Jordan correspondence and block distribution of characters
Radha Kessar, Gunter Malle
TL;DR
The paper resolves the $\,\ell$-block distribution problem for quasi-simple exceptional groups, including bad primes, by developing a Jordan-block framework that links unipotent blocks in centralisers to global blocks via the map $\bar{J}_t^{\mathbf{G}}$ and by handling $t$-twins in $E_8$. It extends KM’s $e$-Harish-Chandra theory to connected-centre settings, establishes parametrisations of blocks through $e$-cuspidal data, and proves a compatibility between Lusztig induction and Jordan decomposition across Levi subgroups, aided by BR-correspondence and inductive arguments. The work also corrects and completes earlier results (En00, KM13, KM15), provides a descent to adjoint types $E_6$ and $E_7$, and gives a comprehensive treatment of isolated blocks and specific $5$-blocks in $E_8$, culminating in a proof of Robinson's conjecture for all blocks. Overall, the methods yield a complete, symmetry-respecting description of the modular representation theory for these groups and establish block-level compatibilities essential for broader conjectures in finite group representation theory.
Abstract
We complete the determination of the $\ell$-block distribution of characters for quasi-simple exceptional groups of Lie type up to some minor ambiguities relating to non-uniqueness of Jordan decomposition. For this, we first determine the $\ell$-block distribution for finite reductive groups whose ambient algebraic group defined in characteristic different from $\ell$ has connected centre. As a consequence we derive a compatibility between $\ell$-blocks, $e$-Harish-Chandra series and Jordan decomposition. Further we apply our results to complete the proof of Robinson's conjecture on defects of characters.
