Professional Documents
Culture Documents
Darmon Diamond Taylor Fermats Last Theorem
Darmon Diamond Taylor Fermats Last Theorem
Henri Darmon
([email protected])
Department of Mathematics
McGill University
Montreal, QC
Canada H3A 2K6
Fred Diamond
([email protected])
D.P.M.M.S.
Cambridge University
Cambridge, CB2 1SB
United Kingdom
Richard Taylor
([email protected])
Mathematics Institute
Oxford University
24-29 St. Giles
Oxford, OX1 3LB
United Kingdom
May 24, 2000
The authors would like to give special thanks to N. Boston, K. Buzzard, and
B. Conrad for providing so much valuable feedback on earlier versions of this
paper. They are also grateful to A. Agboola, M. Bertolini, B. Edixhoven, J.
Fearnley, R. Gross, L. Guo, F. Jarvis, H. Kisilevsky, E. Liverance, J. Manoharmayum, K. Ribet, D. Rohrlich, M. Rosen, R. Schoof, J.-P. Serre, C. Skinner,
D. Thakur, J. Tilouine, J. Tunnell, A. Van der Poorten, and L. Washington
for their helpful comments.
Darmon thanks the members of CICMA and of the Quebec-Vermont Number Theory Seminar for many stimulating conversations on the topics of this
paper, particularly in the Spring of 1995. For the same reason Diamond is
grateful to the participants in an informal seminar at Columbia University
in 1993-94, and Taylor thanks those attending the Oxford Number Theory
Seminar in the Fall of 1995.
1
Parts of this paper were written while the authors held positions at other
institutions: Darmon at Princeton University, Diamond at the Institute for
Advanced Study, and Taylor at Cambridge University. During some of the period, Diamond enjoyed the hospitality of Princeton University, and Taylor that
of Harvard University and MIT. The writing of this paper was also supported
by research grants from NSERC (Darmon), NSF # DMS 9304580 (Diamond)
and by an advanced fellowship from EPSRC (Taylor).
This article owes everything to the ideas of Wiles, and the arguments presented here are fundamentally his [W3], though they include both the work
[TW] and several simplifications to the original arguments, most notably that
of Faltings. In the hope of increasing clarity, we have not always stated
theorems in the greatest known generality, concentrating instead on what is
needed for the proof of the Shimura-Taniyama conjecture for semi-stable elliptic curves. This article can serve as an introduction to the fundamental papers
[W3] and [TW], which the reader is encouraged to consult for a different, and
often more in-depth, perspective on the topics considered. Another useful
more advanced reference is the article [Di2] which strengthens the methods of
[W3] and [TW] to prove that every elliptic curve that is semistable at 3 and 5
is modular.
Introduction
Fermats Last Theorem
Fermats Last Theorem states that the equation
xn + y n = z n ,
xyz 6= 0
Peter Lejeune Dirichlet [Dir] and Adrien Marie Legendre [Leg] in 1825. Their
elementary arguments are quite involved. (Cf. [Edw], sec. 3.3.)
In 1839, Fermats equation for exponent 7 also yielded to elementary methods, through the heroic efforts of Gabriel Lame. Lames proof was even more
intricate than the proof for exponent 5, and suggested that to go further, new
theoretical insights would be needed.
The work of Sophie Germain: Around 1820, in a letter to Gauss, Sophie
Germain proved that if ` is a prime and q = 2` + 1 is also prime, then Fermats
equation x` + y ` = z ` with exponent ` has no solutions (x, y, z) with xyz 6= 0
(mod `). Germains theorem was the first really general proposition on Fermats Last Theorem, unlike the previous results which considered the Fermat
equation one exponent at a time.
The case where the solution (x, y, z) to x` + y ` = z ` satisfies xyz 6= 0 (mod
`) was called the first case of Fermats Last Theorem, and the case where `
divides xyz, the second case. It was realized at that time that the first case
was generally easier to handle: Germains theorem was extended, using similar
ideas, to cases where k` + 1 is prime and k is small, and this led to a proof that
there were no first case solutions to Fermats equation with prime exponents
` 100, which in 1830 represented a significant advance. The division between
first and second case remained fundamental in much of the later work on the
subject. In 1977, Terjanian [Te] proved that if the equation x2` + y 2` = z 2` has
a solution (x, y, z), then 2` divides either x or y, i.e., the first case of Fermats
Last Theorem is true for even exponents. His simple and elegant proof used
only techniques that were available to Germain and her contemporaries.
The work of Kummer: The work of Ernst Eduard Kummer marked the
beginning of a new era in the study of Fermats Last Theorem. For the first
time, sophisticated concepts of algebraic number theory and the theory of
L-functions were brought to bear on a question that had until then been
addressed only with elementary methods. While he fell short of providing
a complete solution, Kummer made substantial progress. He showed how
Fermats Last Theorem is intimately tied to deep questions on class numbers
of cyclotomic fields which are still an active subject of research. Kummers
approach relied on the factorization
(x + y)(x + ` y) (x + ``1 y) = z `
of Fermats equation over the ring Z[` ] generated by the `th roots of unity.
One observes that the greatest common divisor of any two factors in the product on the left divides the element (1 ` ), which is an element of norm `.
4
Since the product of these numbers is a perfect `-th power, one is tempted to
conclude that (x + y), . . . , (x + ``1 y) are each `-th powers in the ring Z[` ] up
to units in this ring, and up to powers of (1 ` ). Such an inference would be
valid if one were to replace Z[` ] by Z, and is a direct consequence of unique
factorization of integers into products of primes. We say that a ring R has
property U F if every non-zero element of R is uniquely a product of primes,
up to units. Mathematicians such as Lame made attempts at proving Fermats Last Theorem based on the mistaken assumption that the rings Z[` ]
had property U F . Legend even has it that Kummer fell into this trap, although this story now has been discredited; see for example [Edw], sec. 4.1. In
fact, property U F is far from being satisfied in general: one now knows that
the rings Z[` ] have property U F only for ` < 23 (cf. [Wa], ch. 1).
It turns out that the full force of property U F is not really needed in the
applications to Fermats Last Theorem. Say that a ring R has property U F`
if the following inference is valid:
ab = z ` , and gcd(a, b) = 1 a and b are `th powers up to units of R.
If a ring R has property U F , then it also has property U F` , but the converse
need not be true. Kummer showed that Fermats last theorem was true for
exponent ` if Z[` ] satisfied the property U F` (cf. [Wa]). The proof is far from
trivial, because of difficulties arising from the units in Z[` ] as well as from
the possible failure of property U F . (A number of Kummers contemporaries,
such as Cauchy and Lame, seem to have overlooked both of these difficulties
in their attempts to prove Fermats Last Theorem.)
Kummer then launched a systematic study of the property U F` for the
rings Z[` ]. He showed that even if Z[` ] failed to have unique factorization,
it still possessed unique factorization into prime ideals. He defined the ideal
class group as the quotient of the group of fractional ideals by its subgroup
consisting of principal ideals, and was able to establish the finiteness of this
class group. The order of the class group of Z[` ], denoted h` , could be taken
as a measure of the failure of the ring Z[` ] to satisfy U F . It was rather
straightforward to show that if ` did not divide h` , then Z[` ] satisfied the
property U F` . In this case, one called ` a regular prime. Kummer thus showed
that Fermats last theorem is true for exponent ` if ` is a regular prime.
He did not stop here. For it remained to give an efficient means of computing h` , or at least an efficient way of checking when ` divides h` . The class
number h` can be factorized as a product
h` = h +
` h` ,
+
where h+
` is the class number of the real subfield Q(` ) , and h` is defined as
+
+
+
h` /h` . Essentially because of the units in Q(` ) , the factor h` is somewhat
difficult to compute, while, because the units in Q(` )+ generate the group of
units in Q(` ) up to finite index, the term h
` can be expressed in a simple
L(s, ) =
(n)ns
n=1
There were many other refinements of similar criteria for Fermats Last
theorem to be true. Computer calculations based on these criteria led to a
verification that Fermats Last theorem is true for all odd prime exponents less
than four million [BCEM], and that the first case is true for all ` 8.858 10 20
[Su].
The condition that ` is a regular prime seems to hold heuristically for about
61% of the primes. (See the discussion on p. 63, and also p. 108, of [Wa], for
example.) In spite of the convincing numerical evidence, it is still not known
if there are infinitely many regular primes. Ironically, it is not too difficult to
show that there are infinitely many irregular primes. (Cf. [Wa].)
Thus the methods introduced by Kummer, after leading to very strong
results in the direction of Fermats Last theorem, seemed to become mired in
difficulties, and ultimately fell short of solving Fermats conundrum1 .
Faltings proof of the Mordell conjecture: In 1985, Gerd Faltings [Fa]
proved the very general statement (which had previously been conjectured
by Mordell) that any equation in two variables corresponding to a curve of
genus strictly greater than one had (at most) finitely many rational solutions.
In the context of Fermats Last Theorem, this led to the proof that for each
exponent n 3, the Fermat equation xn + y n = z n has at most finitely many
integer solutions (up to the obvious rescaling). Andrew Granville [Gra] and
Roger Heath-Brown [HB] remarked that Faltings result implies Fermats Last
Theorem for a set of exponents of density one.
However, Fermats Last Theorem was still not known to be true for an
infinite set of prime exponents. In fact, the theorem of Faltings seemed illequipped for dealing with the finer questions raised by Fermat in his margin,
namely of finding a complete list of rational points on all of the Fermat curves
xn + y n = 1 simultaneously, and showing that there are no solutions on these
curves when n 3 except the obvious ones.
Mazurs work on Diophantine properties of modular curves: Although
it was not realized at the time, the chain of ideas that was to lead to a proof
of Fermats Last theorem had already been set in motion by Barry Mazur
in the mid seventies. The modular curves X0 (`) and X1 (`) introduced in
section 1.2 and 1.5 give rise to another naturally occurring infinite family
of Diophantine equations. These equations have certain systematic rational
solutions corresponding to the cusps that are defined over Q, and are analogous
1
finite field F` with ` elements, and the absolute Galois group GQ = Gal (Q/Q)
acts F` -linearly on E[`]. Choosing an F` -basis for E[`], the action is described
by a representation
E,` : Gal (L/Q) , GL2 (F` ).
Mazurs results in [Maz1] and [Maz2] imply that E,` is irreducible if ` > 7
(using the fact that E is semi-stable). In fact, combined with earlier results
of Serre [Se6], Mazurs results imply that for ` > 7, the representation E,` is
surjective, so that Gal (L/Q) is actually isomorphic to GL2 (F` ) in this case.
Serres conjectures: In [Se7], Jean-Pierre Serre made a careful study of mod
` Galois representations : GQ GL2 (F` ) (and, more generally, of representations into GL2 (k), where k is any finite field). He was able to make very
precise conjectures (see section 3.2) relating these representations to modular
forms mod `. In the context of the representations E,` that occur in Freys
construction, Serres conjecture predicted that they arose from modular forms
(mod `) of weight two and level two. Such modular forms, which correspond to
differentials on the modular curve X0 (2), do not exist because X0 (2) has genus
0. Thus Serres conjecture implied Fermats Last Theorem. The link between
fields with Galois groups contained in GL2 (F` ) and modular forms mod ` still
appears to be very deep, and Serres conjecture remains a tantalizing open
problem.
Ribets work: lowering the level: The conjecture of Shimura and Taniyama (cf. section 1.8) provides a direct link between elliptic curves and modular
forms. It predicts that the representation E,` obtained from the `-division
points of the Frey curve arises from a modular form of weight 2, albeit a form
whose level is quite large. (It is the product of all the primes dividing abc,
where a` + b` = c` is the putative solution to Fermats equation.) Ribet [R5]
proved that, if this were the case, then E,` would also be associated with a
modular form mod ` of weight 2 and level 2, in the way predicted by Serres
conjecture. This deep result allowed him to reduce Fermats Last Theorem to
the Shimura-Taniyama conjecture.
Wiles work: proof of the Shimura-Taniyama conjecture: In [W3]
Wiles proves the Shimura-Taniyama conjecture for semi-stable elliptic curves,
providing the final missing step and proving Fermats Last Theorem. After
more than 350 years, the saga of Fermats Last theorem has come to a spectacular end.
The relation between Wiles work and Fermats Last Theorem has been
very well documented (see, for example, [R8], and the references contained
therein). Hence this article will focus primarily on the breakthrough of Wiles
[W3] and Taylor-Wiles [TW] which leads to the proof of the Shimura-Taniyama
conjecture for semi-stable elliptic curves.
From elliptic curves to `-adic representations: Wiles opening gambit
for proving the Shimura-Taniyama conjecture is to view it as part of the more
9
general problem of relating two-dimensional Galois representations and modular forms. The Shimura-Taniyama conjecture states that if E is an elliptic
curve over Q, then E is modular. One of several equivalent definitions
of modP
n
ularity is that for some integer N there is an eigenform f =
an q of weight
two on 0 (N ) such that
#E(Fp ) = p + 1 ap
for all but finitely primes p. (By an eigenform, here we mean a cusp form
which is a normalized eigenform for the Hecke operators; see section 1 for
definitions.)
This conjecture acquires a more Galois theoretic flavour when one considers
the two dimensional `-adic representation
E,` : GQ GL2 (Z` )
obtained from the action of GQ on the `-adic Tate module of E: T` E =
An `-adic representation of GQ is said to arise from an eigenlim E[ln ](Q).
P
form f =
an q n with integer coefficients an if
tr ((Frob p )) = ap ,
for all but finitely many primes p at which is unramified. Here Frob p is a
Frobenius element at p (see section 2), and its image under is a well-defined
conjugacy class.
A direct computation shows that #E(Fp ) = p + 1 tr (E,` (Frob p )) for
all primes p at which E,` is unramified, so that E is modular (in the sense
defined above) if and only if for some `, E,` arises from an eigenform. In
fact the Shimura-Taniyama conjecture can be generalized to a conjecture that
every `-adic representation, satisfying suitable local conditions, arises from a
modular form. Such a conjecture was proposed by Fontaine and Mazur [FM].
that was introduced before). Given a holomorphic newform f one can attach
to f a system of `-adic representations, following Eichler, Shimura, Deligne and
Serre. These `-adic representations are called modular. The Fontaine-Mazur
conjecture (see [FM]) predicts if is an odd, irreducible, `-adic representation
whose restriction to the decomposition group at ` is well enough behaved,
then is modular. (The restriction on the behaviour of the representation
on the decomposition group at ` is essential in this conjecture; it is not true
that all odd, irreducible two dimensional `-adic representation are modular.)
Before Wiles work almost nothing was known about this conjecture, except
that certain very special cases could be deduced from the work of Hecke,
Langlands and Tunnell.
Mod ` representations and Serres conjecture: A mod ` representation
` ). For example if E/Q is an
is a continuous representation : GQ GL2 (F
elliptic curve then the action of GQ on the `-division points of E gives rise to a
mod ` representation E,` which is just the reduction modulo ` of E,` . One can
use the work of Eichler, Shimura, Deligne and Serre to associate to each mod
` eigenform a mod ` representation of GQ . The mod ` representations which
arise in this way are called modular. Serre has conjectured [Se7] that every
odd (absolutely) irreducible mod ` representation is modular and should arise
from a mod ` eigenform with certain very specific properties. This conjecture
can be thought of as having two parts.
The first asserts that every odd irreducible mod ` representation is modular.
About this very little is known. It is known for : GQ GL2 (F2 ) by work
of Hecke. It is also known for : GQ GL2 (F3 ). This latter result is an
application of the Langlands-Tunnell theorem using
the two accidents that
Maps of this kind were already considered in [Maz3] and [BM], and it is conjectured in
[MT] that these maps are isomorphisms in certain cases, though not in exactly the situations
considered by Wiles.
13
ficulty, and the real challenge is to prove injectivity, i.e., to show, in essence,
that R is not larger than T .
By an ingenious piece of commutative algebra, Wiles found a numerical
criterion for this map to be an isomorphism, and for the ring T to be a
local complete intersection. This numerical criterion seems to be very close
to a special case of the Bloch-Kato conjecture [BK]. Wiles further showed
(by combining arguments from Galois cohomology and from the theory of
congruences between modular forms) that this numerical criterion was satisfied
if the minimal version T of this Hecke algebra (obtained by taking = , i.e.,
allowing the least possible amount of ramification in the deformations) was a
complete intersection. Finally in [TW] it was proved that T is a complete
intersection.
14
Contents
1 Elliptic curves and modular forms
1.1 Elliptic curves . . . . . . . . . . . . . . . .
1.2 Modular curves and modular forms over C
1.3 Hecke operators and Hecke theory . . . . .
1.4 The L-function associated to a cusp form .
1.5 Modular curves and modular forms over Q
1.6 The Hecke algebra . . . . . . . . . . . . .
1.7 The Shimura construction . . . . . . . . .
1.8 The Shimura-Taniyama conjecture . . . .
.
.
.
.
.
.
.
.
2 Galois theory
2.1 Galois representations . . . . . . . . . . . .
2.2 Representations associated to elliptic curves
2.3 Galois cohomology . . . . . . . . . . . . . .
2.4 Representations of GQ` . . . . . . . . . . . .
2.5 The theory of Fontaine and Laffaille . . . . .
2.6 Deformations of representations . . . . . . .
2.7 Deformations of Galois representations . . .
2.8 Special cases . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
16
16
22
28
33
34
39
44
47
.
.
.
.
.
.
.
.
50
50
55
59
62
69
73
76
80
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
84
84
88
93
98
100
102
4 Hecke algebras
4.1 Full Hecke algebras . . .
4.2 Reduced Hecke algebras
4.3 Proof of theorem 3.31 . .
4.4 Proof of theorem 3.36 . .
4.5 Homological results . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
106
106
111
119
124
133
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5 Commutative algebra
137
5.1 Wiles numerical criterion . . . . . . . . . . . . . . . . . . . . 138
5.2 Basic properties of A and A . . . . . . . . . . . . . . . . . . 140
15
5.3
5.4
5.5
5.6
5.7
5.8
5.9
5.10
1
1.1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
143
148
149
152
153
153
154
155
We begin with a brief review of elliptic curves. A general reference for the
results discussed in this section is [Si1] and [Si2].
An elliptic curve E over a field F is a proper smooth curve over F of genus
one with a distinguished F -rational point. If E/F is an elliptic curve and if
is a non-zero holomorphic differential on E/F then E can be realised in the
projective plane by an equation (called a Weierstrass equation) of the form
(W )
Y 2 Z + a1 XY Z + a3 Y Z 2 = X 3 + a2 X 2 Z + a4 XZ 2 + a6 Z 3
One can check that the equation (W ) defines an elliptic curve if and only if
is nonzero. One can also check that such equations define elliptic curves
which are isomorphic over F if and only if they give the same quantity j. Thus
j only depends on E so we will denote it jE . The quantity depends only
on the pair (E, ) so we shall denote it (E, ). If u belongs to F then
u12 (E, u) = (E, ).
An elliptic curve E/F has a natural structure of a commutative algebraic
group with the distinguished F -rational point as the identity element.
An algebraic map between two elliptic curves which sends the distinguished
point of one to the distinguished point of the other is automatically a morphism
16
of algebraic groups. A map between elliptic curves which has finite kernel (and
hence, is generically surjective) is called an isogeny.
Elliptic curves over C: If F = C, then the curve E is isomorphic as a
complex analytic manifold to the complex torus C/, where is a lattice
in C, i.e., a discrete Z-submodule of C of rank 2. The group law on E(C)
corresponds to the usual addition in C/. In terms of , an affine equation
for E in A2 (C) is given by
y 2 = 4x3 + g2 x + g3 ,
where
g2 = 60
{0}
1
,
z4
g3 = 140
{0}
1
.
z6
(1.1.1)
The following basic facts are a direct consequence of the analytic theory:
17
Proposition 1.1 The subgroup E[n](C) of points of order n on E(C) is isomorphic (non-canonically) to Z/nZ Z/nZ. More generally, if F is any field
of characteristic zero, the subgroup E[n](F ) is contained in Z/nZ Z/nZ.
Proof: The analytic theory shows that E(C) is isomorphic as an abstract group
to a product of two circle groups, and the first statement follows. The second
statement follows from the Lefschetz principle (cf. [Si1], ch. VI, 6).
2
Proposition 1.2 The endomorphism ring End C (E) of an elliptic curve over
C is isomorphic either to Z or to an order in a quadratic imaginary field. The
same is true if one replaces C by any field of characteristic 0.
Proof: An endomorphism of E(C) ' C/ induces multiplication by complex
number on the tangent space. Hence End C (E) is isomorphic to the ring
of C satisfying . Such a ring is isomorphic either to Z or to a
quadratic imaginary order. The corresponding statement for fields of characteristic 0 follows as in the proof of proposition 1.1.
2
If End C (E) Q is a quadratic imaginary field, we say that E has complex
multiplication.
Remark 1.3 It follows from the arithmetic theory of complex multiplication
(cf. [Si2], ch. 1) that any elliptic curve E with complex multiplication is defined
over an abelian extension of the quadratic imaginary field K = End C (E) Q.
If E is defined over Q, then K has class number one. There are only finitely
many elliptic curves over Q with complex multiplication, up to twists (i.e.,
C-isomorphism).
Elliptic curves over Qp : Now suppose that E is an elliptic curve defined
over the p-adic field Qp . There is an equation
(W min )
Y 2 Z + a1 XY Z + a3 Y Z 2 = X 3 + a2 X 2 Z + a4 XZ 2 + a6 Z 3
for E with the property ai Zp for all i and || is minimal amongst all such
equations for E. Although (W min ) is not unique, the associated discriminant
min
depends only on E and is denoted min
)
E . Moreover the reduction of (W
(1.1.2)
19
(1.1.3)
GQp ,
where : GQp 1 is
the trivial character, if E has split multiplicative reduction;
the unique unramified quadratic character of GQp , if E has non-split
multiplicative reduction.
The proof of this proposition is explained in [Si2], ch. V, for example.
We define the L-function L(E/Qp , s) to be
(1.1.4)
(1.1.5)
20
1 n 10,
Z/2nZ Z/2Z,
n = 12,
1 n 4.
The proof is given in [Maz1] (see also [Maz2]). Thanks to this result, the
structure of the torsion subgroup T is well understood. (Recently, the techniques of Mazur have been extended by Kamienny [Kam] and Merel [Mer] to
prove uniform boundedness results on the torsion of elliptic curves over general
number fields.)
Much more mysterious is the behaviour of the rank r. It is not known if r
can be arbitrarily large, although results of Mestre [Mes] and Nagao [Na] show
that it is greater or equal to 13 for infinitely many elliptic curves over Q. It
turns out that many of the deep results on E(Q) and on r are based on the
relation with L-functions.
We define the global L-function of the complex variable s by:
Y
L(E/Q, s) =
L(E/Qp , s).
(1.1.6)
p
21
Exercise 1.8 Using theorem 1.4, show that the infinite product defining the
L-function L(E/Q, s) converges absolutely on the right half plane Real(s) >
3/2.
Conjecture 1.9 (Birch-Swinnerton-Dyer) The L-function L(E/Q, s) has
an analytic continuation to the entire complex plane,and in particular is analytic at s = 1. Furthermore:
ords=1 L(E/Q, s) = r.
There is also a more precise form of this conjecture, which expresses the leading
coefficient of L(E/Q, s) at s = 1 in terms of certain arithmetic invariants of
E/Q. For more details, see [Si1], conj. 16.5.
As we will explain in more detail in section 1.8, the analytic continuation
of L(E/Q, s) now follows from the work of Wiles and Taylor-Wiles and a
strengthening by Diamond [Di2], for a very large class of elliptic curves over
Q, which includes all the semi-stable ones.
Abelian varieties: Elliptic curves admit higher-dimensional analogues, called
abelian varieties, which also play a role in our discussion. Analytically, the set
of complex points on an abelian variety is isomorphic to a quotient Cg /,
where is a lattice in Cg of rank 2g, satisfying the so-called Riemann period
relations. A good introduction to the basic theory of abelian varieties can be
found in [CS] and [We1].
1.2
Modular curves: The group SL2 (Z) of two by two integer matrices of determinant one acts by fractional linear (Mobius) transformations on the complex
upper half plane
H = {z C | Im (z) > 0},
equipped with its standard complex analytic structure. The principal congruence group (N ) of level N is the subgroup of matrices in SL2 (Z) which
reduce to the identity matrix modulo the positive integer N . A subgroup of
SL2 (Z) is called a congruence group if it contains (N ) for some N . The level
of is the smallest N for which this is true. The most important examples of
congruence groups are:
The group 0 (N ) consisting of all matrices that reduce modulo N to an
upper triangular matrix.
22
a b
7 d mod N.
c d
For any subgroup H of (Z/N Z) , we denote by H (N ) the group of matrices
in 0 (N ) whose image in 0 (N )/1 (N ) belongs to H.
If is a congruence subgroup of SL2 (Z), define Y to be the quotient of
the upper half plane H by the action of . One equips Y with the analytic
structure coming from the projection map : H Y . (More precisely, if
y = ( ), and G is the stabilizer of in , then the local ring OY ,y is
identified with the local ring of germs of holomorphic functions at which are
invariant under the action of G .) This makes Y into a connected complex
analytic manifold of dimension one, i.e., a Riemann surface. If is 0 (N ) (resp.
1 (N ), or (N )), we will also denote Y by Y0 (N ) (resp. Y1 (N ), or Y (N )). One
compactifies Y by adjoining a finite set of cusps which correspond to orbits
of P1 (Q) = Q {} under . Call X the corresponding compact Riemann
surface. (For more details, notably on the definition of the analytic structure
on X at the cusps, see for example [Kn], p. 311, or [Shi2], ch. 1.) It follows
from the definition of this analytic structure that the field K of meromorphic
functions on X is equal to the set of meromorphic functions on H satisfying
(Transformation property): f ( ) = f ( ), for all ;
(Behaviour at the cusps):P
For all SL2 (Z), the function f ( ) has a
n/h
Puiseux series expansion
in fractional powers of q = e2i .
m an q
Riemanns existence theorem (cf. for example [For], ch. 2) asserts that the
analytic structure on X comes from an algebraic one, i.e., the field K is a
finitely generated extension of C of transcendence degree 1. Thus we can, and
will, view X as a complex algebraic curve over C. If is 0 (N ) (resp. 1 (N ),
or (N )), we will also denote X by X0 (N ) (resp. X1 (N ), or X(N )).
Examples and exercises:
23
q = e2i .
of in H,
Using the
Riemann-Hurwitz formula (cf. [Ki], sec. 4.3) show that the genus of X is
given by
1 X
g() = 1 [P SL2 (Z) : ] +
(e 1).
2
H/
Use this to compute the genus of X0 (p), X1 (p), and X(p) for p prime. For
details, see [Shi2], sec. 1.6 or [Ogg].
3. For = (2), show that X is isomorphic to P1 , and that Y is isomorphic
to P1 {0, 1, }. Show
that
/h1i is the free group on the two generators
1
2
1
0
g1 = 0 1 and g2 = 2 1 .
projection Y1 (N ) Y0 (N ) sending 1 (N ) to 0 (N ) becomes the forgetful map sending (E, P ) to (E, hP i).
Remark 1.11 (This remark will be used in section 1.3 when discussing Hecke
operators.) Define an n-isogeny of -structures to be an n-isogeny of the
underlying elliptic curves which sends one -structure to the other. If p is a
prime not dividing N , then there are exactly p + 1 distinct p-isogenies from
(C/h, 1i, N1 ), whose images are the pairs:
p
1
+i
(i = 0, . . . , p 1),
C/hp, 1i,
.
,1 ,
C/
p
N
N
If p divides N , then there are only p distinct p-isogenies from (C/h, 1i, N1 ),
since (C/hp, 1i, Np ) is not a 1 (N )-structure (the point p/N not being of exact
order N on the complex torus C/hp, 1i).
Modular forms: Let k be an even positive integer. A modular form of weight
k on is a holomorphic function f on H satisfying:
n/h
d)k f ( ) has a Puiseux
series
expansion
in fractional powers
0 an q
P
2i
n/h
of q = e . We call
an q
the Fourier expansion of f at the cusp
1 (i).
A modular form which satisfies the stronger property that the constant
coefficient of its Fourier expansion at each cusp vanishes is called a cusp form.
We denote by Mk () the complex vector space of modular forms of weight k
on , and by Sk () the space of cusp forms on . (For examples, see [DI],
sec. 2.2 and the references therein, especially, [Shi2], ch. 2.)
This article is mainly concerned with modular forms of weight 2, and hence
we will focus our attention on these from now on. A pleasant feature of the case
k = 2 is that the cusp forms in S2 () admit a direct geometric interpretation
as holomorphic differentials on the curve X .
Lemma 1.12 The map f ( ) 7 f := 2if ( )d is an isomorphism between
the space S2 () and the space 1 (X ) of holomorphic differentials on the curve
X .
25
The Petersson inner product: The spaces S2 () are also equipped with a
natural Hermitian inner product given by
Z
Z
i
f
g =
f ( )
g ( )dxdy,
hf, gi = 2
8 X
H/
where = x + iy. This is called the Petersson inner product.
The diamond operators: Suppose now that = 1 (N ) and let d be an element of (Z/N Z) . The map hdi which sends an elliptic curve with -structure
(E, P ) to the pair (E, dP ) gives an automorphism of Y which
extends to X .
a
b
It is called the diamond operator. For in H and = c d in 0 (N ), we
have
hdi( ) = ( ).
26
a + b
2
.
hdif ( ) = (c + d) f
c + d
a + b
a
b
2
f
= (d)(c + d) f ( ), for all c d 0 (N ).
c + d
Note that if 1 is the trivial character, then S2 (N, 1) is canonically identified
with S2 (0 (N )), which we will also denote by S2 (N ). Finally note the direct
sum decomposition:
S2 (1 (N )) = S2 (N, ),
where the sum ranges over all the even Dirichlet characters modulo N .
Exercise 1.14 Show that if f ( ) belongs to S2 (N ), then f (a ) belongs to
S2 (mN ), for each integer a dividing m.
Jacobians of modular curves: Let V be the dual space
V = S2 () := Hom(S2 (), C).
It is a complex vector space of dimension g = genus(X ). The integral homology = H1 (X , Z) maps naturally to
R V by sending a homology cycle c
to the functional c defined by c (f ) = c f . The image of is a lattice in
V , i.e., a Z-module of rank 2g which is discrete (cf. [Mu1], cor. 3.8). Fix a
base point 0 H, and define the Abel-Jacobi map AJ : X (C) V / by
RP
AJ (P )(f ) = 0 f . Note that this is well-defined, i.e., it does not depend on
the choice of path on X from 0 to P , up to elements in .
We extend the map AJ by linearity to the group Div (X ) of divisors on
X , and observe that the restriction of AJ to the group Div 0 (X ) of degree
0 divisors does not depend on the choice of base-point 0 . Moreover we have
the Abel-Jacobi theorem:
27
1.3
p1
+i
1 X
Tp (f ) =
f
+ phpif (p ).
p i=0
p
We give a more conceptual description of Tp in terms of remark 1.11 in the
case = 1 (N ). We have
X
Tp (f ) =
i (f ),
where i ( ) = +i
, and ( ) = hpip represent the p + 1 curves with p
structure that are images of (C/h, 1i, N1 ) by a p-isogeny, and the i are the
pull-back maps on differential forms on H. (An isogeny of elliptic curves with
-structure is simply an isogeny between the underlying curves which sends
one -structure to the other.) Such a description makes it evident that
PTp (fn)
belongs to S2 (), if f does. In terms of the Fourier expansion of f =
an q ,
the formula for the operator Tp on S2 (N, ) is given by:
X
X
Tp (f ) =
an q n/p + p(p)
an q pn .
p|n
28
p1
1 X
+i
Up (f ) =
=
an q n/p .
f
p i=0
p
p|n
The reader is invited to check that the Hecke operators of the form Tp or
Uq commute with each other, and also that they commute with the diamond
operators introduced in the previous section.
We extend the definition of the Hecke operators to operators Tpn , with
n > 1, by the inductive formulae
Tpn+1 = Tp Tpn hpipTpn1 ,
if (p, N ) = 1,
pei i is
The reader can consult [DI], sec. 3 and the references therein (especially [Shi2],
ch. 3 or [Kn]) for more details and different points of view on Hecke operators. Let T be the subring of EndC (S2 ()) generated over C by all the Hecke
operators Tp for p6 |N , Uq for q|N , and hdi acting on S2 ().
Definition 1.16 A modular form f is an eigenform if it is a simultaneous
eigenvector for all the Hecke operators in T, i.e., if there exists a C-algebra
homomorphism : T C such that T f = (T )f , for all T T.
A direct calculation shows that the coefficients an of an eigenform f can be
recovered from the homomorphism by the formula:
an (f ) = a1 (f )(Tn ).
29
30
Thus, T0 has the merit of being semi-simple, while T is not in general. The
cost of replacing T by T0 , however, is that one loses multiplicity one, i.e.,
the eigenspaces for T0 need not be one-dimensional. For example, the space
Sf defined in the previous exercise is contained in a single eigenspace for T0 .
The theory of Atkin-Lehner [AL] gives essentially a complete understanding
of the structure of the algebra T, and the structure of the space of eigenforms.
To motivate the main result, observe that the problem in the exercise above
seems to be caused by forms of level N that are coming from forms of lower
level N/p3 by a straightforward operation, and are therefore not genuinely
of level N . They are the analogues, in the context of modular forms, of nonprimitive Dirichlet characters.
Definition 1.21 We define the old subspace of S2 () to be the space spanned
by those functions which are of the form g(az), where g is in S2 (1 (M )) for
some M < N and aM divides N . We define the new subspace of S2 ()
to be the orthogonal complement of the old subspace with respect to the
Petersson scalar product. A normalized eigenform in the new subspace is
called a newform of level N .
The following result is the main consequence of the theory of Atkin-Lehner. It
gives a complete answer to the question of what is the structure of the algebra
T acting on S2 ().
Theorem 1.22 If f is in the new subspace of S2 () and is an eigenvector for
all the operators in T0 , then it is also an eigenform for T, and hence is unique
up to scaling. More generally, if f is a newform of level Nf |N , then the space
Sf defined by
Sf = {g S2 () such that T g = f (T )g, for all T T0 }
is stable under the action of all the Hecke operators in T. It is spanned by the
linearly independent forms f (az) where a ranges over the divisors of N/N f .
Furthermore, we have
S2 () = f Sf ,
where the sum is taken over all newforms f of some level Nf dividing N .
See [AL] for the proof in the case = 0 (N ), and [La2], ch. VIII for the
general case. (See also [DI], sec. 6 for an overview.)
Exercise 1.23 Consider the case where = 0 (22). Show that X0 (22) is of
genus 2, and hence that S2 (22) has dimension 2. Show that S2 (22) is equal
to Sf , where f = (( )(11 ))2 is a newform of level 11, so that in particular
there are no newforms of level 22 on . Show that T0 is isomorphic to C in
this case, and that T is isomorphic to the semisimple algebra C C.
31
Action on homology and Jacobians: Note that the Hecke operators act
on V = S2 () by duality. One checks (cf. [Kn], props. 11.23, 11.24) that the
sublattice of V is stable under the action of all the Hecke operators Tn , and
of the diamond operators hdi. Therefore the operators Tn and hdi give rise
to endomorphisms of the torus V /, and hence the Jacobian variety J , in
a natural way. The involution 7
gives rise to an involution on X (C)
(which is complex conjugation on the model of X over R deduced from the
Q- model defined in section 1.5). Since complex conjugation is continuous it
also acts on the integral homology = H1 (X (C), Z). Let + and be the
sublattices of on which complex conjugation acts by +1 and 1. These are
sublattices of of rank g which are stable under the Hecke operators, since
complex conjugation commutes with the Hecke action.
A more algebraic description of the action of Tp on J is given via the
notion of an algebraic correspondence. A correspondence on a curve X is a
divisor C on X X which is taken modulo divisors of the form {P } X and
X {Q}. Let 1 and 2 denote the projections of X X onto each factor.
Then the correspondence C induces a map on divisors of X, by setting
C(D) = 2 (11 (D) C).
(For the definition of the intersection D1 D2 of two divisors, see [We1].) The
map C preserves divisors of degree 0, and sends principal divisors on X to
principal divisors. It gives a well defined algebraic endomorphism of the Jacobian variety Jac (X). Given a correspondence C, its transpose C is defined to
be the divisor of X X obtained by interchanging the two factors of X X.
One can define a natural notion of composition for correspondences, and the
set of correspondences forms a ring. The general theory of correspondences
and the proofs of the above facts are given in [We1], particularly the second
chapter.
The Hecke correspondence Tn is defined as the closure in X X of the
locus of points (A, B) in Y Y such that there is a degree n cyclic isogeny of
elliptic curves with -structure from A to B. For example, if p is a prime not
dividing N , then Tp is an algebraic curve in X1 (N )X1 (N ) which is birational
to X1 (N )0 (p) . The induced map on divisors in this case satisfies
X
Tp ((E, P )) =
(E/C, P mod C)
where the sum runs over the subgroups C of E having order p. Note also that
if (A, B) belongs to Tp , then the isogeny dual to A B gives a p-isogeny
from B to hpiA, so that
Tp = hpi1 Tp .
32
1.4
As in exercise 1.8, one can show that the infinite sum defining L(f, s) converges
absolutely in the right half-plane Re (s) > 23 . A much better insight is gained
into the function L(f, s) by noting that it is essentially the Mellin transform of
the modular form f . More precisely, if we set (f, s) = N s/2 (2)s (s)L(f, s),
then we have
Z
s/2
(f, s) = N
f (iy)y s dy/y
(1.4.1)
0
1.5
curves together with a cyclic subgroup of order N , given by sending the point
H/0 (N ) to the pair (C/h1, i, h1/N i).
Consider the universal elliptic curve
36
1
Ej : y 2 + xy = x3
x
.
j 1728
j 1728
It is an elliptic curve over the function field Q(j), with j-invariant j. Let d be
the order of P1 (Z/N Z), and let C1 , . . . , Cd denote the set of all cyclic subgroups
of Ej of order N , defined over Q(j), an algebraic closure of Q(j). Fix one of
these subgroups, C. The Galois group Gal (Q(j)/Q(j)) permutes the Ci in a
natural way. Let FN be the smallest extension of Q(j) (viewed as embedded in
Q(j)) with the property that (C) = C, for all Gal (Q(j)/F N ). It can be
seen (cf. [Shi2], thm. 6.6) that the Galois action on the Ci is transitive so that
= Q.
FN /Q(j) is of degree d, and that it is a regular extension, i.e., FN Q
Geometrically, FN can be viewed as the function field of a curve X/Q over Q,
with the inclusion of Q(j) corresponding to a map from X/Q to the projective
j-line over Q. The pair (Ej , C) is an elliptic curve over Q(j) with a subgroup
of order N defined over FN . Using (Ej , C), each complex point of X gives an
elliptic curve over C with a subgroup of order N , provided the point does
not lie over j = 0, 1728 or . The resulting map to X0 (N ) extends to an
isomorphism from X to X0 (N ). The curve X thus constructed, together with
this identification, is the desired model of X0 (N ) over Q.
More concretely, the functions j = j( ) and jN = j(N ) are related by a
polynomial equation N (j, jN ) = 0 with coefficients in Q, of bidegree d. The
field FN is the function field of the affine curve N (X, Y ) = 0, and the mapping
7 (j( ), j(N )) gives a birational equivalence between H/0 (N ) and the
complex curve defined by the equation N . In practice it is not feasible to
write down the polynomial N , except for certain very small values of N . To
study the models over Q of X0 (N ), more indirect methods are needed, which
rely crucially on the moduli interpretation of X0 (N ). Similar remarks hold for
X1 (N ).
Models over Z: The work of Igusa [Ig], Deligne-Rapoport [DR], Drinfeld [Dr],
and Katz-Mazur [KM] uses the moduli-theoretic interpretation to describe a
canonical proper model for X over Spec Z. These models allow us to talk
about the reduction of X over finite fields Fp , for p prime. The curve has
good reduction at primes p not dividing N , with the non-cuspidal points of
p with -structure. The singular
X/Fp corresponding to elliptic curves over F
fibers at primes p dividing N can also be described precisely; an important
special case (see [DR]) is that of 0 (N ) with p exactly dividing N . For further
discussion and references, see [DI], sec. 8, 9.
35
point of X,Q
but not necessarily of X .
Jacobians: Weils theory [We1] of the Jacobian shows that the Jacobians J
defined in section 1.2 as complex tori also admit models over Q. When we
speak of J , J0 (N ) and J1 (N ) from now on, we will refer to these as abelian
varieties defined over Q. Thus, the points in J (K), for any Q-algebra K, are
identified with the divisor classes on X of degree 0, defined over K.
We let J/Z , denote the Neron model of the Jacobian J over Spec (Z).
Using this model we define J/A for arbitrary rings A. In particular we can
consider J/Fp , the reduction of the Jacobian in characteristic p, which is closely
related to the reduction of the integral model of the curve X mentioned above.
In particular, if p does not divide the level of , then J/Fp can be identified
with the Jacobian of X/Fp . For a treatment of the case = 0 (N ) with p
exactly dividing N , see the appendix of [Maz1]; for more general discussion
and references, see [DI], sec. 10, especially sec. 10.3.
Hecke operators: The Hecke operators have a natural moduli interpretation,
which was already touched upon in section 1.3. In particular, one finds that
the operator Tn arises from a correspondence which is defined over Q, and
gives rise to an endomorphism of the Jacobian J which is defined over Q.
This in turn gives rise to an endomorphism of the Neron model J/Z , and we
can then consider the endomorphism Tn on the reduction of the Jacobian in
characteristic p. Recall that if p is a prime not dividing N , we may identify
this reduction with the Jacobian of X/Fp . (Cf. [MW], ch. 2, sec. 1, prop. 2.)
Furthermore, one can show that the moduli-theoretic interpretation of the
Hecke operator remains valid in characteristic p; i.e., the endomorphism Tn of
36
J/Fp is induced by a map on divisors satisfying, for all ordinary elliptic curves
A with -structure:
X
Tn (A) =
i(A),
i
where the sum is taken over all cyclic isogenies of degree n. (See [DI], sec. 10.2
for further discussion and references.)
This description allows one to analyse, for example, the Hecke operator
Tp over Fp , when (p, N ) = 1. Let us work with = 1 (N ), to illustrate
p , let X be the Frobenius morphism on X
the idea. For a variety X over F
defined by raising coordinates to the pth power. Thus if (E, P ) corresponds
to a point of X1 (N )/Fp , then E is an isogeny of degree p from (E, P ) to the
pair (E , P ) = X1 (N ) (E, P ). The graph of X1 (N ) in (X1 (N ) X1 (N ))/Fp is
a correspondence of degree p, which we call F . Let F 0 be the transpose of this
correspondence. The endomorphism F of J induced by F is the Frobenius
endomorphism J , and the endomorphism F 0 is the dual endomorphism (in
the sense of duality of abelian varieties). Now consider the divisor
F 0 ((E, P )) = (E1 , P1 ) + + (Ep , Pp ),
where the (Ei , Pi ) are elliptic curves with -structure in characteristic p. Since
Ei is an isogeny of degree p from (Ei , Pi ) to (E, P ), we also have the dual
isogeny from (E, P ) to (Ei , pPi ). If E is ordinary at p, then the points
(E , P ), (E1 , pP1 ), . . . , (Ep , pPp ) are a complete list of the distinct curves
with -structure which are p-isogenous to (E, P ). Hence one has the equality
of divisors on X1 (N )/Fp :
Tp ((E, P )) = (E , P ) + (E1 , pP1 ) + + (Ep , pPp ) = (F + hpiF 0 )((E, P )).
Since the ordinary points are dense on X1 (N )/Fp , we deduce that Tp = (F +
hpiF 0 ) as endomorphisms of J1 (N )/Fp . This equation, known as EichlerShimura congruence relation, plays a central role in the theory. (For more
details, see [DI], sec. 10.2, 10.3.)
Theorem 1.29 If p6 |N then the endomorphism Tp of J/Fp satisfies
Tp = F + hpiF 0 .
Remark 1.30 This was proved by Eichler [Ei] to hold for all but finitely many
p in the case of 0 (N ), and by Shimura ([Shi1], see also [Shi2], ch. 7) in the
case of 1 (N ). The fact that it holds for all p not dividing N follows from
work of Igusa [Ig].
37
Modular forms: In the same way that modular curves have models over
Q and over Z, the Fourier coefficients of modular forms also have natural
rationality and integrality properties. We start by sketching the proof of:
Theorem 1.31 The space S2 () has a basis consisting of modular forms with
integer Fourier coefficients.
Proof: The Hecke operators act on the integral homology + in a way that
is compatible with the action on S2 () and respects the natural (Poincare)
duality between these two spaces. Hence, if {n }nN is a system of eigenvalues
for the Tn , then the n are algebraic integers in some finite extension K of Q,
and the system {n }nN is a system of eigenvalues for the Tn for any Galois
automorphism of Q/Q.
Hence, we have shown:
Proposition 1.32 If f S2 (M, ) is a newform of some level M dividing
N , then its Fourier coefficients lie in a finite extension K of Q. Moreover,
if Gal (Q/Q)
is any Galois automorphism, then the Fourier series f
obtained by applying to the Fourier coefficients is a newform in S2 (M, ).
The explicit description of S2 () given in section 1.3 implies that S2 () is
spanned by forms having Fourier coefficients which are algebraic integers in
some finite (Galois) extension K of Q, and that the space of forms with Fourier
coefficients in K is stable under the natural action of Gal (K/Q) on Fourier
expansions. An application of Hilberts theorem 90 shows that S2 () has a
basis consisting of forms with rational Fourier expansions, and the integrality
of the Fourier coefficients of eigenforms yields the integrality statement of
theorem 1.31.
2
We define S2 (, Z) to be the space of modular forms with integral Fourier
coefficients in S2 (). Theorem 1.31 states that S2 (, Z) C = S2 (). Given
any ring A, we define
S2 (, A) = S2 (, Z) A,
and define S2 (N, A) and S2 (N, , A) (where now is a character with values
in A ) in the obvious way. If A is contained in C, the q-expansion principle
below allows us to identify S2 (, A) with the set of modular forms in S2 ()
with Fourier coefficients in A.
The q-expansion principle: Because the modular curve X0 (N ) has a model
over Q, the space of modular forms S2 (N ) = 1 (X0 (N )) has a natural rational
structure, given by considering the differential forms on X0 (N ) defined over
Q. The fundamental q-expansion principle (see [DR], ch. 7, or [Kat], sec. 1.6)
says that these algebraic structures are the same as those obtained analytically
38
1.6
It follows directly from the formulas for the Hecke operators acting on qexpansions that the Tn leave S2 (0 (N ), Z) stable, as well as the subspace
of S2 (N, ) with coefficients in Z[]. Using the q-expansion principle (theorem 1.33), one can also show ([DI], sec. 12.4) that the diamond operators
preserve the spaces of cusp forms on with integral Fourier expansions, and
hence that the space S2 (, Z) is preserved by all the Hecke operators.
We define TZ to be the ring generated over Z by the Hecke operators Tn
and hdi acting on the space S2 (, Z). More generally, if A is any ring, we define
TA to be the A-algebra TZ A. This Hecke ring acts on the space S2 (, A)
in a natural way. Before studying the structure of the Hecke rings TA as we
vary the rings A, we note the following general result (Cf. [Shi2], ch. 3.):
Lemma 1.34 The space S2 (, A) = HomA (S2 (, A), A) is a free TA -module
of rank one.
Sketch of proof: One checks that the pairing TZ S2 (, Z) Z defined by
(T, f ) 7 a1 (T f ) sets up a perfect, TZ -equivariant duality between TZ and
S2 (, Z). The result for arbitrary A follows.
2
Hecke rings over C: If A = C, then the structure of the ring T = TC
is completely described by theorem 1.22. More precisely, if TC,f denotes the
image of the Hecke algebra acting on the space Sf defined in section 1.3, then
TC = f TC,f ,
where the direct sum ranges over all distinct newforms f of some level Nf
dividing N . Furthermore, the algebra TC,f can be described explicitly. In
particular, if f is a newform of level N then TC,f is isomorphic to C, but if Nf
is not equal to N then the ring TC,f need not be a semi-simple algebra over C.
Lemma 1.34 in the case A = C says that V = S2 () is a free TC -module
of rank one, but we also have:
39
Gal (Q/Q))
of a normalized newform f of some level Nf dividing N , and let
Kf be the field extension of Q generated by the Fourier coefficients of f . The
space g[f ] Sg is a vector space of dimension [Kf : Q]0 (N/Nf ), which is
spanned by modular forms with rational Fourier coefficients. Let S[f ] be the
Q-subspace of forms in g[f ] Sg with rational Fourier coefficients. The space
S[f ] is stable under the action of TQ , and letting TQ,[f ] be the image of TQ
acting on S[f ] , we obtain the direct sum decomposition
TQ = [f ] TQ,[f ] ,
where the sum is taken over the distinct GQ -orbits of normalized newforms
f of some level Nf dividing N . If Nf is equal to N , then the algebra TQ,[f ]
is isomorphic to the field Kf . If Nf is a proper divisor of N , then, as in the
complex case, the algebra TQ,[f ] is a (not necessarily semi-simple) algebra over
Q of rank 0 (N/Nf )[Kf : Q]. The nature of the fields Kf , and in general the
structure of TQ , is very poorly understood at this stage; for example, one does
not know how to characterize the number fields that occur as a Kf for some
f (but they are all known to be totally real or CM fields).
The ring TQ acts naturally on the rational homology H1 (X , Q) = Q,
and we have
Lemma 1.37 The module Q is free of rank two over TQ .
Sketch of proof: The modules + C ' V and C ' V are free of rank
one over TC , by lemma 1.34. This implies that + Q and Q are both
free of rank one over TQ .
2
40
Hecke rings over Z: The ring TZ is a certain (not necessarily maximal) order
in TQ . One still has an injection
TZ , [f ] TZ,[f ] ,
where now TZ,[f ] denotes the ring generated over Z by the Hecke operators
acting on S[f ] . Of course the structure of TZ is even more mysterious than
that of TQ ! The ring TZ acts naturally on , but it is not the case in general
that is free of rank two over TZ , i.e., that the integral analogue of lemma 1.37
is true. (See remark 1.42 below.)
Hecke rings over Q` : The study of the algebras TZ` and TQ` arises naturally
because of the Hecke action on the Tate module T` (J ),
T` (J ) := lim(J )[`n ],
where the inverse limit is taken with respect to the multiplication by ` maps.
The action of TZ` on T` (J ) is compatible with that of GQ , and it is this pair of
actions on the Tate module which is used to associate two-dimensional Galois
representations to modular forms.
It will sometimes be more convenient to consider the ring TQ` and its action
on
V = T` (J ) Z` Q` .
We first record a useful duality property enjoyed by the Tate modules. The
Weil pairings on the groups J [`n ] for n 1 induce a perfect pairing
h , i : T` (J ) T` (J ) Z` .
Since each Hecke operator T is adjoint to wT w where w = wN is the AtkinLehner involution, we have the following lemma.
Lemma 1.38 The map x 7 x where x (y) = hx, wyi defines an isomorphism of TZ` -modules,
T` (J )
= T` (J ) = HomZ` (T` (J ), Z` ),
and hence an isomorphism of TQ` -modules
V
= V = HomQ` (V, Q` ).
The following lemma allows us to regard TQ` as a coefficient ring for a
two-dimensional Galois representation
Gal (Q/Q)
Aut TQ` (V)
= GL2 (TQ` ).
41
Q
Decomposing TQ` as i Ri where each factor Ri is a finite-dimensional local
Q` -algebra, we obtain an isomorphism
Ri R i
= Ri Ri
for each i. At least one of the four maps Ri Ri deduced from this
isomorphism must be surjective, and by counting dimensions, we see that it
2
must be injective as well. It follows that TQ` is isomorphic to TQ` .
Recall that for primes p not dividing N , the Jacobian J has good reduction
mod p, and the Eichler-Shimura relation, theorem 1.29, states that on J/Fp ,
we have
Tp = F + hpiF 0 .
For primes p not dividing N `, we may identify T` (J ) with the `-adic Tate
module of the reduction (see [ST]) and consider the Frobenius endomorphism
F on the free rank two TQ` -module V. As a consequence of the Eichler-Shimura
relation, we find:
Theorem 1.41 For p not dividing N `, the characteristic polynomial of F on
the TQ` -module V is
X 2 Tp X + hpip.
Proof: (We are grateful to Brian Conrad for showing us this argument.) Since
F F 0 = p, it follows from the Eichler-Shimura relation that
F 2 Tp F + hpip = 0.
To conclude that this is in fact the characteristic polynomial, it suffices to
compute the trace of F . To do so, we use the TQ` isomorphism
V V
42
where the product runs over the maximal ideals m of TZ` , and Tm is the
localization of TZ` at m. For each m, Tm is a complete local Z` -algebra, free
of finite rank as a Z` -module.
Remark 1.42 While the analogue of lemma 1.39 does not always hold for
TZ` (see [MRi], sec. 13) we shall see that it holds for certain localizations T m .
Results of this type are much deeper than lemma 1.39 and were first obtained
by Mazur [Maz1], sec. 14, 15. We shall return in chapter 4 to explain Mazurs
result and its generalizations, which play a role in the arguments of [W3] and
[TW].
Example 1.43 The curve X0 (19) has genus 1, and X0 (57) has genus 5. By
consulting the tables in the Antwerp volume [Ant4] or Cremonas book [Cr],
one finds that there is exactly one newform of level 19, and that there are
three newforms of level 57, which all have rational Fourier coefficients. Their
Fourier coefficients ap , for the first few primes p, are listed in the following
table:
19A
57A
57B
57C
2
3
5
7 11 13 17 19 23
29 31
0 2
3 1
3 4 3
1
0
6 4
2 1 3 5
1
2 1 1 4 2 6
1
1 2
0
0
6 6 1
4
2
8
2
1
1
3 3 6
3 1
4 10
2
finds that the Hecke ring TZ3 generated over Z3 by the Hecke operators acting
on S2 (57, Z3 ) is isomorphic to the subalgebra of Z53 :
{(x, y, z, t, w) such that t w
(mod 3)}.
(mod 3)},
1.7
P
Let f =
an q n be an eigenform on with (not necessarily rational) Fourier
coefficients, corresponding to a surjective algebra homomorphism f : TQ
Kf , where Kf is the field generated over Q by the Fourier coefficients of f . We
briefly review in this section a construction of Shimura ([Shi2], ch. 7) which
associates to f (or rather, to the orbit [f ] of f under GQ ) an abelian variety
Af defined over Q and of dimension [Kf : Q].
Let If TZ be the ideal ker (f ) TZ . The image If (J ) is a (connected)
subabelian variety of J which is stable under TZ and is defined over Q.
Definition 1.44 The abelian variety Af associated to f is the quotient
Af = J /If J
From this definition one sees that Af is defined over Q and depends only on
[f ], and that its endomorphism ring contains TZ /If which is isomorphic to an
order in Kf .
Remark 1.45 Using theorem 1.22, one can show that J0 (N ) is isogenous to
Q
0 (N/Nf )
.
[f ] A[f ]
Let [f ] be the set of all eigenforms whose Fourier coefficients are Galois
conjugate to those of f . The number of forms in [f ] is equal to the degree d
of Kf over Q. Now, we set
X
V[f ] = g[f ] Vg , [f ] =
g .
g[f ]
Note that [f ] is simply the orthogonal projection of V to V[f ] . Note also that
[f ] belongs to the Hecke algebra TQ .
Lemma 1.46 The abelian variety Af is isomorphic over C to the complex
torus V[f ] /[f ] (), with the map [f ] : V / V[f ] /[f ] () corresponding to
the natural projection from J to Af .
In particular, one sees that Af is an abelian variety of dimension d = [Kf : Q].
Hence if f has rational Fourier coefficients, then the abelian variety Af is an
elliptic curve. This elliptic curve is called the strong modular elliptic curve
associated to f if also f is a newform of level N and = 0 (N ).
Example 1.47 If = 0 (26), one checks that the genus of X is two, and
that there are two distinct normalized eigenforms f1 and f2 in S2 (26). From
the tables in [Ant4] or [Cr], one sees that f1 and f2 have integral Fourier
coefficients, whose values for the primes 31 are:
f1
f2
2
3
5
7 11 13 17 19 23 29 31
1
1 3 1
6
1 3 2
0 6 4
1 3 1
1 2 1 3 6 4 2
4
Hence the abelian varieties Af1 and Af2 are elliptic curves. The above table
suggests (and this can be checked directly by looking at the equations for these
curves given in [Ant4] or [Cr], or by using the discussion in [DO], lemma 2.1)
that the Fourier coefficients of f1 and f2 are congruent modulo 2. The natural
projection J0 (26) Af1 Af2 is an isogeny whose kernel is isomorphic to
Z/2Z Z/2Z, and J0 (26) is not isomorphic to Af1 Af2 . More generally, one
knows that a Jacobian can never decompose as a non-trivial direct product of
two principally polarized abelian varieties, cf. [Maz1], prop. 10.6. (There are
no non-zero homomorphisms from Af1 to Af2 , and so a non-trivial product
decomposition of J would have to induce a decomposition as a product of
principally polarized abelian varieties.)
Let T` (Af ) be the Tate module of the abelian variety Af ,
T` (Af ) := lim(Af )[`n ],
45
where the inverse limit is taken with respect to the multiplication by ` maps.
This module is naturally a module for Tf Z` , and T` (Af ) Q` is a module
for Kf Q` as well.
Lemma 1.48 The module T` (Af )Q` is a free module of rank 2 over Kf Q` .
Proof: This follows directly from lemma 1.39.
Proposition 1.49 The algebra End Q (Af ) Q is isomorphic to Kf . In particular Af is simple over Q.
The proof is given in [R2], cor. 4.2. The main ingredient is the irreducibility of
the Galois representation attached to Af as in section 3.1. (Cf. theorem 3.1.)
Properties of Af : good reduction:
Theorem 1.50 If f is an eigenform of level N , and p is a prime that does
not divide N , then the abelian variety Af has good reduction at p.
Proof: This follows from the fact that J has good reduction at such primes,
which in turn is a consequence of the good reduction of X .
2
If p is a prime not dividing N , it then becomes natural to study the number
Nf,p of points on the abelian variety Af over the finite field Fp . It is given by
the following formula:
Proposition 1.51 The number of points Nf,p is given by the formula
Nf,p = NormKf /Q (f (1 ap (f ) + hpip)),
where af (p) Kf is the p-th Fourier coefficient of f .
Proof: By Weils theory [We1], this number is given by the determinant
Nf,p = det(1 F ),
where F is the Frobenius endomorphism acting on the `-adic Tate module
T` (Af ). So theorem 1.41 implies
det(1 F ) = NormKf /Q (f (1 Tp + hpip)),
and proposition 1.51 follows.
2
Defining the local Hasse-Weil L-function of Af over Fp by the formula
L(Af /Fp , s) = det(1 F ps )1 ,
46
Lp (f , s),
(1.7.1)
1.8
Let E be any elliptic curve defined over Q, and let N denote its arithmetic
conductor, defined as in section 1.1. Then we have:
Proposition 1.53 The following are equivalent:
(a) The curve E is isogenous over Q to Af , for some newform f on some
congruence group .
(b) The curve E is isogenous over Q to Af , for a newform f on 0 (N ).
(c) There is a non-constant morphism defined over Q, from X0 (N ) to E.
47
Sketch of proof: The implication (b) (a) is immediate, and (a) (b) follows
from the work of Carayol [Ca3] (cf. proposition 3.20 below). If (b) holds, then
there is a surjective map J0 (N ) E. Composing with the Abel-Jacobi map
X0 (N ) J0 (N ) gives a map to E which is not constant, by the definition of
the Jacobian, and hence (c) holds. Conversely, a non-constant map of curves
: X0 (N ) E induces, by Albanese (covariant) functoriality, a surjective
map of Jacobians : J0 (N ) E, where we have identified E with its own
Jacobian in the natural way. Since J0 (N ) is isogenous to a product (possibly
with multiplicities) of abelian varieties Af where f run over newforms of level
Nf dividing N (cf. remark 1.45), it follows that there is a surjective map
Af E for some Af . Since Af is simple (proposition 1.49), this map is an
isogeny, and part (b) follows.
2
We call an elliptic curve over Q satisfying the equivalent properties above
a modular elliptic curve. A startling conjecture that was first proposed by
Taniyama in the 1950s and made more precise by Shimura predicts that the
Shimura construction explained in the previous section is surjective:
Conjecture 1.54 All elliptic curves defined over Q are modular.
This conjecture is now known to be true for a very wide class of elliptic curves.
The results of Wiles [W3] and Taylor-Wiles [TW] imply it is true for all semistable elliptic curves, and a strengthening of the method, [Di2], implies the
conjecture for all elliptic curves which have semi-stable reduction at 3 and 5.
Remark 1.55 See [DI], sec. 13 for a more thorough list of equivalent forms
of the conjecture.
Relation with L-functions: Define numbers ap = ap (E) as in section 1.1.
Recall the (global) Hasse-Weil L-function defined in section 1.1 by equations
(1.1.2), (1.1.4), (1.1.5), and (1.1.6). The following proposition reveals some of
the importance of the Shimura-Taniyama conjecture:
Proposition 1.56 If E is modular, i.e., is associated to a newform f by the
Shimura construction, then
L(E/Q, s) = L(f, s).
In particular, L(E/Q, s) has an analytic continuation to the entire complex
plane, and a functional equation.
48
Conjecture 1.58 Every simple abelian variety A over Q which is of GL2 -type
is modular.
This generalization of the Shimura-Taniyama conjecture is far from being
proved, and tackling it still seems to require some major new ideas, even
after the ideas introduced by Wiles.
2
2.1
Galois theory
Galois representations
p and
to a valuation vp : Qp
Q. Let OQ p denote the ring of integers of Q
mQ p its maximal ideal. The residue field OQ p /mQ p is an algebraic closure of
p . The valuation vp is compatible with the
Fp , which we will identify with F
action of GQp , and hence OQ p and mQ p are stable under GQp . In particular
GQp acts on OQ p /mQ p and we obtain a map % : GQp GFp , which is in fact
a surjection. We call the kernel the inertia group and denote it Ip . It has a
unique maximal pro-p subgroup: the wild inertia group, denoted Pp . There is
a canonical isomorphism
Y
Z` (1),
Ip /Pp
=
`6=p
where the (1) indicates that if f GQp /Pp is a lifting of Frobenius and if
Ip /Pp then f f 1 = p .
The inertia group Ip is filtered by a series of subgroups called the higher
ramification groups. More precisely, there are closed normal subgroups Ipu in
50
These groups are defined in section 3 of chapter IV of [Se2], where they are
1
denoted GuQp (except that there, G1
Qp = GQp whereas we have Ip = Ip ).
The local Kronecker-Weber theorem ([Se2], ch. 14, sec. 7, for example)
asserts that the map
% p : Gab
Qp GFp Zp .
is an isomorphism. (We will use Gab to denote the abelianisation of a profinite
group G, i.e. the unique maximal abelian continuous image.) Under it Ip goes
due
u
Zp ) Z
to Z
p , where due denotes the least
p and for u > 0, Ip goes to (1 + p
integer greater than or equal to u. If ` 6= p then ` is trivial on Ip and takes
Frob p to p Z
` .
51
p : Gab
Z
Q
p.
p
where the product is over all primes p 6= ` in the case of an `-adic or mod `
representation and over all primes in the case of an Artin representation. This
makes sense because is unramified almost everywhere.
The following lemma is an exercise.
R
0
codimIp du =
R
0
codim
Ip du. Thus
N () = N (
)
Ip
p(dim
dim Ip )
p6=`
and in particular N (
) divides N ().
We take this opportunity to introduce some important notation. Let G
denote a group and R a ring. If : G GLd (R) is a representation we
shall let M denote the underlying R[G]-module (so that M
= Rd as an Rmodule). If M is an R[G]-module we shall let End (M ) denote the module of
R-linear endomorphisms of M . It is also an R[G]-module with the G-action
being defined by
(g())(m) = g((g 1 m)).
If M is a finitely generated free R-module we will let End 0 (M ) denote the
sub-R[G]-module of End (M ) consisting of endomorphisms of trace zero. We
will use ad to denote End (M ) and ad0 to denote End 0 (M ). Note that if
d is invertible in R then ad
= ad0 R as R[G]-modules, where R has the
trivial action of G. Note also then that the kernel of ad0 is the same as the
= Ind G
H . In the latter case ad = Ind H (/ ), where is the nontrivial
character of G/H and 0 is the composite of and conjugation by an element
0
0
of G H. Moreover, either Ind G
H (/ ) is irreducible or ad = 1 2 3
where the i are distinct characters of G of order 2. In the latter case, is
induced from a character from each of the subgroups Hi = ker i of index 2 in
G. We leave the verification of these facts as an exercise to the reader.
2.2
Furthermore, E[n](Q) carries a natural action of GQ and so we get a representation (defined up to conjugation)
E,n : GQ GL2 (Z/nZ).
55
(b) The representation E,` is absolutely irreducible for all `. For fixed E,
E,` is absolutely irreducible for all but finitely many `.
(c) If E does not have complex multiplication then E,` (and hence E,` ) is
surjective for all but finitely many `.
Proof: Part (a) follows from the existence of the non-degenerate alternating
Galois-equivariant Weil pairing
T` E T` E Z` (1) := lim `n .
Part (b) is proved in [Se5], ch. IV. Part (c) is the main result of [Se6].
2
The following deep result, which is stronger than the second assertion of
part (b) of proposition 2.8, is due to Mazur [Maz2], thms. 1 and 2. (Parts (b)
and (c) can actually be deduced directly from theorem 1.7, using the fact that
a semi-stable curve with a rational subgroup of order ` is necessarily isogenous
to a curve with a rational point of order `.)
Theorem 2.9 Suppose that E/Q is an elliptic curve.
(a) If ` > 163 is a prime then E,` is irreducible.
(b) If E is semistable then E,` is irreducible for ` > 7.
(c) If E is semistable and E,2 is trivial then E,` is irreducible for ` > 3.
Remark 2.10 Combined with the results of Serre [Se6], Mazurs results imply
that if E is semistable everywhere then E,` is surjective for ` > 7 ([Maz2],
thm. 4).
Local behaviour of E,` and E,` :
56
as Gp -modules.
(c)
p
E,p |Ip 0 1 ;
p ) is trivial and
If E has supersingular reduction at p, then Ep [p](F
E,p |Gp is irreducible.
Proofs: The proof of part (a) can be found in chapters V and VII of [Si1].
For part (b) one considers the finite flat group scheme Fn = E[pn ], where E
is the model for E over Zp defined by (W min ). (See [Sha] for an introduction
to theory of finite flat group schemes.) Part (c) follows from the results in
chapters V and VII of [Si1] together with basic results on finite flat group
schemes.
Proposition 2.12 Suppose that E has multiplicative reduction at p. Let q =
qE,p pZp be the multiplicative Tate period attached to E, defined as in section 1.1. Let : Gp /Ip {1} be the unique non-trivial unramified quadratic
character of Gp if E has non-split multiplicative reduction, and let be the
trivial character if E has split multiplicative reduction. Then
(a)
E,` |Gp
and if ` 6= p then mp (E,` ) = 1.
`
0 1 ,
(b)
E,` |Gp 0`
1 ,
`
and H 1 (GQp , Z/`Z(1))
= Q
p /(Qp ) corresponds to the Tate period
`
qE,p Q
p /(Qp ) .
57
(c) If ` 6= p then mp (
E,` ) = 0 (i.e., E,` is unramified at p) if and only if
min
`|vp (E ) = vp (jE ).
p ) if
p)
(d) There is a finite flat group scheme F/Zp such that E[p](Q
= F(Q
min
and only if p|vp (E ) = vp (jE ).
Sketch of proof: By Tates proposition 1.5 of section 1.1, we have
p)
/q Z )()
E(Q
= (Q
p
/q Z )() by : x 7
as Gp -modules. (The () indicates that Gp acts on (Q
p
()
(x) .) The proposition follows directly from this, together with [Edi], prop.
8.2. for the last part. (For further details, see ch. VII, prop. 5.1 of [Si1], ch.
IV, thm. 10.2. and ch. V of [Si2], and sec. 2.9 of [Se7].)
2
Proposition 2.13 Suppose that E has additive reduction at p. Then for any
` 6= p the conductor of E,` |Gp is at least 2, and if p > 3 then it is equal to 2.
Remark 2.14 For an elliptic curve E with any type of reduction at p, the
conductor of E,` |Gp (for ` 6= p) coincides with the local conductor of E at p
given by a formula of Ogg (proved by Saito [Sa] in the case p = 2). In particular
the conductor of E,` |Gp is independent of ` 6= p. For further discussion and
references, see ch. IV, secs. 10 and 11 of [Si2].
The Frey curve: We now consider the elliptic curve
EA,B : Y 2 Z = X(X AZ)(X + BZ),
where A and B are non-zero coprime integers with A + B 6= 0. Set C =
A + B. This equation has discriminant = 16(ABC)2 . If p is an odd prime
then this curve has good reduction at p if p6 |(ABC) and has multiplicative
412n
reduction at p if p|ABC. Thus min
(ABC)2 for some n Z0 .
EA,B = 2
If for instance A 1 mod 4 and B 0 mod 16 then EA,B has a minimal
Weierstrass equation
Y 2 Z + XY Z = X 3 +
BA1 2
X Z (AB/16)XZ 2 .
4
In this special case EA,B has semi-stable reduction everywhere and min
EA,B =
8
2
2 (ABC) .
The connection with Fermats Last Theorem is provided by the elliptic
curve
E : Y 2 Z = X(X a` Z)(x + b` Z)
58
considered by Hellegouarch [He] and Frey [Fr], where a` + b` = c` is a hypothetical non-trivial solution to Fermats Last theorem, and where we have
supposed without loss of generality that a, b and c are pairwise coprime, that
b is even and that a 1 mod 4. Frey suggested that properties of this hypothetical elliptic curve E could lead to a contradiction if the curve were also
known to be modular. This suggestion was made precise by Serre [Se7], who
observed that the following list of properties would yield the desired contradiction, when combined with his conjectures on Galois representations associated
to modular forms (to be discussed in section 3.2).
Theorem 2.15 Suppose that ` > 3 is prime. The representation E,` has the
following properties.
(a) E,` is irreducible.
(b) E,` is unramified outside ` and 2.
`)
` ) as
(c) There is a finite flat group scheme F/Z` such that F(Q
= E[`](Q
G` -modules.
(d) #
E,` (I2 ) = `.
Proof: Part (a) follows from Mazurs theorem (theorem 2.9, (c)). To prove (b),
8
2`
(c), and (d), we note the key fact that min
E = 2 (abc) is a perfect `-th power,
up to powers of 2. The fact that E,` is unramified at p 6= `, 2 (part (b)) follows
from this, from proposition 2.11, (a) (for primes p 6= ` of good reduction for E)
and from proposition 2.12 (c) (for primes p 6= `, 2 of multiplicative reduction
for E). Part (c) is a consequence of proposition 2.12 (d), and part (d) follows
from proposition 2.12 (c).
2
Remark 2.16 Part (a) implies that the image of E,` is large, while parts (b)
and (c) state that the image has very limited ramification. We remark that
the conclusion of the theorem ensures that mp (
E,` ) = 0 for p 6= 2, ` and that
m2 (
E,` ) = 1.
2.3
Galois cohomology
In this section M will denote a continuous discrete GQ -module of finite cardinality. If M and N are two such modules we will endow the space Hom(M, N )
of homomorphisms of abelian groups with an action of GQ via the formula
(g())(m) = g((g 1 m)).
59
We will use M to denote Hom(M, n (Q)),
where n is an integer such that
nM = (0) and where n (Q) denotes the group of nth roots of unity in Q.
We will use to denote Pontryagin duals. All cohomology groups of profinite groups will mean continuous cohomology. The general references for this
section are Milne [Mi] and Serre [Se1].
We start by recalling Tates local duality theorem (see chapter 1, section 2
of [Mi]).
Theorem 2.17 Let v be a place of Q (i.e. either a prime number or ).
(a) H i (Gv , M ) is finite for all i.
,
(b) For all integers n there are compatible embeddings H 2 (Gv , n (Q))
, m (Q)
if
Q/Z (i.e. compatible with the maps coming from n (Q)
n|m).
(c) For v 6= and i = 0, 1, 2 the cup product and the above embeddings give
rise to a perfect pairing
, Q/Z.
H i (Gv , M ) H 2i (Gv , M ) H 2 (Gv , n (Q))
(This also holds when v = , if one replaces H 0 (Gv , M ) = H 0 (R, M ) by
0 (R, M ) := M =1 /(1 + )M , where
Tates modified cohomology group: H
is complex conjugation.)
(d) If v 6= is a prime, then H i (Gv , M ) = (0) for i > 2 and
#H 1 (Gv , M ) = #H 0 (Gv , M )#H 2 (Gv , M )#(M Zv ).
(e) If v 6= and M Zv = 0 then H 1 (Gv /Iv , M Iv ) and H 1 (Gv /Iv , M Iv )
are the annihilators of each other under the pairing
H 1 (Gv , M ) H 1 (Gv , M ) Q/Z.
By a collection of local conditions for M we shall mean a collection L =
{Lv } of subgroups Lv H 1 (Gv , M ) as v runs over all places of Q, such that
Lv = H 1 (Gv /Iv , M Iv ) for all but finitely many v. Note that if L is a collection
of local conditions for M then L = {L
v } is a collection of local conditions for
v
M Iv H 1 (Gv /Iv , M Iv ) (0),
(0) H 0 (Gv , M ) M Iv
0
#HL (Q, M )
#H (GQ , M ) v #H (Gv , M )
We note that all but finitely many terms in the product are 1 so that it makes
sense.
Proof of theorem 2.18: Choose a finite set, S, of places of Q, which contains ,
all the places whose residue characteristic divides the order of M , all places
at which M is ramified and all places at which Lv 6= H 1 (Gv /Iv , M Iv ). Let
QS denote the maximal extension of Q unramified outside S and let GS =
Gal (QS /Q). Then we have an exact sequence
M
(0) HL1 (Q, M ) H 1 (GS , M )
H 1 (Gv , M )/Lv .
vS
Because H 1 (GS , M ) is finite (see cor. 4.15 of ch. 1 of [Mi]), we see that
HL1 (Q, M ) is finite.
Dualising this sequence for M and using theorem 2.17, we see that we
have an exact sequence
M
Lv H 1 (GS , M ) HL1 (Q, M ) (0).
vS
On the other hand we have the following part of the Poitou-Tate nine term
sequence (see thm. 4.10 of ch. 1 of [Mi]):
L
0
H
(G
,
M
)
/(1 + c)M H 2 (GS , M )
(0) H 0 (GS , M )
v
vS
L
1
1
H 1 (GS , M )
H
(G
,
M
)
H
(G
v
S , M ),
vS
61
0
(0)
H 0 (GS , M )
vS H (Gv , M ) /(1 + c)M
L
1
2
L
H
(Q,
M
)
H
(G
, M )
v
S
L
vS
(0).
H 1 (GS , M ) HL1 (Q, M )
Theorem 2.18 follows from this and the global Euler characteristic formula
(see thm. 5.1 of ch. 1 of [Mi]):
#H 1 (GS , M )
(#M )
=
= #(1 + c)M.
#H 0 (GS , M )#H 2 (GS , M )
#H 0 (G , M )
2
We remark, but will not use elsewhere in this article, that the above argument works over any number field:
Theorem 2.19 Let F be a number field and M a discrete GF module of finite
cardinality. For each place v of F let Gv denote a decomposition group at v
and if v is finite let Iv Gv denote the inertia group. Fix subgroups Lv
H 1 (Gv , M ), and such that for all but finitely many v, Lv = H 1 (Gv /Iv ,Q
M Iv ).
1
1
Let HL (F, M )Q(respectively, HL (F, M )) denote the inverse
Q 1 image of v Lv
1
(respectively, v Lv ) under
Q the map H (GF , M ) v H (Gv , M ) (respectively, H 1 (GF , M ) v H 1 (Gv , M )). Then HL1 (G, M ) and HL1 (G, M )
are finite and we have the formula
#Lv
#HL1 (F, M )
#H 0 (GF , M ) Y
.
=
1
0
#HL (F, M )
#H (GF , M ) v #H (Gv , M )
2.4
Representations of GQ`
62
(2.4.1)
and if the underlying profinite O[G` ]-module, M is good, ordinary or semistable. We write for mod mR . We record the following consequence of
Nakayamas lemma.
(1)
(0)
and M
Remark 2.21 (a) These definitions are somewhat ad hoc, but at the moment that is all that seems to be available (though the work of Fontaine
and Laffaille [FL] and its generalisations may well provide a more systematic setting).
(b) For part of the motivation for our definitions, see proposition 2.23 and
remark 2.24. For further motivation, we shall see in the next chapter
that representations arising from certain modular forms are semistable.
Moreover the Fontaine-Mazur conjecture predicts that, conversely, any
representation of GQ with semistable restriction to G` arises from such
a modular form. We shall state a weak form of the Fontaine-Mazur
conjecture below (conjecture 3.17).
(c) The terminology in [W3] to describe representations of G` is slightly
different. In particular, we impose the condition (2.4.1) in our definitions
of good and ordinary, as this is all we shall need here. Assuming
satisfies this condition, the notion of ordinary in [W3] coincides with the
one here, and is flat in the sense of [W3] if and only if it is good but not
ordinary. (Note that a representation may be both good and ordinary,
for instance E,` for an elliptic curve with good, ordinary reduction; see
proposition 2.23.)
63
2
the character 7 ($)/$ mod $ where $ = ` 1 `. If F is a field of
characteristic other than ` and if M is a profinite Z` [G` ]-module then we
set M (1) = lim M Z` `n (F ). If : G` GL2 (R) is ordinary the extension
of R[I` ]-modules gives rise to a class c H 1 (I` , R(1)). Kummer theory and
I` ) give rise to a map v : H 1 (I` , R(1)) H Z R R,
the valuation on (Q
`
`
I` ) . Then we also have the
where H denotes the `-adic completion of (Q
`
following lemma.
Lemma 2.25 (a) If : G` GL2 (k) is good then either is ordinary or
|I` k = ` .
(b) If : G` GL2 (R) is such that M is good and is ordinary, then is
good and ordinary.
(c) If : G` GL2 (R) is ordinary then is good if and only if v(c ) = 0.
Note that we need only consider the case that R has finite cardinality to
prove the lemma. Parts (a) and (b) again follow from Raynauds results [Ray]
(for part (b) consider the connected-etale sequences for M and M).
We sketch the proof of part (c) using an argument suggested to us by
Edixhoven. As in [Edi], prop. 8.2 it suffices to determine which extensions
(0) R(1) M R (0)
(2.4.2)
of R[I` ]-modules arise from finite flat group schemes over Znr
` , the ring of
integers of the maximal unramified extension of Z` . By prop. 17.4 of [Oo],
the extension (2.4.2) arises from a finite flat group scheme if and only if it
corresponds to an extension of sheaves of R-modules in the fpqc topology over
Znr
` . Therefore we must compute the image of
Ext1Rmod/fpqc (R, R(1)) Ext1R[I` ] (R, R(1)).
Since the sheaf Ext1 of R by R(1) vanishes, this is equivalent to computing
the image of
1
1
Hfpqc
(Znr
` , R(1)) H (I` , R(1)).
Part (c) follows from the fact this is precisely the kernel of v.
Remark 2.26 The authors expect that the theory of Fontaine and Laffaille
[FL] discussed in the next section could be used to prove that if : G`
GL2 (R) is such that M is good and is good, then det |I` is cyclotomic, i.e.,
is good.
For semistable representations : G` GL2 (O/n ) we shall define Osubmodules
Hf1 (G` , ad) Hss1 (G` , ad) H 1 (G` , ad).
65
where M
denote the maximal submodule of M where I` acts by . Let
(0)
(1)
(1)
ad denote the submodule Hom(M , M ) of ad0 . Then
Hss1 (G` , ad) = ker (H 1 (G` , ad) H 1 (I` , ad/ad(1) )).
The same is true with ad0 replacing ad.
If : G` GL2 (O) is semi-stable then we set
Hf1 (G` , ad K/O) = lim Hf1 (G` , ad (n /O)) H 1 (G` , ad K/O).
1
be the unramified characters such that 0 2 . Let
c` = (1 /2 )(Frob ` ) 1.
If c` 6= 0, then
Hss1 (G` , ad0 K/O)/Hf1 (G` , ad0 K/O)
3
The authors expect equality to hold without this hypothesis; cf. remark 2.26.
67
(a) We have
of lemma 2.29 (and this is all that is proved in [W3]). We have included the
precise formulas for the sake of completeness, since they are not much more
difficult to obtain. The only additional observation required is that if is
good, then the composite of the natural maps
H 1 (G` /I` , (ad0 /ad(1) )I` ) H 1 (G` , ad0 /ad(1) ) H 2 (G` , ad(1) )
is trivial. To prove this, rewrite the composite as
H 1 (G` /I` , O/n ) H 1 (G` , O/n ) H 2 (G` , ad(1) )
with the second map given by c where c is the class in H 1 (G` , ad(1) )
defining the extension M and apply lemma 2.25 (c). (In fact, one only needs
the easier half of lemma 2.25 (c): if is good then v(c ) = 0.)
2.5
1 |D0 = `0 ,
Im 1 + Im 0 = D.
The morphisms are O-linear maps compatible with the additional data of the
distinguished submodules and maps .
It is useful to note that if D is an object of MFO , then there is a surjection
1 0 : D/D0 D0 /`D0
D/`0 (D0 ),
and on counting orders we see that this is in fact an isomorphism. Thus there
is an isomorphism
Remark 2.32 (a) It follows on applying part (a) to M/i M for each i that
M and D(M ) are in fact (non-canonically) isomorphic as O-modules.
(b) As mentioned above, our formulation differs from that of Fontaine and
Laffaille in that we are using covariant functors. To deduce theorem 2.31
from the results in [FL], sec. 9, we define D as a quasi-inverse of the
functor U S ( [1]), where U S is defined in [FL] and [1] indicates a shift
by 1 in filtration degrees. In particular, if F is a finite flat group scheme
` ))
over Z` with an action of O, then the underlying O-module of D(F(Q
can be identified with the covariant Dieudonne module of F/F` , and 1
with F .
Suppose now that : G` GL2 (O/n ) is a good continuous representation. Let D = D(M ) be the corresponding object of MFO . Then
D
= O/n .
= (O/n )2 as an O-module while D0
Lemma 2.33 The following are isomorphic.
(a) The group of extensions of M by itself in the category of good finite
O/n [G` ]-modules.
(b) The group of extensions of D by itself in the full subcategory of MFO
consisting of objects which are O/n -modules.
(c) Pairs (1 , 0 ) where 1 HomO (D , D ), 0 HomO (D0 , D ) and
`0 = 1 |D0 , modulo pairs of the form (a,1 ,1 a, a,0 ,0 a|D0 )
where a HomO (D , D ) and aD0 D0 .
Proof: The first two groups of extensions are isomorphic by the FontaineLaffaille theorem. Following Ramakrishna [Ram] we explain how to calculate
the second group of extensions. We will write D for D . If
(0) D E D (0)
and
1a
01
0 E,0
0 0
0
= 00 E ,0
0
1 a|D0 ;
0 1
`x y . If
respect to this basis let 0 have matrix x
and
have
matrix
1
z
`z w
a1 a2
a has matrix 0 a3 then a direct calculation shows that
(a1 1 a, a0 0 a|D0 ) =
a2 z
`a2 z
(a1 a3 )y + a2 (w `x)
.
, (a3 a1 )z
(a3 a1 )`z
`za2
Thus the Ext group we want is the quotient of the set of pairs of matrices
(1 , 0 ) as above by the submodule generated by
`z w `x , z ) and ( 0 y , 0 .
z
0
`z 0
0 `z
Note that either z or w is a unit in O/ from which it follows that the Ext
group we want is isomorphic to (O/n )2 O/(z, n ). On the other hand
H 0 (I` , M ) corresponds to the largest submodule C D 0 such that 0 C = C,
i.e. C = {d D 0 : zd = 0}
2
= O/(z, n ).
Corollary 2.35 Suppose that : G` GL2 (O/n ) is a continuous good
representation. Then Hf1 (G` , ad) is isomorphic to (O/n )2 H 0 (G` , ad0 ).
Proof: If is not ordinary then H 0 (G` , ad0 ) and H 0 (I` , M ) are both trivial, so suppose that is ordinary. In this case, let 0 denote the ordinary
representation defined by
M0 = Hom(M(0) , M ) ad0 ()
(1)
and let M0
(1)
Since H 0 (I` , M0
H 1 (G` , M0
(1)
) H 1 (I` , M0
(0) M0
M0 O/n 0
2.6
Deformations of representations
In this section we shall review Mazurs theory of deformations of representations of profinite groups (see [Maz3]).
Let CO denote the category whose objects are complete noetherian local
O-algebras with residue field k and whose morphisms are O-algebra homomorphisms which are local (i.e. take maximal ideals into maximal ideals). (The
structure maps from O to every object of CO are also assumed to be local.) Let
G denote a topologically finitely generated profinite group and let denote an
absolutely irreducible representation of G into GLd (k). (In fact all we shall
use in the sequel is that k is the centraliser in Md (k) of the image of .) Let
D0 denote the category of profinite O[G]-modules with continuous morphisms.
We will let D denote a full subcategory of D0 which is closed under taking subobjects, quotients and direct products and which contains M. Note that if
M is an object of D0 and Mi is a collection of subobjects which have trivial
intersection and such
Q that each M/Mi is an object of D, then M is an object
of D, since M i M/Mi .
Let : G O be a continuous character such that det = mod . By
a lifting of of type D = (O, , D) we shall mean an object R of CO and a
continuous representation : G GLd (R) such that:
(a) mod mR = ,
(b) det = ,
(c) M is an object of D.
Note that if : R R0 and : G GLd (R) is of type D so is .
Theorem 2.36 There is a lifting univ
: G GLd (RD ) of of type D such
D
that if : G GLd (R) is any lifting of of type D then there is a unique
homomorphism of O-algebras : RD R such that is conjugate to univ
D .
73
74
D0
We define HD1 (G, End (M )) to be the image of Ext1D(n) (M, M ) in the group
H 1 (G, End (M )), and HD1 (G, End 0 (M )) to be the intersection
H 1 (G, End 0 (M )) HD1 (G, End (M )).
Lemma 2.39 There is a canonical isomorphism of k-vector spaces
Homk (mR /(, m2 ), k)
= H 1 (G, ad0 ).
RD
Proof: There is a natural bijection between Homk (mRD /(, m2RD ), k) and the
set of O-algebra homomorphisms from RD to the algebra k[] where 2 = 0 (the
correspondence associates : RD k[] to |mRD ). Hence there is a bijection
to the set of liftings : G GLd (k[]) of of type D, modulo conjugation by
elements of 1 + Md (k). On the other hand, recall from section 2.4 that there
is a natural bijection between Ext1D(1) (M, M) and the set of of all continuous
0
75
Proof: One shows in a very similar manner to the proof of lemma 2.39 that
for all n there is a natural isomorphism.
HomO (/2 , O/n )
= HD1 (G, ad0 O/n ).
2
2.7
Again in this section we assume that ` is an odd prime. Let : GQ GL2 (k)
denote a continuous absolutely irreducible representation. Suppose moreover
that det = and that is semi-stable in the sense that
|G` is semi-stable,
and if p 6= ` then #
(Ip )|`.
Note that if E/Q is a semistable elliptic curve then E,` : GQ GL2 (F` ), satisfies these conditions if it is irreducible. By a theorem of Mazur (theorem 2.9)
this will be the case if ` > 7.
Let denote a finite set of prime numbers. If R is an object of CO then
we say that a continuous lifting : GQ GL2 (R) of is of type if the
following hold.
det = .
|G` is semi-stable.
If ` 6 and |G` is good then |G` is good.
If p 6 {`} and is unramified at p then is unramified at p.
If p 6 {`} then |Ip 10 1 .
ad0 (1)
= HomO (ad0 , O/n )(1) Hom(ad0 , Q` /Z` )(1) = (ad0 ) .
The O-submodule of H 1 (Q, ad0 (1)) corresponding to HL1 (Q, (ad0 ) ) is independent of the choice of generator and we denote it H1 (Q, ad0 (1)). Thus
H1 (Q, ad0 (1)) is defined by the local conditions {L
,v } where the orthogonality is with respect to the pairing
H 1 (Gv , ad0 ) H 1 (Gv , ad0 (1)) Q` /Z`
arising from the above isomorphism. We may equivalently regard the orthogonality as being with respect to the natural perfect O-bilinear pairing defined
by the composition
H 1 (Gv , ad0 ) H 1 (Gv , ad0 (1)) H 2 (Gv , ad0 O ad0 (1))
H 2 (Gv , O/n (1))
= O/n .
Note that if p 6= ` we have that
0
1
Ip
L
,p = H (Gp /Ip , (ad )(1) ) if p 6 ;
L
,p = (0) if p .
If : GQ GL2 (O) is a lifting of of type , we will write H1 (Q, ad0
K/O) for the direct limit lim H1 (Q, ad0 n /O), and H1 (Q, ad0 (1)K/O)
univ
type , i.e. is a lifting of type and if : GQ GL2 (R) is any lifting
of type then there is a unique O-algebra homomorphism : R R such
that univ
. Moreover we have the following.
77
0
(a) If K 0 /K is a finite extension and R
is the corresponding deformation
0
0
univ 0
1.
ring then R = R O O and ( ) = univ
78
The existence part of theorem 2.41 now follows from theorem 2.36 and part (a)
follows from lemma 2.38. Parts (b) and (c) follow from lemmas 2.39 and 2.40,
and the following observation:
If : GQ GL2 (O/n ) is a lifting of type and if p 6 {l} is a prime
where ramifies then
ker (H 1 (Gp , ad0 ) H 1 (Ip , ad0 /(ad0 )Ip )) = H 1 (Gp /Ip , (ad0 )Ip ).
(2.7.1)
Equation (2.7.1) follows from the fact that the natural map H 1 (Ip , ad0 )
H 1 (Ip , ad0 /(ad0 )Ip ) is an injection (in fact an isomorphism). It is nothing
other than the map (ad0 )Ip (ad0 /(ad0 )Ip )Ip . This completes the proof
of theorem 2.41.
2
Corollary 2.43 Suppose that if ` = 3 then |GQ(3) is absolutely irreducible.
Then R can be topologically generated as an O-algebra by
X
dimk H1 (Q, ad0 (1)) + d` +
dimk H 0 (Qp , ad0 (1))
p{`}
elements, where d` = dimk Hss1 (Q` , ad0 ) dimk Hf1 (Q` , ad0 ) if ` , while
d` = 0 if ` 6 .
Proof: Note that H 0 (Q, ad0 (1)) = 0 unless ` = 3 and |GQ(3) is not absolutely irreducible. Thus according to theorem 2.18, dimk H1 (Q, ad0 ) is the
sum of the following terms.
dimk H1 (Q, ad0 (1)).
dimk Hf1 (Q` , ad0 )dimk H 0 (Q` , ad0 )dimk H 0 (R, ad0 ) 0 by proposition 2.27.
d` .
For each p {l}, dimk H 1 (Qp , ad0 ) dimk H 0 (Qp , ad0 ) =
dimk H 0 (Qp , ad0 (1)) by the local Euler characteristic formula (see theorem 2.17).
2
79
2.8
Special cases
0
is
conjugate
to
Lemma 2.44 If q Q then univ
|
for some char1
G
q
Q
0
0
q
where
q and q reduce to q
choose a basis such that univ
Q (f ) =
0 q
and q in k. It suffices to show that for any Iq , univ
Q () is diagonal in this
a
b
univ
basis. Suppose Q () = 12 + c d with a, b, c, d mRQ . Because univ
is
Q
univ
univ
univ
1
univ
q
tamely ramified at q we see that Q (f )Q ()Q (f ) = Q () . Thus
(
q /q q)b and (q /
q q)c lie in mRQ (b, c) so that (b, c) = mRQ (b, c) and
(by Nakayamas lemma) b = c = 0.
2
character GQ RQ
which is unramified outside the primes of Q and whose
restriction to Iq is q,Q for each q in Q. Thus Q factors through Q . We wish
to regard RQ as an O[Q ]-algebra, and it will be most convenient to do so
80
2
via the map which gives rise to Q
. Note the following consequence of the
lemma 2.44.
RQ /aQ RQ R .
Lemma 2.46 (a) If q Q then H 0 (Fq , ad0 ) = H 0 (Fq , ad0 (1)) = k and
H 1 (Fq , ad0 ) = H 1 (Fq , ad0 (1)) = k.
(b) RQ can be topologically generated as an O-algebra by
#Q + dimk HQ1 (Q, ad0 (1))
elements.
(c) If
As each H 1 (Fq , ad0 (1)) is one dimensional we may replace the last condition
with
M
H 1 (Fq , ad0 (1)).
H1 (Q, ad0 (1)) ,
qQn
Thus what we need show is that for each non-zero class [] H1 (Q, ad0 (1))
there is a prime q (depending on []) such that
(a) q 1 mod `n ,
(b) is unramified at q and (Frob q ) has distinct eigenvalues,
(c) resq [] H 1 (Fq , ad0 (1)) is nontrivial.
83
(2.8.1)
3
3.1
P
We suppose in this section that f =
an (f )q n is a newform of weight two and
level Nf (see definition 1.21). Let Kf denote the number field in C generated
by the Fourier coefficients an (f ). Let f denote the character of f , i.e., the
homomorphism (Z/Nf Z) Kf defined by mapping d to the eigenvalue of
hdi on f .
Recall that a construction of Shimura (section 1.7) associates to f an
abelian variety Af of dimension [Kf : Q]. This abelian variety is a certain
quotient of J1 (Nf ), and the action of the Hecke algebra on J1 (Nf ) provides an
embedding
Kf , End Q (Af ) Q.
84
We saw also that for each prime ` the Tate module T` (Af ) Z` Q` becomes a
free module of rank two over Kf Q` (lemma 1.48). The action of the Galois
group GQ on the Tate module commutes with that of Kf , so that a choice of
basis for the Tate module provides a representation
GQ GL2 (Kf Q` ).
(3.1.1)
0 .
If p divides the conductor of , then |Gp is of the form
1 0 |Gp .
(f ) If `6 |2N , then |G` is good. Moreover |G` is ordinary if and only if a` is
a unit in the ring of integers of K 0 , in which case I` (Frob ` ) is the unit
root of the polynomial X 2 a` X + `(`).
(g) If ` is odd and `||N , but the conductor of is not divisible by `, then
|G` is ordinary and I` (Frob ` ) = a` .
Proof: Part (a) was established by Shimura ([Shi2], [Shi3]). The key ingredient is the Eichler-Shimura congruence relation, theorem 1.29. Recall that
J1 (N ) has good reduction at primes p not dividing N . So the action of
Gp on T` (Af ) Z` Q` is unramified and is in fact described by the action of
Frob p GFp on the Tate module of the reduction. But this is given by the
Frobenius endomorphism F whose characteristic polynomial is computed in
corollary 1.41.
The first assertion of (b) follows from (a) on applying the Chebotarev
density theorem. The second assertion then follows on noting that (1) = 1.
Part (c) was proved by Ribet (see section 2 of [R3]). Assuming reducibility
of the representation, he applies algebraicity results of Lang and Serre to
obtain a contradiction to the estimate on the Fourier coefficients stated in
theorem 1.24.
86
Parts (d) and (e) follow from a deep result of Carayol [Ca1], Thm. (A),
building on the work of Langlands [Ll1], Deligne and others. In fact, this
result and the local Langlands correspondence characterize |Gp in terms of
|Gp and the L- and -factors at p of twists of f . The descriptions in the case
of p||N are based on the analysis of Deligne-Rapoport of the reduction mod p
of J1 (N ) (see [DR], [Ll1]).
The first assertion of (f) follows from the fact that Af has good reduction
at ` if ` does not divide N . The second assertion of (f) (respectively, all of (g))
follows from the Eichler-Shimura congruence relation (respectively, the work of
Deligne-Rapoport), and general results on `-divisible groups and the reduction
of abelian varieties; see thm. 2.2 of [W2], lemma 2.1.5 of [W1], 12 of [Gro]
and thms. 2.5 and 2.6 of [Edi]. The restriction to odd ` in (f) and (g) is made
primarily out of lack of suitable definitions.
2
Mod ` representations: We maintain the notation used in the discussion of
`-adic representations. Define
f : GQ GL2 (kf )
to be the semi-simplification of the reduction of f . (See the discussion following proposition 2.6.) Assertions analogous to those in theorem 3.1 hold for
= f , except that
The representation need not be absolutely irreducible (as in (c)). However if ` is odd, one checks using (b) that is irreducible if and only if
it is absolutely irreducible.
In (d) one only has divisibility of the prime-to-` part of Nf by N (
) in
general.
The various possibilities for mp (
) to be strictly less than the exponent of
p in N (where p 6= `) were classified independently by Carayol [Ca2] and Livne
[Liv]. We record the following consequence of their results (cf. the introduction
of [DT1]):
Proposition 3.2 Suppose that p is a prime such that p|Nf , p 6 1 mod ` and
f is unramified at p. Then tr (
f (Frob p ))2 = (p + 1)2 in kf .
Artin representations: The theory of Hecke operators and newforms (see
section 1.3) extends to modular forms on 1 (N ) of arbitrary weight. The
construction of `-adic representations associated to newforms was generalized
to weight greater than 1 by Deligne [De] using etale cohomology. There are
also Galois representations associated to newforms of weight 1 by Deligne and
Serre [DS], but an essential difference is that these are Artin representations.
87
P
Theorem 3.3 If g =
an (g)q n is a newform of weight one, level Ng and
character g , then there is an irreducible Artin representation
g : GQ GL2 (C)
of conductor Ng with the following property: If p6 |Ng , then the characteristic
polynomial of g (Frob p ) is
X 2 ap (g)X + g (p).
Remark 3.4 Note that
det(
g ) is the character of GQ corresponding to and
0 (see theorem 3.1).
g (c) is conjugate to 10 1
Remark 3.5 A basis can be chosen so that the representation g takes values
in GL2 (Kg ) (where Kg is the number field generated by the an (g)). Moreover
` and we have fixed embeddings
suppose that K is a finite extension of Q` in Q
g (p)) pf (p)
for all p6 |N `, the congruence being modulo the maximal ideal of the ring
of integers of Kf0 . Thus f is the semisimplification of the desired mod `
representation (with scalars extended to kf ).
3.2
s : GL2 (F3 ) GL2 (Z[ 2]) and applies theorem 3.9 to s . The resulting
representation arises from a weight one newform, and hence its reduction is
equivalent to f for some f (see remark 3.6).
G
In case (b), is equivalent to a representation of the form Ind GQF where
F is a quadratic extension of Q and is a character GF k . (We have
choose an embedding
here enlarged K if necessary.) Let n be the order of ,
Q(e2i/n ) , K and lift to a character : GF Z[e2i/n ] . We may
G
always choose so that the Artin representation = Ind GQF is odd, i.e.,
det((c)) = 1. (In the case ` = 2 and F real quadratic, we may have to
multiply by a suitable quadratic character of GF .) We then apply theorem 3.9
to and deduce as in case (a) that is modular.
2
Serre also proposed a refinement ([Se7], (3.2.4)) of the conjecture which
predicts that is associated to a newform of specified weight, level and character. Through work of Mazur, Ribet [R5], Carayol [Ca2], Gross [Gro] and
others, this refinement is now known to be equivalent to conjecture 3.13 if ` is
odd. (One also needs to impose a mild restriction in the case ` = 3.) See [R6]
and [Di1] for statements of the results and further references; here we give a
variant which applies to newforms of weight two. Before doing so, we assume
` is odd and define an integer (
) as follows:
(
) = 0 if |G` is good;
(
) = 1 if |G` is not good and |I` k k is of the form
a
a
0
,
or
`a
0
0 1
0
for some positive integer a < `. (Recall that is the cyclotomic character
and is the character of I` defined after lemma 2.22.)
90
(
) = 2 otherwise.
Theorem 3.15 Suppose that ` is odd and is absolutely irreducible and modular. If ` = 3, then suppose also that |GQ(3) is absolutely irreducible. Then
there exists a newform f of weight two such that
is equivalent over kf to f ;
Nf = N (
)`() ;
the order of f is not divisible by `.
Proof: The existence of such an f follows from [Di1] thm. 1.1, thm. 5.1 and
lemma 2.1, but with Nf dividing N (
)`() . By lemma 2.7 above, we see that
Nf is divisible by N (
). The divisibility of Nf by (
) follows from results in
sec. 8 of [Gro] and sec. 2.4 of [Edi].
2
`-adic representations: We again use the notation of section 3.1. Let :
GQ GL2 (K) be an `-adic representation.
Definition 3.16 We say that is modular if for some weight 2 newform f ,
is equivalent over Kf0 to f .
The notion is independent of the choices of embeddings and well-behaved under
extension of scalars by proposition 1.32 (cf. definition 3.12).
The following is a special case of a conjecture of Fontaine and Mazur [FM].
Conjecture 3.17 If : GQ GL2 (K) is an absolutely irreducible `-adic
representation and |G` is semistable (in the sense of section 2.4), then is
modular.
(Recall that for us `-adic representations are defined to be unramified at all but
finitely many primes. Recall also that if |G` is semistable, then by definition
det |I` is the cyclotomic character .)
Remark 3.18 Relatively little was known about this conjecture before Wiles
work [W3]. Wiles proves that under suitable hypotheses, the modularity of
implies that of .
Remark 3.19 The conjecture stated in [FM] is stronger than the one here;
in particular, the semistability hypothesis could be replaced with a suitable
notion of potential semistability. On the other hand, one expects that if | G` is
semistable, then it is equivalent to f (over Kf0 ) for some f on 1 (N ())0 (`)
(and on 1 (N ()) if |G` is good).
91
(3.2.1)
Applying proposition 2.6, we find that for each prime `, E,` is equivalent to
f and is therefore modular.
We finally show that (b) (a). The equality (3.2.1) holds for all primes
p not dividing Nf , which by theorem 3.1, part (d), is the conductor of E.
Since det(f ) = det(E,` ) = , we see by 3.1, Part (b) that f is trivial. By
theorem 1.27 parts (b) and (d) (or [AL] thm. 3), ap is in {0, 1} for primes
p dividing Nf . Thus Kf = Q and Af is an elliptic curve. Faltings isogeny
theorem (see [CS], sec. II.5) now tells us that E and Af are isogenous and we
conclude that E is modular.
2
Remark 3.21 Note that the equivalence (b) (c) does not require Faltings
isogeny theorem.
Proposition 3.22 If the Fontaine-Mazur conjecture (conjecture 3.17) holds
for some prime `, then the Shimura-Taniyama conjecture (conjecture 1.54)
holds. If Serres conjecture (conjecture 3.13) holds for infinitely many `, then
conjecture 1.54 holds.
92
Proof: The first assertion is immediate from proposition 3.20 and the irreducibility of E,` . See [Se7], sec. 4.6 for a proof of the second. (We have
implicitly chosen the field K to be Q` in the statements of conjectures 3.17
and 3.13, but it may be replaced by a finite extension.)
2
Remark 3.23 Note that to prove a given elliptic curve E is modular, it suffices to prove that conjecture 3.17 holds for a single ` at which E has semistable
reduction. Wiles approach is to show that certain cases of conjecture 3.13 imply cases of conjecture 3.17 and hence cases of conjecture 1.54.
3.3
Hecke algebras
In this section fix the following notation. Let ` be an odd prime, let K be a
finite extension of Q` , let O denote the ring of integers of K, let denote its
`, Q
, Q
` and
maximal ideal and k its residue field. Fix embeddings K , Q
, C. Let : GQ GL2 (k) denote a continuous representation with the
Q
following properties
(a) is irreducible,
(b) is modular,
(c) det = ,
(d) |G` is semi-stable,
(e) and if p 6= ` then #
(Ip )|`.
Let us first record the following lemma.
Lemma
3.24 The representation |GL is absolutely irreducible where L =
p
Q( (1)(`1)/2 `).
0
all p 6= `. Moreover we can check that |I` 0 1 . If ` > 3 we can use
theorem 3.15 to deduce that is modular of weight 2 and level 1 and hence
obtain a contradiction.
If ` = 3 we see that the splitting
field of is everywhere
unramified over Q( 3) and hence must equal Q( 3), a contradiction. 2
Let denote a finite set of primes. For the application to modularity of
semistable elliptic curves, it suffices to consider sets contained in where
is defined as follows
93
p{`}
Ip
As Q
is modular it follows from theorem 3.15 that for all , N 6= . Set
), we let Tp
T = f N Of0 . If p is a prime not in and not dividing `N (
mod
tr (Frob p ) = Tp . Moreover we have the following.
T
is a lift of of type and there is a unique surjection : R
(a) mod
univ
such that mod
)
and
p
.
p
p
94
1
0
jugation, c, and conjugate so that (c) = 0 1 . By conjugating mod
we may assume
that
modulo every maximal ideal it reduces to and that
1
0
mod
(c) = 0 1 . Call the corresponding basis {e+ , e }. Note that by con-
(g) and
mod
tr (cg) both do.
b
By irreducibility of we can find some GQ such that () =
cd
mod
g GQ the lower left entry of (g) lies in T (look at the upper left entry
of mod
using the irreducibility of we can find GQ such that
(g)).
Again
mod
mod
( ) = e where e T
( g)
. Looking at the lower right entry of
mod
we see that for any g GQ the upper right entry of (g) lies in T . Thus
mod
is now in fact valued in GL2 (T ) and will be our candidate for mod
. We
mod
leave the verification of the other properties of as an exercise.
2
Example 3.28 Let = f57B ,3 where f57B is the newform of level 57 discussed
in the example of section 1.6. As f57B is not congruent modulo 3 to any form of
level 19 or 3 we see by theorem 3.15 that is ramified at 19 and |G3 is not good.
On the other hand is semi-stable. The facts that (Frob 2 ) has order 8 (see the
table below) and 3|#
(I19 ) (from the discussion above) imply that : GQ
GL2 (F3 ). The table below also shows that T = {(x, y) Z3 : x y mod 3}.
There is a unique continuous representation mod
: GQ GL2 (T ) such that
mod
mod
if p6 |57 then is unramified at p and tr (Frob p ) = Tp . We have an
2
5
7
11
13
17
23
29
1
1
0
0
0
0
1
1
mod
(1, 2) (2, 1) (0, 3) (0, 3) (6, 6) (6, 3) (4, 4) (2, 10)
1 1+
0
1 1
Exercise 3.29 Show that there are three algebra homomorphisms T
Z/9Z and hence show that there are at least three liftings of of type
to Z/9Z.
95
f : GQ GL2 (O).
Remark 3.33 Most of the objects considered in the rest of the section will
depend on the choice of newform f . We also remark that to give an O-algebra
homomorphism T O is equivalent to giving a newform f in N with
coefficients in O. Indeed given such a homomorphism, there exists a newform
f in N such that the homomorphism is defined by Tp 7 ap (f ) for all p
/
with p not dividing `N (
). The fact that ap (f ) O for such p implies (using
for example parts (b) and (d) of theorem 1.27 and lemma 4.1 below) that all
the Fourier coefficients of f are in O. The uniqueness of f is a consequence of
the theory of newforms, theorem 1.22.
96
For 0 satisfying
0 ,
(3.3.1)
Note that because T0 is reduced, 0 6= (0). Also let 0 denote the kernel of
0 0 . Recall that
#0 /20 = #H1 0 (Q, ad0 f O K/O).
Remark 3.34 We have not yet shown that these groups are finite, but if
either is finite, then so is the other and their cardinality is the same.
is actually in T .
Note that if ` is in then T` := (a` (g))g T
This follows from theorem 3.1(f), which shows that T` = ` + ``1 where `
is the eigenvalue of Frob ` on the free rank one T -module M (0) where M is
the module underlying mod
(see lemma 2.20). Now that we have shown that
(3.3.2)
cp )).
p0
p0
Hp (0)
is exact.
Finally we state the second theorem on Hecke algebras which shall be
proved in section 4.4.
Theorem 3.36 If 0 and f is a newform in N with coefficients
in O, then we have
Y
0 (
cp ) .
p0
3.4
Isomorphism criteria
The main thrust of Wiles approach is to prove that in many circumstances the
map : R T is an isomorphism. For this we will need two criteria from
commutative algebra. The first was found by Wiles [W3] (but is presented here
in a slightly stronger form due to Lenstra [Len]); the second was developed by
Faltings, from the original arguments of [TW]. In both criteria the notion of
complete intersection plays a vital part.
Proofs of all the results in this section, together with some background,
references and examples, is given in chapter 5.
Definition 3.38 Suppose that A is an object of CO which is finitely generated
and free as an O-module. Then we call A a complete intersection if and only
if for some r Z0 and some f1 , ..., fr O[[X1 , ..., Xr ]] we have
A
= O[[X1 , ..., Xr ]]/(f1 , ..., fr )
(i.e. there are the same number of generators as relations).
98
O[[X1 , ..., Xr ]]
R0
T0
T,
such that
3.5
We are now in a position to deduce the main theorems. We will keep the
notation and assumptions from the start of section 3.3.
Theorem 3.42 Keep the notation and assumptions of section 3.3. Then, for
all finite sets , : R T is an isomorphism and these rings are
complete intersections.
Remark 3.43 There seems to be a deep link between the fact that is an
isomorphism and the fact that T is a complete intersection. The proof of the
theorem divides into two parts. One first proves it in the minimal case where
= . One then deduces the full theorem from this special case by a different
argument. In both these steps the facts that is an isomorphism and that
T is a complete intersection are proved simultaneously.
Proof of theorem 3.42: Note that to prove the theorem we may extend scalars
if necessary (by lemma 3.39) and hence assume that both of the following hold:
The eigenvalues of all elements of the image of are rational over k.
There is a newform f in N with coefficients in O, hence an O-algebra
homomorphism T O.
We first prove that : R T is an isomorphism and that the rings
R and T are complete intersections.
Note that according to theorem 2.49 and lemma 3.24 we can find an integer
r 0 and for each n Z>0 we can find a set Qn of r primes such that
if q Qn then q 1 mod `n ;
if q Qn then is unramified at q and (Frob q ) has distinct eigenvalues;
RQn can be topologically generated by r elements as a O-algebra.
n
O[[X1 , . . . , Xr ]]
R Qn
T Qn
where
100
T ,
Let Jn = ((Si + 1)` 1 : i = 1, . . . , r). Replacing TQn and RQn by TQn /Jn
and RQn /Jn we see that we have a Jn -structure for every n. Theorem 3.42 for
= now follows from the criterion of theorem 3.41.
We now turn to the proof of theorem 3.42 in the general case. By theorem 2.41 and theorem 3.40 we see that
#H1 (Q, ad0 f O K/O) = #O/ ,
(3.5.1)
(3.5.2)
A second application of theorems 2.41 and 3.40 allows us to deduce theorem 3.42.
2
Remark 3.44 In certain cases where = (1), the bound (3.5.1) on the order
of the Selmer group H1 (Q, ad0 f O K/O) also follows from the previous work
of Flach, by a different method. See [Fl1] for details.
Corollary 3.45 Keep the notation of theorem 3.42 and suppose that f is a
newform in N with coefficients in O.
(a) We have
#H1 (Q, ad0 f O K/O) = #O/ < ,
where was defined in equation (3.3.1) after remark 3.33.
(b) If 0 then
0
1
(0) H1 (Q, ad0 f O K/O) H
L0 (Q, ad f O K/O)
p0 Hp (0)
is exact, where the groups Hp and H` were defined before proposition 3.35.
Proof: The first part now follows as a direct consequence of theorem 3.42 and
another application of theorems 2.41 and 3.40. The second part follows from
the first, together with proposition 3.35 and theorem 3.36.
2
101
(d) det = .
Then is modular.
Proof: We let denote the set of primes in at which is ramified. Then
: GQ GL2 (O) is a deformation of of type , so there is an O-algebra
homomorphism R O such that = univ
R O. Since : R T is an
3.6
Applications
103
Remark 3.52 Rubin and Silverberg observed that an elliptic curve of the
form y 2 = x(x a)(x + b) has a twist with semi-stable reduction at all odd
primes, hence is modular by [Di2]. In fact it is shown in [DK] that their
observation together with the general results of [W3] and [TW] already imply
modularity.
Fermats Last Theorem: As was already mentioned in the introduction,
the Shimura-Taniyama conjecture for semi-stable elliptic curves (and, more
precisely, for the elliptic curves that arise in Freys construction explained in
section 2.2) implies Fermats Last Theorem.
More precisely, suppose that there is a non-trivial solution to the Fermat
equation x` + y ` = z ` , with ` > 3. By theorem 2.15 the Frey curve constructed
from this solution (cf. section 2.2) is a semistable elliptic curve E/Q whose
associated mod ` representation E,` is irreducible, unramified outside 2` and is
good at `. Serres conjecture predicts that E,` arises from a newform of weight
2 and level 2; the lowering the level result of Ribet [R5] (cf. theorem 3.15)
actually proves this, once we know that E is modular, i.e., E,` arises from a
modular form of weight 2 and some level. But this is a contradiction, since
there are no modular forms of weight two and level two: such forms would
correspond to holomorphic differentials on the modular curve X0 (2) which is
of genus 0. This contradiction completes the proof.
Values of L-functions: Also mentioned in the introduction was the relationship between the calculation of the Selmer group (3.5.1) and certain cases of
a conjecture of Bloch-Kato [BK], called the Tamagawa number conjecture. It
was in this context that partial results were obtained by Flach in [Fl1] (cf.
remark 3.44).
If f is a newform of weight 2, then one can associate to f a certain symmetric square L-function L(Symm 2 f, s). We shall recall the definition in section 4.4 and explain how a method of Hida establishes a relationship between
L(Symm 2 f, 2) and O/ in the setting of corollary 3.45. We may therefore
regard part (a) of that corollary as a relationship between L(Symm 2 f, 2) and
the size of a Selmer group. While the result is in the spirit of the Tamagawa
number conjecture of [BK], we have not verified that the relevant cases of the
conjecture can be deduced from it. We shall however state a partial result
in the context of semistable elliptic curves. The reader can consult [Fl1] and
[Fl2] for a discussion of the relation to the Tamagawa number conjecture.
104
E(C)
ENeron
ENeron
ENeron
where
is the Neron differential defined in section 1.1. Since E is modular
by theorem 3.48, a method of Shimura (see [Shi4] and the introduction of [St])
establishes the analytic continuation of L(Symm 2 E, s) to an entire function
and shows that L(Symm 2 E, 2) is a non-zero rational multiple of iE . We
now explain how to deduce the following theorem from Wiles results and a
formula of Hida, corollary 4.21.
Theorem 3.53 Suppose that E is a semistable elliptic curve and ` is a prime
such that
E,` is irreducible, and
Q
` does not divide 2
dp .
p|NE
NE L(Symm 2 E, 2)
iE
is the order of
H1 (Q, ad0 E,` Z` Q` /Z` ).
105
Hecke algebras
4.1
the reduced Hecke algebras of the form T as localizations of the full Hecke
algebras of the form TO . Before doing so we need to recall some fundamental
properties of the algebras TK , TO and Tk .
Let = H (N ) for some positive integer N and subgroup H of (Z/N Z)
(see section 1.2). We let TZ denote the subring of End (S2 ()) generated by
the operators Tn for all positive integers n and hdi for all d (Z/N Z) . If R
is a ring, then TR denotes the R-algebra TZ R. Recall that TR acts faithfully
on S2 (, R) and is finitely generated and free as an R-module. (This holds for
R = Z, hence for arbitrary R.)
We first record the following lemma:
Lemma 4.1 (a) TR is generated as an R-algebra by either of the following
sets of elements:
Tn for all positive integers n.
Tp for all primes p not dividing D, and hdi for all d in (Z/N Z) .
For a proof of (a), see [DI], prop. 3.5.1; for (b), see p. 491 of [W3].
The spectrum of TO : First note that TK and Tk are Artinian, hence have
only a finite number of prime ideals, all of which are maximal. Since TO is
finitely generated and free as an O-module, its maximal (resp. minimal) prime
ideals are those lying over the prime (resp. (0)) of O. (This follows from the
going-up and going-down theorems, [Mat] thms. 9.4 and 9.5, for example.) It
follows that the natural maps
TO , TO O K
= TK ;
and TO
T O O k
= Tk
induce bijections
{ maximal ideals of TK } { minimal primes of TO } and
{ maximal ideals of Tk } { maximal primes of TO }.
Moreover, since O is complete we have (by [Mat] thms. 8.7 and 8.15, for
example) that the natural map
Y
TO
Tm
m
107
is an isomorphism, where the product is over the finite set of maximal ideals m
of TO and Tm denotes the localization of TO at m. Furthermore each Tm is a
complete local O-algebra which is finitely generated and free as an O-module,
and each minimal prime P P
of TO is contained in a unique m.
for
Now suppose that f =
an q n is a normalized eigenform in S2 (, K)
modulo Gk -conjugacy
modulo GK -conjugacy
S2 (, K)
S2 (, k)
{ maximal ideals of TK }
{ maximal ideals of Tk }
(4.1.1)
l
l
{ minimal primes of TO }
{ maximal primes of TO }.
Proposition 4.2 The vertical maps are bijective, and the horizontal maps are
surjective.
Proof: For the injectivity of the upper-left vertical arrow, note that if p is
are
a maximal ideal of TK , then all K-algebra homomorphisms TK /p , K
obtained from a single one by composing with an element of GK . For the
surjectivity, let K 0 = TK /p and p0 denote the kernel of the natural K 0 -algebra
homomorphism TK 0
K 0 . Since TK 0 acts faithfully on S2 (, K 0 ), the localization S2 (, K 0 )p is non-zero, hence so is S2 (, K 0 )[p0 ]. (For an R-module M
and an ideal I of R, we write M [I] for the intersection over the elements r in I
of the kernels of r : M M .) It follows that there is a normalized eigenform
f in S2 (, K 0 ) so that p0 is the kernel of the f0 : TK 0 K 0 , and therefore p is
the kernel of f . To prove that the upper-right vertical arrow is bijective, note
that the above arguments carry over with K replaced by k.
2
Remark 4.3 The surjectivity of the top arrow is called the Deligne-Serre
lifting lemma ([DS], lemma 6.11).
108
m0 $ 1 (m)
m0 $ 1 (m)
m0 .
Suppose that g is in the GK -conjugacy class of eigenforms in S2 () corresponding to p. If g is a newform then TK /p is isomorphic to the field denoted
Kg0 in section 3.1 and p can be identified with the representation
g : GQ GL2 (Kg0 )
considered in theorem 3.1. If also ` is odd then g is obtained from m by
extending scalars. More generally suppose that g is not necessarily a newform
and consider the associated newform f . Let D denote the product of the primes
(D)
which divide N but not Nf . Let TK denote the K-subalgebra of TK generated
by the operators Tn for n relatively prime to D and hdi for d in (Z/N Z) . Let
0 = H 0 (Nf ) where H 0 is the image of H in (Z/Nf Z) , and let T0K = T0Z K
where T0Z is the Hecke algebra acting on S2 (0 ). Restriction of operators defines
(D)
a natural homomorphism TK T0K which is surjective by lemma 4.1. The
(D)
(D)
Kf0 factors through the field TK /(p TK ), so we may
composite with T0K
identify Kf0 with a subfield of TK /p and p is then equivalent to the extension
of scalars of f . If ` is odd, then m and f are defined and equivalent over a
common subfield of TO /m and kf .
The structure of TK : We now give an explicit description of TK in the case
that K contains the coefficients of all eigenforms of level dividing N . Let N
denote the set of newforms in S2 (), i.e., the set of newforms f of level Nf
dividing N such that H is contained in the kernel of the character
f
(Z/N Z) (Z/Nf Z) K .
By theorem 1.22
S2 (, K) =
SK,f ,
f N
For each f =
an (f )q n in N , let
p
Pf,p (uf,p ) = uf,p
TK,f .
f
dimK TK = dimC S2 (, C) =
0 (N/Nf ) =
we conclude
Lemma 4.4 There is an isomorphism of K-algebras:
Y
: TK
AK,f /IK,f
f N
defined by
(Tp )f = ap (f ) if p is a prime not dividing N/Nf ;
(Tp )f = uf,p mod IK,f if p is a prime dividing N/Nf ;
(hdi)f = f (d) if d is relatively prime to N .
Remark 4.5 It follows that the algebra TK is a complete intersection over
K in the sense that it is a finite-dimensional K-algebra of the form
K[X1 , . . . , Xr ]/(P1 , . . . , Pr )
for some r.
4.2
p|N (
)
(4.2.1)
p{`}
for all p - N `,
for all p - N `.
K,
: Tm K
f N
Tm K
(AK,f /IK,f )p ,
f N pMf
dividing N (
)` , the image of (mod
)Ip (Frob p ) in K is (f )Ip (Frob p ) = ap (f )f
and therefore
(mod
)Ip (Frob p ) = (Tp ).
Finally if = 0, then we conclude from lemma 4.1 that T contains (T` ). 2
A twisted variant: Suppose now that is a finite set of primes contained
in {`} such that if p is in then (Frob p ) has eigenvalues p and p in
k satisfying
p /p 6= p1 .
(4.2.2)
(4.2.3)
Let Q denote the set of primes in such that q 1 mod `. The set of primes Q
is therefore as in section 2.8. Choosing an eigenvalue q of (Frob q ) for each
q Q as in section 2.8, we regard RQ and hence TQQas an O[Q ]-algebra.
(Recall that Q is the the maximal quotient of (Z/( qQ q)Z) of `-power
order.)
Instead of working with 0 (N ) as in proposition 4.7, we shall now work
with the group
= 0 (N ) 1 (M ),
where
M=
pQ
p2
(4.2.4)
q.
115
f,q 0
f | G q
0 1 ,
f,q
where f,q : Gq K is a character whose reduction is the unramified character sending Frob q to q . Note that the characters f,q |Iq have `-power order. In
particular Nf /N is Q2f where Qf is the product of the primes q Q such that
f,q is ramified (theorem 3.1 (d)). Note also that there is a unique character
f : (Z/Qf Z) K
such that
Iq G Q
Gal (Q(Qf )/Q) K
coincides with the restriction of f,q to Iq for each q|Qf . (We have written f
for the character of GQ as well as the corresponding Dirichlet character. We
shall also write f for the corresponding character of Q .) Let g denote the
newform associated to the eigenform
X
f1 (n)an (f )q n S2 (0 (N ) 1 (Q2f ), K).
By proposition 2.6 and theorem 3.1 we have
g f f1 ,
g = f2 ,
Ng = N Qf and
aq (g) = (f |1
Gq f,q )(Frob q ) has reduction f,q (Frob q ) = q for all q dividing Qf .
Therefore g is in N 0 and f = g g . In particular g 7 g g is surjective and
g = gg . Injectivity follows as well on noting that if g g = g 0 g0 then
g = g0 , hence g = g 0 .
2
Recall that TO contains the operators hdi for d in (Z/M Z) , and note that
hdi 1 m for all d. Let H 0 denote the kernel of
Y
(Z/M Z)
(Z/(
q)Z)
Q .
qQ
116
P
/ m. It
Since the order of H 0 is not divisible by `, we find that dH 0 hdi
0
follows that if d H then hdi = 1 in Tm . We may therefore regard Tm
as an O[Q ]-algebra via the map d 7 hdi. Recall that TQ is considered an
O[Q ]-algebra via the map
O[Q ] RQ TQ .
If p is a prime not in Q, let xp denote the unique element of Q such that
x2
.
p = p
: Tm K
K
gN 0
(AK,g /IK,g )p ,
Tm K
gN pMg
117
such that
0 (Tp ) = (g (p)ap (g))g = (xp Tp )
for primes p
/ such that p - `N . Since TQ is generated over O by the set
of such Tp , we see that 0 (TQ ) is contained in (Tm ). On the other hand, the
image of 0 contains the image of Q , hence contains (hdi) for d in (Z/M Z)
as well as (Tp ) for p
/ with p - N `.
ItQremains to prove that (Tp ) is in the image of 0 (TQ ) for p dividing
N ` qQ q. Consider the composite
Q
RQ TQ
K.
gN 0
g g and therefore
g is
1
Q
univ
Q
(4.2.5)
has finite order then the roots of X 2 (a + d)X + 1 are roots of unity and we
deduce that this matrix is the identity.
In order to show that can be so chosen, we appeal to the following lemma
(cf. [DT2], lemma 3).
118
(4.2.6)
4.3
We shall give a proof of theorem 3.31 which is based on the q-expansion principle rather than the method of de Shalit [dS] employed in [TW]. It will be more
convenient to consider the action of the Hecke operators on the full space of
modular forms M2 (). The Riemann-Roch theorem shows that the dimension
of M2 () is g + s 1 where g is the genus of the modular curve X associated
to and s is the number of cusps on X (see for example [Shi2] thm. 2.23 or
[DI] (12.1.5)).
Eisenstein maximal ideals: One can give an explicit description of a space
of Eisenstein series G2 () so that
M2 () = S2 () G2 ().
(See for example [Hi3], lemma 5.2.)
There is a natural action on M2 () by the Hecke operators hdi for all d
(Z/N Z) and Tp for all primes p. We shall only need to use the operators hdi
Z denote the subring of End (M2 ())
and Tp for p not dividing N , and we let T
these generate. The ring TZ is commutative and is finitely generated and free
Z is finitely generated is proved for example by showing
as a Z-module. That T
that M2 (, Z), the set of forms with integer Fourier coefficients at , is stable
under the Hecke operators and contains a basis for M2 () (see [DI], cor. 12.3.12
Z acts faithfully on the
and prop. 12.4.1). Alternatively, one can show that T
cohomology of the non-compact modular curve Y (cf. section 1.3).
119
in TO .
O is Eisenstein if
We say that a maximal ideal n of T
Tp p + 1 mod n for all p 1 mod N .
One sees from the explicit description of the Eisenstein series that Tp = p+1 on
G2 () if p 1 mod N . We shall be interested in the non-Eisenstein maximal
ideals because of the following lemma (see [R5], thm. 5.2(c)).
Lemma 4.12 Suppose that ` is odd. The representation
m : GQ GL2 (TO /m)
is not Eisenstein.
is absolutely irreducible if and only if m
Proof: If m is not absolutely irreducible then m 1 2 with N divisible by
the product of the conductors of 1 and 2 . Furthermore if ` does not divide
N , then one of the characters is unramified at ` while the other coincides with
is Eisenstein. Conversely if m
is
the cyclotomic character on I` . Therefore m
Eisenstein then proposition 2.6 implies that m restricted to GQ(N ` ) has trivial
semisimplification from which it follows that m is also reducible.
2
Differentials: Suppose now that is a finite set of primes satisfying (4.2.2)
Moreover we assume that contains a prime p > 3 with p 6 1 mod `. Let
and 0 be defined as in (4.2.4) and (4.2.5); i.e.
= 0 (N ) 1 (M )
0 = 0 (N ) H (M )
A = H 0 (XA , 1XA /A ),
0A = H 0 (XA0 , 1X 0 /A ),
A
A = H 0 (XA , 1
XA /A (D)),
0 = H 0 (X 0 , 1 0 (D0 )),
A
A
X /A
A
where D (resp. D 0 ) denotes the reduced divisor defined by the cusps of X (resp.
X 0 ). The q-expansion principle and standard base-change arguments allow us
to identify these with S2 (, A), S2 (0 , A), M2 (, A) and M2 (0 , A) (see [Kat],
sec. 1.6, 1.7). We may therefore regard the first two of these as modules for
A.
TA and all of them as modules for T
Lemma 4.13 Suppose that A = k or K.
O O A
A and
0 O A
0 .
(a)
=
=
O
A
(b) The natural map X X 0 induces an isomorphism
H
0A
A .
121
Remark 4.15 The approach in [TW] to proving theorem 3.31 is based on the
result that under certain hypotheses (T` (J) Z` O)m is free of rank two over
Tm (see [W3], thm. 2.1 and its corollaries). This result generalizes work of
Mazur Ribet, and Edixhoven (see secs. 14 and 15 of ch. II of [Maz1], thm. 5.2
of [R5], [MRi] and sec. 9 of [Edi]), and the key to its proof is lemmas 4.13(d)
and 4.14. We shall instead give a proof of theorem 3.31 which is based directly
on lemma 4.13, (d).
The fact that (T` (J) Z` O)m is free of rank two over Tm actually underlies
Wiles approach to many intermediate results along the way to proving the
Shimura-Taniyama conjecture for semistable elliptic curves. We shall appeal
to a special case in the course of proving theorem 3.36 below.
Proof of the theorem: Suppose we are given a representation as in section 3.3 and a set of primes Q as in the statement of theorem 3.31. We apply
lemma
Q4.11 with G = GQ and = to choose an auxiliary prime p not dividing
6N ` qQ q such that
p 6 1 mod `
tr (
(Frob p ))2 6= (p + 1)2 .
To prove theorem 3.31 we may replace K by a larger field and assume that k
contains the eigenvalues of (Frob p ).
Let = Q {p}. Thus is a set of primes as in proposition 4.10 and
contains a prime p > 3 such that p 6 1 mod `. We let be as in (4.2.4); thus
= 0 (N ) 1 (M )
Q
where M = p2 qQ q, and we choose the maximal ideal m of TO as in proposition 4.10. Let H denote the `-Sylow subgroup of (Z/M Z) and regard TO
and hence Tm as an O[H]-algebra via d 7 hdi. In view of proposition 4.10,
theorem 3.31 is equivalent to the following:
Theorem 4.16 Tm is free over O[H].
O -module
Proof: Consider the T
= HomO (
O , O)
L
and the TO -module
L = HomO (O , O).
123
(4.3.1)
H
0
L/(a,
)L
= Homk (
k , k) = Homk (k , k)
(4.3.2)
O[H]d L
4.4
where
Lp (Symm 2 f, s) = (1 p2 (f )ps )1 (1 p (f )p (f )ps )1 (1 p2 (f )ps )1 .
(We caution the reader we have defined the Euler factors rather naively at
primes p such that p2 divides Nf .) The product converges absolutely for real
part of s > 2 and can be analytically continued to an entire function.
The calculation of : We write for the kernel of and I for the annihilator
of in T . Since T is reduced, we have I = 0 and I has finite index
in T . Note that
O/
= T /( I).
Suppose that we are given a T -module L which is finitely generated and free
over O, and such that L K is free of rank d over T O K. Suppose also
that L is endowed with a perfect O-bilinear pairing
LL O
(x, y) 7 hx, yi
L HomO (L, O)
x 7
hx, i
125
(4.4.1)
2
and
(with (e2i/` )n as the chosen generator for lim `n (C)). The composite
(4.4.2)
TO T m
= T O
arises as Tn 7 an (g) for a normalized eigenform g in S2 (0 (N ), K) whose associated newform is f . We shall write P for the kernel of this homomorphism;
thus P is the preimage in TO of the ideal of T . Note that Pm corresponds
to under the isomorphism Tm
= T .
127
where
cp (f ) =
if p 6= `
pp Lp (Symm 2 f, 2)1
2
(` (f ) ` (f ))(1 ` (f )) if p = `.
Proof: We have
hx, yiH det A =
g
wgc = 8 2 ihg, (wg)c i
where the last h, i denotes the Petersson inner product and we have used that
wg c = (wg)c for forms on 0 (N ). We then appeal to a formula of Shimura
(see [Shi5] (2.5) and [Hi1] (5.13)) to obtain
hg, (wg)c i = (48)1 [SL2 (Z) : 0 (N )]ress=2 D(g, wg, s).
128
P
By [Shi5] lemma 1, the Dirichlet series D(g, h, s) is defined by an (g)an (h)ns
and if g and h are normalized eigenforms then this has an Euler product
expression in which the factors are
(1 p (g)p (g)p (h)p (h)p2s ) (1 p (g)p (h)ps )1
(1 p (g)p (h)ps )1 (1 p (g)p (h)ps )1 (1 p (g)p (h)ps )1 ,
(where the p s and p s are defined as they were for f ). Using the recipe for
obtaining g from f (see lemma 4.6) we find that if ` is not in then
Y
D(g, wg, s) = D (f, f, s)
p1 ,
p|N/Nf
(where the subscript indicates that the primes in are removed from the
defining Euler product). If ` is in , then we must also multiply by the factor
(1 `s )1 if ` divides Nf ;
(1 `1s )1 (1 `2 (f )`s )1 (` (f ) `2s (f )) otherwise.
To obtain the formula in this last case, consider the eigenform
f = f ` (f )f (` )
on 0 (Nf `). Since
and
wNf ` f = (`f (` ) `1 (f )f )
D(f , `f (` ), s) = ` (f )`1s D(f , f, s),
we get
D(f , wNf ` f , s) = `1 (` (f ) ` (f )`2s )D(f , f, s)
and use the Euler product. The Euler product expression also gives
D (f, f, s) =
N (s 1)L (Symm 2 f, s)
,
N (2s 2)
where N (s) is the Riemann zeta function with the Euler factors removed at
primes dividing N . Thus
Q
2
6 N L (Symm f, 2) p|N/Nf p
,
ress=2 D(g, wg, s) = 2
[SL2 (Z) : 0 (N )]
with the extra factor of (1 `2 )1 or
(` (f ) ` (f ))(1 `2 (f ))1 (1 `1 )1
2
129
0, if `
/ ,
2, if ` and `|Nf ,
3, if ` and ` - Nf .
(4.4.3)
(4.4.4)
of T0m0 -modules. Note that T is also the composite of the canonical splitting
L0T , T` (J 0 ) Z` O
with (4.4.4).
The construction of T is similar for p = ` except that we have only two
copies of J and the map L2T LT is given by the matrix (1, `1 ) where ` is
the unit root of (3.3.2). For arbitrary 0 satisfying 0 we define T
as a composite of the maps defined above. It is independent of the choice of
ordering of 0 . Recall that 0T is used to denote the dual map LT L0T .
We define H : L0H LH as the adjoint of the map 0H which renders the
diagram
0H : H 1 (X, O)m H 1 (X 0 , O)m0
commutative (where the vertical maps are the natural isomorphisms and the
bottom one is dual to T ).
The second crucial result whose discussion we postpone until the next
section is the following generalization of a lemma of Ihara.
131
LH
0H
L0H ,
2
Remark 4.25 Note that to obtain the inclusion stated in theorem 3.36 it suffices to apply theorem 4.18 for , rather than both and 0 (cf. remark 4.22).
Wiles argument: The method of [W3] sec. 2.2, like that of Ribet in [R4],
is more direct than the one given above but does not explicitly illustrate the
relation with values of L-functions. The approach is simply to compute the
composite T 0T . It suffices to consider the case 0 = {p} and we suppose
first that p 6= `. Let denote the endomorphism of J 3 defined by the matrix
= w0 ( , , )w .
!
= ,
132
which can be computed for example by considering its effect on the cotangent
space S2 (0 (N ))3 . The result is that
p(p + 1)
Tp2 (p + 1) pTp
,
pTp
p(p + 1)
pTp
=
2
p(p + 1)
pTp Tp (p + 1)
which we note commutes with the action of TZ on J 3 . One then checks that
!
1
p1 Tp
T 0T = (1, p1 Tp , p1 ) (T` () Z` O)m
p1
2
2
2
= p (p 1)(Tp (p + 1) ).
The case of p = ` is similar but simpler. One uses
T
`
+
1
0
`
= w ( , )w = ` + 1 T
`
and gets
T 0T = (` ``1 )(1 `2 ).
4.5
Homological results
In this section we sketch the proofs of theorem 4.18 and lemma 4.24, but shall
often refer the reader to ch. 2 of [W3] for more details.
Multiplicities: We first consider theorem 4.18, generalizing a result proved
by Mazur in sections II.14 and II.15 of [Maz1]. Recall that LT (resp. LH )
is defined as (T` (J0 (N ))O )m (resp. H 1 (X0 (N ), O)m ) where m is a maximal
of the ideal of the Hecke algebra TO generated by the Hecke operators on
S2 (0 (N )). We are assuming moreover that m is irreducible and that one of
the following holds:
` does not divide N ;
`2 does not divide N and T`
/ m.
We wish to prove that LT and LH are free over Tm .
Since LH and LT are isomorphic as Tm -modules, it suffices to prove that
LT is free. Furthermore, to prove that LT is free, it suffices to prove that the
133
as TF` -modules, it suffices to prove the following cases of [W3] thm. 2.1.
Theorem 4.26 Let m be a maximal ideal of TZ` . Suppose that m is irreducible and that either
(a) ` does not divide N , or
(b) `2 - N and T`
/ m.
Then
The result in this case is due to Mazur and Ribet (the main result of [MRi])
and the proof is slightly easier than in the case of (b).
Returning to case (a), we appeal to a general property of the functor D of
theorem 2.31 which follows from cor. 5.11 of [Oda].
Theorem 4.27 Suppose that A is an abelian variety over Z` (i.e., the Neron
model of an abelian variety over Q` with good reduction). There is a canonical
and functorial isomorphism of vector spaces over F`
` )[`] )0
D(A(Q
= Cot 0 (A/F` ),
where Cot 0 denotes the cotangent space at the origin.
134
(4.5.1)
(b) If N > 3 and ` is an odd prime different from p, then the sequence
H1 (Y1 (N p, p2 ), Z` ) H1 (Y1 (N p), Z` )2 H1 (Y1 (N ), Z` )
x
7 (1 x, 1 x); (y, z) 7 2 y 2 z
is exact, where the maps 1 and 2 are defined by 7 and 1 and 2
by 7 p .
To prove the lemma, one first translates the statement into one about the
homology (or cohomology) of the congruence subgroups involved. Part (a)
of the lemma is due to Ihara [Ih], but see also the proof of thm. 4.1 of [R4].
Part (b) due to Wiles is more elementary and is established in the course of
proving lemma 2.5 of [W3] (see the sequence (2.13)). The result stated there
is in terms of cohomology rather than homology, but the one given here is
immediate from it.
Remark 4.29 A method of Khare [Kh], sec. 2 using modular symbols yields
an alternate proof of part (b) (after localization at a non-Eisenstein maximal
ideal).
To deduce lemma 4.24 from lemma 4.28 we also need the following result:
Lemma 4.30 Suppose that O is the ring of integers of a finite extension of Q` .
O = T
Z O
Let M be a positive integer and H a subgroup of (Z/M Z) . Let T
Z is the subring of endomorphisms of M2 (1 (M )) generated by the
where T
operators hdi and Tr for primes r - M . If n is a non-Eisenstein maximal ideal
then the localization at n of the natural map
(a) H1 (Y1 (M ), O) H1 (X1 (M ), O) is an isomorphism;
(b) H1 (X1 (M ), O) H1 (XH (M ), O) is surjective.
To prove part (a), one checks that
H1 (Y1 (M ), Z) H1 (X1 (M ), Z)
is surjective and that Tr = r + 1 on the kernel for primes r 1 mod M (cf.
proposition 4.13, part (c)).
Part (b) is most easily proved by showing that GQ acts trivially on the
cokernel of the natural map
T` (J1 (M )) T` (JH (M ))
(using [LO], prop. 6 for example) and then applying lemma 4.12.
136
Finally we explain how to use lemmas 4.28 and 4.30 to prove lemma 4.24.
It suffices to consider the case 0 = {p}. We discuss only the more difficult
case of p 6= `. We define N = N , N 0 = N p2 , m and m0 as in the preceding
section. Thus m0 is a maximal ideal of T0O where T0Z acts on S2 (0 (N 0 )). All the
maps we consider in the argument below will respect the action of the Hecke
00 where T
00 is the subring of endomorphisms of M2 (1 (N p, p2 ))
operators in T
Z
Z
generated by the operators hdi and Tr for primes r not dividing N p. We let n
00 = T
00 O. We first apply lemma 4.28(b) to
denote the preimage of m0 in T
O
Z
show that
H1 (Y1 (N p, p2 ), O)n H1 (Y1 (N p), O)2n H1 (Y1 (N ), O)n
is exact. Next we apply lemma 4.30(a) to obtain the exactness of
H1 (X1 (N p, p2 ), O)n H1 (X1 (N p), O)2n H1 (X1 (N ), O)n .
Applying lemma 4.30(b) with M = N p, we find that
H1 (X1 (N, p2 ), O)n H1 (X1 (N, p), O)2n H1 (X1 (N ), O)n
is exact. We then apply lemma 4.28(a) to conclude that the map
H1 (X1 (N, p2 ), O)n H1 (X1 (N ), O)3n
x
7 ( x, x, x)
is surjective where , and are defined respectively by 7 , 7 p and
7 p2 . Finally, we use lemma 4.30(b) again (now with M = N ) to get the
surjectivity of
H1 (X0 (N p2 ), O)n H1 (X0 (N ), O)3n ,
Commutative algebra
In this section we collect some basic facts of commutative algebra that are
used in the proof. We recall that O is the ring of integers in a finite extension
K of Q` , and that O has residue field k. Let CO denote as in section 2.6 the
category of complete noetherian local O-algebras with residue field k.
137
5.1
The category CO
: The numerical criterion of Wiles is stated more naturally
in terms of rings A in the category CO which are endowed with some extra
structure: namely, a surjective O-algebra homomorphism A : A O. Let
CO
be the category whose objects are pairs (A, A ), where A is an object of
CO and A : A O is a surjective O-algebra homomorphism, also called the
The invariant A seems less familiar at first sight. It is called the congruence
ideal. (The reason for this terminology should become clearer shortly.) We
are now ready to state Wiles numerical criterion:
Theorem 5.3 Let : R T be a surjective morphism of augmented rings.
Assume that T is finitely generated and torsion-free as an O-module, and that
T 6= (0). (And hence, in particular, #(O/T ) < .) Then the following are
equivalent:
(a) The inequality #R #(O/T ) is satisfied.
(b) The equality #R = #(O/T ) is satisfied.
(c) The rings R and T are complete intersections, and the map : R T
is an isomorphism.
Remark 5.4 The above theorem is slightly different from the one that appears in [W3], where it is assumed from the outset that the ring T is Gorenstein. In the form in which we state it above, the theorem is due to H. Lenstra
[Len]. Our presentation follows Lenstras very closely. For the original (and
slightly different) point of view, the reader should consult the appendix of
[W3].
Some examples: Before going further, it may be good to pause and consider
a b (mod n )}
with A (a, b) := a.
A = (n ).
Example 2:
A = {(a, b, c) O O O, a b c (mod )}
' O[[X, Y ]]/(X(X ), Y (Y ), XY ),
A ' O/O O/O,
139
with A (a, b, c) := a.
A = ().
Example 3:
A = O[[X]]/(X 2 ),
with A (f ) := f (0).
A ' O,
A = 0.
Example 4:
a b c d (mod ),
A = (a, b, c, d) O O,
a + d b + c (mod 2 )
' O[[X, Y ]]/(X(X ), Y (Y )),
with A (a, b, c, d) := a.
A = (2 ).
with A (f ) := f (0).
A ' On ,
A = (0).
O = Z` ).
a b (mod `2 ),
c0
(mod `)}
A ' Z/`2 Z,
A = (`2 ).
Example 7:
A = O[[T ]]/(T ) ' O kT kT 2 ,
A ' k,
5.2
with A (f ) = f (0).
A = ().
In this section we collect some of the basic properties of the invariants A and
A , and prove the equivalence of (a) and (b) in theorem 5.3.
Behaviour of A under morphisms: The assignment A 7 A is a functor
(5.2.1)
(5.2.2)
(5.2.3)
(5.2.4)
(5.2.5)
(5.2.6)
(5.2.7)
where the tensor product is taken with respect to the augmentation map A .
For more details and references on the Fitting ideal, see [Len] for example.
Now we are ready to prove equation (5.2.3). For, noting that A =
ker A A O, where the tensor product is taken with respect to A , and applying
equation (5.2.7) with M = ker A , we have:
Fitt O (A ) = A (Fitt A (ker A )) A (AnnA ker A ) = A ,
(5.2.8)
where the containment follows from equation (5.2.6). Now the inequality
(5.2.3) follows by combining (5.2.5) and (5.2.8).
As a consequence, we have
Corollary 5.6 The statements (a) and (b) in theorem 5.3 are equivalent.
Proof: If R T is a surjective map of augmented rings, then #R
#T , by equation (5.2.1). But equation (5.2.3) gives the inequality #T
#(O/T ). Hence the inequality #R #(O/T ) always holds, so that (a)
implies (b) in theorem 5.3. The reverse implication is clear.
2
(5.2.9)
5.3
In this section we show that complete intersections satisfy a Gorenstein condition, and that (c) implies (a) and (b) in the statement of Wiles isomorphism
criterion (theorem 5.3).
Definition 5.8 An object A in CO which is finite flat is said to be Gorenstein
if
HomO (A, O) ' A as A-modules.
Proposition 5.9 Suppose A in CO is finite flat. If A is a complete intersection, then A is Gorenstein.
The remainder of this section will be devoted to proving proposition 5.9. Since
the proof is a bit long and involved, and the concepts it uses are not used
elsewhere, the reader is advised on a first reading to take it on faith and skip
to the next section. A more direct proof of proposition 5.9 which bypasses the
arguments of this section is explained in [Len].
We let A be a ring which is finite flat, and is a complete intersection.
(Hence, A can be written as O[[X1 , . . . , Xn ]]/(f1 , . . . , fn ).) We assume that
the augmentation map for A is induced from the map on O[[X1 , . . . , Xn ]]
sending f to f (0). This implies that fi (0) = 0, i.e., the fi have no constant
term.
We recall some definitions from commutative algebra that we will need.
An ideal I of a local ring R is said to be primary if I 6= R and every zero
divisor in R/I is nilpotent. If (x1 , . . . , xn ) generates a primary ideal of R, and
n = dim R, then (x1 , . . . , xn ) is called a system of parameters for R.
Lemma 5.10 The sequence (f1 , . . . , fn , ) is a system of parameters for U =
O[[X1 , . . . , Xn ]].
Proof: The quotient ring U/(f1 , . . . , fn , ) is local and is finitely generated as
a k-vector space; therefore every element in its maximal ideal is nilpotent. 2
A sequence (x1 , . . . , xn ) in a ring R is said to be a regular sequence if xi is
not a zero-divisor in R/(x1 , . . . , xi1 ) for i = 1, . . . , n.
Lemma 5.11 The sequence (f1 , . . . , fn ) is a regular sequence for U .
143
p
X
t=1
(d) If x is a regular sequence, then Hp (x; R) = 0 for all p > 0 (i.e., the
complex Kp (x; R) is a free resolution of R/(x).)
Proof: The first assertion follows directly from the definition. For (b) and (c),
see [Mat], th. 16.4. The assertion (d) can be proved by a direct induction
argument on n, using the long exact homology sequence: For p > 1, this
sequence becomes
0 Hp (x, xn+1 ; R) 0,
and for p = 1, it is
xn+1
(5.3.1)
n (f ) = (f (D)).
2
We finally come to the proof of proposition 5.9, which we can state in a
more precise form.
146
The reader can check directly that is indeed an A-linear isomorphism from
the A-module HomO (A, O) to A.
147
5.4
Let A be an object of CO
, which is finite flat and is a complete intersection as
in the previous section, so that A ' O[[X1 , . . . , Xn ]]/(f1 , . . . , fn ).
Using the result of the previous section, we will give an explicit formula
for computing A in this case, and prove that (c) implies (b) in theorem 5.3.
Let A := HomO (A, O), and let A : O A be the dual map. From
the Gorenstein property of A, we may identify A with A (as A-modules). Fix
'
any identification : A A. (Any two such differ by a unit in A.) One
checks that
A (O ) = AnnA ker A .
2
148
5.5
Isomorphism theorems
The usefulness of the notion of complete intersections comes from the following
two (vaguely stated) principles:
1. Isomorphisms to complete intersections can often be recognized by looking
at their effects on the tangent spaces.
2. Isomorphisms from complete intersections can often be recognized by looking at their effects on the invariants .
These vague principles are made precise in theorems 5.21 and 5.24 respectively.
Theorem 5.21 Let : A B be a surjective morphism of augmented rings,
with B a (finite, flat) complete intersection. If induces an isomorphism from
A to B , and these modules are finite, then is an isomorphism.
Remark 5.22 Let A = O[[X, Y ]]/(X(X ), Y (Y )) be the ring of example 4, let B = O[[X, Y ]]/(X(X ), Y (Y ), XY ) be the ring of example
2, and let : A B be the natural projection. The map induces an
isomorphism A B , even though is not an isomorphism. The assumption that B is a complete intersection is crucial for concluding that is an
isomorphism.
Remark 5.23 The natural map
O[[X]]/(X 3 ) O[[X]]/(X 2 )
is a surjective morphism inducing an isomorphism on tangent spaces, and the
target ring is a complete intersection. Yet this map is not an isomorphism.
This shows that the assumption on the finiteness of the tangent spaces cannot
be dispensed with.
Proof of theorem 5.21: Recall that U = O[[X1 , . . . , Xn ]] is the augmented ring
of example 5 of section 5.1. Let
B : U B
be a surjective morphism of augmented rings with ker B = (f1 , . . . , fn ). Let
b1 , . . . , bn ker B denote the images of X1 , . . . , Xn by B , and let a1 , . . . , an
ker A denote inverse images of b1 , . . . , bn by . Since is an isomorphism
on tangent spaces, the elements ai generate (ker A )/(ker A )2 . Hence the
morphism
A : O[[X1 , . . . , Xn ]] A
149
(5.5.1)
(5.5.2)
151
5.6
A resolution lemma
Theorem 5.26 Let A be an augmented ring which is finite flat over O. Then
5.7
A = A ,
so that is an isomorphism by theorem 5.24. It follows that A is a complete
intersection.
2
5.8
Proof: We have:
#(O/T ) #T #R #(O/T ),
where the first inequality is by equation (5.2.3), the second follows from the
surjectivity of , and the third is by hypothesis. Therefore,
#T = #(O/T ),
and hence T is a complete intersection by theorem 5.27. Since the orders of
R and T are the same, induces an isomorphism between them. Hence
is an isomorphism R T , by theorem 5.21. This completes the proof.
2
153
Theorem 5.28 shows that the statements (a) and (b) in theorem 5.3 imply
the statement (c). Combining corollaries 5.6, 5.20, and theorem 5.28 completes
the proof of theorem 5.3.
5.9
A reduction to characteristic `
Lemma 5.30 Suppose that K K 0 are local fields with rings of integers
O O0 and that A is an object of CO which is finitely generated and free as
an O-module. Then A is a complete intersection if and only if A O O0 is.
Proof: One implication is clear. Let k and k 0 be the residue fields of O and O 0
respectively. By lemma 5.29 it is enough to prove that, if R is an object of Ck
which is finite dimensional as a k-vector space, then
R0 = R k k 0 is a complete intersection
R is a complete intersection.
m0
=
m0
commutes. Extending this map to an isomorphism of k 0 -algebras,
: k 0 [[X1 , . . . , Xr ]] k 0 [[Y1 , . . . , Yr ]],
one sees that I 0 = 1 (J), and hence can be generated by r elements, as
claimed.
In particular we have dimk0 (I 0 /m0 I 0 ) r, and
dimk (I/mI) = dimk0 ((I/mI) k k 0 ) dimk0 (I 0 /m0 I 0 ) r.
Nakayamas lemma now implies that I can be generated by r elements, and
hence R is a complete intersection. This proves the lemma.
2
5.10
J-structures
We now turn to the proof of theorem 3.41. In view of the last section we will
work in characteristic `. Thus let : R
T be a surjective morphism in the
category Ck , where R and T are finite dimensional as k-vector spaces. Let r
be a non-negative integer. If J k[[S1 , ..., Sr ]] and J (S1 , ..., Sr ) then by a
strong J-structure we shall mean a commutative diagram in Ck
k[[S1 , ..., Sr ]]
k[[X1 , ..., Xr ]]
R0
T0
T,
such that
155
O[[X1 , ..., Xr ]]
R0
T0
T,
is a J-structure for R T , let a = ker (R T ), and let R 00 denote the ring
Im (R0 (R/mR a (T 0 /J)) O k). Then S given by
k[[S1 , ..., Sr ]]
k[[X1 , ..., Xr ]]
R00
(T 0 /J) O k
(R/mR a) O k
T O k,
is a strong J-structure
for (R/mR a)O k T O k. If Jn is a sequence of ideals
as inTtheorem 3.41 then Jn is a sequence of nested ideals with J0 = (S1 , ..., Sr )
and n Jn = (0). Moreover if for each n there is a Jn -structure Sn for R T
then for each n there is a strong Jn -structure Sn for
(R/mR a) O k T O k.
Then by theorem 5.31 we see that this map is an isomorphism of complete
intersections. Lemma 5.29 shows that theorem 3.41 follows.
2
Before returning to the proof of theorem 5.31, let us first make some remarks about strong J-structures.
156
The set of strong J structures for all ideals J (S1 , ..., Sr ) k[[S1 , ..., Sr ]]
forms a category, with the obvious notion of morphism.
If S is a strong J-structure and if (S1 , ..., Sr ) J 0 J then there is a natural
J 0 -structure S mod J 0 obtained by replacing T 0 by T 0 /J 0 and R0 by the image
of R0 R (T 0 /J 0 ).
If R is a finite dimensional k-vector space and if J has finite index in
k[[S1 , ..., Sr ]] then there are only finitely many isomorphism classes of strong
J-structure. (This follows because we can bound the order of R 0 in any Jstructure. Explicitly we must have #R0 (#R)(#k[[S1 , ..., Sr ]]/J)dimk T .)
Lemma 5.32 Suppose that R is a finite dimensional k vector space. Suppose
T
also that {Jn } is a nested (decreasing) sequence of ideals and that J = n Jn .
If for each n a strong Jn structure exists then a strong J structure exists.
Proof: We may suppose that each Jn has finite index in k[[S1 , ..., Sr ]]. Let Sn
denote a strong Jn -structure. Let Sn,m = Sn mod Jm if m n. Because there
are only finitely many isomorphism classes of strong Jm structure, we may
recursively choose integers n(m) such that
Sn(m),m
= Sn,m for infinitely many n,
0
0
Let Sm
= Sn(m),m . Then Sm
is a strong Jm structure and if m m1 then
0
0
0
Lemma 5.33 Suppose that a strong (0) structure exists. Then the map R
T is an isomorphism, and these rings are complete intersections.
Proof: Because k[[S1 , ..., Sr ]] , T 0 (and T 0 is a finitely generated k[[S1 , ..., Sr ]]module by Nakayamas lemma, cf. [Mat], thm. 8.4) we see that the Krull
dimension of T 0 is at least r. On the other hand k[[X1 , ..., Xr ]]
T 0 and so
by Krulls principal ideal theorem this map must be an isomorphism. Thus
k[[X1 , ..., Xr ]] R0 T 0 .
Hence we have that
2
2
References
[Ant2]
[Ant3]
[Ant4]
[Ant5]
[AL]
[BLR]
[BM]
N. Boston, B. Mazur, Explicit universal deformations of Galois representations, Algebraic Number Theory, 121, Adv. Stud. Pure Math.
17, Acad. Press, Boston MA, 1989.
[Bour]
[Ca1]
[Ca2]
[Ca3]
[Co]
J. Coates, S.T. Yau, eds., Elliptic Curves, Modular Forms and Fermats Last Theorem, International Press, Cambridge, 1995.
[CPS]
[Cr]
[CR]
[CS]
G. Cornell and J.H. Silverman, eds., Arithmetic Geometry, SpringerVerlag, New York, Berlin, Heidelberg, 1986.
[CW]
[Da1]
[Da2]
[De]
P. Deligne, Formes modulaires et representations `-adiques, in: Lecture Notes in Math. 179, Springer-Verlag, New York, Berlin, Heidelberg, 1971, pp. 139172.
[DR]
P. Deligne, M. Rapoport, Les schemas de modules de courbes elliptiques, in [Ant2], pp. 143316.
[dS]
[dSL]
B. de Smit, H.W. Lenstra, Explicit construction of universal deformation rings, to appear in the Proceedings of the Conference on
Fermats Last Theorem, Boston University, August 918, 1995.
159
[DS]
[Di1]
[Di2]
[Dic1]
[Dic2]
[Dir]
[DI]
[DK]
[DO]
[Dr]
[DT1]
[DT2]
[Edi]
[Edw]
H.M. Edwards, Fermats Last Theorem: A genetic introduction to algebraic number theory, Graduate Texts in Math. 50, Springer-Verlag,
New York, Berlin, Heidelberg, 1977.
160
[Ei]
[Fa]
G. Faltings, Endlichkeitssatze f
ur abelsche Varietaten u
ber Zahlkorpern, Inv. Math. 73 (1983), 349366. (English translation in [CS].)
[Fl1]
[Fl2]
[FL]
J.-M. Fontaine and G. Laffaille, Construction de representations padiques, Ann. Sci. Ec. Norm. Super. 15 (1982), 547608.
[FM]
[Fo1]
[Fo2]
[For]
[Fr]
[Gra]
[Gre]
R. Greenberg, Iwasawa theory for p-adic representations, in: Algebraic Number Theory, in honor of K. Iwasawa, Adv. Studies in Pure
Math. 17, Academic Press, San Diego, 1989, pp. 97137.
[Gro]
B.H. Gross, A tameness criterion for Galois representations associated to modular forms modp, Duke Math. J. 61 (1990), 445517.
161
[Ha]
[HB]
D.R. Heath-Brown, Fermats Last Theorem for almost all exponents, Bull. London. Math. Soc. 17 (1985), 15-16
[He]
[Hi1]
[Hi2]
[Hi3]
H. Hida, Galois representations into GL2 (Zp [[X]]) attached to ordinary cusp forms, Inv. Math. 85 (1986), 545613.
[Hu]
[Ig]
[Ih]
Y. Ihara, On modular curves over finite fields, in: Proc. Intl. Coll.
on Discrete Subgroups of Lie Groups and Applications to Moduli,
Bombay, 1973, pp. 161202.
[Kam]
[Kat]
[Kh]
[Ki]
F. Kirwan, Complex Algebraic Curves, LMS Student Texts 23, Cambridge Univ. Press, Cambridge, 1993.
[KL]
[KM]
[Kn]
[Koc]
[Kol]
[Ku]
E. Kunz, Introduction to Commutative Algebra and Algebraic Geometry, Birkhauser, Boston, Basel, Berlin, 1985.
[La1]
[La2]
[Leg]
A.M. Legendre, Sur quelques objets danalyse indeterminee et particuli`erement sur le theor`eme de Fermat, Mem. Acad. R. Sc. de
lInstitut de France 6, Paris, 1827.
[Len]
[Liv]
[Ll1]
[Ll2]
[LO]
[Mat]
[Maz1]
163
[Maz2]
[Maz3]
[Mc]
[Mer]
L. Merel, Bornes pour la torsion des courbes elliptiques sur les corps
de nombres, Inventiones Math., to appear.
[Mes]
J-F. Mestre, Constructions polynomiales et theorie de Galois, Proceedings of the ICM, Zurich, August 1994.
[Mi]
[MRi]
B. Mazur, K. Ribet, Two-dimensional representations in the arithmetic of modular curves, Asterisque 196197 (1991), 215255.
[MRo]
[MT]
[Mu1]
[Mu2]
V.K. Murty, ed., Seminar on Fermats Last Theorem, CMS Conference Proceedings 17, AMS Publ., Providence, 1995.
[MW]
[Na]
[Oda]
T. Oda, The first De Rham cohomology group and Dieudonne modules, Ann. Sci. Ec. Norm. Super. 2 (1969), 63135.
164
[Oe]
[Ogg]
[Oo]
[Ram]
[Ray]
M. Raynaud, Schemas en groupes de type (p, p, ..., p), Bull. Soc. Math.
France 102 (1974), 241280.
[R1]
[R2]
[R3]
[R4]
[R5]
[R6]
[R7]
K.A. Ribet, Abelian varieties over Q and modular forms, 1992 Proceedings of KAIST Mathematics Workshop, Korea Advanced Institute of Science and Technology, Taejon, 1992, pp. 5379.
[R8]
[Ri]
P. Ribenboim, 13 Lectures on Fermats Last Theorem, SpringerVerlag, New York, Berlin, Heidelberg, 1979.
[Sa]
T. Saito, Conductor, discriminant and the Noether formula for arithmetic surfaces, Duke Math. J. 57 (1988), 151173.
165
[Se1]
[Se2]
[Se3]
[Se4]
[Se5]
J.-P. Serre, Abelian `-adic Representations and Elliptic Curves, Benjamin, New York, 1968.
[Se6]
J.-P. Serre, Proprietes galoisiennes des points dordre fini des courbes
elliptiques, Inv. Math. 15 (1972), 259331.
[Se7]
Gal (Q/Q),
Duke Math. J. 54 (1987), 179230.
[Sha]
[Shi1]
[Shi2]
[Shi3]
G. Shimura, On the factors of the jacobian variety of a modular function field, J. Math. Soc. Japan 25 (1973), 523544.
[Shi4]
[Shi5]
[Si1]
[Si2]
[St]
[ST]
[Su]
[Ta]
[Te]
[Th]
[Tu]
[TW]
R. Taylor, A. Wiles, Ring theoretic properties of certain Hecke algebras, Annals of Math. 141 (1995), 553572.
[Wa]
[We1]
[We2]
A. Weil, Uber
die Bestimmung Dirichletscher Reihen durch Funktionalgleichungen, Math. Ann. 168 (1967), 165172.
[W1]
[W2]
[W3]
167