The author was partially supported by NSF grants DMS-0100537 and DMS-0354772.
<ph f="cmbx">A Mordell-Weil theorem for abelian varieties over fields generated by torsion points</ph>

Michael Larsen

E-mail address : larsen@math.indiana.edu Department of Mathematics, Indiana University, Bloomington, IN 47405, U.S.A.

1 Introduction

Let A   be an abelian variety over a number field K   and F   an algebraic extension of K   . The quotient A ( F ) / A ( F ) t o r   is always torsion-free and abelian. When F   is a number field, the Mordell-Weil theorem asserts that it is finitely generated and therefore free. At the opposite extreme, when F   is algebraically closed, the quotient is divisible and is therefore a Q   -vector space. In this paper, we show that when F   is any field generated over K   by the coordinates of torsion points on A   , the quotient is again a free abelian group, though it may not be finitely generated. The idea is that dividing points of infinite order in A ( K )   gives rise to fields which are in some weak sense linearly disjoint from the fields obtained by dividing torsion points.
This is essentially a question of Kummer theory of the abelian variety and is closely related to ideas of Bašmikov [1, Ribet [8, and Jacquinot-Ribet [5.
Let T   denote the   -adic Tate module of A   over a number field K   . Let G   denote the image of the Galois group in A u t ( T )   . The crucial point in this paper is that for all i 0   , H i ( G , T )   is finite, and when 0   , it vanishes for all i   . In particular, this is true for i = 1   and i = 2   .
Throughout the paper, K   is a number field, K ¯   an algebraic closure, and G K = G a l ( K ¯ / K )   . We write ρ   for any continuous   -adic representation of G K   . (In the later part of the paper, ρ   comes from the action of G K   on the   -adic Tate module T   of an abelian variety A / K   .) The image of ρ   is always denoted G   . It is a closed subgroup of G L n ( Q )   for some n   and is therefore an   -adic Lie group. The fixed field of the kernel of ρ   is denoted K   . Group cohomology is always understood to mean continuous cohomology.

2 Galois cohomology

Let G   be a compact   -adic Lie group and L   the Lie algebra of G   ([7 V 2.4.2).
Let V   be a finite-dimensional vector space over Q   on which G   operates continuously. There is a corresponding “infinitesimal” action of L   on V   ([7 V 2.4.6). By a theorem of D. Lazard ([7 V 2.4.10 (ii)), if H   is a small enough open subgroup of G   ,
H p ( H , V ) ~ H p ( L , V ) , (2.1)
where the right hand side denotes Lie algebra cohomology.
Proposition 2.1. Let G   be a compact topological group and V   a finite dimensional vector space over Q   on which G   admits a faithful and semisimple action. Suppose that no subrepresentation of V   factors through a finite quotient of G   . Then for all i 0   , H i ( G , V ) = 0 .  
  • Proof. Let H   be an open normal subgroup of G   . The restriction of a semisimple representation to such a subgroup is again semisimple, and by hypothesis, V H = 0   . Suppose that for some integer i   and some normal open subgroup H G   , we have H q ( H , V ) = 0   for all q i   . The Lyndon-Hochschild-Serre spectral sequence for modules finite dimensional over Q   (which follows from the usual spectral sequence with finite coefficients together with [11 Corollary 2.2 and [11 Proposition 2.3) asserts E 2 p q = H p ( G / H , H q ( H , V ) ) H p + q ( G , V ) .   Thus, H i ( G , V ) = 0   for this value of i   . To prove H i ( G , V ) = 0   for all i 0   , we may therefore assume G = H   is such that ( 2.1 ) holds for all p   . Every G   -subrepresentation is also an L   -subrepresentation, and every L   -subrepresentation comes from a H   -subrepresentation for some open subgroup H   of G   ([7 V 2.4.6). Without loss of generality, therefore, we may assume that the G   -subrepresentations and L   -subrepresentations of V   are in one-to-one correspondence and therefore V   is a semisimple representation of L   . Applying ( 2.1 ) for p = 0   , we see that V H = 0   implies V L = 0   . The proposition now follows from the vanishing of the cohomology of a reductive Lie algebra L   with coefficients in a semisimple module V   such that V L = 0   [4 Th. 10.
Proposition 2.2. Let K   be any number field,   a rational prime, V   a finite dimensional Q   -vector space, and ρ : G K A u t ( V )   a continuous semisimple representation which is pure of weight 0   . Let G = i m ρ   . Then for all i 0   ,
H i ( G , V ) = 0 (2.2)
  • Proof. We claim that V H 0   for all open subgroups H   of G   . Indeed, let m a t h f r a k p   be any prime ideal of K   which is unramified in ρ   and whose Frobenius lies in H   . As ρ   is pure of non-zero weight, ρ ( F r o b m a t h f r a k p )   cannot have 1 as eigenvalue. Therefore, V H = 0   . Applying Proposition  2.1 with G = G   , we obtain ( 2.2 ).
Proposition 2.3. Let T   be a free Z   -module on which G K   acts continuously in such a way that V = T Q   satisfies the hypotheses of Theorem  2.2 . Then H i ( G , T )   is a finite   -group for all i 0   .
  • Proof. By Lyndon-Hochschild-Serre, it suffices to check that G   has an open subgroup H   such that H q ( H , T )   is finite for all q i   . By [7 III 3.1.3, G   contains an open subgroup H   which is equi-   -valued, and by [7 V 2.5.8, such a group is a Poincaré group in the sense of [9 I 4.5. In particular, H q ( H , T / k T )   is finite for all q , k 0   . From the cohomology sequence for 0 T T T F 0   it follows that that H q ( H , T ) F   is finite for all q 0   . Since H q ( H , T ) = l i m k H q ( H , T / k T )   is an abelian pro-   group whose tensor product with F   is finite, it must be a finitely generated Z   -module. Finally, H q ( H , T ) Z Z [ 1 / ] = H q ( H , T Z [ 1 / ] ) = H q ( H , V ) = 0 .   We conclude that H q ( H , T )   is finite.
If X   is a non-singular projective variety over a number field K   , i   is a positive integer, and   is a rational prime, we expect V = H i ( X , Q )   to satisfy the hypotheses of Theorem  2.2 . This is known to be true in a few important cases, notably when X   is an abelian variety and i = 1   ([3 Satz 3).
It is this case, or more precisely, its dual, which interests us for the rest of the paper.
Theorem 2.4. Let K   be a number field and A   an abelian variety over K   . For each rational prime   different from the characteristic of K   , we define T   to be the   -adic Tate module of A   and G   the image of G K   in A u t ( T )   . Then for all 0   , H i ( G , T ) = 0   for all i   .
  • Proof. By a theorem of Bogomolov [2, the group C   of homotheties in G   is open in Z *   . By a refinement of Bogomolov's result due to Serre [10 §2, the index of C   in Z *   is bounded independently of   ; in particular, for 0   , it contains some element z   not congruent to 1   (mod   ). Thus lim k z k   is a central torsion point generating a non-trivial finite subgroup Z   of prime-to-   order. We apply the Lyndon-Hochschild-Serre spectral sequence for Z G   . Now H q ( Z , T )   is a | Z |   -torsion group but | Z |   is invertible in Z   , so H q ( Z , T ) = 0   for q 0   . Moreover, H 0 ( Z , T ) = 0   , since Z   has non-trivial scalar action on T   . The theorem follows.
Corollary 2.5. Let K   be a number field and A   an abelian variety over K   . For every i 0   , there exists a uniform bound N   such that | H i ( G , T / n T ) | < N   for all prime powers n   .
  • Proof. The cohomology sequence of the short exact sequence 0 T n T T / n T 0   gives | H i ( G , T / n T ) | | H i ( G , T ) | | H i + 1 ( G , T ) | ,   whence the bound follows from Theorem  2.4 .

3 Rational points over the field of division points

In this section, we apply the computations of the first section to show that the group of points on an abelian variety over the field K t o r   generated by all of its torsion is free modulo torsion. Given Corollary  2.5 , it is not difficult to control   -divisibility of points on the abelian variety in the towers of number fields arising from   -power torsion. What remains is to control   -divisibility in the towers arising from prime-to-   torsion.
Theorem 3.1. Let A   be an abelian variety over a number field K   . Let K t o r : = K ,   the field generated over K   by the coordinates of all torsion points of A   . Then A ( K t o r ) = M ( Q / Z ) 2 dim A   , where M   is a free abelian group.
For any abelian group M   , we write M t f   for the (torsion-free) quotient M / M t o r   . If M N   , then M t o r = N t o r M   , so the natural homomorphism M t f N t f   is injective. We use the following criterion for M t f   to be free:
Lemma 3.2. Let M   be an abelian group. Then M t f   is a free abelian group if and only if for every finite-dimensional subspace V M Q = M t f Q   , the group { m M t f | m 1 V }   is finitely generated.
  • Proof. If M t f   is free, we choose a Z   -basis S   and write each element of a Q   -basis of V   as a Q   -linear combination of elements of S   . Thus only finitely many elements of S   are involved in V   , so replacing M   by the preimage of S   , we may assume M t f   is finitely generated. In this case, every subgroup of M t f   is again finitely generated.
    In the other direction, we define a partial basis of M t f   to be any subset S M t f   which maps to a linearly independent set in M Q   and such that for m M t f   , m 1 S p a n ( S 1 )   implies m S   . By Zorn's lemma, there exists a maximal partial basis for M t f   , which is then necessarily a basis.
Definition 3.3. If Λ   is a subgroup of an abelian group M   and   is a rational prime, we say that the   -saturation of Λ   is the subgroup of M   consisting of elements m   such that k m Λ   for some non-negative integer k   . We say Λ   is   -saturated (in M   ) if it equals its own   -saturation.
Lemma 3.4. Let K   denote the fixed field of G   . Then A ( K ) t f   is of finite index in its   -saturation in A ( K ) t f   . If 0   , A ( K ) t f   is   -saturated.
  • Proof. Let P A ( K )   represent a class belonging to the   -saturation of A ( K ) t f   , and let k   be the smallest positive integer such that k [ P ] A ( K ) t f   , i.e., such that k P A ( K ) + A ( K ) t o r   . As A ( K ) t o r   is   -divisible, without loss of generality we may assume that Q : = k P A ( K )   . Consider the cohomology sequence of the short exact sequence of G   -modules 0 A [ k ] A ( K ) k k A ( K ) 0 .   The image of Q H 0 ( G , k A ( K ) )   in H 1 ( G , A [ k ] ) = H 1 ( G , T / k T )   has order k   ; otherwise k 1 Q = k R   for some R A ( K )   , which means k 1 P R A ( K ) t o r   and therefore k 1 [ P ] = 0   in A ( K ) t f   , contrary to the definition of k   . The lemma now follows from Corollary  2.5 .
We have already remarked that the torsion in A ( K )   is   -divisible. It follows that if Λ A ( K ) t f   has inverse image Λ ~   in A ( K )   , then the index of Λ   in its   -saturation in A ( K t o r ) t f   is the same as the index of Λ ~   in its   -saturation in A ( K t o r )   . In particular, Λ   is   -saturated if and only if Λ ~   is so.
Note also that by Kummer theory, if L K   , then every Galois extension of L   obtained by adjoining coordinates of points which are obtained from points of A ( K )   by   -division is a pro-   extension.
We can now prove the main theorem.
  • Proof of Theorem  3.1 . By definition of K t o r   , we have A ( K t o r ) t o r = A ( K ¯ ) t o r = ( Q / Z ) 2 dim A .   We will prove that A ( K t o r ) t f   is a free Z   -module, and this will imply that the short exact sequence 0 A ( K t o r ) t o r A ( K t o r ) A ( K t o r ) t f 0   splits, which then implies the theorem.
    Without loss of generality, we may replace K   by a larger number field K   , since A ( K t o r ) t f A ( K t o r ) t f   is injective, and a subgroup of a free abelian group is again free abelian. In order that the extensions K   of K   have the good linear disjointness property we will need below, we assume that K   is large enough for the purposes of [10 §1. For any fixed finite dimensional V A ( K t o r ) Q   , we choose K   large enough that V A ( K ) Q   . Let Λ ~ = { P A ( K ) | P 1 V }   . By Lemma  3.2 , it suffices to prove that Λ ~   is of finite index in its   -saturation in A ( K t o r )   for all   and   -saturated for all 0   .
    We now fix a prime   . By Lemma  3.4 , there exists a finite extension K / K   such that the   -saturation of Λ ~   in A ( K )   is contained in A ( K )   ; moreover, K = K   if 0   . Replacing K   by K   and Λ ~   by its   -saturation in A ( K )   before examining   -saturations in K t o r   , we may therefore assume without loss of generality that Λ ~   is already   -saturated in A ( K )   .
    For every finite set T   of primes, let K T   denote the compositum of K   for T   . By [10 Lemma 1.2.1, there exists a finite set T   of primes such that G a l ( K T / K ) = T G .   By Serre's uniform version of Bogomolov's theorem [10, if 0   , G   contains a homothety in Z *   not congruent to 1   (mod   ). Thus G a l ( K T / K )   has an element τ   mapping to such a homothety but acting trivially on   -torsion for all / { } T   . We also assume   is larger than the order of T G L 2 dim A ( F )   ; in particular, T   . We claim first that Λ ~   is   -saturated in K T   . If it were not, we could choose σ G a l ( K T / K ) = T { } G   and P A ( K T )   such that P A ( K )   but σ ( P ) P   . As σ   and τ   act (respectively) trivially and without fixed points on   -torsion, we have
    τ σ 1 τ 1 ( P ) P = τ ( σ 1 τ 1 ( P ) ) P = τ ( σ 1 ( τ 1 ( P ) P ) + σ 1 ( P ) ) P = τ ( τ 1 ( P ) P + σ 1 ( P ) ) P = τ ( σ 1 ( P ) P ) σ 1 ( P ) P , (3.1)
    so σ τ σ 1 τ 1   acts non-trivially on P   . This is impossible since σ   and τ   commute, being supported on complementary index sets in the product T G   . Now, K t o r   is obtained from K T   by adjoining coordinates of points of order T   in A   and then by adjoining higher   -power points for T   . A pro-   extension of a field never contains a non-trivial   -extension of that field for   , and a Galois extension of degree less than   cannot contain a non-trivial   -extension, so the saturation of Λ ~   does not change in passing from K T K   to K t o r   . Thus Λ ~   is   -saturated in A ( K t o r )   .
    Finally, we need to show that Λ ~   is of finite index in its   -saturation in A ( K t o r )   for every   . It is enough to show that it is of finite index in its   -saturation in A ( K T K )   since the passage from K T K   to K t o r   involves a finite extension followed by a prime-to-   profinite extension. The proof proceeds exactly as before except that Bogomolov's theorem [2guarantees only the existence of τ ¯ 0   in G   which is a non-trivial scalar. By the first theorem in [10, G a l ( K T K / K )   contains an open subgroup of G × T { } G   , so setting τ ¯ = τ ¯ 0 k   , we can lift to τ G a l ( K T K / K )   whose image in G   a homothety not congruent to 1   (mod m   ) and whose image in G a l ( K T / K )   is trivial. If σ G a l ( K T / K )   acts non-trivially on P A ( K T )   such that m 1 P A ( K )   and m P A ( K )   , we can again apply ( 3.1 ) to conclude that σ   and τ   do not commute, which is impossible. Thus, Λ ~   is of index ( m 1 ) dim V   in its   -saturation in A ( K T K )   . The theorem follows.
Corollary 3.5. Let A   and B   be abelian varieties over a number field K   . Let K A t o r   denote the extension of K   generated by the torsion points of A   . If K L K A t o r   , then B ( L ) t f   is free abelian.
  • Proof. Applying Theorem  3.1 to A × B   , we see that ( A × B ) ( K A × B t o r ) t f   is free abelian. Therefore, B ( L ) t f B ( K A t o r ) t f ( A × B ) ( K A × B t o r ) t f   is also free abelian.
We remark that if A   is an elliptic curve and K = Q   , then A ( K t o r ) t f   can be determined up to isomorphism: it is Z ω   . In general, the rank of A ( K t o r )   must be countable; the difficulty lies in proving that it is infinite.
For elliptic curves over Q   written in Weierstrass form, it is clear that points with rational x   -coordinates are defined over Q a b Q t o r   , and by the method of [6, it is easy to see that they generate an infinite-dimensional subspace of A ( Q ¯ ) Q   . This raises the following question: is the rank of A ( K t o r )   infinite for all abelian varieties over number fields?
References

  1. Bašmakov, Marc: Un théorème de finitude sur la cohomologie des courbes elliptiques. C. R. Acad. Sci. Paris Sér. A-B 270 (1970), A999–A1001.
  2. Bogomolov, Fedor Alekseivich: Sur l'algébricité des représentations l   -adiques. C. R. Acad. Sci. Paris Sér. A-B 290 (1980), no. 15, A701–A703.
  3. Faltings, Gerd: Endlichkeitssẗze für abelsche Varietäten über Zahlkörpern. Invent. Math. 73 (1983), no. 3, 349–366.
  4. Hochschild, Gerhard; Serre, Jean-Pierre: Cohomology of Lie algebras. Ann. of Math. 57 (1953), 591–603.
  5. Jacquinot, Olivier; Ribet, Kenneth A.: Deficient points on extensions of abelian varieties by G m   . J. Number Theory 25 (1987), no. 2, 133–151.
  6. Larsen, Michael: Rank of elliptic curves over almost separably closed fields. Bull. London Math. Soc. 35 (2003), 817–820.
  7. Lazard, Michel: Groupes analytiques p   -adiques. Inst. Hautes Études Sci. Publ. Math. 26 (1965), 389–603.
  8. Ribet, Kenneth A.: Kummer theory on extensions of abelian varieties by tori. Duke Math. J. 46 (1979), no. 4, 745–761.
  9. Serre, Jean-Pierre: Cohomologie Galoisienne. Lecture Notes in Mathematics, Vol. 5. Springer-Verlag, Berlin, 1965.
  10. Serre, Jean-Pierre: Lettre à Ken Ribet du 7/3/1986, Œvres Volume IV 1985–1998, pp. 56–65. Springer-Verlag, Berlin, 2000.
  11. Tate, John: Relations between K 2   and Galois cohomology, Invent. Math. 36 (1976), 257–274.

E-mail address : larsen@math.indiana.edu Department of Mathematics, Indiana University, Bloomington, IN 47405, U.S.A.