Deforming abelian elliptic $\mathrm{SL}(2,\mathbb{R})$--representations of knot groups
Yi Liu
TL;DR
This work establishes a criterion linking zeros of the knot's Alexander polynomial on the unit circle to deformations of abelian elliptic representations of knot groups into $SL(2,\mathbb{R})$, producing continuous families of irreducible representations that approach the abelian limit and lift to the universal cover $\widetilde{SL}(2,\mathbb{R})$. The authors develop a unified quaternionic framework and a detailed trace-slice/intersection approach, leveraging equivariant transversality to obtain local modulo-$2$ intersection numbers that guarantee deformation when the zero is of odd order. As a key application, every nontrivial $L$-space knot has an Alexander polynomial with an odd-order unit-circle zero, yielding irreducible $SL(2,\mathbb{R})$ representations for its knot group and, via small-slope surgery, potential implications for left-orderability of surgery groups. The paper also provides a coefficient-based criterion (via Konvalina–Matache) that ensures the existence of such odd-order unit-circle zeros, broadening accessible checks beyond simple zeros. Overall, the results connect knot invariants, representation varieties, and 3-manifold topology, offering a general framework to detect nonabelian real representations from Alexander-polynomial data and to relate these to orderability phenomena after Dehn surgery.
Abstract
The following criterion is proved in this paper. If the Alexander polynomial of a knot $K\subset S^3$ has a zero of odd order on the complex unit circle, then there exists a continuous family of irreducible representations $π_1(S^3\setminus K)\to \mathrm{SL}(2,\mathbb{R})$ converging to an abelian representation of noncentral elliptic type. As an application, the author shows that the Alexander polynomial of any nontrivial L-space knot satisfies the condition of the criterion. In particular, it follows that the fundamental group of any nontrivial L-space knot complement admits an irreducible $\mathrm{SL}(2,\mathbb{R})$--representation.
