Chapter 1 What is a Ring?

1.1 Some Examples

Consider the following sets : 1. Z = {. . . , 1, 0, 1, 2, . . . } the set of integers. 2. Q = { a | a, b Z, b = 0} the set of rational numbers. b 3. M2 (R) the set of 2 2 matrices with real numbers as entries. 4. 2Z = {. . . , 2, 0, 2, 4, . . . } the set of even integers. 5. C(R) the set of continuous functions from R to R. 6. Q[x] = {an xn + an1 xn1 + + a1 x + a0 | an , . . . , a0 Q} the set of polynomials with rational coefcients. 7. Z/6Z = {0, 1, 2, 3, 4, 5} the set of congruence classes in Z modulo 6. Remember the very general denition of an algebraic structure as a set equipped with a binary operation, that is a scheme for combining any pair of elements of the set to produce a new element of the same set. All of the sets in our list above have binary operations dened on them in natural and probably mostly familiar ways. Of course it is possible for a set to have more than one natural binary operation dened on it. Algebra, in its broadest sense, is the study of algebraic structures. What do all the six sets described above have in common as algebraic structures? Each of them is equipped with two binary operations called addition and multiplication. In Z, Q and 2Z we have the usual addition and multiplication of integers and rational numbers. In M2 (R) we have matrix addition and matrix multiplication. In C(R) we have addition and multiplication dened by ( f + g )(x) = f(x) + g(x) , for all x R and all f, g C(R)
+ in C(R) +inR

( f g )(x) = f(x) g(x) , for all x R and all f, g C(R).

in C(R) inR

In Q[x] we have the usual addition and multiplication of polynomials, e.g. (x2 + 2x + 4) + (x3 3x + 2) = x3 + x2 x + 6, (x2 2x + 1)(x + 5) = x3 + 5x2 2x2 10x + x + 5 = x3 + 3x2 9x + 5. In Z/6Z the addition and multiplication are dened modulo 6, e.g. 4 + 5 = 3; 4 5 = 2, etc. Note: In each case the set under consideration is closed under the relevant operations of addition and multiplication; this means that in each case the product and sum of a pair of elements in a particular set also belong to that set. For example the set of odd integers is not closed under addition, since the sum of two odd integers is not odd. ADDITION
IN OUR

EXAMPLES

All the above examples contain an identity element for addition, which we refer to as the zero element and write as 0. This element has the property that adding it to another element has no effect. The zero elements in our examples are 1. The integer 0 2. The rational number 0 0 0 3. The zero matrix 0 0 4. The integer 0 5. The function f0 : R R dened by f(x) = 0, x R 6. The zero polynomial 0 7. The congruence class 0 modulo 6 In each of our sets, every element has an additive inverse or negative. Two elements are additive inverses each other if their sum is the zero element. The fact that every element of a set has an additive inverse means that subtraction can be dened in the set. In all of our sets, addition is commutative, i.e. a + b = b + a for all pairs a and b of elements. MULTIPLICATION
IN OUR

EXAMPLES

The multiplication is commutative in all these examples except for M2 (R). For 2 2 matrices A and B, the products AB and BA need not be equal. 2

 Except for 2Z each of these examples contains an identity element for multiplication, i.e. an element e for which e a = a e = a for all elements a of the set; multiplying by e has no effect. The multiplicative identities are 1. The integer 1 2. The rational number 1 1 0 3. The matrix 0 1 4. No identity element for multiplication 5. The function f1 : R R dened by f(x) = 1 for all x R 7. The congruence class 1 modulo 6 Two elements are multiplicative inverses of each other if their product is the multiplicative identity element. In Q, every element except 0 has a multiplicative inverse, namely its reciprocal. All the other examples contain non-zero elements without multiplicative inverses. The seven algebraic structures mentioned in this section are all examples of rings. 6. The polynomial 1

1.2 The Axioms of a Ring

NOTE: In this section and throughout these lecture notes, please do not confuse the symbol R, which is used for a general ring, with the symbol R which is used for the set of real numbers. Denition 1.2.1 A ring is a non-empty set R equipped with two binary operations called addition (+) and multiplication (), satisfying the following properties : The rst four are concerned with the operation that is called addition. A1 Addition is associative. (r + s) + t = r + (s + t) for all r, s, t R. A2 Addition is commutative. r + s = s + r for all r, s R. A3 R contains an identity element for addition, denoted by 0R and called the zero element of R. r + 0R = 0R + r = r for all r R. A4 Every element of R has an inverse with respect to addition. (The additive inverse of r is often denoted by r). For every r R, there exists an element r R for which r + (r) = 0R . NOTE : Axioms A1 to A4 could be summarized by saying that R is an abelian group under addition. (If this remark is not helpful for you, disregard it for now). The multiplication operation is required only to satisfy one special condition : M1 Multiplication is associative. (r s) t = r (s t) for all r, s, t R. The last two axioms are concerned with the manner in which the two operations must interact. D1 r (s + t) = (r s) + (r t) for all r, s, t R. D2 (r + s) t = (r t) + (s t) for all r, s, t R. -Distributive laws for multiplication over addition. REMARKS 1. A ring is called commutative if its multiplication is commutative.

2. A ring R is called unital or referred to as a ring with identity if it contains an identity element for multiplication. In this case we will denote the multiplicative identity by 1R or just 1. We have already met one example of a ring without identity, namely the ring 2Z of even integers. 3. The term ring was introduced by David Hilbert in the late 19th century, when he referred to a Zahlring or number ring. Our rst theorem about rings is the following consequence of the ring axioms. Theorem 1.2.2 Let R be a ring. Then for all elements r of R we have 0R r = 0R and r 0R = 0R . i.e. multiplying any element of R by the zero element results in the zero element as the product. Proof : Let r R. We have (0R r) + (0R r) = (0R + 0R ) r = 0R r. Adding the additive inverse of the element 0R r to both sides of this equation gives 0R r = 0 R . A similar argument shows that r 0R = 0R . THREE
REMARKS

1. The problem of deducing the truth of a statement like Theorem 1.2.2 from the axioms of a ring might be somewhat daunting. The proof may not be too hard to follow, but could you have come up with it yourself? If you were trying to, and you didnt know where to start, there are certain observations you could make that might help. There are nine axioms for rings - which might be likely to be helpful in proving the two (left and right) statements of Theorem 1.2.2? Well, the statement is about multiplication and about the zero element. According to the ring axioms, what is special about the zero element has to do with addition not multiplication. So it might seem likely that the statement in the theorem is essentially connected to the interaction of the addition and multiplication - the two axioms that deal with that are the distributive laws, so maybe we should not be so surprised that these have a crucial role in the proof.

2. The next two remarks are about the philosophy of abstract algebra and the mechanisms by which the subject progresses. The denition of a ring consists of a list of technical properties, but the motivation for this denition is the ubiquity of objects having these properties, like the ones in Section 1.1. When making a denition like that of a ring (or group or vector space), the goal is to arrive at a set of axioms that exactly captures the crucial unifying properties of those objects that you wish to study. In familiar number systems like the integers, the rational numbers and the real numbers, we are all used to the fact with which Theorem 1.2.2 is concerned, namely that multiplying by zero gives zero. The same fact is easily observed to hold in the polynomial ring Q[x] and in the ring of matrices M2 (R). We might well speculate that in any ring, it is probably the case that multiplying by the zero element always results in the zero element. But before we can assume that this property holds in every ring and incorporate it into our mental scheme for thinking about rings we must deduce this property as a consequence of the ring axioms. If we were unable to do this, but we only wanted to study rings with the property described in Theorem 1.2.2, we could an extra axiom to our denition of a ring insisting on this multiplication by zero property. However the fact that this property does turn out to follow from the standard ring axioms means that it does not need to be included in the denition. 3. On looking at Denition 1.2.1, you may wonder why these nine axioms in particular are chosen to comprise the denition of a ring. Does it look like an arbitrary selection of rules? Why do we insist that the addition have an identity element and that every element have an inverse for addition, but where the multiplication is concerned only ask that it be associative? What happens if we add more axioms about how the multiplication should behave, or drop some of the axioms about addition? The answer is that people do these things and they lead to different areas of study within abstract algebra. Relaxing the addition axioms in various ways leads to different types of algebraic structures such as nearrings and semirings. If you drop the requirement that multiplication must be associative then you are studying non-associative rings people do study all of these variants and some of them have important connections to other areas of mathematics. You can even relax the distributive laws and people do this too. However rings themselves as dened in Denition 1.2.1 are of paramount importance in mathematics. On the other hand, if you want more instead of fewer axioms, you can insist that multiplication be commutative as well as associative, then you are studying commutative rings. In fact much of this course will be concerned with commutative rings. If you further insist that you want an identity element for multiplication and that every (non-zero) element have an in6

verse for multiplication, then you are studying elds. Fields are examples of rings, and eld theory itself is a vast area of mathematical activity. A crucial practice in studying abstract algebra is to be absolutely clear on the precise axioms that determine the class of objects that you are studying.

1.3 Units in Rings

As we have already mentioned, the axioms of a ring are not very restrictive concerning how the operation of multiplication should behave - all we ask is that it should be associative. We do not even insist that every ring should contain an identity element for multiplication (although incidentally some authors in ring theory do). If a ring does contain an identity element for multiplication, then we can enter a discussion about whether or not something like division is possible in the ring; we can try to identify pairs of elements that are related to each other in the way that a rational number is related to its reciprocal or in the way that a non-singular matrix is related to its inverse. Denition 1.3.1 Let R be a ring with identity element 1R for multiplication. An element r R is called a unit in R if there exists s R for which r s = 1R and s r = 1R . In this case r and s are (multiplicative) inverses of each other. Example 1.3.2 1. In Q every element except 0 is a unit; the inverse of a non-zero rational number is its reciprocal. 2. In Z the only units are 1 and 1 : no other integer can be multiplied by an integer to give 1. 3. In M2 (R), the units are the 2 2 matrices with non-zero determinant, and 1 0 the identity element is . 0 1 4. In Z/6Z the only units are 1 and 5; each of these is its own inverse. 5. Question for discussion in the seminar : what are the units in M2 (Z), the ring of 2 2 matrices with integer entries? NOTATION : We will denote the set of units in a ring R with identity by U(R). REMARKS 1. If R is a unital ring having two or more elements then it follows from Theorem 1.2.2 that the zero element of R and the multiplicative identity in R cannot be the same element. 2. If R has two or more elements then 0R cannot be a unit in R, again by Theorem 1.2.2. 8

3. It is possible for a ring to have only one element; for example the subset of Z containing only 0 is a ring. (This is called the zero ring and as an example of a ring it is not very instructive) 4. 1R is always a unit in R since it is its own inverse. The next theorem is concerned with a special property of the subset of a ring consisting of the units. Suppose that R is a unital ring. Then from the above comments it follows that U(R) is a subset of R that includes the (multiplicative) identity element but not the zero element. Is U(R) just a set, or does it have algebraic structure of its own? The full ring R has addition and multiplication dened on it. If we take two units of R we can add them in R; will the result be a unit? If we take two units of R and multiply them (in R), will the result be a unit? If the answer to this second question is yes, then the set of units of R is itself an algebraic structure with respect to the multiplication of R, and we can study its properties. Algebraists are always on the lookout for substructures of the objects that they are studying, which are themselves algebraic structures with respect to the operation(s) of the larger object. The general thinking behind this practice is that small things are usually easier to understand than big things, and that we have some chance of understanding (at least partically) a large complicated algebraic structure if we can identify smaller parts of it that are themselves algebraic structures. Theorem 1.3.3 Let R be a ring with identity element 1R . Then U(R) is a group under the multiplication of R. (U(R) is called the unit group of R). Note : The statement that U(R) is a group under multiplication means that : U(R) is closed under multiplication - whenever elements a and b belong to U(R), so does their product ab. U(R) contains an identity element for multiplication. U(R) contains a multiplicative inverse for each of its elements. Proof of Theorem 1.3.3: We need to show 1. U(R) is closed under the multiplication of R; i.e. that rs is a unit in R whenever r and s are units in R. So assume that r and s belong to U(R) and let r1 and s1 denote their respective inverses in R. Then (rs)(s1 r1 ) = = = = 9 r(ss1 )r1 r1R r1 rr1 1R .

Similarly (s1 r1 )(rs) = 1R and so s1 r1 is an inverse in R for rs, and rs U(R). 2. U(R) contains an identity element for multiplication. This is true since 1R U(R). 3. U(R) contains an inverse for each of its elements. To see this, suppose r U(R), and let r1 be the inverse of r in R. Then r1 r = 1R and rr1 = 1R , so r is the inverse of r1 , and r1 is in U(R). This proves the theorem. EXAMPLES 1. U(Z) = {1, 1} is a cyclic group of order 2. 2. The unit group of the matrix ring Mn (R) is the general linear group GL(n, R) of n n invertible matrices over R. 3. The unit group of Q is denoted Q and consists of all non-zero rational numbers. QUESTION FOR DISCUSSION IN THE SEMINAR: In general, is there anything to be said about the behaviour of U(R) with respect to addition in R? Suppose that R is a ring with identity. Then we know that the unit group of R cannot include the zero element of R, but any non-zero element of R could potentially be a unit. A particularly nice thing to happen is for every non-zero element of R to be a unit. Rings in which this occurs are worthy of special study. Denition 1.3.4 A ring with identity is called a eld if it is commutative and every non-zero element is a unit (so we can divide by every non-zero element). Examples of elds include Q, R, C and Z/5Z (check). A ring with identity in which every non-zero element is a unit is called a division ring. Commutative division rings are elds. Examples of non-commutative division rings are not easy to nd, but we will see at least one in this course.

10

1.4 Integral Domains and ZeroDivisors

We saw in Theorem 1.2.2 that whenever an element of a ring is multiplied by zero, the result is zero. When working in the set of real numbers we often use the converse of this - a product ab can be zero in R only if at least one of a and b is equal to zero. QUESTION FOR THE SEMINAR : When/how do we use this? QUESTION: Is it true in every ring that the product of two elements can be zero only if at least one of the elements is zero? To think about this question, look at some examples. Example 1.4.1 1. In M2 (Q) 1 1 2 2 1 1 1 1 = 0 0 0 0 .

i.e. the product of two non-zero matrices in M2 (Q) can be the zero matrix. 2. In Z/6Z, 2 3 = 0 So the answer to the question is no in general. However, it is of interest to study the class of rings in which the property described in the question holds. Denition 1.4.2 Let R be a ring with zero element 0R . An element a of R is called a (left) zerodivisor in R if a = 0R and there exists an element b = 0R of R for which ab = 0R . (In this case b is a right zerodivisor). NOTE: If R is commutative then ab = ba and we just talk about zerodivisors (not left and right zerodivisors). Denition 1.4.3 A commutative ring with identity that contains no zero-divisors is called an integral domain (or just a domain). In an integral domain, the product of two elements can be zero only if one of the elements is zero. EXAMPLES 1. Z is an integral domain. Somehow it is the primary example - it is from the ring of integers that the term integral domain is derived. The adjective integral in this context is related to integer (nothing to do with integrals in the calculus sense!). 11

2. Every eld is an integral domain. For let F be a eld and suppose that a, b are elements of F for which ab = 0F . Assume a = 0. Then a has a multiplicative inverse in F and ab = a (ab) = (a1 a)b = 1F b = b
1

= = = = =

0F a1 0F 0F by Theorem 1.2.2 0F 0F .

REMARK : It follows from the above argument that no unit can be a (left or right) zero-divisor in any ring. EXERCISE: Write down a proof of the statement of the above remark. 3. An example of a commutative ring with identity that is not an integral domain is Z/6Z (or Z/nZ for any composite natural number n). QUESTIONS
FOR THE

SEMINAR :

1. For which natural numbers n is Z/nZ a eld? 2. For which natural numbers n is Z/nZ an integral domain? 3. For a natural number n, which elements of Z/nZ are units? 4. Is it true for every natural number n that every non-zero element of Z/nZ is either a unit or a zero-divisor? Can we prove this? 5. Suppose that R is a commutative ring with identity that is not an integral domain. Must it be true that every non-zero element of R is either a zero-divisor or a unit?

12