The rank of Mazur's Eisenstein ideal (Q2178445): Difference between revisions

Let \(N\) and \(p>3\) be prime numbers. Let \(\mathbb{T}\) denote the completion of the Hecke algebra with weight \(2\) and level \(\Gamma_0(N)\) at the Eisenstein maximal ideal with residual characteristic \(p\), and let \(\mathbb{T}_0\) denote the cuspidal quotient of \(\mathbb{T}\). \textit{B. Mazur} [Publ. Math., Inst. Hautes Étud. Sci. 47, 33--186 (1977; Zbl 0394.14008)] posed the following question: Is there anything general that can be said about the Newton polygon of \(\mathbb{T}^0\), or even about \(\text{rank}_{\mathbb{Z}_p} \mathbb{T}^0\)? In the present paper, the authors use pseudo-deformation theory to study Mazur's Eisenstein ideal. They study the Eisenstein part of the \(p\)-adic Hecke algebra for \(\Gamma_0(N)\), compute the rank of this Hecke algebra and its Newton polygon in terms of Massey products in Galois cohomology. Therefore, they give a complete answer to Mazur's question. Note that Mazur gave a different criterion for \(\text{rank}_{\mathbb{Z}_p} \mathbb{T}^0=1\) in terms of the geometry of modular curves and \textit{L. Merel} [J. Reine Angew. Math. 477, 71--115 (1996; Zbl 0859.11036)] gave a number-theoretic interpretation of this criterion. There is an alternate formulation by Calegari and Venkatesh of this theorem in terms of zeta values. The authors restate Merel's theorem and give a new proof in terms of the cup products.