A mapping theorem for topological complexity (Q2351933): Difference between revisions

Let \(\text{TC}_n(X)\) denote the (normalized) \(n\)th topological complexity of a space \(X\). Among several nice results in this paper on \(\text{TC}\), I consider the following two as most interesting: Theorem 1. Let \(f_j: Y_j\to X, j=1,2\) be two maps of connected spaces with each \((f_i)_*: \pi_i(Y_j)\to\pi_i(X)\) injective for all \(i\). Suppose also that the images subgroups of \((f_i)_*\) are complementary in \(\pi_i(X)\) for all \(i\). The for all \(n\geq 2\) we have \(\text{cat}(Y_1\times Y_2\times X^{n-2})\leq\text{TC}_n(X)\). Here we say that two subgroups \(A\) and \(B\) of a group \(G\) are complementary if \(A\cap B=\emptyset\) and if every element of \(G\) can be written as a product \(ab\) with \(a\in A\), \(b\in B\). As a corollary, the authors establish the inequality \(\text{cat}(F\times B)\leq\text{TC}(E)\) where \(F\to E \to B\) is a fibration that admits a homotopy section. The Avramov-Félix conjecture, cf. \textit{Y. Félix} et al. [Rational homotopy theory. New York, NY: Springer (2001; Zbl 0961.55002)] says that the rational homotopy Lie algebra \(\pi_*(\Omega X)\otimes \mathbb Q\) of a simply connected hyperbolic finite complex should have a Lie subgroup that is free on two generators. Félix-Halperin-Thomas proved the theorem for \(X\) with \(\text{cat} X=2\). The authors prove the conjecture in case \(\text{TC}(X_{\mathbb Q})=2\) (a short version of Theorem 7.)