Random matrices, the Cohen-Lenstra heuristics, and roots of unity (Q2018351)

The Cohen-Lenstra heuristic [\textit{H. Cohen} and \textit{H. W. Lenstra jun.}, Lect. Notes Math. 1068, 33--62 (1984; Zbl 0558.12002)] predicts the frequency with which a fixed finite abelian \(\ell\)-group appears as the \(\ell\)-part of the ideal class group of an imaginary quadratic number field. \textit{G. Malle} [J. Number Theory 128, No. 10, 2823--2835 (2008; Zbl 1225.11143)] provided some numerical evidence that a generalization of Cohen-Lenstra heuristic formulated in [\textit{H. Cohen} and \textit{J. Martinet}, J. Reine Angew. Math. 404, 39--76 (1990; Zbl 0699.12016)] for relative class groups of finite extensions of arbitrary number fields does not hold when an \(\ell\)-th root of unity is contained in the base field. Malle also proposed some new heuristics which are in very good accordance to his numerical calculations. The paper under review addresses the function field case. The heuristic propesed in [\textit{E. Friedman} and \textit{L. C. Washington}, in: Théorie des nombres, C. R. Conf. Int., Québec/Can. 1987, 227--239 (1989; Zbl 0693.12013)] does not take into account roots of unity, so in analogy to the number field case, it is not expected to hold in general. A revised version of this heuristic is formulated in [\textit{J. D. Achter}, Contemp. Math. 463, 1--7 (2008; Zbl 1166.11018)]. It relates the distribution of class groups of quadratic extensions of \(\mathbb F_{p^n}(t)\) with the sizes of certain sets of matrices. The main results of the paper under review concern these sizes, making the results of Achter explicit when the base field has \(\ell\)-th roots of unity but not \(\ell^2\)-th roots of unity and in the case when the base field has \(\ell^2\)-th roots of unity but not \(\ell^3\)-th roots of unity. This might be seen as a function field version of (a generalization of) Malles's conjecture.