Dear all,
About the Sanctuary-Gill debate that I suggested.
Richard Gill, the faint-hearted gatekeeper of Bell, will not defend his ideas because he knows he will lose.
Richard did not like me calling a spade black, with my characterization of him:
“He is prone to discrediting and dismissing Bell dissenters without substance, rather than engaging”.
Gill shouted out: “This is libellous and untrue”. But read what he said recently, and this is one of many:
“I suspect that there could be some gaps (or loss of memory) in Bryan’s basic math education as well as in Fred’s - cf. the confusion about whether you can add correlations or take convex combinations of them. Go back to basics! The textbooks you learnt university level exact sciences math from.”
He, and other gatekeepers, say things like I failed probability 101, I lack integrity and honesty. I cannot do HS math, my mistakes are obvious, etc. The list is long and my archivist is extracting it all and having Chat GPT analyse them to reveal Gill’s academic dishonesty.
But my case is just one of many. One of those he railed against, W. Philiipp, had a small brain tumor, and later died. Gill did not like his work and allegedly suggested there was a correlation. Gill is aggressive in his condemnation of anyone who opposes Bell, for example Joy’s work, which is wrong, but Gill went overboard. Then there are Han Geurds, Marian Kupczynski, and Karl Hess, copied here, eminent scientists that have found credible errors in Bell. I suspect they all think Gill is academically dishonest, often discrediting them by innuendo and out-of-scope mathematics to obfuscate.
This is Gill consistent pattern with me and my ideas (see his quote above for example):
· Questions my credentials
· States I do not understand Bell
· Deflects to his own work as the true answer
· Condescends to me by pandering
· Makes disingenuous remarks to “improve my work”.
He uses his credibility to assert I have no credibility, not just to me, but behind my back. It works. I have no defense, except to have a debate on a level playing field. But Gill is too scared. He will not defend his ideas, cannot show mine wrong, and that makes him an academic coward. He can talk behind my back, but he cannot face me.
My goals are academic and fundamental. I prove Bell’s theorem is wrong, and Gill dismisses it. Gill and I should debate or he must capitulate.
Bell’s theorem assumes a single Kolmogorov probability space. I found a second, corresponding to vector and bivector degrees of freedom. This disproves Bell’s theorem.
So, Richard the Faintheart, which is it? Face and defend your life long positive stance, or be like a poltergeist behind the tulips?
I personally think the latter.
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.
To view this discussion visit https://groups.google.com/d/msgid/Bell_quantum_foundations/CALLw9YzyOUOhsgBf177rmOAsVB6tmNcaJU1074RSZX0qAokVXg%40mail.gmail.com.
--
Dear Bryan,
I am not engaging with personal accusations, character assessments, or speculation about motives. That line of discussion is not productive and does not advance your claims.
What is relevant — and remains unresolved — is the set of specific technical gaps and category errors that have been identified across your papers. These concerns are all listed below (and in atached PDF) with the help of chatGPT and thus reproducible by you; nothing here relies on authority or interpretation.
Until these items are addressed explicitly and point-by-point, further claims about having disproven Bell’s theorem, or about paradigm shifts, are premature. Continuing to escalate rhetorically without resolving these foundations only obscures the issues and makes constructive engagement increasingly difficult.
If you wish to proceed, the path forward is straightforward: take the identified points one at a time and show precisely where each is resolved in the mathematics or referenced in the academic literature. Otherwise, I see no value in continuing the exchange.
preprints.org/manuscript/202312.1277/v1
| Claim / Assumption | Status | Gap / Failure |
|---|---|---|
| Spin is a classical bivector with internal structure | Mathematically coherent | No issue at this stage; ontology clearly defined |
| Bivectors possess an intrinsic geometric “quantum limit” | Heuristic | “Quantum limit” introduced descriptively; not yet defined as a mathematical constraint or theorem |
| Parity separation (even/odd) distinguishes matter/force | Heuristic | Parity defined geometrically but not yet tied to invariant classification or conserved quantity |
| Fermions are emergent polarized blades | Ontological claim | No inconsistency, but no derivation or uniqueness shown |
| Probability implicitly geometric rather than Kolmogorov | Unstated assumption | Becomes critical later; not addressed here |
| Claim / Assumption | Status | Gap / Failure |
|---|---|---|
| Double-helix structure defines the quantum limit | Heuristic but structured | No proof that helix topology enforces discrete measurement outcomes |
| SU(2) symmetry arises geometrically from Cl(2,2) | Partially justified | Symmetry shown algebraically, but no proof of equivalence to SU(2) representations acting on observables |
| Spin is not quantized | Internally consistent | Creates tension with later use of quantum-correlated statistics |
| Measurement interpreted as geometric projection | Contextual | Projection depends on measurement setting → later violates Bell factorization |
Mathematics 12(13), 1962
| Claim / Assumption | Status | Gap / Failure |
|---|---|---|
| Use of convex geometry to model spin states | Valid framework | Acceptable mathematically |
| “Between dual convex sets, you must add (the Minkowski sum)” | Incorrect / heuristic | No theorem supports necessity of Minkowski sum between dual sets; duality does not imply additive closure |
| Minkowski sum preserves physical realism | Unsupported | Addition of state spaces alters probability structure without justification |
| Geometric averaging replaces probabilistic expectation | Out of scope for Bell | Later conflated with probabilistic averages |
Entropy 6(3), 26
| Claim / Assumption | Status | Gap / Failure |
|---|---|---|
| Bell assumes scalar spin ontology | Incorrect | Bell assumes scalar outcomes, not ontology |
| Bivector hidden variables invalidate Bell | Category error | Violating assumptions ≠ disproof |
| Correlations computed algebraically replace Bell correlations | Invalid substitution | Algebra-valued “averages” are not expectations in Kolmogorov probability |
| Single global bivector preserves locality | Incorrect | Shared algebraic object breaks Bell factorization |
| Contextual projections are local | Incorrect | Bell locality is statistical, not spacetime |
Entropy 6(3), 28
| Claim / Assumption | Status | Gap / Failure |
|---|---|---|
| Bell’s theorem is “not about reality” | Philosophical, not mathematical | Bell’s theorem is about probability models |
| Bell fails due to noncommutativity | Incorrect | Bell never assumes commutativity of hidden variables |
| Bivector algebra restores realism | Ontological claim | Bell realism is statistical, not metaphysical |
| Bell inequalities violated legitimately | Trivially true | Any non-Kolmogorov model violates them by construction |
Axioms 14(9), 668
| Claim / Assumption | Status | Gap / Failure |
|---|---|---|
| Paper is classical with a quantum limit | Correct | Scope clarification accepted |
| Quantum limit separates parity sectors | Geometrically coherent | Still not formalized as invariant or theorem |
| Parity explains matter/force distinction | Heuristic | No classification proof or necessity shown |
| SU(2) geometry fully explained | Partially justified | No demonstration that it constrains measurement statistics |
| Classical bivectors are fundamental | Ontological claim | Consistent internally |
| Observables fully addressed | Disputed | Addressed geometrically, but probabilistic extraction remains contextual |
These appear once and propagate forward:
Category Error: Ontology vs probability
Replacement of expectation values with algebraic averages
Contextual projection treated as outcome function
Invalid necessity claim (Minkowski sum of dual convex sets)
Equating violation of assumptions with refutation
There is no internal algebraic inconsistency in the bivector mechanics itself.
There are repeated instances of:
heuristic necessity claims,
category mistakes between geometry and probability,
and logical misinterpretation of Bell’s theorem.
A coherent classical geometric model in which Bell’s assumptions do not apply.
That Bell’s theorem is false, inconsistent, or mathematically incomplete.
It is my feeling that many discussions have been made on Bryan’s model involving “geometric blades”, including ChatGPT analyses. In my view, the discussions have been quite complete and, fundamentally, it shows that Bryan produced a geometric or mathematical model with extensive implications. As already pointed out by others (e.g. Anton), the fragility of Bryan’s model is reflected by the lack of correspondence with “physical reality” (in my view, the “reality” of a model demands support of the model with outcomes from repeated measurements).
Just to emphasize some comments I made in this thread:
In Bryan’s model, “In isotropy, the electron has mass only. It has no charge, no helicity”.
It is “standard” knowledge that charge is invariant, but Bryan’s model disagrees with this. There are many experiments on the invariance of the electron charge. For example, the old Millikan oil-drop experiment demonstrated charges on the oil droplets always in integer multiples of the electron charge. One may argue that oil droplets or the geometry involved in the experiment violates Bryan’s assumption of “isotropy”. One should understand what is lack of isotropy according to Bryan, trying to understand his whole idea.
In practice, isotropy may be seen as a relative concept as well: two electrons in free space separated by kilometers away may be seen as completely separated, such that each electron could be assumed in an isotropic condition?
What kind of interaction should not exist to define “isotropy”?
An experiment that could falsify a model positing non-invariant electron charge (e.g., charge varying under Lorentz transformations or in different reference frames) is the precision measurement of atomic neutrality. In heavy atoms, electrons orbit at relativistic speeds while protons in the nucleus do not, yet atoms remain electrically neutral to an extremely high degree of precision (better than 1 part in 10^21). If electron charge were not Lorentz invariant, the effective charge of these fast-moving electrons would differ from that of the slower protons, leading to a net charge on the atom that would be detectable, but that is not observed. This supports the standard invariant charge model and contradicts any non-invariant alternative.
More direct tests include hydrogen spectroscopy experiments comparing the 2S-1S transition frequencies in boosted frames (e.g., using high-precision optical clocks on moving platforms or in Earth's orbital motion), which probe electron boost invariance and have confirmed no violations to parts in 10^17. Any deviation would indicate charge non-invariance, but none has been found.
How could one create an “isotropic” situation to verify that the electron has only mass? How do you see the transition from having just mass to also having charge?
Geraldo
To view this discussion visit https://groups.google.com/d/msgid/Bell_quantum_foundations/em176683d1-1cde-4cd6-b7f4-e6ac775570b1%409efdb2c4.com.
To view this discussion visit https://groups.google.com/d/msgid/Bell_quantum_foundations/em176683d1-1cde-4cd6-b7f4-e6ac775570b1%409efdb2c4.com.