View Full Version : Quantum entanglement and information transfer
Paul Stewart Snyder
Jun29-04, 05:38 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>Having read the recent news releases about quantum entanglement of\natoms, by tweaking one of three ions so that it has a particular\nquantum state and entangling the second and third ions, it seems to me\nthat the argument against instantaneous information transmission is\nlooking more and more like an egg or chicken argument. It is true that\nto communicate information we must know the meaning of a 0 and a 1,\nand that such knowledge must be transmitted by sub-light-speed methods\nbefore (or after) the quantum "teleportation". Yet this seems to be\nmore an issue of language than an issue of information transportation.\nIf in your dictionary you define the character 0 to mean "decrease"\nand the character 1 to mean "increase", then any time a 1 or a 0 is\nteleported the information "increase" or "decrease" would seem to be\nteleported?\n\nThis seems no different to the fact that we define English words, and\nthen communicate information by telephone using those words. No one\nwould question that the information being transmitted is communicated\nat the time of the call, and not when the words are defined. Indeed\nthe words may be defined after the communication, and the receiver can\nthen translate the words that were previously received by them,\nleaving the question whether words come before information or\ninformation before words, yet in either case we have both a language\nand an instantaneous transfer of information, a chicken and an egg.\n\nIn Bub, Jeffrey, "Quantum Entanglement and Information", The Stanford\nEncyclopedia of Philosophy (Winter 2002 Edition), Edward N. Zalta\n(ed.), URL = http://plato.stanford.edu/archives/win2002/entries/qt-entangle/\nthe entry concludes that: "In effect, if Bob (the receiver) can obtain\nno information about the bit in the safe, then entanglement will allow\nAlice (the sender) to ‘steer\' the bit to either 0 or 1 at will." It\nseems to me that the fact that Alice can steer the bit at will does\nnot mean that instantaneous information is not communicated, if for no\nother reason than she is in control, she decides the 0 or 1 "at will".\nEven though we can do all kinds of math that seems to deny\ninstantaneous transfer of information, I do not see how we can deny\nthat the ability to instantaneously send a 0 or 1 is communication of\ninformation, even if it is just a single letter of an unknown word?\n\nPS\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>Having read the recent news releases about quantum entanglement of
atoms, by tweaking one of three ions so that it has a particular
quantum state and entangling the second and third ions, it seems to me
that the argument against instantaneous information transmission is
looking more and more like an egg or chicken argument. It is true that
to communicate information we must know the meaning of a and a 1,
and that such knowledge must be transmitted by sub-light-speed methods
before (or after) the quantum "teleportation". Yet this seems to be
more an issue of language than an issue of information transportation.
If in your dictionary you define the character to mean "decrease"
and the character 1 to mean "increase", then any time a 1 or a is
teleported the information "increase" or "decrease" would seem to be
teleported?
This seems no different to the fact that we define English words, and
then communicate information by telephone using those words. No one
would question that the information being transmitted is communicated
at the time of the call, and not when the words are defined. Indeed
the words may be defined after the communication, and the receiver can
then translate the words that were previously received by them,
leaving the question whether words come before information or
information before words, yet in either case we have both a language
and an instantaneous transfer of information, a chicken and an egg.
In Bub, Jeffrey, "Quantum Entanglement and Information", The Stanford
Encyclopedia of Philosophy (Winter 2002 Edition), Edward N. Zalta
(ed.), URL = http://plato.stanford.edu/archives/win2002/entries/qt-entangle/
the entry concludes that: "In effect, if Bob (the receiver) can obtain
no information about the bit in the safe, then entanglement will allow
Alice (the sender) to ‘steer' the bit to either or 1 at will." It
seems to me that the fact that Alice can steer the bit at will does
not mean that instantaneous information is not communicated, if for no
other reason than she is in control, she decides the or 1 "at will".
Even though we can do all kinds of math that seems to deny
instantaneous transfer of information, I do not see how we can deny
that the ability to instantaneously send a or 1 is communication of
information, even if it is just a single letter of an unknown word?
PS
Peter Shor
Jul9-04, 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>\nps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0406260830.1ef36b5f@posting.google. com>...\n> Having read the recent news releases about quantum entanglement of\n> atoms, by tweaking one of three ions so that it has a particular\n> quantum state and entangling the second and third ions, it seems to me\n> that the argument against instantaneous information transmission is\n> looking more and more like an egg or chicken argument. It is true that\n> to communicate information we must know the meaning of a 0 and a 1,\n> and that such knowledge must be transmitted by sub-light-speed methods\n> before (or after) the quantum "teleportation". Yet this seems to be\n> more an issue of language than an issue of information transportation.\n> If in your dictionary you define the character 0 to mean "decrease"\n> and the character 1 to mean "increase", then any time a 1 or a 0 is\n> teleported the information "increase" or "decrease" would seem to be\n> teleported?\n\nYou can\'t communicate information faster than light using quantum\nentanglement, even if you have shared a dictionary by sub-light speeds\nbeforehand. I think you\'re reading the new release about a quantum\nteleportation experiment. Here, you transmit a quantum state over a\nclassical channel, but to do this you need to use the classical\ninformation transmitted over the channel to rectify the quantum state.\nWithout this, if you try to teleport a qubit you will get a\nprobabilistic mixture of four possible states of the qubit which are\narranged so that without the classical message, they will reveal no\ninformation about the original qubit to be teleported.\n\n> This seems no different to the fact that we define English words, and\n> then communicate information by telephone using those words. No one\n> would question that the information being transmitted is communicated\n> at the time of the call, and not when the words are defined. Indeed\n> the words may be defined after the communication, and the receiver can\n> then translate the words that were previously received by them,\n> leaving the question whether words come before information or\n> information before words, yet in either case we have both a language\n> and an instantaneous transfer of information, a chicken and an egg.\n\nThis isn\'t the same thing.\n\n> In Bub, Jeffrey, "Quantum Entanglement and Information", The Stanford\n> Encyclopedia of Philosophy (Winter 2002 Edition), Edward N. Zalta\n> (ed.), URL = http://plato.stanford.edu/archives/win2002/entries/qt-entangle/\n> the entry concludes that: "In effect, if Bob (the receiver) can obtain\n> no information about the bit in the safe, then entanglement will allow\n> Alice (the sender) to ?steer\' the bit to either 0 or 1 at will." It\n> seems to me that the fact that Alice can steer the bit at will does\n> not mean that instantaneous information is not communicated, if for no\n> other reason than she is in control, she decides the 0 or 1 "at will".\n> Even though we can do all kinds of math that seems to deny\n> instantaneous transfer of information, I do not see how we can deny\n> that the ability to instantaneously send a 0 or 1 is communication of\n> information, even if it is just a single letter of an unknown word?\n\nYou\'re misreading the article, which I\'ve skimmed briefly and which\nlooks correct, if somewhat confusing. The bit that Alice can steer at\nwill is not half of an EPR pair, but the outcome of a hypothesized\nquantum bit committment protocol. Such a quantum bit committment\nprotocol is impossible, and this "steering" argument is trying to\nexplain why.\n\n> PS\n\nPeter Shor\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>ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0406260830.1ef36b5f@posting.google.com>...
> Having read the recent news releases about quantum entanglement of
> atoms, by tweaking one of three ions so that it has a particular
> quantum state and entangling the second and third ions, it seems to me
> that the argument against instantaneous information transmission is
> looking more and more like an egg or chicken argument. It is true that
> to communicate information we must know the meaning of a and a 1,
> and that such knowledge must be transmitted by sub-light-speed methods
> before (or after) the quantum "teleportation". Yet this seems to be
> more an issue of language than an issue of information transportation.
> If in your dictionary you define the character to mean "decrease"
> and the character 1 to mean "increase", then any time a 1 or a is
> teleported the information "increase" or "decrease" would seem to be
> teleported?
You can't communicate information faster than light using quantum
entanglement, even if you have shared a dictionary by sub-light speeds
beforehand. I think you're reading the new release about a quantum
teleportation experiment. Here, you transmit a quantum state over a
classical channel, but to do this you need to use the classical
information transmitted over the channel to rectify the quantum state.
Without this, if you try to teleport a qubit you will get a
probabilistic mixture of four possible states of the qubit which are
arranged so that without the classical message, they will reveal no
information about the original qubit to be teleported.
> This seems no different to the fact that we define English words, and
> then communicate information by telephone using those words. No one
> would question that the information being transmitted is communicated
> at the time of the call, and not when the words are defined. Indeed
> the words may be defined after the communication, and the receiver can
> then translate the words that were previously received by them,
> leaving the question whether words come before information or
> information before words, yet in either case we have both a language
> and an instantaneous transfer of information, a chicken and an egg.
This isn't the same thing.
> In Bub, Jeffrey, "Quantum Entanglement and Information", The Stanford
> Encyclopedia of Philosophy (Winter 2002 Edition), Edward N. Zalta
> (ed.), URL = http://plato.stanford.edu/archives/win2002/entries/qt-entangle/
> the entry concludes that: "In effect, if Bob (the receiver) can obtain
> no information about the bit in the safe, then entanglement will allow
> Alice (the sender) to ?steer' the bit to either or 1 at will." It
> seems to me that the fact that Alice can steer the bit at will does
> not mean that instantaneous information is not communicated, if for no
> other reason than she is in control, she decides the or 1 "at will".
> Even though we can do all kinds of math that seems to deny
> instantaneous transfer of information, I do not see how we can deny
> that the ability to instantaneously send a or 1 is communication of
> information, even if it is just a single letter of an unknown word?
You're misreading the article, which I've skimmed briefly and which
looks correct, if somewhat confusing. The bit that Alice can steer at
will is not half of an EPR pair, but the outcome of a hypothesized
quantum bit committment protocol. Such a quantum bit committment
protocol is impossible, and this "steering" argument is trying to
explain why.
> PS
Peter Shor
Paul Stewart Snyder
Jul11-04, 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\n\npeterwshor@aol.com (Peter Shor) wrote in message news:<9b2e17b4.0407071445.6580c78e@posting.google. com>...\n> ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0406260830.1ef36b5f@posting.google. com>...\n> > Having read the recent news releases about quantum entanglement of\n> > atoms, by tweaking one of three ions so that it has a particular\n> > quantum state and entangling the second and third ions, it seems to me\n> > that the argument against instantaneous information transmission is\n> > looking more and more like an egg or chicken argument. It is true that\n> > to communicate information we must know the meaning of a 0 and a 1,\n> > and that such knowledge must be transmitted by sub-light-speed methods\n> > before (or after) the quantum "teleportation". Yet this seems to be\n> > more an issue of language than an issue of information transportation.\n> > If in your dictionary you define the character 0 to mean "decrease"\n> > and the character 1 to mean "increase", then any time a 1 or a 0 is\n> > teleported the information "increase" or "decrease" would seem to be\n> > teleported?\n>\n> You can\'t communicate information faster than light using quantum\n> entanglement, even if you have shared a dictionary by sub-light speeds\n> beforehand. I think you\'re reading the new release about a quantum\n> teleportation experiment. Here, you transmit a quantum state over a\n> classical channel, but to do this you need to use the classical\n> information transmitted over the channel to rectify the quantum state.\n> Without this, if you try to teleport a qubit you will get a\n> probabilistic mixture of four possible states of the qubit which are\n> arranged so that without the classical message, they will reveal no\n> information about the original qubit to be teleported.\n>\nI see - but when two entangled particles are separated, can you not\nknow the states of both particles so that if the state of one changes\nyou at least know that the state of the other changed - which would in\nitself be a piece of information? Or are you saying that you cannot\neven tell if entangled particle A changes state without knowing (by\nclassical transfer of information) particle B\'s state? Can you not\ndetect a change in the state of particle A and then "know" that B has\nchanged as well, even if you have no idea what that change is?\n\nPS\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>peterwshor@aol.com (Peter Shor) wrote in message news:<9b2e17b4.0407071445.6580c78e@posting.google.com>...
> ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0406260830.1ef36b5f@posting.google.com>...
> > Having read the recent news releases about quantum entanglement of
> > atoms, by tweaking one of three ions so that it has a particular
> > quantum state and entangling the second and third ions, it seems to me
> > that the argument against instantaneous information transmission is
> > looking more and more like an egg or chicken argument. It is true that
> > to communicate information we must know the meaning of a and a 1,
> > and that such knowledge must be transmitted by sub-light-speed methods
> > before (or after) the quantum "teleportation". Yet this seems to be
> > more an issue of language than an issue of information transportation.
> > If in your dictionary you define the character to mean "decrease"
> > and the character 1 to mean "increase", then any time a 1 or a is
> > teleported the information "increase" or "decrease" would seem to be
> > teleported?
>
> You can't communicate information faster than light using quantum
> entanglement, even if you have shared a dictionary by sub-light speeds
> beforehand. I think you're reading the new release about a quantum
> teleportation experiment. Here, you transmit a quantum state over a
> classical channel, but to do this you need to use the classical
> information transmitted over the channel to rectify the quantum state.
> Without this, if you try to teleport a qubit you will get a
> probabilistic mixture of four possible states of the qubit which are
> arranged so that without the classical message, they will reveal no
> information about the original qubit to be teleported.
>
I see - but when two entangled particles are separated, can you not
know the states of both particles so that if the state of one changes
you at least know that the state of the other changed - which would in
itself be a piece of information? Or are you saying that you cannot
even tell if entangled particle A changes state without knowing (by
classical transfer of information) particle B's state? Can you not
detect a change in the state of particle A and then "know" that B has
changed as well, even if you have no idea what that change is?
PS
Peter Shor
Jul13-04, 02:37 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>ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407092005.48183cf5@posting.google. com>...\n> peterwshor@aol.com (Peter Shor) wrote in message news:<9b2e17b4.0407071445.6580c78e@posting.google. com>...\n> > ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0406260830.1ef36b5f@posting.google. com>...\n> > > Having read the recent news releases about quantum entanglement of\n> > > atoms, by tweaking one of three ions so that it has a particular\n> > > quantum state and entangling the second and third ions, it seems to me\n> > > that the argument against instantaneous information transmission is\n> > > looking more and more like an egg or chicken argument. It is true that\n> > > to communicate information we must know the meaning of a 0 and a 1,\n> > > and that such knowledge must be transmitted by sub-light-speed methods\n> > > before (or after) the quantum "teleportation". Yet this seems to be\n> > > more an issue of language than an issue of information transportation.\n> > > If in your dictionary you define the character 0 to mean "decrease"\n> > > and the character 1 to mean "increase", then any time a 1 or a 0 is\n> > > teleported the information "increase" or "decrease" would seem to be\n> > > teleported?\n> >\n> > You can\'t communicate information faster than light using quantum\n> > entanglement, even if you have shared a dictionary by sub-light speeds\n> > beforehand. I think you\'re reading the new release about a quantum\n> > teleportation experiment. Here, you transmit a quantum state over a\n> > classical channel, but to do this you need to use the classical\n> > information transmitted over the channel to rectify the quantum state.\n> > Without this, if you try to teleport a qubit you will get a\n> > probabilistic mixture of four possible states of the qubit which are\n> > arranged so that without the classical message, they will reveal no\n> > information about the original qubit to be teleported.\n> >\n> I see - but when two entangled particles are separated, can you not\n> know the states of both particles so that if the state of one changes\n> you at least know that the state of the other changed - which would in\n> itself be a piece of information? Or are you saying that you cannot\n> even tell if entangled particle A changes state without knowing (by\n> classical transfer of information) particle B\'s state? Can you not\n> detect a change in the state of particle A and then "know" that B has\n> changed as well, even if you have no idea what that change is?\n>\n> PS\n\nYou can know the joint state of the two entangled particles. However,\nyou cannot monitor the state of a quantum particle without changing it,\nso trying to send information instantaneously this way doesn\'t work.\nYou can learn a little bit about the state of a particle without changing\nits state too much, so you might think that maybe by weakly monitoring\nthe state of particle B, then when particle A\'s state is changed, you\nwould be able to detect this with some small probability by looking at B.\nHowever, it doesn\'t work that way. If you try to work it out mathematically,\nit may seem as though the laws of quantum mechanics conspire to foil all\nattempts at instantaneous communication. Of course it\'s not really a\nconspiracy, it\'s built into the basic rules of quantum mechanics. A good\nparallel is people trying to build a perpetual motion machine which can\nprovide useful work. No matter how ingenious they are, their designs don\'t\nwork --- even though it might be hard to actually find the flaw in some\nof them; and this is because the laws of thermodynamics are built into\nthe basic rules of classical mechanics.\n\nPeople learning physics have had a difficult time trying to get their\nminds around entanglement and "spooky action at a distance" since quantum\nmechanics was developed. And now, with the rise of quantum information\nand quantum computation, this has become no longer just a curious property\nof the quantum world, but a potentially useful resource, and so we have\na lot more journalists and websites trying to explain it unsuccessfully.\n(And with the burgeoning use of entangled qubits for instantenous\ncommunication in science fiction books, the lack of general understanding\nis only going to get worse.)\n\nPeter Shor\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>ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407092005.48183cf5@posting.google.com>...
> peterwshor@aol.com (Peter Shor) wrote in message news:<9b2e17b4.0407071445.6580c78e@posting.google.com>...
> > ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0406260830.1ef36b5f@posting.google.com>...
> > > Having read the recent news releases about quantum entanglement of
> > > atoms, by tweaking one of three ions so that it has a particular
> > > quantum state and entangling the second and third ions, it seems to me
> > > that the argument against instantaneous information transmission is
> > > looking more and more like an egg or chicken argument. It is true that
> > > to communicate information we must know the meaning of a and a 1,
> > > and that such knowledge must be transmitted by sub-light-speed methods
> > > before (or after) the quantum "teleportation". Yet this seems to be
> > > more an issue of language than an issue of information transportation.
> > > If in your dictionary you define the character to mean "decrease"
> > > and the character 1 to mean "increase", then any time a 1 or a is
> > > teleported the information "increase" or "decrease" would seem to be
> > > teleported?
> >
> > You can't communicate information faster than light using quantum
> > entanglement, even if you have shared a dictionary by sub-light speeds
> > beforehand. I think you're reading the new release about a quantum
> > teleportation experiment. Here, you transmit a quantum state over a
> > classical channel, but to do this you need to use the classical
> > information transmitted over the channel to rectify the quantum state.
> > Without this, if you try to teleport a qubit you will get a
> > probabilistic mixture of four possible states of the qubit which are
> > arranged so that without the classical message, they will reveal no
> > information about the original qubit to be teleported.
> >
> I see - but when two entangled particles are separated, can you not
> know the states of both particles so that if the state of one changes
> you at least know that the state of the other changed - which would in
> itself be a piece of information? Or are you saying that you cannot
> even tell if entangled particle A changes state without knowing (by
> classical transfer of information) particle B's state? Can you not
> detect a change in the state of particle A and then "know" that B has
> changed as well, even if you have no idea what that change is?
>
> PS
You can know the joint state of the two entangled particles. However,
you cannot monitor the state of a quantum particle without changing it,
so trying to send information instantaneously this way doesn't work.
You can learn a little bit about the state of a particle without changing
its state too much, so you might think that maybe by weakly monitoring
the state of particle B, then when particle A's state is changed, you
would be able to detect this with some small probability by looking at B.
However, it doesn't work that way. If you try to work it out mathematically,
it may seem as though the laws of quantum mechanics conspire to foil all
attempts at instantaneous communication. Of course it's not really a
conspiracy, it's built into the basic rules of quantum mechanics. A good
parallel is people trying to build a perpetual motion machine which can
provide useful work. No matter how ingenious they are, their designs don't
work --- even though it might be hard to actually find the flaw in some
of them; and this is because the laws of thermodynamics are built into
the basic rules of classical mechanics.
People learning physics have had a difficult time trying to get their
minds around entanglement and "spooky action at a distance" since quantum
mechanics was developed. And now, with the rise of quantum information
and quantum computation, this has become no longer just a curious property
of the quantum world, but a potentially useful resource, and so we have
a lot more journalists and websites trying to explain it unsuccessfully.
(And with the burgeoning use of entangled qubits for instantenous
communication in science fiction books, the lack of general understanding
is only going to get worse.)
Peter Shor
<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>\nPaul Stewart Snyder:\n\n> Can you not detect a change in the state of particle A\n> and then "know" that B has changed as well, even if you\n> have no idea what that change is?\n\nMaybe the problem is that observer at A "knows", but the\nobserver at B don\'t.\n\nJ.B. Kennedy ["On the Empirical Foundations of the Quantum\nNo-Signalling Proofs", Phil. Sci., 62, 543-560 (1995)]\nthinks that any proof that QM cannot support superluminal\nsignalling, by using entanglement, is essentially circular.\nKennedy also suggested that von Neumann\'s original motivation\nfor introducing the tensor product rule, was essentially\nto block the possibility of superluminal signalling.\n\nC.A. Fuchs wrote something about the tensor product\nrule, for combining quantum systems:\nhttp://www.arxiv.org/abs/quant-ph/0106166\n\ns.\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>Paul Stewart Snyder:
> Can you not detect a change in the state of particle A
> and then "know" that B has changed as well, even if you
> have no idea what that change is?
Maybe the problem is that observer at A "knows", but the
observer at B don't.
J.B. Kennedy ["On the Empirical Foundations of the Quantum
No-Signalling Proofs", Phil. Sci., 62, 543-560 (1995)]
thinks that any proof that QM cannot support superluminal
signalling, by using entanglement, is essentially circular.
Kennedy also suggested that von Neumann's original motivation
for introducing the tensor product rule, was essentially
to block the possibility of superluminal signalling.
C.A. Fuchs wrote something about the tensor product
rule, for combining quantum systems:
http://www.arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/0106166
s.
Paul Stewart Snyder
Jul15-04, 03:57 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"scerir" <scerir@libero.it> wrote in message news:<XmTIc.48796\\$GQ3.1368186@twister2.libero.it >...\n>... Kennedy also suggested that von Neumann\'s original motivation\n> for introducing the tensor product rule, was essentially\n> to block the possibility of superluminal signalling.\n\nAs Peter Shor said: "Of course it\'s not really a conspiracy, it\'s\nbuilt into the basic rules of quantum mechanics." Honestly that is\nwhat bothers me. There seems to be an assumption underlying QM that\nnothing can exceed the speed of light, one question would be how\nfundamental and necessary is that assumption? We have A\ninstantaneously changing state when B does, no matter what the\ndistance, and presumably even if no measurement can be made. I see\nshadows of the Copenhagen Interpretation, if we moved toward a many\nworlds model would the idea of instantaneous change be more likely to\nhave a physical reality?\n\nAre Allan Goff and Dale Lehmann suggesting that QM may be consistent\nwith FTL interactions in:\nhttp://content.aip.org/APCPCS/v699/i1/1182_1.html\nwhen they say:\n"A protocol using cross-entangled independent Einstein-Podolsky-Rosen\n(EPR) beams is developed as a means of sending information faster than\nlight (FTL) by taking advantage of quantum nonlocality and\nindistinguishable particle statistics. Two observers bracket a central\nmidpoint transmitter that contains dual EPR sources from which bits\nare encoded in packets of photon pairs. FTL communication occurs\nbetween the observers in a simplex mode. A reformulation of quantum\nmechanics is proposed that permits such communications, as well as\nwave function collapse, to be relativistically consistent, while also\nresolving the problem of causal ordering normally associated with\nspacelike connections. The issue of temporal paradox is handled\nseparately at both the quantum and classical levels. Spacelike causal\nconnections lead to closed loops in spacetime and causal\nself-reference. It is shown that such self-reference leads to\nnonlinearities in the evolution of the wave function that may be\nsufficient to lead to wave function collapse. ©2004 American Institute\nof Physics"\n\nPS\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>"scerir" <scerir@libero.it> wrote in message news:<XmTIc.48796$GQ3.1368186@twister2.libero.it>...
>... Kennedy also suggested that von Neumann's original motivation
> for introducing the tensor product rule, was essentially
> to block the possibility of superluminal signalling.
As Peter Shor said: "Of course it's not really a conspiracy, it's
built into the basic rules of quantum mechanics." Honestly that is
what bothers me. There seems to be an assumption underlying QM that
nothing can exceed the speed of light, one question would be how
fundamental and necessary is that assumption? We have A
instantaneously changing state when B does, no matter what the
distance, and presumably even if no measurement can be made. I see
shadows of the Copenhagen Interpretation, if we moved toward a many
worlds model would the idea of instantaneous change be more likely to
have a physical reality?
Are Allan Goff and Dale Lehmann suggesting that QM may be consistent
with FTL interactions in:
http://content.aip.org/APCPCS/v699/i1/1182_1.html
when they say:
"A protocol using cross-entangled independent Einstein-Podolsky-Rosen
(EPR) beams is developed as a means of sending information faster than
light (FTL) by taking advantage of quantum nonlocality and
indistinguishable particle statistics. Two observers bracket a central
midpoint transmitter that contains dual EPR sources from which bits
are encoded in packets of photon pairs. FTL communication occurs
between the observers in a simplex mode. A reformulation of quantum
mechanics is proposed that permits such communications, as well as
wave function collapse, to be relativistically consistent, while also
resolving the problem of causal ordering normally associated with
spacelike connections. The issue of temporal paradox is handled
separately at both the quantum and classical levels. Spacelike causal
connections lead to closed loops in spacetime and causal
self-reference. It is shown that such self-reference leads to
nonlinearities in the evolution of the wave function that may be
sufficient to lead to wave function collapse. ©2004 American Institute
of Physics"
PS
<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\nPaul Stewart Snyder:\n> As Peter Shor said: "Of course it\'s not really a conspiracy, it\'s\n> built into the basic rules of quantum mechanics." Honestly that is\n> what bothers me. There seems to be an assumption underlying QM that\n> nothing can exceed the speed of light, one question would be how\n> fundamental and necessary is that assumption?\n\nThe assumption that no signal can exceed the speed of light?\nAccording to the philosophers it is as fundamental and necessary\nas indeterminism, and viceversa.\n\nBell imposed (as far as I remember) few assumptions.\n1) Hidden variables, that is to say, determinism.\n2) Local realism, local causality (but see point 3).\n3) The same "lambda" (hidden variables description) for both\nspace-like separated wings of his set-up.\n4) Reality is "single valued" (no MWI).\n5) Observers are "free", i.e. to choose observables, parameters,\netc. (If observers were not free, that is to say if they too\nwere pre-determined, i.e. by hidden variables, Bell\'s result\nis completely meaningless. Hence: no super-determinism).\n\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----------\n\nUnder those assumptions, if A is one of the two wings of a\ntypical Bell apparatus, i the observable to be measured in A\nand x its possible value, and if B is the other of the two wings,\nj is the observable to be measured in B and y its possible value,\nand if Lambda is the hidden variables joint state description of\nthe composite system, we can write, following Bell .....\n\np_A,B,Lambda (x,y|i,j) = p_A,Lambda (x|i) p_B,Lambda (y|j)\n\nwhich is the so called "Local Realism" factorizability condition or,\nin Bell\'s own terms, the "Local Causality" condition or, in Jarrett\'s\nterms, the "Strong Causality" condition. And it just means that the\njoint probability of outcomes x and y, for measurements of observables\ni and j, at the A and B wings, is equal to the product of the the\nseparate probabilities. (We know that so many experiments have shown\nthe expression above is far from reality).\n\nThe above condition is equivalent (after Jarrett, 1983/1984) to the\nconjunction of two (double) independent conditions ....\n\nLocality Condition (or parameter independence, simple locality)\np_A,Lambda (x|i,j) = p_A,Lambda (x|i)\np_B,Lambda (y|i,j) = p_B,Lambda (y|j)\n\nSeparability Condition (or outcome independence, completeness)\np_A,Lambda (x|i,j,y) = p_A,Lambda (x|i,j)\np_B,Lambda (y|i,j,x) = P_B,Lambda (y|i,j)\n\nIt is possible to show (following Jarrett, Shimony, Ghirardi, Howard,\nCushing, Eberhard, maybe van Fraassen, maybe Fine, etc.) that QM\nviolates the Separability Condition but does *not* violate the Locality\nCondition. In physical terms the above means (Eberhard, Nuovo Cimento,\n46B, 1978, 392; Ghirardi et al., Found. Phys., 23, 1993, 341) that QM\ndoes not allow controllable FTL signalling (which may be different\nfrom *uncontrollable* actions, influences, "passions", "fashions"\nat a distance).\n\nIt is possible to show (following Jarrett, Shimony, Ghirardi, Howard,\nEberhard, Cushing, maybe van Fraassen, maybe Fine, etc.) that a\na (phantomatic) deterministic theory (i.e. one in which the range\nof any probability distribution of outcomes is the set: 0 or 1)\nreproducing all the predictions of QM (and thus necessarily violating\nBell\'s original "Local Causality" condition) can *not* violate the\nSeparability Condition, and, thus, must violate the Locality\nCondition. Just the reverse of what the usual indeterministic\nversion of QM does. (Better not to ask what kind of physical theory\ncan violate both the Separability Condition and the Locality\nCondition!)\n\nIt seems simple to realize that, at least in a naive mode.\nThe Separability Condition (or outcome independence, or completeness)\nmeans that ...\n\np_A,Lambda (x|i,j,y) = p_A,Lambda (x|i,j)\np_B,Lambda (y|i,j,x) = P_B,Lambda (y|i,j)\n\n..... so, if the specification of Lambda, i, j, in principle determines\ncompletely the outcomes x, y, then any additional conditioning on\nx or y is superfluous, having x and y just one value allowed, so they\ncannot affect the probability, which (in a deterministic theory) can\ntake just the values 0 or 1. Thus a (phantomatic) *deterministic* QM\ncannot violate - I would say by definition - the Separability Condition,\nand thus it must violate the Locality Condition (which means FTL\nsignalling).\n\nSomebody may ask: if QM (the usual orthodox indeterministic version)\nviolates just the Separability Condition (and does not violate\nthe Locality Condition) can the original stronger Bell\'s "Local\nCausality" condition be replaced by this Separability Condition alone?\nGood question. With a poor answer! The Separability Condition alone\nis not sufficient to prove, mathematically, Bell\'s inequalities, as\nfar as I know.\n\ns.\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>Paul Stewart Snyder:
> As Peter Shor said: "Of course it's not really a conspiracy, it's
> built into the basic rules of quantum mechanics." Honestly that is
> what bothers me. There seems to be an assumption underlying QM that
> nothing can exceed the speed of light, one question would be how
> fundamental and necessary is that assumption?
The assumption that no signal can exceed the speed of light?
According to the philosophers it is as fundamental and necessary
as indeterminism, and viceversa.
Bell imposed (as far as I remember) few assumptions.
1) Hidden variables, that is to say, determinism.
2) Local realism, local causality (but see point 3).
3) The same "\lambda" (hidden variables description) for both
space-like separated wings of his set-up.
4) Reality is "single valued" (no MWI).
5) Observers are "free", i.e. to choose observables, parameters,
etc. (If observers were not free, that is to say if they too
were pre-determined, i.e. by hidden variables, Bell's result
is completely meaningless. Hence: no 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)
----------
Under those assumptions, if A is one of the two wings of a
typical Bell apparatus, i the observable to be measured in A
and x its possible value, and if B is the other of the two wings,
j is the observable to be measured in B and y its possible value,
and if \Lambda is the hidden variables joint state description of
the composite system, we can write, following Bell .....
p_A,B,\Lambda (x,y|i,j) = p_A,\Lambda (x|i) p_B,\Lambda (y|j)
which is the so called "Local Realism" factorizability condition or,
in Bell's own terms, the "Local Causality" condition or, in Jarrett's
terms, the "Strong Causality" condition. And it just means that the
joint probability of outcomes x and y, for measurements of observables
i and j, at the A and B wings, is equal to the product of the the
separate probabilities. (We know that so many experiments have shown
the expression above is far from reality).
The above condition is equivalent (after Jarrett, 1983/1984) to the
conjunction of two (double) independent conditions ....
Locality Condition (or parameter independence, simple locality)
p_A,\Lambda (x|i,j) = p_A,\Lambda (x|i)p_B,\Lambda (y|i,j) = p_B,\Lambda (y|j)
Separability Condition (or outcome independence, completeness)
p_A,\Lambda (x|i,j,y) = p_A,\Lambda (x|i,j)p_B,\Lambda (y|i,j,x) = P_B,\Lambda (y|i,j)
It is possible to show (following Jarrett, Shimony, Ghirardi, Howard,
Cushing, Eberhard, maybe van Fraassen, maybe Fine, etc.) that QM
violates the Separability Condition but does *not* violate the Locality
Condition. In physical terms the above means (Eberhard, Nuovo Cimento,
46B, 1978, 392; Ghirardi et al., Found. Phys., 23, 1993, 341) that QM
does not allow controllable FTL signalling (which may be different
from *uncontrollable* actions, influences, "passions", "fashions"
at a distance).
It is possible to show (following Jarrett, Shimony, Ghirardi, Howard,
Eberhard, Cushing, maybe van Fraassen, maybe Fine, etc.) that a
a (phantomatic) deterministic theory (i.e. one in which the range
of any probability distribution of outcomes is the set: or 1)
reproducing all the predictions of QM (and thus necessarily violating
Bell's original "Local Causality" condition) can *not* violate the
Separability Condition, and, thus, must violate the Locality
Condition. Just the reverse of what the usual indeterministic
version of QM does. (Better not to ask what kind of physical theory
can violate both the Separability Condition and the Locality
Condition!)
It seems simple to realize that, at least in a naive mode.
The Separability Condition (or outcome independence, or completeness)
means that ...
p_A,\Lambda (x|i,j,y) = p_A,\Lambda (x|i,j)p_B,\Lambda (y|i,j,x) = P_B,\Lambda (y|i,j)
..... so, if the specification of \Lambda, i, j, in principle determines
completely the outcomes x, y, then any additional conditioning on
x or y is superfluous, having x and y just one value allowed, so they
cannot affect the probability, which (in a deterministic theory) can
take just the values or 1. Thus a (phantomatic) *deterministic* QM
cannot violate - I would say by definition - the Separability Condition,
and thus it must violate the Locality Condition (which means FTL
signalling).
Somebody may ask: if QM (the usual orthodox indeterministic version)
violates just the Separability Condition (and does not violate
the Locality Condition) can the original stronger Bell's "Local
Causality" condition be replaced by this Separability Condition alone?
Good question. With a poor answer! The Separability Condition alone
is not sufficient to prove, mathematically, Bell's inequalities, as
far as I know.
s.
Tom Trotter
Jul16-04, 08:19 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>\nps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407141925.2054c9a0@posting.google. com>...\n\n[... snip]\n\n> We have A instantaneously changing state when\n> B does, no matter what the distance, and presumably\n> even if no measurement can be made. I see\n> shadows of the Copenhagen Interpretation, if we\n> moved toward a many worlds model would the idea\n> of instantaneous change be more likely to\n> have a physical reality?\n\n[ ... snip]\n\nThe idea of instantaneous (simultaneous) change\n*does* have a physical reality (without MWI),\ndoesn\'t it?\n\nConsider two photons emitted from\nthe same atom and correlated in polarization.\nIf you learn the polarization of one of the\nphotons, then you instantly know the polarization\nof the other.\n\nConsider a wheel with two marks on opposite\nsides of it. As the wheel revolves, the\nopposing marks simultaneously change position.\n\nConsider two colliding objects whose momenta\nare known prior to their collision. In a\nsufficiently closed/isolated setting, subsequently\nlearning the momentum of one of the objects\nwill allow you to deduce the momentum of the\nother.\n\nThe entities in the above examples have certain\nrelationships to each other that allow their\nbehavior to be correlated in some encompassing\nobservational context. These correlations\nrequire no communication between the correlated\nentities.\n\nOr, am I missing your point?\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>ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407141925.2054c9a0@posting.google.com>...
[... snip]
> We have A instantaneously changing state when
> B does, no matter what the distance, and presumably
> even if no measurement can be made. I see
> shadows of the Copenhagen Interpretation, if we
> moved toward a many worlds model would the idea
> of instantaneous change be more likely to
> have a physical reality?
[ ... snip]
The idea of instantaneous (simultaneous) change
*does* have a physical reality (without MWI),
doesn't it?
Consider two photons emitted from
the same atom and correlated in polarization.
If you learn the polarization of one of the
photons, then you instantly know the polarization
of the other.
Consider a wheel with two marks on opposite
sides of it. As the wheel revolves, the
opposing marks simultaneously change position.
Consider two colliding objects whose momenta
are known prior to their collision. In a
sufficiently closed/isolated setting, subsequently
learning the momentum of one of the objects
will allow you to deduce the momentum of the
other.
The entities in the above examples have certain
relationships to each other that allow their
behavior to be correlated in some encompassing
observational context. These correlations
require no communication between the correlated
entities.
Or, am I missing your point?
Tom Trotter
Jul16-04, 01:50 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"scerir" <scerir@libero.it> wrote in message news:<BdzJc.54288\\$c_1.1744453@twister1.libero.it >...\n\n[... snip interesting discussion on orders of moderator]\n\n> Somebody may ask: if QM (the usual orthodox indeterministic version)\n> violates just the Separability Condition (and does not violate\n> the Locality Condition) can the original stronger Bell\'s "Local\n> Causality" condition be replaced by this Separability Condition alone?\n> Good question. With a poor answer! The Separability Condition alone\n> is not sufficient to prove, mathematically, Bell\'s inequalities, as\n> far as I know.\n\nIs there perhaps a simple and clear way to resolve\nthe conundra associated with EPRBell considerations?\n\nThe variable, lambda, which, if its value were\nknown would allow more accurate predictions of\nindividual results, simply isn\'t relevant wrt the\ndetermination of correlations in the combined\ncontext.\n\n"Nonseparability" is, imo, a vague way of\nreferring to what *is* relevant (wrt paired\nphotons) in the combined context -- and\nthat is the *relationship* of their\nemission-produced polarizations (ie., they\nare polarized identically via the emission\nprocess).\n\nThe combined context isn\'t a local hidden\nvariable context. This doesn\'t mean that\nthere aren\'t local hidden variables operating\nin \'reality\'. It just means that these variables\naren\'t relevant to the determination of results\nin combined contexts.\n\nAnd, insofar as qm deals with combined\ncontexts (ie., correlations of entangled\nentities), then it would of course follow\nthat, at least wrt these contexts, qm is\nincompatible with local hidden variable\nformulations.\n\nExperimental tests of Bell inequalities\nare just comparing formulations that\nare relevant to the observational context,\n(qm), to formulations that aren\'t, (lhv).\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>"scerir" <scerir@libero.it> wrote in message news:<BdzJc.54288$c_1.1744453@twister1.libero.it>...
[... snip interesting discussion on orders of moderator]
> Somebody may ask: if QM (the usual orthodox indeterministic version)
> violates just the Separability Condition (and does not violate
> the Locality Condition) can the original stronger Bell's "Local
> Causality" condition be replaced by this Separability Condition alone?
> Good question. With a poor answer! The Separability Condition alone
> is not sufficient to prove, mathematically, Bell's inequalities, as
> far as I know.
Is there perhaps a simple and clear way to resolve
the conundra associated with EPRBell considerations?
The variable, \lambda, which, if its value were
known would allow more accurate predictions of
individual results, simply isn't relevant wrt the
determination of correlations in the combined
context.
"Nonseparability" is, imo, a vague way of
referring to what *is* relevant (wrt paired
photons) in the combined context -- and
that is the *relationship* of their
emission-produced polarizations (ie., they
are polarized identically via the emission
process).
The combined context isn't a local hidden
variable context. This doesn't mean that
there aren't local hidden variables operating
in 'reality'. It just means that these variables
aren't relevant to the determination of results
in combined contexts.
And, insofar as qm deals with combined
contexts (ie., correlations of entangled
entities), then it would of course follow
that, at least wrt these contexts, qm is
incompatible with local hidden variable
formulations.
Experimental tests of Bell inequalities
are just comparing formulations that
are relevant to the observational context,
(qm), to formulations that aren't, (lhv).
Caroline Thompson
Jul19-04, 03:08 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"Tom Trotter" <tom129@juno.com> wrote in message\nnews:29df3039.0407152247.ba0eee4@posting. google.com...\n\n> The idea of instantaneous (simultaneous) change\n> *does* have a physical reality (without MWI),\n> doesn\'t it?\n>\n> Consider two photons emitted from\n> the same atom and correlated in polarization.\n> If you learn the polarization of one of the\n> photons, then you instantly know the polarization\n> of the other.\n\nYes, but the case of interest is when the polarisation is a real "hidden\nvariable" and is in a random direction. Since you can only measure the\n*component* of polarisation in the direction of your polariser axis, you\ndon\'t have full information about the original "photon". Quantum mechanics\nimplies that you *do* have full information even then, but this has not been\nproved in real life. There are some very neat sources of "photons" around\nthese days that do seem to naturally support QM, but have they really proved\nthat they were initially of random direction? It\'s not easy to tell the\ndifference between a mixture of signals, some of one polarisation, some\northogonal, and a genuinely random set. [See\nhttp://arXiv.org/abs/quant-ph/9912082]\n\nCaroline\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Tom Trotter" <tom129@juno.com> wrote in message
news:29df3039.0407152247.ba0eee4@posting.google.co m...
> The idea of instantaneous (simultaneous) change
> *does* have a physical reality (without MWI),
> doesn't it?
>
> Consider two photons emitted from
> the same atom and correlated in polarization.
> If you learn the polarization of one of the
> photons, then you instantly know the polarization
> of the other.
Yes, but the case of interest is when the polarisation is a real "hidden
variable" and is in a random direction. Since you can only measure the
*component* of polarisation in the direction of your polariser axis, you
don't have full information about the original "photon". Quantum mechanics
implies that you *do* have full information even then, but this has not been
proved in real life. There are some very neat sources of "photons" around
these days that do seem to naturally support QM, but have they really proved
that they were initially of random direction? It's not easy to tell the
difference between a mixture of signals, some of one polarisation, some
orthogonal, and a genuinely random set. [See
http://arXiv.org/abs/http://www.arxiv.org/abs/quant-ph/9912082]
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Caroline Thompson
Jul19-04, 03:08 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"scerir" <scerir@libero.it> wrote in message\nnews:BdzJc.54288\\$c_1.1744453@twister1.l ibero.it...\n\n> Bell imposed (as far as I remember) few assumptions.\n> 1) Hidden variables, that is to say, determinism.\n\nHmmm ... Yes, but, as Clauser and Horne discovered, and expressed very\nclearly in their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful\nkind of hidden variable model for real (optical) Bell test experiments is\none in which the hidden variable set at the source is not in itself enough\nto determine (in conjuction with the detector setting) whether or not any\ngiven "photon" is detected. Other hidden variables, local to the detector\nregion, are needed. The one set at the source determines only the\n*probability* of detection.\n\n> 2) Local realism, local causality (but see point 3).\n> 3) The same "lambda" (hidden variables description) for both space-like\nseparated wings of his set-up.\n\nIf the lambda\'s are not exactly equal (or, in his original paper, opposite),\nBell\'s inequalities should still be true. The correlation in coincidence\ncounts should merely be less than it otherwise would have been.\n\n> 4) Reality is "single valued" (no MWI).\n> 5) Observers are "free", i.e. to choose observables, parameters, etc. (If\nobservers were not free, that is to say if they too were pre-determined,\ni.e. by hidden variables, Bell\'s result is completely meaningless. Hence: no\nsuper-determinism).\n>\n> ----------\n> "It has been argued that quantum mechanics is not locally causal and\ncannot be embedded in a locally causal theory. That conclusion depends on\ntreating certain experimental parameters, typically the orientations of\npolarization filters, as free variables. But it might be that this apparent\nfreedom is illusory. Perhaps experimental parameters and experimental\nresults are both consequences, or partially so, of some common hidden\nmechanism. Then the apparent non-locality could be simulated."\n> - John Bell, "Free Variables and Local Causality", \'Epistemological\nLetters\', 15, (1977)\n\nThanks for the quote. Bell was right in that QM cannot be embedded in a\nlocally causal theory, but I think he never quite understood the seriousness\nof the loopholes in the experiments. As far as experiment is concerned,\nhidden variable theories remain a possibility. There is no need to assume\nanything out of the ordinary, such as the experimenters not really being\nable to make free choices. In the sense required for the experiments, their\nchoice is completely free.\n\n[skip]\n> ... in Bell\'s own terms, the "Local Causality" condition or, in Jarrett\'s\n> terms, the "Strong Causality" condition. And it just means that the\n> joint probability of outcomes x and y, for measurements of observables\n> i and j, at the A and B wings, is equal to the product of the the\n> separate probabilities. (We know that so many experiments have shown\n> the expression above is far from reality).\n\nThis is not necessarily true. The usual tests these days are true only if\nthere is "fair sampling", and there is every reason (from the point of view\nof a realist) to think that this will not be the case. The correlation that\nis quoted has been normalised in a manner that would not, back in 1978, have\nbeen approved by Clauser and Shimony [J. F. Clauser and A. Shimony, "Bell\'s\ntheorem: experimental tests and implications", Reports on Progress in\nPhysics 41, 1881 (1978)]. They argue that the CHSH test (-2 <= S <= 2) can\nonly be used if the number of emitted pairs N is known and, one presumes,\nused in place of the total observed coincidence count. It is not commonly\nrealised that the derivation given by Clauser and Horne in 1974 of an\nequality for use in single-channel experiments shows that this inequality\ndoes not require knowledge of N. It does require the assumption of "no\nenhancement", but this seems reasonable and certainly preferable to the\nalmost certain bias of the CHSH test.\n\nCaroline\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"scerir" <scerir@libero.it> wrote in message
news:BdzJc.54288$c_1.1744453@twister1.libero.it...
> Bell imposed (as far as I remember) few assumptions.
> 1) Hidden variables, that is to say, determinism.
Hmmm ... Yes, but, as Clauser and Horne discovered, and expressed very
clearly in their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful
kind of hidden variable model for real (optical) Bell test experiments is
one in which the hidden variable set at the source is not in itself enough
to determine (in conjuction with the detector setting) whether or not any
given "photon" is detected. Other hidden variables, local to the detector
region, are needed. The one set at the source determines only the
*probability* of detection.
> 2) Local realism, local causality (but see point 3).
> 3) The same "\lambda" (hidden variables description) for both space-like
separated wings of his set-up.
If the \lambda's are not exactly equal (or, in his original paper, opposite),
Bell's inequalities should still be true. The correlation in coincidence
counts should merely be less than it otherwise would have been.
> 4) Reality is "single valued" (no MWI).
> 5) Observers are "free", i.e. to choose observables, parameters, etc. (If
observers were not free, that is to say if they too were pre-determined,
i.e. by hidden variables, Bell's result is completely meaningless. Hence: no
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)
Thanks for the quote. Bell was right in that QM cannot be embedded in a
locally causal theory, but I think he never quite understood the seriousness
of the loopholes in the experiments. As far as experiment is concerned,
hidden variable theories remain a possibility. There is no need to assume
anything out of the ordinary, such as the experimenters not really being
able to make free choices. In the sense required for the experiments, their
choice is completely free.
[skip]
> ... in Bell's own terms, the "Local Causality" condition or, in Jarrett's
> terms, the "Strong Causality" condition. And it just means that the
> joint probability of outcomes x and y, for measurements of observables
> i and j, at the A and B wings, is equal to the product of the the
> separate probabilities. (We know that so many experiments have shown
> the expression above is far from reality).
This is not necessarily true. The usual tests these days are true only if
there is "fair sampling", and there is every reason (from the point of view
of a realist) to think that this will not be the case. The correlation that
is quoted has been normalised in a manner that would not, back in 1978, have
been approved by Clauser and Shimony [J. F. Clauser and A. Shimony, "Bell's
theorem: experimental tests and implications", Reports on Progress in
Physics 41, 1881 (1978)]. They argue that the CHSH test (-2 <= S <= 2) can
only be used if the number of emitted pairs N is known and, one presumes,
used in place of the total observed coincidence count. It is not commonly
realised that the derivation given by Clauser and Horne in 1974 of an
equality for use in single-channel experiments shows that this inequality
does not require knowledge of N. It does require the assumption of "no
enhancement", but this seems reasonable and certainly preferable to the
almost certain bias of the CHSH test.
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
<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\nTom Trotter:\n\n> "Nonseparability" is, imo, a vague way of\n> referring to what *is* relevant (wrt paired\n> photons) in the combined context -- and\n> that is the *relationship* of their\n> emission-produced polarizations (ie., they\n> are polarized identically via the emission\n> process).\n\nMaybe. But it seems to me that here too (entanglement)\nis going on the old debate about the nature of "psi"\n("It" or "bit"?) and about "indistinguishability".\n\nImagine you have two atoms: A and B, situated in\ndistant locations, both in an excited state |0>.\nThese atoms may both decay to the state |1>,\ndue to spontaneous emission, producing one photon.\nA (360°, unfocused) detector is placed at half way,\nbetween the two atoms. After some time the dectector\nclicks. But we cannot distinguish from which atom\nthe detected photon came.\n\nWe have thus (apparently) produced this entangled state:\n|psi> = 2^(-1/2) [|0>_A |1>_B + e^(i phi)|1>_A |0>_B]\nwhere phi is a fixed phase.\n\nThe point here is the impossibility to determine which\natom emitted the photon. Can we produce entanglements\nnot just during emissions, but also during detections?\n\ns.\n\n"In an experiment the state reflects not what is actually\nknown about the system, but rather what is knowable,\nin principle, with the help of auxiliary measurements\nthat do not disturb the original experiment. By focusing\non what is knowable in principle, and treating what is known\nas largely irrelevant, one completely avoids the\nanthropomorphism and any reference to consciousness that\nsome physicists have tried to inject into quantum mechanics."\n- Leonard Mandel (Rev.Mod.Phys.,1999,p.S-274)\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>Tom Trotter:
> "Nonseparability" is, imo, a vague way of
> referring to what *is* relevant (wrt paired
> photons) in the combined context -- and
> that is the *relationship* of their
> emission-produced polarizations (ie., they
> are polarized identically via the emission
> process).
Maybe. But it seems to me that here too (entanglement)
is going on the old debate about the nature of "\psi"
("It" or "bit"?) and about "indistinguishability".
Imagine you have two atoms: A and B, situated in
distant locations, both in an excited state |0>.
These atoms may both decay to the state |1>,
due to spontaneous emission, producing one photon.
A (360°, unfocused) detector is placed at half way,
between the two atoms. After some time the dectector
clicks. But we cannot distinguish from which atom
the detected photon came.
We have thus (apparently) produced this entangled state:
|\psi> = 2^(-1/2) [|0>_A |1>_B + e^(i \phi)|1>_A |0>_B]
where \phi is a fixed phase.
The point here is the impossibility to determine which
atom emitted the photon. Can we produce entanglements
not just during emissions, but also during detections?
s.
"In an experiment the state reflects not what is actually
known about the system, but rather what is knowable,
in principle, with the help of auxiliary measurements
that do not disturb the original experiment. By focusing
on what is knowable in principle, and treating what is known
as largely irrelevant, one completely avoids the
anthropomorphism and any reference to consciousness that
some physicists have tried to inject into quantum mechanics."
- Leonard Mandel (Rev.Mod.Phys.,1999,p.S-274)
Paul Stewart Snyder
Jul19-04, 03:09 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\ntom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407152247.ba0eee4@posting.google.c om>...\n> ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407141925.2054c9a0@posting.google. com>...\n>\n> [... snip]\n>\n> > We have A instantaneously changing state when\n> > B does, no matter what the distance, and presumably\n> > even if no measurement can be made. I see\n> > shadows of the Copenhagen Interpretation, if we\n> > moved toward a many worlds model would the idea\n> > of instantaneous change be more likely to\n> > have a physical reality?\n>\n> [ ... snip]\n>\n> The idea of instantaneous (simultaneous) change\n> *does* have a physical reality (without MWI),\n> doesn\'t it?\n>\n> Consider two photons emitted from\n> the same atom and correlated in polarization.\n> If you learn the polarization of one of the\n> photons, then you instantly know the polarization\n> of the other.\n>\n> Consider a wheel with two marks on opposite\n> sides of it. As the wheel revolves, the\n> opposing marks simultaneously change position.\n>\n> Consider two colliding objects whose momenta\n> are known prior to their collision. In a\n> sufficiently closed/isolated setting, subsequently\n> learning the momentum of one of the objects\n> will allow you to deduce the momentum of the\n> other.\n>\n> The entities in the above examples have certain\n> relationships to each other that allow their\n> behavior to be correlated in some encompassing\n> observational context. These correlations\n> require no communication between the correlated\n> entities.\n>\n> Or, am I missing your point?\n\nIt is true that if we know that A is opposite B, whenever we determine\nA we know B, but that is not what I mean. In quantum entanglement when\nA and B are entangled, if we determine A we know B – which is nothing\nspecial – what really matters is that if we change the state of A we\ninstantaneously know that the state of B is changed. The argument is\nthat no information is transferred, which from an information theory\nview is correct, yet I still feel that there is a communication of\nsome sort that we can recognize, a transfer that we are defining away\nwith classical models. We know that when A is changed B is\ninstantaneously changed, we know that we can cause an instantaneous\nchange in B, that should count for something when we consider the\ninformation state of an observer at B (even if the only information is\nthat we know we caused the observer at B to observe something\ndifferent than he or she would otherwise have seen).\n\nI agree when you say "The assumption that no signal can exceed the\nspeed of light?\nAccording to the philosophers it is as fundamental and necessary as\nindeterminism, and viceversa." However that does not mean that the\n"deck is not stacked" with classical assumptions that require the\nconclusions. As far as I can tell (and I may be very wrong) the\nstructure of arguments against a signal exceeding the speed of light\ninclude assumptions that beg the question. Essentially relativity is\nimposed on QM in such a way as to negate the argument that meaningful\nFTL interactions are possible. The fact that quantum entanglement is\ndenied to be any kind of communication seems to be based on math and\nlogic that has built into it assumptions that FTL cannot exist?\n\nPS\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>tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407152247.ba0eee4@posting.google.com>...
> ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407141925.2054c9a0@posting.google.com>...
>
> [... snip]
>
> > We have A instantaneously changing state when
> > B does, no matter what the distance, and presumably
> > even if no measurement can be made. I see
> > shadows of the Copenhagen Interpretation, if we
> > moved toward a many worlds model would the idea
> > of instantaneous change be more likely to
> > have a physical reality?
>
> [ ... snip]
>
> The idea of instantaneous (simultaneous) change
> *does* have a physical reality (without MWI),
> doesn't it?
>
> Consider two photons emitted from
> the same atom and correlated in polarization.
> If you learn the polarization of one of the
> photons, then you instantly know the polarization
> of the other.
>
> Consider a wheel with two marks on opposite
> sides of it. As the wheel revolves, the
> opposing marks simultaneously change position.
>
> Consider two colliding objects whose momenta
> are known prior to their collision. In a
> sufficiently closed/isolated setting, subsequently
> learning the momentum of one of the objects
> will allow you to deduce the momentum of the
> other.
>
> The entities in the above examples have certain
> relationships to each other that allow their
> behavior to be correlated in some encompassing
> observational context. These correlations
> require no communication between the correlated
> entities.
>
> Or, am I missing your point?
It is true that if we know that A is opposite B, whenever we determine
A we know B, but that is not what I mean. In quantum entanglement when
A and B are entangled, if we determine A we know B – which is nothing
special – what really matters is that if we change the state of A we
instantaneously know that the state of B is changed. The argument is
that no information is transferred, which from an information theory
view is correct, yet I still feel that there is a communication of
some sort that we can recognize, a transfer that we are defining away
with classical models. We know that when A is changed B is
instantaneously changed, we know that we can cause an instantaneous
change in B, that should count for something when we consider the
information state of an observer at B (even if the only information is
that we know we caused the observer at B to observe something
different than he or she would otherwise have seen).
I agree when you say "The assumption that no signal can exceed the
speed of light?
According to the philosophers it is as fundamental and necessary as
indeterminism, and viceversa." However that does not mean that the
"deck is not stacked" with classical assumptions that require the
conclusions. As far as I can tell (and I may be very wrong) the
structure of arguments against a signal exceeding the speed of light
include assumptions that beg the question. Essentially relativity is
imposed on QM in such a way as to negate the argument that meaningful
FTL interactions are possible. The fact that quantum entanglement is
denied to be any kind of communication seems to be based on math and
logic that has built into it assumptions that FTL cannot exist?
PS
Tom Trotter
Jul20-04, 03:25 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"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<zxZJc.103\\$wq5.95@newsfe3-gui.ntli.net>...\n> "scerir" <scerir@libero.it> wrote in message\n> news:BdzJc.54288\\$c_1.1744453@twister1.libero.it. ..\n>\n> > Bell imposed (as far as I remember) few assumptions.\n> > 1) Hidden variables, that is to say, determinism.\n>\n> Hmmm ... Yes, but, as Clauser and Horne discovered, and expressed very\n> clearly in their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful\n> kind of hidden variable model for real (optical) Bell test experiments is\n> one in which the hidden variable set at the source is not in itself enough\n> to determine (in conjuction with the detector setting) whether or not any\n> given "photon" is detected. Other hidden variables, local to the detector\n> region, are needed. The one set at the source determines only the\n> *probability* of detection.\n\nCaroline, I appreciate your work, but I think you\'ve\n(along with many commentators) overcomplicated the problem.\nBell-type inequalities are simply inapplicable (wrt\nthe usual considerations of locality, reality, hidden\nvariables, etc.) to the combined contexts exemplified by\nBell-type experiments.\n\nThe formulations on which the inequalities are based contain\na variable, lambda (representing, in effect, the polarization\nof the photons), which is irrelevant wrt the determination of\ncoincidental detection.\n\nThis doesn\'t mean that lambda doesn\'t exist. Indeed, as Bell\nacknowledged, if lambda were known, then the results of individual\nmeasurements could be more accurately predicted.\n\nBut, Bell-type experiments aren\'t individual contexts,\nthey\'re combined contexts, and the thing that\nproduces the correlations between the data streams in\ncombined contexts isn\'t the same thing that would produce\nmore accurate predictions of individual measurements.\n\nIn the combined context, it\'s the *relationship* between\nthe photons of a pair that matters. The assumption in the\nqm formulation is that they\'re polarized identically via\nemission, based on an emission model, and this is supported\nby experiments.\n\n>\n> > 2) Local realism, local causality (but see point 3).\n> > 3) The same "lambda" (hidden variables description) for both space-like\n> separated wings of his set-up.\n>\n> If the lambda\'s are not exactly equal (or, in his original paper, opposite),\n> Bell\'s inequalities should still be true. The correlation in coincidence\n> counts should merely be less than it otherwise would have been.\n\nIf the lambdas aren\'t exactly equal via emission, the photons\nof a pair can still be correlated, if the *relationship*\nbetween them doesn\'t change from emission to detection.\nAnd, as the lambdas diverge, then rates of coincidental\ndetection should decrease.\n\nBut the emission model says that photons emitted in opposite\ndirections (and these are the ones chosen for eventual\nmutual analysis by the polarizers) must be polarized identically\ndue to conservation of angular momentum. And, experiments\nseem to me to support this model.\n\nAs for qm being a \'nonlocal\' theory, yes, insofar as it\'s\ndealing with combined contexts rather than individual ones.\nBell tests are, by definition, nonlocal experimental contexts.\n\nNote that none of this has anything to do with the existence\n(or non-existence) of hidden variables, or signals moving\nfaster than light.\n\nBell tests are superfluous wrt these considerations. The only\nthing that results of Bell tests will give you is some idea wrt\ndegree of entanglement.\n\nThe subtle assumption that causes the confusion surrounding\nthese issues is that lambda is a factor that\'s relevant wrt\nthe determination of coincidental detection in combined\ncontexts. If you omit lambda from the formulation (and, note\nthat this doesn\'t mean that it doesn\'t exist), then you\'re\nleft with one relevant variable, the angular difference of\nthe polarizer settings, and, thus, the qm formulation.\n\n[... snip]\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>"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<zxZJc.103$wq5.95@newsfe3-gui.ntli.net>...
> "scerir" <scerir@libero.it> wrote in message
> news:BdzJc.54288$c_1.1744453@twister1.libero.it...
>
> > Bell imposed (as far as I remember) few assumptions.
> > 1) Hidden variables, that is to say, determinism.
>
> Hmmm ... Yes, but, as Clauser and Horne discovered, and expressed very
> clearly in their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful
> kind of hidden variable model for real (optical) Bell test experiments is
> one in which the hidden variable set at the source is not in itself enough
> to determine (in conjuction with the detector setting) whether or not any
> given "photon" is detected. Other hidden variables, local to the detector
> region, are needed. The one set at the source determines only the
> *probability* of detection.
Caroline, I appreciate your work, but I think you've
(along with many commentators) overcomplicated the problem.
Bell-type inequalities are simply inapplicable (wrt
the usual considerations of locality, reality, hidden
variables, etc.) to the combined contexts exemplified by
Bell-type experiments.
The formulations on which the inequalities are based contain
a variable, \lambda (representing, in effect, the polarization
of the photons), which is irrelevant wrt the determination of
coincidental detection.
This doesn't mean that \lambda doesn't exist. Indeed, as Bell
acknowledged, if \lambda were known, then the results of individual
measurements could be more accurately predicted.
But, Bell-type experiments aren't individual contexts,
they're combined contexts, and the thing that
produces the correlations between the data streams in
combined contexts isn't the same thing that would produce
more accurate predictions of individual measurements.
In the combined context, it's the *relationship* between
the photons of a pair that matters. The assumption in the
qm formulation is that they're polarized identically via
emission, based on an emission model, and this is supported
by experiments.
>
> > 2) Local realism, local causality (but see point 3).
> > 3) The same "\lambda" (hidden variables description) for both space-like
> separated wings of his set-up.
>
> If the \lambda's are not exactly equal (or, in his original paper, opposite),
> Bell's inequalities should still be true. The correlation in coincidence
> counts should merely be less than it otherwise would have been.
If the lambdas aren't exactly equal via emission, the photons
of a pair can still be correlated, if the *relationship*
between them doesn't change from emission to detection.
And, as the lambdas diverge, then rates of coincidental
detection should decrease.
But the emission model says that photons emitted in opposite
directions (and these are the ones chosen for eventual
mutual analysis by the polarizers) must be polarized identically
due to conservation of angular momentum. And, experiments
seem to me to support this model.
As for qm being a 'nonlocal' theory, yes, insofar as it's
dealing with combined contexts rather than individual ones.
Bell tests are, by definition, nonlocal experimental contexts.
Note that none of this has anything to do with the existence
(or non-existence) of hidden variables, or signals moving
faster than light.
Bell tests are superfluous wrt these considerations. The only
thing that results of Bell tests will give you is some idea wrt
degree of entanglement.
The subtle assumption that causes the confusion surrounding
these issues is that \lambda is a factor that's relevant wrt
the determination of coincidental detection in combined
contexts. If you omit \lambda from the formulation (and, note
that this doesn't mean that it doesn't exist), then you're
left with one relevant variable, the angular difference of
the polarizer settings, and, thus, the qm formulation.
[... snip]
Tom Trotter
Jul20-04, 03:25 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>\nps@ws5.com (Paul Stewart Snyder) wrote in message\n> It is true that if we know that A is opposite B, whenever we determine\n> A we know B, but that is not what I mean. In quantum entanglement when\n> A and B are entangled, if we determine A we know B ? which is nothing\n> special ? what really matters is that if we change the state of A we\n> instantaneously know that the state of B is changed.\n\nNo, it isn\'t always known that if A is changed, then\nB is changed. It depends on the experimental context.\nIn Aspect-type setups the changes at A and B are\nindependent of each other. The detections of paired photons\nare correlated because they\'re being analyzed with polarizers,\nand they\'re identically polarized via emission.\n\nThe relationship between paired photons doesn\'t change\nin this observational context, no matter when you change\na polarizer setting.\n\n> The argument is that no information is transferred,\n> which from an information theory view is correct,\n\nThat no information is transferred between correlated\nentities, A and B, is true in any (correct) view or experimental\ncontext.\n\n> yet I still feel that there is a communication of\n> some sort that we can recognize, a transfer that we are\n> defining away with classical models.\n\nNope. It\'s just not happening. And, the only reason\nit was considered in the first place was because of\nmisinterpretations of EPR-Bell.\n\n> We know that when A is changed B is instantaneously\n> changed,\n\nOnly in certain contexts, but they don\'t imply signal\ntransfer between A and B.\n\n> we know that we can cause an instantaneous change in B,\n> that should count for something when we consider the\n> information state of an observer at B (even if the only\n> information is that we know we caused the observer at B\n> to observe something different than he or she would\n> otherwise have seen).\n\nIn Aspect-type experiments, A and B are effectively\nisolated from each other. In the context of individual\nmeasurement, the data streams at A and B are independent\nof each other. In the combined context the data streams\nare correlated wrt some circular function of the angular\ndifference of the polarizers that are analyzing the\nphoton pairs, because the photons of any given pair\nare related via identical emission polarization -- and\nthis relationship doesn\'t vary from pair to pair.\n\n>\n> I agree when you say "The assumption that no signal can exceed the\n> speed of light?\n> According to the philosophers it is as fundamental and necessary as\n> indeterminism, and viceversa." However that does not mean that the\n> "deck is not stacked" with classical assumptions that require the\n> conclusions. As far as I can tell (and I may be very wrong) the\n> structure of arguments against a signal exceeding the speed of light\n> include assumptions that beg the question.\n\nThe arguments *for* ftl or instantaneous signalling are\nbased on misconceptions. There\'s nothing in the current\nliterature, afaik, that requires ftl signalling -- at least\nwrt EPR-Bell stuff.\n\n> Essentially relativity is imposed on QM in such a way as to\n> negate the argument that meaningful FTL interactions are possible.\n\nI don\'t know if ftl interactions are possible. But the point\nis that they\'re not necessary to explain the data. So, the\ndefault position is the current formulation of standard physics,\nwhich entails Lorentz invariance, afaik.\n\n> The fact that quantum entanglement is denied to be any kind\n> of communication seems to be based on math and\n> logic that has built into it assumptions that FTL\n> cannot exist?\n\nThat\'s the assumption, and it seems to be holding up ok so far.\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>ps@ws5.com (Paul Stewart Snyder) wrote in message
> It is true that if we know that A is opposite B, whenever we determine
> A we know B, but that is not what I mean. In quantum entanglement when
> A and B are entangled, if we determine A we know B ? which is nothing
> special ? what really matters is that if we change the state of A we
> instantaneously know that the state of B is changed.
No, it isn't always known that if A is changed, then
B is changed. It depends on the experimental context.
In Aspect-type setups the changes at A and B are
independent of each other. The detections of paired photons
are correlated because they're being analyzed with polarizers,
and they're identically polarized via emission.
The relationship between paired photons doesn't change
in this observational context, no matter when you change
a polarizer setting.
> The argument is that no information is transferred,
> which from an information theory view is correct,
That no information is transferred between correlated
entities, A and B, is true in any (correct) view or experimental
context.
> yet I still feel that there is a communication of
> some sort that we can recognize, a transfer that we are
> defining away with classical models.
Nope. It's just not happening. And, the only reason
it was considered in the first place was because of
misinterpretations of EPR-Bell.
> We know that when A is changed B is instantaneously
> changed,
Only in certain contexts, but they don't imply signal
transfer between A and B.
> we know that we can cause an instantaneous change in B,
> that should count for something when we consider the
> information state of an observer at B (even if the only
> information is that we know we caused the observer at B
> to observe something different than he or she would
> otherwise have seen).
In Aspect-type experiments, A and B are effectively
isolated from each other. In the context of individual
measurement, the data streams at A and B are independent
of each other. In the combined context the data streams
are correlated wrt some circular function of the angular
difference of the polarizers that are analyzing the
photon pairs, because the photons of any given pair
are related via identical emission polarization -- and
this relationship doesn't vary from pair to pair.
>
> I agree when you say "The assumption that no signal can exceed the
> speed of light?
> According to the philosophers it is as fundamental and necessary as
> indeterminism, and viceversa." However that does not mean that the
> "deck is not stacked" with classical assumptions that require the
> conclusions. As far as I can tell (and I may be very wrong) the
> structure of arguments against a signal exceeding the speed of light
> include assumptions that beg the question.
The arguments *for* ftl or instantaneous signalling are
based on misconceptions. There's nothing in the current
literature, afaik, that requires ftl signalling -- at least
wrt EPR-Bell stuff.
> Essentially relativity is imposed on QM in such a way as to
> negate the argument that meaningful FTL interactions are possible.
I don't know if ftl interactions are possible. But the point
is that they're not necessary to explain the data. So, the
default position is the current formulation of standard physics,
which entails Lorentz invariance, afaik.
> The fact that quantum entanglement is denied to be any kind
> of communication seems to be based on math and
> logic that has built into it assumptions that FTL
> cannot exist?
That's the assumption, and it seems to be holding up ok so far.
Tom Trotter
Jul20-04, 03:25 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"scerir" <scerir@libero.it> wrote in message news:<PbWJc.54264\\$GQ3.1515854@twister2.libero.it >...\n> Tom Trotter:\n>\n> > "Nonseparability" is, imo, a vague way of\n> > referring to what *is* relevant (wrt paired\n> > photons) in the combined context -- and\n> > that is the *relationship* of their\n> > emission-produced polarizations (ie., they\n> > are polarized identically via the emission\n> > process).\n>\n> Maybe. But it seems to me that here too (entanglement)\n> is going on the old debate about the nature of "psi"\n> ("It" or "bit"?) and about "indistinguishability".\n>\n> Imagine you have two atoms: A and B, situated in\n> distant locations, both in an excited state |0>.\n> These atoms may both decay to the state |1>,\n> due to spontaneous emission, producing one photon.\n> A (360°, unfocused) detector is placed at half way,\n> between the two atoms. After some time the dectector\n> clicks. But we cannot distinguish from which atom\n> the detected photon came.\n>\n> We have thus (apparently) produced this entangled state:\n> |psi> = 2^(-1/2) [|0>_A |1>_B + e^(i phi)|1>_A |0>_B]\n> where phi is a fixed phase.\n>\n> The point here is the impossibility to determine which\n> atom emitted the photon. Can we produce entanglements\n> not just during emissions, but also during detections?\n>\n\nAs I currently see it, here are three underlying causes\nof entanglement. The entangled entities:\n(1) have a common origin.\nor\n(2) have interacted.\nor\n(3) are being observed in the context of the motion\nof a system that includes them both\n\nEntanglement involves some fixed quantity that\nrepresents an unchanging physical relationship\nbetween the correlated entities.\n\nHow are the photons in your above example entangled?\nHow are they correlated? Istm that just not being\nable to determine which atom emitted the photon\nresponsible for a detector click isn\'t really an\nexample of entanglement.\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>"scerir" <scerir@libero.it> wrote in message news:<PbWJc.54264$GQ3.1515854@twister2.libero.it>...
> Tom Trotter:
>
> > "Nonseparability" is, imo, a vague way of
> > referring to what *is* relevant (wrt paired
> > photons) in the combined context -- and
> > that is the *relationship* of their
> > emission-produced polarizations (ie., they
> > are polarized identically via the emission
> > process).
>
> Maybe. But it seems to me that here too (entanglement)
> is going on the old debate about the nature of "\psi"
> ("It" or "bit"?) and about "indistinguishability".
>
> Imagine you have two atoms: A and B, situated in
> distant locations, both in an excited state |0>.
> These atoms may both decay to the state |1>,
> due to spontaneous emission, producing one photon.
> A (360°, unfocused) detector is placed at half way,
> between the two atoms. After some time the dectector
> clicks. But we cannot distinguish from which atom
> the detected photon came.
>
> We have thus (apparently) produced this entangled state:
> |\psi> = 2^(-1/2) [|0>_A |1>_B + e^(i \phi)|1>_A |0>_B]
> where \phi is a fixed phase.
>
> The point here is the impossibility to determine which
> atom emitted the photon. Can we produce entanglements
> not just during emissions, but also during detections?
>
As I currently see it, here are three underlying causes
of entanglement. The entangled entities:
(1) have a common origin.
or
(2) have interacted.
or
(3) are being observed in the context of the motion
of a system that includes them both
Entanglement involves some fixed quantity that
represents an unchanging physical relationship
between the correlated entities.
How are the photons in your above example entangled?
How are they correlated? Istm that just not being
able to determine which atom emitted the photon
responsible for a detector click isn't really an
example of entanglement.
<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>\nTom Trotter asked:\n\n> > We have thus (apparently) produced this entangled state:\n> > |psi> = 2^(-1/2) [|0>_A |1>_B + e^(i phi)|1>_A |0>_B]\n> > where phi is a fixed phase.\n\n> How are the photons in your above example entangled?\n> How are they correlated?\n\nNot the photon(s), the atoms are entangled\nhttp://www.arxiv.org/abs/quant-ph/0205182\nand this seems even more strange than the\nusual entanglement.\nRegards,\ns.\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>Tom Trotter asked:
> > We have thus (apparently) produced this entangled state:
> > |\psi> = 2^(-1/2) [|0>_A |1>_B + e^(i \phi)|1>_A |0>_B]
> > where \phi is a fixed phase.
> How are the photons in your above example entangled?
> How are they correlated?
Not the photon(s), the atoms are entangled
http://www.arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/0205182
and this seems even more strange than the
usual entanglement.
Regards,
s.
Caroline Thompson
Jul20-04, 09:49 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\n"Tom Trotter" <tom129@juno.com> wrote in message\nnews:29df3039.0407191115.539d5a7a@posting .google.com...\n>\n> "Caroline Thompson" <ch.thompson1@virgin.net> wrote\n\n> > ... as Clauser and Horne discovered, and expressed very clearly in\n> > their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful\n> > kind of hidden variable model for real (optical) Bell test experiments\nis\n> > one in which the hidden variable set at the source is not in itself\nenough\n> > to determine (in conjuction with the detector setting) whether or not\n> > any given "photon" is detected. Other hidden variables, local to the\n> > detector region, are needed. The one set at the source determines only\n> > the *probability* of detection.\n>\n> Caroline, I appreciate your work, but I think you\'ve\n> (along with many commentators) overcomplicated the problem.\n> Bell-type inequalities are simply inapplicable (wrt\n> the usual considerations of locality, reality, hidden\n> variables, etc.) to the combined contexts exemplified by\n> Bell-type experiments.\n\nAgreed, Bell\'s original (1964) inequality is inapplicable, but the CHSH one\n*would* be applicable if the fair sampling assumption were correct (and\nother experimental flaws not present!) and (with the same proviso) the CH74\none is almost certain to be applicable (since nobody can think of any real\nphysical reason why there should be "enhancement", the presence of a\npolarisers increasing the probability of detection for certain values of\nlambda).\n\n> The formulations on which the inequalities are based contain\n> a variable, lambda (representing, in effect, the polarization\n> of the photons), which is irrelevant wrt the determination of\n> coincidental detection.\n\nHow can you say this? The difference between lamba and the polariser\nsettings are the *only* parameters available that could possibly control the\ncoincidences. As I\'ve said before, they only partially control it, but, as\nevidence by the observed high "correlations", this is enough.\n\n[Incidentally, all this discussion is done on the assumption that the idea\nof "quantum correlation" makes sense and that Bell inequalities have to be\ncouched in terms of estimates of it. The CH74 inequality does not in fact\nrequire the quantum correlation concept. It relies solely on\nstraightforward probabilities. And by looking only at \'+\' results, not at\nboth \'+\' and \'-\', it avoids an otherwise awkward situation that can occur in\nreal experiments, when both \'+\' and \'-\' detections are made from the same\npolariser. "Quantum correlation" is not even defined in such cases, as far\nas I know. [See my new paper,\nhttp://freespace.virgin.net/ch.thompson1/Papers/CH74/CH74assumptions.htm ]]\n\n> This doesn\'t mean that lambda doesn\'t exist. Indeed, as Bell\n> acknowledged, if lambda were known, then the results of\n> individual measurements could be more accurately predicted.\n>\n> But, Bell-type experiments aren\'t individual contexts,\n> they\'re combined contexts, and the thing that\n> produces the correlations between the data streams in\n> combined contexts isn\'t the same thing that would produce\n> more accurate predictions of individual measurements.\n\nWhat is it, then?\n\n> In the combined context, it\'s the *relationship* between\n> the photons of a pair that matters.\n\nHow can you have a relationship without both individuals having their own\nattributes?\n\n> ... the emission model says that photons emitted in opposite\n> directions (and these are the ones chosen for eventual\n> mutual analysis by the polarizers) must be polarized identically\n> due to conservation of angular momentum. And, experiments\n> seem to me to support this model.\n\nOK. This does not conflict with local realism.\n\n> As for qm being a \'nonlocal\' theory, yes, insofar as it\'s\n> dealing with combined contexts rather than individual ones.\n> Bell tests are, by definition, nonlocal experimental contexts.\n\nHow so? They are perfectly ordinary lab experiments. What is a "nonlocal\nexperimental context"?\n\n> Note that none of this has anything to do with the existence\n> (or non-existence) of hidden variables, or signals moving\n> faster than light.\n>\n> Bell tests are superfluous wrt these considerations. The only\n> thing that results of Bell tests will give you is some idea wrt\n> degree of entanglement.\n\nAs a convinced local realist, this is not my problem. This represents a\nvast overcomplication on the part of quantum theorists.\n\n> The subtle assumption that causes the confusion surrounding\n> these issues is that lambda is a factor that\'s relevant wrt\n> the determination of coincidental detection in combined\n> contexts. If you omit lambda from the formulation (and, note\n> that this doesn\'t mean that it doesn\'t exist), then you\'re\n> left with one relevant variable, the angular difference of\n> the polarizer settings, and, thus, the qm formulation.\n\nAnd if they were to look just a little bit harder at the evidence they\'d\nfind that this was not enough! Especially in situations where the source is\nnot rotationally invariant, which includes, as far as I can tell, all\nexperiments using parametric down conversion sources. Check how the Bell\ntests are applied. They don\'t often use the simplified version that applies\nwhen the source *is* rotationally invariant.\n\nCaroline\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>"Tom Trotter" <tom129@juno.com> wrote in message
news:29df3039.0407191115.539d5a7a@posting.google.c om...
>
> "Caroline Thompson" <ch.thompson1@virgin.net> wrote
> > ... as Clauser and Horne discovered, and expressed very clearly in
> > their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful
> > kind of hidden variable model for real (optical) Bell test experiments
is
> > one in which the hidden variable set at the source is not in itself
enough
> > to determine (in conjuction with the detector setting) whether or not
> > any given "photon" is detected. Other hidden variables, local to the
> > detector region, are needed. The one set at the source determines only
> > the *probability* of detection.
>
> Caroline, I appreciate your work, but I think you've
> (along with many commentators) overcomplicated the problem.
> Bell-type inequalities are simply inapplicable (wrt
> the usual considerations of locality, reality, hidden
> variables, etc.) to the combined contexts exemplified by
> Bell-type experiments.
Agreed, Bell's original (1964) inequality is inapplicable, but the CHSH one
*would* be applicable if the fair sampling assumption were correct (and
other experimental flaws not present!) and (with the same proviso) the CH74
one is almost certain to be applicable (since nobody can think of any real
physical reason why there should be "enhancement", the presence of a
polarisers increasing the probability of detection for certain values of
\lambda).
> The formulations on which the inequalities are based contain
> a variable, \lambda (representing, in effect, the polarization
> of the photons), which is irrelevant wrt the determination of
> coincidental detection.
How can you say this? The difference between lamba and the polariser
settings are the *only* parameters available that could possibly control the
coincidences. As I've said before, they only partially control it, but, as
evidence by the observed high "correlations", this is enough.
[Incidentally, all this discussion is done on the assumption that the idea
of "quantum correlation" makes sense and that Bell inequalities have to be
couched in terms of estimates of it. The CH74 inequality does not in fact
require the quantum correlation concept. It relies solely on
straightforward probabilities. And by looking only at '+' results, not at
both '+' and '-', it avoids an otherwise awkward situation that can occur in
real experiments, when both '+' and '-' detections are made from the same
polariser. "Quantum correlation" is not even defined in such cases, as far
as I know. [See my new paper,
http://freespace.virgin.net/ch.thompson1/Papers/CH74/CH74assumptions.htm ]]
> This doesn't mean that \lambda doesn't exist. Indeed, as Bell
> acknowledged, if \lambda were known, then the results of
> individual measurements could be more accurately predicted.
>
> But, Bell-type experiments aren't individual contexts,
> they're combined contexts, and the thing that
> produces the correlations between the data streams in
> combined contexts isn't the same thing that would produce
> more accurate predictions of individual measurements.
What is it, then?
> In the combined context, it's the *relationship* between
> the photons of a pair that matters.
How can you have a relationship without both individuals having their own
attributes?
> ... the emission model says that photons emitted in opposite
> directions (and these are the ones chosen for eventual
> mutual analysis by the polarizers) must be polarized identically
> due to conservation of angular momentum. And, experiments
> seem to me to support this model.
OK. This does not conflict with local realism.
> As for qm being a 'nonlocal' theory, yes, insofar as it's
> dealing with combined contexts rather than individual ones.
> Bell tests are, by definition, nonlocal experimental contexts.
How so? They are perfectly ordinary lab experiments. What is a "nonlocal
experimental context"?
> Note that none of this has anything to do with the existence
> (or non-existence) of hidden variables, or signals moving
> faster than light.
>
> Bell tests are superfluous wrt these considerations. The only
> thing that results of Bell tests will give you is some idea wrt
> degree of entanglement.
As a convinced local realist, this is not my problem. This represents a
vast overcomplication on the part of quantum theorists.
> The subtle assumption that causes the confusion surrounding
> these issues is that \lambda is a factor that's relevant wrt
> the determination of coincidental detection in combined
> contexts. If you omit \lambda from the formulation (and, note
> that this doesn't mean that it doesn't exist), then you're
> left with one relevant variable, the angular difference of
> the polarizer settings, and, thus, the qm formulation.
And if they were to look just a little bit harder at the evidence they'd
find that this was not enough! Especially in situations where the source is
not rotationally invariant, which includes, as far as I can tell, all
experiments using parametric down conversion sources. Check how the Bell
tests are applied. They don't often use the simplified version that applies
when the source *is* rotationally invariant.
Caroline
Tom Trotter
Jul21-04, 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\n"scerir" <scerir@libero.it> wrote in message news:<9o8Lc.1622\\$1V3.44751@twister2.libero.it>.. .\n> Tom Trotter asked:\n>\n> > > We have thus (apparently) produced this entangled state:\n> > > |psi> = 2^(-1/2) [|0>_A |1>_B + e^(i phi)|1>_A |0>_B]\n> > > where phi is a fixed phase.\n>\n> > How are the photons in your above example entangled?\n> > How are they correlated?\n>\n> Not the photon(s), the atoms are entangled\n> http://www.arxiv.org/abs/quant-ph/0205182\n> and this seems even more strange than the\n> usual entanglement.\n> Regards,\n> s.\n\nAh yes, another addition to "the future did it"\nschool of thought. Note that the term, BS, appears\nfrequently in the paper (the authors are of course\nreferring to "beam splitter", but I\'m thinking of\nthe term in another sense).\n\nIf you think that their discussion is illuminating,\nrather than obfuscating, then you\'ll have to\nexplain to me why. I don\'t think it amounts\nto much.\n\nOn the other hand, atomic emissions might\nperhaps be correlated (in a more meaningful way)\nwrt some, more encompassing, context --\nlike, say, configurational changes on a\ncosmic scale (ultimately wrt the isotropic\nexpansion maybe?). Now, I think that question\nis very interesting.\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>"scerir" <scerir@libero.it> wrote in message news:<9o8Lc.1622$1V3.44751@twister2.libero.it>...
> Tom Trotter asked:
>
> > > We have thus (apparently) produced this entangled state:
> > > |\psi> = 2^(-1/2) [|0>_A |1>_B + e^(i \phi)|1>_A |0>_B]
> > > where \phi is a fixed phase.
>
> > How are the photons in your above example entangled?
> > How are they correlated?
>
> Not the photon(s), the atoms are entangled
> http://www.arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/0205182
> and this seems even more strange than the
> usual entanglement.
> Regards,
> s.
Ah yes, another addition to "the future did it"
school of thought. Note that the term, BS, appears
frequently in the paper (the authors are of course
referring to "beam splitter", but I'm thinking of
the term in another sense).
If you think that their discussion is illuminating,
rather than obfuscating, then you'll have to
explain to me why. I don't think it amounts
to much.
On the other hand, atomic emissions might
perhaps be correlated (in a more meaningful way)
wrt some, more encompassing, context --
like, say, configurational changes on a
cosmic scale (ultimately wrt the isotropic
expansion maybe?). Now, I think that question
is very interesting.
Tom Trotter
Jul21-04, 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\n\n"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<S%8Lc.131\\$2i.87@newsfe2-gui.ntli.net>...\n> "Tom Trotter" <tom129@juno.com> wrote in message\n> news:29df3039.0407191115.539d5a7a@posting.google.c om...\n> >\n> > "Caroline Thompson" <ch.thompson1@virgin.net> wrote\n>\n> > > ... as Clauser and Horne discovered, and expressed very clearly in\n> > > their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful\n> > > kind of hidden variable model for real (optical) Bell test experiments\n> is\n> > > one in which the hidden variable set at the source is not in itself\n> enough\n> > > to determine (in conjuction with the detector setting) whether or not\n> > > any given "photon" is detected. Other hidden variables, local to the\n> > > detector region, are needed. The one set at the source determines only\n> > > the *probability* of detection.\n> >\n> > Caroline, I appreciate your work, but I think you\'ve\n> > (along with many commentators) overcomplicated the problem.\n> > Bell-type inequalities are simply inapplicable (wrt\n> > the usual considerations of locality, reality, hidden\n> > variables, etc.) to the combined contexts exemplified by\n> > Bell-type experiments.\n>\n> Agreed, Bell\'s original (1964) inequality is inapplicable, but the CHSH one\n> *would* be applicable if the fair sampling assumption were correct (and\n> other experimental flaws not present!) and (with the same proviso) the CH74\n> one is almost certain to be applicable (since nobody can think of any real\n> physical reason why there should be "enhancement", the presence of a\n> polarisers increasing the probability of detection for certain values of\n> lambda).\n>\n> > The formulations on which the inequalities are based contain\n> > a variable, lambda (representing, in effect, the polarization\n> > of the photons), which is irrelevant wrt the determination of\n> > coincidental detection.\n>\n> How can you say this? The difference between lamba and the polariser\n> settings are the *only* parameters available that could possibly control the\n> coincidences.\n\nI think that\'s incorrect. There\'s another parameter that\nseems to have gotten lost in the mix of discussions about\nthis stuff -- the *relationship* between photon 1 and\nphoton 2 of any given pair.\n\nLambda and the polarizer settings are the effective\ndeterminers of individual results. But we\'re concerned\nwith the combined context, and in this context the relevant\nfactor is the relationship between paired photons. This\nrelationship is not lambda. This relationship refers to\nthe theoretical determination that opposite moving photons\nemitted from the same atom must be polarized identically\nvia emission. This relationship doesn\'t vary from pair to\npair as lambda does. You don\'t need to know anything about\nthe varying specific polarization of pairs of opposite moving\nphotons in order to predict rates of coincidental detection.\nAs long as photon 1 and photon 2 of any given pair are\nidentically polarized, then the orientation of the polarizers\nwrt each other is the only relevant variable in the\nobservational context, and a circular function of this\nchanging angular difference is what determines rates of\ncoincidental detection.\n\n> As I\'ve said before, they only partially control it, but, as\n> evidence by the observed high "correlations", this is enough.\n>\n> [Incidentally, all this discussion is done on the assumption that the idea\n> of "quantum correlation" makes sense and that Bell inequalities have to be\n> couched in terms of estimates of it. The CH74 inequality does not in fact\n> require the quantum correlation concept. It relies solely on\n> straightforward probabilities. And by looking only at \'+\' results, not at\n> both \'+\' and \'-\', it avoids an otherwise awkward situation that can occur in\n> real experiments, when both \'+\' and \'-\' detections are made from the same\n> polariser. "Quantum correlation" is not even defined in such cases, as far\n> as I know. [See my new paper,\n> [ http://freespace.virgin.net/ch.thompson1/Papers/CH74/CH74assumptions.htm ]]\n>\n\nOk, I will. But I think you\'re barking up the wrong tree.\n(Not that your concerns wrt experimental refinements aren\'t\nrelevant in a broader sense.)\n\n> > This doesn\'t mean that lambda doesn\'t exist. Indeed, as Bell\n> > acknowledged, if lambda were known, then the results of\n> > individual measurements could be more accurately predicted.\n> >\n> > But, Bell-type experiments aren\'t individual contexts,\n> > they\'re combined contexts, and the thing that\n> > produces the correlations between the data streams in\n> > combined contexts isn\'t the same thing that would produce\n> > more accurate predictions of individual measurements.\n>\n> What is it, then?\n>\n> > In the combined context, it\'s the *relationship* between\n> > the photons of a pair that matters.\n>\n> How can you have a relationship without both individuals\n> having their own attributes?\n\nThey do have their own attributes. They have a specific\nwavelength, direction, and polarization. But wrt calculating\naccurate expectation values of rates of coincidental detection,\nthe only attributes that are important are the ones they\nhave *in common.* Afaik, they have only one thing in common,\nthey\'re *polarized identically* via the emission process -- which\nis all you need to know (along with the polarizer settings) to\nmake accurate predictions regarding coincidental detection.\n\nEven if you separate detections and changes in polarizer\nsettings by a spacelike interval, even if you change polarizer\nsettings in mid-flight, even if A has registered a detection and\nyou then change B\'s polarizer setting -- for any given pair\nof photons there\'s still only one Theta. And, so long as\nphoton 1 and photon 2 of any given pair are identically\npolarized via emission, then, as Theta increases, then rate\nof coincidental detection must decrease, and as Theta decreases,\nthen rate of coincidental detection must increase, as\na circular function of Theta.\n\nThe experimental evidence for this view is pretty good, I think.\n\n>\n> > ... the emission model says that photons emitted in opposite\n> > directions (and these are the ones chosen for eventual\n> > mutual analysis by the polarizers) must be polarized identically\n> > due to conservation of angular momentum. And, experiments\n> > seem to me to support this model.\n>\n> OK. This does not conflict with local realism.\n>\n\nAs I\'ve previously noted, there\'s nothing in the qm formulation\nthat conflicts with local realism (in the sense that local refers\nto a limiting speed for signals, and that events at the A-end\nof experiments are effectively isolated from, and independent of,\nevents at the B-end.\n\nIt\'s just that interpretations of the formulation have become\nmuddled following misinterpretations of Bell\'s work.\n\n> > As for qm being a \'nonlocal\' theory, yes, insofar as it\'s\n> > dealing with combined contexts rather than individual ones.\n> > Bell tests are, by definition, nonlocal experimental contexts.\n>\n> How so? They are perfectly ordinary lab experiments. What is a "nonlocal\n> experimental context"?\n\nCorrelated events at A and B are spacelike separated. Separating\nevents at A and B by a spacelike interval was necessary in order\nto experimentally rule out any communication between the\nthe correlated photons.\n\n>\n> > Note that none of this has anything to do with the existence\n> > (or non-existence) of hidden variables, or signals moving\n> > faster than light.\n> >\n> > Bell tests are superfluous wrt these considerations. The only\n> > thing that results of Bell tests will give you is some idea wrt\n> > degree of entanglement.\n>\n> As a convinced local realist, this is not my problem. This\n> represents a vast overcomplication on the part of quantum theorists.\n>\n\nI agree, at least wrt *interpretations* of quantum theory.\n\n> > The subtle assumption that causes the confusion surrounding\n> > these issues is that lambda is a factor that\'s relevant wrt\n> > the determination of coincidental detection in combined\n> > contexts. If you omit lambda from the formulation (and, note\n> > that this doesn\'t mean that it doesn\'t exist), then you\'re\n> > left with one relevant variable, the angular difference of\n> > the polarizer settings, and, thus, the qm formulation.\n>\n> And if they were to look just a little bit harder at the\n> evidence they\'d find that this was not enough!\n\nBut it is enough, as long as experimenters are making sure\nthat they\'re analyzing pairs of photons correlated in polarization\nvia the emission process. And, I think they make sufficient\nefforts to ensure that this is what they\'re doing.\n\n> Especially in situations where the source is\n> not rotationally invariant, which includes, as far as I can tell, all\n> experiments using parametric down conversion sources. Check how the Bell\n> tests are applied. They don\'t often use the simplified version that applies\n> when the source *is* rotationally invariant.\n\nI don\'t know about pdc, I\'ve so far only studied the Aspect\nexperiments, atomic calcium cascades, and Bell\'s papers,\nbut the principle should be the same no matter what the\nsource.\n\nTake a while to think about the *relationship* of photons\nthat are correlated in polarization,as opposed to their varying\nlambdas, and what this means wrt different experimental contexts.\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>"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<S%8Lc.131$2i.87@newsfe2-gui.ntli.net>...
> "Tom Trotter" <tom129@juno.com> wrote in message
> news:29df3039.0407191115.539d5a7a@posting.google.c om...
> >
> > "Caroline Thompson" <ch.thompson1@virgin.net> wrote
>
> > > ... as Clauser and Horne discovered, and expressed very clearly in
> > > their 1974 paper (Phys. Rev. D 10, 526-35 (1974)) the most useful
> > > kind of hidden variable model for real (optical) Bell test experiments
> is
> > > one in which the hidden variable set at the source is not in itself
> enough
> > > to determine (in conjuction with the detector setting) whether or not
> > > any given "photon" is detected. Other hidden variables, local to the
> > > detector region, are needed. The one set at the source determines only
> > > the *probability* of detection.
> >
> > Caroline, I appreciate your work, but I think you've
> > (along with many commentators) overcomplicated the problem.
> > Bell-type inequalities are simply inapplicable (wrt
> > the usual considerations of locality, reality, hidden
> > variables, etc.) to the combined contexts exemplified by
> > Bell-type experiments.
>
> Agreed, Bell's original (1964) inequality is inapplicable, but the CHSH one
> *would* be applicable if the fair sampling assumption were correct (and
> other experimental flaws not present!) and (with the same proviso) the CH74
> one is almost certain to be applicable (since nobody can think of any real
> physical reason why there should be "enhancement", the presence of a
> polarisers increasing the probability of detection for certain values of
> \lambda).
>
> > The formulations on which the inequalities are based contain
> > a variable, \lambda (representing, in effect, the polarization
> > of the photons), which is irrelevant wrt the determination of
> > coincidental detection.
>
> How can you say this? The difference between lamba and the polariser
> settings are the *only* parameters available that could possibly control the
> coincidences.
I think that's incorrect. There's another parameter that
seems to have gotten lost in the mix of discussions about
this stuff -- the *relationship* between photon 1 and
photon 2 of any given pair.
\Lambda and the polarizer settings are the effective
determiners of individual results. But we're concerned
with the combined context, and in this context the relevant
factor is the relationship between paired photons. This
relationship is not \lambda. This relationship refers to
the theoretical determination that opposite moving photons
emitted from the same atom must be polarized identically
via emission. This relationship doesn't vary from pair to
pair as \lambda does. You don't need to know anything about
the varying specific polarization of pairs of opposite moving
photons in order to predict rates of coincidental detection.
As long as photon 1 and photon 2 of any given pair are
identically polarized, then the orientation of the polarizers
wrt each other is the only relevant variable in the
observational context, and a circular function of this
changing angular difference is what determines rates of
coincidental detection.
> As I've said before, they only partially control it, but, as
> evidence by the observed high "correlations", this is enough.
>
> [Incidentally, all this discussion is done on the assumption that the idea
> of "quantum correlation" makes sense and that Bell inequalities have to be
> couched in terms of estimates of it. The CH74 inequality does not in fact
> require the quantum correlation concept. It relies solely on
> straightforward probabilities. And by looking only at '+' results, not at
> both '+' and '-', it avoids an otherwise awkward situation that can occur in
> real experiments, when both '+' and '-' detections are made from the same
> polariser. "Quantum correlation" is not even defined in such cases, as far
> as I know. [See my new paper,
> [ http://freespace.virgin.net/ch.thompson1/Papers/CH74/CH74assumptions.htm ]]
>
Ok, I will. But I think you're barking up the wrong tree.
(Not that your concerns wrt experimental refinements aren't
relevant in a broader sense.)
> > This doesn't mean that \lambda doesn't exist. Indeed, as Bell
> > acknowledged, if \lambda were known, then the results of
> > individual measurements could be more accurately predicted.
> >
> > But, Bell-type experiments aren't individual contexts,
> > they're combined contexts, and the thing that
> > produces the correlations between the data streams in
> > combined contexts isn't the same thing that would produce
> > more accurate predictions of individual measurements.
>
> What is it, then?
>
> > In the combined context, it's the *relationship* between
> > the photons of a pair that matters.
>
> How can you have a relationship without both individuals
> having their own attributes?
They do have their own attributes. They have a specific
wavelength, direction, and polarization. But wrt calculating
accurate expectation values of rates of coincidental detection,
the only attributes that are important are the ones they
have *in common.* Afaik, they have only one thing in common,
they're *polarized identically* via the emission process -- which
is all you need to know (along with the polarizer settings) to
make accurate predictions regarding coincidental detection.
Even if you separate detections and changes in polarizer
settings by a spacelike interval, even if you change polarizer
settings in mid-flight, even if A has registered a detection and
you then change B's polarizer setting -- for any given pair
of photons there's still only one \Theta. And, so long as
photon 1 and photon 2 of any given pair are identically
polarized via emission, then, as \Theta increases, then rate
of coincidental detection must decrease, and as \Theta decreases,
then rate of coincidental detection must increase, as
a circular function of \Theta.
The experimental evidence for this view is pretty good, I think.
>
> > ... the emission model says that photons emitted in opposite
> > directions (and these are the ones chosen for eventual
> > mutual analysis by the polarizers) must be polarized identically
> > due to conservation of angular momentum. And, experiments
> > seem to me to support this model.
>
> OK. This does not conflict with local realism.
>
As I've previously noted, there's nothing in the qm formulation
that conflicts with local realism (in the sense that local refers
to a limiting speed for signals, and that events at the A-end
of experiments are effectively isolated from, and independent of,
events at the B-end.
It's just that interpretations of the formulation have become
muddled following misinterpretations of Bell's work.
> > As for qm being a 'nonlocal' theory, yes, insofar as it's
> > dealing with combined contexts rather than individual ones.
> > Bell tests are, by definition, nonlocal experimental contexts.
>
> How so? They are perfectly ordinary lab experiments. What is a "nonlocal
> experimental context"?
Correlated events at A and B are spacelike separated. Separating
events at A and B by a spacelike interval was necessary in order
to experimentally rule out any communication between the
the correlated photons.
>
> > Note that none of this has anything to do with the existence
> > (or non-existence) of hidden variables, or signals moving
> > faster than light.
> >
> > Bell tests are superfluous wrt these considerations. The only
> > thing that results of Bell tests will give you is some idea wrt
> > degree of entanglement.
>
> As a convinced local realist, this is not my problem. This
> represents a vast overcomplication on the part of quantum theorists.
>
I agree, at least wrt *interpretations* of quantum theory.
> > The subtle assumption that causes the confusion surrounding
> > these issues is that \lambda is a factor that's relevant wrt
> > the determination of coincidental detection in combined
> > contexts. If you omit \lambda from the formulation (and, note
> > that this doesn't mean that it doesn't exist), then you're
> > left with one relevant variable, the angular difference of
> > the polarizer settings, and, thus, the qm formulation.
>
> And if they were to look just a little bit harder at the
> evidence they'd find that this was not enough!
But it is enough, as long as experimenters are making sure
that they're analyzing pairs of photons correlated in polarization
via the emission process. And, I think they make sufficient
efforts to ensure that this is what they're doing.
> Especially in situations where the source is
> not rotationally invariant, which includes, as far as I can tell, all
> experiments using parametric down conversion sources. Check how the Bell
> tests are applied. They don't often use the simplified version that applies
> when the source *is* rotationally invariant.
I don't know about pdc, I've so far only studied the Aspect
experiments, atomic calcium cascades, and Bell's papers,
but the principle should be the same no matter what the
source.
Take a while to think about the *relationship* of photons
that are correlated in polarization,as opposed to their varying
lambdas, and what this means wrt different experimental contexts.
Caroline Thompson
Jul21-04, 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"scerir" <scerir@libero.it> wrote in message\nnews:9o8Lc.1622\\$1V3.44751@twister2.libe ro.it...\n\n> Not the photon(s), the atoms are entangled\n> http://www.arxiv.org/abs/quant-ph/0205182\n> and this seems even more strange than the\n> usual entanglement.\n\nAnd what proof did they give that there really was "entanglement"? What\ntest was used, with what assumptions? I\'ve only looked at the abstract:\n\n"When a single photon is split by a beam splitter, its two `halves\' can\nentangle two distant atoms into an EPR pair. We discuss a time-reversed\nanalogue of this experiment where two distant sources cooperate so as to\nemit a single photon. The two `half photons,\' having interacted with two\natoms, can entangle these atoms into an EPR pair once they are detected as a\nsingle photon. Entanglement occurs by creating indistinguishabilility\nbetween the two mutually exclusive histories of the photon. This\nindistinguishabilility can be created either at the end of the two histories\n(by `erasing\' the single photon\'s path) or at their beginning (by `erasing\'\nthe two atoms\' positions)."\n\nbut thought perhaps you might be able to save the trouble of reading the\narticle?\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"scerir" <scerir@libero.it> wrote in message
news:9o8Lc.1622$1V3.44751@twister2.libero.it...
> Not the photon(s), the atoms are entangled
> http://www.arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/0205182
> and this seems even more strange than the
> usual entanglement.
And what proof did they give that there really was "entanglement"? What
test was used, with what assumptions? I've only looked at the abstract:
"When a single photon is split by a beam splitter, its two `halves' can
entangle two distant atoms into an EPR pair. We discuss a time-reversed
analogue of this experiment where two distant sources cooperate so as to
emit a single photon. The two `half photons,' having interacted with two
atoms, can entangle these atoms into an EPR pair once they are detected as a
single photon. Entanglement occurs by creating indistinguishabilility
between the two mutually exclusive histories of the photon. This
indistinguishabilility can be created either at the end of the two histories
(by `erasing' the single photon's path) or at their beginning (by `erasing'
the two atoms' positions)."
but thought perhaps you might be able to save the trouble of reading the
article?
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Joe Rongen
Jul22-04, 04:16 AM
<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<3KZJc.104\\$wq5.31@newsfe3-gui.ntli.net>...\n[snip]\n\n>There are some very neat sources of "photons" around these days\n>that do seem to naturally support QM, but have they really proved\n>that they were initially of random direction? It\'s not easy to tell the\n>difference between a mixture of signals, some of one polarisation, some\n>orthogonal, and a genuinely random set.\n\nOne can find a practical example here:\n\nhttp://xxx.lanl.gov/abs/quant-ph?0404115\n\n"We present an entangled-state quantum cryptography system that operated for\nthe first time in a real world application scenario. The full key generation\nprotocol was performed in real time between two distributed embedded\nhardware devices, which were connected by 1.45 km of optical fiber,\ninstalled for this experiment in the Vienna sewage system. The generated\nquantum key was immediately handed over and used by a secure communication\napplication."\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>"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<3KZJc.104$wq5.31@newsfe3-gui.ntli.net>...
[snip]
>There are some very neat sources of "photons" around these days
>that do seem to naturally support QM, but have they really proved
>that they were initially of random direction? It's not easy to tell the
>difference between a mixture of signals, some of one polarisation, some
>orthogonal, and a genuinely random set.
One can find a practical example here:
http://xxx.lanl.gov/abs/quant-ph?0404115
"We present an entangled-state quantum cryptography system that operated for
the first time in a real world application scenario. The full key generation
protocol was performed in real time between two distributed embedded
hardware devices, which were connected by 1.45 km of optical fiber,
installed for this experiment in the Vienna sewage system. The generated
quantum key was immediately handed over and used by a secure communication
application."
Caroline Thompson
Jul22-04, 04:17 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"Tom Trotter" <tom129@juno.com> wrote in message\nnews:29df3039.0407202211.3857ba19@posting .google.com...\n> "Caroline Thompson" <ch.thompson1@virgin.net> wrote\n\n> > ... The differences between lamba and the polariser\n> > settings are the *only* parameters available that could\n> > possibly control the coincidences.\n>\n> I think that\'s incorrect. There\'s another parameter that\n> seems to have gotten lost in the mix of discussions about\n> this stuff -- the *relationship* between photon 1 and\n> photon 2 of any given pair.\n\nBut the assumed relationship is merely that lambda is the same for each\nphoton in a pair: that lambda_A = lambda_B. Knowing this has no effect on\nwhat happens to the photons individually. I know QM says there *is*\nsomething extra at work, but the matter really has not been adequately\ninvestigated. If QM was the correct model in all Bell test experiments we\nwould never have a case that was not rotationally invariant. But even\nAspect\'s famous experiments were not, and for this reason he used the\ngeneral form of the Bell tests (the CHSH one in one experiment, the CH74 in\nthe other two). There are good physical reasons to suppose that there would\nhave been bias towards vertical polarisation (both stimulating polarisers\nwere polarised vertically and the ion stream was vertical). And there is\nevidence in the data. See\nhttp://freespace.virgin.net/ch.thompson1/Tables/tab7sel.txt. This is some\nof Aspect\'s raw data, together with a few derived variables. Unfortunately\nthere were large numbers of "accidentals" (the true nature of which is still\na matter for debate). If they were not present then one would definitely\nexpect, if the paired photons were part of a rotationally invariant\nensemble, that the singles counts should not vary with detector setting. As\nit is, the photons in the singles counts may very likely be from a mixture\nof two different populations -- one the population that is paired at the\nsource, the other unwanted stray light. There is no reason to suppose the\npolarisation characteristics of the two populations to be identical. The\npossibility is clearly present, though, that the source was not producing\npairs that had an equal chance of polarisation in any direction.\n\nAs I said, there are other more recent experiments where rotational\ninvariance is unlikely. It is no problem so far as local realism and the\nBell inqualities are concerned but does mean that you can\'t assume in real\nlife that it is only the *difference* in polariser settings that matters.\n(See more on this in http://arxiv.org/abs/quant-ph/9912082 )\n\n> ... As long as photon 1 and photon 2 of any given pair are\n> identically polarized, then the orientation of the polarizers\n> wrt each other is the only relevant variable in the\n> observational context, and a circular function of this\n> changing angular difference is what determines rates of\n> coincidental detection.\n\nBut this is true only if we have rotational invariance. In most real\nexperiments we find in practice that they keep one polariser fixed and just\nvary the other one. The onlooker is left assuming that it would not have\nmattered if they\'d chosen a different orientation for the fixed polariser.\nBut sometimes this orientation seems to be carefully chosen. Kwiat, for\ninstance, in one of his well-known experiments\n<http://arXiv.org/abs/quant-ph/9810003>, kept one polariser fixed at 45 deg\nand varied the other. What would have happened if he had chosen some other\nangle? Most models lead to at least an approximately sinusoidal curve, but\nthe amplitude of that curve can vary with the choice of the fixed angle. It\nis *not* just the difference in detector settings that matters.\n\n> Even if you separate detections and changes in polarizer\n> settings by a spacelike interval, even if you change polarizer\n> settings in mid-flight, even if A has registered a detection and\n> you then change B\'s polarizer setting -- for any given pair\n> of photons there\'s still only one Theta. And, so long as\n> photon 1 and photon 2 of any given pair are identically\n> polarized via emission, then, as Theta increases, then rate\n> of coincidental detection must decrease, and as Theta decreases,\n> then rate of coincidental detection must increase, as\n> a circular function of Theta.\n\nTrue, but the rates of change can vary with a and b separately, not just\nwith the difference, a - b = theta.\n\n> The experimental evidence for this view is pretty good, I think.\n\nI\'m not disputing the general pattern, only the detail. Do we or do we not\nalways have rotational invariance? If we sometimes do not (and the fact\nthat experimenters often feel they need to use the non-rotationally\ninvariant form of the Bell test) can the QM model be correct for the\nexperiments concerned?\n\n> As I\'ve previously noted, there\'s nothing in the qm formulation\n> that conflicts with local realism (in the sense that local refers\n> to a limiting speed for signals, and that events at the A-end\n> of experiments are effectively isolated from, and independent of,\n> events at the B-end.\n\nWhat was Bell worried about then? He had proved that there were predictions\nof QM that were incompatible with the existence of local hidden variables.\nTo me that is matter of concern, an indicator that QM is not a satisfactory\ntheory. EPR thought it was incomplete, but perhaps they should rather have\nsaid that what they had shown was that it cannot be completed. This is what\nBell had confirmed.\n\n> It\'s just that interpretations of the formulation have become\n> muddled following misinterpretations of Bell\'s work.\n\nI agree that there has been a lot of misinterpretation of Bell\'s work, but\nthis is mainly on the behalf of science journalists. Judging by his PhD\nthesis, and known communication between them, I think Aspect understood Bell\nvery well.\n\n> > > As for qm being a \'nonlocal\' theory, yes, insofar as it\'s\n> > > dealing with combined contexts rather than individual ones.\n> > > Bell tests are, by definition, nonlocal experimental contexts.\n> >\n> > How so? They are perfectly ordinary lab experiments. What\n> > is a "nonlocal experimental context"?\n>\n> Correlated events at A and B are spacelike separated. Separating\n> events at A and B by a spacelike interval was necessary in order\n> to experimentally rule out any communication between the\n> the correlated photons.\n\nYes, but that does not make the experiment "nonlocal". "Nonlocal", to me,\nmeans instantaneous action at a distance. All we have in the Bell tests is\nlambda values set at the source and carried with the photons to their\nrespective detectors. A purely local action takes place at each detector.\n\n> I don\'t know about pdc, I\'ve so far only studied the Aspect\n> experiments, atomic calcium cascades, and Bell\'s papers,\n> but the principle should be the same no matter what the\n> source.\n\nQM covers all these experiments by the same theory. Local realist theories\nare not so restricted. They can more readily accomodate real differences in\nthe degree of departure from rotational invariance.\n\n> Take a while to think about the *relationship* of photons\n> that are correlated in polarization,as opposed to their varying\n> lambdas, and what this means wrt different experimental\n> contexts.\n\nRest assured I have done. It is very unfortunate that, as far as I know, no\nexperimenter has ever seen fit to publish coincidence curves for different\nfixed positions for one of the detectors. I think were they to do so there\nwould be relatively few instances in which the choice made no difference.\n[Come to think of it, there is a great deal more information that\nexperimenters could usefully publish! They have concentrated on the Bell\ntest itself, forgetting that they are trying to fit a model. The behaviour\nof the model as conditions are varied needs to be known if we are to make\ninformed decisions on its suitability for the job.]\n\nCaroline\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Tom Trotter" <tom129@juno.com> wrote in message
news:29df3039.0407202211.3857ba19@posting.google.c om...
> "Caroline Thompson" <ch.thompson1@virgin.net> wrote
> > ... The differences between lamba and the polariser
> > settings are the *only* parameters available that could
> > possibly control the coincidences.
>
> I think that's incorrect. There's another parameter that
> seems to have gotten lost in the mix of discussions about
> this stuff -- the *relationship* between photon 1 and
> photon 2 of any given pair.
But the assumed relationship is merely that \lambda is the same for each
photon in a pair: that \lambda_A = \lambda_B. Knowing this has no effect on
what happens to the photons individually. I know QM says there *is*
something extra at work, but the matter really has not been adequately
investigated. If QM was the correct model in all Bell test experiments we
would never have a case that was not rotationally invariant. But even
Aspect's famous experiments were not, and for this reason he used the
general form of the Bell tests (the CHSH one in one experiment, the CH74 in
the other two). There are good physical reasons to suppose that there would
have been bias towards vertical polarisation (both stimulating polarisers
were polarised vertically and the ion stream was vertical). And there is
evidence in the data. See
http://freespace.virgin.net/ch.thompson1/Tables/tab7sel.txt. This is some
of Aspect's raw data, together with a few derived variables. Unfortunately
there were large numbers of "accidentals" (the true nature of which is still
a matter for debate). If they were not present then one would definitely
expect, if the paired photons were part of a rotationally invariant
ensemble, that the singles counts should not vary with detector setting. As
it is, the photons in the singles counts may very likely be from a mixture
of two different populations -- one the population that is paired at the
source, the other unwanted stray light. There is no reason to suppose the
polarisation characteristics of the two populations to be identical. The
possibility is clearly present, though, that the source was not producing
pairs that had an equal chance of polarisation in any direction.
As I said, there are other more recent experiments where rotational
invariance is unlikely. It is no problem so far as local realism and the
Bell inqualities are concerned but does mean that you can't assume in real
life that it is only the *difference* in polariser settings that matters.
(See more on this in http://arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/9912082 )
> ... As long as photon 1 and photon 2 of any given pair are
> identically polarized, then the orientation of the polarizers
> wrt each other is the only relevant variable in the
> observational context, and a circular function of this
> changing angular difference is what determines rates of
> coincidental detection.
But this is true only if we have rotational invariance. In most real
experiments we find in practice that they keep one polariser fixed and just
vary the other one. The onlooker is left assuming that it would not have
mattered if they'd chosen a different orientation for the fixed polariser.
But sometimes this orientation seems to be carefully chosen. Kwiat, for
instance, in one of his well-known experiments
<http://arXiv.org/abs/http://www.arxiv.org/abs/quant-ph/9810003>, kept one polariser fixed at 45 deg
and varied the other. What would have happened if he had chosen some other
angle? Most models lead to at least an approximately sinusoidal curve, but
the amplitude of that curve can vary with the choice of the fixed angle. It
is *not* just the difference in detector settings that matters.
> Even if you separate detections and changes in polarizer
> settings by a spacelike interval, even if you change polarizer
> settings in mid-flight, even if A has registered a detection and
> you then change B's polarizer setting -- for any given pair
> of photons there's still only one \Theta. And, so long as
> photon 1 and photon 2 of any given pair are identically
> polarized via emission, then, as \Theta increases, then rate
> of coincidental detection must decrease, and as \Theta decreases,
> then rate of coincidental detection must increase, as
> a circular function of \Theta.
True, but the rates of change can vary with a and b separately, not just
with the difference, a - b = \theta.
> The experimental evidence for this view is pretty good, I think.
I'm not disputing the general pattern, only the detail. Do we or do we not
always have rotational invariance? If we sometimes do not (and the fact
that experimenters often feel they need to use the non-rotationally
invariant form of the Bell test) can the QM model be correct for the
experiments concerned?
> As I've previously noted, there's nothing in the qm formulation
> that conflicts with local realism (in the sense that local refers
> to a limiting speed for signals, and that events at the A-end
> of experiments are effectively isolated from, and independent of,
> events at the B-end.
What was Bell worried about then? He had proved that there were predictions
of QM that were incompatible with the existence of local hidden variables.
To me that is matter of concern, an indicator that QM is not a satisfactory
theory. EPR thought it was incomplete, but perhaps they should rather have
said that what they had shown was that it cannot be completed. This is what
Bell had confirmed.
> It's just that interpretations of the formulation have become
> muddled following misinterpretations of Bell's work.
I agree that there has been a lot of misinterpretation of Bell's work, but
this is mainly on the behalf of science journalists. Judging by his PhD
thesis, and known communication between them, I think Aspect understood Bell
very well.
> > > As for qm being a 'nonlocal' theory, yes, insofar as it's
> > > dealing with combined contexts rather than individual ones.
> > > Bell tests are, by definition, nonlocal experimental contexts.
> >
> > How so? They are perfectly ordinary lab experiments. What
> > is a "nonlocal experimental context"?
>
> Correlated events at A and B are spacelike separated. Separating
> events at A and B by a spacelike interval was necessary in order
> to experimentally rule out any communication between the
> the correlated photons.
Yes, but that does not make the experiment "nonlocal". "Nonlocal", to me,
means instantaneous action at a distance. All we have in the Bell tests is
\lambda values set at the source and carried with the photons to their
respective detectors. A purely local action takes place at each detector.
> I don't know about pdc, I've so far only studied the Aspect
> experiments, atomic calcium cascades, and Bell's papers,
> but the principle should be the same no matter what the
> source.
QM covers all these experiments by the same theory. Local realist theories
are not so restricted. They can more readily accomodate real differences in
the degree of departure from rotational invariance.
> Take a while to think about the *relationship* of photons
> that are correlated in polarization,as opposed to their varying
> lambdas, and what this means wrt different experimental
> contexts.
Rest assured I have done. It is very unfortunate that, as far as I know, no
experimenter has ever seen fit to publish coincidence curves for different
fixed positions for one of the detectors. I think were they to do so there
would be relatively few instances in which the choice made no difference.
[Come to think of it, there is a great deal more information that
experimenters could usefully publish! They have concentrated on the Bell
test itself, forgetting that they are trying to fit a model. The behaviour
of the model as conditions are varied needs to be known if we are to make
informed decisions on its suitability for the job.]
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Paul Stewart Snyder
Jul25-04, 08:16 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\ntom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google. com>...\n> ps@ws5.com (Paul Stewart Snyder) wrote in message\n> > We know that when A is changed B is instantaneously\n> > changed,\n>\n> Only in certain contexts, but they don\'t imply signal\n> transfer between A and B.\n>\nThis is what continues to bother me - there are at least some contexts\nin which there are instantaneous changes to B when A changes - even if\nA and B are spatially separated. There may not be a signal transfer in\nthe formal terms of information transfer, and EPR may not require an\nFTL event, yet there still seems to be a strong intuitive basis\n(Einstein loved intuition) for thinking that something FTL is going on\nhere.\n\nThanks, PS\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>tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google.com>...
> ps@ws5.com (Paul Stewart Snyder) wrote in message
> > We know that when A is changed B is instantaneously
> > changed,
>
> Only in certain contexts, but they don't imply signal
> transfer between A and B.
>
This is what continues to bother me - there are at least some contexts
in which there are instantaneous changes to B when A changes - even if
A and B are spatially separated. There may not be a signal transfer in
the formal terms of information transfer, and EPR may not require an
FTL event, yet there still seems to be a strong intuitive basis
(Einstein loved intuition) for thinking that something FTL is going on
here.
Thanks, PS
Tom Trotter
Jul25-04, 08:16 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"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<eXALc.411\\$5R5.370@newsfe5-gui.ntli.net>...\n> "Tom Trotter" <tom129@juno.com> wrote in message\n> news:29df3039.0407202211.3857ba19@posting.google.c om...\n> > "Caroline Thompson" <ch.thompson1@virgin.net> wrote\n>\n> > > ... The differences between lamba and the polariser\n> > > settings are the *only* parameters available that could\n> > > possibly control the coincidences.\n> >\n> > I think that\'s incorrect. There\'s another parameter that\n> > seems to have gotten lost in the mix of discussions about\n> > this stuff -- the *relationship* between photon 1 and\n> > photon 2 of any given pair.\n>\n> But the assumed relationship is merely that lambda is the same for each\n> photon in a pair: that lambda_A = lambda_B. Knowing this has no effect on\n> what happens to the photons individually.\n\nThat\'s right. Knowing this relationship will not allow\nyou to more accurately predict individual results,\nif you\'re just considering A, or just considering B.\n\nAnd, knowing the polarization for photons incident\non the polarizers won\'t increase your ability\nto predict coincidental detections, as long as\nyou know how the photons are related.\n\n\n> I know QM says there *is* something extra at work,\n> but the matter really has not been adequately\n> investigated.\n\nKeep in mind that we\'re considering \'ideal\' scenarios in\norder to understand which variables would be relevant wrt\npredicting rates of coincidental detection.\n\nThe \'something extra\' is the emission-dependent *relationship*\nbetween photon 1 and photon 2, if what\'s being analyzed is\nentangled. But, it isn\'t really extra. It\'s just the only\nthing circumscribed by Bell\'s lambda that\'s relevant\nto coincidental detection. You just eliminate from\nconsideration all the extraneous stuff, like how the\nphotons are polarized. Then you\'re left with the qm\nformulation.\n\n> If QM was the correct model in all Bell test experiments we\n> would never have a case that was not rotationally invariant. But even\n> Aspect\'s famous experiments were not, and for this reason he used the\n> general form of the Bell tests (the CHSH one in one experiment, the CH74 in\n> the other two). There are good physical reasons to suppose that there would\n> have been bias towards vertical polarisation (both stimulating polarisers\n> were polarised vertically and the ion stream was vertical). And there is\n> evidence in the data. See\n> http://freespace.virgin.net/ch.thompson1/Tables/tab7sel.txt. This is some\n> of Aspect\'s raw data, together with a few derived variables. Unfortunately\n> there were large numbers of "accidentals" (the true nature of which is still\n> a matter for debate). If they were not present then one would definitely\n> expect, if the paired photons were part of a rotationally invariant\n> ensemble, that the singles counts should not vary with detector setting. As\n> it is, the photons in the singles counts may very likely be from a mixture\n> of two different populations -- one the population that is paired at the\n> source, the other unwanted stray light. There is no reason to suppose the\n> polarisation characteristics of the two populations to be identical. The\n> possibility is clearly present, though, that the source was not producing\n> pairs that had an equal chance of polarisation in any direction.\n>\n> As I said, there are other more recent experiments where rotational\n> invariance is unlikely. It is no problem so far as local realism and the\n> Bell inqualities are concerned but does mean that you can\'t assume in real\n> life that it is only the *difference* in polariser settings that matters.\n> (See more on this in http://arxiv.org/abs/quant-ph/9912082 )\n>\n\nIf photon 1 and photon 2 were polarized differently, then the\nangular difference of their polarizations would be a factor\nin calculating rates of coincidental detection.\n\nBut photons that are emitted in opposite directions are\nwhat are selected. And, if they\'re emitted by the same\natom, then they have to be polarized identically,\naccording to the emission model, in order to comply with\nthe law of conservation of angular momentum.\n\nWrt this scenario, in the combined context, the only variable\nthat matters is the angular difference in polarizer settings.\n\n> > ... As long as photon 1 and photon 2 of any given pair are\n> > identically polarized, then the orientation of the polarizers\n> > wrt each other is the only relevant variable in the\n> > observational context, and a circular function of this\n> > changing angular difference is what determines rates of\n> > coincidental detection.\n>\n> But this is true only if we have rotational invariance.\n\nI don\'t think so. Remember, Bell\'s unknown parameter\naffecting individual results can have any value. It can\nhave one value (where singles counts will vary with polarizer\nsettings), or it can be random, ie., have an infinite range\nof values (where singles counts will not vary with\npolarizer settings).\n\nBell\'s analysis says that the coincidence rate, cr, for 2Theta,\ncan\'t exceed cr/2 for Theta. But, experimentally, this doesn\'t\nhold for all tested values of Theta. The qm formulation, on the\nother hand, says that cr will vary as a circular, rather than\na linear, function of Theta -- and experiments bear this out.\n\n> In most real\n> experiments we find in practice that they keep one polariser fixed and just\n> vary the other one. The onlooker is left assuming that it would not have\n> mattered if they\'d chosen a different orientation for the fixed polariser.\n\nIt wouldn\'t. Not for Bell tests anyway.\n\n> But sometimes this orientation seems to be carefully chosen. Kwiat, for\n> instance, in one of his well-known experiments\n> <http://arXiv.org/abs/quant-ph/9810003>, kept one polariser fixed at 45 deg\n> and varied the other. What would have happened if he had chosen some other\n> angle? Most models lead to at least an approximately sinusoidal curve, but\n> the amplitude of that curve can vary with the choice of the fixed angle. It\n> is *not* just the difference in detector settings that matters.\n>\n\nAs long as you\'re dealing with opposite moving photons emitted\nby the same atom, then the angular difference of the settings\nof the analyzers is the only thing that matters wrt predicting\nrates of coincidental detection in the combined context.\n\n> > Even if you separate detections and changes in polarizer\n> > settings by a spacelike interval, even if you change polarizer\n> > settings in mid-flight, even if A has registered a detection and\n> > you then change B\'s polarizer setting -- for any given pair\n> > of photons there\'s still only one Theta. And, so long as\n> > photon 1 and photon 2 of any given pair are identically\n> > polarized via emission, then, as Theta increases, then rate\n> > of coincidental detection must decrease, and as Theta decreases,\n> > then rate of coincidental detection must increase, as\n> > a circular function of Theta.\n>\n> True, but the rates of change can vary with a and b separately, not just\n> with the difference, a - b = theta.\n>\n\nI sense that you still haven\'t grabbed hold of the idea that\nwe\'re dealing with two, entirely different, experimental contexts\nhere. This is the key to understanding what Bell\'s analysis, and\nexperimental tests of Bell inequalities, actually tell you (and\nnot what most commentators, including some statements in the\nsci.physics faq, say they tell you).\n\n> > The experimental evidence for this view is pretty good, I think.\n>\n> I\'m not disputing the general pattern, only the detail. Do we or do we not\n> always have rotational invariance? If we sometimes do not (and the fact\n> that experimenters often feel they need to use the non-rotationally\n> invariant form of the Bell test) can the QM model be correct for the\n> experiments concerned?\n>\n\nRotational *variance* wouldn\'t alter what Bell\'s analysis reveals\n(ie., that the parameters that determine individual results are\nnot the same parameters that determine combined results).\n\n> > As I\'ve previously noted, there\'s nothing in the qm formulation\n> > that conflicts with local realism (in the sense that local refers\n> > to a limiting speed for signals, and that events at the A-end\n> > of experiments are effectively isolated from, and independent of,\n> > events at the B-end.\n>\n> What was Bell worried about then? He had proved that there were predictions\n> of QM that were incompatible with the existence of local hidden variables.\n\nMaybe that\'s what he was worried about, but that\'s not what he proved.\nHe proved that if you include parameters that would affect individual\nmeasurements in a formulation concerning combined measurements, then\nyou get results that are inconsistent with qm and experiments, or\n(after tweaking so that it\'s consistent with the statistical\npredictions of qm)\nresults that are inconsistent with the assumptions of standard physics\nregarding a limitation on signal velocity (thereby violating Einstein\ncausality or Lorentz invariance).\n\nEither way, you get nonsense. Sort of a reduction to absurdities.\nThe conclusion that follows from this is that the added parameters,\nwhile relevant wrt individual contexts, simply aren\'t relevant wrt\ncombined contexts. That\'s all. Nothing has been said about the\npredictions of qm being inconsistent with the *existence*\nof hidden variables, or being consistent with the *necessity* of\nftl signals.\n\nThe mechanism for A and B \'instantaneously\' communicating is\nof course \'necessary\' if you want to keep the tweaked-up,\nsr-violating,\nqm(exp)-compatible, added-parameter formulation for the combined\ncontext.\nBut, why on earth would anyone want to do *that*? :-)\n\n> > It\'s just that interpretations of the formulation have become\n> > muddled following misinterpretations of Bell\'s work.\n>\n> I agree that there has been a lot of misinterpretation of Bell\'s work, but\n> this is mainly on the behalf of science journalists. Judging by his PhD\n> thesis, and known communication between them, I think Aspect understood Bell\n> very well.\n>\n\nI\'m not sure he understood that it\'s *not* their obedience to\nEinstein causality that makes supplementary-parameter theories\nincompatible with qm and experiment. The qm formulation, and\nthe experiments, also obey Einstein causality.\n\n\n> > > > As for qm being a \'nonlocal\' theory, yes, insofar as it\'s\n> > > > dealing with combined contexts rather than individual ones.\n> > > > Bell tests are, by definition, nonlocal experimental contexts.\n> > >\n> > > How so? They are perfectly ordinary lab experiments. What\n> > > is a "nonlocal experimental context"?\n> >\n> > Correlated events at A and B are spacelike separated. Separating\n> > events at A and B by a spacelike interval was necessary in order\n> > to experimentally rule out any communication between the\n> > the correlated photons.\n>\n> Yes, but that does not make the experiment "nonlocal". "Nonlocal", to me,\n> means instantaneous action at a distance.\n\n"Nonlocal" has at least four meanings that I know of. The way you\'re\nusing it above is a connotation that, in connection with Bell\'s work,\nevolved from misinterpretations of it\'s standard meaning -- which is\nthat\ntwo events are nonlocally correlated if they\'re functionally related\nand separated by a spacelike interval. Note that this doesn\'t entail\nthat they\'re communicating in any way. They\'re just related by\nsome function in some, nonlocal, observational context.\n\nThe *context* of Aspect\'s 1982 experiment makes it a nonlocal\nexperiment. And, it would seem to follow that theories applicable\nto nonlocal contexts, would be called nonlocal theories.\n\n> All we have in the Bell tests is\n> lambda values set at the source and carried with the photons to their\n> respective detectors. A purely local action takes place at each detector.\n>\n\nWell, that\'s not *all* you have. You\'re forgetting the observational\ncontext. :-)\n\n\n> > I don\'t know about pdc, I\'ve so far only studied the Aspect\n> > experiments, atomic calcium cascades, and Bell\'s papers,\n> > but the principle should be the same no matter what the\n> > source.\n>\n> QM covers all these experiments by the same theory. Local realist theories\n> are not so restricted. They can more readily accomodate real differences in\n> the degree of departure from rotational invariance.\n>\n\nCertain formulations aren\'t applicable to certain contexts.\n\nRotational variance doesn\'t affect what Bell\'s analysis\nreveals.\n\n> > Take a while to think about the *relationship* of photons\n> > that are correlated in polarization,as opposed to their varying\n> > lambdas, and what this means wrt different experimental\n> > contexts.\n>\n> Rest assured I have done. It is very unfortunate that, as far as I know, no\n> experimenter has ever seen fit to publish coincidence curves for different\n> fixed positions for one of the detectors.\n\nBecause it\'s not relevant to the tests being done.\nThere\'s two main things they concentrate on afaik:\n(1) dealing with photons emitted by the same\natom in any given coincidence/correlation window, and\n(2) an accurate assessment of the efficiencies\nof the polarizers and detectors.\n\n> I think were they to do so there\n> would be relatively few instances in which the choice made no difference.\n> [Come to think of it, there is a great deal more information that\n> experimenters could usefully publish! They have concentrated on the Bell\n> test itself, forgetting that they are trying to fit a model. The behaviour\n> of the model as conditions are varied needs to be known if we are to make\n> informed decisions on its suitability for the job.]\n\nThat\'s why they\'re concerned about polarizer/detector\nefficiencies, isn\'t it?\n\n\n>\n> Caroline\n>\n> Caroline H Thompson\n>\n> ch.thompson1@virgin.net\n> http://freespace.virgin.net/ch.thompson1/\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>"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<eXALc.411$5R5.370@newsfe5-gui.ntli.net>...
> "Tom Trotter" <tom129@juno.com> wrote in message
> news:29df3039.0407202211.3857ba19@posting.google.c om...
> > "Caroline Thompson" <ch.thompson1@virgin.net> wrote
>
> > > ... The differences between lamba and the polariser
> > > settings are the *only* parameters available that could
> > > possibly control the coincidences.
> >
> > I think that's incorrect. There's another parameter that
> > seems to have gotten lost in the mix of discussions about
> > this stuff -- the *relationship* between photon 1 and
> > photon 2 of any given pair.
>
> But the assumed relationship is merely that \lambda is the same for each
> photon in a pair: that \lambda_A = \lambda_B. Knowing this has no effect on
> what happens to the photons individually.
That's right. Knowing this relationship will not allow
you to more accurately predict individual results,
if you're just considering A, or just considering B.
And, knowing the polarization for photons incident
on the polarizers won't increase your ability
to predict coincidental detections, as long as
you know how the photons are related.
> I know QM says there *is* something extra at work,
> but the matter really has not been adequately
> investigated.
Keep in mind that we're considering 'ideal' scenarios in
order to understand which variables would be relevant wrt
predicting rates of coincidental detection.
The 'something extra' is the emission-dependent *relationship*
between photon 1 and photon 2, if what's being analyzed is
entangled. But, it isn't really extra. It's just the only
thing circumscribed by Bell's \lambda that's relevant
to coincidental detection. You just eliminate from
consideration all the extraneous stuff, like how the
photons are polarized. Then you're left with the qm
formulation.
> If QM was the correct model in all Bell test experiments we
> would never have a case that was not rotationally invariant. But even
> Aspect's famous experiments were not, and for this reason he used the
> general form of the Bell tests (the CHSH one in one experiment, the CH74 in
> the other two). There are good physical reasons to suppose that there would
> have been bias towards vertical polarisation (both stimulating polarisers
> were polarised vertically and the ion stream was vertical). And there is
> evidence in the data. See
> http://freespace.virgin.net/ch.thompson1/Tables/tab7sel.txt. This is some
> of Aspect's raw data, together with a few derived variables. Unfortunately
> there were large numbers of "accidentals" (the true nature of which is still
> a matter for debate). If they were not present then one would definitely
> expect, if the paired photons were part of a rotationally invariant
> ensemble, that the singles counts should not vary with detector setting. As
> it is, the photons in the singles counts may very likely be from a mixture
> of two different populations -- one the population that is paired at the
> source, the other unwanted stray light. There is no reason to suppose the
> polarisation characteristics of the two populations to be identical. The
> possibility is clearly present, though, that the source was not producing
> pairs that had an equal chance of polarisation in any direction.
>
> As I said, there are other more recent experiments where rotational
> invariance is unlikely. It is no problem so far as local realism and the
> Bell inqualities are concerned but does mean that you can't assume in real
> life that it is only the *difference* in polariser settings that matters.
> (See more on this in http://arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/9912082 )
>
If photon 1 and photon 2 were polarized differently, then the
angular difference of their polarizations would be a factor
in calculating rates of coincidental detection.
But photons that are emitted in opposite directions are
what are selected. And, if they're emitted by the same
atom, then they have to be polarized identically,
according to the emission model, in order to comply with
the law of conservation of angular momentum.
Wrt this scenario, in the combined context, the only variable
that matters is the angular difference in polarizer settings.
> > ... As long as photon 1 and photon 2 of any given pair are
> > identically polarized, then the orientation of the polarizers
> > wrt each other is the only relevant variable in the
> > observational context, and a circular function of this
> > changing angular difference is what determines rates of
> > coincidental detection.
>
> But this is true only if we have rotational invariance.
I don't think so. Remember, Bell's unknown parameter
affecting individual results can have any value. It can
have one value (where singles counts will vary with polarizer
settings), or it can be random, ie., have an infinite range
of values (where singles counts will not vary with
polarizer settings).
Bell's analysis says that the coincidence rate, cr, for 2Theta,
can't exceed cr/2 for \Theta. But, experimentally, this doesn't
hold for all tested values of \Theta. The qm formulation, on the
other hand, says that cr will vary as a circular, rather than
a linear, function of \Theta -- and experiments bear this out.
> In most real
> experiments we find in practice that they keep one polariser fixed and just
> vary the other one. The onlooker is left assuming that it would not have
> mattered if they'd chosen a different orientation for the fixed polariser.
It wouldn't. Not for Bell tests anyway.
> But sometimes this orientation seems to be carefully chosen. Kwiat, for
> instance, in one of his well-known experiments
> <http://arXiv.org/abs/http://www.arxiv.org/abs/quant-ph/9810003>, kept one polariser fixed at 45 deg
> and varied the other. What would have happened if he had chosen some other
> angle? Most models lead to at least an approximately sinusoidal curve, but
> the amplitude of that curve can vary with the choice of the fixed angle. It
> is *not* just the difference in detector settings that matters.
>
As long as you're dealing with opposite moving photons emitted
by the same atom, then the angular difference of the settings
of the analyzers is the only thing that matters wrt predicting
rates of coincidental detection in the combined context.
> > Even if you separate detections and changes in polarizer
> > settings by a spacelike interval, even if you change polarizer
> > settings in mid-flight, even if A has registered a detection and
> > you then change B's polarizer setting -- for any given pair
> > of photons there's still only one \Theta. And, so long as
> > photon 1 and photon 2 of any given pair are identically
> > polarized via emission, then, as \Theta increases, then rate
> > of coincidental detection must decrease, and as \Theta decreases,
> > then rate of coincidental detection must increase, as
> > a circular function of \Theta.
>
> True, but the rates of change can vary with a and b separately, not just
> with the difference, a - b = \theta.
>
I sense that you still haven't grabbed hold of the idea that
we're dealing with two, entirely different, experimental contexts
here. This is the key to understanding what Bell's analysis, and
experimental tests of Bell inequalities, actually tell you (and
not what most commentators, including some statements in the
sci.physics faq, say they tell you).
> > The experimental evidence for this view is pretty good, I think.
>
> I'm not disputing the general pattern, only the detail. Do we or do we not
> always have rotational invariance? If we sometimes do not (and the fact
> that experimenters often feel they need to use the non-rotationally
> invariant form of the Bell test) can the QM model be correct for the
> experiments concerned?
>
Rotational *variance* wouldn't alter what Bell's analysis reveals
(ie., that the parameters that determine individual results are
not the same parameters that determine combined results).
> > As I've previously noted, there's nothing in the qm formulation
> > that conflicts with local realism (in the sense that local refers
> > to a limiting speed for signals, and that events at the A-end
> > of experiments are effectively isolated from, and independent of,
> > events at the B-end.
>
> What was Bell worried about then? He had proved that there were predictions
> of QM that were incompatible with the existence of local hidden variables.
Maybe that's what he was worried about, but that's not what he proved.
He proved that if you include parameters that would affect individual
measurements in a formulation concerning combined measurements, then
you get results that are inconsistent with qm and experiments, or
(after tweaking so that it's consistent with the statistical
predictions of qm)
results that are inconsistent with the assumptions of standard physics
regarding a limitation on signal velocity (thereby violating Einstein
causality or Lorentz invariance).
Either way, you get nonsense. Sort of a reduction to absurdities.
The conclusion that follows from this is that the added parameters,
while relevant wrt individual contexts, simply aren't relevant wrt
combined contexts. That's all. Nothing has been said about the
predictions of qm being inconsistent with the *existence*
of hidden variables, or being consistent with the *necessity* of
ftl signals.
The mechanism for A and B 'instantaneously' communicating is
of course 'necessary' if you want to keep the tweaked-up,
sr-violating,
qm(\exp)-compatible, added-parameter formulation for the combined
context.
But, why on earth would anyone want to do *that*? :-)
> > It's just that interpretations of the formulation have become
> > muddled following misinterpretations of Bell's work.
>
> I agree that there has been a lot of misinterpretation of Bell's work, but
> this is mainly on the behalf of science journalists. Judging by his PhD
> thesis, and known communication between them, I think Aspect understood Bell
> very well.
>
I'm not sure he understood that it's *not* their obedience to
Einstein causality that makes supplementary-parameter theories
incompatible with qm and experiment. The qm formulation, and
the experiments, also obey Einstein causality.
> > > > As for qm being a 'nonlocal' theory, yes, insofar as it's
> > > > dealing with combined contexts rather than individual ones.
> > > > Bell tests are, by definition, nonlocal experimental contexts.
> > >
> > > How so? They are perfectly ordinary lab experiments. What
> > > is a "nonlocal experimental context"?
> >
> > Correlated events at A and B are spacelike separated. Separating
> > events at A and B by a spacelike interval was necessary in order
> > to experimentally rule out any communication between the
> > the correlated photons.
>
> Yes, but that does not make the experiment "nonlocal". "Nonlocal", to me,
> means instantaneous action at a distance.
"Nonlocal" has at least four meanings that I know of. The way you're
using it above is a connotation that, in connection with Bell's work,
evolved from misinterpretations of it's standard meaning -- which is
that
two events are nonlocally correlated if they're functionally related
and separated by a spacelike interval. Note that this doesn't entail
that they're communicating in any way. They're just related by
some function in some, nonlocal, observational context.
The *context* of Aspect's 1982 experiment makes it a nonlocal
experiment. And, it would seem to follow that theories applicable
to nonlocal contexts, would be called nonlocal theories.
> All we have in the Bell tests is
> \lambda values set at the source and carried with the photons to their
> respective detectors. A purely local action takes place at each detector.
>
Well, that's not *all* you have. You're forgetting the observational
context. :-)
> > I don't know about pdc, I've so far only studied the Aspect
> > experiments, atomic calcium cascades, and Bell's papers,
> > but the principle should be the same no matter what the
> > source.
>
> QM covers all these experiments by the same theory. Local realist theories
> are not so restricted. They can more readily accomodate real differences in
> the degree of departure from rotational invariance.
>
Certain formulations aren't applicable to certain contexts.
Rotational variance doesn't affect what Bell's analysis
reveals.
> > Take a while to think about the *relationship* of photons
> > that are correlated in polarization,as opposed to their varying
> > lambdas, and what this means wrt different experimental
> > contexts.
>
> Rest assured I have done. It is very unfortunate that, as far as I know, no
> experimenter has ever seen fit to publish coincidence curves for different
> fixed positions for one of the detectors.
Because it's not relevant to the tests being done.
There's two main things they concentrate on afaik:
(1) dealing with photons emitted by the same
atom in any given coincidence/correlation window, and
(2) an accurate assessment of the efficiencies
of the polarizers and detectors.
> I think were they to do so there
> would be relatively few instances in which the choice made no difference.
> [Come to think of it, there is a great deal more information that
> experimenters could usefully publish! They have concentrated on the Bell
> test itself, forgetting that they are trying to fit a model. The behaviour
> of the model as conditions are varied needs to be known if we are to make
> informed decisions on its suitability for the job.]
That's why they're concerned about polarizer/detector
efficiencies, isn't it?
>
> Caroline
>
> Caroline H Thompson
>
> ch.thompson1@virgin.net
> http://freespace.virgin.net/ch.thompson1/
Caroline Thompson
Jul25-04, 08:16 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"Joe Rongen" <joe@alpha.to> wrote in message\nnews:a345818b.0407211703.4801f5f0@posting .google.com...\n> "Caroline Thompson" <ch.thompson1@virgin.net> wrote in message\nnews:<3KZJc.104\\$wq5.31@newsfe3-gui.ntli.net>...\n>\n> >There are some very neat sources of "photons" around these\n> >days that do seem to naturally support QM, but have they really\n> >proved that they were initially of random direction? It\'s not easy\n> >to tell the difference between a mixture of signals, some of one\n> >polarisation, some orthogonal, and a genuinely random set.\n>\n> One can find a practical example here:\n> http://xxx.lanl.gov/abs/quant-ph?0404115\n>\n> "We present an entangled-state quantum cryptography system ...\n\nThey don\'t tell us anything about how they generated the "entangled" pairs\nhere, though they reference Kwiat\'s 1995 paper. This is the source with the\noverlapping cones of light, one polarised vertically, the other\nhorizontally, if I remember rightly. The "photons" used are taken from the\npoints A and B where cones intersect. The QM story is that you have either\nan H at A and a V at B or a V A and an H at B.\n\nI\'m rather sticking my neck out here, I know, but I\'ve studied quite a lot\nof experiments that use this source and am fairly sure that what really\nhappens is that each of the supposedly single photons is really a\nsuperposition of two signals, one H and one V. The difference between the\nemissions at A and B is not really a matter of them being of orthogonal\npolarisation directions but of the H and V components having related phase\ndifferences. I\'d love to have more info on this!\n\nThere are a lot of questions I\'d like to ask about just what the role of the\nvarious half-wave plates etc that are used is supposed to be. A half-wave\nplate changes the phase difference by 180 deg. It can change the\npolarisation of plane polarised light by 90 deg, or reverse the direction of\npolarisation of circularly polarised light. What I suspect happens when\nKwiat\'s source is used in QKD is that small frequency differences from one\npair to the next cause difference in the phase difference between V and H\ncomponents, so that some pairs are approximately circularly polarised, some\napproximately plane. The choice by Alice of whether to try and measure\nplane or circular polarisation (by inserting an extra half-wave plate or\nnot) results, effectively, in selection of the pairs that are most nearly\nperfectly in the selected categories, pairs that otherwise polarised having\nmuch less chance of detection.\n\nMuch of the above is guesswork -- and I may have got some of the jargon\nwrong -- but I do know that Kwiat never proved conclusively that his photon\npairs *were* entangled since he used the CHSH Bell inequality and so had to\nassume "fair sampling". There are real correlations between the emitted\nphotons, and I think, for various reasons, that the correlation is in a\ncombination of phase and frequency. If it really was a matter of getting\neither H or V then we would *not* have the rotational invariance that QM\nassumes, since the axes of the nonlinear crystal source would provide\npreferred directions. As it is, we may have rotational invariance (in that\nall phase differences may be almost equally possible) but nobody has proved\nthat the results cannot be explained by ordinary correlations in the manner\nI\'ve suggested: the pairs have identical phase differences. Or possibly, as\nI suggest in quant-ph/9912082 in relation to Weihs\' Bell test experiment,\nabout half have one phase difference, the other half another, differing by\n180 deg. It all depends how accurately frequency is controlled.\n\nIn the QKD demonstration, it is possible that the method works by selecting\none of four distinguishable subsets of the emitted signals: circular L or R\nand linear +45 deg or -45 deg. The signals received by Alice are always\nidentical (or always conjugate?) to the ones received by Bob.\n\nCaroline\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Joe Rongen" <joe@\alpha.to> wrote in message
news:a345818b.0407211703.4801f5f0@posting.google.c om...
> "Caroline Thompson" <ch.thompson1@virgin.net> wrote in message
news:<3KZJc.104$wq5.31@newsfe3-gui.ntli.net>...
>
> >There are some very neat sources of "photons" around these
> >days that do seem to naturally support QM, but have they really
> >proved that they were initially of random direction? It's not easy
> >to tell the difference between a mixture of signals, some of one
> >polarisation, some orthogonal, and a genuinely random set.
>
> One can find a practical example here:
> http://xxx.lanl.gov/abs/quant-ph?0404115
>
> "We present an entangled-state quantum cryptography system ...
They don't tell us anything about how they generated the "entangled" pairs
here, though they reference Kwiat's 1995 paper. This is the source with the
overlapping cones of light, one polarised vertically, the other
horizontally, if I remember rightly. The "photons" used are taken from the
points A and B where cones intersect. The QM story is that you have either
an H at A and a V at B or a V A and an H at B.
I'm rather sticking my neck out here, I know, but I've studied quite a lot
of experiments that use this source and am fairly sure that what really
happens is that each of the supposedly single photons is really a
superposition of two signals, one H and one V. The difference between the
emissions at A and B is not really a matter of them being of orthogonal
polarisation directions but of the H and V components having related phase
differences. I'd love to have more info on this!
There are a lot of questions I'd like to ask about just what the role of the
various half-wave plates etc that are used is supposed to be. A half-wave
plate changes the phase difference by 180 deg. It can change the
polarisation of plane polarised light by 90 deg, or reverse the direction of
polarisation of circularly polarised light. What I suspect happens when
Kwiat's source is used in QKD is that small frequency differences from one
pair to the next cause difference in the phase difference between V and H
components, so that some pairs are approximately circularly polarised, some
approximately plane. The choice by Alice of whether to try and measure
plane or circular polarisation (by inserting an extra half-wave plate or
not) results, effectively, in selection of the pairs that are most nearly
perfectly in the selected categories, pairs that otherwise polarised having
much less chance of detection.
Much of the above is guesswork -- and I may have got some of the jargon
wrong -- but I do know that Kwiat never proved conclusively that his photon
pairs *were* entangled since he used the CHSH Bell inequality and so had to
assume "fair sampling". There are real correlations between the emitted
photons, and I think, for various reasons, that the correlation is in a
combination of phase and frequency. If it really was a matter of getting
either H or V then we would *not* have the rotational invariance that QM
assumes, since the axes of the nonlinear crystal source would provide
preferred directions. As it is, we may have rotational invariance (in that
all phase differences may be almost equally possible) but nobody has proved
that the results cannot be explained by ordinary correlations in the manner
I've suggested: the pairs have identical phase differences. Or possibly, as
I suggest in http://www.arxiv.org/abs/quant-ph/9912082 in relation to Weihs' Bell test experiment,
about half have one phase difference, the other half another, differing by
180 deg. It all depends how accurately frequency is controlled.
In the QKD demonstration, it is possible that the method works by selecting
one of four distinguishable subsets of the emitted signals: circular L or R
and linear +45 deg or -45 deg. The signals received by Alice are always
identical (or always conjugate?) to the ones received by Bob.
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Tom Trotter
Jul26-04, 03:41 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\nps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google. com>...\n> tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google. com>...\n> > ps@ws5.com (Paul Stewart Snyder) wrote in message\n> > > We know that when A is changed B is instantaneously\n> > > changed,\n> >\n> > Only in certain contexts, but they don\'t imply signal\n> > transfer between A and B.\n> >\n> This is what continues to bother me - there are at least some contexts\n> in which there are instantaneous changes to B when A changes - even if\n> A and B are spatially separated. There may not be a signal transfer in\n> the formal terms of information transfer, and EPR may not require an\n> FTL event, yet there still seems to be a strong intuitive basis\n> (Einstein loved intuition) for thinking that something FTL is going on\n> here.\n\nInclude in your intuitive database this example of instantaneous\n(simultaneous) change wrt separated points A and B that doesn\'t\nrequire communication between them.\n\nDraw a circle and put two points at different locations on it.\nRotate the circle. A changes with B,and vice versa, but A\ndoesn\'t affect B, and vice versa. There\'s a non-varying\nphysical connection between A and B, but no \'communication\'.\nTheir motion will be correlated in contexts that refer to their\nphysical connection.\n\nNote that this is not a model of what\'s happening in Bell-test\nphoton experiments. In these experiments the entanglement is of\na different sort. A and B don\'t change together. They don\'t\naffect each other. They\'re not physically connected in any way.\nThey do, however, have a motional property in common (polarization)\ndue to their common source. In observational contexts that refer to\nthis common property, the photons will produce predictable (correlated)\ndetection patterns, functionally related to the combined settings\nof the analyzers.\n\nOr, perhaps you can tell me exactly what your "strong intuitive basis"\nfor considering flt signals is.\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>ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google.com>...
> tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google.com>...
> > ps@ws5.com (Paul Stewart Snyder) wrote in message
> > > We know that when A is changed B is instantaneously
> > > changed,
> >
> > Only in certain contexts, but they don't imply signal
> > transfer between A and B.
> >
> This is what continues to bother me - there are at least some contexts
> in which there are instantaneous changes to B when A changes - even if
> A and B are spatially separated. There may not be a signal transfer in
> the formal terms of information transfer, and EPR may not require an
> FTL event, yet there still seems to be a strong intuitive basis
> (Einstein loved intuition) for thinking that something FTL is going on
> here.
Include in your intuitive database this example of instantaneous
(simultaneous) change wrt separated points A and B that doesn't
require communication between them.
Draw a circle and put two points at different locations on it.
Rotate the circle. A changes with B,and vice versa, but A
doesn't affect B, and vice versa. There's a non-varying
physical connection between A and B, but no 'communication'.
Their motion will be correlated in contexts that refer to their
physical connection.
Note that this is not a model of what's happening in Bell-test
photon experiments. In these experiments the entanglement is of
a different sort. A and B don't change together. They don't
affect each other. They're not physically connected in any way.
They do, however, have a motional property in common (polarization)
due to their common source. In observational contexts that refer to
this common property, the photons will produce predictable (correlated)
detection patterns, functionally related to the combined settings
of the analyzers.
Or, perhaps you can tell me exactly what your "strong intuitive basis"
for considering flt signals is.
Caroline Thompson
Jul26-04, 03:41 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"Tom Trotter" <tom129@juno.com> wrote in message\nnews:29df3039.0407240733.405f6c1a@posting .google.com...\n> "Caroline Thompson" <ch.thompson1@virgin.net> wrote\n\n> > But the assumed relationship is merely that lambda is the\n> > same for each photon in a pair: that lambda_A = lambda_B.\n> > Knowing this has no effect on what happens to the photons\n> > individually.\n>\n> That\'s right. Knowing this relationship will not allow\n> you to more accurately predict individual results,\n> if you\'re just considering A, or just considering B.\n>\n> And, knowing the polarization for photons incident\n> on the polarizers won\'t increase your ability\n> to predict coincidental detections, as long as\n> you know how the photons are related.\n\nBut knowing the result of the measurement at A *does* give you a little\ninformation re the likely result at B. Suppose that you know both\npolarisation directions were the same. If you see H at A then, if\npolarisers are parallel, you know the probability of seeing V at B is low.\nIf you assume Malus Law you can predict using ordinary correlations a\ncoincidence curve that is the same as the QM curve only has half the\nvisibility. (See quant-ph/9903066.) In order to predict a curve nearer to\nthe QM prediction you need to take account of other factors relevant to the\nparticular experiment, though in some instances the predicted (0.5\nvisibility) curve is just what is needed to match the experimental results.\n\n> The \'something extra\' is the emission-dependent *relationship*\n> between photon 1 and photon 2, if what\'s being analyzed is\n> entangled. But, it isn\'t really extra. It\'s just the only\n> thing circumscribed by Bell\'s lambda that\'s relevant\n> to coincidental detection. You just eliminate from\n> consideration all the extraneous stuff, like how the\n> photons are polarized ...\n\nSorry, Tom, but the polarisation direction of the individual photon is all\nyou\'ve got. We\'re not making progress, are we?\n\n> ... photons that are emitted in opposite directions are\n> what are selected. And, if they\'re emitted by the same\n> atom, then they have to be polarized identically,\n> according to the emission model, in order to comply with\n> the law of conservation of angular momentum.\n\nI\'m not disputing this.\n\n> Wrt this scenario, in the combined context, the only variable\n> that matters is the angular difference in polarizer settings.\n\nNot necessarily so. There may, for instance, be more photon pairs polarised\nvertically than at any other angle. The distribution of polarisation angles\nat the source can influence the coincidence pattern. In other words, as I\ntried to explain in the section I\'ve snipped, we may not have rotational\ninvariance. It is usual in practice in Bell test experiments to keep one\npolariser fixed and rotate the other to get a coincidence curve. If we\ndon\'t have rotational invariance then the visibility of the curve will vary\naccording to our choice for the fixed detector orientation. See\nquant-ph/9912082.\n\n> > > ... As long as photon 1 and photon 2 of any given pair are\n> > > identically polarized, then the orientation of the polarizers\n> > > wrt each other is the only relevant variable in the\n> > > observational context, and a circular function of this\n> > > changing angular difference is what determines rates of\n> > > coincidental detection.\n> >\n> > But this is true only if we have rotational invariance.\n>\n> I don\'t think so. Remember, Bell\'s unknown parameter\n> affecting individual results can have any value.\n\nYes, but it can also have any distribution. In the general formulation of a\nBell test setup there is a density function rho controlling this\ndistribution. Though it is often set to a constant, in the most general\nlocal hidden variable model (and in real experiments) it does not have to\nbe. Do have a look at Clauser and Horne\'s 1974 derivation of their\nsingle-channel Bell test. We integrate the product of the two singles\nprobabilities using a weighting factor rho(lambda) that can be *any*\nfunction.\n\n[The CH74 derivation is reproduced in\nhttp://freespace.virgin.net/ch.thompson1/Papers/CH74/CH74assumptions.htm ]\n\n> It can have one value (where singles counts will vary with\n> polarizer settings), or it can be random, ie., have an infinite range\n> of values (where singles counts will not vary with\n> polarizer settings).\n\nThis sounds about right ...\n\n> Bell\'s analysis says that the coincidence rate, cr, for 2Theta,\n> can\'t exceed cr/2 for Theta.\n\nNo, Bell never said that! I think maybe you are following Nick Herbert? I\nhaven\'t read his book but keep on coming upon misconceptions on behalf of\nthose who have.\n\n> But, experimentally, this doesn\'t\n> hold for all tested values of Theta. The qm formulation, on the\n> other hand, says that cr will vary as a circular, rather than\n> a linear, function of Theta -- and experiments bear this out.\n\nYes, and the local realist model appropriate to optical Bell tests says the\nsame. (See quant-ph/9903066)\n\n> > In most real\n> > experiments we find in practice that they keep one\n> > polariser fixed and just vary the other one. The\n> > onlooker is left assuming that it would not have\n> > mattered if they\'d chosen a different orientation for\n> > the fixed polariser.\n>\n> It wouldn\'t. Not for Bell tests anyway.\n\nReal Bell tests allow for the possibility that we don\'t have rotational\ninvariance. The data that forces them to allow for this is not generally\npublished. It would be interesting to see it, but sometimes the very nature\nof the apparatus means that rotational invariance is unlikely.\n\n> > But sometimes this orientation seems to be carefully chosen.\n> > Kwiat, for instance, in one of his well-known experiments\n> > <http://arXiv.org/abs/quant-ph/9810003>, kept one polariser\n> > fixed at 45 deg and varied the other. What would have\n> > happened if he had chosen some other angle? Most models\n> > lead to at least an approximately sinusoidal curve, but\n> > the amplitude of that curve can vary with the choice of the fixed\n> > angle. It is *not* just the difference in detector settings that\n> > matters.\n> >\n>\n> As long as you\'re dealing with opposite moving photons emitted\n> by the same atom, then the angular difference of the settings\n> of the analyzers is the only thing that matters wrt predicting\n> rates of coincidental detection in the combined context.\n\n>From physical considerations this does not make sense. Suppose *all*\nphotons are V. Then if our fixed polariser is set vertically we get a\nbeautiful full-visibility coincidence curve as we vary the other. If, on\nthe other hand, we set it horizontally, we don\'t get any coincidences at\nall, whatever the angle of the other polariser. Another interesting\npossiblity is discussed in quant-ph/9912082 -- that of a population half of\nwhich are polarised vertically, half horizontally.\n\n> > > Even if you separate detections and changes in polarizer\n> > > settings by a spacelike interval, even if you change polarizer\n> > > settings in mid-flight, even if A has registered a detection and\n> > > you then change B\'s polarizer setting -- for any given pair\n> > > of photons there\'s still only one Theta. And, so long as\n> > > photon 1 and photon 2 of any given pair are identically\n> > > polarized via emission, then, as Theta increases, then rate\n> > > of coincidental detection must decrease, and as Theta decreases,\n> > > then rate of coincidental detection must increase, as\n> > > a circular function of Theta.\n> >\n> > True, but the rates of change can vary with a and b separately, not just\n> > with the difference, a - b = theta.\n>\n> I sense that you still haven\'t grabbed hold of the idea that\n> we\'re dealing with two, entirely different, experimental contexts\n> here ...\n\nI\'m afraid I sense that you are not reading what I write! I have studied\nhow various Bell inequalities are derived, and studied quite a number of\nactual experiments.\n\n> This is the key to understanding what Bell\'s analysis, and\n> experimental tests of Bell inequalities, actually tell you (and\n> not what most commentators, including some statements in the\n> sci.physics faq, say they tell you).\n\nI think I know what it\'s all about -- and, perhaps I should remind you, Bell\nwas intending to work only from local realist assumptions. He accidentally\nmade one QM one in his 1964 paper (the assumption that when detectors are\nparallel you *always* get identical results) but later it was found that\nmore general inequalities that did not depend on this assumption can be\nderived.\n\n> Rotational *variance* wouldn\'t alter what Bell\'s analysis\n> reveals (ie., that the parameters that determine individual\n> results are not the same parameters that determine combined\n> results).\n\nWhat derivation of a Bell test are you thinking of? I\'m seriously beginning\nto doubt if it is a valid one.\n\n> > What was Bell worried about then? He had proved that\n> > there were predictions of QM that were incompatible with\n> > the existence of local hidden variables.\n>\n> Maybe that\'s what he was worried about, but that\'s not what\n> he proved. He proved that if you include parameters that\n> would affect individual measurements in a formulation concerning\n> combined measurements, then you get results that are inconsistent\n> with qm\n\nYes, so far so good.\n\n> and experiments,\n\nNo. This is what he hoped could be found out, one way or the other.\n\n> or (after tweaking so that it\'s consistent with the statistical\n> predictions of qm)\n> results that are inconsistent with the assumptions of standard physics\n> regarding a limitation on signal velocity (thereby violating Einstein\n> causality or Lorentz invariance).\n\nNo, his inequality is one that is obeyed under standard physics. No\nloophole-free experiment has ever violated it. Bell had no obligation to\nmake his prediction agree exactly with qm though, as I mentioned earlier,\nhis first inequality was restricted to those local HV theories that did\nagree with qm as regards observations for parallel detectors.\n\n> Either way, you get nonsense. Sort of a reduction to absurdities.\n> The conclusion that follows from this is that the added parameters,\n> while relevant wrt individual contexts, simply aren\'t relevant wrt\n> combined contexts. That\'s all. Nothing has been said about the\n> predictions of qm being inconsistent with the *existence*\n> of hidden variables, or being consistent with the *necessity* of\n> ftl signals.\n\nSorry, Tom, but a lot has been said about the above! The whole point of\nBell\'s theorem is that the predictions of qm are incompatible with the\nexistence of hidden variables.\n\n> The mechanism for A and B \'instantaneously\' communicating is\n> of course \'necessary\' if you want to keep the tweaked-up,\n> sr-violating, qm(exp)-compatible,\n\nYou can\'t equate "qm-compatible" with "exp-compatible". You\'ve jumping the\ngun as regards the experimental evidence.\n\n> added-parameter formulation for the combined context.\n> But, why on earth would anyone want to do *that*? :-)\n\nThere is no need for anything this complicated or unrealistic to explain the\nreal experimental results.\n\n> > > It\'s just that interpretations of the formulation have become\n> > > muddled following misinterpretations of Bell\'s work.\n> >\n> > I agree that there has been a lot of misinterpretation of Bell\'s\n> > work, but this is mainly on the behalf of science journalists.\n> > Judging by his PhD thesis, and known communication between\n> > them, I think Aspect understood Bell very well.\n>\n> I\'m not sure he understood that it\'s *not* their obedience to\n> Einstein causality that makes supplementary-parameter theories\n> incompatible with qm and experiment. The qm formulation, and\n> the experiments, also obey Einstein causality.\n\nHmmm ... If so why did Richard Feynman and Niels Bohr both come out with\nstatements to the effect that nobody understands qm -- that if you think you\ndo then you are mistaken?\n\nAnd I do wish you would not keep on implying that experiments have been\nproved to back qm. They have not. The matter is still open.\n\n> "Nonlocal" has at least four meanings that I know of. The way\n> you\'re using it above is a connotation that, in connection with\n> Bell\'s work, evolved from misinterpretations of it\'s standard\n> meaning -- which is that two events are nonlocally correlated\n> if they\'re functionally related and separated by a spacelike\n> interval. Note that this doesn\'t entail that they\'re communicating\n> in any way. They\'re just related by some function in some,\n> nonlocal, observational context.\n\nTrue. The term is used in relation to the Bell tests in rather a different\nway. It implies something magical -- effects that cannot be explained\nsolely in terms of common causes nor by exchange of signals at any\nreasonable speed.\n\n> The *context* of Aspect\'s 1982 experiment makes it a nonlocal\n> experiment. And, it would seem to follow that theories applicable\n> to nonlocal contexts, would be called nonlocal theories.\n\nNo, this is not a reasonable use of the term. There is a lot of difference\nbetween a nonlocal HV theory such as David Bohm\'s and a local one. Only the\nlatter qualify to be called "realist". It is these that are of interest for\nexplaining real experiments.\n\n> > All we have in the Bell tests is lambda values set at the source\n> > and carried with the photons to their respective detectors. A\n> > purely local action takes place at each detector.\n>\n> Well, that\'s not *all* you have. You\'re forgetting the observational\n> context. :-)\n\nWhat is this "observational context"? Remember that in real life the two\nstations may, as in Tittel\'s experiments at Geneva, be separated by several\nmiles.\n\n> > ... It is very unfortunate that, as far as I know, no\n> > experimenter has ever seen fit to publish coincidence curves for\n> > different fixed positions for one of the detectors.\n>\n> Because it\'s not relevant to the tests being done.\n> There\'s two main things they concentrate on afaik:\n> (1) dealing with photons emitted by the same\n> atom in any given coincidence/correlation window, and\n> (2) an accurate assessment of the efficiencies\n> of the polarizers and detectors.\n\nNot all experimenters seem to be aware of what rotational invariance means.\nThey do not look at things from the realist point of view, which makes the\nexperiments almost inevitably biased towards qm. Think of medical trials\n....\n\nIncidentally, though they carefully try and assess their quantum\nefficiencies, they do not, as far as I can tell, carry out the studies that\na local realist would like to see: they do not study the relationship\nbetween input intensity and probability of detection. If such studies are\nperformed I\'d be interested to know where they are published. It is clear\nthat the relationship cannot be a matter of perfect linear correlation. The\ncurve cannot pass through the origin since there are known "dark counts"\neven when the input intentity is zero. It cannot be linear except\napproximately, over a limited range, since the detector will get saturated\nat high intensities.\n\nCheers\nCaroline\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Tom Trotter" <tom129@juno.com> wrote in message
news:29df3039.0407240733.405f6c1a@posting.google.c om...
> "Caroline Thompson" <ch.thompson1@virgin.net> wrote
> > But the assumed relationship is merely that \lambda is the
> > same for each photon in a pair: that \lambda_A = \lambda_B.
> > Knowing this has no effect on what happens to the photons
> > individually.
>
> That's right. Knowing this relationship will not allow
> you to more accurately predict individual results,
> if you're just considering A, or just considering B.
>
> And, knowing the polarization for photons incident
> on the polarizers won't increase your ability
> to predict coincidental detections, as long as
> you know how the photons are related.
But knowing the result of the measurement at A *does* give you a little
information re the likely result at B. Suppose that you know both
polarisation directions were the same. If you see H at A then, if
polarisers are parallel, you know the probability of seeing V at B is low.
If you assume Malus Law you can predict using ordinary correlations a
coincidence curve that is the same as the QM curve only has half the
visibility. (See http://www.arxiv.org/abs/quant-ph/9903066.) In order to predict a curve nearer to
the QM prediction you need to take account of other factors relevant to the
particular experiment, though in some instances the predicted (0.5
visibility) curve is just what is needed to match the experimental results.
> The 'something extra' is the emission-dependent *relationship*
> between photon 1 and photon 2, if what's being analyzed is
> entangled. But, it isn't really extra. It's just the only
> thing circumscribed by Bell's \lambda that's relevant
> to coincidental detection. You just eliminate from
> consideration all the extraneous stuff, like how the
> photons are polarized ...
Sorry, Tom, but the polarisation direction of the individual photon is all
you've got. We're not making progress, are we?
> ... photons that are emitted in opposite directions are
> what are selected. And, if they're emitted by the same
> atom, then they have to be polarized identically,
> according to the emission model, in order to comply with
> the law of conservation of angular momentum.
I'm not disputing this.
> Wrt this scenario, in the combined context, the only variable
> that matters is the angular difference in polarizer settings.
Not necessarily so. There may, for instance, be more photon pairs polarised
vertically than at any other angle. The distribution of polarisation angles
at the source can influence the coincidence pattern. In other words, as I
tried to explain in the section I've snipped, we may not have rotational
invariance. It is usual in practice in Bell test experiments to keep one
polariser fixed and rotate the other to get a coincidence curve. If we
don't have rotational invariance then the visibility of the curve will vary
according to our choice for the fixed detector orientation. See
http://www.arxiv.org/abs/quant-ph/9912082.
> > > ... As long as photon 1 and photon 2 of any given pair are
> > > identically polarized, then the orientation of the polarizers
> > > wrt each other is the only relevant variable in the
> > > observational context, and a circular function of this
> > > changing angular difference is what determines rates of
> > > coincidental detection.
> >
> > But this is true only if we have rotational invariance.
>
> I don't think so. Remember, Bell's unknown parameter
> affecting individual results can have any value.
Yes, but it can also have any distribution. In the general formulation of a
Bell test setup there is a density function \rho controlling this
distribution. Though it is often set to a constant, in the most general
local hidden variable model (and in real experiments) it does not have to
be. Do have a look at Clauser and Horne's 1974 derivation of their
single-channel Bell test. We integrate the product of the two singles
probabilities using a weighting factor \rho(\lambda) that can be *any*
function.
[The CH74 derivation is reproduced in
http://freespace.virgin.net/ch.thompson1/Papers/CH74/CH74assumptions.htm ]
> It can have one value (where singles counts will vary with
> polarizer settings), or it can be random, ie., have an infinite range
> of values (where singles counts will not vary with
> polarizer settings).
This sounds about right ...
> Bell's analysis says that the coincidence rate, cr, for 2Theta,
> can't exceed cr/2 for \Theta.
No, Bell never said that! I think maybe you are following Nick Herbert? I
haven't read his book but keep on coming upon misconceptions on behalf of
those who have.
> But, experimentally, this doesn't
> hold for all tested values of \Theta. The qm formulation, on the
> other hand, says that cr will vary as a circular, rather than
> a linear, function of \Theta -- and experiments bear this out.
Yes, and the local realist model appropriate to optical Bell tests says the
same. (See http://www.arxiv.org/abs/quant-ph/9903066)
> > In most real
> > experiments we find in practice that they keep one
> > polariser fixed and just vary the other one. The
> > onlooker is left assuming that it would not have
> > mattered if they'd chosen a different orientation for
> > the fixed polariser.
>
> It wouldn't. Not for Bell tests anyway.
Real Bell tests allow for the possibility that we don't have rotational
invariance. The data that forces them to allow for this is not generally
published. It would be interesting to see it, but sometimes the very nature
of the apparatus means that rotational invariance is unlikely.
> > But sometimes this orientation seems to be carefully chosen.
> > Kwiat, for instance, in one of his well-known experiments
> > <http://arXiv.org/abs/http://www.arxiv.org/abs/quant-ph/9810003>, kept one polariser
> > fixed at 45 deg and varied the other. What would have
> > happened if he had chosen some other angle? Most models
> > lead to at least an approximately sinusoidal curve, but
> > the amplitude of that curve can vary with the choice of the fixed
> > angle. It is *not* just the difference in detector settings that
> > matters.
> >
>
> As long as you're dealing with opposite moving photons emitted
> by the same atom, then the angular difference of the settings
> of the analyzers is the only thing that matters wrt predicting
> rates of coincidental detection in the combined context.
>From physical considerations this does not make sense. Suppose *all*
photons are V. Then if our fixed polariser is set vertically we get a
beautiful full-visibility coincidence curve as we vary the other. If, on
the other hand, we set it horizontally, we don't get any coincidences at
all, whatever the angle of the other polariser. Another interesting
possiblity is discussed in http://www.arxiv.org/abs/quant-ph/9912082 -- that of a population half of
which are polarised vertically, half horizontally.
> > > Even if you separate detections and changes in polarizer
> > > settings by a spacelike interval, even if you change polarizer
> > > settings in mid-flight, even if A has registered a detection and
> > > you then change B's polarizer setting -- for any given pair
> > > of photons there's still only one \Theta. And, so long as
> > > photon 1 and photon 2 of any given pair are identically
> > > polarized via emission, then, as \Theta increases, then rate
> > > of coincidental detection must decrease, and as \Theta decreases,
> > > then rate of coincidental detection must increase, as
> > > a circular function of \Theta.
> >
> > True, but the rates of change can vary with a and b separately, not just
> > with the difference, a - b = \theta.
>
> I sense that you still haven't grabbed hold of the idea that
> we're dealing with two, entirely different, experimental contexts
> here ...
I'm afraid I sense that you are not reading what I write! I have studied
how various Bell inequalities are derived, and studied quite a number of
actual experiments.
> This is the key to understanding what Bell's analysis, and
> experimental tests of Bell inequalities, actually tell you (and
> not what most commentators, including some statements in the
> sci.physics faq, say they tell you).
I think I know what it's all about -- and, perhaps I should remind you, Bell
was intending to work only from local realist assumptions. He accidentally
made one QM one in his 1964 paper (the assumption that when detectors are
parallel you *always* get identical results) but later it was found that
more general inequalities that did not depend on this assumption can be
derived.
> Rotational *variance* wouldn't alter what Bell's analysis
> reveals (ie., that the parameters that determine individual
> results are not the same parameters that determine combined
> results).
What derivation of a Bell test are you thinking of? I'm seriously beginning
to doubt if it is a valid one.
> > What was Bell worried about then? He had proved that
> > there were predictions of QM that were incompatible with
> > the existence of local hidden variables.
>
> Maybe that's what he was worried about, but that's not what
> he proved. He proved that if you include parameters that
> would affect individual measurements in a formulation concerning
> combined measurements, then you get results that are inconsistent
> with qm
Yes, so far so good.
> and experiments,
No. This is what he hoped could be found out, one way or the other.
> or (after tweaking so that it's consistent with the statistical
> predictions of qm)
> results that are inconsistent with the assumptions of standard physics
> regarding a limitation on signal velocity (thereby violating Einstein
> causality or Lorentz invariance).
No, his inequality is one that is obeyed under standard physics. No
loophole-free experiment has ever violated it. Bell had no obligation to
make his prediction agree exactly with qm though, as I mentioned earlier,
his first inequality was restricted to those local HV theories that did
agree with qm as regards observations for parallel detectors.
> Either way, you get nonsense. Sort of a reduction to absurdities.
> The conclusion that follows from this is that the added parameters,
> while relevant wrt individual contexts, simply aren't relevant wrt
> combined contexts. That's all. Nothing has been said about the
> predictions of qm being inconsistent with the *existence*
> of hidden variables, or being consistent with the *necessity* of
> ftl signals.
Sorry, Tom, but a lot has been said about the above! The whole point of
Bell's theorem is that the predictions of qm are incompatible with the
existence of hidden variables.
> The mechanism for A and B 'instantaneously' communicating is
> of course 'necessary' if you want to keep the tweaked-up,
> sr-violating, qm(\exp)-compatible,
You can't equate "qm-compatible" with "\exp-compatible". You've jumping the
gun as regards the experimental evidence.
> added-parameter formulation for the combined context.
> But, why on earth would anyone want to do *that*? :-)
There is no need for anything this complicated or unrealistic to explain the
real experimental results.
> > > It's just that interpretations of the formulation have become
> > > muddled following misinterpretations of Bell's work.
> >
> > I agree that there has been a lot of misinterpretation of Bell's
> > work, but this is mainly on the behalf of science journalists.
> > Judging by his PhD thesis, and known communication between
> > them, I think Aspect understood Bell very well.
>
> I'm not sure he understood that it's *not* their obedience to
> Einstein causality that makes supplementary-parameter theories
> incompatible with qm and experiment. The qm formulation, and
> the experiments, also obey Einstein causality.
Hmmm ... If so why did Richard Feynman and Niels Bohr both come out with
statements to the effect that nobody understands qm -- that if you think you
do then you are mistaken?
And I do wish you would not keep on implying that experiments have been
proved to back qm. They have not. The matter is still open.
> "Nonlocal" has at least four meanings that I know of. The way
> you're using it above is a connotation that, in connection with
> Bell's work, evolved from misinterpretations of it's standard
> meaning -- which is that two events are nonlocally correlated
> if they're functionally related and separated by a spacelike
> interval. Note that this doesn't entail that they're communicating
> in any way. They're just related by some function in some,
> nonlocal, observational context.
True. The term is used in relation to the Bell tests in rather a different
way. It implies something magical -- effects that cannot be explained
solely in terms of common causes nor by exchange of signals at any
reasonable speed.
> The *context* of Aspect's 1982 experiment makes it a nonlocal
> experiment. And, it would seem to follow that theories applicable
> to nonlocal contexts, would be called nonlocal theories.
No, this is not a reasonable use of the term. There is a lot of difference
between a nonlocal HV theory such as David Bohm's and a local one. Only the
latter qualify to be called "realist". It is these that are of interest for
explaining real experiments.
> > All we have in the Bell tests is \lambda values set at the source
> > and carried with the photons to their respective detectors. A
> > purely local action takes place at each detector.
>
> Well, that's not *all* you have. You're forgetting the observational
> context. :-)
What is this "observational context"? Remember that in real life the two
stations may, as in Tittel's experiments at Geneva, be separated by several
miles.
> > ... It is very unfortunate that, as far as I know, no
> > experimenter has ever seen fit to publish coincidence curves for
> > different fixed positions for one of the detectors.
>
> Because it's not relevant to the tests being done.
> There's two main things they concentrate on afaik:
> (1) dealing with photons emitted by the same
> atom in any given coincidence/correlation window, and
> (2) an accurate assessment of the efficiencies
> of the polarizers and detectors.
Not all experimenters seem to be aware of what rotational invariance means.
They do not look at things from the realist point of view, which makes the
experiments almost inevitably biased towards qm. Think of medical trials
....
Incidentally, though they carefully try and assess their quantum
efficiencies, they do not, as far as I can tell, carry out the studies that
a local realist would like to see: they do not study the relationship
between input intensity and probability of detection. If such studies are
performed I'd be interested to know where they are published. It is clear
that the relationship cannot be a matter of perfect linear correlation. The
curve cannot pass through the origin since there are known "dark counts"
even when the input intentity is zero. It cannot be linear except
approximately, over a limited range, since the detector will get saturated
at high intensities.
Cheers
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Paul Stewart Snyder
Jul28-04, 03:58 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>\n\n\n\ntom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407251124.1cd7a038@posting.google. com>...\n> ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google. com>...\n> > tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google. com>...\n> > > ps@ws5.com (Paul Stewart Snyder) wrote in message\n> > > > We know that when A is changed B is instantaneously\n> > > > changed,\n> > >\n> > > Only in certain contexts, but they don\'t imply signal\n> > > transfer between A and B.\n> > >\n> > This is what continues to bother me - there are at least some contexts\n> > in which there are instantaneous changes to B when A changes - even if\n> > A and B are spatially separated. There may not be a signal transfer in\n> > the formal terms of information transfer, and EPR may not require an\n> > FTL event, yet there still seems to be a strong intuitive basis\n> > (Einstein loved intuition) for thinking that something FTL is going on\n> > here.\n>\n> Draw a circle and put two points at different locations on it.\n> Rotate the circle. A changes with B,and vice versa, but A\n> doesn\'t affect B, and vice versa. There\'s a non-varying\n> physical connection between A and B, but no \'communication\'.\n> Their motion will be correlated in contexts that refer to their\n> physical connection.\n>\n> Note that this is not a model of what\'s happening in Bell-test\n> photon experiments.\n\nYou are right, the example is not related to quantum entanglement.\n\n> In these experiments the entanglement is of\n> a different sort. A and B don\'t change together. They don\'t\n> affect each other. They\'re not physically connected in any way.\n> They do, however, have a motional property in common (polarization)\n> due to their common source. In observational contexts that refer to\n> this common property, the photons will produce predictable (correlated)\n> detection patterns, functionally related to the combined settings\n> of the analyzers.\n\nYou are describing one way to understand entanglement, if I understand\ncorrectly you require observation of both A and B to produce entanglement,\nso that the correlated patterns are related to the "combined settings of the\nanalyzers", being the "observers" in a Copenhagen Interpretation sense?\nThat may be the case – however, unless I have missed something, I do\nnot think we have a definitive understanding of the role of observation,\nif any, in quantum measurement? As far as I can tell even Cramer\'s\nunusual Transactional Interpretation of Quantum Mechanics has not been\nruled out as a possibility. So I don\'t believe we can say that the polarization\nproperty may be limited to "observational contexts". If state vector reduction\nturns out to be unrelated to observation (and/or if many-universe models are\ncorrect), then there is an instantaneous physical change in B when A changes\nthat is separate and apart from any observation. That possibility is my source\nfor my intuitive feelings that something FTL may be happening.\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>tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407251124.1cd7a038@posting.google.com>...
> ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google.com>...
> > tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google.com>...
> > > ps@ws5.com (Paul Stewart Snyder) wrote in message
> > > > We know that when A is changed B is instantaneously
> > > > changed,
> > >
> > > Only in certain contexts, but they don't imply signal
> > > transfer between A and B.
> > >
> > This is what continues to bother me - there are at least some contexts
> > in which there are instantaneous changes to B when A changes - even if
> > A and B are spatially separated. There may not be a signal transfer in
> > the formal terms of information transfer, and EPR may not require an
> > FTL event, yet there still seems to be a strong intuitive basis
> > (Einstein loved intuition) for thinking that something FTL is going on
> > here.
>
> Draw a circle and put two points at different locations on it.
> Rotate the circle. A changes with B,and vice versa, but A
> doesn't affect B, and vice versa. There's a non-varying
> physical connection between A and B, but no 'communication'.
> Their motion will be correlated in contexts that refer to their
> physical connection.
>
> Note that this is not a model of what's happening in Bell-test
> photon experiments.
You are right, the example is not related to quantum entanglement.
> In these experiments the entanglement is of
> a different sort. A and B don't change together. They don't
> affect each other. They're not physically connected in any way.
> They do, however, have a motional property in common (polarization)
> due to their common source. In observational contexts that refer to
> this common property, the photons will produce predictable (correlated)
> detection patterns, functionally related to the combined settings
> of the analyzers.
You are describing one way to understand entanglement, if I understand
correctly you require observation of both A and B to produce entanglement,
so that the correlated patterns are related to the "combined settings of the
analyzers", being the "observers" in a Copenhagen Interpretation sense?
That may be the case – however, unless I have missed something, I do
not think we have a definitive understanding of the role of observation,
if any, in quantum measurement? As far as I can tell even Cramer's
unusual Transactional Interpretation of Quantum Mechanics has not been
ruled out as a possibility. So I don't believe we can say that the polarization
property may be limited to "observational contexts". If state vector reduction
turns out to be unrelated to observation (and/or if many-universe models are
correct), then there is an instantaneous physical change in B when A changes
that is separate and apart from any observation. That possibility is my source
for my intuitive feelings that something FTL may be happening.
Tom Trotter
Jul28-04, 03:59 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"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<GvTMc.551\\$Go2.507@newsfe3-gui.ntli.net>...\n> "Tom Trotter" <tom129@juno.com> wrote in message\n> news:29df3039.0407240733.405f6c1a@posting.google.c om...\n> > "Caroline Thompson" <ch.thompson1@virgin.net> wrote\n>\n> > > But the assumed relationship is merely that lambda is the\n> > > same for each photon in a pair: that lambda_A = lambda_B.\n> > > Knowing this has no effect on what happens to the photons\n> > > individually.\n> >\n> > That\'s right. Knowing this relationship will not allow\n> > you to more accurately predict individual results,\n> > if you\'re just considering A, or just considering B.\n> >\n> > And, knowing the polarization for photons incident\n> > on the polarizers won\'t increase your ability\n> > to predict coincidental detections, as long as\n> > you know how the photons are related.\n>\n> But knowing the result of the measurement at A *does*\n> give you a little information re the likely result at B.\n\nThen you wouldn\'t be just considering A or just\nconsidering B, would you?\n\n[snip]\n\n> > The \'something extra\' is the emission-dependent *relationship*\n> > between photon 1 and photon 2, if what\'s being analyzed is\n> > entangled. But, it isn\'t really extra. It\'s just the only\n> > thing circumscribed by Bell\'s lambda that\'s relevant\n> > to coincidental detection. You just eliminate from\n> > consideration all the extraneous stuff, like how the\n> > photons are polarized ...\n>\n> Sorry, Tom, but the polarisation direction of the individual\n> photon is all you\'ve got. We\'re not making progress, are we?\n\nWhat you have (empirically) is that a detector registered\nduring a coincidence window, or it didn\'t.\n\nFor increasing the accuracy of predictions of individual\nresults, you would need to know the emission-produced polarization.\nBut, this is never known.\n\nFor accurately predicting coincidence rates in the combined\ncontext, you don\'t need to know the emission-produced\npolarization. You just need to know that they\'re polarized\nidentically -- hence the efforts made to ensure that\ncoincidence counters are dealing with photons emitted by\nthe same atom wrt any given coincidence window.\n\n>\n> > ... photons that are emitted in opposite directions are\n> > what are selected. And, if they\'re emitted by the same\n> > atom, then they have to be polarized identically,\n> > according to the emission model, in order to comply with\n> > the law of conservation of angular momentum.\n>\n> I\'m not disputing this.\n>\n> > Wrt this scenario, in the combined context, the only variable\n> > that matters is the angular difference in polarizer settings.\n>\n> Not necessarily so. There may, for instance, be more photon\n> pairs polarised vertically than at any other angle.\n\n[snip]\n\nIt doesn\'t matter. In the combined context, it\'s not the\npolarization, but the relationship between photon 1 and\nphoton 2 that\'s being observed.\n\n>\n> > > > ... As long as photon 1 and photon 2 of any given pair are\n> > > > identically polarized, then the orientation of the polarizers\n> > > > wrt each other is the only relevant variable in the\n> > > > observational context, and a circular function of this\n> > > > changing angular difference is what determines rates of\n> > > > coincidental detection.\n> > >\n> > > But this is true only if we have rotational invariance.\n> >\n> > I don\'t think so. Remember, Bell\'s unknown parameter\n> > affecting individual results can have any value.\n>\n> Yes, but it can also have any distribution.\n\n[snip]\n\nThat\'s right, and the point is that this distribution\n(re polarization) is irrelevant wrt determining coincidence\nrates as long as photon 1 and photon 2 of an emitted\npair are polarized identically.\n\n[snip]\n\n> > Bell\'s analysis says that the coincidence rate, cr, for 2Theta,\n> > can\'t exceed cr/2 for Theta.\n>\n> No, Bell never said that! I think maybe you are following Nick Herbert? I\n> haven\'t read his book but keep on coming upon misconceptions on behalf of\n> those who have.\n\nIt\'s a corollary of Bell\'s theorem.\n\n>\n> > But, experimentally, this doesn\'t\n> > hold for all tested values of Theta. The qm formulation, on the\n> > other hand, says that cr will vary as a circular, rather than\n> > a linear, function of Theta -- and experiments bear this out.\n>\n> Yes, and the local realist model appropriate to optical Bell tests says the\n> same. (See quant-ph/9903066)\n\nThat\'s because the qm formulation doesn\'t violate local\nreality. It\'s just that some interpretations of it say\nthat it does.\n\n[snip]\n\n> > As long as you\'re dealing with opposite moving photons emitted\n> > by the same atom, then the angular difference of the settings\n> > of the analyzers is the only thing that matters wrt predicting\n> > rates of coincidental detection in the combined context.\n>\n> >From physical considerations this does not make sense. Suppose *all*\n> photons are V. Then if our fixed polariser is set vertically we get a\n> beautiful full-visibility coincidence curve as we vary the other. If, on\n> the other hand, we set it horizontally, we don\'t get any coincidences at\n> all, whatever the angle of the other polariser.\n\nNot so. The coincidence curve will still be a circular function\nof the angular difference between the polarizer settings (Theta), as\nlong\nas photon 1 and photon 2 of any given pair are polarized identically.\nA mutual non-detection is also a coincidence.\n\nNo matter what the polarization of any given pair is, the\nrate of coincidental detection attributes will increase as\nTheta decreases, and will decrease as Theta increases.\n\n[snip]\n\n\n> > Maybe that\'s what he was worried about, but that\'s not what\n> > he proved. He proved that if you include parameters that\n> > would affect individual measurements in a formulation concerning\n> > combined measurements, then you get results that are inconsistent\n> > with qm\n> > or (after tweaking so that it\'s consistent with the statistical\n> > predictions of qm)\n> > results that are inconsistent with the assumptions of standard physics\n> > regarding a limitation on signal velocity (thereby violating Einstein\n> > causality or Lorentz invariance).\n>\n> No, his inequality is one that is obeyed under standard physics.\n\nBell adjusted his formulation so that it would be\nconsistent with the predictions of qm. In this formulation\nthe setting at a can affect the result at B, and the setting\nat b can affect the result at A -- which would require it, in\nBell\'s words, to be "non Lorentz invariant".\n\n[snip]\n\n> > Either way, you get nonsense. Sort of a reduction to absurdities.\n> > The conclusion that follows from this is that the added parameters,\n> > while relevant wrt individual contexts, simply aren\'t relevant wrt\n> > combined contexts. That\'s all. Nothing has been said about the\n> > predictions of qm being inconsistent with the *existence*\n> > of hidden variables, or being consistent with the *necessity* of\n> > ftl signals.\n>\n> Sorry, Tom, but a lot has been said about the above! The whole point of\n> Bell\'s theorem is that the predictions of qm are incompatible with the\n> existence of hidden variables.\n\nThat\'s the normal interpretation, and, as stated, it\'s just\nwrong.\n\nThe hidden variables that would allow for more accurate\npredictions of individual results don\'t vanish from existence\njust because you\'re looking at things wrt a different\ncontext. It\'s just that they\'re not relevant in the combined\ncontext.\n\nPredictions of qm wrt combined contexts such as Bell tests\nare inconsistent with those of formulations that include a term,\neg., lambda, that is irrelevant wrt determining coincidental\ndetection. The qm formulation, taken by itself without reference to\nmisinterpretations of Bell\'s work, isn\'t incompatible with\nEinstein locality or the existence of hidden variables.\n(See the emission models.)\n\n[snip]\n\n> > >\n> > > I agree that there has been a lot of misinterpretation of Bell\'s\n> > > work, but this is mainly on the behalf of science journalists.\n> > > Judging by his PhD thesis, and known communication between\n> > > them, I think Aspect understood Bell very well.\n> >\n> > I\'m not sure he understood that it\'s *not* their obedience to\n> > Einstein causality that makes supplementary-parameter theories\n> > incompatible with qm and experiment. The qm formulation, and\n> > the experiments, also obey Einstein causality.\n>\n> Hmmm ... If so why did Richard Feynman and Niels Bohr both come out with\n> statements to the effect that nobody understands qm -- that if you think you\n> do then you are mistaken?\n\nWhat\'s this got to do with ... anything?\n\n[snip]\n\n> > > All we have in the Bell tests is lambda values set at the source\n> > > and carried with the photons to their respective detectors. A\n> > > purely local action takes place at each detector.\n> >\n> > Well, that\'s not *all* you have. You\'re forgetting the observational\n> > context. :-)\n>\n> What is this "observational context"? Remember that in real life the two\n> stations may, as in Tittel\'s experiments at Geneva, be separated by several\n> miles.\n\nAs long as you can preserve the emission-produced relationship\n(the entanglement of photons emitted by the same atom), then\nit doesn\'t matter how far apart they are when you analyze\nthem. Observation context doesn\'t necessarily have anything\nto do with distance.\n\nThe "observational context" is (1) opposite moving photons correlated\nin polarization via emission, and (2) analysis of this common\npolarization\n(this *relationship") for paired photons via corresponding, joint\nsettings of separated polarizers.\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>"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<GvTMc.551$Go2.507@newsfe3-gui.ntli.net>...
> "Tom Trotter" <tom129@juno.com> wrote in message
> news:29df3039.0407240733.405f6c1a@posting.google.c om...
> > "Caroline Thompson" <ch.thompson1@virgin.net> wrote
>
> > > But the assumed relationship is merely that \lambda is the
> > > same for each photon in a pair: that \lambda_A = \lambda_B.
> > > Knowing this has no effect on what happens to the photons
> > > individually.
> >
> > That's right. Knowing this relationship will not allow
> > you to more accurately predict individual results,
> > if you're just considering A, or just considering B.
> >
> > And, knowing the polarization for photons incident
> > on the polarizers won't increase your ability
> > to predict coincidental detections, as long as
> > you know how the photons are related.
>
> But knowing the result of the measurement at A *does*
> give you a little information re the likely result at B.
Then you wouldn't be just considering A or just
considering B, would you?
[snip]
> > The 'something extra' is the emission-dependent *relationship*
> > between photon 1 and photon 2, if what's being analyzed is
> > entangled. But, it isn't really extra. It's just the only
> > thing circumscribed by Bell's \lambda that's relevant
> > to coincidental detection. You just eliminate from
> > consideration all the extraneous stuff, like how the
> > photons are polarized ...
>
> Sorry, Tom, but the polarisation direction of the individual
> photon is all you've got. We're not making progress, are we?
What you have (empirically) is that a detector registered
during a coincidence window, or it didn't.
For increasing the accuracy of predictions of individual
results, you would need to know the emission-produced polarization.
But, this is never known.
For accurately predicting coincidence rates in the combined
context, you don't need to know the emission-produced
polarization. You just need to know that they're polarized
identically -- hence the efforts made to ensure that
coincidence counters are dealing with photons emitted by
the same atom wrt any given coincidence window.
>
> > ... photons that are emitted in opposite directions are
> > what are selected. And, if they're emitted by the same
> > atom, then they have to be polarized identically,
> > according to the emission model, in order to comply with
> > the law of conservation of angular momentum.
>
> I'm not disputing this.
>
> > Wrt this scenario, in the combined context, the only variable
> > that matters is the angular difference in polarizer settings.
>
> Not necessarily so. There may, for instance, be more photon
> pairs polarised vertically than at any other angle.
[snip]
It doesn't matter. In the combined context, it's not the
polarization, but the relationship between photon 1 and
photon 2 that's being observed.
>
> > > > ... As long as photon 1 and photon 2 of any given pair are
> > > > identically polarized, then the orientation of the polarizers
> > > > wrt each other is the only relevant variable in the
> > > > observational context, and a circular function of this
> > > > changing angular difference is what determines rates of
> > > > coincidental detection.
> > >
> > > But this is true only if we have rotational invariance.
> >
> > I don't think so. Remember, Bell's unknown parameter
> > affecting individual results can have any value.
>
> Yes, but it can also have any distribution.
[snip]
That's right, and the point is that this distribution
(re polarization) is irrelevant wrt determining coincidence
rates as long as photon 1 and photon 2 of an emitted
pair are polarized identically.
[snip]
> > Bell's analysis says that the coincidence rate, cr, for 2Theta,
> > can't exceed cr/2 for \Theta.
>
> No, Bell never said that! I think maybe you are following Nick Herbert? I
> haven't read his book but keep on coming upon misconceptions on behalf of
> those who have.
It's a corollary of Bell's theorem.
>
> > But, experimentally, this doesn't
> > hold for all tested values of \Theta. The qm formulation, on the
> > other hand, says that cr will vary as a circular, rather than
> > a linear, function of \Theta -- and experiments bear this out.
>
> Yes, and the local realist model appropriate to optical Bell tests says the
> same. (See http://www.arxiv.org/abs/quant-ph/9903066)
That's because the qm formulation doesn't violate local
reality. It's just that some interpretations of it say
that it does.
[snip]
> > As long as you're dealing with opposite moving photons emitted
> > by the same atom, then the angular difference of the settings
> > of the analyzers is the only thing that matters wrt predicting
> > rates of coincidental detection in the combined context.
>
> >From physical considerations this does not make sense. Suppose *all*
> photons are V. Then if our fixed polariser is set vertically we get a
> beautiful full-visibility coincidence curve as we vary the other. If, on
> the other hand, we set it horizontally, we don't get any coincidences at
> all, whatever the angle of the other polariser.
Not so. The coincidence curve will still be a circular function
of the angular difference between the polarizer settings (\Theta), as
long
as photon 1 and photon 2 of any given pair are polarized identically.
A mutual non-detection is also a coincidence.
No matter what the polarization of any given pair is, the
rate of coincidental detection attributes will increase as
\Theta decreases, and will decrease as \Theta increases.
[snip]
> > Maybe that's what he was worried about, but that's not what
> > he proved. He proved that if you include parameters that
> > would affect individual measurements in a formulation concerning
> > combined measurements, then you get results that are inconsistent
> > with qm
> > or (after tweaking so that it's consistent with the statistical
> > predictions of qm)
> > results that are inconsistent with the assumptions of standard physics
> > regarding a limitation on signal velocity (thereby violating Einstein
> > causality or Lorentz invariance).
>
> No, his inequality is one that is obeyed under standard physics.
Bell adjusted his formulation so that it would be
consistent with the predictions of qm. In this formulation
the setting at a can affect the result at B, and the setting
at b can affect the result at A -- which would require it, in
Bell's words, to be "non Lorentz invariant".
[snip]
> > Either way, you get nonsense. Sort of a reduction to absurdities.
> > The conclusion that follows from this is that the added parameters,
> > while relevant wrt individual contexts, simply aren't relevant wrt
> > combined contexts. That's all. Nothing has been said about the
> > predictions of qm being inconsistent with the *existence*
> > of hidden variables, or being consistent with the *necessity* of
> > ftl signals.
>
> Sorry, Tom, but a lot has been said about the above! The whole point of
> Bell's theorem is that the predictions of qm are incompatible with the
> existence of hidden variables.
That's the normal interpretation, and, as stated, it's just
wrong.
The hidden variables that would allow for more accurate
predictions of individual results don't vanish from existence
just because you're looking at things wrt a different
context. It's just that they're not relevant in the combined
context.
Predictions of qm wrt combined contexts such as Bell tests
are inconsistent with those of formulations that include a term,
eg., \lambda, that is irrelevant wrt determining coincidental
detection. The qm formulation, taken by itself without reference to
misinterpretations of Bell's work, isn't incompatible with
Einstein locality or the existence of hidden variables.
(See the emission models.)
[snip]
> > >
> > > I agree that there has been a lot of misinterpretation of Bell's
> > > work, but this is mainly on the behalf of science journalists.
> > > Judging by his PhD thesis, and known communication between
> > > them, I think Aspect understood Bell very well.
> >
> > I'm not sure he understood that it's *not* their obedience to
> > Einstein causality that makes supplementary-parameter theories
> > incompatible with qm and experiment. The qm formulation, and
> > the experiments, also obey Einstein causality.
>
> Hmmm ... If so why did Richard Feynman and Niels Bohr both come out with
> statements to the effect that nobody understands qm -- that if you think you
> do then you are mistaken?
What's this got to do with ... anything?
[snip]
> > > All we have in the Bell tests is \lambda values set at the source
> > > and carried with the photons to their respective detectors. A
> > > purely local action takes place at each detector.
> >
> > Well, that's not *all* you have. You're forgetting the observational
> > context. :-)
>
> What is this "observational context"? Remember that in real life the two
> stations may, as in Tittel's experiments at Geneva, be separated by several
> miles.
As long as you can preserve the emission-produced relationship
(the entanglement of photons emitted by the same atom), then
it doesn't matter how far apart they are when you analyze
them. Observation context doesn't necessarily have anything
to do with distance.
The "observational context" is (1) opposite moving photons correlated
in polarization via emission, and (2) analysis of this common
polarization
(this *relationship") for paired photons via corresponding, joint
settings of separated polarizers.
Caroline Thompson
Jul29-04, 04:59 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\n"Paul Stewart Snyder" <ps@ws5.com> wrote in message\nnews:d16eb5af.0407272017.2eb35f58@posting .google.com...\n> tom129@juno.com (Tom Trotter) wrote\n\n> > Note that this is not a model of what\'s happening in Bell-test\n> > photon experiments.\n>\n> You are right, the example is not related to quantum entanglement.\n>\n> > In these experiments the entanglement is of\n> > a different sort. A and B don\'t change together. They don\'t\n> > affect each other. They\'re not physically connected in any way.\n> > They do, however, have a motional property in common (polarization)\n> > due to their common source. In observational contexts that refer to\n> > this common property, the photons will produce predictable\n> > (correlated) detection patterns, functionally related to the\n> > combined settings of the analyzers.\n>\n> You are describing one way to understand entanglement ...\n\nYes, but how about understanding the actual experiments? If you look at\nany real experiment it will surely give you a totally different perspective\non things. You\'ve got real light, interacting one "photon" at a time with a\ndetector, real electrical signals going from the detector to the coincidence\nmonitor, real decisions being made on the basis of individual photon pairs.\nUnless you can account for how the final statistical results are built up\nfrom these individual events you do not have a sound physical theory for the\nexperiment.\n\n> if I understand correctly you require observation of both A and B\n> to produce entanglement, so that the correlated patterns are related\n> to the "combined settings of the analyzers", being the "observers"\n> in a Copenhagen Interpretation sense? That may be the case -\n> however, unless I have missed something, I do not think we have a\n> definitive understanding of the role of observation,\n> if any, in quantum measurement?\n\n[Who am I to argue with that?]\n\n> As far as I can tell even Cramer\'s\n> unusual Transactional Interpretation of Quantum Mechanics has\n> not been ruled out as a possibility. So I don\'t believe we can\n> say that the polarization property may be limited to "observational\n> contexts".\n\n[whatever that may mean]\n\n> If state vector reduction turns out to be unrelated to observation\n> (and/or if many-universe models are correct), then there is\n> an instantaneous physical change in B when A changes\n> that is separate and apart from any observation. That possibility\n> is my source for my intuitive feelings that something FTL may\n> be happening.\n\nWhy so little interest in the *evidence*? Why envisage all these weird and\nwonderful possibilities without first checking what the experiments do, and\nwhether or not the tests used to establish entanglement are valid?\n\nI\'ve spent the past few weeks studying the evolution of the Bell tests that\nare used in real experiments. Some of my most interesting findings concern\nClauser and Horne\'s 1974 single-channel inequality. I\'ve just put a page in\nwikipedia on the subject:\n\nhttp://en.wikipedia.org/wiki/Clauser_and_Horne%27s_1974_Bell_test\n\nYou might do worse than look at it.\n\nCaroline\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Paul Stewart Snyder" <ps@ws5.com> wrote in message
news:d16eb5af.0407272017.2eb35f58@posting.google.c om...
> tom129@juno.com (Tom Trotter) wrote
> > Note that this is not a model of what's happening in Bell-test
> > photon experiments.
>
> You are right, the example is not related to quantum entanglement.
>
> > In these experiments the entanglement is of
> > a different sort. A and B don't change together. They don't
> > affect each other. They're not physically connected in any way.
> > They do, however, have a motional property in common (polarization)
> > due to their common source. In observational contexts that refer to
> > this common property, the photons will produce predictable
> > (correlated) detection patterns, functionally related to the
> > combined settings of the analyzers.
>
> You are describing one way to understand entanglement ...
Yes, but how about understanding the actual experiments? If you look at
any real experiment it will surely give you a totally different perspective
on things. You've got real light, interacting one "photon" at a time with a
detector, real electrical signals going from the detector to the coincidence
monitor, real decisions being made on the basis of individual photon pairs.
Unless you can account for how the final statistical results are built up
from these individual events you do not have a sound physical theory for the
experiment.
> if I understand correctly you require observation of both A and B
> to produce entanglement, so that the correlated patterns are related
> to the "combined settings of the analyzers", being the "observers"
> in a Copenhagen Interpretation sense? That may be the case -
> however, unless I have missed something, I do not think we have a
> definitive understanding of the role of observation,
> if any, in quantum measurement?
[Who am I to argue with that?]
> As far as I can tell even Cramer's
> unusual Transactional Interpretation of Quantum Mechanics has
> not been ruled out as a possibility. So I don't believe we can
> say that the polarization property may be limited to "observational
> contexts".
[whatever that may mean]
> If state vector reduction turns out to be unrelated to observation
> (and/or if many-universe models are correct), then there is
> an instantaneous physical change in B when A changes
> that is separate and apart from any observation. That possibility
> is my source for my intuitive feelings that something FTL may
> be happening.
Why so little interest in the *evidence*? Why envisage all these weird and
wonderful possibilities without first checking what the experiments do, and
whether or not the tests used to establish entanglement are valid?
I've spent the past few weeks studying the evolution of the Bell tests that
are used in real experiments. Some of my most interesting findings concern
Clauser and Horne's 1974 single-channel inequality. I've just put a page in
wikipedia on the subject:
http://en.wikipedia.org/wiki/Clauser_and_Horne%27s_1974_Bell_test
You might do worse than look at it.
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Caroline Thompson
Jul29-04, 04:59 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\n"Tom Trotter" <tom129@juno.com> wrote in message\nnews:29df3039.0407271103.333d99a3@posting .google.com...\n>\n> "Caroline Thompson" <ch.thompson1@virgin.net> wrote >\n\n> > > [TT:] Wrt this scenario, in the combined context, the only\n> > > variable that matters is the angular difference in polarizer settings.\n> >\n> > [CT:] Not necessarily so. There may, for instance, be more photon\n> > pairs polarised vertically than at any other angle.\n>\n> [snip]\n>\n> It doesn\'t matter. In the combined context, it\'s not the\n> polarization, but the relationship between photon 1 and\n> photon 2 that\'s being observed.\n\nI\'m afraid I can prove you\'re wrong, Tom. Have you read my paper on\nrotational invariance, http://arxiv.org/abs/quant-ph/9912082 ?\n\n> > > > > ... As long as photon 1 and photon 2 of any given pair are\n> > > > > identically polarized, then the orientation of the polarizers\n> > > > > wrt each other is the only relevant variable in the\n> > > > > observational context, and a circular function of this\n> > > > > changing angular difference is what determines rates of\n> > > > > coincidental detection.\n> > > >\n> > > > But this is true only if we have rotational invariance.\n> > >\n> > > I don\'t think so. Remember, Bell\'s unknown parameter\n> > > affecting individual results can have any value.\n\nIt\'s all very well you "thinking so", but I have actually worked through the\ndetails, assuming that we are dealing with light that comes in pulses, each\nwith a definite polarisation direction (the same on both sides of the\nexperiment, only varying between pairs).\n\n> > > Bell\'s analysis says that the coincidence rate, cr, for 2Theta,\n> > > can\'t exceed cr/2 for Theta.\n> >\n> > No, Bell never said that! I think maybe you are following\n> > Nick Herbert? I haven\'t read his book but keep on coming\n> > upon misconceptions on behalf of those who have.\n>\n> It\'s a corollary of Bell\'s theorem.\n\nWhat inequality are you assuming? There a quite a few different ones all\ngoing by that name, all derived with the same intention but in fact often\ndiffering in their assumptions. Many are true only if there are no\nnon-detections.\n\n> > > But, experimentally, this doesn\'t hold for all tested values\n> > > of Theta. The qm formulation, on the\n> > > other hand, says that cr will vary as a circular, rather than\n> > > a linear, function of Theta -- and experiments bear this out.\n> >\n> > Yes, and the local realist model appropriate to optical Bell tests says\nthe\n> > same. (See quant-ph/9903066)\n>\n> That\'s because the qm formulation doesn\'t violate local\n> reality. It\'s just that some interpretations of it say\n> that it does.\n\nBut the QM formula violates Bell\'s inequality! If the real world actually\ndid this it would mean the end of rational physics. It would mean that you\ncan\'t depend on being able to multiply independent probabilities to get the\ncombined one. It\'s not just a matter of interpretation.\n\n> > >From physical considerations this does not make sense.\n> > >Suppose *all* photons are V. Then if our fixed polariser\n> > >is set vertically we get a beautiful full-visibility coincidence\n> > >curve as we vary the other. If, on the other hand, we set\n> > > it horizontally, we don\'t get any coincidences at\n> > all, whatever the angle of the other polariser.\n>\n> Not so. The coincidence curve will still be a circular function\n> of the angular difference between the polarizer settings (Theta),\n> as long as photon 1 and photon 2 of any given pair are polarized\n> identically. A mutual non-detection is also a coincidence.\n\nBut hang on a minute, Tom! Do read that paper of mine. The situation I\'ve\njust described is one in which on one side you get *no* detections,\ntherefore you get no positive coincidences. But neither do you get "mutual\nnon-detections", since, as you vary the angle, you do get some detections on\nthe other side.\n\nIn point of fact, though, it is not safe to count mutual non-detections as\ncoincidences. That is the path to confusion.\n\n> No matter what the polarization of any given pair is, the\n> rate of coincidental detection attributes will increase as\n> Theta decreases, and will decrease as Theta increases.\n\nNot if it is fixed at zero! But you get more interesting possibilities\narising if your source is in fact producing a 50-50 mix of photons, half in\none state, half the opposite. The case likely to arise in real life is when\nthe hidden variable is the phase difference between vertically and\nhorizontally polarised components.\n\n> > > Maybe that\'s what he was worried about, but that\'s not what\n> > > he proved. He proved that if you include parameters that\n> > > would affect individual measurements in a formulation concerning\n> > > combined measurements, then you get results that are inconsistent\n> > > with qm or (after tweaking so that it\'s consistent with the\nstatistical\n> > > predictions of qm) results that are inconsistent with the assumptions\n> > > of standard physics regarding a limitation on signal velocity\n> > > (thereby violating Einstein causality or Lorentz invariance).\n> >\n> > No, his inequality is one that is obeyed under standard physics.\n>\n> Bell adjusted his formulation so that it would be\n> consistent with the predictions of qm.\n\nPerhaps you are looking only at Bell\'s 1964 paper? The inequalities used in\nreal experiments do not require his assumption that when detectors are\nparallel there will be perfect correlation.\n\n> > ... The whole point of Bell\'s theorem is that the predictions\n> > of qm are incompatible with the existence of hidden variables.\n>\n> That\'s the normal interpretation, and, as stated, it\'s just\n> wrong.\n>\n> The hidden variables that would allow for more accurate\n> predictions of individual results don\'t vanish from existence\n> just because you\'re looking at things wrt a different\n> context. It\'s just that they\'re not relevant in the combined\n> context.\n>\n> Predictions of qm wrt combined contexts such as Bell tests\n> are inconsistent with those of formulations that include a term,\n> eg., lambda, that is irrelevant wrt determining coincidental\n> detection. The qm formulation, taken by itself without reference to\n> misinterpretations of Bell\'s work, isn\'t incompatible with\n> Einstein locality or the existence of hidden variables.\n> (See the emission models.)\n\nWhere do I find these? Do I take it that you can give me a definite example\nof a hidden variable model that agrees with QM?\n\n> > > ... The qm formulation, and\n> > > the experiments, also obey Einstein causality.\n\nI agree that the experiments do (by virtue of the looholes, and because I\ncan give an actual local realist explanation), but can you prove that qm\ndoes? As far as I can see your argument amounts to hand-waving about this\n"relationship" that is something more than a mere zero difference in\npolarisation directions.\n\n> As long as you can preserve the emission-produced relationship\n> (the entanglement of photons emitted by the same atom), then\n> it doesn\'t matter how far apart they are when you analyze\n> them. Observation context doesn\'t necessarily have anything\n> to do with distance.\n\nPolarisation direction has no physical reason to change with distance ...\n\n> The "observational context" is (1) opposite moving photons correlated\n> in polarization via emission, and (2) analysis of this common\n> polarization (this *relationship") for paired photons via corresponding,\n> joint settings of separated polarizers.\n\nBut can you actually show me the details? What are you saying is the\nphysical cause of the high correlations that violate various Bell\'s\ninequalities? I can show you the details of local hidden variable models\nthat, in perfect conditions, can\'t violate a Bell inequality. Given\nappropriate imperfections, though, they can.\n\nCaroline\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Tom Trotter" <tom129@juno.com> wrote in message
news:29df3039.0407271103.333d99a3@posting.google.c om...
>
> "Caroline Thompson" <ch.thompson1@virgin.net> wrote >
> > > [TT:] Wrt this scenario, in the combined context, the only
> > > variable that matters is the angular difference in polarizer settings.
> >
> > [CT:] Not necessarily so. There may, for instance, be more photon
> > pairs polarised vertically than at any other angle.
>
> [snip]
>
> It doesn't matter. In the combined context, it's not the
> polarization, but the relationship between photon 1 and
> photon 2 that's being observed.
I'm afraid I can prove you're wrong, Tom. Have you read my paper on
rotational invariance, http://arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/9912082 ?
> > > > > ... As long as photon 1 and photon 2 of any given pair are
> > > > > identically polarized, then the orientation of the polarizers
> > > > > wrt each other is the only relevant variable in the
> > > > > observational context, and a circular function of this
> > > > > changing angular difference is what determines rates of
> > > > > coincidental detection.
> > > >
> > > > But this is true only if we have rotational invariance.
> > >
> > > I don't think so. Remember, Bell's unknown parameter
> > > affecting individual results can have any value.
It's all very well you "thinking so", but I have actually worked through the
details, assuming that we are dealing with light that comes in pulses, each
with a definite polarisation direction (the same on both sides of the
experiment, only varying between pairs).
> > > Bell's analysis says that the coincidence rate, cr, for 2Theta,
> > > can't exceed cr/2 for \Theta.
> >
> > No, Bell never said that! I think maybe you are following
> > Nick Herbert? I haven't read his book but keep on coming
> > upon misconceptions on behalf of those who have.
>
> It's a corollary of Bell's theorem.
What inequality are you assuming? There a quite a few different ones all
going by that name, all derived with the same intention but in fact often
differing in their assumptions. Many are true only if there are no
non-detections.
> > > But, experimentally, this doesn't hold for all tested values
> > > of \Theta. The qm formulation, on the
> > > other hand, says that cr will vary as a circular, rather than
> > > a linear, function of \Theta -- and experiments bear this out.
> >
> > Yes, and the local realist model appropriate to optical Bell tests says
the
> > same. (See http://www.arxiv.org/abs/quant-ph/9903066)
>
> That's because the qm formulation doesn't violate local
> reality. It's just that some interpretations of it say
> that it does.
But the QM formula violates Bell's inequality! If the real world actually
did this it would mean the end of rational physics. It would mean that you
can't depend on being able to multiply independent probabilities to get the
combined one. It's not just a matter of interpretation.
> > >From physical considerations this does not make sense.
> > >Suppose *all* photons are V. Then if our fixed polariser
> > >is set vertically we get a beautiful full-visibility coincidence
> > >curve as we vary the other. If, on the other hand, we set
> > > it horizontally, we don't get any coincidences at
> > all, whatever the angle of the other polariser.
>
> Not so. The coincidence curve will still be a circular function
> of the angular difference between the polarizer settings (\Theta),
> as long as photon 1 and photon 2 of any given pair are polarized
> identically. A mutual non-detection is also a coincidence.
But hang on a minute, Tom! Do read that paper of mine. The situation I've
just described is one in which on one side you get *no* detections,
therefore you get no positive coincidences. But neither do you get "mutual
non-detections", since, as you vary the angle, you do get some detections on
the other side.
In point of fact, though, it is not safe to count mutual non-detections as
coincidences. That is the path to confusion.
> No matter what the polarization of any given pair is, the
> rate of coincidental detection attributes will increase as
> \Theta decreases, and will decrease as \Theta increases.
Not if it is fixed at zero! But you get more interesting possibilities
arising if your source is in fact producing a 50-50 mix of photons, half in
one state, half the opposite. The case likely to arise in real life is when
the hidden variable is the phase difference between vertically and
horizontally polarised components.
> > > Maybe that's what he was worried about, but that's not what
> > > he proved. He proved that if you include parameters that
> > > would affect individual measurements in a formulation concerning
> > > combined measurements, then you get results that are inconsistent
> > > with qm or (after tweaking so that it's consistent with the
statistical
> > > predictions of qm) results that are inconsistent with the assumptions
> > > of standard physics regarding a limitation on signal velocity
> > > (thereby violating Einstein causality or Lorentz invariance).
> >
> > No, his inequality is one that is obeyed under standard physics.
>
> Bell adjusted his formulation so that it would be
> consistent with the predictions of qm.
Perhaps you are looking only at Bell's 1964 paper? The inequalities used in
real experiments do not require his assumption that when detectors are
parallel there will be perfect correlation.
> > ... The whole point of Bell's theorem is that the predictions
> > of qm are incompatible with the existence of hidden variables.
>
> That's the normal interpretation, and, as stated, it's just
> wrong.
>
> The hidden variables that would allow for more accurate
> predictions of individual results don't vanish from existence
> just because you're looking at things wrt a different
> context. It's just that they're not relevant in the combined
> context.
>
> Predictions of qm wrt combined contexts such as Bell tests
> are inconsistent with those of formulations that include a term,
> eg., \lambda, that is irrelevant wrt determining coincidental
> detection. The qm formulation, taken by itself without reference to
> misinterpretations of Bell's work, isn't incompatible with
> Einstein locality or the existence of hidden variables.
> (See the emission models.)
Where do I find these? Do I take it that you can give me a definite example
of a hidden variable model that agrees with QM?
> > > ... The qm formulation, and
> > > the experiments, also obey Einstein causality.
I agree that the experiments do (by virtue of the looholes, and because I
can give an actual local realist explanation), but can you prove that qm
does? As far as I can see your argument amounts to hand-waving about this
"relationship" that is something more than a mere zero difference in
polarisation directions.
> As long as you can preserve the emission-produced relationship
> (the entanglement of photons emitted by the same atom), then
> it doesn't matter how far apart they are when you analyze
> them. Observation context doesn't necessarily have anything
> to do with distance.
Polarisation direction has no physical reason to change with distance ...
> The "observational context" is (1) opposite moving photons correlated
> in polarization via emission, and (2) analysis of this common
> polarization (this *relationship") for paired photons via corresponding,
> joint settings of separated polarizers.
But can you actually show me the details? What are you saying is the
physical cause of the high correlations that violate various Bell's
inequalities? I can show you the details of local hidden variable models
that, in perfect conditions, can't violate a Bell inequality. Given
appropriate imperfections, though, they can.
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Paul Stewart Snyder
Jul30-04, 03:21 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"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<inXNc.543\\$Lz.461@newsfe4-gui.ntli.net>...\n> Yes, but how about understanding the actual experiments? If you look at\n> any real experiment it will surely give you a totally different perspective\n> on things. You\'ve got real light, interacting one "photon" at a time with a\n> detector, real electrical signals going from the detector to the coincidence\n> monitor, real decisions being made on the basis of individual photon pairs.\n> Unless you can account for how the final statistical results are built up\n> from these individual events you do not have a sound physical theory for the\n> experiment.\n\nNo disrespect meant, but I don\'t believe that what you are saying has\nanything to do with quantum entanglement, perhaps you are leaving\nsomething out? If a single photon hits a detector, and is recorded by\na coincidence monitor, what does that have to do with a pair of\nphotons, one always having the opposite polarization of the other? If\nyou are telling me that you can select the polarization of one of a\npair of quantum entangled photons AND the other single photon that is\nmeasured does NOT assume the opposite polarization every time (within\nexperimental limits) – then I would be interested in your objections.\n\nPS\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>"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<inXNc.543$Lz.461@newsfe4-gui.ntli.net>...
> Yes, but how about understanding the actual experiments? If you look at
> any real experiment it will surely give you a totally different perspective
> on things. You've got real light, interacting one "photon" at a time with a
> detector, real electrical signals going from the detector to the coincidence
> monitor, real decisions being made on the basis of individual photon pairs.
> Unless you can account for how the final statistical results are built up
> from these individual events you do not have a sound physical theory for the
> experiment.
No disrespect meant, but I don't believe that what you are saying has
anything to do with quantum entanglement, perhaps you are leaving
something out? If a single photon hits a detector, and is recorded by
a coincidence monitor, what does that have to do with a pair of
photons, one always having the opposite polarization of the other? If
you are telling me that you can select the polarization of one of a
pair of quantum entangled photons AND the other single photon that is
measured does NOT assume the opposite polarization every time (within
experimental limits) – then I would be interested in your objections.
PS
Tom Trotter
Jul30-04, 03:21 AM
<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<U9XNc.535\\$Lz.139@newsfe4-gui.ntli.net>...\n> > [snip]\n> >\n> > ... In the combined context, it\'s not the\n> > polarization, but the relationship between photon 1 and\n> > photon 2 that\'s being observed.\n>\n> I\'m afraid I can prove you\'re wrong, Tom.\n> Have you read my paper on rotational invariance,\n> http://arxiv.org/abs/quant-ph/9912082 ?\n\nI downloaded it, and will read it along with a\nfew others when I get time. I suppose that I disagree,\nat least in part, with what I take to be it\'s main\npremise -- that the polarization of the photons\nincident on the polarizers will affect the\nrate of coincidental detection.\n\nThis is true only if the detection attributes\nbeing paired via the coincidence circuitry are\neither not emitted from the same atomic process\n(that is, the experimental setup is not\nanalyzing what it\'s supposed to be analyzing),\nor the emission model is wrong (and opposite-moving\nphotons emitted by the same atom are simply not\ncorrelated in polarization via the emission\nprocess).\n\n[snip]\n\n> > > ... (See quant-ph/9903066)\n\nIt\'s in the queue. :-)\n\n[snip]\n\n> >\n> > ... the qm formulation doesn\'t violate local\n> > reality. It\'s just that some interpretations\n> > of it say that it does.\n>\n> But the QM formula violates Bell\'s inequality!\n> If the real world actually did this it would\n> mean the end of rational physics. It would mean\n> that you can\'t depend on being able to multiply\n> independent probabilities to get the combined one.\n> It\'s not just a matter of interpretation.\n\nThe probabilities pertaining to coincidental detection\ndon\'t depend on the same thing as the probabilities\npertaining to the individual data streams do.\n\nAssuming the emission model is correct and an\nideal setup, there is no *variable* in the\n*combined context* other than the settings of the\npolarizers.\n\nThis doesn\'t mean that hidden variables (eg.,\npolarization of photons incident on a and b) don\'t\nexist. That would be a silly conclusion. We can\nstill *see* the light from the emitter to the polarizers.\n\n> > > >From physical considerations this does not make sense.\n> > > >Suppose *all* photons are V. Then if our fixed polariser\n> > > >is set vertically we get a beautiful full-visibility coincidence\n> > > >curve as we vary the other. If, on the other hand, we set\n> > > > it horizontally, we don\'t get any coincidences at\n> > > all, whatever the angle of the other polariser.\n> >\n> > Not so. The coincidence curve will still be a circular function\n> > of the angular difference between the polarizer settings (Theta),\n> > as long as photon 1 and photon 2 of any given pair are polarized\n> > identically. A mutual non-detection is also a coincidence.\n>\n> But hang on a minute, Tom! Do read that paper of mine.\n> The situation I\'ve just described is one in which on one\n> side you get *no* detections, therefore you get no positive\n> coincidences.\n\nI thought we were dealing with all coincidental\n(identical) detection attributes, not just positive\nones. In your above example with the fixed polarizer\nset to the angle at which all photons are emitted, as\nyou rotate the other polarizer away from it, then\nthe cr will decrease, and as you rotate it toward\nthe fixed polarizer, then cr will increase. Are you\ndisputing this?\n\n> But neither do you get "mutual non-detections",\n> since, as you vary the angle, you do get some\n> detections on the other side.\n\nIn the case we\'re considering, you get no *mutual*\nnon-detections because the detector behind the *fixed*\npolarizer will register during every coincidence\nwindow. (Remember, we\'re considering this in the\nideal, which is the way you sort out the *logic* of\nthis stuff.)\n\nAs you rotate the other polarizer away from the fixed\none, then the intensity of the light that it transmits\nwill decrease as a cos function of the angular\ndifference between it and the fixed polarizer\n(as you rotate it away from alignment with\nthe fixed polarizer, then the rate at which it\'s\ndetector will register must decrease, as we\'re assuming\nthat all of the emitted light is polarized in alignment\nwith the fixed polarizer).\n\n> In point of fact, though, it is not safe to count\n> mutual non-detections as coincidences. That is the\n> path to confusion.\n\nHow so?\n\n>\n> > No matter what the polarization of any given pair is, the\n> > rate of coincidental detection attributes will increase as\n> > Theta decreases, and will decrease as Theta increases.\n>\n> Not if it is fixed at zero!\n\nNot if *what* is fixed at zero?\n\n> > > ... [Bell\'s] inequality is one that is obeyed under\n> > > standard physics.\n\nI consider the Aspect (et.al.) experiments to be\nstandard physics, and their results violate the\ninequality.\n\n> > The qm formulation, taken by itself, without reference to\n> > misinterpretations of Bell\'s work, isn\'t incompatible with\n> > Einstein locality or the existence of hidden variables.\n> > (See the emission models.)\n>\n> Where do I find these? Do I take it that you can give\n> give a definite example of a hidden variable model that\n> agrees with QM?\n\nGreenstein and Zajonc (The Quantum Challenge, 1997) has\na nice discussion of how entangled photons are emitted\nvia atomic calcium cascades.\n\nThe term "local hidden variable" simply\ndoesn\'t apply to what\'s being observed in, for\nexample, the Aspect experiments. They\'re concerned\nwith the correlation of spacelike separated detection\nevents. These events, and the connection between\nthe analyzed objects (the photons emitted by the\nsame atom) are all due to localized interactions,\nbut the connection or relationship itself doesn\'t\ndiminish with distance (in the ideal) and is the\ndeterminer, along with the corresponding mutual\nanalyzer settings, of *coincidental* detection.\n\nBut since this relationship (identical emission-\npolarization of photon 1 and photon 2 of any given\npair) doesn\'t vary from pair to pair, it isn\'t\nincluded in calculations of variable coincidence\nrates. So, what\'s left to include in the\nquantitative formula for predicting how the\ncoincidence rates will vary as they do?\nThere\'s only one thing left, Theta, the angular\ndifference of the polarizer settings.\n\n>\n> > > > ... The qm formulation, and\n> > > > the experiments, also obey Einstein causality.\n>\n> I agree that the experiments do (by virtue of the\n> looholes, and because I can give an actual local\n> realist explanation), but can you prove that qm\n> does? As far as I can see your argument amounts\n> to hand-waving about this "relationship" that is\n> something more than a mere zero difference in\n> polarisation directions.\n\nIt\'s not something more than a *mere* zero difference\nin polarizations. That\'s what it *is*. Try producing\nthat with polarizers. I\'ll bet that you can\'t.\n\nLet me summarize my position, and then take some time\nto get caught up on my reading.\n\n(1) Local hidden variable formulations are\napplicable in contexts where the *variability*\nof the unknown parameter is a determining factor.\nTheir inapplicability to certain contexts\ntells nothing about the existence or not\nof local hidden variables. Of course, they exist.\n\n(2) Bell\'s formulation is inapplicable to the\ncontext for which it\'s intended. Wrt this\nformulation, the assumption that is contradicted\nby the qm formulation and experimental results\nis not the locality of the physical interactions\ninvolved, or the reality of photon polarization\nprior to detection, but rather that the hidden\nparameter (variable polarization) that\'s applicable\nto individual measurement contexts is also applicable\nin the combined context. (But, as long as the\nphotons of any given pair are polarized identically\nvia emission, then the hidden parameter that\'s\napplicable in the combined context does not\nvary. Therefore it\'s inclusion in the\nformulation is obviated, and one is left with\nonly the variable mutual polarizer settings as the\neffective determiner of coincidental detection.)\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>"Caroline Thompson" <ch.thompson1@virgin.net> wrote in message news:<U9XNc.535$Lz.139@newsfe4-gui.ntli.net>...
> > [snip]
> >
> > ... In the combined context, it's not the
> > polarization, but the relationship between photon 1 and
> > photon 2 that's being observed.
>
> I'm afraid I can prove you're wrong, Tom.
> Have you read my paper on rotational invariance,
> http://arxiv.org/abs/http://www.arxiv.org/abs/quant-ph/9912082 ?
I downloaded it, and will read it along with a
few others when I get time. I suppose that I disagree,
at least in part, with what I take to be it's main
premise -- that the polarization of the photons
incident on the polarizers will affect the
rate of coincidental detection.
This is true only if the detection attributes
being paired via the coincidence circuitry are
either not emitted from the same atomic process
(that is, the experimental setup is not
analyzing what it's supposed to be analyzing),
or the emission model is wrong (and opposite-moving
photons emitted by the same atom are simply not
correlated in polarization via the emission
process).
[snip]
> > > ... (See http://www.arxiv.org/abs/quant-ph/9903066)
It's in the queue. :-)
[snip]
> >
> > ... the qm formulation doesn't violate local
> > reality. It's just that some interpretations
> > of it say that it does.
>
> But the QM formula violates Bell's inequality!
> If the real world actually did this it would
> mean the end of rational physics. It would mean
> that you can't depend on being able to multiply
> independent probabilities to get the combined one.
> It's not just a matter of interpretation.
The probabilities pertaining to coincidental detection
don't depend on the same thing as the probabilities
pertaining to the individual data streams do.
Assuming the emission model is correct and an
ideal setup, there is no *variable* in the
*combined context* other than the settings of the
polarizers.
This doesn't mean that hidden variables (eg.,
polarization of photons incident on a and b) don't
exist. That would be a silly conclusion. We can
still *see* the light from the emitter to the polarizers.
> > > >From physical considerations this does not make sense.
> > > >Suppose *all* photons are V. Then if our fixed polariser
> > > >is set vertically we get a beautiful full-visibility coincidence
> > > >curve as we vary the other. If, on the other hand, we set
> > > > it horizontally, we don't get any coincidences at
> > > all, whatever the angle of the other polariser.
> >
> > Not so. The coincidence curve will still be a circular function
> > of the angular difference between the polarizer settings (\Theta),
> > as long as photon 1 and photon 2 of any given pair are polarized
> > identically. A mutual non-detection is also a coincidence.
>
> But hang on a minute, Tom! Do read that paper of mine.
> The situation I've just described is one in which on one
> side you get *no* detections, therefore you get no positive
> coincidences.
I thought we were dealing with all coincidental
(identical) detection attributes, not just positive
ones. In your above example with the fixed polarizer
set to the angle at which all photons are emitted, as
you rotate the other polarizer away from it, then
the cr will decrease, and as you rotate it toward
the fixed polarizer, then cr will increase. Are you
disputing this?
> But neither do you get "mutual non-detections",
> since, as you vary the angle, you do get some
> detections on the other side.
In the case we're considering, you get no *mutual*
non-detections because the detector behind the *fixed*
polarizer will register during every coincidence
window. (Remember, we're considering this in the
ideal, which is the way you sort out the *logic* of
this stuff.)
As you rotate the other polarizer away from the fixed
one, then the intensity of the light that it transmits
will decrease as a cos function of the angular
difference between it and the fixed polarizer
(as you rotate it away from alignment with
the fixed polarizer, then the rate at which it's
detector will register must decrease, as we're assuming
that all of the emitted light is polarized in alignment
with the fixed polarizer).
> In point of fact, though, it is not safe to count
> mutual non-detections as coincidences. That is the
> path to confusion.
How so?
>
> > No matter what the polarization of any given pair is, the
> > rate of coincidental detection attributes will increase as
> > \Theta decreases, and will decrease as \Theta increases.
>
> Not if it is fixed at zero!
Not if *what* is fixed at zero?
> > > ... [Bell's] inequality is one that is obeyed under
> > > standard physics.
I consider the Aspect (et.al.) experiments to be
standard physics, and their results violate the
inequality.
> > The qm formulation, taken by itself, without reference to
> > misinterpretations of Bell's work, isn't incompatible with
> > Einstein locality or the existence of hidden variables.
> > (See the emission models.)
>
> Where do I find these? Do I take it that you can give
> give a definite example of a hidden variable model that
> agrees with QM?
Greenstein and Zajonc (The Quantum Challenge, 1997) has
a nice discussion of how entangled photons are emitted
via atomic calcium cascades.
The term "local hidden variable" simply
doesn't apply to what's being observed in, for
example, the Aspect experiments. They're concerned
with the correlation of spacelike separated detection
events. These events, and the connection between
the analyzed objects (the photons emitted by the
same atom) are all due to localized interactions,
but the connection or relationship itself doesn't
diminish with distance (in the ideal) and is the
determiner, along with the corresponding mutual
analyzer settings, of *coincidental* detection.
But since this relationship (identical emission-
polarization of photon 1 and photon 2 of any given
pair) doesn't vary from pair to pair, it isn't
included in calculations of variable coincidence
rates. So, what's left to include in the
quantitative formula for predicting how the
coincidence rates will vary as they do?
There's only one thing left, \Theta, the angular
difference of the polarizer settings.
>
> > > > ... The qm formulation, and
> > > > the experiments, also obey Einstein causality.
>
> I agree that the experiments do (by virtue of the
> looholes, and because I can give an actual local
> realist explanation), but can you prove that qm
> does? As far as I can see your argument amounts
> to hand-waving about this "relationship" that is
> something more than a mere zero difference in
> polarisation directions.
It's not something more than a *mere* zero difference
in polarizations. That's what it *is*. Try producing
that with polarizers. I'll bet that you can't.
Let me summarize my position, and then take some time
to get caught up on my reading.
(1) Local hidden variable formulations are
applicable in contexts where the *variability*
of the unknown parameter is a determining factor.
Their inapplicability to certain contexts
tells nothing about the existence or not
of local hidden variables. Of course, they exist.
(2) Bell's formulation is inapplicable to the
context for which it's intended. Wrt this
formulation, the assumption that is contradicted
by the qm formulation and experimental results
is not the locality of the physical interactions
involved, or the reality of photon polarization
prior to detection, but rather that the hidden
parameter (variable polarization) that's applicable
to individual measurement contexts is also applicable
in the combined context. (But, as long as the
photons of any given pair are polarized identically
via emission, then the hidden parameter that's
applicable in the combined context does not
vary. Therefore it's inclusion in the
formulation is obviated, and one is left with
only the variable mutual polarizer settings as the
effective determiner of coincidental detection.)
Tom Trotter
Jul31-04, 09:16 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>\nps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407272017.2eb35f58@posting.google. com>...\n> tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407251124.1cd7a038@posting.google. com>...\n> > ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google. com>...\n> > > tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google. com>...\n> > > > ps@ws5.com (Paul Stewart Snyder) wrote in message\n> > > > > We know that when A is changed B is instantaneously\n> > > > > changed,\n> > > >\n> > > > Only in certain contexts, but they don\'t imply signal\n> > > > transfer between A and B.\n> > > >\n> > > This is what continues to bother me - there are at least some contexts\n> > > in which there are instantaneous changes to B when A changes - even if\n> > > A and B are spatially separated. There may not be a signal transfer in\n> > > the formal terms of information transfer, and EPR may not require an\n> > > FTL event, yet there still seems to be a strong intuitive basis\n> > > (Einstein loved intuition) for thinking that something FTL is going on\n> > > here.\n> >\n> > Draw a circle and put two points at different locations on it.\n> > Rotate the circle. A changes with B,and vice versa, but A\n> > doesn\'t affect B, and vice versa. There\'s a non-varying\n> > physical connection between A and B, but no \'communication\'.\n> > Their motion will be correlated in contexts that refer to their\n> > physical connection.\n> >\n> > Note that this is not a model of what\'s happening in Bell-test\n> > photon experiments.\n>\n> You are right, the example is not related to quantum entanglement.\n>\n\nI think the example is related to a type of entanglement,\nbut not the type of entanglement that optical Bell tests have dealt\nwith.\n\n> > In these experiments the entanglement is of\n> > a different sort. A and B don\'t change together. They don\'t\n> > affect each other. They\'re not physically connected in any way.\n> > They do, however, have a motional property in common (polarization)\n> > due to their common source. In observational contexts that refer to\n> > this common property, the photons will produce predictable (correlated)\n> > detection patterns, functionally related to the combined settings\n> > of the analyzers.\n>\n> You are describing one way to understand entanglement, if I understand\n> correctly you require observation of both A and B to produce entanglement,\n\nObservation of both A and B wrt corresponding joint analyzer settings\nis required to *observe* entanglement, not produce it, per se. The\nquantum entanglement itself (between photons emitted by the same atom)\nis assumed to be produced via the emission process, which is never\ndirectly\nobserved.\n\n> so that the correlated patterns are related to the "combined settings of the\n> analyzers", being the "observers" in a Copenhagen Interpretation sense?\n\nThe correlated detection patterns that violate Bell inequalities\nemerge in the observational context of combined analyzer settings.\nI don\'t know what the Copenhagen Interpretation is.\n\n> That may be the case ? however, unless I have missed something, I do\n> not think we have a definitive understanding of the role of observation,\n> if any, in quantum measurement?\n\nDifferent observational contexts require different formulations.\nThat\'s all. This is true whether you\'re dealing with quantum\nor classical phenomena.\n\n[snip]\n\n> So I don\'t believe we can say that the polarization property\n> may be limited to "observational contexts".\n\nThe relevant observational (and therefore the relevant\nformal) parameters are certainly a function of the observational\ncontext.\n\n> If state vector reduction turns out to be unrelated to observation\n> (and/or if many-universe models are correct), then there is an\n> instantaneous physical change in B when A changes that is separate\n> and apart from any observation. That possibility is my source\n> for my intuitive feelings that something FTL may be happening.\n\nThis approach won\'t help wrt to understanding what\'s happening\nin the EPRBell experiments, and it won\'t help wrt to correctly\ninterpreting their physical meaning, imho.\n\nThis \'something FTL\' remains an entirely metaphysical construction\nuntil somebody produces something FTL. The default position\nis the assumption that the speed of light circumscribes any and\nall physical processes or effects. And, so far, this assumption\nhasn\'t been contradicted by experiment.\n\nThere are good reasons (via actual experimental designs) to\nbelieve that, in optical EPRBell experiments, there\'s no\ninstantaneous physical change in B when A changes, and\nvice versa. And, there are no good reasons to believe that\nA and B are instantaneously \'communicating\'.\n\nSo, if you want to hang onto those intuitive feelings about\nthe possibility that \'something FTL may be happening\', then\nyou\'ll be doing it in the face of some pretty overwhelming\nevidence to the contrary, imho.\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>ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407272017.2eb35f58@posting.google.com>...
> tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407251124.1cd7a038@posting.google.com>...
> > ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google.com>...
> > > tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google.com>...
> > > > ps@ws5.com (Paul Stewart Snyder) wrote in message
> > > > > We know that when A is changed B is instantaneously
> > > > > changed,
> > > >
> > > > Only in certain contexts, but they don't imply signal
> > > > transfer between A and B.
> > > >
> > > This is what continues to bother me - there are at least some contexts
> > > in which there are instantaneous changes to B when A changes - even if
> > > A and B are spatially separated. There may not be a signal transfer in
> > > the formal terms of information transfer, and EPR may not require an
> > > FTL event, yet there still seems to be a strong intuitive basis
> > > (Einstein loved intuition) for thinking that something FTL is going on
> > > here.
> >
> > Draw a circle and put two points at different locations on it.
> > Rotate the circle. A changes with B,and vice versa, but A
> > doesn't affect B, and vice versa. There's a non-varying
> > physical connection between A and B, but no 'communication'.
> > Their motion will be correlated in contexts that refer to their
> > physical connection.
> >
> > Note that this is not a model of what's happening in Bell-test
> > photon experiments.
>
> You are right, the example is not related to quantum entanglement.
>
I think the example is related to a type of entanglement,
but not the type of entanglement that optical Bell tests have dealt
with.
> > In these experiments the entanglement is of
> > a different sort. A and B don't change together. They don't
> > affect each other. They're not physically connected in any way.
> > They do, however, have a motional property in common (polarization)
> > due to their common source. In observational contexts that refer to
> > this common property, the photons will produce predictable (correlated)
> > detection patterns, functionally related to the combined settings
> > of the analyzers.
>
> You are describing one way to understand entanglement, if I understand
> correctly you require observation of both A and B to produce entanglement,
Observation of both A and B wrt corresponding joint analyzer settings
is required to *observe* entanglement, not produce it, per se. The
quantum entanglement itself (between photons emitted by the same atom)
is assumed to be produced via the emission process, which is never
directly
observed.
> so that the correlated patterns are related to the "combined settings of the
> analyzers", being the "observers" in a Copenhagen Interpretation sense?
The correlated detection patterns that violate Bell inequalities
emerge in the observational context of combined analyzer settings.
I don't know what the Copenhagen Interpretation is.
> That may be the case ? however, unless I have missed something, I do
> not think we have a definitive understanding of the role of observation,
> if any, in quantum measurement?
Different observational contexts require different formulations.
That's all. This is true whether you're dealing with quantum
or classical phenomena.
[snip]
> So I don't believe we can say that the polarization property
> may be limited to "observational contexts".
The relevant observational (and therefore the relevant
formal) parameters are certainly a function of the observational
context.
> If state vector reduction turns out to be unrelated to observation
> (and/or if many-universe models are correct), then there is an
> instantaneous physical change in B when A changes that is separate
> and apart from any observation. That possibility is my source
> for my intuitive feelings that something FTL may be happening.
This approach won't help wrt to understanding what's happening
in the EPRBell experiments, and it won't help wrt to correctly
interpreting their physical meaning, imho.
This 'something FTL' remains an entirely metaphysical construction
until somebody produces something FTL. The default position
is the assumption that the speed of light circumscribes any and
all physical processes or effects. And, so far, this assumption
hasn't been contradicted by experiment.
There are good reasons (via actual experimental designs) to
believe that, in optical EPRBell experiments, there's no
instantaneous physical change in B when A changes, and
vice versa. And, there are no good reasons to believe that
A and B are instantaneously 'communicating'.
So, if you want to hang onto those intuitive feelings about
the possibility that 'something FTL may be happening', then
you'll be doing it in the face of some pretty overwhelming
evidence to the contrary, imho.
Paul Stewart Snyder
Aug7-04, 05:08 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>Tom Trotter <tom129@juno.com> wrote in message news:<410ba9a3\\$1@news.sentex.net>...\n\n> There are good reasons (via actual experimental designs) to\n> believe that, in optical EPRBell experiments, there\'s no\n> instantaneous physical change in B when A changes, and\n> vice versa. And, there are no good reasons to believe that\n> A and B are instantaneously \'communicating\'.\n\nIf that is true - then there seeme ot be nothing that special about\nquantum entanglement. A and B change when some unknown mechanism\ntransmits the state of A to B - that is closer to the hidden variable\nexplanation but does not seem to be consistent with the idea of\nquantum entanglement?\n\nPS\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>Tom Trotter <tom129@juno.com> wrote in message news:<410ba9a3$1@news.sentex.net>...
> There are good reasons (via actual experimental designs) to
> believe that, in optical EPRBell experiments, there's no
> instantaneous physical change in B when A changes, and
> vice versa. And, there are no good reasons to believe that
> A and B are instantaneously 'communicating'.
If that is true - then there seeme ot be nothing that special about
quantum entanglement. A and B change when some unknown mechanism
transmits the state of A to B - that is closer to the hidden variable
explanation but does not seem to be consistent with the idea of
quantum entanglement?
PS
Italo Vecchi
Aug12-04, 08:29 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\nTom Trotter <tom129@juno.com> wrote in message news:<410ba9a3\\$1@news.sentex.net>...\n\n....\n\n > There are good reasons (via actual experimental designs) to\n> believe that, in optical EPRBell experiments, there\'s no\n> instantaneous physical change in B when A changes, and\n> vice versa. And, there are no good reasons to believe that\n> A and B are instantaneously \'communicating\'.\n\nI think this is not true, although that may depend on what we mean by\n"physical change". Experiments (see [1] for a good survey ) confirm\nthat state vector reduction of A and B is simultaneous and consistent\nwith QM, although the observers may verify it only "a posteriori" at\nluminal speed or less. Of course you may always resort to "hidden\nvariables" or some other unverifiable ginmickry pre-orchestrating the\nshow.\n\nThe right setting to view EPR is, imo, Rovelli\'s Relational QM, which\ndoes away with the notion of observer-independent physical quantities.\n\n> So, if you want to hang onto those intuitive feelings about\n> the possibility that \'something FTL may be happening\', then\n> you\'ll be doing it in the face of some pretty overwhelming\n> evidence to the contrary, imho.\n\nYou might provide a reference for your "overwhelming evidence".\n\nIV\n\n[1] A. Zeilinger, Experiment and the foundations of quantum physics,\nRev. Mod. Phys. 71.2 S288-S297 (1999)\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>Tom Trotter <tom129@juno.com> wrote in message news:<410ba9a3$1@news.sentex.net>...
....
> There are good reasons (via actual experimental designs) to
> believe that, in optical EPRBell experiments, there's no
> instantaneous physical change in B when A changes, and
> vice versa. And, there are no good reasons to believe that
> A and B are instantaneously 'communicating'.
I think this is not true, although that may depend on what we mean by
"physical change". Experiments (see [1] for a good survey ) confirm
that state vector reduction of A and B is simultaneous and consistent
with QM, although the observers may verify it only "a posteriori" at
luminal speed or less. Of course you may always resort to "hidden
variables" or some other unverifiable ginmickry pre-orchestrating the
show.
The right setting to view EPR is, imo, Rovelli's Relational QM, which
does away with the notion of observer-independent physical quantities.
> So, if you want to hang onto those intuitive feelings about
> the possibility that 'something FTL may be happening', then
> you'll be doing it in the face of some pretty overwhelming
> evidence to the contrary, imho.
You might provide a reference for your "overwhelming evidence".
IV
[1] A. Zeilinger, Experiment and the foundations of quantum physics,
Rev. Mod. Phys. 71.2 S288-S297 (1999)
Tom Trotter
Aug12-04, 08:31 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\n\nps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407272017.2eb35f58@posting.google. com>...\n> tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407251124.1cd7a038@posting.google. com>...\n> > ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google. com>...\n> > > tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google. com>...\n> > > > ps@ws5.com (Paul Stewart Snyder) wrote in message\n> > > > > We know that when A is changed B is instantaneously\n> > > > > changed,\n> > > >\n> > > > Only in certain contexts, but they don\'t imply signal\n> > > > transfer between A and B.\n> > > >\n> > > This is what continues to bother me - there are at least some contexts\n> > > in which there are instantaneous changes to B when A changes - even if\n> > > A and B are spatially separated. There may not be a signal transfer in\n> > > the formal terms of information transfer, and EPR may not require an\n> > > FTL event, yet there still seems to be a strong intuitive basis\n> > > (Einstein loved intuition) for thinking that something FTL is going on\n> > > here.\n> >\n> > Draw a circle and put two points at different locations on it.\n> > Rotate the circle. A changes with B,and vice versa, but A\n> > doesn\'t affect B, and vice versa. There\'s a non-varying\n> > physical connection between A and B, but no \'communication\'.\n> > Their motion will be correlated in contexts that refer to their\n> > physical connection.\n> >\n> > Note that this is not a model of what\'s happening in Bell-test\n> > photon experiments.\n>\n> You are right, the example is not related to quantum entanglement.\n>\n\nI think the example is related to a type of entanglement,\nbut not the type of entanglement that optical Bell tests have dealt\nwith.\n\n> > In these experiments the entanglement is of\n> > a different sort. A and B don\'t change together. They don\'t\n> > affect each other. They\'re not physically connected in any way.\n> > They do, however, have a motional property in common (polarization)\n> > due to their common source. In observational contexts that refer to\n> > this common property, the photons will produce predictable (correlated)\n> > detection patterns, functionally related to the combined settings\n> > of the analyzers.\n>\n> You are describing one way to understand entanglement, if I understand\n> correctly you require observation of both A and B to produce entanglement,\n\nObservation of both A and B wrt corresponding joint analyzer settings\nis required to *observe* entanglement, not produce it, per se. The\nquantum entanglement itself (between photons emitted by the same atom)\nis assumed to be produced via the emission process, which is never\ndirectly\nobserved.\n\n> so that the correlated patterns are related to the "combined settings of the\n> analyzers", being the "observers" in a Copenhagen Interpretation sense?\n\nThe correlated detection patterns that violate Bell inequalities\nemerge in the observational context of combined analyzer settings.\nI don\'t know what the Copenhagen Interpretation is.\n\n> That may be the case ? however, unless I have missed something, I do\n> not think we have a definitive understanding of the role of observation,\n> if any, in quantum measurement?\n\nDifferent observational contexts require different formulations.\nThat\'s all. This is true whether you\'re dealing with quantum\nor classical phenomena.\n\n[snip]\n\n> So I don\'t believe we can say that the polarization property\n> may be limited to "observational contexts".\n\nThe relevant observational (and therefore the relevant\nformal) parameters are certainly a function of the observational\ncontext.\n\n> If state vector reduction turns out to be unrelated to observation\n> (and/or if many-universe models are correct), then there is an\n> instantaneous physical change in B when A changes that is separate\n> and apart from any observation. That possibility is my source\n> for my intuitive feelings that something FTL may be happening.\n\nThis approach won\'t help wrt to understanding what\'s happening\nin the EPRBell experiments, and it won\'t help wrt to correctly\ninterpreting their physical meaning, imho.\n\nThis \'something FTL\' remains an entirely metaphysical construction\nuntil somebody produces something FTL. The default position\nis the assumption that the speed of light circumscribes any and\nall physical processes or effects. And, so far, this assumption\nhasn\'t been contradicted by experiment.\n\nThere are good reasons (via actual experimental designs) to\nbelieve that, in optical EPRBell experiments, there\'s no\ninstantaneous physical change in B when A changes, and\nvice versa. And, there are no good reasons to believe that\nA and B are instantaneously \'communicating\'.\n\nSo, if you want to hang onto those intuitive feelings about\nthe possibility that \'something FTL may be happening\', then\nyou\'ll be doing it in the face of some pretty overwhelming\nevidence to the contrary, imho.\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>ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407272017.2eb35f58@posting.google.com>...
> tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407251124.1cd7a038@posting.google.com>...
> > ps@ws5.com (Paul Stewart Snyder) wrote in message news:<d16eb5af.0407240954.2eebbbe9@posting.google.com>...
> > > tom129@juno.com (Tom Trotter) wrote in message news:<29df3039.0407191024.38f00d8f@posting.google.com>...
> > > > ps@ws5.com (Paul Stewart Snyder) wrote in message
> > > > > We know that when A is changed B is instantaneously
> > > > > changed,
> > > >
> > > > Only in certain contexts, but they don't imply signal
> > > > transfer between A and B.
> > > >
> > > This is what continues to bother me - there are at least some contexts
> > > in which there are instantaneous changes to B when A changes - even if
> > > A and B are spatially separated. There may not be a signal transfer in
> > > the formal terms of information transfer, and EPR may not require an
> > > FTL event, yet there still seems to be a strong intuitive basis
> > > (Einstein loved intuition) for thinking that something FTL is going on
> > > here.
> >
> > Draw a circle and put two points at different locations on it.
> > Rotate the circle. A changes with B,and vice versa, but A
> > doesn't affect B, and vice versa. There's a non-varying
> > physical connection between A and B, but no 'communication'.
> > Their motion will be correlated in contexts that refer to their
> > physical connection.
> >
> > Note that this is not a model of what's happening in Bell-test
> > photon experiments.
>
> You are right, the example is not related to quantum entanglement.
>
I think the example is related to a type of entanglement,
but not the type of entanglement that optical Bell tests have dealt
with.
> > In these experiments the entanglement is of
> > a different sort. A and B don't change together. They don't
> > affect each other. They're not physically connected in any way.
> > They do, however, have a motional property in common (polarization)
> > due to their common source. In observational contexts that refer to
> > this common property, the photons will produce predictable (correlated)
> > detection patterns, functionally related to the combined settings
> > of the analyzers.
>
> You are describing one way to understand entanglement, if I understand
> correctly you require observation of both A and B to produce entanglement,
Observation of both A and B wrt corresponding joint analyzer settings
is required to *observe* entanglement, not produce it, per se. The
quantum entanglement itself (between photons emitted by the same atom)
is assumed to be produced via the emission process, which is never
directly
observed.
> so that the correlated patterns are related to the "combined settings of the
> analyzers", being the "observers" in a Copenhagen Interpretation sense?
The correlated detection patterns that violate Bell inequalities
emerge in the observational context of combined analyzer settings.
I don't know what the Copenhagen Interpretation is.
> That may be the case ? however, unless I have missed something, I do
> not think we have a definitive understanding of the role of observation,
> if any, in quantum measurement?
Different observational contexts require different formulations.
That's all. This is true whether you're dealing with quantum
or classical phenomena.
[snip]
> So I don't believe we can say that the polarization property
> may be limited to "observational contexts".
The relevant observational (and therefore the relevant
formal) parameters are certainly a function of the observational
context.
> If state vector reduction turns out to be unrelated to observation
> (and/or if many-universe models are correct), then there is an
> instantaneous physical change in B when A changes that is separate
> and apart from any observation. That possibility is my source
> for my intuitive feelings that something FTL may be happening.
This approach won't help wrt to understanding what's happening
in the EPRBell experiments, and it won't help wrt to correctly
interpreting their physical meaning, imho.
This 'something FTL' remains an entirely metaphysical construction
until somebody produces something FTL. The default position
is the assumption that the speed of light circumscribes any and
all physical processes or effects. And, so far, this assumption
hasn't been contradicted by experiment.
There are good reasons (via actual experimental designs) to
believe that, in optical EPRBell experiments, there's no
instantaneous physical change in B when A changes, and
vice versa. And, there are no good reasons to believe that
A and B are instantaneously 'communicating'.
So, if you want to hang onto those intuitive feelings about
the possibility that 'something FTL may be happening', then
you'll be doing it in the face of some pretty overwhelming
evidence to the contrary, imho.
Caroline Thompson
Aug19-04, 04:51 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>\n"Paul Stewart Snyder" <ps@ws5.com> wrote in message\nnews:d16eb5af.0408071126.68c585b7@posting .google.com...\n> Tom Trotter <tom129@juno.com> wrote in message\nnews:<410ba9a3\\$1@news.sentex.net>...\n> \n> > There are good reasons (via actual experimental designs) to\n> > believe that, in optical EPRBell experiments, there\'s no\n> > instantaneous physical change in B when A changes, and\n> > vice versa. And, there are no good reasons to believe that\n> > A and B are instantaneously \'communicating\'.\n>\n> If that is true - then there seems to be nothing that special\n> about quantum entanglement. A and B change when some\n> unknown mechanism transmits the state of A to B....\n\nIt\'s not that there is an unknown mechanism transfering states but that\nthere are all the loopholes and they mean that the Bell tests that are\nviolated are not valid ones. They do not discriminate between QM and local\nhidden variable theories.\n\n> ... - that is closer to the hidden variable\n> explanation but does not seem to be consistent with the idea of\n> quantum entanglement?\n\nI don\' t know what Tom had in mind, but do know that the actual experiments\ncan be explained without assuming entanglement. This is why experimenters\nsuch as Fry and Walther are *still* trying to find loophole-free\nexperiments. See http://plato.stanford.edu/entries/bell-theorem/\n\nCaroline\n\nCaroline H Thompson\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Paul Stewart Snyder" <ps@ws5.com> wrote in message
news:d16eb5af.0408071126.68c585b7@posting.google.c om...
> Tom Trotter <tom129@juno.com> wrote in message
news:<410ba9a3$1@news.sentex.net>...
>
> > There are good reasons (via actual experimental designs) to
> > believe that, in optical EPRBell experiments, there's no
> > instantaneous physical change in B when A changes, and
> > vice versa. And, there are no good reasons to believe that
> > A and B are instantaneously 'communicating'.
>
> If that is true - then there seems to be nothing that special
> about quantum entanglement. A and B change when some
> unknown mechanism transmits the state of A to B....
It's not that there is an unknown mechanism transfering states but that
there are all the loopholes and they mean that the Bell tests that are
violated are not valid ones. They do not discriminate between QM and local
hidden variable theories.
> ... - that is closer to the hidden variable
> explanation but does not seem to be consistent with the idea of
> quantum entanglement?
I don' t know what Tom had in mind, but do know that the actual experiments
can be explained without assuming entanglement. This is why experimenters
such as Fry and Walther are *still* trying to find loophole-free
experiments. See http://plato.stanford.edu/entries/bell-theorem/
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
Caroline Thompson
Aug19-04, 04:51 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>\n"Italo Vecchi" <vecchi@weirdtech.com> wrote in message\nnews:61789046.0408092353.113a4d7e@posting .google.com...\n\n> that state vector reduction of A and B is simultaneous and consistent\n> with QM, although the observers may verify it only "a posteriori" at\n> luminal speed or less. Of course you may always resort to "hidden\n> variables" or some other unverifiable ginmickry pre-orchestrating the\n> show.\n\nHmmm ... Perhaps you\'ve forgotten that hidden variable theory was\nconsidered the natural one before Bohr managed to win the EPR debate (which\nI am not alone in thinking he won unfairly). There is no gimmickry in the\nhidden variable theories described at:\n\nhttp://en.wikipedia.org/wiki/Local_hidden_variable_theory\n\n\n\nIt\'s all just ordinary physics, together with the basic facts of probability\ntheory -- the fact that you can multiply independent probabilities to get\njoint ones.\n\n\n\nCaroline\n\n\n\nCaroline H Thompson\n\n\n\nch.thompson1@virgin.net\nhttp://freespace.virgin.net/ch.thompson1/\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>"Italo Vecchi" <vecchi@weirdtech.com> wrote in message
news:61789046.0408092353.113a4d7e@posting.google.c om...
> that state vector reduction of A and B is simultaneous and consistent
> with QM, although the observers may verify it only "a posteriori" at
> luminal speed or less. Of course you may always resort to "hidden
> variables" or some other unverifiable ginmickry pre-orchestrating the
> show.
Hmmm ... Perhaps you've forgotten that hidden variable theory was
considered the natural one before Bohr managed to win the EPR debate (which
I am not alone in thinking he won unfairly). There is no gimmickry in the
hidden variable theories described at:
http://en.wikipedia.org/wiki/Local_hidden_variable_theory
It's all just ordinary physics, together with the basic facts of probability
theory -- the fact that you can multiply independent probabilities to get
joint ones.
Caroline
Caroline H Thompson
ch.thompson1@virgin.net
http://freespace.virgin.net/ch.thompson1/
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.