MODULARITY LIFTING THEOREMS - NOTES FOR POSTECH ...
MODULARITY LIFTING THEOREMS - NOTES FOR POSTECH ...
MODULARITY LIFTING THEOREMS - NOTES FOR POSTECH ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> - <strong>NOTES</strong> <strong>FOR</strong><br />
<strong>POSTECH</strong> WINTER SCHOOL ON SERRE’S CONJECTURE<br />
TOBY GEE<br />
1. The local Langlands correspondence for GL2<br />
Let K be a finite extension of Qp, with ring of integers OK, uniformiser<br />
πK and residue field OK/πK = k. Let valK : K × ↠ Z, and let | · |K =<br />
(#k) − valK(α) . Let IK be the inertia subgroup of GK, the absolute Galois<br />
group of K, and let WK be the Weil group of K. By definition, WK is the<br />
subgroup of GK consisting of elements which map to an integral power of<br />
the Frobenius in Gk (the absolute Galois group of k). The group WK is<br />
a topological group, but its topology is not the subspace topology of GK;<br />
rather, the topology is determined by demanding that IK is open, and has<br />
its usual topology. Let FrobK ∈ GK/IK ∼ = Gk be the geometric Frobenius<br />
element.<br />
Let Knr = K IK<br />
be the maximal unramified extension of K, and let<br />
Ktame = ∪ (n,p)=1K nr (π 1/n<br />
K ) be the maximal tamely ramified extension. Then<br />
PK := Gal(K/Ktame ) is the unique Sylow pro-p subgroup of IK. Let<br />
ζ = (ζN) be a compatible system of primitive roots of unity (i.e. ζa ab = ζb).<br />
Then we have a character<br />
tζ : IK/PK ∼ −→ �<br />
Zl,<br />
defined by<br />
σ(π 1/n<br />
K )<br />
π 1/n<br />
K<br />
l�=p<br />
= ζ (tζ(σ) (mod n))<br />
n<br />
.<br />
If σ ∈ WK, then tζ(στσ −1 ) = ɛ(σ)tζ(τ), where ɛ is the cyclotomic character.<br />
We let tζ,l be the composite of tζ and the projection to Zl.<br />
Local class field theory is summarised in the following statement.<br />
Theorem 1.1. Let W ab<br />
K denote the group WK/[WK, WK]. Then there are<br />
unique isomorphisms ArtK : K × → W ab<br />
K such that<br />
(i) if K ′ /K is a finite extension, then ArtK ′ = ArtK ◦N K ′ /K, and<br />
These notes are based on a lecture course given by Richard Taylor at Harvard University<br />
in 2009.<br />
1
2 TOBY GEE<br />
(ii) we have a commutative square<br />
K × ArtK ��<br />
valK<br />
����<br />
Z<br />
W ab<br />
K<br />
����<br />
��<br />
Frob Z K<br />
where the bottom arrow is the isomorphism sending a ↦→ Frob a K .<br />
The irreducible representations of the group W ab<br />
K are just the characters<br />
of WK, and local class field theory gives a simple description of them, as<br />
representations of K × = GL1(K). The local Langlands correspondence for<br />
GLn is a kind of n-dimensional generalisation of this, giving a description<br />
of the n-dimensional representations of WK in terms of the representation<br />
theory of GLn(K).<br />
Let L be a field of characteristic 0.<br />
Definition 1.2. A representation of WK over L is a representation (on a<br />
finite-dimensional L-vector space) which is continuous if L has the discrete<br />
topology; that is, it is a representation with open kernel.<br />
A Weil-Deligne representation of WK on a finite-dimensional L-vector<br />
space V is a pair (r, N) consisting of a representation r : WK → GL(V ),<br />
and an endomorphism N ∈ End(V ) such that for all σ ∈ WK,<br />
r(σ)Nr(σ) −1 = (#k) −vK(σ) N,<br />
where vK : WK → Z is determined by σ|K<br />
nr = FrobvK(σ) K .<br />
Remark 1.3. (i) Since IK is compact and open in WK, if r is a representation<br />
of WK then r(IK) is finite.<br />
(ii) N is necessarily nilpotent.<br />
Exercise 1.4. (i) Show that if (r, V ) is a representation of WK and<br />
m ≥ 1 then the following defines a Weil-Deligne representation<br />
Sp m(r) with underlying vector space V m : we let WK act via<br />
r| Art −1<br />
K |m−1<br />
K<br />
⊕ r| Art−1 K |m−2<br />
K ⊕ · · · ⊕ r,<br />
and let N be the natural isomorphism from r| Art −1<br />
K |i−1<br />
K<br />
for each i < m − 1, and be 0 on r| Art −1<br />
K |m−1<br />
K .<br />
to r| Art−1<br />
K |i K<br />
(ii) Show that every Weil-Deligne representation is isomorphic to a direct<br />
sum of representations Sp mi (ri).<br />
(iii) Show that if (r, V, N) is a Weil-Deligne representation of WK, and<br />
K ′ /K is a finite extension, then (r|G K ′ , V, N) is a Weil-Deligne<br />
representation of WK ′.<br />
(iv) Suppose that r is a representation of WK. Show that if σ ∈ WK<br />
then for some positive integer n, r(σ n ) is in the centre of r(WK).<br />
(v) Assume further that σ /∈ IK. Show that for any τ ∈ WK there exists<br />
n ∈ Z and m > 0 such that r(σ n ) = r(τ m ).
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 3<br />
(vi) Show that for a representation r of WK, the following conditions<br />
are equivalent:<br />
(a) r is semisimple.<br />
(b) r(σ) is semisimple for all σ ∈ WK.<br />
(c) r(σ) is semisimple for some σ /∈ IK.<br />
(vii) Let (r, N) be a Weil-Deligne representation of WK. Set ˜r(σ) =<br />
r(σ) ss . Prove that (˜r, N) is also a Weil-Deligne representation of<br />
WK.<br />
Definition 1.5. We say that a Weil-Deligne representation (r, N) is Frobeniussemsimple<br />
if r is semisimple. With notation as above, we say that (˜r, N) is<br />
the Frobenius semisimplification of (r, N).<br />
Definition 1.6. If L is an algebraic extension of Ql for some l, then we say<br />
that an element A ∈ GLn(L) is bounded if it has determinant in O ×<br />
L , and<br />
characteristic polynomial in OL[X].<br />
Exercise 1.7. A is bounded if and only if it stabilises an OL-lattice in L n .<br />
Definition 1.8. Let L be an algebraic extension of Ql for some l. Then we<br />
say that r is bounded if r(σ) is bounded for all σ ∈ WK.<br />
Exercise 1.9. Show r is bounded if and only if r(σ) is bounded for some<br />
σ /∈ IK.<br />
Proposition 1.10. (Grothendieck’s monodromy theorem) Suppose that l �=<br />
p, that K/Qp is finite, and that V is a finite-dimensional L-vector space,<br />
with L an algebraic extension of Qp. Fix φ ∈ WK a lift of FrobK and a<br />
compatible system (ζN) of primitive roots of unity. If ρ : GK → GL(V )<br />
is a continuous representation then there is a finite extension K ′ /K and a<br />
uniquely determined nilpotent N ∈ End(V ) such that for all σ ∈ IK ′,<br />
ρ(σ) = exp(Ntζ,l(σ)).<br />
For all σ ∈ WK, we have ρ(σ)Nρ(σ) −1 = #k −v(σ) N. In fact, we have an<br />
equivalence of categories WD = WDζ,φ from the category of continuous representations<br />
of GK on finite-dimensional L-vector spaces to the category of<br />
bounded Weil-Deligne representations on finite-dimensional L-vector spaces,<br />
taking<br />
ρ ↦→ (V, r, N), r(τ) := ρ(τ) exp(−tζ,l(φ −vK(τ) τ)N).<br />
The functors WDζ ′ ,φ ′ and WDζ,φ are naturally isomorphic.<br />
Remark 1.11. Note that since N is nilpotent, the exponential here is just a<br />
polynomial - there are no convergence issues!<br />
We sketch the proof as a series of exercises.<br />
Exercise 1.12. (i) Show that there exists a GK-stable OL-lattice Λ ⊂<br />
V . Show that the if GK ′ is the kernel of the induced map GK →<br />
Aut(Λ/lΛ), then K ′ /K is a finite extension, and ρ(GK ′) is pro-l.<br />
Show that ρ|IK ′ factors through tζ,l : IK ′ → Zl.
4 TOBY GEE<br />
(ii) Choose σ ∈ IK ′ such that tζ,l(σ) generates tζ,l(IK ′). By considering<br />
the action of conjugation by φ, show that the eigenvalues of ρ(σ)<br />
are all l-power roots of unity. Hence show that one may make a<br />
further finite extension K ′′ /K ′ such that the elements of ρ(IK ′′) are<br />
all unipotent.<br />
(iii) Deduce the existence of a unique nilpotent N ∈ End(V ) such that<br />
for all σ ∈ IK ′′, ρ(σ) = exp(Ntζ,l(σ)). [Hint: use the logarithm map<br />
(why are there no convergence issues?).]<br />
(iv) Complete the proof of the proposition, by showing that (r, N) is a<br />
Weil-Deligne representation. Where does the condition that r is<br />
bounded come in?<br />
One advantage of Weil-Deligne representations over Galois representations<br />
is that there are no subtle topological issues: the topology on the<br />
Weil-Deligne representation is the discrete topology. This allows one to<br />
describe representations in a way that is “independent of L”.<br />
Weil-Deligne representations are the objects on the “Galois” side of the<br />
local Langlands correspondence. We now describe the objects on the “automorphic”<br />
side. These will be representations (π, V ) of GLn(K) on (usually<br />
infinite-dimensional) C-vector spaces.<br />
Definition 1.13. We say that (π, V ) is smooth if for any vector v ∈ V , the<br />
stabiliser of v in GLn(K) is open. We say that (π, V ) is admissible if it is<br />
smooth, and for any open subgroup K ⊂ V , V K is finite-dimensional.<br />
For example, a smooth one-dimensional representation of K × is the same<br />
thing as a continuous character.<br />
Fact 1.14. (i) If π is smooth and irreducible then it is admissible.<br />
(ii) Schur’s lemma is true, and in particular if π is admissible and irreducible<br />
then it has a central character χπ : K × → C × .<br />
In general these representations are classified in terms of the (super)cuspidal<br />
representations. We won’t need the details of this classification, and accordingly<br />
we won’t define the cuspidal representations.<br />
Let B be the subgroup of GL2 consisting of upper-triangular matrices.<br />
Define δ : B → K × by<br />
δ<br />
�� ��<br />
a ∗<br />
= ad<br />
0 d<br />
−1 .<br />
Given two characters χ1, χ2 : K × → C × , we may view χ1 ⊗ χ2 as a representation<br />
of B by<br />
�<br />
a<br />
χ1 ⊗ χ2 :<br />
0<br />
�<br />
∗<br />
↦→ χ1(a)χ2(d).<br />
d<br />
Then we define a representation χ1 × χ2 of GL2(K) by<br />
χ1 × χ2 = n-Ind GL2(K)<br />
B (χ1 ⊗ χ2)<br />
:= {φ : GL2(K) → C|φ(hg) = (χ1 ⊗ χ2)(h)|δ(h)| 1/2<br />
K φ(g) for all h ∈ B, g ∈ GL2(K)}
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 5<br />
where GL2(K) acts by (gφ)(g ′ ) = φ(g ′ g), and we only allow smooth φ,<br />
i.e. functions for which there is an open subgroup U of GL2(K) such that<br />
φ(gu) = φ(g) for all g ∈ GL2(K), u ∈ U.<br />
The representation χ1 × χ2 has length at most 2, but is not always irreducible.<br />
It is always the case that χ1 × χ2 and χ2 × χ1 have the same<br />
Jordan-Hölder factors. If χ1 × χ2 is irreducible then we say that it is a<br />
principal series representation.<br />
Fact 1.15. (i) χ1 × χ2 is irreducible unless χ1/χ2 = | det | ±1 .<br />
(ii) χ×χ| det | has a one-dimensional irreducible subrepresentation, and<br />
the corresponding quotient is irreducible. We denote this quotient<br />
by Sp 2(χ).<br />
We will let χ1 ⊞ χ2 denote the unique irreducible quotient of χ1 × χ2.<br />
These, and one-dimensional characters, are all the non-cuspidal irreducible<br />
admissible representations of GL2(K). We say that an irreducible smooth<br />
representation π of GL2(K) is discrete series if it is of the form Sp 2(χ) or<br />
is cuspidal.<br />
The local Langlands correspondence provides a unique family of bijections<br />
recK from the set of irreducible smooth representations of GLn(K) to the set<br />
of n-dimensional Frobenius semisimple Weil-Deligne representations of WK<br />
over C, satisfying a list of properties. In order to be uniquely determined, one<br />
needs to formulate the correspondence for all n at once, and the properties<br />
are expressed in terms of L- and ɛ-factors, neither of which we have defined.<br />
Accordingly, we will not make a complete statement of the local Langlands<br />
correspondence, but will rather state the properties of the correspondence<br />
that we will need to use. It is also possible to define the correspondence in<br />
global terms, as we will see later, and indeed at present the only proof of<br />
the correspondence is global.<br />
Fact 1.16. We now list some properties of recK for n = 1, 2.<br />
(i) If n = 1 then recK(π) = π ◦ Art −1<br />
K .<br />
(ii) If χ is a smooth character, recK(π ⊗(χ◦det)) = recK(π)⊗recK(χ).<br />
(iii) recK(Sp2(χ)) = Sp2(recK(χ)). (iv) If χ1/χ2 �= | det | ±1 , then recK(χ1 ⊞ χ2) = recK(χ1) ⊕ recK(χ2).<br />
(v) recK(π) is unramified (i.e. N = 0 and the restriction to IK is trivial)<br />
if and only if π = χ1 ⊞χ2 with χ1/χ2 �= | det | ±1 both unramified<br />
characters (i.e. trivial on O ×<br />
K ), or π = χ ◦ det for some unramified<br />
character χ. These conditions are equivalent to πGL2(OK) �= 0, in<br />
which case it is one-dimensional.<br />
(vi) π is discrete series if and only if recK(π) is indecomposable.<br />
1.1. Hecke operators. Let φ be a compactly supported C-valued function<br />
on GL2(OK)\ GL2(K)/ GL2(OK). Concretely, these are functions which<br />
vanish outside of a finite number of double cosets GL2(OK)g GL2(OK). The<br />
set of such functions is in fact a ring, with the multiplication being given by<br />
convolution. To be precise, we fix µ the (left and right) Haar measure on
6 TOBY GEE<br />
GL2(K) such that µ(GL2(OK)) = 1, and we define<br />
�<br />
(φ1 ∗ φ2)(x) = φ1(g)φ2(g −1 x)dµx.<br />
GL2(K)<br />
Of course, this integral is really just a finite sum. One can check without too<br />
much difficulty that the ring H of these Hecke operators is just C[Tp, S ±1<br />
p ],<br />
where Tp is the characteristic function of<br />
� �<br />
πK 0<br />
GL2(OK) GL2(OK)<br />
0 1<br />
and Sp is the characteristic function of<br />
� �<br />
πK 0<br />
GL2(OK)<br />
GL2(OK).<br />
0 πK<br />
The algebra H acts on an irreducible admissible GL2(K)-representation<br />
π. Given φ ∈ H, we obtain a linear map π(φ) : π → πGL2(OK) , by<br />
�<br />
π(φ)(v) = φ(g)π(g)vdµg.<br />
GL2(K)<br />
In particular, if π is unramified then π(φ) acts via a scalar on the onedimensional<br />
C-vector space π GL2(OK) . We will now compute this scalar<br />
explicitly.<br />
Exercise 1.17. (i) Show that we have decompositions<br />
�<br />
πK<br />
GL2(OK)<br />
0<br />
�<br />
�<br />
0<br />
πK<br />
GL2(OK) =<br />
πK<br />
0<br />
�<br />
0<br />
GL2(OK),<br />
πK<br />
GL2(OK)<br />
and<br />
� �<br />
πK 0<br />
GL2(OK) =<br />
0 1<br />
⎛<br />
⎝<br />
�<br />
α∈OK (mod πK)<br />
⎞<br />
� �<br />
πK α<br />
GL2(OK) ⎠<br />
0 1<br />
� � �<br />
1 0<br />
GL2(OK).<br />
0 πK<br />
(ii) Suppose that π = (χ|·|)◦det with χ unramified. Show that πGL2(OK) =<br />
π, and that Sp acts via χ(πK) 2 (#k) −1 , and that Tp acts via (#k1/2 +<br />
#k−1/2 )χ(πK).<br />
(iii) Suppose that χ1, χ2 are unramified characters and that χ1 �= χ2|·| ±1<br />
K .<br />
Let π = χ1 ⊕ χ2. Using the Iwasawa decomposition GL2(K) =<br />
B(K) GL2(OK), check that πGL2(OK) �� is �� one-dimensional, and is<br />
a b<br />
spanned by a function φ0 with φ0<br />
= χ1(a)χ2(d). Show<br />
0 d<br />
that π(Sp) acts on πGL2(OK) via (χ1χ2)(πK), and that π(Tp) acts<br />
via #k1/2 (χ1(πK) + χ2(πK)).
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 7<br />
2. Automorphic representations and Galois representations<br />
Let F be a totally real field, and let D/F be a central simple algebra of<br />
dimension 4. Letting S(D) be the set of places v of F at which D is ramified,<br />
i.e. for which D ⊗F Fv is a division algebra (equivalently, is not isomorphic<br />
to M2(Fv)), it is known that S(D) classifies D up to isomorphism, and<br />
that S(D) can be any finite set of places of D of even cardinality (so for<br />
example S(D) is empty if D = M2(F )). We will now define some spaces of<br />
automorphic forms on D × .<br />
For each v|∞ fix kv ≥ 2 and ηv ∈ Z such that kv + 2ηv − 1 = w is<br />
independent of v. These will be the weights of our modular forms. Let GD<br />
be the algebraic group over Q such that for any R-algebra Q, GD(R) =<br />
(D ⊗Q R) × . For v|∞, we define a compact mod centre subgroup of GD(Fv)<br />
as follows: if v ∈ S(D) we let Uv = GD(Fv) ∼ = H × , and if v /∈ S(D), so<br />
that GD(Fv) ∼ = GL2(R), we take Uv = R × � �<br />
a b<br />
SO(2). If γ = ∈ GL2(R)<br />
c d<br />
and z ∈ C − R, we let j(γ, z) = cz + d. One checks easily that j(γδ, z) =<br />
j(γ, δz)j(δ, z).<br />
We now define a representation (τv, Wv) of Uv over C for each v|∞. If<br />
v ∈ S(D), we have GD(Fv) ↩→ GL2(Fv) ∼ = GL2(C) which acts on C 2 , and<br />
we let (τv, Wv) be the representation<br />
(Sym kv−2 C 2 ) ⊗ (∧ 2 C 2 ) ηv .<br />
If v /∈ S(D), then we have Uv ∼ = R × SO(2), and we take Wτ = C, with<br />
τv(γ) = j(γ, i) kv (det γ) ηv−1 .<br />
We write U∞ = �<br />
v|∞ Uv, W∞ = ⊗v|∞Wv, τ∞ = ⊗v|∞τv. Let A = AQ be<br />
the adeles of Q, and let A∞ be the finite adeles. We then define SD,k,η to<br />
be the space of functions φ : GD(Q)\GD(A) → W∞ which satisfy<br />
(i) φ(gu∞) = τ∞(u∞) −1 φ(g) for all u∞ ∈ U∞ and g ∈ GD(A).<br />
(ii) There is a nonempty open subset U ∞ ⊂ GD(A ∞ ) such that φ(gu) =<br />
φ(g) for all u ∈ U ∞ , g ∈ GD(A).<br />
(iii) Let S∞ denote the infinite places of F . If g ∈ GD(A ∞ ) then the<br />
function<br />
defined by<br />
(C − R) S∞−S(D) → W∞<br />
h∞(i, . . . , i) ↦→ ⊗v∈S∞τv(hv)φ(ghv)<br />
is holomorphic. [Note that this function is well-defined by the first<br />
condition, as U∞ is the stabiliser (i, . . . , i).]<br />
(iv) If S(D) = ∅ then for all g ∈ GD(A) = GL2(AF ), then we have<br />
� � �<br />
1 x<br />
φ( g)dx = 0.<br />
0 1<br />
F \AF
8 TOBY GEE<br />
If in addition we have F = Q, then we furthermore demand that for<br />
all g ∈ GD(A ∞ ), h∞ ∈ GL2(R) + the function φ(gh∞)|ℑ(h∞i)| k/2<br />
is bounded on C − R.<br />
There is a natural action of GD(A ∞ ) on SD,k,η by right-translation, i.e.<br />
(gφ)(x) := φ(xg).<br />
Exercise 2.1. While this definition may at first sight appear rather mysterious,<br />
it is just a generalisation of the familiar spaces of cuspidal modular<br />
forms. For example, take F = Q, S(D) = ∅, k∞ = k, and η∞ = 0. Define<br />
U1(N) = {g ∈ GL2( ˆ � �<br />
∗ ∗<br />
Z)|g ≡ (mod N)}.<br />
0 1<br />
(i) Let GL2(Q) + be the subgroup of GL2(Q) consisting of matrices with<br />
positive determinant. Show that the intersection of GL2(Q) + and<br />
U1(N) inside GL2(A∞ � �<br />
) is Γ1(N), the matrices in SL2(Z) congruent<br />
1 ∗<br />
to (mod N). [Hint: what is<br />
0 1<br />
ˆ Z × ∩ Q × ?]<br />
(ii) Use the facts that GL2(A) = GL2(Q)U1(N) GL2(R) + [which follows<br />
from strong approximation for SL2 and the fact that det U1(N) =<br />
ˆZ × ] and that A × = Q × Z ˆ× ×<br />
R >0 to show that S U1(N)<br />
D,k,0 can naturally be<br />
identified with a space of functions<br />
satisfying<br />
φ : Γ1(N)\ GL2(R) + → C<br />
φ(gu∞) = j(u∞, i) −k φ(g)<br />
for all g ∈ GL2(R) + , u∞ ∈ R × >0 SO(2).<br />
(iii) Show that the stabiliser of i ∈ in GL2(R) + is R × >0 SO(2). Hence deduce<br />
a natural isomorphism between S U1(N)<br />
D,k,0 and Sk(Γ1(N)), which<br />
takes a function φ as above to the function (gi ↦→ j(g, i) kφ(g), g ∈ GL2(R) + .<br />
The case that S∞ ⊂ S(D) is particularly simple; then if U ⊂ GD(A∞ ) is<br />
open, SU D,2,0 is just the set of C-valued functions on<br />
GD(Q)/GD(A)/GD(R)U,<br />
which is a finite set.<br />
We will now examine the action of Hecke operators on these spaces.<br />
Choose an order OD ⊂ D (that is, a Z-subalgebra of D which is finitely<br />
generated as a Z-module and for which OD ⊗Z Q ∼ −→ D). For example, if<br />
D = M2(F ), one may take OD = M2(OF ).<br />
For all but finite many finite places v of F we can choose an isomorphism<br />
Dv ∼ = M2(Fv) such that this isomorphism induces an isomorphism OD ⊗OF<br />
∼<br />
OFv −→ M2(OFv). Then GD(A∞ �<br />
) is the subset of elements g = (gv) ∈<br />
v∤∞ GD(Fv) such that gv ∈ GL2(OFv) for almost all v.
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 9<br />
We now wish to describe certain irreducible representations of GD(A) in<br />
terms of irreducible representations of the GL2(Fv). More generally, we have<br />
the following construction. Let I be an indexing set and let Vi be a C-vector<br />
space. Suppose that we are given 0 �= ei ∈ Vi for almost all i. Then we<br />
define the restricted tensor product<br />
⊗ ′ {ei} Vi := lim<br />
−→<br />
J⊂I<br />
⊗i∈JVi,<br />
where the sets J contain all the places for which ei is not defined, and where<br />
the transition maps for the colimit are given by “tensoring with the ei”. It<br />
can be checked that ⊗ ′ {ei} Vi ∼ = ⊗ ′ {fi} Vi if for almost all i, ei and fi span the<br />
same line.<br />
Definition 2.2. We call a representation (π, V ) of GD(A ∞ ) admissible if<br />
(i) for any v ∈ V , the stabiliser of v is open, and<br />
(ii) for any U ⊂ GD(A ∞ ) an open subgroup, dimC V U < ∞.<br />
Fact 2.3. If πv is an irreducible smooth (so admissible) representation of<br />
GL2(OFv )<br />
GD(Fv) with πv �= 0 for almost all v, then ⊗ ′ πv := ⊗ ′ �<br />
π GL2 (OF,v )<br />
�πv<br />
v<br />
is an irreducible admissible representation of GD(A ∞ ), and any irreducible<br />
representation of GD(A ∞ ) arises in this way for unique πv.<br />
We have a global Hecke algebra, which decomposes as a restricted product<br />
of the local Hecke algebras in the following way. For each finite place v of F<br />
we choose Uv ⊂ GD(Fv) a compact open subgroup, such that Uv = GL2(OFv)<br />
for almost all v. Let µv be a Haar measure on GD(Fv), chosen such that<br />
for almost all v we have µv(GL2(OFv)) = 1. Then there is a unique Haar<br />
measure µ on GD(A∞ ) such that for any Uv as above, if we set U = �<br />
v Uv ⊂<br />
GD(A∞ ), then µ(U) = �<br />
v µv(Uv). Then there is a decomposition<br />
Cc(U\GD(A ∞ )/U)µ ∼ = ⊗ ′ {1Uv µv}Cc(Uv\GD(Fv)/Uv)µv,<br />
and the actions of these Hecke algebras are compatible with the decomposition<br />
π = ⊗ ′ πv.<br />
Fact 2.4. SD,k,η is a semisimple admissible representation of GD(A ∞ ).<br />
Definition 2.5. The irreducible constituents of SD,k,η are called the cuspidal<br />
automorphic representations of GD(A ∞ ) of weight (k, η).<br />
Fact 2.6. (Strong multiplicity one for GL2) Suppose that S(D) = ∅. Then<br />
every irreducible consituent of SD,k,η has multiplicity one. In fact if π (respectively<br />
π ′ ) is a cuspidal automorphic representation of weight (k, η) (respectively<br />
(k ′ , η ′ )) such that πv ∼ = π ′ v for almost all v then k = k ′ , η = η ′ ,<br />
and π = π ′ .<br />
Fact 2.7. (The theory of newforms) If n is an ideal of OF , write<br />
U1(n) = {g ∈ GL2( ÔF<br />
� �<br />
∗ ∗<br />
)|g ≡ (mod n)}.<br />
0 1
10 TOBY GEE<br />
If π is a cuspidal automorphic representation of GD(A ∞ ) then there is a<br />
unique ideal n such that π U1(n) is one-dimensional, and π U1(m) �= 0 if and<br />
only if n|m. We call n the conductor of π.<br />
Analogous to the theory of admissible representations of GL2(K), K/Qp<br />
finite that we sketched above, there is a theory of admissible representations<br />
of D × , D a nonsplit quaternion algebra over K. Since D × /K × is<br />
compact, any irreducible smooth representation of D × is finite-dimensional.<br />
There is a bijection JL from the irreducible smooth representations of D ×<br />
to the discrete series representations of GL2(K), determined by a character<br />
identity.<br />
Fact 2.8. (The global Jacquet-Langlands correspondence)<br />
(i) The only finite-dimensional cuspidal automorphic representations of<br />
GD(A∞ ) occur if S(D) ⊃ S∞ and kv = 2 for all v ∈ S∞, in which<br />
case there are 1-dimensional representations, which factor through<br />
the reduced determinant.<br />
(ii) There is a bijection JL from the infinite-dimensional cuspidal automorphic<br />
representations of GD(A∞ ) of weight (k, η) to the cuspidal<br />
automorphic representations π of GL2(A∞ F ) of weight (k, η)<br />
such that πv is discrete series for all finite places v ∈ S(D). Furthermore<br />
if v /∈ S(D) then JL(π)v = πv, and if v ∈ S(D) then<br />
JL(π)v = JL(πv).<br />
Fact 2.9. (The existence of Galois representations associated to cuspidal<br />
modular forms) Let π be an irreducible admissible representation of GL2(AF ).<br />
Then there is a CM field Lπ and for each finite place λ of Lπ a continuous<br />
irreducible Galois representation<br />
such that<br />
rλ(π) : GF → GL2(Lπ,λ)<br />
(i) if πv is unramified and v does not divide the residue characteristic of<br />
λ, then rλ(π)|GFv is unramified, and the characteristic polynomial<br />
of Frobv is X2 − tvX + (#kv)sv, where tv and sv are the eigen-<br />
GL2(OFv )<br />
values of Tv and Sv respectively on πv . [Note that by the<br />
Chebotarev density theorem, this already characterises rλ(π) up to<br />
isomorphism.]<br />
(ii) More generally, WD(rλ(π)|GFv )F −ss ∼ = recFv(πv ⊗ | det | −1/2 ).<br />
(iii) If v divides the residue characteristic of λ then rλ(π)|GFv is de Rham<br />
with τ-Hodge-Tate weights ητ , ητ + kτ − 1, where τ : F ↩→ Lπ ⊂ C<br />
is an embedding lying over v. If πv is unramified then rλ(π)|GFv is<br />
crystalline.<br />
(iv) If cv is a complex conjugation, then det rλ(π)(cv) = −1.<br />
Fact 2.10. (Base change) Let E/F be a cyclic extension of totally real fields<br />
of prime degree. Let Gal(E/F ) = 〈σ〉 and let Gal(E/F ) ∨ = 〈δ E/F 〉. Let π be
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 11<br />
a cuspidal automorphic representation of GL2(A∞ F ) of weight (k, η). Then<br />
there is a cuspidal automorphic representation BCE/F (π) of GL2(AE) of<br />
weight (BCE/F (k), BCE/F (η)) such that<br />
(i) for all finite places v of E, recEv(BCE/F (π)v) = (recF (π v|F v|F ))|WEv .<br />
In particular, rλ(BCE/F (π)) ∼ = rλ(π)|GE .<br />
(ii) BCE/F (k)v = kv|F , BCE/F (η)v = ηv|F .<br />
(iii) BCE/F (π) ∼ = BCE/F (π ′ ) if and only if π ∼ = π ′ ⊗ (δi E/F ◦ ArtF ◦ det)<br />
for some i.<br />
(iv) A cuspidal automorphic representation π of GL2(A∞ E ) is in the image<br />
of BCE/F if and only if π ◦ σ ∼ = π.<br />
Definition 2.11. We say that r : GF → GL2(Q p) is modular (of weight (k, η))<br />
if it is isomorphic to i(rλ(π)) for some cuspidal automorphic representation<br />
π (of weight (k, η)) and some i : Lπ ↩→ Q p lying over λ.<br />
Proposition 2.12. Suppose that r : GF → GL2(Qp) is a continuous representation,<br />
and that E/F is a solvable Galois extension of totally real fields.<br />
Then r|GE is modular if and only if r is modular.<br />
Exercise 2.13. Prove the above proposition as follows.<br />
(i) Use induction to reduce to the case that E/F is cyclic of prime<br />
degree.<br />
(ii) Suppose that r|GE is modular, say ∼= r|GE i(rλ(π)). Use strong<br />
multiplicity one to show that π ◦ σ ∼ = π. Deduce that there is an<br />
automorphic representation π ′ such that BCE/F (π ′ ) = π.<br />
(iii) Use Schur’s lemma to deduce that there is a character χ of GF such<br />
that r ∼ = i(rλ(π ′ )) ⊗ χ. Conclude that r is modular.<br />
We can make use of this result to make considerable simplifications in our<br />
proofs of modularity lifting theorems. It is frequently employed in conjunction<br />
with the following fact from class field theory.<br />
Fact 2.14. Let K be a number field, and let S be a finite set of places of<br />
K. For each v ∈ S, let Lv be a finite Galois extension of Kv. Then there<br />
is a finite solvable Galois extension M/K such that for each place w of M<br />
above a place v ∈ S there is an isomorphism Lv ∼ = Mw of Kv-algebras.<br />
Note that we are allowed to have infinite places in S, so that if K is totally<br />
real we may choose to make L totally real by an appropriate choice of the<br />
Lv.<br />
3. Modularity lifting theorems<br />
Assume that p > 2 is prime, let F/Q be totally real, and let ρ : GF →<br />
GL2(Q p) be a continuous irreducible representation. We will now be a little<br />
slack with our notation for Galois representations attached to automorphic<br />
representations, being content to write rp(π) for some i(rλ(π)) as above. We
12 TOBY GEE<br />
want to prove that ρ is modular. For this to be true, we will have to assume<br />
some conditions. Firstly:<br />
◦ ρ is unramified at almost all primes, and is de Rham at places<br />
dividing p, with distinct Hodge-Tate weights.<br />
◦ ρ is odd, i.e. det ρ(cv) = −1 for any complex conjugation cv.<br />
These conditions are conjecturally enough. However, we can’t prove a theorem<br />
anything like as strong as this. There are a number of assumptions<br />
that we need to make that appear to be intrinsic to the method.<br />
◦ ρ is modular, say ρ ∼ = ¯rp(π).<br />
◦ ρ|G F (ζp) is irreducible.<br />
◦ π has the same weight as ρ, in a sense that we will make precise<br />
below.<br />
There is another assumption that we will make that can be considerably<br />
relaxed, but not as yet removed.<br />
◦ If v|p is a place of F , then Fv/Qp is unramified, and the Hodge-Tate<br />
weights of ρ are all {0, k − 1} with 0 ≤ k ≤ p − 1. Furthermore πv<br />
is unramified for all v|p, and π has weight (k, 0).<br />
We begin by making a number of simplifications. Firstly, we use Grothendieck’s<br />
monodromy theorem, local-global compatibility, and solvable base change,<br />
which together show that we may assume that:<br />
◦ At every finite place v ∤ p of F , ρ|GFv is either unramified, or has<br />
unipotent ramification.<br />
◦ Each πv is either unramified or is Sp2(χ) with χ unramified.<br />
Less obviously, but explained in Ken Ribet’s talk, we can assume that<br />
◦ Each πv is unramified.<br />
[Sketch: base change so that if v is a place with πv Steinberg, than Nv ≡ 1<br />
(mod p). Then produce a congruence between the Steinberg and a ramified<br />
principal series, and then make another base change to replace this with an<br />
unramified principal series.] For the method we’re going to use, we will also<br />
make a further base change and assume that:<br />
◦ If v ∤ p is a place at which ρ|GFv is ramified, then Nv ≡ 1 (mod p),<br />
and ρ(GFv) = 1.<br />
We can finally also assume that det ρ = det rp(π); the ratio of these characters<br />
is everywhere unramified and residually trivial, thus has finite p-power<br />
order, and in particular is necessarily trivial on each complex conjugation,<br />
so can be trivialised by base change.<br />
By a standard argument using the Baire category theorem and the fact<br />
that GF is compact, we may assume that ρ : GF → GL2(K), K/Qp finite.<br />
Again using the compactness of GF , after conjugation we may assume that<br />
ρ : GF → GL2(O), where O is the ring of integers of K, with maximal ideal λ<br />
and residue field O/λ = k. We will sometimes assume that K is sufficiently
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 13<br />
large (so that various polynomials have roots in K, for example), but the<br />
reader can check that we only ever require a finite extension.<br />
For simplicity of exposition, we will now restrict to a special case. This<br />
is purely to make our notation simpler; the arguments that we make go<br />
through in exactly the same way in the general case. We will restrict to the<br />
case F = Q, and will assume that all of the above assumptions apply, and<br />
that ρ is only ramified at one prime other than p. Thus we want to prove<br />
the following theorem.<br />
Theorem 3.1. Let p > 2 be prime and let ρ : GQ → GL2(O) be irreducible.<br />
Suppose that<br />
◦ ρ|G is irreducible.<br />
F (ζp)<br />
◦ ρ is unramified except for at p and some prime l.<br />
◦ ρ|GQ has unipotent ramification, l ≡ 1 (mod p), and ρ(GQl ) = 1.<br />
l<br />
◦ There is a cuspidal automorphic representation π of GL2(A) of<br />
weight (k, 0), with πv unramified for all primes v and 2 ≤ k ≤ p−1,<br />
such that ρ ∼ = ¯rp(π), and det ρ = det rp(π).<br />
is crystalline with Hodge-Tate weights (0, k − 1).<br />
◦ r|GQp<br />
Then ρ is modular.<br />
Write χ = det ρ from now on.<br />
3.1. Local deformation theory. We will need to consider several lifting<br />
rings for ρ|GQ (which is the trivial 2-dimensional representation of GQl l ,<br />
with l ≡ 1 (mod p)). Write rloc l : GQl → GL2(Rloc l ) for the universal lifting<br />
with determinant χ. Choose φ ∈ GQl a lift of the Frobenius, and σ ∈<br />
IQl a generator of tame inertia, so that φσφ−1 = σl . Note that rloc l is<br />
determined by the data of the two matrices rloc l (φ), rloc l (σ) satisfying this<br />
relation; the deformation theory is genuinely explicit in this case! With some<br />
(non-trivial) messing around with pairs of two by two matrices, one can prove<br />
the following facts (or for a more sophisticated approach, which establishes<br />
the analogous results for n-dimensional representations, see [Tay08]).<br />
Definition 3.2. (i) Let Pnr be the minimal ideal of R loc<br />
r loc<br />
l (σ) = 12.<br />
l<br />
modulo which<br />
(ii) For any root of unity ζ, we let Pnr be the minimal ideal of R loc<br />
l<br />
modulo which r loc<br />
l (σ) has characteristic polynomial (X − ζ)(X −<br />
ζ−1 ).<br />
(iii) Let Pmult be the minimal ideal of Rloc modulo which rloc<br />
l (σ) has<br />
l<br />
characteristic polynomial (X−1) 2 , and l(tr rloc l (φ)) 2 = (1+l) 2 det rloc l (φ).<br />
[The motivation for the definition of Pmult is that we are attempting to<br />
describe the unipotent liftings, and if you assume that rloc � �<br />
1 1<br />
l (σ) = ,<br />
0 1<br />
this is the relation forced on rloc l (φ). All but one Qp-point of Rloc l /Pmult will<br />
have N �= 0, in fact.]
14 TOBY GEE<br />
Fact 3.3. The minimal primes of Rloc l are precisely √ Pnr, √ �<br />
Pmult, and the<br />
Pζ for ζ �= 1. We have √ P1 = √ Pnr ∩ √ Pmult.<br />
Write Rl,1, Rl,ζ, Rl,nr, Rl,mult for the corresponding quotients of R loc<br />
l .<br />
Fact 3.4. (i) Rl,ζ[1/p] is formally smooth and geometrically irreducible.<br />
(ii) Rl,nr is formally smooth over O (and thus geometrically irreducible).<br />
(iii) Rl,mult[1/p] is formally smooth and geometrically irreducible.<br />
(iv) Spec Rl,1 = Spec Rl,nr∪Spec Rl,mult and Spec Rl,1/λ = Spec Rl,nr/λ∪<br />
Spec Rl,mult/λ are both a union of two irreducible components.<br />
We will make use of these facts later, and also of the trivial but crucial<br />
fact that Rl,1/λ = Rl,ζ/λ.<br />
3.2. Global deformations. Let ? denote one of 1, ζ, nr, mult. Let R be a<br />
complete local Noetherian O-algebra with residue field k. By a deformation<br />
to R of ρ of type ? we mean a ker(GL2(R) → GL2(k))-conjugacy class of<br />
liftings of ρ to r : GQ → GL2(R) such that:<br />
(i) r unramified outside {l, p}, and det r = χ.<br />
(ii) The map Rloc l → R determined by ρ|GQ factors through Rl,?.<br />
l<br />
(iii) The map Rloc p → R determined by ρ|GQp factors through Rp,k, the<br />
universal lifting ring coming from Fontaine-Laffaille theory.<br />
A universal deformation of type ? exists, which we denote ρuniv ? : GQ →<br />
GL2(Runiv ? ). Note that ρ, the deformation that we wish to prove is modular,<br />
is a deformation of type 1.<br />
A framed deformation of type ? is a ker(GL2(R) → GL2(k))-conjugacy<br />
class of (r, {rq}q=l,p, {αq}q=l,p) where r lifts ρ and is as above, the rq are<br />
= αqrqα −1<br />
q .<br />
lifts of r|GQq , and the αq ∈ ker(GL2(R) → GL2(k)) satisfy r|GQq<br />
Here conjugacy by β replaces r by βrβ −1 , leaves rq alone, and replaces αq<br />
by βαq.<br />
There is a universal framed deformation over R� ? , and there are natural<br />
maps Rloc q → R� ? , determined by the rq. Let Rloc ? = Rp,k ˆ⊗Rl,?, which has<br />
relative dimension 7 over O. We also have a map Runiv ? → R� ? , and a non-<br />
canonical isomorphism Runiv ? [[A1, . . . , A7]] ∼ = R� ? , given by choosing matrices<br />
αq.<br />
A standard calculation, together with the Poitou-Tate sequence, shows<br />
that R� ? is topologically generated over Rloc<br />
? by<br />
dim H 1 l,p,⊥ (GQ, ad 0 ρ(1))<br />
elements, where “H1 l,p,⊥ ” denotes the cohomology classes that are unramified<br />
outside l and p. We denote this quantity by u.<br />
Lemma 3.5. For all N ≥ 1 there is a finite set QN of primes such that<br />
#QN = u, and for all q ∈ QN,<br />
(i) q ≡ 1 (mod p N ),<br />
(ii) ρ(Frobq) has distinct eigenvalues,
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 15<br />
and H 1 l,p,⊥,QN (GQ, (ad 0 ρ)(1)) = (0) (this is defined in the same way as above,<br />
except that we demand that our classes are trivial at places in QN).<br />
This may be proved by the Cebotarev density theorem.<br />
We need to understand the deformations that we’re permitting at places<br />
in QN.<br />
Lemma 3.6. Suppose that (q − 1) is exactly divisible by pm , and that<br />
ρ(Frobq) has distinct eigenvalues. Then the universal lifting ring for lifts of<br />
ρ|GQq with determinant χ|GQq is isomorphic to O[[x, y, B, u]]/((1+u)pm −1),<br />
with the universal lift being given by<br />
� �−1 � � � �<br />
1 y α + B 0 1 y<br />
φ ↦→<br />
,<br />
x 1 0 χ(φ)/(α + B) x 1<br />
where α is a lift to O of one of the eigenvalues of ρ(φ), and<br />
�<br />
1<br />
σ ↦→<br />
x<br />
�−1 �<br />
y 1 + u<br />
1 0<br />
0<br />
(1 + u) −1<br />
� �<br />
1<br />
x<br />
�<br />
y<br />
.<br />
1<br />
Exercise 3.7. Prove this lemma as follows.<br />
� �<br />
α 0<br />
(i) Suppose that ρ(φ) = . Let ρ : GQq 0 β<br />
→ GL2(A) be a lift<br />
of ρ. By Hensel’s lemma, there are a, b ∈ mA such that ρ(φ) has<br />
characteristic polynomial (X − (α + a))(X − (β + b)). Show that<br />
there are x, y ∈ mA such that<br />
� � � �<br />
1<br />
1<br />
ρ(φ) = (α + a)<br />
x<br />
x<br />
and<br />
ρ(φ)<br />
� � � �<br />
y<br />
y<br />
= (β + b)<br />
1<br />
1<br />
(ii) Since ρ|GQq is unramified, ρ(σ) = 1, so we may write<br />
�<br />
1<br />
x<br />
�−1 �<br />
y 1<br />
ρ(σ)<br />
1 x<br />
� �<br />
y 1 + u<br />
=<br />
1 w<br />
�<br />
v<br />
1 + z<br />
with u, v, w, z ∈ mA. Use the commutation relation between ρ(φ)<br />
and ρ(σ) to show that v = w = 0.<br />
(iii) Use the fact that χ|GQq is unramified to show that 1+z = (1+u)−1 .<br />
(iv) Show that (1 + u) q = 1 + u, and deduce that (1 + u) q−1 = 1.<br />
(v) Deduce that (1 + u) pm = 1.<br />
(vi) Complete the proof of the lemma.<br />
The rather vague moral to take away from this calculation is that this<br />
local lifting ring is not smooth, but that if we were able to pass to a limit<br />
with m we would obtain a smooth ring. This in effect is what we will do.
16 TOBY GEE<br />
Define R� as above, except that we no longer require the deformations<br />
?,QN<br />
to be unramified at q ∈ QN. R� is still generated by u elements over<br />
?,QN<br />
Rloc ? , by another duality calculation.<br />
�<br />
Write ∆QN = q∈QN Sylp((Z/qZ) × ). For simplicity of notation, let’s suppose<br />
that this is isomorphic to (Z/pNZ) u (again, the general case works in<br />
the same way, but the notation is slightly more complicated). By the above,<br />
∼= χα ⊕ χβ, and we have a homomor-<br />
runiv |GQq is reducible, say runiv<br />
?,QN ?,QN |GQq<br />
phism Sylp((Z/qZ) × ) → (Runiv ?,QN )× induced by local class field theory and<br />
χα|IQq . Taking the product over q ∈ QN, we have a map ∆QN → (Runiv<br />
?,QN )× .<br />
So R� ?,QN is a module over O[∆QN ], and if we let aN denote the augmentation<br />
ideal of O[∆QN ], then R� ?,QN /aN = Runiv ? .<br />
We now have to introduce modular forms into the picture. At this point<br />
we will be rather brief, as to give the full details would take a prohibitive<br />
amount of space. This is also the one place where the theory is better<br />
developed over a totally real field other than Q; indeed, the optimal choice<br />
of field is a field F with [F : Q] even. We then choose a division algebra<br />
D with S(D) = ∞. We now define spaces of “algebraic modular forms” for<br />
GD(A∞ ) ∼ = GL2(A∞ F ). These are an integral p-adic avatar of the spaces of<br />
modular forms considered above, and we can use them to study congruences<br />
between modular forms, and in particular to relate the global deformation<br />
rings we are considering to Hecke algebras.<br />
Let U = �<br />
v Uv ⊂ GL2(A∞ F ) be a compact open subgroup, with Uv =<br />
GL2(OFv) for all v|p. Fix χ0 : (A∞ F )× /F × → O × a continuous character.<br />
Let Λk denote the representation of GL2(OF,l) given by<br />
⊗τ:F ↩→K Sym k−2 O 2 .<br />
If A is a finite O-module, we let Sk,χ0 (U, A) denote the space of functions<br />
satisfying<br />
φ : D × \ GL2(A ∞ F ) → Λ ⊗O A<br />
φ(guz) = χ0(z)u −1<br />
p φ(g)<br />
for all g ∈ GL2(A∞ F ), u ∈ U, z ∈ (A∞ F )× . We will sometimes write S(U, A)<br />
for Sk,χ0 (U, A). Note that S(U, O) is a finite free O-module.<br />
We have an action of Hecke operators on S(U, A); if we let S be the set<br />
of places v of F which either divide p or for which Uv �= GL2(OFv), then<br />
we have an action of the Hecke operators Tv, Sv at all places v /∈ S, via<br />
the usual formulas we found above. Let TU denote the O-subalgebra of<br />
EndO(Sk,χ0 (U, O)) generated by these operators; this is a commutative Oalgebra,<br />
and is finite and free as an O-module. It is not so hard to relate<br />
these spaces to the spaces of modular forms we considered previously.<br />
∼<br />
Fact 3.8. Fix an isomorphism ı : Qp −→ C. Then the ı-linear ring homomorphisms<br />
θ : TU → C are in natural bijection with the cuspidal automorphic<br />
representations of GD(A∞ F ) of weight (k, 0) such that πv is unramified
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 17<br />
for all v /∈ S and for all v|p, and such that the central character of π is<br />
z ↦→ ı((Nzp) k−2χ0(z ∞ ))(Nz∞) 2−k . The correspondence is given by π ↦→ θ,<br />
where for example θ(Tv) is the eigenvalue of Tv on ı−1 � �<br />
GL2(OFv )<br />
π .<br />
We wish to construct a Galois representation ρU : GF → GL2(TU). This<br />
should be compatible with the above bijection in the sense that if π corresponds<br />
to θ : TU → C as above, then the composite GF → GL2(TU) →<br />
GL2(Q p) (given by ı −1 θ) is just rp(π) up to isomorphism.<br />
This representation is constructed by “patching together” the representations<br />
rp(π). We will not give the details, noting only that the key input<br />
is the following result of Carayol.<br />
Fact 3.9. Let R be a complete local Noetherian O-algebra with residue field<br />
k, and let ρ1, ρ2 : Γ → GLn(R) be continuous representations, where Γ is a<br />
topologically finitely generated profinite group. Assume that ρ 1 = ρ 2 and is<br />
absolutely irreducible. Then:<br />
(i) if a ∈ GLn(R) and aρ1a −1 = ρ1 then a ∈ R × .<br />
(ii) if tr ρ1 = tr ρ2 then there exists a ∈ ker(GLn(R) → GLn(k)) such<br />
that aρ1a −1 = ρ2.<br />
(iii) if S is a closed subring of R and is also a complete local Noetherian<br />
O-algebra with residue field k, with mS = mR ∩ S, and tr(ρ(GK)) ⊂<br />
S, then there is an a ∈ ker(GLn(R) → GLn(k)) such that aρa −1 :<br />
GK → GLn(S).<br />
The proof of this is elementary but somewhat involved; one reduces to<br />
the Artinian case and argues by induction on the length of R, reducing to<br />
(in the third case) R = k[ɛ]/(ɛ2 ), S = k, where one argues directly (note<br />
though that even this case is not trivial).<br />
We will now return to the main argument, and will abuse the truth by<br />
pretending that Q is a totally real field of even degree, allowing us to use<br />
the above spaces and notation [of course, in reality we would make a base<br />
change to put ourselves in that situation]. Now choose χ0 to correspond<br />
to the central character of our given π. For each finite set of places Q<br />
not containing l or p, let UQ = � UQ,v, where UQ,v = GL2(OF,v) unless<br />
either v = l �or v �∈<br />
Q. If v = l we take the subset U1(v) of matrices<br />
1 ∗<br />
congruent to (mod v), and if v ∈ Q we take the subset of matrices<br />
0 1<br />
� �<br />
a b<br />
∈ U0(v) (the matrices which are upper-triangular modulo v) such<br />
c d<br />
that a/d ↦→ 1 ∈ ∆v, the maximal p-power quotient of k × v . Define U ′ Q in the<br />
same way, but using U0(v) at v = l. If ζ is a p-th root of unity (possibly equal<br />
to 1), we define SQ,ζ(U, O) to be the subspace of Sk,χ0 (UQ, O) consisting of<br />
(ul)u−1 p φ(g) for all u ∈ U ′ Q , where<br />
functions φ which satisfy φ(gu) = ψ −1<br />
ζ<br />
v
18 TOBY GEE<br />
ψζ : U0(l) → O × � �<br />
a b<br />
is the composite of the character ↦→ a/d (mod l)<br />
c d<br />
with the character corresponding under local class field theory to σ ↦→ ζ.<br />
Now let m∅ denote the maximal ideal of the Hecke algebra T1,U (which<br />
∅<br />
acts on S1,∅(U, O)) corresponding to the Hecke eigenvalues of π mod p, and<br />
write S1,∅ := S(U∅, O)m , T ∅ 1,∅ := TU∅,m . Let ζ be a primitive p-th root<br />
∅<br />
of unity, and define Sζ,∅ and Tζ,∅ in the same way. Note that we have<br />
Sζ,∅/λ = S1,∅/λ. In the same way as above, we find that there are Galois<br />
representations ρmod 1,∅ : GQ → GL2(T1,∅) and ρmod ζ,∅ : GQ → GL2(Tζ,∅) lifting<br />
ρ.<br />
Furthermore, it follows from local-global compatibility that there are natural<br />
maps Runiv 1 → T1,∅ and Runiv ζ → Tζ,∅. These maps are actually surjective<br />
(consider the images of the characteristic polynomials of the Frobenius<br />
elements). So to complete the proof it suffices to show that<br />
(as we would then have that ker(R univ<br />
1<br />
SuppRuniv(S1,∅) = Spec R<br />
1<br />
univ<br />
1<br />
→ T1,∅) is nilpotent (see the material<br />
on nearly faithful modules below), so any O-point of Runiv 1 is a O-point of<br />
T1,∅ and is thus modular).<br />
It turns out to be easier to prove the corresponding statement for Runiv ζ ;<br />
this is the “minimal” case. Before 2006, the statement for Runiv 1 was deduced<br />
from this one by Ihara’s lemma, which is unknown in dimension greater than<br />
two, and was a significant obstacle to extending the theory. The argument<br />
that we give here is Richard Taylor’s version of the Taylor-Wiles-Kisin argument,<br />
which avoids the use of Ihara’s lemma.<br />
Now, for each Q = QN as above, we consider a Hecke algebra TUQ acting<br />
on S(UQ, O). It is generated by the operators Tv, Sv for all places v /∈ Q,<br />
v ∤ l, p, and by the Hecke operators<br />
� � � �<br />
πv 0<br />
Uv = UQ,v UQ,v<br />
0 1<br />
for v ∈ Q, where πv is a uniformiser for OFv (as the notation suggests, Uv is<br />
independent of the choice of uniformiser). There is a maximal ideal mQ of<br />
TUQ , which is induced by m∅ and the requirement that it also contains Uv−αv<br />
for each v ∈ Q. We write S1,Q = SQ,1(U, O)mQ , T1,Q = (TUQ )mQ , and define<br />
Sζ,Q, Tζ,Q similarly. Again, we have a natural surjection Runiv 1,Q → T1,Q, and<br />
thus we can view S1,Q as an Runiv 1,Q -module.<br />
Fact 3.10. S1,Q is finite and free over O[∆Q], and the quotient S1,Q/aQ ∼ =<br />
S 1,∅.<br />
[This is not so hard to prove; one uses local-global compatibility to express<br />
the action of ∆Q in terms of explicit Hecke operators, and reduces to a<br />
concrete statement. It is worth noting however that this statement is far
<strong>MODULARITY</strong> <strong>LIFTING</strong> <strong>THEOREMS</strong> 19<br />
from formal in general, and e.g. would fail if we were working with modular<br />
forms over an imaginary quadratic field.]<br />
Define S � 1,Q := SQ ⊗ R univ<br />
1,Q R� 1,Q , an R� 1,Q -module. Similarly, define T� 1,Q =<br />
T1,Q ⊗ R univ<br />
1,Q R� 1,Q = T1,Q[[A1, . . . , A7]]. For each N, choose O[[T1, . . . , Tu]] ↠<br />
O[∆QN ] (which we can do because #QN = u). This gives us a map<br />
O[[T1, . . . , Tu, A1, . . . , A7]] → R� 1,QN . Since R� is topologically generated<br />
1,QN<br />
over R loc<br />
1<br />
by u elements, we can also choose a surjection Rloc<br />
1 [[X1, . . . , Xu]] →<br />
R � 1,QN .<br />
We now wish to “take the limit as N → ∞”. It’s not immediately<br />
clear that we can do this, because the sets of primes QN are unrelated<br />
for different N. However, this isn’t a problem; the trick is to consider each<br />
S � 1,QM /(pN , A N 1 , . . . , AN 7 ) as a module for each (O/pN )[(Z/p N ) u ][A1, . . . , A7]/(A N 1 , . . . , AN 7 ),<br />
M ≥ N. For fixed N there are only finitely many isomorphism classes of<br />
such modules, and this allows us to choose a sequence of pairs (M, N) for<br />
which the resulting modules can be patched together.<br />
In fact, we also need to patch the actions of R loc<br />
1 [[X1, . . . , Xu]], and we<br />
also want to do the same thing simultaneously for the ζ-objects. Having<br />
carried this out, we find ourselves in the following situation.<br />
◦ S � 1,∞ and S� ζ,∞ are finite free O[[T1, . . . , Tu, A1, . . . , A7]]-modules,<br />
and are also Rloc 1 [[X1, . . . , Xu]] (respectively Rloc ζ [[X1, . . . , Xu]])-modules.<br />
In fact, the action of O[[T1, . . . , Tu, A1, . . . , A7]] on S� 1,∞ (respectively<br />
on S� ζ,∞ ) may be factored through Rloc 1 [[X1, . . . , Xu]] (respectively<br />
Rloc ζ [[X1, . . . , Xu]]). [For this last point, note that it suffices to<br />
observe that the image of O[[T1, . . . , Tu, A1, . . . , A7]] in EndO(S� ?,∞ )<br />
is contained in the image of Rloc ? [[X1, . . . , Xu]], which follows from<br />
the corresponding result at each finite level.]<br />
◦ S� 1,∞ /λ and S� ζ,∞λ are isomorphic as modules for k[[T1, . . . , Tu, A1, . . . , A7]]<br />
and for (Rloc 1 /λ)[[X1, . . . , Xu]] = (Rloc ζ /λ)[[X1, . . . , Xu]].<br />
◦ S� 1,∞ /(T1, . . . , Tu, A1, . . . , A7) ∼ = S1,∅ as an Runiv 1 -module, and<br />
S� ζ,∞ /(T1, . . . , Tu, A1, . . . , A7) ∼ = Sζ,∅ as an Runiv ζ -module.<br />
We now employ some commutative algebra.<br />
Definition 3.11. Let A be a Noetherian ring, and let M be a finitely generated<br />
A-module. Say that M is nearly faithful over A if any of the following<br />
equivalent conditions hold:<br />
(i) Supp M = Spec A.<br />
(ii) Supp M contains all the minimal primes of Spec A.<br />
(iii) AnnA M is nilpotent.<br />
This terminology was introduced by Richard Taylor. We will need 3<br />
general results about near faithfulness.<br />
Lemma 3.12. (i) If M is a nearly faithful A-module and I is an ideal<br />
of A, then M/IM is a nearly faithful A/I-module.
20 TOBY GEE<br />
(ii) Suppose that A is a catenary local O-algebra, that each irreducible<br />
component of Spec A has the same dimension, and that each generic<br />
point of Spec A has characteristic 0. Suppose that every prime of A<br />
minimal over λA contains a unique minimal prime of A. Let M be<br />
a finitely generated O-torsion free A-module. If M/λM is a nearly<br />
faithful A/λA-module, then M is a nearly faithful A-module.<br />
(iii) Suppose that A is a Noetherian local ring, that Spec A is irreducible,<br />
and that M is a finitely generated A-module. Suppose that the mAdepth<br />
of M is greater than or equal to the Krull dimension of A.<br />
Then M is a nearly faithful A-module.<br />
Proof. (i) We need to show that (M/IM)P �= 0 for every prime I ⊂ P .<br />
If not, then we have MP /IP MP = 0, and by Nakayama’s lemma we<br />
conclude that MP = 0, a contradiction.<br />
(ii) This requires slightly more work, but is still elementary. See Lemma<br />
2.2(2) of [Tay08].<br />
(iii) This uses some homological algebra - see Lemma 2.3 of [Tay08].<br />
�<br />
We now apply these results as follows. Consider firstly S � ζ,∞<br />
as a mod-<br />
ule for R loc<br />
ζ [[X1, . . . , Xu]]. We employ the third criterion above with A =<br />
Rloc ζ [[X1, . . . , Xu]]. Since Spec Rloc ζ,l<br />
is irreducible, we see that Spec Rloc<br />
ζ [[X1, . . . , Xu]]<br />
is irreducible, so that it is enough to check that the mA-depth of S� ζ,∞ is at<br />
least dim A. But A has the same dimension as O[[T1, . . . , Tu, A1, . . . , A7]],<br />
considered as an<br />
so it is enough to prove the corresponding assertion for S� ζ,∞<br />
O[[T1, . . . , Tu, A1, . . . , A7]]-module. But S� ζ,∞ is a finite free O[[T1, . . . , Tu, A1, . . . , A7]]module,<br />
so this is immediate, and we conclude that S� ζ,∞ is a nearly faithful<br />
Rloc ζ [[X1, . . . , Xu]]-module.<br />
From the third criterion, we see that S� ζ,∞ /λ is a nearly faithful (Rloc<br />
ζ λ)[[X1, . . . , Xu]]module,<br />
so that S� 1,∞ /λ is a nearly faithful (Rloc 1 λ)[[X1, . . . , Xu]]-module.<br />
We now use the second criterion to conclude that S� 1,∞ is a nearly faithful<br />
Rloc 1 [[X1, . . . , Xu]]-module (note that any complete local noetherian ring is<br />
catenary (even excellent)). Finally, using the first criterion again we see that<br />
S� 1,∞ /(T1, . . . , A1, . . . , A7) ∼ = S1,∅ is a nearly faithful Runiv 1 -module.<br />
As above, this means that ker(Runiv 1 → T1,∅) is nilpotent, so any O-point<br />
of Runiv 1 is a O-point of T1,∅ and is thus modular. In particular ρ is modular,<br />
as required.<br />
References<br />
[Tay08] Richard Taylor, Automorphy for some l-adic lifts of automorphic mod l Galois<br />
representations. II, Pub. Math. IHES 108 (2008), 183–239.<br />
Department of Mathematics, Northwestern University, 2033 Sheridan Road,<br />
Evanston, IL 60208-2730, USA<br />
E-mail address: gee@math.northwestern.edu