Table of Contents
Fetching ...

The Goldman bracket characterizes homeomorphisms between non-compact surfaces

Sumanta Das, Siddhartha Gadgil, Ajay Kumar Nair

TL;DR

The paper provides a sharp criterion for when a homotopy equivalence between non-compact orientable surfaces without boundary is homotopic to a homeomorphism: this holds exactly when the map commutes with the Goldman bracket, excluding the plane and cylinder. The authors develop an exhaustion-based construction that yields a proper map $g$ homotopic to the given $f$, by controlling loop images via fillings and splitting arguments rooted in the Goldman bracket. They then upgrade $g$ to a (proper) homeomorphism using a standard rigidity result for non-compact surfaces, with orientation forcing explained through bracket considerations. The approach ties a Lie-algebraic structure on curves directly to topological rigidity, providing a practical criterion and a robust technique that also yields a related intersection-number characterization.

Abstract

We show that a homotopy equivalence between two non-compact orientable surfaces is homotopic to a homeomorphism if and only if it preserves the Goldman bracket, provided our surfaces are neither the plane nor the punctured plane.

The Goldman bracket characterizes homeomorphisms between non-compact surfaces

TL;DR

The paper provides a sharp criterion for when a homotopy equivalence between non-compact orientable surfaces without boundary is homotopic to a homeomorphism: this holds exactly when the map commutes with the Goldman bracket, excluding the plane and cylinder. The authors develop an exhaustion-based construction that yields a proper map homotopic to the given , by controlling loop images via fillings and splitting arguments rooted in the Goldman bracket. They then upgrade to a (proper) homeomorphism using a standard rigidity result for non-compact surfaces, with orientation forcing explained through bracket considerations. The approach ties a Lie-algebraic structure on curves directly to topological rigidity, providing a practical criterion and a robust technique that also yields a related intersection-number characterization.

Abstract

We show that a homotopy equivalence between two non-compact orientable surfaces is homotopic to a homeomorphism if and only if it preserves the Goldman bracket, provided our surfaces are neither the plane nor the punctured plane.
Paper Structure (18 sections, 10 theorems, 6 equations, 3 figures)

This paper contains 18 sections, 10 theorems, 6 equations, 3 figures.

Key Result

Theorem 1.1

A homotopy equivalence $f\colon \Sigma'\to \Sigma$ between two non-compact oriented surfaces without boundary is homotopic to an orientation-preserving homeomorphism if and only if it commutes with the Goldman bracket, i.e., for all $x',y'\in\mathbb Z[\widehat{\pi}(\Sigma')]$, we have, where $f_*\colon \mathbb Z[\widehat{\pi}(\Sigma')]\to\mathbb Z[\widehat{\pi}(\Sigma)]$ is the function induced b

Figures (3)

  • Figure 1: $\left\{\alpha_1,...,\alpha_7\right\}$ is a filling system of $K_m$ and $\left\{\alpha_1,...,\alpha_9\right\}$ is an extended filling system for $K_m$ because $I_\Sigma\left(C_1,\alpha_9\right)\neq 0\neq I_\Sigma\left(C_2,\alpha_8\right)$.
  • Figure 2: If $\widehat{V}\neq V$, $f_*([\gamma'])\cdot f_*([\delta])=[\gamma*\theta*\eta*\bar{\theta}]$ is not homotopic to a curve disjoint from $\partial\widehat{V}$, giving a contradiction.
  • Figure 3: Description of $g^{(j)}_m\colon V'=V^{(j)}_m\to V\subset\Sigma\setminus K_{m-1}$. On the top, i.e., in $V'$, the purple, blue, and green portions (arcs and circles) form $\Gamma'$, and blue and black arcs form $T'^{(j)}_m$. At the bottom: the grey and yellow shades indicate $g^{(j)}_m(V')$ and $g^{(j)}_m(C_i'\times [0,1])$, respectively, and red and green loops denote $g^{(j)}_m(C_i'\times \{1\})$ and $g^{(j)}_m(\delta')$, respectively, where $i=1,2$.

Theorems & Definitions (15)

  • Theorem 1.1
  • Theorem 1.2
  • Lemma 2.1
  • Lemma 2.2
  • Theorem 2.3: Goldman MR0846929
  • Lemma 3.1
  • proof
  • Remark 3.2
  • Lemma 3.3
  • proof
  • ...and 5 more