scientific article; zbMATH DE number 1134571

From MaRDI portal
Publication:4382046





Let \({\mathcal H} {\mathcal L} (k,M)\) denote the set of homotopy (not isotopy) classes of \(k\)-component links in a 3-manifold \(M\). The author proves the following theorem, using techniques which have been developed by him and \textit{X. S. Lin} in a previous paper [J. Am. Math. Soc. 3, No. 2, 389-419 (1990; Zbl 0704.57016)]: Theorem 1: Let \(M\subset N\) be an inclusion of simply connected 3-manifolds. Then the induced mapping \({\mathcal H} {\mathcal L} (k,M) \to {\mathcal H} {\mathcal L} (k,N)\) is a bijection. For non-simply connected manifolds, this statement is not true, as an example given by the author shows. The author names the following applications of this theorem: (1) Together with results of [\textit{Habegger} and \textit{Lin}, loc. cit.], links in simply connected 3-manifolds can be classified up to homotopy. (2) Link homotopy cannot be used to detect a fake 3-sphere. (3) It has been asked by Milnor whether for compact \(M\), \({\mathcal H}{\mathcal L}(k,M)\) was determined by the homotopy type of \((M,\partial M)\). The theorem shows that if one was to construct a counterexample, this should be non-simply connected. For the entire collection see [Zbl 0882.00042].











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4382046)