Fundamental Lemma for Hecke Correspondences

• The Fundamental Lemma for Hecke correspondences is a local matching statement that equates spherical Hecke algebras on PGL₃(F) with those for anti-genuine functions on its cubic metaplectic cover.
• It constructs an explicit Hecke algebra isomorphism indexed by pairs (m, n) and utilizes cubic Hilbert symbols as transfer factors to relate unit and non-unit elements.
• The matching extends to all spherical Hecke functions, providing the unramified backbone for global Shimura lift comparisons and applications in relative trace formulas.

The Fundamental Lemma for Hecke correspondences is a local matching statement between orbital-type distributions attached to corresponding Hecke operators on two different spaces. In the cubic setting studied in "On the cubic Shimura lift to $PGL(3)$: Hecke correspondences" (Friedberg et al., 9 Jul 2025), it concerns $PGL_3(F)$ and the cubic metaplectic cover of $SL_3(F)$ over a non-archimedean local field $F$ containing the cube roots of unity. The result identifies the spherical Hecke algebra of $PGL_3(F)$ with the spherical Hecke algebra of anti-genuine functions on the triple cover and proves that the relative distributions on the $PGL_3$ side match the metaplectic Kuznetsov distributions on the covering side, up to explicit transfer factors, for all corresponding spherical Hecke functions (Friedberg et al., 9 Jul 2025). This extends an earlier unit-element matching in the same cubic Shimura-lift program (Friedberg et al., 2022) and belongs to a broader family of fundamental lemmas for spherical or parahoric Hecke algebras, relative trace formulas, and arithmetic intersections (Feng, 2017, Lemaire et al., 2015, Wang et al., 2024, Chen, 10 Feb 2025).

1. Local cubic setting and the two sides of the correspondence

The local framework in (Friedberg et al., 9 Jul 2025) assumes that $F$ is a non-archimedean local field with residual characteristic greater than $3$, that $F$ contains a cube root of unity $\rho$, and that $PGL_3(F)$0. The notation is fixed by taking $PGL_3(F)$1 as the ring of integers, $PGL_3(F)$2 as the maximal ideal, $PGL_3(F)$3, and an additive character $PGL_3(F)$4 of conductor $PGL_3(F)$5.

On the linear side, the group is

$PGL_3(F)$6

with hyperspecial maximal compact subgroup $PGL_3(F)$7, where $PGL_3(F)$8. The upper triangular unipotent subgroup $PGL_3(F)$9 is embedded in $SL_3(F)$0, and if

$SL_3(F)$1

This character $SL_3(F)$2 is the Whittaker-type character used throughout the relative and Kuznetsov constructions (Friedberg et al., 9 Jul 2025).

On the covering side, one considers the cubic metaplectic cover

$SL_3(F)$3

The paper realizes this extension by means of the block-compatible $SL_3(F)$4-cocycle $SL_3(F)$5 of Banks–Levy–Sepanski, forming first

$SL_3(F)$6

on $SL_3(F)$7, and then restricting to $SL_3(F)$8 (Friedberg et al., 9 Jul 2025). The trivial section splits $SL_3(F)$9, so the same subgroup $F$0 and the same character $F$1 appear on the metaplectic side as well.

A central notion is that of an anti-genuine function: a function $F$2 is anti-genuine if

$F$3

This convention is intrinsic to the Hecke algebra used in the paper (Friedberg et al., 9 Jul 2025).

Because the residual characteristic is not $F$4, the hyperspecial subgroup $F$5 splits in the cover. A specific splitting $F$6 is chosen, and the image of $F$7 in $F$8 is denoted $F$9. The subgroup $PGL_3(F)$0 is a maximal compact subgroup of $PGL_3(F)$1 (Friedberg et al., 9 Jul 2025).

2. Spherical Hecke algebras and the cubic Satake relation

The spherical Hecke algebra on the $PGL_3(F)$2 side is

$PGL_3(F)$3

with convolution. On the covering side, the relevant algebra is the spherical Hecke algebra of anti-genuine functions,

$PGL_3(F)$4

Both are equipped with explicit bases indexed by pairs $PGL_3(F)$5 (Friedberg et al., 9 Jul 2025).

For $PGL_3(F)$6, the standard diagonal representatives are

$PGL_3(F)$7

and $PGL_3(F)$8 denotes the characteristic function of $PGL_3(F)$9. For the cubic cover, the diagonal representatives are

$PGL_3$0

and $PGL_3$1 is the unique anti-genuine function supported on $PGL_3$2 satisfying

$PGL_3$3

The families $PGL_3$4 and $PGL_3$5 are $PGL_3$6-bases of the two Hecke algebras (Friedberg et al., 9 Jul 2025).

The key algebraic statement is the explicit Hecke algebra isomorphism

$PGL_3$7

The paper denotes this map by $PGL_3$8 and proves that it is an algebra isomorphism by following Kazhdan–Patterson's method (Friedberg et al., 9 Jul 2025). In Satake terms, this intertwines the unramified Hecke algebras in the cubic setting and is consistent with the generalized Shimura picture in which Satake parameters are related by a cubic power map. The paper notes that the explicit basis relation is sufficient for the Fundamental Lemma itself (Friedberg et al., 9 Jul 2025).

This passage from the unit element to all spherical Hecke elements is one of the decisive features of the result. In the cubic $PGL_3$9–triple-cover setting, earlier work had established the unit-element matching only (Friedberg et al., 2022). The extension from the unit to the full spherical Hecke algebra parallels other developments in the subject, such as the reduction from unit elements to all spherical Hecke elements in twisted endoscopy (Lemaire et al., 2015), the parahoric base-change fundamental lemma for $F$0 (Feng, 2017), and the Jacquet–Rallis relative fundamental lemma for spherical Hecke algebras (Wang et al., 2024).

3. Relative distributions on $F$1 and metaplectic Kuznetsov distributions

On the $F$2 side, the local distributions arise from a period involving the minimal representation on $F$3. The paper introduces

$F$4

acting on $F$5 via explicit coordinate formulas, together with a cocycle

$F$6

satisfying the cocycle relation

$F$7

An element $F$8 is called relevant when

$F$9

and for such $3$0 the relative orbital integral is

$3$1

Using the adjoint embedding $3$2 and the Weil representation $3$3, the paper rewrites this as

$3$4

and takes test functions of the form $3$5 with $3$6 and $3$7 (Friedberg et al., 9 Jul 2025).

On the metaplectic side, the local distributions are Kuznetsov-type orbital integrals on $3$8. If $3$9 is relevant for the $F$0 action, meaning that

$F$1

then the metaplectic orbital integral is

$F$2

The quotient measure is normalized following Jacquet–Ye, with explicit coordinates for the relevant orbits given in the paper (Friedberg et al., 9 Jul 2025).

The matching problem is organized by orbit representatives. On the $F$3 side, the relevant orbits consist of a generic family $F$4, two one-parameter families $F$5 and $F$6, and three isolated orbits $F$7. On the metaplectic side, the corresponding orbit representatives are $F$8 in the big Bruhat cell, two one-parameter families $F$9 in small Bruhat cells, and three isolated central orbits $\rho$0 (Friedberg et al., 9 Jul 2025). The paper states that there is a non-canonical bijection of orbit spaces

$\rho$1

and the Fundamental Lemma is the assertion that the two families of local distributions match under that bijection for corresponding Hecke functions.

4. The Fundamental Lemma and its transfer factors

The main local theorem of (Friedberg et al., 9 Jul 2025) matches the relative distributions on $\rho$2 with the metaplectic Kuznetsov distributions on $\rho$3 for all spherical Hecke functions related by $\rho$4. The statement is divided according to the orbit type.

For the generic family, with parameters related by

$\rho$5

the theorem states

$\rho$6

The transfer factor here is the cubic Hilbert symbol $\rho$7 (Friedberg et al., 9 Jul 2025).

For the one-parameter families,

$\rho$8

where

$\rho$9

and

$PGL_3(F)$00

For the isolated orbits, there is no extra factor:

$PGL_3(F)$01

These three formulas constitute the local Fundamental Lemma in the paper (Friedberg et al., 9 Jul 2025).

Two structural features are emphasized. First, the matching holds for every spherical Hecke element $PGL_3(F)$02, not merely for the unit element. Second, the product over all places of the local transfer factors is $PGL_3(F)$03, which is essential for the intended global trace formula comparison (Friedberg et al., 9 Jul 2025).

This kind of statement is relative rather than endoscopic in the classical sense. The paper explicitly contrasts it with the classical endoscopic Fundamental Lemma, which usually concerns matching orbital integrals across endoscopic groups and is often first established for unit elements. Here the setting is relative, the Hecke algebra is matched in full, and one side is a higher-degree metaplectic cover with anti-genuine test functions (Friedberg et al., 9 Jul 2025). A comparable enlargement from unit elements to all spherical Hecke elements appears in twisted endoscopy (Lemaire et al., 2015), while relative Hecke-algebra fundamental lemmas have also been proved in the Jacquet–Rallis setting over local function fields (Wang et al., 2024).

5. Structure of the proof and arithmetic input

The proof in (Friedberg et al., 9 Jul 2025) proceeds in six stages. The first is the construction of the Hecke algebra isomorphism

$PGL_3(F)$04

The second is an expansion of orbital integrals on both sides in terms of diagonal translates of the unit element, using right $PGL_3(F)$05-invariance on $PGL_3(F)$06 and right $PGL_3(F)$07-invariance on the cover. On the $PGL_3(F)$08 side, the expansion has the form

$PGL_3(F)$09

while on the cover side one has

$PGL_3(F)$10

These reductions isolate the unit-element orbital integrals together with explicit Whittaker coefficients (Friedberg et al., 9 Jul 2025).

The third stage computes those Whittaker coefficients. On $PGL_3(F)$11 they are determined by the Cartan decomposition in $PGL_3(F)$12, and on the cover side they are computed in the metaplectic setting, with careful tracking of the cocycle $PGL_3(F)$13 and the compact splitting $PGL_3(F)$14 (Friedberg et al., 9 Jul 2025).

The fourth stage is the explicit evaluation of the unit-element orbital integrals. On the relative side this involves recursive decomposition of complicated $PGL_3(F)$15-adic domains and identities for cubic exponential sums $PGL_3(F)$16 and $PGL_3(F)$17. On the metaplectic side it requires Kloosterman sums with cubic character together with cocycle computations (Friedberg et al., 9 Jul 2025).

The fifth stage is the number-theoretic bridge between the two sides. The paper uses the identity

$PGL_3(F)$18

in the cases specified by Duke–Iwaniec and by earlier work of the same authors, together with Gauss sums, Jacobi sums, and stationary phase (Friedberg et al., 9 Jul 2025). This identity explains the appearance of the cubic Hilbert-symbol transfer factor and allows one to compare the generic-orbit calculations exactly.

The sixth stage assembles the expansions, the Whittaker coefficients, the explicit orbital integrals, and the number-theoretic identities to prove the three matching formulas orbit by orbit (Friedberg et al., 9 Jul 2025).

The paper characterizes this as a first direct proof, by explicit $PGL_3(F)$19-adic computation rather than by transfer from function fields, of a Fundamental Lemma for the full spherical Hecke algebra in a relative rank-two setting with a higher-degree cover (Friedberg et al., 9 Jul 2025). This suggests a methodological contrast with geometric proofs of Hecke-algebra fundamental lemmas via shtukas or Hitchin fibrations, such as the parahoric base-change fundamental lemma for $PGL_3(F)$20 (Feng, 2017), the standard spherical Hecke-algebra fundamental lemma via multiplicative Hitchin fibrations (Wang, 2024), and the Jacquet–Rallis relative fundamental lemma for spherical Hecke algebras (Wang et al., 2024).

6. Global role, conceptual significance, and related directions

The local theorem of (Friedberg et al., 9 Jul 2025) is presented as the geometric input needed for a relative trace formula comparison between $PGL_3(F)$21 and the triple cover of $PGL_3(F)$22. Its intended global consequence is a new Shimura lift from genuine automorphic representations on the cubic cover of $PGL_3(F)$23 to automorphic representations on $PGL_3(F)$24, together with a characterization of the image by a period involving the minimal representation on $PGL_3(F)$25 (Friedberg et al., 9 Jul 2025).

This fits directly into the broader program initiated in the earlier paper "On the Cubic Shimura lift to $PGL_3(F)$26: The Fundamental Lemma" (Friedberg et al., 2022). That earlier work analyzed global distributions on both sides, showed that they decompose into factorizable orbital integrals, and established the local matching for the unit elements of the spherical Hecke algebras. The 2025 paper extends that local matching to all spherical Hecke operators and therefore supplies the Hecke correspondence statement required for a full unramified comparison (Friedberg et al., 2022, Friedberg et al., 9 Jul 2025).

Within the wider literature, the phrase "Fundamental Lemma for Hecke correspondences" covers several closely related but distinct phenomena. In twisted endoscopy, one can pass from the unit element to all spherical Hecke elements through spectral transfer and descent (Lemaire et al., 2015), with the torus endoscopic unit case supplying the remaining local input in characteristic zero (Lemaire et al., 2015). In base change, one matches twisted and ordinary orbital integrals for centers of parahoric or Iwahori Hecke algebras (Feng, 2017, Ray-Dulany, 2017). In relative trace formulas, one proves corresponding identities for spherical Hecke algebras on symmetric spaces or spherical varieties, as in unitary Friedberg–Jacquet periods (Leslie, 2020) and Jacquet–Rallis (Wang et al., 2024). In arithmetic variants, one replaces orbital-integral identities by identities between derivatives of weighted orbital integrals and intersection numbers of Hecke correspondences on Rapoport–Zink spaces or integral models (Li et al., 2023, Chen, 10 Feb 2025, He et al., 26 Feb 2025).

The cubic $PGL_3(F)$27–triple-cover theorem belongs to this relative and metaplectic branch. Its distinguishing features are the cubic cover, the anti-genuine spherical Hecke algebra, the period involving the minimal representation on $PGL_3(F)$28, and the explicit transfer factors built from the cubic Hilbert symbol and oscillatory characters (Friedberg et al., 9 Jul 2025). A plausible implication is that it provides the unramified Hecke-theoretic backbone for a new global functoriality statement of Shimura type in rank two, but the paper itself presents this as a step toward, rather than the completion of, the global comparison (Friedberg et al., 9 Jul 2025).