Algebraic topology of finite topological spaces and applications (Q2276273)

There is a one-to-one correspondence between the topologies on a finite set \(X\) and the partial preorders on \(X\) which induces a one-to-one correspondence between \(T_0\)-topologies and partial order relations; so, finite topological spaces, which are studied in this book, are mainly posets (partially ordered sets). Following the common idea that spaces with a finite number of points are irrelevant for topological or homotopical considerations, it is usual to replace a poset by (the geometric realization of) its order complex (the simplicial complex whose simplices are given by the chains of the poset). But a poset can also be considered as a \(T_0\)-space and its homotopy type is different from the homotopy type of its order complex. This book illustrates convincingly the idea that the study of finite non-Hausdorff spaces from a homotopical point of view is useful in many areas and can even be used to study well-known problems in classical algebraic topology. In the first chapter, the author recalls important classical results about the homotopy type of finite posets which go back essentially to two articles published in 1966 by \textit{R. E. Stong} [``Finite topological spaces'', Trans. Am. Math. Soc. 123, 325--340 (1966; Zbl 0151.29502)] and \textit{M. C. McCord} [``Singular homology groups and homotopy groups of finite topological spaces'', Duke Math. J. 33, 465--474 (1966; Zbl 0142.21503)]. Following Stong, a finite poset \(P\) has a unique core (a strong deformation retract which has no beat points) up to isomorphism; this core is obtainable by removing beat points. Following McCord, a poset and its order complex have the same weak homotopy type. In particular, this shows that a compact polyhedron can be modeled, up to weak homotopy, by a finite space; this result will be extended in Chapter 7 to \(h\)-regular CW-complexes (which are an extension of regular CW-complexes). Two posets having the same weak homotopy type may have very different homotopy types and, in Chapter 5, the author develops the notion of strong homotopy based on the notion of contiguity between simplicial maps for defining a strong homotopy type in the framework of simplicial complexes corresponding to the homotopy type in the framework of posets. In Chapter 2, various usual constructions of algebraic topology (fundamental group, Euler characteristic, join, product, quotient, wedge, mapping cylinder) are described in the context of finite spaces and the author gives another proof of the fact that a finite group can be realized as the automorphism group of a finite poset, improving earlier results of Birkhoff (1946) and Thornton (1972). In Chapter 3, the notion of minimal finite model is introduced (a minimal finite model of a polyhedron \(Y\) is a finite space of minimum cardinality weakly homotopy equivalent to \(Y\)); it includes the case of spheres (answering a conjecture of J.P. May) and the case of finite graphs (or 1-dimensional CW-complexes). The homotopy type of a poset may be obtained by addition or removal of beat points. In Chapter 4, it is shown that addition or deletion of weak points (beat points are weak points) give rise to a notion called simple homotopy which corresponds to the simplicial simple homotopy type under the correspondence between a poset and its order complex. So, the situation is much easier in the finite setting in the sense that elementary moves are given simply by the addition or the deletion of (weak) points. Nevertheless, the parallel between the two situations is not total because in the finite framework and contrary to the case of CW-complexes, a homotopy equivalence is a simple homotopy equivalence. The general procedure of reducing a poset by considering certain points (as beat points for homotopy, weak points for simple homotopy) is explored in Chapter 6 where it is proved that one-point reduction methods are not sufficient to describe weak homotopy types of finite spaces. The last chapters present various applications of these results to well known classical problems. In the celebrated paper [Adv. Math. 28, 101--128 (1978; Zbl 0388.55007)], \textit{D. Quillen} considered the poset \(S_p(G)\) of nontrivial \(p\)-subgroups of a group \(G\) ordered by inclusion but didn't consider it for itself and studied instead its order complex and formulated his conjecture: if the order complex of \(S_p(G)\) is contractible, then \(G\) has a nontrivial normal \(p\)-subgroup. By considering directly \(S_p(G)\) as a topological space, the author gives various alternative presentations of the Quillen conjecture (in Chapter 8, by following previous results of Stong and by the consideration of equivariant homotopy for posets and in Chapter 9, by considering \(S_p(G)\) as a reduced lattice). Chapter 10 is devoted to the fixed point property for posets and, providing an alternative approach to Oliver's work, a stronger version of the Lefschetz Theorem for simplicial automorphisms is obtained. In Chapter 11, the author introduces the notion of quasi constructible complexes and shows that these 2-dimensional complexes satisfy the geometric Andrews-Curtis conjecture if they are contractible (i.e., they can be deformed into a point by a sequence of collapsing and anticollapsing which involve complexes of dimension less than or equal to 3). The book ends with a chapter collecting various useful results on simplicial complexes and CW-complexes. This book is a revised version of the PhD Thesis of the author. Most of the results have appeared in various articles, while others are yet unpublished. All the concepts introduced with the chapters are usefully illustrated by examples and the recollection of all these results gives a very nice introduction to a domain of growing interest.