dutroduction to TOPOLOGY AND MODERN ANALYSISJutroduction to TOPOLOGY AND MODERN ANALYSIS GEORGE F. SIMMONS Associate Professor of Mathematics Colorado College ® ROBERT E. KRIEGER PUBLISHING COMPANY MALABAR, FLORIDAFor Virgie May Hatcher and Elizabeth B. Blossom TO EACH OF WHOM 1 OWE MORE THAN | CAN POSSIBLY EXPRESS Original Edition 1963 Reprint Edition 1983 Printed and Published by ROBERT E. KRIEGER PUBLISHING COMPANY, INC. KRIEGER DRIVE MALABAR, FLORIDA 32950 Copyright © 1963 by McGraw-Hill, Inc. Reprinted by arrangement All rights reserved. No part of this book may be reproduced in any form or by any electronic or mechanical means including information ‘storage and retrieval systems without permission in writing from the publisher. Printed in the United States of America Library of Congress Cataloging in Publication Data Simmons, George Finlay, 1925- Introduction to topology and modern analysis. Reprint. Originally published: New York: McGraw- Hill, 1963 (International series in pure and applied mathematics) Bibliography: p. Includes index. 1. Topology. . 2. Mathematical analysis. I. Title. Il. Title: Topology and modern analysis. IIL. Series: International series in pure and applied mathematics. QA6I1.S49 1983 514 82-14845 ISBN 0-89874—551-9 1098765Preface For some time now, topology has been firmly established as one of the basic disciplines of pure mathematics. Its ideas and methods have transformed large parts of geometry and analysis almost beyond recogni- tion. It has also greatly stimulated the growth of abstract algebra. As things stand today, much of modern pure mathematics must remain a closed book to the person who does not acquire a working knowledge of at. least the elements of topology. There are many domains in the broad field of topology, of which the following are only a few: the homology and cohomology theory of com- plexes, and of more general spaces as well; dimension theory; the theory of differentiable and Riemannian manifolds and of Lie groups; the theory of continuous curves; the theory of Banach and Hilbert spaces and their operators, and of Banach algebras; and abstract harmonic analysis on locally compact groups. Each of these subjects starts from roughly the same body of fundamental knowledge and develops its own methods of dealing with its own characteristic problems. The purpose of Part 1 of this book is to make available to the student this “hard core” of funda- mental topology; specifically, to make it available in a form which is general enough to meet the needs of modern mathematics, and yet is unburdened by excess baggage best left in the research journals. A topological space can be thought of as a set from which has been swept away all structure irrelevant to the continuity of functions defined on it. Part 1 therefore begins with an informal (but quite extensive) treatment of sets and functions. Some writers deal with the theory of metric spaces as if it were merely a fragment of the general theory of topological spaces. This practice is no doubt logically correct, but it seems to me to violate the natural relation between these topics, in which metric spaces motivate the more general theory. Metric spaces are therefore discussed rather fully in Chapter 2, and topological spaces are introduced in Chapter 3. The remaining four chapters in Part 1 are concerned with various kinds of topological spaces of special importance in applications and with the continuous functions carried by them. It goes without saying that one aspect of this type of mathematics is its logical precision. Too many writers, however, are content with this, and make little effort to help the reader maintain his orientation in vilPreface the midst of mazes of detail. One of the main features of this book is the attention given to motivating the ideas under discussion. On every possible occasion I have tried to make clear the intuitive meaning of what is taking place, and diagrams are provided, whenever it seems feasible, to help the reader develop skill in using his imagination to visualize abstract ideas. Also, each chapter begins with a brief introduction which describes its main theme in general terms. Courses in topology are being taught more and more widely on the undergraduate level in our colleges and universities, and I hope that these features, which tend to soften the austere framework of definitions, theorems, and proofs, will make this book readable and easy to use as a text. Historically speaking, topology has followed two principal lines of development. In homology theory, dimension theory, and the study of manifolds, the basic motivation appears to have come from geometry. In these fields, topological spaces are looked upon as generalized geometric configurations, and the emphasis is placed on the structure of the spaces themselves. In the other direction, the main stimulus has been analysis. Continuous functions are the chief objects of interest here, and topological spaces are regarded primarily as carriers of such functions and as domains over which they can be integrated. These ideas lead naturally into the theory of Banach and Hilbert spaces and Banach algebras, the modern theory of integration, and abstract harmonic analysis on locally compact. groups. In Part 1 of this book, I have attempted an even balance between these two points of view. This part is suitable for a basic semester course, and most of the topics treated are indispensable for further study in almost any direction. If the instructor wishes to devote a second semester to some of the extensions and applications of the theory, many possibilities are open. If he prefers applications in modern analysis, he can continue with Part 2 of this book, supplemented, perhaps, with a brief treatment of measure and integration aimed at the general form of the Riesz representation theorem. Or if his tastes incline him toward the geometric aspects of topology, he can switch over to one of the many excellent books which deal with these matters. The instructor who intends to continue with Part 2 must face a question which only he can answer. Do his students know enough about algebra? This question is forced to the surface by the fact that Chapters 9 to 11 are as much about algebra as they are about topology and analy- sis. If his students know little or nothing about modern algebra, then a careful and detailed treatment of Chapter 8 should make it possible to proceed without difficulty. And if they know a good deal, then a quick survey of Chapter 8 should suffice. It is my own opinion that education in abstract mathematics ought to begin on the junior level with a course in modern algebra, and that topology should be offered only to studentsPreface ix who have acquired some familiarity, through such a course, with abstract methods. Part 3 is intended for individual study by exceptionally well-qualified students with a reasonable knowledge of complex analysis. Its principal purpose is to unify Parts 1 and 2 into a single body of thought, along the lines mapped out in the last section of Chapter 11. Taken as a whole, the present work stands at the threshold of the more advanced books by Rickart [34], Loomis [27], and Naimark [32]; and much of its subject matter can be found (in one form or another and with innumerable applications to analysis) in the encyclopedic treatises of Dunford and Schwartz (8) and Hille and Phillips [20].!_ This book is intended to be elementary, in the sense of being accessible to well-trained undergraduates, while those just mentioned are not. Its prerequisites are almost negligible. Several facts about determinants are used without proof in Chapter 11, and Chapter 12 leans heavily on Liouville’s theorem and the Laurent expansion from complex analysis. With these excep- tions, the book is essentially self-contained. It seems to me that a worthwhile distinction can be drawn between two types of pure mathematics. The first—which unfortunately is somewhat out of style at present—centers attention on particular func- tions and theorems which are rich in meaning and history, like the gamma function and the prime number theorem, or on juicy individual facts, like Euler’s wonderful formula 14 M4MH 0+: = 0/6. ‘The second is concerned primarily with form and structure. The present book belongs to this camp; for its dominant theme can be expressed in just two words, continuity and linearity, and its purpose is to illuminate the meanings of these words and their relations to each other. Mathe- matics of this kind hardly ever yields great and memorable results like the prime number theorem and Euler's formula. On the contrary, its theorems are generally small parts of a much larger whole and derive their main significance from the place they occupy in that whole. In my opinion, if a body of mathematics like this is to justify itself, it must possess aesthetic qualities akin to those of a good piece of architecture. It should have a solid foundation, its walls and beams should be firmly and truly placed, each part should bear a meaningful relation to every other part, and its towers and pinnacles should exalt the mind. It is my hope that this book can contribute to a wider appreciation of these mathe- matical values. George F. Simmons 1 The numbers in brackets refer to works listed in the Bibliography.A Note to the Keader Two matters call for special comment: the problems and the proofs. The majority of the problems are corollaries and extensions of theorems proved in the text, and are freely drawn upon at all later stages of the book. In general, they serve as a bridge between ideas just treated and developments yet: to come, and the reader is strongly urged to master them as he goes along. In the earlier chapters, proofs are given in considerable detail, in an effort to smooth the way for the beginner. As our subject unfolds through the successive chapters and the reader acquires experience in following abstract mathematical arguments, the proofs become briefer and minor details are more and more left for the reader to fill in for himself. The serious student will train himself to look for gaps in proofs, and should regard them as tacit invitations to do a little thinking on his own. Phrases like “it is easy to see,” “one can easily show,” “evidently,” “clearly,” and so on, are always to be taken as warning signals which indicate the presence of gaps, and they should put the reader on his guard. It is a basic principle in the study of mathematics, and one too seldom emphasized, that a proof is not really understood until the stage is reached at which one can grasp it as a whole and see it as a single idea. In achieving this end, much more is necessary than merely follow- ing the individual steps in the reasoning. This is only the beginning. A proof should be chewed, swallowed, and digested, and this process of assimilation should not be abandoned until it yields a full comprehension of the overall pattern of thought.Contents Preface A Note to the Reader xi PART ONE: TOPOLOGY Chapter One SETS AND FUNCTIONS 3 ONO e OE . Sets and set inclusion 3 . The algebra of sets 7 . Functions 14 Products of sets 21 Partitions and equivalence relations 25 Countable sets 31 . Uncountable sets 36 . Partially ordered sets and lattices 43 Chapter Two METRIC SPACES 49 9. The definition and some examples 51 10. Open sets 59 11. Closed sets 65 12. Convergence, completeness, and Baire’s theorem 70 13. Continuous mappings 75 14, Spaces of continuous functions 80 15. Euclidean and unitary spaces 85 Chapter Three TOPOLOGICAL SPACES 91 16. The definition and some examples 92 17. Elementary concepts 95 18. Open bases and open subbases 99 19. Weak topologies 104 20. The function algebras €(X,R) and e(X,C) 106 Chapter Four COMPACTNESS 110 21, Compact spaces 111 22, Products of spaces 115 xillxiv Contents 23, Tychonofi’s theorem and locally compact spaces 118 24. Compactness for metric spaces 120 25, Ascoli’s theorem 124 Chapter Five SEPARATION 26. Trspaces and Hausdorff spaces 130 27. Completely regular spaces and normal spaces 132 28. Urysohn’s lemma and the Tietze extension theorem 135 29, The Urysohn imbedding theorem 137 30. The Stone-Cech compactification 139 Chapter Six CONNECTEDNESS 31. Connected spaces 143 32. The components of a space 146 33, Totally disconnected spaces 149 34, Locally connected spaces 150 Chapter Seven APPROXIMATION 35. The Weierstrass approximation theorem 153 36. The Stone-Weierstrass theorems 157 37, Locally compact Hausdorff spaces 162 38, The extended Stone-Weierstrass theorems 165 PART TWO: Chapter Eight ALGEBRAIC SYSTEMS 39. Groups 172 40. Rings 181 41, The structure of rings 184 42, Linear spaces 191 43, The dimension of a linear space 196 44, Linear transformations 203 45, Algebras 208 Chapter Nine BANACH SPACES 46. The definition and some examples 212 47, Continuous linear transformations 219 48, The Hahn-Banach theorem 224 49, The natural imbedding of Nin N** 231 50. The open mapping theorem 235 51, The conjugate of an operator 239 Chapter Ten HILBERT SPACES 52. The definition and some simple properties 244 53. Orthogonal complements 249 54. Orthonormal sets 251 129 142 153 OPERATORS 171 211 243Contents x” 55. The conjugate space H* 260 56. The adjoint of an operator 262 57. Self-adjoint operators 266 58. Normal and unitary operators 269 59. Projections 273 Chapter Eleven FINITE-DIMENSIONAL SPECTRAL THEORY 278 60. Matrices 280 61. Determinants and the spectrum of an operator 287 62. The spectral theorem 290 63. A survey of the situation 295 PART THREE: ALGEBRAS OF OPERATORS Chapter Twelve GENERAL PRELIMINARIES ON BANACH ALGEBRAS 301 64. The definition and some examples 302 65. Regular and singular elements 305 66. Topological divisors of zero 307 67. The spectrum 308 68, The formula for the spectral radius 312 69. The radical and semi-simplicity 313 Chapter Thirteen THE STRUCTURE OF COMMUTATIVE BANACH ALGEBRAS 318 70. The Gelfand mapping 318 71. Applications of the formula r(x) = lim |[z"|]™ 323 72, Involutions in Banach algebras 324 73. The Gelfand-Neumark theorem 325 Chapter Fourteen SOME SPECIAL COMMUTATIVE BANACH ALGEBRAS 327 74. Ideals in €(X) and the Banach-Stone theorem 327 75. The Stone-Cech compactification (continued) 330 76. Commutative C*-algebras 332 APPENDICES ONE _ Fixed point theorems and some applications to analysis 337 TWO Continuous curves and the Hahn-Mazurkiewicz theorem 341 THREE Boolean algebras, Boolean rings, and Stone’s theorem 344 Bibliography 355 Index of Symbols 359 Subject Index 363PART ONE CopologyCHAPTER ONE Sets and Functions It is sometimes said that mathematics is the study of sets and func- tions. Naturally, this oversimplifies matters; but it does come as close to the truth as an aphorism can. The study of sets and functions leads two ways. One path goes down, into the abysses of logic, philosophy, and the foundations of mathematics. The other goes up, onto the highlands of mathematics itself, where these concepts are indispensable in almost all of pure mathe- matics as it is today. Needless to say, we follow the latter course. We regard sets and functions as tools of thought, and our purpose in this chapter is to develop these tools to the point where they are sufficiently powerful to serve our needs through the rest of this book. As the reader proceeds, he will come to understand that the words set and function are not as simple as they may seem. In a sense, they are simple; but they are potent words, and the quality of simplicity they possess is that which lies on the far side of complexity. They are like seeds, which are primitive in appearance but have the capacity for vast and intricate development. 1, SETS AND SET INCLUSION We adopt a naive point of view in our discussion of sets and assume that the concepts of an element and of a set of elements are intuitively clear. By an element we mean an object or entity of some sort, as, for example, a positive integer, a point on the real line (= real number), 34 Topology or a point in the complex plane (= a complex number). A set is a collec- tion or aggregate of such elements, considered together or as a whole. Some examples are furnished by the set of all even positive integers, the set of all rational points on the real line, and the set of all points in the complex plane whose distance from the origin is 1 (= the unit circle in the plane). We reserve the word class to refer to a set of sets. We might speak, for instance, of the class of all circles in a plane (thinking of each circle as a set of points). It will be useful in the work we do if we carry this hierarchy one step further and use the term family for a set: of classes. One more remark: the words element, set, class, and family are not intended to be rigidly fixed in their usage; we use them fluidly, to express varying attitudes toward the mathematical objects and systems we study. It is entirely reasonable, for instance, to think of a circle not as a set of points, but as a single entity in itself, in which case we might justifiably speak of the set of all circles in a plane. There are two standard notations available for designating a par- ticular set. Whenever it is feasible to do so, we can list its elements between braces. Thus {1, 2, 3} signifies the set consisting of the first three positive integers, {1, 7, —1, —1} is the set of the four fourth roots of unity, and {+1, +3, +5, . . .} is the set of all odd integers. This manner of specifying a set, by listing its elements, is unworkable in many circumstances. We are then obliged to fall back on the second method, which is to use a property or attribute that characterizes the elements of the set in question. If P denotes a certain property of elements, then {z:P} stands for the set of all elements xz for which the property P is meaningful and true. For example, the expression {z:2 is real and irrational}, which we read the set of all x such that x is real and irrational, denotes the set of all real numbers which cannot be written as the quotient of two integers. The set under discussion contains all those elements (and no others) which possess the stated property. The three sets of numbers described at the beginning of this paragraph can be written either way: {1, 2, 3} = {minis an integer and 0 A and BD A. It will often be convenient to have a symbol for logical implication, and = is the symbol we use. If p and gq are statements, then p= q means that p implies q, or that if p is true, then gis also true. Similarly, « is our symbol for two-way implication or logical equivalence. It means that the statement on each side implies the statement on the other, and is usually read if and only if, or is equivalent to. The main properties of set inclusion are obvious. They are the following: (1) A CA for every A; (2) ACBandBCA>A =B; (3) AC Band BOCS>ACC. It is quite important to observe that (1) and (2) can be combined into the single statement that A= BACB and BCA. This remark contains a useful principle of proof, namely, that the only way to show that two sets are equal, apart from merely inspecting them, is to show that each is a subset of the other. Problems 1. Perhaps the most famous of the logical difficulties referred to in the text is Russell’s paradoz. To explain what this is, we begin by observing that a set can easily have elements which are themselves sets, e.g., {1, {2,3}, 4}. This raises the possibility that a set might well contain itself as one of its elements. We call such a set an abnormal set, and any set which does not contain itself as an element we call a normal set. Most sets are normal, and if we suspect that abnormal sets are in some way undesirable, we might try to confine our attention to the set N of all normal sets. Someone is now sure to ask, Is N itself normal or abnormal? It is evidently one or the other, and it cannot be both. Show that if N is normal, then it must be abnormal. Show also that if N is abnormal, then it must be normal. We see in this way that each of our two alternatives is self-contradictory, and it seems to be the assumption that N exists asa set which has brought us to this impasse. For further discussion of these matters, we refer the interested reader to Wilder (42, p. 55]Sets and Functions 7 or Fraenkel and Bar-Hillel (10, p. 6]. Russell’s own account of the discovery of his paradox can be found in Russell (36, p. 75]. 2. The symbol we have used for set inclusion is similar to that used for the familiar order relation on the real line: if x and y are real numbers, a x=y; (3) 22AUBEA; (2) AandBeA>ANBeEA; (3) AtA=SA’cA. Since A is assumed to be non-empty, it must contain at least one set A. Property (3) shows that A’ is in A along with A, and since A A’ = 0 and A U A’ = JU, (1) and (2) guarantee that A contains the empty set and the universal set. Since the class consisting only of the empty set and the universal set is clearly a Boolean algebra of sets, these two distinct sets are the only ones which every Boolean algebra of sets mustSets and Functions. 13 contain. It is equally clear that the class of all subsets of U is also a Boolean algebra of sets. There are many other less trivial kinds, and their applications are manifold in fields of study as diverse as statistics and electronics. Let A be a Boolean algebra of sets. It is obvious that if {Ai, 42, . . . , An} is a non-empty finite subclass of A, then A,\UA2U++*UAn and Ai M AN ++ > ON An are both sets in A; and since A contains the empty set and the universal set, it is easy to see that A is a class of sets which is closed under the formation of finite unions, finite intersections, and complements. We now go in the other direction, and let A be a class of sets which is closed under the formation of finite unions, finite intersections, and comple- ments. By these assumptions, A automatically contains the empty set and the universal set, so it is non-empty and is easily seen to be a Boolean algebra of sets. We conclude from these remarks that Boolean algebras of sets can be described alternatively as classes of sets which are closed under the formation of finite unions, finite intersections, and comple- ments. It should be emphasized once again that when discussing Boolean algebras of sets we always assume that the universal set is non- empty. One final comment. We speak of Boolean algebras of sets because there are other kinds of Boolean algebras than those which consist of sets, and we wish to preserve the distinction. We explore this topic further in our Appendix on Boolean algebras. Problems 1. If {A.J and {B;} are two classes of sets such that {A.J C {Bj}, show that U.A; C UB; and 0B; © OAs 2. The difference between two sets A and B, denoted by A — B, is the set of all elements in A and not in B; thus A — B = AB’. Show the following: A-~B=A-(ANB) = (AUB) -B; (A -B)-C=A-(BUC); A-(B-C)=(A-B)U(ANO); (AUB) -C=(A-C)UB-O); A-(BUC)=(A-B)N(A- 0). 3. The symmetric difference of two sets A and B, denoted by AA B, is defined by A AB = (A — B) U (B — A); it is thus the union of14 Topology their differences in opposite orders. Show the following: AA(BAC) = (AAB)AC; AAG=A;AAA =H; AAB=BAA; AN (BAC) = (ANB)A(ANC). 4, A ring of sets is a non-empty class A of sets such that if A and B are in A, then A A B and AB arealsoin A. Show that A must also contain the empty set, AU B, and A — B. Show that if a non- empty class of sets contains the union and difference of any pair of its sets, then it isa ring of sets. Show that a Boolean algebra of sets is a ring of sets. 5. Show that the class of all finite subsets (including the empty set) of an infinite set is a ring of sets but is not a Boolean algebra of sets. 6. Show that the class of all finite unions of closed-open intervals on the real line is a ring of sets but is not a Boolean algebra of sets. 7. Assuming that the universal set U is non-empty, show that Boolean algebras of sets can be described as rings of sets which contain U. o . FUNCTIONS Many kinds of functions occur in topology, in a great variety of situations. In our work we shall need the full power of the general con- cept of a function, and since its modern meaning is much broader and deeper than its elementary meaning, we discuss this concept in con- siderable detail and develop its main abstract properties. Let us begin with a brief inspection of some simple examples. Con- sider the elementary function yon? of the real variable z. What do we have in mind when we call this a function and say that y is a function of z? In a nutshell, we are drawing attention to the fact that each real number z has linked to it a specific real number y, which can be calculated according to the rule (or law of correspondence) given by the formula. We have here a process which, applied to any real number z, does something to it (squares it) to produce another real number y (the square of x). Similarly, =x—32 and y= (e?+1)7 are two other simple functions of the real variable 2, and each is given by a rule in the form of an algebraic expression which specifies the exact manner in which the value of y depends on the value of z.Sets and Functions = 15 The rules for the functions we have just mentioned are expressed by formulas. In general, this is possible only for functions of a very simple kind or for those which are sufficiently important to deserve special symbols of their own. Consider, for instance, the function of the real variable x defined as follows: for each real number z, write x as an infinite decimal (using the scheme of decimal expansion in which infinite chains of 9’s are avoided—in which, for example, 14 is represented by .25000 . . . rather than by .24999 . ; then let y be the fifty-ninth digit after the decimal point. There is of course no standard formula for this, but nevertheless it is a perfectly respectable function whose rule is given by a verbal description. On the other hand, the function y = sin z of the real variable x is so important that its rule, though fully as compli- cated as the one just defined, is assigned the special symbol sin. When discussing functions in general, we want to allow for all sorts of rules and to talk about them all at once, so we usually employ noncommittal notations like y = f(x), y = g(x), and so on. Each of the functions mentioned above is defined for all real numbers x. The example y = 1/z shows that this restriction is much too severe, for this function is defined only for non-zero values of z. Similarly, y = log z is defined only for positive values of x, and y = sin-! z only for values of x which lie in the interval [—1,1]. Whatever our conception of a function may be, it should certainly be broad enough to include examples like these, which are defined only for some values of the real variable x. In real analysis the notion of function is introduced in the following way. Let X be any non-empty set of real numbers. We say that a function y = f(z) is defined on X if the rule f associates a definite real number y with each real number z in X. The specific nature of the tule f is totally irrelevant to the concept of a function. The set X is called the domain of the given function, and the set Y of all the values it assumes is called its range. If we speak of complex numbers here instead of real numbers, we have the notion of function as it is used in complex analysis. This point of view toward functions is actually a bit more general than is needed for the aims of analysis, but it isn’t nearly general enough for our purposes. The sets X and Y above were taken to be sets of numbers. If we now remove even this restriction and allow X and Y to be completely arbitrary non-empty sets, then we arrive at the most inclusive concept of a function. By way of illustration, suppose that X is the set of all squares in a plane and that Y is the set of all circles in the same plane. We can define a function y = f(x) by requiring that the rule f associate with each square z that circle y which is inscribed in it. In general, there is no need at all for either X or Y to be a set of16 Topology numbers. All that is really necessary for a function is two non-empty sets X and Y and a rule f which is meaningful and unambiguous in assigning to each element x in X a specific element y in Y. With these preliminary descriptive remarks, we now turn to the rather abstract but very precise ideas they are intended to motivate. A function consists of three objects: two non-empty sets X and Y (which may be equal, but need not be) and a rule f which assigns to each element x in X a single fully determined element y in Y. The y which corresponds in this way to a given z is usually written f(x), and is called the image of x under the rule f, or the value of f at the element z. This f Fig. 6. A way of visualizing mappings. notation is supposed to be suggestive of the idea that the rule f takes the element z and does something to it to produce the element y = f(z). The rule f is often called a mapping, or transformation, or operator, to amplify this concept of it. We then think of f as mapping 2’s to y’s, or transforming z’s into y’s, or operating on 2’s to produce y’s. The set X is called the domain of the function, and the set of all f(z)’s for all z’s in X is called its range. A function whose range consists of just one element is called a constant function. We often denote by f:X — Y the function with rule f, domain X, and range contained in Y. This notation is useful because the essential parts of the function are displayed in a manner which emphasizes that it is a composite object, the central thing being the rule or mapping f. Figure 6 gives a convenient way of picturing this function. On the left, X and Y are different sets, and on the right, they are equal—in which case we usually refer to f as a mapping of X into itself. If it is clear from the context what the sets X and ¥Y are, or if there is no real need to specify them explicitly, it is common practice to identify the function :X — Y with the rule f, and to speak of f alone as if it were the function under consideration (without mentioning the sets X and Y). It sometimes happens that two perfectly definite sets X and Y are under discussion and that a mapping of X into Y arises which has nu natural symbol attached to it. If there is no necessity to invent @Sets and Functions = 17 symbol for this mapping, and if it is quite clear what the mapping is, it is often convenient to designate it by x— y. Accordingly, the function y = 2? mentioned at the beginning of this section can be written as x— x* or z— y (where y is understood to be the square of z). A function f is called an eztension of a function g (and g is called a restriction of f) if the domain of f contains the domain of g and f(z) = g(x) for each x in the domain of g. Most of mathematical analysis, both classical and modern, deals with functions whose values are real numbers or complex numbers. ~~ Fig. 7. The inverse of a mapping. This is also true of those parts of topology which are concerned with the foundations of analysis. If the range of a function consists of real numbers, we call it a real function; similarly, a complex function is one whose range consists of complex numbers. Obviously, every real function is also complex. We lay very heavy emphasis on real and complex func- tions throughout our work. As a matter of usage, we generally prefer to reserve the term function for real or complex functions and to speak of mappings when dealing with functions whose values are not necessarily numbers. Consider a mapping f:X — Y. When we call f a mapping of X into Y, we mean to suggest by this that the elements f(x)—as z varies over all the elements of X—need not fill up Y; but if it definitely does happen that the range of f equals Y, or if we specifically want to assume this, then we call f a mapping of X onto Y. If two different elements in X always have different images under f, then we call f a one-to-one mapping of X into Y. If f:X — Y is both onto and one-to-one, then we can define its inverse mapping f-!:Y + X as follows: for each y in Y, we find that unique element z in X such that f(x) = y (z exists and is unique since f is onto and one-to-one); we then define z to be f-'(y). The equation z = f-"(y) is the result of solving y = f(z) for z in just the same way as x = log y is the result of solving y = e* for x. Figure 7 illustrates the concept of the inverse of a mapping.18 Topology If f is a one-to-one mapping of X onto Y, it will sometimes be con- venient to subordinate the conception of f as a mapping sending 2’s over to y’s and to emphasize its role as a link between 2’s and y’s. Each z has linked to it (or has corresponding to it) precisely one y = f(z); and, turning the situation around, each y has linked to it (or has corre- sponding to it) exactly one z = f(y). When we focus our attention on this aspect of a mapping which is one-to-one onto, we usually call it a one-to-one correspondence. Thus f is a one-to-one correspondence between X and Y, and f-! is a one-to-one correspondence between Y and X. Now consider an arbitrary mapping f:X—> Y. The mapping f, which sends each element of X over to an element of Y, induces the following two important set mappings. If A is a subset of X, then its image f(A) is the subset of Y defined by S(A) = {f@):xe A}, and our first set mapping is that which sends each A over to its corre- sponding f(A). Similarly, if B is a subset of Y, then its inverse image J-1(B) is the subset of X defined by JB) = {x:f(@) € By, and the second set mapping pulls each B back to its corresponding J-(B). It is often essential for us to know how these set mappings behave with respect to set inclusion and operations on sets. We develop most of their significant features in the following two paragraphs. The main properties of the first set mapping are: S90) =9; SX) CY; Ar S As=f(Ai) C f(A»); HUA) = Uifl(Ad; qd) SOWA) S OG(Ad)- The reader should convince himself of the truth of these statements. For instance, to prove (1) we would have to prove first that f(U;A,) isa subset of Usf(Ai), and second that U,f(A,) is a subset of f(U:A,). A proof of the first of these set inclusions might run as follows: an element in f(U,A,) is the image of some element in U;A;, therefore it is the image of an element in some Aj, therefore it is in some f(A,), and so finally it is in Uyf(Ad. The irregularities and gaps which the reader will notice in the above statements are essential features of this set mapping. For exam- ple, the image of an intersection need not equal the intersection of the images, because two disjoint sets can easily have images which are not disjoint. Furthermore, without special assumptions (see Problem 6) nothing can be said about the relation between f(A)’ and f(A’).Sets and Functions Wv The second set mapping is much better behaved. Its properties are satisfyingly complete, and can be stated as follows: f1@) = 9; fr) = xX; B,C Br >f-(Bi) C f-*(Bs); SMB) = Uif(Bd; (2) SOB) = OS (Bi); (3) SB) = PBy. (4) Again, the reader should verify each of these statements for himself. Fig. 8. Multiplication of mappings. We discuss one more concept in this section, that of the multiplication (or composition) of mappings. If y = f(z) = 2? + 1 and z=gy) =siny, then these two functions can be put together to form a single function defined by z = (gf)(z) = g(f(z)) = g(a? + 1) = sin (2? +1). One of the most important tools of calculus (the chain rule) explains how to dif- ferentiate functions of this kind. This manner of multiplying functions together is of basic importance for us as well, and we formulate it in general as follows. Suppose that f:X — Y and g:Y — Z are any two mappings. We define the product of these mappings, denoted by gf:X — Z, by (gf)(z) = g(f(z)). In words: an element z in X is taken by f to the element f(x) in Y, and then g maps f(z) to g(f(x)) in Z. Figure 8isa picture of this process. We observe that the two mappings involved here are not entirely arbitrary, for the set Y which contains the range of the first equals the domain of the second. More generally, the product of two mappings is meaningful whenever the range of the first is con- tained in the domain of the second. We have regarded f as the first mapping and g as the second, and in forming their product gf, their symbols have gotten turned around. This is a rather unpleasant phenomenon, for which we blame the occasional perversity of mathe- matical symbols. Perhaps it will help the reader to keep this straight20 Topology in his mind if he will remember to read the product gf from right to left: first apply f, then g. Problems 1. Two mappings f:X + Y and g:X — Y are said to be equal (and we write this f = 9) if f(z) = g(z) for every z in X. Let f, g, and h be any three mappings of a non-empty set X into itself, and show that multiplication of mappings is associative in the sense that S(gh) = (foyh. 2. Tet X be a non-empty set. The identity mapping ix on X is the mapping of X onto itself defined by ix(z) = 2 for every z. ‘Thus ix sends each element of X to itself; that is, it leaves fixed each element of X. Show that fix = ixf = f for any mapping f of X into itself. If f is one-to-one onto, so that its inverse f-! exists, show that Jf) = f/f = ix. Show further that f- is the only mapping of X into itself which has this property; that is, show that if g is a mapping of X into itself such that fg = gf = ix, then g =f-) (hint: 9 = gix = gf) = (off = ixf = J, or 9 = ixg = (Fg = Sg) = Stix =f). 3. Let X and Y be non-empty sets and fa mapping of X into Y. Show the following: (a) f is one-to-one <= there exists a mapping g of Y into X such that gf = ix; (b) f is onto © there exists a mapping h of Y into X such that fh = ty. 4, Let X be a non-empty set and f a mapping of X into itself. Show that f is one-to-one onto © there exists a mapping g of X into itself such that fg = gf = ix. If there exists a mapping g with this property, then there is only one such mapping. Why? 5, Let X be a non-empty set, and let f and g be one-to-one mappings of X onto itself. Show that fg is also a one-to-one mapping of X onto itself and that (fg)-! = g-f, 6. Let X and Y be non-empty sets and f a mapping of X into Y. If A and B are, respectively, subsets of X and Y, show the following: (a) ff-(B) © B, and ff-"(B) = B is true for all B & f is onto; (0) A Cf-¥(A), and A = f-Y¥(A) is true for all A © fis one-to-one; (©) (AiO Az) = f(Ay) Of(A2) is true for all Ai and Arf is one-to-one; (a) (AY Cf(A’) is true for all A & f is onto; (e) if f is onto—so that f(A)’ C f(A’) is true for all A—then J(A)’ = f(A’) is true for all A © is also one-to-one.Sets and Functions 21 4, PRODUCTS OF SETS We shall often have occasion to weld together the sets of a given class into a single new set called their product (or their Cartesian product). The ancestor of this concept is the coordinate plane of analytic geometry, that is, a plane equipped with the usual rectangular coordinate system. We give a brief description of this fundamental idea with a view to paving the way for our discussion of products of sets in general. First, a few preliminary comments about the real line. We have already used this term several times without any explanation, and of course what we mean by it is an ordinary geometric straight line (see Fig. 9) whose points have been identified with—or coordinatized by—the V3 1 v2 er 1 0 1 2 3 -3 Fig. 9. The real line. set R of all real numbers. We use the letter R to denote the real line as well as the set of all real numbers, and we often speak of real numbers as if they were points on the real line, and of points on the real line as if they were real numbers. Let no one be deceived into thinking that the real line is a simple thing, for its structure is exceedingly intricate. Our present view of it, however, is as naive and uncomplicated as the picture of it given in Fig. 9. Generally speaking, we assume that the reader is familiar with the simpler properties of the real line—those relating to inequalities (see Problem 1-2) and the basic algebraic operations of addition, subtraction, multiplication, and division. One of the most significant facts about the real number system is perhaps less well known. This is the so-called least upper bound property, which asserts that every non-empty set of real numbers which has an upper bound has a least upper bound. It is an easy consequence of this that every non- empty set of real numbers which has a lower bound has a greatest lower bound. All these matters can be developed rigorously on the basis of a small number of axioms, and detailed treatments can often be found in books on elementary abstract elgebra. To construct the coordinate plane, we now proceed as follows. We take two identical replicas of the real line, which we call the x azis and the y avis, and paste them on a plane at right angles to one another in such a way that they cross at the zero point on each. The usual picture is given in Fig. 10. Now let P be a point in the plane. We project P perpendicularly onto points P, and P, on the axes. If z and y are the coordinates of P, and P, on their respective axes, this process22 Topology leads us from the point P to the uniquely determined ordered pair (z,y) of real numbers, where x and y are called the x coordinate and y coordi- nate of P. We can reverse the process, and, starting with the ordered pair of real numbers, we can recapture the point. This is the manner in which we establish the familiar one-to-one correspondence between points P in the plane and ordered pairs (2,y) of real numbers. In fact, we think of a point in the plane (which is a geometric object) and its corresponding ordered pair of real numbers (which is an algebraic object) as being—to all intents and purposes—identical with one another. The essence of analytic geometry lies y axis in the possibility of exploiting this identification by using algebraic tools in geometric arguments and giving geometric interpretations to Pay) algebraic calculations. The conventional attitude to- ward the coordinate plane in ana- lytic geometry is that the geometry is the focus of interest and the alge- bra of ordered pairs is only a con- «axis venient tool. Here we reverse this point of view. For us, the coordinate Fig. 10. The coordinate plane. plane is defined to be the set of all ordered pairs (z,y) of real numbers. We can satisfy our desire for visual images by using Fig. 10 as a picture of this set and by calling such an ordered pair a point, but this geo- metric language is more a convenience than a necessity. Our notation for the coordinate plane is R XR, or R*. This symbolism reflects the idea that the coordinate plane is the result of “multiplying together” two replicas of the real line R. It is perhaps necessary to comment on one possible source of mis- understanding. When we speak of R? as a plane, we do so only to establish an intuitive bond with the reader’s previous experience in ana- lytic geometry. Our present: attitude is that R? is a pure set and has no structure whatever, because no structure has yet been assigned to it. We remarked earlier (with deliberate vagueness) that a space is a set to which has been added some kind of algebraic or geometric structure. In Sec. 15 we shall convert the set 2? into the space of analytic geometry by defining the distance between any two points (,y1) and (x2,y2) to be Ve = 22)® + = y2)* This notion of distance endows the set /¢? with a certain “spatial” char- acter, which we shall recognize by calling the resulting space the Euclidean plane instead of the coordinate plane.Sets and Functions 23 We assume that the reader is fully acquainted with the way in which the set C of all complex numbers can be identified (as a set) with the coordinate plane R%. If z is a complex number, and if z has the standard form x + iy where z and y are real numbers, then we identify z with the ordered pair (x,y), and thus with an element of R?. The complex numbers, however, are much more than merely a set. ‘They constitute a number system, with operations of addition, multi- plication, conjugation, etc. When the coordinate plane R? is thought of as consisting of complex numbers and is enriched by the algebraic structure it acquires in this way, it is called the complex plane. The y,| letter C is used to denote either the set of all complex numbers or the complex plane. Weshallmake *? a space out of the complex plane in Sec. 9. Suppose now that X, and X2 are any two non-empty sets. By analogy with our above discussion, their product X, X X» is defined to A x be the set of allordered pairs (21,22), Fig. 11. A way of visualizing X; X Xx. where x, is in X, and 2, is in X2. In spite of the arbitrary nature of X; and X», their product can be repre- sented by a picture (see Fig. 11) which is loosely similar to the usual picture of the coordinate plane. The term product is applied to this set, and it is thought of as the result of “multiplying together” X; and X», for the following reason: if X, and X; are finite sets with m and n elements, then (clearly) X: X X» has mn elements. If f:X:—> X» is a mapping with domain X; and range in Xz, its graph is that subset of X, X Xz which consists of all ordered pairs of the form (a1,f(z1)). We observe that this is an appropriate generalization of the concept of the graph of a function as it occurs in elementary mathematics. This definition of the product of two sets extends easily to the case of n sets for any positive integer n. If Xi, Xx, ... , X, are non-empty sets, then their product X: X Xz: X +++ X Xn is the set of all ordered n-tuples (x1, %2, . . . , Zn), Where 2; is in X, for each subscript 7. If the Xs are all replicas of a single set X, that is, if XioX= °° =X, =X, then their product is usually denoted by the symbol X*. These ideas specialize directly to yield the important sets R* and C*. R'is just R, the real line, and R? is the coordinate plane. R*—the set of all ordered triples of real numbers—is the set which underlies solid analytic geometry, and we assume that the reader is familiar with24 Topology the manner in which this set arises, through the introduction of a rec- tangular coordinate system into ordinary three-dimensional space. We can draw pictures here just as in the case of the coordinate plane, and we can use geometric language as much as we please, but it must be under- stood that the mathematics of this set is the mathematics of ordered triples of real numbers and that the pictures are merely an aid to the intuition. Once we fully grasp this point of view, there is no difficulty whatever in advancing at once to the study of the set R* of all ordered n-tuples (21, 22, .. . , Za) of real numbers for any positive integer n. It is quite true that when 7 is greater than 3 it is no longer possible to draw the same kinds of intuitively rich pictures, but at worst this is merely an inconvenience. We can (and do) continue to use suggestive geometric language, so all is not lost. The set C* is defined similarly: it is the set of all ordered n-tuples (21, 22, . . . , 2n) of complex numbers, Each of the sets R* and C* plays a prominent part in our later work. We emphasized above that for the present the coordinate plane is to be considered as merely a set, and not a space. Similar remarks apply to R* and C". In due course (in Sec. 15) we shall impart form and content to each of these sets by suitable definitions. We shall convert them into the Euclidean and unitary n-spaces which underlie and motivate so many developments in modern pure mathematics, and we shall explore some aspects of their algebraic and topological structure to the very last pages of this book. But as of now—and this is the point we insist on—neither one of these sets has any structure at all. As the réader doubtless suspects, it is not enough that we consider only products of finite classes of sets. The needs of topology compel us to extend these ideas to arbitrary classes of sets. We defined the product X; X Xz: X --- X Xn to be the set of all ordered n-tuples (z;, 22, ..., tn) such that 2; is in X, for each subscript 7. To see how to extend this definition, we reformulate it as follows. We have an index set J, consisting of the integers from 1 to n, and corresponding to each index (or subscript) i we have a non-empty set X;. The n-tuple (21, 22, ... , 2») is simply a function (call it z) defined on the index set J, with the restriction that its value z(¢) = 2; is an element of the set X; for each iin J. Our point of view here is that the function z is completely determined by, and is essentially equivalent to, the array (21, 22, . . . , 2») of its values. The way is now open for the definition of products in their full generality. Let {X;} be a non-empty class of non-empty sets, indexed by the elements 7 of an index set J. The sets X; need not be different from one another; indeed, it may happen that they are all identical replicas of a single set, distinguished only by different indices. The product of the sets Xi, written P,r X,, is defined to be the set of all functions z defined on J such that z(z) is an element of the set X; forSets and Functions 25 each index i. We call X; the ith coordinate set. When there can be no misunderstanding about the index set, the symbol P;, X; isoften abbrevi- ated to P:X;. The definition we have just given requires that each coordinate set be non-empty before the product car be formed. It will be useful if we extend this definition slightly by agreeing that if any of the X,’s are empty, then P,X; is also empty. This approach to the idea of the product of a class of sets, by means of functions defined on the index set, is useful mainly in giving the definition. In practice, it is much more convenient to use the subscript notation 2; instead of the function notation z(i). We then interpret the product P.X, as made up of elements z, each of which is specified by the exhibited array {z;} of its values in the respective coordinate sets X;. We call a; the ith coordinate of the element z = {zi}. The mapping p; of the product P,X; onto its ith coordinate set X; which is defined by p(x) = x,—that is, the mapping whose value at.an arbitrary element of the product is the ith coordinate of that element —is called the projection onto the ith coordinate set. The projection p; selects the ith coordinate of each element in its domain. There is clearly one projection for each element of the index set I, and the set of all projections plays an important role in the general theory of topological spaces. Problems 1, The graph of a mapping f: X — Y is a subset of the product X X Y. What properties characterize the graphs of mappings among all subsets of X X Y? 2. Let X and Y be non-empty sets. If A; and A; are subsets of X, and B, and B; subsets of Y, show the following: (Ar X Bi) A (Az X Br) = (Ar O A2) X (Bi O Ba); (Ay X Bi) — (Az X Ba) = (Ai — Az) X (Bi — Ba) U (AL Az) X (Bi — Bs) U (Ai — As) X (Bi A B:). 3. Let X and Y be non-empty sets, and let A and B be rings of subsets of X and Y, respectively. Show that the class of all finite unions of sets of the form A X B with A ¢ A and B ¢ Bisa ring of subsets of oY: 5. PARTITIONS AND EQUIVALENCE RELATIONS In the first part of this section we consider a non-empty set X, and we study decompositions of X into non-empty subsets which fill it out26 Topology and have no elements in common with one another. We give special attention to the tools (equivalence relations) which are normally used to generate such decompositions. A partition of X is a disjoint class {X,} of non-empty subsets of X whose union is the full set X itself. The X,’s are called the partition sets. Expressed somewhat differently, a partition of X is the result of splitting it, or subdividing it, into non-empty subsets in such a way that each element of X belongs to one and only one of the given subsets. If Xis the set {1, 2, 3, 4,5}, then {1, 3, 5}, {2,4} and {1, 2, 3}, {4,5} are two different partitions of X. If X is the set F of all real numbers, then we can partition X into the set of all rationals and the set of all irrationals, or into the infinitely many closed-open intervals of the form [n, n + 1) where n is an integer. If X is the set of all points in the coordinate plane, then we can partition X in such a way that each partition set consists of all points with the same x coordinate (vertical lines), or so that each partition set consists of all points with the same y coordinate (horizontal lines). Other partitions of each of these sets will readily occur to the reader. In general, there are many different ways in which any given set can be partitioned. These manufactured examples are admittedly rather uninspiring and serve only to make our ideas more concrete. Later in this section we consider some others which are more germane to our present purposes. A binary relation in the set X is a mathematical symbol or verbal phrase, which we denote by R in this paragraph, such that for each ordered pair (z,y) of elements of X the statement z Ry is meaningful, in the sense that it can be classified definitely as true or false. For such a binary relation, z R y symbolizes the assertion that x is related by R to y, and x Ry the negation of this, namely, the assertion that x is not related by R to y Many examples of binary relations can be given, some familiar and others less so, some mathematical and others not. For instance, if X is the set of all integers and R is interpreted to mean “4s less than,” which of course is usually denoted by the symbol <, then we clearly have 4 <7 and 5< 2. We have been speaking of binary relations, which are so named because they apply only to ordered pairs of elements, rather than to ordered triples, etc. In our work we drop the qualifying adjective and speak simply of a relation in X, since we shall have occasion to consider only relations of this kind.' We now assume that a partition of our non-empty set X is given, 1 Some writers prefer to regard a relation R in X as a subset Rof X X X. From this point of view, z Ry and z Ity are simply equivalent ways of writing (z,y)e R and (z,y)¢R. This definition has the advantage of being more tangible than ours, and the disadvantage that few people really think of a relation in this way.Sets and Functions 7 and we associate with this partition a relation in X. This relation is defined in the following way: we say that z is equivalent to y and write this 2 ~ y (the symbol ~ is pronounced “wiggle”), if z and y belong to the same partition set. It is obvious that the relation ~ has the follow- ing properties: (1) 2~ 2 for every z (reflexivity); (2) 2~y=y~z (symmetry); (3) r~yandy~z=2 ~2z (transitivity). This particular relation in X arose in a special way, in connection with a given partition of X, and its properties are immediate consequences of its definition. Any relation whatever in X which possesses these three properties is called an equivalence relation in X. We have just seen that each partition of X has associated with it a natural equivalence relation in X. We now reverse the situation and show that a given equivalence relation in X determines a natural partition of X. Let ~ be an equivalence relation in X; that is, assume that it is reflexive, symmetric, and transitive in the sense described above. If z is an element of X, the subset of X defined by [2] = {y:y ~ 2} is called the equivalence set of x. The equivalence set of z is thus the set of all elements which are equivalent to z. We show that the class of all distinct equivalence sets forms a partition of X. By reflexivity, x ¢ [2] for each element 2 in X, so each equivalence set is non-empty and their union is X. It remains to be shown that any two equivalence sets [2] and [z.] are either disjoint or identical. We prove this by showing that if [x1] and [z2] are not disjoint, then they must be identical. Sup- pose that [21] and [2s] are not disjoint; that is, suppose that they have a common clement z. Since z belongs to both equivalence sets, 2 ~ 21 and z~ 22, and by symmetry, z;~2z. Let y be any element of [21], so that y~z. Since y~2; and x ~z, transitivity shows that y~z. By another application of transitivity, y ~ z and z ~ 2» imply that y ~ 2», so that yisin [22]. Since y was chosen arbitrarily in [21], we see by this that [x1] C [x]. The same reasoning shows that [22] C [1], and from this we conclude (sce the last paragraph of Sec. 1) that [21] = [2a]. The above discussion demonstrates that there is no real distinction (other than a difference in language) between partitions of a set and equivalence relations in the set. If we start with a partition, we get an equivalence relation by regarding elements as equivalent if they belong to the same partition set, and if we start with an equivalence relation, we get a partition by grouping together into subsets all elements which are equivalent to one another. We have here a single mathematical idea, which we have been considering from two different points of view, and the approach we choose in any particular application depends entirely28 ‘Topology on our own convenience. In practice, it is almost invariably the case that we use equivalence relations (which are usually easy to define) to obtain partitions (which are sometimes difficult to describe fully). We now turn to several of the more important simple examples of equivalence relations. Let I be the set of all integers. If a and 6 are elements of this set, we write a = b (and say that a equals 6) if a and b are the same integer. Thus 2+ 3 = 5 means that the expressions on the left and right are simply different ways of writing the same integer. It is apparent that = used in this sense is an equivalence relation in the set J: (1) a =a for every a; (2) a=b=>b=a; (83) a=bandb=c>a=c. Clearly, each equivalence set consists of precisely one integer. Another familiar example is the relation of equality commonly used for fractions. We remind the reader that, strictly speaking, a fraction is merely a symbol of the form a/b, where a and 6 are integers and b is not zero. The fractions 24 and 4 are obviously not identical, but nevertheless we consider them to be equal. In general, we say that two fractions a/b and ¢/d are equal, written a/b = c/d, if ad and be are equal as integers in the usual sense (see the above paragraph). We leave it to the reader to show that this is an equivalence relation in the set of all fractions. An equivalence set of fractions is what we call a rational number. Everyday usage ignores the distinction between fractions and rational numbers, but it is important to recognize that from the strict point of view it is the rational numbers (and not the fractions) which form part of the real number system. Our final example has a deeper significance, for it provides us with the basic tool for our work of the next two sections. For the remainder of this section we consider a relation between pairs of non-empty sets, and each set mentioned (whether we say so explicitly or not) is assumed to be non-empty. If X and Y are two sets, we say that X is numerically equivalent to Y if there exists a one-to- one correspondence between X and Y, ie., if there exists a one-to-one mapping of X onto Y. This relation is reflexive, since the identity mapping 7x: X — X is one-to-one onto; it is symmetric, since if f:X + Y is one-to-one onto, then its inverse mapping f~': Y — X is also one-to-one onto; and it is transitive, since if f:X — Y and g:Y — Z are one-to-one onto, then gf:X — Z is also one-to-one onto. Numerical equivalence has all the properties of an equivalence relation, and if we consider it as an equivalence relation in the class of all non-empty subsets of some universal set U, it groups together into equivalence sets all those subsets of U which have the same number of elements. After we state and prove the