>NEW MATHEMATICAL LIBRARY PUBLISHED BY ‘Tue MATHEMATICAL ASSOCIATION OF AMERICA Editorial Committee Basil Gordon, Chairman (1975-76) Anneli Lax, Editor University of California, L.A. New York University Ivan Niven (1975-77) University of Oregon M. M. Schiffer (1975-77) Stanford University The New Mathematical Library (NML) was begun in 1961 by the School Mathematics Study Group to make available to high school students short expository books on various topics not usually covered in the high school syllabus, In a decade the NML matured into a steadily growing series of some twenty titles of interest not only to the originally intended audience, but to college students and teachers at all levels, Previously published by Random House and L. W. Singer, the NML became a publication series of the Mathematical Association of America (MAA) in 1975. Under the auspices of the MAA the NML will continue to grow and will remain dedicated to its original and expanded purposes.GROUPS AND THEIR GRAPHS by Israel Grossman Albert Leonard Junior High School and Wilhelm Magnus New York University 14 THE MATHEMATICAL ASSOCIATION OF AMERICAIlustrations by Carl Bass ‘©Copyright 1964, 1992 by The Mathematical Association of America Alll rights reserved under International and Pan-American Copyright Conventions Published in Washington, D.C. by The Mathematical Association of America Library of Congress Catalog Card Number: 64-8512 Print ISBN 978-0-88385-614-7 Electronic ISBN 978-0-88385-929-2 ‘Manufactured in the United States of AmericaANNELI LAX NEW MATHEMATICAL LIBRARY 1. Numbers: Rational and Irrational by Ivan Niven 2, What is Calculus About? by W. W. Sawyer 3. An Introduction to Inequalities by E. F. Beckenbach and R. Bellman 4, Geometric Inequalities by N. D. Kazarinoff 5. The Contest Problem Book I Annual High School Mathematics Examinations 1950-1960, Compiled and with solutions by Charles T. Salkind 6. The Lore of Large Numbers by P. J. Davis 7. Uses of Infinity by Leo Zippin 8 Geometric Transformations I by J. M. Yaglom, translated by A. Shields 9. Continued Fractions by Car! D, Olds Replaced by NML-34 Hungarian Problem Books I and Il, Based on the Etvis Competitions 1894-1905 and 1906-1928, translated by E. Rapaport Episodes from the Early History of Mathematics by A. Aaboe Groups and Their Graphs by £, Grossman and W. Magnus The Mathematics of Choice by Ivan Niven From Pythagoras to Einstein by K. O. Friedrichs The Contest Problem Book II Annual High School Mathematics Examinations 1961-1965. Compiled and with solutions by Charles T. Salkind 18. First Concepts of Topology by W. G. Chinn and N. E. Steenrod 19. Geometry Revisited by H. S. M. Coxeter and S. L. Greitzer 20. Invitation to Number Theory by Oystein Ore 21. Geometric Transformations Il by JM. Yaglom, translated by A. Shields 22, Elementary Cryptanalysis—A Mathematical Approach by 4. Sinkov 23. Ingenuity in Mathematics by Rass Honsberger 24. Geometric Transformations III by I. M. Yaglom, translated by A. Shenitzer 25. The Contest Problem Book II Annual High School Mathematics Examinations 1966-1972, Compiled and with solutions by C. T. Salkind and J. M. Earl 26. Mathematical Methods in Science by George Pélya 27. International Mathematical Olympiads—1959-1977. Compiled and with solutions by S. L. Greitzer 28. The Mathematics of Games and Gambling by Edward W. Packel 29, The Contest Problem Book IV Annual High School Mathematics Examinations 1973-1982. Compiled and with solutions by R. A. Artino, A. M. Gaglione, and N. Shell 30, The Role of Mathematics in Science by M. M. Schiffer and L. Bowden 31. International Mathematical Olympiads 1978-1985 and forty supplementary problems. Compiled and with solutions by Murray S. Klamkin 32. Riddles of the Sphinx by Martin Gardner 33, U.S.A. Mathematical Olympiads 1972-1986. Compiled and with solutions by Murray S. Klamkin 34, Graphs and Their Uses by Oystein Ore. Revised and updated by Robin J. Wilson 35, Exploring Mathematics with Your Computer by Arthur Engel 36. Game Theory and Strategy by Philip D. Straffin, Jr37. 38. 39. 41 42, Episodes in Nineteenth and Twenthieth Century Euclidean Geometry by Ross Honsberger The Contest Problem Book V American High School Mathematics Examinations and American Invitational Mathematics Examinations 1983-1988. Compiled and augmented by George Berzsenyi and Stephen B. Maurer Over and Over Again by Gengzhe Chang and Thomas W. Sederberg The Contest Problem Book VI American High School Mathematics Examinations 1989-1994, Compiled and augmented by Leo J. Schneider The Geometry of Numbers by C. D. Olds, Anneli Lax, and Giuliana P. Davidoff Hungarian Problem Book III Based on the Eétvos Competitions 1929-1943 translated by Andy Liu Other titles in preparation.Preface Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 Chapter 12 Chapter 13 Chapter 14 Chapter 15 Appendix Solutions Bibliography Index Contents Introduction to Groups Group Axioms Examples of Groups Multiplication Table of a Group Generators of a Group Graph of a Group Definition of a Group by Generators and Relations Subgroups Mappings Permutation Groups Normal Subgroups The Quaternion Group Symmetric and Alternating Groups Path Groups Groups and Wallpaper Designs Group of the Dodecahedron and the Icosahedron 10 26 41 56 7 89 107 120 137 141 150 160 167 170 189 192Preface Asstudent in the primary or secondary schools frequently has the no- tion that mathematics is concerned solely with number and measure. However, mathematics has always been much more than merely a quan- titative science with applications to activities such as bookkeeping and money-changing; it is deeply concerned with logic and structure. ‘The theory of groups is one of the important non-quantitative branches of mathematics. The concept of a group, although comparatively recent in the development of mathematics, has been most fruitful; for example, it has been a powerful tool in the investigation of algebraic equations, of geometric transformations, and of problems in topology and number theory. ‘Two features of group theory have traditionally made it advisable to postpone its study until rather late in a student’s mathematical educa- tion. First, a high degree of abstractness is inherent in group theoretical ideas, and ability to cope with abstract concepts comes with mathe- matical maturity. Second, the ways in which group theory interacts with other fields of study to illuminate and advance them can be seen only after long and elaborate development of the theory, and then only by students acquainted with the other fields. In this book we have aimed. at a presentation suitable for students at a relatively early stage of mathematical growth. To bypass the difficulties stemming from ab- stractness, we have used geometric pictures of groups—graphs of groups. In this way, abstract groups are made concrete in visual patterns that correspond to group structure. However, we cannot hope to provide a substitute for the protonged reading and study necessary to grasp the concepts of varied mathematical fields of inquiry. We have tried to make the best of this situation by indicating the broader significance of some of the theorems and concepts presented. We acknowledge that we cannot always motivate the reader with “practical” applications. Ultimately, we have to rely on the appeal of the mathematical content in and of itself. Of course, the most effective spur comes from the reader himself; this has to be his contribution. 12 PREFACE We wish to thank the editors of the New Mathematical Library who contributed to this book through their advice. Also, we acknowledge gratefully the technical help given to us by Dr. Anneli Lax and Miss Arlys Stritzel, and the support granted to us by the National Science Foundation.CHAPTER ONE Introduction to Groups A theory of groups first began to take form at the end of the eighteenth century. It developed slowly and.attracted very little notice during the first decades of the nineteenth century. Then, in a few years centering about 1830, the theory of groups took a giant leap forward and made a major contribution to the general development of mathematics in the work of Galois and Abel on the solvability of algebraic equations. Since then, the concepts underlying the theory of groups have been elaborated and extended into many branches of mathematics. There have been applications to such diverse fields as quantum mechanics, crystal- lography, and the theory of knots. This book is concerned with groups and their graphical representation. Our first task is to clarify what is meant by a group. One basic idea that reaches to the very essence of the group concept is the notion of structure, or pattern. In what follows, the reader will sce the unfolding of a succession of examples and explanations, definitions and theorems, all calculated to be variations on one fundamental theme: how groups and their graphs embody and illustrate one kind of mathe- matical structure. So far, we have been using the word “group” without giving the reader any idea of what the word means. To present a complete formal defini- tion at one fell swoop might leave the reader as mystified as he was to start with. We shall therefore develop the concept of a group gradually, and we begin by presenting two examples. It is expected that the reader will keep these in mind during the following introductory discussion of the structural features of a group.4 GROUPS AND THEIR GRAPHS Group A: The se¢ of all integers considered as numbers that can be added to each other. In other words, the elements of Group A are the integers {-++, —3, —2, —1, 0, 1, 2, 3, +++}, and the only operation we are interested in performing is that of addition of any fwo elements of the set; for example, 2+ 5 = 7. Group B: The sel of all positive rational numbers considered as numbers that can be mulliplied by each other. In this case, the elements of the sef are all numbers that can be represented in the form a/b, where a and & are positive integers, and the only operation we are interested in performing is that of multiplication of any two elements of the sel; for example, 3- $= +. Now that the reader has been exposed to examples of a group, he may still not be much further along the road to understanding what a group is, since he may not immediately recognize which features of these ex- amples are significant in the essential structure of a group. In presenting the descriptions of Groups A and B, certain words were italicized to stress a basic structural pattern present in all groups. We can isolate two features. Group A: all integers 1. A set of elements Group B: all positive rational numbers Group A: addition of any two integers 2. A binary operation on the set) Group B; multiplication of any two positive rational numbers We called the operations in Groups A and B binary operations because each involves fwo elements at a time. A binary operation on a sel is a correspondence that assigns to each ordered pair of elements of the set a uniquely determined element of the set. Thus, in Group A, addition is a binary operation on the set of integers, for, if r and s are any two elements of our set then r + s is also an element of the set, If we denote the element r + s by the symbol f, we can rephrase our description in this way. If r and s are any two ele- ments of our set then there is one, and only one, element # of the setINTRODUCTION 5 such that r+ s = 4, For example, if we select 2 and 5-as two elements of our set, there is the unigue set element 7 such that 2 + 5 = 7. Multiplication is the binary operation of Group B; for, if r and s are any two elements of our set (positive rationals) then there is one, and only one, element ¢ of the set such that r+ s = ¢, (The uniqueness of the element / follows from the understanding that equivalent rationals such as 4 and 4 represent the same number.) If we choose 3 and § as two elements of our set, there exists a unique element 75 of the set such that $- $= ry. Notice that the concept “binary operation” involves an associated set. That is why we have used the words “binary operation on a se”. The pair of elements and the corresponding element assigned by the binary operation must all be elements of the same set. Thus, we see that two related features of a group are (1) a set of elements and (2) a binary operation on this set. These two features are indissolubly intertwined and cannot be separated, though we may sometimes find it convenient to shift our focus from one to the other. The examples of group binary operations that we have considered so far are ordinary addition of integers, denoted by the symbol +, and multiplication of positive rational numbers, denoted by +. We shall see that there are many different binary operations associated with different groups, and it will sometimes be convenient to use a single symbol for any of these. We shall let the symbol @ denote a binary operation that is unspecified. This notation enables us to describe the structural features (1) and (2) exhibited by Groups A and B as a set S together with a binary operation @ on S. If r and s are any two elements of S, then there isa unique element ¢ in S$ such that r@s=t. For Group A, ® denotes the specific operation “addition of integers”; for Group B, ® denotes “multiplication of positive rational numbers”. To stress the idea that a binary operation is a correspondence, we can describe the groups we have been examining in yet another way. In the case of Group A, we can say that corresponding to any pair of integers r and s, there is a unique integer #. In symbols, we can write (ys) 4 where the arrow denotes “corresponds to”. In the case of Group B, we can say that corresponding to any pair of positive rational numbers r and s, there isa unique positive rational ¢.6 GROUPS AND THEIR GRAPHS To gain a broader view of a binary operation on a set, we shall consider this question: can a binary operation on a set also be a binary operation. on a subsel? (We say that set U is a subset of set S if every element of U is also an element of S.) For example, suppose S is the set of all positive rational numbers, and U is the subset consisting of all positive integers. Let us first determine whether division is a binary operation on S. The reader can readily satisfy himself that division is a binary operation on. the set S of positive rational numbers. If 7 and s are any two positive rational numbers, there exists a unique positive rational number ¢ such that res=h Now, let us examine whether division, a binary operation on the set SS, is also a binary operation on the subset U of positive integers. It is evident that if we choose, say, 2 and 3 as two elements of our subset U, then there does nof exist any positive integer ¢ such that 2+35t Division, then, is xo! a binary operation on the subset U of positive inte- gers, since there are pairs of positive integers that do #of correspond to a third positive integer. In contrast to this situation, let us now consider the set S of all inte- gers and the subset U of all even integers. We have seen that addition is a binary operation on the set S of all integers. What happens on the subset U of even integers under the operation of addition? When two even integers are added, the result is an even integer. In other words, addition is a binary operation on the subset U of even integers as well as on the set S of all integers. Whenever two elements of the subset U are added, the sum is always an element of U. This property is de- scribed by saying: the subset U of even integers is closed under the binary operation of addition. The reader can verify that the subset T of odd integers is not closed under this operation. In more general terms, we describe the closure property of a subset under a binary operation in this way: if @ is a binary operation on a set S, and if U is a subset of S with the property that 4 @ » is an element of U whenever u and » are in the subset U, we say that U is closed under the operation @. The word “closed” suggests that the operation ®, when restricted to pairs of elements in U, does not take us out of U; hence we may think of @ as a binary operation on the set U. We shall see in Chapter 8 how this closure property of a subset under a binary operation plays a central role in the discussion of “subgroups”.INTRODUCTION 7 Exercise 1: (a) Is addition a binary operation on the set of odd positive integers? (b) For the same set as in (a), is multiplication a binary operation? (c) Let the elements of the set be 1, 4, —1, —i, where i = . Is addition a binary operation on this set? (d) For the same set as in (c), is multiplica- tion a binary operation? So far we have seen that a group is a set logether with a binary operation on the set. If r and s are any two elements of the set, there exists a unique element 1 of the set such that r@s=t or (7,8) >t The wording “if r and s are any two elements of the set” does not exclude the possibility that r and s denote the same element; nor does it presuppose any specific ordering of r and s. Thus, if r and s are any two elements of the set, then 7@s, r@r, s@s, s@r are also elements of the set (not necessarily all distinct). “ ‘The question now arises: in a group, can r @ s and s @ r ever be different elements of the set? For Groups A and B, it is clear that it is always true that r @ s = s @ r. For example, in Group A we have 3+5=5+3, and in Group B, 3- $= $- %. But on the set of positive rational numbers with division as the binary operation, we see, for example, that + | 4 J+ 3. In general, r@s#5@r for this set. Thus, the order of the elements is significant; in some sets, inter- changing or commuting the elements can lead to different results, i.e., is possible that (a,b)+¢ and (ba) 4d, where a, 6, c,d are elements of a group and ¢ # d. Incase 7 ® s = $@r, we say the elements r and s commute (with respect to the specific operation denoted by @); if r@s ~s@r, we say the elements 7 and s do not commule (with respect to the specific operation). From now on, we must not take it for granted in advance that under ® the ordered pair (r,s) corresponds to the same element as the ordered pair (s, 7). Each situation must be examined separately for commutativity. Taking account of the need, in general, to distinguish between r @ 5 and s @ , we restate our characterization of a set with an associated binary operation in this way: for every ordered pair of elements r and8 GROUPS AND THEIR GRAPHS S of our set, there exists a unique element ¢ of the set such that r@s=t or, (r,s) 4 So far, all examples of sets with their associated binary operations have involved numbers as elements and one of the familiar operations of arithmetic as the binary operation. But we shall see that the elements of a group can also be non-numerical entities such as motions, permutations, functions, geometric transformations, or a set of symbols; and in these cases the associated binary operation is not arithmetic in nature. 2 3 1 2 Rotation 90° clockwise into + Figure 1.1 For example, consider a square which is free to rotate in its plane about an axis through its center, with the restriction that the only per- missible rotations are those that bring the square into coincidence with itself. Then one permissible rotation would be through an angle of 90° in a clockwise direction. (See Figure 1.1.) Let us designate this rotation by a. Some other possible rotations could be: (1) a clockwise rotation of 180°, which we denote by 5; (2) a clockwise rotation of 270°, which we denote by c. We can view these rotations a, b, and ¢ as: possible elements of a group. Can we define a binary operation so that ¢ @ 6 = ¢ makes sense? One way to do so is to think along these lines: A clockwise rotation 90° followed by a clockwise rotation 180° is equivalent to a clockwise rotation of 270°, or clement a followed by element 6 equals element c, or a@b The operation that associates the two elements a and } with element cis “followed by”. This operation of succession makes sense for rotations.INTRODUCTION 9 It will be seen that it can also make sense for other kinds of possible group elements. Exercise 2: With this conception of the binary operation as “followed by”, or succession, what element of the set of rotations of the square does 8 @c represent? What rotation does a @ ¢ represent?CHAPTER TWO Group Axioms Although we have so far concentrated on discussing the concept of binary operation on a set. the reader should not conclude that this is the single defining characteristic of a group. For a set with a binary operation to constitute a group, we postulate that the binary operation possesses certain properties in relation to the set elements. Such postulates, or axioms, describe these basic properties, and we shall need three such axioms. They will be referred to as (1) Associativity, (2) Unit Element (or Identity), (3) Inverses. Associativily. The associative property requires that if 7, s, are any three set elements, then 7® (s@!) = (r@s) OF; that is, if s @ ¢ is element x of the set and r @ s is element y, then r@x=y@l. Let us consider Groups A and B (p. 4). In Group A, the associative property requires that for any three integers 1. s, ¢, r+(sti=(rts) th For example, we have S+ (348) =5+11= 16 and (+3) +8=8+8= 16. 10GROUP AXIOMS Mw In the case of Group B, we should have re(ss) = (resheh For example, and We know from our experience with elementary algebra that the binary operations of Groups A and B are associative. However, let us now consider division as a binary operation on the set of positive rationals and test to see if the associative property holds. We have i $+(3+D while (+3) 43 4 so r+(st) #(r45) Division is not an associative binary operation on the set of positive rationals. What meaning, if any, shall we attach to the expression r @ s @ 1 ? If @ denotes a binary operation on a set, how can we use it when ¢hree elements of the set are involved? We can give a definite meaning to the expression r @ s @ ¢ either by inserting parentheses around the first two symbols or around the last two. In the first case the expression would appear as (r ® s) @ #, and in the second case as r ® (s @ t). Since @ isa binary operation on our set, y = (r ® s) and x = (s@#) are elements of our set. Therefore (r ® s) @ i and r ® (s @ é) may each be thought of as involving only two elements of the set, namely y and ¢ in the first case, and r and x in the second. If the binary operation @ is not associative, the elements r @ x and y @ f are, in general, distinct, and the expression r @ s @ / has no unique meaning. For example, in the case of division on the set of posi- tive rationals, the expression $ + 3 + } is ambiguous because (2+ 3)+3=4, and $+ (3+) =% If the binary operation ® is associative, the elements r @ x and y @ ¢ are identical and so it makes no difference which of the two modes of inserting parentheses we adopt. In either case we shall have repre- sentations of the same element. It is precisely because of the associative2 GROUPS AND THEIR GRAPHS property that we can agree that the expressions 7O@s@t, 71r@(s@t, (r@®s\@r all represent the same element. Axiom 1 (Associativity): In a group, a binary operation is defined such that if r,s, and¢ are any three elements, then r@ (sO) = (Os) Ot Is the operation we have described as “followed by” associative? Con- sider a circular disc that can be rotated around an axis through its center —like a bicycle wheel. Suppose a, 6, ¢ are any set of rotations of the disc. Then, if @ denotes the operation “followed by”, or succession of rota- tions, is it always true that (¢ @ 6) @c = a @ (b @c)? It can be seen that the parentheses serve merely as pauses in a steady sequence: first ¢, then 4, then c. The operation is associative for rotations or for any set of motions and is, to that extent, a permissible group operation. Unit element or identity. The two remaining axioms deal with concepts that are extensions of ideas connected with the number t. These axioms seem quite natural if we think of ordinary multiplication as our binary operation. First, we examine the property of the number 1 in multiplica- tion. If » is a number, then n-l=ien=n; that is, the product of m and 1is #. Extending this idea to group ele- ments and a group operation, we arrive at Axiom 2. Axiom 2 (Identity) : There exists a unique group element J such that, for any group element a, ao@l=1@a=a4. Under the binary operation, any element paired with the element I corresponds to itself. The element J is called the unit element or the identity of the group. The use of the letter J. suggests the analogy with the number 1 of ordinary arithmetic. Exercise 3: Suppose a set consists of the real numbers, and the binary is addition, What element is the unit element?GROUP AXIOMS 13 Reciprocals or inverse elements. The second idea related to the number 1 that will be generalized and extended to groups is the concept of re- ciprocals. If « and » are any two numbers such that uv = 1, we say that « and v are reciprocals of each other. The next axiom is a generali- zation of this notion. Axiom 3 (Inverses): If @ is any element of a group, then there exists a unique element a~ of the group such that @O@e'=a'@a=l, The element o-! is called the inverse of a, Clearly, the inverse of a is (a) = a, The symbol for the inverse of a uses the negative exponent as an extension of the situation in ordinary algebra where, if 4 x 0, its inverse (reciprocal) is denoted by u—. Let us summarize our definition of a group. A group is a set G and a Binary operation ®@ on G such that the following axioms are satisfied: Axiom 1 (Associalivity). For any elements r,s, of G, 7@(s@i) = (r@s) OL. Axiom 2 (Identity). There is a unique element I in G such that, for every element r of G, r@l=1@r=r. Axiom 3 (Jnverses): For any element r of G, there exists a unique clement 1 of G such that r@rta@ral. The reader should not assume that this axiomatic definition of a group sprang full-grown from the brain of a single mathematician. Mathe- matical concepts often are developed by many mathematicians in an irregular fashion, by fits and starts, with dead-ends as well as revolu- tionary discoveries. Formal axioms that underlie a group were not ex- plicitly stated until after almost a century of work in group theory. The first important theorem was stated and proved in 1771 by Lagrange. (We shall consider this theorem in a later section.)rc GROUPS AND THEIR GRAPHS Cauchy,t whose contributions to group theory started in 1815, con- sidered only groups whose elements are represented in terms of permuta- tions. The word “group” was introduced in 1832 by Galois, the first to show that groups can be defined without using permutations as elements. Tt was not until 1854 that the process of emphasizing structure was carried to the point where Cayleyt was able to show that a group can be defined without reference to the specific, concrete nature of the ele- ments. The essential structure of a group, Cayley showed, depends solely on the way in which the binary operation on pairs of elements is prescribed. Before we go on to give additional examples of groups, we shall simplify and generalize the notation we have been using to denote the binary group operation. The experience of elementary algebra suggests that in- stead of a @ b = c we write ab = c, which we read as: element a multiplied by element 6 corresponds to the element ab, called the product of a and & (and also designated as c). Hereafter, we shall not always use the symbol ® for a general binary operation; instead we shall often rely on the notation ab to signify group multiplication of @ and b. We shall sometimes also write the group product ab in the form a + }. “Multiptication” as a general term for a group binary operation should not be confused with multiplication in ordinary arithmetic. As a special case, it might turn out that the elements of a group are numbers, and the associated group binary operation is ordinary multiplication. But, in general, group multiplication should be viewed as an abstract generaliza- tion of arithmetic multiplication. CAUTION: Although there are many operations that might be de- fined on elements of a set, in any specific group there is a definite single operation that is the group operation. t Augustin-Louis Cauchy (1789-1857) made a major contribution to the develop- ment of mathematics by emphasizing rigor in analysis. His presentations of “limit”, “continuity” and ‘‘convergence” still form the basis of modem concepts in analysis. Cauchy was one of the pioneers in the systematic development of a theory of groups, particularly groups of permutations. He is also remembered for his fundamental ‘theorems on functions of a complex variable. } Arthur Cayley (1821-1808) made contributions to many branches of mathematics, ranging from geometry and algebra to theoretical dynamics and physical astronomy. He also found time to practice law for fourteen years, Cayley is best known today for his creation of the theory of matrices and his work in group theory.CHAPTER THREE Examples of Groups If we want to decide whether a given set of elements with a specific binary operation constitutes a group, we must test to see whether the axioms are satisfied. Let us examine the following sets for eligibility as groups. We begin with Group A (p. 4). Example 1 ‘Set of elements: All integers (positive, negative, and zero). Binary operation: Addition. Associativity: Addition of numbers is associative, Identity: The set contains zero as an element and, for every integer u, u+0=0+4 =. Zerois the identity element. Tnverses: If u is an’ integer, its negative —x is an integer and u+ (—u) = (—u) + u = 0; —u is the inverse of #, or, in group notation, uw"! = —w, Thus, the set under test is a group. Since this group has infinitely many elements, we say the group is infinite. This group will sometimes be referred to as an infinite additive group or the additive group of integers. Example 2 Let the set be the same as in Example 1, but now consider multiplica- tion. The reader can check for himself that multiplication is a binary operation on the set of all integers and that the axioms on associativity and the existence of an identity element are satisfied. To see if the set satisfies Axiom 3, we try to determine the inverse of the element 2. We need an integer # such that 2@ «=I, or 2u = 1. There is no such integer, so we do not have a group. 1516 GROUPS AND THEIR GRAPHS Example 3 ‘The set consists of the two numbers 1 and —1, with multiplication as the binary operation: 1 (Da) 5 (-)(-)=t ((-1) = (-)M) = 1. Associativity: Undoubtedly. Identity: The identity element is 1. Inverses: (1)(1) = 1 and (—1)(-1) = 1, so (1)! = 1 and (—1)+1 = —1. Each element is its own inverse, ‘Thus we have a group. The number of elements in this group is finite. We say we have a finite group. The order of a finite group is the number of elements in the set. This group is of order 2. Example 4 Is there a group of order 1? Does the set containing only the number 1 constitute a group with respect to multiplication as the operation? Checking the three axioms shows that this is a group of order 1. Example 5 Next we examine a group whose elements are motions of a geometric figure. The discussion of this group will go into considerable detail on matters that are inherently associated with such motions. The concepts presented will recur often in the following pages, and therefore such detailed presentation is justified. By the same token, the reader should be prepared to lay a firm foundation for what will follow. Consider the motions of an equilateral triangle that rotates in its plane about an axis through its center. Our proposed group will have as ele- ments certain motions selected from the totality of these rotations, and the binary operation on this set will be “followed by” or “succession” (p. 8). We shall be interested in those motions that bring the triangle into coincidence with itself. Such a motion is called a congruence motion. 2 3@) 1 4 20 10) Figure 3.1 Figure 3.2EXAMPLES 7 To give ourselves a concrete picture of congruence motions, we first arbitrarily choose a particular position in the plane as the initial position of our equilateral triangle, before there are any rotations. We next assign a number to each vertex as an identification label. Our equilateral tri- angle might then look like the one in Figure 3.1. The dot in the center represents the intersection of the axis of rotation with the plane of the triangle, and the labeling of the vertices will help us locate them when they are displaced by the motions of our set. We must keep in mind that for the triangle to coincide with itself it is not necessary that each in- dividual (labeled) vertex coincide with itself, but only that the set of points making up the sides of the triangle after rotation coincide with the set that made up its sides in the initial position. For example, if the tri- angle of Figure 3.1 is rotated 120° counter-clockwise about the axis, we can view the rotated triangle as if it were a second triangle superimposed on the triangle in its initial position. This situation is depicted in Figure 3.2. The symbols in parentheses correspond to the positions occupied by the vertices of the equilateral triangle when it was in the initial position. ‘We can see that this rotation is associated with an interchange of vertices, namely 1 is replaced by 2, 2 is replaced by 3, 3 is replaced by 1. 2 3 1 3 2 1 Tnitial position «Position after rotation 120° ‘counter-clockwise Figure 3.3, It will be convenient to represent this particular manner of bringing about coincidence of the triangle with itself by “separating” the two posi- tions of the triangle; see Figure 3.3, Notice that a region near vertex 1 has been shaded to help visualize the motions of the triangle. Although the positions of the triangle have been drawn side by side, we must not forget that the picture is only a convenient representation of a super- position resulting from a rotation that brings the triangle into coincidence with itself. Are there any other rotations that bring the triangle from its original position into the second position pictured above? Surely a clockwise rotation of 240° does, as does a counter-clockwise rotation of 480° or18 GROUPS AND THEIR GRAPHS 840°. The reader can satisfy himself that any one of the rotations in the infinite set A = {counter-clockwise rotations of 120° -+ (360k)°, & = 0,1,2,+++} has the same effect. (A negative counter-clockwise rotation is to be inter- preted as a clockwise rotation.) The motions of set A have in common the property that each pairs off the vertices of our triangle in the initial position with the vertices after rotation in this specific way: Initial position Position after rotation 1 — 2 2 — 3 3 _ 1 The reader should notice that the rotations of set A have this properly no matter which position we choose as the initial position of the triangle. Now let @ denote any element of set A. The motion ¢ can be viewed as representative of the set A in the sense that rotation @ moves the triangle from the (arbitrarily) selected initial position into coincidence with itself in the position corresponding to this pairing-off of vertices: a: 132, 253, 3-1; see Figure 3.4. (Remember that a// motions of set A have this effect.) 2 3 1 a 2 1 Initial position Position resulting from motion Figure 3.4 In a given situation, we might find it convenient to let @ denote a particular motion in the set A; for example, a might be the counter- clockwise rotation of 120°. This choice corresponds to = 0 in 120° = (360k)°. If the reader prefers some other specific meaning for a, he can choose, say, & = 13, and keep in mind that the counter-clockwise rota-EXAMPLES 19 tion of 4800° is his private representative of set A. The particular choice is only a matter of convenience. The important thing is that all motions chosen from the set A pair off vertices of our triangle in the same way regardless of its initial position, Using our identifying labels for the vertices, we designate this pairing off by 152, 233, 31. Are there other rotations, aside from those in set A, that are congru- ence motions of the triangle? Consider the set of rotations B = {counter-clockwise rotations of 240° + (360k)°, & = 0,1, 2, +++}. Any motion of this set results in the superposition shown in Figure 3.5. Figure 3.6 shows these in “separated” form. 1a 0 20 Figure 3.5 2 1 1 : 3 2 b Initial position Figure 3.6 As above, 6 denotes any element of set B and is a “representative” of that set. For the sake of convenience, we have labeled the position re- sulting from this motion by the symbol 4. No matter which rotation is selected from the set B, its effect is to pair off the vertices of our triangle as follows: bs 193, 3-42, 291 (that is, 1 is replaced by 3, 3 is replaced by 2, 2 is replaced by 1).20 GROUPS AND THEIR GRAPHS There is yet another set of motions that bring the triangle into coin- cidence with itself—the set C = {counter-clockwise rotations of 0° =: (360k)°, # = 0,1, 2, +++}. 2 2 1 3 1 3 Initial position Figure 3.7 In Figure 3.7, ¢ is any element of set C. Notice that the effect of the motion ¢ is to bring the triangle back to its original position, with the vertices paired off in this way: @ 131, 252, 333, Our objective is to arrive at a group of motions; and since a group must contain a unit element (Identity Axiom), we should be alert to recognize that any motion c of set C has the earmarks of a unit element. In fact, if x is any element of sets A, B or C, then “‘x followed by c” is a rotation of the seme set as x, and “c followed by x” is a rotation of the same se! as x. To see this, recall that the rotations in set A, for example, are through angles of 120° + (360k)°, & = 0, 1, 2, «++, and those in C are through angles of 0° +: (360m)°, m = 0,1, 2, +++. If one rota- tion is followed by another, then the angle through which the triangle has been rotated is the sum of the angles in the separate rotations. Thus “g followed by c”’is a rotation through an angle 120° = (360k)° + 0° st (360m) ° or 120° & 360(& + m)°. Since &and m are integers, & + m is an integer, and this is all we need to know to place the rotation “a followed by c” into the set A. Similarly “¢ followed by a” is a rotation through 120° + 360(m -+ k)° in A. In the notation of group multiplication (p. 14), we have aw=a=-a bk=db=b c=eEXAMPLES a and these results are valid no matter which elements of sets 4, B and C are represented by a, b, c, respectively. These relations justify our using the symbol J (denoting a unit element) to represent any element of set C. aA A Figure 3.8 We have exhausted all possible rotations around our axis that are congruence motions of the triangle. Every such rotation is contained in one of the three sets A, B, C, with representative elements a, b, J associ- ated with the corresponding “separated” positions shown in Figure 3.8. Notice that each of the three positions of the triangle is labeled with the symbol denoting a motion that will bring the triangle from the given initial position to the depicted position. We claim that the set consisting of the three classes of congruence motions with representative rotations I, a, & forms a group with “followed by” as the binary operation. To show that “followed by” is a binary operation on this set, and to verify the group axioms, we find all products of two elements. For example, let us find ab by determining the pairing-off of vertices associated with the motion “a followed by b.” a: 132 b t3 2-3 332 31 2-1. The motion a pairs vertex 1 with vertex 2 (in the manner described on P. 18); the motion & pairs vertex 2 with vertex 1; so the effect of per- forming motion a followed by motion 6 is to pair vertex 1 with itself. Similarly, ¢ pairs 2 with 3, 6 pairs 3 with 2, so ab pairs 2 with itself, etc. Thus a: 1421 or 1-1 23-2 or 2-2 331>3 or 33.2 GROUPS AND THEIR GRAPHS Clearly, ab= 1. ‘The reader can easily verify that the remaining products are ae = 8 6 = a, andba = J. Now that we have established that “followed by” is a binary opera- tion on our set, we need only show that the group axioms are satisfied. Associativity: We have already pointed out (p. 12) that the operation of succession is associative when the elements of the set are motions. Identity: The discussion above has shown that the set C with repre- sentative rotation / is the unit element. Inverses: Since ab = I, and ba = I (and, of course, I - J = 1), each element has an inverse in the set. Example 6 Suppose that for any integer we consider only the remainder resulting from division by 2, and we define two integers to be equivalent if they have the same remainder; two integers are equivalent if both are even, or both are odd. We express that 8 and 6 both have the same remainder when divided by 2 by writing 8 = 6 (mod 2), where = denotes “equivalent” and “mod” is an abbreviation for “modulo”. Similarly, we can write 7 = 3 (mod 2) since 7 and 3 have the same remainder when divided by 2. Thus, if x denotes any even integer, and y denotes any odd integer, then x= 0(mod2) and y= 1 (mod 2). In effect, this concept of “equivalence modulo 2” enables us to take 0 and 1 as “representatives” of the even and odd integers, respectively. We are now in a position to examine a group with elements 0 and 1, and with binary operation “addition modulo 2”. We define addition modulo 2 (denoted by ©) of two integers a and b as follows: eO@b=0 if @+b=0(mod2), that is, if the ordinary sum of a and 0 is even; andEXAMPLES 23 e@b=1 if a+5=1(mod2). Addition modulo 2 is a well-defined binary operation on the set (0, 1} since 0+0=0(mod2), 0+1=1 (mod 2), 1+ 0 = 1 (mod 2), 1+ 1 0 (mod 2), or 0@0=0, 001-1, 1@0=1, 1@1=0 Associativity: It is easy to verify that addition modulo 2 is associative; for example, 1+ (141) =1+0=1 (mod 2), (+1) +1=0+1 = 1 (mod 2); Identity: 0 is the identity element. Inverses: Each element is its own inverse, since 0 + 0 = 0 (mod 2) and 1 + 1 = 0 (mod 2). We have just partitioned the set of all integers into two classes, the even integers with representative 0 and the odd integers with representa- tive 1. We can also partition the set of all integers into three classes by considering the remainders upon division by 3. All integers with remainder 0 upon division by 3 are in one class, all those with remainder 1 are in another, and all those with remainder 2 are in the third. We write, for example, 12 = 15 (mod 3), 7= 1 (mod 3), 5 = 8 (mod 3); that is, integers with the same remainder upon division by 3 are equiva- lent modulo 3, Similarly, we may consider classes of equivalent integers modulo 4, by considering remainders upon division by 4, and, in general, classes of equivalent integers modulo any integer n. Since the possible remainders upon division by # are 0, 1, +++, — 1, we obtain m classes which we may represent by 0, 1, +++, # — The reader should try to satisfy himself that the set (0, 1,2, «++, — 4} with binary operation “addition modulo n” constitutes a group. [If z is4 GROUPS AND THEIR GRAPHS any element of our set, what is its inverse? We seek the element y such that x+y = 0 (mod n), and we observe that = 0 (mod n).] Example 7 Now we consider the set of integers {1, 2, 3, 4] with the binary opera~ tion “multiplication modulo 5”. Thus, for any two integersr, s in our set, r@s=t or (7,8)—¢ if r+s = 4 (mod 5), ive., if the integers r+ s and ¢ have the same remainder when divided by 5. For example, 3@4=2, or (3,4)-+2 since 3+4= 12 = 2(mod5). The reader should verify that multiplication modulo 5 is a binary opera- tion on our set by showing that the product of any pair of elements is equivalent to one of the integers in our set. Associativity: It follows from the associativity of ordinary multiplica- tion of integers that multiplication modulo 5 is associative. (Verify this.) Identity: The identity element is 1. Inverses: 1 is its own inverse; 2, 3 and 4 satisfy the following relations which determine their inverses. 2+3 = 1 (mod 5), inverse of 2 is 3; 3+2= 1 (mod 5), inverse of 3 is 2; 4+4= 1 (mod5), inverse of 4 is 4. The reader should try to decide whether the omission of 0 from our set is necessary; that is, is the set {0, 1, 2, 3, 4} a group under the opera- tion multiplication modulo 5? Another question that the reader might want to tackle is this: Does the set {1, 2, 3} constitute a group under the binary operation multiplica- tion modulo 4? First, let the reader try to find the inverse of 2, that is, find the set element x such that 2x = 1 (mod 4). Example 8 Let p> 1 bea prime number, i.e., a number with precisely two posi- tive integral divisors, 1 and p, and consider the set {1, 2, 3, ++, p — 1}.EXAMPLES 25 We claim that “multiplication modulo p” is a binary operation on this set and that the group axioms are satisfied. We leave it to the reader to show that the axioms on associativity and the unit element are satisfied; the proof that the axiom on inverses is also satisfied is left as an exercise. Exercise 4: Consider the set {1, 2, 3, «++, #— 1}, # a prime number, with binary operation “multiplication modulo p”. Show that for any ele ment x of the set there is an element y of the set such that xy = 1 (mod p).CHAPTER FOUR Multiplication Table of a Group We must undertake the consideration of this question: How can we define a specific group? In other words, how many bits of information will suffice to determine a group as a mathematical entity? And how shall we display the data which define a particular group? ‘An answer to these questions was given by Cayley in 1854 when he introduced the multiplication table of a group. This is an arrangement similar to the familiar multiplication tables of arithmetic. The elements of the group are displayed in the top row and, in the same order, in the left column of the table, and the entries in the table are the group products, Consider first the group of order 2, consisting of the elements 1 and —1, with ordinary multiplication as the binary operation. Table 4.1 exhibits all possible products of two elements of the group. Since ordinary multiplication is commutative, any two elements of this group commute with each other. Table 4.1 Next, we shall construct the multiplication table for the group of con- gruence rotations in a plane of an equilateral triangle (see Example 5, p. 16), Using 7, a, b to represent the three elements of this group, we 26MULTIPLICATION TABLE 7 display them and their products in Table 4.2. Some explanations and simplifications are in order. We cannot take for granted in advance that any two elements of this group commute with each other; for this reason the factors in each product are written in the order in which the multiplication is to be performed, and the first factor is listed in the column on the left, the second in the row on top. 2nd factor I jar | te | Astfactor a | af | oo | ab b | or | oo | Table 4.2 ‘We recall that in our detailed discussion of this group we found on p. 22 that aa=b, ab=be=1, bb=a. Using these results, and the properties of the identity Z, we may write the multiplication table as follows: 2nd factor rlael]o Istfactor a | o | b | 7 Table 4.3 Many of the properties of this rotation group can be read directly from its multiplication table. Inverses can be found by observing in which row and column J occurs in the table. Notice the interesting “coincidence” whereby the rows are rearrangements (or permutations) of the top row, and the columns are rearrangements of the left column. The table also shows that all elements of the group commute with each other, since all products located symmetrically with respect to the2B GROUPS AND THEIR GRAPHS main diagonal are the same. The main diagonal runs from the upper left- hand corner to the lower right-hand corner, and is, in this case, I b In any multiplication table, if one product is rs, then the symmetrically situated product is sr. We call a group commutative if any two elements commute. Therefore, we can say: A finite group is commutative if, and only if, its multiplication table has the property that products located symmelrically with respect to the main diagonal represent the same group element. Another important property of the group of congruence rotations of an equilateral triangle cannot be read off from the multiplication table in its present form, but will become evident after we introduce some new notation and use it to write our table in yet another form. In keeping with the idea that group multiplication is a generalization of ordinary multiplication, we shail designate the group element aa by e?, aca by a’, and, in general, a product of & a’s will be written as a*. Similarly, we shall write (¢-") (a~!) as a-*, and the product of & factors, each a~!, as a-*, Since a* - a-* = J, it is natural to define a® = I. Group elements @", where n is any integer, are referred to as powers of @. The reader can verify for himself that the usual rules for multiplying powers also hold for group multiplication of powers of a group element. rlaola@ ririlale ~ afa|a@ eyeytye Table 44 In the group under discussion, b= aa=a, b= aa=a=T,MULTIPLICATION TABLE 29 so its multiplication table can be written as in Table 4.4. In this last form, the table shows that every element of the group is a power of the single element a. A group with this property is said to be generated by the element a, and a is called a generator. This concept will be developed later in a section on group generators. A non-commutative group. Although we have met examples of non- commutative pairs of elements, we have not yet seen a non-commutative group. We should recall that we defined a commutative group as a group in which any two elements commute. Such groups are also called Abelian, in honor of Abel,f who first applied such groups to the theory of equations. If a group has two elements that do not commute, then the group is called non-commutative, regardless of how many other pairs of elements do commute. Can there be a group in which no two elements commute? The answer is clearly “no”, since every group contains the unit element T which commutes with every element. ‘We shall now construct a non-commutative group of order 6. As we proceed, it will become evident that this is the smallest possible order for a non-commutative group. To construct our group, we consider the motions of an equilateral triangle that bring it into coincidence with itself. We have already examined such a set of motions subject to the restriction that the triangle rotate in its own plane, and we saw that these rotations constitute a group of order 3. If we remove this restriction, other motions become permissible, since the triangle can be turned over. For example, flipping the triangle about one of its altitudes brings it into coincidence with itself, but is not one of the rotations studied in Example §, p. 16, We shall see that there are now six positions in which the triangle coincides with itself. We label these I, r, 7%, f, fr, fr? for reasons that will become clearer as the reader proceeds, The positions are illustrated by the six diagrams in Figure 4.1. (From time to time there will be geometric representations of group properties that call upon the intuitive space perception of the reader. It is recommended that the reader take advantage of the assistance of a physical model. For example, a cut-out equilateral triangle will help in visualizing the motions to be described here.) + The Norwegian mathematician Niels Hentik Abel (1802-1829) proved the impossi- bility of solving the general algebraic equation of fifth degree by radicals. In his work with algebraic equations, he used the concept of commutative groups, now called “Abelian groups”. He also opened new fields of inquiry in the theory of functions, particularly elliptic functions, Abel died of tuberculosis at the age of 26.30 GROUPS AND THEIR GRAPHS 2 a 1 1 ey mr vat 2 3 1 3 1 t ‘tf? pie F Figure 4.1 In constructing our group we shall use procedures similar to those followed in Example S on p. 16. Such procedures are convenient whenever we deal with a group of motions. ‘A symbol used to denote a motion will frequently be given a specific meaning. Thus, in this section, 7 will denote a counter-clockwise rota- tion of 120° around the axis through the center of the equilateral triangle. But it can also represent any element of the set A of counter-clockwise rotations through angles of 120° + (360)°, & = 0, 1,2, +++. Similarly, we shall later introduce a motion f that will be given a specific meaning for convenience in visualizing the motions, but that will also represent any element of a certain set of motions, The essential feature is that we shall consider as the “same” motion all those that pair off vertices in the same way. We should like to have a pictorial representation of the motions of our group, but the static diagrams of this book cannot portray motions directly. Therefore, we enlist the cooperation of the reader in interpreting a static diagram as representing a motion in the same way as indicated on p. 21: namely, if a symbol « denoting a motion is assigned to a diagram showing the position of a figure, we shall understand the diagram fo represent a motion x thal brings the figure from a given initial position lo the position labeled x. In what follows we shall find it convenient to suppose that r represents a counter-clockwise rotation of 120° about an axis perpendicular to the plane of the triangle and through its center. Then the first three positions can be reached by the motions J, r, r2, as we have already seen. (Remem- ber that J is any rotation through an angle of 0° 4: (360k)°.)MULTIPLICATION TABLE at 2 2 3 1 1 33 1 2 12 3 I f t f Figure 4.2 To reach one of the new positions we must somehow flip the triangle over. We can accomplish this by rotating the triangle 180° around an altitude through one of the vertices. We choose the altitude through vertex 2 as our axis of rotation. The rotation of 180° around this altitude as axis will be denoted by f. Of course, f also represents any rotation through an angle of 180° + (360k)° about this axis. Thus we have the diagrams in Figure 4.2. 2 3 A 3 1 1 2 { fr Figure 4.3, We should try to clarify what we mean by the symbol fr. The posi- tions labeled f and fr in Figure 4.1 are shown again in Figure 4.3. From these, the rotation 7 scems to be 120° clockwise, instead of the prescribed 120° counter-clockwise. This apparent discrepancy is cleared up if we notice that lipping the triangle has inverted the axis of the rotation r. First, we need a more detailed description of the rotation r. We have taken as the axis of rotation the line through the center of the equilateral tri- angle and perpendicular to its plane. A direction is assigned to this axis as shown by the arrow-head in Figure 4.4, and our rotation ris associ- ated with this pairing-off of vertices: 1-> 2, 2~» 3, 3-1 (that is, 1 is replaced by 2, 2 is replaced by 3, 3 is replaced by 1). Figure 4.432 GROUPS AND THEIR GRAPHS Imagine, now, that the arrow-head of the axis is the threaded tip of a right-hand screw; to effect the rotation r, turn the triangle 120° in the direction you would turn to tighten a right-hand screw. If the first triangle in Figure 4.4 is subjected to the motion f, it will be in the posi- tion labeled f in Figure 4.5. Notice that the axis has been inverted by the flip. If, subsequently, the triangle is subjected to a rotation 7, where + is interpreted as the tightening of a right-hand screw, then the Position labeled fr in Figure 4.5 is obtained, Thus, whether the rolation r acts on the triangle in the position I or in that labeled f, it pairs off vertices ‘in the same way: 1—+ 2, 2 3, 3 1, Figure 4.5 The set of the six classes of motions illustrated by the six possible Positions of the triangle, with the binary operation of succession, or “followed by”, forms a group. We know that the operation is associative, and the unit element J is an element of the set. That the axiom on in- verses also holds can be seen intuitively by considering that, if there is a motion which transforms one position into another, then there is also one reversing the transformation (inverse). 2nd factor rilrlels | aloe ryhr|rlelsr [ae | oe Astfactor ot ff} # | rf or | vf | vf | ripe ay | | OS | AD | Sehr | fe fe | we | of | fe | SAF | See | pre Table 4.5MULTIPLICATION TABLE 33 It will be instructive to demonstrate some of the group properties by means of the multiplication table. Note that 71 = J and f? = J, by the very meaning of these motions. Using these special properties, we con- struct Table 4.5. 3 1 2 fr frf frfreI Figure 4.6 To complete the construction of the multiplication table for our group, ‘we must express each of the entries in Table 4.5 as one of the six elements I,7,7,f, fr, Jr. We shall work through the details of simplifying two of the products and leave the others to the reader. First, we show that frfr = I. Consider the sequence of diagrams in Figure 4.6. The first represents fr, Beginning with the triangle in this position, we rotate 180° around the altitude through vertex 2. The result frf (fr followed by f) is pictured in the second diagram, Then we rotate 120° in the direction of a right hand screw around the axis through the center; the result, ‘ifr, is pictured in the last diagram and is seen to be the same as the initial position, designated by I. Thus jrfr = I. a 1 3 r ifr fr Figure 4.7 Next we show that fr? = fr by means of the diagrams in Figure 4.7. Using all such simplified products, we form the multiplication table 4.6. The table reveals that (1) “Followed by” is a binary operation on our elements. (2) Since I occurs precisely once in each row and column, the axiom on inverses is satisfied. We can determine at once the inverse of any group element. For example, the configuration fr freee dl shows us that. (fr)-! = fr.cr GROUPS AND THEIR GRAPHS 2nd factor Ast factor yr) rl) st f T r mie | tl) oe fir bade Table 4.6 (3) The group is non-commutative. A glance at elements located sym- metrically with respect to the main diagonal shows, for example, that (ff =P #1 = {(fr). (4) The rows and columns are permutations or rearrangements of the elements in the top row and left column, respectively—the “‘coincidence”” previously observed. (5) The 3 x 3 square in the upper left-hand corner is precisely the multiplication table of the group of order 3 of the rotations of an equi- lateral triangle in its plane. At the other end of the main diagonal, in the lower right-hand corner, there is a 3 x 3 square which is a rearrange- ment of the upper left-hand square. But in the lower left and upper right 3 x 3 squares, we have the two main diagonal squares reproduced with each product prefixed by f. If M represents the set of elements in the upper left 3 x 3 square, then Table 4.7 symbolizes this pattern-within- a-pattern in the group multiplication table and offers a hint for making further discoveries to refine our analysis of the structure of a group. We shall explore these possibilities in a later section on normal subgroups and factor groups. mM | M ‘Table 4.7MULTIPLICATION TABLE 35 Structure of the multiplication table of a group. We now take a closer look at the internal structure of the multiplication table of a group. First, we examine the “coincidence” mentioned in (4) above, namely, that the rows and columns of a group multiplication table are permutations of the top row and left column, respectively. We shall show that this is not at all a coincidence, but rather a characteristic property of the multiplica- tion table of any group. After having done this, we shall view a group multiplication table as a pattern exhibited by an array of symbols ar- ranged in a square. Within this array, we shall observe spatial patterns of symbols and indicate how they correspond to group relationships. In this way, the structure of a group is reflected in “geometric” properties of its multiplication table. It can be shown that, conversely, a square array with these “geometric” properties is the multiplication table of a group. “Solving” a group “equation”. In dealing with group elements and their relations, it is sometimes necessary to be able to answer this ques- tion: If ¢ and & are known clements of a group, is there an element = of the group such that av = 6? We claim that x = art is the group ele- ment we seek, for a(a-®) = (aa) = 1b that is, x = a satisfies the group “equation” ax = 6. Are other solutions possible? We shall answer this question by showing that, whenever y is @ solution of ax = 6, then y = ab; in other words, a7 is a unique solution. We first assume that there is a group element y for which ay = b. We know that a” exists (Inverses). We can multiply each member of ay = 5 on the left by o~! and find that a~(ay) and a-1(b) represent the same group element, that is, a*(ay) = a-'(6). Consequently (o"a)y = a (Associativity) , Ty = 0%, or y= od (Identity). Since we have already verified by substitution that the element a7" satisfies the group equation ax = 6, our claim that ob is a unique6 GROUPS AND THEIR GRAPHS solution has been proved. Notice that all group axioms were needed in the proof. The solution of an equation of the form aa = b, where a and & are group elements, may be treated analogously. Mul- tiplying on the right by a, we obtain the solution waa = ba, We formulate our results as a “rule”: To “solve” ax = 6, multiply on the deft by a~ jo find x = a~"b; to “solve” za = 8, multiply on the right by a~ to find = ba, Exercise 5: In each of the following, find 2: (a) obe=c = (d) @= bat and =I (b)ex=c () #=0 and =I () mb=¢ (wt = abe As a first application of the preceding discussion, we shall prove a relationship about group elements and their inverses that will be useful later on. Suppose we have an element of a group represented as a product of other group elements. We might have, for example, d= ab. The question is: how can we represent d+, the inverse of d? An equivalent question is: how can we find a group element x such that dx = I or abz = I? We know from the preceding discussion that this group equa- tion has a unique solution. To find it explicitly, we first multiply on the left by o-! obtaining eax=a'T or be =a, and then we multiply on the left by 5~ obtaining be = aor = ta, To verify that d-! = ba we show that d(daMULTIPLICATION TABLE 37 d(b0-!) = ab(b-a-) = aldo) = of (ba) = oat =. Similarly, if d = abc, then d- = cba“, The pattern is clear, and we can make the general statement that, if d = qya2+-+dq, then d+ = aa;},++-az4a;1, In words, the inverse of a product is the product of the inverses taken in reverse order. As an additional application of the procedures for solving a group equation, we shall prove a theorem that explains why any row (or col- umn) of the multiplication table of a group is a rearrangement of the elements in any other row (or column). Suppose we have a group of order m consisting of the elements 4G, a2, ***, @, (of course, one of these elements is the identity element, I, but it is not specifically labeled as such). Take any one of these # elements, say aj, and for convenience call it 5. Form the set of m products bay, bar, +++, ban by multiplying on the left by 5. It is claimed that these products are the original n group elements, possibly rearranged. To prove this claim, we shall show that no two elements of the set of products can be the same group element. Suppose, for example, ba; = ba; where i + j. Then, multiplying on the left by 5-1, we have bba; = bay or ae = as (i 9). But a, and a, are different group elements if i # j; so the assumption that ba; = ba; has led toa contradiction, and we conclude that ba; # ba; whenever i # 7. Therefore the products of the set are distinct. Since each of the distinct products bai, ba, ---, bda isan element of the original group, together they must be all » group elements. This completes the proof. We have proved our claim for multiplication on the left. The same argument may be applied to the set of products resulting from right multiplication of the group elements by a fixed element to complete the proof of this theorem for finite groups: