Pith. sign in

REVIEW 3 major objections 5 minor 44 references

Luis Santal\'o and classical field theory

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read This historical review argues that Santaló's classification theorem completely determines the possible generalized Ricci tensors of Einstein's asymmetric unified field theory: they form an eight-parameter family, and the field equations…

desk verdict A clear historical review of Santaló's classification of Ricci-type tensors in Einstein's asymmetric unified field theory, with the central theorem cited rather than proved. read the letter →

arxiv 1908.02881 v2 pith:WU6MBZUH submitted 2019-08-08 physics.hist-ph gr-qcmath-phmath.MP

classification physics.hist-phgr-qcmath-phmath.MP MSC 53B0583C05 PACS 01.65.+g04.20.-q04.50.-h
keywords unifiedfieldtheoryasymmetricaffineconnectionRicci-typetensorclassificationtorsionEinstein-SchrödingervariationalprinciplehistoryofgeneralrelativitySantalótheorem
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper revisits a mathematical result that settled a long-standing ambiguity in Einstein's classical unified field theory. It argues that Luis Santaló classified every rank-2 tensor that can be constructed from an asymmetric affine connection and its first derivatives, subject to being at most quadratic in the connection, and found exactly eight independent candidates. The most general Ricci-type tensor is therefore a linear combination $R^*_{\mu\nu} = \sum_{i=1}^8 c_i L^{(i)}_{\mu\nu}$ with eight arbitrary coefficients. Santaló then showed that the Euler-Lagrange equations coming from such a general tensor are generically incompatible, meaning no variational principle of this type can uniquely select the asymmetric field equations. The paper presents that theorem, the geometry behind it, and its place in the history of unified field theory.

What carries the argument

The load-bearing object is the eight-element basis $L^{(1)}_{\mu\nu}, \ldots, L^{(8)}_{\mu\nu}$ of Ricci-type tensors in a space with asymmetric connection. The list includes the ordinary Ricci tensor $R_{\mu\nu}$, the alternative contraction $\Sigma_{\mu\nu} = \partial_\nu S^\lambda_{\lambda\mu} - \partial_\mu S^\lambda_{\lambda\nu}$, torsion divergences $\nabla_\lambda T^\lambda_{\mu\nu}$, gradients of the torsion trace $\Gamma_\mu$, and quadratic contractions of torsion. The theorem that this list is complete under the stated restrictions is what converts the proliferation of candidate field equations into a finite eight-parameter family. The generic incompatibility of the resulting Euler-Lagrange equations is then the mathematical mechanism that establishes the paper's central historical lesson about the unavoidable arbitrariness of Einstein's asymmetric unified field theory.

What would settle it

Inspect the algebra of rank-2 tensors built from $\Gamma^\alpha_{\mu\nu}$ and its first derivatives in a torsionful affine space: finding any such tensor that is not a linear combination of the eight $L^{(i)}$ would refute the classification. Alternatively, choose numerical coefficients $c_i$ and exhibit a metric and connection satisfying both $R^*_{\mu\nu}=0$ and $K^\alpha_{\mu\nu}=0$, which would show that the incompatibility claim is not generic.

Watch

Extended reading notes

Core claim

The central claim is that in an affine space with connection $\Gamma^\alpha_{\mu\nu}$ and torsion $T^\alpha_{\mu\nu}$, the only rank-2 tensors that depend only on $\Gamma$ and $\partial\Gamma$ and are at most quadratic in $\Gamma$ are the eight tensors $L^{(1)}_{\mu\nu}$ through $L^{(8)}_{\mu\nu}$ listed in Eq. (9). These include the Ricci tensor, an alternative contraction of the Riemann tensor, covariant divergences of torsion, gradients of the torsion trace, and quadratic torsion terms. Santaló forms the most general action $S = \frac{1}{4\pi} \int d^n x \sqrt{-g} g^{\mu\nu} R^*_{\mu\nu}$, whose variation yields the field equations $R^*_{\mu\nu}=0$ plus a rank-3 condition $K^\alpha_{\mu\nu}=0$. Santaló proved that these conditions are generically incompatible: while $R^*_{\mu\nu}=0$ holds only under restrictive conditions, the additional rank-3 equation imposes constraints that cannot be satisfied together. The upshot is that the geometry itself does not single out one asymmetric field theory; different choices of the eight coefficients $c_i$ produce different theories, including those of Einstein, Tonnelat, and Winogradzki, and the paper's historical claim is that this arbitrariness is structural, not a failure of imagination.

Load-bearing premise

The load-bearing premise is that Santaló's enumeration of eight tensors is genuinely complete under the stated restrictions, a fact the paper quotes from his earlier papers rather than proves; if a legitimate rank-2 tensor built only from the connection and its first derivatives and at most quadratic in the connection is missing, then the claim that $R^*_{\mu\nu}$ is the most general Ricci-type tensor, and with it the conclusion of unavoidable arbitrariness, does not follow.

Editorial extensions

If this is right

  • If the classification is correct, any asymmetric unified field theory whose Lagrangian is a function of the metric, the connection, and the first derivatives of the connection, at most quadratic in the connection, is a special case of the eight-parameter family $R^*_{\mu\nu} = \sum_i c_i L^{(i)}_{\mu\nu}$.
  • The generic incompatibility of $R^*_{\mu\nu}=0$ and $K^\alpha_{\mu\nu}=0$ means that no Palatini-type variational principle built from these ingredients can yield a unique set of field equations.
  • The theories of Einstein, Tonnelat, and Winogradzki all correspond to particular choices of the coefficients $c_i$, so the historical proposals are organized as special cases of a single classification.
  • The number of independent tensors is the same in any dimension $n$, so the classification is dimension-independent.
  • Within the classical unified field program, the theorem closes the question of uniqueness: the remaining ambiguities are not a technical gap but an unavoidable feature of the asymmetric geometric setup.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct computer-algebra check of Santaló's list, treating the space of rank-2 tensors generated by $\Gamma$ and $\partial\Gamma$ under the allowed identities of the torsionful affine geometry, could confirm the completeness of the eight tensors or reveal a counterexample; the paper does not perform such a check.
  • The same enumeration strategy could be applied to modern metric-affine and modified-gravity theories, where the analogous question is whether the space of Ricci-type or energy-momentum-like tensors has a finite basis under similar restrictions.
  • If the incompatibility is generic, it offers a structural explanation for why the classical unified field program never converged on a single theory: the multiplicity of possible equations is a theorem rather than an accident of the choices made by Einstein and his contemporaries.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper is a historical review of Luis Santaló's contributions to classical unified field theory, focusing on his classification of rank-2 tensors in the non-symmetric affine geometry underlying Einstein's unified theory. Section 2 summarizes the Einstein-Schrödinger formulation, including the definition of torsion, the two independent Ricci contractions, and the derivation of the field equations from an action. Section 3 presents Santaló's classification theorem, Eq. (9), stating that the most general Ricci-type tensor constructed from the connection and its first derivatives, at most quadratic in the connection, is a linear combination of eight basis tensors L^(1) through L^(8). The paper further claims that the Euler-Lagrange equations derived from the corresponding action are generically incompatible, and concludes that the asymmetric field theory inevitably retains an irreducible arbitrariness. The manuscript is a conference transcript and does not claim to prove the classification theorem, which is attributed to Santaló's original works.

Significance. If the mathematical statements are accurate, the paper serves a useful historical purpose by documenting Santaló's often-overlooked work in unified field theory and placing it in context with the Einstein-Schrödinger program. The exposition of the Einstein-Schrödinger theory in Section 2 is clear and internally consistent, and the paper correctly attributes the central classification theorem to Santaló rather than claiming originality. However, because the classification theorem and the incompatibility result are not proved in the manuscript, the paper's value is primarily historical; its utility depends on the correctness and faithful transcription of Santaló's original results. The paper would be strengthened by making explicit which statements are direct quotations from the historical sources and by providing the necessary cross-checks to ensure internal consistency.

major comments (3)
  1. [Section 3, Eq. (9)] The classification theorem is stated in the text as a fact but is not proved in the manuscript; it is only cited to Santaló [16]. Since the later conclusion of generic incompatibility relies entirely on the exhaustiveness of the list of eight tensors, the authors should explicitly state that the theorem is quoted from the original source and not proved here. A more precise citation, such as a theorem number or page reference, would also help readers verify the attribution.
  2. [Section 3, Eq. (12) and surrounding text] The claim that conditions (12) are necessary and sufficient for the vanishing of R*_{μν} is not substantiated and appears questionable: the tensor L^(3) = ∇_μ T^μ_{νη} is not obviously forced to vanish under the stated conditions, since neither Γ_μ=0 nor T^ξ_{νρ}T^ρ_{ηξ}=0 implies that the divergence of the torsion vanishes. If this statement is a result of Santaló, the authors should provide a proof or a direct quotation; if it is an independent claim, it requires a derivation.
  3. [Section 2 vs. Section 3] The second Ricci tensor R^ρ_{ρμν} introduced in Section 2 is a rank-2 tensor that depends only on the connection and its first derivatives and is quadratic in the connection, so it satisfies the conditions of Santaló's theorem. For the classification in Eq. (9) to be complete, this tensor must be a linear combination of the eight tensors L^(1) through L^(8). The authors should either identify the combination explicitly or explain why the second Ricci tensor is not a counterexample to completeness.
minor comments (5)
  1. [Equation (9)] The term L^(8)_{νη} = Γ^μ T^μ_{νη} has an undefined index contraction if Γ^μ is treated as a covector, as it is elsewhere in the paper; it should read Γ_μ T^μ_{νη}.
  2. [Section 2, notation] The definition Γ^μ ≡ T^ν_{μν} uses an upper index on Γ while later expressions (e.g., Eq. (4) with δ^μ_ρ Γ_λ) treat Γ_λ as a covector. The index positioning should be made consistent throughout.
  3. [After Eq. (9)] The phrase "two derivatives of the metric" is misleading because the tensors in Eq. (9) depend only on the connection, not on the metric; the sentence should refer to derivatives of the connection.
  4. [References] Some references contain typos: [36] "Kaufan" should be "Kaufman", "struture" should be "structure", and "thory" should be "theory".
  5. [Equation (2)] The notation in Eq. (2) is introduced without explanation; a brief comment that the second equation is the cyclic identity for the Ricci tensor would aid readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is a historical review whose central classification is attributed to external sources.

full rationale

This paper does not claim to derive Santaló's classification from first principles; it explicitly presents the eight tensors of Eq. (9) as Santaló's theorem and cites his papers [15, 16] and extensions [17, 18]. The subsequent construction R*_mu_nu = sum_i c_i L^(i)_mu_nu is presented as a general linear combination, with Einstein, Tonnelat, and Winogradzki cases given as particular coefficient choices; this is a unification statement, not a fitted parameter renamed as a prediction. No load-bearing step is defined in terms of its own conclusion, and no substantive claim depends on a self-citation by the present authors. The skeptical concern that the completeness of the classification is cited rather than proved in this paper is a legitimate verification or correctness risk, but it is not circularity: invoking an external theorem is independent support under the stated rules. The paper also derives the Schrödinger form of the field equations from the action (3) in a self-contained way, but that derivation is not used to justify the classification. Accordingly, the circularity score is 0.

Assumptions & free parameters 1 free parameters · 2 assumptions · 0 invented entities

The central claim rests on Santaló's completeness theorem, imported from [16] and [17,18], and on standard differential-geometry background. The c_i are unconstrained coefficients parameterizing the family, but they are not fitted to data. No new entities are introduced.

free parameters (1)
  • Coefficients c_i (i=1,...,8) in R* = Not fitted; e.g., c_i = δ_{1i} for Einstein-Schrödinger, with other values for Einstein, Tonnelat, and Winogradzki…
    They parameterize the family of Ricci-type tensors in Eq. (10). The paper uses them only to show known proposals are special cases; it does not fit them to data, but they are free constants of the generalized action.
assumptions (2)
  • domain assumption The only rank-2 tensors satisfying the stated conditions in Section 3 are the eight tensors in Eq. (9).
    This is the load-bearing mathematical premise of the paper's account of Santaló's contribution. It is cited to references [16-18] and not proved here, so the present paper's central claim inherits its correctness.
  • standard math In the symmetric case, the Einstein tensor is the unique rank-2 conserved tensor satisfying the stated derivative conditions (Cartan's theorem).
    Used in Section 3 to contrast the symmetric theory with the asymmetric one; cited to reference [32].

how reviews work

0 comments
Cite this review

Pith. "Pith review of Luis Santal\'o and classical field theory." pith.science (2026). https://pith.science/paper/WU6MBZUH

@misc{pith2026190802881,
  author       = {Pith},
  title        = {Pith review of: Luis Santal\'o and classical field theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WU6MBZUH}},
  note         = {Machine review of arXiv:1908.02881}
}
read the original abstract

Considered one of the founding fathers of integral geometry, Luis Santal\'o has contributed to various areas of mathematics. His work has applications in number theory, in the theory of differential equations, in stochastic geometry, in functional analysis, and also in theoretical physics. Between the 1950's and the 1970's, he wrote a series of papers on general relativity and on the attempts at generalizing Einstein's theory to formulate a unified field theory. His main contribution in this subject was to provide a classification theorem for the plethora of tensors that were populating Einstein's generalized theory. This paper, which is the transcript of the conferences delivered by the authors in occasion of the celebration of the 100th anniversary of Luis Santal\'o's birth, revisits his work on theoretical physics.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 39 canonical work pages

  1. [16]

    On Einstein’s unified field theory

    Santaló L., “On Einstein’s unified field theory”, in Pros pects in Geometry and Relativity (1966) 343

  2. [1]

    Un invariante afin para los cuerpos convexos del espacio des n dimensiones

    Santaló L., “Un invariante afin para los cuerpos convexos del espacio des n dimensiones”, Portugaliae Math. 8 (1949) 155

  3. [2]

    Review: Luis A. Santaló, Integral geometry an d geometric probability

    Chern S., “Review: Luis A. Santaló, Integral geometry an d geometric probability”, Bull. Amer. Math. Soc. 83 (6) (1977) 1289

  4. [3]

    Holographic Subregion Complexity from Kinematic Space

    Abt R., Erdmenger J., Gerbershagen M., Melby-Thompson C ., Northe C., “Holographic Subregion Complexity from Kinematic Space”, JHEP 1901 (201 9) 12

  5. [4]

    The dual of non-extre mal area: differential entropy in higher dimensions

    Balasubramanian V., Rabideau C., “The dual of non-extre mal area: differential entropy in higher dimensions”, [arXiv:1812.06985]

  6. [5]

    Integral Geometry and Holography

    Czech B., Lamprou L., McCandlish S., Sully J., “Integral Geometry and Holography”, JHEP 1510 (2015) 175

  7. [6]

    Geometric tomography

    Gardner R., “Geometric tomography”, Cambridge Univers ity Press (1995). 7

  8. [7]

    Girbau J., Discurs llegit en la cerimònia del doctorat ho noris causa de Luis Antoni Santaló, Universitat Autònoma de Barcelona (1986)

Show all 44 references
  1. [8]

    Luís Antonio Santaló : Selected works

    Naveira A. and Reventós A. (Eds.), “Luís Antonio Santaló : Selected works”, Springer Collected Works in Mathematics (2009)

  2. [9]

    La última teoría del campo unificado de Einst ein

    Santaló L., “La última teoría del campo unificado de Einst ein”, Ciencia e Investigación 9 (1953) 300

  3. [10]

    El problema de la unificación de los campos

    Santaló L., “El problema de la unificación de los campos” , Mundo Atómico 4 (1953) 11

  4. [11]

    Sobre algunos tensores análogos al de curv atura en espacios con una conex- ión afín no-simétrica

    Santaló L., “Sobre algunos tensores análogos al de curv atura en espacios con una conex- ión afín no-simétrica”, Revista de la Universidad Nacional de Tucumán 10 (1954) 19

  5. [12]

    El legado einsteiniano

    Santaló L., “El legado einsteiniano”, Ciencia e Invest igación 7 (1955) 289

  6. [13]

    Influencia de Einstein en el campo matemáti co

    Santaló L., “Influencia de Einstein en el campo matemáti co”, Ciencia e Investigación 11 (1955) 304

  7. [14]

    Sobre las ecuaciones del campo unificado de Einstein

    Santaló L., “Sobre las ecuaciones del campo unificado de Einstein”, Revista de la Uni- versidad Nacional de Tucumán 12 (1959) 31

  8. [15]

    Sobre las ecuaciones del campo unificado de Einstein

    Santaló L., “Sobre las ecuaciones del campo unificado de Einstein”, Revista de la Unión Matemática Argentina 19 (1960) 196

  9. [17]

    Sobre algunas teorías asimétricas del cam po unificado

    Santaló L., “Sobre algunas teorías asimétricas del cam po unificado”, Revista de la Real Academia de Ciencias Exactas de Madrid 66 (1972) 395

  10. [18]

    Unified field theory of Einstein’s type dedu ced from variational principle

    Santaló L., “Unified field theory of Einstein’s type dedu ced from variational principle”, Tensor 25 (1972) 383

  11. [19]

    The meaning of relativity

    Einstein A., “The meaning of relativity”, 4 thedition, Princeton University Press (1953) 321

  12. [20]

    On the History of Unified Field Theories. Pa rt II

    Goenner H., “On the History of Unified Field Theories. Pa rt II”, Living Rev. Rel. 17 (2014) 5

  13. [21]

    A generalized theory of gravitation

    Einstein A., “A generalized theory of gravitation”, Re v. Mod. Phys. 20 (1948) 35

  14. [22]

    The Bianchi identities in the generalize d theory of gravitation

    Einstein A., “The Bianchi identities in the generalize d theory of gravitation”, Canadian J. of Math. 2 (1950) 120

  15. [23]

    A generalization of the relativistic the ory of gravitation

    Einstein A., “A generalization of the relativistic the ory of gravitation”, Ann. of Math. 46 (1945) 578

  16. [24]

    A generalization of the relati vistic theory of gravitation II

    Einstein A., Straus E., “A generalization of the relati vistic theory of gravitation II”, Ann. of Math. 47 (1946) 731. 8

  17. [25]

    Sur l’état actuel de la théori e générale de la gravitation

    Einstein A., Kaufman B., “Sur l’état actuel de la théori e générale de la gravitation” in Louis de Broglie, Physicien et Penseur, Volume in honor to Louis de Broglie, Paris (1952) 321

  18. [26]

    Algebraic properties of the fi eld theory of the asymmetric field

    Einstein A., Kaufman B., “Algebraic properties of the fi eld theory of the asymmetric field”, Ann. of Math. 59 (1954) 230

  19. [27]

    A new form of the general relat ivistic field equations

    Einstein A., Kaufman B., “A new form of the general relat ivistic field equations”, Ann. of Math. 62 (1955) 128

  20. [28]

    Integral Geometry on surfaces

    Santaló L., “Integral Geometry on surfaces”, Duke Math . J. 16 (1949) 361

  21. [29]

    On parallel hypersurfaces in the elliptic and hyperbolic n-dimensional space

    Santaló L., “On parallel hypersurfaces in the elliptic and hyperbolic n-dimensional space”, Proc. Amer. Math. Soc. 1 (1950) 325

  22. [30]

    Integral Geometry in general spaces

    Santaló L., “Integral Geometry in general spaces”, Pro ceedings of the International Congress of Mathematicians, Cambridge Mass. Amer. Math. So c. R. I.1 (1950) 482

  23. [31]

    Integral Geometry in projective and affine s paces

    Santaló L., “Integral Geometry in projective and affine s paces”, Ann. of Math. 51 (1950) 739

  24. [32]

    Sur les équations de la gravitation d’Einst ein

    Cartan E., “Sur les équations de la gravitation d’Einst ein”, J. Math. Pures et Appliqueés 1 (1922) 141

  25. [33]

    Generalizations of Einstein theory

    Schrödinger E., “Generalizations of Einstein theory” , in Space-Time Structure, Cam- bridge University Press (1950), 106

  26. [34]

    The Einstein generalized Riemannian ge ometry

    Eisenhart L., “The Einstein generalized Riemannian ge ometry”, Proceedings of the National Academy of Sciences of USA 50 (1963) 190

  27. [35]

    Geometry of Einstein’s Unified Field Theor y

    Hlavatý V., “Geometry of Einstein’s Unified Field Theor y”, Noordhoff, Groningen (1957)

  28. [36]

    Mathematical struture of th non-symmetric field thory

    Kaufan B., “Mathematical struture of th non-symmetric field thory”, Proceeding of the 50 anniversary conference on relativity, Bern 1950, Helvet ia Physica Atca IV (1956) 227

  29. [37]

    Théories relativistes de la gravita tion et l’electromagnétisme

    Lichnerowicz A., “Théories relativistes de la gravita tion et l’electromagnétisme”, Paris, Masson (1955)

  30. [38]

    La théorie du champ unifié d’Einstein et qu elquesuns de ses développe- ments

    Tonnelat M., “La théorie du champ unifié d’Einstein et qu elquesuns de ses développe- ments”, Paris, Gauthier-Villars (1955)

  31. [39]

    Le group relativiste de la théorie uni taire d’Einstein-Schrödinger

    Winogradzki J., “Le group relativiste de la théorie uni taire d’Einstein-Schrödinger”, J. Phys. Radium 16 (1955) 438

  32. [40]

    Non-Riemannian Geometry

    Eisenhart L., “Non-Riemannian Geometry”, Am. Math. So c. Coll. Pubns VIII (1927)

  33. [41]

    Generalized complex geometry

    Gualtieri M., “Generalized complex geometry”, Oxford University D. Phil thesis. [arXiv:math/0401221 [math.DG]]. 9

  34. [42]

    Superspace duality in low-energy superstr ings

    Siegel W., “Superspace duality in low-energy superstr ings”, Phys. Rev. D 48 (1993) 2826 [hep-th/9305073]

  35. [43]

    Two vierbein formalism for string inspired axionic gravity

    Siegel W., “Two vierbein formalism for string inspired axionic gravity”, Phys. Rev. D 47 (1993) 5453 [hep-th/9302036]

  36. [44]

    Double Field Theory

    Hull C., Zwiebach B., “Double Field Theory”, JHEP 0909 ( 2009) 099 [arXiv:0904.4664 [hep-th]]. 10

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.