On the Mumford-Tate conjecture for 1-motives (Q2436639)

A 1-motive over a field \(k\) is given by a pair \((G,Y)\) with \(G\) an extension of an abelian variety by a torus over \(k\) and \(Y\) is étale locally constant, locally isomorphic to a finitely generated free \(\mathbb{Z}\)-module. There is an analogue for 1-motives of the classical Mumford-Tate conjecture for abelian varieties. The main result of the paper is the following theorem: Let \(M= (G,Y)\) be a 1-motive over a finitely generated subfield \(k\) of the field of complex numbers. Then the Mumford-Tate conjecture holds for \(M\) if and only if it holds for the quotient abelian variety of \(G\). As an immediate consequence one gets the following interesting corollary: Let \(X\) be a variety over \(k\). The Mumford-Tate conjecture holds for cohomology in degree 1 of \(X\) if and only if the classical Mumford-Tate conjecture holds for the Albanese variety of a smooth projective variety birational to \(X\). Although there is no Mumford-Tate group in positive characteristics, there is a result for these fields concerning the unipotent motivic fundamental group \(P(M)\) of a 1-motive \(M\). The appendix contains an explanation, due to Deligne, of the construction of the Lie algebra of \(P(M)\) in terms of Tannakian formalism.