Local Tate Duality

Local Tate duality perfectly pairs complementary-degree Galois cohomology of a finite module and its dual over a p-adic local field.

Version

v1 · 2026-10-04 · History

Domain-specific #

13746

Domain group

Formal Sciences

Origin domain

Mathematics

Subdomain

Arithmetic Geometry → Mathematics

Aliases

Tate local duality, Local Galois duality

Core Idea

Local Tate duality, in the finite-module form stated here, pairs complementary-degree Galois cohomology over a p-adic local field when the coefficient module is replaced by its appropriate dual. Let \(K\) be a finite extension of \(\mathbb Q_p\), let \(M\) be a finite continuous \(G_K\)-module, and put \(M^D=\operatorname{Hom}(M,\overline K^\times)\) with its induced Galois action. For \(r=0,1,2\), the cup product and local invariant give a perfect pairing

\[ H^r(K,M)\ \times\ H^{2-r}(K,M^D) \longrightarrow H^2(K,\mathbb G_m) \xrightarrow{\operatorname{inv}_K}\mathbb Q/\mathbb Z. \]

The word perfect matters: each finite cohomology group is determined as the character dual of the group in the complementary degree, not merely connected by an arbitrary bilinear map.^[1]

It would be incorrect to identify \(H^2(K,\mu)\) itself with all of \(\mathbb Q/\mathbb Z\). In the displayed theorem the full target is \(H^2(K,\mathbb G_m)\), the local Brauer group. With finite \(\mu_n\) coefficients one obtains an \(n\)-torsion piece, not the whole divisible group. This coefficient distinction is central, not typographical.^[1]

Structural Signature

Sig role-phrases:

1. Local arithmetic setting: a specified local field \(K\) and its absolute Galois group.
2. Finite coefficient: a finite \(G_K\)-module \(M\) with continuous action.
3. Tate/Cartier dual: \(M^D=\operatorname{Hom}(M,\overline K^\times)\), carrying the appropriate twist.
4. Complementary degrees: cohomology in degree \(r\) pairs with degree \(2-r\).
5. Cup product and evaluation: the coefficient pairing produces a class in the local Brauer group.
6. Invariant and perfectness: the Brauer invariant maps into \(\mathbb Q/\mathbb Z\), and the resulting finite pairing is nondegenerate on both sides.^[1]

Condensed: local Galois group + finite module and its Tate dual + cup product in degrees summing to two → perfect arithmetic pairing.

What It Is Not

• Not an equality between \(H^2(K,\mu_n)\) and all of \(\mathbb Q/\mathbb Z\). The coefficient and torsion level matter.^[1]
• Not a theorem pairing a module with itself in general. The dual is twisted by roots of unity; special self-dual examples require an additional pairing.
• Not ordinary vector-space duality. The construction uses Galois cohomology, cup product and the local Brauer invariant.
• Not automatically the same theorem for \(p\)-adic vector-space representations. Topology and coefficient hypotheses change.
• Not global Poitou–Tate duality. Global statements assemble contributions over many places and have different exact-sequence structure.^[1][2]
• Not a blanket claim that all local conditions are self-annihilating. Annihilator relations require specified modules, pairings and conditions.

Scope of Application

In the stated finite-module case over a finite extension of \(\mathbb Q_p\), the degrees \(0,1,2\) have complementary partners \(2,1,0\). Degree-zero invariants are dual to degree-two cohomology classes of the dual module; degree one is self-complementary in degree, but still pairs \(M\) with \(M^D\). Milne's account explicitly presents this finite-module theorem separately from local duality for abelian varieties, for which the complementary degrees and groups are formulated differently.^[1]

For a concrete coefficient boundary, take \(M=\mathbb Z/n\mathbb Z\) with trivial action. Its Tate dual is \(\mu_n\), the \(n\)-th roots of unity, not simply another un-twisted copy of \(M\) over every \(K\). The degree-zero/two pairing is finite. The local invariant on \(H^2(K,\mathbb G_m)\) can take values in all of \(\mathbb Q/\mathbb Z\), while the \(\mu_n\) contribution lands in its \(n\)-torsion. This example shows why the target groups must be distinguished.^[1]

In arithmetic applications, a degree-one pairing can define orthogonal local conditions: given a subgroup in \(H^1(K,M)\), its annihilator is a subgroup of \(H^1(K,M^D)\). This is an operation enabled by perfectness, not the assertion that every naturally chosen subgroup equals its own annihilator. Specialist notes use this form when discussing global deformation and Selmer-style arguments.^[2]

Clarity

Always specify field, coefficient category, dual convention, degree and target. Calling both \(M^D\) and \(V^*(1)\) “the Tate dual” can be reasonable in their respective settings, but substituting a \(p\)-adic representation \(V\) into the finite-module theorem without changing topology is not. Likewise write the output as \(H^2(K,\mathbb G_m)\cong\mathbb Q/\mathbb Z\) if that is the step being used; writing \(H^2(K,\mu_n)=\mathbb Q/\mathbb Z\) hides a false enlargement.^[1]

Manages Complexity

Local cohomology groups can be difficult to compute separately. The theorem links one degree to a dual complementary degree and turns some local-condition questions into orthogonality questions. This compresses several calculations into one pairing architecture. But the compression only works when the coefficient, twist and invariant map are correct. A compact formula with the wrong target can produce incorrect cardinalities or annihilator claims.^[1][2]

Abstract Reasoning

Begin with a finite continuous Galois module over a specified \(p\)-adic local field. Construct its Tate dual and the evaluation pairing of coefficients. Use cup product to obtain a degree-two class, then apply the local invariant. Check that cohomology degrees sum to two and invoke perfectness only under the theorem's hypotheses. For a local condition \(L\subset H^1(K,M)\), define \(L^\perp\) in the dual degree-one group; do not call it \(L\) without an independently justified identification.^[1][2]

Knowledge Transfer

The complementary-degree/cup-product pattern informs local-global arithmetic and theories of Selmer groups. What transfers is the design of dual coefficients and orthogonal conditions. A finite-module pairing, abelian-variety duality, equal-characteristic flat duality and a \(p\)-adic representation pairing have related but nonidentical hypotheses; results cannot be moved among them by notation alone.

Examples

Constant module and roots of unity

Let \(K\) be a finite extension of \(\mathbb Q_p\) and take the constant finite module \(M=\mathbb Z/n\mathbb Z\). Its dual is \(\mu_n\). The degree-zero/degree-two pairing runs through \(H^2(K,\mu_n)\) and then \(H^2(K,\mathbb G_m)=\operatorname{Br}(K)\) before the local invariant. The coefficient-level group is finite \(n\)-torsion; it is not all of \(\mathbb Q/\mathbb Z\). This is a worked specialization of Milne's theorem, not an independently observed physical case.^[1]

Mapped back: local field \(K\) → finite module \(\mathbb Z/n\mathbb Z\) → dual \(\mu_n\) → complementary degrees 0 and 2 → cup product, Brauer invariant and perfect finite pairing.

Orthogonal degree-one condition in deformation theory

The cited Galois-representation course notes take a finite adjoint module \(M=\operatorname{ad}^0\bar\rho\) for an \(n\)-dimensional residual representation over characteristic \(\ell\), under their simplifying assumption \(\ell\nmid n\) (and, for the surrounding global calculation, \(\ell>2\)). This ensures the relevant scalar/trace-zero splitting; the notes then use the nondegenerate trace pairing on \(M\) to identify its Tate dual with \(M(1)\). For a non-archimedean local field, the degree-one local-duality pairing sends a deformation condition \(\mathcal L(\mathcal D_v)\subseteq H^1(K,M)\) to its orthogonal complement in \(H^1(K,M^D)\). This is a documented mathematical use of the theorem under stated coefficients, not a proof that arbitrary named local conditions are self-dual.^[2]

Mapped back: local field \(K\) → finite adjoint coefficient module \(M\) and trace-dual \(M^D\) → complementary degrees 1 and 1 → invariant-valued perfect pairing → annihilator \(\mathcal L(\mathcal D_v)^\perp\) on the dual side.

Structural Tensions

Concise invariant target versus finite coefficient level. Saying “pairs into \(\mathbb Q/\mathbb Z\)” makes the perfect duality memorable but can falsely equate \(H^2(K,\mu_n)\) with the entire Brauer group. Writing only the finite coefficient group prevents that mistake but may hide why the invariant yields a common target. Diagnostic: which coefficient group receives the cup product, and through which map does the local invariant act?^[1]

General template versus exact theorem. A broad “dual objects pair” template aids transfer but may silently extend this finite-module theorem to abelian varieties or \(p\)-adic representations with different topologies. Stating only the narrow theorem avoids that overreach but may conceal separately proved extensions. Diagnostic: what field, coefficient category, topology and degrees does the invoked theorem actually specify?^[1]

Orthogonality versus self-duality. Taking an annihilator gives a precise dual local condition but does not identify it with the original one. Calling a condition “self-dual” too quickly can invalidate a global deformation comparison; refusing any comparison until literal equality is shown can discard a legitimate pairing between distinct coefficient modules. Diagnostic: what trace or other map identifies the dual modules, and what theorem identifies the two subgroups?^[2]

Structural–Framed Character

Local Tate duality lies toward the structural end of the spectrum because complementary-degree perfectness is a mathematical relation, yet it is framed by local-field cohomology and exact coefficient hypotheses. Evaluative weight is low: the theorem is true or false under its hypotheses regardless of whether an arithmetic application is useful. Human proof practice determines what has been established, but does not create the pairing; Tate's work and later expositions give its historical and institutional provenance without making any one institution part of its statement. The vocabulary travels literally to the same theorem in local arithmetic settings, while extensions to abelian varieties or \(p\)-adic representations require their own named hypotheses. Calling an arbitrary software pairing “local Tate duality” imports a metaphor; recognizing generic duality there does not instantiate this theorem. Its character: an exact arithmetic duality with a portable perfect-pairing skeleton and nonportable cohomological premises.

Structural Core vs. Domain Accent

The skeletal relation is a perfect pairing between degree-complementary invariants of dual objects. The domain-bound mechanism is cup product in absolute-Galois cohomology, Cartier/Tate dual coefficients, the local field's cohomological dimension and the Brauer invariant. The named local theorem fails the prime bar because its finite-module, local-field and degree-two tests do not remain valid in unrelated domains. A general perfect-pairing pattern does not by itself license a parent edge. Local Field is a strict composition prerequisite for the stated p-adic theorem, not its taxonomic genus.

Instantiates / Related Primes

This entry presupposes Local field.

Duality, complementarity and invariant pairings may be broad prime themes. The strict structural link to Local Field records the necessary p-adic carrier under the finite-module statement, without claiming that every local field satisfies this theorem or that the theorem is a field.

Relationships to Other Abstractions

Current abstraction Local Tate Duality Domain-specific

Parents (1) — more general patterns this builds on

• Local Tate Duality presupposes Local field Domain-specific

  The stated local Galois-cohomology pairing requires a p-adic local field.

Hierarchy paths (6) — routes to 5 parentless roots

• Local Tate Duality → Local field → Locality Of Reference → Recurrence

Show alternative paths (5)

Neighborhood in Abstraction Space

Local Tate Duality sits in a sparse region of the domain-specific corpus (62^nd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Sheaves & Birational Geometry (22 abstractions)

Nearest neighbors

• Structure Theorem for Finitely Generated Modules over a Principal Ideal Domain — 0.85
• Zariski Tangent Space — 0.85
• McKay Graph — 0.84
• Cohomological dimension — 0.84
• Sheaf of Modules — 0.84

Computed from structural-signature embeddings · 2026-10-08

Not to Be Confused With

Poitou–Tate duality is global and relates local terms through global exact sequences. Tate local duality for abelian varieties is a related formulation with different objects and degree relation. Cartier duality names the coefficient-side dual operation, not the whole cohomological perfect-pairing theorem. A Hilbert pairing can instantiate a local degree-one phenomenon but is not the full \(r=0,1,2\) statement.^[1]

References

[1] J. S. Milne, “Arithmetic Duality”, §1 (2026), later expert historical exposition of Tate's local finite-module theorem and its Brauer target, not Tate's original 1957 source. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n

[2] Specialist Galois-representation notes: local duality and annihilator conditions, theorem and application section. registry ↩a ↩b ↩c ↩d ↩e ↩f