On the fundamental 3-classes of knot group representations (Q2290840)

Taking a link exterior and the peripheral system as a model, a topological study of a more general connected \(3\)-manifold \(M\) whose boundary is a union of tori aims at investigating a pairing given by the following data: a group representation \(\pi_1 M \to G\) mapping the image of \(\pi_1 \partial M\) to a subgroup \(K\), a \({\mathbb Z}[G]\)-module \(A\), and a \(3\)-cocycle \(\theta\) from \(H^3(G,K;A)\); then \(\theta\) and a pushforward \(f_*[M,\partial M] \in H_3(M,\partial M; {\mathbb Z})\) are paired to \(\langle \theta, f_*[M,\partial M] \rangle \in A_G = H_0(G;A)\). This pairing occurs in the context of the volume and the Chern-Simons invariant of a hyperbolic manifold as well as of the Dijkgraaf-Witten invariant. So far, the pairing was thought inaccessible to explicit computations. Here, for rather general types of knots, a calculable description is presented that starts out with geometric configurations such as diagrams. Based on key ideas from [Geom. Dedicata 171, 265--292 (2014; Zbl 1300.57012)] by \textit{A. Inoue} and \textit{Y. Kabaya} the author uses quandles as a bridge to knot theory. So he applies the findings of \textit{D. Joyce} [J. Pure Appl. Algebra 23, 37--65 (1982; Zbl 0474.57003)] that the knot quandle classifies knots, which in turn rests upon Waldhausen's deep result on knot classification by knot group plus peripheral system.