Combinatorial homotopy and 4-dimensional complexes (Q1188855)

This book is addressed to the following (intimately related) classification problems of low dimensional homotopy theory, namely: (i) Classify homotopy types of 4-dimensional CW-complexes by means of algebraic invariants. (ii) Compute the set of homotopy classes [X;Y] for CW-complexes of dimensions \(\leq 3.\) The new invariants which are used for the classification are quadratic modules and quadratic chain complexes. The quadratic chain complex associated to a CW-complex X is a chain complex \[ ...\sigma_ 5\to \sigma_ 4\to \sigma_ 3\to \sigma_ 2\to \sigma_ 1 \] together with a map \(\omega: C_ 2\otimes C_ 2\to \sigma_ 3\) satisfying several identities. Here the \(\sigma_ n\) are groups (abelian for \(n\geq 4)\) on which \(\sigma_ 1\) acts. This quadratic chain complex determines the homology groups of X as well as the homotopy groups \(\pi_ n(X)\), \(n\leq 3\), and \(image(\pi_ 4(X)\to H_ 4(\hat X))\), where \(\hat X\) is the universal cover of X. The ``quadratic structure'' of this complex is contained in the map \(\omega\) above, where \(C_ 2\) is the 2-dimensional cellular chain group of X. Roughly speaking, the importance of this quadratic structure stems from the fact that the Hopf map \(\eta: {\mathbb{S}}^ 3\to {\mathbb{S}}^ 2\) satisfies a quadratic distributivity law. On the category of quadratic chain complexes there is the notion of homotopy between maps and the main results of the book which are new in literature can be stated as follows: Theorem: The full homotopy category of connected 3-dimensional CW- complexes is equivalent to the homotopy category of 3-dimensional totally free quadratic chain complexes. Theorem: The homotopy types of connected 4-dimensional CW-complexes are in 1-1-correspondence with the homotopy types of 4-dimensional totally free quadratic chain complexes. It is also shown that any 5-dimensional totally free quadratic chain complex (resp. any morphism of 4-dimensional totally free quadratic chain complexes) is realizable by a 5-dimensional CW-complex resp. by a map between 4-dimensional CW-complexes. Besides these new results, this book is a rather self contained Introduction into low dimensional homotopy theory. It contains J. H. C. Whiteheads classification of simply connected 4-dimensional CW-complexes resp. the classification of connected 3-dimensional CW-complexes. Other topics discussed in this book are the CW-tower [see \textit{H. J. Baues}, Algebraic homotopy (1989; Zbl 0688.55001)]; and the concept of crossed modules and crossed chain complexes, which are important tools in unstable homotopy theory. One chapter discusses the problems of classifying homotopy types by means of cohomological invariants. In particular the Pontryagin square which can be used in the classification of simply connected 4-complexes is examined in detail. It turns out that there is no universal Pontryagin square with local coefficients. The obstruction for the existence of such an operation is related to the exotic element in the algebraic K-theory \(K_ 3({\mathbb{Z}})\). The final chapter is concerned with the calculus of tracks (i.e. homotopy classes of homotopies). This calculus is related to the concepts of linear extensions of categories resp. the cohomology theory of categories which generalize the well known concepts of central extensions of groups and the cohomology modules of groups. There are several appendices containing degressions from the main stream of the book (e.g., classification up to stable homotopy type) or explicit examples (e.g., an algebraic description of the homotopy category (stable and unstable) of pseudo projective planes, i.e. spaces of the form \({\mathbb{S}}^ 1\cup_ ne^ 2\) where n denotes a self map of degree n).