Counterexamples for Türkelli's Modification on Malle's Conjecture
Jiuya Wang
TL;DR
The paper presents counterexamples to Türkelli's modification of Malle's conjecture, showing that the proposed $b$-constant $b_T(G,k)$ can differ from the true $b(G,k)$ and from Malle's original prediction, with explicit instances where function-field and number-field constants diverge. It analyzes the embedding problem of cyclotomic extensions, introduces a refined, invariant-sensitive framework that partitions G-extensions by cyclotomic data, and defines a lifting-based counting function $N_{Q,\pi,\phi}(G^{inv},X)$ to capture these subtleties. The authors compute $b$-constants for specific wreath-product groups over both number fields and function fields, demonstrating infinite families where $b_T$ and $b_M$ bound the true constant in different ways and where $b$-constants diverge across base fields. They propose a refined Malle conjecture that aggregates over finitely many embedding data pairs $(\pi,\phi)$ and accommodates function-field behavior, supported by function-field results and recent advances on embedding problems. Overall, the work indicates that a universal, single-constant formulation of Malle’s conjecture is unlikely and motivates a composition-by-embedding approach to capture cyclotomic interactions and constants more accurately, potentially guiding future conjectures and proofs in both arithmetic and function-field contexts.
Abstract
We give counterexamples for the modification on Malle's Conjecture given by Türkelli. Türkelli's modification on Malle's conjecture is inspired by an analogue of Malle's conjecture over a function field. As a consequence, our counterexamples demonstrate that the $b$ constant can differ between function fields and number fields. We also show that Klüners' counterexamples give counterexamples for a natural extension of Malle's conjecture to counting number fields by product of ramified primes. We then propose a refined version of Malle's conjecture which implies a new conjectural value for the constant $b$ for number fields.
