On Atkin--Swinnerton-Dyer congruence relations (Q2566168)

In [Invent. Math. 79, 49--77 (1985; Zbl 0553.10023)] \textit{A. J. Scholl} constructed Galois representations (motives) attached to the space of cusp forms of weight \(k\geq 3\) of non-congruence subgroups of \(\text{SL}(2, \mathbb Z).\) It remains unclear whether these motives are modular, i.e.\ they coincide with motives for congruence subgroups. The reviewed article provides a new example for this situation. Using the elliptic modular surface defined by \(y^2+(1-t^3)xy-t^3y=x^3-t^3x^2\) the authors study a non-congruence genus zero subgroup of index 3 in \(\Gamma^1(5)\) with cusp-widths 15 at the cusps \(\infty\) and \(-2\). Using heavily the theory of elliptic modular surfaces, and in particular \textit{T. Shioda}'s [J. Math. Soc. Japan 24, 20--59 (1972; Zbl 0226.14013)], they show, that \(\Gamma\backslash {\mathbb H}\) has a model over \({\mathbb Q}\) and that there is a basis of the 2-dimensional space of weight 3 cusp forms for \(\Gamma\) which satisfies certain congruence relations with cusp forms for some (specified) congruence subgroup. This provides a modular interpretation for the motive associated to \(S_3(\Gamma)\) by Scholl [loc. cit.]. It also leads to a kind of congruence relations for the given basis of \(S_3(\Gamma)\) which replace the recursive relations on Hecke-eigenvalues for cusp-forms on congruence subgroups. Those relations had first been observed and conjectured more generally by \textit{A. O. L. Atkin} and \textit{H. P. F. Swinnerton-Dyer} in [Combinatorics, Proc. Symp. Pure Math. 19, 1--25 (1971; Zbl 0235.10015)].