Bell and quantum philosophy

52 views
Skip to first unread message

Mark Hadley

unread,
Dec 4, 2025, 8:43:50 AMDec 4
to Jarek Duda, Bell Inequalities and quantum foundations, Álvaro García, nature of time, Models of particles, Michael Masters


For those who are interested in a philosophical analysis. I found this...



Álvaro García

unread,
Dec 6, 2025, 3:40:10 AMDec 6
to Mark Hadley, Jarek Duda, Bell Inequalities and quantum foundations, nature of time, Models of particles, Michael Masters
Thanks, Mark. 

Some references that I was not aware of look interesting. I'll check it out, since I am working on Bell's stuff. Best,

Álvaro
 
******************************************************************************************************* 
Dr. Álvaro García López
Profesor Contratado Doctor
Nonlinear Dynamics, Chaos and Complex Systems Group, Universidad Rey Juan Carlos
Tulipán s/n, 28933 Mostoles, Madrid, Spain
Tel.: (+34) 916648921
*******************************************************************************************************
*******************************************************************************************************

Marian Kupczynski

unread,
Dec 6, 2025, 6:48:27 AMDec 6
to Mark Hadley, Jarek Duda, Bell Inequalities and quantum foundations, Álvaro García, nature of time, Models of particles, Michael Masters
In spite of  90  years of discussions of Bell Inequalities many authors still do not understand what is their meaning of Bell inequalities and what really was tested in Bell Tests. Here is a very short comment in PRA clarifying some of these misunderstandings.  The authors replied to this comment but their misreported and misunderstood the content , what may notice easily,   so they should   publish  an erratum 
I wish you a White Christmas and a Happy New Year 2026 for all of you. In Canada is already everything white and so it will stay like this for Christmas.
Marian



On Thu, Dec 4, 2025 at 8:43 AM Mark Hadley <drmark...@gmail.com> wrote:


For those who are interested in a philosophical analysis. I found this...



--
You received this message because you are subscribed to the Google Groups "nature of time" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nature-of-tim...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/nature-of-time/CAHfX0DppDdX0vsiMitKS%3D8V8NrB%3DQsr-_f9-USm3xOvawNKS-Q%40mail.gmail.com.

Fred Diether

unread,
Dec 6, 2025, 2:42:17 PMDec 6
to Bell inequalities and quantum foundations
90 years???  1964 gives 36 plus 25 for 61 years for Bell's junk.  The inequalities are broken physics and don't mean anything.  All the quantum experiments prove is that the predictions of quantum mechanics are correct.

Fred Diether

unread,
Dec 6, 2025, 4:13:01 PMDec 6
to Bell inequalities and quantum foundations
Mark sent me an email in reply to my post that said,

"Experiments do indeed confirm that QM is correct. As do a hundred experiments every day.

But the experiments do more than that, and this is where the controversy comes in. They prove non locality WITHOUT assuming QM"

For some reason, Mark didn't want the group to see what he said.  Anyways, we would like to know how a quantum experiment proves non-locality without assuming QM?

Triple LOL!

Mark Hadley

unread,
Dec 6, 2025, 4:14:45 PMDec 6
to Fred Diether, Bell inequalities and quantum foundations
Thanks Fred, it was intended to go to everyone 

--
You received this message because you are subscribed to the Google Groups "Bell inequalities and quantum foundations" group.
To unsubscribe from this group and stop receiving emails from it, send an email to Bell_quantum_found...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/Bell_quantum_foundations/16d8071f-c772-42ef-b42d-becbeed0ac76n%40googlegroups.com.

Fred Diether

unread,
Dec 6, 2025, 9:22:06 PMDec 6
to Bell inequalities and quantum foundations
You're welcome.  The truth will set you free.  

Austin Fearnley

unread,
Dec 13, 2025, 7:01:37 AM (8 days ago) Dec 13
to Bell inequalities and quantum foundations
Table 1 on page 24 makes no mention of retrocausality in its taxonomy of nineteen QM interpretations (unless it is subsumed into one of the nineteen categories?)
My own interest is in quantum retrocausality and its implications, where those implications are very different from those of macroscopic retrocausality.
I am piqued by the idea of weak measurement as it is sponsored by Aharonov who is also promoting Two State Vector Formalism which uses retrocausality. One state forwards in time and the other, second, state backwards.  Yet weak measurement seems to me to be in contradiction to the quantum nature of particles.

On retirement 20 years ago I took up an interest in fundamental particles and posted for a few years on the unmoderated sci.physics website.  Rather like the website here which also appears to be unmoderated. Anyway, I did learn a lot there.  In particular, I learned about the importance of quanta and realised that I had nowhere near appreciated their extent previously.  I gave up that site as my skin was not thick enough when being called a liar by Androcles as I had the cheek to accept the correctness of special relativity.  That site was shut down in 2024 along with all other google sites, even some which were moderated. Not sure why our current site is still running?

So I bought the 'dummies' book for quantum mechanics and also started Leonard Susskind's online courses from Stanford University.  The book, and some others, were not sufficent for me to learn using books alone, but the online courses were very easy.  Almost immediately I came across the supposed strangeness (causes a click or does not on the measuring device) of an electron's behaviour when approaching a magnet. I never appreciated why it was considered to be so strange.

Next, I very slowly developed my preon model. My preons are quantised particles with four preons comprising an electron.  So the electron is still apparently quantised in flight except that at an interaction where the electron changes/swaps some of its preons and carries on after the interaction still having four preons, though not all the same preons as before the interaction.

The quantised nature of the electron, at least in flight, makes the idea of weak measurement very strange to me.  A 'strong' measurement in my model is one like I have described above.  But what changes within the electron in a weak measurement?  In my model, the four preons should not change or swap in a weak measurement.  And they definitely should not change or swap in flight before the interaction as that would destroy the quantum nature of the electron. My preon model also includes other layers such as hexarks which are also quantised with 24 hexarks comprising one preon.  But I do not see any mechanism for changing hexarks at a weak interaction.  If it were possible to change or swap hexarks then I could more accept weak measurements.  This train of ideas leads to Quantum Field Theory for which Susskind has an excellent online course.  I do like the maths and ideas in QFT although I still think in terms of using the approach of quantised particles and preons.

Enough for now.  Thanks to Mark for the reference to 'heptalemma'.

Fred Diether

unread,
Dec 13, 2025, 1:15:34 PM (8 days ago) Dec 13
to Bell inequalities and quantum foundations
Sci.physics was not shut down.  It's a Usenet group not a Google group.  Google stopped supporting Usenet last year.  You need a Usenet newsgroup provider to access it.  I wonder if there is as much nonsense on the sci.physics Usenet groups now that you can't post via Google groups.

Yeah, thanks to Mark for adding to the nonsense here.

Local Quantum Mechanical Prediction of the Singlet State Using Geometric Algebra
https://www.scirp.org/journal/paperinformation?paperid=147271

On Saturday, December 13, 2025 at 4:01:37 AM UTC-8 ben...@hotmail.com wrote:
snip...


On retirement 20 years ago I took up an interest in fundamental particles and posted for a few years on the unmoderated sci.physics website.  Rather like the website here which also appears to be unmoderated. Anyway, I did learn a lot there.  In particular, I learned about the importance of quanta and realised that I had nowhere near appreciated their extent previously.  I gave up that site as my skin was not thick enough when being called a liar by Androcles as I had the cheek to accept the correctness of special relativity.  That site was shut down in 2024 along with all other google sites, even some which were moderated. Not sure why our current site is still running?

snip....

Bryan Sanctuary

unread,
Dec 14, 2025, 12:22:47 PM (7 days ago) Dec 14
to Fred Diether, Bell inequalities and quantum foundations
Hi All,

Fred has published a paper.  It is based upon Joy Christian's work.  In their papers, they always paste in a set of equations that Joy says are the key to his process.  Fred parrots these in Equations 20 to 27.  Years ago, I amicably asked Joy to explain a problem I had with them, and he replied that there were no errors, but did not answer my question. So I had to figure it out.

Now I see that Fred's work is based on the same equations, and Eq 25 to 26 is wrong.  That is, r_o is not zero when averaged, unless one first assumes isotropy and independence of . Hclassifies the remainder  as a removable “cross-product” and sets it to zero. However,   contains non trivial scalar couplings, and r_0 has a scalar component that does not average away, so the step cannot be justified.

There are other errors in the paper, but the above is enough to kill it. 

Bryan

--
You received this message because you are subscribed to the Google Groups "Bell inequalities and quantum foundations" group.
To unsubscribe from this group and stop receiving emails from it, send an email to Bell_quantum_found...@googlegroups.com.

Fred Diether

unread,
Dec 14, 2025, 1:02:03 PM (7 days ago) Dec 14
to Bell inequalities and quantum foundations
Hmm.... I wonder what the average of thousands of zeroes is?  Oh yeah, it's zero!

BS is still completely clueless and now has shown us he can't even do simple math.

Mark Hadley

unread,
Dec 14, 2025, 1:05:38 PM (7 days ago) Dec 14
to Fred Diether, Bell inequalities and quantum foundations
Neither you nor Bryan can do simple maths. You could make a great team.

Mark

Bryan Sanctuary

unread,
Dec 14, 2025, 1:35:23 PM (7 days ago) Dec 14
to Mark Hadley, Fred Diether, Bell inequalities and quantum foundations
 Mark, 

So according to you, both Fred and I are wrong, then you are saying that r_0 is equal and not equal to a scalar.  We cannot both be wrong.  Which is it Mark, do the math and show me?

Bryan

Mark Hadley

unread,
Dec 14, 2025, 1:41:06 PM (7 days ago) Dec 14
to Bryan Sanctuary, Fred Diether, Bell inequalities and quantum foundations
I haven't read the r_0 paper. I'll trust your criticism  on that.

I just know that you and Fred have have insurmountable problems understanding small simple maths steps. You not being able to add correlations, Fred not understanding steps in BI or CHSH.
Absurdly so.

Fred Diether

unread,
Dec 14, 2025, 2:10:12 PM (7 days ago) Dec 14
to Bell inequalities and quantum foundations
And... here comes Mark with more of his typical nonsense since he offers no evidence of math errors.

Fred Diether

unread,
Dec 14, 2025, 5:44:33 PM (7 days ago) Dec 14
to Bell inequalities and quantum foundations
LOL!  Mark is way too lazy and incompetent to be able to understand the math involved in r_0 and probably any of my paper.  Take the limits; r_0 is zero all day long.  Guess what?  You are wrong big time and still clueless to Joy's work.  It proves that Bell was wrong.

Bryan Sanctuary

unread,
Dec 15, 2025, 1:52:51 AM (6 days ago) Dec 15
to Mark Hadley, Fred Diether, Bell inequalities and quantum foundations
Dear Mark,

Thanks for the explanation about my arithmetic. . 

1. I calculated Fred's r_0 and I get a scalar before averaging so 
54066396-5542-49b8-a4ac-c856d2dd8942.png
It is incumbent upon Fred to prove this wrong,  If I am correct, then Fred's paper is wrong.. Also Fred's simulation violates the rules and his function  predetermines the outcome.  He does not use binary pairs.  

2. ADD/AVERAGE  Mark: you seem obsessed with the add-average issue, and I thought we had come to an impasse and decided to agree to disagree. I try again to give an explanation
.
a. I agree that if you have different particles which are correlated, that we take the average.  
But
b. I ask you how to average only one particle?  That is, I take spin as not a POINT but a structured BIVECTOR.  That electron has two dual sectors.Contextuality determines the sector of one spin>
image.png
Now Mark, please tell me how you average over one particle pair with each having different outcomes from the SAME particle?  That is
Particle A has two possible outcomes and so does B, so you have A_1B_1   or    A_1B_2   or  A_2B_1  or  A_2B_2.  My cross terms vanish so you have either  A_1B_1  or  A_2B_2.  One or the other.. Each makes an independent contribution.  You do not average them, you add them because they exist independently.

As I have said, Bell ignored complementarity and dealt with one convex set.   

Bryan

Mark Hadley

unread,
Dec 15, 2025, 4:52:33 AM (6 days ago) Dec 15
to Bryan Sanctuary, Fred Diether, Bell inequalities and quantum foundations
Bryan,
We had not agreed to disagree. You are wrong and you have fraudulently claimed to have won a bet that you lost.

For one particle pair there is nothing to average. In the epr setup the results are either correlated (+1) or anti correlated (-1) regardless of the underlying mechanism. If you have two possible mechanisms you do not get +2 or -2

There is no convexity assumption in BI or CHSH

Mark

Austin Fearnley

unread,
Dec 15, 2025, 5:48:09 AM (6 days ago) Dec 15
to Bell inequalities and quantum foundations
Dear all

Bryan disagreed with Christian's old model achieving -a.b.

I have a note of mine commenting on Joy's one page paper written in 2019 and then tucked away on my wordpress blog site.
https://ben6993.wordpress.com/commentary-on-joy-christians-model-of-correlation-a-b-in-his-one-page-paper/

In view of Bryan's comments on Fred's paper, I will update my paper to cover Fred's model.  This may take some time, and it is a busy time of year.  I have Christmas cards to write and some to paint.  The method in my paper was to go back to the raw data as the need to make the formulae work on raw data can be overlooked in an abstract mathematical paper.

I am not completely clear about Bryan's 'I3 (a x b) being not equal to zero' but, in Joy's paper, there was an approximately equal mix of I3 being +1 and I3 being -1 in a similar calculation.  This would let the sum approximately approach zero, if one were allowed to treat I3 as varying within a single calculation.  In my commentary I was not too concerned about that, though most other critics were against using +1 and -1 in the same calculation.

--------

Hidden variables are hidden to Alice and Bob.  Alice and Bob calculate the correlation very simply based on paired outcomes from their measuring devices. The hidden variables are not hidden to the simulation program but the simulation should lead directly to pairs of outcomes corresponding to the measurements made by Alice and Bob.  Then the correlation is simple to calculate in the usual way.  

Bryan Sanctuary

unread,
Dec 15, 2025, 7:18:08 AM (6 days ago) Dec 15
to Mark Hadley, Fred Diether, Bell inequalities and quantum foundations
Mark,  

 I tried again to explain and you do not want to know what the second linearization of the KG  equation gives,  so I will give up.  

You said "There is no convexity assumption in BI or CHSH" which is exactly my point.  Bell is classical and quantum has complementary sectors.  You want me to average two distinct orthogonal contributions:  \sigma and i \sigma.  Do you know how to average position and momentum, which are also  complementary?  If so let me know

Best wishes

Bryan

Austin Fearnley

unread,
Dec 15, 2025, 7:51:55 AM (6 days ago) Dec 15
to Bell inequalities and quantum foundations
Dear all

In my previous post I wrote "Hidden variables are hidden to Alice and Bob.  Alice and Bob calculate the correlation very simply based on paired outcomes from their measuring devices. The hidden variables are not hidden to the simulation program but the simulation should lead directly to pairs of outcomes corresponding to the measurements made by Alice and Bob.  Then the correlation is simple to calculate in the usual way."

It seems that I need to amplify this.  A simulation should use hidden variables to generate the table of unit integer outcomes of Alice and Bob.  As I see it, Bryan has not done this.  That is a failure or rather an incompleteness in using his model.  Alice and Bob need pairwise final measurements to calculate the correlation in a very simple manner.  There is no point in calculating the correlation on hidden variable data which have not first been brought to real observable values. So the order of the procedure should be to calculate the observables and next use the observables to calculate the correlation.  It should not be to calculate the correlations and then calculate the observables.  At some point in earlier posts Bryan said that observables would depend about 40% of the time on p and 60% of the time on c (not sure if 40% or p or c are the correct terms).  I suggest that Bryan uses, in a simulation, pseudo random techniques to allocate pseudo random values to get observables, based on the 40% estimate.  Then there will be a table of real observables.  Then calculate the correlation in a very simple manner. 

Mark Hadley

unread,
Dec 15, 2025, 8:07:07 AM (6 days ago) Dec 15
to Bryan Sanctuary, Fred Diether, Bell inequalities and quantum foundations
I am not interested in your linearization if KG. It has no relevance to BI or to CHSH and I don't trust any new work by you. I studied the paper that you asked the group to study. It was fatally flawed.

Bell is about correlations between A and B in an EPR experiment, those are what need to me combined.

BI and CHSH allow the hidden variables to be complementary, orthogonal or non complimentary in any way you define it so long as the probability at B does not depend on settings at A and viceversa .

As I understand your work some events are from one mechanism and some clicks are from a different mechanism that you call irthogonal. That is quite compatible with Bells derivation. The clicks are from an unremarkabke combinations of two populations. Adding the word orthogonal adds nothing to the derivation or the statistics.

Averaging position and momentum.,
Well if some events were due to particle position and some events were due to particle momentum. The two populations can be combined in the usual way. Totally compatible with BI and with the formulas that you were given.

You can also run the experiment on a spreadsheet. Get some typical +1 -1 results and work out the correlations yourself.

Mark

Bryan Sanctuary

unread,
Dec 15, 2025, 9:03:02 AM (6 days ago) Dec 15
to Austin Fearnley, Bell inequalities and quantum foundations
Dear Austin,

  Thank you for your comment. In my simulation,   https://doi.org/10.3390/quantum6030026 hidden variables are used only to generate the final ±1 outcomes for Alice and Bob, and the correlations are then calculated in the standard way from those observable outcomes. The procedure therefore follows the usual rules.  The only difference from Bell is the underlying duality. That is,  I calculate the correlation from the even parity sector of my bivector, and then I calculate the correlation for the odd parity sector. The two contributions are distinct and dual, so I add them.  I agree with the experiments, with no need for  non-local events.

If you want,  I can generate a table of ±1 outcomes for each convex sector and compute in the  way you describe; but the difference is there is a table for each set.  So I do not assume a single global table assigning outcomes for all incompatible settings simultaneously.  

I think you miss understood a bit:  p means polarization and c means coherence.  Contextuality shows that from a random source, 25% of the time the correlation is from coherence and 75% the correlation is from polarization. This appears from the simulation, the 25:75 split comes from the program and is not imposed.

So this is certainly not a failure of my model or its incompleteness. I can send you the program if you wish.

 I appreciate your input and the opportunity to clarify this, and would welcome your response.

Bryan



Bryan Sanctuary

unread,
Dec 15, 2025, 9:41:39 AM (6 days ago) Dec 15
to Mark Hadley, Fred Diether, Bell inequalities and quantum foundations

Mark,

If you think my work is flawed, please point to the specific step(s) where the reasoning fails, ideally with reference to an equation or section.
https://doi.org/10.3390/axioms14090668 

You may choose to ignore the second linearization of the KG equation, but its existence is a mathematical fact independent of my views or yours. Once an alternative is known to exist, it cannot be dismissed by assumption; it must be assessed on its merits. Historically, Dirac’s linearization became standard, but it was not known that it was not unique until my work.

It is like discovering relativity but insisting on Galilian invariance.

The relevance of the second way is that it gives the classical origin of spin, resolves all problems with the SM (that I know of), and gives a geometrical basis for our observations which Dirac's does not give. It replaces QFT with geometric algebra.  The relevance to the EPR paradox is that it resolves quantum weirdness.

Bryan








Mark Hadley

unread,
Dec 15, 2025, 9:49:52 AM (6 days ago) Dec 15
to Bryan Sanctuary, Fred Diether, Bell inequalities and quantum foundations
Dear Bryan,

I have no interest in looking at more work from you.

Your paper claiming to resolve BI was fatally flawed and your response when people showed you the mistake and corrected it was a disgrace. And discredited you. You know where the mistake is. Acknowledge that, before asking anyone to look at subsequent work.

Austin Fearnley

unread,
Dec 15, 2025, 10:28:01 AM (6 days ago) Dec 15
to Bell inequalities and quantum foundations
Hi Bryan

Don't send the computer program as I only nowadays use Visual Basic.  Also, although years ago I familiarised myself with Cl(3,1) algebra, I am not familiar with Cl(2,2) algebra which is very different.

You could display for me the various (two) tables of results that you have.  You could simply send p++ for each of the two 2x2 results tables.  No need to look at individual pairs at this stage.  Also give n1 and n2 and confirm that n1 and n2 add up to the total number of pairs in the experiment.

Thanks

Fred Diether

unread,
Dec 15, 2025, 10:46:54 AM (6 days ago) Dec 15
to Bell inequalities and quantum foundations
Since you didn't pay attention to what I said, it is incumbent upon you to prove that your result for r_0 is correct.  It is pure junk just like your papers.

Local Quantum Mechanical Prediction of the Singlet State Using Geometric Algebra
Bell was wrong big time.
On Sunday, December 14, 2025 at 10:52:51 PM UTC-8 bryancs...@gmail.com wrote:
Dear Mark,

Thanks for the explanation about my arithmetic. . 

1. I calculated Fred's r_0 and I get a scalar before averaging so 
54066396-5542-49b8-a4ac-c856d2dd8942.png
It is incumbent upon Fred to prove this wrong,  If I am correct, then Fred's paper is wrong.. Also Fred's simulation violates the rules and his function  predetermines the outcome.  He does not use binary pairs.  

snip

Bryan Sanctuary

unread,
Dec 15, 2025, 11:26:09 AM (6 days ago) Dec 15
to Fred Diether, Bell inequalities and quantum foundations
Fred,

You should find my error, which to me seems the best approach.  It is a well defined vector algebra problem. Otherwise, I will write a rebuttal to your editor, and you can reply in the press.  Either way, you are going to have to do the calculation, since your whole paper rests on r_0 \to zero. Why are you reluctant to do the calculation?  IF you want, I can give you some guidance.  Let me know,
 
Bryan

--
You received this message because you are subscribed to the Google Groups "Bell inequalities and quantum foundations" group.
To unsubscribe from this group and stop receiving emails from it, send an email to Bell_quantum_found...@googlegroups.com.

Bryan Sanctuary

unread,
Dec 15, 2025, 11:38:27 AM (6 days ago) Dec 15
to Austin Fearnley, Bell inequalities and quantum foundations
Hi Austin,

Thanks for that, and I want to help, but I have not got the time, nor patience, right now, to spend a day or so generating the tables.  It is also unlikely they will change anyone's views.  However, you will likely find your answer if you read the algorithm in section 3.3 of the attached.  It is only one page that shows how I program vector-vector correlation and bivector-bivector correlation.  

Hope this helps

Bryan

quantumrep-06-00026.pdf

Fred Diether

unread,
Dec 15, 2025, 12:03:16 PM (6 days ago) Dec 15
to Bell inequalities and quantum foundations
I didn't think you would be able to prove your result for r_0 because I know it is pure junk just like your junk papers.  Maybe some day you will pay attention.

Local Quantum Mechanical Prediction of the Singlet State Using Geometric Algebra
Bell was wrong big time.

Austin Fearnley

unread,
Dec 15, 2025, 12:34:56 PM (6 days ago) Dec 15
to Bell inequalities and quantum foundations
Hi Bryan

No, a reference to section 3.3 does not help.

We do not seem to be on the same page as to the way you should conduct the simulation.  Like Mark I think you are just not accepting the issue.  Not surprising since we are all or mostly all quantum crackpots in Richard's view and we all think we are correct and are somehow blind to our own faulty thinking.  There is no need for Alice and Bob to go into technical details of the hidden variables when performing the correlation calculation.  They simply have a 2xn table of measurements and a very straightforward calculation of the correlation.  Your simulation needs to provide the 2xn table.  If you cannot do that then your use of your method is incomplete.  You can of course provide two separate and independent 2xn1 and 2xn2 tables and a correlation can very simply be calculated from that.  But you will not accept the proper method of combining the two tables of raw data into one table.

Bryan Sanctuary

unread,
Dec 15, 2025, 12:46:31 PM (6 days ago) Dec 15
to Austin Fearnley, Bell inequalities and quantum foundations
Dear Austin,

Sorry it is not enough, but it is incumbent on you to find how my program works, and I do not want to do your work. Your conclusions are speculative without you knowing my program, so you should not dismiss it.  All I might do is give you the algorithm I used in the program, which I have somewhere.

  Right now I am visiting family in Suisse for a month, and I would prefer to go skiing. So all I can do is help Fred with his calculation, but I will try to find the algorithm.

So Merry Christmas or Happy Holidays to all,

Bryan



 

 

Austin Fearnley

unread,
Dec 15, 2025, 1:43:24 PM (6 days ago) Dec 15
to Bell inequalities and quantum foundations
Hi Bryan

I agree that it is up to me or others to decide our opinions of your model.  However, it is not worth it for me to spend more time on your model when you are incorrectly calculating a correlation coefficient.  You are ignoring what a number of statisticians have said about how to calculate a correlation coeffficient.  

Here is an example of adding correlations:  (this is a re-post from some months ago)
Take a coincidence bin with N1=200 readings in it.
Say N++ = 75, which is the count of paired observations by Alice and Bob.  So p++ = 75/200.
N+-, N-+ and N-- could easily be calculated as there is only one degree of freedom in the bin, though it is unnecessary as there is a simple formula to calculate the correlation for the bin:
correl = 4 * p++ - 1.
So for this bin, correl = 0.5.

Take a second bin with N++ = 50 with N2 = 200 and p++ = 50/200.  Correl =  0.0.

Add the numbers in bins 1 and 2 to give N++ = 125 with N3 = N1 + N2 = 400 and p++ = 125/400 for a combined bin 3.  where bin 3 = bin 1 + bin 2 and has correl = 0.25.

Bin 1 has the higher correlation, 0.5, which is dragged down when accompanied by the zero correlation in bin 2, to give an intermediate correlation of 0.25 for the combined bin.  I would never add the correlations 0.5 and 0.0 (that is, 0.5) to get the correlation of 0.5 for bin 3.  That is why correl = -0.375 for your method.

I hope this example helps.  The calculation of the correlation on measurements or observables has nothing to do with your particular model.

Fred Diether

unread,
Dec 17, 2025, 12:13:36 PM (4 days ago) Dec 17
to Bell inequalities and quantum foundations
No worries Austin, because BS's work is a pile of junk and weird fantasies.  Looks like we are going to have to publish papers demonstrating how he is a completely clueless jerk.  It will be really easy.  LOL!

Bryan Sanctuary

unread,
Dec 17, 2025, 12:31:51 PM (4 days ago) Dec 17
to Austin Fearnley, Bell inequalities and quantum foundations
Dear Austin,

In fact I agree with you that my expression in that simulation  paper is not convincing the way I tried to explain the two sources of correlation. I will, soon, write a n erratum, but that section does not detract from the bigger picture.  I recently gave for the first time the geometric basis for the FSC; I also found the classical origin of spin. And I hope my newest paper that describes the ZBW shows the quaternion approach is viable. It is a new approach and I know it is hard to accept new ideas that fly in the face of present approaches. 

So our ideas evolve, and I ask you to accept that since writing that, I got it clearer.  In a  nutshell please consider this statement.

Bivector spin has two sectors. That is duality on the SAME particle. That is, sometimes the correlation is from the vector sector (the body fixed frame of the bivetor).  The two sectors cannot appear simultaneously, (like position and momentum) and I cannot average two dual spaces on the same particle.  It is one or the other, and does not follow the Bell statistics which are from one sector.  It is that simple.  Duality of spin.

I hope that this helps that we have something new and the violation of BI is evidence of duality.

I appreciate the time you have taken and I respect your position.  

Regards,

Bryan

Reply all
Reply to author
Forward
0 new messages