Hecke character explained

In number theory, a Hecke character is a generalisation of a Dirichlet character, introduced by Erich Hecke to construct a class ofL-functions larger than Dirichlet L-functions, and a natural setting for the Dedekind zeta-functions and certain others which have functional equations analogous to that of the Riemann zeta-function.

Definition

A Hecke character is a character of the idele class group of a number field or global function field. It corresponds uniquely to a character of the idele group which is trivial on principal ideles, via composition with the projection map.

This definition depends on the definition of a character, which varies slightly between authors: It may be defined as a homomorphism to the non-zero complex numbers (also called a "quasicharacter"), or as a homomorphism to the unit circle in

C

("unitary"). Any quasicharacter (of the idele class group) can be written uniquely as a unitary character times a real power of the norm, so there is no big difference between the two definitions.

The conductor of a Hecke character

\chi

is the largest ideal

ak{m}

such that

\chi

is a Hecke character mod

ak{m}

. Here we say that

\chi

is a Hecke character mod

ak{m}

if

\chi

(considered as a character on the idele group) is trivial on the group of finite ideles whose every

\nu

-adic component lies in

1+ak{m}O\nu

.

GrĂ¶ĂŸencharakter

K

, let

ak{m}=ak{m}fak{m}infty

be a

K

-modulus, with

ak{m}f

, the "finite part", being an integral ideal of

K

and

ak{m}infty

, the "infinite part", being a (formal) product of real places of

K

. Let

Iak{m}

denote the group of fractional ideals of

K

relatively prime to

ak{m}f

andlet

Pak{m}

denote the subgroup of principal fractional ideals

(a)

where

a

is near

1

at each place of

ak{m}

in accordance with the multiplicities ofits factors. That is, for each finite place

\nu

in

ak{m}f

, the order

ord\nu(a-1)

is at least as large as the exponent for

\nu

in

ak{m}f

, and

a

is positive under each real embedding in

ak{m}infty

. A GrĂ¶ĂŸencharakter with modulus

ak{m}

is a group homomorphism from

Iak{m}

into the nonzero complex numbers such that on ideals

(a)

in

Pak{m}

its value is equal to thevalue at

a

of a continuous homomorphism to the nonzero complex numbers from the product of the multiplicative groups of all Archimedean completions of

K

where each local component of the homomorphism has the same real part (in the exponent). (Here we embed

a

into the product of Archimedean completions of

K

using embeddings corresponding to the various Archimedean places on

K

.) Thus a GrĂ¶ĂŸencharakter may be defined on the ray class group modulo

ak{m}

, which is the quotient

Iak{m}/Pak{m}

.

Strictly speaking, Hecke made the stipulation about behavior on principal ideals for those admitting a totally positive generator. So, in terms of the definition given above, he really only worked with moduli where all real places appeared.The role of the infinite part m∞ is now subsumed under the notion ofan infinity-type.

Relationship between GrĂ¶ĂŸencharakter and Hecke character

A Hecke character and a GrĂ¶ĂŸencharakter are essentially the same notion with a one-to-one correspondence. The ideal definition is much more complicated than the idelic one, and Hecke's motivation for his definition was to construct L-functions (sometimes referred to as Hecke L-functions)[1] that extend the notion of a Dirichlet L-function from the rationals to other number fields. For a GrĂ¶ĂŸencharakter χ, its L-function is defined to be the Dirichlet series

\sum(I,m)=1\chi(I)N(I)-s=L(s,\chi)

carried out over integral ideals relatively prime to the modulus

ak{m}

of the GrĂ¶ĂŸencharakter.Here

N(I)

denotes the ideal norm. The common real part condition governing the behavior of GrĂ¶ĂŸencharakter on the subgroups

Pak{m}

implies theseDirichlet series are absolutely convergent in some right half-plane. Hecke proved these L-functions have a meromorphic continuation to the whole complex plane, being analytic except for a simple pole of order 1 at '

s=1

when the character is trivial. For primitive GrĂ¶ĂŸencharakter (defined relative to a modulus in a similar manner to primitive Dirichlet characters), Hecke showed these L-functions satisfy a functional equation relating the values of the L-function of a character and the L-function of its complex conjugate character.

Consider a character

\psi

of the idele class group, taken to be a map into the unit circle which is 1 on principal ideles and on an exceptional finite set

S

containing all infinite places. Then

\psi

generates a character

\chi

of the ideal group

IS

, which is the free abelian group on the prime ideals not in

S

.[2] Take a uniformising element

\pi

for each prime

ak{p}

not in

S

and define a map

\Pi

from

IS

to idele classes by mapping each

ak{p}

to the class of the idele which is

\pi

in the

ak{p}

coordinate and

1

everywhere else. Let

\chi

be the composite of

\Pi

and

\psi

. Then

\chi

is well-defined as a character on the ideal group.[3]

In the opposite direction, given an admissible character

\chi

of

IS

there corresponds a unique idele class character

\psi

.[4] Here admissible refers to the existence of a modulus

ak{m}

based on the set

S

such that the character

\chi

evaluates to

1

on the ideals which are 1 mod

ak{m}

.[5]

The characters are 'big' in the sense that the infinity-type when present non-trivially means these characters are not of finite order. The finite-order Hecke characters are all, in a sense, accounted for by class field theory: their L-functions are Artin L-functions, as Artin reciprocity shows. But even a field as simple as the Gaussian field has Hecke characters that go beyond finite order in a serious way (see the example below). Later developments in complex multiplication theory indicated that the proper place of the 'big' characters was to provide the Hasse–Weil L-functions for an important class of algebraic varieties (or even motives).

Special cases

Examples

R+

with all the unit groups of the p-adic integers. So a quasicharacter can be written as product of a power of the norm with a Dirichlet character.

χ((a)) = |a|s(a/|a|)4n

for s imaginary and n an integer, where a is a generator of the ideal (a). The only units are powers of i, so the factor of 4 in the exponent ensures that the character is well defined on ideals.

Tate's thesis

See main article: Tate's thesis. Hecke's original proof of the functional equation for L(s,χ) used an explicit theta-function. John Tate's 1950 Princeton doctoral dissertation, written under the supervision of Emil Artin, applied Pontryagin duality systematically, to remove the need for any special functions. A similar theory was independently developed by Kenkichi Iwasawa which was the subject of his 1950 ICM talk. A later reformulation in a Bourbaki seminar by showed that parts of Tate's proof could be expressed by distribution theory: the space of distributions (for Schwartz–Bruhat test functions) on the adele group of K transforming under the action of the ideles by a given χ has dimension 1.

Algebraic Hecke characters

An algebraic Hecke character is a Hecke character taking algebraic values: they were introduced by Weil in 1947 under the name type A0. Such characters occur in class field theory and the theory of complex multiplication.[6]

Indeed let E be an elliptic curve defined over a number field F with complex multiplication by the imaginary quadratic field K, and suppose that K is contained in F. Then there is an algebraic Hecke character χ for F, with exceptional set S the set of primes of bad reduction of E together with the infinite places. This character has the property that for a prime ideal p of good reduction, the value χ(p) is a root of the characteristic polynomial of the Frobenius endomorphism. As a consequence, the Hasse–Weil zeta function for E is a product of two Dirichlet series, for χ and its complex conjugate.[7]

References

Notes and References

  1. As in
  2. Heilbronn (1967) p.204
  3. Heilbronn (1967) p. 205
  4. Tate (1967) p.169
  5. Heilbronn (1967) p.207
  6. Husemoller (1987) pp. 299–300; (2002) p.320
  7. Husemoller (1987) pp. 302–303; (2002) pp. 321–322