No label defined (Q3840286)

This paper systematically studies quantum invariants from the skein theoretical point of view. By a skein theory the author means a monoidal and involutive functor from the category of compact oriented 3-manifolds with isotopy classes of oriented embeddings to the category of \(k\)-modules. Here \(k\) is a commutative ring with unit. This definition unifies many known skein theories. The main examples considered are the Kauffman bracket skein theory, the skein theory associated with linking, and the HOMFLY skein theory. Starting from the formal definition of a skein theory, the paper develops a method of constructing invariants of 3-manifolds, generalizing Lickorish's method of constructing the \(SU(2)\)-invariants from the Kauffman bracket skein theory. In this way one can obtain the Murakami-Ohtsuki-Okada invariants from the skein theory associated with linking, and the \(SU(n)\)-invariants from the HOMFLY skein theory. The latter was done by \textit{Y. Yokota} [in `Skeins and quantum \(SU(N)\) invariants of 3-manifolds', Math. Ann. 307, No. 1, 109-138 (1997; Zbl 0953.57009)]. Further, the author studies common properties of the invariants obtained in this way. Section 2 deals with cohomological refinements of the invariants. It is shown how to obtain invariants of a 3-manifold \(M\) equipped with a cohomology class \(\sigma\in H^1(M,{\mathbb{Z}}/n)\). Similar results are discussed in Section 3, for manifolds equipped with spin structures. This section starts from studying of the invariants obtained from the HOMFLY skein theory. In [Quantum invariants for 3-manifolds, Proc. of the 3rd Korea-Japan School of knots and links (to appear)], \textit{H. Murakami} decomposed these invariants using some spine type structures. He observed that for \(n=2\) these are the spin structures, and the corresponding refinements were studied by other people, including the author. For \(n>2\), he only gave a combinatorial description of the structures and asked for a topological interpretation. The author provides such interpretation. He introduces the notion of a \(\text{Spin}({\mathbb{Z}}/n)\)-structure on a manifold \(M\) as a homotopy class of fiber maps from the stable tangent bundle \(\tau _M\) to \(\pi^*_n(\gamma _{SO})\), the pull-back of the canonical vector bundle over \(BSO\). In conclusion, in Section 4 the author discusses topological quantum field theories associated with the invariants studied in the paper. For the entire collection see [Zbl 0890.00048].