Algebraic Topology A Structural Introduction 1st Edition Marco Grandis

Full download ebook at ebookname.
com
Algebraic Topology A Structural Introduction 1st

Edition Marco Grandis
https://ebookname.com/product/algebraic-topology-a-
structural-introduction-1st-edition-marco-grandis/
OR CLICK BUTTON
DOWLOAD NOW
Download more ebook from https://ebookname.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Introduction to Algebraic Topology 1st Edition Holger

Kammeyer
https://ebookname.com/product/introduction-to-algebraic-
topology-1st-edition-holger-kammeyer/
Algebraic Topology A Primer 1st Edition Satya Deo

(Auth.)
https://ebookname.com/product/algebraic-topology-a-primer-1st-
edition-satya-deo-auth/
Homology Cohomology and Sheaf Cohomology for Algebraic

Topology Algebraic Geometry and Differential Geometry
1st Edition Jean H. Gallier
https://ebookname.com/product/homology-cohomology-and-sheaf-
cohomology-for-algebraic-topology-algebraic-geometry-and-
differential-geometry-1st-edition-jean-h-gallier/
Introduction to Algebraic Geometry Justin R. Smith
https://ebookname.com/product/introduction-to-algebraic-geometry-
justin-r-smith/
Introduction to Aircraft Structural Analysis 1st
Edition T.H.G. Megson
https://ebookname.com/product/introduction-to-aircraft-
structural-analysis-1st-edition-t-h-g-megson/
Introduction to Algebraic and Constructive Quantum

Field Theory John C. Baez
https://ebookname.com/product/introduction-to-algebraic-and-
constructive-quantum-field-theory-john-c-baez/
Local Cohomology An Algebraic Introduction with

Geometric Applications 2ed. Edition Brodmann M.P.
https://ebookname.com/product/local-cohomology-an-algebraic-
introduction-with-geometric-applications-2ed-edition-brodmann-
m-p/
Dopamine 1st Edition Marco Diana
https://ebookname.com/product/dopamine-1st-edition-marco-diana/
An Invitation To Algebraic Numbers And Algebraic

Functions 1st Edition Franz Halter-Koch (Author)
https://ebookname.com/product/an-invitation-to-algebraic-numbers-
and-algebraic-functions-1st-edition-franz-halter-koch-author/
12586_9789811248351_tp.indd 1 3/12/21 4:15 PM
Other World Scientific Titles by the Author
An Elementary Overview of Mathematical Structures:

Algebra, Topology and Categories
ISBN: 978-981-122-031-9
Manifolds and Local Structures: A General Theory

ISBN: 978-981-123-399-9
Category Theory and Applications: A Textbook for Beginners

Second Edition
ISBN: 978-981-123-608-2
Yumeng - 12586 - Algebraic Topology.indd 1 15/11/2021 9:03:06 am

12586_9789811248351_tp.indd 2 3/12/21 4:15 PM
Published by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.
ALGEBRAIC TOPOLOGY
A Structural Introduction
Copyright © 2022 by World Scientific Publishing Co. Pte. Ltd.
All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy
is not required from the publisher.
ISBN 978-981-124-835-1 (hardcover)

ISBN 978-981-124-836-8 (ebook for institutions)
ISBN 978-981-124-837-5 (ebook for individuals)
For any available supplementary material, please visit

https://www.worldscientific.com/worldscibooks/10.1142/12586#t=suppl
Printed in Singapore
Yumeng - 12586 - Algebraic Topology.indd 2 15/11/2021 9:03:06 am

November 2, 2021 9:51 bk-9x6 12586-main page v
To my colleagues and friends

B1948 Governing Asia
This page intentionally left blank
B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

November 2, 2021 9:51 bk-9x6 12586-main page vii
Preface
Algebraic Topology is a system and strategy of partial translations, aiming

to reduce difficult topological problems to algebraic facts that can be more
easily solved. This interface role, between topology and algebra, is the
source of its power and attraction.
Each of these translations views a topological space in a particular light,
and codifies this picture into an algebraic structure — typically a group,
or a graded abelian group, or a graded ring. In this view, a huge amount
of information is dropped, and a manageable part is kept. Choosing an
appropriate translation, we can hope to solve the topological problem we
are analysing.
The main subject of this book is singular homology, the simplest of these
translations. Studying this theory and its applications we also investigate
parts of other disciplines, which form its underlying structural layout (and
in part grew out of it):
- Homological Algebra, for tensor products, the Hom functor and their
derived functors,
- Homotopy Theory, for the homotopy groups and fundamental groupoid,
- Category Theory, the grammar of translations between mathematical
fields.
This book is an introduction to a complex domain, with references to
its advanced parts and ramifications. It is written with minimal prerequi-
sites — basic general topology and little else — and a moderate progression,
starting with a very elementary beginning. A consistent part of the exposi-
tion is organised in the form of exercises, with suitable hints and solutions.
It can be used as a textbook for a first course on Algebraic Topology or
self-study, and a guidebook for further study.
vii
B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

November 2, 2021 9:51 bk-9x6 12586-main page ix
Contents
Preface page vii

Introduction 1
0.1 Investigating spaces with algebraic structures 1
0.2 Homology and cohomology theories 2
0.3 An outline 3
0.4 Translations and underlying affinities 4
0.5 An inductive approach on structural bases 5
0.6 Prerequisites and literature 5
0.7 Notation and conventions 6
0.8 Acknowledgements 7
1 Introducing Algebraic Topology 9
1.1 Classifying spaces and maps 9
1.2 Transforming problems 17
1.3 The terminology of categories and functors 25
1.4 Paths and homotopy 34
1.5 Natural transformations and equivalence of categories 44
1.6 Products, sums and universal properties 50
2 Singular homology 61
2.1 Chain complexes and their homology 62
2.2 The singular homology groups 72
2.3 Mayer–Vietoris and subdivision 82
2.4 The homology of the spheres 95
2.5 Computing homology 108
2.6 Compact surfaces and projective spaces 115
2.7 Diagram lemmas in Homological Algebra 123
2.8 Complements 127
3 Relative singular homology and homology theories 137
3.1 Main definitions 137
ix
November 2, 2021 9:51 bk-9x6 12586-main page x
x Contents
3.2 Excision and compact pairs 144
3.3 Local homology and orientable manifolds 149
3.4 Eilenberg–Steenrod axioms 155
3.5 Alexander–Spanier cohomology 162
3.6 The product in Alexander–Spanier cohomology 169
3.7 Hints at de Rham cohomology 171
4 Singular homology with coefficients 177
4.1 Tensor product of modules and Hom 177
4.2 Additive and exact functors 185
4.3 Relative singular homology with coefficients 192
4.4 Relative singular cohomology with coefficients 197
4.5 *Changing the coefficient group 202
5 Derived functors, universal coefficients and products 207
5.1 Derived functors 207
5.2 Torsion product and universal coefficients 216
5.3 Ext functor and universal coefficients 221
5.4 Tensor product of chain complexes 225
5.5 Products in singular cohomology 230
5.6 *Acyclic models and products of spaces 240
6 An introduction to homotopy groups 251
6.1 The fundamental groupoid 251
6.2 The fundamental group 258
6.3 Moving or dropping the basepoint 266
6.4 The van Kampen Theorem 272
6.5 Hints at the higher homotopy groups 280
7 Complements on categories and topology 285
7.1 An overview of categorical limits and adjoint functors 285
7.2 The compact-open topology 293
8 Solution of the exercises 299
8.1 Exercises of Chapter 1 299
References 353
Index 357
November 2, 2021 9:51 bk-9x6 12586-main page 1
Introduction
0.1 Investigating spaces with algebraic structures

Algebraic Topology studies topological spaces and continuous mappings,
transforming spaces into algebraic structures (like groups and rings) and
maps into homomorphisms. The general aim is to reduce difficult topolog-
ical problems to simpler algebraic facts.
For instance, the following figure shows two surfaces, the sphere S2 and
the ‘torus’ T (the surface of a lifebuoy)
. . . . . . . . . ..
... . (0.1)
We have also drawn a great circle on the sphere, to suggest its curvature.
Plainly, the sphere and the torus have a different shape: we cannot
continuously deform one space into the other. However, proving that these
subspaces of R3 are not homeomorphic is not an elementary thing, like
many seemingly obvious facts, in Topology.
To distinguish these spaces, we can think of a ‘loop’ (a closed path) on
these surfaces. On the sphere, every loop can be continuously shrunk to a
point, but a tight string around the lifebuoy cannot be further contracted —
staying in the surface we are considering.
All this can be made precise with elementary results of Algebraic Topol-
ogy: the homology group H1 (S2 ) is trivial, while H1 (T) is not. This
is sufficient to prove that the sphere and the torus are not homeomor-
phic: two homeomorphic spaces X, Y have isomorphic homology groups
Hn (X) ∼ = Hn (Y ), in every degree n > 0.
1
2 Introduction
The homology groups of a space have a geometrical meaning, which is
often fairly intuitive: for instance, H1 (T) is a free abelian group with two
generators, which can be realised as the homology classes [a], [b] of two
loops a and b, around the ‘equator’ and a ‘meridian’ of the torus (with
respect to the axis of rotation)
(0.2)
b
a
The homology class 3[a] + 2[b] can be realised as a loop which — glob-
ally — turns thrice around the equator and twice around a meridian. Neg-
ative coefficients turn up by reversing a loop.
Loosely speaking, the homology group H1 (X) explores 1-dimensional
properties of a space, and can detect 1-dimensional holes, as we have seen.
Similarly, the homology group H2 (X) is related with 2-dimensional fea-
tures. For the sphere and the torus, the groups H2 (S2 ) and H2 (T) are
infinite cyclic, detecting a 2-dimensional ‘cavity’ in each space; moreover,
the choice of a generator in H2 (S2 ) or H2 (T) amounts to choosing an ori-
entation, in each of these manifolds.
The n-dimensional sphere Sn (for n > 0) has only two non trivial groups,
namely H0 (Sn ) and Hn (Sn ), both isomorphic to the additive group of the
integers. This shows that all the spheres are really distinct spaces: Sm and
Sn are only homeomorphic if m = n, and justifies our idea of topological
dimension. It is again an important fact, which looks obvious in low di-
mension, but is not easily proved with ordinary topological tools (after the
easy cases, for m 6 1 or n 6 1).
There are diverse homology theories of topological spaces, but all of
them coincide on ‘reasonably good’ spaces. More precisely, all homology
theories satisfying Eilenberg–Steenrod axioms, with the same coefficient
group, give the same result on every triangulable space, up to isomorphism
(see Section 3.4).
0.2 Homology and cohomology theories

Basically, a homology theory is a sequence of transformations Hn (−); each
of them turns a topological space X into an abelian group Hn (X), and a
continuous mapping f : X → Y into a homomorphism Hn (f ) : Hn (X) →
Hn (Y ). In this transformation, composition of maps and identity maps are
preserved:
0.3 An outline 3
- for consecutive maps f : X → Y and g : Y → Z, we have: Hn (gf ) =
Hn (g)Hn (f ),
- for a space X we have: Hn (id X) = id Hn (X).
In the basic terminology of category theory, each Hn is a covariant functor,
from the category Top of topological spaces (and continuous mappings) to the
category Ab of abelian groups (and homomorphisms). As an obvious conse-
quence, each Hn preserves invertible arrows: it transforms a homeomorphism
of spaces into an isomorphism of groups.
In fact, a homology theory turns up in a richer form: for every ‘relative
pair’ (X, A) of topological spaces (where A is a subspace of X), and a fixed
abelian group G, we have a sequence of groups Hn (X, A; G) of relative
homology, with coefficients in G. The basic case is absolute homology with
integral coefficients: Hn (X) = Hn (X, ∅; Z).
On the other hand, a cohomology theory H n (−) takes a continuous map-
ping f : X → Y to a homomorphism H n (f ) : H n (Y ) → H n (X), reversing
the direction of arrows, reversing composition and preserving identities.
Here also we have an enriched version, for relative pairs of topological
spaces and a coefficient group G.
An important fact appears: the cohomology groups with coefficients in
a ring R have a graded multiplication, and their family forms a graded
R-algebra. We shall study singular cohomology (a by-product of singular
homology), and briefly review other theories, namely Alexander–Spanier
cohomology and de Rham cohomology, which naturally arise in contravari-
ant form.
0.3 An outline
Chapter 1 is an introduction to the goals and methods of Algebraic Topol-
ogy, with a brief analysis of the elementary issues of category theory in-
volved in the interface between topology and algebra.
Chapter 2 introduces singular homology, the simplest homology theory of
general topological spaces, showing how to compute the homology groups,
and get information on spaces and maps. Section 2.4 is devoted to the
homology groups of the spheres, and their consequences: for instance,
the Theorem of Topological Dimension, the topological degree of a map
Sn → Sn , and the theorem about vector fields on even-dimensional spheres.
Several computations of homology groups can be found in Sections 2.5, 2.6.
Singular homology is constructed in the cubical form, that — in the author’s
opinion — gives a simpler approach. The more usual simplicial form is briefly
presented in Section 2.8; their equivalence is proved in Section 5.6.
4 Introduction
In Chapter 3 we extend singular homology to relative pairs (X, A). Sec-
tion 3.3 shows how relative homology is able to investigate local features,
for instance the local and global orientation of topological manifolds. The
Eilenberg–Steenrod axioms for relative homology theories are listed in Sec-
tion 3.4, and verified for relative singular homology. Alexander–Spanier
cohomology, described in Section 3.5, satisfies the dual axioms, for a rel-
ative cohomology theory; its graded product is dealt with in Section 3.6.
Similarly, de Rham cohomology of differentiable manifolds gives a graded
algebra on the real field (see Section 3.7).
Singular homology and cohomology with a coefficient group is studied
in Chapters 4 and 5. Varying the coefficient group we may be able to get
results that the ordinary theory (with integral coefficients) cannot obtain,
as shown in Subsection 4.3.7. All this requires some tools of Homological
Algebra, namely the tensor and torsion products for homology, or the Hom
and Ext functors for cohomology (in Sections 4.1 and 5.1–5.3).
Chapter 5 also explores the multiplicative structure of singular cohomol-
ogy with coefficients in a ring (in Section 5.5), and the homology groups of a
product of spaces (in Section 5.6). The latter are detemined by Eilenberg–
Mac Lane’s Acyclic Model Theorem, which is also used to prove the equiv-
alence of the cubical and simplicial constructions of singular homology.
Chapter 6 is an introduction to the fundamental groupoid and homotopy
groups, with their relationship to singular homology.
Chapter 7 briefly reviews issues that have already appeared in the previ-
ous chapters, on category theory (categorical limits and adjoint functors)
and general topology (the compact-open topology). Finally Chapter 8 gath-
ers most solutions to the exercises of the previous chapters.
0.4 Translations and underlying affinities

As already said, Algebraic Topology is a ‘discipline of translations’, from
topology to algebra. Each translation gives a partial view of spaces and
maps, in a particular perspective. This simplified picture may be able to
solve the problem we are studying; otherwise, another translation might
do.
The grammar of this discipline is provided by elementary category theory,
reviewed here in some sections of Chapter 1, as far as this makes clear how
the translation works.
As a simple Sinstance, the fact thatLthe functor Hn transforms a disjoint
union of spaces Xi into a direct sum Hn (Xi ) of abelian groups is the outer
appearance of an underlying property, the preservation of categorical sums: a
disjoint union of spaces and a direct sum of abelian groups satisfy the same
categorical universal property (see 1.6.2).
0.5 An inductive approach on structural bases 5

The effectiveness of a translation likely depends on the affinity between
the two languages involved, in our case the source category of topological
spaces versus the algebraic category of destination. In this regard, the cate-
gory Ab is sufficiently flexible to offer ‘covariant translations’, by homology
theories, and ‘contravariant translations’, by cohomology theories.
Less flexible categories, like those of rings and graded algebras, can only
give a ‘mirror-image’, by a cohomology functor H ∗ . But this translation is
far richer than viewing H ∗ (X) as a mere graded abelian group.
The roots of this phenomenon can be found in an affinity between Top
and Rngop , the dual of the category of rings; an affinity expressed in the-
orems, like Gelfand duality, and in formal categorical aspects sketched
in 5.5.9.
0.5 An inductive approach on structural bases

Notions will be presented in a concrete, ‘inductive’ way, starting from ele-
mentary examples. Whenever possible, the reader will be guided to build
the theory through a series of exercises.
On the other hand, the roots of the interplay between Topology and
Algebra, formalised in Homological Algebra and Category Theory, will be
investigated more deeply than usual in an introductory book.
Algebraic Topology is a complex domain, with ramifications in diverse
fields. After covering the basic parts, and some more advanced ones, we
shall sketch several developments that do not have a place here, giving
extensive references for further study.
The exposition will be particularly elementary and detailed in Chapter 1
and the first part of Chapter 2: this is not a book on general topology or
abelian groups, but the reader will be invited (and guided) to check the
topological or algebraic ground on which we are building. This point made,
we shall go on more quickly.
Chapters 1 and 2 form a basic introduction to singular homology; Chap-
ters 1 and 6 play the same role for homotopy groups; the three of them can
give a preliminary view of Algebraic Topology.
0.6 Prerequisites and literature

The only prerequisites are general topology and the very basic theory of
groups, abelian groups and rings. Categories and functors are introduced
when their need arises. The same holds for the part of Homological Alge-
bra used here: exact sequences, chain complexes of abelian groups, tensor
products and Hom-functors, their derived functors.
6 Introduction
For the fascinating field of Algebraic Topology there are elementary text-
books, like Vick [Vi] and Massey [Mas3], and more advanced ones, like
Hilton–Wylie [HiW], Spanier [Sp2] and Hatcher [Ha]; the last is freely
downloadable. The history of this discipline is dealt with in Dieudonné
[Di].
For Homological Algebra we shall refer to Cartan–Eilenberg [CE] and
Mac Lane [M1]; for General Topology to Kelley [Ke], Munkres [Mu] and
Bourbaki [Bou2].
Category Theory is exposed in well-known books, like Mac Lane [M2],
Borceux [Bo1, Bo2, Bo3], Adámek, Herrlich and Strecker [AHS]. The au-
thor’s [G5] is a textbook for beginners, also devoted to applications in
Algebra, Topology, Algebraic Topology and Homological Algebra.
0.7 Notation and conventions

Weak inclusion of sets is denoted by the symbol ⊂, instead of the more
usual ⊆.
As usual, the symbols N, Z, Q, R, C denote the sets of natural, integral,
rational, real or complex numbers; N∗ is the subset of the positive inte-
gers. Open and semi-open real intervals are denoted as ]a, b[, [a, b[, etc. —
a Bourbaki’s notation. The standard euclidean interval [0, 1] is written as
I, the standard n-sphere as Sn (see 1.4.1).
The identity mapping of a (possibly structured) set X is written as id X.
A singleton set is generally written as {∗}. The equivalence class of an
element x with respect to an assigned equivalence relation is denoted as
[x], by default. A bullet in a diagram stands for an object. The two-valued
index α (or β) varies in the set {0, 1}, also written as {−, +}.
A topological space is generally called a ‘space’, provided no ambiguity
can arise. In this domain, a ‘map’ is a continuous mapping, and ‘nbd’ stays
for neighbourhood.
In a space X, the interior of a subset A is written as intA or intX A, the
closure as clA, or clX A, or A; the subset ∂A = clA \ intA will be called the
border of A in X. (The more usual term ‘boundary’ is here reserved for
chain complexes and manifolds with boundary.)
A euclidean space is any subspace of some Rn with euclidean topology.
A ring is assumed to be unital, by default.
A part marked with * is out of the main line of exposition. It may
refer to issues dealt with further on, or be addressed to readers with some
knowledge of the subject which is being discussed, or give references for
higher topics. Marginal remarks are written in small characters.
0.8 Acknowledgements 7
Each chapter and section has its own introduction. Many subsections end
with a list of exercises, whose solution is generally deferred to Chapter 8;
but the solution can follow the exercise, when it is important for the sequel.
Easy exercises are generally left to the reader.
0.8 Acknowledgements
This book is dedicated to the colleagues and dear friends with whom I have
worked and am working.
In particular to the memory of Gabriele Darbo, my professor and friend.
To Tim Porter and Ettore Carletti, with whom I discussed many topics
which are developed here. To Bob Paré, George Janelidze and many others,
recalling the happy hours we spent together when travelling was easier than
it is today.
Diagrams and figures are composed with ‘xy-pic’, by K.H. Rose and
R. Moore — a free package.
The cover comes out of a collaboration. A spider prepared the installa-
tion; the night dew arranged crystal-clear balls, in rhythmic progressions;
the early morning light drew bright spheres out of them; I took the picture;
Mr Loo Chuan Ming gave the final design.
I would also like to thank Dr Lim Swee Cheng and Ms Liu Yumeng, at
World Scientific Publishing Co., for their kind help in the publication of
this book.
B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

1
Introducing Algebraic Topology
Algebraic Topology studies spaces and continuous mappings, transforming

them into algebraic structures and homomorphisms, in order to reduce
difficult topological problems to algebraic facts that can be more easily
solved.
For instance, proving that the homology group H1 of the circle is the
additive group Z of the integers has various consequences, from straight-
forward ones (the circle is not a retract of the plane) to others, requiring
additional work, as the Fundamental Theorem of Algebra. Computing the
homology groups of the spheres has far-reaching results; several of them
can be found in 1.1.5 and Section 2.4.
This introductory chapter is written in a very elementary way, and also
reviews well-known issues about topological spaces and abelian groups.
The reader will know which parts to skip, or browse, or read.
1.1 Classifying spaces and maps

We begin by considering the shape and the ‘homotopy shape’ of a space,
hinting at their investigation with homology groups. The homotopy rela-
tion will be further examined in Section 1.4.
1.1.1 The shape of a space

Two topological spaces X, Y are homeomorphic (written as X ∼ = Y ) if
there exists a homeomorphism X → Y , i.e. a continuous mapping which
has a continuous inverse. We also say that X, Y have the same (topological)
shape, or the same homeomorphism-type.
To prove that two given spaces have the same shape can require long
computations, in order to build a homeomorphism between them, but is
generally a ‘confined’ problem.
9
10 Introducing Algebraic Topology

To prove that they are not homeomorphic can be quite difficult, and
enticing, even when we clearly ‘see’ that they have a different shape: we
must prove that there cannot exist a homeomorphism between them.
The proof is generally based on a ‘topological property’ (invariant up to
homeomorphism) which holds in one of them but not in the other: con-
nectedness, compactness, separation or countability axioms, higher forms
of connectedness, etc. When these topological properties are explored
throughout algebraic structures we enter in the domain of Algebraic Topol-
ogy.
The following exercises show two families of spaces which is important
to classify; the first classification can be achieved with the usual means
of General Topology, the second will be completed later, using singular
homology.
Exercises and complements. The solutions to these exercises can be found
in Chapter 8.
(a) (The shape of the intervals, I ) Any non-degenerate interval of the eu-
clidean line is homeomorphic to ]0, 1[, or [0, 1[, or [0, 1].
Hints. It is an easy exercise of Calculus, which can be solved using affine linear
functions and some elementary transcendental functions: see 8.1.1. (One can also
use rational functions and their pastings, but the argument would be longer.)
(b) (The shape of the intervals, II ) These three homeomorphism-types are
distinct. Adding the degenerate cases, the intervals of R are thus classified
in five types of homeomorphism:
(i) open non-degenerate intervals,

(ii) semiopen non-degenerate intervals,
(iii) compact non-degenerate intervals,
(iv) singletons,
(v) the empty space.
Hints: consider the connected components of an interval deprived of a point.
(c) (Topological Dimension in low degree) Classifying the euclidean spaces
Rn , by proving that they represent distinct homeomorphism-types, is a cru-
cial fact in Topology, without which there would be no notion of topological
dimension.
This point will be reviewed in 1.1.5(b), but the reader can solve as of
now the lowest degrees: the singleton R0 = {0} and the euclidean line R
are not homeomorphic to any higher euclidean space Rn , for n > 1.
(d) An interval of the line is not homeomorphic to any sphere Sn (n > 0).

1.1.2 Classifying maps
Studying the maps f : X → Y between two topological spaces, we are also
interested in classifying them, ‘up to continuous deformation’.
Let us recall that two maps f, g : X → Y are said to be homotopic
(written as f ' g) if there is a map ϕ : X × I → Y defined on the cylinder
I(X) = X × I (with the product topology), that coincides with f on the
lower basis of the cylinder and with g on the upper one
ϕ : X × I → Y,
(1.1)
ϕ(x, 0) = f (x), ϕ(x, 1) = g(x) (for x ∈ X),
forming a continuous deformation of f into g. This is easily proved to be
an equivalence relation (see 1.4.4). The homotopy will also be written as
ϕ: f ' g : X → Y .
A point x ∈ X can be identified with the corresponding map x : {∗} → X,
defined on the singleton space. A homotopy a : x ' x0 : {∗} → X is then
the same as a path in X from x to x0 , i.e. a map such that
a : I → X, a(0) = x, a(1) = x0 . (1.2)
On the other hand, a homotopy ϕ : f ' g : X → Y gives a family of
paths ϕ(x, −) : I → Y from f (x) to g(x), indexed by x ∈ X and varying
continuously on X.
Exercises and complements. (a) The classification of the maps {∗} → X
up to homotopy amounts to the partition of X in path components.
(b) We have two maps f, g : X → Y with values in a euclidean space Y
(i.e. a subspace of some Rn ).
If Y is a convex subset of Rn (i.e. for all y, y 0 ∈ Y , the line segment from
y to y 0 is contained in Y ), the maps f and g are always homotopic. More
generally, the same holds if, for every x ∈ X, the line segment from f (x)
to g(x) is contained in Y .
Hints. We can use the linear structure of Rn to define the affine homotopy
ϕ : f ' g : X → Y, ϕ(x, t) = (1 − t)f (x) + tg(x), (1.3)
which describes, at each x ∈ X, the line segment from f (x) to g(x), and
is indeed an affine linear map in the variable t ∈ I. Now we only have to
verify the continuity of the mapping ϕ on X × I.
1.1.3 Classifying the homotopy type

The homotopy of maps leads to a possibility of deforming spaces, called
homotopy equivalence, which is far wider than homeomorphism.

Namely, a map f : X → Y is a homotopy equivalence if it admits a weak
inverse g : Y → X, up to homotopy equivalence: this means that gf ' id X
and f g ' id Y . Then we say that the spaces X, Y are homotopy equivalent,
or that they have the same homotopy type; it is again an equivalence relation
between spaces (see 1.4.4), written as X ' Y .
A space is said to be contractible if it has the homotopy type of the
singleton.
A subspace A ⊂ X is a deformation retract of X if there is a map p such
that
p : X → A, pm = id A, mp ' id X, (1.4)
where m : A → X is the inclusion. Forgetting the homotopy mp ' id X,

we have a simple retract; the map p is called a retraction of m, and m is
called a section of p.
We speak of a strong deformation retract if in (1.4) we can find a homo-
topy which does not move the points of A
ϕ : mp ' id X, ϕ(x, t) = x (x ∈ A, t ∈ I). (1.5)
Exercises and complements. (a) Every euclidean space Rn is contractible,

and admits {0} as a deformation retract. (Any other point would do, but
this is a general fact: see 1.4.3(b).)
(b) Every non-empty convex subspace X ⊂ Rn is contractible. More gener-
ally, the same is true of every subset X ⊂ Rn which is starred with respect
to some x0 ∈ X; this means that: if x ∈ X, the line segment from x0 to x
is contained in X.
(c) Let X be a non-empty space. The following conditions are equivalent:
(i) X is contractible,
(ii) id X is homotopic to a constant map,
(iii) for any space Y , each map Y → X is homotopic to a constant map,
(iv) for any space Y , each map X → Y is homotopic to a constant map.
(d) The sphere Sn is a deformation retract of the pierced space X =
Rn+1 \ {0}. Hints: use the normalisation mapping N
N : X → Sn , N (x) = x/||x|| (x 6= 0). (1.6)
(e) The subspace A ⊂ X is a deformation retract of X if and only if there

is a map ϕ such that
ϕ : X × I → X, ϕ(x, 0) ∈ A,
(1.7)
ϕ(x, 1) = x, ϕ(a, 0) = a (for x ∈ X, a ∈ A).

For a strong deformation retract we replace the last condition with:
ϕ(a, t) = a (for a ∈ A, t ∈ I).
1.1.4 Hints at singular homology

Singular homology will be defined in Chapter 2. The construction requires
some work, and we try to explain why it is worthwhile.
Singular homology associates:
- to a topological space X an abelian group Hk (X), for each k > 0,
- to a map f : X → Y a homomorphism Hk (f ) : Hk (X) → Hk (Y ) of groups,
for each k > 0,
so that composition and identity mappings are preserved:
(i) Hk (gf ) = Hk (g)Hk (f ), for all maps f : X → Y and g : Y → Z,
(ii) Hk (id X) = id Hk (X), for any space X.
Because of these properties the ‘transformation’ Hk is called a ‘functor’,
from spaces and maps to abelian groups and homomorphisms. Hk (X) is
called the k-th group of singular homology of the space X; Hk (f ) is called
the associated homomorphism, and is usually written as f∗ : Hk (X) →
Hk (Y ), or f∗k .
As a consequence of (i), (ii), if the spaces X, Y are homeomorphic, the
homology groups Hk (X) and Hk (Y ) are isomorphic, for all k > 0.
In fact we have two maps, inverse to each other
f : X Y : g, gf = id X, f g = id Y,
transformed by Hk into homomorphisms f∗ : Hk (X) Hk (Y ) : g∗ which
are again inverse to each other.
As a typical property of the transformations of Algebraic Topology, we
shall see that the homology functors are homotopy invariant: this means
that whenever two maps f, f 0 : X → Y are homotopic, then Hk (f ) =
Hk (f 0 ), for every k > 0.
Therefore, a homotopy equivalence f : X Y : g, with gf ' id X and
f g ' id Y , is also transformed by Hk into an isomorphism Hk (X) ∼
= Hk (Y )
of abelian groups.
The homology groups are thus invariants of the homotopy type of a space;
directly, they cannot distinguish homotopy equivalent spaces. In particular,
all the spaces Rn have the same homology groups as the singleton R0 = {∗},
which will be (easily) computed as:
H0 (Rn ) ∼
= Z, Hk (Rn ) = 0 (k > 0), (1.8)
the trivial group being (always) written as 0.

Yet, the distinction of the diverse Rn can be achieved, by suitable mod-
ifications of the spaces (which only preserve the homeomorphism type).
1.1.5 The homology of the spheres

Developing the theory, we shall learn to compute the homology groups and
the associated homomorphisms. For instance, on the standard spheres Sn
we get (in 2.4.1 and Theorem 2.4.2):
H0 (S0 ) ∼
= Z2 , Hk (S0 ) = 0 (k > 0),
(1.9)
= Hn (Sn ) ∼
H0 (Sn ) ∼ = Z, Hk (Sn ) = 0 (n > 0, 0 6= k 6= n).
Roughly speaking, a non-trivial group Hk (X) detects ‘holes’ of dimension

k in the space X.
This has many important consequences, and we begin by presenting some
straightforward ones.
(a) First, these results readily prove that no sphere is contractible. More-
over the sphere Sn is not homeomorphic, nor homotopy equivalent, to any
other sphere Sm (for m 6= n).
In fact, if the space X is homotopy equivalent to Sn then Hk (X) ∼ =
Hk (Sn ) (for all k > 0), and the integer n is determined by the homology
groups of X as: max{k ∈ N | Hk (X) 6= 0}.
(b) (Theorem of Topological Dimension) If the euclidean spaces Rm and
Rn are homeomorphic, then m = n.
In fact, the one-point compactifications Sm and Sn are homeomorphic.
Let us note the analogy with the theorem of linear dimension of vector spaces.
(c) We already remarked that we cannot get this result applying the ho-
mology functors, directly, to Rm and Rn .
Loosely speaking, the n-dimensional character of the topology of Rn ap-
pears more clearly in its Alexandrov compactification — a transformation
which preserves homeomorphism but does not preserve homotopy equiva-
lence.
(d) The argument of (b) can also be achieved replacing Alexandrov com-
pactification with a more elementary procedure, still invariant up to home-
omorphism and not to homotopy equivalence: taking out a point.
Suppose for simplicity that m, n > 0. If the euclidean spaces Rm and
R are homeomorphic, the pierced space Rm \ {0} is homeomorphic to a
n
pierced space Rn \ {x0 }, whence Sm−1 and Sn−1 are homotopy equivalent
(by 1.1.3(d)), and again m = n.

For m = 1 we are reviewing the solution to Exercise 1.1.1(c): connectedness
does not distinguish R from any other Rn , but does distinguish the pierced line
R \ {0} from any higher pierced space Rn \ {x0 } with n > 1.
(e) We already remarked, since the beginning of the general Introduction,
how the homology groups of the sphere and the torus show that these spaces
are not homeomorphic; we can now add that they are not even homotopy
equivalent.
1.1.6 Studying retracts

Showing that a subspace A ⊂ X is a retract of X is a ‘confined’ problem:
we are to find a (continuous) retraction p : X → A, such that p(x) = x
for all x ∈ A. Showing that it is not, i.e. that there is no (continuous)
retraction, can be an important result.
For instance, saying that the subspace S0 = {−1, 1} of the euclidean
line R is not a retract is an obvious consequence of the Intermediate Value
Theorem, and can be viewed as a way of stating the theorem. (The general
statement is an elementary consequence of the previous property.)
More generally, saying that the sphere Sn is not a retract of Rn+1 is
intuitively clear (at least in low dimension), but not easy to prove with the
usual tools of general topology. This fact has important consequences, like
the Brouwer fixed-point theorem (see 2.4.4(a)).
Exercises and complements. (a) For n > 0, Sn is not a retract of Rn+1 , nor
of any contractible space. Hints: use the functor Hn and its values on our
spaces, mentioned in (1.8) and (1.9).
(b) These facts can be easily extended: for a subspace A ⊂ X, the inclusion
m : A → X gives a homomorphism m∗ : Hn (A) → Hn (X) (for every n > 0),
and:
- this homomorphism need not be injective,
- if A is a retract of X, then m∗ : Hn (A) → Hn (X) has a left inverse, and
is injective.
1.1.7 Topological embeddings and projections

Even though general topology is a prerequisite of this book, we end this
section by reviewing a few elementary issues of frequent use, to fix our
terminology.
A map f : X → Y between topological spaces has a canonical factorisa-
tion
X
f
/ Y
O
p f = mgp, (1.10)
m

X/Rf / Im f
g
where:
- p is the canonical projection onto the topological quotient of X modulo
the equivalence relation associated to f (xRf x0 means that f (x) = f (x0 )),
- m is the inclusion of the subspace Im f = f (X) into the codomain Y ,
- g : X/Rf → Im f is the bijective map defined by g[x] = f (x) (x ∈ X).
If f is open (or closed), the same holds for g, which is a homeomorphism.
Here we are interested in the following cases where g is a homeomorphism.
(a) The map f : X → Y is said to be a topological embedding if it is an
injective mapping and X has the initial topology for f : the open sets of
X are (precisely) the preimages f −1 (V ) of the open sets of Y . Then f
restricts to a homeomorphism g : X → f (X) between X and a subspace of
Y
X
f
/ Y
O
(1.11)
g % m
Im f
In this case we may want to ‘identify’ X with its image in Y , so that

it becomes a subspace. Plainly, topological embeddings are stable under
composition. An injective map which is either open or closed is always a
topological embedding.
(b) The map f : X → Y is said to be a topological projection if it is a
surjective mapping and Y has the final topology for f : the open sets of Y
are (precisely) the subsets V whose preimage f −1 (V ) is open in X. Then
f induces a homeomorphism g : X/Rf → Y between a quotient of X and
the space Y
X
f
/ Y
9
p (1.12)
g
X/Rf
Here we may want to identify Y with this quotient of X. Topological

projections are stable under composition, an advantage with respect to
‘strict’ topological quotients. (Note that, in Set Theory, a quotient of a
quotient of the set X is not a quotient of X.)

A surjective map which is either open or closed is always a topological
projection. In particular, a surjective map from a compact space to a
Hausdorff space is closed, and a topological projection.
1.1.8 Cover Lemmas

S
Let (Xi )i∈I be a cover of the space X, that is X = Xi . The continuity of
a mapping f : X → Y with values in a topological space Y can be reduced
to the continuity of its restrictions fi : Xi → Y in two main cases.
Exercises and complements. (a) (The Open Cover Lemma) Let (Xi )i∈I be
an open cover of the space X (i.e. all Xi are open subsets of X). Then
f : X → Y is continuous if and only if all its restrictions fi : Xi → Y are.
(b) (The Finite Closed Cover Lemma) Let (Xi )i∈I be a finite closed cover
of the space X (i.e. all Xi are closed subsets of X, and I is finite). Then
f : X → Y is continuous if and only if all its restrictions fi : Xi → Y are.
(c) The general situation is a space X and a cover (Xi )i∈I such that X has
the finest topology that makes all inclusions ui : Xi → X continuous.
*Categorically, this means that X is the colimit of the inclusions Xi → X.*
1.2 Transforming problems

We want to transform topological spaces into algebraic structures (typically
groups) and continuous mappings into homomorphisms. Technically, this
means to construct functors, from a category of topological spaces to a
category of algebraic structures.
Here we give a brief presentation of these topics, making informal use of
issues that will be made precise in the sequel.
The set of maps X → Y between two topological spaces is written as
Top(X, Y ).
1.2.1 Hints at categories and functors

We transform topological problems into algebraic ones setting up a proce-
dure F that transforms
- a topological space X into an abelian group F (X) (or some other algebraic
structure),
- a map f : X → Y into a homomorphism F (f ) : F (X) → F (Y ),
so that composition and identities are preserved — as in 1.1.4, in the case
of singular homology.

Such a procedure is denoted as F : Top → Ab and called a (covariant)
functor, from the category Top of topological spaces and continuous map-
pings to the category Ab of abelian groups and homomorphisms. These
issues will be defined in Section 1.3.
Here it will be sufficient to say that a category C consists of objects
X, Y, ... and morphisms f : X → Y, ... between objects; the latter have a
(partial) composition law: for each pair of ‘consecutive’ morphisms f : X →
Y and g : Y → Z there is a composed morphism gf : X → Z. This partial
operation is associative (whenever composition is legitimate) and every
object X has an identity, written as id X : X → X, or 1X , which acts as a
unit for legitimate compositions. The notation C(X, Y ) denotes the set of
morphisms X → Y in C.
The prime example is the category Set of sets (and mappings), where:
- an object is a set,
- the morphisms : X → Y between two sets are the (set-theoretical) map-
pings from X to Y , and form the set Set(X, Y ),
- the composition law is the usual composition of mappings: (gf )(x) =
g(f (x)).
We also need the following categories of structured sets and structure-
preserving mappings (with the usual composition):
- the category Top of topological spaces (and continuous mappings),
- the category Ab of abelian groups (and their homomorphisms),
- the category Gp of groups (and their homomorphisms),
- the category R Mod of R-modules on a commutative ring R (and their
homomorphisms),
- the category Rng of rings (and their homomorphisms),
- the category Set• of pointed sets (and pointed mappings),
- the category Top• of pointed topological spaces (and pointed maps).
Other categories of structured sets will be introduced later on.
A functor F : Top → Ab is defined as above. More generally, we define
in a similar way a functor F : C → D, from a category C to a category D.
(Contravariant functors are introduced in 1.2.2(iii).)
Remarks and complements. (a) By default, a ring R is assumed to be
associative and unital; an R-module is assumed to be unitary.
(b) We recall that a pointed set is a pair (X, x0 ) consisting of a set X and
a basepoint x0 ∈ X; a pointed mapping f : (X, x0 ) → (Y, y0 ) is a mapping
f : X → Y such that f (x0 ) = y0 ; the composition is the usual one.

Similarly, a pointed topological space (X, x0 ) is a space with a basepoint,
and a pointed map f : (X, x0 ) → (Y, y0 ) is a continuous mapping from X
to Y such that f (x0 ) = y0 . (We shall see that the fundamental group
π1 (X, x0 ) is defined for a pointed topological space.)
(c) When a category is named after its objects alone (e.g. the ‘category of
topological spaces’), this means that the morphisms are understood to be
the obvious ones (in this case the continuous mappings), with the obvious
composition law. Of course, different categories with the same objects are
given different names.
1.2.2 Exploring connectedness

We begin by constructing some elementary functors Top → D, with values
in a category of algebraic structures. The important ones, the homol-
ogy and homotopy functors, will require more work, but will appear to
be ‘higher dimensional versions’ of some of the present ones, related to
properties of ‘higher connectedness’.
As in 1.1.4, a functor F : Top → D is homotopy invariant if, for every
pair of homotopic maps f ' g, we have F (f ) = F (g).
(i) A set has a trivial algebraic structure, with no operations under no
axioms, and all mappings preserve this structure: Set is itself a category of
algebraic structures, the simplest.
There is an obvious forgetful functor
U : Top → Set, U (X) = |X|, U (f ) = |f |, (1.13)
that sends each space X to its underlying set |X|, and each map f : X → Y
to the underlying mapping |f | : |X| → |Y |.
Each category of structured sets has a similar forgetful functor with values in
Set (see 1.3.7). Such functors are so obvious that we tend to forget them! The
notation | − | will be used when useful.
(ii) We also consider two less obvious functors with values in Set
Cc : Top → Set, Π0 : Top → Set. (1.14)
The former sends a space X to the set Cc(X) of its connected compo-
nents, and a map f : X → Y to the mapping Cc(f ) : Cc(X) → Cc(Y ) which
sends a component A of X to the component of Y containing f (A).
Π0 is constructed in the same way, using the path components; it will be
reviewed in (1.69).

(iii) (A contravariant functor) Here Z is the ring of integers equipped with
the discrete topology. We form a contravariant functor (denoted by a
dashed arrow)
A : Top 99K Rng, A(X) = Top(X, Z). (1.15)
It sends a space X to the (commutative) ring of all maps a : X → Z,
endowed with the pointwise operations:
(a + b)(x) = a(x) + b(x), (ab)(x) = a(x).b(x), 1(x) = 1Z . (1.16)
A map f : X → Y is sent to a homomorphism of rings
A(f ) : A(Y ) → A(X), A(f )(a : Y → Z) = (af : X → Z). (1.17)
The reader will note that A reverses the direction of the map f ; consis-
tently, this transformation preserves the identities but reverses the compo-
sition: a map g : Y → T will give a homomorphism A(g) : A(T ) → A(Y ),
and
A(gf ) = A(f )A(g) : Top(T, Z) −→ Top(X, Z),
(1.18)
(a : T → Z) 7→ (agf : X → Z).
Exercises and complements. (a) The forgetful functor U : Top → Set is not
homotopy invariant.
(b) The functors Cc : Top → Set, Π0 : Top → Set and A : Top 99K Ab
defined above are homotopy invariant.
*(c) More generally, in the construction of A one can replace Z with any
ring, equipped with the discrete topology.
1.2.3 Other exercises

The reader may find it interesting to explore the contravariant functor A
(in (1.15)), either independently or with the following exercises.
(a) A function a : X → Z is continuous if and only if each preimage Xk =
a−1 {k} (for k ∈ Z) is open in X, if and only the function a is locally
constant (i.e. constant on some neighbourhood of each point).
This implies that a is constant on each connected component of X; the
converse certainly holds if the connected components of X are open, but
plainly fails for the rational line Q whose connected components are sin-
gletons.
S
(b) A map a : X → Z amounts to a partition X = k∈Z Xk in disjoint open
subsets (automatically closed in X). But this description is ineffective with
respect to the operations of the ring A(X), defined in (1.16).

(c) The space X is empty if and only if the ring A(X) is trivial.
(d) The following conditions on the space X are equivalent:
(i) X is non-empty and connected,
(ii) A(X) ∼= Z (as rings),
(iii) A(X) is an integral domain (i.e. a non-trivial ring without proper
zero divisors),
(iv) the underlying additive group of A(X) is infinite cyclic.
S
(e) Let X = i∈I Xi be the partition of X in connected components, and
suppose that each of them is open in X (which is certainly true if I is finite).
Then the ring A(X) is canonically isomorphic to the cartesian-power ring
ZI = i∈I Z. (Note that we can identify the index-set I with the set Cc(X)
Q
of connected components of X.)

*(f) For the euclidean space
X = {x ∈ R | x = 1/n, for some n ∈ N∗ } ∪ {0}, (1.19)
the ring A(X) is isomorphic to a countable subring of ZN .
*(g) The ring A(Q) of the rational line (with euclidean topology) is un-
countable. (In fact, it has the cardinal of the continuum, i.e. the cardinal
of R.)
1.2.4 Distinguishing shape and homotopy type

In a category C, an isomorphism is a morphism f : X → Y which has an
inverse, i.e. a morphism g : Y → X such that gf = 1X and f g = 1Y . The
inverse is determined by f (by the usual verification) and written as f −1 .
We say that the objects X, Y are isomorphic, written as X ∼ = Y , if there
is an isomorphism X → Y (in the category that we are considering).
In the categories listed in 1.2.1 this definition gives the usual isomor-
phisms of each structure: bijective mappings in Set, homeomorphisms in
Top, isomorphisms of groups in Gp and Ab, etc.
Plainly:
(i) a functor F : C → D takes any isomorphism of C to an isomorphism
of D,
(ii) a homotopy invariant functor F : Top → D transforms a homotopy
equivalence f : X → Y into an isomorphism of the category D; in other
words, X ' Y implies F (X) ∼ = F (Y ).
Therefore, a functor F : Top → D can be used to distinguish spaces, up to
homeomorphism: two spaces X, Y cannot be homeomorphic if F (X) and

F (Y ) are not isomorphic in D. A homotopy invariant functor can be used
to distinguish homotopy types.
Important results in this sense will come out of complex functors, like
singular homology. The elementary examples of 1.2.2 give trivial or ele-
mentary results.
Exercises and complements. (a) Applying principle (i) to the functors U ,
Cc, Π0 of 1.2.2 gives obvious facts, for which there would be no need of
categories and functors: two homeomorphic spaces have equipotent under-
lying sets, the same number (or cardinal) of connected components, and
the same number (or cardinal) of path components. We shall see that Π0
is closely related to singular homology in degree zero.
(b) Applying principle (ii) to the homotopy invariant functors Cc and Π0
gives elementary, less obvious facts: two homotopy equivalent spaces have
the same number (or cardinal) of connected components, and the same
number (or cardinal) of path components.
*(c) The functor A is able to distinguish the euclidean space
X = {x ∈ R | x = 1/n, for some n ∈ N∗ } ∪ {0}, (1.20)
of Exercise 1.2.3(f) from any infinite discrete space (while the functors U ,
Cc and Π0 give isomorphic results on X and the discrete N).
1.2.5 Free abelian groups

We recall here some points about abelian groups, that will be frequently
used throughout this book. (Everything can be extended to modules on a
commutative ring R, that replaces the ring Z of integers: see 4.1.1.)
P
(a) In an abelian group A, a finite linear combination i∈I λi xi of elements
xi ∈ A with integral coefficients λi makes sense, without any ordering of
the set I of indices, because of the associative and commutative property
of addition.
P
More generally, a linear combination i∈I λi xi makes sense for an arbi-
trary set of indices I, if the family of coefficients (λi )i∈I ∈ ZI is essentially
finite: there are finitely many indices i such that λi 6= 0; then the sum is
computed on these indices.
(b) A basis of an abelian group A, is a family (xi )i∈I of elements of A such
that every x ∈ A can be uniquely written as a linear combination with
essentially finite integral coefficients
P
x= i∈I λ i xi . (1.21)

The abelian group A is free (on the ring Z) if it has a basis. The latter
is an injective family, and we can embed I in A, identifying i = xi .
In fact, if i 6= j in I, the difference xi − xj = 1.xi + (−1).xj cannot be
0A = 0.xi + 0.xj .
The universal property of the basis says that every mapping f : I → B
with values in an abelian group can be uniquely extended to a group-
homomorphism g : A → B
P P
g( i∈I λi xi ) = i∈I λi f (i), (1.22)
computed by an (essentially finite) linear combination in B.

(c) Every set I is the basis of a free abelian group. We construct the latter
as the direct sum
Z ⊂ ZI ,
L
ZI = i∈I (1.23)
namely the subgroup of ZI formed by the essentially finite families (λi )i∈I
of integers.
In fact ZI has a canonical basis ei (i ∈ I), given by the Kronecker families
eii = 1, eij = 0 (j 6= i), (1.24)
since the general element (λi )i∈I ∈ ZI coincides with the essentially finite
P
linear combination i∈I λi ei .
Identifying any i ∈ I with the element ei of the canonical basis, the
P
elements of ZI can be written as formal linear combinations i∈I λi i of the
elements of I, with essentially finite integral coefficients.
(d) Finally we have the free abelian group functor
L
F : Set → Ab, F (I) = ZI = i∈I Z,
(1.25)
F (f ) : ZI → ZJ, F (f )(i) = f (i) (for f : I → J),
where the homomorphism F (f ) is defined on the canonical basis I; its value

on a general essentially finite family of ZI is:
P P
F (f )(λi )i∈I = i λi f (i) = ( f (i)=j λi )j∈J . (1.26)
The functorial properties of F are readily verified on the bases: for in-
stance, for a composed mapping gf : I → J → K
F (gf )(i) = (gf )(i) = g(f (i)) = F (g)F (f )(i).
(e) We have studied in 1.2.2 and 1.2.3 the contravariant functor A =

Top(−, Z) : Top 99K Rng.

Forgetting topology, we have a contravariant functor P = Set(−, Z) de-
fined on sets:
P (I) = ZI =
Q
P = Set(−, Z) : Set 99K Rng, i∈I Z,
(1.27)
P (f ) : P (J) → P (I), P (f )(λj )j∈J = (λf (i) )i∈I ,
where f : I → J is a mapping. The functor P is the same as the composite
of the discrete-topology functor D : Set → Top with A : Top 99K Rng.
1.2.6 A structure theorem

We shall also use some other classical results on abelian groups. This part
can be skipped, until referred to. For integers m, n the notation m | n
means that m divides n, that is n = km for some k ∈ Z.
(a) In an abelian group A, the elements of finite order form a subgroup
tA = {x ∈ A | ∃ n ∈ N∗ such that nx = 0}, (1.28)
called the torsion subgroup of A. We say that A is a torsion group if A = tA;

we say that it is torsion free if tA = 0.
The quotient A/tA is obviously torsion free.
(b) An abelian group A is finitely generated if there is a finite family x1 , ...,
xn of elements that generates it: every x ∈ A can be written as a linear
P
combination i∈I λi xi of the generators, with integral coefficients.
We recall a classical theorem, that determines the structure of such
groups. The proof can be found in any book on abelian groups (for in-
stance [Fu]). Its extension to modules on principal ideal domains can be
found in [Bou3] Section VII.4.4.
(c) (The Structure Theorem of Finitely Generated Abelian Groups) A
finitely generated abelian group A is (non-canonically) isomorphic to a di-
rect sum
A ∼
= Zn ⊕ Zτ1 ⊕ ... ⊕ Zτk , τ1 | τ2 | ... | τk−1 | τk , (1.29)
that determines the following natural numbers:

- the rank n = rkA of A,
- the torsion coefficients τ1 , ..., τk ∈ N∗ , ordered by divisibility (k > 0).
The sequence (n, τ1 , ..., τk ) classifies A, up to isomorphism.
(d) For a finitely generated abelian group A, the decomposition (1.28) can
plainly be restricted to an isomorphism of the torsion subgroups
tA ∼
= Zτ1 ⊕ ... ⊕ Zτk , τ1 | τ2 | ... | τk−1 | τk . (1.30)

(e) A can thus be decomposed in the direct sum A ∼ = Zn ⊕ tA. But note
that there is no canonical subgroup of A isomorphic to Zn .
(f) A finitely generated abelian group A which is torsion free is necessarily a
free abelian group. All its bases have the same number of elements, namely
n = rkA.
(g) The additive group Q of rational numbers is not finitely generated. It is
torsion free, and is not free: no element is a generator, but any two distinct
elements are linearly dependent on Z.
(h) The cyclic group Zk has one torsion coefficient, namely k. The Klein
four-element group Z2 ⊕ Z2 has torsion coefficients (2, 2), which distin-
guishes it from the cyclic group Z4 . The direct sum Z2 ⊕ Z3 has torsion
coefficient 6.
1.3 The terminology of categories and functors

Categories and functors were introduced by Eilenberg and Mac Lane [EM1]
in 1945. We review here these issues, after the hints given in the previous
section.
A difficulty of category theory, for a beginner, is the multi-layered struc-
ture of this discipline: three layers for the theory itself (categories, functors
and their transformations), and two additional layers inside each category
(objects and arrows). Starting from examples (as in the previous section)
and working on exercises is the best way to master this feature.
Dealing with categories, one has to avoid the usual contradictions related
to ‘the set of all sets’. Here we make use of a particular set theory where
there are sets and classes; every set is also a class, while a proper class is
not a set: for instance the class of all sets and the class of all groups are
proper (see [AHS] or the Appendix of [Ke]). This approach is followed in
[Mi, Fr, AHS].
*When two levels, like sets and classes, are insufficient, one can introduce a
third level of ‘hyperclasses’. Alternatively, a more flexible setting widely used
for categories is ordinary Set Theory (say ZFC) together with the assumption of
the existence of a Grothendieck universe, or of a suitable hierarchy of universes
(cf. [M2], Section I.6).*
1.3.1 Definition
A category C consists of the following data:
(a) a class Ob C, whose elements are called objects of C,
(b) for every pair X, Y of objects, a set C(X, Y ) (called a hom-set) whose

elements are called morphisms (or arrows) of C from X to Y and denoted
as f : X → Y ,
(c) for every triple X, Y, Z of objects of C, a mapping of composition
C(X, Y ) × C(Y, Z) → C(X, Z), (f, g) 7→ gf,
where gf is also written as g.f .

These data must satisfy the following axioms.
(i) (Associativity) Given three consecutive arrows f : X → Y , g : Y → Z
and h : Z → W , one has: h(gf ) = (hg)f .
(ii) (Identities) Every object X has an endomorphism e : X → X which
acts as an identity whenever composition makes sense: if f : X 0 → X and
g : X → X 00 , then ef = f and ge = g.
One shows, in the usual way, that e is determined by X; it is called the
identity of X and written as id X, or 1X .
We generally assume that the following condition is also satisfied:
(iii) (Separation) For X, X 0 , Y, Y 0 objects of C, the sets C(X, Y ) and
C(X 0 , Y 0 ) are disjoint, unless X = X 0 and Y = Y 0 .
Therefore a morphism f : X → Y has a well-determined domain and
codomain
Dom (f ) = X, Cod (f ) = Y.
Concretely, when constructing a category, one can forget about condition

(iii), since one can always satisfy it by redefining a morphism fˆ: X → Y as
a triple (X, Y, f ) where f is a morphism from X to Y in the original sense
(possibly not satisfying the Separation axiom).
Mor C denotes the class of all the morphisms of C, i.e. the disjoint union
of its hom-sets. Two morphisms f, g of C are said to be parallel when they
have the same domain and the same codomain. For every object X, the set
C(X) = C(X, X) of endomorphisms X → X forms a monoid (i.e. a unital
semigroup).
If C is a category, the opposite (or dual) category, written as Cop , has the
same objects as C, reversed arrows and reversed composition
Cop (X, Y ) = C(Y, X), g ∗ f = f g, id op (X) = id X. (1.31)
Every topic of category theory has a dual instance, which comes from the
opposite category (or categories). A dual issue is generally distinguished
by the prefix ‘co-’.
The term ‘map’ is generally used as equivalent to morphism; here it will
always mean ‘continuous mapping’ between topological spaces.

Complements. A category is discrete if the only arrows are its identities.
A category is indiscrete if for any two objects X, Y there is precisely one
arrow X → Y .
A category is connected if it is not empty and any two objects X, Y are
linked by some ‘zig-zag sequence’ of arrows
X → X1 ← X2 → ... ← Xn−1 → Y. (1.32)
We follow here the usual convention, excluding the empty category from the
connected ones. On the other hand, for connected and path-connected spaces,
we are not excluding the empty space, although this might be preferable in the
domain of Algebraic Topology (see 1.2.3(d), 1.4.6(i), etc.).
1.3.2 Small and large categories

We have assumed that a category C has a class Ob C of objects (e.g. the
class of all sets, or the class of all topological spaces) and, for every pair
X, Y of objects, a set C(X, Y ) of morphisms from X to Y .
The categories of structured sets that we shall consider, like the examples
of 1.2.1, are generally large categories, where the objects (or equivalently
the morphisms) form a proper class (i.e. not a set). A category C is said
to be small if the class Ob C is a set. A finite category is a small category
whose set of morphisms is finite; then the same is true of its set of objects,
since an object is determined by its identity.
Examples and complements. (a) A set S will often be viewed as a discrete
small category: its objects are the elements of S, and the only arrows are
their (formal) identities; here S op = S.
(b) A preordered set X is equipped with a relation x ≺ x0 which is assumed
to be reflexive and transitive. It will be viewed as a small category, where
the objects are the elements of X and the set X(x, x0 ) contains a unique
arrow if x ≺ x0 (which can be written as (x, x0 ) : x → x0 ), and no arrow
otherwise.
Composition and identities are uniquely determined, as follows
(x0 , x00 ).(x, x0 ) = (x, x00 ), id x = (x, x). (1.33)
In this sense, categories generalise preordered sets. Their dualities agree:
the category associated to the opposite preordered set X op (with reversed
preordering) is dual to the category associated to X. Loosely speaking, the
extension from preordered sets (or classes) to categories consists in allowing
different arrows between two given objects.
In particular, each finite ordinal defines a category, which is often written
as 0, 1, 2... Thus, 0 is the empty category; 1 is the singleton category,

i.e. the discrete category on one object; 2 is the arrow category, with two
objects (0 and 1), and one non-identity arrow, 0 → 1.
A (partial) order relation is a preordering which is anti-symmetric: x ≺ x0
and x0 ≺ x0 implies x = x0 .
(c) A monoid M can be viewed as a small category with one object, say ∗,
and one set of morphisms M , from ∗ to ∗; composition is the multiplication
of M . In this sense, categories generalise monoids. Again, their dualities
coincide: the opposite monoid M op corresponds to the opposite category.
Loosely speaking, a category can be viewed as a ‘many-object’ extension
of a monoid.
(d) (Commutative diagrams) Objects and morphisms of a category are often
represented by vertices and arrows in a diagram, as in the examples below,
to make evident their relationship and which compositions are legitimate
X
f
/ Y A
f
/ B
g h d k X o
u / Y (1.34)
h
" " v
Z C / D
g
We say that such a diagram is commutative if:

(i) whenever we have two ‘paths’ of consecutive arrows, from a certain
object to another, the two composed morphisms are the same,
(ii) whenever we have a ‘loop’ of consecutive arrows, from an object to
itself, the composed morphism is the identity of that object.
Thus, the first diagram above is commutative if and only if gf = h. For
the second, commutativity means that kf = d = gh. For the third, it
means that vu = id X and uv = id Y (so that these morphisms are inverse
to each other).
(e) A small category is (associated to) a preordered set (as in point (b)) if
and only if all diagrams in such a category commute.
1.3.3 Isomorphisms and retracts

As already said in 1.2.4, a morphism f : X → Y in a category C is said
to be invertible, or an isomorphism, if it has an inverse, i.e. a morphism
g : Y → X such that gf = 1X and f g = 1Y . The latter is determined
by f ; it is called the inverse of f and written as f −1 . (Isomorphisms of
topological spaces are called homeomorphisms.)
The isomorphism relation X ∼ = Y between objects of C (meaning that
there is an isomorphism X → Y ) is an equivalence relation: an obvious con-
sequence of identity arrows, inverting isomorphisms and composing them.

A groupoid is a category where every map is invertible. Every discrete
or indiscrete category is a groupoid.
The fundamental groupoid of a space X contains all the fundamental groups
π1 (X, x), for x ∈ X; it will be constructed in Section 6.1.
One-side inverses are also important. If, in a category C, we have two
morphisms m and p such that
m : A X : p, pm = id A, (1.35)
then m is called a section (of p), p is called a retraction (of m), and A is
called a retract of X.
In the usual categories of structured sets and structure-preserving map-
pings, m is injective and p is surjective: we can replace A with its image
m(A) ⊂ X, endowed with the restricted structure. (This is common prac-
tice, as in 1.1.3 and in the examples below.)
Exercises and complements. Retracts are easily characterised, below, in
Set and Ab. Examples in Top have been examined in 1.1.3 and 1.1.6; later,
this issue will be studied with the help of homology functors.
(a) In Set a retract of a set X 6= ∅ is any non-empty subset. The empty
set is only a retract of itself. In Set• a retract of an object (X, x0 ) is any
pointed subset (A, x0 ).
(b) In Ab, retracts coincide with ‘direct summands’. More precisely, a
subgroup A of an abelian group X is a retract if and only if there exists
a subgroup B ⊂ X such that A ∩ B = {0} and A + B = X. Then X is
said to be the internal direct sum of A, B and is canonically isomorphic to
A ⊕ B.
Note that Z is isomorphic to 2Z ⊕ 0, but 2Z is not a direct summand of Z:
the previous directed sum is not an internal one.
(c) A cyclic group X of infinite or prime order has no other retracts than
the trivial, unavoidable ones: {0} and X.
(d) In Top, suppose that A is a retract of X, as defined in (1.35). Prove that
the map m is a topological embedding, i.e. an injective map m : A → X
that makes A homeomorphic to the subspace m(A) ⊂ X (see 1.1.7(a)).
Identifying A with this subspace we recover the usual definition of re-
tracts in Topology: a subspace A ⊂ X that admits a retraction, i.e. a map
p : X → A such that p(x) = x for all x ∈ A.
(e) In the same hypothesis, prove that p is a topological projection, i.e.
a surjective map p : X → A that makes A homeomorphic to the quotient
X/Rp , modulo the equivalence relation associated to p (see 1.1.7(b)).
We can also view a retract of X as a quotient p : X → X/R that admits a
(continuous) section, i.e. a map m : X/R → X such that pm = id X/R.

1.3.4 Subcategories, quotients and products of categories
(a) Let C be a category. A subcategory D is defined by assigning:
- a subclass Ob D ⊂ Ob C, whose elements are called objects of D,
- for every pair of objects X, Y of D, a subset D(X, Y ) ⊂ C(X, Y ), whose
elements are called morphisms of D, from X to Y ,
so that the following conditions hold:
(i) the composite in C of every pair of consecutive morphisms of D belongs
to D,
(ii) the identity in C of every object of D belongs to D.
Then D, equipped with the restricted composition law, is a category.
We say that D is a full subcategory of C if, for every pair of objects X, Y
of D, we have D(X, Y ) = C(X, Y ), so that D is determined by assigning a
subclass of objects. We say that D is a wide subcategory of C if it has the
same objects.
For instance, Ab is a full subcategory of Gp. The open continuous mappings
form a wide subcategory of Top.
(b) A congruence R = (RXY ) in a category C consists of a family of equiva-
lence relations RXY (also written as R) in each set of morphisms C(X, Y ),
that is consistent with composition:
(iii) if f RXY f 0 and gRY Z g 0 , then gf RXZ g 0 f 0 .
Equivalently (by transitivity) it is sufficient to say that:
(iii0 ) if f Rf 0 then f h Rf 0 h and kf R kf 0 (for all legitimate compositions).
Then one defines the quotient category D = C/R: the objects are those of
C, and D(X, Y ) = C(X, Y )/RXY ; in other words, a morphism [f ] : X → Y
in D is an equivalence class of morphisms X → Y in C. The composition
is induced by that of C, which is legitimate because of condition (iii):
[g][f ] = [gf ]. (1.36)
For instance, in Top the homotopy relation f ' f 0 is a congruence of

categories: see 1.4.4.
The relation of isomorphism is wider in a quotient category (but it may
coincide with the original one, also in a non-trivial quotient).
(c) If C and D are categories, one defines the product category C × D. An
object is a pair (X, Y ) where X is in C and Y in D; a morphism is a pair
of morphisms
(f, g) : (X, Y ) → (X 0 , Y 0 ), f ∈ C(X, X 0 ), g ∈ D(Y, Y 0 ). (1.37)


The composition of (f, g) with a consecutive morphism (f 0 , g 0 ) is (obvi-
ously) defined componentwise: (f 0 , g 0 )(f, g) = (f 0 f, g 0 g).
Q
More generally one defines the cartesian product C = i∈I Ci of a family
of categories (Ci )i∈I indexed by a set I: an object of C is a family (Ai )i∈I
where Ai ∈ Ob (Ci ) (for every i), and a morphism f = (fi ) : (Ai ) → (Bi ) is
a family of morphisms fi ∈ Ci (Ai , Bi ); the composition is componentwise
and id ((Ai )i∈I ) = (id Ai )i∈I . A product of groupoids is a groupoid.
Cartesian powers, like C2 = C × C and CI = i∈I C, will also be used.
Q
1.3.5 Functors
A (covariant) functor F : C → D consists of the following data:
(a) a mapping F0 : Ob C → Ob D, whose action is generally written as
X 7→ F (X),
(b) for every pair of objects X, X 0 in C, a mapping
FXX 0 : C(X, X 0 ) −→ D(F (X), F (X 0 )),
whose action is generally written as f 7→ F (f ).

Composition and identities must be preserved:
(i) if f, g are consecutive maps in C, then F (gf ) = F (g).F (f ),
(ii) if X is in C, then F (id X) = id (F (X)).
For a second functor G : D → E, one defines in the obvious way the
composed functor GF : C → E. The composition of consecutive functors is
associative and has identities: the identity functor of each category
id C : C → C, X 7→ X, f 7→ f.
A functor preserves commutative diagrams, isomorphisms, retracts, sec-

tions and retractions.
A contravariant functor F : C 99K D can be defined as a covariant functor
C → D, or equivalently C → Dop . Composing two contravariant functors
op
one gets a covariant one.

A functor in two variables is an ordinary functor F : C × D → E defined
on the product of two categories. Fixing an object X0 in C, we have a
functor F (X0 , −) : D → E; and symmetrically.
Cat (resp. Gpd) will denote the category of small categories (resp. small
groupoids) and their functors. (These categories also have a 2-dimensional
structure, with higher morphisms ϕ : F → G between functors: see Sec-
tion 1.5.)

The action of a functor F on objects and morphisms will often be written
without parentheses: F X and F f for F (X) and F (f ).
A functor (resp. a contravariant functor) between two preordered sets,
viewed as categories, is a monotone (resp. antitone) mapping. A functor
between two monoids, viewed as categories, is a homomorphism of monoids.
1.3.6 Isomorphic categories

An isomorphism of categories is a functor F : C → D which is invertible:
there is a functor G : D → C such that GF = id C and F G = id D. The
functor G is determined by F ; it is called the inverse functor, and written
as F −1 .
An easy verification proves that the functor F is an isomorphism if and
only if all the mappings F0 and FXX 0 considered above are bijective.
Being isomorphic categories is plainly an equivalence relation, written as
C ∼= D.
Categories linked by an obvious isomorphism are often viewed as ‘the
same thing’. For instance, everybody knows that an abelian group has a
unique structure of module on the ring Z of integers (where 2x = x + x,
and so on), preserved by every homomorphism of abelian groups. This
fact readily shows that the category Ab is canonically isomorphic to the
category Z Mod of modules on the ring of integers; one generally makes no
distinction between these categories.
A topological space can be defined in many equivalent ways (based on
open sets, closed sets, neighbourhoods, a closure operator, etc.); from a
formal point of view, this yields many isomorphic categories which is rarely
convenient to distinguish.
1.3.7 Forgetful and structural functors

(a) Forgetting the structure, or part of it, yields various examples of func-
tors between categories of structured sets, like the following obvious in-
stances
Top → Set, Top• → Top, Top• → Set• ,
(1.38)
Set• → Set, Rng → Ab, Ab → Set.
The first, for instance, forgets the topology, taking a topological space
X to its underlying set |X|, and a continuous mapping X → Y to the
underlying mapping |X| → |Y | of sets. The second forgets the basepoint;
and so on.

These forgetful functors are often denoted by the letter U , which refers
to the underlying set, or underlying space, or underlying abelian group, etc.
(b) A subcategory D of C yields an inclusion functor D → C, which we also
write as D ⊂ C. For instance, Ab ⊂ Gp. These functors forget properties,
of objects or morphisms, rather than structure.
(c) A congruence R in a category C yields an obvious projection functor
P : C → C/R, which is the identity on objects and sends a morphism f to
its equivalence class [f ]. For instance, we are interested in the projection
functor Top → hoTop = Top/' (see (1.62)).
Q
(d) A product category C = i∈I Ci has a family of projection functors
Pi : C → Ci .
1.3.8 Hom functors

(a) Every category C has a functor of morphisms, or hom-functor:
MorC : Cop × C → Set, (X, Y ) 7→ C(X, Y ), (h, k) 7→ k. − .h, (1.39)
where, if h : X 0 → X and k : Y → Y 0 are in C, the mapping k. − .h
k. − .h : C(X, Y ) → C(X 0 , Y 0 ), f 7→ kf h, (1.40)
means pre-composing with h and post-composing with k.

The hom-functor (1.39) is viewed as a functor on two variables of C,
‘contravariant on the first and covariant on the second’. Note that it takes
values in Set because of our assumption, in 1.3.1(b), that all classes C(X, Y )
be sets.
(b) Fixing the first variable we get a covariant functor on C
C(X0 , −) : C → Set,
(1.41)
k 7→ k ] = k.− : C(X0 , Y ) → C(X0 , Y 0 ) (k : Y → Y 0 in C),
that is said to be representable, and represented by the object X0 . This
notion will be extended to any functor C → Set isomorphic to C(X0 , −), as
defined in 1.5.2.
Dually, fixing the second variable we get a functor
C(−, Y0 ) : Cop → Set,
(1.42)
h 7→ h] = −.h : C(X, Y0 ) → C(X 0 , Y0 ) (h : X 0 → X in C),
that is contravariant on C, and representable as such. Again, we say the
same of a functor Cop → Set isomorphic to the former. Contravariant
functors are often constructed in this way.

(c) In the category Ab of abelian groups, the set of homomorphisms A →
B has a canonical structure of abelian group, by the pointwise addition
(f + g)(x) = f (x) + g(x). This abelian group is denoted as Hom(A, B).
The category Ab has thus an enriched version of the hom-functor, with
values in the category itself (instead of Set)
Hom : Abop × Ab → Ab,
(1.43)
(A, B) 7→ Hom(A, B), (h, k) 7→ k. − .h.
In fact one can easily verify that the mapping
k. − .h : Hom(A, B) → Hom(A0 , B 0 )
preserves the addition of homomorphisms: k(f + g)h = kf h + kgh.

(d) We have already used representable contravariant functors in the pre-
vious section, in an enriched form (with values in rings): the functors
P = Set(−, Z) in (1.27), and A = Top(−, Z) in (1.15). This comes out of
the fact that Z is a ring, and a topological ring (with the discrete topology).
1.4 Paths and homotopy

We review here the elementary issues of homotopy theory, also to fix the
terminology we are using.
Typically, the ‘invariants’ of Algebraic Topology are invariant up to ho-
motopy — a stronger property than invariance up to homeomorphism; a
homology functor Hk extracts from the complexity of a space, or a map, a
limited information about k-dimensional connectedness, which we can hope
to compute; choosing k, or collecting all the information provided by these
functors (or others), we may be able to solve the problem we are studying.
The two-valued index α (or β) varies in the set {0, 1}, also written as
{−, +}.
1.4.1 Euclidean spaces

We use the following notation, for spaces which play an important role in
Topology and Algebraic Topology.
The space Rn of n-tuples of real numbers x = (x1 , ..., xn ) is equipped
with the euclidean topology, determined by the euclidean norm and the
euclidean distance
||x|| = (x21 + ... + x2n )1/2 , d(x, y) = ||x − y||, (1.44)
or equivalently as a cartesian power of the euclidean line R.


Identifying Rn with the hyperplane Rn × {0} ⊂ Rn+1 , these spaces form
a nested family, and the singleton R0 = {0} is included in all of them
{0} = R0 ⊂ R1 ⊂ R2 ⊂ ... Rn ⊂ ... (1.45)
A euclidean space is any subspace of some Rn , and an open euclidean
space is any open subspace of some Rn .
I denotes the standard compact interval [0, 1] of R, on which paths and
homotopies are parametrised. Its cartesian power In is the standard n-cube,
a subspace of Rn . We are also interested in the standard n-disc (or closed
ball)
Dn = {x ∈ Rn | ||x|| 6 1}. (1.46)
Inside Rn+1 , we also have the standard n-sphere
Sn = {x ∈ Rn+1 | ||x|| = 1} (n > 0), (1.47)
with basepoint at e1 = (1, 0, ..., 0), when useful.
(We can specify en1 ∈ Sn ⊂ Rn+1 , but in fact all these points are identified
with 1 ∈ R, in the nesting (1.45); and also with the unit of C.)
In particular, S0 = {−1, 1} ⊂ R is a discrete space on two points, pointed
at 1. Adding a lower sphere S−1 = ∅ as the sphere of R0 = {0}, the discs
and the spheres form a nested family of subspaces of the family (1.45)
{0} = D0 ⊂ D1 ⊂ D2 ⊂ ... Dn ⊂ ...
(1.48)
∅ = S−1 ⊂ S0 ⊂ S1 ⊂ ... Sn−1 ⊂ ....
where Sn−1 = ∂Dn is the border of Dn in Rn , for all n > 0. (The term
‘boundary’ will be used in other meanings.)
The convenience of adding S−1 will also appear elsewhere. Its topological
dimension should be set at −∞, to agree with the formula dim(X × Y ) =
dimX + dimY .
The standard circle S1 can also be viewed as a subspace of the complex
plane, which topologically is the same as R2
S1 = {z ∈ C | |z| = 1}. (1.49)
The ordinary parametrisation of the circle on the standard interval
p : I → S1 , p(t) = (cos 2πt, sin 2πt) = e2πit , (1.50)
called the exponential map, covers the circle counterclockwise (in the ori-
ented plane), identifying the two points of the border ∂I = {0, 1} in R.
Writing I/∂I the quotient space, the map p induces a bijective continuous
mapping
p0 : I/∂I → S1 , p0 [t] = (cos 2πt, sin 2πt) = e2πit , (1.51)

which is a homeomorphism (as any bijective map from a compact to a
Hausdorff space).
1.4.2 Paths and loops

A path in the space X is a map a : I → X defined on the standard euclidean
interval I = [0, 1].
The basic structure of the interval I consists of four maps, linking the
interval to the singleton I0 = {∗}, its 0-th cartesian power:
∂ α : {∗} I, ∂ − (∗) = 0, ∂ + (∗) = 1 (f aces),

e : I → {∗}, e(t) = ∗ (degeneracy), (1.52)
r : I → I, r(t) = 1 − t (reversion).
Identifying a point x ∈ X with the corresponding map x : {∗} → X, this
basic structure determines:
(i) the endpoints of a path a : I → X, namely the points a∂ − = a(0) and
a∂ + = a(1),
(ii) the trivial path, or constant path, at the point x, written as ex = xe,
(iii) the reversed path of a, written as a] = ar.
We write as P (X) the set of paths of the space X, and
P (X, x0 , x1 ) = {a ∈ P (X) | a(0) = x0 , a(1) = x1 }, (1.53)
will be the set of paths of X, from x0 to x1 . (These sets are canonically

equipped with the compact-open topology, reviewed in Section 7.2.)
Two paths a, b : I → X are consecutive if a(1) = b(0), that is a∂ + = b∂ − ;
then they have a concatenated path a ∗ b : I → X (see Exercise (a))

a(2t), for 0 6 t 6 1/2,
(a ∗ b)(t) = (1.54)
b(2t − 1), for 1/2 6 t 6 1.
This partial operation is not associative, generally; we shall see that it

works well up to homotopy with fixed endpoints.
A loop at x0 ∈ X is a path from x0 to x0 , called the basepoint of the
loop. The set of loops of X at x0
Ω(X, x0 ) = P (X, x0 , x0 ) = {a ∈ Top(I, X) | a(0) = x0 = a(1)}, (1.55)
inherits a binary operation of concatenation, which again is not associative

(except trivial cases).

Exercises and complements. (a) Verify that the mapping a ∗ b : I → X is
well defined in (1.54), and continuous.
Although this fact is quite elementary, the Finite Closed Cover Lemma, in
1.1.8, is the standard tool for concatenations of paths, homotopies, etc.
(b) For two points x, x0 ∈ X, we write x ' x0 if there is a path in X from x
to x0 . It is an equivalence relation, because of trivial paths, reversed paths
and path-concatenation.
(c) The path component [x]P of the point x ∈ X is its equivalence class for
this relation. A space X is path connected if any two points are linked by
a path; this implies that X is connected.
The empty space has no path component, and no connected component.
(d) A loop a : I → X amounts to map â : S1 → X, where â(e2πit ) = a(t).
We say that a is a simple loop if â is injective; then, if X is Hausdorff, â
is a topological embedding (see 1.1.7) and gives a homeomorphism of the
circle onto its image, a subspace of X.
A loop at x0 amounts to a pointed map â : (S1 , e1 ) → (X, x0 ).
1.4.3 Cylinder and homotopies

The cylinder IX = X × I on the space X has the product topology.
Given two maps f, g : X → Y (in Top), a homotopy ϕ : f ' g ‘is’ a map
ϕ : X × I → Y defined on the cylinder IX, that coincides with f on the
lower basis and with g on the upper one
ϕ : X × I → Y,
(1.56)
ϕ(x, 0) = f (x), ϕ(x, 1) = g(x) (for x ∈ X),
forming a continuous deformation of f into g.
The homotopy will also be written as ϕ : f ' g : X → Y . The map (1.56)
will be written as ϕ̂ when we want to distinguish it from the homotopy
which it represents.
All this is based on the basic structure of the cylinder IX = X × I,
inherited from the structural maps (1.52) of the standard interval, and
written as the latter
∂ α : X → IX, ∂ − (x) = (x, 0), ∂ + (x) = (x, 1) (f aces),
e : IX → X, e(x, t) = x (degeneracy), (1.57)
r : IX → I), r(x, t) = (x, 1 − t) (reversion).
This structure gives:
(i) the faces ϕ∂ α : X → Y of the homotopy, namely the maps ϕ∂ − = f
and ϕ∂ + = g,

(ii) the trivial homotopy ef = f e of a map f ,
(iii) the reversed homotopy ϕ] = ϕr.
More precisely, the homotopy ϕ] : g ' f is represented by the map
ϕ̂r : IX → Y .
Identifying I{∗} = I, a path a : I → X is the same as a homotopy
a : x ' x0 : {∗} → X between its endpoints (consistently with the notation
x ' x0 we are using). Moreover the structural maps of the standard interval
coincide with the previous maps (1.57), when X is the singleton.
Two consecutive homotopies ϕ : f ' g and ψ : g ' h of maps X → Y
have a concatenation ϕ ∗ ψ : f ' h

ϕ(x, 2t), for (x, t) ∈ X × [0, 1/2],
(ϕ ∗ ψ)(x, t) = (1.58)
ψ(x, 2t − 1), for (x, t) ∈ X × [1/2, 1],
which extends the concatenation of paths (see Exercise (a)).

Homotopy equivalences and deformation retracts have already been dealt
with, at a basic level, in 1.1.3.
Exercises and complements. (a) Verify that the mapping ϕ ∗ ψ : X × I → Y
is well defined in (1.58), and continuous.
(b) Let X be a contractible space, i.e. homotopy equivalent to the singleton
(see 1.1.3). Then X is path-connected and non-empty, and any singleton
{x} of X is a deformation retract.
(c) The following conditions on a space Y are equivalent:
- Y is totally path disconnected, i.e. every path in Y is trivial,
- for every space X, every homotopy ϕ : f ' g : X → Y is trivial.
*(d) The interval I (and therefore the cylinder I(X)) has a higher structure
g − : I2 → I, g − (t, t0 ) = max(t, t0 ),
g + : I2 → I, g + (t, t0 ) = min(t, t0 ) (connections), (1.59)
2 2 0 0
s: I → I , s(t, t ) = (t , t) (transposition),
which governs double homotopies (i.e. homotopies of homotopies) through
the double cylinder I 2 (X) = X × I2 : see [G5], Subsection 5.2.2.
*The whole structure satisfies the axioms of a ‘diad’ — a cubical version of
a monad [G1]. Double homotopies will be frequently used in Chapter 6. The
formal structure of the cylinder is also investigated in [KP].*
1.4.4 The homotopy category

Two parallel maps f, g : X → Y in Top are said to be homotopic, written
as f ' g, if there exists a homotopy ϕ : f ' g. This is an equivalence

relation in Top(X, Y ), because of trivial homotopies, reversed homotopies
and homotopy concatenation.
Furthermore, homotopies have a whisker composition with maps
f
/
X0
h / X ↓ϕ / Y
k / Y0 (1.60)
g
kϕh : kf h → kgh : X 0 → Y 0 .
where kϕh is represented on the cylinder by the map:
k.ϕ̂.I(h) : IX 0 → Y 0 , (x0 , t) 7→ k(ϕ(h(x0 ), t), (1.61)
for (x0 , t) ∈ X 0 ×I. The homotopy relation is thus a congruence of categories

in Top.
The quotient category hoTop = Top/' is called the homotopy category of
topological spaces: its objects are the topological spaces, and a morphism
[f ] : X → Y is a homotopy class of continuous mappings X → Y . We have
a projection functor
P : Top → hoTop, P (X) = X, P (f ) = [f ]. (1.62)
The classical notation [X, Y ] stands for the set of these homotopy classes
[f ] : X → Y
[X, Y ] = hoTop(X, Y ) = Top(X, Y )/'. (1.63)
An isomorphism in hoTop is the same as a homotopy equivalence in Top,

namely a pair of maps f : X Y : g such that gf ' id X and f g ' id Y .
A functor F : Top → D with values in an arbitrary category is said to
be homotopy invariant if F (f ) = F (g) whenever f ' g in Top, so that F
induces a functor
F : hoTop → D, F (X) = F (X), F (f ) = F [f ], (1.64)
determined by the relation F = F P . Typically, this will be the case of

all homology and homotopy functors. For simplicity, we may use the same
symbol for F and F .
A homotopy invariant functor Top → D takes a homotopy equivalence
f : X → Y in Top to an isomorphism in D.
*The weak homotopy category HoTop is obtained by a more complex proce-
dure, formally inverting all the continuous mappings that — loosely speaking —
induce isomorphisms in all homotopy sets and groups [GaZ, Bo1]. It can also
be obtained as the full subcategory of hoTop whose objects are the spaces ho-
motopically equivalent to CW-complexes.*
Another random document with
no related content on Scribd:
Die Predigt Christi
Auch bei einem Bild, wie die Auferstehung von 1639 (Pinakothek)
beruht die Wirkung durchaus auf der im Grund doch virtuosen
Gegenüberstellung der flimmernden Engelwolke und der Finsternis
des Vordergrundes mit den taumelnden Wächtern; ein peinlicher
Anblick, wie Christus, eine wächserne Puppe, mühsam im
Sarkophag, dessen Deckel der Engel hochwälzte, sich erhebt. Mit
welch schlagender unerbittlicher Eindringlichkeit wurde das Wunder
solchen Geschehens bei Grünewald malerische Tat: unten die
niederstürzenden Wächter, in der Höhe schwebend Christus, breit
und mächtig in leuchtender Herrlichkeit, im triumphierend
flatternden Gewand, mit feierlich erhobenen Händen, die die
Wundmale zeigen. Ist das Licht für Rembrandt (in diesen Jahren) ein
zu vollkommener Virtuosität ausgebildetes Kunstmittel, so bei
Grünewald notwendiger sinnlicher Ausdruck metaphysischer
Verklärung.
In der Mitte des Lebens geschieht der entscheidende Durchstoß
zur Religiosität: In den Bildern der heiligen Familie, deren schönstes
die Casseler Holzhackerfamilie ist. Es heißt nur den Anteil der
holländischen Zeitstimmung erörtern, wenn man bemerkt, daß die
Szene jetzt in das gemütliche Stubeninnere verlegt wird. Religiös ist
hier, wie menschliches Gefühl in seiner innerlichsten Regung
bloßgelegt wird: wie die Mutter, selig das Haupt geneigt, das Kind
mit beiden Händen an sich drückt; wie alles übrige in Dämmerung
versinkt, um das sanfte Wunder nicht zu stören. Aus dem Geheimnis
der Mütterlichkeit fließt das Erlebnis der Religiosität. Dies ist das
Neue, bisher nicht Dagewesene. Immer ist in den mittelalterlichen
Marien ein Rest von Scheu, von Feierlichkeit, von Unnahbarkeit. Nur
Leonardos traumhaft sich neigenden Madonnen und der Maria im
Rosenhag des Isenheimer Altars scheint eine Ahnung von solch
strömender Hingabe innezuwohnen.
In den letzten fünfzehn Jahren seines Lebens hat Rembrandt
keine profane Historie mehr gemalt oder radiert. Mit zwei
Ausnahmen: das Civilisbild, das nicht aus eigener Initiative, sondern
auf Grund eines Auftrages der Stadt Amsterdam entstanden ist. Und
selbst in dieser weltlichen Darstellung tritt die politische Aktion hinter
der feierlichen liturgischen Stimmung gänzlich zurück. Und ähnliches
gilt von dem andern Gemälde „Jupiter und Merkur bei Philemon und
Baucis zu Gast“. Das Verhältnis hat sich umgekehrt: konnte man
bisweilen stofflich religiöse Darstellungen der Jugend für profane
Geschichten halten (den drohenden Simson beispielsweise), so wird
jetzt der weltliche Stoff von religiöser Stimmung erfüllt.
Weitergehend die Frage: welche religiösen Stoffe haben
Rembrandt in diesen fünfzehn Jahren angezogen? Wenn man die
Reihe dieser reichlich zwei Dutzend Stücke vorüberziehen läßt, sieht
man, daß es sich fast ausschließlich um Themen handelt, die zu
getragener feierlicher Stimmung neigen. Alles Handlungsmäßige, alle
Aktion wird gleichgiltig. Keine Auferstehung, keine Himmelfahrt,
keine Wunderwirkung. Wo sie, die Aktion, ausnahmsweise — bei
dem die Tafeln zerschmetternden Moses, beim Jakobskampf —
Bildinhalt ist, erscheint ihre Äußerung unwichtig gegenüber dem
gewaltigen Sprung innerer seelischer Kraft, die das Bild baut.
Was bleibt, ist das tiefst Menschliche, im läuternden Feuerofen
des Leidens und der Ergebung religiös und Gleichnis geworden. Es
ist nicht ohne Wichtigkeit, den spezifisch religiösen Antrieb zu
betonen; man erlebt es (bei den Zeichnungen etwa) vor jedem
einzelnen Blatt von neuem: wieviel an Innigkeit, Gelöstheit,
Blutwärme hat jede (stofflich) religiöse Handzeichnung den
Civilisentwürfen etwa gegenüber voraus, die doch an sich zu den
ausdrucksvollsten zeichnerischen Dokumenten dieser Jahre gehören.
Baldinucci ist es, glaube ich, der berichtet, Rembrandt habe der
Sekte der Menoniten zugehört. Anderen erscheint er im kalvinischen
Holland tief verwurzelt und wieder andere wollen in seinen
Spätwerken Strömungen einer katholischen Frömmigkeit sehen. Dies
ist alles irgendwie richtig und erschöpft doch das Wesentliche nicht.
Wir wissen aus gelegentlichen Überlieferungen, daß er in gleicher
Weise freundschaftlich mit Juden, Katholiken und Kalvinern
verkehrte. Dieses Werk, gläubig wurzelnd im biblischen Wort, ist
gleich weit entfernt von modernem pantheistischen Gottgefühl wie
von mittelalterlicher fußfällig demütiger Gottergebung. Anbetung ist
Andacht geworden. Ist nicht das Gefühl der Gottnähe, der
Brüderlichkeit mit Gott, und das schweigende starke Verstummen
vor seinem unergründlichen Geheimnis der eigentliche Antrieb dieser
Gebärde?
Denn nunmehr ist die feierliche, aus sanft verklärter Dunkelheit
tauchende Gebärde alles geworden. In ihr festigt und klärt sich zu
gleichnishafter Form: mitleidendes Gefühl, geballter Wille und die
geläuterte gelassene Erkenntnis des gotterschaffenen, im
Unendlichen, Unbegreiflichen schwingenden Alls.
S oll schließlich ein Wort den Sinn dieses Schaffens
zusammenfassen, wie er uns Heutigen sich darstellt, wie er hier
in andeutendem Umriß gezeigt wurde — schließlich vermag
Kunstbetrachtung nichts anderes als Einstellung zu vermitteln; nicht
Vollständigkeit ist das Ziel, sondern Vorbereitung — so ist es das
Wort Religiosität.
Unnötig, zu versichern, daß es hier nicht als banale, modische
Laune, nicht als unklar schwärmende Stimmung begriffen werden
darf, daß es, da ja nur die Dialekte andere sind, aber über die
Jahrhunderte hin der Urlaut des Gefühls unverändert schwingt, für
Rembrandt Wesensgleiches bedeutet, wie für den romanischen
Miniaturenmaler, die Mystiker, Grünewald: Inniges Leben in Gott.
Lobpreis der gottgeschaffenen Kreatur, sei sie Acker, Mensch oder
Tier.
Vielheitlichkeit äußerer Erscheinung wird in diesem Zeichen
geläutert und, ungeheuerlich gesteigert, in ihren unvergänglich
gleichnishaften Sinn gefaßt. Am Ende dieses Lebens wiegt es
gleichviel: ob Landschaft, ob Bildnis, ob religiöses Geschehen Thema
ist.
Worte und Zeichen versinken. Die Liebe, in Gott brennender
Dornbusch, flammt; verwirklicht in Fläche und Form durch die
Verklärung aller Erscheinung.
Phot. F. Hanfstaengl, München
Die Anbetung der Könige
Phot. F. Hanfstaengl, München
Die Rückkehr des verlorenen Sohnes
AUGUST L. MAYER
Matthias Grünewald
Mit 68 meist ganzseitigen Abbildungen, zum Teil nach
neuen Gesamt- und Detailaufnahmen
Geheftet 11 Mark
Pappband 13 Mark. Halbleinenband 15 Mark
„Das Buch ist in doppelter Hinsicht eine bemerkenswerte
Erscheinung: In klarer gedrängter Form bietet es die
wissenschaftlichen Ergebnisse der weitläufigen Werke von H. A.
Schmid und Hagen und ergänzt sie glücklich durch eigene
Schlüsse und Beobachtungen. Ferner aber gibt es mit seinen
durchweg musterhaften Abbildungen dem Leser beinahe das
gesamte malerische und zeichnerische Werk Grünewalds an die
Hand.“
R. Oldenbourg in Münchner Neueste Nachrichten.
„Sein Bestes gibt Mayer in der systematischen Analyse, die zu der

Kompositionsweise und zum Farbenaufbau Grünewalds viel
Treffendes und stets Selbstbeobachtetes sagt, geistvoll das
Verhältnis Grünewalds zur Naturstudie, zu der Kunst Dürers und
zur italienischen erörtert.“
H. Eßwein in Münchner Post.
„Ein ausgezeichnetes, wohlfeiles, neues Grünewaldbuch, klar und

mit besonnener Begeisterung geschrieben. Keine deutsche Familie
darf künftig ganz ohne Grünewald sein. Gerade heute ist
Grünewald uns näher denn je, einmal, weil er wie keiner Leiden
und maßlosen Schmerz, wie keiner aber auch die Erlösung des
Geistes von der Materie zu künden weiß.“
R. Paulsen in Deutsche Zeitung.
„Wenn ich dieses Buch als Erinnerungs- und Gedächtniswerk
rühme, so tue ich es nicht zum wenigsten wegen des reichen und
trefflichen Bildermaterials. In 68 ausgezeichneten Photo-
Reproduktionen ersteht Grünewalds Werk, soweit es uns
überliefert ist. Der „Isenheimer Altar“ findet besondere
Berücksichtigung, und es ist dem Buche vorzüglich zu danken,
daß dem Leser neben dem Gesamteindrucke jeder Tafel und jeder
thematischen Folge auch Gelegenheit geboten ist, an Einzelheiten
die malerische Kraft Grünewalds zu studieren. Er erkennt die
wundersame Belebung des Details, die psychologischen
Erleuchtungen im Ausdruck nicht nur des Gesichtes, nein, auch
der Haltung, der Hände, der Finger neben den genialen
malerischen Gesichtspunkten des Arrangements, der linearen
Komposition und der Bildausfüllung.“
Richard Rieß in Freie Bahn, München.
D e l p h i n -Ve r l a g / M ü n c h e n
AUGUST L. MAYER
El Greco
Eine Einführung in das Leben und Wirken des
Domenico Theotocopuli genannt El Greco
Mit 70 Abbildungen. Zweite Auflage
Geheftet 7 Mark
Halbleinenband 9 Mark
„Eine sachliche, ruhig abwägende Arbeit, die unbekümmert um
Modeströmungen ein lebendiges Bild vom künstlerischen Wollen
des Greco zu entrollen versucht.“
Kunstwart.
Kleine Velazquez-Studien
Mit 15 Abbildungen
Geheftet 7 Mark. Pappband 9 Mark
„Diese Studien zeigen an einigen besonders interessanten Werken
aus dem Oeuvre des Velazquez, wieviel trotz der Arbeiten von
Justi und Beruete noch in der Chronologie der Bilder des Künstlers
zu tun bleibt. Der Verfasser zieht nicht nur ein reiches
künstlerisches, sondern auch ein umfassendes archivalisch-
genealogisches Material zur Prüfung heran.“
Kunstchronik.
Expressionistische Miniaturen des
deutschen Mittelalters
Mit 32 Tafeln
Geheftet 21 Mark. Pappband 27 Mark
„Da eine populäre Darstellung mittelalterlicher Miniaturen fehlt,
wird das Buch sicherlich große Verbreitung finden. Dank und Lob
gebühren dem Verlag für die gute Wiedergabe und Ausstattung.“
Das Kunstblatt.
FRITZ BURGER
Cézanne und Hodler
Einführung in die Probleme der Malerei der Gegenwart
Mit 195 Abbildungen. Zwei Bände
Dritte Auflage
Geheftet 26 Mark. Halbleinenband 34 Mark
„Seit Lessings Laookon ist vielleicht nicht wieder ein so
tiefgründiges Werk über die Probleme der Malerei geschrieben
worden.“
Rheinische Hochschulzeitung.
„Burger bemüht sich, mit dem ganzen Apparat des

Kunstwissenschaftlers, dem Gesetz der Gestaltung
nahezukommen. Er rückt manchmal etwas stofflich aber niemals
falsch Kunstwerke aus entlegenen Sphären zusammen, vergleicht
einen Hodler mit einem Klinger, Böcklin, Renoir, Runge, einen
Cézanne mit Rembrandt, Delacroix, Greco, Tizian, Rujsdael,
Marées, Trübner, stellt einen Genin neben Millet, einen Gauguin
neben Corot, Böcklin neben Courbet, Monet neben Matisse,
immer mit der Absicht, dadurch etwas vom Wesen der Gestaltung
und des Gestalters anschaulich zu machen.“
Paul Westheim im Berliner Tageblatt.
Weltanschauungsprobleme und
Lebenssysteme in der Kunst der
Vergangenheit
Mit 65 Abbildungen
Kartoniert 12 Mark. Halbleinenband 16 Mark
„Diese erste Niederschrift eines unvollendet gebliebenen Buches
hat den vollen Reiz einer Auseinandersetzung mit den tiefsten und
drängendsten Problemen, die mehr erlebt als durchdacht, mehr
gesprochen als geschrieben ist.“
Die bildenden Künste.
„Der erste Versuch, die europäische Kunst in den Komplex der

Gesamtkunst aller Länder und Zeiten einzuordnen.“
Der neue Tag.
MAX RAPHAEL
Von Monet zu Picasso
Grundzüge einer Ästhetik und Entwicklung der
modernen Malerei
Mit 30 zum Teil unveröffentlichten Abbildungen
Zweite Auflage
Geheftet 12 Mark. Halbpergament 16 Mark
„Trotz allem aber halte ich, ohne den Ausführungen des Autors in
allem zuzustimmen, neben den Schriften Kandinskys dieses Buch
für die beste Orientierungsmöglichkeit über die Bestrebungen der
neuen Kunst.“
Dr. H. Sauermann i. d. Krit. Rundschau.
HERMANN BAHR
Expressionismus
Mit 20 Kupferdrucktafeln. Vierte Auflage
Geheftet 10 Mark. Pappband 13.50 Mark
„Er scheut nicht die Behauptung, daß in dem Expressionismus die
tiefste Sehnsucht unserer Zeit nach Ausdruck ringt. Die Welt ist
mechanisiert. Der Mensch schreit nach seiner Seele. Die Kunst
schreit mit. Der Schrei nach Geist — das ist der Expressionismus.“
Neues Wiener Tageblatt.
MAX PICARD
Expressionistische Bauernmalerei
Mit 24 Lichtdrucktafeln. Zweite Auflage.
Kartoniert 20 Mark. Halbleinenband 24 Mark.
„Ein Buch, das seinen Wert schon dadurch hat, daß es ein nicht
beachtetes Gebiet des Kunstschaffens in die Literatur einführt und
damit überhaupt der Beachtung zuführt. Es handelt sich um die
Hinterglasbilder, die von einfältig naiven Dorfmalern als
Votivstücke geschaffen worden sind. Diese primitiven
Gestaltungen ursprünglicher Menschen, die nichts anderes als
ergreifend wirkende Expressionen sein konnten, nennt Picard
ebenso schön wie treffend, das Volkslied in der Geschichte der
Malerei.“
Das Kunstblatt.
KURT GERSTENBERG
Deutsche Sondergotik
Eine Untersuchung über das Wesen der deutschen
Baukunst im späten Mittelalter
Mit 41 Abbildungen
Geheftet 14 Mark. Halbpergament 17 Mark
„Das Buch hat außergewöhnliche Qualitäten, ist von
ungewöhnlicher Intensität und größtem inneren Reichtum. Es
gehört zu jenen besten Büchern unserer Wissenschaft, die jeder
Fachmann lesen muß. — Der Verfasser will nicht die Geschichte
der deutschen Spätgotik schreiben, sondern ihr Wesen
untersuchen und es auf Bekanntes zurückführen, auf die
Sonderart deutschen Wesens.“
Dozent Dr. Paul Frankl.
ERNST TROSS
Das Raumproblem in der bildenden Kunst
Mit 14 Abbildungen
Geheftet 7.50 Mark. Ganzleinenband 10 Mark
„Es handelt sich also nun nicht mehr um das Problem der Form,
sondern um das Problem des Raumes, — in der Tat eine
wesentliche Klärung und Vertiefung jener Theorie. Sein nicht
geringes Verdienst liegt aber darin, die Fiedler-Hildebrandische
Lehre in den philosophischen Grundlagen, besonders der
Erkenntnistheorie fest verankert zu haben. Wer sich um ihre
Erfassung und Verarbeitung bemüht, für den wird das vorliegende
Buch unentbehrlich sein.“
Monatsheft für Kunstwissenschaft.
JULIE VOGELSTEIN
Von französischer Buchmalerei
Mit 77 Abbildungen auf 28 Lichtdrucktafeln
Geheftet 25 Mark. Gebunden 30 Mark
„Es wird hier versucht, das Wesen der Illumination als
Buchschmuck zu erkennen und in diesem Sinne gesondert zu
betrachten. In systematischer Gruppierung des Materials wird an
prägnanten Typen die Beziehung der Buchmalerei zur Buchseite in
ideeller wie technischer Hinsicht gezeigt.“
Der Cicerone.
ARTHUR RÜMANN
Honoré Daumier als Illustrator
Mit 150 Abbildungen
Pappband 16 Mark
„Das geschmackvolle Buch zeigt drei Jahrzehnte französischen
Bürgertums in einem unvergleichlichen künstlerischen
Niederschlag. Ein schlichter, anspruchsloser Text verbindet die
Fülle ausgezeichneter Illustrationsproben, die eine Vorstellung von
der unerschöpflichen Fruchtbarkeit des Meisters erwecken.“
Die bildenden Künste.
JOSEFA DÜRCK-KAULBACH
Erinnerungen an Wilhelm von Kaulbach
und sein Haus
Mit Briefen, 160 Zeichnungen und Bildern
Zweite Auflage
Pappband 12 Mark. Halbleinenband 15 Mark
„Der Wert dieses sehr hübschen Buches liegt im rein
Menschlichen. Die Verfasserin, eine Tochter Kaulbachs, zeichnet
die Bilder ihrer Kindheit und Jugend, den Kreis ihrer Eltern und
Freunde, das alte München vor siebzig Jahren ... Es ist ein Stück
Großvaterzeit, ein Stück Biedermeier, ein Stück Kindheit und
verlorenes Paradies.“
H. Hesse.
LOTHAR BRIEGER
Das Kunstsammeln / Theorie und Technik
Mit 16 Abbildungen. Zweite Auflage
Pappband 5.50 Mark
„Es handelt sich hier um ein Nachschlagewerk für den
Kunstkenner, wie er es kürzer, inhaltreicher und handlicher kaum
finden dürfte. Für den Laien ist es eine Einführung, wie er sie
besser gar nicht haben könnte.“
Werkstatt der Kunst.
Kleine Delphin-Kunstbücher
Die hübschen Bändchen bringen jeweils mindestens 24
Bilder in guter Wiedergabe und geben nach einer
knappen Einleitung glücklich ausgewählte Proben von
Briefen oder auch poetischen oder gedanklichen
Äußerungen der Meister.
Jedes Bändchen fest kartoniert 1.35 Mark

Gesamtauflage eine Million
Bis jetzt sind erschienen:
1. Folge
S p i t z w e g , Reime und Bilder
S c h w i n d , Briefe und Bilder
W a l d m ü l l e r , Bilder und Erlebnisse
F e u e r b a c h , Bilder u.
Bekenntnisse
R i c h t e r , Aus beschaulicher Zeit
O b e r l ä n d e r , Heiteres und Ernstes
2. Folge
R e t h e l , Der Künstler und Mensch
R u b e n s , Der große Flame
T h o m a , Der Malerpoet
M e n z e l , Werke und Dokumente
G r ü n e w a l d , Der Romantiker d.
Schmerzes
C o r i n t h , Ein Maler unserer Zeit
3. Folge
L e i b l , Ein deutscher Maler
M u r i l l o , Der Maler der
Betteljungen und
Madonnen
W i l h e l m B u s c h , Lustige
Kleinigkeiten
D a u m i e r , Der Meister der
Karikatur
L i o n a r d o , Bilder und Gedanken
H o s e m a n n , Der Maler des
Berliner Volkes
„Zu den lieblichsten Erscheinungen des Büchermarktes zählen die

seit einiger Zeit erscheinenden Delphin-Kunst-Bücher, kleine
Künstlermonographien in Wort und Bild, mit dem Bestreben, Seele
und Können des betreffenden Künstlers in Kürze, Klarheit und
Schönheit zu offenbaren.“
Augsburger Post-Zeitung.
„Hier ist mit feinstem Geschmack in der Bilderauswahl die

trefflichste Ausgestaltung des begleitenden und einführenden
Textes zu einem eindringlichen und harmonischen Ganzen
gestaltet, sodaß ein lichtvolles und wirkliches Bild der
Künstlerpersönlichkeit entsteht.“
Mannheimer Tageblatt.
Romane und Novellen
HERMANN ESSWEIN. M e g a n d e r, der Mann mit
den zween Köpfen. Geheftet 3.50 Mark,
Halblederband 5.50 Mark.
„Ein köstlicher Schalksgeist.“
Eduard Müller im Gothaischen Tageblatt.
ALFRED FANKHAUSER. Peter der To r und

seine Liebe. Geheftet 7 Mark, Pappband 9
Mark.
„Alle die suchen, werden von diesem Tagebuch erschüttert,
aber auch bereichert und vertieft ihre Lebenswanderung
fortsetzen.“
O. v. Greyerz im Berner Bund.
HANS FRANCK. Das Pentagramm der Liebe.

Geheftet 7 Mark, Pappband 9 Mark.
„Die Urströme weiblicher Empfindung an den Quellen gefaßt.“
Leipz. N. Nachr.
A. M. FREY. Kastan und die Dirnen. Geheftet 7

Mark, Pappband 9 Mark.
„Verblüffend ist die Mischung von Brutalität und Weichheit,
Grauen und Komik, Perversität und Keuschheit ..., die dem Leser
fast den Atem nehmen.“
A. Bechthold in „Das neue Buch“.
A. M. FREY. Solneman der Unsichtbare. Dritte

Auflage. Mit 12 Holzschnitten von Nückel.
Geheftet 9 Mark, gebunden 12.50 Mark.
„Gehört zum Allerbesten, was die phantastische Literatur von
heute hervorgebracht hat.“
Thomas Mann.
LUDWIG GORM. Päpstin Johanna. Zweite

Auflage. Geheftet 5 Mark, Pappband 7 Mark.
„Ein klingendes und duftendes Gedicht menschlicher Liebe und
Leidenschaft und sündigen Ehrgeizes.“
Die Mainbrücke.
LUDWIG GORM. D i e K i n d e r v o n G e n f. G e h e f t e t
7 Mark, Pappband 9 Mark.
„Ein schönes und reiches Buch, voll starker Farben und
kraftvoller Linien, voll tiefer Gedanken und eindringlicher Bilder.“
Fritz Böhme in Warschauer Zeitung.
HANNS JOHST. Der Anfang. Geheftet 7 Mark,

Pappband 9 Mark.
„Ein Dichter mit durchaus eigener Sprache, ein Lohefeuer von
Jugend, Begeisterung und einer strahlend frischen Güte!“
Zwiebelfisch.
REINHARD KOESTER. Der Gang des

Gottlosen. Geheftet 7 Mark, Pappband 9 Mark.
Ein Entwurzelter und Entgötterter ringt, an allem zweifelnd,
nach neuem Halt. Freude ist letztes Ziel, dem der Gang des
Gottlosen gilt.
*** END OF THE PROJECT GUTENBERG EBOOK REMBRANDT ***
Updated editions will replace the previous one—the old editions

will be renamed.
Creating the works from print editions not protected by U.S.

copyright law means that no one owns a United States
copyright in these works, so the Foundation (and you!) can copy
and distribute it in the United States without permission and
without paying copyright royalties. Special rules, set forth in the
General Terms of Use part of this license, apply to copying and
distributing Project Gutenberg™ electronic works to protect the
PROJECT GUTENBERG™ concept and trademark. Project
Gutenberg is a registered trademark, and may not be used if
you charge for an eBook, except by following the terms of the
trademark license, including paying royalties for use of the
Project Gutenberg trademark. If you do not charge anything for
copies of this eBook, complying with the trademark license is
very easy. You may use this eBook for nearly any purpose such
as creation of derivative works, reports, performances and
research. Project Gutenberg eBooks may be modified and
printed and given away—you may do practically ANYTHING in
the United States with eBooks not protected by U.S. copyright
law. Redistribution is subject to the trademark license, especially
commercial redistribution.
START: FULL LICENSE

THE FULL PROJECT GUTENBERG LICENSE
PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK
To protect the Project Gutenberg™ mission of promoting the

free distribution of electronic works, by using or distributing this
work (or any other work associated in any way with the phrase
“Project Gutenberg”), you agree to comply with all the terms of
the Full Project Gutenberg™ License available with this file or
online at www.gutenberg.org/license.
Section 1. General Terms of Use and

Redistributing Project Gutenberg™
electronic works
1.A. By reading or using any part of this Project Gutenberg™
electronic work, you indicate that you have read, understand,
agree to and accept all the terms of this license and intellectual
property (trademark/copyright) agreement. If you do not agree
to abide by all the terms of this agreement, you must cease
using and return or destroy all copies of Project Gutenberg™
electronic works in your possession. If you paid a fee for
obtaining a copy of or access to a Project Gutenberg™
electronic work and you do not agree to be bound by the terms
of this agreement, you may obtain a refund from the person or
entity to whom you paid the fee as set forth in paragraph 1.E.8.
1.B. “Project Gutenberg” is a registered trademark. It may only

be used on or associated in any way with an electronic work by
people who agree to be bound by the terms of this agreement.
There are a few things that you can do with most Project
Gutenberg™ electronic works even without complying with the
full terms of this agreement. See paragraph 1.C below. There
are a lot of things you can do with Project Gutenberg™
electronic works if you follow the terms of this agreement and
help preserve free future access to Project Gutenberg™
electronic works. See paragraph 1.E below.
1.C. The Project Gutenberg Literary Archive Foundation (“the
Foundation” or PGLAF), owns a compilation copyright in the
collection of Project Gutenberg™ electronic works. Nearly all the
individual works in the collection are in the public domain in the
United States. If an individual work is unprotected by copyright
law in the United States and you are located in the United
States, we do not claim a right to prevent you from copying,
distributing, performing, displaying or creating derivative works
based on the work as long as all references to Project
Gutenberg are removed. Of course, we hope that you will
support the Project Gutenberg™ mission of promoting free
access to electronic works by freely sharing Project Gutenberg™
works in compliance with the terms of this agreement for
keeping the Project Gutenberg™ name associated with the
work. You can easily comply with the terms of this agreement
by keeping this work in the same format with its attached full
Project Gutenberg™ License when you share it without charge
with others.
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside
the United States, check the laws of your country in addition to
the terms of this agreement before downloading, copying,
displaying, performing, distributing or creating derivative works
based on this work or any other Project Gutenberg™ work. The
Foundation makes no representations concerning the copyright
status of any work in any country other than the United States.
1.E. Unless you have removed all references to Project

Gutenberg:
1.E.1. The following sentence, with active links to, or other

immediate access to, the full Project Gutenberg™ License must
appear prominently whenever any copy of a Project
Gutenberg™ work (any work on which the phrase “Project
Gutenberg” appears, or with which the phrase “Project
Gutenberg” is associated) is accessed, displayed, performed,
viewed, copied or distributed:
This eBook is for the use of anyone anywhere in the United

States and most other parts of the world at no cost and
with almost no restrictions whatsoever. You may copy it,
give it away or re-use it under the terms of the Project
Gutenberg License included with this eBook or online at
www.gutenberg.org. If you are not located in the United
States, you will have to check the laws of the country
where you are located before using this eBook.
1.E.2. If an individual Project Gutenberg™ electronic work is

derived from texts not protected by U.S. copyright law (does not
contain a notice indicating that it is posted with permission of
the copyright holder), the work can be copied and distributed to
anyone in the United States without paying any fees or charges.
If you are redistributing or providing access to a work with the
phrase “Project Gutenberg” associated with or appearing on the
work, you must comply either with the requirements of
paragraphs 1.E.1 through 1.E.7 or obtain permission for the use
of the work and the Project Gutenberg™ trademark as set forth
in paragraphs 1.E.8 or 1.E.9.
1.E.3. If an individual Project Gutenberg™ electronic work is

posted with the permission of the copyright holder, your use and
distribution must comply with both paragraphs 1.E.1 through
1.E.7 and any additional terms imposed by the copyright holder.
Additional terms will be linked to the Project Gutenberg™
License for all works posted with the permission of the copyright
holder found at the beginning of this work.
1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files
containing a part of this work or any other work associated with
Project Gutenberg™.
1.E.5. Do not copy, display, perform, distribute or redistribute

this electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1
with active links or immediate access to the full terms of the
Project Gutenberg™ License.
1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if
you provide access to or distribute copies of a Project
Gutenberg™ work in a format other than “Plain Vanilla ASCII” or
other format used in the official version posted on the official
Project Gutenberg™ website (www.gutenberg.org), you must,
at no additional cost, fee or expense to the user, provide a copy,
a means of exporting a copy, or a means of obtaining a copy
upon request, of the work in its original “Plain Vanilla ASCII” or
other form. Any alternate format must include the full Project
Gutenberg™ License as specified in paragraph 1.E.1.
1.E.7. Do not charge a fee for access to, viewing, displaying,

performing, copying or distributing any Project Gutenberg™
works unless you comply with paragraph 1.E.8 or 1.E.9.
1.E.8. You may charge a reasonable fee for copies of or

providing access to or distributing Project Gutenberg™
electronic works provided that:
• You pay a royalty fee of 20% of the gross profits you derive
from the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”
• You provide a full refund of any money paid by a user who

notifies you in writing (or by e-mail) within 30 days of receipt
that s/he does not agree to the terms of the full Project
Gutenberg™ License. You must require such a user to return or
destroy all copies of the works possessed in a physical medium
and discontinue all use of and all access to other copies of
Project Gutenberg™ works.
• You provide, in accordance with paragraph 1.F.3, a full refund of

any money paid for a work or a replacement copy, if a defect in
the electronic work is discovered and reported to you within 90
days of receipt of the work.
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.E.9. If you wish to charge a fee or distribute a Project

Gutenberg™ electronic work or group of works on different
terms than are set forth in this agreement, you must obtain
permission in writing from the Project Gutenberg Literary
Archive Foundation, the manager of the Project Gutenberg™
trademark. Contact the Foundation as set forth in Section 3
below.
1.F.
1.F.1. Project Gutenberg volunteers and employees expend

considerable effort to identify, do copyright research on,
transcribe and proofread works not protected by U.S. copyright

Algebraic Topology A Structural Introduction 1st Edition Marco Grandis

Uploaded by

Copyright:

Available Formats

Algebraic Topology A Structural Introduction 1st Edition Marco Grandis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Algebraic Topology A Structural Introduction 1st Edition Marco Grandis

Uploaded by

Copyright:

Available Formats

Full download ebook at ebookname.

Algebraic Topology A Structural Introduction 1st

Download more ebook from https://ebookname.com

Introduction to Algebraic Topology 1st Edition Holger

Algebraic Topology A Primer 1st Edition Satya Deo

Homology Cohomology and Sheaf Cohomology for Algebraic

Introduction to Algebraic Geometry Justin R. Smith

Introduction to Algebraic and Constructive Quantum

Local Cohomology An Algebraic Introduction with

Dopamine 1st Edition Marco Diana

An Invitation To Algebraic Numbers And Algebraic

An Elementary Overview of Mathematical Structures:

Manifolds and Local Structures: A General Theory

Category Theory and Applications: A Textbook for Beginners

Yumeng - 12586 - Algebraic Topology.indd 1 15/11/2021 9:03:06 am

British Library Cataloguing-in-Publication Data

ISBN 978-981-124-835-1 (hardcover)

For any available supplementary material, please visit

Yumeng - 12586 - Algebraic Topology.indd 2 15/11/2021 9:03:06 am

To my colleagues and friends

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

Algebraic Topology is a system and strategy of partial translations, aiming

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

Preface page vii

0.1 Investigating spaces with algebraic structures

0.2 Homology and cohomology theories

0.4 Translations and underlying affinities

0.5 An inductive approach on structural bases 5

0.5 An inductive approach on structural bases

0.6 Prerequisites and literature

0.7 Notation and conventions

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

Algebraic Topology studies spaces and continuous mappings, transforming

1.1 Classifying spaces and maps

1.1.1 The shape of a space

10 Introducing Algebraic Topology

(i) open non-degenerate intervals,

1.1 Classifying spaces and maps 11

1.1.3 Classifying the homotopy type

12 Introducing Algebraic Topology

where m : A → X is the inclusion. Forgetting the homotopy mp ' id X,

ϕ : mp ' id X, ϕ(x, t) = x (x ∈ A, t ∈ I). (1.5)

Exercises and complements. (a) Every euclidean space Rn is contractible,

N : X → Sn , N (x) = x/||x|| (x 6= 0). (1.6)

(e) The subspace A ⊂ X is a deformation retract of X if and only if there

1.1 Classifying spaces and maps 13

1.1.4 Hints at singular homology

14 Introducing Algebraic Topology

1.1.5 The homology of the spheres

Roughly speaking, a non-trivial group Hk (X) detects ‘holes’ of dimension

1.1 Classifying spaces and maps 15

1.1.6 Studying retracts

1.1.7 Topological embeddings and projections

16 Introducing Algebraic Topology

In this case we may want to ‘identify’ X with its image in Y , so that

Here we may want to identify Y with this quotient of X. Topological

∂ α : {∗} I, ∂ − (∗) = 0, ∂ + (∗) = 1 (f aces),