Fixed-point-free elements of iterated monodromy groups

TL;DR

This paper investigates the fixed-point properties of elements in iterated monodromy groups of complex polynomials, showing that fixed points on the boundary are rare for most polynomials, with specific notable exceptions.

Contribution

It demonstrates that fixed points on the boundary are measure-zero for nearly all such groups, and characterizes the exceptional cases, using automaton analysis and martingale convergence.

Findings

Fixed points are measure-zero for most polynomials.

Exceptions include Chebyshev-related polynomials and an unresolved case.

Automaton analysis and martingale theorems underpin the proof.

Abstract

The iterated monodromy group of a post-critically finite complex polynomial of degree d \geq 2 acts naturally on the complete d-ary rooted tree T of preimages of a generic point. This group, as well as its pro-finite completion, act on the boundary of T, which is given by extending the branches to their "ends" at infinity. We show that for nearly all polynomials, elements that have fixed points on the boundary are rare, in that they belong to a set of Haar measure zero. The exceptions are those polynomials linearly conjugate to multiples of Chebyshev polynomials and a case that remains unresolved, where the polynomial has a non-critical fixed point with many critical pre-images. The proof involves a study of the finite automaton giving the action of generators of the iterated monodromy group, and an application of a martingale convergence theorem. Our result is motivated in part by applications to arithmetic dynamics, where iterated monodromy groups furnish the "geometric part" of certain Galois extensions encoding information about densities of dynamically interesting sets of prime ideals.