Yes, I think that is a fair summary. As long as one uses the standard rules of orthodox quantum theory, i.e. linearity of the operators in Hilbert-Fock spaces, unitarity in the dynamics between von-Neumann strong measurements (including only Hermitian observables, one will get no-signaling in the sense that there is no dependence on distant settings in the local probabilities computed according to standard tracing of the total entangled density matrix (over configuration and/orWigner phase space) over the distant eigenstates.
What I, Antony Valentini, Brian Josephson, Henry Stapp, Steven Weinberg and others have all independently suggested in different variations is a violation of orthodox quantum theory in a more general theory (like Einstein's 1916 GR is to his earlier 1905 SR) allowing non-linear & non-unitary dynamics with a complete breakdown of the Born probability rule. Emergence of new order, as in ground state spontaneous symmetry breaking with Higgs & Goldstone modes, means that the original space of possibilities is changed and there is no reason to expect conservation of probabilities in the original space of possibilities.
On Jun 3, 2013, at 1:53 PM, Ruth Kastner <This email address is being protected from spambots. You need JavaScript enabled to view it. > wrote:
As I understand it, John's point is that DK's approximation, though it may appear valid and could be considered acceptable in some contexts, cannot be used for FTL signalling -- because Nature does not truncate at that level and the terms that Nature keeps in play serve to eliminate the interference DK needs for the signal. So, for purposes of FTL signalling, DK's approximation is not a valid one. This seems to me to address the requirement for a specific refutation of DK's scheme: once Nature's actual detailed behavior is taken into account, the interference goes away.
Ruth
> Date: Mon, 3 Jun 2013 16:14:56 -0400
> Subject: Re: The end of the problem, hopefully
> From: This email address is being protected from spambots. You need JavaScript enabled to view it.
> To: This email address is being protected from spambots. You need JavaScript enabled to view it.
> CC: This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ;This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it.
> > Nick, I would say that so far the approximations are what have lead to the
> errors.
> Cheers
> John
> > > > John
> >
> > "We will mess things up if we do anything
> > other than an exact calculation."
> >
> > This is a rather pessimistic view, John, and amounts
> > to abandoning the Kalamidas Scheme without any explanation
> > of where it fails except: "Well it's just an approximation".
> >
> > Since the approximations rA < 1 is used all the time in quantum optics,
> > it seems we owe Kalamidas and the quantum optics community at least
> > the favor
> > of showing them how to make a "correct approximation" in this matter
> > of single photon/Coherent state mixing.
> >
> > Nick
> >
> > PS: I've uncoupled G & C.
> >
> >
> >
> >
> >
> >
> > On Jun 3, 2013, at 10:43 AM, John Howell wrote:
> >
> >> Hello Everyone,
> >> I just have a few comments
> >>
> >> 1) I think we should respect Giancarlo's and Chris's desire to
> >> decouple
> >> from this conversation. So, I think they should not be copied in on
> >> further emails.
> >>
> >> 2) I have done the full calculation without any approximations,
> >> expansions
> >> etc. for the PACS and DFS, and as expected, there is no
> >> interference. I
> >> have already shown the DFS, so the PACS is Attached.
> >>
> >> 3) The second order cross correlation for the evolution of the field
> >> operators vs the Suda state evolution yield different results. I
> >> need to
> >> double check my answers (long calculation).
> >>
> >> 4) I like Chris's approach, which is basically to consider a
> >> binomially
> >> distributed photon number outcome interfering with a photon from
> >> the other
> >> port. That will take me a while, but it should corroborate the Suda's
> >> state evolution paper.
> >>
> >> Cheers
> >> John<FullCalculationNoSignaling.pdf>
> >
> >
>
On Jun 3, 2013, at 12:46 PM, Suda Martin <This email address is being protected from spambots. You need JavaScript enabled to view it. > wrote:
Nick, thanks for nice comment!
As regards the |00> term I am not at all surprised. In fact, because of the following considerations:
Each coherent state (CS) consists of an infinite sum of Fock states of certain probabilities, the vacuum state included. If these infinite many terms are taken into account this state has more or less classical properties (fully contrary to a Fock state), even though a CS is a regular quantum state! A CS = D|0>. D is the well-known exponential operator where a and a+ appear in the exponent. A DFS = D|1>. Both states (of different modes 3 and 4, in our case) can therefore be expanded in (infinite) Taylor series. The product of such a series expansion inevitably includes a |00> term. An artificial truncation of the series after few terms (2 in our case) contains automatically a |00> term at a prominent position. Therefore a physical interpretation becomes difficult and is in a certain manner misleading. So don't attach too great importance to such a |00> state. It's a result of the early truncation of the Taylor expansion. And it has to be considered whatsoever. Martin
________________________________________
Von: nick herbert [This email address is being protected from spambots. You need JavaScript enabled to view it. ]
Gesendet: Montag, 3. Juni 2013 18:12
An: Suda Martin
Cc: JACK SARFATTI; Demetrios Kalamidas; Ghirardi Giancarlo; CHRISTOPHER GERRY; John Howell; Ruth Elinor Kastner; Romano This email address is being protected from spambots. You need JavaScript enabled to view it. [MATH]; David Kaiser; S-P Sirag; Brian Josephson; Fred Wolf
Betreff: Re: AW: Martin Suda's Refutation? Wait a minute Nick your 11 & 00 amplitudes do not cancel to zero!
Martin--
This is a nice summary of your work.
But could you say a bit more about
where the |0, 0> term comes from?
Does it emerge naturally
from the renormalization procedure.
Nick
PS. Nick has been calling result #1
(PACS_DFS_BS.pdf) the Martin Suda Paradox
because its conclusion is rather counter-intuiyive.
On Jun 3, 2013, at 10:44 AM, nick herbert <This email address is being protected from spambots. You need JavaScript enabled to view it. > wrote:
Demetrios--
Indeed. Right now it doesn't add up.
Once the pros are able to clearly explain
the physical origin of the high amplitude |00> term
the refutation is airtight and complete.
But minus an understanding
of how this term physically arises
at the beamsplitter
Suda's wonderful (and surely correct) refutation seems
mere sleight of math.
Nick
On Jun 3, 2013, at 9:26 AM, Demetrios Kalamidas wrote:
Hi all,
Here is my concise understanding of the |00> term:
The probability of the right-going Fock photon being reflected is proportional to |r|^2, with |r|-->0. Thus, this reflection probability is vanishing.
However, as everybody can plainly see, the probability for the |00> outcome to occur is proportional to |r*alpha|^2, which is never equal |r|^2, and can be made far larger.
So it doesn't add up....you can't explain the missing right-going Fock photon as that being reflected by the highly transmissive beam splitters.
Probability |r|^2 is vanishing, and can be made as small as we wish (infinitesimal), while the product |r*alpha| can be maintained at any value we want just by increasing 'alpha' accordingly, and therefore the probability |r*alpha|^2 is always finite.
Demetrios
On Mon, 3 Jun 2013 09:14:53 -0700
nick herbert <This email address is being protected from spambots. You need JavaScript enabled to view it. > wrote:
GianCarlo--
It's important that all aspects of Martin's proof be examined to make certain that what we have is a true refutation and not a
mere pseudo-refutation motivated by what we know the answer has to be.
Nick
On Jun 3, 2013, at 5:28 AM, ghirardi wrote:
Dear all,
I have no doubts now that Kalamidas' proposal does not work and its refutation does not require any new insight in subtle quantum problems.
Accordingly I will write a precise comment and I invite everybody to consider it seriously and not to go on suggesting strange effects and so on to overcome difficulties which do not exist.
GianCarlo
Il giorno Jun 3, 2013, alle ore 6:06 AM, nick herbert ha scritto:
The problem here, as in summing Feynman diagrams, is to account for all possible outcomes. One possible outcome is that lower path is EMPTY and the
upper photon "goes down the hole", that is, it's reflected instead of being transmitted. Have you calculated the amplitude of this "down the hole" event and compared its magnitude with the amplitudes of all the other events you are looking at, especially the amplitude |1, 1>. Every photon that goes "down the hole" contributes to |0, 0>. So how big is this term?
On Jun 2, 2013, at 4:54 PM, Demetrios Kalamidas wrote:
Indeed Jack, but it seems that this term is quite problematic: the |00> term means that there is a left-going photon present in a superposition of modes a1 and b1 BUT its right-going partner has vanished! I am studying this and I don't think it is trivial or easily explained. Last, the PACS formulation only contains terms that make physical sense. This |00> is a surprising feature that arose out of the discussion surrounding my scheme.
Demetrios
On Sun, 02 Jun 2013 15:42:41 -0700
JACK SARFATTI <This email address is being protected from spambots. You need JavaScript enabled to view it. > wrote:
These amplitudes, as you wrote them, do not cancel as you claim - see below.
Summing them ~ 2iIm{alpha} =/= 0
On Jun 2, 2013, at 12:56 AM, nick herbert <This email address is being protected from spambots. You need JavaScript enabled to view it. > wrote:
However--and this is the gist of the Suda refutation--the additional Suda term |0.0> has precisely the right amplitude
to EXACTLY CANCEL the effect of the Kalamidas |1,1> term. Using A (Greek upper-case alpha) to represent "alpha",
Martin calculates that the amplitude of the Kalamidas |1,1> term is A. And that the amplitude of the Suda |0,0> term is -A*.
And if these amplitudes are correct, the total interference at Alice's detectors completely disappears.
Kalamidas Fans--
I have looked over Martin Suda's two papers entitled 1. Taylor expansion of Output States and 2. Interferometry at the 50/50 BS.
My conclusion is that Martin is within one millimeter of a solid refutation of the kalamidas scheme. Congratulations, Martin, on
achieving this result and on paying so much close attention to kalamidas's arguments.
The result, as expected, comes from a very strange direction. In particular, the approximation does not enter into Suda's refutation.
Martin accepts all of kalamidas's approximations and refutes him anyway.
I have not followed the math in detail but I have been able to comprehend the essential points.
First, on account of the Martin Suda paradox, either PACS or DFS can be correctly used at this stage of the argument. So martin
derives the kalamidas result both ways using PACS (Kalamidas's Way) and then DFS (Howell's Way). Both results are the same.
Then Martin calculates the signal at the 50/50 beam splitter (Alice's receiver) due to Bob's decision to mix his photon with a coherent state |A>.
Not surprisingly Martin discovers lots of interference terms.
So Kalamidas is right.
However all of these interference terms just happen to cancel out.
So Kalamidas is wrong.
Refutation Complete. Martin Suda Wins.
This is a very elegant refutation and if it can be sustained, then Kalamidas's Scheme has definitively
entered the Dustbin of History. And GianCarlo can add it to his upcoming review of refuted FTL schemes.
But before we pass out the medals, there is one feature of the Suda Refutation that needs a bit of justification.
Suda's formulation of the Kalamidas Scheme differs in one essential way from Demetrios's original presentation.
And it is this difference between the two presentations that spells DOOM FOR DEMETRIOS.
Kalamidas has ONE TERM |1,1> that erases which-way information and Suda has two. Suda's EXTRA TERM is |0,0>
and represents the situation where neither of Bob's primary counters fires.
Having another term that erases which-way information would seem to be good, in that the Suda term might be expected to increase
the strength of the interference term.
However--and this is the gist of the Suda refutation--the additional Suda term |0.0> has precisely the right amplitude
to EXACTLY CANCEL the effect of the Kalamidas |1,1> term. Using A (Greek upper-case alpha) to represent "alpha",
Martin calculates that the amplitude of the Kalamidas |1,1> term is A. And that the amplitude of the Suda |0,0> term is -A*.
And if these amplitudes are correct, the total interference at Alice's detectors completely disappears.
Congratulations, Martin. I hope I have represented your argument correctly.
The only task remaining is to justify the presence (and the amplitude) of the Suda term. Is it really physically reasonable,
given the physics of the situation, that so many |0,0> events can be expected to occur in the real world?
I leave that subtle question for the experts to decide.
Wonderful work, Martin.
Nick Herbert
GianCarlo Ghirardi
Emeritus
University of Trieste
Italy
Begin forwarded message:
From: Suda Martin <This email address is being protected from spambots. You need JavaScript enabled to view it. >
Subject: AW: The end of the problem, hopefully
Date: June 3, 2013 11:10:24 AM PDT
To: John Howell , Demetrios Kalamidas <This email address is being protected from spambots. You need JavaScript enabled to view it. >
Thanks, John, for "Full calculation, no approximation". Somewhere the phase exp(i Phi) is missing in Eq.(2)? And you forgot perhaps the different adjustments of 1,0 and 0,1 in Eq.(2)? But I am sure the results are the same as in Eqs.(3) and (4). Great!
Martin
________________________________________
Von: John Howell [This email address is being protected from spambots. You need JavaScript enabled to view it. ]
Gesendet: Montag, 3. Juni 2013 19:43
An: Demetrios Kalamidas
Cc: nick herbert; ghirardi; JACK SARFATTI; CHRISTOPHER GERRY; John Howell; Suda Martin; Ruth Elinor Kastner; Romano This email address is being protected from spambots. You need JavaScript enabled to view it. [MATH]; David Kaiser; S-P Sirag; Brian Josephson; Fred Wolf
Betreff: The end of the problem, hopefully
Hello Everyone,
I just have a few comments
1) I think we should respect Giancarlo's and Chris's desire to decouple
from this conversation. So, I think they should not be copied in on
further emails.
2) I have done the full calculation without any approximations, expansions
etc. for the PACS and DFS, and as expected, there is no interference. I
have already shown the DFS, so the PACS is Attached.
3) The second order cross correlation for the evolution of the field
operators vs the Suda state evolution yield different results. I need to
double check my answers (long calculation).
4) I like Chris's approach, which is basically to consider a binomially
distributed photon number outcome interfering with a photon from the other
port. That will take me a while, but it should corroborate the Suda's
state evolution paper.
Cheers
John
Jack Sarfatti
Kalamidas Affair update June 3, 2013
Jack Sarfatti Begin forwarded message:
From: nick herbert <This email address is being protected from spambots. You need JavaScript enabled to view it. >
Subject: Re: AW: Martin Suda's Refutation? Wait a minute Nick your 11 & 00 amplitudes do not cancel to zero!
Date: June 3, 2013 9:11:17 AM PDT
To: Suda Martin
Martin--
This is a nice summary of your work.
But could you say a bit more about
where the |0, 0> term comes from?
Does it emerge naturally
from the renormalization procedure.
Nick
PS. Nick has been calling result #1
(PACS_DFS_BS.pdf) the Martin Suda Paradox
because its conclusion is rather counter-intuitive.
On Jun 3, 2013, at 3:46 AM, Suda Martin wrote:
Dear all,
Thank you very much for emails and discussion!
Let me summarize my results so far which are seen in the attachment. They demonstrate that it is unlikely to be FTL signaling in the system of DK.
4 files are attached:
1) PACS_DFS_BS.pdf
2) PACS_DFS_Howell_Suda.pdf
3) Taylor-Exp-PACS_DFS_Howell_Suda.pdf
4) Interf_BS_50_50_Suda.pdf
I would like to discuss these 4 short statements sequentially.
1) In PACS_DFS_BS.pdf I showed that for input |1>|alpha> or |alpha>|1>, behind a BS both the PACS-formulation of the output state and the DFS-formulation of the output state are identical. This can be shown using the relation a^{+}D = Da^{+} + alpha^{*}D and, in addition, using the well-known Stokes relations of a BS.
2) In PACS_DFS_Howell_Suda.pdf I have demonstrated (and this is only a supplement to John Howells paper) that the normalizations of both, the input wave function |psi_{0}> and the output wave function |psi'_{0}>, are exactly = 1. The orthogonality between DFS and the coherent state |alpha> is thereby crucial. This applies for the PACS-formulation as well as for the DFS-formulation. Because of this orthogonality no interference can appear.
3) In Taylor-Exp-PACS_DFS_Howell_Suda.pdf the Taylor expansion of the displacement operator D has been introduced in order to follow DK's calculation procedure. PACS as well as DFS are taken into account. The approximation |r alpha|<
4) In Interf_BS_50_50_Suda.pdf a more complete T series expansion of D and DFS is used (see Eq.27 and Eq.28 of John's paper) and the normalization of the wave function |psi'_{0}> behind the BS yields 1 + 2|r alpha|^{2} + |r alpha|^{4} instead of being exactly=1. The wave function after the 50/50 BS on the left side produces therefore an "interference term" with a probability |p_{10}|^{2} = 4|r alpha|^{2} [1-sin(Phi)] and this probability is proportional to
|r alpha|^{2}. This is not a miracle because of the modified normalization. The additional term appearing in the norm is proportional to |r alpha|^{2} as well!
As a result one can say that the whole problem is up to the T expansion of the D operator and hence of the modification of the normalization condition.
Nice regards,
Peter Lynn Martin
From: nick herbert <
Subject: Re: AW: Martin Suda's Refutation? Wait a minute Nick your 11 & 00 amplitudes do not cancel to zero!
Date: June 3, 2013 9:11:17 AM PDT
To: Suda Martin
Martin--
This is a nice summary of your work.
But could you say a bit more about
where the |0, 0> term comes from?
Does it emerge naturally
from the renormalization procedure.
Nick
PS. Nick has been calling result #1
(PACS_DFS_BS.pdf) the Martin Suda Paradox
because its conclusion is rather counter-intuitive.
On Jun 3, 2013, at 3:46 AM, Suda Martin wrote:
Dear all,
Thank you very much for emails and discussion!
Let me summarize my results so far which are seen in the attachment. They demonstrate that it is unlikely to be FTL signaling in the system of DK.
4 files are attached:
1) PACS_DFS_BS.pdf
2) PACS_DFS_Howell_Suda.pdf
3) Taylor-Exp-PACS_DFS_Howell_Suda.pdf
4) Interf_BS_50_50_Suda.pdf
I would like to discuss these 4 short statements sequentially.
1) In PACS_DFS_BS.pdf I showed that for input |1>|alpha> or |alpha>|1>, behind a BS both the PACS-formulation of the output state and the DFS-formulation of the output state are identical. This can be shown using the relation a^{+}D = Da^{+} + alpha^{*}D and, in addition, using the well-known Stokes relations of a BS.
2) In PACS_DFS_Howell_Suda.pdf I have demonstrated (and this is only a supplement to John Howells paper) that the normalizations of both, the input wave function |psi_{0}> and the output wave function |psi'_{0}>, are exactly = 1. The orthogonality between DFS and the coherent state |alpha> is thereby crucial. This applies for the PACS-formulation as well as for the DFS-formulation. Because of this orthogonality no interference can appear.
3) In Taylor-Exp-PACS_DFS_Howell_Suda.pdf the Taylor expansion of the displacement operator D has been introduced in order to follow DK's calculation procedure. PACS as well as DFS are taken into account. The approximation |r alpha|<
4) In Interf_BS_50_50_Suda.pdf a more complete T series expansion of D and DFS is used (see Eq.27 and Eq.28 of John's paper) and the normalization of the wave function |psi'_{0}> behind the BS yields 1 + 2|r alpha|^{2} + |r alpha|^{4} instead of being exactly=1. The wave function after the 50/50 BS on the left side produces therefore an "interference term" with a probability |p_{10}|^{2} = 4|r alpha|^{2} [1-sin(Phi)] and this probability is proportional to
|r alpha|^{2}. This is not a miracle because of the modified normalization. The additional term appearing in the norm is proportional to |r alpha|^{2} as well!
As a result one can say that the whole problem is up to the T expansion of the D operator and hence of the modification of the normalization condition.
Nice regards,
Peter Lynn Martin