > to compute the observed probabilities\n\n> But to give the same probabilities, that theory has to be local (in the\n> sense of forbidding FTL information transfer) as well. It may not *look*\n> local the way it is formulated, but its nonlocal features can never have\n> any observable effect.\n\nSorry, but we can observe violations of Bell\'s inequalities. And\nviolations of Bell\'s inequalities are nonlocal effects.\n\n> It can\'t permit an FTL phone, since by definition it\n> makes the same predictions as a local theory.\n> BM is only OK because it has that property.\n>\n> > Then we remove the realistic, interpretational part from the theory,\n> > that means everything except the directly observable predictions.\n>\n> So it\'s back where it started, isn\'t it?\n\nNo, in this thought experiment it started with a realistic theory,\nthat means with a theory which describes a model of reality.\nObservable probabilities are derived from this model.\n\nAfter the removal, the only thing which survives is a mathematical\napparatus without interpretation and the resulting observable\nprobabilities.\n\nThese are two different classes of theories.\n\n> > As part of this scheme, we can interpret the preferred time T as a\n> > physical field T(x) which fulfills a relativistic wave equation.\n>\n> You can introduce any (unobservable) fields you like.\n> An FTL phone is still forbidden by the theory.\n\nSeems you have not understood the point of this thought\nexperiment. It is about a hypothetical world where FTL phones\nexist, and hypothetical scientists who develop a hypothetical\nmathematical theory which allows to compute the hypothetical\nobservable probabilities observed in the FTL phone.\n\nThe thesis is that it is possible to develop such a theory.\nThen, the hypothetical situation is almost analoguous to the\ncurrent situation with QM.\n\nThe only difference is that the hypothetical theory allows the\nprobabilities obbservable in FTL phones, but QM doesn\'t.\n\nThesis: The existence of a nonrealistic mathematical theory\nwhich allows to compute the observable probabilies proves\nnothing about the local or nonlocal character of the observed\neffects.\n\n> > Using this quite artificial "theory" a relativistic dogmatic is able\n> > to defend relativity in a similar way even if confronted with\n> > an FTL phone.\n>\n> I don\'t see how. I don\'t even see how that is *supposed* to follow.\n\nHe applies the following argument:\n\n> > "We have a mathematical theory which allows to compute\n> > the probabilities we observe. It is not a realistic theory,\n> > it does not explain how these probabilities appear,\n> > but that does not matter."\n\n> You can always formulate things in a way that must be rejected. What does\n> that prove?\n\nIt proves what it proves. If you know yourself that the argument\nhas to be rejected, fine, my argumentation was not necessary.\n\n> But I still don\'t understand which argument you *claim* must be rejected.\n\nThe claim is:\n\n> > "We have a mathematical theory which allows to compute\n> > the probabilities we observe. It is not a realistic theory,\n> > it does not explain how these probabilities appear,\n> > but that does not matter."\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb im Newsbeitrag
news:416DA0AD.4000808@aic.nrl.navy.mil...
> Ilja Schmelzer wrote:
> > "Ralph Hartley" :
> >>Quantum mechanics (together with special relativity) *does* have a
> >>particular kind of locality. That kind of locality *is* falsifiable (an
> >>FTL phone would do quite nicely), so what's the problem?
> >
> > The let's apply the FTL phone test to the following argument:
> >
> > "We have a mathematical theory which allows to compute
> > the probabilities we observe. It is not a realistic theory,
> > it does not explain how these probabilities appear, but
> > that does not matter."

> But it *is* a local theory. It absolutely forbids any FTL phone.

This is not a criterion of being a local theory. BM also absolutely
forbids any FTL phone but is obviously nonlocal.

Once QM is not a realistic theory (in the sense of EPRB realism)
I don't think it is reasonable to name it "local" or "nonlocal".

> > Can this argument be applied to an FTL phone? I think yes.
> > We use some realistic theory with preferred frame which allows
> > to compute the observed probabilities

> But to give the same probabilities, that theory has to be local (in the
> sense of forbidding FTL information transfer) as well. It may not *look*
> local the way it is formulated, but its nonlocal features can never have
> any observable effect.

Sorry, but we can observe violations of Bell's inequalities. And
violations of Bell's inequalities are nonlocal effects.

> It can't permit an FTL phone, since by definition it
> makes the same predictions as a local theory.
> BM is only OK because it has that property.
>
> > Then we remove the realistic, interpretational part from the theory,
> > that means everything except the directly observable predictions.
>
> So it's back where it started, isn't it?

No, in this thought experiment it started with a realistic theory,
that means with a theory which describes a model of reality.
Observable probabilities are derived from this model.

After the removal, the only thing which survives is a mathematical
apparatus without interpretation and the resulting observable
probabilities.

These are two different classes of theories.

> > As part of this scheme, we can interpret the preferred time T as a
> > physical field T(x) which fulfills a relativistic wave equation.
>
> You can introduce any (unobservable) fields you like.
> An FTL phone is still forbidden by the theory.

Seems you have not understood the point of this thought
experiment. It is about a hypothetical world where FTL phones
exist, and hypothetical scientists who develop a hypothetical
mathematical theory which allows to compute the hypothetical
observable probabilities observed in the FTL phone.

The thesis is that it is possible to develop such a theory.
Then, the hypothetical situation is almost analoguous to the
current situation with QM.

The only difference is that the hypothetical theory allows the
probabilities obbservable in FTL phones, but QM doesn't.

Thesis: The existence of a nonrealistic mathematical theory
which allows to compute the observable probabilies proves
nothing about the local or nonlocal character of the observed
effects.

> > Using this quite artificial "theory" a relativistic dogmatic is able
> > to defend relativity in a similar way even if confronted with
> > an FTL phone.
>
> I don't see how. I don't even see how that is *supposed* to follow.

He applies the following argument:

> > "We have a mathematical theory which allows to compute
> > the probabilities we observe. It is not a realistic theory,
> > it does not explain how these probabilities appear,
> > but that does not matter."

> You can always formulate things in a way that must be rejected. What does
> that prove?

It proves what it proves. If you know yourself that the argument
has to be rejected, fine, my argumentation was not necessary.

> But I still don't understand which argument you *claim* must be rejected.

The claim is:

> > "We have a mathematical theory which allows to compute
> > the probabilities we observe. It is not a realistic theory,
> > it does not explain how these probabilities appear,
> > but that does not matter."

Ilja

rof@maths.tcd.ie

Oct15-04, 01:55 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n\n>> For me, nonlocality itself is rather odd.\n\n>Not that odd. Last not least, Newtonian mechanics is\n>nonlocal too.\n\nTrue; it\'s relativity which makes nonlocality really strange.\n\n>> If the world is nonlocal, then how does locality emerge?\n\n>As the rather small speed of sound emerges in a world\n>where the speed of light is the limiting velocity.\n\nThere\'s obviously a bit more to explain than that (decoherence\nwould need to be appealed to before one can even talk about\nindividual objects which could even have a well-defined or\napproximately well-defined velocity). It also seems circular\nto start with a spatial description (even a nonlocal one) and\nthen deduce later that a spatial description is appropriate.\nA more honest approach would be to start with an abstract description\nwhich didn\'t have space in it anywhere and then deduce that\na spatial representation is appropriate for describing some\ncoarse-grained features. That\'s something that hasn\'t been done,\nwith the possible exception of matrix models in string theory,\nin which I\'m not an expert.\n\n>> It\'s not satisfying to merely say\n>> that we can\'t use it to send signals, where what is really meant\n>> is that it can\'t be used to *deliberately* send signals; signals\n>> can be sent, but only accidentally. If one wants a realistic theory,\n>> the appearance of such a human notion as deliberate intention in\n>> such an important role is distateful, to say the least.\n\n>I don\'t think so. You don\'t have to introduce deliberate action\n>or intention. It is sufficient to prove that some part of the really\n>existing information is not observable for internal observers.\n>For an example of such a proof see (in the context of the\n>preferred frame in a classical ether) my paper gr-qc/0205035.\n\n>> But the arguments in favour of relativity are very strong.\n>> Specifically, the notion that if two observers\n>> get the same results for all experiments, then the description\n>> of nature that they use should be the same, seems self-evident.\n\n>But it isn\'t.\n\n>Look at it from another point. Imagine some world. Imagine\n>some internal creatures of this world. Is it obvious that these\n>internal observers are able to distinguish by observation all\n>situations which are really different? Not at all. I would even\n>say that in some sense it is impossible at all.\n\n>Thus, there will be different situations which cannot be\n>distinguished by observation. Internal relativists will\n>certainly fail to describe their world correctly.\n\nHere is the nasty face of realism. If two experimenters get\nthe same results for all experiments, and use the same\n(non-stochastic or non-opinionated) reasoning, then they\nmust inevitably come to the same description. What you\nare saying is that their descriptions will probably be wrong,\nand, if the situation is as you described it, then we have\nto face the fact that our descriptions are, in the same\nsense, probably wrong, including yours.\n\n>> Combine this with the fact that no known or conceivable experiment\n>> can distinguish between rest and "uniform motion", and there\'s\n>> a powerful argument for special relativity, based on simplicity.\n\n>A symmetry group for observables is, indeed, a powerful argument\n>for proposing the same symmetry group as fundamental. But if\n>we have the following two things:\n\n>1.) There is no fundamental theory with this symmetry group\n>(the failure of quantum GR)\n\n>2.) The appearence of this symmetry group as an approximate\n>symmetry of some (large distance) limit is quite natural, that means,\n>there exists theories with different fundamental symmetry which\n>have the observable symmetry in some limit\n\n>then I would no longer say that the argument is strong.\n\nRight, but the lack of a fundamental theory isn\'t quite\nthe failure of quantum gravity. Instead it\'s the failure\nof the mathematical machinery of normal quantum mechanics in\ngeneral, and (special) relativistic quantum mechanics in\nparticular, to incorporate information about the actual\nreults of measurements.\n\nBut I think we can keep the symmetry group by abandoning\ndeterminism without abandoning realism (although this\nis not a position that I advocate). That is, we can say\nthat qm is a framework which describes the statistical correlations\nbetween real events and other real events, without telling\nus anything about which events caused which other one, or\nindeed whether it did. Causation need not necessarily be\nan ingredient in our theories - after all, we didn\'t discover\ncauses by doing experiments and finding them; rather we repeatedly\nsee A following B and conjecture that A necessarily follows B,\nand the addition of necessity to the correlation is the cause.\n\n>> There may also be independent reasons to abandon realism, but\n>> discussions of these quickly degenerate into nonsense.\n\n>There may be but at least I don\'t know any\n>independent reasons.\n\nRather than get deeply into the realism debate (a discussion of\nwhich would be off-topic for spr), I\'ll say that what physicists\nmean by real is actually a confused mixture of several distinct\nideas which can be useful on their own, but are not useful when mixed\ntogether in this way. The concepts mixed together are something\nlike: externality (as in "the external world"), truth (real\nas oppposed to unreal), fundamental as opposed to derivative,\nindependence of which individual observer makes the observation,\nindependence of the act of observation altogether, referring\nto a property of an object as opposed to a property of the\nrelationship between the object and the measuring device, and\nfinally the difference between something which is actually\npresent and something which is merely referred to.\n\nWhen these are mixed together into a single concept of\nreality, we get confusion, such as "Real pain is different\nfrom the mere thought of pain, but isn\'t real, since it\'s not\nexternal." Also, merely because something is the same for\nall observers, people can infer (by supposing themselves justified\nin assigning the "reality" tag) that it is independent of\nthe act of observation altogether, and should therefore\nappear in an ontological theory, which certainly doesn\'t\nfollow.\n\n>Instead, I believe that realism (that means all what is necessary to\n>prove Bell\'s inequality except Einstein causality) should be\n>postulated, considered as part of the scientific method, as some\n>extension of logic.\n\n>If we reject logic, we give up science.\n>If we reject realism, we also give up some part of\n>science, namely the search for realistic explanations\n>of observable correlations.\n\nI\'d rather not adopt beliefs when there\'s still some more figuring\nout left to do. I suspect the criterion for an explanation to\nbe \'realistic\' isn\'t very clearly defined.\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb

>> For me, nonlocality itself is rather odd.

>Not that odd. Last not least, Newtonian mechanics is
>nonlocal too.

True; it's relativity which makes nonlocality really strange.

>> If the world is nonlocal, then how does locality emerge?

>As the rather small speed of sound emerges in a world
>where the speed of light is the limiting velocity.

There's obviously a bit more to explain than that (decoherence
would need to be appealed to before one can even talk about
individual objects which could even have a well-defined or
approximately well-defined velocity). It also seems circular
to start with a spatial description (even a nonlocal one) and
then deduce later that a spatial description is appropriate.
A more honest approach would be to start with an abstract description
which didn't have space in it anywhere and then deduce that
a spatial representation is appropriate for describing some
coarse-grained features. That's something that hasn't been done,
with the possible exception of matrix models in string theory,
in which I'm not an expert.

>> It's not satisfying to merely say
>> that we can't use it to send signals, where what is really meant
>> is that it can't be used to *deliberately* send signals; signals
>> can be sent, but only accidentally. If one wants a realistic theory,
>> the appearance of such a human notion as deliberate intention in
>> such an important role is distateful, to say the least.

>I don't think so. You don't have to introduce deliberate action
>or intention. It is sufficient to prove that some part of the really
>existing information is not observable for internal observers.
>For an example of such a proof see (in the context of the
>preferred frame in a classical ether) my paper http://www.arxiv.org/abs/gr-qc/0205035.

>> But the arguments in favour of relativity are very strong.
>> Specifically, the notion that if two observers
>> get the same results for all experiments, then the description
>> of nature that they use should be the same, seems self-evident.

>But it isn't.

>Look at it from another point. Imagine some world. Imagine
>some internal creatures of this world. Is it obvious that these
>internal observers are able to distinguish by observation all
>situations which are really different? Not at all. I would even
>say that in some sense it is impossible at all.

>Thus, there will be different situations which cannot be
>distinguished by observation. Internal relativists will
>certainly fail to describe their world correctly.

Here is the nasty face of realism. If two experimenters get
the same results for all experiments, and use the same
(non-stochastic or non-opinionated) reasoning, then they
must inevitably come to the same description. What you
are saying is that their descriptions will probably be wrong,
and, if the situation is as you described it, then we have
to face the fact that our descriptions are, in the same
sense, probably wrong, including yours.

>> Combine this with the fact that no known or conceivable experiment
>> can distinguish between rest and "uniform motion", and there's
>> a powerful argument for special relativity, based on simplicity.

>A symmetry group for observables is, indeed, a powerful argument
>for proposing the same symmetry group as fundamental. But if
>we have the following two things:

>1.) There is no fundamental theory with this symmetry group
>(the failure of quantum GR)

>2.) The appearence of this symmetry group as an approximate
>symmetry of some (large distance) limit is quite natural, that means,
>there exists theories with different fundamental symmetry which
>have the observable symmetry in some limit

>then I would no longer say that the argument is strong.

Right, but the lack of a fundamental theory isn't quite
the failure of quantum gravity. Instead it's the failure
of the mathematical machinery of normal quantum mechanics in
general, and (special) relativistic quantum mechanics in
particular, to incorporate information about the actual
reults of measurements.

But I think we can keep the symmetry group by abandoning
determinism without abandoning realism (although this
is not a position that I advocate). That is, we can say
that qm is a framework which describes the statistical correlations
between real events and other real events, without telling
us anything about which events caused which other one, or
indeed whether it did. Causation need not necessarily be
an ingredient in our theories - after all, we didn't discover
causes by doing experiments and finding them; rather we repeatedly
see A following B and conjecture that A necessarily follows B,
and the addition of necessity to the correlation is the cause.

>> There may also be independent reasons to abandon realism, but
>> discussions of these quickly degenerate into nonsense.

>There may be but at least I don't know any
>independent reasons.

Rather than get deeply into the realism debate (a discussion of
which would be off-topic for spr), I'll say that what physicists
mean by real is actually a confused mixture of several distinct
ideas which can be useful on their own, but are not useful when mixed
together in this way. The concepts mixed together are something
like: externality (as in "the external world"), truth (real
as oppposed to unreal), fundamental as opposed to derivative,
independence of which individual observer makes the observation,
independence of the act of observation altogether, referring
to a property of an object as opposed to a property of the
relationship between the object and the measuring device, and
finally the difference between something which is actually
present and something which is merely referred to.

When these are mixed together into a single concept of
reality, we get confusion, such as "Real pain is different
from the mere thought of pain, but isn't real, since it's not
external." Also, merely because something is the same for
all observers, people can infer (by supposing themselves justified
in assigning the "reality" tag) that it is independent of
the act of observation altogether, and should therefore
appear in an ontological theory, which certainly doesn't
follow.

>Instead, I believe that realism (that means all what is necessary to
>prove Bell's inequality except Einstein causality) should be
>postulated, considered as part of the scientific method, as some
>extension of logic.

>If we reject logic, we give up science.
>If we reject realism, we also give up some part of
>science, namely the search for realistic explanations
>of observable correlations.

I'd rather not adopt beliefs when there's still some more figuring
out left to do. I suspect the criterion for an explanation to
be 'realistic' isn't very clearly defined.

R.

Ralph Hartley

Oct15-04, 01:55 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nPatrick Van Esch wrote:\n\n> The trick is that there is ONE true measurement apparatus in this\n> universe, namely my personal consciousness, and only mine, and it is\n> my own observation which makes the collapse.\n\nI\'m sure they must have had this exact same argument after painters\ndiscovered perspective:\n\nQ: Where is the vanishing point?\nA: It depends on the observer.\nQ: What counts as an observer?\nA: Not *an* observer, *the* observer.\nQ: How do you tell who *the* observer is?\nA: It can be anyone you want.\nA: When I\'m painting, it would be me.\nQ: But where is the vanishing point *really* located?\nA: The question doesn\'t even make sense without the observer.\nQ: Why the hell not? All points have positions.\nA: It has a position, but it depends on the observer.\nQ: Perhaps there is one *true* observer that determines the\nQ: where vanishing point *really* is. God perhaps?\nA: If there is one true observer, that would have to be me.\nQ: You? Why you?\nA: I need perspective to paint things as they appear.\nQ: Why paint things as they appear?\nQ: Why not paint things just the way they are?\nA: I can\'t. It\'s impossible.\nQ: Then you don\'t believe in objective reality?\nA: I do.\nQ: Then why don\'t you paint that?\nA: How? I can only paint what I can see.\nQ: And others see something different?\nA: Yes, unless they look from the same place.\nQ: But which view is right?\nA: Pardon?\nQ: Who sees what\'s really there?\nA: No one\'s view is more real than anyone else\'s.\nQ: So it\'s all subjective?\nQ: You can paint anything you want?\nA: No. Not if I want to paint what I see,\nA: that\'s completely objective.\nQ: But it depends on the observer?\nA: Yes.\n....\nand on and on\n....\n\nChange the question to "When does the wave function collapse?" and you have\na discussion of the interpretation of QM.\n\nThe analogy my not be perfect, but it\'s pretty good.\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch wrote:

> The trick is that there is ONE true measurement apparatus in this
> universe, namely my personal consciousness, and only mine, and it is
> my own observation which makes the collapse.

I'm sure they must have had this exact same argument after painters
discovered perspective:

Q: Where is the vanishing point?
A: It depends on the observer.
Q: What counts as an observer?
A: Not *an* observer, *the* observer.
Q: How do you tell who *the* observer is?
A: It can be anyone you want.
A: When I'm painting, it would be me.
Q: But where is the vanishing point *really* located?
A: The question doesn't even make sense without the observer.
Q: Why the hell not? All points have positions.
A: It has a position, but it depends on the observer.
Q: Perhaps there is one *true* observer that determines the
Q: where vanishing point *really* is. God perhaps?
A: If there is one true observer, that would have to be me.
Q: You? Why you?
A: I need perspective to paint things as they appear.
Q: Why paint things as they appear?
Q: Why not paint things just the way they are?
A: I can't. It's impossible.
Q: Then you don't believe in objective reality?
A: I do.
Q: Then why don't you paint that?
A: How? I can only paint what I can see.
Q: And others see something different?
A: Yes, unless they look from the same place.
Q: But which view is right?
A: Pardon?
Q: Who sees what's really there?
A: No one's view is more real than anyone else's.
Q: So it's all subjective?
Q: You can paint anything you want?
A: No. Not if I want to paint what I see,
A: that's completely objective.
Q: But it depends on the observer?
A: Yes.
....
and on and on
....

Change the question to "When does the wave function collapse?" and you have
a discussion of the interpretation of QM.

The analogy my not be perfect, but it's pretty good.

Ralph Hartley

Thomas Trotter

Oct15-04, 01:55 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<ckdinq\\$dgc\\$1@beech.fernuni-hagen.de>...\n> The point is that QM is not a realistic theory, it does not\n> describe what really happens, it allows to compute only\n> some probabilities without explaining them. But there is\n> no necessity to reject classical realism, realistic theories\n> which are compatible with observation exist\n> (Bohmian mechanics).\n\nOne must, in light of Bell\'s theorem, reject\nthe usual local realist version of the\ncoincidence curves being due to some common\nplane of polarization via paired photons\nemitted by the same atom.\n\nBut, is this because the understanding of\nthe emission and detection processes (still\nholding to locality) is incomplete? Or is\nit because there is some instantaneous\nmechanism at work?\n\nWhich would you choose as a working\nhypothetical approach?\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<ckdinq$dgc$1@beech.fernuni-hagen.de>...
> The point is that QM is not a realistic theory, it does not
> describe what really happens, it allows to compute only
> some probabilities without explaining them. But there is
> no necessity to reject classical realism, realistic theories
> which are compatible with observation exist
> (Bohmian mechanics).

One must, in light of Bell's theorem, reject
the usual local realist version of the
coincidence curves being due to some common
plane of polarization via paired photons
emitted by the same atom.

But, is this because the understanding of
the emission and detection processes (still
holding to locality) is incomplete? Or is
it because there is some instantaneous
mechanism at work?

Which would you choose as a working
hypothetical approach?

Patrick Van Esch

Oct17-04, 01:46 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nRalph Hartley <hartley@aic.nrl.navy.mil> wrote in message news:<ckouol\\$gtr\\$1@ra.nrl.navy.mil>...\n> Patrick Van Esch wrote:\n>\n> > The trick is that there is ONE true measurement apparatus in this\n> > universe, namely my personal consciousness, and only mine, and it is\n> > my own observation which makes the collapse.\n>\n> I\'m sure they must have had this exact same argument after painters\n> discovered perspective:\n[...]\n> The analogy my not be perfect, but it\'s pretty good.\n>\n\nIn fact I think your analogy is very good, and it seems to me that the\nway we have quantum theory right now gives the wave function this\nintermediate character, namely it is on one hand part of a reality,\nand on the other hand it describes the relationship between the\nobserver and nature, in a very close analogy to your description of\nperspective. I\'m not particularly keen on my explanation, but I\nwanted to point out that it solves the following issues:\n\n1) it preserves the so-cherished locality: the EPR paradox disappears\nbecause it is only upon the local comparison that the measurements\n"interfere" and produce the correlations. From the moment you\nintroduce two spacelike separated true measurements, there is no way\nin which you cannot violate locality. The only way to make sure you\ndon\'t is by having only ONE measurement.\n2) it preserves Born\'s rule, which is difficult (I think even\nimpossible) to deduce from hard relative-state interpretations.\n3) it defines exactly what *IS* a measurement, and it is (at least to\nme) somehow comforting that this quantummechanical definition of\nmeasurement (which is an ultimately subjective experience) coincides\nwith an epistemological definition of conscious observation as the\nultimate and only source of knowledge (in the sense of "I think\ntherefor I am"). After all, there is NO WAY for me to find out that\nyou, or anything else, has a consciousness. In the same way, there is\nno way for me to find out whether you "collapsed" the wavefunction, or\nwhether you only "decohered" in the same branch as I. So the hard\nproblem of consiousness in philosophy finds finally its counterpart in\nphysics. I find this attractive as an idea.\n\nBut there are two reasons to be wary of it: first of all, I think that\nit is a bad idea to do philosophy based upon a certain state of\nknowledge of physical laws ; after all, a lot of philosophy was\nuselessly inspired by a deterministic and mechanistic view of the\nuniverse a few centuries ago. And second, it lets go too much\n"reality". But it is the price I prefer to pay.\n\nSo it is only here as long as quantum mechanics in its current state\nis here, but in the mean time it gives me peace of mind with it. I\nnever found another explanation that didn\'t have a serious problem\nsomewhere which I couldn\'t accept (such as this half-baked locality,\nor the undefiniteness of what constitutes a measurement).\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Ralph Hartley <hartley@aic.nrl.navy.mil> wrote in message news:<ckouol$gtr$1@ra.nrl.navy.mil>...
> Patrick Van Esch wrote:
>
> > The trick is that there is ONE true measurement apparatus in this
> > universe, namely my personal consciousness, and only mine, and it is
> > my own observation which makes the collapse.
>
> I'm sure they must have had this exact same argument after painters
> discovered perspective:
[...]
> The analogy my not be perfect, but it's pretty good.
>

In fact I think your analogy is very good, and it seems to me that the
way we have quantum theory right now gives the wave function this
intermediate character, namely it is on one hand part of a reality,
and on the other hand it describes the relationship between the
observer and nature, in a very close analogy to your description of
perspective. I'm not particularly keen on my explanation, but I
wanted to point out that it solves the following issues:

1) it preserves the so-cherished locality: the EPR paradox disappears
because it is only upon the local comparison that the measurements
"interfere" and produce the correlations. From the moment you
introduce two spacelike separated true measurements, there is no way
in which you cannot violate locality. The only way to make sure you
don't is by having only ONE measurement.
2) it preserves Born's rule, which is difficult (I think even
impossible) to deduce from hard relative-state interpretations.
3) it defines exactly what *IS* a measurement, and it is (at least to
me) somehow comforting that this quantummechanical definition of
measurement (which is an ultimately subjective experience) coincides
with an epistemological definition of conscious observation as the
ultimate and only source of knowledge (in the sense of "I think
therefor I am"). After all, there is NO WAY for me to find out that
you, or anything else, has a consciousness. In the same way, there is
no way for me to find out whether you "collapsed" the wavefunction, or
whether you only "decohered" in the same branch as I. So the hard
problem of consiousness in philosophy finds finally its counterpart in
physics. I find this attractive as an idea.

But there are two reasons to be wary of it: first of all, I think that
it is a bad idea to do philosophy based upon a certain state of
knowledge of physical laws ; after all, a lot of philosophy was
uselessly inspired by a deterministic and mechanistic view of the
universe a few centuries ago. And second, it lets go too much
"reality". But it is the price I prefer to pay.

So it is only here as long as quantum mechanics in its current state
is here, but in the mean time it gives me peace of mind with it. I
never found another explanation that didn't have a serious problem
somewhere which I couldn't accept (such as this half-baked locality,
or the undefiniteness of what constitutes a measurement).

cheers,
Patrick.

Ralph Hartley

Oct19-04, 03:29 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\nIlja Schmelzer wrote:\n\n> Seems you have not understood the point of this thought\n> experiment.\n\nThat is correct, but I think I\'m starting to understand now.\n\n> It is about a hypothetical world where FTL phones\n> exist, and hypothetical scientists who develop a hypothetical\n> mathematical theory which allows to compute the hypothetical\n> observable probabilities observed in the FTL phone.\n>\n> The thesis is that it is possible to develop such a theory.\n> Then, the hypothetical situation is almost analoguous to the\n> current situation with QM.\n>\n> The only difference is that the hypothetical theory allows the\n> probabilities obbservable in FTL phones, but QM doesn\'t.\n\nBut that is an *important* difference. That theory wouldn\'t have the\nlimitations on causality (expressed statistically) that our theory has, so\nin that hypothetical world, Einstein causality *would* be falsified.\n\n> Thesis: The existence of a nonrealistic mathematical theory\n> which allows to compute the observable probabilies proves\n> nothing about the local or nonlocal character of the observed\n> effects.\n\nThat is true, but the existence of a *local* (in the proper sense, QM is\nlocal) nonrealistic mathematical theory which allows to compute the\nobservable probabilities proves *something* about the local or nonlocal\ncharacter of the observed effects.\n\nWhat matters is that QM has a local interpretation, it is irrelevent that\nit also has a nonlocal one.\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Ilja Schmelzer wrote:

> Seems you have not understood the point of this thought
> experiment.

That is correct, but I think I'm starting to understand now.

> It is about a hypothetical world where FTL phones
> exist, and hypothetical scientists who develop a hypothetical
> mathematical theory which allows to compute the hypothetical
> observable probabilities observed in the FTL phone.
>
> The thesis is that it is possible to develop such a theory.
> Then, the hypothetical situation is almost analoguous to the
> current situation with QM.
>
> The only difference is that the hypothetical theory allows the
> probabilities obbservable in FTL phones, but QM doesn't.

But that is an *important* difference. That theory wouldn't have the
limitations on causality (expressed statistically) that our theory has, so
in that hypothetical world, Einstein causality *would* be falsified.

> Thesis: The existence of a nonrealistic mathematical theory
> which allows to compute the observable probabilies proves
> nothing about the local or nonlocal character of the observed
> effects.

That is true, but the existence of a *local* (in the proper sense, QM is
local) nonrealistic mathematical theory which allows to compute the
observable probabilities proves *something* about the local or nonlocal
character of the observed effects.

What matters is that QM has a local interpretation, it is irrelevent that
it also has a nonlocal one.

Ralph Hartley

Arnold Neumaier

Oct19-04, 03:29 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nPatrick Van Esch wrote:\n> Ralph Hartley <hartley@aic.nrl.navy.mil> wrote in message news:<ckouol\\$gtr\\$1@ra.nrl.navy.mil>...\n>\n>>P atrick Van Esch wrote:\n>>\n>>>The trick is that there is ONE true measurement apparatus in this\n>>>universe, namely my personal consciousness, and only mine, and it is\n>>>my own observation which makes the collapse.\n>>\n>>I\'m sure they must have had this exact same argument after painters\n>>discovered perspective:\n>\n> 3) it defines exactly what *IS* a measurement, and it is (at least to\n> me) somehow comforting that this quantummechanical definition of\n> measurement (which is an ultimately subjective experience) coincides\n> with an epistemological definition of conscious observation as the\n> ultimate and only source of knowledge (in the sense of "I think\n> therefor I am"). After all, there is NO WAY for me to find out that\n> you, or anything else, has a consciousness.\n\nHow does your consciousness make measurements, in an exactly defined way??\n\nIn my opinion, consciousness is a very unreliable measurement apparatus,\nwhich has even difficulties detecting the difference between a dream and\nreality.\n\nStandards of physics are today such that subjective aspects like\nconsciousness are excluded as much as possible from measurement,\nto get objective data.\n\nMost large-scale measurements are done in a completely automatic way,\nwhere human observers play no role at all except for discussing\nautomatically produced summaries of the data...\n\n\nArnold Neumaier\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch wrote:
> Ralph Hartley <hartley@aic.nrl.navy.mil> wrote in message news:<ckouol$gtr$1@ra.nrl.navy.mil>...
>
>>Patrick Van Esch wrote:
>>
>>>The trick is that there is ONE true measurement apparatus in this
>>>universe, namely my personal consciousness, and only mine, and it is
>>>my own observation which makes the collapse.
>>
>>I'm sure they must have had this exact same argument after painters
>>discovered perspective:
>
> 3) it defines exactly what *IS* a measurement, and it is (at least to
> me) somehow comforting that this quantummechanical definition of
> measurement (which is an ultimately subjective experience) coincides
> with an epistemological definition of conscious observation as the
> ultimate and only source of knowledge (in the sense of "I think
> therefor I am"). After all, there is NO WAY for me to find out that
> you, or anything else, has a consciousness.

How does your consciousness make measurements, in an exactly defined way??

In my opinion, consciousness is a very unreliable measurement apparatus,
which has even difficulties detecting the difference between a dream and
reality.

Standards of physics are today such that subjective aspects like
consciousness are excluded as much as possible from measurement,
to get objective data.

Most large-scale measurements are done in a completely automatic way,
where human observers play no role at all except for discussing
automatically produced summaries of the data...

Arnold Neumaier

Ilja Schmelzer

Oct19-04, 03:29 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Patrick Van Esch" <vanesch@ill.fr> schrieb\n> 1) it preserves the so-cherished locality: the EPR paradox disappears\n> because it is only upon the local comparison that the measurements\n> "interfere" and produce the correlations. From the moment you\n> introduce two spacelike separated true measurements, there is no way\n> in which you cannot violate locality. The only way to make sure you\n> don\'t is by having only ONE measurement.\n\nNo, it doesn\'t. The "explanation" is a pseudo-explanation,\nbecause it would allow you to "explain away" even a working\nFTL phone. Indeed, if you have only one measurement, you\ncannot prove that an FTL phone really is an FTL phone.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Patrick Van Esch" <vanesch@ill.fr> schrieb
> 1) it preserves the so-cherished locality: the EPR paradox disappears
> because it is only upon the local comparison that the measurements
> "interfere" and produce the correlations. From the moment you
> introduce two spacelike separated true measurements, there is no way
> in which you cannot violate locality. The only way to make sure you
> don't is by having only ONE measurement.

No, it doesn't. The "explanation" is a pseudo-explanation,
because it would allow you to "explain away" even a working
FTL phone. Indeed, if you have only one measurement, you
cannot prove that an FTL phone really is an FTL phone.

Ilja

Ilja Schmelzer

Oct19-04, 03:29 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Thomas Trotter" <thomastrotter2005@juno.com> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> One must, in light of Bell\'s theorem, reject\n> the usual local realist version ...\n> But, is this because the understanding of\n> the emission and detection processes (still\n> holding to locality) is incomplete? Or is\n> it because there is some instantaneous\n> mechanism at work?\n>\n> Which would you choose as a working\n> hypothetical approach?\n\nI would choose nonlocality, as it appears\nin Bohmian mechanics or Nelsonian stochastics.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Thomas Trotter" <thomastrotter2005@juno.com> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> One must, in light of Bell's theorem, reject
> the usual local realist version ...
> But, is this because the understanding of
> the emission and detection processes (still
> holding to locality) is incomplete? Or is
> it because there is some instantaneous
> mechanism at work?
>
> Which would you choose as a working
> hypothetical approach?

I would choose nonlocality, as it appears
in Bohmian mechanics or Nelsonian stochastics.

Ilja

Ilja Schmelzer

Oct19-04, 03:29 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> ><rof@maths.tcd.ie> schrieb\n> >> For me, nonlocality itself is rather odd.\n\n> >Not that odd. Last not least, Newtonian mechanics is\n> >nonlocal too.\n>\n> True; it\'s relativity which makes nonlocality really strange.\n\nIn a Lorentz ether there is nothing strange with nonlocality\ntoo.\n\n> >> If the world is nonlocal, then how does locality emerge?\n>\n> >As the rather small speed of sound emerges in a world\n> >where the speed of light is the limiting velocity.\n\n> There\'s obviously a bit more to explain than that (decoherence\n> would need to be appealed to before one can even talk about\n> individual objects which could even have a well-defined or\n> approximately well-defined velocity).\n\nNo necessity. Use Bohmian mechanics.\n\n> It also seems circular\n> to start with a spatial description (even a nonlocal one) and\n> then deduce later that a spatial description is appropriate.\n\nI don\'t understand. A theory always starts with some postulates.\nThen it has to derive something observable. If observation\nsupports this predictions, the postulates seem appropriate.\n\n> A more honest approach would be to start with an abstract\n> description\n> which didn\'t have space in it anywhere and then deduce that\n> a spatial representation is appropriate for describing some\n> coarse-grained features.\n\nAgain, I don\'t see why Newtonian mechanics is "dishonest"\nif it postulates an Euclidean space.\n\n> >Look at it from another point. Imagine some world. Imagine\n> >some internal creatures of this world. Is it obvious that these\n> >internal observers are able to distinguish by observation all\n> >situations which are really different? Not at all. I would even\n> >say that in some sense it is impossible at all.\n>\n> >Thus, there will be different situations which cannot be\n> >distinguished by observation. Internal relativists will\n> >certainly fail to describe their world correctly.\n>\n> Here is the nasty face of realism. If two experimenters get\n> the same results for all experiments, and use the same\n> (non-stochastic or non-opinionated) reasoning, then they\n> must inevitably come to the same description. What you\n> are saying is that their descriptions will probably be wrong,\n> and, if the situation is as you described it, then we have\n> to face the fact that our descriptions are, in the same\n> sense, probably wrong, including yours.\n\nYou forget that their guesses about the laws of nature may\nbe correct, and their conclusions about the actual initial values\nmay be incomplete. If they tell us "unfortunately, we are not able\nto measure our absolute velocity against the ether" they may not\nbe wrong.\n\n> >A symmetry group for observables is, indeed, a powerful argument\n> >for proposing the same symmetry group as fundamental. But if\n> >we have the following two things:\n>\n> >1.) There is no fundamental theory with this symmetry group\n> >(the failure of quantum GR)\n>\n> >2.) The appearence of this symmetry group as an approximate\n> >symmetry of some (large distance) limit is quite natural, that means,\n> >there exists theories with different fundamental symmetry which\n> >have the observable symmetry in some limit\n>\n> >then I would no longer say that the argument is strong.\n>\n> Right, but the lack of a fundamental theory isn\'t quite\n> the failure of quantum gravity. Instead it\'s the failure\n> of the mathematical machinery of normal quantum mechanics in\n> general, and (special) relativistic quantum mechanics in\n> particular, to incorporate information about the actual\n> results of measurements.\n\nThat\'s your guess about the reasons for failure. I disagree.\nIn Bohmian mechanics, the "results of measurements" are\nwell incorporated, but the mathematical machinery is\nessentially the same.\n\n> But I think we can keep the symmetry group by abandoning\n> determinism without abandoning realism (although this\n> is not a position that I advocate).\n\nNo. The proof of Bell\'s inequality is based on Einstein-causal\nrealism, not determinism.\n\n> >> There may also be independent reasons to abandon realism, but\n> >> discussions of these quickly degenerate into nonsense.\n>\n> >There may be but at least I don\'t know any\n> >independent reasons.\n>\n> Rather than get deeply into the realism debate (a discussion of\n> which would be off-topic for spr), I\'ll say that what physicists\n> mean by real is actually a confused mixture of several distinct\n> ideas which can be useful on their own, but are not useful when mixed\n> together in this way. The concepts mixed together are something\n> like: externality (as in "the external world"), truth (real\n> as oppposed to unreal), fundamental as opposed to derivative,\n> independence of which individual observer makes the observation,\n> independence of the act of observation altogether, referring\n> to a property of an object as opposed to a property of the\n> relationship between the object and the measuring device, and\n> finally the difference between something which is actually\n> present and something which is merely referred to.\n>\n> When these are mixed together into a single concept of\n> reality, we get confusion, such as "Real pain is different\n> from the mere thought of pain, but isn\'t real, since it\'s not\n> external."\n\nI prefare a rather simple and well-defined concept of realism.\nIt contains the assumptions which, together with Einstein causality,\nare necessary to prove Bell\'s inequality.\n\nThat\'s not much and well-defined.\n\n> I suspect the criterion for an explanation to\n> be \'realistic\' isn\'t very clearly defined.\n\nYou need some set of "possible realities" Lambda\nwith a probability distribution rho(lambda) on it.\nThen you need a function X(x,lambda) which defines\nthe outcome of the experiment X as depending from\nthe state of reality lamdba and the decisions of the\nexperimenters x. Then the probability of f(X) is\n\n<f(X)> = int f(X(x,lambda)) rho(lambda) d lambda.\n\nIts clearly defined, not?\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> ><rof@maths.tcd.ie> schrieb
> >> For me, nonlocality itself is rather odd.

> >Not that odd. Last not least, Newtonian mechanics is
> >nonlocal too.
>
> True; it's relativity which makes nonlocality really strange.

In a Lorentz ether there is nothing strange with nonlocality
too.

> >> If the world is nonlocal, then how does locality emerge?
>
> >As the rather small speed of sound emerges in a world
> >where the speed of light is the limiting velocity.

> There's obviously a bit more to explain than that (decoherence
> would need to be appealed to before one can even talk about
> individual objects which could even have a well-defined or
> approximately well-defined velocity).

No necessity. Use Bohmian mechanics.

> It also seems circular
> to start with a spatial description (even a nonlocal one) and
> then deduce later that a spatial description is appropriate.

I don't understand. A theory always starts with some postulates.
Then it has to derive something observable. If observation
supports this predictions, the postulates seem appropriate.

> A more honest approach would be to start with an abstract
> description
> which didn't have space in it anywhere and then deduce that
> a spatial representation is appropriate for describing some
> coarse-grained features.

Again, I don't see why Newtonian mechanics is "dishonest"
if it postulates an Euclidean space.

> >Look at it from another point. Imagine some world. Imagine
> >some internal creatures of this world. Is it obvious that these
> >internal observers are able to distinguish by observation all
> >situations which are really different? Not at all. I would even
> >say that in some sense it is impossible at all.
>
> >Thus, there will be different situations which cannot be
> >distinguished by observation. Internal relativists will
> >certainly fail to describe their world correctly.
>
> Here is the nasty face of realism. If two experimenters get
> the same results for all experiments, and use the same
> (non-stochastic or non-opinionated) reasoning, then they
> must inevitably come to the same description. What you
> are saying is that their descriptions will probably be wrong,
> and, if the situation is as you described it, then we have
> to face the fact that our descriptions are, in the same
> sense, probably wrong, including yours.

You forget that their guesses about the laws of nature may
be correct, and their conclusions about the actual initial values
may be incomplete. If they tell us "unfortunately, we are not able
to measure our absolute velocity against the ether" they may not
be wrong.

> >A symmetry group for observables is, indeed, a powerful argument
> >for proposing the same symmetry group as fundamental. But if
> >we have the following two things:
>
> >1.) There is no fundamental theory with this symmetry group
> >(the failure of quantum GR)
>
> >2.) The appearence of this symmetry group as an approximate
> >symmetry of some (large distance) limit is quite natural, that means,
> >there exists theories with different fundamental symmetry which
> >have the observable symmetry in some limit
>
> >then I would no longer say that the argument is strong.
>
> Right, but the lack of a fundamental theory isn't quite
> the failure of quantum gravity. Instead it's the failure
> of the mathematical machinery of normal quantum mechanics in
> general, and (special) relativistic quantum mechanics in
> particular, to incorporate information about the actual
> results of measurements.

That's your guess about the reasons for failure. I disagree.
In Bohmian mechanics, the "results of measurements" are
well incorporated, but the mathematical machinery is
essentially the same.

> But I think we can keep the symmetry group by abandoning
> determinism without abandoning realism (although this
> is not a position that I advocate).

No. The proof of Bell's inequality is based on Einstein-causal
realism, not determinism.

> >> There may also be independent reasons to abandon realism, but
> >> discussions of these quickly degenerate into nonsense.
>
> >There may be but at least I don't know any
> >independent reasons.
>
> Rather than get deeply into the realism debate (a discussion of
> which would be off-topic for spr), I'll say that what physicists
> mean by real is actually a confused mixture of several distinct
> ideas which can be useful on their own, but are not useful when mixed
> together in this way. The concepts mixed together are something
> like: externality (as in "the external world"), truth (real
> as oppposed to unreal), fundamental as opposed to derivative,
> independence of which individual observer makes the observation,
> independence of the act of observation altogether, referring
> to a property of an object as opposed to a property of the
> relationship between the object and the measuring device, and
> finally the difference between something which is actually
> present and something which is merely referred to.
>
> When these are mixed together into a single concept of
> reality, we get confusion, such as "Real pain is different
> from the mere thought of pain, but isn't real, since it's not
> external."

I prefare a rather simple and well-defined concept of realism.
It contains the assumptions which, together with Einstein causality,
are necessary to prove Bell's inequality.

That's not much and well-defined.

> I suspect the criterion for an explanation to
> be 'realistic' isn't very clearly defined.

You need some set of "possible realities" \Lambda
with a probability distribution \rho(\lambda) on it.
Then you need a function X(x,\lambda) which defines
the outcome of the experiment X as depending from
the state of reality lamdba and the decisions of the
experimenters x. Then the probability of f(X) is

<f(X)> = \int f(X(x,\lambda)) \rho(\lambda) d \lambda.

Its clearly defined, not?

Ilja

Ilja Schmelzer

Oct20-04, 03:18 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb\n> Ilja Schmelzer wrote:\n> > The only difference is that the hypothetical theory allows the\n> > probabilities observable in FTL phones, but QM doesn\'t.\n\n> But that is an *important* difference.\n\nIndeed. But important for what?\n\n> That theory wouldn\'t have the\n> limitations on causality (expressed statistically) that our theory\n> has, so in that hypothetical world, Einstein causality\n> *would* be falsified.\n\nThat\'s the open question.\n\nI claim the main difference is the following: QM\nallows FTL only in such a way that we observe\n"A->B or B->A" but cannot decide which\nalternative happens, while the other theory\nallows to observe the direction of FTL causal\ninfluences A->B.\n\nIn this interpretation, the only difference is the\ndifference between direct and indirect observation.\n\nAnd Einstein causality is falsified in above theories.\n>From A->B => Einstein causality falsified\nand B->A => Einstein causality falsified\nfollows\n"A->B or B->A" => Einstein causality falsified.\n\n> > Thesis: The existence of a nonrealistic mathematical theory\n> > which allows to compute the observable probabilies proves\n> > nothing about the local or nonlocal character of the observed\n> > effects.\n\n> That is true, but the existence of a *local* (in the proper sense, QM is\n> local) nonrealistic mathematical theory which allows to compute the\n> observable probabilities proves *something* about the local or nonlocal\n> character of the observed effects.\n\nOk, it proves *something*. Nice. An interesting\nproperty of QM. But this does not save\nEinstein causality.\n\n> What matters is that QM has a local interpretation, it is irrelevent that\n> it also has a nonlocal one.\n\nQM does not have any local realistic interpretation.\nMy thesis: whatever you name "local interpretation"\nis a mixture of the difference I have mentioned\n("A->B or B->A" in comparison with "A->B") and\nthings which can be done analogically in above theories.\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ralph Hartley" <hartley@aic.nrl.navy.mil> schrieb
> Ilja Schmelzer wrote:
> > The only difference is that the hypothetical theory allows the
> > probabilities observable in FTL phones, but QM doesn't.

> But that is an *important* difference.

Indeed. But important for what?

> That theory wouldn't have the
> limitations on causality (expressed statistically) that our theory
> has, so in that hypothetical world, Einstein causality
> *would* be falsified.

That's the open question.

I claim the main difference is the following: QM
allows FTL only in such a way that we observe
"A->B or B->A" but cannot decide which
alternative happens, while the other theory
allows to observe the direction of FTL causal
influences A->B.

In this interpretation, the only difference is the
difference between direct and indirect observation.

And Einstein causality is falsified in above theories.
>From A->B => Einstein causality falsified
and B->A => Einstein causality falsified
follows
"A->B or B->A" => Einstein causality falsified.

> > Thesis: The existence of a nonrealistic mathematical theory
> > which allows to compute the observable probabilies proves
> > nothing about the local or nonlocal character of the observed
> > effects.

> That is true, but the existence of a *local* (in the proper sense, QM is
> local) nonrealistic mathematical theory which allows to compute the
> observable probabilities proves *something* about the local or nonlocal
> character of the observed effects.

Ok, it proves *something*. Nice. An interesting
property of QM. But this does not save
Einstein causality.

> What matters is that QM has a local interpretation, it is irrelevent that
> it also has a nonlocal one.

QM does not have any local realistic interpretation.
My thesis: whatever you name "local interpretation"
is a mixture of the difference I have mentioned
("A->B or B->A" in comparison with "A->B") and
things which can be done analogically in above theories.

Ilja

Patrick Van Esch

Oct20-04, 03:18 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<ckvvm1\\$3k2\\$1@beech.fernuni-hagen.de>...\n> "Patrick Van Esch" <vanesch@ill.fr> schrieb\n> > 1) it preserves the so-cherished locality:\n\n[...]\n\n>\n> No, it doesn\'t. The "explanation" is a pseudo-explanation,\n> because it would allow you to "explain away" even a working\n> FTL phone. Indeed, if you have only one measurement, you\n> cannot prove that an FTL phone really is an FTL phone.\n\nNot really. A working FTL phone would allow me, as a single observer,\nto conclude that the signal I sent went up and down Titan in 0.5\nseconds. After all, there is no quantum interference working here,\nthis can be considered classically (or semiclassically, in that a\nnarrow wavepacket of states went up there and came back).\n\nHowever, the interference of results in an EPR like setup are not more\nsurprising than the interference of a split light beam that went\naround the building left and right and came back. What is surprising\nin EPR is that if you consider that there has been a measurement at\nAlice\'s and Bob\'s, then there ought to be a breaking of locality (I\nknow you are in favor of that conclusion). But then it is surprising\nthat we cannot build an FTL phone using that procedure. We can only\nnotice the effect when, through normal, sub-luminal signalling, we\nhave access to both measurements. So what is the conspiracy ? I\nthink the point is that there hasn\'t been a measurement at Alice\'s and\nBob\'s, but that there is only an interference of results after they\nhave been transported (in superposed state) in a locality-preserving,\nsubluminal way up to the common point where they can be compared.\nSuch an interference (like the interference of split light beams)\nexplains these results, and also explains why you cannot build an FTL\nphone using the idea.\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<ckvvm1$3k2$1@beech.fernuni-hagen.de>...
> "Patrick Van Esch" <vanesch@ill.fr> schrieb
> > 1) it preserves the so-cherished locality:

[...]

>
> No, it doesn't. The "explanation" is a pseudo-explanation,
> because it would allow you to "explain away" even a working
> FTL phone. Indeed, if you have only one measurement, you
> cannot prove that an FTL phone really is an FTL phone.

Not really. A working FTL phone would allow me, as a single observer,
to conclude that the signal I sent went up and down Titan in .5
seconds. After all, there is no quantum interference working here,
this can be considered classically (or semiclassically, in that a
narrow wavepacket of states went up there and came back).

However, the interference of results in an EPR like setup are not more
surprising than the interference of a split light beam that went
around the building left and right and came back. What is surprising
in EPR is that if you consider that there has been a measurement at
Alice's and Bob's, then there ought to be a breaking of locality (I
know you are in favor of that conclusion). But then it is surprising
that we cannot build an FTL phone using that procedure. We can only
notice the effect when, through normal, sub-luminal signalling, we
have access to both measurements. So what is the conspiracy ? I
think the point is that there hasn't been a measurement at Alice's and
Bob's, but that there is only an interference of results after they
have been transported (in superposed state) in a locality-preserving,
subluminal way up to the common point where they can be compared.
Such an interference (like the interference of split light beams)
explains these results, and also explains why you cannot build an FTL
phone using the idea.

cheers,
Patrick.

Patrick Van Esch

Oct20-04, 03:18 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nArnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in message news:<4173C2A3.7010000@univie.ac.at>...\n>\n> How does your consciousness make measurements, in an exactly defined way??\n\nWhat else, except for conscious observation, constitutes a measurement\n?\n\n> Most large-scale measurements are done in a completely automatic way,\n> where human observers play no role at all except for discussing\n> automatically produced summaries of the data...\n\n\nThe measurement then consists in reading the paper in Physical review.\nConsider Physical review to be in a superposed state until I read it\n:-)\n\nBTW, I know this sounds crazy. It probably even is, and I\'m not\nparticularly attached to this view. I just wanted to say that it is\nthe way I found around the conflict between locality, the measurement\nproblem, EPR states and so on. As I said, to me it lacks too much\n"reality".\nBut I have the following boundary conditions:\n- somehow "locality" still seems to be a basic principle\n- EPR seems to violate locality, but in such a way that it doesn\'t\nallow for local observers to notice, until they got news from their\npeers.\n- I think that there are fundamental problems with a hard many-worlds\nscheme (in that at some point you\'ll need to introduce a true\nBorn-style measurement)\n\nThe view I expressed seems to be compatible with these issues. I\'m\nnot saying there aren\'t other ways around it.\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in message news:<4173C2A3.7010000@univie.ac.at>...
>
> How does your consciousness make measurements, in an exactly defined way??

What else, except for conscious observation, constitutes a measurement
?

> Most large-scale measurements are done in a completely automatic way,
> where human observers play no role at all except for discussing
> automatically produced summaries of the data...

The measurement then consists in reading the paper in Physical review.
Consider Physical review to be in a superposed state until I read it
:-)

BTW, I know this sounds crazy. It probably even is, and I'm not
particularly attached to this view. I just wanted to say that it is
the way I found around the conflict between locality, the measurement
problem, EPR states and so on. As I said, to me it lacks too much
"reality".
But I have the following boundary conditions:
- somehow "locality" still seems to be a basic principle
- EPR seems to violate locality, but in such a way that it doesn't
allow for local observers to notice, until they got news from their
peers.
- I think that there are fundamental problems with a hard many-worlds
scheme (in that at some point you'll need to introduce a true
Born-style measurement)

The view I expressed seems to be compatible with these issues. I'm
not saying there aren't other ways around it.

cheers,
Patrick.

rof@maths.tcd.ie

Oct20-04, 03:18 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n\n>> True; it\'s relativity which makes nonlocality really strange.\n\n>In a Lorentz ether there is nothing strange with nonlocality\n>too.\n\nNor is there anything strange about FTL phones, apparently.\n\n>> >> If the world is nonlocal, then how does locality emerge?\n>>\n>> >As the rather small speed of sound emerges in a world\n>> >where the speed of light is the limiting velocity.\n\n>> There\'s obviously a bit more to explain than that (decoherence\n>> would need to be appealed to before one can even talk about\n>> individual objects which could even have a well-defined or\n>> approximately well-defined velocity).\n\n>No necessity. Use Bohmian mechanics.\n\n>> It also seems circular\n>> to start with a spatial description (even a nonlocal one) and\n>> then deduce later that a spatial description is appropriate.\n\n>I don\'t understand. A theory always starts with some postulates.\n>Then it has to derive something observable. If observation\n>supports this predictions, the postulates seem appropriate.\n\nThis is related to the Bohmian point above. A theory has some\naxioms or postulates. These are either postulates about what\nthe results of certain measurements will be (in which case it\nis an "epistemological" theory), or the postulates are about\nwhat "really exists" (an "ontological" theory). There is a challenge\nwhich ontological theories must meet which epistemological\nones don\'t need to. That is to say what will be observed. There\'s\nno guarantee that if we start from an ontological theory which\nhas some kind of spatial manifold as one of its features then\nthe little creatures living in the theory will observe that\nmanifold.\n\nThe observer in the theory has access only to his own internal\nstates (if we restrict our attention to theories without\nextra-sensory perception). From the modifications to his internal\nstate, he is supposed to generate a picture of the world outside\nhimself. Suppose, for example, in discrete time, he receives\nthe "symbol" E_n at time t_n. In order to predict what will\nhappen at the next timestep t_{n+1}, he will need to reflect\non many of the previous symbols he has received, E_{n-1}, E_{n-2} ...\n\nThis looks like an N-step Markov process, with the probability\nof receiving a symbol at t_n being dependent on the N symbols\nreceived before this time. To generate a representation of a\nworld around him, in which the next symbol he receives is\nconsidered to be determined (perhaps probabilistically) by the\n"present" state of the world, rather than by the history of\nsymbols he has received, he must convert this N-step process to\na one-step Markov process.\n\nThat is, he must form a data representation which compresses\nthe history of symbols he has received into a single "state\nof the world", call it X_n, where\nP(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)\n(something which was not true of the symbols E_n).\n\nNow, reception of the symbols E_n is called sensation; the\nsubsequent construction of the X_n is called perception,\nand the transition probabilities P(X_{n+1}|X_n) are called\nthe laws of physics (as far as this observer is concerned).\n\nAn epistemological theory deals with the same stuff as the\nobserver. In an ontological theory which merely postulates\nthe existence of space, it is necessary (before one can say\nthat it predicts that space will be observed) to describe the\ninternal structure of the observer, catalogue the E_n\'s and the\nN-step transition probabilities P(E_{n+1}|E_n, E_{n-1}, ...), and\nthen show that it admits a representation X_n which splits up\nX_n = \\prod_a x^a_n, where a indexes local "positions" and x^a_n\nis the "mini-state of affairs" at position a, along\nwith some kind of locality. The locality is necessary because\nthere\'s no point in representing positions a and b as being\nfar away from each other if the state of affairs at b at time\nn+1 always depends strongly on the state of affairs at a at time\nn. Hence locality need not necessarily be strict, but it should\nbe "usual", in the sense that it\'s useful to distinguish between\none place and another.\n\nBohmian mechanics can, I think, possibly achieve this, but it\'s\nnot obvious, and it hasn\'t been proven.\n\n>> A more honest approach would be to start with an abstract\n>> description\n>> which didn\'t have space in it anywhere and then deduce that\n>> a spatial representation is appropriate for describing some\n>> coarse-grained features.\n\n>Again, I don\'t see why Newtonian mechanics is "dishonest"\n>if it postulates an Euclidean space.\n\nNewtonian mechanics is descriptive of the world that we see - that\nis, the perceived world, and doesn\'t pretend to describe something\nunderneath it from which the perceived world emerges.\n\n>> >A symmetry group for observables is, indeed, a powerful argument\n>> >for proposing the same symmetry group as fundamental. But if\n>> >we have the following two things:\n>>\n>> >1.) There is no fundamental theory with this symmetry group\n>> >(the failure of quantum GR)\n>>\n>> >2.) The appearence of this symmetry group as an approximate\n>> >symmetry of some (large distance) limit is quite natural, that means,\n>> >there exists theories with different fundamental symmetry which\n>> >have the observable symmetry in some limit\n>>\n>> >then I would no longer say that the argument is strong.\n>>\n>> Right, but the lack of a fundamental theory isn\'t quite\n>> the failure of quantum gravity. Instead it\'s the failure\n>> of the mathematical machinery of normal quantum mechanics in\n>> general, and (special) relativistic quantum mechanics in\n>> particular, to incorporate information about the actual\n>> results of measurements.\n\n>That\'s your guess about the reasons for failure. I disagree.\n>In Bohmian mechanics, the "results of measurements" are\n>well incorporated, but the mathematical machinery is\n>essentially the same.\n\nWhich is why there isn\'t a special relativistic Bohmian\nmechanics.\n\n>> But I think we can keep the symmetry group by abandoning\n>> determinism without abandoning realism (although this\n>> is not a position that I advocate).\n\n>No. The proof of Bell\'s inequality is based on Einstein-causal\n>realism, not determinism.\n\nI was talking about keeping the symmetry group (the Lorentz group) -\nnot about keeping locality. Nonlocality has been experimentally\nconfirmed, so there\'s no point in trying to salvage locality.\n\n>I prefare a rather simple and well-defined concept of realism.\n>It contains the assumptions which, together with Einstein causality,\n>are necessary to prove Bell\'s inequality.\n\n>That\'s not much and well-defined.\n\n>> I suspect the criterion for an explanation to\n>> be \'realistic\' isn\'t very clearly defined.\n\n>You need some set of "possible realities" Lambda\n>with a probability distribution rho(lambda) on it.\n>Then you need a function X(x,lambda) which defines\n>the outcome of the experiment X as depending from\n>the state of reality lamdba and the decisions of the\n>experimenters x. Then the probability of f(X) is\n\n><f(X)> = int f(X(x,lambda)) rho(lambda) d lambda.\n\n>Its clearly defined, not?\n\nThis looks more like a definition of determinism. Do\nyou consider determinism and realism to be the same thing?\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb

>> True; it's relativity which makes nonlocality really strange.

>In a Lorentz ether there is nothing strange with nonlocality
>too.

Nor is there anything strange about FTL phones, apparently.

>> >> If the world is nonlocal, then how does locality emerge?
>>
>> >As the rather small speed of sound emerges in a world
>> >where the speed of light is the limiting velocity.

>> There's obviously a bit more to explain than that (decoherence
>> would need to be appealed to before one can even talk about
>> individual objects which could even have a well-defined or
>> approximately well-defined velocity).

>No necessity. Use Bohmian mechanics.

>> It also seems circular
>> to start with a spatial description (even a nonlocal one) and
>> then deduce later that a spatial description is appropriate.

>I don't understand. A theory always starts with some postulates.
>Then it has to derive something observable. If observation
>supports this predictions, the postulates seem appropriate.

This is related to the Bohmian point above. A theory has some
axioms or postulates. These are either postulates about what
the results of certain measurements will be (in which case it
is an "epistemological" theory), or the postulates are about
what "really exists" (an "ontological" theory). There is a challenge
which ontological theories must meet which epistemological
ones don't need to. That is to say what will be observed. There's
no guarantee that if we start from an ontological theory which
has some kind of spatial manifold as one of its features then
the little creatures living in the theory will observe that
manifold.

The observer in the theory has access only to his own internal
states (if we restrict our attention to theories without
extra-sensory perception). From the modifications to his internal
state, he is supposed to generate a picture of the world outside
himself. Suppose, for example, in discrete time, he receives
the "symbol" E_n at time t_n. In order to predict what will
happen at the next timestep t_{n+1}, he will need to reflect
on many of the previous symbols he has received, E_{n-1}, E_{n-2} ...

This looks like an N-step Markov process, with the probability
of receiving a symbol at t_n being dependent on the N symbols
received before this time. To generate a representation of a
world around him, in which the next symbol he receives is
considered to be determined (perhaps probabilistically) by the
"present" state of the world, rather than by the history of
symbols he has received, he must convert this N-step process to
a one-step Markov process.

That is, he must form a data representation which compresses
the history of symbols he has received into a single "state
of the world", call it X_n, where
P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)
(something which was not true of the symbols E_n).

Now, reception of the symbols E_n is called sensation; the
subsequent construction of the X_n is called perception,
and the transition probabilities P(X_{n+1}|X_n) are called
the laws of physics (as far as this observer is concerned).

An epistemological theory deals with the same stuff as the
observer. In an ontological theory which merely postulates
the existence of space, it is necessary (before one can say
that it predicts that space will be observed) to describe the
internal structure of the observer, catalogue the E_n's and the
N-step transition probabilities P(E_{n+1}|E_n, E_{n-1}, ...), and
then show that it admits a representation X_n which splits up
X_n = \prod_a x^{a_n}, where a indexes local "positions" and x^{a_n}
is the "mini-state of affairs" at position a, along
with some kind of locality. The locality is necessary because
there's no point in representing positions a and b as being
far away from each other if the state of affairs at b at time
n+1 always depends strongly on the state of affairs at a at time
n. Hence locality need not necessarily be strict, but it should
be "usual", in the sense that it's useful to distinguish between
one place and another.

Bohmian mechanics can, I think, possibly achieve this, but it's
not obvious, and it hasn't been proven.

>> A more honest approach would be to start with an abstract
>> description
>> which didn't have space in it anywhere and then deduce that
>> a spatial representation is appropriate for describing some
>> coarse-grained features.

>Again, I don't see why Newtonian mechanics is "dishonest"
>if it postulates an Euclidean space.

Newtonian mechanics is descriptive of the world that we see - that
is, the perceived world, and doesn't pretend to describe something
underneath it from which the perceived world emerges.

>> >A symmetry group for observables is, indeed, a powerful argument
>> >for proposing the same symmetry group as fundamental. But if
>> >we have the following two things:
>>
>> >1.) There is no fundamental theory with this symmetry group
>> >(the failure of quantum GR)
>>
>> >2.) The appearence of this symmetry group as an approximate
>> >symmetry of some (large distance) limit is quite natural, that means,
>> >there exists theories with different fundamental symmetry which
>> >have the observable symmetry in some limit
>>
>> >then I would no longer say that the argument is strong.
>>
>> Right, but the lack of a fundamental theory isn't quite
>> the failure of quantum gravity. Instead it's the failure
>> of the mathematical machinery of normal quantum mechanics in
>> general, and (special) relativistic quantum mechanics in
>> particular, to incorporate information about the actual
>> results of measurements.

>That's your guess about the reasons for failure. I disagree.
>In Bohmian mechanics, the "results of measurements" are
>well incorporated, but the mathematical machinery is
>essentially the same.

Which is why there isn't a special relativistic Bohmian
mechanics.

>> But I think we can keep the symmetry group by abandoning
>> determinism without abandoning realism (although this
>> is not a position that I advocate).

>No. The proof of Bell's inequality is based on Einstein-causal
>realism, not determinism.

I was talking about keeping the symmetry group (the Lorentz group) -
not about keeping locality. Nonlocality has been experimentally
confirmed, so there's no point in trying to salvage locality.

>I prefare a rather simple and well-defined concept of realism.
>It contains the assumptions which, together with Einstein causality,
>are necessary to prove Bell's inequality.

>That's not much and well-defined.

>> I suspect the criterion for an explanation to
>> be 'realistic' isn't very clearly defined.

>You need some set of "possible realities" \Lambda
>with a probability distribution \rho(\lambda) on it.
>Then you need a function X(x,\lambda) which defines
>the outcome of the experiment X as depending from
>the state of reality lamdba and the decisions of the
>experimenters x. Then the probability of f(X) is

><f(X)> = \int f(X(x,\lambda)) \rho(\lambda) d \lambda.

>Its clearly defined, not?

This looks more like a definition of determinism. Do
you consider determinism and realism to be the same thing?

R.

Arnold Neumaier

Oct20-04, 10:48 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nPatrick Van Esch wrote:\n> Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in message news:<4173C2A3.7010000@univie.ac.at>...\n>\n>>How does your consciousness make measurements, in an exactly defined way??\n>\n>\n> What else, except for conscious observation, constitutes a measurement\n> ?\n\nI was asking the question to get at waht you mean by \'exact\';\nnow you avoid my question.\n\nA measurement is any permanent record of an event, whether or not anyone\nhas seen it. Thus the terabytes of collision data collected by CERN are\nmeasurements, although most of them have never been looked at by anybody.\nWe only look at crude summaries of such high tech data,\nbut the collapse (which gives rise to individual particle tracks) is\nclearly independent of whether or when we look at them.\n\n\n>>Most large-scale measurements are done in a completely automatic way,\n>>where human observers play no role at all except for discussing\n>>automatically produced summaries of the data...\n>\n> The measurement then consists in reading the paper in Physical review.\n> Consider Physical review to be in a superposed state until I read it\n> :-)\n\n> BTW, I know this sounds crazy. It probably even is, and I\'m not\n> particularly attached to this view. I just wanted to say that it is\n> the way I found around the conflict between locality, the measurement\n> problem, EPR states and so on. As I said, to me it lacks too much\n> "reality".\n\nI prefer unresolved tension to crazy solutions that pretend to give\nunderstanding but only dull the mind.\n\n\n> The view I expressed seems to be compatible with these issues. I\'m\n> not saying there aren\'t other ways around it.\n\nBut it introduces a host of other problems - such as how do you\ndistinguish between an obervation and a dream? How much do you\nactually observe, and how much do you simply interpolate between\nsome key observations and believe the result? The capacity of the\nconsciousness of an individual seems to be extremely small.\nMoreover, what happens if different consciousnesses collapse the\nworld? You\'d need a theory of interaction of consciousnesses,\nunless you go solipsist...\n\n\nArnold Neumaier\n\n\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch wrote:
> Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in message news:<4173C2A3.7010000@univie.ac.at>...
>
>>How does your consciousness make measurements, in an exactly defined way??
>
>
> What else, except for conscious observation, constitutes a measurement
> ?

I was asking the question to get at waht you mean by 'exact';
now you avoid my question.

A measurement is any permanent record of an event, whether or not anyone
has seen it. Thus the terabytes of collision data collected by CERN are
measurements, although most of them have never been looked at by anybody.
We only look at crude summaries of such high tech data,
but the collapse (which gives rise to individual particle tracks) is
clearly independent of whether or when we look at them.

>>Most large-scale measurements are done in a completely automatic way,
>>where human observers play no role at all except for discussing
>>automatically produced summaries of the data...
>
> The measurement then consists in reading the paper in Physical review.
> Consider Physical review to be in a superposed state until I read it
> :-)

> BTW, I know this sounds crazy. It probably even is, and I'm not
> particularly attached to this view. I just wanted to say that it is
> the way I found around the conflict between locality, the measurement
> problem, EPR states and so on. As I said, to me it lacks too much
> "reality".

I prefer unresolved tension to crazy solutions that pretend to give
understanding but only dull the mind.

> The view I expressed seems to be compatible with these issues. I'm
> not saying there aren't other ways around it.

But it introduces a host of other problems - such as how do you
distinguish between an obervation and a dream? How much do you
actually observe, and how much do you simply interpolate between
some key observations and believe the result? The capacity of the
consciousness of an individual seems to be extremely small.
Moreover, what happens if different consciousnesses collapse the
world? You'd need a theory of interaction of consciousnesses,
unless you go solipsist...

Arnold Neumaier

Ralph Hartley

Oct20-04, 10:48 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nPatrick Van Esch wrote:\n> After all, there is NO WAY for me to find out that\n> you, or anything else, has a consciousness.\n\nYou may be trying to say the same thing, but the point of my argument is\nthat Quantum Mechanics gives me no reason to *care*.\n\nTo continue the conversation:\n\nQ: But there is NO WAY to find out that you or\nQ: anything else, has a consciousness, and thus\nQ: qualifies as an observer, with vanishing points etc.\nA: Consciousness has nothing to do with it!\nA: The observer could be a purely mechanical\nA: device like a camera.\nQ: This is the renaissance, have cameras been invented?\nA: Cameras, yes, but not film.\n\n> But there are two reasons to be wary of it: first of all, I think that\n> it is a bad idea to do philosophy based upon a certain state of\n> knowledge of physical laws\n\nThe only connection to philosophy is that Quantum Mechanics says *nothing*\nmore about philosophy than perspective did.\n\nIn both cases that isn\'t much. For example, neither says anything at all\nabout the nature of consciousness.\n\nThey are both about something that was assumed to be absolute turning out\nto be relative, but still objective.\n\nThat happened at *least* four times since the renaissance, and twice in\njust the 20th century, so people should get used to it.\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch wrote:
> After all, there is NO WAY for me to find out that
> you, or anything else, has a consciousness.

You may be trying to say the same thing, but the point of my argument is
that Quantum Mechanics gives me no reason to *care*.

To continue the conversation:

Q: But there is NO WAY to find out that you or
Q: anything else, has a consciousness, and thus
Q: qualifies as an observer, with vanishing points etc.
A: Consciousness has nothing to do with it!
A: The observer could be a purely mechanical
A: device like a camera.
Q: This is the renaissance, have cameras been invented?
A: Cameras, yes, but not film.

> But there are two reasons to be wary of it: first of all, I think that
> it is a bad idea to do philosophy based upon a certain state of
> knowledge of physical laws

The only connection to philosophy is that Quantum Mechanics says *nothing*
more about philosophy than perspective did.

In both cases that isn't much. For example, neither says anything at all
about the nature of consciousness.

They are both about something that was assumed to be absolute turning out
to be relative, but still objective.

That happened at *least* four times since the renaissance, and twice in
just the 20th century, so people should get used to it.

Ralph Hartley

Ilja Schmelzer

Oct21-04, 02:47 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> ><rof@maths.tcd.ie> schrieb\n> >> There\'s obviously a bit more to explain than that (decoherence\n> >> would need to be appealed to before one can even talk about\n> >> individual objects which could even have a well-defined or\n> >> approximately well-defined velocity).\n\n> >No necessity. Use Bohmian mechanics.\n>\n> >> It also seems circular\n> >> to start with a spatial description (even a nonlocal one) and\n> >> then deduce later that a spatial description is appropriate.\n>\n> >I don\'t understand. A theory always starts with some postulates.\n> >Then it has to derive something observable. If observation\n> >supports this predictions, the postulates seem appropriate.\n>\n> This is related to the Bohmian point above. A theory has some\n> axioms or postulates. These are either postulates about what\n> the results of certain measurements will be (in which case it\n> is an "epistemological" theory),\n\nI would have called it "phenomenological".\n\n> or the postulates are about\n> what "really exists" (an "ontological" theory). There is a challenge\n> which ontological theories must meet which epistemological\n> ones don\'t need to. That is to say what will be observed. There\'s\n> no guarantee that if we start from an ontological theory which\n> has some kind of spatial manifold as one of its features then\n> the little creatures living in the theory will observe that\n> manifold.\n\nIndeed.\n\n>\n> The observer in the theory has access only to his own internal\n> states (if we restrict our attention to theories without\n> extra-sensory perception). From the modifications to his internal\n> state, he is supposed to generate a picture of the world outside\n> himself. Suppose, for example, in discrete time, he receives\n> the "symbol" E_n at time t_n. In order to predict what will\n> happen at the next timestep t_{n+1}, he will need to reflect\n> on many of the previous symbols he has received, E_{n-1}, E_{n-2} ...\n>\n> This looks like an N-step Markov process, with the probability\n> of receiving a symbol at t_n being dependent on the N symbols\n> received before this time. To generate a representation of a\n> world around him, in which the next symbol he receives is\n> considered to be determined (perhaps probabilistically) by the\n> "present" state of the world, rather than by the history of\n> symbols he has received, he must convert this N-step process to\n> a one-step Markov process.\n>\n> That is, he must form a data representation which compresses\n> the history of symbols he has received into a single "state\n> of the world", call it X_n, where\n> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)\n> (something which was not true of the symbols E_n).\n\nI don\'t think he is obligued to do such a thing. The state\nof the world contains lot of information, for example about\nthe inner parts of the Mars, which are not fixed in any way\nby my personal sequence of E_n.\n\n> Now, reception of the symbols E_n is called sensation; the\n> subsequent construction of the X_n is called perception,\n> and the transition probabilities P(X_{n+1}|X_n) are called\n> the laws of physics (as far as this observer is concerned).\n\nThe laws of physics in an ontological theory do not act\nbetween anything worth to be named "perception". They\nact between states of reality.\n\nAn ontological theory is falsified in your language if there\nexists no such sequence X_n which is compatible with\nthe proposed laws of physics and the observed sequence\nE_n. Thus, falsification does not need any map E_i -> X_j.\n\n> An epistemological theory deals with the same stuff as the\n> observer. In an ontological theory which merely postulates\n> the existence of space, it is necessary (before one can say\n> that it predicts that space will be observed) to describe the\n> internal structure of the observer, catalogue the E_n\'s and the\n> N-step transition probabilities P(E_{n+1}|E_n, E_{n-1}, ...), and\n> then show that it admits a representation X_n which splits up\n> X_n = \\prod_a x^a_n, where a indexes local "positions" and x^a_n\n> is the "mini-state of affairs" at position a, along\n> with some kind of locality. The locality is necessary because\n> there\'s no point in representing positions a and b as being\n> far away from each other if the state of affairs at b at time\n> n+1 always depends strongly on the state of affairs at a at time\n> n. Hence locality need not necessarily be strict, but it should\n> be "usual", in the sense that it\'s useful to distinguish between\n> one place and another.\n>\n> Bohmian mechanics can, I think, possibly achieve this, but it\'s\n> not obvious, and it hasn\'t been proven.\n\nI think the order of what has to be shown is reverse.\nThe laws are postulated, the theory is falsifiable if there\nexists sequences of E_i so that we can prove that no\nsequence of X_j compatible with these E_i exists,\nand falsified if we observe such a sequence.\n\nIn BM, the splitting is already postulated, and in the\nextremely weak sense of locality you use (useful to\ndistinguish between one place and another) it is local.\n\n> >Again, I don\'t see why Newtonian mechanics is "dishonest"\n> >if it postulates an Euclidean space.\n\n> Newtonian mechanics is descriptive of the world that we see - that\n> is, the perceived world, and doesn\'t pretend to describe something\n> underneath it from which the perceived world emerges.\n\nOf course not. The gravitational force cannot be "seen". We see only\nits consequences. Newton himself distinguishes between "true time"\nand "apparent time" as measured by human beings.\n\n> >That\'s your guess about the reasons for failure. I disagree.\n> >In Bohmian mechanics, the "results of measurements" are\n> >well incorporated, but the mathematical machinery is\n> >essentially the same.\n>\n> Which is why there isn\'t a special relativistic Bohmian\n> mechanics.\n\nThere is. Of course, it uses a preferred frame, thus, would\nbe better named "Lorentzian Bohmian mechanics". But it works\ncorrectly in the domain of applicability of special relativity.\n\n> >> But I think we can keep the symmetry group by abandoning\n> >> determinism without abandoning realism (although this\n> >> is not a position that I advocate).\n>\n> >No. The proof of Bell\'s inequality is based on Einstein-causal\n> >realism, not determinism.\n>\n> I was talking about keeping the symmetry group (the Lorentz group) -\n> not about keeping locality. Nonlocality has been experimentally\n> confirmed, so there\'s no point in trying to salvage locality.\n\nIn this case, you have to give up causality. This seems as unjustified\nas giving up realism.\n\n> >> I suspect the criterion for an explanation to\n> >> be \'realistic\' isn\'t very clearly defined.\n>\n> >You need some set of "possible realities" Lambda\n> >with a probability distribution rho(lambda) on it.\n> >Then you need a function X(x,lambda) which defines\n> >the outcome of the experiment X as depending from\n> >the state of reality lamdba and the decisions of the\n> >experimenters x. Then the probability of f(X) is\n>\n> ><f(X)> = int f(X(x,lambda)) rho(lambda) d lambda.\n>\n> >Its clearly defined, not?\n>\n> This looks more like a definition of determinism. Do\n> you consider determinism and realism to be the same thing?\n\nAlmost the same. Note that there is a probability distribution\nin the definition. A classical stochastic process fits under\nthe definition of realism, but is not deterministic. But, on the\nother hand, for a realistic theory we can always find a\ndeterministic model (using deterministic chaos). And reverse.\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> ><rof@maths.tcd.ie> schrieb
> >> There's obviously a bit more to explain than that (decoherence
> >> would need to be appealed to before one can even talk about
> >> individual objects which could even have a well-defined or
> >> approximately well-defined velocity).

> >No necessity. Use Bohmian mechanics.
>
> >> It also seems circular
> >> to start with a spatial description (even a nonlocal one) and
> >> then deduce later that a spatial description is appropriate.
>
> >I don't understand. A theory always starts with some postulates.
> >Then it has to derive something observable. If observation
> >supports this predictions, the postulates seem appropriate.
>
> This is related to the Bohmian point above. A theory has some
> axioms or postulates. These are either postulates about what
> the results of certain measurements will be (in which case it
> is an "epistemological" theory),

I would have called it "phenomenological".

> or the postulates are about
> what "really exists" (an "ontological" theory). There is a challenge
> which ontological theories must meet which epistemological
> ones don't need to. That is to say what will be observed. There's
> no guarantee that if we start from an ontological theory which
> has some kind of spatial manifold as one of its features then
> the little creatures living in the theory will observe that
> manifold.

Indeed.

>
> The observer in the theory has access only to his own internal
> states (if we restrict our attention to theories without
> extra-sensory perception). From the modifications to his internal
> state, he is supposed to generate a picture of the world outside
> himself. Suppose, for example, in discrete time, he receives
> the "symbol" E_n at time t_n. In order to predict what will
> happen at the next timestep t_{n+1}, he will need to reflect
> on many of the previous symbols he has received, E_{n-1}, E_{n-2} ...
>
> This looks like an N-step Markov process, with the probability
> of receiving a symbol at t_n being dependent on the N symbols
> received before this time. To generate a representation of a
> world around him, in which the next symbol he receives is
> considered to be determined (perhaps probabilistically) by the
> "present" state of the world, rather than by the history of
> symbols he has received, he must convert this N-step process to
> a one-step Markov process.
>
> That is, he must form a data representation which compresses
> the history of symbols he has received into a single "state
> of the world", call it X_n, where
> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)
> (something which was not true of the symbols E_n).

I don't think he is obligued to do such a thing. The state
of the world contains lot of information, for example about
the inner parts of the Mars, which are not fixed in any way
by my personal sequence of E_n.

> Now, reception of the symbols E_n is called sensation; the
> subsequent construction of the X_n is called perception,
> and the transition probabilities P(X_{n+1}|X_n) are called
> the laws of physics (as far as this observer is concerned).

The laws of physics in an ontological theory do not act
between anything worth to be named "perception". They
act between states of reality.

An ontological theory is falsified in your language if there
exists no such sequence X_n which is compatible with
the proposed laws of physics and the observed sequence
E_n. Thus, falsification does not need any map E_i -> X_j.

> An epistemological theory deals with the same stuff as the
> observer. In an ontological theory which merely postulates
> the existence of space, it is necessary (before one can say
> that it predicts that space will be observed) to describe the
> internal structure of the observer, catalogue the E_n's and the
> N-step transition probabilities P(E_{n+1}|E_n, E_{n-1}, ...), and
> then show that it admits a representation X_n which splits up
> X_n = \prod_a x^{a_n}, where a indexes local "positions" and x^{a_n}
> is the "mini-state of affairs" at position a, along
> with some kind of locality. The locality is necessary because
> there's no point in representing positions a and b as being
> far away from each other if the state of affairs at b at time
> n+1 always depends strongly on the state of affairs at a at time
> n. Hence locality need not necessarily be strict, but it should
> be "usual", in the sense that it's useful to distinguish between
> one place and another.
>
> Bohmian mechanics can, I think, possibly achieve this, but it's
> not obvious, and it hasn't been proven.

I think the order of what has to be shown is reverse.
The laws are postulated, the theory is falsifiable if there
exists sequences of E_i so that we can prove that no
sequence of X_j compatible with these E_i exists,
and falsified if we observe such a sequence.

In BM, the splitting is already postulated, and in the
extremely weak sense of locality you use (useful to
distinguish between one place and another) it is local.

> >Again, I don't see why Newtonian mechanics is "dishonest"
> >if it postulates an Euclidean space.

> Newtonian mechanics is descriptive of the world that we see - that
> is, the perceived world, and doesn't pretend to describe something
> underneath it from which the perceived world emerges.

Of course not. The gravitational force cannot be "seen". We see only
its consequences. Newton himself distinguishes between "true time"
and "apparent time" as measured by human beings.

> >That's your guess about the reasons for failure. I disagree.
> >In Bohmian mechanics, the "results of measurements" are
> >well incorporated, but the mathematical machinery is
> >essentially the same.
>
> Which is why there isn't a special relativistic Bohmian
> mechanics.

There is. Of course, it uses a preferred frame, thus, would
be better named "Lorentzian Bohmian mechanics". But it works
correctly in the domain of applicability of special relativity.

> >> But I think we can keep the symmetry group by abandoning
> >> determinism without abandoning realism (although this
> >> is not a position that I advocate).
>
> >No. The proof of Bell's inequality is based on Einstein-causal
> >realism, not determinism.
>
> I was talking about keeping the symmetry group (the Lorentz group) -
> not about keeping locality. Nonlocality has been experimentally
> confirmed, so there's no point in trying to salvage locality.

In this case, you have to give up causality. This seems as unjustified
as giving up realism.

> >> I suspect the criterion for an explanation to
> >> be 'realistic' isn't very clearly defined.
>
> >You need some set of "possible realities" \Lambda
> >with a probability distribution \rho(\lambda) on it.
> >Then you need a function X(x,\lambda) which defines
> >the outcome of the experiment X as depending from
> >the state of reality lamdba and the decisions of the
> >experimenters x. Then the probability of f(X) is
>
> ><f(X)> = \int f(X(x,\lambda)) \rho(\lambda) d \lambda.
>
> >Its clearly defined, not?
>
> This looks more like a definition of determinism. Do
> you consider determinism and realism to be the same thing?

Almost the same. Note that there is a probability distribution
in the definition. A classical stochastic process fits under
the definition of realism, but is not deterministic. But, on the
other hand, for a realistic theory we can always find a
deterministic model (using deterministic chaos). And reverse.

Ilja

Patrick Van Esch

Oct21-04, 02:47 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nArnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in message news:<41765176.5010607@univie.ac.at>...\n\n\n> Moreover, what happens if different consciousnesses collapse the\n> world? You\'d need a theory of interaction of consciousnesses,\n> unless you go solipsist...\n\nExactly ! That\'s the whole point, what I propose IS solipsism, or it\nalmost is. It also solves the problem of the dream: if you never wake\nup from it, then there\'s no difference between nature and that dream.\nWhat I wanted to say is that all you\'ll ever know and do is through\n_your_ conscious observation. If it weren\'t for that conscious\nobservation, there would be no need for any physical theory explaining\nanything. Now we\'d all like to think that there\'s an independent\nreality out there (and I agree that I\'d prefer to think that too), but\nthat isn\'t a strict necessity.\nSo, given the current state of affairs, namely that on one hand,\nlocality seems to be a strong principle, and on the other hand, that\nEPR like situations, together with 2 independent observers, violate\nthat locality in a very peculiar way (so that they cannot find out by\nthemselves, but that they do find out when they compare their data),\nthere seems to be a contradiction.\n\nThe fundamental difficulty comes from the fact that you have to allow\nfor a quantum measurement to be performed at space-like separated\nevents. If you take the collapse of the wavefunction as an element of\nreality (called a "measurement") then you cannot avoid collapsing over\nspace-like intervals, and hence a non-local phenomenon. So is\nlocality wrong then ? Some like to say that, but I find it too\nsurprising that locality is such a strong guiding principle, and that,\nfor all we know, we cannot use this manifest non-local measurement to\nmake an FTL phone. We\'ve much less certainty about what exactly is a\nmeasurement than about locality.\n\nAnd the way I weasel out is by saying that you cannot have that, that\nall observation has to be time-like connected. How can that be ?\nWell, if there\'s only one observer. Where ? Which one ? Well, then\nit is quite obvious that this leads to a solipsist viewpoint, because\nthe only observations I\'m truely aware of are my own ; I could deny\nall yours, or all of CERN\'s observations as true observations, but I\ncannot deny mine. All the others are "observations of observations",\nand the decoherence program shows us that in that case you cannot\ndistinguish between true observations, or decohered superpositions.\n\nBut I think I\'ll stop with this discussion, because people will start\nto think I\'m completely nuts :-) I repeat: I don\'t really "believe"\nwhat I write here. I just wanted to indicate a point of view, as a\nmental exercise, that can resolve the tension between the EPR-style\nnon-locality of the collapse and locality as a fundamental principle.\nAs I said in the first post: the price to pay can seem high; it is\nsolipsism!\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in message news:<41765176.5010607@univie.ac.at>...

> Moreover, what happens if different consciousnesses collapse the
> world? You'd need a theory of interaction of consciousnesses,
> unless you go solipsist...

Exactly ! That's the whole point, what I propose IS solipsism, or it
almost is. It also solves the problem of the dream: if you never wake
up from it, then there's no difference between nature and that dream.
What I wanted to say is that all you'll ever know and do is through
_your_ conscious observation. If it weren't for that conscious
observation, there would be no need for any physical theory explaining
anything. Now we'd all like to think that there's an independent
reality out there (and I agree that I'd prefer to think that too), but
that isn't a strict necessity.
So, given the current state of affairs, namely that on one hand,
locality seems to be a strong principle, and on the other hand, that
EPR like situations, together with 2 independent observers, violate
that locality in a very peculiar way (so that they cannot find out by
themselves, but that they do find out when they compare their data),
there seems to be a contradiction.

The fundamental difficulty comes from the fact that you have to allow
for a quantum measurement to be performed at space-like separated
events. If you take the collapse of the wavefunction as an element of
reality (called a "measurement") then you cannot avoid collapsing over
space-like intervals, and hence a non-local phenomenon. So is
locality wrong then ? Some like to say that, but I find it too
surprising that locality is such a strong guiding principle, and that,
for all we know, we cannot use this manifest non-local measurement to
make an FTL phone. We've much less certainty about what exactly is a
measurement than about locality.

And the way I weasel out is by saying that you cannot have that, that
all observation has to be time-like connected. How can that be ?
Well, if there's only one observer. Where ? Which one ? Well, then
it is quite obvious that this leads to a solipsist viewpoint, because
the only observations I'm truely aware of are my own ; I could deny
all yours, or all of CERN's observations as true observations, but I
cannot deny mine. All the others are "observations of observations",
and the decoherence program shows us that in that case you cannot
distinguish between true observations, or decohered superpositions.

But I think I'll stop with this discussion, because people will start
to think I'm completely nuts :-) I repeat: I don't really "believe"
what I write here. I just wanted to indicate a point of view, as a
mental exercise, that can resolve the tension between the EPR-style
non-locality of the collapse and locality as a fundamental principle.
As I said in the first post: the price to pay can seem high; it is
solipsism!

cheers,
Patrick.

Thomas Trotter

Oct21-04, 02:47 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nvanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410190714.24fe4752@posting.google. com>...\n> However, the interference of results in an EPR like setup are not more\n> surprising than the interference of a split light beam that went\n> around the building left and right and came back. What is surprising\n> in EPR is that if you consider that there has been a measurement at\n> Alice\'s and Bob\'s, then there ought to be a breaking of locality (I\n> know you are in favor of that conclusion). But then it is surprising\n> that we cannot build an FTL phone using that procedure. We can only\n> notice the effect when, through normal, sub-luminal signalling, we\n> have access to both measurements. So what is the conspiracy ? I\n> think the point is that there hasn\'t been a measurement at Alice\'s and\n> Bob\'s, but that there is only an interference of results after they\n> have been transported (in superposed state) in a locality-preserving,\n> subluminal way up to the common point where they can be compared.\n> Such an interference (like the interference of split light beams)\n> explains these results, and also explains why you cannot build an FTL\n> phone using the idea.\n\nCan you explain this more precisely? (I\'ve always had the intuitive\nfeeling that there was no instantaneous mechanism required, but Bell\'s\ntheorem has me stuck.)\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410190714.24fe4752@posting.google.com>...
> However, the interference of results in an EPR like setup are not more
> surprising than the interference of a split light beam that went
> around the building left and right and came back. What is surprising
> in EPR is that if you consider that there has been a measurement at
> Alice's and Bob's, then there ought to be a breaking of locality (I
> know you are in favor of that conclusion). But then it is surprising
> that we cannot build an FTL phone using that procedure. We can only
> notice the effect when, through normal, sub-luminal signalling, we
> have access to both measurements. So what is the conspiracy ? I
> think the point is that there hasn't been a measurement at Alice's and
> Bob's, but that there is only an interference of results after they
> have been transported (in superposed state) in a locality-preserving,
> subluminal way up to the common point where they can be compared.
> Such an interference (like the interference of split light beams)
> explains these results, and also explains why you cannot build an FTL
> phone using the idea.

Can you explain this more precisely? (I've always had the intuitive
feeling that there was no instantaneous mechanism required, but Bell's
theorem has me stuck.)

Patrick Van Esch

Oct22-04, 12:01 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nthomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410200738.1297c60c@posting.google. com>...\n\n> Can you explain this more precisely? (I\'ve always had the intuitive\n> feeling that there was no instantaneous mechanism required, but Bell\'s\n> theorem has me stuck.)\n\nThe initial state of the 2-photon state, and Alice and Bob before\nmeasurement:\n1/sqrt(2) [ |+>|-> - |->|+> ] |A0> |B0>\n\nThe first photon arrives at Alice, and Alice entangles with that\nphoton spin, on Alice\'s axis under angle ta. So we first write:\n|+> = cos(ta) |a+> + sin(ta) |a-> and |-> = - sin(ta) |a+> + cos(ta)\n|a->\nThe state |a+> will, under "observation" make evolve A0 into A+, and\nthe state |a-> will, under observation, make evolve A0 into A-. We\ncan also forget about the photon state once it has been observed.\n\nSo after "observation" by Alice, we have:\n\n1/sqrt(2) [ (cos(ta) |A+>|-> + sin(ta) |A->|-> +\nsin(ta) |A+>|+> - cos(ta) |A->|+> ] |B0>\n\nNote that this interaction is local at Alice\'s (it only used the first\nphoton state).\n\nThe observation at Bob (which looks along the z-axis) is also local,\nand this is easier because we already have the right basis for the\nsecond photon:\n\n1/sqrt(2) [ cos(ta) |A+>|B-> + sin(ta) |A->|B-> +\nsin(ta) |A+>|B+> - cos(ta) |A->|B+> ]\n\nIf we now assume that Bob and Alice, and their notes of experiments,\nremain in such a superposed state UNTIL we can make a measurement on\nthe whole system, which means: when their notes are transported to\nacommon place, then we find the probabilities of measurement results\nusing Born\'s rule, which are applied locally where the notes are\ncompared:\n\nAlice up and Bob down: cos^2(ta)/2\nAlice down and Bob down: sin^2(ta)/2\nAlice up and Bob up: sin^2(ta)/2\nAlice down and Bob up: cos^2(ta)/2\n\n>From these:\nAlice up: 1/2\nBob up: 1/2\nAlice opposite of Bob: cos^2(ta)/2\n\nThese are the same results as what we would have obtained by\n"projecting at a distance", as you can easily work out. I\'ll take one\nexample:\n\nIf Alice measures (using Born\'s rule) then she has 1/2 chance to find\n"up" and the state at Bob\'s is (remotely projected) cos(ta) |-> +\nsin(ta) |-> ;\nshe has also 1/2 chance to find "down" and then the state at Bob\'s is:\n-sin(ta) |+> + cos(ta) |->. It is this "projection at a distance"\nwhich is usually considered non-local, and which you have to consider\nif Alice and Bob performed true measurements.\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410200738.1297c60c@posting.google.com>...

> Can you explain this more precisely? (I've always had the intuitive
> feeling that there was no instantaneous mechanism required, but Bell's
> theorem has me stuck.)

The initial state of the 2-photon state, and Alice and Bob before
measurement:
1/\sqrt(2) [ |+>|-> - |->|+> ] |A0> |B0>

The first photon arrives at Alice, and Alice entangles with that
photon spin, on Alice's axis under angle ta. So we first write:
|+> = cos(ta) |a+> + sin(ta) |a-> and |-> = - sin(ta) |a+> + cos(ta)|a->
The state |a+> will, under "observation" make evolve A0 into A+, and
the state |a-> will, under observation, make evolve A0 into A-. We
can also forget about the photon state once it has been observed.

So after "observation" by Alice, we have:

1/\sqrt(2) [ (cos(ta) |A+>|-> + sin(ta) |A->|-> +sin(ta) |A+>|+> - cos(ta) |A->|+> ] |B0>

Note that this interaction is local at Alice's (it only used the first
photon state).

The observation at Bob (which looks along the z-axis) is also local,
and this is easier because we already have the right basis for the
second photon:

1/\sqrt(2) [ cos(ta) |A+>|B-> + sin(ta) |A->|B-> +sin(ta) |A+>|B+> - cos(ta) |A->|B+> ]

If we now assume that Bob and Alice, and their notes of experiments,
remain in such a superposed state UNTIL we can make a measurement on
the whole system, which means: when their notes are transported to
acommon place, then we find the probabilities of measurement results
using Born's rule, which are applied locally where the notes are
compared:

Alice up and Bob down: cos^2(ta)/2
Alice down and Bob down: sin^2(ta)/2
Alice up and Bob up: sin^2(ta)/2
Alice down and Bob up: cos^2(ta)/2

>From these:
Alice up: 1/2
Bob up: 1/2
Alice opposite of Bob: cos^2(ta)/2

These are the same results as what we would have obtained by
"projecting at a distance", as you can easily work out. I'll take one
example:

If Alice measures (using Born's rule) then she has 1/2 chance to find
"up" and the state at Bob's is (remotely projected) cos(ta) |-> +sin(ta) |-> ;
she has also 1/2 chance to find "down" and then the state at Bob's is:
-sin(ta) |+> + cos(ta) |->. It is this "projection at a distance"
which is usually considered non-local, and which you have to consider
if Alice and Bob performed true measurements.

cheers,
Patrick.

Joe Rongen

Oct22-04, 12:31 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Patrick Van Esch" <vanesch@ill.fr> wrote in message\nnews:c23e597b.0410201955.1b684c56@posting .google.com...\n>\n>\n> Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in\nmessage news:<41765176.5010607@univie.ac.at>...\n>\n>\n> > Moreover, what happens if different consciousnesses collapse the\n> > world? You\'d need a theory of interaction of consciousnesses,\n> > unless you go solipsist...\n>\n> Exactly ! That\'s the whole point, what I propose IS solipsism, or it\n> almost is. It also solves the problem of the dream: if you never wake\n> up from it, then there\'s no difference between nature and that dream.\n> What I wanted to say is that all you\'ll ever know and do is through\n> _your_ conscious observation. If it weren\'t for that conscious\n> observation, there would be no need for any physical theory\n> explaining anything.\n\nMaybe now is a good time to read (if you haven\'t read it yet)\nRoger Penrose, "Shadows of the Mind" ISBN 0-09-958211-2\n"A Search for the Missing Science of Consciousness" 1995\n\n.... 421 pages and Penrose is still searching. :-)\n\nRegards Joe\n\n\n---\nOutgoing mail is certified Virus Free.\nChecked by AVG anti-virus system (http://www.grisoft.com).\nVersion: 6.0.779 / Virus Database: 526 - Release Date: 10/19/04\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Patrick Van Esch" <vanesch@ill.fr> wrote in message
news:c23e597b.0410201955.1b684c56@posting.google.c om...
>
>
> Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in
message news:<41765176.5010607@univie.ac.at>...
>
>
> > Moreover, what happens if different consciousnesses collapse the
> > world? You'd need a theory of interaction of consciousnesses,
> > unless you go solipsist...
>
> Exactly ! That's the whole point, what I propose IS solipsism, or it
> almost is. It also solves the problem of the dream: if you never wake
> up from it, then there's no difference between nature and that dream.
> What I wanted to say is that all you'll ever know and do is through
> _your_ conscious observation. If it weren't for that conscious
> observation, there would be no need for any physical theory
> explaining anything.

Maybe now is a good time to read (if you haven't read it yet)
Roger Penrose, "Shadows of the Mind" ISBN 0-09-958211-2
"A Search for the Missing Science of Consciousness" 1995

.... 421 pages and Penrose is still searching. :-)

Regards Joe

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6..779 / Virus Database: 526 - Release Date: 10/19/04

rof@maths.tcd.ie

Oct24-04, 09:04 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n\n><rof@maths.tcd.ie> schrieb\n>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n>> ><rof@maths.tcd.ie> schrieb\n>>\n>> That is, he must form a data representation which compresses\n>> the history of symbols he has received into a single "state\n>> of the world", call it X_n, where\n>> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)\n>> (something which was not true of the symbols E_n).\n\n>I don\'t think he is obligued to do such a thing. The state\n>of the world contains lot of information, for example about\n>the inner parts of the Mars, which are not fixed in any way\n>by my personal sequence of E_n.\n\nWell, everything that you know about Mars, including the fact\nthat there is a Mars, you found out about through the E_n. You\'ve\ntaken this information and formed for yourself a representation, M_n,\nwhich we could call "the current attributes of Mars", whose evolution\nfrom now onwards doesn\'t depend on events which happened a year ago;\nM_{n+1} depends only on M_n, the rest of the X_n, and the laws of\nphysics P(X_{n+1}|X_n).\n\nThe evolution of the E_n, however, certainly does depend on what\nhappened years ago (that is, those E_m which arrived years ago),\nsince if you had learned long ago that Mars was likely to explode,\nthen the probability that you would see a bright flash in the\npart of the sky where Mars is (a certain E_{n+1}), would be higher\nthan if you hadn\'t learned that, even though all of the intervening\nE_n had been the same.\n\nThat is,\nP(E_{n+1}|E_n, E_{n-1} ... E_m="Teacher tells me that Mars is likely\nto explode", E_{m-1}, ...)\nis not equal to\nP(E_{n+1}|E_n, E_{n-1} ... E\'_m="Teacher tells me that Mars is not likely\nto explode", E_{m-1}, ...)\nhence P(E_{n+1}|E_n, E_{n-1} ...) is not equal to P(E_{n+1}|E_n)\n\nIt is by collecting together the E_n as they arrive, and generating\nfor yourself the representation M_n, which you require to satisfy\nthe condition P(M_{n+1}|X_n)=P(M_{n+1}|X_n, X_{n-1}, ...), that\nyou get to perceive (or rather, generate) a world in which what\nhappens next (in the world X_{n+1}) depends only on what happened\na moment before (in the world X_n).\n\nSo if Mars explodes, you will not suspect that the state of affairs\na year ago is responsible, but will always attribute it to the\nstate of affairs just before the explosion. There are certainly\nthing which you don\'t know about Mars (its internal composition\nand so on), which increase your uncertainty about what will happen\nnext. After the explosion at time n+1 you can modify your representation\nM_n to include some additional information, such as a footnote\nsaying "about to explode" (which is called retrodiction), but mere\nuncertainty about what will happen next doesn\'t prevent you from\ncollecting all the information which you have previously acquired,\nthe E_n, and generating a single representation M_n, on which to\nbase your expectations of what will happen next.\n\n>> Now, reception of the symbols E_n is called sensation; the\n>> subsequent construction of the X_n is called perception,\n>> and the transition probabilities P(X_{n+1}|X_n) are called\n>> the laws of physics (as far as this observer is concerned).\n\n>The laws of physics in an ontological theory do not act\n>between anything worth to be named "perception". They\n>act between states of reality.\n\n>An ontological theory is falsified in your language if there\n>exists no such sequence X_n which is compatible with\n>the proposed laws of physics and the observed sequence\n>E_n.\n\nThe construction of the X_n is perception (of the world). The "states\nof reality" are idealised X_n\'s, which, along with some\nproposed laws of physics are maximally predictive, containing\n"all the information in the world". We can\'t, of course, acquire\nenough information to predict the future exactly, so the X_n\nthat we construct (the part of the world that we perceive) is\nonly a part of the idealised X_n (which is the world that we\nimagine to be there, but do not perceive).\n\n>> An epistemological theory deals with the same stuff as the\n>> observer. In an ontological theory which merely postulates\n>> the existence of space, it is necessary (before one can say\n>> that it predicts that space will be observed) to describe the\n>> internal structure of the observer, catalogue the E_n\'s and the\n>> N-step transition probabilities P(E_{n+1}|E_n, E_{n-1}, ...), and\n>> then show that it admits a representation X_n which splits up\n>> X_n = \\prod_a x^a_n, where a indexes local "positions" and x^a_n\n>> is the "mini-state of affairs" at position a, along\n>> with some kind of locality. The locality is necessary because\n>> there\'s no point in representing positions a and b as being\n>> far away from each other if the state of affairs at b at time\n>> n+1 always depends strongly on the state of affairs at a at time\n>> n. Hence locality need not necessarily be strict, but it should\n>> be "usual", in the sense that it\'s useful to distinguish between\n>> one place and another.\n>>\n>> Bohmian mechanics can, I think, possibly achieve this, but it\'s\n>> not obvious, and it hasn\'t been proven.\n\n>I think the order of what has to be shown is reverse.\n>The laws are postulated, the theory is falsifiable if there\n>exists sequences of E_i so that we can prove that no\n>sequence of X_j compatible with these E_i exists,\n>and falsified if we observe such a sequence.\n\n>In BM, the splitting is already postulated, and in the\n>extremely weak sense of locality you use (useful to\n>distinguish between one place and another) it is local.\n\nIn an ontological theory the splitting needs to be deduced,\nrather than postulated. This is because an ontological theory\nis one in which the putative underlying objects and dynamics\nare postulated, and the structure of the E_n and transition\nprobabilities P(E_n|E_{n-1},...) need to be deduced, since they\nare considered to be vulgar "subjective" things, whose\nproperties and existence are determined by the far more\ndignified "objective" things with which the ontological\ntheory deals directly. So an ontological theory can\'t declare\nin advance that the observers (namely, us) will perceive\nspace - a decomposable set of X_n.\n\nBM is local in my weak sense, from all appearances, but that\'s\na theorem which hasn\'t been proven yet (although compatibility\nwith quantum mechanics indicates that it will be able to describe\nan observer as complicated as a human body, and we can expect\nthat that\'s enough).\n\n>> >> But I think we can keep the symmetry group by abandoning\n>> >> determinism without abandoning realism (although this\n>> >> is not a position that I advocate).\n>>\n>> >No. The proof of Bell\'s inequality is based on Einstein-causal\n>> >realism, not determinism.\n>>\n>> I was talking about keeping the symmetry group (the Lorentz group) -\n>> not about keeping locality. Nonlocality has been experimentally\n>> confirmed, so there\'s no point in trying to salvage locality.\n\n>In this case, you have to give up causality. This seems as unjustified\n>as giving up realism.\n\nYou do, but giving up either realism or causality won\'t affect your\nability to predict X_{n+1} from X_n, since they are merely rules of\nthumb for generating representations.\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb
>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
>> ><rof@maths.tcd.ie> schrieb
>>
>> That is, he must form a data representation which compresses
>> the history of symbols he has received into a single "state
>> of the world", call it X_n, where
>> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)
>> (something which was not true of the symbols E_n).

>I don't think he is obligued to do such a thing. The state
>of the world contains lot of information, for example about
>the inner parts of the Mars, which are not fixed in any way
>by my personal sequence of E_n.

Well, everything that you know about Mars, including the fact
that there is a Mars, you found out about through the E_n. You've
taken this information and formed for yourself a representation, M_n,
which we could call "the current attributes of Mars", whose evolution
from now onwards doesn't depend on events which happened a year ago;
M_{n+1} depends only on M_n, the rest of the X_n, and the laws of
physics P(X_{n+1}|X_n).

The evolution of the E_n, however, certainly does depend on what
happened years ago (that is, those E_m which arrived years ago),
since if you had learned long ago that Mars was likely to explode,
then the probability that you would see a bright flash in the
part of the sky where Mars is (a certain E_{n+1}), would be higher
than if you hadn't learned that, even though all of the intervening
E_n had been the same.

That is,
P(E_{n+1}|E_n, E_{n-1} ... E_m="Teacher tells me that Mars is likely
to explode", E_{m-1}, ...)
is not equal to
P(E_{n+1}|E_n, E_{n-1} ... E'_m="Teacher tells me that Mars is not likely
to explode", E_{m-1}, ...)
hence P(E_{n+1}|E_n, E_{n-1} ...) is not equal to P(E_{n+1}|E_n)

It is by collecting together the E_n as they arrive, and generating
for yourself the representation M_n, which you require to satisfy
the condition P(M_{n+1}|X_n)=P(M_{n+1}|X_n, X_{n-1}, ...), that
you get to perceive (or rather, generate) a world in which what
happens next (in the world X_{n+1}) depends only on what happened
a moment before (in the world X_n).

So if Mars explodes, you will not suspect that the state of affairs
a year ago is responsible, but will always attribute it to the
state of affairs just before the explosion. There are certainly
thing which you don't know about Mars (its internal composition
and so on), which increase your uncertainty about what will happen
next. After the explosion at time n+1 you can modify your representation
M_n to include some additional information, such as a footnote
saying "about to explode" (which is called retrodiction), but mere
uncertainty about what will happen next doesn't prevent you from
collecting all the information which you have previously acquired,
the E_n, and generating a single representation M_n, on which to
base your expectations of what will happen next.

>> Now, reception of the symbols E_n is called sensation; the
>> subsequent construction of the X_n is called perception,
>> and the transition probabilities P(X_{n+1}|X_n) are called
>> the laws of physics (as far as this observer is concerned).

>The laws of physics in an ontological theory do not act
>between anything worth to be named "perception". They
>act between states of reality.

>An ontological theory is falsified in your language if there
>exists no such sequence X_n which is compatible with
>the proposed laws of physics and the observed sequence
>E_n.

The construction of the X_n is perception (of the world). The "states
of reality" are idealised X_n's, which, along with some
proposed laws of physics are maximally predictive, containing
"all the information in the world". We can't, of course, acquire
enough information to predict the future exactly, so the X_n
that we construct (the part of the world that we perceive) is
only a part of the idealised X_n (which is the world that we
imagine to be there, but do not perceive).

>> An epistemological theory deals with the same stuff as the
>> observer. In an ontological theory which merely postulates
>> the existence of space, it is necessary (before one can say
>> that it predicts that space will be observed) to describe the
>> internal structure of the observer, catalogue the E_n's and the
>> N-step transition probabilities P(E_{n+1}|E_n, E_{n-1}, ...), and
>> then show that it admits a representation X_n which splits up
>> X_n = \prod_a x^{a_n}, where a indexes local "positions" and x^{a_n}
>> is the "mini-state of affairs" at position a, along
>> with some kind of locality. The locality is necessary because
>> there's no point in representing positions a and b as being
>> far away from each other if the state of affairs at b at time
>> n+1 always depends strongly on the state of affairs at a at time
>> n. Hence locality need not necessarily be strict, but it should
>> be "usual", in the sense that it's useful to distinguish between
>> one place and another.
>>
>> Bohmian mechanics can, I think, possibly achieve this, but it's
>> not obvious, and it hasn't been proven.

>I think the order of what has to be shown is reverse.
>The laws are postulated, the theory is falsifiable if there
>exists sequences of E_i so that we can prove that no
>sequence of X_j compatible with these E_i exists,
>and falsified if we observe such a sequence.

>In BM, the splitting is already postulated, and in the
>extremely weak sense of locality you use (useful to
>distinguish between one place and another) it is local.

In an ontological theory the splitting needs to be deduced,
rather than postulated. This is because an ontological theory
is one in which the putative underlying objects and dynamics
are postulated, and the structure of the E_n and transition
probabilities P(E_n|E_{n-1},...) need to be deduced, since they
are considered to be vulgar "subjective" things, whose
properties and existence are determined by the far more
dignified "objective" things with which the ontological
theory deals directly. So an ontological theory can't declare
in advance that the observers (namely, us) will perceive
space - a decomposable set of X_n.

BM is local in my weak sense, from all appearances, but that's
a theorem which hasn't been proven yet (although compatibility
with quantum mechanics indicates that it will be able to describe
an observer as complicated as a human body, and we can expect
that that's enough).

>> >> But I think we can keep the symmetry group by abandoning
>> >> determinism without abandoning realism (although this
>> >> is not a position that I advocate).
>>
>> >No. The proof of Bell's inequality is based on Einstein-causal
>> >realism, not determinism.
>>
>> I was talking about keeping the symmetry group (the Lorentz group) -
>> not about keeping locality. Nonlocality has been experimentally
>> confirmed, so there's no point in trying to salvage locality.

>In this case, you have to give up causality. This seems as unjustified
>as giving up realism.

You do, but giving up either realism or causality won't affect your
ability to predict X_{n+1} from X_n, since they are merely rules of
thumb for generating representations.

R.

Ilja Schmelzer

Oct25-04, 08:09 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> ><rof@maths.tcd.ie> schrieb\n> >> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> >> That is, he must form a data representation which compresses\n> >> the history of symbols he has received into a single "state\n> >> of the world", call it X_n, where\n> >> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)\n> >> (something which was not true of the symbols E_n).\n>\n> >I don\'t think he is obligued to do such a thing. The state\n> >of the world contains lot of information, for example about\n> >the inner parts of the Mars, which are not fixed in any way\n> >by my personal sequence of E_n.\n>\n> Well, everything that you know about Mars, including the fact\n> that there is a Mars, you found out about through the E_n. You\'ve\n> taken this information and formed for yourself a representation, M_n,\n> which we could call "the current attributes of Mars", whose evolution\n> from now onwards doesn\'t depend on events which happened a year ago;\n> M_{n+1} depends only on M_n, the rest of the X_n, and the laws of\n> physics P(X_{n+1}|X_n).\n\n> >> Now, reception of the symbols E_n is called sensation; the\n> >> subsequent construction of the X_n is called perception,\n> >> and the transition probabilities P(X_{n+1}|X_n) are called\n> >> the laws of physics (as far as this observer is concerned).\n>\n> >The laws of physics in an ontological theory do not act\n> >between anything worth to be named "perception". They\n> >act between states of reality.\n>\n> >An ontological theory is falsified in your language if there\n> >exists no such sequence X_n which is compatible with\n> >the proposed laws of physics and the observed sequence\n> >E_n.\n>\n> The construction of the X_n is perception (of the world).\n\nI\'m not a native speaker, but this sounds strange. I would\naccept it if you call the M_n "perception".\n\n> >In BM, the splitting is already postulated, and in the\n> >extremely weak sense of locality you use (useful to\n> >distinguish between one place and another) it is local.\n>\n> In an ontological theory the splitting needs to be deduced,\n> rather than postulated. This is because an ontological theory\n> is one in which the putative underlying objects and dynamics\n> are postulated, and the structure of the E_n and transition\n> probabilities P(E_n|E_{n-1},...) need to be deduced, since they\n> are considered to be vulgar "subjective" things, whose\n> properties and existence are determined by the far more\n> dignified "objective" things with which the ontological\n> theory deals directly. So an ontological theory can\'t declare\n> in advance that the observers (namely, us) will perceive\n> space - a decomposable set of X_n.\n\nThe ontological theory postulates the structure and the laws of\nthe X_n. If these have the form of something localized in a space,\nspace is postulated. What you perceive is another question,\nin the E_n we have, for example, nothing three-dimensional.\nWe see only two 2D movies. So it is completely unclear what\nmeans spatial structure in the E_n.\n\nAnd, moreover, there is no necessity for this.\n\n> BM is local in my weak sense, from all appearances, but that\'s\n> a theorem which hasn\'t been proven yet (although compatibility\n> with quantum mechanics indicates that it will be able to describe\n> an observer as complicated as a human body, and we can expect\n> that that\'s enough).\n\nI cannot see anything worth to be "proven" in this sense.\n\n> >> I was talking about keeping the symmetry group (the Lorentz group) -\n> >> not about keeping locality. Nonlocality has been experimentally\n> >> confirmed, so there\'s no point in trying to salvage locality.\n>\n> >In this case, you have to give up causality. This seems as unjustified\n> >as giving up realism.\n>\n> You do, but giving up either realism or causality won\'t affect your\n> ability to predict X_{n+1} from X_n, since they are merely rules of\n> thumb for generating representations.\n\nNo, the rules of thumb are questioned. I have no way to construct\nsome X_n in the usual (causal, realistic) meaning which has the\nproperty of Lorentz-invariance.\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> ><rof@maths.tcd.ie> schrieb
> >> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> >> That is, he must form a data representation which compresses
> >> the history of symbols he has received into a single "state
> >> of the world", call it X_n, where
> >> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)
> >> (something which was not true of the symbols E_n).
>
> >I don't think he is obligued to do such a thing. The state
> >of the world contains lot of information, for example about
> >the inner parts of the Mars, which are not fixed in any way
> >by my personal sequence of E_n.
>
> Well, everything that you know about Mars, including the fact
> that there is a Mars, you found out about through the E_n. You've
> taken this information and formed for yourself a representation, M_n,
> which we could call "the current attributes of Mars", whose evolution
> from now onwards doesn't depend on events which happened a year ago;
> M_{n+1} depends only on M_n, the rest of the X_n, and the laws of
> physics P(X_{n+1}|X_n).

> >> Now, reception of the symbols E_n is called sensation; the
> >> subsequent construction of the X_n is called perception,
> >> and the transition probabilities P(X_{n+1}|X_n) are called
> >> the laws of physics (as far as this observer is concerned).
>
> >The laws of physics in an ontological theory do not act
> >between anything worth to be named "perception". They
> >act between states of reality.
>
> >An ontological theory is falsified in your language if there
> >exists no such sequence X_n which is compatible with
> >the proposed laws of physics and the observed sequence
> >E_n.
>
> The construction of the X_n is perception (of the world).

I'm not a native speaker, but this sounds strange. I would
accept it if you call the M_n "perception".

> >In BM, the splitting is already postulated, and in the
> >extremely weak sense of locality you use (useful to
> >distinguish between one place and another) it is local.
>
> In an ontological theory the splitting needs to be deduced,
> rather than postulated. This is because an ontological theory
> is one in which the putative underlying objects and dynamics
> are postulated, and the structure of the E_n and transition
> probabilities P(E_n|E_{n-1},...) need to be deduced, since they
> are considered to be vulgar "subjective" things, whose
> properties and existence are determined by the far more
> dignified "objective" things with which the ontological
> theory deals directly. So an ontological theory can't declare
> in advance that the observers (namely, us) will perceive
> space - a decomposable set of X_n.

The ontological theory postulates the structure and the laws of
the X_n. If these have the form of something localized in a space,
space is postulated. What you perceive is another question,
in the E_n we have, for example, nothing three-dimensional.
We see only two 2D movies. So it is completely unclear what
means spatial structure in the E_n.

And, moreover, there is no necessity for this.

> BM is local in my weak sense, from all appearances, but that's
> a theorem which hasn't been proven yet (although compatibility
> with quantum mechanics indicates that it will be able to describe
> an observer as complicated as a human body, and we can expect
> that that's enough).

I cannot see anything worth to be "proven" in this sense.

> >> I was talking about keeping the symmetry group (the Lorentz group) -
> >> not about keeping locality. Nonlocality has been experimentally
> >> confirmed, so there's no point in trying to salvage locality.
>
> >In this case, you have to give up causality. This seems as unjustified
> >as giving up realism.
>
> You do, but giving up either realism or causality won't affect your
> ability to predict X_{n+1} from X_n, since they are merely rules of
> thumb for generating representations.

No, the rules of thumb are questioned. I have no way to construct
some X_n in the usual (causal, realistic) meaning which has the
property of Lorentz-invariance.

Ilja

Ilja Schmelzer

Oct25-04, 08:10 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Patrick Van Esch" <vanesch@ill.fr> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> > "Patrick Van Esch" <vanesch@ill.fr> schrieb\n> > > 1) it preserves the so-cherished locality:\n> > > because it is only upon the local comparison that the measurements\n> > > "interfere" and produce the correlations. From the moment you\n> > > introduce two spacelike separated true measurements, there is no way\n> > > in which you cannot violate locality. The only way to make sure you\n> > > don\'t is by having only ONE measurement.\n\n> > No, it doesn\'t. The "explanation" is a pseudo-explanation,\n> > because it would allow you to "explain away" even a working\n> > FTL phone. Indeed, if you have only one measurement, you\n> > cannot prove that an FTL phone really is an FTL phone.\n>\n> Not really. A working FTL phone would allow me, as a single observer,\n> to conclude that the signal I sent went up and down Titan in 0.5\n> seconds.\n\nHow?\n\n> After all, there is no quantum interference working here,\n> this can be considered classically (or semiclassically, in that a\n> narrow wavepacket of states went up there and came back).\n\nWould you believe that a phone works if you are allowed to tell\nby phone only some preexisting text written by somebody else?\nNo. Thus, you use your freedom of the experimenter to generate\ninput. Would you believe if what you hear is in no way a reaction\non what you have told? No. Thus, you base your decision on\nsome observed correlation. But, last not least, even if you have\na nice conversation, how can you be sure that the person you talk\nto is not sitting next door but really on Titan? I suspect you have to\ncompare this with some other information from Titan which you\nconsider to be more reliable. Now, whatever information you have\nabout Titan is available next door too, thus, may be faked. I would\nsay you need a friend on Titan who also creates some information,\nand then, if you meet later, you can compare your records and be\nsure that the phone works.\n\nBut in this case you use essentially the same method as in the\nEPR experiments: free choices of experimenters on above ends,\nmeasurements, later comparison and theoretical conclusion that\nno other realistic explanation is possible.\n\n> However, the interference of results in an EPR like setup are not more\n> surprising than the interference of a split light beam that went\n> around the building left and right and came back.\n\nNo. This type of (classical wave) interference may be explained in\na classical realistic local theory, violations of Bell\'s inequality not.\n\n> What is surprising\n> in EPR is that if you consider that there has been a measurement at\n> Alice\'s and Bob\'s, then there ought to be a breaking of locality (I\n> know you are in favor of that conclusion). But then it is surprising\n> that we cannot build an FTL phone using that procedure.\n\nIt is not very surprising. It is a consequence of the possibility of two\nexplanations: A->B or B->A. Whenever we have two explanations\nfor some observed correlation, we cannot use the observation for\nour own information transfer, because the transfer of information\nA->B would contradict the possible explanation B->A.\n\nThus, every causal connection which does not allow us to observe\nits direction has this property.\n\n> We can only\n> notice the effect when, through normal, sub-luminal signalling, we\n> have access to both measurements. So what is the conspiracy ?\n\nI don\'t see much conspiracy. I see that we observe a causal\nconnection but are unable to observe its direction. That\'s the whole\nconspiracy: A typical case of indirect, incomplete observation.\nThe normal conclusion is that there is a causal connection\nwith some causal direction which is hidden from observation.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Patrick Van Esch" <vanesch@ill.fr> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> > "Patrick Van Esch" <vanesch@ill.fr> schrieb
> > > 1) it preserves the so-cherished locality:
> > > because it is only upon the local comparison that the measurements
> > > "interfere" and produce the correlations. From the moment you
> > > introduce two spacelike separated true measurements, there is no way
> > > in which you cannot violate locality. The only way to make sure you
> > > don't is by having only ONE measurement.

> > No, it doesn't. The "explanation" is a pseudo-explanation,
> > because it would allow you to "explain away" even a working
> > FTL phone. Indeed, if you have only one measurement, you
> > cannot prove that an FTL phone really is an FTL phone.
>
> Not really. A working FTL phone would allow me, as a single observer,
> to conclude that the signal I sent went up and down Titan in .5
> seconds.

How?

> After all, there is no quantum interference working here,
> this can be considered classically (or semiclassically, in that a
> narrow wavepacket of states went up there and came back).

Would you believe that a phone works if you are allowed to tell
by phone only some preexisting text written by somebody else?
No. Thus, you use your freedom of the experimenter to generate
input. Would you believe if what you hear is in no way a reaction
on what you have told? No. Thus, you base your decision on
some observed correlation. But, last not least, even if you have
a nice conversation, how can you be sure that the person you talk
to is not sitting next door but really on Titan? I suspect you have to
compare this with some other information from Titan which you
consider to be more reliable. Now, whatever information you have
about Titan is available next door too, thus, may be faked. I would
say you need a friend on Titan who also creates some information,
and then, if you meet later, you can compare your records and be
sure that the phone works.

But in this case you use essentially the same method as in the
EPR experiments: free choices of experimenters on above ends,
measurements, later comparison and theoretical conclusion that
no other realistic explanation is possible.

> However, the interference of results in an EPR like setup are not more
> surprising than the interference of a split light beam that went
> around the building left and right and came back.

No. This type of (classical wave) interference may be explained in
a classical realistic local theory, violations of Bell's inequality not.

> What is surprising
> in EPR is that if you consider that there has been a measurement at
> Alice's and Bob's, then there ought to be a breaking of locality (I
> know you are in favor of that conclusion). But then it is surprising
> that we cannot build an FTL phone using that procedure.

It is not very surprising. It is a consequence of the possibility of two
explanations: A->B or B->A. Whenever we have two explanations
for some observed correlation, we cannot use the observation for
our own information transfer, because the transfer of information
A->B would contradict the possible explanation B->A.

Thus, every causal connection which does not allow us to observe
its direction has this property.

> We can only
> notice the effect when, through normal, sub-luminal signalling, we
> have access to both measurements. So what is the conspiracy ?

I don't see much conspiracy. I see that we observe a causal
connection but are unable to observe its direction. That's the whole
conspiracy: A typical case of indirect, incomplete observation.
The normal conclusion is that there is a causal connection
with some causal direction which is hidden from observation.

Ilja

Malcolm

Oct26-04, 12:57 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Joe Rongen" <joe@alpha.to> wrote in message news:<004001c4b78f\\$418ecb20\\$2723fea9@joerongen >...\n> "Patrick Van Esch" <vanesch@ill.fr> wrote in message\n> news:c23e597b.0410201955.1b684c56@posting.google.c om...\n> >\n> >\n> > Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in\n> message news:<41765176.5010607@univie.ac.at>...\n> >\n> >\n> > > Moreover, what happens if different consciousnesses collapse the\n> > > world? You\'d need a theory of interaction of consciousnesses,\n> > > unless you go solipsist...\n> >\n> > Exactly ! That\'s the whole point, what I propose IS solipsism, or it\n> > almost is. It also solves the problem of the dream: if you never wake\n> > up from it, then there\'s no difference between nature and that dream.\n> > What I wanted to say is that all you\'ll ever know and do is through\n> > _your_ conscious observation. If it weren\'t for that conscious\n> > observation, there would be no need for any physical theory\n> > explaining anything.\n>\n> Maybe now is a good time to read (if you haven\'t read it yet)\n> Roger Penrose, "Shadows of the Mind" ISBN 0-09-958211-2\n> "A Search for the Missing Science of Consciousness" 1995\n>\n> ... 421 pages and Penrose is still searching. :-)\n>\n> Regards Joe\n>\n>\n> ---\n> Outgoing mail is certified Virus Free.\n> Checked by AVG anti-virus system (http://www.grisoft.com).\n> Version: 6.0.779 / Virus Database: 526 - Release Date: 10/19/04\n\n.... then you can read Roger Penrose "The Road to Reality":\n\nhttp://www.321books.co.uk/reviews/the-road-to-reality-by-roger-penrose.htm\n\n.... 1094 pages and Penrose IS still searching...but is on hold in this book :-)\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Joe Rongen" <joe@\alpha.to> wrote in message news:<004001c4b78f$418ecb20$2723fea9@joerongen>...
> "Patrick Van Esch" <vanesch@ill.fr> wrote in message
> news:c23e597b.0410201955.1b684c56@posting.google.c om...
> >
> >
> > Arnold Neumaier <Arnold.Neumaier@univie.ac.at> wrote in
> message news:<41765176.5010607@univie.ac.at>...
> >
> >
> > > Moreover, what happens if different consciousnesses collapse the
> > > world? You'd need a theory of interaction of consciousnesses,
> > > unless you go solipsist...
> >
> > Exactly ! That's the whole point, what I propose IS solipsism, or it
> > almost is. It also solves the problem of the dream: if you never wake
> > up from it, then there's no difference between nature and that dream.
> > What I wanted to say is that all you'll ever know and do is through
> > _your_ conscious observation. If it weren't for that conscious
> > observation, there would be no need for any physical theory
> > explaining anything.
>
> Maybe now is a good time to read (if you haven't read it yet)
> Roger Penrose, "Shadows of the Mind" ISBN 0-09-958211-2
> "A Search for the Missing Science of Consciousness" 1995
>
> ... 421 pages and Penrose is still searching. :-)
>
> Regards Joe
>
>
> ---
> Outgoing mail is certified Virus Free.
> Checked by AVG anti-virus system (http://www.grisoft.com).
> Version: 6..779 / Virus Database: 526 - Release Date: 10/19/04

.... then you can read Roger Penrose "The Road to Reality":

http://www.321books.co.uk/reviews/the-road-to-reality-by-roger-penrose.htm

.... 1094 pages and Penrose IS still searching...but is on hold in this book :-)

seratend

Oct26-04, 12:57 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cl7p80\\$dck\\$1@beech.fernuni-hagen.de>...\n> > However, the interference of results in an EPR like setup are not more\n> > surprising than the interference of a split light beam that went\n> > around the building left and right and came back.\n>\n> No. This type of (classical wave) interference may be explained in\n> a classical realistic local theory, violations of Bell\'s inequality not.\n>\n> > What is surprising\n> > in EPR is that if you consider that there has been a measurement at\n> > Alice\'s and Bob\'s, then there ought to be a breaking of locality (I\n> > know you are in favor of that conclusion). But then it is surprising\n> > that we cannot build an FTL phone using that procedure.\n>\n> It is not very surprising. It is a consequence of the possibility of two\n> explanations: A->B or B->A. Whenever we have two explanations\n> for some observed correlation, we cannot use the observation for\n> our own information transfer, because the transfer of information\n> A->B would contradict the possible explanation B->A.\n>\n> Thus, every causal connection which does not allow us to observe\n> its direction has this property.\n>\n> > We can only\n> > notice the effect when, through normal, sub-luminal signalling, we\n> > have access to both measurements. So what is the conspiracy ?\n>\n> I don\'t see much conspiracy. I see that we observe a causal\n> connection but are unable to observe its direction. That\'s the whole\n> conspiracy: A typical case of indirect, incomplete observation.\n> The normal conclusion is that there is a causal connection\n> with some causal direction which is hidden from observation.\n>\n> Ilja\n\nAfter all these posts, as an external reader, I still not understand\nthe point of view of Ilia nor what could be its FTL phone.\nFirst what do you intend by causality? For me, like with the QM\nmeasurement, it is very difficult to understand what it is or what it\nsays (human interpretation).\n\nTherefore, I always try to restrict its meaning to each experiment I\ndo or think: I prefer to consider causality, when I can, with an\ninteraction model and thus a time evolution: a near mathematical\nformulation.\n\nThe most striking feature with the explanations you give, is the need\nof the "causality" to explain "physically" your possible FTL phone.\n(as you have removed all the physical model tools such as\ninteractions, … that may help us to understand what do you intend\nreally by causality and what is your FTL phone).\n\nSo, Do you intend "->" (your causality of the FTL phone) as the\nmathematical logic symbol "=>" ?\nIf yes, when I read "It is not very surprising. It is a consequence of\nthe possibility of two explanations: A->B or B->A", I understand (A=>\nB) logical OR (B => A). Thus, the statement "Whenever we have two\nexplanations for some observed correlation" I understand you intend\nthat you have both (A=> B) and (B => A) true.\nThus A <=> B. So if I reinterpret <=>, in your logical context I would\nsay "A is B" thus explaining the mysterious "we cannot use the\nobservation for our own information transfer" because in that case we\nare dealing with the same object (for an information transfer we need\n2 points).\n\nWhat is indeed, more difficult to understand with your point of view\nis that a logical "A => B" is formally equivalent to the statement\n"non B => non A": the point of view of 2 possible observers.\n\n> Thus, every causal connection which does not allow us to observe\n> its direction has this property.\n\nThus the "causal connection direction" is very difficult to define in\nyour statements if you do not add an extra hypothesis like a before,\nafter, parameterisation etc … (the direction itself).\nIt becomes even more difficult, when we use conditional probabilities\nas in QM measurements to describe the outcomes.\n\nNow, to end, I will try to show a possible implementation example of a\nformal FTL phone based solely on your "causal" argument.\n\nLet\'s take 2 classical independent particles (no interaction) with a\ntotal kinetic energy E= E1+E2.\nNow if we change the kinetic energy of particle 1, we thus induce\ninstantaneously (causality) a change on the kinetic energy of the 2\nparticles (E): We have formally a FTL phone with the total Energy E.\n(but we don\'t change the kinetic energy of particle 2).\n\nAt the location of particle 2, we still have the right to say, yes,\nthe kinetic energy E has changed FTL due to a local interaction on\nparticle 1 (particle that may be at "the end" of the universe):\n(Modification of E1) => (Modification of E) (A => B)\n\nNow, I may even refine my FTL phone and say that I have a logical\nblack box that measures the total kinetic energy E. Therefore, I can\nuse this black box at the location of particle 2 to detect an FTL\nchange.\n\nAs everybody has already noticed, I have moved the problem to the\nmeasurement of a global variable E.\n\nThis seems obvious, but now we can go back to the EPR state |Psi> of 2\nparticles. An EPR state is a global state. The main difficulty with\nthis global state is to understand its "internal statistics" (as it\nmay no be derived from a simple Local hidden variable statistical\ntheory).\n\nSo when we say that a local measurement on one EPR particle implies\nthe change of the global state |psi> we are very close to the previous\nexample. To detect the change of |psi> at the other particle location,\nwe should have a black box (like the previous one for the total energy\nE) that is able to measure the global state (and not the local one).\n\nSeratend.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cl7p80$dck$1@beech.fernuni-hagen.de>...
> > However, the interference of results in an EPR like setup are not more
> > surprising than the interference of a split light beam that went
> > around the building left and right and came back.
>
> No. This type of (classical wave) interference may be explained in
> a classical realistic local theory, violations of Bell's inequality not.
>
> > What is surprising
> > in EPR is that if you consider that there has been a measurement at
> > Alice's and Bob's, then there ought to be a breaking of locality (I
> > know you are in favor of that conclusion). But then it is surprising
> > that we cannot build an FTL phone using that procedure.
>
> It is not very surprising. It is a consequence of the possibility of two
> explanations: A->B or B->A. Whenever we have two explanations
> for some observed correlation, we cannot use the observation for
> our own information transfer, because the transfer of information
> A->B would contradict the possible explanation B->A.
>
> Thus, every causal connection which does not allow us to observe
> its direction has this property.
>
> > We can only
> > notice the effect when, through normal, sub-luminal signalling, we
> > have access to both measurements. So what is the conspiracy ?
>
> I don't see much conspiracy. I see that we observe a causal
> connection but are unable to observe its direction. That's the whole
> conspiracy: A typical case of indirect, incomplete observation.
> The normal conclusion is that there is a causal connection
> with some causal direction which is hidden from observation.
>
> Ilja

After all these posts, as an external reader, I still not understand
the point of view of Ilia nor what could be its FTL phone.
First what do you intend by causality? For me, like with the QM
measurement, it is very difficult to understand what it is or what it
says (human interpretation).

Therefore, I always try to restrict its meaning to each experiment I
do or think: I prefer to consider causality, when I can, with an
interaction model and thus a time evolution: a near mathematical
formulation.

The most striking feature with the explanations you give, is the need
of the "causality" to explain "physically" your possible FTL phone.
(as you have removed all the physical model tools such as
interactions, … that may help us to understand what do you intend
really by causality and what is your FTL phone).

So, Do you intend "->" (your causality of the FTL phone) as the
mathematical logic symbol "=>" ?
If yes, when I read "It is not very surprising. It is a consequence of
the possibility of two explanations: A->B or B->A", I understand (A=>
B) logical OR (B => A). Thus, the statement "Whenever we have two
explanations for some observed correlation" I understand you intend
that you have both (A=> B) and (B => A) true.
Thus A <=> B. So if I reinterpret <=>, in your logical context I would
say "A is B" thus explaining the mysterious "we cannot use the
observation for our own information transfer" because in that case we
are dealing with the same object (for an information transfer we need
2 points).

What is indeed, more difficult to understand with your point of view
is that a logical "A => B" is formally equivalent to the statement
"non B => non A": the point of view of 2 possible observers.

> Thus, every causal connection which does not allow us to observe
> its direction has this property.

Thus the "causal connection direction" is very difficult to define in
your statements if you do not add an extra hypothesis like a before,
after, parameterisation etc … (the direction itself).
It becomes even more difficult, when we use conditional probabilities
as in QM measurements to describe the outcomes.

Now, to end, I will try to show a possible implementation example of a
formal FTL phone based solely on your "causal" argument.

Let's take 2 classical independent particles (no interaction) with a
total kinetic energy E= E1+E2.
Now if we change the kinetic energy of particle 1, we thus induce
instantaneously (causality) a change on the kinetic energy of the 2
particles (E): We have formally a FTL phone with the total Energy E.
(but we don't change the kinetic energy of particle 2).

At the location of particle 2, we still have the right to say, yes,
the kinetic energy E has changed FTL due to a local interaction on
particle 1 (particle that may be at "the end" of the universe):
(Modification of E1) => (Modification of E) (A => B)

Now, I may even refine my FTL phone and say that I have a logical
black box that measures the total kinetic energy E. Therefore, I can
use this black box at the location of particle 2 to detect an FTL
change.

As everybody has already noticed, I have moved the problem to the
measurement of a global variable E.

This seems obvious, but now we can go back to the EPR state |\Psi> of 2
particles. An EPR state is a global state. The main difficulty with
this global state is to understand its "internal statistics" (as it
may no be derived from a simple Local hidden variable statistical
theory).

So when we say that a local measurement on one EPR particle implies
the change of the global state |\psi> we are very close to the previous
example. To detect the change of |\psi> at the other particle location,
we should have a black box (like the previous one for the total energy
E) that is able to measure the global state (and not the local one).

Seratend.

Ralph Hartley

Oct27-04, 10:55 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n\nPatrick Van Esch wrote:\n> If we now assume that Bob and Alice, and their notes of experiments,\n> remain in such a superposed state UNTIL we can make a measurement on\n> the whole system, which means: when their notes are transported to\n> acommon place, then we find the probabilities of measurement results\n> using Born\'s rule, which are applied locally where the notes are\n> compared\n\nSo the experiment can be viewed as involving *three* measurements, not two.\nOne each performed by Alice and Bob, and another when they compare notes.\n\nIn this view you can think of the whole experiment as a very big\ninterferometer, that is so well balanced that even apparently classical\nstates can interfere.\n\nIt works only because time evolution is *perfectly* unitary (or at least\nvery *very* close). |A+> and |A-> are orthogonal because they evolved from\n|+>|A0> and |->|A0>, and they *stay* orthogonal.\n\nAny *local* objective collapse process would ruin the correlation between\nBob and Alice. I think this may be a very strong test of Penrose\'s\n"gravitational wavefunction collapse". OK, not if it isn\'t local, I was\nnever too clear on that point.\n\nFrom Bob\'s point of view, his measurement partially collapses Alice\'s\nstate. He knows the polarization of her photon, but may not what her\nmeasurement or result were. The collapse is instantaneous. It doesn\'t\nmatter what reference frame it is instantaneous in, since Bob sees the same\nthing in any frame.\n\nFrom Alice\'s point of view the situation is reversed.\n\nFrom the point of view of Eve, located half way in between, observing\nAlice and Bob with telescopes, they both remain in a superposition until\nshe observes them making their measurements.\n\nAll three points of view are completely objective, but none is more "true"\nthan the others. (Except for *mine*, which is the *absolute* truth :-))\n\nYou *can* view the experiment that way, but there is nothing to say that\nyou *must* do so. Interpretations of QM are interchangeable, since they\nnever disagree on results. You should always use the one that works best\nfor you in the situation at hand.\n\nRalph Hartley\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch wrote:
> If we now assume that Bob and Alice, and their notes of experiments,
> remain in such a superposed state UNTIL we can make a measurement on
> the whole system, which means: when their notes are transported to
> acommon place, then we find the probabilities of measurement results
> using Born's rule, which are applied locally where the notes are
> compared

So the experiment can be viewed as involving *three* measurements, not two.
One each performed by Alice and Bob, and another when they compare notes.

In this view you can think of the whole experiment as a very big
interferometer, that is so well balanced that even apparently classical
states can interfere.

It works only because time evolution is *perfectly* unitary (or at least
very *very* close). |A+> and |A-> are orthogonal because they evolved from
|+>|A0> and |->|A0>, and they *stay* orthogonal.

Any *local* objective collapse process would ruin the correlation between
Bob and Alice. I think this may be a very strong test of Penrose's
"gravitational wavefunction collapse". OK, not if it isn't local, I was
never too clear on that point.

From Bob's point of view, his measurement partially collapses Alice's
state. He knows the polarization of her photon, but may not what her
measurement or result were. The collapse is instantaneous. It doesn't
matter what reference frame it is instantaneous in, since Bob sees the same
thing in any frame.

From Alice's point of view the situation is reversed.

From the point of view of Eve, located half way in between, observing
Alice and Bob with telescopes, they both remain in a superposition until
she observes them making their measurements.

All three points of view are completely objective, but none is more "true"
than the others. (Except for *mine*, which is the *absolute* truth :-))

You *can* view the experiment that way, but there is nothing to say that
you *must* do so. Interpretations of QM are interchangeable, since they
never disagree on results. You should always use the one that works best
for you in the situation at hand.

Ralph Hartley

Patrick Van Esch

Oct27-04, 10:56 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cl7p80\\$dck\\$1@beech.fernuni-hagen.de>...\n> "Patrick Van Esch" <vanesch@ill.fr> schrieb\n\n> > Not really. A working FTL phone would allow me, as a single observer,\n> > to conclude that the signal I sent went up and down Titan in 0.5\n> > seconds.\n>\n> How?\n>\n\nBy triggering the nuclear bombs on titan with the special code I have.\nIf I see a big fireball in the sky after 1 light-speed propagation\ntime after I sent the message, I know my code arrived almost\nimmediately.\nYou cannot do that with an EPR like set-up, and what I wanted to\nillustrate was (see my other post in this thread) that if you consider\nthe measurements at Bob and Alice as just decohered superpositions,\nthen there\'s a very natural and local explanation of the so-called\nnon-locality: namely a state that went though two different paths, and\nits amplitudes for each path interfere when they are added together.\nAs such it is also obvious that you can never get this FTL phone in\nthis way. You shouldn\'t take my stuff on solipsism too seriously. It\nwas just meant to illustrate that (the way I understand current QM)\nthe wavefunction has only a reality with respect to one single\nobserver (which turns out to be indistinguishable to all timelike\nconnected events - there might be something in that, but I don\'t know\nwhat). It is only when we take the same wavefunction to describe an\nindependent reality on space-like connected events that we seem to run\ninto non-locality problems when taking the concept of measurement\n(probably too) so far. I\'m not religiously attached to anything\n(locality, linearity of QM, whatever). I\'m not trying to modify\nanything to the currently accepted principles ; after all they work\nwell in practice. I\'m just trying to make sense of that situation\nwithin the framework as presented, and I think I can, up to a point,\nwhen you consider that all an observer will ever do is make\nobservations which are timelike connected.\nBut this doesn\'t mean - as you seem to suggest - that he cannot make\ntests about a genuine non-locality. These tests are then of course\nlimited to matter or information-carrying faster-than-light transport,\nbut they explain in a natural way why this\n"state-projection-at-a-distance" cannot be used for that, simply\nbecause there is no such thing: the only projections that occur are\nlocal to the observer, and as such they explain completely the EPR\nresults as LOCAL interference between two different paths taken by a\nquantum system and then brought together.\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cl7p80$dck$1@beech.fernuni-hagen.de>...
> "Patrick Van Esch" <vanesch@ill.fr> schrieb

> > Not really. A working FTL phone would allow me, as a single observer,
> > to conclude that the signal I sent went up and down Titan in .5
> > seconds.
>
> How?
>

By triggering the nuclear bombs on titan with the special code I have.
If I see a big fireball in the sky after 1 light-speed propagation
time after I sent the message, I know my code arrived almost
immediately.
You cannot do that with an EPR like set-up, and what I wanted to
illustrate was (see my other post in this thread) that if you consider
the measurements at Bob and Alice as just decohered superpositions,
then there's a very natural and local explanation of the so-called
non-locality: namely a state that went though two different paths, and
its amplitudes for each path interfere when they are added together.
As such it is also obvious that you can never get this FTL phone in
this way. You shouldn't take my stuff on solipsism too seriously. It
was just meant to illustrate that (the way I understand current QM)
the wavefunction has only a reality with respect to one single
observer (which turns out to be indistinguishable to all timelike
connected events - there might be something in that, but I don't know
what). It is only when we take the same wavefunction to describe an
independent reality on space-like connected events that we seem to run
into non-locality problems when taking the concept of measurement
(probably too) so far. I'm not religiously attached to anything
(locality, linearity of QM, whatever). I'm not trying to modify
anything to the currently accepted principles ; after all they work
well in practice. I'm just trying to make sense of that situation
within the framework as presented, and I think I can, up to a point,
when you consider that all an observer will ever do is make
observations which are timelike connected.
But this doesn't mean - as you seem to suggest - that he cannot make
tests about a genuine non-locality. These tests are then of course
limited to matter or information-carrying faster-than-light transport,
but they explain in a natural way why this
"state-projection-at-a-distance" cannot be used for that, simply
because there is no such thing: the only projections that occur are
local to the observer, and as such they explain completely the EPR
results as LOCAL interference between two different paths taken by a
quantum system and then brought together.

cheers,
Patrick.

Thomas Trotter

Oct27-04, 10:56 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n[snip]\n\nPatrick Van Esch:\n> > However, the interference of results in an EPR like setup are not more\n> > surprising than the interference of a split light beam that went\n> > around the building left and right and came back.\n\nIlja Schmelzer:\n> No. This type of (classical wave) interference may be explained in\n> a classical realistic local theory, violations of Bell\'s inequality not.\n\nTT:\nNot yet anyway. :-) Maybe it\'s just that knowledge of what is\nactually happening wrt the emission and detection of the opposite\nmoving light beams is incomplete (ie., that there *is* a local,\ncommon-property reason for the correlations, and not enough is\nknown about the physical process to talk about it in a straighforward\nlocal realistic way yet).\n\n[snip]\n\nPatrick Van Esch:\n> > We can only notice the effect when, through normal, sub-luminal\n> > signalling, we have access to both measurements. So what is the\n> > conspiracy ?\n\nIlja Schmelzer:\n> I don\'t see much conspiracy. I see that we observe a causal\n> connection but are unable to observe its direction. That\'s the whole\n> conspiracy: A typical case of indirect, incomplete observation.\n> The normal conclusion is that there is a causal connection\n> with some causal direction which is hidden from observation.\n\nTT:\nIt seems clear to me that we don\'t observe a causal\nconnection between A and B (though it might be inferred\nvia the absence of a non-contradictory local model).\nThe causal contexts are emitter-->polarizer-->detectorA\nand emitter-->polarizer-->detectorB. The correlations\naren\'t produced in those contexts.\n\nIn the correlational context there are two\nalternative explanations: either the crossed linear\npolarizers are analyzing a common property of\nthe opposite moving light beams due to their\nemission from the same oscillator (and, via Bell\'s\nanalysis we don\'t know enough about the process\nto formulate it realistically in a non-contradictory\nway), or there is some sort of instantaneous physical\nmechanism, A-->B and B-->A.\n\nSince all observational evidence suggests a\nlocal reality, I think it\'s too early too assume\nthat instantaneous mechanisms or FTL processes\n*exist* in any sense other than as artifacts\nof our imagination and/or ignorance.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>[snip]

Patrick Van Esch:
> > However, the interference of results in an EPR like setup are not more
> > surprising than the interference of a split light beam that went
> > around the building left and right and came back.

Ilja Schmelzer:
> No. This type of (classical wave) interference may be explained in
> a classical realistic local theory, violations of Bell's inequality not.

TT:
Not yet anyway. :-) Maybe it's just that knowledge of what is
actually happening wrt the emission and detection of the opposite
moving light beams is incomplete (ie., that there *is* a local,
common-property reason for the correlations, and not enough is
known about the physical process to talk about it in a straighforward
local realistic way yet).

[snip]

Patrick Van Esch:
> > We can only notice the effect when, through normal, sub-luminal
> > signalling, we have access to both measurements. So what is the
> > conspiracy ?

Ilja Schmelzer:
> I don't see much conspiracy. I see that we observe a causal
> connection but are unable to observe its direction. That's the whole
> conspiracy: A typical case of indirect, incomplete observation.
> The normal conclusion is that there is a causal connection
> with some causal direction which is hidden from observation.

TT:
It seems clear to me that we don't observe a causal
connection between A and B (though it might be inferred
via the absence of a non-contradictory local model).
The causal contexts are emitter-->polarizer-->detectorA
and emitter-->polarizer-->detectorB. The correlations
aren't produced in those contexts.

In the correlational context there are two
alternative explanations: either the crossed linear
polarizers are analyzing a common property of
the opposite moving light beams due to their
emission from the same oscillator (and, via Bell's
analysis we don't know enough about the process
to formulate it realistically in a non-contradictory
way), or there is some sort of instantaneous physical
mechanism, A-->B and B-->A.

Since all observational evidence suggests a
local reality, I think it's too early too assume
that instantaneous mechanisms or FTL processes
*exist* in any sense other than as artifacts
of our imagination and/or ignorance.

Ilja Schmelzer

Oct27-04, 10:59 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"seratend" <ser_monmail@yahoo.fr> schrieb\n> After all these posts, as an external reader, I still not understand\n> the point of view of Ilia nor what could be its FTL phone.\n> First what do you intend by causality? For me, like with the QM\n> measurement, it is very difficult to understand what it is or what it\n> says (human interpretation).\n\nI press a button by my free choice, and whenever I press it the\nlight bulb changes its state. In this case, I conclude that there is\na causal influence from my behaviour to the state of the light bulb.\n\nI make this conclusion based on the observation of some correlations.\nOf course, I also use some other assumptions: That there is some\nreality which defines the state of the light bulb, and certain assumptions\nabout independence of my free choices from external influences.\n\nI believe that these basic assumptions are the foundation of the\nscientific method, that it is reasonable to hold them even in\ncircumstances where no simple realistic explanation is known.\n\n> Therefore, I always try to restrict its meaning to each experiment I\n> do or think: I prefer to consider causality, when I can, with an\n> interaction model and thus a time evolution: a near mathematical\n> formulation.\n\nI think about causality on a more fundamental level. The observation\nof a correlation between A and B is sufficient to search for a realistic\ncausal explanation. Such explanations are a common cause\nC->A, C->B or a direct causal influence A->B or B->A. Time has\nnot been mentioned in this consideration.\n\n> The most striking feature with the explanations you give, is the need\n> of the "causality" to explain "physically" your possible FTL phone.\n> (as you have removed all the physical model tools such as\n> interactions, . that may help us to understand what do you intend\n> really by causality and what is your FTL phone).\n\nMy FTL phone is only a hypothetical feature.\n\nI see, I have not sufficiently explained that it is based on Popper\'s\nfalsifiability of scientific, physical theory. To distinghish science from\npseudosciences Popper has proposed that the distinguishing\nproperty of physical theories is that they make nontrivial predictions,\npredictions which may appear false if we test them. Such theories\nare called falsifiable. Theories which cannot be falsified are not\nempirical theories, they are metaphysical. This point of view on science\nis IMHO correct and widely accepted among scientists.\n\nNow, Einstein causality is, in the general understanding, a physical\ntheory. It predicts that no FTL phones are possible. At least, this is\nthe usual interpretation of Einstein causality.\n\nWhat would happen if we would really observe some FTL phone?\nVery simple, Einstein causality would be falsified. Nobody would\ndefend relativity in this case. It would survive in physics as a limit\nof some other theory which allows FTL phones.\n\nThis is not the case today. Using the strange correlations of EPR\nand Bell we cannot build FTL phones. This is a simple theorem.\n\nNonetheless, that does not mean there is no problem for relativity.\nThe only realistic explanations of these correlations are A->B or\nB->A, above are in contradiction with relativity. Now, defenders\nof relativity use various argumentations. But, as I have observed,\ncould be applied as well in the hypothetical case of falsification of\nrelativity. That means, if we take them seriously, we can no longer\npredict that no FTL phone exists - because such an FTL phone\ncould be explained away if we observe it.\n\n> So, Do you intend "->" (your causality of the FTL phone) as the\n> mathematical logic symbol "=>" ?\n\nNo. In mathematical terms it means absence of independence.\nRealism assumes\n\nResult_at_A = f_a(reality, decisions_at_A,decisions_at_B)\nResult_at_B = f_b(reality, decisions_at_A,decisions_at_B)\n\nCausality gives the additional restriction\n\nResult_at_A = f_a(reality, decisions_at_A)\nResult_at_B = f_b(reality, decisions_at_B)\n\n> Now, to end, I will try to show a possible implementation example of a\n> formal FTL phone based solely on your "causal" argument.\n>\n> Let\'s take 2 classical independent particles (no interaction) with a\n> total kinetic energy E= E1+E2.\n> Now if we change the kinetic energy of particle 1, we thus induce\n> instantaneously (causality) a change on the kinetic energy of the 2\n> particles (E): We have formally a FTL phone with the total Energy E.\n\nThat\'s not a phone because we cannot use it to talk with each other,\nnor formal nor informal.\n\n> Now, I may even refine my FTL phone and say that I have a logical\n> black box that measures the total kinetic energy E. Therefore, I can\n> use this black box at the location of particle 2 to detect an FTL\n> change.\n\nHaving such a black box would allow to construct an FTL phone.\n\n> As everybody has already noticed, I have moved the problem to the\n> measurement of a global variable E.\n>\n> This seems obvious, but now we can go back to the EPR state |Psi> of 2\n> particles. An EPR state is a global state. The main difficulty with\n> this global state is to understand its "internal statistics" (as it\n> may no be derived from a simple Local hidden variable statistical\n> theory).\n>\n> So when we say that a local measurement on one EPR particle implies\n> the change of the global state |psi> we are very close to the previous\n> example. To detect the change of |psi> at the other particle location,\n> we should have a black box (like the previous one for the total energy\n> E) that is able to measure the global state (and not the local one).\n\nThe whole point is a different one. We really have some black box.\nThis black box allows to do very strange things. Unfortunately it\ndoes not allow us to build an FTL phone. But there exists no realistic\nexplanation of what is inside the black box which does not use some\nsort of hidden FTL information transfer.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"seratend" <ser_monmail@yahoo.fr> schrieb
> After all these posts, as an external reader, I still not understand
> the point of view of Ilia nor what could be its FTL phone.
> First what do you intend by causality? For me, like with the QM
> measurement, it is very difficult to understand what it is or what it
> says (human interpretation).

I press a button by my free choice, and whenever I press it the
light bulb changes its state. In this case, I conclude that there is
a causal influence from my behaviour to the state of the light bulb.

I make this conclusion based on the observation of some correlations.
Of course, I also use some other assumptions: That there is some
reality which defines the state of the light bulb, and certain assumptions
about independence of my free choices from external influences.

I believe that these basic assumptions are the foundation of the
scientific method, that it is reasonable to hold them even in
circumstances where no simple realistic explanation is known.

> Therefore, I always try to restrict its meaning to each experiment I
> do or think: I prefer to consider causality, when I can, with an
> interaction model and thus a time evolution: a near mathematical
> formulation.

I think about causality on a more fundamental level. The observation
of a correlation between A and B is sufficient to search for a realistic
causal explanation. Such explanations are a common cause
C->A, C->B or a direct causal influence A->B or B->A. Time has
not been mentioned in this consideration.

> The most striking feature with the explanations you give, is the need
> of the "causality" to explain "physically" your possible FTL phone.
> (as you have removed all the physical model tools such as
> interactions, . that may help us to understand what do you intend
> really by causality and what is your FTL phone).

My FTL phone is only a hypothetical feature.

I see, I have not sufficiently explained that it is based on Popper's
falsifiability of scientific, physical theory. To distinghish science from
pseudosciences Popper has proposed that the distinguishing
property of physical theories is that they make nontrivial predictions,
predictions which may appear false if we test them. Such theories
are called falsifiable. Theories which cannot be falsified are not
empirical theories, they are metaphysical. This point of view on science
is IMHO correct and widely accepted among scientists.

Now, Einstein causality is, in the general understanding, a physical
theory. It predicts that no FTL phones are possible. At least, this is
the usual interpretation of Einstein causality.

What would happen if we would really observe some FTL phone?
Very simple, Einstein causality would be falsified. Nobody would
defend relativity in this case. It would survive in physics as a limit
of some other theory which allows FTL phones.

This is not the case today. Using the strange correlations of EPR
and Bell we cannot build FTL phones. This is a simple theorem.

Nonetheless, that does not mean there is no problem for relativity.
The only realistic explanations of these correlations are A->B or
B->A, above are in contradiction with relativity. Now, defenders
of relativity use various argumentations. But, as I have observed,
could be applied as well in the hypothetical case of falsification of
relativity. That means, if we take them seriously, we can no longer
predict that no FTL phone exists - because such an FTL phone
could be explained away if we observe it.

> So, Do you intend "->" (your causality of the FTL phone) as the
> mathematical logic symbol "=>" ?

No. In mathematical terms it means absence of independence.
Realism assumes

Result_at_A = f_a(reality, decisions_at_A,decisions_at_B)Result_at_B = f_b(reality, decisions_at_A,decisions_at_B)

Causality gives the additional restriction

Result_at_A = f_a(reality, decisions_at_A)Result_at_B = f_b(reality, decisions_at_B)

> Now, to end, I will try to show a possible implementation example of a
> formal FTL phone based solely on your "causal" argument.
>
> Let's take 2 classical independent particles (no interaction) with a
> total kinetic energy E= E1+E2.
> Now if we change the kinetic energy of particle 1, we thus induce
> instantaneously (causality) a change on the kinetic energy of the 2
> particles (E): We have formally a FTL phone with the total Energy E.

That's not a phone because we cannot use it to talk with each other,
nor formal nor informal.

> Now, I may even refine my FTL phone and say that I have a logical
> black box that measures the total kinetic energy E. Therefore, I can
> use this black box at the location of particle 2 to detect an FTL
> change.

Having such a black box would allow to construct an FTL phone.

> As everybody has already noticed, I have moved the problem to the
> measurement of a global variable E.
>
> This seems obvious, but now we can go back to the EPR state |\Psi> of 2
> particles. An EPR state is a global state. The main difficulty with
> this global state is to understand its "internal statistics" (as it
> may no be derived from a simple Local hidden variable statistical
> theory).
>
> So when we say that a local measurement on one EPR particle implies
> the change of the global state |\psi> we are very close to the previous
> example. To detect the change of |\psi> at the other particle location,
> we should have a black box (like the previous one for the total energy
> E) that is able to measure the global state (and not the local one).

The whole point is a different one. We really have some black box.
This black box allows to do very strange things. Unfortunately it
does not allow us to build an FTL phone. But there exists no realistic
explanation of what is inside the black box which does not use some
sort of hidden FTL information transfer.

Ilja

Ilja Schmelzer

Oct28-04, 01:28 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Thomas Trotter" <thomastrotter2005@juno.com> schrieb\n> Patrick Van Esch:\n> > > However, the interference of results in an EPR like setup are not more\n> > > surprising than the interference of a split light beam that went\n> > > around the building left and right and came back.\n>\n> Ilja Schmelzer:\n> > No. This type of (classical wave) interference may be explained in\n> > a classical realistic local theory, violations of Bell\'s inequality not.\n>\n> TT:\n> Not yet anyway. :-) Maybe it\'s just that knowledge of what is\n> actually happening wrt the emission and detection of the opposite\n> moving light beams is incomplete (ie., that there *is* a local,\n> common-property reason for the correlations, and not enough is\n> known about the physical process to talk about it in a straighforward\n> local realistic way yet).\n\nGiven Bell\'s theorem, there is no such explanation. This is independent\nof our state of knowledge or state of our theories.\n\n> Ilja Schmelzer:\n> > I don\'t see much conspiracy. I see that we observe a causal\n> > connection but are unable to observe its direction. That\'s the whole\n> > conspiracy: A typical case of indirect, incomplete observation.\n> > The normal conclusion is that there is a causal connection\n> > with some causal direction which is hidden from observation.\n>\n> TT:\n> It seems clear to me that we don\'t observe a causal\n> connection between A and B (though it might be inferred\n> via the absence of a non-contradictory local model).\n> The causal contexts are emitter-->polarizer-->detectorA\n> and emitter-->polarizer-->detectorB. The correlations\n> aren\'t produced in those contexts.\n\nThe causal context is\ndecision of the experimenter at A -> observed result at B\nor\ndecision of the experimenter at B -> observed result at A\n\nIt is only these things which are used in Bell\'s inequality.\n\n> In the correlational context there are two\n> alternative explanations: either the crossed linear\n> polarizers are analyzing a common property of\n> the opposite moving light beams due to their\n> emission from the same oscillator (and, via Bell\'s\n> analysis we don\'t know enough about the process\n> to formulate it realistically in a non-contradictory\n> way),\n\nNo. We know enough to exclude this possibility.\n\n> or there is some sort of instantaneous physical\n> mechanism, A-->B and B-->A.\n\n> Since all observational evidence suggests a\n> local reality, I think it\'s too early too assume\n> that instantaneous mechanisms or FTL processes\n> *exist* in any sense other than as artifacts\n> of our imagination and/or ignorance.\n\nWe are discussing here the observational evidence\nwhich strongly falsifies local reality which would be\ngiven by an ideal Bell device. Existing evidence\nstrongly suggests that such a device will definitely\nfalsify Einstein-local reality.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Thomas Trotter" <thomastrotter2005@juno.com> schrieb
> Patrick Van Esch:
> > > However, the interference of results in an EPR like setup are not more
> > > surprising than the interference of a split light beam that went
> > > around the building left and right and came back.
>
> Ilja Schmelzer:
> > No. This type of (classical wave) interference may be explained in
> > a classical realistic local theory, violations of Bell's inequality not.
>
> TT:
> Not yet anyway. :-) Maybe it's just that knowledge of what is
> actually happening wrt the emission and detection of the opposite
> moving light beams is incomplete (ie., that there *is* a local,
> common-property reason for the correlations, and not enough is
> known about the physical process to talk about it in a straighforward
> local realistic way yet).

Given Bell's theorem, there is no such explanation. This is independent
of our state of knowledge or state of our theories.

> Ilja Schmelzer:
> > I don't see much conspiracy. I see that we observe a causal
> > connection but are unable to observe its direction. That's the whole
> > conspiracy: A typical case of indirect, incomplete observation.
> > The normal conclusion is that there is a causal connection
> > with some causal direction which is hidden from observation.
>
> TT:
> It seems clear to me that we don't observe a causal
> connection between A and B (though it might be inferred
> via the absence of a non-contradictory local model).
> The causal contexts are emitter-->polarizer-->detectorA
> and emitter-->polarizer-->detectorB. The correlations
> aren't produced in those contexts.

The causal context is
decision of the experimenter at A -> observed result at B
or
decision of the experimenter at B -> observed result at A

It is only these things which are used in Bell's inequality.

> In the correlational context there are two
> alternative explanations: either the crossed linear
> polarizers are analyzing a common property of
> the opposite moving light beams due to their
> emission from the same oscillator (and, via Bell's
> analysis we don't know enough about the process
> to formulate it realistically in a non-contradictory
> way),

No. We know enough to exclude this possibility.

> or there is some sort of instantaneous physical
> mechanism, A-->B and B-->A.

> Since all observational evidence suggests a
> local reality, I think it's too early too assume
> that instantaneous mechanisms or FTL processes
> *exist* in any sense other than as artifacts
> of our imagination and/or ignorance.

We are discussing here the observational evidence
which strongly falsifies local reality which would be
given by an ideal Bell device. Existing evidence
strongly suggests that such a device will definitely
falsify Einstein-local reality.

Ilja

rof@maths.tcd.ie

Oct28-04, 01:28 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n>> ><rof@maths.tcd.ie> schrieb\n>> >> That is, he must form a data representation which compresses\n>> >> the history of symbols he has received into a single "state\n>> >> of the world", call it X_n, where\n>> >> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)\n>> >> (something which was not true of the symbols E_n).\n\n>> The construction of the X_n is perception (of the world).\n\n>I\'m not a native speaker, but this sounds strange. I would\n>accept it if you call the M_n "perception".\n\nThere is a problem with cognitive dissonance here, since, according\nto popular wisdom, perception is a passive, receptive act wherein\nthe observer merely receives perceptions which have been thrown\nat him from somewhere. This view is incorrect, since passive\nreceptivity is called sensation, and presents to the observer only\nsounds and colours and so on. Perception is what leads to awareness\nof objects, and requires that the data taken from sensation (the E_n)\nbe actively processed in order to generate the representations of\nobjects (the X_n). Hence, perception is an act which involves\nthe active construction of the thing which is perceived, and not\nmerely the passive reception of an already-formed representation.\n\nWhen this obviously true fact is presented to somebody unused to\nthinking about these matters, the confusion which arises from\nthe clash with their existing way of understanding the process\nof perception makes them quickly hurry on to the next sentence,\nhoping that the confusion will go away.\n\nThese matters are relevant for physics because the statistics of\nthe E_n either are or are not representable as a Markov process\nwhich factorises in a spatial way, and so these considerations\nmust arise whenever one wants to ask whether an observer will\nsee space.\n\n>> In an ontological theory the splitting needs to be deduced,\n>> rather than postulated.\n\n>The ontological theory postulates the structure and the laws of\n>the X_n. If these have the form of something localized in a space,\n>space is postulated. What you perceive is another question,\n>in the E_n we have, for example, nothing three-dimensional.\n>We see only two 2D movies. So it is completely unclear what\n>means spatial structure in the E_n.\n\nNo; the ontological theory can\'t make any postulates about the X_n,\nbecause the X_n exist only in the mind of the observer, who has\nconstructed them. What the observer does is say to himself "Good\ngracious - I appear to be receiving symbols E_n and they appear to\nbe somewhat predictable. It\'s very tedious to predict them based\non the previous E_n that have been received, because I have to look\nfar into the past in order to do that. Instead, I\'ll make a new\nrepresentation, X_n, which, at time n, incorporates all the information\nfrom the previous E_n which is relevant for predicting the future\nE_n. I\'ll call X_n \'my representation of the state of the world at\npresent.\'"\n\nIf a theory makes postulates about the X_n, then it is making\npostulates about the kind of data processing done by a\nvery complicated object (the observer). This may be compared\nto postulates about the shape of bananas, insofar as both\nsets of postulates (about what kind of processing the observer\ndoes and what shape a banana has) have no place in an ontological\ntheory. If one wants to have an ontological physical theory and\nsay something about what kind of data processing is done by\nsuch and such an object, or what shape such and such an object\nwill have, one must explain the data processing or the shape\nin terms of the underlying particles or whatever else the\ntheory postulates. That is, such statements must be theorems\nand not postulates.\n\n>> >> I was talking about keeping the symmetry group (the Lorentz group) -\n>> >> not about keeping locality. Nonlocality has been experimentally\n>> >> confirmed, so there\'s no point in trying to salvage locality.\n>>\n>> >In this case, you have to give up causality. This seems as unjustified\n>> >as giving up realism.\n>>\n>> You do, but giving up either realism or causality won\'t affect your\n>> ability to predict X_{n+1} from X_n, since they are merely rules of\n>> thumb for generating representations.\n\n>No, the rules of thumb are questioned. I have no way to construct\n>some X_n in the usual (causal, realistic) meaning which has the\n>property of Lorentz-invariance.\n\nLorentz invariance and relativity are important when one needs to\ncompare one observer\'s description (the series X_n) with another\'s.\nIn my description this is not a simple matter, since there is no\nclear way to perform the comparison. Various things, induction\nand so on, lead us to suppose that other people will behave\nas though they are generating their own X_n, with space and\nso forth, but that entire thought process is pretty difficult\nto model. I\'ll get around to it sometime.\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb
>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
>> ><rof@maths.tcd.ie> schrieb
>> >> That is, he must form a data representation which compresses
>> >> the history of symbols he has received into a single "state
>> >> of the world", call it X_n, where
>> >> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)>> >> (something which was not true of the symbols E_n).

>> The construction of the X_n is perception (of the world).

>I'm not a native speaker, but this sounds strange. I would
>accept it if you call the M_n "perception".

There is a problem with cognitive dissonance here, since, according
to popular wisdom, perception is a passive, receptive act wherein
the observer merely receives perceptions which have been thrown
at him from somewhere. This view is incorrect, since passive
receptivity is called sensation, and presents to the observer only
sounds and colours and so on. Perception is what leads to awareness
of objects, and requires that the data taken from sensation (the E_n)
be actively processed in order to generate the representations of
objects (the X_n). Hence, perception is an act which involves
the active construction of the thing which is perceived, and not
merely the passive reception of an already-formed representation.

When this obviously true fact is presented to somebody unused to
thinking about these matters, the confusion which arises from
the clash with their existing way of understanding the process
of perception makes them quickly hurry on to the next sentence,
hoping that the confusion will go away.

These matters are relevant for physics because the statistics of
the E_n either are or are not representable as a Markov process
which factorises in a spatial way, and so these considerations
must arise whenever one wants to ask whether an observer will
see space.

>> In an ontological theory the splitting needs to be deduced,
>> rather than postulated.

>The ontological theory postulates the structure and the laws of
>the X_n. If these have the form of something localized in a space,
>space is postulated. What you perceive is another question,
>in the E_n we have, for example, nothing three-dimensional.
>We see only two 2D movies. So it is completely unclear what
>means spatial structure in the E_n.

No; the ontological theory can't make any postulates about the X_n,
because the X_n exist only in the mind of the observer, who has
constructed them. What the observer does is say to himself "Good
gracious - I appear to be receiving symbols E_n and they appear to
be somewhat predictable. It's very tedious to predict them based
on the previous E_n that have been received, because I have to look
far into the past in order to do that. Instead, I'll make a new
representation, X_n, which, at time n, incorporates all the information
from the previous E_n which is relevant for predicting the future
E_n. I'll call X_n 'my representation of the state of the world at
present.'"

If a theory makes postulates about the X_n, then it is making
postulates about the kind of data processing done by a
very complicated object (the observer). This may be compared
to postulates about the shape of bananas, insofar as both
sets of postulates (about what kind of processing the observer
does and what shape a banana has) have no place in an ontological
theory. If one wants to have an ontological physical theory and
say something about what kind of data processing is done by
such and such an object, or what shape such and such an object
will have, one must explain the data processing or the shape
in terms of the underlying particles or whatever else the
theory postulates. That is, such statements must be theorems
and not postulates.

>> >> I was talking about keeping the symmetry group (the Lorentz group) -
>> >> not about keeping locality. Nonlocality has been experimentally
>> >> confirmed, so there's no point in trying to salvage locality.
>>
>> >In this case, you have to give up causality. This seems as unjustified
>> >as giving up realism.
>>
>> You do, but giving up either realism or causality won't affect your
>> ability to predict X_{n+1} from X_n, since they are merely rules of
>> thumb for generating representations.

>No, the rules of thumb are questioned. I have no way to construct
>some X_n in the usual (causal, realistic) meaning which has the
>property of Lorentz-invariance.

Lorentz invariance and relativity are important when one needs to
compare one observer's description (the series X_n) with another's.
In my description this is not a simple matter, since there is no
clear way to perform the comparison. Various things, induction
and so on, lead us to suppose that other people will behave
as though they are generating their own X_n, with space and
so forth, but that entire thought process is pretty difficult
to model. I'll get around to it sometime.

R.

Ilja Schmelzer

Oct29-04, 02:07 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> ><rof@maths.tcd.ie> schrieb\n> >> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> >> ><rof@maths.tcd.ie> schrieb\n> >> >> That is, he must form a data representation which compresses\n> >> >> the history of symbols he has received into a single "state\n> >> >> of the world", call it X_n, where\n> >> >> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)\n> >> >> (something which was not true of the symbols E_n).\n>\n> >> The construction of the X_n is perception (of the world).\n>\n> >I\'m not a native speaker, but this sounds strange. I would\n> >accept it if you call the M_n "perception".\n>\n> There is a problem with cognitive dissonance here, since, according\n> to popular wisdom, perception is a passive, receptive act wherein\n> the observer merely receives perceptions which have been thrown\n> at him from somewhere. This view is incorrect, since passive\n> receptivity is called sensation, and presents to the observer only\n> sounds and colours and so on. Perception is what leads to awareness\n> of objects, and requires that the data taken from sensation (the E_n)\n> be actively processed in order to generate the representations of\n> objects (the X_n).\n\nI understand this difference between sensation E_n and perception.\nI have acknowledged it alerady by proposing that the M_n are\nperceptions. But not the X_n.\n\n> Hence, perception is an act which involves\n> the active construction of the thing which is perceived, and not\n> merely the passive reception of an already-formed representation.\n\nYep. The construction of the M_n is of this type. But the X_n\\M_n,\nthe inner part of the moon, is not perceived. Even if I have some\nperception M_moon of the surface of the moon and some vague\ntheoretical ideas what may be inside.\n\n> These matters are relevant for physics because the statistics of\n> the E_n either are or are not representable as a Markov process\n> which factorises in a spatial way, and so these considerations\n> must arise whenever one wants to ask whether an observer will\n> see space.\n\nSorry, but you cannot derive if observers will "see" space.\n"Seeing" space means developing a certain realistic theory,\nmay be implicit (as a hidden prejudice, instinct, created by\nevolution).\n\n> >> In an ontological theory the splitting needs to be deduced,\n> >> rather than postulated.\n>\n> >The ontological theory postulates the structure and the laws of\n> >the X_n. If these have the form of something localized in a space,\n> >space is postulated. What you perceive is another question,\n> >in the E_n we have, for example, nothing three-dimensional.\n> >We see only two 2D movies. So it is completely unclear what\n> >means spatial structure in the E_n.\n\n> No; the ontological theory can\'t make any postulates about the X_n,\n> because the X_n exist only in the mind of the observer, who has\n> constructed them.\n\nThat\'s a confusion of language levels. A believes that there are\nsome X_n in reality. This is an ontological theory proposed by A.\nWe have to decide which language level we want to use - the level\nof A, then we talk about X_n as real states, or on a metalevel where\nwe talk about A\'s X_n as constructions of his mind. Of course, on\nthis metalevel we have our metatheory about reality with our own\nX\'_n. (This consideration was on meta-meta-level)\n\n> What the observer does is say to himself "Good\n> gracious - I appear to be receiving symbols E_n and they appear to\n> be somewhat predictable. It\'s very tedious to predict them based\n> on the previous E_n that have been received, because I have to look\n> far into the past in order to do that. Instead, I\'ll make a new\n> representation, X_n, which, at time n, incorporates all the information\n> from the previous E_n which is relevant for predicting the future\n> E_n. I\'ll call X_n \'my representation of the state of the world at\n> present.\'"\n\nFirst, distinguish the X_n from the M_n. Then, that\'s not what we do.\n\n> If a theory makes postulates about the X_n, then it is making\n> postulates about the kind of data processing done by a\n> very complicated object (the observer).\n\nSorry, but the state of a realistic theory X_n cannot be derived\nfrom observations E_n. The miserable part of them which can be\nhas been denoted by M_n. (BTW, there is no such process\nE_n => M_n, the creation of the M_n essentially depends on\nthe general ontological theory about reality.)\n\nThe game is reverse. The theory is prior. Then, a complicated\ndata processing starts which makes the E_n compatible with the\ntheory using some M_n. A failure of this process is named\nfalsification.\n\nTheories about the data processing are part of some historical\nsciences.\n\n> >> You do, but giving up either realism or causality won\'t affect your\n> >> ability to predict X_{n+1} from X_n, since they are merely rules of\n> >> thumb for generating representations.\n>\n> >No, the rules of thumb are questioned. I have no way to construct\n> >some X_n in the usual (causal, realistic) meaning which has the\n> >property of Lorentz-invariance.\n>\n> Lorentz invariance and relativity are important when one needs to\n> compare one observer\'s description (the series X_n) with another\'s.\n\nCompletely wrong. To compare descriptions of different observers\nis complicate enough in everyday life. Relativity is important in\nstrong gravitational fields or for high velocities. Lorentz invariance\nis useful to prove some otherwise strange impossibility results\n(impossibility to measure absolute time) and otherwise useless.\nProof: Lorentz-symmetric Maxwell theory was successful long\nbefore its Lorentz-symmetry has been observed.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> ><rof@maths.tcd.ie> schrieb
> >> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> >> ><rof@maths.tcd.ie> schrieb
> >> >> That is, he must form a data representation which compresses
> >> >> the history of symbols he has received into a single "state
> >> >> of the world", call it X_n, where
> >> >> P(X_{n+1}|X_n, X_{n-1}, ...) = P(X_{n+1}|X_n)
> >> >> (something which was not true of the symbols E_n).
>
> >> The construction of the X_n is perception (of the world).
>
> >I'm not a native speaker, but this sounds strange. I would
> >accept it if you call the M_n "perception".
>
> There is a problem with cognitive dissonance here, since, according
> to popular wisdom, perception is a passive, receptive act wherein
> the observer merely receives perceptions which have been thrown
> at him from somewhere. This view is incorrect, since passive
> receptivity is called sensation, and presents to the observer only
> sounds and colours and so on. Perception is what leads to awareness
> of objects, and requires that the data taken from sensation (the E_n)
> be actively processed in order to generate the representations of
> objects (the X_n).

I understand this difference between sensation E_n and perception.
I have acknowledged it alerady by proposing that the M_n are
perceptions. But not the X_n.

> Hence, perception is an act which involves
> the active construction of the thing which is perceived, and not
> merely the passive reception of an already-formed representation.

Yep. The construction of the M_n is of this type. But the X_n\M_n,
the inner part of the moon, is not perceived. Even if I have some
perception M_{moon} of the surface of the moon and some vague
theoretical ideas what may be inside.

> These matters are relevant for physics because the statistics of
> the E_n either are or are not representable as a Markov process
> which factorises in a spatial way, and so these considerations
> must arise whenever one wants to ask whether an observer will
> see space.

Sorry, but you cannot derive if observers will "see" space.
"Seeing" space means developing a certain realistic theory,
may be implicit (as a hidden prejudice, instinct, created by
evolution).

> >> In an ontological theory the splitting needs to be deduced,
> >> rather than postulated.
>
> >The ontological theory postulates the structure and the laws of
> >the X_n. If these have the form of something localized in a space,
> >space is postulated. What you perceive is another question,
> >in the E_n we have, for example, nothing three-dimensional.
> >We see only two 2D movies. So it is completely unclear what
> >means spatial structure in the E_n.

> No; the ontological theory can't make any postulates about the X_n,
> because the X_n exist only in the mind of the observer, who has
> constructed them.

That's a confusion of language levels. A believes that there are
some X_n in reality. This is an ontological theory proposed by A.
We have to decide which language level we want to use - the level
of A, then we talk about X_n as real states, or on a metalevel where
we talk about A's X_n as constructions of his mind. Of course, on
this metalevel we have our metatheory about reality with our own
X'_n. (This consideration was on meta-meta-level)

> What the observer does is say to himself "Good
> gracious - I appear to be receiving symbols E_n and they appear to
> be somewhat predictable. It's very tedious to predict them based
> on the previous E_n that have been received, because I have to look
> far into the past in order to do that. Instead, I'll make a new
> representation, X_n, which, at time n, incorporates all the information
> from the previous E_n which is relevant for predicting the future
> E_n. I'll call X_n 'my representation of the state of the world at
> present.'"

First, distinguish the X_n from the M_n. Then, that's not what we do.

> If a theory makes postulates about the X_n, then it is making
> postulates about the kind of data processing done by a
> very complicated object (the observer).

Sorry, but the state of a realistic theory X_n cannot be derived
from observations E_n. The miserable part of them which can be
has been denoted by M_n. (BTW, there is no such process
E_n => M_n, the creation of the M_n essentially depends on
the general ontological theory about reality.)

The game is reverse. The theory is prior. Then, a complicated
data processing starts which makes the E_n compatible with the
theory using some M_n. A failure of this process is named
falsification.

Theories about the data processing are part of some historical
sciences.

> >> You do, but giving up either realism or causality won't affect your
> >> ability to predict X_{n+1} from X_n, since they are merely rules of
> >> thumb for generating representations.
>
> >No, the rules of thumb are questioned. I have no way to construct
> >some X_n in the usual (causal, realistic) meaning which has the
> >property of Lorentz-invariance.
>
> Lorentz invariance and relativity are important when one needs to
> compare one observer's description (the series X_n) with another's.

Completely wrong. To compare descriptions of different observers
is complicate enough in everyday life. Relativity is important in
strong gravitational fields or for high velocities. Lorentz invariance
is useful to prove some otherwise strange impossibility results
(impossibility to measure absolute time) and otherwise useless.
Proof: Lorentz-symmetric Maxwell theory was successful long
before its Lorentz-symmetry has been observed.

Ilja

seratend

Oct29-04, 02:08 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<clnh1g\\$eti\\$1@beech.fernuni-hagen.de>...\n> "seratend" <ser_monmail@yahoo.fr> schrieb\n> > After all these posts, as an external reader, I still not understand\n> > the point of view of Ilia nor what could be its FTL phone.\n> > First what do you intend by causality? For me, like with the QM\n> > measurement, it is very difficult to understand what it is or what it\n> > says (human interpretation).\n>\n> I press a button by my free choice, and whenever I press it the\n> light bulb changes its state. In this case, I conclude that there is\n> a causal influence from my behaviour to the state of the light bulb.\n>\n> I make this conclusion based on the observation of some correlations.\n> Of course, I also use some other assumptions: That there is some\n> reality which defines the state of the light bulb, and certain assumptions\n> about independence of my free choices from external influences.\n>\n> I believe that these basic assumptions are the foundation of the\n> scientific method, that it is reasonable to hold them even in\n> circumstances where no simple realistic explanation is known.\n\nWell, as a simple-minded person (in a non-pejorative sense :) in\nphysics, I am always afraid with arguments/words that are too far from\nmathematics and too philosophical and too general such as the reality\none. Just because, when I want to make some "physical-mathematical"\nlogical deduction that use such a word/argument, I have to interpret\nit (ie, I have to attach to this word at least a coherent mathematical\nrepresentation) and thus I risk to choose the definition that arrange\nmy logical deduction.\n\nThe EPR 1935 paper is the example of a definition of this "reality":\nThey start with this sentence: "The elements of the physical reality\ncannot be determined by a priori philosophical considerations, but\nmust be found by an appeal to results of experiments and\nmeasurements." At least they agree that it seems to be a need for a\nkind of formal logical deduction.\nIn addition, they follow by their own definition of reality "if,\nwithout in any way disturbing a system, we can predict with certainty\n(i.e. with probability equal to unity) the value of a physical\nquantity, then there exists an element of physical reality\ncorresponding to this physical quantity."\n\nTherefore, we may see that even this earlier definition of physical\nreality is very fuzzy and in fact is selected, a priori, for their\npaper argumentation. Remark that this definition does not exclude a\nglobally defined physical quantity (i.e. a reality is not assumed\nlocal by this definition).\n\nIf I go back to this FTL phone, I would like to underline the dangers\nof using black boxes where there concepts are too fuzzy (not directly\nconnected to a "physico-mathematical formulation", a very fuzzy term\n;).\n\n> I see, I have not sufficiently explained that it is based on Popper\'s\n> falsifiability of scientific, physical theory. To distinghish science from\n> pseudosciences Popper has proposed that the distinguishing\n> property of physical theories is that they make nontrivial predictions,\n> predictions which may appear false if we test them. Such theories\n> are called falsifiable. Theories which cannot be falsified are not\n> empirical theories, they are metaphysical. This point of view on science\n> is IMHO correct and widely accepted among scientists.\n\nI understand it and I agree partly with this statement. The "partly"\nis due to the implicit hypothesis that the theories may be falsified.\nThe main problem with modern theories (modern in the sense the few\nones I know) is that they are constructed to support everything. It is\nalways the external assumptions, sometimes implicit, that restricts\nthem and thus may be falsified:\ne.g Newtons\' law: ma=force: if I authorise any force, I may have any\nacceleration, and any x(t) if I enlarge the derivation. So, I can say,\nformally in the sense I may recover "ma"= "force", that newton\'s law\nis always true even at the quantum level: it is the bohemian\ninterpretation.\n\nSometimes, I prefer to say a theory is "wrong" because its formulation\nis more complicated than another formulation (I am lazy person with a\nlimited brain power : ), but it is a dangerous thought...\n\n> My FTL phone is only a hypothetical feature. [...]\n\nThe only difficult point with the FTL phone argumentation resides in\nits "fuzzy nature" that does not allow one to make reasonable logical\ndeductions based on a logical model. The main point I currently\nidentify with this fuzzy model is its possible lack of\n"reality/locality" (my reality/locality : ), because you just removed\ntoo many properties of your black box.\nE.g. your sentence "I press a button by my free choice", so if I\nassume that the button is a local one (an extra hypothesis to your\nmodel).\nI may also suppose that your FTL model is based on local interactions\n(the "press" interaction). With this assumption and mathematical\n"causality" (the logical "=>" symbol), I think I can deduce that this\nmodel works if there exists an FTL interaction on at least a frame.\nIf it is your assumption, I prefer to start with the possibility of\nthe existence of an interaction with an infinite speed in at least one\nframe: this model is much well defined (at a logical point of view),\nthan a FTL phone. I may also start with a non-local interaction model\n(the button is a universe button that I can press locally).\n\n>\n> I think about causality on a more fundamental level. The observation\n> of a correlation between A and B is sufficient to search for a realistic\n> causal explanation. Such explanations are a common cause\n> C->A, C->B or a direct causal influence A->B or B->A. Time has\n> not been mentioned in this consideration.\n>\n\nThis is the problem, it seems to be too fundamental. You are removing\ntoo many things. At least to realise an experiment (i.e. If you want a\ntheory that may be falsified), you need time, at least in one step of\nthe experiment, to make a "causal" interpretation.\n\n>\n> Now, Einstein causality is, in the general understanding, a physical\n> theory. It predicts that no FTL phones are possible. At least, this is\n> the usual interpretation of Einstein causality.\n\nI will say that Einstein causality is an extra postulate. And, I am\nafraid of the multiple interpretations we give to these words\n"Einstein causality".\nFor example with the SR, we may deduce a causality interpretation as\nthat "reality" belongs to the light cone of the space-time (tb>ta\nwhatever SR frame we select) and from the updated ma=force equation.\nI do not even speak about GR, where you need to add the locality to\ndefine a local causality.\n\nIn QFT, the causality is interpreted in a slightly different sense:\nwhether a measurement performed at one point can affect a measurement\nat another point whose separation from the first is spacelike\n(Peskin-Schroeder QFT reference).\nThere are other postulates such as Weinberg cluster decomposition that\ndefines its own causality (weaker), and in fine, it gives almost the\nsame "results".\n\n>\n> What would happen if we would really observe some FTL phone?\n> Very simple, Einstein causality would be falsified. Nobody would\n> defend relativity in this case. It would survive in physics as a limit\n> of some other theory which allows FTL phones.\n>\n> This is not the case today. Using the strange correlations of EPR\n> and Bell we cannot build FTL phones. This is a simple theorem.\n\nWell, I should first know what a FTL phone is. For example if the\nphenomenon is an interaction FTL I would rather first keep relativity\nbut say that maybe "reality" also exists in space like intervals (the\nfamous "tachyons" or whatever else). After, I will try to make a\nworking model (with experiments). After all, I will see what type of\ncausality is broken (if it is really broken).\n\n>\n> Nonetheless, that does not mean there is no problem for relativity.\n> The only realistic explanations of these correlations are A->B or\n> B->A, above are in contradiction with relativity. Now, defenders\n> of relativity use various argumentations. But, as I have observed,\n> could be applied as well in the hypothetical case of falsification of\n> relativity. That means, if we take them seriously, we can no longer\n> predict that no FTL phone exists - because such an FTL phone\n> could be explained away if we observe it.\n>\nThis is the main problem with general arguments. You can adapt them as\na theory to what you want.\nI agree, and I think every physicist, that SR or GR have a limited\ndomain of application. The main problem is that the mathematical\nformulation of these theories as well a QM does not really need\ncausality. It is always an external restriction in order to say this\nis the reality (i.e. to recover the experiments).\n\n> > So, Do you intend "->" (your causality of the FTL phone) as the\n> > mathematical logic symbol "=>" ?\n>\n> No. In mathematical terms it means absence of independence.\n\nWhow, what a very weak and fuzzy definition.\n\nSo I will try to precise the implicit context of this definition.\nI assume that we use the zermello-frankel axiomatic set theory at\nleast with the choice axiom: we assume that A and B may be points\nbelonging to a set (the "points" may be another sets if we want). We\ncall this set the universe.\nI assume you are not using a weaker theory like the category theory\nfor the initial definition of the mathematical objects (i.e what is A,\nB, o need for a universe that is a set).\n\nI thus interpret the "absence of independence" of "->" by the\nexistence of an application f (we can instead choose a relation, but\nit gives no more information and may be more difficult to interpret)\nsuch as we have:\n\nB -> A <=> A= f(B, other points if we want= context).\n\nWe can also complicate the formulation (closer to a relation):\nB->A <=> (A,context2)= f(B, context1) (\n\nThe important point is the surjection of the application on its domain\nof definition (if A1=f(input) and A2=f(input), A1=A2)\n\nSo B -> A <=> (context1, B) => (A, context2); without loss of\ngenerality.\n(context1 may be for example the domain of validity of the logical\n"=>" relation)\n\nWe may also rename them, i.e. we redefine a new set (context1, B)=B,\n(A, context2)=B in order to recover a simple logical A=>B.\n\nThus in my previous post we just need to add the "context" property to\nthe A and B points.\n\n> Realism assumes\n>\n> Result_at_A = f_a(reality, decisions_at_A,decisions_at_B)\n\nfor B -> A?\n\n> Result_at_B = f_b(reality, decisions_at_A,decisions_at_B)\n\nfor A -> B?\n\n>\n> Causality gives the additional restriction\n>\n> Result_at_A = f_a(reality, decisions_at_A)\n> Result_at_B = f_b(reality, decisions_at_B)\n>\nYou are introducing three new notions: decisions, reality and results\nthat are difficult to map into a physical theory. But, your are giving\nsome details of some of your assumptions (in the current hypothesis:\nthe FTL phone is described by an unknown physics).\n\nFirst what is the Result_at_A? Because you have not defined the\nspace-time object, I assume result_at_A is not obliged to be at a\npoint in space: it is only attached to A, to be as general as\npossible.\nI have defined a set of points containing A and B, … (the universe)\nAnd your text above says that A and B are also sets [A=set of (all\npossible results), B= set of (all possible results)]. (here I add a\nnew postulate: all possible results are a set).\n\nSo your are building applications to these sets (f_a for the set A,\nf_b for the set B): ok.\nNow, what represents the inputs of these applications: decisions_at_A,\ndecision_at_B ? Does it exist any relation between decisions and\nresults in a large sense? I need it in order to understand what your\ndeductions may involve.\n\nI do not understand your "reality":\n-- We may take some reality aspects as the domain of validity of the\nfunctions f_a and f_b . (some values of\n(decisions_at_A,decisions_at_B) does not belong to the reality).\n-- reality may also be the other decisions_at_other_objects or\nresults_at_other_objects (the additional context).\n-- is it an object of the universe (the set containing the points A,B\n....)\n-- ???\n\n> Causality gives the additional restriction\n>\n> Result_at_A = f_a(reality, decisions_at_A)\n> Result_at_B = f_b(reality, decisions_at_B)\n\n\nNow with this definition and the previous assumptions, I think here\nyou are mixing causality with locality (I still have questions about\nwhat is "reality" and "decisions_at_object"). Even if we don\'t have\nspecified what we intend by locality, I don\'t think that we can say\nthat "causality" => "locality" without additional assumptions (e.g. QM\ncan be said causal but not local to some extent).\n\n\n> > Now, to end, I will try to show a possible implementation example of a\n> > formal FTL phone based solely on your "causal" argument.\n> >\n> > Let\'s take 2 classical independent particles (no interaction) with a\n> > total kinetic energy E= E1+E2.\n> > Now if we change the kinetic energy of particle 1, we thus induce\n> > instantaneously (causality) a change on the kinetic energy of the 2\n> > particles (E): We have formally a FTL phone with the total Energy E.\n>\n>\n> That\'s not a phone because we cannot use it to talk with each other,\n> nor formal nor informal.\n\nWhy not?\nFormally, we have first to define what is to talk to each other (I do\nnot involve necessarily human objects, nor local objects: it is my\nfreedom choice with your weak model).\nI think you are implicitly restricting the FTL phone model to local\nvariables and thus you are mixing causality with locality.\nIt is what I want to underline with this FTL phone fuzzy model: we do\nnot specify enough to make reasonable logical deductions (i.e. all the\ndeductions have many implicit assumptions).\n\n> > So when we say that a local measurement on one EPR particle implies\n> > the change of the global state |psi> we are very close to the previous\n> > example. To detect the change of |psi> at the other particle location,\n> > we should have a black box (like the previous one for the total energy\n> > E) that is able to measure the global state (and not the local one).\n>\n> The whole point is a different one. We really have some black box.\n> This black box allows to do very strange things. Unfortunately it\n> does not allow us to build an FTL phone. But there exists no realistic\n> explanation of what is inside the black box which does not use some\n> sort of hidden FTL information transfer.\n>\n> Ilja\n\nI agree we may not be able to build a "local" FTL phone with QM\nmechanics, even if I am afraid I am taking implicit hypotheses to say\nthat ; ). But I a FTL phone is possible I quite sure that an adequate\nSchroedinger equation may describe its behaviour.\nI also say that a QM state is not interpreted into an obvious\nkolgomorovian probabilistic event (whatever probabilistic set and\nsigma-algebra is chosen). The things even become rather difficult when\nwe take a global state with global observables and want to deduce if\nthere is a hidden FTL local channel.\nThe EPR state is a global state where the 2 particles have the same\nspin axis (rotationally invariant state): this is the global property\ncommon to the 2 EPR particles. We just can say that a classical\nstatistical hidden variable model, without an hidden FTL channel is\nnot able to represent the statistical results given by this state.\n\nFinally, for the ones interested by a hidden variable model of EPR\nstate with a hidden FTL channel, I recommend the last Cerf, Gisisn\nMassar and Popescu quant-ph/0410027 paper (4 pages – a short concise\none). I think it is a good one (with the pointers it gives).\nIt introduces a possible (logical) implementation of what they call\nthe non-local machines (PR- machines, I like the "machine" term they\nchosen rather than "interaction"). We have another different look to\nthe EPR like states.\n\nI would like to know if you would you say that this model is causal or\nnot and why?\n\nSeratend\n\nHi everybody,\n\nFor the ones interested by a hidden variable model of EPR state with a\nhidden communication channel , I recommend the last Cerf, Gisisn\nMassar and Popescu quant-ph/0410027 paper (4 pages – a short concise\none). I think it is a good one (with the pointers it gives).\nIt introduces a possible (logical) implementation of the EPR like\nstate with what they call the non-local machines (PR- machines, I like\nthe "machine" term they chosen rather than "interaction"): We have\nanother different look to the EPR like states.\n\nAll the comments are welcome.\n\nSeratend.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<clnh1g$eti$1@beech.fernuni-hagen.de>...
> "seratend" <ser_monmail@yahoo.fr> schrieb
> > After all these posts, as an external reader, I still not understand
> > the point of view of Ilia nor what could be its FTL phone.
> > First what do you intend by causality? For me, like with the QM
> > measurement, it is very difficult to understand what it is or what it
> > says (human interpretation).
>
> I press a button by my free choice, and whenever I press it the
> light bulb changes its state. In this case, I conclude that there is
> a causal influence from my behaviour to the state of the light bulb.
>
> I make this conclusion based on the observation of some correlations.
> Of course, I also use some other assumptions: That there is some
> reality which defines the state of the light bulb, and certain assumptions
> about independence of my free choices from external influences.
>
> I believe that these basic assumptions are the foundation of the
> scientific method, that it is reasonable to hold them even in
> circumstances where no simple realistic explanation is known.

Well, as a simple-minded person (in a non-pejorative sense :) in
physics, I am always afraid with arguments/words that are too far from
mathematics and too philosophical and too general such as the reality
one. Just because, when I want to make some "physical-mathematical"
logical deduction that use such a word/argument, I have to interpret
it (ie, I have to attach to this word at least a coherent mathematical
representation) and thus I risk to choose the definition that arrange
my logical deduction.

The EPR 1935 paper is the example of a definition of this "reality":
They start with this sentence: "The elements of the physical reality
cannot be determined by a priori philosophical considerations, but
must be found by an appeal to results of experiments and
measurements." At least they agree that it seems to be a need for a
kind of formal logical deduction.
In addition, they follow by their own definition of reality "if,
without in any way disturbing a system, we can predict with certainty
(i.e. with probability equal to unity) the value of a physical
quantity, then there exists an element of physical reality
corresponding to this physical quantity."

Therefore, we may see that even this earlier definition of physical
reality is very fuzzy and in fact is selected, a priori, for their
paper argumentation. Remark that this definition does not exclude a
globally defined physical quantity (i.e. a reality is not assumed
local by this definition).

If I go back to this FTL phone, I would like to underline the dangers
of using black boxes where there concepts are too fuzzy (not directly
connected to a "physico-mathematical formulation", a very fuzzy term
;).

> I see, I have not sufficiently explained that it is based on Popper's
> falsifiability of scientific, physical theory. To distinghish science from
> pseudosciences Popper has proposed that the distinguishing
> property of physical theories is that they make nontrivial predictions,
> predictions which may appear false if we test them. Such theories
> are called falsifiable. Theories which cannot be falsified are not
> empirical theories, they are metaphysical. This point of view on science
> is IMHO correct and widely accepted among scientists.

I understand it and I agree partly with this statement. The "partly"
is due to the implicit hypothesis that the theories may be falsified.
The main problem with modern theories (modern in the sense the few
ones I know) is that they are constructed to support everything. It is
always the external assumptions, sometimes implicit, that restricts
them and thus may be falsified:
e.g Newtons' law: ma=force: if I authorise any force, I may have any
acceleration, and any x(t) if I enlarge the derivation. So, I can say,
formally in the sense I may recover "ma"= "force", that newton's law
is always true even at the quantum level: it is the bohemian
interpretation.

Sometimes, I prefer to say a theory is "wrong" because its formulation
is more complicated than another formulation (I am lazy person with a
limited brain power : ), but it is a dangerous thought...

> My FTL phone is only a hypothetical feature. [...]

The only difficult point with the FTL phone argumentation resides in
its "fuzzy nature" that does not allow one to make reasonable logical
deductions based on a logical model. The main point I currently
identify with this fuzzy model is its possible lack of
"reality/locality" (my reality/locality : ), because you just removed
too many properties of your black box.
E.g. your sentence "I press a button by my free choice", so if I
assume that the button is a local one (an extra hypothesis to your
model).
I may also suppose that your FTL model is based on local interactions
(the "press" interaction). With this assumption and mathematical
"causality" (the logical "=>" symbol), I think I can deduce that this
model works if there exists an FTL interaction on at least a frame.
If it is your assumption, I prefer to start with the possibility of
the existence of an interaction with an infinite speed in at least one
frame: this model is much well defined (at a logical point of view),
than a FTL phone. I may also start with a non-local interaction model
(the button is a universe button that I can press locally).

>
> I think about causality on a more fundamental level. The observation
> of a correlation between A and B is sufficient to search for a realistic
> causal explanation. Such explanations are a common cause
> C->A, C->B or a direct causal influence A->B or B->A. Time has
> not been mentioned in this consideration.
>

This is the problem, it seems to be too fundamental. You are removing
too many things. At least to realise an experiment (i.e. If you want a
theory that may be falsified), you need time, at least in one step of
the experiment, to make a "causal" interpretation.

>
> Now, Einstein causality is, in the general understanding, a physical
> theory. It predicts that no FTL phones are possible. At least, this is
> the usual interpretation of Einstein causality.

I will say that Einstein causality is an extra postulate. And, I am
afraid of the multiple interpretations we give to these words
"Einstein causality".
For example with the SR, we may deduce a causality interpretation as
that "reality" belongs to the light cone of the space-time (tb>ta
whatever SR frame we select) and from the updated ma=force equation.
I do not even speak about GR, where you need to add the locality to
define a local causality.

In QFT, the causality is interpreted in a slightly different sense:
whether a measurement performed at one point can affect a measurement
at another point whose separation from the first is spacelike
(Peskin-Schroeder QFT reference).
There are other postulates such as Weinberg cluster decomposition that
defines its own causality (weaker), and in fine, it gives almost the
same "results".

>
> What would happen if we would really observe some FTL phone?
> Very simple, Einstein causality would be falsified. Nobody would
> defend relativity in this case. It would survive in physics as a limit
> of some other theory which allows FTL phones.
>
> This is not the case today. Using the strange correlations of EPR
> and Bell we cannot build FTL phones. This is a simple theorem.

Well, I should first know what a FTL phone is. For example if the
phenomenon is an interaction FTL I would rather first keep relativity
but say that maybe "reality" also exists in space like intervals (the
famous "tachyons" or whatever else). After, I will try to make a
working model (with experiments). After all, I will see what type of
causality is broken (if it is really broken).

>
> Nonetheless, that does not mean there is no problem for relativity.
> The only realistic explanations of these correlations are A->B or
> B->A, above are in contradiction with relativity. Now, defenders
> of relativity use various argumentations. But, as I have observed,
> could be applied as well in the hypothetical case of falsification of
> relativity. That means, if we take them seriously, we can no longer
> predict that no FTL phone exists - because such an FTL phone
> could be explained away if we observe it.
>
This is the main problem with general arguments. You can adapt them as
a theory to what you want.
I agree, and I think every physicist, that SR or GR have a limited
domain of application. The main problem is that the mathematical
formulation of these theories as well a QM does not really need
causality. It is always an external restriction in order to say this
is the reality (i.e. to recover the experiments).

> > So, Do you intend "->" (your causality of the FTL phone) as the
> > mathematical logic symbol "=>" ?
>
> No. In mathematical terms it means absence of independence.

Whow, what a very weak and fuzzy definition.

So I will try to precise the implicit context of this definition.
I assume that we use the zermello-frankel axiomatic set theory at
least with the choice axiom: we assume that A and B may be points
belonging to a set (the "points" may be another sets if we want). We
call this set the universe.
I assume you are not using a weaker theory like the category theory
for the initial definition of the mathematical objects (i.e what is A,
B, o need for a universe that is a set).

I thus interpret the "absence of independence" of "->" by the
existence of an application f (we can instead choose a relation, but
it gives no more information and may be more difficult to interpret)
such as we have:

B -> A <=> A= f(B,[/itex] other points if we want= context).

We can also complicate the formulation (closer to a relation):
B->A <=> (A,context2)= f(B, context1) (

The important point is the surjection of the application on its domain
of definition (if A1=f(input) and A2=f(input), A1=A2)

So B -> A <=> (context1, B) => (A, context2); without loss of
generality.
(context1 may be for example the domain of validity of the logical
"=>" relation)

We may also rename them, i.e. we redefine a new set (context1, B)=B,
(A, context2)=B in order to recover a simple logical A=>B.

Thus in my previous post we just need to add the "context" property to
the A and B points.

> Realism assumes
>
> Result_at_A = f_a(reality, decisions_at_A,decisions_at_B)

for B -> A?

> Result_at_B = f_b(reality, decisions_at_A,decisions_at_B)

for A -> B?

>
> Causality gives the additional restriction
>
> Result_at_A = f_a(reality, decisions_at_A)
> Result_at_B = f_b(reality, decisions_at_B)
>
You are introducing three new notions: decisions, reality and results
that are difficult to map into a physical theory. But, your are giving
some details of some of your assumptions (in the current hypothesis:
the FTL phone is described by an unknown physics).

First what is the Result_at_A? Because you have not defined the
space-time object, I assume result_at_A is not obliged to be at a
point in space: it is only attached to A, to be as general as
possible.
I have defined a set of points containing A and B, … (the universe)
And your text above says that A and B are also sets [A=set of (all
possible results), B= set of (all possible results)]. (here I add a
new postulate: all possible results are a set).

So your are building applications to these sets (f_a for the set A,
f_b for the set B): ok.
Now, what represents the inputs of these applications: decisions_at_A,decision_at_B ? Does it exist any relation between decisions and
results in a large sense? I need it in order to understand what your
deductions may involve.

I do not understand your "reality":
-- We may take some reality aspects as the domain of validity of the
functions f_a and f_b . (some values of
(decisions_at_A,decisions_at_B) does not belong to the reality).
-- reality may also be the other decisions_at_other_objects or
results_at_other_objects (the additional context).
-- is it an object of the universe (the set containing the points A,B
....)
-- ???

> Causality gives the additional restriction
>
> Result_at_A = f_a(reality, decisions_at_A)
> Result_at_B = f_b(reality, decisions_at_B)

Now with this definition and the previous assumptions, I think here
you are mixing causality with locality (I still have questions about
what is "reality" and "decisions_at_object"). Even if we don't have
specified what we intend by locality, I don't think that we can say
that "causality" => "locality" without additional assumptions (e.g. QM
can be said causal but not local to some extent).

> > Now, to end, I will try to show a possible implementation example of a
> > formal FTL phone based solely on your "causal" argument.
> >
> > Let's take 2 classical independent particles (no interaction) with a
> > total kinetic energy [itex]E= E1+E2.
> > Now if we change the kinetic energy of particle 1, we thus induce
> > instantaneously (causality) a change on the kinetic energy of the 2
> > particles (E): We have formally a FTL phone with the total Energy E.
>
>
> That's not a phone because we cannot use it to talk with each other,
> nor formal nor informal.

Why not?
Formally, we have first to define what is to talk to each other (I do
not involve necessarily human objects, nor local objects: it is my
freedom choice with your weak model).
I think you are implicitly restricting the FTL phone model to local
variables and thus you are mixing causality with locality.
It is what I want to underline with this FTL phone fuzzy model: we do
not specify enough to make reasonable logical deductions (i.e. all the
deductions have many implicit assumptions).

> > So when we say that a local measurement on one EPR particle implies
> > the change of the global state |\psi> we are very close to the previous
> > example. To detect the change of |\psi> at the other particle location,
> > we should have a black box (like the previous one for the total energy
> > E) that is able to measure the global state (and not the local one).
>
> The whole point is a different one. We really have some black box.
> This black box allows to do very strange things. Unfortunately it
> does not allow us to build an FTL phone. But there exists no realistic
> explanation of what is inside the black box which does not use some
> sort of hidden FTL information transfer.
>
> Ilja

I agree we may not be able to build a "local" FTL phone with QM
mechanics, even if I am afraid I am taking implicit hypotheses to say
that ; ). But I a FTL phone is possible I quite sure that an adequate
Schroedinger equation may describe its behaviour.
I also say that a QM state is not interpreted into an obvious
kolgomorovian probabilistic event (whatever probabilistic set and
\sigma-algebra is chosen). The things even become rather difficult when
we take a global state with global observables and want to deduce if
there is a hidden FTL local channel.
The EPR state is a global state where the 2 particles have the same
spin axis (rotationally invariant state): this is the global property
common to the 2 EPR particles. We just can say that a classical
statistical hidden variable model, without an hidden FTL channel is
not able to represent the statistical results given by this state.

Finally, for the ones interested by a hidden variable model of EPR
state with a hidden FTL channel, I recommend the last Cerf, Gisisn
Massar and Popescu http://www.arxiv.org/abs/quant-ph/0410027 paper (4 pages – a short concise
one). I think it is a good one (with the pointers it gives).
It introduces a possible (logical) implementation of what they call
the non-local machines (PR- machines, I like the "machine" term they
chosen rather than "interaction"). We have another different look to
the EPR like states.

I would like to know if you would you say that this model is causal or
not and why?

Seratend

Hi everybody,

For the ones interested by a hidden variable model of EPR state with a
hidden communication channel , I recommend the last Cerf, Gisisn
Massar and Popescu http://www.arxiv.org/abs/quant-ph/0410027 paper (4 pages – a short concise
one). I think it is a good one (with the pointers it gives).
It introduces a possible (logical) implementation of the EPR like
state with what they call the non-local machines (PR- machines, I like
the "machine" term they chosen rather than "interaction"): We have
another different look to the EPR like states.

All the comments are welcome.

Seratend.

Thomas Trotter

Oct30-04, 08:10 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\nvanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410252242.7a58e2b4@posting.google. com>...\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cl7p80\\$dck\\$1@beech.fernuni-hagen.de>...\n> > "Patrick Van Esch" <vanesch@ill.fr> schrieb\n>\n> > > Not really. A working FTL phone would allow me, as a single observer,\n> > > to conclude that the signal I sent went up and down Titan in 0.5\n> > > seconds.\n> >\n> > How?\n> >\n>\n> By triggering the nuclear bombs on titan with the special code I have.\n> If I see a big fireball in the sky after 1 light-speed propagation\n> time after I sent the message, I know my code arrived almost\n> immediately.\n> You cannot do that with an EPR like set-up, and what I wanted to\n> illustrate was (see my other post in this thread) that if you consider\n> the measurements at Bob and Alice as just decohered superpositions,\n> then there\'s a very natural and local explanation of the so-called\n> non-locality: namely a state that went though two different paths, and\n> its amplitudes for each path interfere when they are added together.\n> As such it is also obvious that you can never get this FTL phone in\n> this way. You shouldn\'t take my stuff on solipsism too seriously. It\n> was just meant to illustrate that (the way I understand current QM)\n> the wavefunction has only a reality with respect to one single\n> observer (which turns out to be indistinguishable to all timelike\n> connected events - there might be something in that, but I don\'t know\n> what). It is only when we take the same wavefunction to describe an\n> independent reality on space-like connected events that we seem to run\n> into non-locality problems when taking the concept of measurement\n> (probably too) so far. I\'m not religiously attached to anything\n> (locality, linearity of QM, whatever). I\'m not trying to modify\n> anything to the currently accepted principles ; after all they work\n> well in practice. I\'m just trying to make sense of that situation\n> within the framework as presented, and I think I can, up to a point,\n> when you consider that all an observer will ever do is make\n> observations which are timelike connected.\n> But this doesn\'t mean - as you seem to suggest - that he cannot make\n> tests about a genuine non-locality. These tests are then of course\n> limited to matter or information-carrying faster-than-light transport,\n> but they explain in a natural way why this\n> "state-projection-at-a-distance" cannot be used for that, simply\n> because there is no such thing: the only projections that occur are\n> local to the observer, and as such they explain completely the EPR\n> results as LOCAL interference between two different paths taken by a\n> quantum system and then brought together.\n\nThe way you are approaching the EPR-Bell stuff makes sense\nto me. But, because of Bell\'s formulation, there is a\nquestion that I don\'t know how to answer. Hopefully I can\nphrase it in a way that effectively communicates the problem:\n\nYou\'ve got the usual A <--light--Emitter--light--> B\nsetup, and you\'re using laser pulsed atomic calcium\ncascades to generate entangled emissions.\n\nLet\'s say A registers a detection. Now the basis for calculating\nthe probability of registering a detection at B (for the\ncoincidence interval corresponding to the detection at A [1])\nis taken as the transmission axis of the polarizer at A for\nthe interval during which the detection at A occured in\nconjunction with the assumption that light emitted during\nthis interval came from the same atomic transition (and,therefore,\nthe oppositely-directed light incident on the polarizers during\nthis interval is identically polarized due to conservation of\nangular momentum). So, for this coincidence interval, the\nprobability of detection at B is cos^2 theta, where theta is\nthe angular difference between the transmission axes of the\npolarizers at A and B. (And, this cos^2 theta function\ngeneralizes to all coincidence intervals -- so that the\nprobability of coincidental detection, in the ideal, is\ncos^2 theta.)\n\nNow, following Bell, formulate the probability of detection\nat A and B *from the outset* (prior to any detection) using\nthe same assumptions as above (identical emission-produced\npolarization of light incident on A and B for any given\ncoincidence interval). Denoting the common polarization as\nlambda, and the transmission axes of the polarizers at A and\nB as a and b, respectively:\n\nP(A) = cos^2 |a - lambda| and P(B) = cos^2 |b - lambda|\n\nIf lambda is not equal to a, then how can P(B) = cos^2 |b - lambda|\nprior to detection at A, and P(B) = cos^2 |b - a| after detection\nat A? [2]\n\nAnyway, there are two incompatible formulations involved here.\nThey both assume that the light incident on the polarizers at\nA and B is correlated in polarization due to emission from\na common atomic transition. But, the first formulation, the\none that is (apparently) supported by experiment, doesn\'t\napply this assumption until a detection has been registered.\nIf, as in the second (Bell\'s) formulation, the assumption is\napplied prior to detection, then contradictions ensue.\n\nWhat\'s wrong with Bell\'s formulation? It doesn\'t seem to be,\nstrictly speaking, the assumption of locality -- since qm also\nuses the assumption of a common emitter as the basis for\ncalculations.\n\nThe two formulations do agree when theta = 0, pi/4, and pi/2.\nPerhaps this is revealing. The qm formulation seems to have\ncaptured something that Bell\'s formulation leaves out.\nBut, qm gives no details as to what that something is.\nOr, does it?\n\nI\'m sort of stuck here. *Any* help will be appreciated.\n\n\nFootnotes:\n\n[1] I don\'t know how the timing stuff actually works in the\ncoincidence circuitry. Like, how the coincidence interval is\ndefined. How is everything ... matched up, etc.? I do know that\nit has to do with the atomic transition time, and the distance\nthat the polarizers are from the emitters, but am not sure\nexactly what sequence of operations a detection triggers.\n\n[2] My answer would be that if there\'s sufficient amplitude in the\ntransmission by the polarizer, a, to generate a detection\nat A for some interval, then, for that interval, and for theta = 0,\nthen there will, with probability 1, be sufficient amplitude\nin the transmission by the polarizer, b, to generate a detection\nat B. And, following Malus\' Law, the probability of detection at\nB will decrease as theta increases from 0 to pi/2.\n\nBut, I don\'t know why it should work that way.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410252242.7a58e2b4@posting.google.com>...
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cl7p80$dck$1@beech.fernuni-hagen.de>...
> > "Patrick Van Esch" <vanesch@ill.fr> schrieb
>
> > > Not really. A working FTL phone would allow me, as a single observer,
> > > to conclude that the signal I sent went up and down Titan in .5
> > > seconds.
> >
> > How?
> >
>
> By triggering the nuclear bombs on titan with the special code I have.
> If I see a big fireball in the sky after 1 light-speed propagation
> time after I sent the message, I know my code arrived almost
> immediately.
> You cannot do that with an EPR like set-up, and what I wanted to
> illustrate was (see my other post in this thread) that if you consider
> the measurements at Bob and Alice as just decohered superpositions,
> then there's a very natural and local explanation of the so-called
> non-locality: namely a state that went though two different paths, and
> its amplitudes for each path interfere when they are added together.
> As such it is also obvious that you can never get this FTL phone in
> this way. You shouldn't take my stuff on solipsism too seriously. It
> was just meant to illustrate that (the way I understand current QM)
> the wavefunction has only a reality with respect to one single
> observer (which turns out to be indistinguishable to all timelike
> connected events - there might be something in that, but I don't know
> what). It is only when we take the same wavefunction to describe an
> independent reality on space-like connected events that we seem to run
> into non-locality problems when taking the concept of measurement
> (probably too) so far. I'm not religiously attached to anything
> (locality, linearity of QM, whatever). I'm not trying to modify
> anything to the currently accepted principles ; after all they work
> well in practice. I'm just trying to make sense of that situation
> within the framework as presented, and I think I can, up to a point,
> when you consider that all an observer will ever do is make
> observations which are timelike connected.
> But this doesn't mean - as you seem to suggest - that he cannot make
> tests about a genuine non-locality. These tests are then of course
> limited to matter or information-carrying faster-than-light transport,
> but they explain in a natural way why this
> "state-projection-at-a-distance" cannot be used for that, simply
> because there is no such thing: the only projections that occur are
> local to the observer, and as such they explain completely the EPR
> results as LOCAL interference between two different paths taken by a
> quantum system and then brought together.

The way you are approaching the EPR-Bell stuff makes sense
to me. But, because of Bell's formulation, there is a
question that I don't know how to answer. Hopefully I can
phrase it in a way that effectively communicates the problem:

You've got the usual A <--light--Emitter--light--> B
setup, and you're using laser pulsed atomic calcium
cascades to generate entangled emissions.

Let's say A registers a detection. Now the basis for calculating
the probability of registering a detection at B (for the
coincidence interval corresponding to the detection at A [1])
is taken as the transmission axis of the polarizer at A for
the interval during which the detection at A occured in
conjunction with the assumption that light emitted during
this interval came from the same atomic transition (and,therefore,
the oppositely-directed light incident on the polarizers during
this interval is identically polarized due to conservation of
angular momentum). So, for this coincidence interval, the
probability of detection at B is cos^2 \theta, where \theta is
the angular difference between the transmission axes of the
polarizers at A and B. (And, this cos^2 \theta function
generalizes to all coincidence intervals -- so that the
probability of coincidental detection, in the ideal, is
cos^2 \theta.)

Now, following Bell, formulate the probability of detection
at A and B *from the outset* (prior to any detection) using
the same assumptions as above (identical emission-produced
polarization of light incident on A and B for any given
coincidence interval). Denoting the common polarization as
\lambda, and the transmission axes of the polarizers at A and
B as a and b, respectively:

P(A) = cos^2 |a - \lambda| and P(B) = cos^2 |b - \lambda|

If \lambda is not equal to a, then how can P(B) = cos^2 |b - \lambda|
prior to detection at A, and P(B) = cos^2 |b - a| after detection
at A? [2]

Anyway, there are two incompatible formulations involved here.
They both assume that the light incident on the polarizers at
A and B is correlated in polarization due to emission from
a common atomic transition. But, the first formulation, the
one that is (apparently) supported by experiment, doesn't
apply this assumption until a detection has been registered.
If, as in the second (Bell's) formulation, the assumption is
applied prior to detection, then contradictions ensue.

What's wrong with Bell's formulation? It doesn't seem to be,
strictly speaking, the assumption of locality -- since qm also
uses the assumption of a common emitter as the basis for
calculations.

The two formulations do agree when \theta = 0, \pi/4, and \pi/2.
Perhaps this is revealing. The qm formulation seems to have
captured something that Bell's formulation leaves out.
But, qm gives no details as to what that something is.
Or, does it?

I'm sort of stuck here. *Any* help will be appreciated.

Footnotes:

[1] I don't know how the timing stuff actually works in the
coincidence circuitry. Like, how the coincidence interval is
defined. How is everything ... matched up, etc.? I do know that
it has to do with the atomic transition time, and the distance
that the polarizers are from the emitters, but am not sure
exactly what sequence of operations a detection triggers.

[2] My answer would be that if there's sufficient amplitude in the
transmission by the polarizer, a, to generate a detection
at A for some interval, then, for that interval, and for \theta = 0,
then there will, with probability 1, be sufficient amplitude
in the transmission by the polarizer, b, to generate a detection
at B. And, following Malus' Law, the probability of detection at
B will decrease as \theta increases from to \pi/2.

But, I don't know why it should work that way.

Ilja Schmelzer

Nov3-04, 09:47 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Patrick Van Esch" <vanesch@ill.fr> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> > > Not really. A working FTL phone would allow me, as a single observer,\n> > > to conclude that the signal I sent went up and down Titan in 0.5\n> > > seconds.\n> >\n> > How?\n> >\n> By triggering the nuclear bombs on titan with the special code I have.\n> If I see a big fireball in the sky after 1 light-speed propagation\n> time after I sent the message, I know my code arrived almost\n> immediately.\n\nOk, that means verification of the information with some other,\nnon-FTL information channel.\n\n> You cannot do that with an EPR like set-up, and what I wanted to\n> illustrate was (see my other post in this thread) that if you consider\n> the measurements at Bob and Alice as just decohered superpositions,\n> then there\'s a very natural and local explanation of the so-called\n> non-locality: namely a state that went though two different paths, and\n> its amplitudes for each path interfere when they are added together.\n> As such it is also obvious that you can never get this FTL phone in\n> this way.\n\nWhy? I explain this with two paths of parts of some mind-force\nequipment which travels non-FTL and forces your mind to send\nthe code. By interference, to use the magic word which explains\neverything.\n\n> You shouldn\'t take my stuff on solipsism too seriously.\n\nI don\'t. I just want to illustrate that the solipsistic explanation is\none which is covered by my FTL-phone-argument too. You cannot\nprove to a solipsist that there really exists some phone.\n\n> It\n> was just meant to illustrate that (the way I understand current QM)\n> the wavefunction has only a reality with respect to one single\n> observer (which turns out to be indistinguishable to all timelike\n> connected events - there might be something in that, but I don\'t know\n> what).\n\nIt doesn\'t work. A solipsist can deny reality without logical\ncontradictions, but he cannot prove that something is not real.\n\n> It is only when we take the same wavefunction to describe an\n> independent reality on space-like connected events that we seem to run\n> into non-locality problems when taking the concept of measurement\n> (probably too) so far.\n\nNo. We run into the non-locality "problem" (I don\'t understand\nwhy it is a problem at all, simply a fact of life) if we want to defend\nEinstein locality and realism. Nor Einstein locality, nor realism,\nnor the description of the observation made in an ideal Bell\nexperiment have to mention the word "quantum" or some details\nof current quantum metaphysics.\n\n> I\'m not religiously attached to anything\n> (locality, linearity of QM, whatever). I\'m not trying to modify\n> anything to the currently accepted principles ; after all they work\n> well in practice.\n\nAgreement. I also don\'t doubt the experimental predictions of\nQM.\n\n> I\'m just trying to make sense of that situation\n> within the framework as presented, and I think I can,\n\nme too. Realistic interpretations like BM are possible. They fit\ninto the framework as presented (give the experimental predictions\nof QM). "Making sense" is something I consider to be equivalent\nto "realistic theory".\n\n> up to a point,\n> when you consider that all an observer will ever do is make\n> observations which are timelike connected.\n> But this doesn\'t mean - as you seem to suggest - that he cannot make\n> tests about a genuine non-locality.\n\nThat\'s not my point. Of course, we can make tests that an FTL phone\nworks. Its quite easy.\n\nThe point is that in these simple and easy tests we nonetheless use\nsome common sense principles, some minimal realism. But if we\naccept this minimal type of realism, most of the argumentation against\nhidden FTL in Bell experiments appears to be nonsensical. The only\nreasonable difference which remains is that a Bell test is an indirect\nobservation, it does not allow to observe the direction of the hidden\ncausal influence. We "see" "A->B or B->A" in Bell tests in the same\nobvious way as we "see" "A->B and B->A" in a working phone.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Patrick Van Esch" <vanesch@ill.fr> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> > > Not really. A working FTL phone would allow me, as a single observer,
> > > to conclude that the signal I sent went up and down Titan in .5
> > > seconds.
> >
> > How?
> >
> By triggering the nuclear bombs on titan with the special code I have.
> If I see a big fireball in the sky after 1 light-speed propagation
> time after I sent the message, I know my code arrived almost
> immediately.

Ok, that means verification of the information with some other,
non-FTL information channel.

> You cannot do that with an EPR like set-up, and what I wanted to
> illustrate was (see my other post in this thread) that if you consider
> the measurements at Bob and Alice as just decohered superpositions,
> then there's a very natural and local explanation of the so-called
> non-locality: namely a state that went though two different paths, and
> its amplitudes for each path interfere when they are added together.
> As such it is also obvious that you can never get this FTL phone in
> this way.

Why? I explain this with two paths of parts of some mind-force
equipment which travels non-FTL and forces your mind to send
the code. By interference, to use the magic word which explains
everything.

> You shouldn't take my stuff on solipsism too seriously.

I don't. I just want to illustrate that the solipsistic explanation is
one which is covered by my FTL-phone-argument too. You cannot
prove to a solipsist that there really exists some phone.

> It
> was just meant to illustrate that (the way I understand current QM)
> the wavefunction has only a reality with respect to one single
> observer (which turns out to be indistinguishable to all timelike
> connected events - there might be something in that, but I don't know
> what).

It doesn't work. A solipsist can deny reality without logical
contradictions, but he cannot prove that something is not real.

> It is only when we take the same wavefunction to describe an
> independent reality on space-like connected events that we seem to run
> into non-locality problems when taking the concept of measurement
> (probably too) so far.

No. We run into the non-locality "problem" (I don't understand
why it is a problem at all, simply a fact of life) if we want to defend
Einstein locality and realism. Nor Einstein locality, nor realism,
nor the description of the observation made in an ideal Bell
experiment have to mention the word "quantum" or some details
of current quantum metaphysics.

> I'm not religiously attached to anything
> (locality, linearity of QM, whatever). I'm not trying to modify
> anything to the currently accepted principles ; after all they work
> well in practice.

Agreement. I also don't doubt the experimental predictions of
QM.

> I'm just trying to make sense of that situation
> within the framework as presented, and I think I can,

me too. Realistic interpretations like BM are possible. They fit
into the framework as presented (give the experimental predictions
of QM). "Making sense" is something I consider to be equivalent
to "realistic theory".

> up to a point,
> when you consider that all an observer will ever do is make
> observations which are timelike connected.
> But this doesn't mean - as you seem to suggest - that he cannot make
> tests about a genuine non-locality.

That's not my point. Of course, we can make tests that an FTL phone
works. Its quite easy.

The point is that in these simple and easy tests we nonetheless use
some common sense principles, some minimal realism. But if we
accept this minimal type of realism, most of the argumentation against
hidden FTL in Bell experiments appears to be nonsensical. The only
reasonable difference which remains is that a Bell test is an indirect
observation, it does not allow to observe the direction of the hidden
causal influence. We "see" "A->B or B->A" in Bell tests in the same
obvious way as we "see" "A->B and B->A" in a working phone.

Ilja

seratend

Nov3-04, 09:51 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<clnh1g\\$eti\\$1@beech.fernuni-hagen.de>...\n> "seratend" <ser_monmail@yahoo.fr> schrieb\n> > After all these posts, as an external reader, I still not understand\n> > the point of view of Ilia nor what could be its FTL phone.\n> > First what do you intend by causality? For me, like with the QM\n> > measurement, it is very difficult to understand what it is or what it\n> > says (human interpretation).\n>\n> I press a button by my free choice, and whenever I press it the\n> light bulb changes its state. In this case, I conclude that there is\n> a causal influence from my behaviour to the state of the light bulb.\n>\n> I make this conclusion based on the observation of some correlations.\n> Of course, I also use some other assumptions: That there is some\n> reality which defines the state of the light bulb, and certain assumptions\n> about independence of my free choices from external influences.\n>\n> I believe that these basic assumptions are the foundation of the\n> scientific method, that it is reasonable to hold them even in\n> circumstances where no simple realistic explanation is known.\n\nWell, as a simple-minded person (in a non-pejorative sense :) in\nphysics, I am always afraid with arguments/words that are too far from\nmathematics and too philosophical and too general such as the reality\none. Just because, when I want to make some "physical-mathematical"\nlogical deduction that use such a word/argument, I have to interpret\nit (ie, I have to attach to this word at least a coherent mathematical\nrepresentation) and thus I risk to choose the definition that arrange\nmy logical deduction.\n\nThe EPR 1935 paper is the example of a definition of this "reality":\nThey start with this sentence: "The elements of the physical reality\ncannot be determined by a priori philosophical considerations, but\nmust be found by an appeal to results of experiments and\nmeasurements." At least they agree that it seems to be a need for a\nkind of formal logical deduction.\nIn addition, they follow by their own definition of reality "if,\nwithout in any way disturbing a system, we can predict with certainty\n(i.e. with probability equal to unity) the value of a physical\nquantity, then there exists an element of physical reality\ncorresponding to this physical quantity."\n\nTherefore, we may see that even this earlier definition of physical\nreality is very fuzzy and in fact is selected, a priori, for their\npaper argumentation. Remark that this definition does not exclude a\nglobally defined physical quantity (i.e. a reality is not assumed\nlocal by this definition).\n\nIf I go back to this FTL phone, I would like to underline the dangers\nof using black boxes where there concepts are too fuzzy (not directly\nconnected to a "physico-mathematical formulation", a very fuzzy term\n;).\n\n> I see, I have not sufficiently explained that it is based on Popper\'s\n> falsifiability of scientific, physical theory. To distinghish science from\n> pseudosciences Popper has proposed that the distinguishing\n> property of physical theories is that they make nontrivial predictions,\n> predictions which may appear false if we test them. Such theories\n> are called falsifiable. Theories which cannot be falsified are not\n> empirical theories, they are metaphysical. This point of view on science\n> is IMHO correct and widely accepted among scientists.\n\nI understand it and I agree partly with this statement. The "partly"\nis due to the implicit hypothesis that the theories may be falsified.\nThe main problem with modern theories (modern in the sense the few\nones I know) is that they are constructed to support everything. It is\nalways the external assumptions, sometimes implicit, that restricts\nthem and thus may be falsified:\ne.g Newtons\' law: ma=force: if I authorise any force, I may have any\nacceleration, and any x(t) if I enlarge the derivation. So, I can say,\nformally in the sense I may recover "ma"= "force", that newton\'s law\nis always true even at the quantum level: it is the bohemian\ninterpretation.\n\nSometimes, I prefer to say a theory is "wrong" because its formulation\nis more complicated than another formulation (I am lazy person with a\nlimited brain power : ), but it is a dangerous thought...\n\n> My FTL phone is only a hypothetical feature. [...]\n\nThe only difficult point with the FTL phone argumentation resides in\nits "fuzzy nature" that does not allow one to make reasonable logical\ndeductions based on a logical model. The main point I currently\nidentify with this fuzzy model is its possible lack of\n"reality/locality" (my reality/locality : ), because you just removed\ntoo many properties of your black box.\nE.g. your sentence "I press a button by my free choice", so if I\nassume that the button is a local one (an extra hypothesis to your\nmodel).\nI may also suppose that your FTL model is based on local interactions\n(the "press" interaction). With this assumption and mathematical\n"causality" (the logical "=>" symbol), I think I can deduce that this\nmodel works if there exists an FTL interaction on at least a frame.\nIf it is your assumption, I prefer to start with the possibility of\nthe existence of an interaction with an infinite speed in at least one\nframe: this model is much well defined (at a logical point of view),\nthan a FTL phone. I may also start with a non-local interaction model\n(the button is a universe button that I can press locally).\n\n>\n> I think about causality on a more fundamental level. The observation\n> of a correlation between A and B is sufficient to search for a realistic\n> causal explanation. Such explanations are a common cause\n> C->A, C->B or a direct causal influence A->B or B->A. Time has\n> not been mentioned in this consideration.\n>\n\nThis is the problem, it seems to be too fundamental. You are removing\ntoo many things. At least to realise an experiment (i.e. If you want a\ntheory that may be falsified), you need time, at least in one step of\nthe experiment, to make a "causal" interpretation.\n\n>\n> Now, Einstein causality is, in the general understanding, a physical\n> theory. It predicts that no FTL phones are possible. At least, this is\n> the usual interpretation of Einstein causality.\n\nI will say that Einstein causality is an extra postulate. And, I am\nafraid of the multiple interpretations we give to these words\n"Einstein causality".\nFor example with the SR, we may deduce a causality interpretation as\nthat "reality" belongs to the light cone of the space-time (tb>ta\nwhatever SR frame we select) and from the updated ma=force equation.\nI do not even speak about GR, where you need to add the locality to\ndefine a local causality.\n\nIn QFT, the causality is interpreted in a slightly different sense:\nwhether a measurement performed at one point can affect a measurement\nat another point whose separation from the first is spacelike\n(Peskin-Schroeder QFT reference).\nThere are other postulates such as Weinberg cluster decomposition that\ndefines its own causality (weaker), and in fine, it gives almost the\nsame "results".\n\n>\n> What would happen if we would really observe some FTL phone?\n> Very simple, Einstein causality would be falsified. Nobody would\n> defend relativity in this case. It would survive in physics as a limit\n> of some other theory which allows FTL phones.\n>\n> This is not the case today. Using the strange correlations of EPR\n> and Bell we cannot build FTL phones. This is a simple theorem.\n\nWell, I should first know what a FTL phone is. For example if the\nphenomenon is an interaction FTL I would rather first keep relativity\nbut say that maybe "reality" also exists in space like intervals (the\nfamous "tachyons" or whatever else). After, I will try to make a\nworking model (with experiments). After all, I will see what type of\ncausality is broken (if it is really broken).\n\n>\n> Nonetheless, that does not mean there is no problem for relativity.\n> The only realistic explanations of these correlations are A->B or\n> B->A, above are in contradiction with relativity. Now, defenders\n> of relativity use various argumentations. But, as I have observed,\n> could be applied as well in the hypothetical case of falsification of\n> relativity. That means, if we take them seriously, we can no longer\n> predict that no FTL phone exists - because such an FTL phone\n> could be explained away if we observe it.\n>\nThis is the main problem with general arguments. You can adapt them as\na theory to what you want.\nI agree, and I think every physicist, that SR or GR have a limited\ndomain of application. The main problem is that the mathematical\nformulation of these theories as well a QM does not really need\ncausality. It is always an external restriction in order to say this\nis the reality (i.e. to recover the experiments).\n\n> > So, Do you intend "->" (your causality of the FTL phone) as the\n> > mathematical logic symbol "=>" ?\n>\n> No. In mathematical terms it means absence of independence.\n\nWhow, what a very weak and fuzzy definition.\n\nSo I will try to precise the implicit context of this definition.\nI assume that we use the zermello-frankel axiomatic set theory at\nleast with the choice axiom: we assume that A and B may be points\nbelonging to a set (the "points" may be another sets if we want). We\ncall this set the universe.\nI assume you are not using a weaker theory like the category theory\nfor the initial definition of the mathematical objects (i.e what is A,\nB, o need for a universe that is a set).\n\nI thus interpret the "absence of independence" of "->" by the\nexistence of an application f (we can instead choose a relation, but\nit gives no more information and may be more difficult to interpret)\nsuch as we have:\n\nB -> A <=> A= f(B, other points if we want= context).\n\nWe can also complicate the formulation (closer to a relation):\nB->A <=> (A,context2)= f(B, context1) (\n\nThe important point is the surjection of the application on its domain\nof definition (if A1=f(input) and A2=f(input), A1=A2)\n\nSo B -> A <=> (context1, B) => (A, context2); without loss of\ngenerality.\n(context1 may be for example the domain of validity of the logical\n"=>" relation)\n\nWe may also rename them, i.e. we redefine a new set (context1, B)=B,\n(A, context2)=B in order to recover a simple logical A=>B.\n\nThus in my previous post we just need to add the "context" property to\nthe A and B points.\n\n> Realism assumes\n>\n> Result_at_A = f_a(reality, decisions_at_A,decisions_at_B)\n\nfor B -> A?\n\n> Result_at_B = f_b(reality, decisions_at_A,decisions_at_B)\n\nfor A -> B?\n\n>\n> Causality gives the additional restriction\n>\n> Result_at_A = f_a(reality, decisions_at_A)\n> Result_at_B = f_b(reality, decisions_at_B)\n>\nYou are introducing three new notions: decisions, reality and results\nthat are difficult to map into a physical theory. But, your are giving\nsome details of some of your assumptions (in the current hypothesis:\nthe FTL phone is described by an unknown physics).\n\nFirst what is the Result_at_A? Because you have not defined the\nspace-time object, I assume result_at_A is not obliged to be at a\npoint in space: it is only attached to A, to be as general as\npossible.\nI have defined a set of points containing A and B, … (the universe)\nAnd your text above says that A and B are also sets [A=set of (all\npossible results), B= set of (all possible results)]. (here I add a\nnew postulate: all possible results are a set).\n\nSo your are building applications to these sets (f_a for the set A,\nf_b for the set B): ok.\nNow, what represents the inputs of these applications: decisions_at_A,\ndecision_at_B ? Does it exist any relation between decisions and\nresults in a large sense? I need it in order to understand what your\ndeductions may involve.\n\nI do not understand your "reality":\n-- We may take some reality aspects as the domain of validity of the\nfunctions f_a and f_b . (some values of\n(decisions_at_A,decisions_at_B) does not belong to the reality).\n-- reality may also be the other decisions_at_other_objects or\nresults_at_other_objects (the additional context).\n-- is it an object of the universe (the set containing the points A,B\n....)\n-- ???\n\n> Causality gives the additional restriction\n>\n> Result_at_A = f_a(reality, decisions_at_A)\n> Result_at_B = f_b(reality, decisions_at_B)\n\n\nNow with this definition and the previous assumptions, I think here\nyou are mixing causality with locality (I still have questions about\nwhat is "reality" and "decisions_at_object"). Even if we don\'t have\nspecified what we intend by locality, I don\'t think that we can say\nthat "causality" => "locality" without additional assumptions (e.g. QM\ncan be said causal but not local to some extent).\n\n\n> > Now, to end, I will try to show a possible implementation example of a\n> > formal FTL phone based solely on your "causal" argument.\n> >\n> > Let\'s take 2 classical independent particles (no interaction) with a\n> > total kinetic energy E= E1+E2.\n> > Now if we change the kinetic energy of particle 1, we thus induce\n> > instantaneously (causality) a change on the kinetic energy of the 2\n> > particles (E): We have formally a FTL phone with the total Energy E.\n>\n>\n> That\'s not a phone because we cannot use it to talk with each other,\n> nor formal nor informal.\n\nWhy not?\nFormally, we have first to define what is to talk to each other (I do\nnot involve necessarily human objects, nor local objects: it is my\nfreedom choice with your weak model).\nI think you are implicitly restricting the FTL phone model to local\nvariables and thus you are mixing causality with locality.\nIt is what I want to underline with this FTL phone fuzzy model: we do\nnot specify enough to make reasonable logical deductions (i.e. all the\ndeductions have many implicit assumptions).\n\n> > So when we say that a local measurement on one EPR particle implies\n> > the change of the global state |psi> we are very close to the previous\n> > example. To detect the change of |psi> at the other particle location,\n> > we should have a black box (like the previous one for the total energy\n> > E) that is able to measure the global state (and not the local one).\n>\n> The whole point is a different one. We really have some black box.\n> This black box allows to do very strange things. Unfortunately it\n> does not allow us to build an FTL phone. But there exists no realistic\n> explanation of what is inside the black box which does not use some\n> sort of hidden FTL information transfer.\n>\n> Ilja\n\nI agree we may not be able to build a "local" FTL phone with QM\nmechanics, even if I am afraid I am taking implicit hypotheses to say\nthat ; ). But I a FTL phone is possible I quite sure that an adequate\nSchroedinger equation may describe its behaviour.\nI also say that a QM state is not interpreted into an obvious\nkolgomorovian probabilistic event (whatever probabilistic set and\nsigma-algebra is chosen). The things even become rather difficult when\nwe take a global state with global observables and want to deduce if\nthere is a hidden FTL local channel.\nThe EPR state is a global state where the 2 particles have the same\nspin axis (rotationally invariant state): this is the global property\ncommon to the 2 EPR particles. We just can say that a classical\nstatistical hidden variable model, without an hidden FTL channel is\nnot able to represent the statistical results given by this state.\n\nFinally, for the ones interested by a hidden variable model of EPR\nstate with a hidden FTL channel, I recommend the last Cerf, Gisisn\nMassar and Popescu quant-ph/0410027 paper (4 pages – a short concise\none). I think it is a good one (with the pointers it gives).\nIt introduces a possible (logical) implementation of what they call\nthe non-local machines (PR- machines, I like the "machine" term they\nchosen rather than "interaction"). We have another different look to\nthe EPR like states.\n\nI would like to know if you would you say that this model is causal or\nnot and why?\n\nSeratend\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<clnh1g$eti$1@beech.fernuni-hagen.de>...
> "seratend" <ser_monmail@yahoo.fr> schrieb
> > After all these posts, as an external reader, I still not understand
> > the point of view of Ilia nor what could be its FTL phone.
> > First what do you intend by causality? For me, like with the QM
> > measurement, it is very difficult to understand what it is or what it
> > says (human interpretation).
>
> I press a button by my free choice, and whenever I press it the
> light bulb changes its state. In this case, I conclude that there is
> a causal influence from my behaviour to the state of the light bulb.
>
> I make this conclusion based on the observation of some correlations.
> Of course, I also use some other assumptions: That there is some
> reality which defines the state of the light bulb, and certain assumptions
> about independence of my free choices from external influences.
>
> I believe that these basic assumptions are the foundation of the
> scientific method, that it is reasonable to hold them even in
> circumstances where no simple realistic explanation is known.

Well, as a simple-minded person (in a non-pejorative sense :) in
physics, I am always afraid with arguments/words that are too far from
mathematics and too philosophical and too general such as the reality
one. Just because, when I want to make some "physical-mathematical"
logical deduction that use such a word/argument, I have to interpret
it (ie, I have to attach to this word at least a coherent mathematical
representation) and thus I risk to choose the definition that arrange
my logical deduction.

The EPR 1935 paper is the example of a definition of this "reality":
They start with this sentence: "The elements of the physical reality
cannot be determined by a priori philosophical considerations, but
must be found by an appeal to results of experiments and
measurements." At least they agree that it seems to be a need for a
kind of formal logical deduction.
In addition, they follow by their own definition of reality "if,
without in any way disturbing a system, we can predict with certainty
(i.e. with probability equal to unity) the value of a physical
quantity, then there exists an element of physical reality
corresponding to this physical quantity."

Therefore, we may see that even this earlier definition of physical
reality is very fuzzy and in fact is selected, a priori, for their
paper argumentation. Remark that this definition does not exclude a
globally defined physical quantity (i.e. a reality is not assumed
local by this definition).

If I go back to this FTL phone, I would like to underline the dangers
of using black boxes where there concepts are too fuzzy (not directly
connected to a "physico-mathematical formulation", a very fuzzy term
;).

> I see, I have not sufficiently explained that it is based on Popper's
> falsifiability of scientific, physical theory. To distinghish science from
> pseudosciences Popper has proposed that the distinguishing
> property of physical theories is that they make nontrivial predictions,
> predictions which may appear false if we test them. Such theories
> are called falsifiable. Theories which cannot be falsified are not
> empirical theories, they are metaphysical. This point of view on science
> is IMHO correct and widely accepted among scientists.

I understand it and I agree partly with this statement. The "partly"
is due to the implicit hypothesis that the theories may be falsified.
The main problem with modern theories (modern in the sense the few
ones I know) is that they are constructed to support everything. It is
always the external assumptions, sometimes implicit, that restricts
them and thus may be falsified:
e.g Newtons' law: ma=force: if I authorise any force, I may have any
acceleration, and any x(t) if I enlarge the derivation. So, I can say,
formally in the sense I may recover "ma"= "force", that newton's law
is always true even at the quantum level: it is the bohemian
interpretation.

Sometimes, I prefer to say a theory is "wrong" because its formulation
is more complicated than another formulation (I am lazy person with a
limited brain power : ), but it is a dangerous thought...

> My FTL phone is only a hypothetical feature. [...]

The only difficult point with the FTL phone argumentation resides in
its "fuzzy nature" that does not allow one to make reasonable logical
deductions based on a logical model. The main point I currently
identify with this fuzzy model is its possible lack of
"reality/locality" (my reality/locality : ), because you just removed
too many properties of your black box.
E.g. your sentence "I press a button by my free choice", so if I
assume that the button is a local one (an extra hypothesis to your
model).
I may also suppose that your FTL model is based on local interactions
(the "press" interaction). With this assumption and mathematical
"causality" (the logical "=>" symbol), I think I can deduce that this
model works if there exists an FTL interaction on at least a frame.
If it is your assumption, I prefer to start with the possibility of
the existence of an interaction with an infinite speed in at least one
frame: this model is much well defined (at a logical point of view),
than a FTL phone. I may also start with a non-local interaction model
(the button is a universe button that I can press locally).

>
> I think about causality on a more fundamental level. The observation
> of a correlation between A and B is sufficient to search for a realistic
> causal explanation. Such explanations are a common cause
> C->A, C->B or a direct causal influence A->B or B->A. Time has
> not been mentioned in this consideration.
>

This is the problem, it seems to be too fundamental. You are removing
too many things. At least to realise an experiment (i.e. If you want a
theory that may be falsified), you need time, at least in one step of
the experiment, to make a "causal" interpretation.

>
> Now, Einstein causality is, in the general understanding, a physical
> theory. It predicts that no FTL phones are possible. At least, this is
> the usual interpretation of Einstein causality.

I will say that Einstein causality is an extra postulate. And, I am
afraid of the multiple interpretations we give to these words
"Einstein causality".
For example with the SR, we may deduce a causality interpretation as
that "reality" belongs to the light cone of the space-time (tb>ta
whatever SR frame we select) and from the updated ma=force equation.
I do not even speak about GR, where you need to add the locality to
define a local causality.

In QFT, the causality is interpreted in a slightly different sense:
whether a measurement performed at one point can affect a measurement
at another point whose separation from the first is spacelike
(Peskin-Schroeder QFT reference).
There are other postulates such as Weinberg cluster decomposition that
defines its own causality (weaker), and in fine, it gives almost the
same "results".

>
> What would happen if we would really observe some FTL phone?
> Very simple, Einstein causality would be falsified. Nobody would
> defend relativity in this case. It would survive in physics as a limit
> of some other theory which allows FTL phones.
>
> This is not the case today. Using the strange correlations of EPR
> and Bell we cannot build FTL phones. This is a simple theorem.

Well, I should first know what a FTL phone is. For example if the
phenomenon is an interaction FTL I would rather first keep relativity
but say that maybe "reality" also exists in space like intervals (the
famous "tachyons" or whatever else). After, I will try to make a
working model (with experiments). After all, I will see what type of
causality is broken (if it is really broken).

>
> Nonetheless, that does not mean there is no problem for relativity.
> The only realistic explanations of these correlations are A->B or
> B->A, above are in contradiction with relativity. Now, defenders
> of relativity use various argumentations. But, as I have observed,
> could be applied as well in the hypothetical case of falsification of
> relativity. That means, if we take them seriously, we can no longer
> predict that no FTL phone exists - because such an FTL phone
> could be explained away if we observe it.
>
This is the main problem with general arguments. You can adapt them as
a theory to what you want.
I agree, and I think every physicist, that SR or GR have a limited
domain of application. The main problem is that the mathematical
formulation of these theories as well a QM does not really need
causality. It is always an external restriction in order to say this
is the reality (i.e. to recover the experiments).

> > So, Do you intend "->" (your causality of the FTL phone) as the
> > mathematical logic symbol "=>" ?
>
> No. In mathematical terms it means absence of independence.

Whow, what a very weak and fuzzy definition.

So I will try to precise the implicit context of this definition.
I assume that we use the zermello-frankel axiomatic set theory at
least with the choice axiom: we assume that A and B may be points
belonging to a set (the "points" may be another sets if we want). We
call this set the universe.
I assume you are not using a weaker theory like the category theory
for the initial definition of the mathematical objects (i.e what is A,
B, o need for a universe that is a set).

I thus interpret the "absence of independence" of "->" by the
existence of an application f (we can instead choose a relation, but
it gives no more information and may be more difficult to interpret)
such as we have:

B -> A <=> A= f(B,[/itex] other points if we want= context).

We can also complicate the formulation (closer to a relation):
B->A <=> (A,context2)= f(B, context1) (

The important point is the surjection of the application on its domain
of definition (if A1=f(input) and A2=f(input), A1=A2)

So B -> A <=> (context1, B) => (A, context2); without loss of
generality.
(context1 may be for example the domain of validity of the logical
"=>" relation)

We may also rename them, i.e. we redefine a new set (context1, B)=B,
(A, context2)=B in order to recover a simple logical A=>B.

Thus in my previous post we just need to add the "context" property to
the A and B points.

> Realism assumes
>
> Result_at_A = f_a(reality, decisions_at_A,decisions_at_B)

for B -> A?

> Result_at_B = f_b(reality, decisions_at_A,decisions_at_B)

for A -> B?

>
> Causality gives the additional restriction
>
> Result_at_A = f_a(reality, decisions_at_A)
> Result_at_B = f_b(reality, decisions_at_B)
>
You are introducing three new notions: decisions, reality and results
that are difficult to map into a physical theory. But, your are giving
some details of some of your assumptions (in the current hypothesis:
the FTL phone is described by an unknown physics).

First what is the Result_at_A? Because you have not defined the
space-time object, I assume result_at_A is not obliged to be at a
point in space: it is only attached to A, to be as general as
possible.
I have defined a set of points containing A and B, … (the universe)
And your text above says that A and B are also sets [A=set of (all
possible results), B= set of (all possible results)]. (here I add a
new postulate: all possible results are a set).

So your are building applications to these sets (f_a for the set A,
f_b for the set B): ok.
Now, what represents the inputs of these applications: decisions_at_A,decision_at_B ? Does it exist any relation between decisions and
results in a large sense? I need it in order to understand what your
deductions may involve.

I do not understand your "reality":
-- We may take some reality aspects as the domain of validity of the
functions f_a and f_b . (some values of
(decisions_at_A,decisions_at_B) does not belong to the reality).
-- reality may also be the other decisions_at_other_objects or
results_at_other_objects (the additional context).
-- is it an object of the universe (the set containing the points A,B
....)
-- ???

> Causality gives the additional restriction
>
> Result_at_A = f_a(reality, decisions_at_A)
> Result_at_B = f_b(reality, decisions_at_B)

Now with this definition and the previous assumptions, I think here
you are mixing causality with locality (I still have questions about
what is "reality" and "decisions_at_object"). Even if we don't have
specified what we intend by locality, I don't think that we can say
that "causality" => "locality" without additional assumptions (e.g. QM
can be said causal but not local to some extent).

> > Now, to end, I will try to show a possible implementation example of a
> > formal FTL phone based solely on your "causal" argument.
> >
> > Let's take 2 classical independent particles (no interaction) with a
> > total kinetic energy [itex]E= E1+E2.
> > Now if we change the kinetic energy of particle 1, we thus induce
> > instantaneously (causality) a change on the kinetic energy of the 2
> > particles (E): We have formally a FTL phone with the total Energy E.
>
>
> That's not a phone because we cannot use it to talk with each other,
> nor formal nor informal.

Why not?
Formally, we have first to define what is to talk to each other (I do
not involve necessarily human objects, nor local objects: it is my
freedom choice with your weak model).
I think you are implicitly restricting the FTL phone model to local
variables and thus you are mixing causality with locality.
It is what I want to underline with this FTL phone fuzzy model: we do
not specify enough to make reasonable logical deductions (i.e. all the
deductions have many implicit assumptions).

> > So when we say that a local measurement on one EPR particle implies
> > the change of the global state |\psi> we are very close to the previous
> > example. To detect the change of |\psi> at the other particle location,
> > we should have a black box (like the previous one for the total energy
> > E) that is able to measure the global state (and not the local one).
>
> The whole point is a different one. We really have some black box.
> This black box allows to do very strange things. Unfortunately it
> does not allow us to build an FTL phone. But there exists no realistic
> explanation of what is inside the black box which does not use some
> sort of hidden FTL information transfer.
>
> Ilja

I agree we may not be able to build a "local" FTL phone with QM
mechanics, even if I am afraid I am taking implicit hypotheses to say
that ; ). But I a FTL phone is possible I quite sure that an adequate
Schroedinger equation may describe its behaviour.
I also say that a QM state is not interpreted into an obvious
kolgomorovian probabilistic event (whatever probabilistic set and
\sigma-algebra is chosen). The things even become rather difficult when
we take a global state with global observables and want to deduce if
there is a hidden FTL local channel.
The EPR state is a global state where the 2 particles have the same
spin axis (rotationally invariant state): this is the global property
common to the 2 EPR particles. We just can say that a classical
statistical hidden variable model, without an hidden FTL channel is
not able to represent the statistical results given by this state.

Finally, for the ones interested by a hidden variable model of EPR
state with a hidden FTL channel, I recommend the last Cerf, Gisisn
Massar and Popescu http://www.arxiv.org/abs/quant-ph/0410027 paper (4 pages – a short concise
one). I think it is a good one (with the pointers it gives).
It introduces a possible (logical) implementation of what they call
the non-local machines (PR- machines, I like the "machine" term they
chosen rather than "interaction"). We have another different look to
the EPR like states.

I would like to know if you would you say that this model is causal or
not and why?

Seratend

Thomas Trotter

Nov3-04, 09:58 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>[snippage throughout]\n\n> > Thomas Trotter (TT)\n> > Maybe it\'s just that knowledge of what is\n> > actually happening wrt the emission and detection of the opposite\n> > moving light beams is incomplete (ie., that there *is* a local,\n> > common-property reason for the correlations, and not enough is\n> > known about the physical process to talk about it in a straighforward\n> > local realistic way yet).\n\nIlja Schmelzer (IS):\n> Given Bell\'s theorem, there is no such explanation. This is independent\n> of our state of knowledge or state of our theories.\n\nTT:\nThen, you are saying that the correlation curve has\nnothing to do with the properties of the light incident\non the polarizers during a given coincidence interval?\n\nBecause, if the correlation curve *does* have something\nto do with this, then it would seem to follow that\nthese properties were produced via emission.\n\nOtherwise, the polarizers aren\'t *analyzing* anything.\nThey\'re just producing some optical state which is then\ninstantaneously transmitted to the other arm of the\nsetup.\n\nIf that\'s the case, then we shouldn\'t need to worry\nabout whether or not we\'re dealing (in a given coincidence\ninterval) with light emitted by the same oscillator.\n\nAnd yet experimenters take great pains to insure that\nthey *are* dealing with light emitted by the same atom\nduring a given coincidence interval.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>[snippage throughout]

> > Thomas Trotter (TT)
> > Maybe it's just that knowledge of what is
> > actually happening wrt the emission and detection of the opposite
> > moving light beams is incomplete (ie., that there *is* a local,
> > common-property reason for the correlations, and not enough is
> > known about the physical process to talk about it in a straighforward
> > local realistic way yet).

Ilja Schmelzer (IS):
> Given Bell's theorem, there is no such explanation. This is independent
> of our state of knowledge or state of our theories.

TT:
Then, you are saying that the correlation curve has
nothing to do with the properties of the light incident
on the polarizers during a given coincidence interval?

Because, if the correlation curve *does* have something
to do with this, then it would seem to follow that
these properties were produced via emission.

Otherwise, the polarizers aren't *analyzing* anything.
They're just producing some optical state which is then
instantaneously transmitted to the other arm of the
setup.

If that's the case, then we shouldn't need to worry
about whether or not we're dealing (in a given coincidence
interval) with light emitted by the same oscillator.

And yet experimenters take great pains to insure that
they *are* dealing with light emitted by the same atom
during a given coincidence interval.

seratend

Nov3-04, 09:58 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>It seems (I may be wrong : )that sometimes you are mixing collapse\nwith the measurement process while "collapse" is only a part of the\nmeasurement. I think my own measurement interpretation is close to the\none of Patrick Van Esh, but I prefer to avoid using words such as\n"consciousness" and use instead relative knowledge or conditional\nstates (as in classical probability).\n\nIn fact, I prefer to separate the measurement results from the quantum\nsystem description where the Quantum system is formally described by\n"quantum probability" (the state |psi> and its evolution in time\nU(t,to)).\nIf we accept the formal quantum probabilistic description, the\nmeasurement results are by construction external to this quantum\ndescription. These measurements results are only conditional states\nthat we may insert in the description of the evolution of the system:\n-- |psi(t)>=U(t,to) |psi> if I do not suppose any constraint on my\nsystem (in fact we are implicitly using the conditional probability\nthat QM system is in state |psi> at to).\n-- |psi(t)>=U(t,to) |measure> if I suppose that at time to my system\nwas in state |measure> (i.e. we are using conditional probability,\n|psi(t)> if (|measure>,to)).\n\ni.e. The QM formalism (minus the measurement) does not intend describe\nthe result of a measurement as in classical probability, it just gives\na probability law (where we can have some outcomes with a 100%\nconfidence).\n\nWith this kind of formulation, we are very close to the classical\nprobabilities. We may even understand better the QM measurement and\ncollapse:\nA classical mechanical system may have its probabilistic state changed\nthrough an interaction (time evolution), for example the evolution due\nto the interaction with a classical measurement apparatus. However,\nthe fact that the classical apparatus always gives a single number\n(the result of a classical measure) is external to the statistical\ndescription of the system in the sense that the statistical state of\nthe classical system may allow several possible measured values\n(different outcomes of an experiment). So when we say that measure=m\nin the probabilistic description of the classical system, we are just\nmaking conditional probabilities:\n-- we can say, I have the probability p_m to get result m\n-- Or we can say, knowing that the measure apparatus gives the m\nresult at time t, what is the probabilistic state evolution of my\nclassical system.\n\nAt the classical level, we are not wondering why a measurement always\ngives a single result or what the collapse is. This is the same for QM\n(I assume it : my formal interpretation :).\n\nThe QM measurement procedure thus gives formally an external piece of\ninformation (the measured value) about the Quantum system like the\nclassical probabilities. Nevertheless, this piece of information\nobtained at the "end" of a measurement process (a real number, e.g.\nthe eigen value of an observable) never represents a well defined\nunknown state but rather another event: we never fix the hidden\nvariables values, if they exist, by a measurement result (naïve point\nof view). We just get an instant access to a naive set of points (an\nevent, the QM state |measure> for the given value "measure") to where\nthe hidden variables belong (note that I consider this as an\napproximation, as quantum probabilities are more complicated than\nclassical probabilities).\n\nTherefore, I always try to separate the quantum measurement in 2\nindependent blocks:\n\n-- the action on the quantum system by the measurement apparatus (i.e.\nthe local measurement interaction on the Particles). It describes at\nthe quantum level what the quantum system becomes under its\ninteraction with the measurement apparatus. We always stay at the\nquantum level.\n\nWe may describe this apparatus interaction with a Hamiltonian or with\nprojectors - we may consider projectors as an approximation of the\naction of a macroscopic system (infinite particles, …). We thus have\nlocal interactions if we believe that the interactions with the\napparatus come from the know interactions (em, …). However, we are\nfree to imagine non-local interactions.\n\n-- The information: the value given by the measurement apparatus\nexperiment. We may use this external information on the QM\nprobabilistic state description and evolution. If we use it, we are in\nfact making conditional statements (the "collapse"): the system is in\nthe state that corresponds to the measures value "m" at time tm.\n\nFor example, we may describe the next state of a QM system if the\nmeasurement has given a peculiar result (a very formal view,\nequivalent to postulating the QM system and the measurement apparatus\nare in a given state). After, we may add beautiful words, such as, the\ndescription (probabilistic) of the evolution of the QM system state if\na human (or a brain, …) has seen a measurement that gives this result.\nWe are just adding conditions, adding humans or non organic entites, …\nMoreover, if we accept that the action of the apparatus is local\n(first block), thus we also accept that the "measured state" (i.e.\ncorresponding to the measured value) is also local (point of view of\nthe QM system locally to this apparatus).\n\nWhat does not explain this point of view, is why measure trials gives\nsuch values (it is outside the scope of the QM formalism as in\nclassical statistical mechanics). However, this is not very important,\nif we "accept" that quantum physics deals only with (special)\nprobabilities. : ).\n\nSo, if you replace in your post, the word "collapse", with\n"conditional state" and independently if you model the interaction of\nthe measurement apparatus as a spin projector that acts only on a very\nshort area of the space, you will get all the results you want. You\nmay add any observer you want. The "at hand" is really what we always\ndo with conditional probabilities/states: "what we get, If we suppose\nwe have this".\n\nSeratend.\n\nRalph Hartley <hartley@aic.nrl.navy.mil> wrote in message news:<cllm4u\\$6db\\$1@ra.nrl.navy.mil>...\n> Patrick Van Esch wrote:\n>\n> Any *local* objective collapse process would ruin the correlation between\n> Bob and Alice. I think this may be a very strong test of Penrose\'s\n> "gravitational wavefunction collapse". OK, not if it isn\'t local, I was\n> never too clear on that point.\n>\n> From Bob\'s point of view, his measurement partially collapses Alice\'s\n> state. He knows the polarization of her photon, but may not what her\n> measurement or result were. The collapse is instantaneous. It doesn\'t\n> matter what reference frame it is instantaneous in, since Bob sees the same\n> thing in any frame.\n>\n> From Alice\'s point of view the situation is reversed.\n>\n> From the point of view of Eve, located half way in between, observing\n> Alice and Bob with telescopes, they both remain in a superposition until\n> she observes them making their measurements.\n>\n> All three points of view are completely objective, but none is more "true"\n> than the others. (Except for *mine*, which is the *absolute* truth :-))\n>\n> You *can* view the experiment that way, but there is nothing to say that\n> you *must* do so. Interpretations of QM are interchangeable, since they\n> never disagree on results. You should always use the one that works best\n> for you in the situation at hand.\n>\n> Ralph Hartley\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>It seems (I may be wrong : )that sometimes you are mixing collapse
with the measurement process while "collapse" is only a part of the
measurement. I think my own measurement interpretation is close to the
one of Patrick Van Esh, but I prefer to avoid using words such as
"consciousness" and use instead relative knowledge or conditional
states (as in classical probability).

In fact, I prefer to separate the measurement results from the quantum
system description where the Quantum system is formally described by
"quantum probability" (the state |\psi> and its evolution in time
U(t,to)).
If we accept the formal quantum probabilistic description, the
measurement results are by construction external to this quantum
description. These measurements results are only conditional states
that we may insert in the description of the evolution of the system:
-- |\psi(t)>=U(t,to) |\psi> if I do not suppose any constraint on my
system (in fact we are implicitly using the conditional probability
that QM system is in state |\psi> at to).
-- |\psi(t)>=U(t,to) |measure> if I suppose that at time to my system
was in state |measure> (i.e. we are using conditional probability,
|\psi(t)> if (|measure>,to)).

i.e. The QM formalism (minus the measurement) does not intend describe
the result of a measurement as in classical probability, it just gives
a probability law (where we can have some outcomes with a 100%
confidence).

With this kind of formulation, we are very close to the classical
probabilities. We may even understand better the QM measurement and
collapse:
A classical mechanical system may have its probabilistic state changed
through an interaction (time evolution), for example the evolution due
to the interaction with a classical measurement apparatus. However,
the fact that the classical apparatus always gives a single number
(the result of a classical measure) is external to the statistical
description of the system in the sense that the statistical state of
the classical system may allow several possible measured values
(different outcomes of an experiment). So when we say that measure=m
in the probabilistic description of the classical system, we are just
making conditional probabilities:
-- we can say, I have the probability p_m to get result m
-- Or we can say, knowing that the measure apparatus gives the m
result at time t, what is the probabilistic state evolution of my
classical system.

At the classical level, we are not wondering why a measurement always
gives a single result or what the collapse is. This is the same for QM
(I assume it : my formal interpretation :).

The QM measurement procedure thus gives formally an external piece of
information (the measured value) about the Quantum system like the
classical probabilities. Nevertheless, this piece of information
obtained at the "end" of a measurement process (a real number, e.g.
the eigen value of an observable) never represents a well defined
unknown state but rather another event: we never fix the hidden
variables values, if they exist, by a measurement result (naïve point
of view). We just get an instant access to a naive set of points (an
event, the QM state |measure> for the given value "measure") to where
the hidden variables belong (note that I consider this as an
approximation, as quantum probabilities are more complicated than
classical probabilities).

Therefore, I always try to separate the quantum measurement in 2
independent blocks:

-- the action on the quantum system by the measurement apparatus (i.e.
the local measurement interaction on the Particles). It describes at
the quantum level what the quantum system becomes under its
interaction with the measurement apparatus. We always stay at the
quantum level.

We may describe this apparatus interaction with a Hamiltonian or with
projectors - we may consider projectors as an approximation of the
action of a macroscopic system (infinite particles, …). We thus have
local interactions if we believe that the interactions with the
apparatus come from the know interactions (em, …). However, we are
free to imagine non-local interactions.

-- The information: the value given by the measurement apparatus
experiment. We may use this external information on the QM
probabilistic state description and evolution. If we use it, we are in
fact making conditional statements (the "collapse"): the system is in
the state that corresponds to the measures value "m" at time tm.

For example, we may describe the next state of a QM system if the
measurement has given a peculiar result (a very formal view,
equivalent to postulating the QM system and the measurement apparatus
are in a given state). After, we may add beautiful words, such as, the
description (probabilistic) of the evolution of the QM system state if
a human (or a brain, …) has seen a measurement that gives this result.
We are just adding conditions, adding humans or non organic entites, …
Moreover, if we accept that the action of the apparatus is local
(first block), thus we also accept that the "measured state" (i.e.
corresponding to the measured value) is also local (point of view of
the QM system locally to this apparatus).

What does not explain this point of view, is why measure trials gives
such values (it is outside the scope of the QM formalism as in
classical statistical mechanics). However, this is not very important,
if we "accept" that quantum physics deals only with (special)
probabilities. : ).

So, if you replace in your post, the word "collapse", with
"conditional state" and independently if you model the interaction of
the measurement apparatus as a spin projector that acts only on a very
short area of the space, you will get all the results you want. You
may add any observer you want. The "at hand" is really what we always
do with conditional probabilities/states: "what we get, If we suppose
we have this".

Seratend.

Ralph Hartley <hartley@aic.nrl.navy.mil> wrote in message news:<cllm4u$6db$1@ra.nrl.navy.mil>...
> Patrick Van Esch wrote:
>
> Any *local* objective collapse process would ruin the correlation between
> Bob and Alice. I think this may be a very strong test of Penrose's
> "gravitational wavefunction collapse". OK, not if it isn't local, I was
> never too clear on that point.
>
> From Bob's point of view, his measurement partially collapses Alice's
> state. He knows the polarization of her photon, but may not what her
> measurement or result were. The collapse is instantaneous. It doesn't
> matter what reference frame it is instantaneous in, since Bob sees the same
> thing in any frame.
>
> From Alice's point of view the situation is reversed.
>
> From the point of view of Eve, located half way in between, observing
> Alice and Bob with telescopes, they both remain in a superposition until
> she observes them making their measurements.
>
> All three points of view are completely objective, but none is more "true"
> than the others. (Except for *mine*, which is the *absolute* truth :-))
>
> You *can* view the experiment that way, but there is nothing to say that
> you *must* do so. Interpretations of QM are interchangeable, since they
> never disagree on results. You should always use the one that works best
> for you in the situation at hand.
>
> Ralph Hartley

Patrick Van Esch

Nov3-04, 10:03 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410291959.28f3e245@posting.google. com>...\n\n> So, for this coincidence interval, the\n> probability of detection at B is cos^2 theta, where theta is\n> the angular difference between the transmission axes of the\n> polarizers at A and B. (And, this cos^2 theta function\n> generalizes to all coincidence intervals -- so that the\n> probability of coincidental detection, in the ideal, is\n> cos^2 theta.)\n>\n> Now, following Bell, formulate the probability of detection\n> at A and B *from the outset* (prior to any detection) using\n> the same assumptions as above (identical emission-produced\n> polarization of light incident on A and B for any given\n> coincidence interval). Denoting the common polarization as\n> lambda, and the transmission axes of the polarizers at A and\n> B as a and b, respectively:\n>\n> P(A) = cos^2 |a - lambda| and P(B) = cos^2 |b - lambda|\n>\n> If lambda is not equal to a, then how can P(B) = cos^2 |b - lambda|\n> prior to detection at A, and P(B) = cos^2 |b - a| after detection\n> at A? [2]\n\nBell is making up a "local realistic" model of the entangled pair. If\nyou say that the photons HAD a specific polarisation direction lambda,\nthen you have to be able to write it down (even if you don\'t know what\nit is). That\'s what he does.\nQuantum mechanics doesn\'t take this approach. It writes down that the\nstate is:\n\n|psi> = 1/sqrt(2) { |z+>|z-> - |z->|z+ }\n\nbut some algebra shows you that you can rewrite this as:\n\n\n|psi> = 1/sqrt(2) { |n+>|n-> - |n->|n+ }\n\nwith n just any direction. So our state is completely "unoriented"\nwhich is not possible in a realistic model.\n\nNow the locality resides in the fact that we have the SAME lambda for\nA and for B, in that there cannot be an extra variable at B that is a\nfunction of a.\nThe photons had to take all their realistic description at their\ncreation and then nothing else for one photon can be dependent on what\nhappens to the other photon.\n\n>\n> Anyway, there are two incompatible formulations involved here.\n\nExactly, that was Bell\'s point.\n\n\n> What\'s wrong with Bell\'s formulation? It doesn\'t seem to be,\n> strictly speaking, the assumption of locality -- since qm also\n> uses the assumption of a common emitter as the basis for\n> calculations.\n\nAs I said, the locality implies that we have to use the same lambda at\nA and at B.\n\n>\n> Footnotes:\n>\n> [1] I don\'t know how the timing stuff actually works in the\n> coincidence circuitry. Like, how the coincidence interval is\n> defined. How is everything ... matched up, etc.? I do know that\n> it has to do with the atomic transition time, and the distance\n> that the polarizers are from the emitters, but am not sure\n> exactly what sequence of operations a detection triggers.\n\nExperimentally, this is quite simple. A photomultiplier gives a\nhit/no hit (pulse or no pulse) in a few nanoseconds. If the incoming\nlight flux is low enough, you have pulse rates of say, 100,000.- hits\nper second. This means that it is rather improbable to have pulses\nwithin a window of a few nanoseconds by chance (the probability is\n~1/10000). So if you have a significant rate of coincidences, this\nsimply means that these photons have a common origin (come from a\n2-photon state).\nExperimentally, one has to choose an optimum between the spurious\ncoincidences (which augment when the lightflux increases) and the dark\ncurrent rate (the pulses a photodetector generates when no light falls\nonto it), which is a nuisance at low lightflux.\n\n\n>\n> [2] My answer would be that if there\'s sufficient amplitude in the\n> transmission by the polarizer, a, to generate a detection\n> at A for some interval, then, for that interval, and for theta = 0,\n> then there will, with probability 1, be sufficient amplitude\n> in the transmission by the polarizer, b, to generate a detection\n> at B. And, following Malus\' Law, the probability of detection at\n> B will decrease as theta increases from 0 to pi/2.\n\nBe careful. I had such discussions before. There are 2 things you\nhave to keep in mind: 1) photodectectors are particle detectors\n(imperfect as they may be), and not "intensity detectors" although\noften these 2 descriptions coincide.\n2) an entangled photon state cannot be described by a maxwell field.\n\n\ncheers,\nPatrick.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410291959.28f3e245@posting.google.com>...

> So, for this coincidence interval, the
> probability of detection at B is cos^2 \theta, where \theta is
> the angular difference between the transmission axes of the
> polarizers at A and B. (And, this cos^2 \theta function
> generalizes to all coincidence intervals -- so that the
> probability of coincidental detection, in the ideal, is
> cos^2 \theta.)
>
> Now, following Bell, formulate the probability of detection
> at A and B *from the outset* (prior to any detection) using
> the same assumptions as above (identical emission-produced
> polarization of light incident on A and B for any given
> coincidence interval). Denoting the common polarization as
> \lambda, and the transmission axes of the polarizers at A and
> B as a and b, respectively:
>
> P(A) = cos^2 |a - \lambda| and P(B) = cos^2 |b - \lambda|
>
> If \lambda is not equal to a, then how can P(B) = cos^2 |b - \lambda|
> prior to detection at A, and P(B) = cos^2 |b - a| after detection
> at A? [2]

Bell is making up a "local realistic" model of the entangled pair. If
you say that the photons HAD a specific polarisation direction \lambda,
then you have to be able to write it down (even if you don't know what
it is). That's what he does.
Quantum mechanics doesn't take this approach. It writes down that the
state is:

|\psi> = 1/\sqrt(2) { |z+>|z-> - |z->|z+ }

but some algebra shows you that you can rewrite this as:

|\psi> = 1/\sqrt(2) { |n+>|n-> - |n->|n+ }

with n just any direction. So our state is completely "unoriented"
which is not possible in a realistic model.

Now the locality resides in the fact that we have the SAME \lambda for
A and for B, in that there cannot be an extra variable at B that is a
function of a.
The photons had to take all their realistic description at their
creation and then nothing else for one photon can be dependent on what
happens to the other photon.

>
> Anyway, there are two incompatible formulations involved here.

Exactly, that was Bell's point.

> What's wrong with Bell's formulation? It doesn't seem to be,
> strictly speaking, the assumption of locality -- since qm also
> uses the assumption of a common emitter as the basis for
> calculations.

As I said, the locality implies that we have to use the same \lambda at
A and at B.

>
> Footnotes:
>
> [1] I don't know how the timing stuff actually works in the
> coincidence circuitry. Like, how the coincidence interval is
> defined. How is everything ... matched up, etc.? I do know that
> it has to do with the atomic transition time, and the distance
> that the polarizers are from the emitters, but am not sure
> exactly what sequence of operations a detection triggers.

Experimentally, this is quite simple. A photomultiplier gives a
hit/no hit (pulse or no pulse) in a few nanoseconds. If the incoming
light flux is low enough, you have pulse rates of say, 100,000.- hits
per second. This means that it is rather improbable to have pulses
within a window of a few nanoseconds by chance (the probability is
~1/10000). So if you have a significant rate of coincidences, this
simply means that these photons have a common origin (come from a
2-photon state).
Experimentally, one has to choose an optimum between the spurious
coincidences (which augment when the lightflux increases) and the dark
current rate (the pulses a photodetector generates when no light falls
onto it), which is a nuisance at low lightflux.

>
> [2] My answer would be that if there's sufficient amplitude in the
> transmission by the polarizer, a, to generate a detection
> at A for some interval, then, for that interval, and for \theta = 0,
> then there will, with probability 1, be sufficient amplitude
> in the transmission by the polarizer, b, to generate a detection
> at B. And, following Malus' Law, the probability of detection at
> B will decrease as \theta increases from to \pi/2.

Be careful. I had such discussions before. There are 2 things you
have to keep in mind: 1) photodectectors are particle detectors
(imperfect as they may be), and not "intensity detectors" although
often these 2 descriptions coincide.
2) an entangled photon state cannot be described by a maxwell field.

cheers,
Patrick.

Ilja Schmelzer

Nov3-04, 10:05 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"seratend" <ser_monmail@yahoo.fr> schrieb\n> > I press a button by my free choice, and whenever I press it the\n> > light bulb changes its state. In this case, I conclude that there is\n> > a causal influence from my behaviour to the state of the light bulb.\n> >\n> > I make this conclusion based on the observation of some correlations.\n> > Of course, I also use some other assumptions: That there is some\n> > reality which defines the state of the light bulb, and certain\nassumptions\n> > about independence of my free choices from external influences.\n> >\n> > I believe that these basic assumptions are the foundation of the\n> > scientific method, that it is reasonable to hold them even in\n> > circumstances where no simple realistic explanation is known.\n>\n> Well, as a simple-minded person (in a non-pejorative sense :) in\n> physics, I am always afraid with arguments/words that are too far from\n> mathematics and too philosophical and too general such as the reality\n> one. Just because, when I want to make some "physical-mathematical"\n> logical deduction that use such a word/argument, I have to interpret\n> it (ie, I have to attach to this word at least a coherent mathematical\n> representation) and thus I risk to choose the definition that arrange\n> my logical deduction.\n\nEPR and Bell are sufficiently formal versions of "realism" to work with.\n\n> The EPR 1935 paper is the example of a definition of this "reality":\n> They start with this sentence: "The elements of the physical reality\n> cannot be determined by a priori philosophical considerations, but\n> must be found by an appeal to results of experiments and\n> measurements." At least they agree that it seems to be a need for a\n> kind of formal logical deduction.\n\n"Appeal" is not a kind of formal logical deduction.\n\n> In addition, they follow by their own definition of reality "if,\n> without in any way disturbing a system, we can predict with certainty\n> (i.e. with probability equal to unity) the value of a physical\n> quantity, then there exists an element of physical reality\n> corresponding to this physical quantity."\n\nThat\'s the EPR criterion. I have proposed a modification of this\ncriterion:\n\nIf we observe an exact correlation between A and B, then the\nfollowing three possibilities may be considered as an explanation:\n\n1. Some element of reality C which acts as a common cause\nC->A, C->B;\n2. A real causal influence A->B;\n3. A real causal influence B->A.\n\nThus, the criterion allows to make a conclusion from observation\n(the correlation) to reality (each of the alternatives introduces some\nelement of reality). The connection is quite simple. The fuzzy\n"without in any way" is replaced by "without any real causal\ninfluence". The reference to ordering in time (implicit in "predict")\nis removed, instead the possibility of the reverse causal influence\nhas been included. That\'s essentially all.\n\n> Therefore, we may see that even this earlier definition of physical\n> reality is very fuzzy and in fact is selected, a priori, for their\n> paper argumentation. Remark that this definition does not exclude a\n> globally defined physical quantity (i.e. a reality is not assumed\n> local by this definition).\n\nIndeed. My modification has the aim to remove the "selected for their\npaper" parts, and, especially, to clarify that realism does not need any\nreference to any spacetime. It is a much more general principle.\n\nBell\'s definition of realism (which is implicit in the proof of his\ninequality,\nIOW what is necessary to prove his inequality except Einstein causality)\nis also general enough but even more formal. You have decisions of\nexperimenters x, observables X, and a realistic explanation is some\nspace of possible realities L with a probability measure rho(l) and\nsome functions X(x,l) which define the results in dependence on l and\nx:\n\n<X> = int X(x,l) rho(l) d l\n\nQuite formal, not?\n\n> If I go back to this FTL phone, I would like to underline the dangers\n> of using black boxes where there concepts are too fuzzy (not directly\n> connected to a "physico-mathematical formulation", a very fuzzy term\n> ;).\n\nI agree. But realism is not fuzzy at all.\n\n> > I see, I have not sufficiently explained that it is based on Popper\'s\n> > falsifiability of scientific, physical theory. To distinghish science\nfrom\n> > pseudosciences Popper has proposed that the distinguishing\n> > property of physical theories is that they make nontrivial predictions,\n> > predictions which may appear false if we test them. Such theories\n> > are called falsifiable. Theories which cannot be falsified are not\n> > empirical theories, they are metaphysical. This point of view on science\n> > is IMHO correct and widely accepted among scientists.\n>\n> I understand it and I agree partly with this statement. The "partly"\n> is due to the implicit hypothesis that the theories may be falsified.\n> The main problem with modern theories (modern in the sense the few\n> ones I know) is that they are constructed to support everything. It is\n> always the external assumptions, sometimes implicit, that restricts\n> them and thus may be falsified:\n> e.g Newtons\' law: ma=force: if I authorise any force, I may have any\n> acceleration, and any x(t) if I enlarge the derivation.\n\nOk, but this problem was known by Popper. It is never a particular\nstatement of a theory which is falsified but the whole theory. And even\nnot the whole theory but only a set of several theories is falsified by a\ngiven observation. (This includes, for example, the theories "experimenters\nand publishers are sufficiently honest people" and so on.) You don\'t\nfalsify f=ma but f=ma + certain assumptions about f = grad V, the\nformula for V as dependent on the positions and masses, the particular\ninitial distribution of the masses in the particular situation and so on.\n\n> > My FTL phone is only a hypothetical feature. [...]\n>\n> The only difficult point with the FTL phone argumentation resides in\n> its "fuzzy nature" that does not allow one to make reasonable logical\n> deductions based on a logical model. The main point I currently\n> identify with this fuzzy model is its possible lack of\n> "reality/locality" (my reality/locality : ), because you just removed\n> too many properties of your black box.\n> E.g. your sentence "I press a button by my free choice", so if I\n> assume that the button is a local one (an extra hypothesis to your\n> model).\n\nI think, instead, that this is the power of my FTL phone argument.\nIt shows that some types of fuzzy argumentation are that fuzzy that\nthey allow to explain away even a working FTL phone. Considering\narguments which work with this type of fuzziness, you should think\nabout the possibility of explaining away the obvious.\n\nLast not least, we know that there are always non-contradictionary\nbut nonsensical theories like consequent solipcism or various\nfundamentalistic\nversions of religions. That means, it is never possible to _prove_\nsomething\nin science in full beauty. Certain types of doubt should be rejected to\nbe able to do science.\n\nIn my phone argument, "to be able to do science" is specified into\n"to be able to reject Einstein causality if confronted with a working\nFTL phone". If we cannot do the last, Einstein causality is no longer\nscientific.\n\n> I may also suppose that your FTL model is based on local interactions\n> (the "press" interaction). With this assumption and mathematical\n> "causality" (the logical "=>" symbol), I think I can deduce that this\n> model works if there exists an FTL interaction on at least a frame.\n> If it is your assumption, I prefer to start with the possibility of\n> the existence of an interaction with an infinite speed in at least one\n> frame: this model is much well defined (at a logical point of view),\n> than a FTL phone. I may also start with a non-local interaction model\n> (the button is a universe button that I can press locally).\n\nFine, no problem. But if there is an FTL phone, it is meaningless to\ndefend Einstein causality. If the causal connection works via local\ninteractions with speed >> c or via global variables is not the point.\nThe point is that Einstein causality is dead, because Einstein causality\nforbids FTL phones.\n\nAs well, I have no problem if you want explain violations of BI via\nspeed >> c or via global variables. The argument is that, whatever,\nEinstein causality is already dead. Because it forbids, together with\nrealism, violations of BI.\n\n> > I think about causality on a more fundamental level. The observation\n> > of a correlation between A and B is sufficient to search for a realistic\n> > causal explanation. Such explanations are a common cause\n> > C->A, C->B or a direct causal influence A->B or B->A. Time has\n> > not been mentioned in this consideration.\n>\n> This is the problem, it seems to be too fundamental. You are removing\n> too many things. At least to realise an experiment (i.e. If you want a\n> theory that may be falsified), you need time, at least in one step of\n> the experiment, to make a "causal" interpretation.\n\nNo. I can, for example, consider observations of correlations between\ndocuments in the American embassy in Moscow and in the KGB centr.\nBased on these correlations I can make conclusions about some types\nof causal interaction. Some restrictions in time (all this happens in the\nperiod of cold war) are part of this, and, of course, the realistic\nexplanations should not contradict the time ordering, but nontrivial\nconclusions are possible even if the uncertainties of related time\nmeasurement are very large.\n\n> > Now, Einstein causality is, in the general understanding, a physical\n> > theory. It predicts that no FTL phones are possible. At least, this is\n> > the usual interpretation of Einstein causality.\n>\n> I will say that Einstein causality is an extra postulate. And, I am\n> afraid of the multiple interpretations we give to these words\n> "Einstein causality".\n\nI see no reason to be afraid. I see only two important\ninterpretations, the realistic one and the one based on\nforbidding FTL phones (directly observable effects) only.\n\n> For example with the SR, we may deduce a causality interpretation as\n> that "reality" belongs to the light cone of the space-time (tb>ta\n> whatever SR frame we select) and from the updated ma=force equation.\n> I do not even speak about GR, where you need to add the locality to\n> define a local causality.\n>\n> In QFT, the causality is interpreted in a slightly different sense:\n> whether a measurement performed at one point can affect a measurement\n> at another point whose separation from the first is spacelike\n> (Peskin-Schroeder QFT reference).\n> There are other postulates such as Weinberg cluster decomposition that\n> defines its own causality (weaker), and in fine, it gives almost the\n> same "results".\n\nRoughly speaking, the classical (SR and GR) definitions are realistic\ndefinitions, which quantum definitions are not. The other technical\ndifferences\ndo not seem to be very important.\n\n> > What would happen if we would really observe some FTL phone?\n> > Very simple, Einstein causality would be falsified. Nobody would\n> > defend relativity in this case. It would survive in physics as a limit\n> > of some other theory which allows FTL phones.\n> >\n> > This is not the case today. Using the strange correlations of EPR\n> > and Bell we cannot build FTL phones. This is a simple theorem.\n>\n> Well, I should first know what a FTL phone is.\n\nA black box which looks like a phone and works like a phone\nbut allows phone calls to Mars without observable time delay.\n(or time delay much smaller than the one predicted by c as the\nmaximal speed).\n\nIt is your choice how to test this black box. You should explain\nhow you propose to test that the black box really violates Einstein\ncausality without opening it. Once you have established, by\nobservation, that, indeed, it violates, then you should be able\nto reject various solipsistic criticism which, nonetheless, tries to\nsave Einstein causality.\n\n> For example if the\n> phenomenon is an interaction FTL I would rather first keep relativity\n> but say that maybe "reality" also exists in space like intervals (the\n> famous "tachyons" or whatever else). After, I will try to make a\n> working model (with experiments). After all, I will see what type of\n> causality is broken (if it is really broken).\n\nFeel free to propose the experiments, I will tell you what happens in\neach of your proposed experiments. (You can do it yourself, based on\na preferred frame with a new interaction with infinite speed.)\n\n> > Nonetheless, that does not mean there is no problem for relativity.\n> > The only realistic explanations of these correlations are A->B or\n> > B->A, above are in contradiction with relativity. Now, defenders\n> > of relativity use various argumentations. But, as I have observed,\n> > could be applied as well in the hypothetical case of falsification of\n> > relativity. That means, if we take them seriously, we can no longer\n> > predict that no FTL phone exists - because such an FTL phone\n> > could be explained away if we observe it.\n> >\n> This is the main problem with general arguments. You can adapt them as\n> a theory to what you want.\n\nThat\'s not a problem but their main advantage.\n\n> I agree, and I think every physicist, that SR or GR have a limited\n> domain of application. The main problem is that the mathematical\n> formulation of these theories as well a QM does not really need\n> causality. It is always an external restriction in order to say this\n> is the reality (i.e. to recover the experiments).\n\nIn some sense, the problems with formulation of causality in\ntime-symmetric theories indicate that there is something wrong with\nthem. Of course, may be there is instead something wrong with\ncausality.\n\n> > > So, Do you intend "->" (your causality of the FTL phone) as the\n> > > mathematical logic symbol "=>" ?\n> >\n> > No. In mathematical terms it means absence of independence.\n>\n> Whow, what a very weak and fuzzy definition.\n\nCertainly not fuzzy. For observables P(AB) = P(A)P(B) means\nindependence. Everything else is a nontrivial correlation which\nrequires a nontrivial realistic explanation.\n\nBut I see I was not precise. "->" is part of the realistic explanation.\nIt means a certain, more specific, direct dependence. C->A, C->B\ncauses a correlation between A and B but does not give A->B or\nB->A.\n\nIn this sense, A->B denotes some fundamental, basic dependence\nas part of some ontological theory. Thus, the theory specifies\nwhat exists (ontology) and what depends on what (causality, usually\ndefined by laws of evolution).\n\n> I thus interpret the "absence of independence" of "->" by the\n> existence of an application f (we can instead choose a relation, but\n> it gives no more information and may be more difficult to interpret)\n> such as we have:\n>\n> B -> A <=> A= f(B, other points if we want= context).\n\nI would say that\'s not all. A->B is a relation, it is transitive,\nand an ordering: A->B and B->A means A=B. There are no\nclosed causal loops except trivial ones A->A.\n\nThe relation has to be defined, postulated, by the realistic causal\ntheory, not derived from observation. It is connected with\nobservation via the EPR principle:\n\nNot A->B and not B->A and not exists C C->B C->A\n=> P(AB)=P(A)P(B).\n\nThis is an exact quantitative prediction about a quite large class\nof observables, that means, a prediction with large empirical content.\n\n> > Causality gives the additional restriction\n> >\n> > Result_at_A = f_a(reality, decisions_at_A)\n> > Result_at_B = f_b(reality, decisions_at_B)\n\n> You are introducing three new notions: decisions, reality and results\n> that are difficult to map into a physical theory.\n\nNot that difficult. Decisions are the same as parameters of\nstate preparation in standard minimal QM, results the same as\nobservables in minimal QM. Reality is some element of\nsome set of possible realities (using set theory) with\nclassical probability measure rho. This is part of the general scheme\nof ontological theories: Each theory worth to be named realistic\n(ontological, having an ontology) has such a set. There is no other\nrestriction on this set.\n\n> But, your are giving\n> some details of some of your assumptions (in the current hypothesis:\n> the FTL phone is described by an unknown physics).\n>\n> First what is the Result_at_A? Because you have not defined the\n> space-time object, I assume result_at_A is not obliged to be at a\n> point in space: it is only attached to A, to be as general as\n> possible.\n\nDoesn\'t matter. If some A ist only weakly attached, chose some\nother consequence which is more attached: Some macroscopic\npointer, some written paper, some memory of some person with\nconsciousness - whatever you accept as being localized. Once the\nresult is localized in one meaning, it is easy to obtain a replacement\nlocalized in some other meaning of localization.\n\n> I have defined a set of points containing A and B, . (the universe)\n> And your text above says that A and B are also sets [A=set of (all\n> possible results), B= set of (all possible results)]. (here I add a\n> new postulate: all possible results are a set).\n\nThis is already part of realism.\n\n> So your are building applications to these sets (f_a for the set A,\n> f_b for the set B): ok.\n> Now, what represents the inputs of these applications: decisions_at_A,\n> decision_at_B ?\n\nIn a similar sense, these are results localized at A and B, they may be\nmemorized, and, moreover, we assume that they do not depend in a\ncausal way on external influences. Something like free will decisions,\nideal independent random devices, or combinations of various such\nthings.\n\n> Does it exist any relation between decisions and\n> results in a large sense?\n\nDecisions are some special types of results.\n\n> I do not understand your "reality":\n> -- We may take some reality aspects as the domain of validity of the\n> functions f_a and f_b . (some values of\n> (decisions_at_A,decisions_at_B) does not belong to the reality).\n> -- reality may also be the other decisions_at_other_objects or\n> results_at_other_objects (the additional context).\n> -- is it an object of the universe (the set containing the points A,B\n> ...)\n> -- ???\n\nA reality l is an element of the set of possible realities L. Think about\nit as the set of all (hidden and observable) variables. Results and\ndecisions are observables. Elements of reality should not be observable.\n\nThe particular set L, the probability distribution rho on it and the\nparticular functions f_A should be postulated by a proposed\nrealistic explanation (realistic theory).\n\n> > Causality gives the additional restriction\n> >\n> > Result_at_A = f_a(reality, decisions_at_A)\n> > Result_at_B = f_b(reality, decisions_at_B)\n\n> Now with this definition and the previous assumptions, I think here\n> you are mixing causality with locality (I still have questions about\n> what is "reality" and "decisions_at_object"). Even if we don\'t have\n> specified what we intend by locality, I don\'t think that we can say\n> that "causality" => "locality" without additional assumptions (e.g. QM\n> can be said causal but not local to some extent).\n\nQM cannot be said to be realistic in the sense of EPR, therefore it does\nnot make much sense to talk about locality and causality. At least,\nQM-compatible definitions of causality and locality should not be mixed\nwith definitions of locality and causality appropriate for realistic\ntheories.\n\nIn a realistic theory, I would name it local if L = some set of functions\non some manifold X named space. This would allow FTL, therefore,\nhas nothing to do with Einstein causality.\n\n> > > Now, to end, I will try to show a possible implementation example of a\n> > > formal FTL phone based solely on your "causal" argument.\n> > >\n> > > Let\'s take 2 classical independent particles (no interaction) with a\n> > > total kinetic energy E= E1+E2.\n> > > Now if we change the kinetic energy of particle 1, we thus induce\n> > > instantaneously (causality) a change on the kinetic energy of the 2\n> > > particles (E): We have formally a FTL phone with the total Energy E.\n> >\n> > That\'s not a phone because we cannot use it to talk with each other,\n> > nor formal nor informal.\n>\n> Why not?\n> Formally, we have first to define what is to talk to each other (I do\n> not involve necessarily human objects, nor local objects: it is my\n> freedom choice with your weak model).\n\nThe FTL phone is not a weak model. It looks like a phone and works\nlike a phone. Without time delay. You cannot create such a device\nusing your consideration about total energy. (This is a falsfiable claim.\nFeel free to falsify this claim by presenting me such a phone.)\n\n> I think you are implicitly restricting the FTL phone model to local\n> variables and thus you are mixing causality with locality.\n\nThe results and decisions which we can use to test if the phone works\nhave local character. What is inside the black box remains open.\n\n> It is what I want to underline with this FTL phone fuzzy model: we do\n> not specify enough to make reasonable logical deductions (i.e. all the\n> deductions have many implicit assumptions).\n\nThat\'s the point. There is no experiment which gives sufficient information\nin itself to make strong logical deductions. To do science, we need\nadditional hypotheses. Based on these hypotheses, we can make\nstrong logical conclusions which, then, may be falsified.\n\nThe FTL phone is a hypothetical device which is sufficient to reject\nEinstein causality. In the optimal way. There is no better way to reject\nEinstein causality by observation.\n\nWhatever you miss in the FTL phone will be always missed, in any\nimaginable empirical falsification of Einstein causality. Thus, you have\nonly two choices: Don\'t care about the missed parts, accepting some\nhypothesis which excludes the loophole, or give up science in the usual\nway, the way of empirical falsification of theories.\n\n> I also say that a QM state is not interpreted into an obvious\n> kolgomorovian probabilistic event (whatever probabilistic set and\n> sigma-algebra is chosen).\n\nI would say a Bohmian interpretation is such a (quite obvious)\nKolmogorovian description.\n\n> We just can say that a classical\n> statistical hidden variable model, without an hidden FTL channel is\n> not able to represent the statistical results given by this state.\n\nWe can do more. We can decide that only the thing named\n"statistical hidden variable model" is worth to be considered\nas a scientific explanation, and once we have found such\nexplanations (like those given by Bohmian mechanics) everything\nwhich gives less (like quantum mechanics) should be rejected.\n\nThis is a methodological decision, a decision about the nature of\nthe scientific method.\n\n> Finally, for the ones interested by a hidden variable model of EPR\n> state with a hidden FTL channel, I recommend the last Cerf, Gisisn\n> Massar and Popescu quant-ph/0410027 paper (4 pages - a short concise\n> one). I think it is a good one (with the pointers it gives).\n\nThanks. I will have a look.\n\n> It introduces a possible (logical) implementation of what they call\n> the non-local machines (PR- machines, I like the "machine" term they\n> chosen rather than "interaction"). We have another different look to\n> the EPR like states.\n>\n> I would like to know if you would you say that this model is causal or\n> not and why?\n\nSimple test question: If it is causal and has FTL causal influences, it\nneeds\na preferred frame (or, for some higher limiting speed C>>c, a class of\npreferred frames). If yes, I see no reason to doubt that it is causal.\nIf not, it is not realistic or not causal in the sense of the definitions I\nhave\ngiven.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"seratend" <ser_monmail@yahoo.fr> schrieb
> > I press a button by my free choice, and whenever I press it the
> > light bulb changes its state. In this case, I conclude that there is
> > a causal influence from my behaviour to the state of the light bulb.
> >
> > I make this conclusion based on the observation of some correlations.
> > Of course, I also use some other assumptions: That there is some
> > reality which defines the state of the light bulb, and certain
assumptions
> > about independence of my free choices from external influences.
> >
> > I believe that these basic assumptions are the foundation of the
> > scientific method, that it is reasonable to hold them even in
> > circumstances where no simple realistic explanation is known.
>
> Well, as a simple-minded person (in a non-pejorative sense :) in
> physics, I am always afraid with arguments/words that are too far from
> mathematics and too philosophical and too general such as the reality
> one. Just because, when I want to make some "physical-mathematical"
> logical deduction that use such a word/argument, I have to interpret
> it (ie, I have to attach to this word at least a coherent mathematical
> representation) and thus I risk to choose the definition that arrange
> my logical deduction.

EPR and Bell are sufficiently formal versions of "realism" to work with.

> The EPR 1935 paper is the example of a definition of this "reality":
> They start with this sentence: "The elements of the physical reality
> cannot be determined by a priori philosophical considerations, but
> must be found by an appeal to results of experiments and
> measurements." At least they agree that it seems to be a need for a
> kind of formal logical deduction.

"Appeal" is not a kind of formal logical deduction.

> In addition, they follow by their own definition of reality "if,
> without in any way disturbing a system, we can predict with certainty
> (i.e. with probability equal to unity) the value of a physical
> quantity, then there exists an element of physical reality
> corresponding to this physical quantity."

That's the EPR criterion. I have proposed a modification of this
criterion:

If we observe an exact correlation between A and B, then the
following three possibilities may be considered as an explanation:

1. Some element of reality C which acts as a common cause
C->A, C->B;
2. A real causal influence A->B;
3. A real causal influence B->A.

Thus, the criterion allows to make a conclusion from observation
(the correlation) to reality (each of the alternatives introduces some
element of reality). The connection is quite simple. The fuzzy
"without in any way" is replaced by "without any real causal
influence". The reference to ordering in time (implicit in "predict")
is removed, instead the possibility of the reverse causal influence
has been included. That's essentially all.

> Therefore, we may see that even this earlier definition of physical
> reality is very fuzzy and in fact is selected, a priori, for their
> paper argumentation. Remark that this definition does not exclude a
> globally defined physical quantity (i.e. a reality is not assumed
> local by this definition).

Indeed. My modification has the aim to remove the "selected for their
paper" parts, and, especially, to clarify that realism does not need any
reference to any spacetime. It is a much more general principle.

Bell's definition of realism (which is implicit in the proof of his
inequality,
IOW what is necessary to prove his inequality except Einstein causality)
is also general enough but even more formal. You have decisions of
experimenters x, observables X, and a realistic explanation is some
space of possible realities L with a probability measure \rho(l) and
some functions X(x,l) which define the results in dependence on l and
x:

<X> = \int X(x,l) \rho(l) d l

Quite formal, not?

> If I go back to this FTL phone, I would like to underline the dangers
> of using black boxes where there concepts are too fuzzy (not directly
> connected to a "physico-mathematical formulation", a very fuzzy term
> ;).

I agree. But realism is not fuzzy at all.

> > I see, I have not sufficiently explained that it is based on Popper's
> > falsifiability of scientific, physical theory. To distinghish science
from
> > pseudosciences Popper has proposed that the distinguishing
> > property of physical theories is that they make nontrivial predictions,
> > predictions which may appear false if we test them. Such theories
> > are called falsifiable. Theories which cannot be falsified are not
> > empirical theories, they are metaphysical. This point of view on science
> > is IMHO correct and widely accepted among scientists.
>
> I understand it and I agree partly with this statement. The "partly"
> is due to the implicit hypothesis that the theories may be falsified.
> The main problem with modern theories (modern in the sense the few
> ones I know) is that they are constructed to support everything. It is
> always the external assumptions, sometimes implicit, that restricts
> them and thus may be falsified:
> e.g Newtons' law: ma=force: if I authorise any force, I may have any
> acceleration, and any x(t) if I enlarge the derivation.

Ok, but this problem was known by Popper. It is never a particular
statement of a theory which is falsified but the whole theory. And even
not the whole theory but only a set of several theories is falsified by a
given observation. (This includes, for example, the theories "experimenters
and publishers are sufficiently honest people" and so on.) You don't
falsify f=ma but f=ma + certain assumptions about f = grad V, the
formula for V as dependent on the positions and masses, the particular
initial distribution of the masses in the particular situation and so on.

> > My FTL phone is only a hypothetical feature. [...]
>
> The only difficult point with the FTL phone argumentation resides in
> its "fuzzy nature" that does not allow one to make reasonable logical
> deductions based on a logical model. The main point I currently
> identify with this fuzzy model is its possible lack of
> "reality/locality" (my reality/locality : ), because you just removed
> too many properties of your black box.
> E.g. your sentence "I press a button by my free choice", so if I
> assume that the button is a local one (an extra hypothesis to your
> model).

I think, instead, that this is the power of my FTL phone argument.
It shows that some types of fuzzy argumentation are that fuzzy that
they allow to explain away even a working FTL phone. Considering
arguments which work with this type of fuzziness, you should think
about the possibility of explaining away the obvious.

Last not least, we know that there are always non-contradictionary
but nonsensical theories like consequent solipcism or various
fundamentalistic
versions of religions. That means, it is never possible to _prove_
something
in science in full beauty. Certain types of doubt should be rejected to
be able to do science.

In my phone argument, "to be able to do science" is specified into
"to be able to reject Einstein causality if confronted with a working
FTL phone". If we cannot do the last, Einstein causality is no longer
scientific.

> I may also suppose that your FTL model is based on local interactions
> (the "press" interaction). With this assumption and mathematical
> "causality" (the logical "=>" symbol), I think I can deduce that this
> model works if there exists an FTL interaction on at least a frame.
> If it is your assumption, I prefer to start with the possibility of
> the existence of an interaction with an infinite speed in at least one
> frame: this model is much well defined (at a logical point of view),
> than a FTL phone. I may also start with a non-local interaction model
> (the button is a universe button that I can press locally).

Fine, no problem. But if there is an FTL phone, it is meaningless to
defend Einstein causality. If the causal connection works via local
interactions with speed >> c or via global variables is not the point.
The point is that Einstein causality is dead, because Einstein causality
forbids FTL phones.

As well, I have no problem if you want explain violations of BI via
speed >> c or via global variables. The argument is that, whatever,
Einstein causality is already dead. Because it forbids, together with
realism, violations of BI.

> > I think about causality on a more fundamental level. The observation
> > of a correlation between A and B is sufficient to search for a realistic
> > causal explanation. Such explanations are a common cause
> > C->A, C->B or a direct causal influence A->B or B->A. Time has
> > not been mentioned in this consideration.
>
> This is the problem, it seems to be too fundamental. You are removing
> too many things. At least to realise an experiment (i.e. If you want a
> theory that may be falsified), you need time, at least in one step of
> the experiment, to make a "causal" interpretation.

No. I can, for example, consider observations of correlations between
documents in the American embassy in Moscow and in the KGB centr.
Based on these correlations I can make conclusions about some types
of causal interaction. Some restrictions in time (all this happens in the
period of cold war) are part of this, and, of course, the realistic
explanations should not contradict the time ordering, but nontrivial
conclusions are possible even if the uncertainties of related time
measurement are very large.

> > Now, Einstein causality is, in the general understanding, a physical
> > theory. It predicts that no FTL phones are possible. At least, this is
> > the usual interpretation of Einstein causality.
>
> I will say that Einstein causality is an extra postulate. And, I am
> afraid of the multiple interpretations we give to these words
> "Einstein causality".

I see no reason to be afraid. I see only two important
interpretations, the realistic one and the one based on
forbidding FTL phones (directly observable effects) only.

> For example with the SR, we may deduce a causality interpretation as
> that "reality" belongs to the light cone of the space-time (tb>ta
> whatever SR frame we select) and from the updated ma=force equation.
> I do not even speak about GR, where you need to add the locality to
> define a local causality.
>
> In QFT, the causality is interpreted in a slightly different sense:
> whether a measurement performed at one point can affect a measurement
> at another point whose separation from the first is spacelike
> (Peskin-Schroeder QFT reference).
> There are other postulates such as Weinberg cluster decomposition that
> defines its own causality (weaker), and in fine, it gives almost the
> same "results".

Roughly speaking, the classical (SR and GR) definitions are realistic
definitions, which quantum definitions are not. The other technical
differences
do not seem to be very important.

> > What would happen if we would really observe some FTL phone?
> > Very simple, Einstein causality would be falsified. Nobody would
> > defend relativity in this case. It would survive in physics as a limit
> > of some other theory which allows FTL phones.
> >
> > This is not the case today. Using the strange correlations of EPR
> > and Bell we cannot build FTL phones. This is a simple theorem.
>
> Well, I should first know what a FTL phone is.

A black box which looks like a phone and works like a phone
but allows phone calls to Mars without observable time delay.
(or time delay much smaller than the one predicted by c as the
maximal speed).

It is your choice how to test this black box. You should explain
how you propose to test that the black box really violates Einstein
causality without opening it. Once you have established, by
observation, that, indeed, it violates, then you should be able
to reject various solipsistic criticism which, nonetheless, tries to
save Einstein causality.

> For example if the
> phenomenon is an interaction FTL I would rather first keep relativity
> but say that maybe "reality" also exists in space like intervals (the
> famous "tachyons" or whatever else). After, I will try to make a
> working model (with experiments). After all, I will see what type of
> causality is broken (if it is really broken).

Feel free to propose the experiments, I will tell you what happens in
each of your proposed experiments. (You can do it yourself, based on
a preferred frame with a new interaction with infinite speed.)

> > Nonetheless, that does not mean there is no problem for relativity.
> > The only realistic explanations of these correlations are A->B or
> > B->A, above are in contradiction with relativity. Now, defenders
> > of relativity use various argumentations. But, as I have observed,
> > could be applied as well in the hypothetical case of falsification of
> > relativity. That means, if we take them seriously, we can no longer
> > predict that no FTL phone exists - because such an FTL phone
> > could be explained away if we observe it.
> >
> This is the main problem with general arguments. You can adapt them as
> a theory to what you want.

That's not a problem but their main advantage.

> I agree, and I think every physicist, that SR or GR have a limited
> domain of application. The main problem is that the mathematical
> formulation of these theories as well a QM does not really need
> causality. It is always an external restriction in order to say this
> is the reality (i.e. to recover the experiments).

In some sense, the problems with formulation of causality in
time-symmetric theories indicate that there is something wrong with
them. Of course, may be there is instead something wrong with
causality.

> > > So, Do you intend "->" (your causality of the FTL phone) as the
> > > mathematical logic symbol "=>" ?
> >
> > No. In mathematical terms it means absence of independence.
>
> Whow, what a very weak and fuzzy definition.

Certainly not fuzzy. For observables P(AB) = P(A)P(B) means
independence. Everything else is a nontrivial correlation which
requires a nontrivial realistic explanation.

But I see I was not precise. "->" is part of the realistic explanation.
It means a certain, more specific, direct dependence. C->A, C->B
causes a correlation between A and B but does not give A->B or
B->A.

In this sense, A->B denotes some fundamental, basic dependence
as part of some ontological theory. Thus, the theory specifies
what exists (ontology) and what depends on what (causality, usually
defined by laws of evolution).

> I thus interpret the "absence of independence" of "->" by the
> existence of an application f (we can instead choose a relation, but
> it gives no more information and may be more difficult to interpret)
> such as we have:
>
> B -> A <=> A= f(B, other points if we want= context).

I would say that's not all. A->B is a relation, it is transitive,
and an ordering: A->B and B->A means A=B. There are no
closed causal loops except trivial ones A->A.

The relation has to be defined, postulated, by the realistic causal
theory, not derived from observation. It is connected with
observation via the EPR principle:

Not A->B and not B->A and not exists C C->B C->A=> P(AB)=P(A)P(B).

This is an exact quantitative prediction about a quite large class
of observables, that means, a prediction with large empirical content.

> > Causality gives the additional restriction
> >
> > Result_at_A = f_a(reality, decisions_at_A)
> > Result_at_B = f_b(reality, decisions_at_B)

> You are introducing three new notions: decisions, reality and results
> that are difficult to map into a physical theory.

Not that difficult. Decisions are the same as parameters of
state preparation in standard minimal QM, results the same as
observables in minimal QM. Reality is some element of
some set of possible realities (using set theory) with
classical probability measure \rho. This is part of the general scheme
of ontological theories: Each theory worth to be named realistic
(ontological, having an ontology) has such a set. There is no other
restriction on this set.

> But, your are giving
> some details of some of your assumptions (in the current hypothesis:
> the FTL phone is described by an unknown physics).
>
> First what is the Result_at_A? Because you have not defined the
> space-time object, I assume result_at_A is not obliged to be at a
> point in space: it is only attached to A, to be as general as
> possible.

Doesn't matter. If some A ist only weakly attached, chose some
other consequence which is more attached: Some macroscopic
pointer, some written paper, some memory of some person with
consciousness - whatever you accept as being localized. Once the
result is localized in one meaning, it is easy to obtain a replacement
localized in some other meaning of localization.

> I have defined a set of points containing A and B, . (the universe)
> And your text above says that A and B are also sets [A=set of (all
> possible results), B= set of (all possible results)]. (here I add a
> new postulate: all possible results are a set).

This is already part of realism.

> So your are building applications to these sets (f_a for the set A,
> f_b for the set B): ok.
> Now, what represents the inputs of these applications: decisions_at_A,
> decision_at_B ?

In a similar sense, these are results localized at A and B, they may be
memorized, and, moreover, we assume that they do not depend in a
causal way on external influences. Something like free will decisions,
ideal independent random devices, or combinations of various such
things.

> Does it exist any relation between decisions and
> results in a large sense?

Decisions are some special types of results.

> I do not understand your "reality":
> -- We may take some reality aspects as the domain of validity of the
> functions f_a and f_b . (some values of
> (decisions_at_A,decisions_at_B) does not belong to the reality).
> -- reality may also be the other decisions_at_other_objects or
> results_at_other_objects (the additional context).
> -- is it an object of the universe (the set containing the points A,B
> ...)
> -- ???

A reality l is an element of the set of possible realities L. Think about
it as the set of all (hidden and observable) variables. Results and
decisions are observables. Elements of reality should not be observable.

The particular set L, the probability distribution \rho on it and the
particular functions f_A should be postulated by a proposed
realistic explanation (realistic theory).

> > Causality gives the additional restriction
> >
> > Result_at_A = f_a(reality, decisions_at_A)
> > Result_at_B = f_b(reality, decisions_at_B)

> Now with this definition and the previous assumptions, I think here
> you are mixing causality with locality (I still have questions about
> what is "reality" and "decisions_at_object"). Even if we don't have
> specified what we intend by locality, I don't think that we can say
> that "causality" => "locality" without additional assumptions (e.g. QM
> can be said causal but not local to some extent).

QM cannot be said to be realistic in the sense of EPR, therefore it does
not make much sense to talk about locality and causality. At least,
QM-compatible definitions of causality and locality should not be mixed
with definitions of locality and causality appropriate for realistic
theories.

In a realistic theory, I would name it local if L = some set of functions
on some manifold X named space. This would allow FTL, therefore,
has nothing to do with Einstein causality.

> > > Now, to end, I will try to show a possible implementation example of a
> > > formal FTL phone based solely on your "causal" argument.
> > >
> > > Let's take 2 classical independent particles (no interaction) with a
> > > total kinetic energy E= E1+E2.
> > > Now if we change the kinetic energy of particle 1, we thus induce
> > > instantaneously (causality) a change on the kinetic energy of the 2
> > > particles (E): We have formally a FTL phone with the total Energy E.
> >
> > That's not a phone because we cannot use it to talk with each other,
> > nor formal nor informal.
>
> Why not?
> Formally, we have first to define what is to talk to each other (I do
> not involve necessarily human objects, nor local objects: it is my
> freedom choice with your weak model).

The FTL phone is not a weak model. It looks like a phone and works
like a phone. Without time delay. You cannot create such a device
using your consideration about total energy. (This is a falsfiable claim.
Feel free to falsify this claim by presenting me such a phone.)

> I think you are implicitly restricting the FTL phone model to local
> variables and thus you are mixing causality with locality.

The results and decisions which we can use to test if the phone works
have local character. What is inside the black box remains open.

> It is what I want to underline with this FTL phone fuzzy model: we do
> not specify enough to make reasonable logical deductions (i.e. all the
> deductions have many implicit assumptions).

That's the point. There is no experiment which gives sufficient information
in itself to make strong logical deductions. To do science, we need
additional hypotheses. Based on these hypotheses, we can make
strong logical conclusions which, then, may be falsified.

The FTL phone is a hypothetical device which is sufficient to reject
Einstein causality. In the optimal way. There is no better way to reject
Einstein causality by observation.

Whatever you miss in the FTL phone will be always missed, in any
imaginable empirical falsification of Einstein causality. Thus, you have
only two choices: Don't care about the missed parts, accepting some
hypothesis which excludes the loophole, or give up science in the usual
way, the way of empirical falsification of theories.

> I also say that a QM state is not interpreted into an obvious
> kolgomorovian probabilistic event (whatever probabilistic set and
> \sigma-algebra is chosen).

I would say a Bohmian interpretation is such a (quite obvious)
Kolmogorovian description.

> We just can say that a classical
> statistical hidden variable model, without an hidden FTL channel is
> not able to represent the statistical results given by this state.

We can do more. We can decide that only the thing named
"statistical hidden variable model" is worth to be considered
as a scientific explanation, and once we have found such
explanations (like those given by Bohmian mechanics) everything
which gives less (like quantum mechanics) should be rejected.

This is a methodological decision, a decision about the nature of
the scientific method.

> Finally, for the ones interested by a hidden variable model of EPR
> state with a hidden FTL channel, I recommend the last Cerf, Gisisn
> Massar and Popescu http://www.arxiv.org/abs/quant-ph/0410027 paper (4 pages - a short concise
> one). I think it is a good one (with the pointers it gives).

Thanks. I will have a look.

> It introduces a possible (logical) implementation of what they call
> the non-local machines (PR- machines, I like the "machine" term they
> chosen rather than "interaction"). We have another different look to
> the EPR like states.
>
> I would like to know if you would you say that this model is causal or
> not and why?

Simple test question: If it is causal and has FTL causal influences, it
needs
a preferred frame (or, for some higher limiting speed C>>c, a class of
preferred frames). If yes, I see no reason to doubt that it is causal.
If not, it is not realistic or not causal in the sense of the definitions I
have
given.

Ilja

Ilja Schmelzer

Nov5-04, 06:10 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Thomas Trotter" <thomastrotter2005@juno.com> schrieb\n> > > Thomas Trotter (TT)\n> > > Maybe it\'s just that knowledge of what is\n> > > actually happening wrt the emission and detection of the opposite\n> > > moving light beams is incomplete (ie., that there *is* a local,\n> > > common-property reason for the correlations, and not enough is\n> > > known about the physical process to talk about it in a straighforward\n> > > local realistic way yet).\n>\n> Ilja Schmelzer (IS):\n> > Given Bell\'s theorem, there is no such explanation. This is independent\n> > of our state of knowledge or state of our theories.\n>\n> TT:\n> Then, you are saying that the correlation curve has\n> nothing to do with the properties of the light incident\n> on the polarizers during a given coincidence interval?\n\nNo. Of course, the correlation has something to do\nwith the properties of light. But to prove Bell\'s\ninequality, we do not have to refer to these properties.\nAll we need is the general hypothesis that there exist\nsome properties which allow to explain the correlations,\nand that in this explanation the decisions of experimenters\nat one place have no causal influence on the observable\nresults at the other place.\n\n> Because, if the correlation curve *does* have something\n> to do with this, then it would seem to follow that\n> these properties were produced via emission.\n\nWhy this? The correlations of the information transferred\nvia usual phones have something to do with the physics\nof the phones, for example some phone lines through the ocean.\n\nIt does not follow that the information has been produced by\nthe producers of the phone lines.\n\n> Otherwise, the polarizers aren\'t *analyzing* anything.\n> They\'re just producing some optical state which is then\n> instantaneously transmitted to the other arm of the\n> setup.\n\nWhich words you use to describe the setup does not matter.\nWhat you need is that they produce an observable result, and\nthat the experimenter can turn them in certain ways. And the\ncorrelations of the results in dependence of the decisions of\nthe two experimenters.\n\n> If that\'s the case, then we shouldn\'t need to worry\n> about whether or not we\'re dealing (in a given coincidence\n> interval) with light emitted by the same oscillator.\n\nWe obviously should worry. If the light is not emitted by the\nsame source, we observe other, trivial correlations which do\nnot violate Bell\'s inequality.\n\nAs well as we are unable to talk using an interrupted phone line.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Thomas Trotter" <thomastrotter2005@juno.com> schrieb
> > > Thomas Trotter (TT)
> > > Maybe it's just that knowledge of what is
> > > actually happening wrt the emission and detection of the opposite
> > > moving light beams is incomplete (ie., that there *is* a local,
> > > common-property reason for the correlations, and not enough is
> > > known about the physical process to talk about it in a straighforward
> > > local realistic way yet).
>
> Ilja Schmelzer (IS):
> > Given Bell's theorem, there is no such explanation. This is independent
> > of our state of knowledge or state of our theories.
>
> TT:
> Then, you are saying that the correlation curve has
> nothing to do with the properties of the light incident
> on the polarizers during a given coincidence interval?

No. Of course, the correlation has something to do
with the properties of light. But to prove Bell's
inequality, we do not have to refer to these properties.
All we need is the general hypothesis that there exist
some properties which allow to explain the correlations,
and that in this explanation the decisions of experimenters
at one place have no causal influence on the observable
results at the other place.

> Because, if the correlation curve *does* have something
> to do with this, then it would seem to follow that
> these properties were produced via emission.

Why this? The correlations of the information transferred
via usual phones have something to do with the physics
of the phones, for example some phone lines through the ocean.

It does not follow that the information has been produced by
the producers of the phone lines.

> Otherwise, the polarizers aren't *analyzing* anything.
> They're just producing some optical state which is then
> instantaneously transmitted to the other arm of the
> setup.

Which words you use to describe the setup does not matter.
What you need is that they produce an observable result, and
that the experimenter can turn them in certain ways. And the
correlations of the results in dependence of the decisions of
the two experimenters.

> If that's the case, then we shouldn't need to worry
> about whether or not we're dealing (in a given coincidence
> interval) with light emitted by the same oscillator.

We obviously should worry. If the light is not emitted by the
same source, we observe other, trivial correlations which do
not violate Bell's inequality.

As well as we are unable to talk using an interrupted phone line.

Ilja

Thomas Trotter

Nov5-04, 06:11 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>[snip]\n\nIlja Schmelzer:\n> ... I have no problem if you want explain violations of BI via\n> speed >> c or via global variables. The argument is that, whatever,\n> Einstein causality is already dead. Because it forbids, together with\n> realism, violations of BI.\n\nThomas Trotter:\nViolations of BI [wrt, eg., typical optical Bell test setups dealing\nwith photons correlated in polarization] are explained via a global\nconstant [identical polarization of light pulses emitted by the same atom]\nthat is being analyzed wrt a global variable [the angular difference\nbetween the transmission axes of crossed linear polarizers].\nI guess you might depict it like this: A <--- C ---> B,\nwhich conforms to standard notions of locality, and causality.\n\n\nBell doesn\'t contradict this. Bell says that the correlational,\nor combined, context can\'t be correctly described (assuming\nthe extant experimental results are correct) as a combination\nof the (unknown local) factor(s), or hidden variable(s), which would,\nif they were known, allow a more accurate prediction of individual\nresults.\n\nWrt the sort of optical tests of BI mentioned above, the hidden\nvariable which determines individual results is the specific\npolarization direction of polarizer-incident light. Knowledge of\nthis, along with the polarizer setting, would facilitate\nan accurate prediction of the individual result. In general then,\nthe probability of individual detection at A and/or B can be\nwritten, P(A) = cos^2 |a - lambda| and P(B) = cos^2 |b - lambda|,\nwhere lambda is the polarization of the polarizer-incident light, and\na and b are the settings of the polarizers at A and B, respectively.\n\nBell shows that if one assumes a common direction of polarization\nfor photon 1 and photon 2 of a pair, thus describing the probability\nof coincidental detection in terms of the product of the individual\nprobabilities at A and B, then one gets a correlation function which\ndisagrees with qm (and, apparently now, with experimental results).\n\nBut if one assumes, instead, that the parameter determining\ncoincidental detection is the emission-produced *relationship*\nbetween the polarization of photon 1 and photon 2 of any given\npair (as opposed to the common *direction* of polarization),\nthen one gets a correlation function which agrees with qm\n(in fact, this is the qm approach) and experiment.\nNote that this parameter is unvarying from pair to pair, unlike\nthe specific direction of polarization which varies randomly\nfrom pair to pair.\n\nThe property of the polarizer-incident light pulse (direction of\npolarization) that determines individual detection, is not the\nproperty of the opposite-moving, polarizer-incident light pulses\n(identical polarization) that determines coincidental detection.\n\nAt the outset, knowing only that a pair of photons are identically\npolarized does not help to accurately describe probability of\nindividual detection, but it does allow an accurate description of\nprobability of coincidental detection.\n\nAgain, to make this clear, in the individual context it\'s\nthe *variable* emission-produced polarization of the incident\nlight pulse that\'s being analyzed by the polarizer -- but in the\ncombined context it\'s the *nonvarying* emission-produced relationship\nbetween the polarizations of the incident light pulses that\'s\nbeing analyzed by the crossed linear polarizers.\n\nViolations of BI don\'t reveal anything about the physical\nnature of the correlations that isn\'t already assumed via the\n(local) emission model. Bell\'s theorem does not allow\nany conclusions about the existence of instantaneous\nmechanisms or the (non)reality of hidden variables. It does\nhelp clarify why certain formulations don\'t work in certain\ncontexts.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>[snip]

Ilja Schmelzer:
> ... I have no problem if you want explain violations of BI via
> speed >> c or via global variables. The argument is that, whatever,
> Einstein causality is already dead. Because it forbids, together with
> realism, violations of BI.

Thomas Trotter:
Violations of BI [wrt, eg., typical optical Bell test setups dealing
with photons correlated in polarization] are explained via a global
constant [identical polarization of light pulses emitted by the same atom]
that is being analyzed wrt a global variable [the angular difference
between the transmission axes of crossed linear polarizers].
I guess you might depict it like this: A <--- C ---> B,
which conforms to standard notions of locality, and causality.

Bell doesn't contradict this. Bell says that the correlational,
or combined, context can't be correctly described (assuming
the extant experimental results are correct) as a combination
of the (unknown local) factor(s), or hidden variable(s), which would,
if they were known, allow a more accurate prediction of individual
results.

Wrt the sort of optical tests of BI mentioned above, the hidden
variable which determines individual results is the specific
polarization direction of polarizer-incident light. Knowledge of
this, along with the polarizer setting, would facilitate
an accurate prediction of the individual result. In general then,
the probability of individual detection at A and/or B can be
written, P(A) = cos^2 |a - \lambda| and P(B) = cos^2 |b - \lambda|,
where \lambda is the polarization of the polarizer-incident light, and
a and b are the settings of the polarizers at A and B, respectively.

Bell shows that if one assumes a common direction of polarization
for photon 1 and photon 2 of a pair, thus describing the probability
of coincidental detection in terms of the product of the individual
probabilities at A and B, then one gets a correlation function which
disagrees with qm (and, apparently now, with experimental results).

But if one assumes, instead, that the parameter determining
coincidental detection is the emission-produced *relationship*
between the polarization of photon 1 and photon 2 of any given
pair (as opposed to the common *direction* of polarization),
then one gets a correlation function which agrees with qm
(in fact, this is the qm approach) and experiment.
Note that this parameter is unvarying from pair to pair, unlike
the specific direction of polarization which varies randomly
from pair to pair.

The property of the polarizer-incident light pulse (direction of
polarization) that determines individual detection, is not the
property of the opposite-moving, polarizer-incident light pulses
(identical polarization) that determines coincidental detection.

At the outset, knowing only that a pair of photons are identically
polarized does not help to accurately describe probability of
individual detection, but it does allow an accurate description of
probability of coincidental detection.

Again, to make this clear, in the individual context it's
the *variable* emission-produced polarization of the incident
light pulse that's being analyzed by the polarizer -- but in the
combined context it's the *nonvarying* emission-produced relationship
between the polarizations of the incident light pulses that's
being analyzed by the crossed linear polarizers.

Violations of BI don't reveal anything about the physical
nature of the correlations that isn't already assumed via the
(local) emission model. Bell's theorem does not allow
any conclusions about the existence of instantaneous
mechanisms or the (non)reality of hidden variables. It does
help clarify why certain formulations don't work in certain
contexts.

Thomas Trotter

Nov5-04, 08:20 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n\n\nPatrick Van Esch:\n> > I\'m just trying to make sense of that situation\n> > within the framework as presented, and I think I can,\n\nIlja Schmelzer:\n> me too. Realistic interpretations like BM are possible. They fit\n> into the framework as presented (give the experimental predictions\n> of QM). "Making sense" is something I consider to be equivalent\n> to "realistic theory".\n\nThomas Trotter:\nIf nonlocality isn\'t a fact of nature, then BM isn\'t\nrealistic. Below is a local realistic explanation\nfor the EPR-Bell correlations, and an explanation\nfor why it isn\'t contradicted by Bell\'s theorem.\n\nThe Local Realistic explanation:\n\nThe correlations are caused by an emission-produced\nnonvarying relationship between paired light pulses,\nin conjunction with the observational context.\nWrt, say, the production of entangled light via\nemission by a common oscillator in a laser-pulsed,\natomic calcium cascade, the nonvarying relationship\nbetween the (paired) opposite moving light pulses is\nthat they\'re *identically polarized* (due to conservation\nof angular momentum) -- and this is the basis for\ncalculating probability of coincidental detection\nvia crossed linear polarizers.\n\nWhy Bell\'s theorem doesn\'t contradict this local\nrealist explanation:\n\nDefine the relevant hidden parameter, denoted as lambda,\nas a nonvarying, emission-produced relationship between\nlight pulses emitted by the same atom.\n\nIn this case, lambda can\'t be used to calculate individual\nresults at A and/or B for some (any) specific setting of\nthe polarizer(s), because lambda is nonvarying.\n\nBell\'s local hidden *variable* formulation is, therefore,\nobviated -- and with it the necessity to invoke FTL or\ninstantaneous mechanisms to explain the observed correlation\ncurves.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch:
> > I'm just trying to make sense of that situation
> > within the framework as presented, and I think I can,

Ilja Schmelzer:
> me too. Realistic interpretations like BM are possible. They fit
> into the framework as presented (give the experimental predictions
> of QM). "Making sense" is something I consider to be equivalent
> to "realistic theory".

Thomas Trotter:
If nonlocality isn't a fact of nature, then BM isn't
realistic. Below is a local realistic explanation
for the EPR-Bell correlations, and an explanation
for why it isn't contradicted by Bell's theorem.

The Local Realistic explanation:

The correlations are caused by an emission-produced
nonvarying relationship between paired light pulses,
in conjunction with the observational context.
Wrt, say, the production of entangled light via
emission by a common oscillator in a laser-pulsed,
atomic calcium cascade, the nonvarying relationship
between the (paired) opposite moving light pulses is
that they're *identically polarized* (due to conservation
of angular momentum) -- and this is the basis for
calculating probability of coincidental detection
via crossed linear polarizers.

Why Bell's theorem doesn't contradict this local
realist explanation:

Define the relevant hidden parameter, denoted as \lambda,
as a nonvarying, emission-produced relationship between
light pulses emitted by the same atom.

In this case, \lambda can't be used to calculate individual
results at A and/or B for some (any) specific setting of
the polarizer(s), because \lambda is nonvarying.

Bell's local hidden *variable* formulation is, therefore,
obviated -- and with it the necessity to invoke FTL or
instantaneous mechanisms to explain the observed correlation
curves.

rof@maths.tcd.ie

Nov5-04, 08:20 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n>Yep. The construction of the M_n is of this type. But the X_n\\M_n,\n>the inner part of the moon, is not perceived. Even if I have some\n>perception M_moon of the surface of the moon and some vague\n>theoretical ideas what may be inside.\n\nOk; I see what you\'re saying. Conceptual information, such as\n"Iron is the stablest nucleus", is not represented spatially,\nand it is most likely ideas like this which guide your thinking\nabout what is going on inside planets. I don\'t disagree with this,\nbut the sense in which I use the X_n is to refer explicitly to the\nprocedure which we use to construct a "world around us" from\nsensation, and which is explicitly spatial.\n\n>> These matters are relevant for physics because the statistics of\n>> the E_n either are or are not representable as a Markov process\n>> which factorises in a spatial way, and so these considerations\n>> must arise whenever one wants to ask whether an observer will\n>> see space.\n\n>Sorry, but you cannot derive if observers will "see" space.\n>"Seeing" space means developing a certain realistic theory,\n>may be implicit (as a hidden prejudice, instinct, created by\n>evolution).\n\nEvolution is another can of worms, best left unopened here, as is\nthe connection between the brain and conscious experience. Animals\nsee space, but they don\'t develop ontological theories. It is only\nconfused humans who see a brick and then imagine that the brick\nisn\'t really there; that it\'s only a perceived brick, and that\nwhat "really" exists are tiny particles swirling around under the\ninfluence of various force fields. It\'s only theoretical physicists\nwho have to face the dilemma of whether to stick the "reality" sticker\nonto the wavefunction or the observables, because the urge to stick\nit somewhere is so strong.\n\n>> >> In an ontological theory the splitting needs to be deduced,\n>> >> rather than postulated.\n>>\n>> >The ontological theory postulates the structure and the laws of\n>> >the X_n. If these have the form of something localized in a space,\n>> >space is postulated. What you perceive is another question,\n>> >in the E_n we have, for example, nothing three-dimensional.\n>> >We see only two 2D movies. So it is completely unclear what\n>> >means spatial structure in the E_n.\n\n>> No; the ontological theory can\'t make any postulates about the X_n,\n>> because the X_n exist only in the mind of the observer, who has\n>> constructed them.\n\n>That\'s a confusion of language levels. A believes that there are\n>some X_n in reality. This is an ontological theory proposed by A.\n\nA may think it\'s on ontological theory, but A is confusing himself.\nImagine there is a digital machine, which takes 1\'s and 0\'s as\ninput. You feed in some sequence of 1\'s and 0\'s, generated by\nsome algorithm. The machine proceeds to construct a Markov model\nto predict the sequence. It is successful to some degree - most\nof the time it correctly predicts the next input. Has it discovered\nreality? Is its Markov model an ontological theory?\n\nThe situation that we find ourselves in is that we know that\nwe cannot perfectly predict the results of measurements (in\nEPR experiments, knowing the outcome of a spin measurement in\nadvance would allow signalling faster than light). This means\nthat our Markov model cannot perfectly predict the future,\neven in principle. We\'re left with a merely statistical prediction,\nalthough we may conjure up fruitless imaginations about what caused\nthe result of a particular experiment to be this or that.\n\nSpecifically, Bohmian mechanics is an attempt to imagine that\nthe technique which we use to order our perceptions (representing\na single state of affairs as lots of little states of affairs,\neach in its own position, in configuration space, for example),\nshould still apply even in the absence of any sensible data\nto represent. That is, it is supposed that whatever is responsible\nfor this experiment giving this result, it must be represented\nas stuff in space, because that\'s what we\'re used to perceiving.\nWhether or not a particle even exists can depend on the motion of\nthe observer (cf. the Unruh effect), so it seems strange to put that\nsticky reality label onto the particles and their positions in space.\n\n>We have to decide which language level we want to use - the level\n>of A, then we talk about X_n as real states, or on a metalevel where\n>we talk about A\'s X_n as constructions of his mind. Of course, on\n>this metalevel we have our metatheory about reality with our own\n>X\'_n. (This consideration was on meta-meta-level)\n\nAll of the X_n are mentally constructed. Putting a sticker on\nthem saying "reality" doesn\'t change that. The difference between\nwhat\'s real and what\'s not is the same as the difference between\nimagined pain and real pain - one is present, the other merely\nreferred to. We may, for the moment, take a particular formulation\nof theoretical physics and work out its consequences, for example\nthe minimally supersymmetric standard model, and say that, within\nthat model, this or that particle is really there (and not merely\nreferred to), but declaring that the model is "out there", beyond\nthe mind of the person who thinks it, governing the evolution\nof the world, is like saying that, since 1 typically follows\n0 in a certain binary sequence, the rule "01" is out there, guiding\nthe sequence.\n\n>The game is reverse. The theory is prior. Then, a complicated\n>data processing starts which makes the E_n compatible with the\n>theory using some M_n. A failure of this process is named\n>falsification.\n\n>Theories about the data processing are part of some historical\n>sciences.\n\nThere\'s a psychological problem which has led theoretical physics\ninto a nasty corner. It\'s the problem of physics students learning\ntoo many theories and doing too many exercises. Once they\'ve grown\nup, the only thing they can do is take a set of axioms and see\nwhat follows from them. Now we have a situation where the physicists\nneed to know "What\'s the Lagrangian?" before they can get any further.\nIt\'s like a McDonald\'s employee starving to death in an orchard\nbecause he can\'t find the McOven and the McFries. Theorizing\nconsists of writing down different Lagrangians and working out\ntheir consequences. "The theory is prior." Compute, compute, compute.\n\nIt\'s usually the case that a species of knowledge has to be\nvery thoroughly understood and the doctrine has to be very mature\nbefore it admits an axiomatic formulation. The process of\ndiscovering, by careful thought and experimentation, what the\nactual properties of the object of study are, is necessary\nbefore one can start thinking of compressing all the relevant\ninformation into some axioms. At the moment, however, physicists\nare still doing their homework, starting from axioms and deriving\nthe consequences. Some of the axioms they think they know (quantum\nmechanics, Lagrangians, manifolds), and some they imaginatively\nproduce on their own, but nobody tries to understand starting from\nthe facts which are right in front of them. That\'s not what they\nwere trained to do.\n\n>> >> You do, but giving up either realism or causality won\'t affect your\n>> >> ability to predict X_{n+1} from X_n, since they are merely rules of\n>> >> thumb for generating representations.\n>>\n>> >No, the rules of thumb are questioned. I have no way to construct\n>> >some X_n in the usual (causal, realistic) meaning which has the\n>> >property of Lorentz-invariance.\n>>\n>> Lorentz invariance and relativity are important when one needs to\n>> compare one observer\'s description (the series X_n) with another\'s.\n\n>Completely wrong. To compare descriptions of different observers\n>is complicate enough in everyday life. Relativity is important in\n>strong gravitational fields or for high velocities. Lorentz invariance\n>is useful to prove some otherwise strange impossibility results\n>(impossibility to measure absolute time) and otherwise useless.\n>Proof: Lorentz-symmetric Maxwell theory was successful long\n>before its Lorentz-symmetry has been observed.\n\nThere are two possibilities. One is that you paid too much attention\nto my (probably inappropriate) use of the word "important", and\ninterpreted it as though I were saying that effects which are related\nto relativity arise only in the circumstances that I was indicating;\nand the other is that you feel as though you\'ve scored a point when\nyou get to say things like "Completely wrong," and couldn\'t resist\ndoing so.\n\nThe point of the Lorentz transformations is to switch between one\nobserver and another, and the essence of relativity is that the\nlaws of physics are the same for different (inertial) observers,\neven the speed of light. Without the idea of changing reference\npoints and considering things from a different observer\'s point\nof view, special relativity would not have the formulation that\nit does. In order to formulate the questions which relativity\naddresses in my language, which is to some extent solipsistic,\none first has to address the question of what relationship\nexists between one observer\'s X_n and another\'s X\'_n, which\ninvolves how one observer manifests himself within the X_n\nof another, and that is a difficult question.\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb
>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

>Yep. The construction of the M_n is of this type. But the X_n\M_n,
>the inner part of the moon, is not perceived. Even if I have some
>perception M_{moon} of the surface of the moon and some vague
>theoretical ideas what may be inside.

Ok; I see what you're saying. Conceptual information, such as
"Iron is the stablest nucleus", is not represented spatially,
and it is most likely ideas like this which guide your thinking
about what is going on inside planets. I don't disagree with this,
but the sense in which I use the X_n is to refer explicitly to the
procedure which we use to construct a "world around us" from
sensation, and which is explicitly spatial.

>> These matters are relevant for physics because the statistics of
>> the E_n either are or are not representable as a Markov process
>> which factorises in a spatial way, and so these considerations
>> must arise whenever one wants to ask whether an observer will
>> see space.

>Sorry, but you cannot derive if observers will "see" space.
>"Seeing" space means developing a certain realistic theory,
>may be implicit (as a hidden prejudice, instinct, created by
>evolution).

Evolution is another can of worms, best left unopened here, as is
the connection between the brain and conscious experience. Animals
see space, but they don't develop ontological theories. It is only
confused humans who see a brick and then imagine that the brick
isn't really there; that it's only a perceived brick, and that
what "really" exists are tiny particles swirling around under the
influence of various force fields. It's only theoretical physicists
who have to face the dilemma of whether to stick the "reality" sticker
onto the wavefunction or the observables, because the urge to stick
it somewhere is so strong.

>> >> In an ontological theory the splitting needs to be deduced,
>> >> rather than postulated.
>>
>> >The ontological theory postulates the structure and the laws of
>> >the X_n. If these have the form of something localized in a space,
>> >space is postulated. What you perceive is another question,
>> >in the E_n we have, for example, nothing three-dimensional.
>> >We see only two 2D movies. So it is completely unclear what
>> >means spatial structure in the E_n.

>> No; the ontological theory can't make any postulates about the X_n,
>> because the X_n exist only in the mind of the observer, who has
>> constructed them.

>That's a confusion of language levels. A believes that there are
>some X_n in reality. This is an ontological theory proposed by A.

A may think it's on ontological theory, but A is confusing himself.
Imagine there is a digital machine, which takes 1's and 0's as
input. You feed in some sequence of 1's and 0's, generated by
some algorithm. The machine proceeds to construct a Markov model
to predict the sequence. It is successful to some degree - most
of the time it correctly predicts the next input. Has it discovered
reality? Is its Markov model an ontological theory?

The situation that we find ourselves in is that we know that
we cannot perfectly predict the results of measurements (in
EPR experiments, knowing the outcome of a spin measurement in
advance would allow signalling faster than light). This means
that our Markov model cannot perfectly predict the future,
even in principle. We're left with a merely statistical prediction,
although we may conjure up fruitless imaginations about what caused
the result of a particular experiment to be this or that.

Specifically, Bohmian mechanics is an attempt to imagine that
the technique which we use to order our perceptions (representing
a single state of affairs as lots of little states of affairs,
each in its own position, in configuration space, for example),
should still apply even in the absence of any sensible data
to represent. That is, it is supposed that whatever is responsible
for this experiment giving this result, it must be represented
as stuff in space, because that's what we're used to perceiving.
Whether or not a particle even exists can depend on the motion of
the observer (cf. the Unruh effect), so it seems strange to put that
sticky reality label onto the particles and their positions in space.

>We have to decide which language level we want to use - the level
>of A, then we talk about X_n as real states, or on a metalevel where
>we talk about A's X_n as constructions of his mind. Of course, on
>this metalevel we have our metatheory about reality with our own
>X'_n. (This consideration was on meta-meta-level)

All of the X_n are mentally constructed. Putting a sticker on
them saying "reality" doesn't change that. The difference between
what's real and what's not is the same as the difference between
imagined pain and real pain - one is present, the other merely
referred to. We may, for the moment, take a particular formulation
of theoretical physics and work out its consequences, for example
the minimally supersymmetric standard model, and say that, within
that model, this or that particle is really there (and not merely
referred to), but declaring that the model is "out there", beyond
the mind of the person who thinks it, governing the evolution
of the world, is like saying that, since 1 typically follows
in a certain binary sequence, the rule "01" is out there, guiding
the sequence.

>The game is reverse. The theory is prior. Then, a complicated
>data processing starts which makes the E_n compatible with the
>theory using some M_n. A failure of this process is named
>falsification.

>Theories about the data processing are part of some historical
>sciences.

There's a psychological problem which has led theoretical physics
into a nasty corner. It's the problem of physics students learning
too many theories and doing too many exercises. Once they've grown
up, the only thing they can do is take a set of axioms and see
what follows from them. Now we have a situation where the physicists
need to know "What's the Lagrangian?" before they can get any further.
It's like a McDonald's employee starving to death in an orchard
because he can't find the McOven and the McFries. Theorizing
consists of writing down different Lagrangians and working out
their consequences. "The theory is prior." Compute, compute, compute.

It's usually the case that a species of knowledge has to be
very thoroughly understood and the doctrine has to be very mature
before it admits an axiomatic formulation. The process of
discovering, by careful thought and experimentation, what the
actual properties of the object of study are, is necessary
before one can start thinking of compressing all the relevant
information into some axioms. At the moment, however, physicists
are still doing their homework, starting from axioms and deriving
the consequences. Some of the axioms they think they know (quantum
mechanics, Lagrangians, manifolds), and some they imaginatively
produce on their own, but nobody tries to understand starting from
the facts which are right in front of them. That's not what they
were trained to do.

>> >> You do, but giving up either realism or causality won't affect your
>> >> ability to predict X_{n+1} from X_n, since they are merely rules of
>> >> thumb for generating representations.
>>
>> >No, the rules of thumb are questioned. I have no way to construct
>> >some X_n in the usual (causal, realistic) meaning which has the
>> >property of Lorentz-invariance.
>>
>> Lorentz invariance and relativity are important when one needs to
>> compare one observer's description (the series X_n) with another's.

>Completely wrong. To compare descriptions of different observers
>is complicate enough in everyday life. Relativity is important in
>strong gravitational fields or for high velocities. Lorentz invariance
>is useful to prove some otherwise strange impossibility results
>(impossibility to measure absolute time) and otherwise useless.
>Proof: Lorentz-symmetric Maxwell theory was successful long
>before its Lorentz-symmetry has been observed.

There are two possibilities. One is that you paid too much attention
to my (probably inappropriate) use of the word "important", and
interpreted it as though I were saying that effects which are related
to relativity arise only in the circumstances that I was indicating;
and the other is that you feel as though you've scored a point when
you get to say things like "Completely wrong," and couldn't resist
doing so.

The point of the Lorentz transformations is to switch between one
observer and another, and the essence of relativity is that the
laws of physics are the same for different (inertial) observers,
even the speed of light. Without the idea of changing reference
points and considering things from a different observer's point
of view, special relativity would not have the formulation that
it does. In order to formulate the questions which relativity
addresses in my language, which is to some extent solipsistic,
one first has to address the question of what relationship
exists between one observer's X_n and another's X'_n, which
involves how one observer manifests himself within the X_n
of another, and that is a difficult question.

R.

Thomas Trotter

Nov5-04, 08:20 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n[snip]\n\nPatrick Van Esch (PVE):\n> Bell is making up a "local realistic" model of the entangled pair. If\n> you say that the photons HAD a specific polarisation direction lambda,\n> then you have to be able to write it down (even if you don\'t know what\n> it is). That\'s what he does.\n\nThomas Trotter (TT):\nOk, but suppose we define lambda as the *relationship between*\nthe entangled photons. That is, after all, what is being\nanalyzed in the correlational context, isn\'t it?\nThe physical basis for this wrt, eg., pairs produced via atomic\ncalcium cascades is that the paired photons correspond to\nopposite-moving light pulses emitted by the same atom, and they\nare identically polarized via emission due to conservation of\nangular momentum.\n\nAt the outset, the state of any individual photon is\nundefined. The specific polarization direction of any\ngiven photon, or pair of photons, is unknown. But the\nrelationship (lambda) between photon 1 and photon 2 of\nany given pair, via the emission model, is unvarying and\nsays that if photon 1 is right (left) circularly polarized,\nthen photon 2 is also right (left) circularly polarized.\n\nIn other words, the correlational state is not undefined,\nprovided that it\'s light pulses emitted from the same atom\nthat are being paired via the coincidence circuitry, and\nwhich correspond to entangled photons.\n\nThis is a local picture, as far as I can tell.\n\nAnd, it follows from this that the probability of\ncoincidental detection (in the ideal) will be cos^2 theta\n(where theta is the angular difference between the transmission\naxes of the crossed linear polarizers).\n\nPVE:\n>\n> Now the locality resides in the fact that we have the SAME lambda for\n> A and for B, in that there cannot be an extra variable at B that is a\n> function of a.\n> The photons had to take all their realistic description at their\n> creation and then nothing else for one photon can be dependent on what\n> happens to the other photon.\n>\n\nTT:\nThat\'s what we have if we take lambda to mean the (nonvarying)\nrelationship between the emission-produced polarization of the\nlight pulses incident on the polarizers during any given\ncorrelation or coincidence window -- rather than the (variable)\nspecific polarization direction. Isn\'t it?\n> >\n> > Anyway, there are two incompatible formulations involved here.\n\nPVE:\n> Exactly, that was Bell\'s point.\n\nTT:\n> > What\'s wrong with Bell\'s formulation? It doesn\'t seem to be,\n> > strictly speaking, the assumption of locality -- since qm also\n> > uses the assumption of a common emitter as the basis for\n> > calculations.\n\nPVE:\n> As I said, the locality implies that we have to use the same\n> lambda at A and at B.\n\nTT:\nOk, we seem to have solved the problem by denoting lambda as\nthe relationship between, rather than the polarization of, paired\nphotons.\n\nSo, the formulations (Bell\'s LHV and QM) would seem to be\nincompatible because the hidden local parameter relevant to\ndetermining coincidental detection isn\'t a variable.\n\nIOW, the specific polarization direction is irrelevant wrt\ndetermining coincidental detection.\n\nWhat do you think?\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>[snip]

Patrick Van Esch (PVE):
> Bell is making up a "local realistic" model of the entangled pair. If
> you say that the photons HAD a specific polarisation direction \lambda,
> then you have to be able to write it down (even if you don't know what
> it is). That's what he does.

Thomas Trotter (TT):
Ok, but suppose we define \lambda as the *relationship between*
the entangled photons. That is, after all, what is being
analyzed in the correlational context, isn't it?
The physical basis for this wrt, eg., pairs produced via atomic
calcium cascades is that the paired photons correspond to
opposite-moving light pulses emitted by the same atom, and they
are identically polarized via emission due to conservation of
angular momentum.

At the outset, the state of any individual photon is
undefined. The specific polarization direction of any
given photon, or pair of photons, is unknown. But the
relationship (\lambda) between photon 1 and photon 2 of
any given pair, via the emission model, is unvarying and
says that if photon 1 is right (left) circularly polarized,
then photon 2 is also right (left) circularly polarized.

In other words, the correlational state is not undefined,
provided that it's light pulses emitted from the same atom
that are being paired via the coincidence circuitry, and
which correspond to entangled photons.

This is a local picture, as far as I can tell.

And, it follows from this that the probability of
coincidental detection (in the ideal) will be cos^2 \theta
(where \theta is the angular difference between the transmission
axes of the crossed linear polarizers).

PVE:
>
> Now the locality resides in the fact that we have the SAME \lambda for
> A and for B, in that there cannot be an extra variable at B that is a
> function of a.
> The photons had to take all their realistic description at their
> creation and then nothing else for one photon can be dependent on what
> happens to the other photon.
>

TT:
That's what we have if we take \lambda to mean the (nonvarying)
relationship between the emission-produced polarization of the
light pulses incident on the polarizers during any given
correlation or coincidence window -- rather than the (variable)
specific polarization direction. Isn't it?
> >
> > Anyway, there are two incompatible formulations involved here.

PVE:
> Exactly, that was Bell's point.

TT:
> > What's wrong with Bell's formulation? It doesn't seem to be,
> > strictly speaking, the assumption of locality -- since qm also
> > uses the assumption of a common emitter as the basis for
> > calculations.

PVE:
> As I said, the locality implies that we have to use the same
> \lambda at A and at B.

TT:
Ok, we seem to have solved the problem by denoting \lambda as
the relationship between, rather than the polarization of, paired
photons.

So, the formulations (Bell's LHV and QM) would seem to be
incompatible because the hidden local parameter relevant to
determining coincidental detection isn't a variable.

IOW, the specific polarization direction is irrelevant wrt
determining coincidental detection.

What do you think?

Oz

Nov7-04, 07:44 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\nIlja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes\n\n>The whole point is a different one. We really have some black box.\n>This black box allows to do very strange things. Unfortunately it\n>does not allow us to build an FTL phone. But there exists no realistic\n>explanation of what is inside the black box which does not use some\n>sort of hidden FTL information transfer.\n\nQuite. The crux of the problem.\n\nOne viable solution is to claim that\n\nFTL is perfectly possible\nBUT ONLY inside the black box.\n\nOne should then examine the box.\nWhat is strange about it?\n\nWell, its easier to reverse the question and ask \'what is strange about\na measurement\'? Well, a measurement is something everyone agrees on. In\nrelativity it is an event. Everyone, everywhere, agrees about an event.\n\nNow that\'s an extraordinary statement when you think about it. Why\n*should* everyone agree about something? Furthermore there is an implied\nconnection between the whole universe (who all agree about the event)\nand the event.\n\nQM, however, says we can never be quite sure. Different observers will\nin fact measure different results for an event, mediated by h. To me\nthis strongly suggests that an event is probabilistic. That is an event\nis simply something with a very high probability of occurring (and I\'m\ntalking about near certainty here).\n\nSo classical physics is simply the physics of (very) probable events.\nThat is the implication of setting h=0.\nSo the implication is that the probability of achieving FTL\ncommunication is related to h.\n\nThat is (macroscopically) you can have a FTL phone, but unfortunately\nyou can\'t expect it to deliver the correct words more than (say)\n1:10^100 times you use it, with the chance of correctness related to h.\nThat is if you listen to noise long enough, you will eventually hear\nwords, very rarely these will actually be a FTL communication. That\'s\nbecause the number of observers is the whole observable universe.\n\nHowever, consider a very small isolated system.\nNow to think inside the box.....\n\nIt has a known (ie known to the universe) particle that decays into two\nparticles A and notA. Because it is isolated it is not connected to the\nrest of the universe. It is not an event, it can be as improbable as it\nlikes. The rest of the universe do NOT (and cannot) agree on A and notA.\nWe send one to jill and one to james, who measure them.\n\nThe measurements are an event. The results must now be highly probable.\nThere are only two solutions, either jill gets A or notA (and vv for\njames). Anything else is by definition improbable, that is it does not\nconform to classical behaviour (which we know to be true to very high\naccuracy).\n\nNow one could say that at this point there are two spherical superposed\nshells of \'results\' that expand at lightspeed. This is, however,\nimplausible to me. That is its not a likely solution, or way of looking\nat it. Far too many observers, far too complicated. The alternative is\nthat A and notA are in fact in *effective* FTL communication until the\nmeasurement. The question is thus how to rationalise this.\n\nOne way is to examine what we mean by \'distance\', the other is what we\nmean by \'time\'. Now, both of these are normally defined macroscopically,\nthat is by chains of events. That is they are classical statements. The\nbox is NOT a classical statement.\n\nThere is no a priori reason why these should be the same for our\nparticle example above. Its by definition isolated. There is by\ndefinition no chain of events to internally define its internal distance\nor time. In a very real sense they are undefined and undefinable.\nWithout any such definition it makes no sense to state anything about\nFTL communication. Excluding gravity they are a mini-universe containing\nA and notA. When \'the rest of the universe\' comes in and takes one, say\nA definitively as it must, then the rest of the universe defines the\nother as notA. Its probably convenient to imagine this happens\nsimultaneously (although time is probably undefined internally) within\nthe \'mini-universe\'. Anything else is (within the greater universe)\nutterly improbable.\n\nI hope this makes a little sense.\n\n--\nOz\nThis post is worth absolutely nothing and is probably fallacious.\n\nUse oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].\nBTOPENWORLD address has ceased. DEMON address has ceased.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes

>The whole point is a different one. We really have some black box.
>This black box allows to do very strange things. Unfortunately it
>does not allow us to build an FTL phone. But there exists no realistic
>explanation of what is inside the black box which does not use some
>sort of hidden FTL information transfer.

Quite. The crux of the problem.

One viable solution is to claim that

FTL is perfectly possible
BUT ONLY inside the black box.

One should then examine the box.
What is strange about it?

Well, its easier to reverse the question and ask 'what is strange about
a measurement'? Well, a measurement is something everyone agrees on. In
relativity it is an event. Everyone, everywhere, agrees about an event.

Now that's an extraordinary statement when you think about it. Why
*should* everyone agree about something? Furthermore there is an implied
connection between the whole universe (who all agree about the event)
and the event.

QM, however, says we can never be quite sure. Different observers will
in fact measure different results for an event, mediated by h. To me
this strongly suggests that an event is probabilistic. That is an event
is simply something with a very high probability of occurring (and I'm
talking about near certainty here).

So classical physics is simply the physics of (very) probable events.
That is the implication of setting h=0.
So the implication is that the probability of achieving FTL
communication is related to h.

That is (macroscopically) you can have a FTL phone, but unfortunately
you can't expect it to deliver the correct words more than (say)
1:10^100 times you use it, with the chance of correctness related to h.
That is if you listen to noise long enough, you will eventually hear
words, very rarely these will actually be a FTL communication. That's
because the number of observers is the whole observable universe.

However, consider a very small isolated system.
Now to think inside the box.....

It has a known (ie known to the universe) particle that decays into two
particles A and notA. Because it is isolated it is not connected to the
rest of the universe. It is not an event, it can be as improbable as it
likes. The rest of the universe do NOT (and cannot) agree on A and notA.
We send one to jill and one to james, who measure them.

The measurements are an event. The results must now be highly probable.
There are only two solutions, either jill gets A or notA (and vv for
james). Anything else is by definition improbable, that is it does not
conform to classical behaviour (which we know to be true to very high
accuracy).

Now one could say that at this point there are two spherical superposed
shells of 'results' that expand at lightspeed. This is, however,
implausible to me. That is its not a likely solution, or way of looking
at it. Far too many observers, far too complicated. The alternative is
that A and notA are in fact in *effective* FTL communication until the
measurement. The question is thus how to rationalise this.

One way is to examine what we mean by 'distance', the other is what we
mean by 'time'. Now, both of these are normally defined macroscopically,
that is by chains of events. That is they are classical statements. The
box is NOT a classical statement.

There is no a priori reason why these should be the same for our
particle example above. Its by definition isolated. There is by
definition no chain of events to internally define its internal distance
or time. In a very real sense they are undefined and undefinable.
Without any such definition it makes no sense to state anything about
FTL communication. Excluding gravity they are a mini-universe containing
A and notA. When 'the rest of the universe' comes in and takes one, say
A definitively as it must, then the rest of the universe defines the
other as notA. Its probably convenient to imagine this happens
simultaneously (although time is probably undefined internally) within
the 'mini-universe'. Anything else is (within the greater universe)
utterly improbable.

I hope this makes a little sense.

--
Oz
This post is worth absolutely nothing and is probably fallacious.

Use oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].
BTOPENWORLD address has ceased. DEMON address has ceased.

Thomas Trotter

Nov7-04, 04:05 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410302355.1727b3ff@posting.google. com>...\n> thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410291959.28f3e245@posting.google. com>...\n\nPatrick Van Esch:\n> Bell is making up a "local realistic" model of the entangled pair. If\n> you say that the photons HAD a specific polarisation direction lambda,\n> then you have to be able to write it down (even if you don\'t know what\n> it is). That\'s what he does.\n\nThomas Trotter:\nYou only have to write it down if it\'s a relevant factor in\nthe experimental context that you\'re modeling. But the specific,\ncommon polarization direction is irrelevant wrt coincidental\ndetection. It can be anything -- doesn\'t matter.\nSo, you wouldn\'t include it in your formulation. Would you?\n\nThe only *variable* determining coincidental detection is\nthe angular difference between the transmission axes of the\ncrossed linear polarizers. Isn\'t it?\n\nPatrick Van Esch:\n> Quantum mechanics doesn\'t take this approach. It writes down that the\n> state is:\n>\n> |psi> = 1/sqrt(2) { |z+>|z-> - |z->|z+ }\n>\n> but some algebra shows you that you can rewrite this as:\n>\n>\n> |psi> = 1/sqrt(2) { |n+>|n-> - |n->|n+ }\n>\n> with n just any direction. So our state is completely "unoriented"\n> which is not possible in a realistic model.\n\nThomas Trotter:\nThe quantum state specifies a *relationship* between\nthe polarizations of paired photons, not a specific\npolarization orientation.\n\nIt\'s this relationship that, along with the joint\npolarizer settings, determines the correlation curve.\nAs the qm notation suggests, the polarization\n*orientation* is irrelevant wrt coincidental\ndetection.\n\nEven a \'realistic\' model has to take this into\naccount. But, Bell\'s model embodies the assumption\nthat the variable *orientation* is a factor in\ndetermining coincidental detection -- which is an\nincorrect assumption.\n\n> > Anyway, there are two incompatible formulations involved here.\n\nPatrick Van Esch:\n> Exactly, that was Bell\'s point.\n\nThomas Trotter:\nYes, but the interpretation that they\'re incompatible because\nof the assumptions of Bell locality and Einstein causality\nis an incorrect interpretation, I think.\n\nThe formulations are incompatible because (assuming that\nexperiments have shown Bell\'s to be wrong and qm to be right)\nBell assumed the relevance of variable polarization *orientation*,\nas opposed to just the nonvarying polarization *sameness* as in the\nqm formulation.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410302355.1727b3ff@posting.google.com>...
> thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410291959.28f3e245@posting.google.com>...

Patrick Van Esch:
> Bell is making up a "local realistic" model of the entangled pair. If
> you say that the photons HAD a specific polarisation direction \lambda,
> then you have to be able to write it down (even if you don't know what
> it is). That's what he does.

Thomas Trotter:
You only have to write it down if it's a relevant factor in
the experimental context that you're modeling. But the specific,
common polarization direction is irrelevant wrt coincidental
detection. It can be anything -- doesn't matter.
So, you wouldn't include it in your formulation. Would you?

The only *variable* determining coincidental detection is
the angular difference between the transmission axes of the
crossed linear polarizers. Isn't it?

Patrick Van Esch:
> Quantum mechanics doesn't take this approach. It writes down that the
> state is:
>
> |\psi> = 1/\sqrt(2) { |z+>|z-> - |z->|z+ }
>
> but some algebra shows you that you can rewrite this as:
>
>
> |\psi> = 1/\sqrt(2) { |n+>|n-> - |n->|n+ }
>
> with n just any direction. So our state is completely "unoriented"
> which is not possible in a realistic model.

Thomas Trotter:
The quantum state specifies a *relationship* between
the polarizations of paired photons, not a specific
polarization orientation.

It's this relationship that, along with the joint
polarizer settings, determines the correlation curve.
As the qm notation suggests, the polarization
*orientation* is irrelevant wrt coincidental
detection.

Even a 'realistic' model has to take this into
account. But, Bell's model embodies the assumption
that the variable *orientation* is a factor in
determining coincidental detection -- which is an
incorrect assumption.

> > Anyway, there are two incompatible formulations involved here.

Patrick Van Esch:
> Exactly, that was Bell's point.

Thomas Trotter:
Yes, but the interpretation that they're incompatible because
of the assumptions of Bell locality and Einstein causality
is an incorrect interpretation, I think.

The formulations are incompatible because (assuming that
experiments have shown Bell's to be wrong and qm to be right)
Bell assumed the relevance of variable polarization *orientation*,
as opposed to just the nonvarying polarization *sameness* as in the
qm formulation.

Oz

Nov9-04, 03:44 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n\nPatrick Van Esch <vanesch@ill.fr> writes\n>Bell is making up a "local realistic" model of the entangled pair. If\n>you say that the photons HAD a specific polarisation direction lambda,\n>then you have to be able to write it down (even if you don\'t know what\n>it is). That\'s what he does.\n\nI have terrible trouble getting anyone to actually answer my question on\nthis clearly. The question is:\n\nIs lambda wrt some apparatus, or wrt the other photon?\n\nThis is a critical difference. The former says the angle is fixed from\nthe start as seen by the rest of the universe, the latter simply says\nthey always have a set relationship.\n\n>Quantum mechanics doesn\'t take this approach. It writes down that the\n>state is:\n>\n>|psi> = 1/sqrt(2) { |n+>|n-> - |n->|n+ }\n>\n>with n just any direction. So our state is completely "unoriented"\n>which is not possible in a realistic model.\n\nIsn\'t this simply saying that we cannot know the absolute orientation,\nbut do know the relative orientation?\n\nI\'m not sure I can see why this is \'unrealistic\'.\n\n>Now the locality resides in the fact that we have the SAME lambda for\n>A and for B, in that there cannot be an extra variable at B that is a\n>function of a.\n>The photons had to take all their realistic description at their\n>creation and then nothing else for one photon can be dependent on what\n>happens to the other photon.\n\nOk. But then you immediately remove any entanglement. You effectively\nsay these are two isolated and unconnected photons, which they cannot\nbe, don\'t you?\n\n\n--\nOz\nThis post is worth absolutely nothing and is probably fallacious.\n\nUse oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].\nBTOPENWORLD address has ceased. DEMON address has ceased.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch <vanesch@ill.fr> writes
>Bell is making up a "local realistic" model of the entangled pair. If
>you say that the photons HAD a specific polarisation direction \lambda,
>then you have to be able to write it down (even if you don't know what
>it is). That's what he does.

I have terrible trouble getting anyone to actually answer my question on
this clearly. The question is:

Is \lambda wrt some apparatus, or wrt the other photon?

This is a critical difference. The former says the angle is fixed from
the start as seen by the rest of the universe, the latter simply says
they always have a set relationship.

>Quantum mechanics doesn't take this approach. It writes down that the
>state is:
>
>|\psi> = 1/\sqrt(2) { |n+>|n-> - |n->|n+ }
>
>with n just any direction. So our state is completely "unoriented"
>which is not possible in a realistic model.

Isn't this simply saying that we cannot know the absolute orientation,
but do know the relative orientation?

I'm not sure I can see why this is 'unrealistic'.

>Now the locality resides in the fact that we have the SAME \lambda for
>A and for B, in that there cannot be an extra variable at B that is a
>function of a.
>The photons had to take all their realistic description at their
>creation and then nothing else for one photon can be dependent on what
>happens to the other photon.

Ok. But then you immediately remove any entanglement. You effectively
say these are two isolated and unconnected photons, which they cannot
be, don't you?

--
Oz
This post is worth absolutely nothing and is probably fallacious.

Use oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].
BTOPENWORLD address has ceased. DEMON address has ceased.

Oz

Nov14-04, 07:04 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes\n>\n>\n>\n>\n>"Oz" <oz@farmeroz.port995.com> schrieb\n>> Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes\n>> >The whole point is a different one. We really have some black box.\n>> >This black box allows to do very strange things. Unfortunately it\n>> >does not allow us to build an FTL phone. But there exists no realistic\n>> >explanation of what is inside the black box which does not use some\n>> >sort of hidden FTL information transfer.\n>>\n>> Quite. The crux of the problem.\n>> One viable solution is to claim that\n>> FTL is perfectly possible\n>> BUT ONLY inside the black box.\n>> One should then examine the box.\n>> What is strange about it?\n>\n>Nothing (if I understand this in a very methaporical way).\n>\n>Translated into our physics "looking into the box" can be understood\n>as developing realistic hidden variable theories. We know from Bell\n>that they need a preferred frame. Therefore we have to develop\n>theories with preferred frames.\n\nThat would seem to be a viable thing to examine.\nIt either works out, or it doesn\'t.\nBut if its never seriously examined, who knows?\n\nIts unfortunate that even mentioning this allocates you a significant\ncrank index, no matter how serious and how well supported by the maths\nit is.\n\n>That\'s what I\'m doing. I wonder why I\'m almost alone.\n\nER, because nobody will discuss it seriously with you?\n\n>> Now that\'s an extraordinary statement when you think about it. Why\n>> *should* everyone agree about something? Furthermore there is an implied\n>> connection between the whole universe (who all agree about the event)\n>> and the event.\n>\n>The fact is that not everyone agrees about it - there are people who\n>believe in the remaining loopholes like detector efficiency. Thus, there is\n>no\n>"should". That almost everybody agrees about it is a nontrivial fact.\n\nThat\'s not what I meant.\nWhat I meant was that, by definition, and event is (in relativity)\nsomething everyone, everywherewhen, agrees on. If two people see\ndifferent results to some experiment then its not an event. What it is\nis noise (or something similar).\n\n>> QM, however, says we can never be quite sure.\n>\n>QM has never told me such things.\n\nEh?\n\n>> Different observers will\n>> in fact measure different results for an event, mediated by h. To me\n>> this strongly suggests that an event is probabilistic. That is an event\n>> is simply something with a very high probability of occurring (and I\'m\n>> talking about near certainty here).\n>\n>Once you measure the same spin direction at A and B, the measurement\n>gives the same result at A and B. There is no more uncertainty about this\n>in QM. There is no h-related bound for the accuracy of this experiment.\n\nI\'ll absolutely bet that isn\'t true. In any real experiment there will\nbe measurements that give the \'wrong\' result. No detector is perfect and\nno environment is noise-free. Surely, ultimately, any measurement must\nhave a limiting accuracy related to h?\n\n\n--\nOz\nThis post is worth absolutely nothing and is probably fallacious.\n\nUse oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].\nBTOPENWORLD address has ceased. DEMON address has ceased.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes
>
>
>
>
>"Oz" <oz@farmeroz.port995.com> schrieb
>> Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes
>> >The whole point is a different one. We really have some black box.
>> >This black box allows to do very strange things. Unfortunately it
>> >does not allow us to build an FTL phone. But there exists no realistic
>> >explanation of what is inside the black box which does not use some
>> >sort of hidden FTL information transfer.
>>
>> Quite. The crux of the problem.
>> One viable solution is to claim that
>> FTL is perfectly possible
>> BUT ONLY inside the black box.
>> One should then examine the box.
>> What is strange about it?
>
>Nothing (if I understand this in a very methaporical way).
>
>Translated into our physics "looking into the box" can be understood
>as developing realistic hidden variable theories. We know from Bell
>that they need a preferred frame. Therefore we have to develop
>theories with preferred frames.

That would seem to be a viable thing to examine.
It either works out, or it doesn't.
But if its never seriously examined, who knows?

Its unfortunate that even mentioning this allocates you a significant
crank index, no matter how serious and how well supported by the maths
it is.

>That's what I'm doing. I wonder why I'm almost alone.

ER, because nobody will discuss it seriously with you?

>> Now that's an extraordinary statement when you think about it. Why
>> *should* everyone agree about something? Furthermore there is an implied
>> connection between the whole universe (who all agree about the event)
>> and the event.
>
>The fact is that not everyone agrees about it - there are people who
>believe in the remaining loopholes like detector efficiency. Thus, there is
>no
>"should". That almost everybody agrees about it is a nontrivial fact.

That's not what I meant.
What I meant was that, by definition, and event is (in relativity)
something everyone, everywherewhen, agrees on. If two people see
different results to some experiment then its not an event. What it is
is noise (or something similar).

>> QM, however, says we can never be quite sure.
>
>QM has never told me such things.

Eh?

>> Different observers will
>> in fact measure different results for an event, mediated by h. To me
>> this strongly suggests that an event is probabilistic. That is an event
>> is simply something with a very high probability of occurring (and I'm
>> talking about near certainty here).
>
>Once you measure the same spin direction at A and B, the measurement
>gives the same result at A and B. There is no more uncertainty about this
>in QM. There is no h-related bound for the accuracy of this experiment.

I'll absolutely bet that isn't true. In any real experiment there will
be measurements that give the 'wrong' result. No detector is perfect and
no environment is noise-free. Surely, ultimately, any measurement must
have a limiting accuracy related to h?

--
Oz
This post is worth absolutely nothing and is probably fallacious.

Use oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].
BTOPENWORLD address has ceased. DEMON address has ceased.

Oz

Nov14-04, 11:59 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n\nBlake Winter <blake.winter@houghton.edu> writes\n\n>Lambda is the "hidden variable" which determines the absolute\n>orientation of each photon - the idea is that as soon as the photons\n>are created, they have a an actual value for their polarization.\n\nAhh. OK.\nSo they are NOT in a mutual superposition, that is each has (in effect)\nboth polarisations and one is (to put it naively) anti-the other (in\npolarisation terms).\n\n>The reason why this is necessary is that otherwise, one would not end\n>up with "polarization conservation".\n\nOK.\n\nShould we not consider them as a single object, in effect a\nsuperposition of both polarisations?\n\n>> Ok. But then you immediately remove any entanglement. You effectively\n>> say these are two isolated and unconnected photons, which they cannot\n>> be, don\'t you?\n\n>Right, that\'s why hidden variables are disproven by the violations of\n>Bell inequalities - at least, naive hidden variables are. By naive\n>hidden variables, I mean that if we still want hidden variables we\'ve\n>got to either violate causality (either by violations of relativistic\n>locality, or else by having closed causal loops), or else we\'ve got to\n>have a nontrivial topology.\n\nOK, so the way I see it would count as a \'nontrivial topology".\nI would be content with that.\nSo Ilja\'s approach is in effect to superimpose a \'nontrivial topology\'\non spacetime. In effect the equivalent of a wormhole joining these two\nparticles, giving (internally) apparent FTL transmission.\n\nThis would be compatible with my current viewpoint that is each particle\ncarries its own spacetime with it, and the macroscopic spacetime we\nobserve is one generated by \'lots of particles interacting\'. That puts\nmomentum as the essence of a particle and distance as, in effect, a\nderived quantity.\n\n>Interestingly Markopoulou and Smolin have written a paper on\n>generating quantum theory from a hidden variable theory involving\n>graphs - they end up with the nonrelativistic Schrodinger equation in\n>a suitable limit as a first or second order approximation, I forget\n>which. www.arxiv.org/abs/gr-qc/0311059. Although it is based on the\n>spin networks from LQG, it doesn\'t involve any quantum superposition\n>of spin networks at all - all the quantum mechanics comes out of a\n>realistic theory. In a sense, the way the spin networks are embedded\n>can result in "nontrivial topologies" on the embedded manifold,\n>thereby allowing seemingly nonlocal correlations.\n\nSounds scary...\n\n>The problem, which no one has ever been able to entirely clear up for\n>me (perhaps simply because I\'m dense), is that in quantum theory you\n>need nonlocality too. Gribben\'s "In Search of Schrodinger\'s Kittens"\n>book explains why this is exactly.\n>The basic idea is this. You start in QM with two photons which are in\n>a superposition of orientations BUT each possible orientation of\n>photon A has photon B in the same orientation. Then, when photon A is\n>measured, it turns out that somehow photon B has to "know" which\n>eigenvalue was returned when it was measured, or else Bell\'s\n>inequality couldn\'t be violated. There are some good descriptions on\n>the web, but I think Gribben\'s book is the best place to look. Most\n>people are at peace with this, because it doesn\'t actually change the\n>local probabilities, because you don\'t actually know what the\n>measurement at the other photon was. This is the famous result that\n>you can\'t send information using EPR effects.\n\nThat seems to be ilja\'s argument, but coming from a different direction.\n\n>Ultimately I\'m of the opinion that we shouldn\'t think that we\n>understand this yet - no doubt there is a sort of way to understand\n>it, but we ought not to be naive about it. Most likely the effect\n>will become better understood\n\nCurrently, I\'d just settle for a clear explanation of what the effect\nactually IS!\n\n--\nOz\nThis post is worth absolutely nothing and is probably fallacious.\n\nUse oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].\nBTOPENWORLD address has ceased. DEMON address has ceased.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Blake Winter <blake.winter@houghton.edu> writes

>\Lambda is the "hidden variable" which determines the absolute
>orientation of each photon - the idea is that as soon as the photons
>are created, they have a an actual value for their polarization.

Ahh. OK.
So they are NOT in a mutual superposition, that is each has (in effect)
both polarisations and one is (to put it naively) anti-the other (in
polarisation terms).

>The reason why this is necessary is that otherwise, one would not end
>up with "polarization conservation".

OK.

Should we not consider them as a single object, in effect a
superposition of both polarisations?

>> Ok. But then you immediately remove any entanglement. You effectively
>> say these are two isolated and unconnected photons, which they cannot
>> be, don't you?

>Right, that's why hidden variables are disproven by the violations of
>Bell inequalities - at least, naive hidden variables are. By naive
>hidden variables, I mean that if we still want hidden variables we've
>got to either violate causality (either by violations of relativistic
>locality, or else by having closed causal loops), or else we've got to
>have a nontrivial topology.

OK, so the way I see it would count as a 'nontrivial topology".
I would be content with that.
So Ilja's approach is in effect to superimpose a 'nontrivial topology'
on spacetime. In effect the equivalent of a wormhole joining these two
particles, giving (internally) apparent FTL transmission.

This would be compatible with my current viewpoint that is each particle
carries its own spacetime with it, and the macroscopic spacetime we
observe is one generated by 'lots of particles interacting'. That puts
momentum as the essence of a particle and distance as, in effect, a
derived quantity.

>Interestingly Markopoulou and Smolin have written a paper on
>generating quantum theory from a hidden variable theory involving
>graphs - they end up with the nonrelativistic Schrodinger equation in
>a suitable limit as a first or second order approximation, I forget
>which. www.arxiv.org/abs/http://www.arxiv.org/abs/gr-qc/0311059. Although it is based on the
>spin networks from LQG, it doesn't involve any quantum superposition
>of spin networks at all - all the quantum mechanics comes out of a
>realistic theory. In a sense, the way the spin networks are embedded
>can result in "nontrivial topologies" on the embedded manifold,
>thereby allowing seemingly nonlocal correlations.

Sounds scary...

>The problem, which no one has ever been able to entirely clear up for
>me (perhaps simply because I'm dense), is that in quantum theory you
>need nonlocality too. Gribben's "In Search of Schrodinger's Kittens"
>book explains why this is exactly.
>The basic idea is this. You start in QM with two photons which are in
>a superposition of orientations BUT each possible orientation of
>photon A has photon B in the same orientation. Then, when photon A is
>measured, it turns out that somehow photon B has to "know" which
>eigenvalue was returned when it was measured, or else Bell's
>inequality couldn't be violated. There are some good descriptions on
>the web, but I think Gribben's book is the best place to look. Most
>people are at peace with this, because it doesn't actually change the
>local probabilities, because you don't actually know what the
>measurement at the other photon was. This is the famous result that
>you can't send information using EPR effects.

That seems to be ilja's argument, but coming from a different direction.

>Ultimately I'm of the opinion that we shouldn't think that we
>understand this yet - no doubt there is a sort of way to understand
>it, but we ought not to be naive about it. Most likely the effect
>will become better understood

Currently, I'd just settle for a clear explanation of what the effect
actually IS!

--
Oz
This post is worth absolutely nothing and is probably fallacious.

Use oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].
BTOPENWORLD address has ceased. DEMON address has ceased.

Oz

Nov14-04, 11:59 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n\nThomas Trotter <thomastrotter2005@juno.com> writes\n>Oz <oz@farmeroz.port995.com> wrote in message news:<T\\$Q7nbQaXIkBFw\\$T@farmeroz.po\n>rt995.com >...\n>> Patrick Van Esch <vanesch@ill.fr> writes\n>> >Bell is making up a "local realistic" model of the entangled pair. If\n>> >you say that the photons HAD a specific polarisation direction lambda,\n>> >then you have to be able to write it down (even if you don\'t know what\n>> >it is). That\'s what he does.\n>>\n>> I have terrible trouble getting anyone to actually answer my question on\n>> this clearly. The question is:\n>>\n>> Is lambda wrt some apparatus, or wrt the other photon?\n>\n>Lambda is the common emission-produced polarization\n>*direction* of a pair of photons emitted by the same\n>atom wrt the setting of some apparatus -- specifically,\n>the transmission axes of the crossed linear polarizers.\n\nAre you saying that lambda is obtained by:\n\n1) Measuring an angle of one particle with one apparatus to give L1.\n\n2) Measuring and angle of the other particle with another apparatus to\ngive L2.\n\n3) Relating angles on one apparatus to the other apparatus to give L3.\n\nThen comparing L1, L3 with L2?\n\n> This is a critical difference. The former says the angle is fixed from\n>> the start as seen by the rest of the universe, the latter simply says\n>> they always have a set relationship.\n>\n>Exactly.\n\nYes, but which is it?\n\nIs the assumption that each particle sets off with some lambda, or is it\nthat the angle between the particles is always lambda?\n\n>>\n>> >Quantum mechanics doesn\'t take this approach. It writes down that the\n>> >state is:\n>> >\n>> >|psi> = 1/sqrt(2) { |n+>|n-> - |n->|n+ }\n>> >\n>> >with n just any direction. So our state is completely "unoriented"\n>> >which is not possible in a realistic model.\n>>\n>> Isn\'t this simply saying that we cannot know the absolute orientation,\n>> but do know the relative orientation?\n>\n>Yes.\n\nOK. Now, given this situation I would immediately want to do the\nfollowing experiment. I would select some transition that emits two\nentangled particles, let us say U (for up) and D (for down).\n\nI would run one through a series of crossed polarisers so as to rotate\none by 180 degrees.\n\nNow I have three contrasting outcomes:\n\n1) The particles are now always measured as both U or both D.\n2) The particles are always still measured as one U and one D.\n3) I never detect two particles.\n\nI have been trying to find out which of these is actually observed for\nyears, nobody will ever tell me.\n\n\n--\nOz\nThis post is worth absolutely nothing and is probably fallacious.\n\nUse oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].\nBTOPENWORLD address has ceased. DEMON address has ceased.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Thomas Trotter <thomastrotter2005@juno.com> writes
>Oz <oz@farmeroz.port995.com> wrote in message news:<T$Q7nbQaXIkBFw$T@farmeroz.po
>rt995.com>...
>> Patrick Van Esch <vanesch@ill.fr> writes
>> >Bell is making up a "local realistic" model of the entangled pair. If
>> >you say that the photons HAD a specific polarisation direction \lambda,
>> >then you have to be able to write it down (even if you don't know what
>> >it is). That's what he does.
>>
>> I have terrible trouble getting anyone to actually answer my question on
>> this clearly. The question is:
>>
>> Is \lambda wrt some apparatus, or wrt the other photon?
>
>\Lambda is the common emission-produced polarization
>*direction* of a pair of photons emitted by the same
>atom wrt the setting of some apparatus -- specifically,
>the transmission axes of the crossed linear polarizers.

Are you saying that \lambda is obtained by:

1) Measuring an angle of one particle with one apparatus to give L1.

2) Measuring and angle of the other particle with another apparatus to
give L2.

3) Relating angles on one apparatus to the other apparatus to give L3.

Then comparing L1, L3 with L2?

> This is a critical difference. The former says the angle is fixed from
>> the start as seen by the rest of the universe, the latter simply says
>> they always have a set relationship.
>
>Exactly.

Yes, but which is it?

Is the assumption that each particle sets off with some \lambda, or is it
that the angle between the particles is always \lambda?

>>
>> >Quantum mechanics doesn't take this approach. It writes down that the
>> >state is:
>> >
>> >|\psi> = 1/\sqrt(2) { |n+>|n-> - |n->|n+ }>> >
>> >with n just any direction. So our state is completely "unoriented"
>> >which is not possible in a realistic model.
>>
>> Isn't this simply saying that we cannot know the absolute orientation,
>> but do know the relative orientation?
>
>Yes.

OK. Now, given this situation I would immediately want to do the
following experiment. I would select some transition that emits two
entangled particles, let us say U (for up) and D (for down).

I would run one through a series of crossed polarisers so as to rotate
one by 180 degrees.

Now I have three contrasting outcomes:

1) The particles are now always measured as both U or both D.
2) The particles are always still measured as one U and one D.
3) I never detect two particles.

I have been trying to find out which of these is actually observed for
years, nobody will ever tell me.

--
Oz
This post is worth absolutely nothing and is probably fallacious.

Use oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].
BTOPENWORLD address has ceased. DEMON address has ceased.

Ilja Schmelzer

Nov16-04, 02:19 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Blake Winter" <blake.winter@houghton.edu> schrieb\n> Oz <oz@farmeroz.port995.com> wrote\n> Right, that\'s why hidden variables are disproven by the violations of\n> Bell inequalities - at least, naive hidden variables are.\n\nSorry, Bohmian mechanics is a hidden variable theory. And a\nquite simple one. That simple that Einstein has considered it\ntoo simple.\n\n> By naive\n> hidden variables, I mean that if we still want hidden variables we\'ve\n> got to either violate causality (either by violations of relativistic\n> locality, or else by having closed causal loops), or else we\'ve got to\n> have a nontrivial topology.\n\nBohmian mechanics has, indeed, hidden causal influences in a hidden\npreferred frame.\n\nOf course, some people do not like hidden preferred frames and hidden\nvariables in general. But sometimes the argumentation is quite absurd:\nThe preferred frame is rejected (in special relativity) because it is\nhidden.\nBohmian mechanics is rejected because it has a preferred frame, which\ndoes not agree with relativistic ideology. And hidden variable theories\nin general are rejected for the same reason, because the violation of Bell\'s\ninequality proves that they need a preferred frame.\n\nBut if we combine these arguments, it appears that the argument against\nhidden variables is that it uses hidden variables. Because there is no\nother argument against the preferred frame.\n\n> Most\n> people are at peace with this, because it doesn\'t actually change the\n> local probabilities, because you don\'t actually know what the\n> measurement at the other photon was. This is the famous result that\n> you can\'t send information using EPR effects.\n\nI don\'t understand why people consider this as an important argument.\nLast not least, every observation which leaves only two explanations\n(A->B or B->A) obviously proves that there exists FTL causal influences.\n\nBut, on the other hand, every effect which leaves these two explanation\nhas the same property: It cannot be used for information transfer.\nIndeed, an application for information transfer A->B would be in\ncontradiction with the possible explanation B->A and reverse.\n\nThus, accepting this argument means preserving Einstein causality\neven in a situation where it can be logically proven to be false.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Blake Winter" <blake.winter@houghton.edu> schrieb
> Oz <oz@farmeroz.port995.com> wrote
> Right, that's why hidden variables are disproven by the violations of
> Bell inequalities - at least, naive hidden variables are.

Sorry, Bohmian mechanics is a hidden variable theory. And a
quite simple one. That simple that Einstein has considered it
too simple.

> By naive
> hidden variables, I mean that if we still want hidden variables we've
> got to either violate causality (either by violations of relativistic
> locality, or else by having closed causal loops), or else we've got to
> have a nontrivial topology.

Bohmian mechanics has, indeed, hidden causal influences in a hidden
preferred frame.

Of course, some people do not like hidden preferred frames and hidden
variables in general. But sometimes the argumentation is quite absurd:
The preferred frame is rejected (in special relativity) because it is
hidden.
Bohmian mechanics is rejected because it has a preferred frame, which
does not agree with relativistic ideology. And hidden variable theories
in general are rejected for the same reason, because the violation of Bell's
inequality proves that they need a preferred frame.

But if we combine these arguments, it appears that the argument against
hidden variables is that it uses hidden variables. Because there is no
other argument against the preferred frame.

> Most
> people are at peace with this, because it doesn't actually change the
> local probabilities, because you don't actually know what the
> measurement at the other photon was. This is the famous result that
> you can't send information using EPR effects.

I don't understand why people consider this as an important argument.
Last not least, every observation which leaves only two explanations
(A->B or B->A) obviously proves that there exists FTL causal influences.

But, on the other hand, every effect which leaves these two explanation
has the same property: It cannot be used for information transfer.
Indeed, an application for information transfer A->B would be in
contradiction with the possible explanation B->A and reverse.

Thus, accepting this argument means preserving Einstein causality
even in a situation where it can be logically proven to be false.

Ilja

Ilja Schmelzer

Nov16-04, 02:56 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"seratend" <ser_monmail@yahoo.fr> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> > Translated into our physics "looking into the box" can be understood\n> > as developing realistic hidden variable theories. We know from Bell\n> > that they need a preferred frame. Therefore we have to develop\n> > theories with preferred frames.\n> >\n> > That\'s what I\'m doing. I wonder why I\'m almost alone.\n\n> I understand that you are saying that hidden variable theories need a\n> preferred frame (I have suppressed the realistic : ).\n> I have not considered this problem but I am surprised as almost all\n> (the one\'s I know) such theories mimics the QM so adding special\n> relativity should not be a problem. Would you mind giving us more\n> details on the preferred frame problem?\n\nA preferred frame we need if we want a realistic interpretation of\nQM. The simplest choice is Bohmian mechanics.\n\n"Realistic" means, roughly speaking, what we have to make a\nhypothesis about reality. The realistic theory has an ontology,\na description of reality. In Bohmian mechanics, the reality\nconsists of two parts, the wave function Psi(q) and the state of the\nuniverse q. Except in the simplest case of one-particle-theory\nq is not the location in space. It is the general configuration of\nthe whole universe, including positions of all particles. Therefore\nPsi(q) is not a local function. But it is local in time, a function\nPsi(q,t). As well, the state of the universe is a function of t q(t).\n\nThe problem with relativistic QFT is that it extends the "shut up and\ncalculate" ideology. It does not care at all about an explanation\nfor the observed probabilities - being able to compute them is\nsufficient. This is not an explanation. Compare with the theory:\n"We hear voices every Friday the 13. at midnight in the castle."\nIt is a falsifiable theory. But, let\'s assume it is supported by\nobservation. Would you accept such a successful prediction\nas an explanation? Certainly not. We would search for other\nexplanations. As physicists we would search for physical\nexplanations. In the worst case, if no physical explanation is\npossible, we would take into consideration explanations using\nghosts, magic, UFOs and so on. That would be really strange,\nonly the last resort, but at least explanations.\n\nInstead, the position "We do not need ghosts for\nexplanation. We have predicted that we hear the voices again\nthis Friday, the prediction was successful, fine. No problem\nat all." would be much more strange, not?\n\nNow, if we reject the "shut up and calculate" ideology as not\ngiving a realistic explanation, we have BM which has a preferred\nframe as a realistic theory.\n\nAnd we have Bell\'s inequality. It shows that for every pair of events\nA, B, such that measurements and observations at these events\nallow to violate Bell\'s inequality, we have a (hidden) causal\ninfluence A->B or B->A. Given the predictions of QM, this holds\nfor almost every pair of events. Now, let\'s consider the question what\nreally happens, A->B or B->A. We cannot measure this. It is\na hidden variable. Nonetheless, in a realistic causal theory we\npostulate that there are no closed causal loops. Thus, A->B xor\nB->A, and the realistic theory has to make a hypothesis which of\nthe claims is the correct one. Combining this with the usual causality\nrelated with light cones we can see that every such consistent set of\nhypotheses about these hidden directions of the causal influence\ndefines a preferred foliation.\n\nNote that this discussion about EPR, Bell and preferred frames is\nonly one part of my general argumentation in favour of a preferred frame.\nIt includes many other, completely independent parts.\n\n* An ether theory of gravity with preferred frame, GR limit, EEP\nderived from ether axioms. It is related with the three-dimensional\ngeometric interpretation of gravity known as ADM decomposition\nin harmonic coordinates.\n\n* Considerations of quantum gravity problems (problem of time,\ninformation loss problem, noncovariance of the renormalization\nof the energy-momentum tensor on curved background,\nquantization of harmonic gauge as constraints).\n\nSee gr-qc/0205035, gr-qc/0001101.\n\nAnd, from the paper I\'m working on now:\n\n* A three-dimensional geometric interpretation of the SM\nfermions as sections of A(3)xCx/\\(R^3)\n\n* A related lattice theory with configuration space\nA(3)(Z^3) or phase space CxA(3)(Z^3) where the\nadditional factor /\\ of the continuous limit appears as\na consequence of the species doubling effect.\n\n* The hypothesis that the gauge fields appear as\nlattice defects of the lattice A(3)(Z^3).\n\nSee hep-th/0310241 (which gives the geometric interpretation\nTx/\\x/\\(R^3) which is close to A(3)xCx/\\(R^3)),\nhep-lat/0311009 (which defines strong interaction on\nan octet /\\(Z^3) which gives /\\x/\\(R^3)) for earlier versions.\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"seratend" <ser_monmail@yahoo.fr> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> > Translated into our physics "looking into the box" can be understood
> > as developing realistic hidden variable theories. We know from Bell
> > that they need a preferred frame. Therefore we have to develop
> > theories with preferred frames.
> >
> > That's what I'm doing. I wonder why I'm almost alone.

> I understand that you are saying that hidden variable theories need a
> preferred frame (I have suppressed the realistic : ).
> I have not considered this problem but I am surprised as almost all
> (the one's I know) such theories mimics the QM so adding special
> relativity should not be a problem. Would you mind giving us more
> details on the preferred frame problem?

A preferred frame we need if we want a realistic interpretation of
QM. The simplest choice is Bohmian mechanics.

"Realistic" means, roughly speaking, what we have to make a
hypothesis about reality. The realistic theory has an ontology,
a description of reality. In Bohmian mechanics, the reality
consists of two parts, the wave function \Psi(q) and the state of the
universe q. Except in the simplest case of one-particle-theory
q is not the location in space. It is the general configuration of
the whole universe, including positions of all particles. Therefore
\Psi(q) is not a local function. But it is local in time, a function
\Psi(q,t). As well, the state of the universe is a function of t q(t).

The problem with relativistic QFT is that it extends the "shut up and
calculate" ideology. It does not care at all about an explanation
for the observed probabilities - being able to compute them is
sufficient. This is not an explanation. Compare with the theory:
"We hear voices every Friday the 13. at midnight in the castle."
It is a falsifiable theory. But, let's assume it is supported by
observation. Would you accept such a successful prediction
as an explanation? Certainly not. We would search for other
explanations. As physicists we would search for physical
explanations. In the worst case, if no physical explanation is
possible, we would take into consideration explanations using
ghosts, magic, UFOs and so on. That would be really strange,
only the last resort, but at least explanations.

Instead, the position "We do not need ghosts for
explanation. We have predicted that we hear the voices again
this Friday, the prediction was successful, fine. No problem
at all." would be much more strange, not?

Now, if we reject the "shut up and calculate" ideology as not
giving a realistic explanation, we have BM which has a preferred
frame as a realistic theory.

And we have Bell's inequality. It shows that for every pair of events
A, B, such that measurements and observations at these events
allow to violate Bell's inequality, we have a (hidden) causal
influence A->B or B->A. Given the predictions of QM, this holds
for almost every pair of events. Now, let's consider the question what
really happens, A->B or B->A. We cannot measure this. It is
a hidden variable. Nonetheless, in a realistic causal theory we
postulate that there are no closed causal loops. Thus, A->B xor
B->A, and the realistic theory has to make a hypothesis which of
the claims is the correct one. Combining this with the usual causality
related with light cones we can see that every such consistent set of
hypotheses about these hidden directions of the causal influence
defines a preferred foliation.

Note that this discussion about EPR, Bell and preferred frames is
only one part of my general argumentation in favour of a preferred frame.
It includes many other, completely independent parts.

* An ether theory of gravity with preferred frame, GR limit, EEP
derived from ether axioms. It is related with the three-dimensional
geometric interpretation of gravity known as ADM decomposition
in harmonic coordinates.

* Considerations of quantum gravity problems (problem of time,
information loss problem, noncovariance of the renormalization
of the energy-momentum tensor on curved background,
quantization of harmonic gauge as constraints).

See http://www.arxiv.org/abs/gr-qc/0205035, http://www.arxiv.org/abs/gr-qc/0001101.

And, from the paper I'm working on now:

* A three-dimensional geometric interpretation of the SM
fermions as sections of A(3)xCx/\(R^3)

* A related lattice theory with configuration space
A(3)(Z^3) or phase space CxA(3)(Z^3) where the
additional factor /\ of the continuous limit appears as
a consequence of the species doubling effect.

* The hypothesis that the gauge fields appear as
lattice defects of the lattice A(3)(Z^3).

See http://www.arxiv.org/abs/hep-th/0310241 (which gives the geometric interpretation
Tx/\x/\(R^3) which is close to A(3)xCx/\(R^3)),
http://www.arxiv.org/abs/hep-lat/0311009 (which defines strong interaction on
an octet /\(Z^3) which gives /\x/\(R^3)) for earlier versions.

Ilja

rof@maths.tcd.ie

Nov16-04, 02:56 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n>> ><rof@maths.tcd.ie> schrieb\n>> >Yep. The construction of the M_n is of this type. But the X_n\\M_n,\n>> >the inner part of the moon, is not perceived. Even if I have some\n>> >perception M_moon of the surface of the moon and some vague\n>> >theoretical ideas what may be inside.\n>>\n>> Ok; I see what you\'re saying. Conceptual information, such as\n>> "Iron is the stablest nucleus", is not represented spatially,\n>> and it is most likely ideas like this which guide your thinking\n>> about what is going on inside planets. I don\'t disagree with this,\n>> but the sense in which I use the X_n is to refer explicitly to the\n>> procedure which we use to construct a "world around us" from\n>> sensation, and which is explicitly spatial.\n\n>My point is that the construction is not complete.\n\n>We have some construction E_n -> M_n but we do not\n>construct a complete world X_n.\n\nI have lost track of what you mean by M_n; it doesn\'t seem to\nbe in the same sense as I introduced it. I agree, though\nthat the construction is not complete - when I see a wall\nin front of me, I don\'t perceive what\'s behind the wall.\nThis isn\'t a problem, though.\n\n>Moreover, it is not given that the construction M_n is spatial.\n>If we construct following Bohmian mechanics we construct\n>some Psi(q) which is not spatial but defined on the configuration\n>space.\n\nSure, but the reason that it\'s an explanation that one feels\ncomfortable with is because it\'s easy to visualise. One imagines\na wave moving in some space (here, it\'s configuration space),\nand that is something that we\'re used to. The state of affairs (the\nwavefunction, Psi, ignoring for a moment the particle positions)\nis represented as being composed of several mini-states of affairs,\nPsi(x), each at a different position. One becomes a lot less comfortable\nwith it once one reflects on the fact that the place where the\nwavefunction lives is not real space at all, but an abstract\nconstruction called configuration space whose dimension is 3N.\n\nWe can make it make a little more sense by noticing that the\nspace over which the wavefunction is defined is the same as\nthe space over which probability distributions (of generic\nstates of affairs of the system, or configurations) are\nnaturally defined. Information that we have might have about\nthe configuration of the system at a given time would naturally\nbe expressed as a real function on this space, and a description\nwhich included the instantaneous rate of change of this\nprobability distribution would naturally be expressed as\ntwo real functions (or one complex function) on this space.\n\n>> >Sorry, but you cannot derive if observers will "see" space.\n>> >"Seeing" space means developing a certain realistic theory,\n>> >may be implicit (as a hidden prejudice, instinct, created by\n>> >evolution).\n\n>> Evolution is another can of worms, best left unopened here, as is\n>> the connection between the brain and conscious experience. Animals\n>> see space, but they don\'t develop ontological theories. ... It\'s only\n>> theoretical physicists\n>> who have to face the dilemma of whether to stick the "reality" sticker\n>> onto the wavefunction or the observables, because the urge to stick\n>> it somewhere is so strong.\n\n>I would not argue that there is some strong urge to develop ontological\n>theories. Ontological theories have been successful, there is no reason\n>to give them up.\n\nThere is no success of any theory which can be ascribed to its\nontological character.\n\n>With Bohmian mechanics they are successful today\n>too.\n\nOnly insofar as Bohmian mechanics has been able to reproduce\nthe predictions of a theory (Copenhagen-style quantum mechanics) which\nwasn\'t ontological. In so far as Bohmian mechanics has not been\nable to follow quantum mechanics (pair creation, for example),\nit has not been successful, although I recognise that there has\nbeen progress in this direction.\n\n>> >A believes that there are\n>> >some X_n in reality. This is an ontological theory proposed by A.\n>>\n>> A may think it\'s on ontological theory, but A is confusing himself.\n>> Imagine there is a digital machine, which takes 1\'s and 0\'s as\n>> input. You feed in some sequence of 1\'s and 0\'s, generated by\n>> some algorithm. The machine proceeds to construct a Markov model\n>> to predict the sequence. It is successful to some degree - most\n>> of the time it correctly predicts the next input. Has it discovered\n>> reality? Is its Markov model an ontological theory?\n\n>Once it predicts successfully, there is some probability that the\n>algorithm has been guessed correctly. In other words, that\n>the algorithm used to create the input sequence and the\n>algorithm used by the machine to predict the input sequence are\n>in some sense equivalent.\n\nBut this doesn\'t answer either of the above questions. There\'s\nalso the problem of the vagueness of the "in some sense" which\nis inevitably reached by proponents of realism in this situation.\n\nThe grand project of ontology is not to guess an algorithm which\npredicts future observations (which is, in fact, the only thing\nwe can do), but to determine what it is that really exists.\nIf a machine develops an algorithm to correctly predict whether\nthe next input will be a 1 or a 0, the most one can say about\nit is that that is what it has achieved. To say that it has\ndiscovered the ontological basis of reality, one has to radically\nchange what one means by "discovered the ontological basis of\nreality."\n\nOne who, observing traffic lights, eventually says "A repeating\npattern of green, then yellow and then red," has hardly reached the\nultimate goal that realists aspire to. He might imagine a circle\nwith a little point on it, moving consecutively through 120-degree\narcs labelled "Green", "Yellow" and "Red" (a Bohmian traffic\nlight, if you will), and declare it to be an ontological theory,\nbut such an individual would have to be regarded as confused.\n\nThis is important, because ontology stands or falls on this\npoint. We are not in a very different position to the digital\nmachine, or the traffic-light observer, with our sensations as the\ninput. Those who want to claim that we are engaged in the business\nof discovering the fundamental ontological consituents of the\nuniverse must make a rigorous link between this process of predicting\nfuture inputs and what they regard as reality.\n\nAlan Sokal, in his "Defense of a Modest Scientific Realism",\nexamines these issues a little bit, but, since he too is\nwedded to realism, doesn\'t quite reach the correct answer,\nbut instead ends with:\n"Since no existing theory purports to be a final theory, there\nis no reason to consider it as literally true or to worry too\nmuch about whether the entities it postulates `really exist\'.\nOr rather, when worrying about whether the unobservable entities\nof a given theory `really exist\', it is important to distinguish\nexistence _as a fundamental constituent of the universe_ from\nexistence _in some coarse-grained sense_. It is a reasonable guess\nthat _none_ of the theoretical entities in our present-day\ntheories are truly fundamental, and that _all_ of the theoretical\nentities in our present-day well-confirmed theories will maintain\nsome status as derived entities in future theories." (His italics)\n\nIn the end, he opts for saying that things exist "in some coarse-grained\nsense", which means that he wants to ascribe to them some property\nwhich he feels is related to existence or reality, but which is so\nvague that he cannot tell us what it is, and, lacking any concrete\nor even specific thing to say, must instead appeal to our sympathy\nto his realist position in order to convince us to accept what he\nis saying as a valid defense of realism.\n\nNotice that he also explicitly subscribes to the view that,\nif an entity from one theory appears as a derived or emergent\nthing in a more mature theory, then that entity cannot be a\n"fundamental constituent of the universe." What he leaves open,\nbut implicitly suggests, is that, in a final theory, which\nmakes perfect predictions and has no further refinement, we\nmight actually consider the entities with which it deals to\nbe fundamental constituents of the universe which really exist.\nThis is, of course, what the whole exercise was intended to\ndeduce, and which it has failed to deduce, so he ends by\nsurreptitiously implying it.\n\nNeither the positing of a kind of equivalence between algorithms,\nnor the relegation of actual existence to existence in some coarse-grained\nsense, actually bridge the gap between ontological existence and the\nmathematical regularities of observed phenomena described by physical\ntheories.\n\nIt is very important in these matters not to decide that we like\nrealism and then accept as valid an otherwise insufficient\nargument in favour of it. It is also important not to decide\nthat it is merely a matter of opinion, or indeed to be influenced\nby opinion at all, since to allow an opinion to direct one\'s\nunderstanding is nothing more than to believe what one likes\nto believe. Thus, if we find that the arguments with which\nwe convince ourselves that realism is necessary amount to\n"In my opinion, realism should be considered a part of logic,"\nthen we must go in search of the reason why that opinion\nwas adopted, and reject it unless there is a rigorous basis\nfor it. As it stands, it looks quite like a demand that\nrealism be adopted as an axiom; that is, it looks like\na dogmatic assertion of a statement which cannot be deduced\nby reason.\n\n>> The situation that we find ourselves in is that we know that\n>> we cannot perfectly predict the results of measurements (in\n>> EPR experiments, knowing the outcome of a spin measurement in\n>> advance would allow signalling faster than light). This means\n>> that our Markov model cannot perfectly predict the future,\n>> even in principle. We\'re left with a merely statistical prediction,\n>> although we may conjure up fruitless imaginations about what caused\n>> the result of a particular experiment to be this or that.\n\n>That\'s not really a problem. Once we don\'t know the complete\n>input data used by the algorithm we cannot predict everything.\n>Nonetheless it is possible that the algorithms we guess and the\n>algorithms used in the "fabric of reality" are equivalent.\n\nThe implicit assertion here is that the fabric of reality\nis implementing some kind of algorithm (recall that an algorithm\nis a list of instructions for computing one thing from another).\nThe equivalence here is again, "in some sense", and has the particular\ndifficulty that one algorithm (the one used by a human) has, as its\noutput, patterns of sensations (as a prediction of future input).\n\nI doubt that an algorithm which has sensations as output is one\nthat most realists would accept as an example of something which\ncould qualify as a fundamental constituent of reality, since the\nthings with which it deals (sensations) are purely mental.\n\n>> Specifically, Bohmian mechanics is an attempt to imagine that\n>> the technique which we use to order our perceptions (representing\n>> a single state of affairs as lots of little states of affairs,\n>> each in its own position, in configuration space, for example),\n>> should still apply even in the absence of any sensible data\n>> to represent.\n\n>BM is not a "technique to order perceptions".\n\nAgreed. I would never suggest that it is.\n\n>It is an ontological theory.\n\nLike the traffic light theory with the circle and the dot.\n\n>> Whether or not a particle even exists can depend on the motion of\n>> the observer (cf. the Unruh effect), so it seems strange to put that\n>> sticky reality label onto the particles and their positions in space.\n\n>I don\'t.\n\nBut Bohmian mechanics does, and that is the context in which I made\nthe remark above.\n\n>The only requirement is that the ontological theory\n>should define something named "reality".\n\nI notice that there is a misunderstanding of the role of definition\namong modern scientists. When one defines an object, one at the same\ntime generates that object. I may, for example, define a vector\nspace as a collection of objects which may be added together and\nmultiplied by the elements of some field, and thereafter I can\nbe completely certain when I make certain judgments about vector\nspaces, because they are not objects which lie outside of me,\nand which I must study through perception and experimentation,\nbut are free inventions of my mind, which have exactly those\nproperties which I gave to them when I defined them. I do not,\nfor example, have the same freedom to define a duck, since\nducks have many properties which must be discovered by examination,\nand are not free inventions of my mind. I may, of course, choose\nto use the word "duck" to refer to something else apart from\nan actual duck, but that is unimportant here. In this case,\nare you asserting that "reality" is something which can be\ndefined, or that its definition can change from theory to theory?\n\n>Some mathematical\n>objects should be given the label "real". And for these really\n>existing objects we can use classical logic and probability theory.\n\nI am always completely in favour of using classical logic and\nprobability theory.\n\n>> ... declaring that the model is "out there", beyond\n>> the mind of the person who thinks it, governing the evolution\n>> of the world, is like saying that, since 1 typically follows\n>> 0 in a certain binary sequence, the rule "01" is out there, guiding\n>> the sequence.\n\n>Sorry, but we use physical theories to predict things which do not\n>happen in our mind, but in reality. (Except you are a pure solipsist.)\n>Of course, I feel free to say that there is a law of gravity out there.\n\n>I may be wrong about the details of the law of gravity. But it is\n>a hypothesis (in my mind) about what is out there.\n\nSo you begin to understand; the law of gravity, as you\nknow it, is not something which exists independently of you, but a\nhypothesis which relies on you for its existence.\n\nThe real difficulty here is (as I mentioned before) the manifold\nuse of the word "real". On the one hand, it is used to refer to\nwhat is actual, rather than merely referred to (real pain versus\nimagined pain). If the use of the word "real" were restricted to\nthis, no confusion would arise. However, there is mixed in with\nthis also a secondary meaning, which causes confusion, and that\nis the idea that the things which are "real" are those things\nwhich would continue to persist in the absence of any perceiving\nobserver (real pain obviously does not satisfy this criterion).\nI prefer to use the term "actual" to refer that which is\nactually there, and not merely referred to, since "real" is\nladen down with so much confusion.\n\nTo add further to this confusion is the additional confusion caused\nby the inappropriate use of the notion of "outside." The relations\nof outside and inside are properly ascribed only to objects which\nare in space. The mind is not in space, although the head is, and\nso, while it is correct to say that a specific object, for example\na computer, is outside of my head, it is not correct to say that\nit is outside of my mind. Indeed, the relations of outside and\ninside do not apply in this case.\n\nNow, taken together, these two confusions lead to the realist\nposition, where the physicist believes that he is examining\nthat which "really" exists, "outside" of his mind. In fact,\nwhat he is doing is examining things which are actual (in\nmy sense of the term) and outside of his head. Perhaps I\nshould emphasise that this is worth thinking about; many\nlives have been wasted chasing ghosts because of this\nmisunderstanding.\n\n>> There\'s a psychological problem which has led theoretical physics\n>> into a nasty corner. It\'s the problem of physics students learning\n>> too many theories and doing too many exercises. Once they\'ve grown\n>> up, the only thing they can do is take a set of axioms and see\n>> what follows from them. .... The process of\n>> discovering, by careful thought and experimentation, what the\n>> actual properties of the object of study are, is necessary\n>> before one can start thinking of compressing all the relevant\n>> information into some axioms.\n>> ... but nobody tries to understand starting from\n>> the facts which are right in front of them. That\'s not what they\n>> were trained to do.\n\n>The point is that this is impossible. It is the old ideal of positivism\n>to derive theories from observable facts. Science works differently.\n>There is no derivation E_n ->X_n, there is guesswork. We guess\n>theories, derive the consequences, and reject the falsified guesses.\n\n>This is what we are trained to, and this is how science works.\n\nThis is what we are trained to do, but it is only a small part of\nhow science works, namely the part that Popper, with what talent\nhe had, was able to identify. Since you are here invoking an argument\nfrom authority (Science), better suited to scaring people who aren\'t\nsure that they know what science is, let me respond with a quote\nfrom Isaac Newton (although you should bear in mind that I don\'t\nactually intend you to take his authority in these matters seriously\nas an indication of the truth of his statements; I merely find that\nhe has nicely expressed what I would have said by myself anyway):\n\n"It is as true in physics as it is in mathematics, that, in\ninvestigating difficult things, the method known as the `analytic\'\nmethod should always precede what is called the `synthetic\' method.\nThe analytic method is to collect observations, to attend\nto the phenomena, and thereby to derive simple things from compound\nthings by reasoning - motive forces from motions, and generally,\ncauses from effects, and general causes from particular causes,\nuntil one eventually arrives at the most general causes. The\nsynthetic method is to take causes which have already been\ninvestigated and proved as one\'s starting point, and use them to\nexplain the phenomena which have arisen from them, and to prove\nthose explanations." Opticks (1704)\n\nThe problem that modern theoretical physics finds itself in is\nthat people are only trained in what Newton calls the synthetic\nmethod, principally because it\'s the easiest thing to test\nin examinations. It would be rather an interesting experiment\nto take a young physicist-to-be and show him all of the phenomena\nof electromagnetism, and assign to him the project of finding\nequations which describe all the observed results.\n\nA physicist with this training would be better equipped for real\ninvestigation than his colleagues, who have merely been trained to\ndo homework assignments in which the axioms are given and the\ntheorems are to be deduced. Theoretical physics is beginning to\nsuffer from a problem that philosophy has always had - that,\nin the absence of a better measure, what is recognized and rewarded\nis impressing your peers, and that is done by being an expert\nin obscure and unintelligible matters of no consequence to\nthose outside your field of investigation.\n\nI should add that it is not the case that the analytic method has\nbeen exhausted due to a lack of performable experiments whose\nresults we do not know in advance. The confusion about realism\nand the many disagreements about quantum mechanics (among those\nwho actually do think about it) are evidence enough that there\nis an opportunity, for one who is prepared to suspend judgment\nand avoid opinion, to discern something previously overlooked.\n\n>> The point of the Lorentz transformations is to switch between one\n>> observer and another, and the essence of relativity is that the\n>> laws of physics are the same for different (inertial) observers,\n>> even the speed of light. Without the idea of changing reference\n>> points and considering things from a different observer\'s point\n>> of view, special relativity would not have the formulation that\n>> it does. In order to formulate the questions which relativity\n>> addresses in my language, which is to some extent solipsistic,\n>> one first has to address the question of what relationship\n>> exists between one observer\'s X_n and another\'s X\'_n, which\n>> involves how one observer manifests himself within the X_n\n>> of another, and that is a difficult question.\n\n>In a realistic theory, the answer is simple: The X_n should be\n>equivalent, or at least one of the observers follows a wrong theory.\n\nThere are several problem with that. One is, as you said yourself,\nthat the X_n are not complete. That is, I perceive a room, with walls\nhere and there, and you perceive a different room (presumably),\nand there is no simple matching of what I perceive to what you\nperceive. The contraints on my X_n and your X\'_n can\'t be one\nof simple identity. There are other constraints, though; if\nwe can signal to each other, then that induces a common\ndirection of time, for example, since I can\'t get you to\nchange my past.\n\nR.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb
>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
>> ><rof@maths.tcd.ie> schrieb
>> >Yep. The construction of the M_n is of this type. But the X_n\M_n,
>> >the inner part of the moon, is not perceived. Even if I have some
>> >perception M_{moon} of the surface of the moon and some vague
>> >theoretical ideas what may be inside.
>>
>> Ok; I see what you're saying. Conceptual information, such as
>> "Iron is the stablest nucleus", is not represented spatially,
>> and it is most likely ideas like this which guide your thinking
>> about what is going on inside planets. I don't disagree with this,
>> but the sense in which I use the X_n is to refer explicitly to the
>> procedure which we use to construct a "world around us" from
>> sensation, and which is explicitly spatial.

>My point is that the construction is not complete.

>We have some construction E_n -> M_n but we do not
>construct a complete world X_n.

I have lost track of what you mean by M_n; it doesn't seem to
be in the same sense as I introduced it. I agree, though
that the construction is not complete - when I see a wall
in front of me, I don't perceive what's behind the wall.
This isn't a problem, though.

>Moreover, it is not given that the construction M_n is spatial.
>If we construct following Bohmian mechanics we construct
>some \Psi(q) which is not spatial but defined on the configuration
>space.

Sure, but the reason that it's an explanation that one feels
comfortable with is because it's easy to visualise. One imagines
a wave moving in some space (here, it's configuration space),
and that is something that we're used to. The state of affairs (the
wavefunction, \Psi, ignoring for a moment the particle positions)
is represented as being composed of several mini-states of affairs,
\Psi(x), each at a different position. One becomes a lot less comfortable
with it once one reflects on the fact that the place where the
wavefunction lives is not real space at all, but an abstract
construction called configuration space whose dimension is 3N.

We can make it make a little more sense by noticing that the
space over which the wavefunction is defined is the same as
the space over which probability distributions (of generic
states of affairs of the system, or configurations) are
naturally defined. Information that we have might have about
the configuration of the system at a given time would naturally
be expressed as a real function on this space, and a description
which included the instantaneous rate of change of this
probability distribution would naturally be expressed as
two real functions (or one complex function) on this space.

>> >Sorry, but you cannot derive if observers will "see" space.
>> >"Seeing" space means developing a certain realistic theory,
>> >may be implicit (as a hidden prejudice, instinct, created by
>> >evolution).

>> Evolution is another can of worms, best left unopened here, as is
>> the connection between the brain and conscious experience. Animals
>> see space, but they don't develop ontological theories. ... It's only
>> theoretical physicists
>> who have to face the dilemma of whether to stick the "reality" sticker
>> onto the wavefunction or the observables, because the urge to stick
>> it somewhere is so strong.

>I would not argue that there is some strong urge to develop ontological
>theories. Ontological theories have been successful, there is no reason
>to give them up.

There is no success of any theory which can be ascribed to its
ontological character.

>With Bohmian mechanics they are successful today
>too.

Only insofar as Bohmian mechanics has been able to reproduce
the predictions of a theory (Copenhagen-style quantum mechanics) which
wasn't ontological. In so far as Bohmian mechanics has not been
able to follow quantum mechanics (pair creation, for example),
it has not been successful, although I recognise that there has
been progress in this direction.

>> >A believes that there are
>> >some X_n in reality. This is an ontological theory proposed by A.
>>
>> A may think it's on ontological theory, but A is confusing himself.
>> Imagine there is a digital machine, which takes 1's and 0's as
>> input. You feed in some sequence of 1's and 0's, generated by
>> some algorithm. The machine proceeds to construct a Markov model
>> to predict the sequence. It is successful to some degree - most
>> of the time it correctly predicts the next input. Has it discovered
>> reality? Is its Markov model an ontological theory?

>Once it predicts successfully, there is some probability that the
>algorithm has been guessed correctly. In other words, that
>the algorithm used to create the input sequence and the
>algorithm used by the machine to predict the input sequence are
>in some sense equivalent.

But this doesn't answer either of the above questions. There's
also the problem of the vagueness of the "in some sense" which
is inevitably reached by proponents of realism in this situation.

The grand project of ontology is not to guess an algorithm which
predicts future observations (which is, in fact, the only thing
we can do), but to determine what it is that really exists.
If a machine develops an algorithm to correctly predict whether
the next input will be a 1 or a 0, the most one can say about
it is that that is what it has achieved. To say that it has
discovered the ontological basis of reality, one has to radically
change what one means by "discovered the ontological basis of
reality."

One who, observing traffic lights, eventually says "A repeating
pattern of green, then yellow and then red," has hardly reached the
ultimate goal that realists aspire to. He might imagine a circle
with a little point on it, moving consecutively through 120-degree
arcs labelled "Green", "Yellow" and "Red" (a Bohmian traffic
light, if you will), and declare it to be an ontological theory,
but such an individual would have to be regarded as confused.

This is important, because ontology stands or falls on this
point. We are not in a very different position to the digital
machine, or the traffic-light observer, with our sensations as the
input. Those who want to claim that we are engaged in the business
of discovering the fundamental ontological consituents of the
universe must make a rigorous link between this process of predicting
future inputs and what they regard as reality.

Alan Sokal, in his "Defense of a Modest Scientific Realism",
examines these issues a little bit, but, since he too is
wedded to realism, doesn't quite reach the correct answer,
but instead ends with:
"Since no existing theory purports to be a final theory, there
is no reason to consider it as literally true or to worry too
much about whether the entities it postulates `really exist'.
Or rather, when worrying about whether the unobservable entities
of a given theory `really exist', it is important to distinguish
existence _as a fundamental constituent of the universe_ from
existence _in some coarse-grained sense_. It is a reasonable guess
that _none_ of the theoretical entities in our present-day
theories are truly fundamental, and that _all_ of the theoretical
entities in our present-day well-confirmed theories will maintain
some status as derived entities in future theories." (His italics)

In the end, he opts for saying that things exist "in some coarse-grained
sense", which means that he wants to ascribe to them some property
which he feels is related to existence or reality, but which is so
vague that he cannot tell us what it is, and, lacking any concrete
or even specific thing to say, must instead appeal to our sympathy
to his realist position in order to convince us to accept what he
is saying as a valid defense of realism.

Notice that he also explicitly subscribes to the view that,
if an entity from one theory appears as a derived or emergent
thing in a more mature theory, then that entity cannot be a
"fundamental constituent of the universe." What he leaves open,
but implicitly suggests, is that, in a final theory, which
makes perfect predictions and has no further refinement, we
might actually consider the entities with which it deals to
be fundamental constituents of the universe which really exist.
This is, of course, what the whole exercise was intended to
deduce, and which it has failed to deduce, so he ends by
surreptitiously implying it.

Neither the positing of a kind of equivalence between algorithms,
nor the relegation of actual existence to existence in some coarse-grained
sense, actually bridge the gap between ontological existence and the
mathematical regularities of observed phenomena described by physical
theories.

It is very important in these matters not to decide that we like
realism and then accept as valid an otherwise insufficient
argument in favour of it. It is also important not to decide
that it is merely a matter of opinion, or indeed to be influenced
by opinion at all, since to allow an opinion to direct one's
understanding is nothing more than to believe what one likes
to believe. Thus, if we find that the arguments with which
we convince ourselves that realism is necessary amount to
"In my opinion, realism should be considered a part of logic,"
then we must go in search of the reason why that opinion
was adopted, and reject it unless there is a rigorous basis
for it. As it stands, it looks quite like a demand that
realism be adopted as an axiom; that is, it looks like
a dogmatic assertion of a statement which cannot be deduced
by reason.

>> The situation that we find ourselves in is that we know that
>> we cannot perfectly predict the results of measurements (in
>> EPR experiments, knowing the outcome of a spin measurement in
>> advance would allow signalling faster than light). This means
>> that our Markov model cannot perfectly predict the future,
>> even in principle. We're left with a merely statistical prediction,
>> although we may conjure up fruitless imaginations about what caused
>> the result of a particular experiment to be this or that.

>That's not really a problem. Once we don't know the complete
>input data used by the algorithm we cannot predict everything.
>Nonetheless it is possible that the algorithms we guess and the
>algorithms used in the "fabric of reality" are equivalent.

The implicit assertion here is that the fabric of reality
is implementing some kind of algorithm (recall that an algorithm
is a list of instructions for computing one thing from another).
The equivalence here is again, "in some sense", and has the particular
difficulty that one algorithm (the one used by a human) has, as its
output, patterns of sensations (as a prediction of future input).

I doubt that an algorithm which has sensations as output is one
that most realists would accept as an example of something which
could qualify as a fundamental constituent of reality, since the
things with which it deals (sensations) are purely mental.

>> Specifically, Bohmian mechanics is an attempt to imagine that
>> the technique which we use to order our perceptions (representing
>> a single state of affairs as lots of little states of affairs,
>> each in its own position, in configuration space, for example),
>> should still apply even in the absence of any sensible data
>> to represent.

>BM is not a "technique to order perceptions".

Agreed. I would never suggest that it is.

>It is an ontological theory.

Like the traffic light theory with the circle and the dot.

>> Whether or not a particle even exists can depend on the motion of
>> the observer (cf. the Unruh effect), so it seems strange to put that
>> sticky reality label onto the particles and their positions in space.

>I don't.

But Bohmian mechanics does, and that is the context in which I made
the remark above.

>The only requirement is that the ontological theory
>should define something named "reality".

I notice that there is a misunderstanding of the role of definition
among modern scientists. When one defines an object, one at the same
time generates that object. I may, for example, define a vector
space as a collection of objects which may be added together and
multiplied by the elements of some field, and thereafter I can
be completely certain when I make certain judgments about vector
spaces, because they are not objects which lie outside of me,
and which I must study through perception and experimentation,
but are free inventions of my mind, which have exactly those
properties which I gave to them when I defined them. I do not,
for example, have the same freedom to define a duck, since
ducks have many properties which must be discovered by examination,
and are not free inventions of my mind. I may, of course, choose
to use the word "duck" to refer to something else apart from
an actual duck, but that is unimportant here. In this case,
are you asserting that "reality" is something which can be
defined, or that its definition can change from theory to theory?

>Some mathematical
>objects should be given the label "real". And for these really
>existing objects we can use classical logic and probability theory.

I am always completely in favour of using classical logic and
probability theory.

>> ... declaring that the model is "out there", beyond
>> the mind of the person who thinks it, governing the evolution
>> of the world, is like saying that, since 1 typically follows
>> in a certain binary sequence, the rule "01" is out there, guiding
>> the sequence.

>Sorry, but we use physical theories to predict things which do not
>happen in our mind, but in reality. (Except you are a pure solipsist.)
>Of course, I feel free to say that there is a law of gravity out there.

>I may be wrong about the details of the law of gravity. But it is
>a hypothesis (in my mind) about what is out there.

So you begin to understand; the law of gravity, as you
know it, is not something which exists independently of you, but a
hypothesis which relies on you for its existence.

The real difficulty here is (as I mentioned before) the manifold
use of the word "real". On the one hand, it is used to refer to
what is actual, rather than merely referred to (real pain versus
imagined pain). If the use of the word "real" were restricted to
this, no confusion would arise. However, there is mixed in with
this also a secondary meaning, which causes confusion, and that
is the idea that the things which are "real" are those things
which would continue to persist in the absence of any perceiving
observer (real pain obviously does not satisfy this criterion).
I prefer to use the term "actual" to refer that which is
actually there, and not merely referred to, since "real" is
laden down with so much confusion.

To add further to this confusion is the additional confusion caused
by the inappropriate use of the notion of "outside." The relations
of outside and inside are properly ascribed only to objects which
are in space. The mind is not in space, although the head is, and
so, while it is correct to say that a specific object, for example
a computer, is outside of my head, it is not correct to say that
it is outside of my mind. Indeed, the relations of outside and
inside do not apply in this case.

Now, taken together, these two confusions lead to the realist
position, where the physicist believes that he is examining
that which "really" exists, "outside" of his mind. In fact,
what he is doing is examining things which are actual (in
my sense of the term) and outside of his head. Perhaps I
should emphasise that this is worth thinking about; many
lives have been wasted chasing ghosts because of this
misunderstanding.

>> There's a psychological problem which has led theoretical physics
>> into a nasty corner. It's the problem of physics students learning
>> too many theories and doing too many exercises. Once they've grown
>> up, the only thing they can do is take a set of axioms and see
>> what follows from them. .... The process of
>> discovering, by careful thought and experimentation, what the
>> actual properties of the object of study are, is necessary
>> before one can start thinking of compressing all the relevant
>> information into some axioms.
>> ... but nobody tries to understand starting from
>> the facts which are right in front of them. That's not what they
>> were trained to do.

>The point is that this is impossible. It is the old ideal of positivism
>to derive theories from observable facts. Science works differently.
>There is no derivation E_n ->X_n, there is guesswork. We guess
>theories, derive the consequences, and reject the falsified guesses.

>This is what we are trained to, and this is how science works.

This is what we are trained to do, but it is only a small part of
how science works, namely the part that Popper, with what talent
he had, was able to identify. Since you are here invoking an argument
from authority (Science), better suited to scaring people who aren't
sure that they know what science is, let me respond with a quote
from Isaac Newton (although you should bear in mind that I don't
actually intend you to take his authority in these matters seriously
as an indication of the truth of his statements; I merely find that
he has nicely expressed what I would have said by myself anyway):

"It is as true in physics as it is in mathematics, that, in
investigating difficult things, the method known as the `analytic'
method should always precede what is called the `synthetic' method.
The analytic method is to collect observations, to attend
to the phenomena, and thereby to derive simple things from compound
things by reasoning - motive forces from motions, and generally,
causes from effects, and general causes from particular causes,
until one eventually arrives at the most general causes. The
synthetic method is to take causes which have already been
investigated and proved as one's starting point, and use them to
explain the phenomena which have arisen from them, and to prove
those explanations." Opticks (1704)

The problem that modern theoretical physics finds itself in is
that people are only trained in what Newton calls the synthetic
method, principally because it's the easiest thing to test
in examinations. It would be rather an interesting experiment
to take a young physicist-to-be and show him all of the phenomena
of electromagnetism, and assign to him the project of finding
equations which describe all the observed results.

A physicist with this training would be better equipped for real
investigation than his colleagues, who have merely been trained to
do homework assignments in which the axioms are given and the
theorems are to be deduced. Theoretical physics is beginning to
suffer from a problem that philosophy has always had - that,
in the absence of a better measure, what is recognized and rewarded
is impressing your peers, and that is done by being an expert
in obscure and unintelligible matters of no consequence to
those outside your field of investigation.

I should add that it is not the case that the analytic method has
been exhausted due to a lack of performable experiments whose
results we do not know in advance. The confusion about realism
and the many disagreements about quantum mechanics (among those
who actually do think about it) are evidence enough that there
is an opportunity, for one who is prepared to suspend judgment
and avoid opinion, to discern something previously overlooked.

>> The point of the Lorentz transformations is to switch between one
>> observer and another, and the essence of relativity is that the
>> laws of physics are the same for different (inertial) observers,
>> even the speed of light. Without the idea of changing reference
>> points and considering things from a different observer's point
>> of view, special relativity would not have the formulation that
>> it does. In order to formulate the questions which relativity
>> addresses in my language, which is to some extent solipsistic,
>> one first has to address the question of what relationship
>> exists between one observer's X_n and another's X'_n, which
>> involves how one observer manifests himself within the X_n
>> of another, and that is a difficult question.

>In a realistic theory, the answer is simple: The X_n should be
>equivalent, or at least one of the observers follows a wrong theory.

There are several problem with that. One is, as you said yourself,
that the X_n are not complete. That is, I perceive a room, with walls
here and there, and you perceive a different room (presumably),
and there is no simple matching of what I perceive to what you
perceive. The contraints on my X_n and your X'_n can't be one
of simple identity. There are other constraints, though; if
we can signal to each other, then that induces a common
direction of time, for example, since I can't get you to
change my past.

R.

Thomas Trotter

Nov16-04, 02:57 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\nOz <oz@farmeroz.port995.com> wrote in message news:<7qRhy3XaEclBFwTJ@farmeroz.port995.com>...\n> Thomas Trotter <thomastrotter2005@juno.com> writes\n> >Oz <oz@farmeroz.port995.com> wrote in message news:<T\\$Q7nbQaXIkBFw\\$T@farmeroz.po\n> >rt995.com>...\n> >> Patrick Van Esch <vanesch@ill.fr> writes\n> >> >Bell is making up a "local realistic" model of the entangled pair. If\n> >> >you say that the photons HAD a specific polarisation direction lambda,\n> >> >then you have to be able to write it down (even if you don\'t know what\n> >> >it is). That\'s what he does.\n> >>\n> >> I have terrible trouble getting anyone to actually answer my question on\n> >> this clearly. The question is:\n> >>\n> >> Is lambda wrt some apparatus, or wrt the other photon?\n> >\n> >Lambda is the common emission-produced polarization\n> >*direction* of a pair of photons emitted by the same\n> >atom wrt the setting of some apparatus -- specifically,\n> >the transmission axes of the crossed linear polarizers.\n>\n> Are you saying that lambda is obtained by:\n>\n> 1) Measuring an angle of one particle with one apparatus to give L1.\n>\n> 2) Measuring and angle of the other particle with another apparatus to\n> give L2.\n>\n> 3) Relating angles on one apparatus to the other apparatus to give L3.\n>\n> Then comparing L1, L3 with L2?\n\nKeep in mind that I\'m trying to figure this Bell (entanglement)\nstuff out also. :-) So, all I can give is my perspective or\ninterpretation of what something means.\n\nLambda, in the optical Bell tests, and wrt some sort of local\nhidden variable model, is a property or characteristic of the\nlight pulses created via some emission process. Since the\npulses are analyzed by linear polarizers, then Lambda has\nsomething to do with polarization (at least that seems like\na good place to start). Specifically, then, Lambda has\nto do with how the expanding (emitted) light wave is rotating\nor spinning.\n\nEach emitted light pulse of a pair is (presumably)\ncircularly polarized, and (presumably) correlated with its\nemission partner, in polarization, via the emission process\ndue to conservation of angular momentum. [So, in effect, they\'re\nidentically polarized, and therefore you would expect\nto get identical detection attributes for any pair with\nthe polarizers *aligned* (theta = 0) -- either both detected\n(denoted by ++) or both not detected (denoted by --).]\n\nIn plane polarization (iiuc) there\'s just one spin or\nrotation direction. In circular polarization there\'s two.\n\nIn either case, the linear polarizer filters this incoming\nwave and transmits a plane polarized wave (or wavetrain) that\nis either intense enough to trigger the detector, or it isn\'t.\n\nSo, strictly speaking, afaik, insofar as Lambda refers to\nthe emission-produced rotations, then Lambda is never known.\n\nBut, if Lambda were known for each polarizer-incident light\npulse, then, presumably, you could predict the results of\nindividual measurements.\n\nThen you might ask the question: If knowledge of Lambda\n(the exact way in which an emitted light pulse is rotating)\ncan be used to accurately predict the individual results\nin setups using a linear polarizer and a detector, then can\nthis hidden variable (Lambda) be used to accurately predict\nthe results of correlational setups where you\'re analyzing\npaired, polarizer-incident light pulses, correlated in\npolarization via emission, wrt crossed linear polarizers?\n\n>\n> > This is a critical difference. The former says the angle is fixed from\n> >> the start as seen by the rest of the universe, the latter simply says\n> >> they always have a set relationship.\n> >\n> >Exactly.\n>\n> Yes, but which is it?\n\nLambda, wrt Bell\'s theorem and tests involving Bell inequalities,\nrefers to the former. It refers to the specific direction(s)\nof rotation of any given pair. Taken as a "single, continuous\nparameter" it (in effect, in the correlational context) refers to\nthe *common* direction(s) of rotation of paired light pulses.\nThe question Bell asked was, in effect, does the emission-produced,\ncommon direction of polarization determine coincidental detection?\nThe answer is no.\n\nThe qm formulation relies only on the latter -- that "they always\nhave a set relationship." The specific values that Lambda might\ntake on via the emission process are irrelevant wrt coincidental\ndetection. It\'s only important that paired light pulses *are*\nrelated wrt their spins or rotations (ie., that they\'re correlated\nin polarization via emission). But, qm provides no details.\nIt simply takes the polarizer transmission axis associated with\na detection as the common plane of polarization for the pair, and\napplies a Malus\' Law angular dependence wrt coincidental\ndetection.\n\n>\n> Is the assumption that each particle sets off with some\n> lambda, or is it that the angle between the particles is\n> always lambda?\n>\n\nIn an lhv model, the assumption is that each light pulse\nis emitted with definite rotational axes.\n\nQm makes no assumptions about the rotational axes, except\nto say that whatever they are, for any given pair they\nwill cancel each other out due to conservation of angular\nmomentum.\n\n> >>\n> >> >Quantum mechanics doesn\'t take this approach. It writes down that the\n> >> >state is:\n> >> >\n> >> >|psi> = 1/sqrt(2) { |n+>|n-> - |n->|n+ }\n> >> >\n> >> >with n just any direction. So our state is completely "unoriented"\n> >> >which is not possible in a realistic model.\n> >>\n> >> Isn\'t this simply saying that we cannot know the absolute orientation,\n> >> but do know the relative orientation?\n> >\n> >Yes.\n>\n> OK. Now, given this situation I would immediately want to do the\n> following experiment. I would select some transition that emits two\n> entangled particles, let us say U (for up) and D (for down).\n>\n> I would run one through a series of crossed polarisers so as to rotate\n> one by 180 degrees.\n>\n> Now I have three contrasting outcomes:\n>\n> 1) The particles are now always measured as both U or both D.\n> 2) The particles are always still measured as one U and one D.\n> 3) I never detect two particles.\n>\n> I have been trying to find out which of these is actually observed for\n> years, nobody will ever tell me.\n\nIn the case of paired light pulses, correlated in\npolarization, and randomly polarized via emission:\n\nStart with an emitter (E) and two ideal detectors (D),\nD <--- E ---> D ,\nand denoting a detection as + and a non-detection\nas - , then for this setup you\'ll always\nget ++ .\n\nNow, add ideal polarizers (P) with transmission axes aligned,\nD <--- P <--- E ---> P ---> D ,\nand for this setup you\'ll get half ++ and half -- .\n\nNow, add another ideal polarizer to one side set so\nthat it\'s transmission axis is perpendicular\nto the polarizer that\'s already there ,\nD <--- P <--- P <--- E ---> P ---> D ,\nand for this setup you\'ll always get - on\nthe the left side. Paired attributes\nwill be half -+ and half -- .\n\nBy the way, remember when we were discussing\nBell\'s inequality? Well, I still don\'t see\nwhat it has to do with locality/nonlocality.\nSo, if you understand it, then please explain\nit to me.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Oz <oz@farmeroz.port995.com> wrote in message news:<7qRhy3XaEclBFwTJ@farmeroz.port995.com>...
> Thomas Trotter <thomastrotter2005@juno.com> writes
> >Oz <oz@farmeroz.port995.com> wrote in message news:<T$Q7nbQaXIkBFw$T@farmeroz.po
> >rt995.com>...
> >> Patrick Van Esch <vanesch@ill.fr> writes
> >> >Bell is making up a "local realistic" model of the entangled pair. If
> >> >you say that the photons HAD a specific polarisation direction \lambda,
> >> >then you have to be able to write it down (even if you don't know what
> >> >it is). That's what he does.
> >>
> >> I have terrible trouble getting anyone to actually answer my question on
> >> this clearly. The question is:
> >>
> >> Is \lambda wrt some apparatus, or wrt the other photon?
> >
> >\Lambda is the common emission-produced polarization
> >*direction* of a pair of photons emitted by the same
> >atom wrt the setting of some apparatus -- specifically,
> >the transmission axes of the crossed linear polarizers.
>
> Are you saying that \lambda is obtained by:
>
> 1) Measuring an angle of one particle with one apparatus to give L1.
>
> 2) Measuring and angle of the other particle with another apparatus to
> give L2.
>
> 3) Relating angles on one apparatus to the other apparatus to give L3.
>
> Then comparing L1, L3 with L2?

Keep in mind that I'm trying to figure this Bell (entanglement)
stuff out also. :-) So, all I can give is my perspective or
interpretation of what something means.

\Lambda,[/itex] in the optical Bell tests, and wrt some sort of local
hidden variable model, is a property or characteristic of the
light pulses created via some emission process. Since the
pulses are analyzed by linear polarizers, then \Lambda has
something to do with polarization (at least that seems like
a good place to start). Specifically, then, \Lambda has
to do with how the expanding (emitted) light wave is rotating
or spinning.

Each emitted light pulse of a pair is (presumably)
circularly polarized, and (presumably) correlated with its
emission partner, in polarization, via the emission process
due to conservation of angular momentum. [So, in effect, they're
identically polarized, and therefore you would expect
to get identical detection attributes for any pair with
the polarizers *aligned* (\theta = 0) -- either both detected
(denoted by ++) or both not detected (denoted by --).]

In plane polarization (iiuc) there's just one spin or
rotation direction. In circular polarization there's two.

In either case, the linear polarizer filters this incoming
wave and transmits a plane polarized wave (or wavetrain) that
is either intense enough to trigger the detector, or it isn't.

So, strictly speaking, afaik, insofar as \Lambda refers to
the emission-produced rotations, then \Lambda is never known.

But, if \Lambda were known for each polarizer-incident light
pulse, then, presumably, you could predict the results of
individual measurements.

Then you might ask the question: If knowledge of \Lambda
(the exact way in which an emitted light pulse is rotating)
can be used to accurately predict the individual results
in setups using a linear polarizer and a detector, then can
this hidden variable (\Lambda) be used to accurately predict
the results of correlational setups where you're analyzing
paired, polarizer-incident light pulses, correlated in
polarization via emission, wrt crossed linear polarizers?

>
> > This is a critical difference. The former says the angle is fixed from
> >> the start as seen by the rest of the universe, the latter simply says
> >> they always have a set relationship.
> >
> >Exactly.
>
> Yes, but which is it?

\Lambda, wrt Bell's theorem and tests involving Bell inequalities,
refers to the former. It refers to the specific direction(s)
of rotation of any given pair. Taken as a "single, continuous
parameter" it (in effect, in the correlational context) refers to
the *common* direction(s) of rotation of paired light pulses.
The question Bell asked was, in effect, does the emission-produced,
common direction of polarization determine coincidental detection?
The answer is no.

The qm formulation relies only on the latter -- that "they always
have a set relationship." The specific values that \Lambda might
take on via the emission process are irrelevant wrt coincidental
detection. It's only important that paired light pulses *are*
related wrt their spins or rotations (ie., that they're correlated
in polarization via emission). But, qm provides no details.
It simply takes the polarizer transmission axis associated with
a detection as the common plane of polarization for the pair, and
applies a Malus' Law angular dependence wrt coincidental
detection.

>
> Is the assumption that each particle sets off with some
> \lambda, or is it that the angle between the particles is
> always \lambda?
>

In an lhv model, the assumption is that each light pulse
is emitted with definite rotational axes.

Qm makes no assumptions about the rotational axes, except
to say that whatever they are, for any given pair they
will cancel each other out due to conservation of angular
momentum.

> >>
> >> >Quantum mechanics doesn't take this approach. It writes down that the
> >> >state is:
> >> >
> >> >|\psi> = 1/\sqrt(2) { |n+>|n-> - |n->|n+ }
> >> >
> >> >with n just any direction. So our state is completely "unoriented"
> >> >which is not possible in a realistic model.
> >>
> >> Isn't this simply saying that we cannot know the absolute orientation,
> >> but do know the relative orientation?
> >
> >Yes.
>
> OK. Now, given this situation I would immediately want to do the
> following experiment. I would select some transition that emits two
> entangled particles, let us say U (for up) and D (for down).
>
> I would run one through a series of crossed polarisers so as to rotate
> one by 180 degrees.
>
> Now I have three contrasting outcomes:
>
> 1) The particles are now always measured as both U or both D.
> 2) The particles are always still measured as one U and one D.
> 3) I never detect two particles.
>
> I have been trying to find out which of these is actually observed for
> years, nobody will ever tell me.

In the case of paired light pulses, correlated in
polarization, and randomly polarized via emission:

Start with an emitter (E) and two ideal detectors (D),
D <--- E ---> D ,
and denoting a detection as + and a non-detection
as - , then for this setup you'll always
get [itex]++ .

Now, add ideal polarizers (P) with transmission axes aligned,
D <--- P <--- E ---> P ---> D ,
and for this setup you'll get half ++ and half -- .

Now, add another ideal polarizer to one side set so
that it's transmission axis is perpendicular
to the polarizer that's already there ,
D <--- P <--- P <--- E ---> P ---> D ,
and for this setup you'll always get - on
the the left side. Paired attributes
will be half -+ and half -- .

By the way, remember when we were discussing
Bell's inequality? Well, I still don't see
what it has to do with locality/nonlocality.
So, if you understand it, then please explain
it to me.

Ilja Schmelzer

Nov17-04, 10:45 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Oz" <oz@farmeroz.port995.com> schrieb im Newsbeitrag\nnews:KNNd7oXoyblBFwnV@farmeroz.port99 5.com...\n> Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes\n> >Translated into our physics "looking into the box" can be understood\n> >as developing realistic hidden variable theories. We know from Bell\n> >that they need a preferred frame. Therefore we have to develop\n> >theories with preferred frames.\n>\n> That would seem to be a viable thing to examine.\n> It either works out, or it doesn\'t.\n> But if its never seriously examined, who knows?\n\nI\'m examining it, with great success. And the only problem\nwith this success is that my papers are simply ignored.\n\nBut at the current moment I don\'t care much about this\nproblem. I have even stopped to submit my papers to\njournals. arxiv.org is sufficient to protect my priority,\nand the results are already too good to be ignored forever.\nSome ignorance during the next few years is not a problem.\nIt allows me to finish the remaining open problems myself.\n(I have already some chiral interaction on EW doublets,\nbut not yet the SM EW theory, and the details of\nsymmetry breaking are open yet.)\n\n> Its unfortunate that even mentioning this allocates you a significant\n> crank index, no matter how serious and how well supported by the\n> maths it is.\n\nYep.\n\n> >That\'s what I\'m doing. I wonder why I\'m almost alone.\n>\n> ER, because nobody will discuss it seriously with you?\n\nYep.\n\n> That\'s not what I meant.\n> What I meant was that, by definition, and event is (in relativity)\n> something everyone, everywherewhen, agrees on. If two people see\n> different results to some experiment then its not an event. What it is\n> is noise (or something similar).\n\nHm, in this sense the violation of Bell\'s inequality is a fact everyone\nagrees on. In, say, relativistic quantum field theory or even\nsemiclassical quantum gravity\n\n> >> QM, however, says we can never be quite sure.\n> >\n> >QM has never told me such things.\n>\n> Eh?\n\n> >Once you measure the same spin direction at A and B, the measurement\n> >gives the same result at A and B. There is no more uncertainty about\nthis\n> >in QM. There is no h-related bound for the accuracy of this experiment.\n\n> I\'ll absolutely bet that isn\'t true. In any real experiment there will\n> be measurements that give the \'wrong\' result. No detector is perfect and\n> no environment is noise-free. Surely, ultimately, any measurement must\n> have a limiting accuracy related to h?\n\nOf course, in every experiment we have some limit of accuracy.\n\nBut, first, for the limiting accuracy of the measurement of spin direction\nwe have no absolute, theory-defined limit related to h. And, second,\nvery reasonable limits of accuracy which are certainly not forbidden in\ntheory (AFAIR something like detector efficiency > 80%) allow to\nestablish with certainty that Bell\'s inequality is violated.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Oz" <oz@farmeroz.port995.com> schrieb im Newsbeitrag
news:KNNd7oXoyblBFwnV@farmeroz.port995.com...
> Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes
> >Translated into our physics "looking into the box" can be understood
> >as developing realistic hidden variable theories. We know from Bell
> >that they need a preferred frame. Therefore we have to develop
> >theories with preferred frames.
>
> That would seem to be a viable thing to examine.
> It either works out, or it doesn't.
> But if its never seriously examined, who knows?

I'm examining it, with great success. And the only problem
with this success is that my papers are simply ignored.

But at the current moment I don't care much about this
problem. I have even stopped to submit my papers to
journals. arxiv.org is sufficient to protect my priority,
and the results are already too good to be ignored forever.
Some ignorance during the next few years is not a problem.
It allows me to finish the remaining open problems myself.
(I have already some chiral interaction on EW doublets,
but not yet the SM EW theory, and the details of
symmetry breaking are open yet.)

> Its unfortunate that even mentioning this allocates you a significant
> crank index, no matter how serious and how well supported by the
> maths it is.

Yep.

> >That's what I'm doing. I wonder why I'm almost alone.
>
> ER, because nobody will discuss it seriously with you?

Yep.

> That's not what I meant.
> What I meant was that, by definition, and event is (in relativity)
> something everyone, everywherewhen, agrees on. If two people see
> different results to some experiment then its not an event. What it is
> is noise (or something similar).

Hm, in this sense the violation of Bell's inequality is a fact everyone
agrees on. In, say, relativistic quantum field theory or even
semiclassical quantum gravity

> >> QM, however, says we can never be quite sure.
> >
> >QM has never told me such things.
>
> Eh?

> >Once you measure the same spin direction at A and B, the measurement
> >gives the same result at A and B. There is no more uncertainty about
this
> >in QM. There is no h-related bound for the accuracy of this experiment.

> I'll absolutely bet that isn't true. In any real experiment there will
> be measurements that give the 'wrong' result. No detector is perfect and
> no environment is noise-free. Surely, ultimately, any measurement must
> have a limiting accuracy related to h?

Of course, in every experiment we have some limit of accuracy.

But, first, for the limiting accuracy of the measurement of spin direction
we have no absolute, theory-defined limit related to h. And, second,
very reasonable limits of accuracy which are certainly not forbidden in
theory (AFAIR something like detector efficiency > 80%) allow to
establish with certainty that Bell's inequality is violated.

Ilja

seratend

Nov17-04, 10:54 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cna7di\\$41f\\$1@beech.fernuni-hagen.de>...\n> > I understand that you are saying that hidden variable theories need a\n> > preferred frame (I have suppressed the realistic : ).\n> > I have not considered this problem but I am surprised as almost all\n> > (the one\'s I know) such theories mimics the QM so adding special\n> > relativity should not be a problem. Would you mind giving us more\n> > details on the preferred frame problem?\n>\n> A preferred frame we need if we want a realistic interpretation of\n> QM. The simplest choice is Bohmian mechanics.\n>\nNow, I may understand that by preferred frame, you may mean a\npreferred observable (the q position, the observable used in bohemian\nmechanics) or a given frame (x1,x2,x3,t), thus which one?\n\n> "Realistic" means, roughly speaking, what we have to make a\n> hypothesis about reality. The realistic theory has an ontology,\n> a description of reality.\n\nPlease see my comments on reality/ontology on my previous post with\nthe date 2004-11-12 11:09:18 PST.\nCurrently, I think we don\'t need to introduce a theory of reality as\nit can move us towards philosophy or metaphysics rather than physics.\nTherefore, I definitively prefer to concentrate on logically\nconsistent models and thereafter interpretation (connection between\nthe model results and experiment results) before speaking of what\nreality is (which I think is out of scope at this step).\n\nI really appreciate your effort (what I seem to understand) to use\nlocal variables with local interactions to describe QM results as long\nas it avoids deductions based on extra interpretations/hypothesis that\nare out of the scope or undefined (on a mathematical level). Reality\nis part of these fuzzy hypotheses.\nYou really do not need it to look for other interesting logical model\nthat can explain more "classically" the results of QM experiments such\nas EPR.\n\nNow, to try to end with the "reality" subject, can I translate what\nyou call "realistic theory" into a self-consistent mathematical model?\nIf yes, I really prefer to read "self-consistent theory" rather than\n"realistic theory" even if it is longer and heavier.\n\n\n> In Bohmian mechanics, the reality\n> consists of two parts, the wave function Psi(q) and the state of the\n> universe q. Except in the simplest case of one-particle-theory\n> q is not the location in space. It is the general configuration of\n> the whole universe, including positions of all particles. Therefore\n> Psi(q) is not a local function. But it is local in time, a function\n> Psi(q,t). As well, the state of the universe is a function of t q(t).\n>\nSee my previous post part on the bohemian mechanics with the date\n2004-11-12 11:09:18 PST.\nThe main problem with bohemian mechanics is the selection of the\nequation of motion of q(t). As long as Bohmian mechanics accept the\nSchrodinger time evolution of the psi(q,t), this theory accepts that\nwe can not see the difference between a QM measurement of the\nobservable q and the bohemian measurement of the path q(t).\n\nIn other words we cannot have in bohemian mechanics\n|psi(q,t)|^2=delta(q-qo(t)) (a known trajectory).Once an infinite\nprecision q measurement occurs with a result qo, the bohemian velocity\nis unknown at this time as in QM. Therefore, the position of the\nbohemian particle for a time t>to is unknown as in QM. It can be\neverywhere in the universe –this is the limit of classical QM\nmeasurement of position: the infinite precision measurement\ninteraction requires infinite energies to give an infinite precision\nmeasurement thus we do not respect the v << c approximation.\nThus, I have a theory with an externally added equation that gives no\nmore information (achievable in an experiment) than the orthodox QM\ntheory.\nTherefore, as long as this new equation of motion cannot provide\nadditional information (existence of an experiment that shows this\nadditional information), I prefer to consider Bohmian mechanics as one\npossible interpretation that is compatible with the orthodox\ninterpretation of QM, with an additionnal cost q(t).\n\nExplanation the previous statement: U(t,to) is the time evolution of\nthe initial state |psi(to)>. We have psi(q,to)=<q|psi(to> for the\nbohemian particle.\n\nIf we define the projector P(t)=|psi(t)><psi(t)| where\n|psi(t)>=U(t,to)|psi(to)> as the measurement at any time t>to, we have\nwith probability 100% the detection of the particle without the change\nof the state psi(t). Thus P(t) defines the trajectory of the particle\nwith the same accuracy as the possible q(t) motions in bohemian\nmechanics compatible with the state psi(q,t).\n\n\n> The problem with relativistic QFT is that it extends the "shut up and\n> calculate" ideology. It does not care at all about an explanation\n> for the observed probabilities - being able to compute them is\n> sufficient. This is not an explanation.\n\nThis is the measurement theory. It tries to explain in a\nself-consistent way the results we have with this mathematical model\n(interpretation consistency problems).\nI think that the "shut up and calculate ideology" is somewhat outside\nthe context. Note that this is what does most of the engineers around\nthe world (I think).\n\nNote that Bohmian meachanics has the same interpretation problem if\nyou do want to make a simple "shut up and calculate". q(t) has\ncurrently no measurable signification. Why not taking (q+p)(t) as the\ntrajectory of new bohemian particle (observable Q+P)? Who is the more\n"real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.\n\nBy preferred frame, If you intend preferred observable (e.g. selection\nof q(t) in BM), I think that bohmian mechanics cannot give you what\nyour are looking for, i.e. you can construct in principle a "A\nbohmian" mechanics, where A is any observable of QM.\nConsidering this aspect (existence of a preferred observable), I\nreally think that the decoherence working group and the work on the\npreferred basis is much closer to the answer to this question (at\nleast logically coherent).\n\n> Compare with the theory:\n> "We hear voices every Friday the 13. at midnight in the castle."\n> It is a falsifiable theory. But, let\'s assume it is supported by\n> observation. Would you accept such a successful prediction\n> as an explanation?\nIf a have the context of this sentence and checked the logical\nconsistency. I may say yes it is true if its interpretation in the\nexperimental results agree. After, I may consider the context and try\nto check if this context can be explained by external theories (i.e.\ntestable). I think that it is exactly we do in our everyday life,\nphysics in this aspect is only an extreme example (but It is a\nparticular point of view, not really important for our subject).\nFor example, I may search for ghost that causes the voice, if I can\ndefine and test what is a ghost (e.g. a human with a white cloth,\nsomething else …)\n\n> Certainly not. We would search for other\n> explanations. As physicists we would search for physical\n> explanations.\n\nAt least logical explanations when possible.\n\n> Now, if we reject the "shut up and calculate" ideology as not\n> giving a realistic explanation, we have BM which has a preferred\n> frame as a realistic theory.\n>\nCurrently with what I have already said on BM, the preferred frame (in\nthe sense of the observable see above) of BM is the same thing as the\nstatement "the position is the preferred basis of QM". Well very\ninteresting property, but what can I do with this extra property? Why\ndo we need this extra property? I thus prefer to look at the\ndecoherence program and the preferred basis research.\nIf you are interested on this aspect, you can look at the paper\nhttp://arxiv.org/PS_cache/quant-ph/pdf/0312/0312059.pdf. It contains\nan exhaustive status (end of 2003) on the decoherence program with\nadditional pointers to other papers.\n\n> And we have Bell\'s inequality. It shows that for every pair of events\n> A, B, such that measurements and observations at these events\n> allow to violate Bell\'s inequality, we have a (hidden) causal\n> influence A->B or B->A. Given the predictions of QM, this holds\n> for almost every pair of events.\nIf we assume one interpretation that (q1+q2)(t)=q1(t)+q2(t). See my\nprevious post section "bohemian mechanics". As long as we cannot have\nan experiment that detects q1(t) as the path of one particle and q2(t)\nas the path of the second particle we stay with a global state\n(q1+q2)(t). If we stay with a global state, we cannot say if there is\na "hidden causal influence" .See my previous post on your causality\ninterpretation section.\n\n> Now, let\'s consider the question what\n> really happens, A->B or B->A. We cannot measure this. It is\n> a hidden variable. Nonetheless, in a realistic causal theory we\n> postulate that there are no closed causal loops.\n\nNever forget that each postulate you add, has to be self-consistent\nwith the rest of your theory. The more postulates you add, the more\nyou increase the inconsistency risk. The logical demonstration of\nconsistency is far from being evident. One famous example is the\naxiomatization of the set theory where the first "naïve" construction\nby Cantor leads to the Russel paradox.\nSee my previous post on the inconsistency of your causality definition\n(A->B => B->A).\n>\n> Note that this discussion about EPR, Bell and preferred frames is\n> only one part of my general argumentation in favour of a preferred frame.\n> It includes many other, completely independent parts.\n>\nI really want to help you in defining a consistent logical substrate\nof your interesting point of view I interpret as trying to build a\nlocal variable extension of QM (with local interactions). However, I\nthink you really need to remove (in a first step) some initial\ninterpretations and some concepts.\n\nMy knowledge in GR is relatively low. I stay with the unresolved\nproblem of inserting the classical electromagnetic tensor in the\nequation : ).\n\nHowever, I question why you seem to consider a preferred frame as\nfundamental and not experiment dependant (i.e. the decoherence program\n– you have a relative frame).\n\nThanks for your papers, I will try to have a look at some of them.\n\nSeratend.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cna7di$41f$1@beech.fernuni-hagen.de>...
> > I understand that you are saying that hidden variable theories need a
> > preferred frame (I have suppressed the realistic : ).
> > I have not considered this problem but I am surprised as almost all
> > (the one's I know) such theories mimics the QM so adding special
> > relativity should not be a problem. Would you mind giving us more
> > details on the preferred frame problem?
>
> A preferred frame we need if we want a realistic interpretation of
> QM. The simplest choice is Bohmian mechanics.
>
Now, I may understand that by preferred frame, you may mean a
preferred observable (the q position, the observable used in bohemian
mechanics) or a given frame (x1,x2,x3,t), thus which one?

> "Realistic" means, roughly speaking, what we have to make a
> hypothesis about reality. The realistic theory has an ontology,
> a description of reality.

Please see my comments on reality/ontology on my previous post with
the date 2004-11-12 11:09:18 PST.
Currently, I think we don't need to introduce a theory of reality as
it can move us towards philosophy or metaphysics rather than physics.
Therefore, I definitively prefer to concentrate on logically
consistent models and thereafter interpretation (connection between
the model results and experiment results) before speaking of what
reality is (which I think is out of scope at this step).

I really appreciate your effort (what I seem to understand) to use
local variables with local interactions to describe QM results as long
as it avoids deductions based on extra interpretations/hypothesis that
are out of the scope or undefined (on a mathematical level). Reality
is part of these fuzzy hypotheses.
You really do not need it to look for other interesting logical model
that can explain more "classically" the results of QM experiments such
as EPR.

Now, to try to end with the "reality" subject, can I translate what
you call "realistic theory" into a self-consistent mathematical model?
If yes, I really prefer to read "self-consistent theory" rather than
"realistic theory" even if it is longer and heavier.

> In Bohmian mechanics, the reality
> consists of two parts, the wave function \Psi(q) and the state of the
> universe q. Except in the simplest case of one-particle-theory
> q is not the location in space. It is the general configuration of
> the whole universe, including positions of all particles. Therefore
> \Psi(q) is not a local function. But it is local in time, a function
> \Psi(q,t). As well, the state of the universe is a function of t q(t).
>
See my previous post part on the bohemian mechanics with the date
2004-11-12 11:09:18 PST.
The main problem with bohemian mechanics is the selection of the
equation of motion of q(t). As long as Bohmian mechanics accept the
Schrodinger time evolution of the \psi(q,t), this theory accepts that
we can not see the difference between a QM measurement of the
observable q and the bohemian measurement of the path q(t).

In other words we cannot have in bohemian mechanics
|\psi(q,t)|^2=\delta(q-qo(t)) (a known trajectory).Once an infinite
precision q measurement occurs with a result qo, the bohemian velocity
is unknown at this time as in QM. Therefore, the position of the
bohemian particle for a time t>to is unknown as in QM. It can be
everywhere in the universe –this is the limit of classical QM
measurement of position: the infinite precision measurement
interaction requires infinite energies to give an infinite precision
measurement thus we do not respect the v << c approximation.
Thus, I have a theory with an externally added equation that gives no
more information (achievable in an experiment) than the orthodox QM
theory.
Therefore, as long as this new equation of motion cannot provide
additional information (existence of an experiment that shows this
additional information), I prefer to consider Bohmian mechanics as one
possible interpretation that is compatible with the orthodox
interpretation of QM, with an additionnal cost q(t).

Explanation the previous statement: U(t,to) is the time evolution of
the initial state |\psi(to)>. We have \psi(q,to)=<q|\psi(to> for the
bohemian particle.

If we define the projector P(t)=|\psi(t)><\psi(t)| where
|\psi(t)>=U(t,to)|\psi(to)> as the measurement at any time t>to, we have
with probability 100% the detection of the particle without the change
of the state \psi(t). Thus P(t) defines the trajectory of the particle
with the same accuracy as the possible q(t) motions in bohemian
mechanics compatible with the state \psi(q,t).

> The problem with relativistic QFT is that it extends the "shut up and
> calculate" ideology. It does not care at all about an explanation
> for the observed probabilities - being able to compute them is
> sufficient. This is not an explanation.

This is the measurement theory. It tries to explain in a
self-consistent way the results we have with this mathematical model
(interpretation consistency problems).
I think that the "shut up and calculate ideology" is somewhat outside
the context. Note that this is what does most of the engineers around
the world (I think).

Note that Bohmian meachanics has the same interpretation problem if
you do want to make a simple "shut up and calculate". q(t) has
currently no measurable signification. Why not taking (q+p)(t) as the
trajectory of new bohemian particle (observable Q+P)? Who is the more
"real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.

By preferred frame, If you intend preferred observable (e.g. selection
of q(t) in BM), I think that bohmian mechanics cannot give you what
your are looking for, i.e. you can construct in principle a "A
bohmian" mechanics, where A is any observable of QM.
Considering this aspect (existence of a preferred observable), I
really think that the decoherence working group and the work on the
preferred basis is much closer to the answer to this question (at
least logically coherent).

> Compare with the theory:
> "We hear voices every Friday the 13. at midnight in the castle."
> It is a falsifiable theory. But, let's assume it is supported by
> observation. Would you accept such a successful prediction
> as an explanation?
If a have the context of this sentence and checked the logical
consistency. I may say yes it is true if its interpretation in the
experimental results agree. After, I may consider the context and try
to check if this context can be explained by external theories (i.e.
testable). I think that it is exactly we do in our everyday life,
physics in this aspect is only an extreme example (but It is a
particular point of view, not really important for our subject).
For example, I may search for ghost that causes the voice, if I can
define and test what is a ghost (e.g. a human with a white cloth,
something else …)

> Certainly not. We would search for other
> explanations. As physicists we would search for physical
> explanations.

At least logical explanations when possible.

> Now, if we reject the "shut up and calculate" ideology as not
> giving a realistic explanation, we have BM which has a preferred
> frame as a realistic theory.
>
Currently with what I have already said on BM, the preferred frame (in
the sense of the observable see above) of BM is the same thing as the
statement "the position is the preferred basis of QM". Well very
interesting property, but what can I do with this extra property? Why
do we need this extra property? I thus prefer to look at the
decoherence program and the preferred basis research.
If you are interested on this aspect, you can look at the paper
http://arxiv.org/PS_cache/quant-ph/pdf/0312/0312059.pdf. It contains
an exhaustive status (end of 2003) on the decoherence program with
additional pointers to other papers.

> And we have Bell's inequality. It shows that for every pair of events
> A, B, such that measurements and observations at these events
> allow to violate Bell's inequality, we have a (hidden) causal
> influence A->B or B->A. Given the predictions of QM, this holds
> for almost every pair of events.
If we assume one interpretation that (q1+q2)(t)=q1(t)+q2(t). See my
previous post section "bohemian mechanics". As long as we cannot have
an experiment that detects q1(t) as the path of one particle and q2(t)
as the path of the second particle we stay with a global state
(q1+q2)(t). If we stay with a global state, we cannot say if there is
a "hidden causal influence" .See my previous post on your causality
interpretation section.

> Now, let's consider the question what
> really happens, A->B or B->A. We cannot measure this. It is
> a hidden variable. Nonetheless, in a realistic causal theory we
> postulate that there are no closed causal loops.

Never forget that each postulate you add, has to be self-consistent
with the rest of your theory. The more postulates you add, the more
you increase the inconsistency risk. The logical demonstration of
consistency is far from being evident. One famous example is the
axiomatization of the set theory where the first "naïve" construction
by Cantor leads to the Russel paradox.
See my previous post on the inconsistency of your causality definition
(A->B => B->A).
>
> Note that this discussion about EPR, Bell and preferred frames is
> only one part of my general argumentation in favour of a preferred frame.
> It includes many other, completely independent parts.
>
I really want to help you in defining a consistent logical substrate
of your interesting point of view I interpret as trying to build a
local variable extension of QM (with local interactions). However, I
think you really need to remove (in a first step) some initial
interpretations and some concepts.

My knowledge in GR is relatively low. I stay with the unresolved
problem of inserting the classical electromagnetic tensor in the
equation : ).

However, I question why you seem to consider a preferred frame as
fundamental and not experiment dependant (i.e. the decoherence program
– you have a relative frame).

Thanks for your papers, I will try to have a look at some of them.

Seratend.

Ilja Schmelzer

Nov17-04, 11:13 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n\n"seratend" <ser_monmail@yahoo.fr> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> On the Causality "->" definition:\n> ====================\n> >>> No. In mathematical terms it means absence of independence.\n> >> Whow, what a very weak and fuzzy definition.\n>\n> > Certainly not fuzzy. For observables P(AB) = P(A)P(B) means\n> > independence.\n>\n> Ok, I will say this relation means probabilistic independence (math\n> definition). I assume that we are speaking about classical\n> probability.\n>\n> So with your definition, if P(AB) = P(A)P(B) then the 2 relations "A\n> -> B" and "B ->A" are not true. ...\n> Thus, we can\'t have (A-> B) and not (A->B) because:\n> not (A -> B) => P(A,B=a,b) = P(A=a)P(B=b) => not (B-> A).\n\nOf course, the symmetric definition P(AB) = P(A)P(B)\ndoes not define the antisymmetric relation A->B.\nStatistical independence is one thing, causal independence\nanother one. P(AB) =/= P(A)P(B) means statistical dependence.\nWhich may be caused by causal dependence A->B or B->A,\nbut as well by a common cause C->A, C->B.\n\nIn a causal realistic theory (which is the context of the definition of\ncausal influence) we have some external reality l in /\\, some\n"decisions of experimenters" or parameters of the experimental\ndevice a_i, and some results of the experiment A_i.\nThe realistic theory should explain the observable results in the\nfollowing way\n\n<A_i> = int A(a1,...,l) rho(l) dl.\n\nNow, we have a_j -/-> A_i if\n\n<A_i> = int A(a_1,...a_j-1,a_j+1,...,l) rho(l) dl.\n\nThis is also precise enough, I hope. This covers only a particular\ncase of causal influences - the one where the cause is some\ndecision of the experimenter a_j - but this one is sufficient for\nthe proof of Bell\'s inequality.\n\nI\'m not sure how to formulate it in the general context, where\nthe causal realistic theory is not specified. Usually the definition\nis quite obvious once a causal realistic theory is given. For\nexample, in classical GR we have A->B => A is in the past\nlight cone of B. And in more detail the causal connections follow\nfrom the evolution equations of the given theory. Thus, "causality"\nis something which has to be defined in detail (specified) only\nby the given theory, for example, via an evolution equation.\n\n> > Everything else is a nontrivial correlation which requires a nontrivial\n> > realistic explanation.\n\nIf we observe a statistical correlation, that means,\nP(AB) =/= P(A)P(B),\nwe need a realistic explanation, that means, an\nexplanation of type\nA->B or B->A or\nexists C with C->A, C->B.\n\nThat\'s my reformulation of the EPR criterion of reality.\n\n> > In this sense, A->B denotes some fundamental, basic\n> > dependence as part of\n> > some ontological theory. Thus, the theory specifies\n> > what exists (ontology)\n> > and what depends on what (causality, usually\n> > defined by laws of evolution).\n\n> I continue to say that you are using too many different concepts, not\n> well defined, to make some logical deductions.\n\nI hope I have now better defined them.\n\n> > I would say that\'s not all. A->B is a relation, it is transitive, and an\n> > ordering: A->B and B->A means A=B. There are no closed\n> > causal loops except trivial ones A->A.\n\nThese are general properties of the relation A->B. As said, the\ndefinition has to be given by the realistic theory.\n\n> About your Reality, ontology :\n> ==============================\n> In your arguments, you are also using many times the word reality. I\n> really think that this term as well as causality is not well defined\n> (in a self consistent mathematical basis).\n>\n> We thus have the same previous danger in using it (risk of\n> inconsistent argumentation).\n\nI think it is enough well-defined, and this well-defined notion of\nrealistic theory is what has been used by Bell in his proof.\n\n> Once again, I do not understand your reality statement/meaning. Bell\'s\n> theorem deals only with the impossibility of a probabilistic space\n> with local non-contextual random variables to reproduce the QM\n> probability results in a particular state. This is the logical\n> statement.\n\nI would agree except that "local" is not the point. Bell\'s theorem\nexcludes large classes of global and/or contextual hidden variables\nas well.\n\n> The word "reality" is an interpretation (it only makes a weak\n> connection between the theorem and the initial EPR paper where the\n> word itself is re-interpreted).\n\nThe initial EPR paper is not that important. In the case you win\nsome part of this discussion and the concept of "realistic causal\ntheory" is not yet enough clarified/specified I would try to correct\nthis using a specification which allows to prove Bell\'s inequality.\nThat means I would use a definition of "realistic theory" which\nallows to prove\n\n<AB> = int A(a,b,l) B(a,b,l) rho(l) dl.\n\nand a definition of "causal theory" which allows to derive from this\nand "not (a->B and b->A)" the formula\n\n<AB> = int A(a,l) B(b,l) rho(l) dl.\n\nThis certainly defines one possible notion of "realistic causal\ntheory". And for this particular notion of realism, (essentially\nby construction) we can prove Bell\'s inequality.\n\nThen, what remains is the metaphysical thesis that this definition\n(specification) of the philosophical notion "realism" is appropriate.\nThis is already something impossible to prove, impossible in\nprinciple. But I would say that from metaphysical point of\nview this is a very weak notion of realism. Last not least, lot\'s\nof "theories" which in common sense are not realistic (theories\ninvolving ghosts, gods and so on) are realistic in this sense.\n\nOf course, it seems possible to think about meaningful extensions\nof this notion of realism, to define some class of "weak realistic\ntheories". But a meaningful notion of weak realism should be\nin some sense stronger than the QM "shut up and calculate"\ninterpretation, in the sense that the proposal of a weak realistic\ntheory which is able to describe a given set of observations\nis an explanation in some meaningful sense. "Shut up and\ncalculate" is not an explanation, but the rejection to give\nexplanations.\n\n> Therefore, argumentation/deduction on a "reality" can only be made\n> with an external self-consistent definition (with its domain of\n> validity). This is almost equivalent to create a physical theory of\n> "reality" even if I still do not know what is "reality".\n> I am not sure if this theory is useful for the Bell theorem and FTL\n> interactions problem.\n\nI\'m not thinking about a physical theory of reality, but about\na characterization of a class of physical theories as realistic theories.\n\nOf course, that\'s not that much different. We can understand GR\nas a class of physical theories with some g_ij(x) and some matter\nfields. The physical theory is defined only if we define the matter fields\nand the matter Lagrangian. In the same sense, "realistic theory"\ndescribes a general scheme which should be used by any realistic\ntheory. This scheme specifies, for example, that some set should\nbe used to describe all possible states of reality, and there should be\nsome probability measure on it. And a causal theory is one where\nwe have some relation A->B defined.\n\n> > Reality is some element of some set of possible realities\n> > (using set theory)\n> > with classical probability measure rho.\n\n> I may reinterpret your reality definition in mathematical words. We\n> keep the consistency with the previous causality definition: we are\n> using the classical probability spaces. I have two possible\n> interpretations:\n>\n> First interpretation:\n> - Classical probability measure: you have a probability space (O,sigma\n> algebra L, probability rho(o).do)\n> - Reality is an event l, an element of the sigma algebra L\n> - set of possible realities= sigma algebra L\n>\n> With this first interpretation, I do not know what an outcome (an\n> element of O) is.\n\nIn the description I have used above, an observable.\n\nThe reality L describes the states of a black box. The experiment consists\nof making some input by free decisions and then observing some output.\n\nThe experimenter is, in this case, not described as part of the reality.\nNor the way he defines the input, nor the way he memorizes the output.\nFor the part of the experimenter we postulate free will and an ideal memory.\n\n(Of course, as realists in the philosophical sense, we would like to have\nalso a theory which is able to describe the whole universe including the\nobserver.\nBM is an example of a theory which is able to solve this problem. You can\ndescribe the whole universe, and you can reduce the problem, consider the\nclassical observer as independent, and the reduced theory, which describes\nonly the black box, has the same form as BM. Nonetheless, considering only\nsuch a special subtype of theory would be too restrictive. There are\nrealistic theories which do not cover the observer too.)\n\n> > This is part of the general scheme of ontological theories: Each theory\n> > worth to be named realistic (ontological, having an ontology) has such a\n> > set. There is no other restriction on this set.\n\n> Always the anylogy wording. : ). Do you simply mean a logical\n> self-consistent theory (mathematical interpretation)?\n\nNo. QM in its minimal interpretation is logical self-consistent\nbut not realistic.\n\n> > In a realistic theory, I would name it local if L = some set of\n> > functions\n> > on some manifold X named space. This would allow FTL, therefore,\n> > has nothing to do with Einstein causality.\n>\n> Well, It seems that you have restricted the possible choices of a\n> possible theory on a theory based on a single classical probability\n> space (with a single probability law).\n> You need to specify the set of outcomes L, the sigma-algebra and the\n> probability law as well as the random variables.\n\nThis is what the realistic theory has to specify. Different realistic\ntheories have different sets of observables, different sets of realities,\nand so on.\n\n> However, with this few assumptions, we can already say that this\n> theory is included in the classical QM probability (we are restricting\n> the set of QM observables to one subset of commuting observables: the\n> random variables on the classical probability space).\n\nQM does not define ontology. If you would like to say that QM can\nbe easily modified in such a way that it fits into the scheme - no problem,\nI have no problem to accept that BM is only a quite trivial "realistic"\nreinterpretation of QM. And that there may be others, for example\nthe collapse interpretation.\n\n> On the classical Bohmian mechanics:\n> ==================================\n> > We can do more. We can decide that only the thing named "statistical\n> > hidden\n> > variable model" is worth to be considered as a scientific explanation,\n> > and\n> > once we have found such explanations (like those given by Bohmian\n> > mechanics)\n> > everything which gives less (like quantum mechanics) should be rejected.\n> > This is a methodological decision, a decision about the nature of the\n> > scientific method.\n\n> First, I have to re-specify what Bohmian mechanics really is (for me,\n> only on the logical mathematic content, not the interpretations we may\n> find around the world).\n> Bohmian mechanics is the re-formulation of a quantum system through a\n> particular observable plus an additional hypothesis on the motion\n> equation of this observable.\n>\n> Bohmian mechanics has selected the Q observable (position), but we can\n> choose, formally, any observable of QM (P, combination of P,Q, .).\n> The bohemian system state (a single particle) is described by the\n> couple (q(t), psi(q,t)).\n\nCorrect. (Except the \'e\' in \'bohemian\')\n\n> -- psi(q,t) is the state in the |q> basis. Its evolution is controlled\n> by the classical QM Schroedinger equation and the QM postulates.\n>\n> -- q(t) in bohemian mechanics is an external random variable. This\n> externally added feature defines a local (relative to the variable q)\n> classical probability space.\n>\n> The equation describing the statistical motion of q(t) is an external\n> addition to QM formulation. Its motion equation is defined by 2\n> external equations in order to keep the continuity for probability\n> density (coherence with a probability space):\n>\n> psi(q,t)=A(q,t)exp(iS(q,t))\n> dq/dt=grad(S)\n\nYep. I prefer the formulation\n\ndq/dt = <psi J psi> / <psi psi>\n\nwhich suggests the way to generalize BM to other more general\nconfiguration spaces (which may be, for example, fields in QFT).\n\n> The interpretation problem of bohemian mechanics is the representation\n> of q(t) as the motion of a "real" vs "fictive" particle: Does this\n> virtual particle give an additional testable information compared with\n> the state |psi> alone? Does this representation is independent from\n> the state psi(t)?\n\nHm, essentially it is the state q which we observe. psi(q) is not what we\nobserve, it is much more our theoretical reconstruction than q.\n\n> If yes, QM mechanics needs an additional equation to be complete.\n\nYep. BTW, in Nelsonian stochastics we use a Marcov process to\ndefine the evolution of q. The predictions are also identical to QM.\nThus, the point is that we do not need the guiding equation to predict\nthe distributions in quantum equilibrium. Which are the observables\nwe use to test QM.\n\n> > I would say a Bohmian interpretation is such a (quite obvious)\n> > Kolmogorovian description.\n>\n> We can formally associate to the bohmian particle state the classical\n> probability space (O_q, sigma algebra F_q_psi(q,t), probability law\n> |psi(q,t)|^2dq) with the question what does q(t) "really" represent?\n\nThe pair (q(t),psi(q,t)) represents the state of reality at the moment t.\n\nThe probability distribution in the quantum equilibrium state defined\nby the wave function psi_0(q) is\n\n|psi_0(q)|^2 delta(psi-psi_0)\n\nwhere delta is a delta functional on the space of wave functions.\n\nThere is not more we can ask from a realistic theory. It specifies\nreality in a mathematical way, using mathematical objects. Once\nthis description is given, the theory remains silent.\n\n> One important point is that this probability space depends on both\n> this external variable q and the probability law "psi(q,t)" attached\n> to this external variable.\n\nThe function psi(q) is (in BM) part of the reality.\nNot a "probability law".\n\nrho(q,psi)=|psi_0(q)|^2 delta(psi-psi_0) is a probability law\nwhich depends on above parameters.\n\n> For example in a general bohemian mechanics, I may decide to use the\n> observable P to describe the motion of a new virtual particle p. I\n> thus define the equations of motion:\n> Psi(p,t)=A(p,t)exp(iS(p,t))\n> dp/dt=grad(S)\n\n> Thus p(t) is also a candidate to describe a virtual bohemian particle.\n> Is this particle much real or less "real" than the bohemian particle\n> q(t)?\n\nIt is simply another realistic theory.\n\nThere is no theory-independent notion of reality. The theory defines\nwhat is real. If BM_q is true, than q is real. If BM_p is true, than\np is real and q not.\n\nTo decide between these two theories seems impossible in quantum\nequilibrium. Nonetheless, that there are other theories which may\nbe used to explain all observations does not give in itself an argument\nagainst a particular theory.\n\nIf there is an alternative theory, which is simpler, then, indeed, it is\nreasonable to prefer this alternative theory. But if there is no simpler\ntheory, and all known alternatives have similar complexity, then\nwe can choose one of them by convention. And it is reasonable to\nremember that these alternatives exist. Each alternative may be,\nfor example, a starting point for different theories.\n\nIn case of comparison of BM_q and BM_p I prefer BM_q simply\nby a feeling that it seems more natural. Moreover, I think about\nalternatives like Nelsonian stochastics. In NS, the property of\nthe momentum part being of the form p^2 is used, thus, there is no\nsimilar p version of NS.\n\n> With these external definitions, I have defined a new "bohemian" like\n> probability space on the p random variable, and I may define many\n> others. The important point is that I have no way to get an\n> independant bijection between this probability set of outcomes (values\n> of p) and the one for q random variable (otherwise, I will have\n> quantum observables that commute).\n\nOk. But I don\'t know why it should be a problem for a particular theory,\nsay BM_q, that there is no simple connection to another, alternative\ntheory as BM_p. If BM_q is true, BM_p is simply false.\n\n> Now, when we apply, formally, the bohemian mechanics to the EPR state,\n> we have two entangled particles, where the statistical motion of q1(t)\n> depends on the motion of q2(t) through the common state psi(t) of the\n> 2 particles (i.e. through a quantum potential U(q1,q2,t)).\n>\n> The main problem come from the bohemian quantum potential U that seems\n> to be a FTL interaction like (in a given frame). Nevertheless, we must\n> not forget that the equation of motion of the bohemian particle is\n> outside the scope of the classical QM (they do not describe "real\n> particles"). This is the interpretation that may lead to the\n> impression that FTL signals exist (like the common energy of 2\n> particles).\n\nIt is simply fact that if BM_q is true then FTL causal influences exist.\n\n(In BM_p it is not straightforward how to define Einstein locality.)\n\n> Therefore, we just can say that we do not know what type of "virtual\n> particle" the equation of motion of q1(t) and q2(t) describes when\n> psi(q1,q2,t) describes an entangled state.\n\n??? The equation of BM_q describes the equation of motion of\nq1 and q2. The configuration space is (q1,q2). Reality is described\nby (q1,q2,psi(q1,q2))\n\nThere exists also a 2-particle-theory BM_p which describes the\nevolution of p1 and p2. Or even theories which describe the evolution\nof q1 and p2 and so on. Theories which we reject as less natural,\nbecause unnecessary asymmetric.\n\n> For example, in an entangled state, I may interpret that q1(t) (and\n> q2(t)) describes the physical path where both particle 1 and particle\n> 2 may be located (exclusive outcomes). Therefore, in this\n> interpretation, it is the bohmian qi(t) random variable that is no\n> more local (i.e. like using the common energy of 2 particles: Ea=E1-E2\n> and Eb=E1+E2)! Therefore U(q1,q2,t) does not describe an FTL\n> interaction between the "real" particles 1 and 2.\n\nI don\'t follow. The complete description of reality contains the position\n(and not the momentum) of above particles q1, q2, and the wavefunction\npsi(q1,q2). The evolution equation dq1/dt = ... depends on the position\nof the other particle q2 at the same moment t and is therefore an\nFTL causal influence.\n\nThe attribute "virtual" is not defined in this theory.\n\n> Note that I prefer this interpretation of Bohmian mechanics (the\n> interpretation of the virtual particles depends on the state psi(t))\n\nOh, I disagree with that interpretation, AFAIU it. BM_q is based\non the interpretation of position q as part of reality. This does not\ndepend on psi. There are states (q, delta(q-q0)) as well as\nstates (q, exp(ihp0q)). The space of q is always the same.\n\n> Cost of implementation of EPR states with classical (for fun :):\n> ================================================== ==============\n> I like their simple model because we have an example of the cost of\n> the implementation of QM properties into "classical" apparatuses.\n>\n> With respect to the PR machine alone, it works either on a Galilean\n> invariance or Lorentz invariance and we can implement it without the\n> requirement of any FTL signal or interaction:\n> x and y are the binary inputs of the PR machine (with distance of\n> light years for example) and a and b are the binary outputs.\n>\n> We have a+b=x.y (modulo 2): definition of PR machine\n>\n> Now, we can redefine the input signals as x\' and y\' and a local\n> external source that generates the signals x\',y\' towards the PR\n> machine (2 different space directions).\n> Now the signals generated by the common source are :\n> x\'=x+y+x.y and\n> y\'= x+y\n>\n> Now the PR machine may be implemented with a=x\' an b=y\' in this\n> context. We have a+b=x.y correlation on 2 points separated by several\n> light years without any FTL signal! (we have a virtual PR machine over\n> a wide space)\n\n>From point of view of classical realism your PR machine contains\nFTL. Last not least, you cannot compute x.y(mod 2) without knowing\nx and y. The information about above values x, y is used in x.y.\nTherefore it is used in x\' as well as y\' too.\n\n>You can see that I have not required causality, nor FTL, etc\n\nNo, I don\'t understand why you claim that it does not need FTL\ninformation transfer.\n\n> On the FTL phone black box\n> ==========================\n> You have used the causality and reality arguments in the hypothetical\n> FTL phone.\n> I think, I have shown that at least your causality definition is at\n> least not logically coherent with your argumentation.\n\nThis was a misunderstanding about the definition of causality.\n\n> Your "Reality" may be interpreted in several ways and I still do not\n> know what logical base you use to build a hypothetical FTL phone.\n\nIt is used only as a counterargument (for some class of arguments\noften used to defend Einstein causality in light of the violation of\nBells inequality).\n\nIt does not work against all counterarguments.\n\n> >> Well, I should first know what a FTL phone is.\n>\n> > A black box which looks like a phone and works like a phone\n> > but allows phone calls to Mars without observable time delay.\n> >(or time delay much smaller than the one predicted by c as the\n> > maximal speed).\n> > It is your choice how to test this black box. You should explain\n> > how you propose to test that the black box really violates Einstein\n> > causality without opening it. Once you have established, by\n> > observation, that, indeed, it violates, then you should be able\n> > to reject various solipsistic criticism which, nonetheless, tries to\n> > save Einstein causality.\n>\n> I cannot resist again :).\n>\n> S.R. and any deterministic theory (including the deterministic time\n> evolution of the QM state) can implement this type of fuzzy definition\n> without requiring FTL interactions: we just require that the 2 talkers\n> have their text given by a single source. This single source is hidden\n> to the experimenter. See the previous section "Cost of implementation\n> of EPR states with classical (for fun :)" for this example.\n\nTo exclude this type of FTL phone which is consistent with SR\nthe experimenter is free to create the input on above sides.\nFor example, by using his own free choice, at least at one end,\nor using a friend of his choice at the other end.\n\nIs it now possible to falsify Einstein causality by observing this\nblack box in such ways?\n\n> > I think, instead, that this is the power of my FTL phone argument.\n> > It shows that some types of fuzzy argumentation are that fuzzy that\n> > they allow to explain away even a working FTL phone. Considering\n> > arguments which work with this type of fuzziness, you should think\n> > about the possibility of explaining away the obvious.\n\n> Well the problem with the fuzziness of models is that somewhere you\n> loose the consistency (i.e. you cannot demonstrate that they are\n> consistent, because you have "nothing").\n\nIn applications of my FTL phone argument its not me who is\nlosing consistency. Instead, I try to show in this way that\nthe argument proposed (by my opponent) to defend relativity\nis inconsistent.\n\n> > Last not least, we know that there are always non-contradictionary\n> > but nonsensical theories like consequent solipcism or various\n> > fundamentalistic versions of religions. That means, it is never\n> > possible to\n> > _prove_ something in science in full beauty. Certain types of doubt\n> > should be rejected to be able to do science.\n\n> I think you really like philosophy and negative logic ; ), and I am\n> always afraid when we are using raw philosophical deductions in\n> physics.\n\nOf course, using philosophy in science is often dangerous. But\nit is the standard interpretation of quantum theory (Bohr and his\ncompany) who have introduced questionable philosophy into\nquantum theory.\n\n> I prefer to stay at a mathematical level and Godel theorem.\n\nBe careful. There is a geneal rule which claims that every\napplication of Goedels theorem outside a very small domain\nof highly specialized pure logic is nonsensical.\n\n> I prefer\n> to say that we cannot prove if a self-consistent mathematic theory is\n> true or false.\n\nIf the theory is a physical theory, it has to make\nnontrivial predictions (Popper\'s criterion).\nIf it does, it may be falsified - if the predicted\nthing does not happen, it is false. If this is\nimpossible, the predictions are (by definition)\ntrivial.\n\nThus, physical theories may be shown to be false\n(of course only if they are false).\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"seratend" <ser_monmail@yahoo.fr> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> On the Causality "->" definition:
> ====================
> >>> No. In mathematical terms it means absence of independence.
> >> Whow, what a very weak and fuzzy definition.
>
> > Certainly not fuzzy. For observables P(AB) = P(A)P(B) means
> > independence.
>
> Ok, I will say this relation means probabilistic independence (math
> definition). I assume that we are speaking about classical
> probability.
>
> So with your definition, if P(AB) = P(A)P(B) then the 2 relations "A
> -> B" and "B ->A" are not true. ...
> Thus, we can't have (A-> B) and not (A->B) because:
> not (A -> B) => P(A,B=a,b) = P(A=a)P(B=b) => not (B-> A).

Of course, the symmetric definition P(AB) = P(A)P(B)
does not define the antisymmetric relation A->B.
Statistical independence is one thing, causal independence
another one. P(AB) =/= P(A)P(B) means statistical dependence.
Which may be caused by causal dependence A->B or B->A,
but as well by a common cause C->A, C->B.

In a causal realistic theory (which is the context of the definition of
causal influence) we have some external reality l in /\, some
"decisions of experimenters" or parameters of the experimental
device a_i, and some results of the experiment A_i.
The realistic theory should explain the observable results in the
following way

<A_i> = \int A(a1,...,l) \rho(l) dl.

Now, we have a_j -/-> A_i if

<A_i> = \int A(a_1,...a_j-1,a_j+1,...[/itex],l) \rho(l) dl.

This is also precise enough, I hope. This covers only a particular
case of causal influences - the one where the cause is some
decision of the experimenter a_j - but this one is sufficient for
the proof of Bell's inequality.

I'm not sure how to formulate it in the general context, where
the causal realistic theory is not specified. Usually the definition
is quite obvious once a causal realistic theory is given. For
example, in classical GR we have A->B => A is in the past
light cone of B. And in more detail the causal connections follow
from the evolution equations of the given theory. Thus, "causality"
is something which has to be defined in detail (specified) only
by the given theory, for example, via an evolution equation.

> > Everything else is a nontrivial correlation which requires a nontrivial
> > realistic explanation.

If we observe a statistical correlation, that means,
P(AB) =/= P(A)P(B),
we need a realistic explanation, that means, an
explanation of type
A->B or B->A or
exists C with C->A, C->B.

That's my reformulation of the EPR criterion of reality.

> > In this sense, A->B denotes some fundamental, basic
> > dependence as part of
> > some ontological theory. Thus, the theory specifies
> > what exists (ontology)
> > and what depends on what (causality, usually
> > defined by laws of evolution).

> I continue to say that you are using too many different concepts, not
> well defined, to make some logical deductions.

I hope I have now better defined them.

> > I would say that's not all. A->B is a relation, it is transitive, and an
> > ordering: A->B and B->A means A=B. There are no closed
> > causal loops except trivial ones A->A.

These are general properties of the relation A->B. As said, the
definition has to be given by the realistic theory.

> About your Reality, ontology :
> ==============================
> In your arguments, you are also using many times the word reality. I
> really think that this term as well as causality is not well defined
> (in a self consistent mathematical basis).
>
> We thus have the same previous danger in using it (risk of
> inconsistent argumentation).

I think it is enough well-defined, and this well-defined notion of
realistic theory is what has been used by Bell in his proof.

> Once again, I do not understand your reality statement/meaning. Bell's
> theorem deals only with the impossibility of a probabilistic space
> with local non-contextual random variables to reproduce the QM
> probability results in a particular state. This is the logical
> statement.

I would agree except that "local" is not the point. Bell's theorem
excludes large classes of global and/or contextual hidden variables
as well.

> The word "reality" is an interpretation (it only makes a weak
> connection between the theorem and the initial EPR paper where the
> word itself is re-interpreted).

The initial EPR paper is not that important. In the case you win
some part of this discussion and the concept of "realistic causal
theory" is not yet enough clarified/specified I would try to correct
this using a specification which allows to prove Bell's inequality.
That means I would use a definition of "realistic theory" which
allows to prove

<AB> = \int A(a,b,l) B(a,b,l) \rho(l) dl.

and a definition of "causal theory" which allows to derive from this
and "not (a->B and b->A)" the formula

<AB> = \int A(a,l) B(b,l) \rho(l) dl.

This certainly defines one possible notion of "realistic causal
theory". And for this particular notion of realism, (essentially
by construction) we can prove Bell's inequality.

Then, what remains is the metaphysical thesis that this definition
(specification) of the philosophical notion "realism" is appropriate.
This is already something impossible to prove, impossible in
principle. But I would say that from metaphysical point of
view this is a very weak notion of realism. Last not least, lot's
of "theories" which in common sense are not realistic (theories
involving ghosts, gods and so on) are realistic in this sense.

Of course, it seems possible to think about meaningful extensions
of this notion of realism, to define some class of "weak realistic
theories". But a meaningful notion of weak realism should be
in some sense stronger than the QM "shut up and calculate"
interpretation, in the sense that the proposal of a weak realistic
theory which is able to describe a given set of observations
is an explanation in some meaningful sense. "Shut up and
calculate" is not an explanation, but the rejection to give
explanations.

> Therefore, argumentation/deduction on a "reality" can only be made
> with an external self-consistent definition (with its domain of
> validity). This is almost equivalent to create a physical theory of
> "reality" even if I still do not know what is "reality".
> I am not sure if this theory is useful for the Bell theorem and FTL
> interactions problem.

I'm not thinking about a physical theory of reality, but about
a characterization of a class of physical theories as realistic theories.

Of course, that's not that much different. We can understand GR
as a class of physical theories with some g_{ij}(x) and some matter
fields. The physical theory is defined only if we define the matter fields
and the matter Lagrangian. In the same sense, "realistic theory"
describes a general scheme which should be used by any realistic
theory. This scheme specifies, for example, that some set should
be used to describe all possible states of reality, and there should be
some probability measure on it. And a causal theory is one where
we have some relation A->B defined.

> > Reality is some element of some set of possible realities
> > (using set theory)
> > with classical probability measure \rho.

> I may reinterpret your reality definition in mathematical words. We
> keep the consistency with the previous causality definition: we are
> using the classical probability spaces. I have two possible
> interpretations:
>
> First interpretation:
> - Classical probability measure: you have a probability space (O,\sigma
> algebra L, probability \rho(o).do)
> - Reality is an event l, an element of the \sigma algebra L
> - set of possible realities= \sigma algebra L
>
> With this first interpretation, I do not know what an outcome (an
> element of O) is.

In the description I have used above, an observable.

The reality L describes the states of a black box. The experiment consists
of making some input by free decisions and then observing some output.

The experimenter is, in this case, not described as part of the reality.
Nor the way he defines the input, nor the way he memorizes the output.
For the part of the experimenter we postulate free will and an ideal memory.

(Of course, as realists in the philosophical sense, we would like to have
also a theory which is able to describe the whole universe including the
observer.
BM is an example of a theory which is able to solve this problem. You can
describe the whole universe, and you can reduce the problem, consider the
classical observer as independent, and the reduced theory, which describes
only the black box, has the same form as BM. Nonetheless, considering only
such a special subtype of theory would be too restrictive. There are
realistic theories which do not cover the observer too.)

> > This is part of the general scheme of ontological theories: Each theory
> > worth to be named realistic (ontological, having an ontology) has such a
> > set. There is no other restriction on this set.

> Always the anylogy wording. : ). Do you simply mean a logical
> self-consistent theory (mathematical interpretation)?

No. QM in its minimal interpretation is logical self-consistent
but not realistic.

> > In a realistic theory, I would name it local if L = some set of
> > functions
> > on some manifold X named space. This would allow FTL, therefore,
> > has nothing to do with Einstein causality.
>
> Well, It seems that you have restricted the possible choices of a
> possible theory on a theory based on a single classical probability
> space (with a single probability law).
> You need to specify the set of outcomes L, the \sigma-algebra and the
> probability law as well as the random variables.

This is what the realistic theory has to specify. Different realistic
theories have different sets of observables, different sets of realities,
and so on.

> However, with this few assumptions, we can already say that this
> theory is included in the classical QM probability (we are restricting
> the set of QM observables to one subset of commuting observables: the
> random variables on the classical probability space).

QM does not define ontology. If you would like to say that QM can
be easily modified in such a way that it fits into the scheme - no problem,
I have no problem to accept that BM is only a quite trivial "realistic"
reinterpretation of QM. And that there may be others, for example
the collapse interpretation.

> On the classical Bohmian mechanics:
> ==================================
> > We can do more. We can decide that only the thing named "statistical
> > hidden
> > variable model" is worth to be considered as a scientific explanation,
> > and
> > once we have found such explanations (like those given by Bohmian
> > mechanics)
> > everything which gives less (like quantum mechanics) should be rejected.
> > This is a methodological decision, a decision about the nature of the
> > scientific method.

> First, I have to re-specify what Bohmian mechanics really is (for me,
> only on the logical mathematic content, not the interpretations we may
> find around the world).
> Bohmian mechanics is the re-formulation of a quantum system through a
> particular observable plus an additional hypothesis on the motion
> equation of this observable.
>
> Bohmian mechanics has selected the Q observable (position), but we can
> choose, formally, any observable of QM (P, combination of P,Q, .).
> The bohemian system state (a single particle) is described by the
> couple (q(t), \psi(q,t)).

Correct. (Except the 'e' in 'bohemian')

> -- \psi(q,t) is the state in the |q> basis. Its evolution is controlled
> by the classical QM Schroedinger equation and the QM postulates.
>
> -- q(t) in bohemian mechanics is an external random variable. This
> externally added feature defines a local (relative to the variable q)
> classical probability space.
>
> The equation describing the statistical motion of q(t) is an external
> addition to QM formulation. Its motion equation is defined by 2
> external equations in order to keep the continuity for probability
> density (coherence with a probability space):
>
> \psi(q,t)=A(q,t)\exp(iS(q,t))
> dq/dt=grad(S)

Yep. I prefer the formulation

dq/dt = <\psi J \psi> / <\psi \psi>

which suggests the way to generalize BM to other more general
configuration spaces (which may be, for example, fields in QFT).

> The interpretation problem of bohemian mechanics is the representation
> of q(t) as the motion of a "real" vs "fictive" particle: Does this
> virtual particle give an additional testable information compared with
> the state |\psi> alone? Does this representation is independent from
> the state \psi(t)?

Hm, essentially it is the state q which we observe. \psi(q) is not what we
observe, it is much more our theoretical reconstruction than q.

> If yes, QM mechanics needs an additional equation to be complete.

Yep. BTW, in Nelsonian stochastics we use a Marcov process to
define the evolution of q. The predictions are also identical to QM.
Thus, the point is that we do not need the guiding equation to predict
the distributions in quantum equilibrium. Which are the observables
we use to test QM.

> > I would say a Bohmian interpretation is such a (quite obvious)
> > Kolmogorovian description.
>
> We can formally associate to the bohmian particle state the classical
> probability space (O_q, \sigma algebra F_{q_psi}(q,t), probability law
> |\psi(q,t)|^2dq) with the question what does q(t) "really" represent?

The pair (q(t),\psi(q,t)) represents the state of reality at the moment t.

The probability distribution in the quantum equilibrium state defined
by the wave function \psi_0(q) is

|\psi_0(q)|^2 \delta(\psi-\psi_0)

where \delta is a \delta functional on the space of wave functions.

There is not more we can ask from a realistic theory. It specifies
reality in a mathematical way, using mathematical objects. Once
this description is given, the theory remains silent.

> One important point is that this probability space depends on both
> this external variable q and the probability law "\psi(q,t)" attached
> to this external variable.

The function \psi(q) is (in BM) part of the reality.
Not a "probability law".

\rho(q,\psi)=|\psi_0(q)|^2 \delta(\psi-\psi_0) is a probability law
which depends on above parameters.

> For example in a general bohemian mechanics, I may decide to use the
> observable P to describe the motion of a new virtual particle p. I
> thus define the equations of motion:
> \Psi(p,t)=A(p,t)\exp(iS(p,t))
> [itex]dp/dt=grad(S)

> Thus p(t) is also a candidate to describe a virtual bohemian particle.
> Is this particle much real or less "real" than the bohemian particle
> q(t)?

It is simply another realistic theory.

There is no theory-independent notion of reality. The theory defines
what is real. If BM_q is true, than q is real. If BM_p is true, than
p is real and q not.

To decide between these two theories seems impossible in quantum
equilibrium. Nonetheless, that there are other theories which may
be used to explain all observations does not give in itself an argument
against a particular theory.

If there is an alternative theory, which is simpler, then, indeed, it is
reasonable to prefer this alternative theory. But if there is no simpler
theory, and all known alternatives have similar complexity, then
we can choose one of them by convention. And it is reasonable to
remember that these alternatives exist. Each alternative may be,
for example, a starting point for different theories.

In case of comparison of BM_q and BM_p I prefer BM_q simply
by a feeling that it seems more natural. Moreover, I think about
alternatives like Nelsonian stochastics. In NS, the property of
the momentum part being of the form p^2 is used, thus, there is no
similar p version of NS.

> With these external definitions, I have defined a new "bohemian" like
> probability space on the p random variable, and I may define many
> others. The important point is that I have no way to get an
> independant bijection between this probability set of outcomes (values
> of p) and the one for q random variable (otherwise, I will have
> quantum observables that commute).

Ok. But I don't know why it should be a problem for a particular theory,
say BM_q, that there is no simple connection to another, alternative
theory as BM_p. If BM_q is true, BM_p is simply false.

> Now, when we apply, formally, the bohemian mechanics to the EPR state,
> we have two entangled particles, where the statistical motion of q1(t)
> depends on the motion of q2(t) through the common state \psi(t) of the
> 2 particles (i.e. through a quantum potential U(q1,q2,t)).
>
> The main problem come from the bohemian quantum potential U that seems
> to be a FTL interaction like (in a given frame). Nevertheless, we must
> not forget that the equation of motion of the bohemian particle is
> outside the scope of the classical QM (they do not describe "real
> particles"). This is the interpretation that may lead to the
> impression that FTL signals exist (like the common energy of 2
> particles).

It is simply fact that if BM_q is true then FTL causal influences exist.

(In BM_p it is not straightforward how to define Einstein locality.)

> Therefore, we just can say that we do not know what type of "virtual
> particle" the equation of motion of q1(t) and q2(t) describes when
> \psi(q1,q2,t) describes an entangled state.

??? The equation of BM_q describes the equation of motion of
q1 and q2. The configuration space is (q1,q2). Reality is described
by (q1,q2,\psi(q1,q2))

There exists also a 2-particle-theory BM_p which describes the
evolution of p1 and p2. Or even theories which describe the evolution
of q1 and p2 and so on. Theories which we reject as less natural,
because unnecessary asymmetric.

> For example, in an entangled state, I may interpret that q1(t) (and
> q2(t)) describes the physical path where both particle 1 and particle
> 2 may be located (exclusive outcomes). Therefore, in this
> interpretation, it is the bohmian qi(t) random variable that is no
> more local (i.e. like using the common energy of 2 particles: Ea=E1-E2
> and Eb=E1+E2)! Therefore U(q1,q2,t) does not describe an FTL
> interaction between the "real" particles 1 and 2.

I don't follow. The complete description of reality contains the position
(and not the momentum) of above particles q1, q2, and the wavefunction
\psi(q1,q2). The evolution equation dq1/dt = ... depends on the position
of the other particle q2 at the same moment t and is therefore an
FTL causal influence.

The attribute "virtual" is not defined in this theory.

> Note that I prefer this interpretation of Bohmian mechanics (the
> interpretation of the virtual particles depends on the state \psi(t))

Oh, I disagree with that interpretation, AFAIU it. BM_q is based
on the interpretation of position q as part of reality. This does not
depend on \psi. There are states (q, \delta(q-q0)) as well as
states (q, \exp(ihp0q)). The space of q is always the same.

> Cost of implementation of EPR states with classical (for fun :):
> ================================================== ==============
> I like their simple model because we have an example of the cost of
> the implementation of QM properties into "classical" apparatuses.
>
> With respect to the PR machine alone, it works either on a Galilean
> invariance or Lorentz invariance and we can implement it without the
> requirement of any FTL signal or interaction:
> x and y are the binary inputs of the PR machine (with distance of
> light years for example) and a and b are the binary outputs.
>
> We have a+b=x.y (modulo 2): definition of PR machine
>
> Now, we can redefine the input signals as x' and y' and a local
> external source that generates the signals x',y' towards the PR
> machine (2 different space directions).
> Now the signals generated by the common source are :
> x'=x+y+x.y and
> y'= x+y
>
> Now the PR machine may be implemented with a=x' an b=y' in this
> context. We have a+b=x.y correlation on 2 points separated by several
> light years without any FTL signal! (we have a virtual PR machine over
> a wide space)

>From point of view of classical realism your PR machine contains
FTL. Last not least, you cannot compute x.y(mod 2) without knowing
x and y. The information about above values x, y is used in x.y.
Therefore it is used in x' as well as y' too.

>You can see that I have not required causality, nor FTL, etc

No, I don't understand why you claim that it does not need FTL
information transfer.

> On the FTL phone black box
> ==========================
> You have used the causality and reality arguments in the hypothetical
> FTL phone.
> I think, I have shown that at least your causality definition is at
> least not logically coherent with your argumentation.

This was a misunderstanding about the definition of causality.

> Your "Reality" may be interpreted in several ways and I still do not
> know what logical base you use to build a hypothetical FTL phone.

It is used only as a counterargument (for some class of arguments
often used to defend Einstein causality in light of the violation of
Bells inequality).

It does not work against all counterarguments.

> >> Well, I should first know what a FTL phone is.
>
> > A black box which looks like a phone and works like a phone
> > but allows phone calls to Mars without observable time delay.
> >(or time delay much smaller than the one predicted by c as the
> > maximal speed).
> > It is your choice how to test this black box. You should explain
> > how you propose to test that the black box really violates Einstein
> > causality without opening it. Once you have established, by
> > observation, that, indeed, it violates, then you should be able
> > to reject various solipsistic criticism which, nonetheless, tries to
> > save Einstein causality.
>
> I cannot resist again :).
>
> S.R. and any deterministic theory (including the deterministic time
> evolution of the QM state) can implement this type of fuzzy definition
> without requiring FTL interactions: we just require that the 2 talkers
> have their text given by a single source. This single source is hidden
> to the experimenter. See the previous section "Cost of implementation
> of EPR states with classical (for fun :)" for this example.

To exclude this type of FTL phone which is consistent with SR
the experimenter is free to create the input on above sides.
For example, by using his own free choice, at least at one end,
or using a friend of his choice at the other end.

Is it now possible to falsify Einstein causality by observing this
black box in such ways?

> > I think, instead, that this is the power of my FTL phone argument.
> > It shows that some types of fuzzy argumentation are that fuzzy that
> > they allow to explain away even a working FTL phone. Considering
> > arguments which work with this type of fuzziness, you should think
> > about the possibility of explaining away the obvious.

> Well the problem with the fuzziness of models is that somewhere you
> loose the consistency (i.e. you cannot demonstrate that they are
> consistent, because you have "nothing").

In applications of my FTL phone argument its not me who is
losing consistency. Instead, I try to show in this way that
the argument proposed (by my opponent) to defend relativity
is inconsistent.

> > Last not least, we know that there are always non-contradictionary
> > but nonsensical theories like consequent solipcism or various
> > fundamentalistic versions of religions. That means, it is never
> > possible to
> > _prove_ something in science in full beauty. Certain types of doubt
> > should be rejected to be able to do science.

> I think you really like philosophy and negative logic ; ), and I am
> always afraid when we are using raw philosophical deductions in
> physics.

Of course, using philosophy in science is often dangerous. But
it is the standard interpretation of quantum theory (Bohr and his
company) who have introduced questionable philosophy into
quantum theory.

> I prefer to stay at a mathematical level and Godel theorem.

Be careful. There is a geneal rule which claims that every
application of Goedels theorem outside a very small domain
of highly specialized pure logic is nonsensical.

> I prefer
> to say that we cannot prove if a self-consistent mathematic theory is
> true or false.

If the theory is a physical theory, it has to make
nontrivial predictions (Popper's criterion).
If it does, it may be falsified - if the predicted
thing does not happen, it is false. If this is
impossible, the predictions are (by definition)
trivial.

Thus, physical theories may be shown to be false
(of course only if they are false).

Ilja

Oz

Nov17-04, 11:13 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nIlja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes\n>"Blake Winter" <blake.winter@houghton.edu> schrieb\n>> Oz <oz@farmeroz.port995.com> wrote\n>> Right, that\'s why hidden variables are disproven by the violations of\n>> Bell inequalities - at least, naive hidden variables are.\n>\n>Sorry, Bohmian mechanics is a hidden variable theory. And a\n>quite simple one. That simple that Einstein has considered it\n>too simple.\n>\n>> By naive\n>> hidden variables, I mean that if we still want hidden variables we\'ve\n>> got to either violate causality (either by violations of relativistic\n>> locality, or else by having closed causal loops), or else we\'ve got to\n>> have a nontrivial topology.\n>\n>Bohmian mechanics has, indeed, hidden causal influences in a hidden\n>preferred frame.\n\nEh? But not hidden acausal influences in a hidden preferred frame?\n\n>Of course, some people do not like hidden preferred frames and hidden\n>variables in general. But sometimes the argumentation is quite absurd:\n>The preferred frame is rejected (in special relativity) because it is\n>hidden.\n>Bohmian mechanics is rejected because it has a preferred frame, which\n>does not agree with relativistic ideology. And hidden variable theories\n>in general are rejected for the same reason, because the violation of Bell\'s\n>inequality proves that they need a preferred frame.\n\nComing from a position of infinite ignorance I have tended to view\nentangelement slightly differently. Its very unclear to me that\nparticles, as extended objects (be they waves or some probabilistic\ndistribution where a particle can be found) exist in global spacetime\nuntil they interact with it \'enough\'. So I tend to think of entangled\nparticles as being a single particle, and thus \'infinitely close to\nitself\' (no matter what the rest of the universe thinks) until it\ninteracts. Well, it saves me from melting my brain cell, anyway. Of\ncourse this view can be extended.\n\n>But if we combine these arguments, it appears that the argument against\n>hidden variables is that it uses hidden variables. Because there is no\n>other argument against the preferred frame.\n\nI don\'t think I know enough to follow your (and others) arguments.\nIf having a hidden frame and hidden particles does not contradict\nexperiment, then it seems to me that it is a valid model.\n\n>> Most\n>> people are at peace with this, because it doesn\'t actually change the\n>> local probabilities, because you don\'t actually know what the\n>> measurement at the other photon was. This is the famous result that\n>> you can\'t send information using EPR effects.\n>\n>I don\'t understand why people consider this as an important argument.\n>Last not least, every observation which leaves only two explanations\n>(A->B or B->A) obviously proves that there exists FTL causal influences.\n>\n>But, on the other hand, every effect which leaves these two explanation\n>has the same property: It cannot be used for information transfer.\n>Indeed, an application for information transfer A->B would be in\n>contradiction with the possible explanation B->A and reverse.\n\nOh. I (think) I see what you mean.\n\n>Thus, accepting this argument means preserving Einstein causality\n>even in a situation where it can be logically proven to be false.\n\n<meltdown...>\n\n--\nOz\nThis post is worth absolutely nothing and is probably fallacious.\n\nUse oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].\nBTOPENWORLD address has ceased. DEMON address has ceased.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Ilja Schmelzer <Ilja.Schmelzer@FernUni-Hagen.de> writes
>"Blake Winter" <blake.winter@houghton.edu> schrieb
>> Oz <oz@farmeroz.port995.com> wrote
>> Right, that's why hidden variables are disproven by the violations of
>> Bell inequalities - at least, naive hidden variables are.
>
>Sorry, Bohmian mechanics is a hidden variable theory. And a
>quite simple one. That simple that Einstein has considered it
>too simple.
>
>> By naive
>> hidden variables, I mean that if we still want hidden variables we've
>> got to either violate causality (either by violations of relativistic
>> locality, or else by having closed causal loops), or else we've got to
>> have a nontrivial topology.
>
>Bohmian mechanics has, indeed, hidden causal influences in a hidden
>preferred frame.

Eh? But not hidden acausal influences in a hidden preferred frame?

>Of course, some people do not like hidden preferred frames and hidden
>variables in general. But sometimes the argumentation is quite absurd:
>The preferred frame is rejected (in special relativity) because it is
>hidden.
>Bohmian mechanics is rejected because it has a preferred frame, which
>does not agree with relativistic ideology. And hidden variable theories
>in general are rejected for the same reason, because the violation of Bell's
>inequality proves that they need a preferred frame.

Coming from a position of infinite ignorance I have tended to view
entangelement slightly differently. Its very unclear to me that
particles, as extended objects (be they waves or some probabilistic
distribution where a particle can be found) exist in global spacetime
until they interact with it 'enough'. So I tend to think of entangled
particles as being a single particle, and thus 'infinitely close to
itself' (no matter what the rest of the universe thinks) until it
interacts. Well, it saves me from melting my brain cell, anyway. Of
course this view can be extended.

>But if we combine these arguments, it appears that the argument against
>hidden variables is that it uses hidden variables. Because there is no
>other argument against the preferred frame.

I don't think I know enough to follow your (and others) arguments.
If having a hidden frame and hidden particles does not contradict
experiment, then it seems to me that it is a valid model.

>> Most
>> people are at peace with this, because it doesn't actually change the
>> local probabilities, because you don't actually know what the
>> measurement at the other photon was. This is the famous result that
>> you can't send information using EPR effects.
>
>I don't understand why people consider this as an important argument.
>Last not least, every observation which leaves only two explanations
>(A->B or B->A) obviously proves that there exists FTL causal influences.
>
>But, on the other hand, every effect which leaves these two explanation
>has the same property: It cannot be used for information transfer.
>Indeed, an application for information transfer A->B would be in
>contradiction with the possible explanation B->A and reverse.

Oh. I (think) I see what you mean.

>Thus, accepting this argument means preserving Einstein causality
>even in a situation where it can be logically proven to be false.

<meltdown...>

--
Oz
This post is worth absolutely nothing and is probably fallacious.

Use oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].
BTOPENWORLD address has ceased. DEMON address has ceased.

seratend

Nov17-04, 11:14 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n\n"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cncbrf\\$esn\\$1@beech.fernuni-hagen.de>...\n> "Blake Winter" <blake.winter@houghton.edu> schrieb\n> > Oz <oz@farmeroz.port995.com> wrote\n> > Right, that\'s why hidden variables are disproven by the violations of\n> > Bell inequalities - at least, naive hidden variables are.\n>\n> Sorry, Bohmian mechanics is a hidden variable theory. And a\n> quite simple one. That simple that Einstein has considered it\n> too simple.\n>\n> > By naive\n> > hidden variables, I mean that if we still want hidden variables we\'ve\n> > got to either violate causality (either by violations of relativistic\n> > locality, or else by having closed causal loops), or else we\'ve got to\n> > have a nontrivial topology.\n>\n> Bohmian mechanics has, indeed, hidden causal influences in a hidden\n> preferred frame.\n\nBohmian Mechanics is mainly QM with the observable position and an\nextra motion equation (q(t)). The observable Q may be expressed in any\nrelativistic frame and thus BM q(t). So I really do not see the\npreferred frame.\n\nAll the problems of causality with BM comes from the interpretation of\nq(t). As long as we cannot get information about q(t) from an\nexperiment different from what a QM Position measurement gives, we can\nquestion the interpretation of q(t) and the deductions based on this\ninterpretation (requirement of a preferred frame?).\n\n>\n> Of course, some people do not like hidden preferred frames and hidden\n> variables in general. But sometimes the argumentation is quite absurd:\n> The preferred frame is rejected (in special relativity) because it is\n> hidden.\n> Bohmian mechanics is rejected because it has a preferred frame, which\n> does not agree with relativistic ideology. And hidden variable theories\n> in general are rejected for the same reason, because the violation of Bell\'s\n> inequality proves that they need a preferred frame.\n>\nIt is one interpretation of q(t) that may require a preferred frame. I\na not sure this is the topic that BM is not widely used (even if I\ndon\'t know what is "widely").\nBut I can say why I am not using BM: BM adds to QM the motion of\nq(t). As long as q(t)cannot add new information that can be tested\n(relatively to QM),I don\'t want to spend time in computing q(t). It is\nvery difficult and I still have to compute psi(q,t) to get q(t). In\nQM, I just need psi(q,t). In both cases, I have the same results (i.e.\ntestable and usable information). This is very pragmatic.\n>\n> But if we combine these arguments, it appears that the argument against\n> hidden variables is that it uses hidden variables. Because there is no\n> other argument against the preferred frame.\n>\n\nWell, I can say because they are untestable variables, I surely can\nselect the interpretation and the hidden variable that is compatible\nor not with a hidden preferred frame concept.\n\n> Thus, accepting this argument means preserving Einstein causality\n> even in a situation where it can be logically proven to be false.\n>\nLogically is based on one interpretation context. So first, we should\nquestion the interpretation consistency.\n\nSeratend.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cncbrf$esn$1@beech.fernuni-hagen.de>...
> "Blake Winter" <blake.winter@houghton.edu> schrieb
> > Oz <oz@farmeroz.port995.com> wrote
> > Right, that's why hidden variables are disproven by the violations of
> > Bell inequalities - at least, naive hidden variables are.
>
> Sorry, Bohmian mechanics is a hidden variable theory. And a
> quite simple one. That simple that Einstein has considered it
> too simple.
>
> > By naive
> > hidden variables, I mean that if we still want hidden variables we've
> > got to either violate causality (either by violations of relativistic
> > locality, or else by having closed causal loops), or else we've got to
> > have a nontrivial topology.
>
> Bohmian mechanics has, indeed, hidden causal influences in a hidden
> preferred frame.

Bohmian Mechanics is mainly QM with the observable position and an
extra motion equation (q(t)). The observable Q may be expressed in any
relativistic frame and thus BM q(t). So I really do not see the
preferred frame.

All the problems of causality with BM comes from the interpretation of
q(t). As long as we cannot get information about q(t) from an
experiment different from what a QM Position measurement gives, we can
question the interpretation of q(t) and the deductions based on this
interpretation (requirement of a preferred frame?).

>
> Of course, some people do not like hidden preferred frames and hidden
> variables in general. But sometimes the argumentation is quite absurd:
> The preferred frame is rejected (in special relativity) because it is
> hidden.
> Bohmian mechanics is rejected because it has a preferred frame, which
> does not agree with relativistic ideology. And hidden variable theories
> in general are rejected for the same reason, because the violation of Bell's
> inequality proves that they need a preferred frame.
>
It is one interpretation of q(t) that may require a preferred frame. I
a not sure this is the topic that BM is not widely used (even if I
don't know what is "widely").
But I can say why I am not using BM: BM adds to QM the motion of
q(t). As long as q(t)cannot add new information that can be tested
(relatively to QM),I don't want to spend time in computing q(t). It is
very difficult and I still have to compute \psi(q,t) to get q(t). In
QM, I just need \psi(q,t). In both cases, I have the same results (i.e.
testable and usable information). This is very pragmatic.
>
> But if we combine these arguments, it appears that the argument against
> hidden variables is that it uses hidden variables. Because there is no
> other argument against the preferred frame.
>

Well, I can say because they are untestable variables, I surely can
select the interpretation and the hidden variable that is compatible
or not with a hidden preferred frame concept.

> Thus, accepting this argument means preserving Einstein causality
> even in a situation where it can be logically proven to be false.
>
Logically is based on one interpretation context. So first, we should
question the interpretation consistency.

Seratend.

Ilja Schmelzer

Nov18-04, 12:52 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"seratend" <ser_monmail@yahoo.fr> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> > A preferred frame we need if we want a realistic interpretation of\n> > QM. The simplest choice is Bohmian mechanics.\n\n> Now, I may understand that by preferred frame, you may mean a\n> preferred observable (the q position, the observable used in bohemian\n> mechanics) or a given frame (x1,x2,x3,t), thus which one?\n\nWith "preferred frame" I usually mean (a little bit sloppy) a preferred\nfoliation of spacetime.\n\n> Currently, I think we don\'t need to introduce a theory of reality as\n> it can move us towards philosophy or metaphysics rather than physics.\n\nThe justification of the scientific method is, indeed, philosophy\nor metaphysics. You cannot derive the scientific method from\nobservation or prove the scientific method by observation.\n\nAnd the discussion of EPR/Bell requires the discussion of\nthe scientific method, because this is a case where the current\nmajority position does not follow the scientific\nmethod (it rejects an essential part of the scientific method -\nthe search for realistic explanations of observed correlations).\nAt least this is my position in this question.\n\n> Therefore, I definitively prefer to concentrate on logically\n> consistent models and thereafter interpretation (connection between\n> the model results and experiment results) before speaking of what\n> reality is (which I think is out of scope at this step).\n\nIn this case, there is nothing to discuss. We agree that the\nQM "shut up and calculate" is a logically consistent way to\nobtain testable predictions which agree with observation.\n\n> Now, to try to end with the "reality" subject, can I translate what\n> you call "realistic theory" into a self-consistent mathematical model?\n> If yes, I really prefer to read "self-consistent theory" rather than\n> "realistic theory" even if it is longer and heavier.\n\nRealistic theories define a subclass of the class of consistent theories.\nQM "shut up and calculate" is a logically consistent but not realistic.\n\nYou can view realism as a restriction of the mathematical form of the\ntheory. As well as, say, relativity defines a subclass of relativistic\ntheories which are also logically consistent but have an additional\nrequirement (some symmetry group).\n\n> The main problem with bohemian mechanics is the selection of the\n> equation of motion of q(t).\n\nNo. For any given BM-like theory the equation of motion\nof q(t) is defined and fixed.\n\n> As long as Bohmian mechanics accept the\n> Schrodinger time evolution of the psi(q,t), this theory accepts that\n> we can not see the difference between a QM measurement of the\n> observable q and the bohemian measurement of the path q(t).\n\nFirst, there is no measurement of a path in BM. Second,\nwe see no difference only in quantum equilibrium. Once\nquantum equilibrium holds in the initial state it holds always.\nBut I would guess that the initial value of the universe is\nnot quantum equilibrium but a well-defined initial value q_0.\n\n> In other words we cannot have in bohemian mechanics\n> |psi(q,t)|^2=delta(q-qo(t)) (a known trajectory).\n\nThis follows from the Schroedinger equation for psi.\n\n> Once an infinite\n> precision q measurement occurs with a result qo, the bohemian velocity\n> is unknown at this time as in QM.\n\nNo. The velocity of the particle d_t q(t) is known if the position\nq(t) is known and the wave function psi(q,t) is known.\n\n> Thus, I have a theory with an externally added equation that\n> gives no more information (achievable in an experiment) than\n> the orthodox QM theory.\n\nYep. By construction. This was the purpose of BM.\n\n> Therefore, as long as this new equation of motion cannot provide\n> additional information (existence of an experiment that shows this\n> additional information), I prefer to consider Bohmian mechanics as one\n> possible interpretation that is compatible with the orthodox\n> interpretation of QM, with an additionnal cost q(t).\n\nFine. I have no problem with this point of view. Essentially, this\nis also Bell\'s who has explicitly named it realistic interpretation\n(AFAIR).\n\n> Explanation the previous statement: U(t,to) is the time evolution of\n> the initial state |psi(to)>. We have psi(q,to)=<q|psi(to)> for the\n> bohemian particle.\n\nSorry, the position of the bohmian particle is not a wave function,\nnot a bra vector <q| or a delta-function delta(q-q0). We have\npsi(q,to)=<q|psi(to)>, but this equation has nothing to do with\nthe bohmian particle.\n\n> If we define the projector P(t)=|psi(t)><psi(t)| where\n> |psi(t)>=U(t,to)|psi(to)> as the measurement at any time t>to, we have\n> with probability 100% the detection of the particle without the change\n> of the state psi(t). Thus P(t) defines the trajectory of the particle\n> with the same accuracy as the possible q(t) motions in bohemian\n> mechanics compatible with the state psi(q,t).\n\nI don\'t understand. Do you want to propose a variant of\na bohmian theory where the "preferred observable" is, instead\nof q or p, P(t)?\n\n> > The problem with relativistic QFT is that it extends the "shut up and\n> > calculate" ideology. It does not care at all about an explanation\n> > for the observed probabilities - being able to compute them is\n> > sufficient. This is not an explanation.\n\n> This is the measurement theory. It tries to explain in a\n> self-consistent way the results we have with this mathematical model\n> (interpretation consistency problems).\n> I think that the "shut up and calculate ideology" is somewhat outside\n> the context. Note that this is what does most of the engineers around\n> the world (I think).\n\nOk. I don\'t claim "shut up and calculate" is meaningless. It gives\ntestable predictions, in a consistent way. It does not give explanations.\n\n> Note that Bohmian meachanics has the same interpretation problem if\n> you do want to make a simple "shut up and calculate".\n\nNo.\n\n> q(t) has currently no measurable signification.\n\nSo what? Theories are obliged to make testable predictions in\ngeneral, but not for all their particular claims, objects and so on.\n\n> Why not taking (q+p)(t) as the\n> trajectory of new bohemian particle (observable Q+P)? Who is the more\n> "real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.\n\nThis is not a problem of a particular theory. This is a question about\npreference for one theory in comparison with another. BM_q is one\ntheory, BM_{p+q} another. BM_q is simpler, therefore preferred\nby Occam\'s razor.\n\n> By preferred frame, If you intend preferred observable (e.g. selection\n> of q(t) in BM), I think that bohmian mechanics cannot give you what\n> your are looking for, i.e. you can construct in principle a "A\n> bohmian" mechanics, where A is any observable of QM.\n\nI know. But that doesn\'t matter. I can prefer among this large\nclass of theories the one I prefer for reasons of simplicity. Moreover,\nI can live quite comfortable having several theories, as long as\nthey all agree with observation and are realistic.\n\nBTW, whatever observable we use here - note that time t is not an\nobservable in the classical sense in Schroedinger theory (different\nfrom x). As BM_q, as BM_p, as BM_A handle time t as a\npreferred absolute time.\n\n> Considering this aspect (existence of a preferred observable), I\n> really think that the decoherence working group and the work on the\n> preferred basis is much closer to the answer to this question (at\n> least logically coherent).\n\nI\'m not sure. Moreover, BM_q is certainly logically coherent.\nThus, there cannot be any advantage of whatever working group\nin logical coherence.\n\n> > Compare with the theory:\n> > "We hear voices every Friday the 13. at midnight in the castle."\n> > It is a falsifiable theory. But, let\'s assume it is supported by\n> > observation. Would you accept such a successful prediction\n> > as an explanation?\n\n> If a have the context of this sentence and checked the logical\n> consistency.\n\nAssume it is. We have a well-defined castle, and lot\'s of observers\nhave heard these voices.\n\n> I may say yes it is true if its interpretation in the\n> experimental results agree.\n\nOk.\n\n> After, I may consider the context and try\n> to check if this context can be explained by external theories\n> (i.e. testable).\n\nOh, here different concepts seem to be mingled. The theory\nis certainly testable. We have lot\'s of Fridays the 13 in future,\nyou can go to the castle and wait.\n\nBut how this is related to "explained by external theories"?\n\n> I think that it is exactly we do in our everyday life,\n> physics in this aspect is only an extreme example (but It is a\n> particular point of view, not really important for our subject).\n> For example, I may search for ghost that causes the voice, if I can\n> define and test what is a ghost (e.g. a human with a white cloth,\n> something else .)\n\nYep, exactly, you search for something. In everyday life as well\nas in physics.\n\nBut the analogon of the QM guy is not the guy who searches for\nsomething. No. It is the guy who thinks that the theory\n"We hear voices every Friday the 13. at midnight in the castle"\nis already sufficient. There is no need to search for ghosts or\ntapes or whatever else. We have already a theory, a testable,\npredictive theory. No need for further explanations.\n\n> > Now, if we reject the "shut up and calculate" ideology as not\n> > giving a realistic explanation, we have BM which has a preferred\n> > frame as a realistic theory.\n\n> Currently with what I have already said on BM, the preferred frame (in\n> the sense of the observable see above) of BM is the same thing as the\n> statement "the position is the preferred basis of QM". Well very\n> interesting property, but what can I do with this extra property? Why\n> do we need this extra property?\n\nIn fundamental science, don\'t ask too early for applications.\nWell, things consists of atoms, fine. An interesting property.\nWhat can we do with this property? Time will show, as it has\nshown that it is possible to build atomic weapons.\n\nNote that nobody has ever claimed that BM is the final theory\nof everything. BM is a theory which invites to think about\nmodifications. For example, one direction is Nelsonian\nstochastics. Whatever.\n\nAssume there is some more fundamental theory, with QM as\nsome limit but itself different from QM. Now, if we look more\ncareful, its quite probable that BM_q is much more close\nto this theory than BM_p. To find this theory, BM_q may be\nthe starting point, BM_p will be a dead end.\n\n> I thus prefer to look at the\n> decoherence program and the preferred basis research.\n> If you are interested on this aspect, you can look at the paper\n> http://arxiv.org/PS_cache/quant-ph/pdf/0312/0312059.pdf. It contains\n> an exhaustive status (end of 2003) on the decoherence program with\n> additional pointers to other papers.\n\nThanks.\n\n> > Now, let\'s consider the question what\n> > really happens, A->B or B->A. We cannot measure this. It is\n> > a hidden variable. Nonetheless, in a realistic causal theory we\n> > postulate that there are no closed causal loops.\n\n> Never forget that each postulate you add, has to be self-consistent\n> with the rest of your theory. The more postulates you add, the more\n> you increase the inconsistency risk. The logical demonstration of\n> consistency is far from being evident. One famous example is the\n> axiomatization of the set theory where the first "naïve" construction\n> by Cantor leads to the Russel paradox.\n\nI\'m not afraid ;-)\n\n> > Note that this discussion about EPR, Bell and preferred frames is\n> > only one part of my general argumentation in favour of a preferred\nframe.\n> > It includes many other, completely independent parts.\n\n> I really want to help you in defining a consistent logical substrate\n> of your interesting point of view I interpret as trying to build a\n> local variable extension of QM (with local interactions).\n\nNo. In the domain of QM foundations I only argue that the violation\nof BI requires the acceptance of a preferred foliation. At least if\nwe do not want to give up realism and/or causality.\n\nIn the context of this discussion I defend BM_q as an example\nof a realistic causal theory with preferred frame. It proves\nthat such theories exist, and, therefore, realism and causality\nare consistent with observation.\n\n> However, I question why you seem to consider a preferred frame as\n> fundamental and not experiment dependant (i.e. the decoherence program\n> - you have a relative frame).\n\nThere is simply too much support in favour of the preferred frame\n;-).\n\nIlja\n\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"seratend" <ser_monmail@yahoo.fr> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> > A preferred frame we need if we want a realistic interpretation of
> > QM. The simplest choice is Bohmian mechanics.

> Now, I may understand that by preferred frame, you may mean a
> preferred observable (the q position, the observable used in bohemian
> mechanics) or a given frame (x1,x2,x3,t), thus which one?

With "preferred frame" I usually mean (a little bit sloppy) a preferred
foliation of spacetime.

> Currently, I think we don't need to introduce a theory of reality as
> it can move us towards philosophy or metaphysics rather than physics.

The justification of the scientific method is, indeed, philosophy
or metaphysics. You cannot derive the scientific method from
observation or prove the scientific method by observation.

And the discussion of EPR/Bell requires the discussion of
the scientific method, because this is a case where the current
majority position does not follow the scientific
method (it rejects an essential part of the scientific method -
the search for realistic explanations of observed correlations).
At least this is my position in this question.

> Therefore, I definitively prefer to concentrate on logically
> consistent models and thereafter interpretation (connection between
> the model results and experiment results) before speaking of what
> reality is (which I think is out of scope at this step).

In this case, there is nothing to discuss. We agree that the
QM "shut up and calculate" is a logically consistent way to
obtain testable predictions which agree with observation.

> Now, to try to end with the "reality" subject, can I translate what
> you call "realistic theory" into a self-consistent mathematical model?
> If yes, I really prefer to read "self-consistent theory" rather than
> "realistic theory" even if it is longer and heavier.

Realistic theories define a subclass of the class of consistent theories.
QM "shut up and calculate" is a logically consistent but not realistic.

You can view realism as a restriction of the mathematical form of the
theory. As well as, say, relativity defines a subclass of relativistic
theories which are also logically consistent but have an additional
requirement (some symmetry group).

> The main problem with bohemian mechanics is the selection of the
> equation of motion of q(t).

No. For any given BM-like theory the equation of motion
of q(t) is defined and fixed.

> As long as Bohmian mechanics accept the
> Schrodinger time evolution of the \psi(q,t), this theory accepts that
> we can not see the difference between a QM measurement of the
> observable q and the bohemian measurement of the path q(t).

First, there is no measurement of a path in BM. Second,
we see no difference only in quantum equilibrium. Once
quantum equilibrium holds in the initial state it holds always.
But I would guess that the initial value of the universe is
not quantum equilibrium but a well-defined initial value q_0.

> In other words we cannot have in bohemian mechanics
> |\psi(q,t)|^2=\delta(q-qo(t)) (a known trajectory).

This follows from the Schroedinger equation for \psi.

> Once an infinite
> precision q measurement occurs with a result qo, the bohemian velocity
> is unknown at this time as in QM.

No. The velocity of the particle d_t q(t) is known if the position
q(t) is known and the wave function \psi(q,t) is known.

> Thus, I have a theory with an externally added equation that
> gives no more information (achievable in an experiment) than
> the orthodox QM theory.

Yep. By construction. This was the purpose of BM.

> Therefore, as long as this new equation of motion cannot provide
> additional information (existence of an experiment that shows this
> additional information), I prefer to consider Bohmian mechanics as one
> possible interpretation that is compatible with the orthodox
> interpretation of QM, with an additionnal cost q(t).

Fine. I have no problem with this point of view. Essentially, this
is also Bell's who has explicitly named it realistic interpretation
(AFAIR).

> Explanation the previous statement: U(t,to) is the time evolution of
> the initial state |\psi(to)>. We have \psi(q,to)=<q|\psi(to)> for the
> bohemian particle.

Sorry, the position of the bohmian particle is not a wave function,
not a bra vector <q| or a \delta-function \delta(q-q0). We have
\psi(q,to)=<q|\psi(to)>, but this equation has nothing to do with
the bohmian particle.

> If we define the projector P(t)=|\psi(t)><\psi(t)| where
> |\psi(t)>=U(t,to)|\psi(to)> as the measurement at any time t>to, we have
> with probability 100% the detection of the particle without the change
> of the state \psi(t). Thus P(t) defines the trajectory of the particle
> with the same accuracy as the possible q(t) motions in bohemian
> mechanics compatible with the state \psi(q,t).

I don't understand. Do you want to propose a variant of
a bohmian theory where the "preferred observable" is, instead
of q or p, P(t)?

> > The problem with relativistic QFT is that it extends the "shut up and
> > calculate" ideology. It does not care at all about an explanation
> > for the observed probabilities - being able to compute them is
> > sufficient. This is not an explanation.

> This is the measurement theory. It tries to explain in a
> self-consistent way the results we have with this mathematical model
> (interpretation consistency problems).
> I think that the "shut up and calculate ideology" is somewhat outside
> the context. Note that this is what does most of the engineers around
> the world (I think).

Ok. I don't claim "shut up and calculate" is meaningless. It gives
testable predictions, in a consistent way. It does not give explanations.

> Note that Bohmian meachanics has the same interpretation problem if
> you do want to make a simple "shut up and calculate".

No.

> q(t) has currently no measurable signification.

So what? Theories are obliged to make testable predictions in
general, but not for all their particular claims, objects and so on.

> Why not taking (q+p)(t) as the
> trajectory of new bohemian particle (observable Q+P)? Who is the more
> "real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.

This is not a problem of a particular theory. This is a question about
preference for one theory in comparison with another. BM_q is one
theory, BM_{p+q} another. BM_q is simpler, therefore preferred
by Occam's razor.

> By preferred frame, If you intend preferred observable (e.g. selection
> of q(t) in BM), I think that bohmian mechanics cannot give you what
> your are looking for, i.e. you can construct in principle a "A
> bohmian" mechanics, where A is any observable of QM.

I know. But that doesn't matter. I can prefer among this large
class of theories the one I prefer for reasons of simplicity. Moreover,
I can live quite comfortable having several theories, as long as
they all agree with observation and are realistic.

BTW, whatever observable we use here - note that time t is not an
observable in the classical sense in Schroedinger theory (different
from x). As BM_q, as BM_p, as BM_A handle time t as a
preferred absolute time.

> Considering this aspect (existence of a preferred observable), I
> really think that the decoherence working group and the work on the
> preferred basis is much closer to the answer to this question (at
> least logically coherent).

I'm not sure. Moreover, BM_q is certainly logically coherent.
Thus, there cannot be any advantage of whatever working group
in logical coherence.

> > Compare with the theory:
> > "We hear voices every Friday the 13. at midnight in the castle."
> > It is a falsifiable theory. But, let's assume it is supported by
> > observation. Would you accept such a successful prediction
> > as an explanation?

> If a have the context of this sentence and checked the logical
> consistency.

Assume it is. We have a well-defined castle, and lot's of observers
have heard these voices.

> I may say yes it is true if its interpretation in the
> experimental results agree.

Ok.

> After, I may consider the context and try
> to check if this context can be explained by external theories
> (i.e. testable).

Oh, here different concepts seem to be mingled. The theory
is certainly testable. We have lot's of Fridays the 13 in future,
you can go to the castle and wait.

But how this is related to "explained by external theories"?

> I think that it is exactly we do in our everyday life,
> physics in this aspect is only an extreme example (but It is a
> particular point of view, not really important for our subject).
> For example, I may search for ghost that causes the voice, if I can
> define and test what is a ghost (e.g. a human with a white cloth,
> something else .)

Yep, exactly, you search for something. In everyday life as well
as in physics.

But the analogon of the QM guy is not the guy who searches for
something. No. It is the guy who thinks that the theory
"We hear voices every Friday the 13. at midnight in the castle"
is already sufficient. There is no need to search for ghosts or
tapes or whatever else. We have already a theory, a testable,
predictive theory. No need for further explanations.

> > Now, if we reject the "shut up and calculate" ideology as not
> > giving a realistic explanation, we have BM which has a preferred
> > frame as a realistic theory.

> Currently with what I have already said on BM, the preferred frame (in
> the sense of the observable see above) of BM is the same thing as the
> statement "the position is the preferred basis of QM". Well very
> interesting property, but what can I do with this extra property? Why
> do we need this extra property?

In fundamental science, don't ask too early for applications.
Well, things consists of atoms, fine. An interesting property.
What can we do with this property? Time will show, as it has
shown that it is possible to build atomic weapons.

Note that nobody has ever claimed that BM is the final theory
of everything. BM is a theory which invites to think about
modifications. For example, one direction is Nelsonian
stochastics. Whatever.

Assume there is some more fundamental theory, with QM as
some limit but itself different from QM. Now, if we look more
careful, its quite probable that BM_q is much more close
to this theory than BM_p. To find this theory, BM_q may be
the starting point, BM_p will be a dead end.

> I thus prefer to look at the
> decoherence program and the preferred basis research.
> If you are interested on this aspect, you can look at the paper
> http://arxiv.org/PS_cache/quant-ph/pdf/0312/0312059.pdf. It contains
> an exhaustive status (end of 2003) on the decoherence program with
> additional pointers to other papers.

Thanks.

> > Now, let's consider the question what
> > really happens, A->B or B->A. We cannot measure this. It is
> > a hidden variable. Nonetheless, in a realistic causal theory we
> > postulate that there are no closed causal loops.

> Never forget that each postulate you add, has to be self-consistent
> with the rest of your theory. The more postulates you add, the more
> you increase the inconsistency risk. The logical demonstration of
> consistency is far from being evident. One famous example is the
> axiomatization of the set theory where the first "naïve" construction
> by Cantor leads to the Russel paradox.

I'm not afraid ;-)

> > Note that this discussion about EPR, Bell and preferred frames is
> > only one part of my general argumentation in favour of a preferred
frame.
> > It includes many other, completely independent parts.

> I really want to help you in defining a consistent logical substrate
> of your interesting point of view I interpret as trying to build a
> local variable extension of QM (with local interactions).

No. In the domain of QM foundations I only argue that the violation
of BI requires the acceptance of a preferred foliation. At least if
we do not want to give up realism and/or causality.

In the context of this discussion I defend BM_q as an example
of a realistic causal theory with preferred frame. It proves
that such theories exist, and, therefore, realism and causality
are consistent with observation.

> However, I question why you seem to consider a preferred frame as
> fundamental and not experiment dependant (i.e. the decoherence program
> - you have a relative frame).

There is simply too much support in favour of the preferred frame
;-).

Ilja

Ilja Schmelzer

Nov19-04, 01:29 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"seratend" <ser_monmail@yahoo.fr> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> > Bohmian mechanics has, indeed, hidden causal influences in a hidden\n> > preferred frame.\n>\n> Bohmian Mechanics is mainly QM with the observable position and an\n> extra motion equation (q(t)). The observable Q may be expressed in any\n> relativistic frame and thus BM q(t). So I really do not see the\n> preferred frame.\n\nThe preferred frame is defined by the wavefunction. In QM, the\nwavefunction is only a mathematical tool, therefore the preferred frame\nof the wavefunction is not considered to be important. (But it is the\nIMHO main reason why people prefer the operator (Heisenberg)\npicture in the relativistic domain.)\n\nBut if the wavefunction is considered to be part of reality, and,\nespecially,\nin combination with the guiding equation, the preferred frame of the\nwavefunction becomes obvious.\n\nIt starts with two-particle theory, and the two-particle wave function\nPsi(x1,x2,t). The guiding equation is\nd_t x1 = something(Psi(x1,x2,t)), where x2 is the position of the\nsecond particle at the same moment of time t.\n\n> All the problems of causality with BM comes from the interpretation of\n> q(t). As long as we cannot get information about q(t) from an\n> experiment different from what a QM Position measurement gives, we can\n> question the interpretation of q(t) and the deductions based on this\n> interpretation (requirement of a preferred frame?).\n\nNo, the problem is certainly not the interpretation of BM, but BM itself.\nThe violation of Einstein causality is a consequence of the BM equations.\n\n> > Bohmian mechanics is rejected because it has a preferred frame, which\n> > does not agree with relativistic ideology. And hidden variable theories\n> > in general are rejected for the same reason, because the violation of\nBell\'s\n> > inequality proves that they need a preferred frame.\n\n> It is one interpretation of q(t) that may require a preferred frame. I\n> a not sure this is the topic that BM is not widely used (even if I\n> don\'t know what is "widely").\n> But I can say why I am not using BM: BM adds to QM the motion of\n> q(t). As long as q(t) cannot add new information that can be tested\n> (relatively to QM), I don\'t want to spend time in computing q(t). It is\n> very difficult and I still have to compute psi(q,t) to get q(t). In\n> QM, I just need psi(q,t). In both cases, I have the same results (i.e.\n> testable and usable information). This is very pragmatic.\n\nSorry, but you are not obliged to compute q(t). If you are not interested\nto know particular particle trajectories you can ignore this equation.\n(I can tell you that I have never really computed a q(t), except in the\nsingle case of the eigenstates of Hamiltonian, where q(t)=const.)\n\nYou can also use the same QM math, once you know that it can be\n(and has been) derived from BM.\n\nThe difference between QM and BM is metaphysical. QM means\nphilosophical confusion, combined with "shut up and calculate".\nInstead, BM is a clear, well-defined theory without any confusion\nin the foundations.\n\n> > But if we combine these arguments, it appears that the argument against\n> > hidden variables is that it uses hidden variables. Because there is no\n> > other argument against the preferred frame.\n>\n> Well, I can say because they are untestable variables, I surely can\n> select the interpretation and the hidden variable that is compatible\n> or not with a hidden preferred frame concept.\n\nThe question is if there is a reasonable version of BM without a\npreferred frame. Bell\'s theorem suggests that there is none.\n\n> > Thus, accepting this argument means preserving Einstein causality\n> > even in a situation where it can be logically proven to be false.\n>\n> Logically is based on one interpretation context. So first, we should\n> question the interpretation consistency.\n\nBM is a consistent interpretation. I see no reason to doubt this.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"seratend" <ser_monmail@yahoo.fr> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> > Bohmian mechanics has, indeed, hidden causal influences in a hidden
> > preferred frame.
>
> Bohmian Mechanics is mainly QM with the observable position and an
> extra motion equation (q(t)). The observable Q may be expressed in any
> relativistic frame and thus BM q(t). So I really do not see the
> preferred frame.

The preferred frame is defined by the wavefunction. In QM, the
wavefunction is only a mathematical tool, therefore the preferred frame
of the wavefunction is not considered to be important. (But it is the
IMHO main reason why people prefer the operator (Heisenberg)
picture in the relativistic domain.)

But if the wavefunction is considered to be part of reality, and,
especially,
in combination with the guiding equation, the preferred frame of the
wavefunction becomes obvious.

It starts with two-particle theory, and the two-particle wave function
\Psi(x1,x2,t). The guiding equation is
d_t x1 = something(\Psi(x1,x2,t)), where x2 is the position of the
second particle at the same moment of time t.

> All the problems of causality with BM comes from the interpretation of
> q(t). As long as we cannot get information about q(t) from an
> experiment different from what a QM Position measurement gives, we can
> question the interpretation of q(t) and the deductions based on this
> interpretation (requirement of a preferred frame?).

No, the problem is certainly not the interpretation of BM, but BM itself.
The violation of Einstein causality is a consequence of the BM equations.

> > Bohmian mechanics is rejected because it has a preferred frame, which
> > does not agree with relativistic ideology. And hidden variable theories
> > in general are rejected for the same reason, because the violation of
Bell's
> > inequality proves that they need a preferred frame.

> It is one interpretation of q(t) that may require a preferred frame. I
> a not sure this is the topic that BM is not widely used (even if I
> don't know what is "widely").
> But I can say why I am not using BM: BM adds to QM the motion of
> q(t). As long as q(t) cannot add new information that can be tested
> (relatively to QM), I don't want to spend time in computing q(t). It is
> very difficult and I still have to compute \psi(q,t) to get q(t). In
> QM, I just need \psi(q,t). In both cases, I have the same results (i.e.
> testable and usable information). This is very pragmatic.

Sorry, but you are not obliged to compute q(t). If you are not interested
to know particular particle trajectories you can ignore this equation.
(I can tell you that I have never really computed a q(t), except in the
single case of the eigenstates of Hamiltonian, where q(t)=const.)

You can also use the same QM math, once you know that it can be
(and has been) derived from BM.

The difference between QM and BM is metaphysical. QM means
philosophical confusion, combined with "shut up and calculate".
Instead, BM is a clear, well-defined theory without any confusion
in the foundations.

> > But if we combine these arguments, it appears that the argument against
> > hidden variables is that it uses hidden variables. Because there is no
> > other argument against the preferred frame.
>
> Well, I can say because they are untestable variables, I surely can
> select the interpretation and the hidden variable that is compatible
> or not with a hidden preferred frame concept.

The question is if there is a reasonable version of BM without a
preferred frame. Bell's theorem suggests that there is none.

> > Thus, accepting this argument means preserving Einstein causality
> > even in a situation where it can be logically proven to be false.
>
> Logically is based on one interpretation context. So first, we should
> question the interpretation consistency.

BM is a consistent interpretation. I see no reason to doubt this.

Ilja

Ilja Schmelzer

Nov19-04, 01:29 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n<rof@maths.tcd.ie> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n> ><rof@maths.tcd.ie> schrieb\n> >> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n> >Moreover, it is not given that the construction M_n is spatial.\n> >If we construct following Bohmian mechanics we construct\n> >some Psi(q) which is not spatial but defined on the configuration\n> >space.\n>\n> Sure, but the reason that it\'s an explanation that one feels\n> comfortable with is because it\'s easy to visualise. One imagines\n> a wave moving in some space (here, it\'s configuration space),\n> and that is something that we\'re used to. The state of affairs (the\n> wavefunction, Psi, ignoring for a moment the particle positions)\n> is represented as being composed of several mini-states of affairs,\n> Psi(x), each at a different position.\n> One becomes a lot less comfortable\n> with it once one reflects on the fact that the place where the\n> wavefunction lives is not real space at all, but an abstract\n> construction called configuration space whose dimension is 3N.\n\nI know. That\'s my point - the Psi(q) isn\'t spatial.\n\n> We can make it make a little more sense by noticing that the\n> space over which the wavefunction is defined is the same as\n> the space over which probability distributions (of generic\n> states of affairs of the system, or configurations) are\n> naturally defined. Information that we have might have about\n> the configuration of the system at a given time would naturally\n> be expressed as a real function on this space, and a description\n> which included the instantaneous rate of change of this\n> probability distribution would naturally be expressed as\n> two real functions (or one complex function) on this space.\n\nThis is the base of the QM interpretation that the wave function\ndoes not describe some state of reality but our knowledge about\nreality. But I think there is no necessity for such a non-realistic\ninterpretation.\n\nI feel uncomfortable with this space too. But I have another\nidea: There is simply our universe, q in Q, and something else,\nunknown. But whatever this unknown thing is, once it interacts\nwith our universe, its action somehow depends on the state\nof our universe and, therefore, it may be described by some\nfunction on all possible states of the universe too.\n\n> >I would not argue that there is some strong urge to develop ontological\n> >theories. Ontological theories have been successful, there is no reason\n> >to give them up.\n\n> There is no success of any theory which can be ascribed to its\n> ontological character.\n\nI disagree, but acknowledge that it\'s hard to prove that\nthe ontological character of certain theories has been\nimportant for their success.\n\n> > With Bohmian mechanics they are successful today\n> > too.\n\n> Only insofar as Bohmian mechanics has been able to reproduce\n> the predictions of a theory (Copenhagen-style quantum mechanics) which\n> wasn\'t ontological. In so far as Bohmian mechanics has not been\n> able to follow quantum mechanics (pair creation, for example),\n> it has not been successful, although I recognise that there has\n> been progress in this direction.\n\nSorry, I don\'t see why you think pair creation is problematic.\nOf course, in field theory you have to use an appropriate\nconfiguration space - a space of functions - and a formulation\nof QFT with an appropriate wave function - which is a wave\nfunctional on this function space.\n\nTo define everything accurately, a discretization or other\ntype of regularization is necessary in QFT as well.\n\nBut, in general, once you have managed to make the\nquantum part mathematically consistent, to add the\nguiding equation is not really problematic.\n\nAll you need is a probability current operator J.\nThen you have\n\nd/dt Q = <Psi J Psi>/<Psi Psi>\n\n\n> >> A may think it\'s on ontological theory, but A is confusing himself.\n> >> Imagine there is a digital machine, which takes 1\'s and 0\'s as\n> >> input. You feed in some sequence of 1\'s and 0\'s, generated by\n> >> some algorithm. The machine proceeds to construct a Markov model\n> >> to predict the sequence. It is successful to some degree - most\n> >> of the time it correctly predicts the next input. Has it discovered\n> >> reality? Is its Markov model an ontological theory?\n>\n> >Once it predicts successfully, there is some probability that the\n> >algorithm has been guessed correctly. In other words, that\n> >the algorithm used to create the input sequence and the\n> >algorithm used by the machine to predict the input sequence are\n> >in some sense equivalent.\n>\n> But this doesn\'t answer either of the above questions.\n\nAnd there is no necessity to answer the first. Of course,\nif we have a realistic theory of everything, we also cannot\nbe sure that the theory is true. And we never will be sure.\nYou want too much certainty.\n\nIf I was not clear enough about the second: A Marcov model is\nan ontological theory (or can be formulated in such a way).\n\n> There\'s also the problem of the vagueness of the "in some\n> sense" which is inevitably reached by proponents of realism\n> in this situation.\n\nSuch is life.\n\n> The grand project of ontology is not to guess an algorithm which\n> predicts future observations (which is, in fact, the only thing\n> we can do), but to determine what it is that really exists.\n\nThis project has been given up long time ago. At least if you\nmean "determine" as giving certainty. The current (not that grand)\nproject of ontology is to guess what it is that really is, and to\ntest this guess using its testable consequences.\n\n> If a machine develops an algorithm to correctly predict whether\n> the next input will be a 1 or a 0, the most one can say about\n> it is that that is what it has achieved. To say that it has\n> discovered the ontological basis of reality, one has to radically\n> change what one means by "discovered the ontological basis of\n> reality."\n\nIts you who wants to say this. I\'m quite satisfied by giving\na reasonable guess which allows to derive predictions\nwhich survive tests.\n\n> One who, observing traffic lights, eventually says "A repeating\n> pattern of green, then yellow and then red," has hardly reached the\n> ultimate goal that realists aspire to.\n\nYou postulate some utopic ultimate goals for realists and\nthen argue that these goals are utopic. So what? These are\nnot my goals.\n\n> Alan Sokal, in his "Defense of a Modest Scientific Realism",\n> examines these issues a little bit, but, since he too is\n> wedded to realism, doesn\'t quite reach the correct answer,\n> but instead ends with:\n> "Since no existing theory purports to be a final theory, there\n> is no reason to consider it as literally true or to worry too\n> much about whether the entities it postulates `really exist\'.\n> Or rather, when worrying about whether the unobservable entities\n> of a given theory `really exist\', it is important to distinguish\n> existence _as a fundamental constituent of the universe_ from\n> existence _in some coarse-grained sense_. It is a reasonable guess\n> that _none_ of the theoretical entities in our present-day\n> theories are truly fundamental, and that _all_ of the theoretical\n> entities in our present-day well-confirmed theories will maintain\n> some status as derived entities in future theories." (His italics)\n\nQuite reasonable.\n\n> In the end, he opts for saying that things exist "in some coarse-grained\n> sense", which means that he wants to ascribe to them some property\n> which he feels is related to existence or reality, but which is so\n> vague that he cannot tell us what it is, and, lacking any concrete\n> or even specific thing to say, must instead appeal to our sympathy\n> to his realist position in order to convince us to accept what he\n> is saying as a valid defense of realism.\n\nWhy do you think this is too vague? We have lot\'s of examples\nwhere one theory appears as a limit of some more fundamental\ntheory. And in all these cases the objects which "exist" in the\napproximation are somehow constructed from the objects\nwhich "exist" in the more fundamental theory. In every such case\nwe can consider this connection in detail, as detailed as we like.\nThus, the "vague" notion is not an appeal to sympathy, but\nan appeal to the large body of experience with such derived,\nnon-fundamental objects.\n\nAnd, of course, a consequence of the fact that the most\nfundamental theory is unknown.\n\n> Notice that he also explicitly subscribes to the view that,\n> if an entity from one theory appears as a derived or emergent\n> thing in a more mature theory, then that entity cannot be a\n> "fundamental constituent of the universe." What he leaves open,\n> but implicitly suggests, is that, in a final theory, which\n> makes perfect predictions and has no further refinement, we\n> might actually consider the entities with which it deals to\n> be fundamental constituents of the universe which really exist.\n> This is, of course, what the whole exercise was intended to\n> deduce, and which it has failed to deduce, so he ends by\n> surreptitiously implying it.\n\nThe idea to _deduce_ the most fundamental theory has been\ngiven up long ago. But there is no reason to give up the\nmodern realism which does not propose such utopic\nnonsense.\n\n> It is very important in these matters not to decide that we like\n> realism and then accept as valid an otherwise insufficient\n> argument in favour of it.\n\nOk. So what?\n\n> Thus, if we find that the arguments with which\n> we convince ourselves that realism is necessary amount to\n> "In my opinion, realism should be considered a part of logic,"\n> then we must go in search of the reason why that opinion\n> was adopted, and reject it unless there is a rigorous basis\n> for it.\n\nI have adopted it because I have found the arguments\nin favour of this position convincing.\n\nI do not believe in such things as "rigorous basis" in\ngeneral and in the most fundamental philosophy especially.\nLogic may not be justified on a rigorous basis. Because\nany possible "rigorous basis" is, itself, based on logic.\nThus, the "rigorous basis" is only circular reasoning.\n\n> As it stands, it looks quite like a demand that\n> realism be adopted as an axiom; that is, it looks like\n> a dogmatic assertion of a statement which cannot be deduced\n> by reason.\n\nIt cannot be deduced, indeed. As well as logic cannot\nbe deduced. It does not mean that there is no justification.\n\n> >That\'s not really a problem. Once we don\'t know the complete\n> >input data used by the algorithm we cannot predict everything.\n> >Nonetheless it is possible that the algorithms we guess and the\n> >algorithms used in the "fabric of reality" are equivalent.\n>\n> The implicit assertion here is that the fabric of reality\n> is implementing some kind of algorithm (recall that an algorithm\n> is a list of instructions for computing one thing from another).\n\nYep. As a realist, I use implicit assertions about reality.\n\n> The equivalence here is again, "in some sense", and has the particular\n> difficulty that one algorithm (the one used by a human) has, as its\n> output, patterns of sensations (as a prediction of future input).\n> I doubt that an algorithm which has sensations as output is one\n> that most realists would accept as an example of something which\n> could qualify as a fundamental constituent of reality, since the\n> things with which it deals (sensations) are purely mental.\n\nHere several things are mingled. The algorithm which is assumed\nto be used by the fabric of reality is one which computes\nX_n+1 = f(X_n).\n\nTo derive testable predictions from this we have to solve some\nother problems. Namely to connect the X_n with the E_n.\n\n> >> Whether or not a particle even exists can depend on the motion of\n> >> the observer (cf. the Unruh effect), so it seems strange to put that\n> >> sticky reality label onto the particles and their positions in space.\n>\n> >I don\'t.\n>\n> But Bohmian mechanics does, and that is the context in which I made\n> the remark above.\n\nBohmian theories supposed to handle the Unruh effect don\'t.\n\n> In this case,\n> are you asserting that "reality" is something which can be\n> defined, or that its definition can change from theory to theory?\n\nThe definition of reality is part of the realistic theory. It\nchanges from theory to theory.\n\nIn NT reality consists of point particles attracted by forces,\nin GR it consists of a manifold with a metric and some\nmatter fields on it, in field theory of fields\n(psi(x),A_i(x),g_ij(x)), in Bohmian theories of\nsome q in Q and some Psi: Q-->C, where Q is\nthe configuration space which is very different in\ndifferent Bohmian theories.\n\nEspecially, a BT which handles the Unruh effect\nmay have, as the configuration space,\nsome fields q=(psi(x),A_i(x),g_mn(x)), or\nsome regularization, as, for example, some\nlattice theory q=(psi(n),A_i(n),g_mn(n)), n in Z^3.\n\nRealism is the preference for theories which\nmake clear definitions about what is real. And fulfill\nsome consistency conditions about these real objects.\n(for example, we can apply probability theory and\nlogic if we work with real objects, "counterfactual\nexistence").\n\n> I am always completely in favour of using classical logic and\n> probability theory.\n\nFine.\n\n> >Sorry, but we use physical theories to predict things which do not\n> >happen in our mind, but in reality. (Except you are a pure solipsist.)\n> >Of course, I feel free to say that there is a law of gravity out there.\n>\n> >I may be wrong about the details of the law of gravity. But it is\n> >a hypothesis (in my mind) about what is out there.\n>\n> So you begin to understand; the law of gravity, as you\n> know it, is not something which exists independently of you, but a\n> hypothesis which relies on you for its existence.\n\nNo. Apples fall down even if I don\'t have any theory about this.\nThe particular law of gravity is a model of this in my mind. But\nits purpose is to describe something which exists really, outside\nof my mind, independent of my existence.\n\n> The real difficulty here is (as I mentioned before) the manifold\n> use of the word "real".\n\nNot that difficult IMHO.\n\nOf course, I\'m sometimes sloppy too, especially in usenet\npostings. But usually I try to label different concepts with\ndifferent labels.\n\nFor example, "realistic theory" is always a theory. Realism\nI use not for a particular theory about reality but a definition\nof a class of theories, namely "realistic theories". Reality\nI use for all this stuff outside, which is not in our minds.\n\nSometimes we have equivalent things denoted by the same\nlabel. For example, the definition of a class of theories,\nnamely the class of realistic theories, is in some sense\nequivalent to a (meta)theory (about reality) that the correct\ntheory is an element of this class. Or to a methodological\ndecision to restrict yourself (or the community which follows\nyour argumentation) to this subclass of theories, rejecting\nother theories from the start. All these conceptually different\nbut equivalent things I name "realism".\n\n> To add further to this confusion is the additional confusion caused\n> by the inappropriate use of the notion of "outside." The relations\n> of outside and inside are properly ascribed only to objects which\n> are in space.\n\nI hope my use of "outside" does not cause confusion.\nIt means "not in my mind or my soul". My mind,\n(at least the material part of it) is sufficiently located in\nspace. If there is something non-material connected with\nconsciousness and so on (soul), it is not part of "outside".\n\n> Now, taken together, these two confusions lead to the realist\n> position, where the physicist believes that he is examining\n> that which "really" exists, "outside" of his mind. In fact,\n> what he is doing is examining things which are actual (in\n> my sense of the term) and outside of his head. Perhaps I\n> should emphasise that this is worth thinking about; many\n> lives have been wasted chasing ghosts because of this\n> misunderstanding.\n\nI see not much advantage in your specification. I propose\ntheories about actual things outside of my head? Fine.\nNo problem. But also no advantage.\n\n> >> ... but nobody tries to understand starting from\n> >> the facts which are right in front of them. That\'s not what they\n> >> were trained to do.\n>\n> >The point is that this is impossible. It is the old ideal of positivism\n> >to derive theories from observable facts. Science works differently.\n> >There is no derivation E_n ->X_n, there is guesswork. We guess\n> >theories, derive the consequences, and reject the falsified guesses.\n> >This is what we are trained to, and this is how science works.\n>\n> This is what we are trained to do, but it is only a small part of\n> how science works, namely the part that Popper, with what talent\n> he had, was able to identify. Since you are here invoking an argument\n> from authority (Science),\n\nMy argument is not from authority (Science or Popper) but\nI have described my theory of science, which agrees in its\nmost essential parts with Popper\'s theory about science.\n\n> The problem that modern theoretical physics finds itself in is\n> that people are only trained in what Newton calls the synthetic\n> method, principally because it\'s the easiest thing to test\n> in examinations. It would be rather an interesting experiment\n> to take a young physicist-to-be and show him all of the phenomena\n> of electromagnetism, and assign to him the project of finding\n> equations which describe all the observed results.\n> A physicist with this training would be better equipped for real\n> investigation than his colleagues, who have merely been trained to\n> do homework assignments in which the axioms are given and the\n> theorems are to be deduced.\n\nMay be, but I don\'t argue about pedagogics.\n\nThere are more serious faults in modern science than this.\nScientists should be independent. At least as independent\nas judges. Even more. Because a judge which makes obvious\nnonsense harms other people, a scientist which makes\nobvious nonsense doesn\'t.\n\nBut current scientific organization is the reverse. If you\nhave a grant for a few years, you have to care about your\nfuture, the next grant. The decisions are made by\nbeaurocrats and other scientists (these two often unified\nin a single person). The predictable result is conformity.\n\nAnother pressure into conformity is "publish or perish".\nPublishing and being cited is easier in a large community\nwith lots of journals to publish and lots of readers.\n\nThe prediction agrees with observation. In the highly\nspeculative domain of modern fundamental physics,\nwhere free science would lead to lot\'s of very different\ndirections (because they are not guided by experiments,\npurely speculative, therefore its extremely hard to reject\na research proposal as false) we observe de facto only\ntwo research directions - strings and LQG.\n\n> I should add that it is not the case that the analytic method has\n> been exhausted due to a lack of performable experiments whose\n> results we do not know in advance. The confusion about realism\n> and the many disagreements about quantum mechanics (among those\n> who actually do think about it) are evidence enough that there\n> is an opportunity, for one who is prepared to suspend judgment\n> and avoid opinion, to discern something previously overlooked.\n\nThe second sentence does not prove the first. The last two\npoints are discussions not about the results of possible\nexperiments.\n\n> >> In order to formulate the questions which relativity\n> >> addresses in my language, which is to some extent solipsistic,\n> >> one first has to address the question of what relationship\n> >> exists between one observer\'s X_n and another\'s X\'_n, which\n> >> involves how one observer manifests himself within the X_n\n> >> of another, and that is a difficult question.\n>\n> >In a realistic theory, the answer is simple: The X_n should be\n> >equivalent, or at least one of the observers follows a wrong theory.\n>\n> There are several problem with that. One is, as you said yourself,\n> that the X_n are not complete.\n\nThe M_n (perceptions) are not complete. The X_n are complete.\n\n> That is, I perceive a room, with walls\n> here and there, and you perceive a different room (presumably),\n> and there is no simple matching of what I perceive to what you\n> perceive.\n\nThis is about the M_n. My X_n contains also complete\ninformation about the inside of your room. I don\'t perceive\nthe X_n. But the laws of physics, the rule which allows\nto compute X_n+1=f(X_n), need the whole information,\nmy room as well as your room.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><rof@maths.tcd.ie> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
> ><rof@maths.tcd.ie> schrieb
> >> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

> >Moreover, it is not given that the construction M_n is spatial.
> >If we construct following Bohmian mechanics we construct
> >some \Psi(q) which is not spatial but defined on the configuration
> >space.
>
> Sure, but the reason that it's an explanation that one feels
> comfortable with is because it's easy to visualise. One imagines
> a wave moving in some space (here, it's configuration space),
> and that is something that we're used to. The state of affairs (the
> wavefunction, \Psi, ignoring for a moment the particle positions)
> is represented as being composed of several mini-states of affairs,
> \Psi(x), each at a different position.
> One becomes a lot less comfortable
> with it once one reflects on the fact that the place where the
> wavefunction lives is not real space at all, but an abstract
> construction called configuration space whose dimension is 3N.

I know. That's my point - the \Psi(q) isn't spatial.

> We can make it make a little more sense by noticing that the
> space over which the wavefunction is defined is the same as
> the space over which probability distributions (of generic
> states of affairs of the system, or configurations) are
> naturally defined. Information that we have might have about
> the configuration of the system at a given time would naturally
> be expressed as a real function on this space, and a description
> which included the instantaneous rate of change of this
> probability distribution would naturally be expressed as
> two real functions (or one complex function) on this space.

This is the base of the QM interpretation that the wave function
does not describe some state of reality but our knowledge about
reality. But I think there is no necessity for such a non-realistic
interpretation.

I feel uncomfortable with this space too. But I have another
idea: There is simply our universe, q in Q, and something else,
unknown. But whatever this unknown thing is, once it interacts
with our universe, its action somehow depends on the state
of our universe and, therefore, it may be described by some
function on all possible states of the universe too.

> >I would not argue that there is some strong urge to develop ontological
> >theories. Ontological theories have been successful, there is no reason
> >to give them up.

> There is no success of any theory which can be ascribed to its
> ontological character.

I disagree, but acknowledge that it's hard to prove that
the ontological character of certain theories has been
important for their success.

> > With Bohmian mechanics they are successful today
> > too.

> Only insofar as Bohmian mechanics has been able to reproduce
> the predictions of a theory (Copenhagen-style quantum mechanics) which
> wasn't ontological. In so far as Bohmian mechanics has not been
> able to follow quantum mechanics (pair creation, for example),
> it has not been successful, although I recognise that there has
> been progress in this direction.

Sorry, I don't see why you think pair creation is problematic.
Of course, in field theory you have to use an appropriate
configuration space - a space of functions - and a formulation
of QFT with an appropriate wave function - which is a wave
functional on this function space.

To define everything accurately, a discretization or other
type of regularization is necessary in QFT as well.

But, in general, once you have managed to make the
quantum part mathematically consistent, to add the
guiding equation is not really problematic.

All you need is a probability current operator J.
Then you have

d/dt Q = <\Psi J \Psi>/<\Psi \Psi>

> >> A may think it's on ontological theory, but A is confusing himself.
> >> Imagine there is a digital machine, which takes 1's and 0's as
> >> input. You feed in some sequence of 1's and 0's, generated by
> >> some algorithm. The machine proceeds to construct a Markov model
> >> to predict the sequence. It is successful to some degree - most
> >> of the time it correctly predicts the next input. Has it discovered
> >> reality? Is its Markov model an ontological theory?
>
> >Once it predicts successfully, there is some probability that the
> >algorithm has been guessed correctly. In other words, that
> >the algorithm used to create the input sequence and the
> >algorithm used by the machine to predict the input sequence are
> >in some sense equivalent.
>
> But this doesn't answer either of the above questions.

And there is no necessity to answer the first. Of course,
if we have a realistic theory of everything, we also cannot
be sure that the theory is true. And we never will be sure.
You want too much certainty.

If I was not clear enough about the second: A Marcov model is
an ontological theory (or can be formulated in such a way).

> There's also the problem of the vagueness of the "in some
> sense" which is inevitably reached by proponents of realism
> in this situation.

Such is life.

> The grand project of ontology is not to guess an algorithm which
> predicts future observations (which is, in fact, the only thing
> we can do), but to determine what it is that really exists.

This project has been given up long time ago. At least if you
mean "determine" as giving certainty. The current (not that grand)
project of ontology is to guess what it is that really is, and to
test this guess using its testable consequences.

> If a machine develops an algorithm to correctly predict whether
> the next input will be a 1 or a 0, the most one can say about
> it is that that is what it has achieved. To say that it has
> discovered the ontological basis of reality, one has to radically
> change what one means by "discovered the ontological basis of
> reality."

Its you who wants to say this. I'm quite satisfied by giving
a reasonable guess which allows to derive predictions
which survive tests.

> One who, observing traffic lights, eventually says "A repeating
> pattern of green, then yellow and then red," has hardly reached the
> ultimate goal that realists aspire to.

You postulate some utopic ultimate goals for realists and
then argue that these goals are utopic. So what? These are
not my goals.

> Alan Sokal, in his "Defense of a Modest Scientific Realism",
> examines these issues a little bit, but, since he too is
> wedded to realism, doesn't quite reach the correct answer,
> but instead ends with:
> "Since no existing theory purports to be a final theory, there
> is no reason to consider it as literally true or to worry too
> much about whether the entities it postulates `really exist'.
> Or rather, when worrying about whether the unobservable entities
> of a given theory `really exist', it is important to distinguish
> existence _as a fundamental constituent of the universe_ from
> existence _in some coarse-grained sense_. It is a reasonable guess
> that _none_ of the theoretical entities in our present-day
> theories are truly fundamental, and that _all_ of the theoretical
> entities in our present-day well-confirmed theories will maintain
> some status as derived entities in future theories." (His italics)

Quite reasonable.

> In the end, he opts for saying that things exist "in some coarse-grained
> sense", which means that he wants to ascribe to them some property
> which he feels is related to existence or reality, but which is so
> vague that he cannot tell us what it is, and, lacking any concrete
> or even specific thing to say, must instead appeal to our sympathy
> to his realist position in order to convince us to accept what he
> is saying as a valid defense of realism.

Why do you think this is too vague? We have lot's of examples
where one theory appears as a limit of some more fundamental
theory. And in all these cases the objects which "exist" in the
approximation are somehow constructed from the objects
which "exist" in the more fundamental theory. In every such case
we can consider this connection in detail, as detailed as we like.
Thus, the "vague" notion is not an appeal to sympathy, but
an appeal to the large body of experience with such derived,
non-fundamental objects.

And, of course, a consequence of the fact that the most
fundamental theory is unknown.

> Notice that he also explicitly subscribes to the view that,
> if an entity from one theory appears as a derived or emergent
> thing in a more mature theory, then that entity cannot be a
> "fundamental constituent of the universe." What he leaves open,
> but implicitly suggests, is that, in a final theory, which
> makes perfect predictions and has no further refinement, we
> might actually consider the entities with which it deals to
> be fundamental constituents of the universe which really exist.
> This is, of course, what the whole exercise was intended to
> deduce, and which it has failed to deduce, so he ends by
> surreptitiously implying it.

The idea to _deduce_ the most fundamental theory has been
given up long ago. But there is no reason to give up the
modern realism which does not propose such utopic
nonsense.

> It is very important in these matters not to decide that we like
> realism and then accept as valid an otherwise insufficient
> argument in favour of it.

Ok. So what?

> Thus, if we find that the arguments with which
> we convince ourselves that realism is necessary amount to
> "In my opinion, realism should be considered a part of logic,"
> then we must go in search of the reason why that opinion
> was adopted, and reject it unless there is a rigorous basis
> for it.

I have adopted it because I have found the arguments
in favour of this position convincing.

I do not believe in such things as "rigorous basis" in
general and in the most fundamental philosophy especially.
Logic may not be justified on a rigorous basis. Because
any possible "rigorous basis" is, itself, based on logic.
Thus, the "rigorous basis" is only circular reasoning.

> As it stands, it looks quite like a demand that
> realism be adopted as an axiom; that is, it looks like
> a dogmatic assertion of a statement which cannot be deduced
> by reason.

It cannot be deduced, indeed. As well as logic cannot
be deduced. It does not mean that there is no justification.

> >That's not really a problem. Once we don't know the complete
> >input data used by the algorithm we cannot predict everything.
> >Nonetheless it is possible that the algorithms we guess and the
> >algorithms used in the "fabric of reality" are equivalent.
>
> The implicit assertion here is that the fabric of reality
> is implementing some kind of algorithm (recall that an algorithm
> is a list of instructions for computing one thing from another).

Yep. As a realist, I use implicit assertions about reality.

> The equivalence here is again, "in some sense", and has the particular
> difficulty that one algorithm (the one used by a human) has, as its
> output, patterns of sensations (as a prediction of future input).
> I doubt that an algorithm which has sensations as output is one
> that most realists would accept as an example of something which
> could qualify as a fundamental constituent of reality, since the
> things with which it deals (sensations) are purely mental.

Here several things are mingled. The algorithm which is assumed
to be used by the fabric of reality is one which computes
X_n+1 = f(X_n).

To derive testable predictions from this we have to solve some
other problems. Namely to connect the X_n with the E_n.

> >> Whether or not a particle even exists can depend on the motion of
> >> the observer (cf. the Unruh effect), so it seems strange to put that
> >> sticky reality label onto the particles and their positions in space.
>
> >I don't.
>
> But Bohmian mechanics does, and that is the context in which I made
> the remark above.

Bohmian theories supposed to handle the Unruh effect don't.

> In this case,
> are you asserting that "reality" is something which can be
> defined, or that its definition can change from theory to theory?

The definition of reality is part of the realistic theory. It
changes from theory to theory.

In NT reality consists of point particles attracted by forces,
in GR it consists of a manifold with a metric and some
matter fields on it, in field theory of fields
(\psi(x),A_i(x),g_{ij}(x)), in Bohmian theories of
some q in Q and some \Psi: Q-->C, where Q is
the configuration space which is very different in
different Bohmian theories.

Especially, a BT which handles the Unruh effect
may have, as the configuration space,
some fields q=(\psi(x),A_i(x),g_{mn}(x)), or
some regularization, as, for example, some
lattice theory q=(\psi(n),A_i(n),g_{mn}(n)), n in Z^3.

Realism is the preference for theories which
make clear definitions about what is real. And fulfill
some consistency conditions about these real objects.
(for example, we can apply probability theory and
logic if we work with real objects, "counterfactual
existence").

> I am always completely in favour of using classical logic and
> probability theory.

Fine.

> >Sorry, but we use physical theories to predict things which do not
> >happen in our mind, but in reality. (Except you are a pure solipsist.)
> >Of course, I feel free to say that there is a law of gravity out there.
>
> >I may be wrong about the details of the law of gravity. But it is
> >a hypothesis (in my mind) about what is out there.
>
> So you begin to understand; the law of gravity, as you
> know it, is not something which exists independently of you, but a
> hypothesis which relies on you for its existence.

No. Apples fall down even if I don't have any theory about this.
The particular law of gravity is a model of this in my mind. But
its purpose is to describe something which exists really, outside
of my mind, independent of my existence.

> The real difficulty here is (as I mentioned before) the manifold
> use of the word "real".

Not that difficult IMHO.

Of course, I'm sometimes sloppy too, especially in usenet
postings. But usually I try to label different concepts with
different labels.

For example, "realistic theory" is always a theory. Realism
I use not for a particular theory about reality but a definition
of a class of theories, namely "realistic theories". Reality
I use for all this stuff outside, which is not in our minds.

Sometimes we have equivalent things denoted by the same
label. For example, the definition of a class of theories,
namely the class of realistic theories, is in some sense
equivalent to a (meta)theory (about reality) that the correct
theory is an element of this class. Or to a methodological
decision to restrict yourself (or the community which follows
your argumentation) to this subclass of theories, rejecting
other theories from the start. All these conceptually different
but equivalent things I name "realism".

> To add further to this confusion is the additional confusion caused
> by the inappropriate use of the notion of "outside." The relations
> of outside and inside are properly ascribed only to objects which
> are in space.

I hope my use of "outside" does not cause confusion.
It means "not in my mind or my soul". My mind,
(at least the material part of it) is sufficiently located in
space. If there is something non-material connected with
consciousness and so on (soul), it is not part of "outside".

> Now, taken together, these two confusions lead to the realist
> position, where the physicist believes that he is examining
> that which "really" exists, "outside" of his mind. In fact,
> what he is doing is examining things which are actual (in
> my sense of the term) and outside of his head. Perhaps I
> should emphasise that this is worth thinking about; many
> lives have been wasted chasing ghosts because of this
> misunderstanding.

I see not much advantage in your specification. I propose
theories about actual things outside of my head? Fine.
No problem. But also no advantage.

> >> ... but nobody tries to understand starting from
> >> the facts which are right in front of them. That's not what they
> >> were trained to do.
>
> >The point is that this is impossible. It is the old ideal of positivism
> >to derive theories from observable facts. Science works differently.
> >There is no derivation E_n ->X_n, there is guesswork. We guess
> >theories, derive the consequences, and reject the falsified guesses.
> >This is what we are trained to, and this is how science works.
>
> This is what we are trained to do, but it is only a small part of
> how science works, namely the part that Popper, with what talent
> he had, was able to identify. Since you are here invoking an argument
> from authority (Science),

My argument is not from authority (Science or Popper) but
I have described my theory of science, which agrees in its
most essential parts with Popper's theory about science.

> The problem that modern theoretical physics finds itself in is
> that people are only trained in what Newton calls the synthetic
> method, principally because it's the easiest thing to test
> in examinations. It would be rather an interesting experiment
> to take a young physicist-to-be and show him all of the phenomena
> of electromagnetism, and assign to him the project of finding
> equations which describe all the observed results.
> A physicist with this training would be better equipped for real
> investigation than his colleagues, who have merely been trained to
> do homework assignments in which the axioms are given and the
> theorems are to be deduced.

May be, but I don't argue about pedagogics.

There are more serious faults in modern science than this.
Scientists should be independent. At least as independent
as judges. Even more. Because a judge which makes obvious
nonsense harms other people, a scientist which makes
obvious nonsense doesn't.

But current scientific organization is the reverse. If you
have a grant for a few years, you have to care about your
future, the next grant. The decisions are made by
beaurocrats and other scientists (these two often unified
in a single person). The predictable result is conformity.

Another pressure into conformity is "publish or perish".
Publishing and being cited is easier in a large community
with lots of journals to publish and lots of readers.

The prediction agrees with observation. In the highly
speculative domain of modern fundamental physics,
where free science would lead to lot's of very different
directions (because they are not guided by experiments,
purely speculative, therefore its extremely hard to reject
a research proposal as false) we observe de facto only
two research directions - strings and LQG.

> I should add that it is not the case that the analytic method has
> been exhausted due to a lack of performable experiments whose
> results we do not know in advance. The confusion about realism
> and the many disagreements about quantum mechanics (among those
> who actually do think about it) are evidence enough that there
> is an opportunity, for one who is prepared to suspend judgment
> and avoid opinion, to discern something previously overlooked.

The second sentence does not prove the first. The last two
points are discussions not about the results of possible
experiments.

> >> In order to formulate the questions which relativity
> >> addresses in my language, which is to some extent solipsistic,
> >> one first has to address the question of what relationship
> >> exists between one observer's X_n and another's X'_n, which
> >> involves how one observer manifests himself within the X_n
> >> of another, and that is a difficult question.
>
> >In a realistic theory, the answer is simple: The X_n should be
> >equivalent, or at least one of the observers follows a wrong theory.
>
> There are several problem with that. One is, as you said yourself,
> that the X_n are not complete.

The M_n (perceptions) are not complete. The X_n are complete.

> That is, I perceive a room, with walls
> here and there, and you perceive a different room (presumably),
> and there is no simple matching of what I perceive to what you
> perceive.

This is about the M_n. My X_n contains also complete
information about the inside of your room. I don't perceive
the X_n. But the laws of physics, the rule which allows
to compute X_n+1=f(X_n), need the whole information,
my room as well as your room.

Ilja

seratend

Nov22-04, 05:40 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cnhmvr\\$j89\\$1@beech.fernuni-hagen.de>...\n\n> > Therefore, I definitively prefer to concentrate on logically\n> > consistent models and thereafter interpretation (connection between\n> > the model results and experiment results) before speaking of what\n> > reality is (which I think is out of scope at this step).\n>\n> In this case, there is nothing to discuss. We agree that the\n> QM "shut up and calculate" is a logically consistent way to\n> obtain testable predictions which agree with observation.\n>\nYes we have to discuss : ).\n\nThe initial advantage of focussing first on the logical content of\nmodels than the interpretation question is first to see/detect their\nequivalence. As long as they give the same result (i.e experiments), I\nprefer to leave the "true" interpretation graal question to\nphilosophers as long as it does not change the models nor the\nexperimental results. In that sense, the question of interpretation is\nsomewhat related to the preferred frame of human logic (e.g. sometimes\nit is simpler to view the em field as a set of photons in order to\nmake some simple logical deductions).\nBM is by construction an equivalent mathematical model of QM but with\nits own interpretation. The BM interpretation statement "q(t)\ndescribes the path of a bohemian particle" is non testable (included\nin the principles of BM formulation). It is thus a philosophy to\nbelieve or not on the reality interpretation of the bohmian particle\nthat has no impact on the achievable results of QM.\n\nNow, I do not mean that having a model that works is sufficient, just\nbecause I do not know where it does not work. Therefore, the search\nfor other models is important as it can detect missing points in\nexisting models.\nThe state |psi> in classical QM cannot simply mapped into classical\nmodels (e.g. with point particles – replacement of |psi>). This is one\nof the difficult points with classical QM. It prevents me to apply\nsimple logic (e.g. deterministic particle or wave view) to classical\nQM models (e.g. how to view the quantum correlation of EPR states\nwithout the state |psi>). I am thus searching for models that allow\nsimpler logic (i.e. close to every day environment).\n\nAnother point for searching new models is the testing and\nunderstanding of the QM statistical statement (i.e we can only compute\nprobabilities whatever interpretation we give). Note that I am using\nstatistical (math) rather than some other interpretations\n(non-determinism and so one ;).\n\nSorry for the "bohemian" in my different posts: my text editor does\nnot like bohmian. It sometimes replaces it randomly. : (\n\n\n> > Now, to try to end with the "reality" subject, can I translate what\n> > you call "realistic theory" into a self-consistent mathematical model?\n> > If yes, I really prefer to read "self-consistent theory" rather than\n> > "realistic theory" even if it is longer and heavier.\n>\n> Realistic theories define a subclass of the class of consistent theories.\n> QM "shut up and calculate" is a logically consistent but not realistic.\n>\n> You can view realism as a restriction of the mathematical form of the\n> theory. As well as, say, relativity defines a subclass of relativistic\n> theories which are also logically consistent but have an additional\n> requirement (some symmetry group).\n>\nyep. The problem remains on the restriction term somewhat fuzzy.\nHowever OK for me ; ) . "realistic theory" = restricted "self\nconsistent math model".\nI give up for the restricted definition ; )\n\nNow let\'s return to the BM.\n\n> > The main problem with bohemian mechanics is the selection of the\n> > equation of motion of q(t).\n>\n> No. For any given BM-like theory the equation of motion\n> of q(t) is defined and fixed.\n>\n> > As long as Bohmian mechanics accept the\n> > Schrodinger time evolution of the psi(q,t), this theory accepts that\n> > we can not see the difference between a QM measurement of the\n> > observable q and the bohemian measurement of the path q(t).\n>\n> First, there is no measurement of a path in BM.\n\nThis is one of the principal questionable BM interpretation. Each time\nwe measure the position of the particle, the path of the bohmian\nparticle is changed (if we admit that this is a "real" path), even if\nthe bohmian particle occurs to be at the same position as the\nmeasurement apparatus during an outcome.\nI know that there are several BM adept explainations. However, it\nillustrates the difficulty and the strange results (non-local\nbehaviour) we get when we try to interpret the "real" path of the\nbohmian particle.\nIf we remove the "reality" of the bohmian particle path, we just have\nQM results with a hypothetic particle following an unknown path. I\naccept BM without q(t) interpretation, there is no problem for me, I\njust do not see the utility of adding the q(t) interpretation (non\nmeasurable effects of an external information).\n\n>Second,\n> we see no difference only in quantum equilibrium. Once\n> quantum equilibrium holds in the initial state it holds always.\n> But I would guess that the initial value of the universe is\n> not quantum equilibrium but a well-defined initial value q_0.\n>\nSorry, but I do not know what is a quantum equilibrium. However, if\nyou accept the mathematical formulation dq/dt=grad(S(q,t)) with the\nclassical Bohmian Hamilton-jacobi equation and the probability\nconservation equation, you accept the statistical results of classical\nQM. If the initial quantum state is a particle with a well know\nposition at a time to, the time after (as close to "to" as you want)\nyou can find the particle everywhere in the universe (at least with\nthe non-relativistic approximation).\n\nRecall that we always have:\n\n(QM) psi(q,t) = Sqrt[rho(q,t)].exp iS(q,t)] (BM)\n\nThus if QM formultation says psi(q,t)= <q|psi(q,t)> <>0, for t>to\nwhatever q that means that we have the same thing for BM rho(q,t) (<>0\nfor all q, t>to). Therefore, we have no way to get a single known path\nq(t) even if we know at the beginning q(to).\n\n> > Once an infinite\n> > precision q measurement occurs with a result qo, the bohemian velocity\n> > is unknown at this time as in QM.\n>\n> No. The velocity of the particle d_t q(t) is known if the position\n> q(t) is known and the wave function psi(q,t) is known.\n>\nIf you measure with an infinite precision the position of a particle\ni.e. (qo,to), you have at to, v(q,to)=0, but you cannot know the\nbohmian velocity at time v(qo,to+) (infinite) nor its position q(t).\nThis is a simple application of the equations of BM and QM.\n\nThis is due to the form of the BM equations that are indeterminate for\nthis kind of state (a rho(q,to) that is a squared dirac delta(q-qo)\nequivalent to corresponding QM position state |qo>).\nA calculus on the BM equations shows this behaviour, with some\nprecautions on the discontinuities. However, it is simpler to recover\nthe results with the schroedinger equation (simple discontinuity in\ntime evolution for a starting state |qo> and applying the equation\npsi=rho.exp(iS/hbar)).\n\nFor a free particle we have:\n\n|psi(0)>=|qo> and for time t>0, we have:\n\n<q|psi(t)>=<q|U(t,0)|qo>=K.Sqrt(m/ht)exp(im(q-qo)^2/hbar.t)\n=Sqrt(rho).exp(iS/hbar)\n\n(this is the classical propagator of a free particle).\n\nThus for a free particle measure @ t=0 at position qo, we have the\nbohmian particle state for t>0:\n\nv(q,t>o)=2(q-qo)/t => q(t=0+) is undefined (|v(q,t=0+)|=oo).\nRho(q,t>0)=K\'/t= constant(q). This is the density of a bohmian\nparticle that shows that the particle may be everywhere in the\nuniverse.\n\nNow, if you do not know where is the "real particle" you can say what\nyou want: for example that the bohmian particle may have a\n(q(t),v(q(t),t)) within the probability domain of rho (i.e. rho\ndifferent from 0). As long as you cannot "see" this bohmian particle,\nyou are free to say almost everything you want as long as it remains\ncompatible with the classical QM results.\n\n>> I prefer to consider Bohmian mechanics as one\n> > possible interpretation that is compatible with the orthodox\n> > interpretation of QM, with an additional cost q(t).\n>\n> Fine. I have no problem with this point of view. Essentially, this\n> is also Bell\'s who has explicitly named it realistic interpretation\n> (AFAIR).\n>\nYes, but he may have called it surrealistic or whatever else. This is\nnot important, only the results are important: some simple\nmathematical models of the quantum probability are not compliant with\nit. You cannot use the word "realistic" itself to make reasonable\nlogical deductions, It is what I want to underline.\nWhat some people are first searching is a kind of "deterministic"\nmodel that complies with the current formulation of QM, and the\n"realistic interpretation" always displace the problem: we still have\nno simple (to be defined ; ) non statistical mathematical model that\ncomplies with all the statistical results of QM (relatively to my\npersonal knowledge : ).\nBM does not solve this issue. It simply introduces q(t) but it fails\nto remove the statistical behaviour (it is formally equivalent to QM).\nIn this aspect, BM is a useful model to explain why classical\nexplanations fail to explain QM statistical behaviour.\n\n> > Explanation the previous statement: U(t,to) is the time evolution of\n> > the initial state |psi(to)>. We have psi(q,to)=<q|psi(to)> for the\n> > bohemian particle.\n>\n> Sorry, the position of the bohmian particle is not a wave function,\n> not a bra vector <q| or a delta-function delta(q-q0). We have\n> psi(q,to)=<q|psi(to)>, but this equation has nothing to do with\n> the bohmian particle.\n>\nThe position q(t) belongs to the spectrum of the observable Q: it is\nan eigen value of Q. The QM-BM correspondence formula imposes this\naspect:\n\n(QM) psi(q,t) = Sqrt[rho(q,t)].exp iS(q,t)] (BM)\nand v(q,t)=grad(S)/m = "dq/dt".\n\nIf you do not accept this, you must accept it formally: the spectrum\nof Q is the |R set. If you are more aware of Hilbert spaces, we are\njust speaking about a representation:\n\nWe have from BM: rho(q,t) and v(q,t)\n\nWe define formally : rho(q,t)=|<q|psi(to)>|^2.\nand Grad(S(q,t))=v(q,t) => S(q,t) is defined up to a constant (not a\nproblem –gauge freedom of QM and BM).\n\nI thus can attach formally a Hilbert space generated by the basis |q>.\nTherefore, we simply define <q|psi>=psi(q,t)=\nSqrt(rho(q,t).exp(iS(q,t)/hbar)\nWe have defined equivalence between the BM state (rho, v) and the QM\nstate up to an isomorphism (gauge freedom).\n\nIn addition, the only thing we know certainly in BM and in QM is the\ndomain where we can find the particle: the set of points q where the\nprobability density rho(q,t)=|<q|psi(q,t)>|^2 is different from 0.\n\n\n> > If we define the projector P(t)=|psi(t)><psi(t)| where\n> > |psi(t)>=U(t,to)|psi(to)> as the measurement at any time t>to, we have\n> > with probability 100% the detection of the particle without the change\n> > of the state psi(t). Thus P(t) defines the trajectory of the particle\n> > with the same accuracy as the possible q(t) motions in bohemian\n> > mechanics compatible with the state psi(q,t).\n>\n> I don\'t understand. Do you want to propose a variant of\n> a bohmian theory where the "preferred observable" is, instead\n> of q or p, P(t)?\n>\nNo. P(t), P for projector, is the observable that always detects the\nparticle. If we construct a measurement apparatus based on this\nobservable, we always see/follow the particle during the time and we\ndo not change neither rho(q,t) nor |psi(t)>.\nIt is only a reformulation of the QM density matrix but on a\nmeasurement view of the path: if the state |psi(t)> is a concentrated\nwave packet (deltap.deltaq>h), we have a measurement apparatus that\nfollows the path of the QM particle without changing its state (we\nhave an uncertainty measurement in q,p).\nP(t) shows what is the most precise measurement you can do to follow\nthe QM particle path without changing its state (and thus its path).\nTherefore, the definition of the bohmian path q(t) and any measurement\nthat preserves this path does not give more information than the\nobservable P(t) (classical particle view: we just need to know where\nthe particle is, to construct its path and speed, etc …).\n\nNow, P(t) observable does not need q(t) but just (rho(q,t),v(q,t))\nknowledge. This was an argument about the utility to compute q(t) if\nwe are not able to "see" it.\n\n>\n> Ok. I don\'t claim "shut up and calculate" is meaningless. It gives\n> testable predictions, in a consistent way. It does not give explanations.\n>\n> > Note that Bohmian meachanics has the same interpretation problem if\n> > you do want to make a simple "shut up and calculate".\n>\n> No.\n\nQM postulates say exactly where there are no explanations (i.e. what\nit does not try to explain). "The shut up and calculate" comes from\nthe measurement postulate and the Born rules. It is well known and\nthere is a lot of work on how QM could explain the measurement results\n(i.e. without the measurement postulate).\nOne tentative of such an explanation is the decoherence program\n(Zurek, ...). It tries to explain the macroscopic results of QM\nexperiments just using all the QM postulates except the measurement\npostulate (explanation/interpretation of the state projection from the\nother postulates of QM).\nSo if it is the shut up and calculate of the QM measurement that is\nnot acceptable for you, may be you should look at the decoherence\nprogram (very instructive, I think).\n\n>\n> > q(t) has currently no measurable signification.\n>\n> So what? Theories are obliged to make testable predictions in\n> general, but not for all their particular claims, objects and so on.\n>\nThe main problem is that BM says that we cannot measure q(t). As long\nas we keep this affirmation, I will say QM=BM.\nIt is like fluid mechanics: do you prefer the Lagrangian (~BM) or the\nEulerian (~QM) point of view? Does the 2 points change the Newtonian\nphysics and interpretation?\nIt is used to prefer the Eulerian point of view in fluid mechanics\n(simpler calculus and models I think, I am not an expert). This is\nalmost the same with QM.\nAdding the virtual path q(t) to BM does not provide new information.\nIt does not try to explain rho(q,t) as q(t) cannot be measured. It is\nthus a postulate (q(t) is the path of a particle) in this theory that\nbrings nothing (testable): we can assume it true or false, it does not\nchange the theory which relies only on rho(q,t) and v(q,t)=grad(S)\n(equivalent to the wave function in QM\npsi(q,t)=Sqrt(Rho).exp(iS/habr)).\n\n> > Why not taking (q+p)(t) as the\n> > trajectory of new bohemian particle (observable Q+P)? Who is the more\n> > "real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.\n>\n> This is not a problem of a particular theory. This is a question about\n> preference for one theory in comparison with another. BM_q is one\n> theory, BM_{p+q} another. BM_q is simpler, therefore preferred\n> by Occam\'s razor.\n>\nNo all these theories are the same. Only the interpretation of the\npossible path q(t) with the allowed density probability may be\ndifferent.\nIn this aspect you are close to the wave QM where the q representation\nis selected (i.e. use of psi(q,t) instead of the vector |psi> in\nanother basis). You say BM_q is simpler. However, may be, haven\'t you\ntried to look to other representations for a given rho(q,t) and given\nsymmetries of the interactions? I am quiet sure, as in QM, that you\ncan find representations (other basis) that are simpler than the q\nrepresentation in such cases (energy representation should be fun).\n\n> > By preferred frame, If you intend preferred observable (e.g. selection\n> > of q(t) in BM), I think that bohmian mechanics cannot give you what\n> > your are looking for, i.e. you can construct in principle a "A\n> > bohmian" mechanics, where A is any observable of QM.\n>\n> I know. But that doesn\'t matter. I can prefer among this large\n> class of theories the one I prefer for reasons of simplicity. Moreover,\n> I can live quite comfortable having several theories, as long as\n> they all agree with observation and are realistic.\n\nIt is what I am also trying to underline. I think you are selecting\nBM_q as a candidate theory, not for simplicity but because it is well\ndescribed today (since 1952). Because BM_q is attached to the\nobservable Q of QM mechanics, we are trying to say that q(t) is the\nsomewhat path of a reality (interpretation). However if we use BM_p,\nthe path p(t) represents another reality (surely it can be interpreted\nas a set of paths q(t)). Why should the path q(t) represents the real\npath while p(t) not?\nAt least, the multiple equivalent Newtonian classical models\n(Hamilton, Hamilton jacobi, Lagrange, ...) always give the same\ninterpretation. For a realistic adept, I think you should solve this\nproblem (may be it is already solved, but I do not know, I am not a BM\nspecialist :).\n\n>\n> BTW, whatever observable we use here - note that time t is not an\n> observable in the classical sense in Schroedinger theory (different\n> from x). As BM_q, as BM_p, as BM_A handle time t as a\n> preferred absolute time.\n>\nThis is the classical approximation (E <<mc^2) in a preferred\nrepresentation (analogue to use the wave function psi(q,t) rather than\nthe vector |psi(t)>). You can change easily change the form of\nclassical QM (implicit time representation) to get time handled in a\nmore symmetrical way as position q (i.e you may define a time\noperator, etc …). In relativistic QM theory, time is handled more\nsymmetrically.\nNote that there is also a relativistic BM. However, the things become\nworse as we have not the particle number conservation in relativistic\nQM (requires the use of QFT instead and all the ugly things associated\nwith, like renormalisation).\n\n\n> Note that nobody has ever claimed that BM is the final theory\n> of everything. BM is a theory which invites to think about\n> modifications. For example, one direction is Nelsonian\n> stochastics. Whatever.\n\nI know. However, I really do not see what additional concrete features\nBM brought. BM_q does not explain the wave function psi(q,t) of QM,\nthe hidden variables that may generate psi(q,t) and all the\nprobability spaces attached to it. It uses it and on the top of it, it\ndefines a virtual path q(t): it is only a q representation of QM\nstatistics that\'s all.\nI do not know Nelsonian stochasitcs , so if you have a pointer to\npaper, I would be very glad to see it.\nHowever, I am afraid that this will be the usual second way of\nrepresenting QM: the use of a Langevin type interaction/field. You\njust replace the psi(q,t) wave function by a Brownian type motion\nnoise. Well, we stay with the same problem: you cannot explain the\nsource of this Brownian noise better than the statistical results of\nQM. We recover the BM_q problems (non locality, existence of a virtual\nparticle, etc ...): we cannot explain the source of randomness of QM\nresults nor the "reality of this noise source".\n\n>\n> Assume there is some more fundamental theory, with QM as\n> some limit but itself different from QM. Now, if we look more\n> careful, its quite probable that BM_q is much more close\n> to this theory than BM_p. To find this theory, BM_q may be\n> the starting point, BM_p will be a dead end.\n>\nI am almost sure that BM_p is as close as BM_q as we want, but\ncurrently I have not the courage to prove such a thing. Only the\nbohmian particle path should be reinterpreted in both cases.\n\n>\n> No. In the domain of QM foundations I only argue that the violation\n> of BI requires the acceptance of a preferred foliation. At least if\n> we do not want to give up realism and/or causality.\n>\nWell all the results of QM tend to show that there is no preferred\nrepresentation. Instead, these results tend to show that there is only\na relativity of representation defined by the interaction between the\nsystem under measurement and the measurement apparatus.\nThe entangled states may help to understand this topic. We have a\nglobal state that has only a spin reference relatively to both the\nentangle particles. The EPR source creates locally this relative\ndependence. A measurement defines the state of one particle relatively\nto the measurement apparatus; we can thus know the state of the other\nparticle relatively to this apparatus and verify it with another\napparatus. The two measures are really independent. We just see the\ncorrelation of the global entangled state by comparing the 2 measures.\n\nNow the usual interpretation of this correlation (the bell "realistic"\nmodel) does not explain how we can get this degree of correlation: we\ncan remove this model (+its context) to explain QM statistical\nresults.\n\nDue to the weak definition of realism (the restriction of the\ntheories), we cannot answer if we give up or not the "realism" without\nthis simple Bell "realistic" model. I think it is not very important.\nOnly the remaining set of models compatible with QM results is\nimportant. You can label such a remaining model "realistic",\n"surrealistic", etc … the name itself is not important just the\ncontent and the results it can give. Note that one result brought by\nthese hypothetic models may the redefinition/philosophy of reality : )\n\nTherefore, whatever is your answer about realism, my question remains:\nwhat type of *new* models can you remove, with such affirmation?\n\nRelativistic causality defined in QFT is compatible with the EPR\nresults. Therefore, I can say QM preserves causality (in the sense of\nthe one defined/understood in QFT). Note that the causality you are\ntrying to define seems to be very close to the cluster decomposition\nof QFT.\n\n> In the context of this discussion I defend BM_q as an example\n> of a realistic causal theory with preferred frame. It proves\n> that such theories exist, and, therefore, realism and causality\n> are consistent with observation.\n>\nSo your current postulate (proof) is realism= yes because there is at\nleast a BM_q theory explaining some results of QM. I assume in that\ncase that any theory postulating a virtual particle path is realist in\nyour definition.\nOk, QM becomes realist if I attach to it an independent postulate: for\neach quantum system, there exists a non measurable path associated to\nthe quantum particle (call it q(t)). QFT may also be amended with such\na postulate and every theory.\n\nCausality=yes. QM already says yes to its causality. Nothing new with\nthis statement.\n\nSeratend.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote in message news:<cnhmvr$j89$1@beech.fernuni-hagen.de>...

> > Therefore, I definitively prefer to concentrate on logically
> > consistent models and thereafter interpretation (connection between
> > the model results and experiment results) before speaking of what
> > reality is (which I think is out of scope at this step).
>
> In this case, there is nothing to discuss. We agree that the
> QM "shut up and calculate" is a logically consistent way to
> obtain testable predictions which agree with observation.
>
Yes we have to discuss : ).

The initial advantage of focussing first on the logical content of
models than the interpretation question is first to see/detect their
equivalence. As long as they give the same result (i.e experiments), I
prefer to leave the "true" interpretation graal question to
philosophers as long as it does not change the models nor the
experimental results. In that sense, the question of interpretation is
somewhat related to the preferred frame of human logic (e.g. sometimes
it is simpler to view the em field as a set of photons in order to
make some simple logical deductions).
BM is by construction an equivalent mathematical model of QM but with
its own interpretation. The BM interpretation statement "q(t)
describes the path of a bohemian particle" is non testable (included
in the principles of BM formulation). It is thus a philosophy to
believe or not on the reality interpretation of the bohmian particle
that has no impact on the achievable results of QM.

Now, I do not mean that having a model that works is sufficient, just
because I do not know where it does not work. Therefore, the search
for other models is important as it can detect missing points in
existing models.
The state |\psi> in classical QM cannot simply mapped into classical
models (e.g. with point particles – replacement of |\psi>). This is one
of the difficult points with classical QM. It prevents me to apply
simple logic (e.g. deterministic particle or wave view) to classical
QM models (e.g. how to view the quantum correlation of EPR states
without the state |\psi>). I am thus searching for models that allow
simpler logic (i.e. close to every day environment).

Another point for searching new models is the testing and
understanding of the QM statistical statement (i.e we can only compute
probabilities whatever interpretation we give). Note that I am using
statistical (math) rather than some other interpretations
(non-determinism and so one ;).

Sorry for the "bohemian" in my different posts: my text editor does
not like bohmian. It sometimes replaces it randomly. : (

> > Now, to try to end with the "reality" subject, can I translate what
> > you call "realistic theory" into a self-consistent mathematical model?
> > If yes, I really prefer to read "self-consistent theory" rather than
> > "realistic theory" even if it is longer and heavier.
>
> Realistic theories define a subclass of the class of consistent theories.
> QM "shut up and calculate" is a logically consistent but not realistic.
>
> You can view realism as a restriction of the mathematical form of the
> theory. As well as, say, relativity defines a subclass of relativistic
> theories which are also logically consistent but have an additional
> requirement (some symmetry group).
>
yep. The problem remains on the restriction term somewhat fuzzy.
However OK for me ; ) . "realistic theory" = restricted "self
consistent math model".
I give up for the restricted definition ; )

Now let's return to the BM.

> > The main problem with bohemian mechanics is the selection of the
> > equation of motion of q(t).
>
> No. For any given BM-like theory the equation of motion
> of q(t) is defined and fixed.
>
> > As long as Bohmian mechanics accept the
> > Schrodinger time evolution of the \psi(q,t), this theory accepts that
> > we can not see the difference between a QM measurement of the
> > observable q and the bohemian measurement of the path q(t).
>
> First, there is no measurement of a path in BM.

This is one of the principal questionable BM interpretation. Each time
we measure the position of the particle, the path of the bohmian
particle is changed (if we admit that this is a "real" path), even if
the bohmian particle occurs to be at the same position as the
measurement apparatus during an outcome.
I know that there are several BM adept explainations. However, it
illustrates the difficulty and the strange results (non-local
behaviour) we get when we try to interpret the "real" path of the
bohmian particle.
If we remove the "reality" of the bohmian particle path, we just have
QM results with a hypothetic particle following an unknown path. I
accept BM without q(t) interpretation, there is no problem for me, I
just do not see the utility of adding the q(t) interpretation (non
measurable effects of an external information).

>Second,
> we see no difference only in quantum equilibrium. Once
> quantum equilibrium holds in the initial state it holds always.
> But I would guess that the initial value of the universe is
> not quantum equilibrium but a well-defined initial value q_0.
>
Sorry, but I do not know what is a quantum equilibrium. However, if
you accept the mathematical formulation dq/dt=grad(S(q,t)) with the
classical Bohmian Hamilton-jacobi equation and the probability
conservation equation, you accept the statistical results of classical
QM. If the initial quantum state is a particle with a well know
position at a time to, the time after (as close to "to" as you want)
you can find the particle everywhere in the universe (at least with
the non-relativistic approximation).

Recall that we always have:

(QM) \psi(q,t) = \Sqrt[\rho(q,t)].\exp iS(q,t)] (BM)

Thus if QM formultation says \psi(q,t)= <q|\psi(q,t)> <>0, for t>to
whatever q that means that we have the same thing for BM \rho(q,t) (<>0
for all q, t>to). Therefore, we have no way to get a single known path
q(t) even if we know at the beginning q(to).

> > Once an infinite
> > precision q measurement occurs with a result qo, the bohemian velocity
> > is unknown at this time as in QM.
>
> No. The velocity of the particle d_t q(t) is known if the position
> q(t) is known and the wave function \psi(q,t) is known.
>
If you measure with an infinite precision the position of a particle
i.e. (qo,to), you have at to, v(q,to)=0, but you cannot know the
bohmian velocity at time v(qo,to+) (infinite) nor its position q(t).
This is a simple application of the equations of BM and QM.

This is due to the form of the BM equations that are indeterminate for
this kind of state (a \rho(q,to) that is a squared dirac \delta(q-qo)
equivalent to corresponding QM position state |qo>).
A calculus on the BM equations shows this behaviour, with some
precautions on the discontinuities. However, it is simpler to recover
the results with the schroedinger equation (simple discontinuity in
time evolution for a starting state |qo> and applying the equation
\psi=\rho.\exp(iS/\hbar)).

For a free particle we have:

|\psi(0)>=|qo>[/itex] and for time t>0, we have:

<q|\psi(t)>=<q|U(t,0)|qo>=K.\Sqrt(m/ht)\exp(im(q-qo)^2/\hbar.t)
=\Sqrt(\rho).\exp(iS/\hbar)

(this is the classical propagator of a free particle).

Thus for a free particle measure @ t=0 at position qo, we have the
bohmian particle state for t>0:

v(q,t>o)=2(q-qo)/t => q(t=0+) is undefined (|v(q,t=0+)|=oo).\Rho(q,t>0)=K'/t= constant(q). This is the density of a bohmian
particle that shows that the particle may be everywhere in the
universe.

Now, if you do not know where is the "real particle" you can say what
you want: for example that the bohmian particle may have a
(q(t),v(q(t),t)) within the probability domain of \rho (i.e. \rho
different from 0). As long as you cannot "see" this bohmian particle,
you are free to say almost everything you want as long as it remains
compatible with the classical QM results.

>> I prefer to consider Bohmian mechanics as one
> > possible interpretation that is compatible with the orthodox
> > interpretation of QM, with an additional cost q(t).
>
> Fine. I have no problem with this point of view. Essentially, this
> is also Bell's who has explicitly named it realistic interpretation
> (AFAIR).
>
Yes, but he may have called it surrealistic or whatever else. This is
not important, only the results are important: some simple
mathematical models of the quantum probability are not compliant with
it. You cannot use the word "realistic" itself to make reasonable
logical deductions, It is what I want to underline.
What some people are first searching is a kind of "deterministic"
model that complies with the current formulation of QM, and the
"realistic interpretation" always displace the problem: we still have
no simple (to be defined ; ) non statistical mathematical model that
complies with all the statistical results of QM (relatively to my
personal knowledge : ).
BM does not solve this issue. It simply introduces q(t) but it fails
to remove the statistical behaviour (it is formally equivalent to QM).
In this aspect, BM is a useful model to explain why classical
explanations fail to explain QM statistical behaviour.

> > Explanation the previous statement: U(t,to) is the time evolution of
> > the initial state |\psi(to)>. We have \psi(q,to)=<q|\psi(to)> for the
> > bohemian particle.
>
> Sorry, the position of the bohmian particle is not a wave function,
> not a bra vector <q| or a \delta-function \delta(q-q0). We have
> \psi(q,to)=<q|\psi(to)>, but this equation has nothing to do with
> the bohmian particle.
>
The position q(t) belongs to the spectrum of the observable Q: it is
an eigen value of Q. The QM-BM correspondence formula imposes this
aspect:

(QM) \psi(q,t) = \Sqrt[\rho(q,t)].\exp iS(q,t)] (BM)
and v(q,t)=grad(S)/m = "dq/dt".

If you do not accept this, you must accept it formally: the spectrum
of Q is the |R set. If you are more aware of Hilbert spaces, we are
just speaking about a representation:

We have from BM: \rho(q,t) and v(q,t)

We define formally : \rho(q,t)=|<q|\psi(to)>|^2.
and Grad(S(q,t))=v(q,t) => S(q,t) is defined up to a constant (not a
problem –gauge freedom of QM and BM).

I thus can attach formally a Hilbert space generated by the basis |q>.
Therefore, we simply define <q|\psi>=\psi(q,t)=\Sqrt(\rho(q,t).\exp(iS(q,t)/\hbar)
We have defined equivalence between the BM state (\rho, v) and the QM
state up to an isomorphism (gauge freedom).

In addition, the only thing we know certainly in BM and in QM is the
domain where we can find the particle: the set of points q where the
probability density \rho(q,t)=|<q|\psi(q,t)>|^2 is different from .

> > If we define the projector [itex]P(t)=|\psi(t)><\psi(t)| where
> > |\psi(t)>=U(t,to)|\psi(to)> as the measurement at any time t>to, we have
> > with probability 100% the detection of the particle without the change
> > of the state \psi(t). Thus P(t) defines the trajectory of the particle
> > with the same accuracy as the possible q(t) motions in bohemian
> > mechanics compatible with the state \psi(q,t).
>
> I don't understand. Do you want to propose a variant of
> a bohmian theory where the "preferred observable" is, instead
> of q or p, P(t)?
>
No. P(t), P for projector, is the observable that always detects the
particle. If we construct a measurement apparatus based on this
observable, we always see/follow the particle during the time and we
do not change neither \rho(q,t) nor |\psi(t)>.
It is only a reformulation of the QM density matrix but on a
measurement view of the path: if the state |\psi(t)> is a concentrated
wave packet (deltap.deltaq>h), we have a measurement apparatus that
follows the path of the QM particle without changing its state (we
have an uncertainty measurement in q,p).
P(t) shows what is the most precise measurement you can do to follow
the QM particle path without changing its state (and thus its path).
Therefore, the definition of the bohmian path q(t) and any measurement
that preserves this path does not give more information than the
observable P(t) (classical particle view: we just need to know where
the particle is, to construct its path and speed, etc …).

Now, P(t) observable does not need q(t) but just (\rho(q,t),v(q,t))
knowledge. This was an argument about the utility to compute q(t) if
we are not able to "see" it.

>
> Ok. I don't claim "shut up and calculate" is meaningless. It gives
> testable predictions, in a consistent way. It does not give explanations.
>
> > Note that Bohmian meachanics has the same interpretation problem if
> > you do want to make a simple "shut up and calculate".
>
> No.

QM postulates say exactly where there are no explanations (i.e. what
it does not try to explain). "The shut up and calculate" comes from
the measurement postulate and the Born rules. It is well known and
there is a lot of work on how QM could explain the measurement results
(i.e. without the measurement postulate).
One tentative of such an explanation is the decoherence program
(Zurek, ...). It tries to explain the macroscopic results of QM
experiments just using all the QM postulates except the measurement
postulate (explanation/interpretation of the state projection from the
other postulates of QM).
So if it is the shut up and calculate of the QM measurement that is
not acceptable for you, may be you should look at the decoherence
program (very instructive, I think).

>
> > q(t) has currently no measurable signification.
>
> So what? Theories are obliged to make testable predictions in
> general, but not for all their particular claims, objects and so on.
>
The main problem is that BM says that we cannot measure q(t). As long
as we keep this affirmation, I will say QM=BM.
It is like fluid mechanics: do you prefer the Lagrangian (~BM) or the
Eulerian (~QM) point of view? Does the 2 points change the Newtonian
physics and interpretation?
It is used to prefer the Eulerian point of view in fluid mechanics
(simpler calculus and models I think, I am not an expert). This is
almost the same with QM.
Adding the virtual path q(t) to BM does not provide new information.
It does not try to explain \rho(q,t) as q(t) cannot be measured. It is
thus a postulate (q(t) is the path of a particle) in this theory that
brings nothing (testable): we can assume it true or false, it does not
change the theory which relies only on \rho(q,t) and v(q,t)=grad(S)
(equivalent to the wave function in QM\psi(q,t)=\Sqrt(\Rho).\exp(iS/habr)).

> > Why not taking (q+p)(t) as the
> > trajectory of new bohemian particle (observable Q+P)? Who is the more
> > "real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.
>
> This is not a problem of a particular theory. This is a question about
> preference for one theory in comparison with another. BM_q is one
> theory, BM_{p+q} another. BM_q is simpler, therefore preferred
> by Occam's razor.
>
No all these theories are the same. Only the interpretation of the
possible path q(t) with the allowed density probability may be
different.
In this aspect you are close to the wave QM where the q representation
is selected (i.e. use of \psi(q,t) instead of the vector |\psi> in
another basis). You say BM_q is simpler. However, may be, haven't you
tried to look to other representations for a given \rho(q,t) and given
symmetries of the interactions? I am quiet sure, as in QM, that you
can find representations (other basis) that are simpler than the q
representation in such cases (energy representation should be fun).

> > By preferred frame, If you intend preferred observable (e.g. selection
> > of q(t) in BM), I think that bohmian mechanics cannot give you what
> > your are looking for, i.e. you can construct in principle a "A
> > bohmian" mechanics, where A is any observable of QM.
>
> I know. But that doesn't matter. I can prefer among this large
> class of theories the one I prefer for reasons of simplicity. Moreover,
> I can live quite comfortable having several theories, as long as
> they all agree with observation and are realistic.

It is what I am also trying to underline. I think you are selecting
BM_q as a candidate theory, not for simplicity but because it is well
described today (since 1952). Because BM_q is attached to the
observable Q of QM mechanics, we are trying to say that q(t) is the
somewhat path of a reality (interpretation). However if we use BM_p,
the path p(t) represents another reality (surely it can be interpreted
as a set of paths q(t)). Why should the path q(t) represents the real
path while p(t) not?
At least, the multiple equivalent Newtonian classical models
(Hamilton, Hamilton jacobi, Lagrange, ...) always give the same
interpretation. For a realistic adept, I think you should solve this
problem (may be it is already solved, but I do not know, I am not a BM
specialist :).

>
> BTW, whatever observable we use here - note that time t is not an
> observable in the classical sense in Schroedinger theory (different
> from x). As BM_q, as BM_p, as BM_A handle time t as a
> preferred absolute time.
>
This is the classical approximation (E <<mc^2) in a preferred
representation (analogue to use the wave function \psi(q,t) rather than
the vector |\psi(t)>). You can change easily change the form of
classical QM (implicit time representation) to get time handled in a
more symmetrical way as position q (i.e you may define a time
operator, etc …). In relativistic QM theory, time is handled more
symmetrically.
Note that there is also a relativistic BM. However, the things become
worse as we have not the particle number conservation in relativistic
QM (requires the use of QFT instead and all the ugly things associated
with, like renormalisation).

> Note that nobody has ever claimed that BM is the final theory
> of everything. BM is a theory which invites to think about
> modifications. For example, one direction is Nelsonian
> stochastics. Whatever.

I know. However, I really do not see what additional concrete features
BM brought. BM_q does not explain the wave function \psi(q,t) of QM,
the hidden variables that may generate \psi(q,t) and all the
probability spaces attached to it. It uses it and on the top of it, it
defines a virtual path q(t): it is only a q representation of QM
statistics that's all.
I do not know Nelsonian stochasitcs , so if you have a pointer to
paper, I would be very glad to see it.
However, I am afraid that this will be the usual second way of
representing QM: the use of a Langevin type interaction/field. You
just replace the \psi(q,t) wave function by a Brownian type motion
noise. Well, we stay with the same problem: you cannot explain the
source of this Brownian noise better than the statistical results of
QM. We recover the BM_q problems (non locality, existence of a virtual
particle, etc ...): we cannot explain the source of randomness of QM
results nor the "reality of this noise source".

>
> Assume there is some more fundamental theory, with QM as
> some limit but itself different from QM. Now, if we look more
> careful, its quite probable that BM_q is much more close
> to this theory than BM_p. To find this theory, BM_q may be
> the starting point, BM_p will be a dead end.
>
I am almost sure that BM_p is as close as BM_q as we want, but
currently I have not the courage to prove such a thing. Only the
bohmian particle path should be reinterpreted in both cases.

>
> No. In the domain of QM foundations I only argue that the violation
> of BI requires the acceptance of a preferred foliation. At least if
> we do not want to give up realism and/or causality.
>
Well all the results of QM tend to show that there is no preferred
representation. Instead, these results tend to show that there is only
a relativity of representation defined by the interaction between the
system under measurement and the measurement apparatus.
The entangled states may help to understand this topic. We have a
global state that has only a spin reference relatively to both the
entangle particles. The EPR source creates locally this relative
dependence. A measurement defines the state of one particle relatively
to the measurement apparatus; we can thus know the state of the other
particle relatively to this apparatus and verify it with another
apparatus. The two measures are really independent. We just see the
correlation of the global entangled state by comparing the 2 measures.

Now the usual interpretation of this correlation (the bell "realistic"
model) does not explain how we can get this degree of correlation: we
can remove this model (+its context) to explain QM statistical
results.

Due to the weak definition of realism (the restriction of the
theories), we cannot answer if we give up or not the "realism" without
this simple Bell "realistic" model. I think it is not very important.
Only the remaining set of models compatible with QM results is
important. You can label such a remaining model "realistic",
"surrealistic", etc … the name itself is not important just the
content and the results it can give. Note that one result brought by
these hypothetic models may the redefinition/philosophy of reality : )

Therefore, whatever is your answer about realism, my question remains:
what type of *new* models can you remove, with such affirmation?

Relativistic causality defined in QFT is compatible with the EPR
results. Therefore, I can say QM preserves causality (in the sense of
the one defined/understood in QFT). Note that the causality you are
trying to define seems to be very close to the cluster decomposition
of QFT.

> In the context of this discussion I defend BM_q as an example
> of a realistic causal theory with preferred frame. It proves
> that such theories exist, and, therefore, realism and causality
> are consistent with observation.
>
So your current postulate (proof) is realism= yes because there is at
least a BM_q theory explaining some results of QM. I assume in that
case that any theory postulating a virtual particle path is realist in
your definition.
Ok, QM becomes realist if I attach to it an independent postulate: for
each quantum system, there exists a non measurable path associated to
the quantum particle (call it q(t)). QFT may also be amended with such
a postulate and every theory.

Causality=yes. QM already says yes to its causality. Nothing new with
this statement.

Seratend.

Ilja Schmelzer

Nov24-04, 01:46 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"seratend" <ser_monmail@yahoo.fr> schrieb\n> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote\n> > > Therefore, I definitively prefer to concentrate on logically\n> > > consistent models and thereafter interpretation (connection between\n> > > the model results and experiment results) before speaking of what\n> > > reality is (which I think is out of scope at this step).\n> >\n> > In this case, there is nothing to discuss. We agree that the\n> > QM "shut up and calculate" is a logically consistent way to\n> > obtain testable predictions which agree with observation.\n> >\n> Yes we have to discuss : ).\n> The initial advantage of focussing first on the logical content of\n> models than the interpretation question is first to see/detect their\n> equivalence. As long as they give the same result (i.e experiments), I\n> prefer to leave the "true" interpretation graal question to\n> philosophers as long as it does not change the models nor the\n> experimental results.\n\nFine. Now, the point of my argumentation is not that I prefer one\nof the versions of BM in comparison to others. But I\'m interested\nin common properties of all (all reasonable in some sense)\ninterpretations. Now, the preferred frame is a common property\nof all interpretations which are realistic and have a notion of\ncausality related with spacetime.\n\n> I am thus searching for models that allow\n> simpler logic (i.e. close to every day environment).\n\nYep. But the details of such a model don\'t really matter, as far as:\n\n1.) We have at least one model close to "simple logic" (EPR realism,\nno "quantum logic" but classical logic, causality), which is reached\nwith BM.\n2.) We can prove general properties of all theories of this class\nwith "simple logic". Violation of Einstein causality is such a common\nproperty, as proven by Bell. The only way to connect this with\ncausality is return to classical causality in a preferred foliation.\n\nOnce we have these two points, we can be quite sure that there\nexists a preferred frame. And this is what I want. (As support\nfor my own ether theories.)\n\nThus, my main argument is that I have a prove using the "simple logic"\nof realism, and BM shows that there is nothing wrong with this "simple\nlogic", at least it is not in contradiction with observation.\n\n> Sorry for the "bohemian" in my different posts: my text editor does\n> not like bohmian. It sometimes replaces it randomly. : (\n\n;-))). As an anarchist, I have no problems with bohemians.\n\n> > Realistic theories define a subclass of the class of consistent\ntheories.\n> > QM "shut up and calculate" is a logically consistent but not realistic.\n> >\n> > You can view realism as a restriction of the mathematical form of the\n> > theory. As well as, say, relativity defines a subclass of relativistic\n> > theories which are also logically consistent but have an additional\n> > requirement (some symmetry group).\n> >\n> yep. The problem remains on the restriction term somewhat fuzzy.\n\nFor the purpose of my argument, the following pre-definition is\nsufficient: Realism is what is necessary to prove, combined\nwith Einstein causality, Bell\'s inequality.\n\nBell\'s inequality is a strong theorem. Every definition in agreement\nwith this pre-definition gives me what I need. You can work out\nyourself the details of the definition, if my definition is too fuzzy for\nyou.\n\n> > First, there is no measurement of a path in BM.\n\n> If we remove the "reality" of the bohmian particle path, we just have\n> QM results with a hypothetic particle following an unknown path. I\n> accept BM without q(t) interpretation, there is no problem for me, I\n> just do not see the utility of adding the q(t) interpretation (non\n> measurable effects of an external information).\n\nThe use of unobservable but real things is that we can apply\nclassical logic to them. If the theory gives me x in {x1,x2,...xn}\nbut I cannot observe which it is, I nonetheless can prove\nsomething about x. If A(x1) and A(x2) and .... A(xn) then\nA(x).\n\nThe particular example I need is x in {A->B, B->A},\nEinstein causality violated(A->B) and\nEinstein causality violated(B->A) then\nEinstein causality violated(x).\n\n> >Second,\n> > we see no difference only in quantum equilibrium. Once\n> > quantum equilibrium holds in the initial state it holds always.\n> > But I would guess that the initial value of the universe is\n> > not quantum equilibrium but a well-defined initial value q_0.\n\n> Sorry, but I do not know what is a quantum equilibrium.\n\nA probabilistic state with some fixed (but arbitrary) wave function\nPsi(q) and the probability distribution rho(q)=<Psi(q)Psi(q)>\nfor the position q.\n\n> However, if\n> you accept the mathematical formulation dq/dt=grad(S(q,t)) with the\n> classical Bohmian Hamilton-jacobi equation and the probability\n> conservation equation, you accept the statistical results of classical\n> QM. If the initial quantum state is a particle with a well know\n> position at a time to, the time after (as close to "to" as you want)\n> you can find the particle everywhere in the universe (at least with\n> the non-relativistic approximation).\n\nNo. If in the initial state the particle has a well-defined position q,\nit has always a well-defined position q(t). But don\'t mingle here\na state with position q0 with a state with wave function delta(q-q0).\nSuch degenerate states are not even allowed in BM.\n\n> Recall that we always have:\n>\n> (QM) psi(q,t) = Sqrt[rho(q,t)].exp iS(q,t)] (BM)\n\nNo. This holds only in quantum equilibrium.\n\n> > > Once an infinite\n> > > precision q measurement occurs with a result qo, the bohemian velocity\n> > > is unknown at this time as in QM.\n> >\n> > No. The velocity of the particle d_t q(t) is known if the position\n> > q(t) is known and the wave function psi(q,t) is known.\n\n> If you measure with an infinite precision the position of a particle\n> i.e. (qo,to), you have at to, v(q,to)=0, but you cannot know the\n> bohmian velocity at time v(qo,to+) (infinite) nor its position q(t).\n> This is a simple application of the equations of BM and QM.\n\nI should have said that there are no infinite precision measurements\nin BM (as well as QM). delta(q-q0) is not in H.\n\n> >> I prefer to consider Bohmian mechanics as one\n> > > possible interpretation that is compatible with the orthodox\n> > > interpretation of QM, with an additional cost q(t).\n> >\n> > Fine. I have no problem with this point of view. Essentially, this\n> > is also Bell\'s who has explicitly named it realistic interpretation\n> > (AFAIR).\n\n> Yes, but he may have called it surrealistic or whatever else. This is\n> not important, only the results are important: some simple\n> mathematical models of the quantum probability are not compliant with\n> it. You cannot use the word "realistic" itself to make reasonable\n> logical deductions, It is what I want to underline.\n\nOf course, the word "realistic" is not the point. The point is that we\ncan apply the math which is necessary to prove such things like\nBell\'s inequality for causally disconnected events.\n\n> What some people are first searching is a kind of "deterministic"\n> model that complies with the current formulation of QM, and the\n> "realistic interpretation" always displace the problem: we still have\n> no simple (to be defined ; ) non statistical mathematical model that\n> complies with all the statistical results of QM (relatively to my\n> personal knowledge : ).\n> BM does not solve this issue. It simply introduces q(t) but it fails\n> to remove the statistical behaviour (it is formally equivalent to QM).\n> In this aspect, BM is a useful model to explain why classical\n> explanations fail to explain QM statistical behaviour.\n\nI don\'t understand this point. As long as statistical behaviour is\nof the same type as in classical thermodynamics, uncertainty\ncaused by uncertainty of the initial values and classical chaos,\nI see no reason to remove it.\n\n> > > Explanation the previous statement: U(t,to) is the time evolution of\n> > > the initial state |psi(to)>. We have psi(q,to)=<q|psi(to)> for the\n> > > bohemian particle.\n\n> > Sorry, the position of the bohmian particle is not a wave function,\n> > not a bra vector <q| or a delta-function delta(q-q0). We have\n> > psi(q,to)=<q|psi(to)>, but this equation has nothing to do with\n> > the bohmian particle.\n\n> The position q(t) belongs to the spectrum of the observable Q: it is\n> an eigen value of Q. The QM-BM correspondence formula imposes this\n> aspect:\n>\n> (QM) psi(q,t) = Sqrt[rho(q,t)].exp iS(q,t)] (BM)\n> and v(q,t)=grad(S)/m = "dq/dt".\n\nThe correct formulation:\npsi(q,t) = R(q,t).exp iS(q,t)\nSchroedinger equation for psi,\nv(q,t)=grad(S)/m = dq/dt.\nComplete configuration space: R(q), S(q), q.\n\nQuantum equilibrium: rho(q) = R^2(q).\n\nThe position of q belongs to IR.\n\n> If you do not accept this, you must accept it formally: the spectrum\n> of Q is the |R set. If you are more aware of Hilbert spaces, we are\n> just speaking about a representation:\n>\n> We have from BM: rho(q,t) and v(q,t)\n\nThis is not the complete BM state. It is only sufficient to define\nthe state in quantum equilibrium.\n\n> [considerations about P(t)]\n> This was an argument about the utility to compute q(t) if\n> we are not able to "see" it.\n\nI see no utility in computing q(t), except for illustration. Or for\nchecking various consistency results.\n\n> > Ok. I don\'t claim "shut up and calculate" is meaningless. It gives\n> > testable predictions, in a consistent way. It does not give\nexplanations.\n> >\n> > > Note that Bohmian meachanics has the same interpretation problem if\n> > > you do want to make a simple "shut up and calculate".\n> >\n> > No.\n>\n> QM postulates say exactly where there are no explanations (i.e. what\n> it does not try to explain). "The shut up and calculate" comes from\n> the measurement postulate and the Born rules.\n\nIn this sense "shut up and calculate" is not an explanation. BM is\ndifferent. It explains the probabilities by evolution equations and\ninitial values, thus, in the same way as classical mechanics.\n\n> It is well known and\n> there is a lot of work on how QM could explain the measurement\n> results (i.e. without the measurement postulate).\n> One tentative of such an explanation is the decoherence program\n> (Zurek, ...). It tries to explain the macroscopic results of QM\n> experiments just using all the QM postulates except the measurement\n> postulate (explanation/interpretation of the state projection from the\n> other postulates of QM).\n> So if it is the shut up and calculate of the QM measurement that is\n> not acceptable for you, may be you should look at the decoherence\n> program (very instructive, I think).\n\nI don\'t believe that decoherence works in this way.\nDecoherence explains human inabilities in a nice way. But\nreality?\n\nImagine a finite computer, with some simulation running on it.\nIf the thing which is simulated is deterministic chaos, you will\nhave some sort of decoherence and you can prove related\nresults. Nonetheless, the whole thing remains deterministic.\nAnd predictable: If you run the program again with the same\ninput data you have the same results.\n\n> The main problem is that BM says that we cannot measure q(t). As long\n> as we keep this affirmation, I will say QM=BM.\n\nThe difference is not about the predictions, but about\nthe form of the theory. QM is not realistic, I repeat myself.\n\n> > > Why not taking (q+p)(t) as the\n> > > trajectory of new bohemian particle (observable Q+P)? Who is the more\n> > > "real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.\n\n> > This is not a problem of a particular theory. This is a question about\n> > preference for one theory in comparison with another. BM_q is one\n> > theory, BM_{p+q} another. BM_q is simpler, therefore preferred\n> > by Occam\'s razor.\n\n> No all these theories are the same. Only the interpretation of the\n> possible path q(t) with the allowed density probability may be\n> different.\n\nSo what? The thing which is different (the interpretation) is the thing\nI use to define my preferences, with the help of Occam\'s razor.\n\nIf you nonetheless insist on "why not use BM_{p+q}" - nice too.\nBecause BM_{p+q} also has an absolute time, the main conclusion\n(the preferred frame) remains valid.\n\n> In this aspect you are close to the wave QM where the q representation\n> is selected (i.e. use of psi(q,t) instead of the vector |psi> in\n> another basis). You say BM_q is simpler. However, may be, haven\'t you\n> tried to look to other representations for a given rho(q,t) and given\n> symmetries of the interactions?\n\nNot really. Feel free to try out.\n\n> It is what I am also trying to underline. I think you are selecting\n> BM_q as a candidate theory, not for simplicity but because it is well\n> described today (since 1952).\n\nFeel free to think so. Indeed, I do not really care. (The reason is that\nI don\'t believe BM is the final theory. It is a starting point. The\ndirection\nof research may be, for example, Nelsonian stochastics. It needs the\nform H=p^2 +V(q), does not work in other representations.)\n\n> > BTW, whatever observable we use here - note that time t is not an\n> > observable in the classical sense in Schroedinger theory (different\n> > from x). As BM_q, as BM_p, as BM_A handle time t as a\n> > preferred absolute time.\n\n> You can change easily change the form of\n> classical QM (implicit time representation) to get time handled in a\n> more symmetrical way as position q (i.e you may define a time\n> operator, etc .). In relativistic QM theory, time is handled more\n> symmetrically.\n\nCan you extend this to some sufficiently simple realistic\ninterpretation of QM?\n\nNote: Whatever we consider, once at least one realistic theory exists,\nI prefer it. I have relativistic BFT with preferred frame, following the\nclassical scheme, with wave functional and time as parameter.\n\n> Note that there is also a relativistic BM. However, the things become\n> worse as we have not the particle number conservation in relativistic\n> QM (requires the use of QFT instead and all the ugly things associated\n> with, like renormalisation).\n\nFor this we have BFT.\n\n> > Note that nobody has ever claimed that BM is the final theory\n> > of everything. BM is a theory which invites to think about\n> > modifications. For example, one direction is Nelsonian\n> > stochastics. Whatever.\n\n> I know. However, I really do not see what additional concrete\n> features BM brought.\n\nCertainty that there is nothing wrong with classical realism.\nAs a consequence, certainty that there is a preferred frame,\nbecause this is a consequence of classical realism and the\nviolation of Bell\'s inequality.\n\n> I do not know Nelsonian stochasitcs , so if you have a pointer to\n> paper, I would be very glad to see it.\n\nSorry, no. (Learned it from a source in German.)\n\n> However, I am afraid that this will be the usual second way of\n> representing QM: the use of a Langevin type interaction/field. You\n> just replace the psi(q,t) wave function by a Brownian type motion\n> noise.\n\nYep.\n\n> Well, we stay with the same problem: you cannot explain the\n> source of this Brownian noise better than the statistical results of\n> QM. We recover the BM_q problems (non locality, existence of a virtual\n> particle, etc ...): we cannot explain the source of randomness of QM\n> results nor the "reality of this noise source".\n\nNonlocality is a property of all realistic theories able to describe\nviolations of Bell\'s inequality. That\'s not a problem but a tool: We have\nbetter observational possibilities - we can detect causal connections\nin an indirect way, observing (A->B or B->A).\n\nIt is not the aim of stochastic theories to explain the source\nof randomness. One possibility is a classical randomness, of the\ntype deterministic chaos.\n\n> > No. In the domain of QM foundations I only argue that the violation\n> > of BI requires the acceptance of a preferred foliation. At least if\n> > we do not want to give up realism and/or causality.\n\n> Well all the results of QM tend to show that there is no preferred\n> representation.\n\nQM does not make claims about reality at all, thus, all it can say is\nthat we cannot measure it.\n\n> Now the usual interpretation of this correlation (the bell "realistic"\n> model) does not explain how we can get this degree of correlation: we\n> can remove this model (+its context) to explain QM statistical\n> results.\n\nThe "Bell realistic model" (if there is one) is BM. It explains\neverything except its axioms.\n\nIf you mean the whole class of theories which allows to prove\nBell-type theorems (the class I name "realistic theories"), then,\nthe class as a whole explains nothing. Only a single theory explains\nsomething.\n\n> Therefore, whatever is your answer about realism, my question remains:\n> what type of *new* models can you remove, with such affirmation?\n\nI don\'t understand this question. Tentative answer:\nI can simply remove all non-realistic theories. Including QM, as\nlong as I have a realistic replacement.\n\nThere is no necessity at all to consider non-realistic theories.\nOk, there was a short period of time, from definition of QM to\ndefinition of BM, when it was reasonable to hold BM. But now?\n\n> Relativistic causality defined in QFT is compatible with the EPR\n> results. Therefore, I can say QM preserves causality (in the sense of\n> the one defined/understood in QFT). Note that the causality you are\n> trying to define seems to be very close to the cluster decomposition\n> of QFT.\n\nI doubt. It is a stronger notion of causality, one which allows to\nprove Bell\'s inequality for causally disconnected events.\n(Bell\'s theorem)\n\n> > In the context of this discussion I defend BM_q as an example\n> > of a realistic causal theory with preferred frame. It proves\n> > that such theories exist, and, therefore, realism and causality\n> > are consistent with observation.\n\n> So your current postulate (proof) is realism= yes because there is at\n> least a BM_q theory explaining some results of QM. I assume in that\n> case that any theory postulating a virtual particle path is realist in\n> your definition.\n\nSimple test question: In a theory which is realistic and causal Bell\'s\ninequality holds for causally disconnected events. And there should\nnot be closed causal loops.\n\n> Ok, QM becomes realist if I attach to it an independent postulate: for\n> each quantum system, there exists a non measurable path associated to\n> the quantum particle (call it q(t)). QFT may also be amended with such\n> a postulate and every theory.\n>\n> Causality=yes. QM already says yes to its causality.\n\nIt is only a weak form of causality. Not the one which is natural for\na realistic theory. A causal connection is part of reality, not of\nour observations.\n\nIlja\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"seratend" <ser_monmail@yahoo.fr> schrieb
> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> wrote
> > > Therefore, I definitively prefer to concentrate on logically
> > > consistent models and thereafter interpretation (connection between
> > > the model results and experiment results) before speaking of what
> > > reality is (which I think is out of scope at this step).
> >
> > In this case, there is nothing to discuss. We agree that the
> > QM "shut up and calculate" is a logically consistent way to
> > obtain testable predictions which agree with observation.
> >
> Yes we have to discuss : ).
> The initial advantage of focussing first on the logical content of
> models than the interpretation question is first to see/detect their
> equivalence. As long as they give the same result (i.e experiments), I
> prefer to leave the "true" interpretation graal question to
> philosophers as long as it does not change the models nor the
> experimental results.

Fine. Now, the point of my argumentation is not that I prefer one
of the versions of BM in comparison to others. But I'm interested
in common properties of all (all reasonable in some sense)
interpretations. Now, the preferred frame is a common property
of all interpretations which are realistic and have a notion of
causality related with spacetime.

> I am thus searching for models that allow
> simpler logic (i.e. close to every day environment).

Yep. But the details of such a model don't really matter, as far as:

1.) We have at least one model close to "simple logic" (EPR realism,
no "quantum logic" but classical logic, causality), which is reached
with BM.
2.) We can prove general properties of all theories of this class
with "simple logic". Violation of Einstein causality is such a common
property, as proven by Bell. The only way to connect this with
causality is return to classical causality in a preferred foliation.

Once we have these two points, we can be quite sure that there
exists a preferred frame. And this is what I want. (As support
for my own ether theories.)

Thus, my main argument is that I have a prove using the "simple logic"
of realism, and BM shows that there is nothing wrong with this "simple
logic", at least it is not in contradiction with observation.

> Sorry for the "bohemian" in my different posts: my text editor does
> not like bohmian. It sometimes replaces it randomly. : (

;-))). As an anarchist, I have no problems with bohemians.

> > Realistic theories define a subclass of the class of consistent
theories.
> > QM "shut up and calculate" is a logically consistent but not realistic.
> >
> > You can view realism as a restriction of the mathematical form of the
> > theory. As well as, say, relativity defines a subclass of relativistic
> > theories which are also logically consistent but have an additional
> > requirement (some symmetry group).
> >
> yep. The problem remains on the restriction term somewhat fuzzy.

For the purpose of my argument, the following pre-definition is
sufficient: Realism is what is necessary to prove, combined
with Einstein causality, Bell's inequality.

Bell's inequality is a strong theorem. Every definition in agreement
with this pre-definition gives me what I need. You can work out
yourself the details of the definition, if my definition is too fuzzy for
you.

> > First, there is no measurement of a path in BM.

> If we remove the "reality" of the bohmian particle path, we just have
> QM results with a hypothetic particle following an unknown path. I
> accept BM without q(t) interpretation, there is no problem for me, I
> just do not see the utility of adding the q(t) interpretation (non
> measurable effects of an external information).

The use of unobservable but real things is that we can apply
classical logic to them. If the theory gives me x in {x1,x2,...xn}
but I cannot observe which it is, I nonetheless can prove
something about x. If A(x1) and A(x2) and .... A(xn) then
A(x).

The particular example I need is x in {A->B, B->A},
Einstein causality violated(A->B) and
Einstein causality violated(B->A) then
Einstein causality violated(x).

> >Second,
> > we see no difference only in quantum equilibrium. Once
> > quantum equilibrium holds in the initial state it holds always.
> > But I would guess that the initial value of the universe is
> > not quantum equilibrium but a well-defined initial value q_0.

> Sorry, but I do not know what is a quantum equilibrium.

A probabilistic state with some fixed (but arbitrary) wave function
\Psi(q) and the probability distribution \rho(q)=<\Psi(q)\Psi(q)>
for the position q.

> However, if
> you accept the mathematical formulation dq/dt=grad(S(q,t)) with the
> classical Bohmian Hamilton-jacobi equation and the probability
> conservation equation, you accept the statistical results of classical
> QM. If the initial quantum state is a particle with a well know
> position at a time to, the time after (as close to "to" as you want)
> you can find the particle everywhere in the universe (at least with
> the non-relativistic approximation).

No. If in the initial state the particle has a well-defined position q,
it has always a well-defined position q(t). But don't mingle here
a state with position q0 with a state with wave function \delta(q-q0).
Such degenerate states are not even allowed in BM.

> Recall that we always have:
>
> (QM) \psi(q,t) = \Sqrt[\rho(q,t)].\exp iS(q,t)] (BM)

No. This holds only in quantum equilibrium.

> > > Once an infinite
> > > precision q measurement occurs with a result qo, the bohemian velocity
> > > is unknown at this time as in QM.
> >
> > No. The velocity of the particle d_t q(t) is known if the position
> > q(t) is known and the wave function \psi(q,t) is known.

> If you measure with an infinite precision the position of a particle
> i.e. (qo,to), you have at to, v(q,to)=0, but you cannot know the
> bohmian velocity at time v(qo,to+) (infinite) nor its position q(t).
> This is a simple application of the equations of BM and QM.

I should have said that there are no infinite precision measurements
in BM (as well as QM). \delta(q-q0) is not in H.

> >> I prefer to consider Bohmian mechanics as one
> > > possible interpretation that is compatible with the orthodox
> > > interpretation of QM, with an additional cost q(t).
> >
> > Fine. I have no problem with this point of view. Essentially, this
> > is also Bell's who has explicitly named it realistic interpretation
> > (AFAIR).

> Yes, but he may have called it surrealistic or whatever else. This is
> not important, only the results are important: some simple
> mathematical models of the quantum probability are not compliant with
> it. You cannot use the word "realistic" itself to make reasonable
> logical deductions, It is what I want to underline.

Of course, the word "realistic" is not the point. The point is that we
can apply the math which is necessary to prove such things like
Bell's inequality for causally disconnected events.

> What some people are first searching is a kind of "deterministic"
> model that complies with the current formulation of QM, and the
> "realistic interpretation" always displace the problem: we still have
> no simple (to be defined ; ) non statistical mathematical model that
> complies with all the statistical results of QM (relatively to my
> personal knowledge : ).
> BM does not solve this issue. It simply introduces q(t) but it fails
> to remove the statistical behaviour (it is formally equivalent to QM).
> In this aspect, BM is a useful model to explain why classical
> explanations fail to explain QM statistical behaviour.

I don't understand this point. As long as statistical behaviour is
of the same type as in classical thermodynamics, uncertainty
caused by uncertainty of the initial values and classical chaos,
I see no reason to remove it.

> > > Explanation the previous statement: U(t,to) is the time evolution of
> > > the initial state |\psi(to)>. We have \psi(q,to)=<q|\psi(to)> for the
> > > bohemian particle.

> > Sorry, the position of the bohmian particle is not a wave function,
> > not a bra vector <q| or a \delta-function \delta(q-q0). We have
> > \psi(q,to)=<q|\psi(to)>, but this equation has nothing to do with
> > the bohmian particle.

> The position q(t) belongs to the spectrum of the observable Q: it is
> an eigen value of Q. The QM-BM correspondence formula imposes this
> aspect:
>
> (QM) \psi(q,t) = \Sqrt[\rho(q,t)].\exp iS(q,t)] (BM)
> and v(q,t)=grad(S)/m = "dq/dt".

The correct formulation:
\psi(q,t) = R(q,t).\exp iS(q,t)
Schroedinger equation for \psi,v(q,t)=grad(S)/m = dq/dt.
Complete configuration space: R(q), S(q), q.

Quantum equilibrium: \rho(q) = R^2(q).

The position of q belongs to IR.

> If you do not accept this, you must accept it formally: the spectrum
> of Q is the |R set. If you are more aware of Hilbert spaces, we are
> just speaking about a representation:
>
> We have from BM: \rho(q,t) and v(q,t)

This is not the complete BM state. It is only sufficient to define
the state in quantum equilibrium.

> [considerations about P(t)]
> This was an argument about the utility to compute q(t) if
> we are not able to "see" it.

I see no utility in computing q(t), except for illustration. Or for
checking various consistency results.

> > Ok. I don't claim "shut up and calculate" is meaningless. It gives
> > testable predictions, in a consistent way. It does not give
explanations.
> >
> > > Note that Bohmian meachanics has the same interpretation problem if
> > > you do want to make a simple "shut up and calculate".
> >
> > No.
>
> QM postulates say exactly where there are no explanations (i.e. what
> it does not try to explain). "The shut up and calculate" comes from
> the measurement postulate and the Born rules.

In this sense "shut up and calculate" is not an explanation. BM is
different. It explains the probabilities by evolution equations and
initial values, thus, in the same way as classical mechanics.

> It is well known and
> there is a lot of work on how QM could explain the measurement
> results (i.e. without the measurement postulate).
> One tentative of such an explanation is the decoherence program
> (Zurek, ...). It tries to explain the macroscopic results of QM
> experiments just using all the QM postulates except the measurement
> postulate (explanation/interpretation of the state projection from the
> other postulates of QM).
> So if it is the shut up and calculate of the QM measurement that is
> not acceptable for you, may be you should look at the decoherence
> program (very instructive, I think).

I don't believe that decoherence works in this way.
Decoherence explains human inabilities in a nice way. But
reality?

Imagine a finite computer, with some simulation running on it.
If the thing which is simulated is deterministic chaos, you will
have some sort of decoherence and you can prove related
results. Nonetheless, the whole thing remains deterministic.
And predictable: If you run the program again with the same
input data you have the same results.

> The main problem is that BM says that we cannot measure q(t). As long
> as we keep this affirmation, I will say QM=BM.

The difference is not about the predictions, but about
the form of the theory. QM is not realistic, I repeat myself.

> > > Why not taking (q+p)(t) as the
> > > trajectory of new bohemian particle (observable Q+P)? Who is the more
> > > "real" q(t) or (q+p)(t)? as the 2 motions cannot be known precisely.

> > This is not a problem of a particular theory. This is a question about
> > preference for one theory in comparison with another. BM_q is one
> > theory, BM_{p+q} another. BM_q is simpler, therefore preferred
> > by Occam's razor.

> No all these theories are the same. Only the interpretation of the
> possible path q(t) with the allowed density probability may be
> different.

So what? The thing which is different (the interpretation) is the thing
I use to define my preferences, with the help of Occam's razor.

If you nonetheless insist on "why not use BM_{p+q}" - nice too.
Because BM_{p+q} also has an absolute time, the main conclusion
(the preferred frame) remains valid.

> In this aspect you are close to the wave QM where the q representation
> is selected (i.e. use of \psi(q,t) instead of the vector |\psi> in
> another basis). You say BM_q is simpler. However, may be, haven't you
> tried to look to other representations for a given \rho(q,t) and given
> symmetries of the interactions?

Not really. Feel free to try out.

> It is what I am also trying to underline. I think you are selecting
> BM_q as a candidate theory, not for simplicity but because it is well
> described today (since 1952).

Feel free to think so. Indeed, I do not really care. (The reason is that
I don't believe BM is the final theory. It is a starting point. The
direction
of research may be, for example, Nelsonian stochastics. It needs the
form H=p^2 +V(q), does not work in other representations.)

> > BTW, whatever observable we use here - note that time t is not an
> > observable in the classical sense in Schroedinger theory (different
> > from x). As BM_q, as BM_p, as BM_A handle time t as a
> > preferred absolute time.

> You can change easily change the form of
> classical QM (implicit time representation) to get time handled in a
> more symmetrical way as position q (i.e you may define a time
> operator, etc .). In relativistic QM theory, time is handled more
> symmetrically.

Can you extend this to some sufficiently simple realistic
interpretation of QM?

Note: Whatever we consider, once at least one realistic theory exists,
I prefer it. I have relativistic BFT with preferred frame, following the
classical scheme, with wave functional and time as parameter.

> Note that there is also a relativistic BM. However, the things become
> worse as we have not the particle number conservation in relativistic
> QM (requires the use of QFT instead and all the ugly things associated
> with, like renormalisation).

For this we have BFT.

> > Note that nobody has ever claimed that BM is the final theory
> > of everything. BM is a theory which invites to think about
> > modifications. For example, one direction is Nelsonian
> > stochastics. Whatever.

> I know. However, I really do not see what additional concrete
> features BM brought.

Certainty that there is nothing wrong with classical realism.
As a consequence, certainty that there is a preferred frame,
because this is a consequence of classical realism and the
violation of Bell's inequality.

> I do not know Nelsonian stochasitcs , so if you have a pointer to
> paper, I would be very glad to see it.

Sorry, no. (Learned it from a source in German.)

> However, I am afraid that this will be the usual second way of
> representing QM: the use of a Langevin type interaction/field. You
> just replace the \psi(q,t) wave function by a Brownian type motion
> noise.

Yep.

> Well, we stay with the same problem: you cannot explain the
> source of this Brownian noise better than the statistical results of
> QM. We recover the BM_q problems (non locality, existence of a virtual
> particle, etc ...): we cannot explain the source of randomness of QM
> results nor the "reality of this noise source".

Nonlocality is a property of all realistic theories able to describe
violations of Bell's inequality. That's not a problem but a tool: We have
better observational possibilities - we can detect causal connections
in an indirect way, observing (A->B or B->A).

It is not the aim of stochastic theories to explain the source
of randomness. One possibility is a classical randomness, of the
type deterministic chaos.

> > No. In the domain of QM foundations I only argue that the violation
> > of BI requires the acceptance of a preferred foliation. At least if
> > we do not want to give up realism and/or causality.

> Well all the results of QM tend to show that there is no preferred
> representation.

QM does not make claims about reality at all, thus, all it can say is
that we cannot measure it.

> Now the usual interpretation of this correlation (the bell "realistic"
> model) does not explain how we can get this degree of correlation: we
> can remove this model (+its context) to explain QM statistical
> results.

The "Bell realistic model" (if there is one) is BM. It explains
everything except its axioms.

If you mean the whole class of theories which allows to prove
Bell-type theorems (the class I name "realistic theories"), then,
the class as a whole explains nothing. Only a single theory explains
something.

> Therefore, whatever is your answer about realism, my question remains:
> what type of *new* models can you remove, with such affirmation?

I don't understand this question. Tentative answer:
I can simply remove all non-realistic theories. Including QM, as
long as I have a realistic replacement.

There is no necessity at all to consider non-realistic theories.
Ok, there was a short period of time, from definition of QM to
definition of BM, when it was reasonable to hold BM. But now?

> Relativistic causality defined in QFT is compatible with the EPR
> results. Therefore, I can say QM preserves causality (in the sense of
> the one defined/understood in QFT). Note that the causality you are
> trying to define seems to be very close to the cluster decomposition
> of QFT.

I doubt. It is a stronger notion of causality, one which allows to
prove Bell's inequality for causally disconnected events.
(Bell's theorem)

> > In the context of this discussion I defend BM_q as an example
> > of a realistic causal theory with preferred frame. It proves
> > that such theories exist, and, therefore, realism and causality
> > are consistent with observation.

> So your current postulate (proof) is realism= yes because there is at
> least a BM_q theory explaining some results of QM. I assume in that
> case that any theory postulating a virtual particle path is realist in
> your definition.

Simple test question: In a theory which is realistic and causal Bell's
inequality holds for causally disconnected events. And there should
not be closed causal loops.

> Ok, QM becomes realist if I attach to it an independent postulate: for
> each quantum system, there exists a non measurable path associated to
> the quantum particle (call it q(t)). QFT may also be amended with such
> a postulate and every theory.
>
> Causality=yes. QM already says yes to its causality.

It is only a weak form of causality. Not the one which is natural for
a realistic theory. A causal connection is part of reality, not of
our observations.

Ilja

rof@maths.tcd.ie

Nov25-04, 03:36 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n\n><rof@maths.tcd.ie> schrieb\n>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:\n>> ><rof@maths.tcd.ie> schrieb\n\n>> We can make it make a little more sense by noticing that the\n>> space over which the wavefunction is defined is the same as\n>> the space over which probability distributions (of generic\n>> states of affairs of the system, or configurations) are\n>> naturally defined. Information that we have might have about\n>> the configuration of the system at a given time would naturally\n>> be expressed as a real function on this space, and a description\n>> which included the instantaneous rate of change of this\n>> probability distribution would naturally be expressed as\n>> two real functions (or one complex function) on this space.\n\n>This is the base of the QM interpretation that the wave function\n>does not describe some state of reality but our knowledge about\n>reality. But I think there is no necessity for such a non-realistic\n>interpretation.\n\n>I feel uncomfortable with this space too. But I have another\n>idea: There is simply our universe, q in Q, and something else,\n>unknown. But whatever this unknown thing is, once it interacts\n>with our universe, its action somehow depends on the state\n>of our universe and, therefore, it may be described by some\n>function on all possible states of the universe too.\n\nSo your position might be summarised as: Whatever is actually\ngoing on is unknowable to us, but we can deal with it by\nconsidering the mathematical regularities in the effects\nwhich it induces on that thing which we can actually observe,\nwhich in this case is the configuration q.\n\nI would say that\'s a thoroughly sensible position. Much more\nsensible than declaring that we should try to guess\nwhat it might be, in the hope that (unknown to us) our\nguess might be right.\n\n>But, in general, once you have managed to make the\n>quantum part mathematically consistent, to add the\n>guiding equation is not really problematic.\n\n>All you need is a probability current operator J.\n>Then you have\n\n>d/dt Q = <Psi J Psi>/<Psi Psi>\n\nThe configuration space for a system with a variable number of\nparticles has disconnected components, one for each number\nof particles. The guiding equation above can only prescribe\nmotion on a (path) connected configuration space.\n\n>> >> A may think it\'s on ontological theory, but A is confusing himself.\n>> >> Imagine there is a digital machine, which takes 1\'s and 0\'s as\n>> >> input. You feed in some sequence of 1\'s and 0\'s, generated by\n>> >> some algorithm. The machine proceeds to construct a Markov model\n>> >> to predict the sequence. It is successful to some degree - most\n>> >> of the time it correctly predicts the next input. Has it discovered\n>> >> reality? Is its Markov model an ontological theory?\n>>\n>> >Once it predicts successfully, there is some probability that the\n>> >algorithm has been guessed correctly. In other words, that\n>> >the algorithm used to create the input sequence and the\n>> >algorithm used by the machine to predict the input sequence are\n>> >in some sense equivalent.\n>>\n>> But this doesn\'t answer either of the above questions.\n\n>And there is no necessity to answer the first. Of course,\n>if we have a realistic theory of everything, we also cannot\n>be sure that the theory is true. And we never will be sure.\n>You want too much certainty.\n\nI don\'t; I just want clarity. The position of a physicist\npredicting an experimental result is very similar to the\nposition of a machine predicting its next input. The two\ncomplete more or less the same job, but with different\ndata (although in both cases the data are in the form of digital\ninputs).\n\nYou want to restrict the set of algorithms from which the\nphysicist can choose, by demanding that they be "realistic".\nPresumably, you would make a similar demand of the machine,\nor else you have some reason to suppose that the situation\nbetween the two differs in some important respect.\n\nMore important is the fact that, whatever you claim about\nthe physicist\'s accomplishments should not be greater than\nyour claim about the machine\'s accomplishments when it\ncompletes the same task. A reasonable thing for the machine\nto declare might be "I have detected the regularities in\nmy observed data, and can use them to predict future data."\n\n>If I was not clear enough about the second: A Marcov model is\n>an ontological theory (or can be formulated in such a way).\n\nSo, "1 comes after 0 and 0 comes after 1", is an ontological\ntheory, or can be formulated as one?\n\nThe speculation that the machine\'s construction "might be right", is\nfine, insofar as we realise that what that means is that\nit might actually be the case that the rule it found accurately\ndescribes its input. That, however, is a statement about the\nregularities in its input and not a statement about what would\nremain in the absence of any machine or any input.\n\n>> The grand project of ontology is not to guess an algorithm which\n>> predicts future observations (which is, in fact, the only thing\n>> we can do), but to determine what it is that really exists.\n\n>This project has been given up long time ago. At least if you\n>mean "determine" as giving certainty. The current (not that grand)\n>project of ontology is to guess what it is that really is, and to\n>test this guess using its testable consequences.\n\nThe problem here is that the goal is not achievable with the\nmethods used. What is wanted is something with property X\n(a theory which tells us what really exists). What can be\nchecked is property Y (the predictions of measurement results\nare correct). X implies Y (suppose). The procedure adopted\nis: take a candidate, check for property Y, and then celebrate,\nproclaiming "Maybe it also has property X!"\n\nThis is analogous to looking for a peanut by taking an\nobject at random, checking if it\'s a small object (peanut\nimplies small), and then, if it is, celebrating, because\nmaybe it is in fact a peanut.\n\nIn this case, certainty isn\'t the issue. The ontological\nproject seeks to remove the observer from the description\nof "what\'s going on", but what\'s going on is that there\nis an observer, making observations.\n\n>> If a machine develops an algorithm to correctly predict whether\n>> the next input will be a 1 or a 0, the most one can say about\n>> it is that that is what it has achieved. To say that it has\n>> discovered the ontological basis of reality, one has to radically\n>> change what one means by "discovered the ontological basis of\n>> reality."\n\n>Its you who wants to say this. I\'m quite satisfied by giving\n>a reasonable guess which allows to derive predictions\n>which survive tests.\n\nNo you\'re not; you want the algorithm to satisfy additional\ncriteria apart from merely giving correct predictions,\nbecause you claim to know about the properties of that\nsomething which would remain in the absence of any\nobservations, for example, that causality would still apply.\n\nContrast this with your position above, where you said,\nvery reasonably, that one can deal with the "something\nelse", which is not the configuration of the system, by\nmodelling it in terms merely of its action on the\nconfiguration, leading naturally to a representation\nof it in terms of a function on configuration space.\nThis is not a prescription for demanding a list of "real"\nobjects, to which causality applies, separate from the system\nand its configuration.\n\n>> Alan Sokal, in his "Defense of a Modest Scientific Realism",\n>> examines these issues a little bit, but, since he too is\n>> wedded to realism, doesn\'t quite reach the correct answer,\n>> but instead ends with:\n>> "Since no existing theory purports to be a final theory, there\n>> is no reason to consider it as literally true or to worry too\n>> much about whether the entities it postulates `really exist\'.\n>> Or rather, when worrying about whether the unobservable entities\n>> of a given theory `really exist\', it is important to distinguish\n>> existence _as a fundamental constituent of the universe_ from\n>> existence _in some coarse-grained sense_. It is a reasonable guess\n>> that _none_ of the theoretical entities in our present-day\n>> theories are truly fundamental, and that _all_ of the theoretical\n>> entities in our present-day well-confirmed theories will maintain\n>> some status as derived entities in future theories." (His italics)\n\n>Quite reasonable.\n\nHe\'s a reasonable man.\n\n>> In the end, he opts for saying that things exist "in some coarse-grained\n>> sense", which means that he wants to ascribe to them some property\n>> which he feels is related to existence or reality, but which is so\n>> vague that he cannot tell us what it is, and, lacking any concrete\n>> or even specific thing to say, must instead appeal to our sympathy\n>> to his realist position in order to convince us to accept what he\n>> is saying as a valid defense of realism.\n\n>Why do you think this is too vague? We have lots of examples\n>where one theory appears as a limit of some more fundamental\n>theory. And in all these cases the objects which "exist" in the\n>approximation are somehow constructed from the objects\n>which "exist" in the more fundamental theory. In every such case\n>we can consider this connection in detail, as detailed as we like.\n>Thus, the "vague" notion is not an appeal to sympathy, but\n>an appeal to the large body of experience with such derived,\n>non-fundamental objects.\n\nThere is one thing which is vague (his "existence in some coarse-grained\nsense), and a different thing which is precise (an object which is\nfundamental in one theory being derived in another). I said that the\nvague thing was vague, and you are now pointing at the precise thing\nand saying that it isn\'t vague.\n\nWhat Sokal is claiming is that the objects which appear as\nderived entities (such as Newtonian space) should not be regarded\nas merely "derived entities", but should be considered to actually\nexist "in some coarse-grained sense." He wants to ascribe to them\nsome property other than, and in addition to, their merely being\nderived in some other theory. This property is vague, and the reader\ncan only allow himself to believe that Sokal has said something\ncorrect and sensible if the reader himself is sympathetic to the\nrealist position that Sokal expresses, and is willing to forgive\nsome sloppiness in Sokal\'s phrasing because he feels that he\nagress with it.\n\n>> It is very important in these matters not to decide that we like\n>> realism and then accept as valid an otherwise insufficient\n>> argument in favour of it.\n\n>Ok. So what?\n\nSo be careful.\n\n>> Thus, if we find that the arguments with which\n>> we convince ourselves that realism is necessary amount to\n>> "In my opinion, realism should be considered a part of logic,"\n>> then we must go in search of the reason why that opinion\n>> was adopted, and reject it unless there is a rigorous basis\n>> for it.\n\n>I have adopted it because I have found the arguments\n>in favour of this position convincing.\n\nWhat are those arguments?\n\n>I do not believe in such things as "rigorous basis" in\n>general and in the most fundamental philosophy especially.\n>Logic may not be justified on a rigorous basis. Because\n>any possible "rigorous basis" is, itself, based on logic.\n>Thus, the "rigorous basis" is only circular reasoning.\n\nIndeed. Logic has this characteristic, but realism, by which\nI mean the assertion that the role of the observer should\nbe ignored in the development of physical theories and only\nadded on at the end as an afterthought, does not.\n\n>> The equivalence here is again, "in some sense", and has the particular\n>> difficulty that one algorithm (the one used by a human) has, as its\n>> output, patterns of sensations (as a prediction of future input).\n>> I doubt that an algorithm which has sensations as output is one\n>> that most realists would accept as an example of something which\n>> could qualify as a fundamental constituent of reality, since the\n>> things with which it deals (sensations) are purely mental.\n\n>Here several things are mingled. The algorithm which is assumed\n>to be used by the fabric of reality is one which computes\n>X_n+1 = f(X_n).\n\nYou have mangled the X_n which I introduced. As I use them, they\nrefer to the object which the observer constructs as his own mental\nrepresentation of a putative state of affairs outside him, which\nhe demands is a Markov sequence, and further breaks down into\nmany miniature, localised states of affairs, in so far as it is\nuseful for him to do so in the process of surveying the regularities\nin his observations. You have taken it and treated it as though\nit were an element of an ontological theory which attempted to\ndescribe something which has nothing to do with observation.\n\n>> >> Whether or not a particle even exists can depend on the motion of\n>> >> the observer (cf. the Unruh effect), so it seems strange to put that\n>> >> sticky reality label onto the particles and their positions in space.\n>>\n>> >I don\'t.\n>>\n>> But Bohmian mechanics does, and that is the context in which I made\n>> the remark above.\n\n>Bohmian theories supposed to handle the Unruh effect don\'t.\n\nYou mean Bohmian field theories. This is true, but the original\nmotivation for Bohmian mechanics was to insist that there are\nindeed particles, with positions, rather than ethereal waves\nwhich collapse. To make the transition to a field theory one\nhas to abandon that, effectively recognizing that the original\nontology was wrong.\n\nWhat survived from the original Bohmian theory was the quantum\nformalism, namely that part of the theory which characterized\nregularities in observed phenomena, and provided a prescription for\nmaking future predictions. At this point one should realise that\nthe insistence on introducing the particle positions in the first\nplace, and insisting that they were ontologically real, was misguided.\n\n>> In this case,\n>> are you asserting that "reality" is something which can be\n>> defined, or that its definition can change from theory to theory?\n\n>The definition of reality is part of the realistic theory. It\n>changes from theory to theory.\n\n>In NT reality consists of point particles attracted by forces,\n>in GR it consists of a manifold with a metric and some\n>matter fields on it, in field theory of fields\n>(psi(x),A_i(x),g_ij(x)), in Bohmian theories of\n>some q in Q and some Psi: Q-->C, where Q is\n>the configuration space which is very different in\n>different Bohmian theories.\n\nSo you don\'t actually mean "define reality", but rather, "list\nsome objects which are to be considered to be real."\n\n>> The real difficulty here is (as I mentioned before) the manifold\n>> use of the word "real".\n\n>Not that difficult IMHO.\n> ...\n\n>Reality\n>I use for all this stuff outside, which is not in our minds.\n\nWhat, then, does actual pain have which imagined pain lacks?\n\n>> To add further to this confusion is the additional confusion caused\n>> by the inappropriate use of the notion of "outside." The relations\n>> of outside and inside are properly ascribed only to objects which\n>> are in space.\n\n>I hope my use of "outside" does not cause confusion.\n>It means "not in my mind or my soul". My mind,\n>(at least the material part of it) is sufficiently located in\n>space. If there is something non-material connected with\n>consciousness and so on (soul), it is not part of "outside".\n\nWell, you have basically defined "outside" as "not inside", but\nperhaps there might be a chance for some progress here. Firstly,\nrealise that in order to consider something to be "inside" something\nelse, both things need to have the property of location, while the\nsecond also needs to have the property of extension.\n\nWhat you refer to as "the material part" of your mind is, presumably,\nyour brain. It is true that this is localised in space, and is inside\nyour head. If, when you say that you are studying things outside of\nyour mind, all that you mean is that you are studying things outside\nof your head, then I would not disagree, except to say that you\nshould use the word "brain" or "head" instead of "mind", which\nalready has a different meaning.\n\nRegarding this "outside of the mind" business, consider any\nexisting physical object (for example, a computer screen). Notice\nthat there are colours on it, and that they are in specific\nlocations. A particular coloured shape (for example, a letter)\nis inside a certain region, which contains many colours. This\nis a legitimate use of the relation called "inside", since the\nshape has a location, and the region has both location and\nextension, and the coloured shape is, in fact, inside the\nregion. Furthermore, the region, and the coloured shape, are\noutside of your head. However, neither the coloured shape\nnor the region (of the image which is presented to your visual\nfield) would exist without a perceiving mind, since colours are\nsensations and sensations are experiences. It is not correct to say\nthat they are inside the mind, since the mind is not a place, but\nthey exist only in virtue of your apprehension of them.\n\n>> Now, taken together, these two confusions lead to the realist\n>> position, where the physicist believes that he is examining\n>> that which "really" exists, "outside" of his mind. In fact,\n>> what he is doing is examining things which are actual (in\n>> my sense of the term) and outside of his head. Perhaps I\n>> should emphasise that this is worth thinking about; many\n>> lives have been wasted chasing ghosts because of this\n>> misunderstanding.\n\n>I see not much advantage in your specification. I propose\n>theories about actual things outside of my head? Fine.\n>No problem. But also no advantage.\n\nBy stating things in this way, the implicit assertion that the\nobjects of investigation would continue to exist and have the\nproperties that they do in the absence of your perception of them\nis dropped.\n\nThis is the actual solution to Sokal\'s problem. He wants to say\nthat, for example, Newtonian absolute space and time exist,\nalthough, since he has learned relativity and believes it to\nmore accurately describe what exists, he can\'t. Rather than\ninventing some vague watered-down type of existence, he should\nhave realised, when relativity (for him) deprived Newtonian space\nand time of their ontological existence, that it meant that\nNewtonian space and time enjoy no existence at all, outside\nof the mind of the observer.\n\nWhen he wants to say that they approximately exist, what he is\ntrying to capture is the idea that they approximately characterize\ncertain aspects of observed phenomena, and are therefore useful\nmental inventions for the observer to employ when surveying\nthe data he has observed.\n\nThe actual situation is that there is an observer, and Newtonian\nspace and various other theories arise as he tries to characterize\nobserved phenomena. The realist, however, pays so much attention\nto the theory and what it describes that he forgets this, thinking\nthat the theory is "What\'s really going on", and that the observer\nis merely some object inside the theory, and not a very interesting\none at that. Having forgotten about the fact that there is an\nobserver who constructs the theories to characterize his observations,\nSokal is left trying to assign some kind of existence to the\ntheoretical objects themselves, while at the same time acknowledging\nthat they do not, in fact, actually exist, since they have been\nsuperceded by different theoretical objects.\n\n>There are more serious faults in modern science than this.\n>Scientists should be independent. At least as independent\n>as judges. Even more. Because a judge which makes obvious\n>nonsense harms other people, a scientist which makes\n>obvious nonsense doesn\'t.\n\nIndeed, but pedagogy, which you don\'t care to argue about, is\nextraordinarily important. Most physicists never think about it,\nbut they do get their interpretation of quantum mechanics from their\nteachers, and if more people had a teacher who presented the actual\nfacts about Bohmian mechanics, rather than saying, for example,\nthat hidden variable theories were ruled out by Bell, then Bohmian\nmechanics would be more popular.\n\nThe good news for you is that most physicists would agree with you,\nif they took the time to examine the issues more carefully. The bad\nnews is that that won\'t happen.\n\n>But current scientific organization is the reverse. If you\n>have a grant for a few years, you have to care about your\n>future, the next grant. The decisions are made by\n>beaurocrats and other scientists (these two often unified\n>in a single person). The predictable result is conformity.\n\n>Another pressure into conformity is "publish or perish".\n>Publishing and being cited is easier in a large community\n>with lots of journals to publish and lots of readers.\n\n>The prediction agrees with observation. In the highly\n>speculative domain of modern fundamental physics,\n>where free science would lead to lots of very different\n>directions (because they are not guided by experiments,\n>purely speculative, therefore its extremely hard to reject\n>a research proposal as false) we observe de facto only\n>two research directions - strings and LQG.\n\nI vaguely remember a letter from Heisenberg to Bohr where he laments\nabout the widespread belief that if a large number of intelligent\npeople are working on a theory then it must be true.\n\n>> I should add that it is not the case that the analytic method has\n>> been exhausted due to a lack of performable experiments whose\n>> results we do not know in advance. The confusion about realism\n>> and the many disagreements about quantum mechanics (among those\n>> who actually do think about it) are evidence enough that there\n>> is an opportunity, for one who is prepared to suspend judgment\n>> and avoid opinion, to discern something previously overlooked.\n\n>The second sentence does not prove the first. The last two\n>points are discussions not about the results of possible\n>experiments.\n\nRight; there is a lack of performable experiments whose results\nwe do not know in advance. However, this does not mean that\nthe analytic method of carefully considering the data that\nwe already have has been exhausted. There are many disagreements\nabout quantum mechanics, for example, which demonstrates that\nnobody has yet presented a clear and indisputable analysis of the\navailable data which leads to a particular point of view.\n\nInstead, physicists have adopted the positions that they have based\non opinions, which are different from one physicist to another,\nhence the disagreements. To understand what\'s happening here, one\nhas to survey the various positions that people have adopted and\nunderstand how they got to that position. The aim is to identify\nthe point at which opinion crept in, either the opinion of the\nindividual physicist, or one which was given to her by those around\nher.\n\nFor example, "The wavefunction is a complete description,"\ncoupled to realism in the form "The observer is just another\nobject in the theory, so observation is just like any other\ninteraction," leads to the many-worlds interpretation. The\nrealisation that the first assertion is merely an opinion, together\nwith unquestioning adherence to the second leads to a Bohmian\ninterpretation.\n\n>> That is, I perceive a room, with walls\n>> here and there, and you perceive a different room (presumably),\n>> and there is no simple matching of what I perceive to what you\n>> perceive.\n\n>This is about the M_n. My X_n contains also complete\n>information about the inside of your room. I don\'t perceive\n>the X_n. But the laws of physics, the rule which allows\n>to compute X_n+1=f(X_n), need the whole information,\n>my room as well as your room.\n\nAs I mentioned, you\'ve hijacked entities which don\'t belong\nin an ontological theory and are treating them as though they\ndo.\n\nR.\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:

><rof@maths.tcd.ie> schrieb
>> "Ilja Schmelzer" <Ilja.Schmelzer@FernUni-Hagen.de> writes:
>> ><rof@maths.tcd.ie> schrieb

>> We can make it make a little more sense by noticing that the
>> space over which the wavefunction is defined is the same as
>> the space over which probability distributions (of generic
>> states of affairs of the system, or configurations) are
>> naturally defined. Information that we have might have about
>> the configuration of the system at a given time would naturally
>> be expressed as a real function on this space, and a description
>> which included the instantaneous rate of change of this
>> probability distribution would naturally be expressed as
>> two real functions (or one complex function) on this space.

>This is the base of the QM interpretation that the wave function
>does not describe some state of reality but our knowledge about
>reality. But I think there is no necessity for such a non-realistic
>interpretation.

>I feel uncomfortable with this space too. But I have another
>idea: There is simply our universe, q in Q, and something else,
>unknown. But whatever this unknown thing is, once it interacts
>with our universe, its action somehow depends on the state
>of our universe and, therefore, it may be described by some
>function on all possible states of the universe too.

So your position might be summarised as: Whatever is actually
going on is unknowable to us, but we can deal with it by
considering the mathematical regularities in the effects
which it induces on that thing which we can actually observe,
which in this case is the configuration q.

I would say that's a thoroughly sensible position. Much more
sensible than declaring that we should try to guess
what it might be, in the hope that (unknown to us) our
guess might be right.

>But, in general, once you have managed to make the
>quantum part mathematically consistent, to add the
>guiding equation is not really problematic.

>All you need is a probability current operator J.
>Then you have

>d/dt Q = <\Psi J \Psi>/<\Psi \Psi>

The configuration space for a system with a variable number of
particles has disconnected components, one for each number
of particles. The guiding equation above can only prescribe
motion on a (path) connected configuration space.

>> >> A may think it's on ontological theory, but A is confusing himself.
>> >> Imagine there is a digital machine, which takes 1's and 0's as
>> >> input. You feed in some sequence of 1's and 0's, generated by
>> >> some algorithm. The machine proceeds to construct a Markov model
>> >> to predict the sequence. It is successful to some degree - most
>> >> of the time it correctly predicts the next input. Has it discovered
>> >> reality? Is its Markov model an ontological theory?
>>
>> >Once it predicts successfully, there is some probability that the
>> >algorithm has been guessed correctly. In other words, that
>> >the algorithm used to create the input sequence and the
>> >algorithm used by the machine to predict the input sequence are
>> >in some sense equivalent.
>>
>> But this doesn't answer either of the above questions.

>And there is no necessity to answer the first. Of course,
>if we have a realistic theory of everything, we also cannot
>be sure that the theory is true. And we never will be sure.
>You want too much certainty.

I don't; I just want clarity. The position of a physicist
predicting an experimental result is very similar to the
position of a machine predicting its next input. The two
complete more or less the same job, but with different
data (although in both cases the data are in the form of digital
inputs).

You want to restrict the set of algorithms from which the
physicist can choose, by demanding that they be "realistic".
Presumably, you would make a similar demand of the machine,
or else you have some reason to suppose that the situation
between the two differs in some important respect.

More important is the fact that, whatever you claim about
the physicist's accomplishments should not be greater than
your claim about the machine's accomplishments when it
completes the same task. A reasonable thing for the machine
to declare might be "I have detected the regularities in
my observed data, and can use them to predict future data."

>If I was not clear enough about the second: A Marcov model is
>an ontological theory (or can be formulated in such a way).

So, "1 comes after and comes after 1", is an ontological
theory, or can be formulated as one?

The speculation that the machine's construction "might be right", is
fine, insofar as we realise that what that means is that
it might actually be the case that the rule it found accurately
describes its input. That, however, is a statement about the
regularities in its input and not a statement about what would
remain in the absence of any machine or any input.

>> The grand project of ontology is not to guess an algorithm which
>> predicts future observations (which is, in fact, the only thing
>> we can do), but to determine what it is that really exists.

>This project has been given up long time ago. At least if you
>mean "determine" as giving certainty. The current (not that grand)
>project of ontology is to guess what it is that really is, and to
>test this guess using its testable consequences.

The problem here is that the goal is not achievable with the
methods used. What is wanted is something with property X
(a theory which tells us what really exists). What can be
checked is property Y (the predictions of measurement results
are correct). X implies Y (suppose). The procedure adopted
is: take a candidate, check for property Y, and then celebrate,
proclaiming "Maybe it also has property X!"

This is analogous to looking for a peanut by taking an
object at random, checking if it's a small object (peanut
implies small), and then, if it is, celebrating, because
maybe it is in fact a peanut.

In this case, certainty isn't the issue. The ontological
project seeks to remove the observer from the description
of "what's going on", but what's going on is that there
is an observer, making observations.

>> If a machine develops an algorithm to correctly predict whether
>> the next input will be a 1 or a 0, the most one can say about
>> it is that that is what it has achieved. To say that it has
>> discovered the ontological basis of reality, one has to radically
>> change what one means by "discovered the ontological basis of
>> reality."

>Its you who wants to say this. I'm quite satisfied by giving
>a reasonable guess which allows to derive predictions
>which survive tests.

No you're not; you want the algorithm to satisfy additional
criteria apart from merely giving correct predictions,
because you claim to know about the properties of that
something which would remain in the absence of any
observations, for example, that causality would still apply.

Contrast this with your position above, where you said,
very reasonably, that one can deal with the "something
else", which is not the configuration of the system, by
modelling it in terms merely of its action on the
configuration, leading naturally to a representation
of it in terms of a function on configuration space.
This is not a prescription for demanding a list of "real"
objects, to which causality applies, separate from the system
and its configuration.

>> Alan Sokal, in his "Defense of a Modest Scientific Realism",
>> examines these issues a little bit, but, since he too is
>> wedded to realism, doesn't quite reach the correct answer,
>> but instead ends with:
>> "Since no existing theory purports to be a final theory, there
>> is no reason to consider it as literally true or to worry too
>> much about whether the entities it postulates `really exist'.
>> Or rather, when worrying about whether the unobservable entities
>> of a given theory `really exist', it is important to distinguish
>> existence _as a fundamental constituent of the universe_ from
>> existence _in some coarse-grained sense_. It is a reasonable guess
>> that _none_ of the theoretical entities in our present-day
>> theories are truly fundamental, and that _all_ of the theoretical
>> entities in our present-day well-confirmed theories will maintain
>> some status as derived entities in future theories." (His italics)

>Quite reasonable.

He's a reasonable man.

>> In the end, he opts for saying that things exist "in some coarse-grained
>> sense", which means that he wants to ascribe to them some property
>> which he feels is related to existence or reality, but which is so
>> vague that he cannot tell us what it is, and, lacking any concrete
>> or even specific thing to say, must instead appeal to our sympathy
>> to his realist position in order to convince us to accept what he
>> is saying as a valid defense of realism.

>Why do you think this is too vague? We have lots of examples
>where one theory appears as a limit of some more fundamental
>theory. And in all these cases the objects which "exist" in the
>approximation are somehow constructed from the objects
>which "exist" in the more fundamental theory. In every such case
>we can consider this connection in detail, as detailed as we like.
>Thus, the "vague" notion is not an appeal to sympathy, but
>an appeal to the large body of experience with such derived,
>non-fundamental objects.

There is one thing which is vague (his "existence in some coarse-grained
sense), and a different thing which is precise (an object which is
fundamental in one theory being derived in another). I said that the
vague thing was vague, and you are now pointing at the precise thing
and saying that it isn't vague.

What Sokal is claiming is that the objects which appear as
derived entities (such as Newtonian space) should not be regarded
as merely "derived entities", but should be considered to actually
exist "in some coarse-grained sense." He wants to ascribe to them
some property other than, and in addition to, their merely being
derived in some other theory. This property is vague, and the reader
can only allow himself to believe that Sokal has said something
correct and sensible if the reader himself is sympathetic to the
realist position that Sokal expresses, and is willing to forgive
some sloppiness in Sokal's phrasing because he feels that he
agress with it.

>> It is very important in these matters not to decide that we like
>> realism and then accept as valid an otherwise insufficient
>> argument in favour of it.

>Ok. So what?

So be careful.

>> Thus, if we find that the arguments with which
>> we convince ourselves that realism is necessary amount to
>> "In my opinion, realism should be considered a part of logic,"
>> then we must go in search of the reason why that opinion
>> was adopted, and reject it unless there is a rigorous basis
>> for it.

>I have adopted it because I have found the arguments
>in favour of this position convincing.

What are those arguments?

>I do not believe in such things as "rigorous basis" in
>general and in the most fundamental philosophy especially.
>Logic may not be justified on a rigorous basis. Because
>any possible "rigorous basis" is, itself, based on logic.
>Thus, the "rigorous basis" is only circular reasoning.

Indeed. Logic has this characteristic, but realism, by which
I mean the assertion that the role of the observer should
be ignored in the development of physical theories and only
added on at the end as an afterthought, does not.

>> The equivalence here is again, "in some sense", and has the particular
>> difficulty that one algorithm (the one used by a human) has, as its
>> output, patterns of sensations (as a prediction of future input).
>> I doubt that an algorithm which has sensations as output is one
>> that most realists would accept as an example of something which
>> could qualify as a fundamental constituent of reality, since the
>> things with which it deals (sensations) are purely mental.

>Here several things are mingled. The algorithm which is assumed
>to be used by the fabric of reality is one which computes
>X_n+1 = f(X_n).

You have mangled the X_n which I introduced. As I use them, they
refer to the object which the observer constructs as his own mental
representation of a putative state of affairs outside him, which
he demands is a Markov sequence, and further breaks down into
many miniature, localised states of affairs, in so far as it is
useful for him to do so in the process of surveying the regularities
in his observations. You have taken it and treated it as though
it were an element of an ontological theory which attempted to
describe something which has nothing to do with observation.

>> >> Whether or not a particle even exists can depend on the motion of
>> >> the observer (cf. the Unruh effect), so it seems strange to put that
>> >> sticky reality label onto the particles and their positions in space.
>>
>> >I don't.
>>
>> But Bohmian mechanics does, and that is the context in which I made
>> the remark above.

>Bohmian theories supposed to handle the Unruh effect don't.

You mean Bohmian field theories. This is true, but the original
motivation for Bohmian mechanics was to insist that there are
indeed particles, with positions, rather than ethereal waves
which collapse. To make the transition to a field theory one
has to abandon that, effectively recognizing that the original
ontology was wrong.

What survived from the original Bohmian theory was the quantum
formalism, namely that part of the theory which characterized
regularities in observed phenomena, and provided a prescription for
making future predictions. At this point one should realise that
the insistence on introducing the particle positions in the first
place, and insisting that they were ontologically real, was misguided.

>> In this case,
>> are you asserting that "reality" is something which can be
>> defined, or that its definition can change from theory to theory?

>The definition of reality is part of the realistic theory. It
>changes from theory to theory.

>In NT reality consists of point particles attracted by forces,
>in GR it consists of a manifold with a metric and some
>matter fields on it, in field theory of fields
>(\psi(x),A_i(x),g_{ij}(x)), in Bohmian theories of
>some q in Q and some \Psi: Q-->C, where Q is
>the configuration space which is very different in
>different Bohmian theories.

So you don't actually mean "define reality", but rather, "list
some objects which are to be considered to be real."

>> The real difficulty here is (as I mentioned before) the manifold
>> use of the word "real".

>Not that difficult IMHO.
> ...

>Reality
>I use for all this stuff outside, which is not in our minds.

What, then, does actual pain have which imagined pain lacks?

>> To add further to this confusion is the additional confusion caused
>> by the inappropriate use of the notion of "outside." The relations
>> of outside and inside are properly ascribed only to objects which
>> are in space.

>I hope my use of "outside" does not cause confusion.
>It means "not in my mind or my soul". My mind,
>(at least the material part of it) is sufficiently located in
>space. If there is something non-material connected with
>consciousness and so on (soul), it is not part of "outside".

Well, you have basically defined "outside" as "not inside", but
perhaps there might be a chance for some progress here. Firstly,
realise that in order to consider something to be "inside" something
else, both things need to have the property of location, while the
second also needs to have the property of extension.

What you refer to as "the material part" of your mind is, presumably,
your brain. It is true that this is localised in space, and is inside
your head. If, when you say that you are studying things outside of
your mind, all that you mean is that you are studying things outside
of your head, then I would not disagree, except to say that you
should use the word "brain" or "head" instead of "mind", which
already has a different meaning.

Regarding this "outside of the mind" business, consider any
existing physical object (for example, a computer screen). Notice
that there are colours on it, and that they are in specific
locations. A particular coloured shape (for example, a letter)
is inside a certain region, which contains many colours. This
is a legitimate use of the relation called "inside", since the
shape has a location, and the region has both location and
extension, and the coloured shape is, in fact, inside the
region. Furthermore, the region, and the coloured shape, are
outside of your head. However, neither the coloured shape
nor the region (of the image which is presented to your visual
field) would exist without a perceiving mind, since colours are
sensations and sensations are experiences. It is not correct to say
that they are inside the mind, since the mind is not a place, but
they exist only in virtue of your apprehension of them.

>> Now, taken together, these two confusions lead to the realist
>> position, where the physicist believes that he is examining
>> that which "really" exists, "outside" of his mind. In fact,
>> what he is doing is examining things which are actual (in
>> my sense of the term) and outside of his head. Perhaps I
>> should emphasise that this is worth thinking about; many
>> lives have been wasted chasing ghosts because of this
>> misunderstanding.

>I see not much advantage in your specification. I propose
>theories about actual things outside of my head? Fine.
>No problem. But also no advantage.

By stating things in this way, the implicit assertion that the
objects of investigation would continue to exist and have the
properties that they do in the absence of your perception of them
is dropped.

This is the actual solution to Sokal's problem. He wants to say
that, for example, Newtonian absolute space and time exist,
although, since he has learned relativity and believes it to
more accurately describe what exists, he can't. Rather than
inventing some vague watered-down type of existence, he should
have realised, when relativity (for him) deprived Newtonian space
and time of their ontological existence, that it meant that
Newtonian space and time enjoy no existence at all, outside
of the mind of the observer.

When he wants to say that they approximately exist, what he is
trying to capture is the idea that they approximately characterize
certain aspects of observed phenomena, and are therefore useful
mental inventions for the observer to employ when surveying
the data he has observed.

The actual situation is that there is an observer, and Newtonian
space and various other theories arise as he tries to characterize
observed phenomena. The realist, however, pays so much attention
to the theory and what it describes that he forgets this, thinking
that the theory is "What's really going on", and that the observer
is merely some object inside the theory, and not a very interesting
one at that. Having forgotten about the fact that there is an
observer who constructs the theories to characterize his observations,
Sokal is left trying to assign some kind of existence to the
theoretical objects themselves, while at the same time acknowledging
that they do not, in fact, actually exist, since they have been
superceded by different theoretical objects.

>There are more serious faults in modern science than this.
>Scientists should be independent. At least as independent
>as judges. Even more. Because a judge which makes obvious
>nonsense harms other people, a scientist which makes
>obvious nonsense doesn't.

Indeed, but pedagogy, which you don't care to argue about, is
extraordinarily important. Most physicists never think about it,
but they do get their interpretation of quantum mechanics from their
teachers, and if more people had a teacher who presented the actual
facts about Bohmian mechanics, rather than saying, for example,
that hidden variable theories were ruled out by Bell, then Bohmian
mechanics would be more popular.

The good news for you is that most physicists would agree with you,
if they took the time to examine the issues more carefully. The bad
news is that that won't happen.

>But current scientific organization is the reverse. If you
>have a grant for a few years, you have to care about your
>future, the next grant. The decisions are made by
>beaurocrats and other scientists (these two often unified
>in a single person). The predictable result is conformity.

>Another pressure into conformity is "publish or perish".
>Publishing and being cited is easier in a large community
>with lots of journals to publish and lots of readers.

>The prediction agrees with observation. In the highly
>speculative domain of modern fundamental physics,
>where free science would lead to lots of very different
>directions (because they are not guided by experiments,
>purely speculative, therefore its extremely hard to reject
>a research proposal as false) we observe de facto only
>two research directions - strings and LQG.

I vaguely remember a letter from Heisenberg to Bohr where he laments
about the widespread belief that if a large number of intelligent
people are working on a theory then it must be true.

>> I should add that it is not the case that the analytic method has
>> been exhausted due to a lack of performable experiments whose
>> results we do not know in advance. The confusion about realism
>> and the many disagreements about quantum mechanics (among those
>> who actually do think about it) are evidence enough that there
>> is an opportunity, for one who is prepared to suspend judgment
>> and avoid opinion, to discern something previously overlooked.

>The second sentence does not prove the first. The last two
>points are discussions not about the results of possible
>experiments.

Right; there is a lack of performable experiments whose results
we do not know in advance. However, this does not mean that
the analytic method of carefully considering the data that
we already have has been exhausted. There are many disagreements
about quantum mechanics, for example, which demonstrates that
nobody has yet presented a clear and indisputable analysis of the
available data which leads to a particular point of view.

Instead, physicists have adopted the positions that they have based
on opinions, which are different from one physicist to another,
hence the disagreements. To understand what's happening here, one
has to survey the various positions that people have adopted and
understand how they got to that position. The aim is to identify
the point at which opinion crept in, either the opinion of the
individual physicist, or one which was given to her by those around
her.

For example, "The wavefunction is a complete description,"
coupled to realism in the form "The observer is just another
object in the theory, so observation is just like any other
interaction," leads to the many-worlds interpretation. The
realisation that the first assertion is merely an opinion, together
with unquestioning adherence to the second leads to a Bohmian
interpretation.

>> That is, I perceive a room, with walls
>> here and there, and you perceive a different room (presumably),
>> and there is no simple matching of what I perceive to what you
>> perceive.

>This is about the M_n. My X_n contains also complete
>information about the inside of your room. I don't perceive
>the X_n. But the laws of physics, the rule which allows
>to compute X_n+1=f(X_n), need the whole information,
>my room as well as your room.

As I mentioned, you've hijacked entities which don't belong
in an ontological theory and are treating them as though they
do.

R.