Three Generations in E7

John C. Baez

REVIEW 2 major objections 6 minor 14 references

Three Generations in E7

T0 review · 2 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read E7 contains three copies of the Standard Model's one-generation fermion representation, right-handed neutrinos included.

desk verdict A transparent, largely sound rederivation of Nasmith's E7 three-generation decomposition, with a couple of unproved classification claims that should be cited or proved before publication. read the letter →

arxiv 2608.06271 v2 pith:F3JQIMD2 submitted 2026-08-06 math-ph math.MP

John C. Baez This is my paper · ORCID

classification math-phmath.MP MSC 17B2517B10

keywords exceptionalLiealgebraE7StandardModelthreegenerationsfermionrepresentationsroot-removalchainexteriorregularsubalgebragrandunifiedtheory

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

This paper proves a structural claim about the exceptional complex Lie algebra $\mathfrak{e}_7$: once the Standard Model gauge algebra $\mathfrak{g}_{\mathrm{SM}} = \mathfrak{sl}_3 \oplus \mathfrak{sl}_2 \oplus \mathbb{C}$ is embedded in $\mathfrak{e}_7$ in a specified 'good' way, the whole algebra splits into a subalgebra $\mathfrak{sl}_6^{\mathrm{SM}} \oplus \mathbb{C}\otimes P$ plus three 32-dimensional subspaces $V_1,V_2,V_3$. Each $V_k$ transforms under $\mathfrak{g}_{\mathrm{SM}}$, by the Lie bracket, exactly as one generation of Standard Model fermions and their antiparticles, including right-handed neutrinos. The reason a sympathetic reader would care is that the three generations are not put in by hand as a triple repetition; they are exhibited inside a single exceptional algebra. The paper is explicit that this is a mathematical pattern, not a theory of physics.

What carries the argument

The load-bearing object is a 'good' embedding of $\mathfrak{g}_{\mathrm{SM}}$ into $\mathfrak{e}_7$, a regular subalgebra obtained by the root-removal chain $\mathfrak{e}_7 \to \mathfrak{e}_6 \to \mathfrak{so}_{10} \to \mathfrak{sl}_5 \to \mathfrak{g}_{\mathrm{SM}}$. With a compatible Cartan subalgebra, the roots of the centralizer $\mathfrak{sl}_3^{\mathrm{gen}}$ span a 2-plane $P$ in the root space, and orthogonal projection onto $P$ yields a trichotomy (Lemma 2): every root of $\mathfrak{e}_7$ projects to $0$, to a weight $\pm w_k$, or to a root of the $A_2$ system. This partitions the 126 roots into $\Phi_0$ (30 roots, type $A_5$, giving $\mathfrak{sl}_6^{\mathrm{SM}}$) and three sets $\Phi_1,\Phi_2,\Phi_3$ of 30 roots each. Each $V_k$ is the span of the root spaces $\{ \pm\beta_k \} \sqcup \Phi_k$, making it a 32-dimensional module; as an $\mathfrak{sl}_6^{\mathrm{SM}}$-module it is $\Lambda^{\mathrm{even}}\mathbb{C}^6$, which restricts under $\mathfrak{sl}_5^{\mathrm{SM}}$ to $\Lambda\mathbb{C}^5$.

What would settle it

Run a concrete root-system computation: build the 126 roots of $\mathfrak{e}_7$, fix the generation plane $P$ from an $A_2$ root subsystem, form the spaces $V_k = \bigoplus_{r \in \{ \pm\beta_k \} \sqcup \Phi_k} (\mathfrak{e}_7)_r$, and compute the $\mathfrak{g}_{\mathrm{SM}}$-module structure of each $V_k$. If any $V_k$ is not isomorphic to $\Lambda\mathbb{C}^5$, so that its weights do not match Table 3, Theorem 13 fails.

Watch

Extended reading notes

Core claim

The central theorem, Theorem 13, is a direct-sum decomposition $$\mathfrak{e}_7 = \mathfrak{sl}$_6^{{\mathrm{SM}}$} \oplus (\mathbb{C}\otimes P) \oplus V_1 \oplus V_2 \oplus V_3,$$ where $\mathfrak{sl}_6^{\mathrm{SM}}$ is the centralizer of the generation subalgebra $\mathfrak{sl}_3^{\mathrm{gen}}$ and $P$ is the generation plane spanned by the roots of that subalgebra. Each $V_k$ is a 32-dimensional $\mathfrak{sl}_6^{\mathrm{SM}}$-submodule, and as a representation of $\mathfrak{g}_{\mathrm{SM}}$ it is the Standard Model representation on the exterior algebra $\Lambda\mathbb{C}^5$: one generation of fermions and their antiparticles, right-handed neutrinos included. The three generations are labelled by three roots $\beta_k$ of $\mathfrak{sl}_3^{\mathrm{gen}}$; the right-handed neutrino and its antiparticle live in the root spaces of $\pm\beta_k$, while the other 30 fermionic states of a generation live in the root spaces whose projection to the generation plane is $\pm w_k$.

Load-bearing premise

The whole construction depends on fixing a 'good' embedding of $\mathfrak{g}_{\mathrm{SM}}$ into $\mathfrak{e}_7$, one obtained from the root-removal chain, together with a compatible Cartan subalgebra and three roots $\beta_k$; if the embedding were not of this regular kind, the three 32-dimensional subspaces need not transform as Standard Model generations.

Editorial extensions

If this is right

The decomposition fills 96 of the 133 dimensions of $\mathfrak{e}_7$ with three linearly independent copies of the 32-dimensional Standard Model fermion representation; the remaining 37 dimensions form $\mathfrak{sl}_6^{\mathrm{SM}} \oplus (\mathbb{C}\otimes P)$, which contains $\mathfrak{g}_{\mathrm{SM}}$ itself.
Right-handed neutrinos and their antiparticles are included in each generation: they occupy the root spaces of $\pm\beta_k$, transform trivially under $\mathfrak{g}_{\mathrm{SM}}$, and can be extracted only after choosing a Cartan subalgebra and naming the roots $\beta_k$.
Without any Cartan choice, the algebra still exhibits a weaker, embedding-independent generation structure: under $\mathfrak{sl}_3^{\mathrm{gen}} \oplus \mathfrak{sl}_6^{\mathrm{SM}}$, $\mathfrak{e}_7$ decomposes as $(\mathbf{3}\otimes\mathbf{15}) \oplus (\mathbf{3}^*\otimes\mathbf{15}^*)$, corresponding to three generations of fermions excluding right-handed neutrinos.
The remaining 37 dimensions reproduce the familiar SU(5) grand-unified content: the $X$ and $Y$ leptoquark directions and a $\mathbf{5}\oplus\overline{\mathbf{5}}$ multiplet, together with three singlet directions.
The triplication is not an added ingredient but a consequence of the count of roots in the trichotomy: 30 roots in each $\Phi_k$, plus the two roots $\pm\beta_k$, give exactly 32 states per generation inside a single $\mathfrak{e}_7$.

Reading between the lines

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

If this pattern is robust, the number of generations would be a property of the Lie algebra rather than an input: one could test whether other exceptional algebras with $A_2$ subsystems produce a different number of 32-dimensional blocks under the same projection construction, making generation count a root-system invariant.
The paper demonstrates the three-generation decomposition only for one 'good' embedding; a natural extension is to classify embeddings of $\mathfrak{g}_{\mathrm{SM}}$ into $\mathfrak{e}_7$ up to automorphism and check whether the trichotomy and the three copies of $\Lambda\mathbb{C}^5$ survive for all of them, which would show how canonical the result is.
The right-handed neutrino sector is the part of the pattern most sensitive to convention: the six-dimensional space spanned by $\pm\beta_k$ cannot be separated from the rest of $\mathfrak{sl}_3^{\mathrm{gen}}$ without extra choices, so any physical use of this construction would need a mechanism that breaks the generation symmetry to name the three generations.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Here's my take. The paper is what it claims to be: a rederivation of Nasmith's result that E7 contains three copies of the Standard Model fermion representation, with cleaner proofs and a nice exterior-algebra framing. The genuinely new parts are the proof strategy that leans only on the embedding gSM ⊂ e7 and the explicit identification of each 32-dimensional subspace with ΛC5. The main decomposition theorem is clearly stated and the supporting root combinatorics (Lemma 2's trichotomy, the counting arguments) is solid and readable.

I want to be fair: the paper does not oversell. It credits Nasmith and Kugo-Yanagida, and it says plainly that no physics follows. That honesty helps. The expository improvement is real—the ΛC5 description in Table 3 makes the Standard Model representation much easier to see.

Now the soft spots. The load-bearing Lemma 9 has two unproved assertions. First, N_k^+ is said to have 15 distinct weights all of the same length; distinctness is claimed but not shown. Lemma 2 controls the projection of a single root onto the generation plane, but it says nothing about whether two different roots in Φ_k^+ can have the same projection. If two did, the weight space would be larger than 1-dimensional and the representation would not be Λ2 or Λ4. Second, the classification of 15-dimensional sl6 representations with these properties is invoked without proof or citation. The stress-test note is right that Theorem 12 and Theorem 13 inherit this dependence. Proposition 5 has the same pattern: the 32 distinct weights claim and the so12 spinor identification are asserted, not proved. These are standard facts, but a referee should ask the author to either prove them or give precise citations. A minor related issue is that the whole construction depends on choosing a 'good' embedding; the paper acknowledges that other embeddings may exist and be non-conjugate, but it doesn't discuss whether the result is invariant under all good embeddings. That is an assumption, not a theorem, and it should be stated more prominently.

None of this looks fatal to me. I did not find a circular argument; the decomposition is derived from the embedding, not used as an input. The self-aware note about the LLM draft is unusual, but the actual mathematics reads like it was checked by a human. I'd send this to a referee. With the classification claims filled in or cited, it would be a solid paper in math-ph.

Referee Report

2 major / 6 minor

Summary. The paper constructs, for a chosen 'good' embedding of the complexified Standard Model Lie algebra gSM = C ⊕ sl2 ⊕ sl3 into e7, a vector-space decomposition e7 = slSM6 ⊕ (C⊗P) ⊕ V1 ⊕ V2 ⊕ V3, where slSM6 is an sl6 subalgebra commuting with gSM, C⊗P is a 2-dimensional abelian complement in the Cartan, and each Vk is a 32-dimensional subspace on which gSM acts by the standard-model representation on ΛC5 (one generation including right-handed neutrinos and their antiparticles). The construction follows a chain gSM ⊂ sl5 ⊂ sl6 ⊂ e7 obtained by a root-removal procedure, and the three generations are associated with three sl2 subalgebras of a unique sl3 subalgebra centralizing gSM. The main technical steps are Lemma 9, which identifies a 30-dimensional root-space sum as Λ2C6 ⊕ Λ4C6 as an slSM6-module, Theorem 12, which upgrades this to ΛevenC6 and hence to ΛC5, and Theorem 13, which gives the final direct-sum decomposition. The paper emphasizes that as many results as possible are derived from the embedding alone and explicitly notes that the final three-generation labeling requires choices of a Cartan subalgebra and roots βk.

Significance. If the results are correct, the paper gives a clean, parameter-free, root-system derivation of Nasmith's observation that e7 contains three linearly independent copies of one Standard Model generation, and it connects that observation to the standard SU(5) exterior-algebra description of one generation. The main strengths are the systematic use of regular subalgebras and root projections, the explicit chain gSM ⊂ sl5 ⊂ sl6 ⊂ e7, the uniqueness arguments for slgen3, slSM6 and slSM5, and the self-contained statement of the final decomposition. The paper does not propose a physical theory, and it is honest about the dependence of the three-generation labeling on auxiliary choices. However, two classification claims in the proofs of Lemma 9 and Proposition 5 are asserted without proof or citation; they are standard and likely true, but they are load-bearing for the central claim and should be supported before the paper is accepted.

major comments (2)

[Section 9, Lemma 9] The proof asserts that N_k^+ has '15 distinct weights all of the same length' and that 'the only representations of sl6 with these properties are its fundamental representations on Λ2C6 and its dual Λ4C6.' The distinctness claim needs an explicit argument: it is true, because if two roots in Φ_k^+ have the same restriction to the Cartan of slSM6, their difference lies in P and also has zero projection to P, hence is zero. The classification claim also needs a proof or a citation to standard representation theory, for example Slansky's tables or Bourbaki. Since Theorem 12 and Theorem 13 depend directly on Lemma 9, this gap should be closed before the central claim is accepted.
[Section 6, Proposition 5] The proof asserts that the 32 projected weights of M^+ are distinct and all have the same length, and then that 'the only representations of so12 with these properties are the two chiral spinor representations.' As in Lemma 9, distinctness follows from a short argument using the orthogonal decomposition, and the classification is a standard but unstated fact about D6 minuscule weights. Please add the argument or a citation. This claim is part of the structural description of the three sl2(βk) ⊕ so12(βk) subalgebras and should be supported.

minor comments (6)

[Introduction, p. 2] The line 'gSM ⊕ V1 ⊕ V2 ⊕ V2 ⊂ e7' should read V3, matching the abstract and Theorem 13.
[Section 2, Table 2] The notation e3 for sl3 ⊕ sl2 is nonstandard and potentially confusing, since e3 is not a conventional exceptional Lie algebra; a footnote or a different notation such as sl3 ⊕ sl2 would help.
[Section 7, Proposition 7] The phrase 'C⊗P ⊂ e6' should be 'C⊗P ⊂ e7'.
[Theorem 11 proof] The symbol 'slgen6' in the proof should be 'slSM6'.
[Theorem 13 proof] The sentence 'C⊗P is annihilated by bracketing with slSM6 because it is spanned by the roots βk' is imprecise: C⊗P is spanned by the Cartan elements corresponding to βk, not by the root vectors themselves. The conclusion is nevertheless correct, since those Cartan elements lie in slgen3, which commutes with slSM6.
[Lemma 3] The statement that a given E7 root is orthogonal to 60 roots is used to determine |Φ0| and |Φk| but is asserted without proof or citation; a one-line derivation or a reference to a root-system table would make the paper more self-contained.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the three-generation decomposition follows from the chosen regular embedding and E7 root-system combinatorics; the exterior-algebra identification is an external benchmark, and the acknowledged choice-dependence affects only the labeling of generations.

full rationale

The central derivation is not circular. The paper fixes a regular ('good') embedding gSM⊂e7 by a root-removal chain (Section 2); the direct sum e7 = slSM6⊕(C⊗P)⊕V1⊕V2⊕V3 is then obtained by partitioning the E7 root system via the orthogonal-projection trichotomy of Lemma 2 and counting root spaces (Lemmas 3 and 4, Propositions 5 and 7, Theorem 13). No parameter is fitted, and no conclusion is used as an input. The identification of each Vk with ΛC5 is introduced only after the modules are computed from root combinatorics (Theorems 10 and 12) and is an external benchmark from the SU(5) literature (Ref. [2]), not a constraint used to construct Vk. The paper's self-citations ([1], [3]) are contextual and not load-bearing; the load-bearing cited facts (A4 subsystem orbit, SU(5) restriction, Dynkin's maximal-subalgebra classification) are independent. The closing caveat that Theorem 13 depends on a choice of Cartan subalgebra and roots βk limits the labeling of generations, not the existence of the three slSM6-submodules, so it is a scope condition rather than circularity. Lemma 9's assertion that a 15-dimensional sl6-module with 15 distinct equal-length weights must be Λ2C6⊕Λ4C6, and the analogous so12 classification in Proposition 5, are left without proof or citation; these are completeness/correctness gaps, not circular reductions, because they do not presuppose the Standard Model representation being derived.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No free numeric parameters are fitted. The paper's construction depends on the choice of a regular 'good' embedding of gSM in e7 (Section 2), on standard representation-theoretic classification facts used as black boxes (Lemma 9, Proposition 5), and on the known SU(5) exterior-algebra model of Standard Model fermions (Section 1, [2]). No new physical entities are postulated.

assumptions (4)

domain assumption A 'good' regular embedding gSM ⊂ e7 exists via the Dynkin root-removal chain e7 → e6 → so10 → sl5 → gSM, and all regular A4 root subsystems of E7 are conjugate under the Weyl group.
Section 2 uses this to define the class of embeddings to which all later constructions apply. The paper explicitly notes other embeddings exist that are not related by automorphisms.
standard math The classification of low-dimensional irreducible representations of sl6 and so12: a 15-dimensional sl6 rep with 15 distinct equal-length weights is Λ2C6 or its dual, and a 32-dimensional so12 rep with 32 distinct equal-length weights is a half-spinor.
Used in Lemma 9 and Proposition 5 to identify the representations; asserted without proof or citation.
domain assumption The exterior algebra ΛC5 with the restriction of the natural SU(5) representation to GSM is the Standard Model representation of one generation of fermions and antiparticles.
Section 1 and Table 3 take this from Baez and Huerta [2]; it is the physical interpretation that gives the paper its meaning.
standard math The maximal subalgebra decomposition of e7 as a representation of sl_gen3 ⊕ sl6, e7 = sl3 ⊕ sl6 ⊕ (3⊗15) ⊕ (3*⊗15*), is valid.
Theorem 11 cites Slansky [14, Table 52] and gives a proof outline; it underlies the 45-dimensional splitting into three generations.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Three Generations in E7." pith.science (2026). https://pith.science/paper/F3JQIMD2

@misc{pith2026260806271,
  author       = {Pith},
  title        = {Pith review of: Three Generations in E7},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F3JQIMD2}},
  note         = {Machine review of arXiv:2608.06271}
}

abstract

Starting from the Standard Model Lie algebra $\mathfrak{g}_{\mathrm{SM}} = \mathfrak{sl}_3 \oplus \mathfrak{sl}_2 \oplus \mathbb{C}$ sitting inside the complex Lie algebra $\mathfrak{e}_7$, we show how to decompose $\mathfrak{e}_7$ into the direct sum of a Lie subalgebra containing $\mathfrak{g}_{\mathrm{SM}}$ and three 32-dimensional subspaces, each of which forms the same representation of $\mathfrak{g}_{\mathrm{SM}}$ as one generation of fermions and their antiparticles. The setting is due to Nasmith, and much of the mathematics is that underlying the $\mathrm{E}_7$ generation unification of Kugo and Yanagida. New features include the derivation, in which as many results as possible rely only on the embedding $\mathfrak{g}_{\mathrm{SM}} \subset \mathfrak{e}_7$, and also the description of each 32-dimensional subspace as a copy of the exterior algebra $\Lambda\mathbb{C}^5$.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 1 canonical work pages

[1]

J. C. Baez, Octonions and the Standard Model (Part 13), 2025. Available at https://golem.ph.utexas.edu/ category/2025/12/octonions and the standard mod 11.html. (Referred to on page 4.)

2025
[2]

J. C. Baez and J. Huerta, The algebra of grand unified theories,Bull. Amer. Math. Soc.47(3) (2010), 483–552. Also available as arXiv:0904.1556. (Referred to on page 2, 14.) THREE GENERATIONS IN E 7 17

arXiv 2010
[3]

J. C. Baez, E. Bokor and L. Boyle, Jordan pair quantum mechanics and the Standard Model, 2026. Available as arXiv:2607.10833. (Referred to on page 3.)

arXiv 2026
[4]

P. J. Cameron, J. M. Goethals, J. J. Seidel and E. E. Shult, Line graphs, root systems, and elliptic geometry, J. Alg.43(1976), 305–327. (Referred to on page 8.)

1976
[5]

Distler and S

J. Distler and S. Garibaldi, There is no ‘Theory of Everything’ inside E 8,Comm. Math. Phys.298(2) (2010), 419–436. Also available as arXiv:0905.2658. (Referred to on page 3.)

arXiv 2010
[6]

E. B. Dynkin, Semisimple subalgebras of semisimple Lie algebras, inFive Papers on Algebra and Group Theory, AMS, Providence, 1957, pp. 111–244. (Referred to on page 4, 6, 10.)

1957
[7]

Georgi and S

H. Georgi and S. Glashow, Unity of all elementary-particle forces,Phys. Rev. Lett.32(8) (1974), 438–441. (Referred to on page 11.)

1974
[8]

Kugo and Y

T. Kugo and Y. Yanagida, Unification of families based on a coset space E 7/SU(5)×SU(3)×U(1),Phys. Lett. B134(5) (1984), 313–317. doi:10.1016/0370-2693(84)90007-8. (Referred to on page 3.)

doi:10.1016/0370-2693(84)90007-8 1984

Show all 14 references

[9]

Lisi, Division algebras, triality, and exceptional magic, 2026

G. Lisi, Division algebras, triality, and exceptional magic, 2026. To appear inAdv. Appl. Clifford Algebras. Available at https://deferentialgeometry.org/works/DATaEM.6.3.pdf. (Referred to on page 3.)

2026
[10]

Oshima, A classification of subsystems of a root system, 2006

T. Oshima, A classification of subsystems of a root system, 2006. Available as arXiv:math/0611904. (Re- ferred to on page 4.)

2006 arXiv
[11]

Nasmith, An exceptional combinatorial sequence and Standard Model particles, 2020

B. Nasmith, An exceptional combinatorial sequence and Standard Model particles, 2020. Available as arXiv:2012.03933. (Referred to on page 1, 2, 3, 6, 8, 15.)

2020 arXiv
[12]

Nasmith,Tight Projective 5-Designs and Exceptional Structures

B. Nasmith,Tight Projective 5-Designs and Exceptional Structures. Ph.D. thesis, Royal Military College of Canada, 2023. Available at https://espace.rmc.ca/jspui/handle/11264/1423. (Referred to on page 1, 3.)

2023
[13]

Sato and T

J. Sato and T. Yanagida, Large lepton mixing in a coset-space family unification on E 7/SU(5)×U(1) 3, Phys. Lett. B430(1998) 127–131. arXiv:hep-ph/9710516. (Referred to on page 3.)

1998 arXiv
[14]

Slansky, Group theory for unified model building,Phys

R. Slansky, Group theory for unified model building,Phys. Rep.79(1) (1981), 1–128. (Referred to on page 14.) School of Mathematics, University of Edinburgh,, James Clerk Maxwell Building, Peter Guthrie Tait Road, Edinburgh, UK EH9 3FD Department of Mathematics, University of C...

1981

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

J. C. Baez, Octonions and the Standard Model (Part 13), 2025. Available at https://golem.ph.utexas.edu/ category/2025/12/octonions and the standard mod 11.html. (Referred to on page 4.)

2025

[2] [2]

J. C. Baez and J. Huerta, The algebra of grand unified theories,Bull. Amer. Math. Soc.47(3) (2010), 483–552. Also available as arXiv:0904.1556. (Referred to on page 2, 14.) THREE GENERATIONS IN E 7 17

arXiv 2010

[3] [3]

J. C. Baez, E. Bokor and L. Boyle, Jordan pair quantum mechanics and the Standard Model, 2026. Available as arXiv:2607.10833. (Referred to on page 3.)

arXiv 2026

[4] [4]

P. J. Cameron, J. M. Goethals, J. J. Seidel and E. E. Shult, Line graphs, root systems, and elliptic geometry, J. Alg.43(1976), 305–327. (Referred to on page 8.)

1976

[5] [5]

Distler and S

J. Distler and S. Garibaldi, There is no ‘Theory of Everything’ inside E 8,Comm. Math. Phys.298(2) (2010), 419–436. Also available as arXiv:0905.2658. (Referred to on page 3.)

arXiv 2010

[6] [6]

E. B. Dynkin, Semisimple subalgebras of semisimple Lie algebras, inFive Papers on Algebra and Group Theory, AMS, Providence, 1957, pp. 111–244. (Referred to on page 4, 6, 10.)

1957

[7] [7]

Georgi and S

H. Georgi and S. Glashow, Unity of all elementary-particle forces,Phys. Rev. Lett.32(8) (1974), 438–441. (Referred to on page 11.)

1974

[8] [8]

Kugo and Y

T. Kugo and Y. Yanagida, Unification of families based on a coset space E 7/SU(5)×SU(3)×U(1),Phys. Lett. B134(5) (1984), 313–317. doi:10.1016/0370-2693(84)90007-8. (Referred to on page 3.)

doi:10.1016/0370-2693(84)90007-8 1984