Modified Heegner Hypothesis in Iwasawa Theory

• The Modified Heegner Hypothesis is an evolution of the classical condition, introducing refinements such as allowing inert primes through a controlled N = N⁺N⁻ factorization.
• It employs a quaternionic framework where N⁺ includes primes splitting in the quadratic field and N⁻, square-free and even in number, handles inert primes for Shimura-curve constructions.
• By relaxing strict splitting rules, the hypothesis facilitates advanced Euler system constructions and Selmer group analyses, underpinning proofs of main conjectures in anticyclotomic Iwasawa theory.

Modified Heegner hypothesis denotes, in the cited literature, a family of refinements of the classical Heegner hypothesis for a pair consisting of a modular object and an imaginary quadratic or CM field. In the classical form, every prime dividing the level $N$ splits in the quadratic field $K$. The modified forms that occur in anticyclotomic Iwasawa theory, Shimura-curve constructions, supersingular theory, Hida families, and Hilbert modular settings either replace that condition by a factorization $N=N^+N^-$ with split primes in $N^+$ and inert primes in $N^-$, typically with $N^-$ square-free of even cardinality, or retain the classical splitting condition while adding local hypotheses at $p$, residual ramification conditions, or Selmer-theoretic refinements (Burungale et al., 2018, Magrone, 2020).

1. Terminology and classical baseline

The classical Heegner hypothesis is the condition that every prime $\ell\mid N$ splits in $K$. In the modular-curve setting, this is equivalent to the existence of an ideal $\mathfrak N\subset \mathcal O_K$ with $K$0, and it ensures the existence of CM points on $K$1 or its $K$2-variants. Howard’s work on Hida families formulates the condition in exactly this ideal-theoretic way, while Matar and Kriz also keep the classical prime-by-prime splitting hypothesis in their basic Heegner constructions (Howard, 2012, Matar, 2015, Kriz, 2015).

The cited literature suggests that the label “modified Heegner hypothesis” does not refer to a single universally fixed axiom. Some papers use the phrase explicitly, as in Longo–Vigni’s Assumption 1.2 for supersingular anticyclotomic Iwasawa theory, where $K$3, the primes dividing $K$4 split in $K$5, and the primes dividing $K$6 are inert in $K$7 (Longo et al., 2015). Other papers instead speak of a “generalized,” “relaxed,” or “weak” Heegner hypothesis, or they impose the classical Heegner hypothesis together with extra local conditions at $K$8, residual image hypotheses, or refined Selmer conditions (Magrone, 2020, Howard, 2012). In this broader usage, “modified” names a controlled departure from the bare modular-curve hypothesis rather than a single canonical formula.

2. Quaternionic reformulations and the factorization $K$9

The most common modified form is quaternionic. One writes

$N=N^+N^-$0

with $N=N^+N^-$1 divisible only by primes that split in $N=N^+N^-$2 and $N=N^+N^-$3 divisible only by primes that are inert in $N=N^+N^-$4. The standard generalized or relaxed Heegner hypothesis then requires that $N=N^+N^-$5 be a square-free product of an even number of primes. In the case $N=N^+N^-$6, the condition reduces to the classical Heegner hypothesis (Burungale et al., 2018, Burungale et al., 2019).

This factorization is not merely formal. If $N=N^+N^-$7 is square-free of even cardinality, there is an indefinite quaternion algebra $N=N^+N^-$8 of discriminant $N=N^+N^-$9, and the relevant CM points live on the Shimura curve attached to $N^+$0 with $N^+$1- or $N^+$2-type level structure. In higher weight, this is the setting of generalized Heegner cycles on quaternionic Kuga–Sato-type varieties. The same pattern appears in work extending Castella–Hsieh to a quaternionic setting, in Magrone’s quaternionic anticyclotomic Iwasawa theory, and in proofs of the Heegner point main conjecture that allow non-squarefree $N^+$3 and include the classical case $N^+$4 (Magrone, 2020, Pati, 2023, Burungale et al., 2018).

The arithmetic role of $N^+$5 is twofold. Geometrically, it is the discriminant of the quaternion algebra. Analytically and Iwasawa-theoretically, it governs the local conditions at the inert primes and the shape of the Euler system. This is why the “modification” is often described as allowing inert primes in the conductor while moving from modular curves to Shimura curves.

3. Local modifications at $N^+$6 and residual hypotheses

A frequent source of modification is local rather than geometric. Several papers keep the classical Heegner hypothesis at primes dividing $N^+$7 and instead strengthen the setup at $N^+$8. Matar’s anticyclotomic study of fine Selmer groups assumes the classical splitting condition at $N^+$9, but adds $N^-$0, the large-image condition $N^-$1, and delicate ordinary-case congruence conditions on $N^-$2 (Matar, 2015). Howard’s Hida-family construction likewise assumes the existence of $N^-$3 with $N^-$4, and then imposes $N^-$5, ordinarity, residual irreducibility, and later $N^-$6 and $N^-$7 (Howard, 2012).

In supersingular anticyclotomic theory, the modification at $N^-$8 becomes structural. Longo–Vigni require that $N^-$9 split in $N^-$0, that the two primes above $N^-$1 be totally ramified in the anticyclotomic $N^-$2-extension, and then define Kobayashi-style plus/minus local conditions and plus/minus Heegner points. Their modified Heegner hypothesis combines mixed split/inert behavior in the conductor with a supersingular local theory at $N^-$3 (Longo et al., 2015).

A further refinement appears in ordinary Shimura-curve settings through residual ramification conditions. In the work on indivisibility of Heegner points, Condition $N^-$4 requires ramification of $N^-$5 at all primes $N^-$6 and all primes $N^-$7, together with additional hypotheses when $N^-$8 is not square-free or when $N^-$9. Castella’s work on $p$0-adic variation similarly requires $p$1 to be ramified at every prime $p$2 nonsplit in $p$3 (Burungale et al., 2018, Castella, 2014). These conditions are not part of the classical Heegner hypothesis, but they are part of the modified Heegner setting in which the Euler-system and Selmer-theoretic arguments close.

4. Moduli interpretations: from CM points to generalized Heegner cycles

Under the classical hypothesis, one works on modular curves and CM elliptic curves. The existence of an ideal $p$4 of norm $p$5 produces cyclic $p$6-isogenies between CM elliptic curves, hence CM points on $p$7 or $p$8, and therefore Heegner points on an elliptic curve via a modular parametrization. Howard’s big Heegner points over Hida families are built exactly in this way on the tower $p$9, and Kriz’s generalized Heegner cycles at Eisenstein primes still begin with the same classical Heegner splitting at level $\ell\mid N$0 (Howard, 2012, Kriz, 2015).

Under quaternionic modifications, the moduli problem changes. One replaces elliptic curves by QM abelian surfaces, modular curves by Shimura curves attached to an indefinite quaternion algebra $\ell\mid N$1, and ordinary Kuga–Sato varieties by quaternionic analogues. In the generalized setting of relaxed Heegner hypotheses, CM points come from optimal embeddings of quadratic orders into Eichler orders in $\ell\mid N$2; generalized Heegner cycles are then defined on varieties such as

$\ell\mid N$3

and their $\ell\mid N$4-adic Abel–Jacobi images give cohomology classes in the Galois representation attached to the modular form (Magrone, 2020, Pati, 2023).

In Hilbert modular and Shimura-curve settings, the same principle persists with a sign condition replacing full splitting at all level primes. Disegni assumes $\ell\mid N$5, chooses a quaternion algebra ramified exactly at the finite places $\ell\mid N$6 with $\ell\mid N$7 and at all infinite places except one, and obtains Heegner points on the resulting Shimura curve. The $\ell\mid N$8-adic Gross–Zagier formula then relates the derivative of a $\ell\mid N$9-adic Rankin–Selberg $K$0-function to the $K$1-adic height of the Heegner point $K$2 (Disegni, 2013).

5. Selmer structures, Euler systems, and main conjectures

Modified Heegner hypotheses are inseparable from the Selmer structures they support. In ordinary anticyclotomic Iwasawa theory one encounters Greenberg local conditions at $K$3, strict or trivial local conditions at selected primes, and $K$4-minimal Selmer groups obtained by imposing trivial local conditions at inert primes dividing $K$5. In the generalized Heegner setting of elliptic curves over $K$6, these local choices are precisely what connect Heegner classes to the Bertolini–Darmon–Prasanna $K$7-adic $K$8-function and to the anticyclotomic control theorems (Burungale et al., 2018).

The main conjectural framework is Perrin–Riou’s Heegner point main conjecture. In Howard’s formulation, the $K$9-adic Heegner class $\mathfrak N\subset \mathcal O_K$0 should generate the rank-one part of the Greenberg Selmer group, while the characteristic ideal of the torsion module $\mathfrak N\subset \mathcal O_K$1 should equal the characteristic ideal of $\mathfrak N\subset \mathcal O_K$2. Howard proved one divisibility; later work upgraded this to equality using Wei Zhang’s proof of Kolyvagin’s conjecture, Kolyvagin’s structure theorem for $\mathfrak N\subset \mathcal O_K$3, and an explicit reciprocity law for Heegner points. A decisive feature is that these arguments now allow non-squarefree conductors and include the classical case $\mathfrak N\subset \mathcal O_K$4 (Burungale et al., 2018).

Castella, Çiperiani, Skinner, Sprung, and Wan prove Perrin–Riou’s Heegner point main conjecture under the generalized Heegner hypothesis by combining Howard’s bipartite Euler systems with Wei Zhang’s work. When $\mathfrak N\subset \mathcal O_K$5 splits in $\mathfrak N\subset \mathcal O_K$6, they also prove the Iwasawa–Greenberg main conjecture for the Bertolini–Darmon–Prasanna $\mathfrak N\subset \mathcal O_K$7-adic $\mathfrak N\subset \mathcal O_K$8-function, showing that the Heegner-point and $\mathfrak N\subset \mathcal O_K$9-adic-$K$00-function formulations are equivalent in that setting (Burungale et al., 2019).

A further reinterpretation views the Heegner construction as a higher-rank Euler system. In the anticyclotomic ordinary setting with the classical Heegner hypothesis, Heegner points can be organized as a rank-two Euler-system special element $K$01, and Perrin–Riou’s main conjecture becomes an Iwasawa main conjecture for this element. Darmon-type derivatives of $K$02 then connect the Heegner-point main conjecture to Bockstein regulators and, conjecturally, to the $K$03-part of Birch–Swinnerton-Dyer over $K$04 (Kataoka et al., 2022).

6. Totally real and Hilbert modular generalizations

Over totally real fields, the modified Heegner hypothesis is usually expressed by a sign condition or by a split/inert factorization of the tame level. Howard’s work on abelian varieties of $K$05-type assumes a weak Heegner hypothesis

$K$06

with $K$07. This determines a quaternion algebra $K$08 ramified exactly at a prescribed set of finite places and at all but one archimedean place, thereby producing Shimura curves with CM points by $K$09 and Heegner points on the quotient $K$10 attached to a Hilbert modular form (Howard, 2012).

Disegni’s $K$11-adic Gross–Zagier formula is another Hilbert modular modification. The global condition $K$12 replaces the prime-by-prime classical splitting statement, while the $K$13-adic formula also assumes that every prime $K$14 is principal in $K$15 and splits in $K$16. The result is a $K$17-adic Rankin–Selberg $K$18-function whose central derivative equals the $K$19-adic height of a Heegner point on the associated abelian variety (Disegni, 2013).

Recent quaternionic work over totally real fields adopts a weak Heegner hypothesis of the form

$K$20

where the primes dividing $K$21 split in $K$22, the primes dividing $K$23 are inert in $K$24, and $K$25 is square-free. The parity of the number of primes in $K$26, together with $K$27, determines whether the quaternion algebra is definite or indefinite. On this basis one constructs towers of Gross curves or Shimura curves, big Heegner points attached to Hida families of Hilbert modular forms, and in the indefinite case big Heegner classes in the sense of Howard (Jiménez, 30 Oct 2025).

7. Scope, variations, and common misconceptions

A common misconception is that a modified Heegner hypothesis always means allowing inert primes in the level. The cited literature shows a broader picture. In several papers the primes dividing $K$28 still all split in $K$29, and the modification lies instead in the local package at $K$30, in residual ramification assumptions, or in the choice of Selmer conditions. This is the case in Matar’s work on fine Selmer groups and in Howard’s study of variation in Hida families (Matar, 2015, Howard, 2012).

Another misconception is that the modification is merely technical. The papers considered here show a more structural phenomenon. The relevant hypotheses are designed so that CM points, Heegner points, or generalized Heegner cycles exist on the appropriate moduli space—modular curve, Shimura curve, Gross curve, or Hilbert modular Shimura curve—and so that their cohomology classes satisfy the local conditions required by Euler-system and Iwasawa-theoretic arguments. The precise form of the modified hypothesis therefore depends on the geometric ambient space, the reduction type at $K$31, and the Selmer-theoretic framework one wants to use (Burungale et al., 2018, Jiménez, 30 Oct 2025).

In that sense, “modified Heegner hypothesis” is best understood as an umbrella term for the controlled replacement of the classical split-at-all-level-primes condition by a package of hypotheses adapted to quaternion algebras, generalized Heegner cycles, plus/minus theory at supersingular primes, Hida families, or totally real base fields. The unifying purpose is stable across the literature: to preserve a Heegner-theoretic source of global cohomology classes while extending the reach of Gross–Zagier, Kolyvagin, and anticyclotomic Iwasawa theory beyond the original modular-curve setting.