Abstract Homotopy Theory: The Interaction of Category Theory and Homotopy theory

Timothy Porter

Abstract

This is the introduction and first section of an expanded version of notes for a series of lectures given at the Corso estivo Categorie e Topologia organised by the Gruppo Nazionale di Topologia del M.U.R.S.T. in Bressanone, 2 - 6 September 1991. Those notes have been brought up to date by the addition of new references and a summary of what has happen in the area in the last ten years. A postscript file of the complete article is available here.

In 1991, the Italian Gruppo Nazionale di Topologia organised a summer course on Category Theory and Topology in the delightful setting of the town of Bressanone. There were several series of lectures planned and they asked me to give some on abstract homotopy theory and its interaction, both with category theory and with topology. The notes for the course were typed up and were made available as preprints from both Genova and Bangor. In the ten years since that meeting, I have been asked on several occasions if there was a published version of the course, or was there a version online somewhere. It seemed that others had found the notes to be of some use as a way into the area, giving, as I had hoped, some intution about what the problems were, what the potential for new applications is and so on. There has been no published version, nor web version available until now and my stock of hard copies was diminishing, and the contents was not as up to date as they might be, so ... .

When it was suggested that I write an `overview' article on my own area of research for Cubo, I realised that there was here a good opportunity to prepare an updated version of the notes. It would fulfil the request from the editor and would make the notes more widely available, as a similar version could be put `on-line'. In fact, in retyping the notes I found that not many changes did need making and this article has retained the form of notes for an informal postgraduate course (complete with suggestions for exercises etc.) The main body of the material has been updated only in as much as the references have been changed to account for publication of material previously in preprint form, but in the last ten years several excellent texts and papers on the subject matter have been published so I have added another section with a discussion of the view of the various topics today and have given a supplementary bibliography.

I hope the notes are still found useful. They were aimed at postgraduate students one year into their studies, so are intended to be approachable.

T.P. Bangor, 2001.

Aims of the course.

To give the background and some historical perspective on Abstract Homotopy Theory.
To introduce some of the key ideas of abstract homotopy, its different `schools' and how they interact and to indicate the sources in which they may be found.
To try to help build up categorical and geometrical intuition on the subject which, by its very abstraction, can seem too unapproachable to be of use.
To provide, where possible, a unifying `overview', by concentrating on certain `themes' that illustrate:

what is abstract homotopy theory,
why the problems that form the subject matter of abstract homotopy theory can be of interest outside the confines of that theory,

Prerequisites

A basic knowledge of homotopy theory with some intuition about the fundamental group (or fundamental groupoid), polyhedra, etc. A knowledge of the higher homotopy groups would be useful, but is not essential. In the last section, some covering space theory is used, but this can be approached via sheaf theory for the reader with a categorical or algebraic geometric background or can be found in books such as Massey []. All the material on groupoids and much more can be found in Brown [].
A first course in category theory covering categories, functors, natural transformations, limits (and colimits) and the idea of adjoint functors. (At the Corso estivo such a concentrated course was given by R. Betti.) This subject matter corresponds approximately to the first five chapters of Mac Lane [], but we will not need the depth given there.
For the last section of these notes, some ideas about sheaves and toposes are assumed. At the Corso estivo, such material was beautifully covered in lectures by F. Borceux and the notes are available (Borceux, []).

1 Introduction to Abstract Homotopy Theory.

To understand the possible aims of abstract homotopy theory, it will help to list some general `problems', some of which are of a philosophical or metatheoretic nature, others are very pragmatic and practical.

Notions of deformation and of homotopy occur in many parts of mathematics. Such ideas are central in topological homotopy theory, but what is homotopy theory? What is `a homotopy theory'? As one talks of homology and cohomology theories, what explicit structures, that exist in the topological case, are needed for homotopy theory? In other words: `what is homotopy theory?' and `what makes homotopy theory work?'

In algebraic topology, one often looks for algebra that models the topology. How can one look for algebraic models of homotopy types and once one has found them, can one `do homotopy theory' with them? This leads to a related problem:

Early models for algebraic information on homotopy types, such as chain complexes, led to the study of chain complexes of modules, etc., which represented `algebraic homotopy types'. Their study led to the growth of homological algebra. Finding newer, fuller algebraic models for homotopy types that take account of non-abelian behavior would seem to open the door to a process of adaptation which could have implications in non-abelian cohomology, homotopical algebra and hence, in particular, in algebraic geometry. This depends on being able to `do' homotopy theory with such models of algebraic homotopy types, and not just at a trivial level.
To manipulate algebra, one uses categorical tools such as limits and colimits. What are the good categorical tools for working with homotopy theory?
In some of the applications of abstract homotopy theory, one is studying geometric objects such as spaces; one is using a geometric notion of homotopy, but the tools used, and even the language used, seems simpler when viewed from the abstract viewpoint (cf. Edwards and Hastings [] for strong shape theory, Porter [] and Herna\'ndez-Porter [], [] for proper homotopy theory using inverse systems and Baues [], [] for an approach to classical homotopy questions with remarkable new results, using in part insight gained by the abstract homotopy theory he develops. There is also work by Baues with Ayala, Marquez, and Quintero which is in preprint form and uses a direct abstract homotopy theory approach to proper homotopy theory.)