A characterization of Jacobi sums

Andrew Snowden

REVIEW 2 major objections 4 minor 4 references

A characterization of Jacobi sums

T0 review · 2 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Three elementary axioms force every Jacobi-like pairing to come from a finite field.

desk verdict A novel and substantially correct characterization of Jacobi sums; the proof has minor normalization typos that should be fixed but are not fatal. read the letter →

arxiv 2411.15011 v1 pith:WMC2RDKC submitted 2024-11-22 math.NT

Andrew Snowden This is my paper · ORCID

classification math.NT MSC 11T2411T30

keywords JacobisumsfinitefieldscharacterFourierinversionconvolutionidentitycharacterizationtheoremabeliangroupsPontryaginduality

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

Jacobi sums are exponential-sum averages over pairs of nonzero field elements adding to one, and this paper sets out to show that their essential structure is purely axiomatic. The main theorem states that if a function on a nontrivial finite abelian group obeys three elementary identities—symmetry, a two-variable product relation, and a convolution relation—then the group is automatically the dual of the multiplicative group of a finite field, and the function is the Jacobi sum of that field. In other words, the concrete formulas of classical number theory are forced by abstract combinatorial conditions, with a unique field structure recovered from the function itself. The only degeneracy is the trivial group, where one extra solution survives and corresponds to a boolean semi-ring rather than a field.

What carries the argument

The engine of the proof is the convolution identity $Q_\alpha * Q_\beta = Q_{\alpha\beta}$ for $Q_\alpha(\beta)=J(\alpha\beta^{-1},\beta)$, which is exactly condition (C) rewritten. This identity turns the Fourier transform $\widehat Q_\alpha$ into a character of $M$ supported on a common set $S$, so $J$ can be expressed as a Fourier sum over $S$ with a bijection $i:S\to\widehat M$; the pair $(S,i)$ then encodes the prospective field addition. The explicit addition law $x\oplus y=x\,i(x/y)^{-1}$ (with the edge case $x\oplus y=0$ when $x=cy$) is the bridge from the abstract axioms to a concrete field structure.

What would settle it

Search for a counterexample among small cyclic groups: write $J$ as the Fourier sum in Proposition 4.1 with unknown support $S$ and bijection $i$, impose (A)–(C) as polynomial equations, and check whether any solution with $|M|\ge2$ yields an $(S,i)$ that does not define a field.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central object is the field structure hidden inside a Jacobi function. Condition (C) is used to define auxiliary functions $Q_\alpha(\beta)=J(\alpha\beta^{-1},\beta)$, whose convolution identity $Q_\alpha*Q_\beta=Q_{\alpha\beta}$ makes the Fourier transform of $Q_\alpha$ factor. From that factorization the paper reconstructs a subset $S \subseteq \widehat M$ and a bijection $i:S\to\widehat M$, so that $J(\alpha,\beta)=\frac1m\sum_{x\in S}\alpha(i(x))\beta(i(x)x^{-1})$; two cases remain, and the first yields an explicit addition law on $F=\widehat M\sqcup\{0\}$, namely $x\oplus y=0$ when $x=cy$ and $x\oplus y=x\,i(x/y)^{-1}$ otherwise. Associativity of $\oplus$ is proved by comparing two triple sums whose equality follows from condition (B), and the resulting structure is a field whose Jacobi sum is exactly $J$. Thus every Jacobi function on a nontrivial finite abelian group is classical, with the trivial group as the single near-exception.

Load-bearing premise

The argument depends on condition (C), the convolution identity $Q_\alpha*Q_\beta=Q_{\alpha\beta}$: without it the Fourier reconstruction of the field addition has no starting point, and the nontriviality assumption $m\ge2$ is also essential because the trivial group has the extra boolean solution.

Editorial extensions

If this is right

If the theorem is correct, a Jacobi function is never exotic: every solution on a nontrivial finite abelian group is the ordinary Jacobi sum of some finite field, so the axioms describe the whole class.
The field addition is recoverable algorithmically from $J$: compute $\widehat Q_\alpha$, read off the support $S$ and the bijection $i$, then define $\oplus$; no search is required.
Precomposing a Jacobi function with a group automorphism or postcomposing with a Galois automorphism of $\mathbf{C}$ does not produce a new type beyond the classical family; the paper shows each such twist is the Jacobi sum for a new field structure on the same underlying set.
The trivial group is the only place where the characterization fails; the extra solution $J(1,1)=1$ is interpreted as the Jacobi sum of the boolean semi-ring, so the theorem is nearly complete.

Reading between the lines

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

The same axiomatic method should apply to other exponential sums defined by algebraic equations over finite rings: if an analogous convolution identity holds, the Fourier support will again carry a ring structure, giving characterizations for Kloosterman-type sums.
The boolean semi-ring exception suggests a degeneration $q\to 1$ in which finite fields collapse to the boolean semi-ring; one can test this by taking a one-parameter family of Jacobi sums and passing to the limit.
Condition (C) is strong enough that it may imply condition (B) in many settings; checking whether (B) is redundant would simplify the axiomatization to two identities.

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.

Desk Editor's Note

read the letter

Snowden proves that three elementary properties (symmetry, the multiplicative identity (B), and the orthogonality relation (C)) characterize classical Jacobi sums on any non-trivial finite abelian group. The converse is new; Ireland-Rosen, Nica, and the Harman–Snowden tensor-category papers don't have it. The proof is self-contained and uses Fourier inversion on the group to extract the set of pairs (x,y) with x+y=1 from J, then reconstructs the addition law of the field. That's a nice structural argument, and the classical case satisfies axioms by straightforward computation—no circularity, no fitted parameters.

The main theorem holds for m≥2, and the trivial group exception is explicitly acknowledged: there is a non-classical solution with J(1,1)=1, which is the 'boolean semiring' artifact. That is the right level of completeness.

Now the soft spots. Lemma 5.5 contains a normalization error. With S(w) = (1/m)∑_{x⊕y=w} α(x)β(y), orthogonality gives S(0) = α(c)δ(αβ), not α(c)δ(αβ)/m as printed. The following line, S(0) = δ(α)δ(β) − J(α,α^{-1})δ(αβ), is actually the value of (1/m)S(0) after substituting J(α,α^{-1}) = δ(α) − α(c)/m. So the printed chain has the factor of m in the wrong place, and the associativity proof (Lemma 5.6) depends on it. But the intended computation goes through cleanly once S(0) is corrected: you get exactly A = J(α,β)J(αβ,γ) − J(α,α^{-1})δ(αβ) + δ(α)δ(β), and B analogously, so (B) gives equality. This is a repairable typo, not a fatal gap. There is also a slip in the definition of multiplication on F in Section 5: it says 0·x = x, which would make 0 an identity rather than an annihilator; the correct field axiom is 0·x = 0. The proof never actually uses that rule, so it is harmless.

I agree with the reader's take. The theorem is within-subfield, but it is clean, new, and the proof strategy is worth knowing. The paper deserves a serious referee; after fixing the two typos it should be publishable. I would cite it if I worked on Jacobi sums or on the tensor-category side.

Referee Report

2 major / 4 minor

Summary. The paper proposes three elementary axioms (A), (B), and (C) for a complex-valued function J on a finite abelian group M, and proves that any such 'Jacobi function' on a nontrivial M arises from a unique finite field structure on the set F = M-hat union {0}, with J equal to the Jacobi sum associated to that field. The proof reconstructs the addition law by Fourier analysis from the function J, then verifies the field axioms. The classical Jacobi sum is shown to satisfy the axioms, and the trivial group is identified as a genuine exception (the 'boolean semi-ring' case).

Significance. If the proof is made fully rigorous, this is an appealing and self-contained converse: three simple identities characterize Jacobi sums among all functions on finite abelian groups. The Fourier-analytic reconstruction of the field addition is elegant, and the paper explicitly verifies that classical Jacobi sums satisfy the axioms, making the characterization concrete. The result also connects to the author's work on oligomorphic groups and tensor categories. The proof is not machine-checked, but the intended argument is conceptually sound; however, two local but load-bearing slips in Section 5 need correction before the theorem is fully established.

major comments (2)

[Section 5, Lemma 5.5] The computation of S(0) is off by a factor of m. The text defines S(0) = (1/m) times the sum over y in M-hat of alpha(cy) beta(y) and then states S(0) = alpha(c) delta(alpha beta) divided by m. But orthogonality gives the sum over y of alpha(cy) beta(y) equal to alpha(c) times m times delta(alpha beta), so the correct value is S(0) = alpha(c) delta(alpha beta). Consequently the displayed identity 'S(0) = delta(alpha) delta(beta) - J(alpha, alpha^{-1}) delta(alpha beta)' is false as written; the right-hand side equals alpha(c) delta(alpha beta) divided by m, not alpha(c) delta(alpha beta). The subsequent derivation of A = J(alpha,beta) J(alpha beta,gamma) - J(beta,beta^{-1}) delta(alpha beta) + delta(alpha) delta(beta) is nonetheless the correct final result, because the w=0 contribution to A is (1/m) S(0), not S(0). This is a repairable normalization typo, but as printed the proof of Lemma 5.5 is internally inconsistent, and since Lemma 5.6 relies on Lemma 5.5 to prove associativity, the central theorem is not fully proven as written until this step is corrected.
[Section 5, definition of multiplication on F] The text extends multiplication to F by declaring '0 · x = x · 0 = x'. In a field with 0 as the additive identity, 0 must be absorbing for multiplication, i.e., 0 · x = 0. The printed definition makes 0 a second multiplicative identity, which is impossible for a nontrivial group M-hat and is inconsistent with Lemma 5.3(c): for x = 0, the element c · 0 would equal c under the printed definition, but then c ⊕ 0 = c, not the additive identity. The intended definition is almost certainly 0 · x = x · 0 = 0; with that correction Lemma 5.3(c) holds (0 ⊕ 0 = 0) and the classical case matches the usual field structure. This typo affects the statement of Proposition 5.1 and must be fixed.

minor comments (4)

[Theorem 1.2] The uniqueness assertion in Theorem 1.2 is not explicitly proved. It would be helpful to state that the set S and function i in Proposition 4.1 are uniquely determined by J (Fourier inversion is injective), and hence the constructed operation is forced, so the field structure is unique.
[Section 5, Lemma 5.5] After correcting the S(0) normalization, the line 'S(0) = delta(alpha) delta(beta) - J(alpha,alpha^{-1}) delta(alpha beta)' should be rewritten as '(1/m) S(0) = ...' to avoid confusion between the normalized inner sum and its contribution to A.
[Section 6, Lemmas 6.6 and 6.7] The symbol 0 is used both for the additive identity produced by Lemma 6.6 and for an element of the group M-hat in the proof of Lemma 6.7; this is confusing. Using a different symbol for the additive identity would improve clarity.
[Introduction, Remark 1.3(a)] The abstract says the characterization is 'very nearly' complete, and Remark 1.3(a) explains the trivial-group exception. It would be helpful to state this exception explicitly in the abstract or in the theorem statement itself.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Jacobi-sum characterization is derived self-containedly from the axioms, with classical facts used only to verify the classical example.

full rationale

The main theorem (Theorem 1.2) is a characterization result: it starts from the three axioms (A)-(C) defining a Jacobi function and reconstructs a field structure on F = hat(M) union {0} whose Jacobi sum reproduces J. The proof is self-contained: Lemma 4.3 derives Q_alpha * Q_beta = Q_{alpha beta} directly from condition (C); Lemmas 4.4-4.7 use Fourier inversion to obtain the structural representation of J; Section 5 defines the addition from the auxiliary bijection i and proves distributivity and associativity from conditions (A), (B), and the reformulation in Proposition 2.1; Section 6 handles the remaining case. The Ireland-Rosen citation appears only in Section 3, where it verifies that classical Jacobi sums satisfy the axioms; it is not used to prove the characterization direction. The self-references [HS] and [Sno] are contextual remarks about related work and carry no load in the proof. No fitted parameter is renamed as a prediction, and no uniqueness claim is imported from the author's prior work. The explicitly acknowledged trivial-group exception (Remark 1.3(a)) is a genuine boundary case, not a concealed circularity. Thus there is no circular step; at most there are independent correctness concerns, such as a possible normalization issue in Lemma 5.5, which are outside the scope of circularity analysis.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the three defining axioms of a Jacobi function; these are assumed, not derived. The proof itself uses standard Fourier analysis and a cited classical computation. No free parameters or new entities are introduced.

assumptions (5)

domain assumption Condition (A): J(alpha, beta) = J(beta, alpha) for all alpha, beta (symmetry).
Stated in Section 1.1 as one of the three defining properties of a Jacobi function; the classification is conditional on it.
domain assumption Condition (B): J*(alpha, beta) J*(alpha beta, gamma) = J*(alpha, beta gamma) J*(beta, gamma) with J* = J - delta(alpha) - delta(beta).
Defining property (B) in Section 1.1; used in Lemma 4.7 and Lemma 5.5 to equate triple-sum expressions.
domain assumption Condition (C): sum_beta J(alpha_1 beta, alpha_2 beta^{-1}) J(alpha_3 beta, alpha_4 beta^{-1}) = J(alpha_1 alpha_4, alpha_2 alpha_3).
Defining property (C) in Section 1.1; this is the key identity producing the convolution structure in Lemma 4.3.
standard math Pontryagin duality and Fourier inversion for finite abelian groups.
Used throughout Sections 4 and 5 to pass between M and \hat(M); standard background.
standard math Ireland-Rosen computation of J(alpha, alpha^{-1}) in the classical case.
Cited in Section 3 (IR, Section 8.3) to verify condition (B) for classical Jacobi sums and to compute J(alpha, alpha^{-1}) = delta(alpha) - alpha(c)/m in Lemma 5.5.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A characterization of Jacobi sums." pith.science (2026). https://pith.science/paper/WMC2RDKC

@misc{pith2026241115011,
  author       = {Pith},
  title        = {Pith review of: A characterization of Jacobi sums},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WMC2RDKC}},
  note         = {Machine review of arXiv:2411.15011}
}

abstract

Let $\mathbb{M}$ be the group of multiplicative characters of a finite field $\mathbb{F}$, and let $\mathbb{J}(\alpha, \beta)$ be the Jacobi sum, for $\alpha, \beta \in \mathbb{M}$. We observe that the function $\mathbb{J} \colon \mathbb{M} \times \mathbb{M} \to \mathbf{C}$ satisfies three elementary properties. We show that these properties (very nearly) characterize Jacobi sums: if $M$ is an arbitrary non-trivial finite abelian group and $J \colon M \times M \to \mathbf{C}$ is a function satisfying these properties then $M$ is naturally the group of multiplicative characters of a finite field and $J$ is the Jacobi sum.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

4 extracted references · 2 canonical work pages

[1]

Oligomorphic groups and tensor categories

Nate Harman, Andrew Snowden. Oligomorphic groups and tensor categories. arXiv:2204.04526

arXiv
[2]

A classical introduction to modern number theory

Kenneth Ireland, Michael Rosen. A classical introduction to modern number theory. Graduate Texts in Mathematics 84, Springer, 1982. doi:10.1007/978-1-4757-2103-4

doi:10.1007/978-1-4757-2103-4 1982
[3]

Kloosterman sums over finite Frobenius rings

Bogdan Nica. Kloosterman sums over finite Frobenius rings. Acta Arith. 201 (2021), pp. 391--420. doi:10.4064/aa200826-15-7 arXiv:1910.00165

work page Pith review arXiv 2021
[4]

Jacobi tensor categories

Andrew Snowden. Jacobi tensor categories. In preparation

work page

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Oligomorphic groups and tensor categories

Nate Harman, Andrew Snowden. Oligomorphic groups and tensor categories. arXiv:2204.04526

arXiv

[2] [2]

A classical introduction to modern number theory

Kenneth Ireland, Michael Rosen. A classical introduction to modern number theory. Graduate Texts in Mathematics 84, Springer, 1982. doi:10.1007/978-1-4757-2103-4

doi:10.1007/978-1-4757-2103-4 1982

[3] [3]

Kloosterman sums over finite Frobenius rings

Bogdan Nica. Kloosterman sums over finite Frobenius rings. Acta Arith. 201 (2021), pp. 391--420. doi:10.4064/aa200826-15-7 arXiv:1910.00165

work page Pith review arXiv 2021

[4] [4]

Jacobi tensor categories

Andrew Snowden. Jacobi tensor categories. In preparation

work page