Take a manifold with a nilpotent fundamental group — the simplest class of nonabelian groups where commutators eventually vanish. Construct random coverings by sampling homomorphisms from this group into symmetric groups. The coverings break into connected components, and the number of components is a random variable. Abdesselam (arXiv: 2603.24499) proves a central limit theorem for this count: as the covering degree grows, the distribution of connected components converges to a Gaussian.
The proof connects probability theory to the zeta functions of subgroup growth. For nilpotent groups, du Sautoy and Grunewald showed that the number of subgroups of each index follows a pattern governed by zeta functions with Euler products. Abdesselam uses Delange's Tauberian theorem — a tool that extracts asymptotic behavior from analytic properties of generating functions — to convert the subgroup growth asymptotics into a CLT for random covering components.
The result extends prior work from abelian to nilpotent fundamental groups. The abelian case was understood: cyclic groups have well-behaved subgroup lattices, and the CLT follows from classical number-theoretic methods. The nilpotent extension requires the deeper machinery because the subgroup structure is more complex — nonabelian commutator relations create dependencies between subgroups that don't exist in the abelian case.
The through-claim: randomness on algebraic topology is governed by number theory. The number of components in a random covering is a topological quantity, defined by the manifold's fundamental group. But its statistical behavior is controlled by the analytic properties of subgroup growth zeta functions — objects from number theory. The bridge between topology and probability runs through arithmetic, and the CLT is the toll.
Abdesselam, 2603.24499. Group theory / random coverings / central limit theorem / nilpotent groups / subgroup growth.