16 CHAPTER 1 PRELIMINARIES

13. Prove that n < 2n for all n 2 N.

14. Prove that 2n < n! for all n 4, n 2 N.
15. Prove that 2n  3  2 n2
for all n 5, n 2 N.
16. Find all natural numbers n such that n2 < 2n. Prove your assertion.
17. Find the largest natural number m such that n3  n is divisible by m for all n 2 N. Prove your
assertion.
pffiffiffi pffiffiffi pffiffiffi pffiffiffi
18. Prove that 1= 1 þ 1= 2 þ þ 1= n > n for all n 2 N, n > 1.
19. Let S be a subset of N such that (a) 2k 2 S for all k 2 N, and (b) if k 2 S and k 2, then
k  1 2 S. Prove that S ¼ N.
20. Let the numbers xn be defined as follows: x1 :¼ 1, x2 :¼ 2, and xnþ2 :¼ 12 ðxnþ1 þ xn Þ for all
n 2 N. Use the Principle of Strong Induction (1.2.5) to show that 1  xn  2 for all n 2 N.

Section 1.3 Finite and Infinite Sets

When we count the elements in a set, we say ‘‘one, two, three, . . . ,’’ stopping when we
have exhausted the set. From a mathematical perspective, what we are doing is defining a
bijective mapping between the set and a portion of the set of natural numbers. If the set is
such that the counting does not terminate, such as the set of natural numbers itself, then we
describe the set as being infinite.
The notions of ‘‘finite’’ and ‘‘infinite’’ are extremely primitive, and it is very likely that
the reader has never examined these notions very carefully. In this section we will define
these terms precisely and establish a few basic results and state some other important
results that seem obvious but whose proofs are a bit tricky. These proofs can be found in
Appendix B and can be read later.

1.3.1 Definition (a) The empty set ; is said to have 0 elements.

(b) If n 2 N, a set S is said to have n elements if there exists a bijection from the set
N n :¼ f1; 2; . . . ; ng onto S.
(c) A set S is said to be finite if it is either empty or it has n elements for some n 2 N.
(d) A set S is said to be infinite if it is not finite.

Since the inverse of a bijection is a bijection, it is easy to see that a set S has n
elements if and only if there is a bijection from S onto the set {1, 2, . . . , n}. Also,
since the composition of two bijections is a bijection, we see that a set S1 has n
elements if and only if there is a bijection from S1 onto another set S2 that has n
elements. Further, a set T1 is finite if and only if there is a bijection from T1 onto
another set T2 that is finite.
It is now necessary to establish some basic properties of finite sets to be sure that the
definitions do not lead to conclusions that conflict with our experience of counting. From
the definitions, it is not entirely clear that a finite set might not have n elements for more
than one value of n. Also it is conceivably possible that the set N :¼ f1; 2; 3; . . .g might be
a finite set according to this definition. The reader will be relieved that these possibilities do
not occur, as the next two theorems state. The proofs of these assertions, which use the
fundamental properties of N described in Section 1.2, are given in Appendix B.

1.3.2 Uniqueness Theorem If S is a finite set, then the number of elements in S is a

unique number in N.
 1.3 FINITE AND INFINITE SETS 17

1.3.3 Theorem The set N of natural numbers is an infinite set.

The next result gives some elementary properties of finite and infinite sets.

1.3.4 Theorem (a) If A is a set with m elements and B is a set with n elements and if
A \ B ¼ ; , then A [ B has m þ n elements.
(b) If A is a set with m 2 N elements and C  A is a set with 1 element, then AnC is a set
with m  1 elements.
(c) If C is an infinite set and B is a finite set, then CnB is an infinite set.

Proof. (a) Let f be a bijection of N m onto A, and let g be a bijection of N n onto B. We

define h on N mþn by h(i) :¼ f (i) for i ¼ 1; . . . ; m and h(i) :¼ g(i  m) for
i ¼ m þ 1; . . . ; m þ n. We leave it as an exercise to show that h is a bijection from
N mþn onto A [ B.
The proofs of parts (b) and (c) are left to the reader, see Exercise 2. Q.E.D.

It may seem ‘‘obvious’’ that a subset of a finite set is also finite, but the assertion must
be deduced from the definitions. This and the corresponding statement for infinite sets are
established next.

1.3.5 Theorem Suppose that S and T are sets and that T  S.

(a) If S is a finite set, then T is a finite set.
(b) If T is an infinite set, then S is an infinite set.

Proof. (a) If T ¼ ;, we already know that T is a finite set. Thus we may suppose that
T 6¼ ;. The proof is by induction on the number of elements in S.
If S has 1 element, then the only nonempty subset T of S must coincide with S, so T is a
finite set.
Suppose that every nonempty subset of a set with k elements is finite. Now let S be a
set having k þ 1 elements (so there exists a bijection f of N kþ1 onto S), and let T  S. If
f ð k þ 1Þ 2= T, we can consider T to be a subset of S1 :¼ Snff ð k þ 1Þg, which has k
elements by Theorem 1.3.4(b). Hence, by the induction hypothesis, T is a finite set.
On the other hand, if f ð k þ 1Þ 2 T, then T 1 :¼ Tnff ð k þ 1Þg is a subset of S1. Since
S1 has k elements, the induction hypothesis implies that T1 is a finite set. But this implies
that T ¼ T 1 [ ff ð k þ 1Þg is also a finite set.
(b) This assertion is the contrapositive of the assertion in (a). (See Appendix A for a
discussion of the contrapositive.) Q.E.D.

Countable Sets
We now introduce an important type of infinite set.

1.3.6 Definition (a) A set S is said to be denumerable (or countably infinite) if there
exists a bijection of N onto S.
(b) A set S is said to be countable if it is either finite or denumerable.
(c) A set S is said to be uncountable if it is not countable.

From the properties of bijections, it is clear that S is denumerable if and only if there
exists a bijection of S onto N. Also a set S1 is denumerable if and only if there exists a
18 CHAPTER 1 PRELIMINARIES

bijection from S1 onto a set S2 that is denumerable. Further, a set T1 is countable if and only
if there exists a bijection from T1 onto a set T2 that is countable. Finally, an infinite
countable set is denumerable.

1.3.7 Examples (a) The set E :¼ f2n : n 2 N g of even natural numbers is denumerable,
since the mapping f : N ! E defined by f (n) :¼ 2n for n 2 N is a bijection of N onto E.
Similarly, the set O :¼ f2n  1 : n 2 N g of odd natural numbers is denumerable.
(b) The set Z of all integers is denumerable.
To construct a bijection of N onto Z, we map 1 onto 0, we map the set of even natural
numbers onto the set N of positive integers, and we map the set of odd natural numbers onto
the negative integers. This mapping can be displayed by the enumeration:
Z ¼ f0; 1;  1; 2;  2; 3;  3; . . .g:
(c) The union of two disjoint denumerable sets is denumerable.
Indeed, if A ¼ fa1 ; a2 ; a3 ; . . .g and B ¼ fb1 ; b2 ; b3 ; . . .g, we can enumerate the
elements of A [ B as:
&
a 1 ; b 1 ; a2 ; b 2 ; a 3 ; b 3 ; . . . :

1.3.8 Theorem The set N  N is denumerable.

Informal Proof. Recall that N  N consists of all ordered pairs (m, n), where m, n 2 N.
We can enumerate these pairs as:
ð1; 1Þ; ð1; 2Þ; ð2; 1Þ; ð1; 3Þ; ð2; 2Þ; ð3; 1Þ; ð1; 4Þ; . . . ;
according to increasing sum m þ n, and increasing m. (See Figure 1.3.1.) Q.E.D.

The enumeration just described is an instance of a ‘‘diagonal procedure,’’ since we

move along diagonals that each contain finitely many terms as illustrated in Figure 1.3.1.
The bijection indicated by the diagram can be derived as follows. We first notice that
the first diagonal has one point, the second diagonal has two points, and so on, with k points
in the kth diagonal. Applying the formula in Example 1.2.4(a), we see that the total number
of points in diagonals 1 through k is given by
cðkÞ ¼ 1 þ 2 þ þ k ¼ 12 kðk þ 1Þ

Figure 1.3.1 The set N  N

 1.3 FINITE AND INFINITE SETS 19

The point (m, n) lies in the kth diagonal when k ¼ m þ n  1, and it is the mth point in
that diagonal as we move downward from left to right. (For example, the point (3, 2) lies
in the 4th diagonal since 3 þ 2  1 ¼ 4, and it is the 3rd point in that diagonal.) Therefore,
in the counting scheme displayed by Figure 1.3.1, we count the point (m, n) by first
counting the points in the first k  1 ¼ m þ n  2 diagonals and then adding m. Therefore,
the counting function h : N  N ! N is given by
hðm; nÞ :¼ cðm þ n  2Þ þ m
¼ 12 ðm þ n  2Þðm þ n  1Þ þ m:
For example, the point (3, 2) is counted as number hð3; 2Þ ¼ 12 3 4 þ 3 ¼ 9, as
shown by Figure 1.3.1. Similarly, the point (17, 25) is counted as number h (17, 25) ¼ c(40)
þ 17 ¼ 837.
This geometric argument leading to the counting formula has been suggestive and
convincing, but it remains to be proved that h is, in fact, a bijection of N  N onto N.
A detailed proof is given in Appendix B.
The construction of an explicit bijection between sets is often complicated. The next
two results are useful in establishing the countability of sets, since they do not involve
showing that certain mappings are bijections. The first result may seem intuitively clear,
but its proof is rather technical; it will be given in Appendix B.

1.3.9 Theorem Suppose that S and T are sets and that T  S.

(a) If S is a countable set, then T is a countable set.
(b) If T is an uncountable set, then S is an uncountable set.

1.3.10 Theorem The following statements are equivalent:

(a) S is a countable set.
(b) There exists a surjection of N onto S.
(c) There exists an injection of S into N.

Proof. (a) ) (b) If S is finite, there exists a bijection h of some set N n onto S and we
define H on N by
hðkÞ for k ¼ 1; . . . ; n;
HðkÞ :¼
hðnÞ for k > n:
Then H is a surjection of N onto S.
If S is denumerable, there exists a bijection H of N onto S, which is also a surjection of
N onto S.
(b) ) (c) If H is a surjection of N onto S, we define H 1 : S ! N by letting H1(s) be the
least element in the set H 1 ðsÞ :¼ fn 2 N : HðnÞ ¼ sg. To see that H1 is an injection of S
into N, note that if s, t 2 S and nst :¼ H 1 ðsÞ ¼ H 1 ðtÞ, then s ¼ H(nst) ¼ t.
(c) ) (a) If H1 is an injection of S into N, then it is a bijection of S onto H 1 ðSÞ  N. By
Theorem 1.3.9(a), H1(S) is countable, whence the set S is countable. Q.E.D.

1.3.11 Theorem The set Q of all rational numbers is denumerable.

Proof. The idea of the proof is to observe that the set Q þ of positive rational numbers is
contained in the enumeration:
1 1 2 1 2 3 1
1; 2; 1; 3; 2; 1; 4;...;
20 CHAPTER 1 PRELIMINARIES

which is another ‘‘diagonal mapping’’ (see Figure 1.3.2). However, this mapping is not an
injection, since the different fractions 12 and 24 represent the same rational number.

Figure 1.3.2 The set Q þ

To proceed more formally, note that since N  N is countable (by Theorem 1.3.8), it
follows from Theorem 1.3.10(b) that there exists a surjection f of N onto N  N. If g :
N  N ! Q þ is the mapping that sends the ordered pair (m, n) into the rational number
having a representation m=n, then g is a surjection onto Q þ . Therefore, the composition g f
is a surjection of N onto Q þ , and Theorem 1.3.10 implies that Q þ is a countable set.
Similarly, the set Q  of all negative rational numbers is countable. It follows as in
Example 1.3.7(b) that the set Q ¼ Q  [ f0g [ Q þ is countable. Since Q contains N, it
must be a denumerable set. Q.E.D.

The next result is concerned with unions of sets. In view of Theorem 1.3.10, we need
not be worried about possible overlapping of the sets. Also, we do not have to construct a
bijection.
S1
1.3.12 Theorem If Am is a countable set for each m 2 N, then the union A :¼ m¼1 Am
is countable.

Proof. For each m 2 N, let wm be a surjection of N onto Am. We define b : N  N ! A by

bðm; nÞ :¼ wm ðnÞ:
We claim that b is a surjection. Indeed, if a 2 A, then there exists a least m 2 N such that
a 2 Am, whence there exists a least n 2 N such that a ¼ wm ðnÞ. Therefore, a ¼ b (m, n).
Since N  N is countable, it follows from Theorem 1.3.10 that there exists a surjection
f : N ! N  N whence b f is a surjection of N onto A. Now apply Theorem 1.3.10 again
to conclude that A is countable. Q.E.D.

Remark A less formal (but more intuitive) way to see the truth of Theorem 1.3.12 is to
enumerate the elements of Am, m 2 N, as:
A1 ¼ fa11 ; a12 ; a13 ; . . .g;
A2 ¼ fa21 ; a22 ; a23 ; . . .g;
A3 ¼ fa31 ; a32 ; a33 ; . . .g;
:
 1.3 FINITE AND INFINITE SETS 21

We then enumerate this array using the ‘‘diagonal procedure’’:

a11 ; a12 ; a21 ; a13 ; a22 ; a31 ; a14 ; . . . ;

as was displayed in Figure 1.3.1.

Georg Cantor
Georg Cantor (1845–1918) was born in St. Petersburg, Russia. His father,
a Danish businessman working in Russia, moved the family to Germany
several years later. Cantor studied briefly at Zurich, then went to the
University of Berlin, the best in mathematics at the time. He received his
doctorate in 1869, and accepted a position at the University of Halle,
where he worked alone on his research, but would occasionally travel the
seventy miles to Berlin to visit colleagues.
Cantor is known as the founder of modern set theory and he was the first to study the
concept of infinite set in rigorous detail. In 1874 he proved that Q is countable and, in
contrast, that R is uncountable (see Section 2.5), exhibiting two kinds of infinity. In a series of
papers he developed a general theory of infinite sets, including some surprising results. In
1877 he proved that the two-dimensional unit square in the plane could be put into one-one
correspondence with the unit interval on the line, a result he sent in a letter to his colleague
Richard Dedekind in Berlin, writing ‘‘I see it, but I do not believe it.’’ Cantor’s Theorem on
sets of subsets shows there are many different orders of infinity and this led him to create a
theory of ‘‘transfinite’’ numbers that he published in 1895 and 1897. His work generated
considerable controversy among mathematicians of that era, but in 1904, London’s Royal
Society awarded Cantor the Sylvester Medal, its highest honor.
Beginning in 1884, he suffered from episodes of depression that increased in severity as the
years passed. He was hospitalized several times for nervous breakdowns in the Halle Nervenklinik
and spent the last seven months of his life there.

We close this section with one of Cantor’s more remarkable theorems.

1.3.13 Cantor’s Theorem If A is any set, then there is no surjection of A onto the set
PðAÞ of all subsets of A.

Proof. Suppose that w : A ! PðAÞ is a surjection. Since w(a) is a subset of A, either a

belongs to w(a) or it does not belong to this set. We let

D :¼ fa 2 A : a 2
= wðaÞg:

Since D is a subset of A, if w is a surjection, then D ¼ wða0 Þ for some a0 2 A.

We must have either a0 2 D or a0 2 = D. If a0 2 D, then since D ¼ wða0 Þ, we must have
a0 2 wða0 Þ, contrary to the definition of D. Similarly, if a0 2
= D, then a0 2
= wða0 Þ so that
a0 2 D, which is also a contradiction.
Therefore, w cannot be a surjection. Q.E.D.

Cantor’s Theorem implies that there is an unending progression of larger and larger
sets. In particular, it implies that the collection PðNÞ of all subsets of the natural numbers
N is uncountable.
22 CHAPTER 1 PRELIMINARIES

Exercises for Section 1.3

1. Prove that a nonempty set T1 is finite if and only if there is a bijection from T1 onto a finite set T2.
2. Prove parts (b) and (c) of Theorem 1.3.4.
3. Let S :¼ {1, 2} and T :¼ {a, b, c}.
(a) Determine the number of different injections from S into T.
(b) Determine the number of different surjections from T onto S.
4. Exhibit a bijection between N and the set of all odd integers greater than 13.
5. Give an explicit definition of the bijection f from N onto Z described in Example 1.3.7(b).
6. Exhibit a bijection between N and a proper subset of itself.
7. Prove that a set T1 is denumerable if and only if there is a bijection from T1 onto a denumerable
set T2.
8. Give an example of a countable collection of finite sets whose union is not finite.
9. Prove in detail that if S and T are denumerable, then S [ T is denumerable.
10. (a) If (m, n) is the 6th point down the 9th diagonal of the array in Figure 1.3.1, calculate its
number according to the counting method given for Theorem 1.3.8.
(b) Given that h(m, 3) ¼ 19, find m.
11. Determine the number of elements in PðSÞ, the collection of all subsets of S, for each of the
following sets:
(a) S :¼ {1, 2},
(b) S :¼ {1, 2, 3},
(c) S :¼ {1, 2, 3, 4}.
Be sure to include the empty set and the set S itself in PðSÞ.
12. Use Mathematical Induction to prove that if the set S has n elements, then PðSÞ has 2n elements.
13. Prove that the collection F ðNÞ of all finite subsets of N is countable.
APPENDIX B

FINITE AND COUNTABLE SETS

We will establish the results that were stated in Section 1.3 without proof. The reader
should refer to that section for the definitions.
The first result is sometimes called the ‘‘Pigeonhole Principle.’’ It may be interpreted
as saying that if m pigeons are put into n pigeonholes and if m > n, then at least two
pigeons must share one of the pigeonholes. This is a frequently used result in combinatorial
analysis. It yields many useful consequences.

B.1 Theorem Let m; n 2 N with m > n. Then there does not exist an injection from N m
into N n .

Proof. We will prove this by induction on n.

If n ¼ 1 and if g is any map of N m ðm > 1Þ into N 1 , then it is clear that gð1Þ ¼    ¼
gðmÞ ¼ 1, so that g is not injective.
Assume that k > 1 is such that if m > k, there is no injection from N m into N k . We
will show that if m > k þ 1, there is no function h : N m ! N kþ1 that is an injection.
Case 1: If the range hðN m Þ  N k  N kþ1 , then the induction hypothesis implies that h is
not an injection of N m into N k , and therefore into N kþ1 .
Case 2: Suppose that hðN m Þ is not contained in N k . If more than one element in N m
is mapped into k þ 1, then h is not an injection. Therefore, we may assume that a single
p 2 N m is mapped into k þ 1 by h. We now define h1 : N m1 ! N k by

hðqÞ if q ¼ 1; . . . ; p  1;
h1 ðqÞ :¼
hðq þ 1Þ if q ¼ p; . . . ; m  1:
Since the induction hypothesis implies that h1 is not an injection into N k , it is easily seen
that h is not an injection into N kþ1 . Q.E.D.

We now show that a finite set determines a unique number in N.

1.3.2 Uniqueness Theorem If S is a finite set, then the number of elements in S is a

unique number in N.

Proof. If the set S has m elements, there exists a bijection f 1 of N m onto S. If S also has n
elements, there exists a bijection f 2 of N n onto S. If m > n, then (by Exercise 21
2  f 1 , is a bijection of N m onto N n , which contradicts Theorem B.1.
of Section 1.1) f 1
1  f 2 is a bijection of N n onto N m , which contradicts Theorem B.1.
If n > m, then f 1
Therefore we have m ¼ n. Q.E.D.

B.2 Theorem If n 2 N, there does not exist an injection from N into N n .

Proof. Assume that f : N ! N n is an injection, and let m :¼ n þ 1. Then the restriction of

f to N m  N is also an injection into N n . But this contradicts Theorem B.1. Q.E.D.

357
358 APPENDIX B FINITE AND COUNTABLE SETS

1.3.3 Theorem The set N of natural numbers is an infinite set.

Proof. If N is a finite set, there exists some n 2 N and a bijection f of N n onto N. In this
case the inverse function f 1 is a bijection (and hence an injection) of N onto N n . But this
contradicts Theorem B.2. Q.E.D.

We will next establish Theorem 1.3.8. In connection with the array displayed in
Figure 1.3.1, the function h was defined by hðm; nÞ ¼ cðm þ n  2Þ þ m, where
cðkÞ ¼ 1 þ 2 þ    þ k ¼ 12 kðk þ 1Þ. We now prove that the function h is a bijection.

1.3.8 Theorem The set N  N is denumerable.

Proof. We will show that the function h is a bijection.

(a) We first show that h is injective. If ðm; nÞ 6¼ ðm0 ; n0 Þ, then either (i) m þ n 6¼
m þ n0 , or (ii) m þ n ¼ m0 þ n0 and m 6¼ m0 .
0

In case (i), we may suppose m þ n < m0 þ n0 . Then, using formula (1), the fact that c
is increasing, and m0 > 0, we have
hðm; nÞ ¼ c ðm þ n  2Þ þ m  c ðm þ n  2Þ þ ðm þ n  1Þ
¼ c ðm þ n  1Þ  c ðm0 þ n0  2Þ
< c ðm0 þ n0  2Þ þ m0 ¼ hðm0 ; n0 Þ:
In case (ii), if m þ n ¼ m0 þ n0 and m 6¼ m0 , then
hðm; nÞ  m ¼ cðm þ n  2Þ ¼ cðm0 þ n0  2Þ ¼ hðm0 ; n0 Þ  m0 ;
whence hðm; nÞ 6¼ hðm0 ; n0 Þ.
(b) Next we show that h is surjective.
Clearly hð1; 1Þ ¼ 1. If p 2 N with p 2, we will find a pair ðmp ; np Þ 2 N  N with
hðmp ; np Þ ¼ p. Since p < c ðpÞ, then the set Ep :¼ fk 2 N : p  c ðkÞg is nonempty.
Using the Well-Ordering Property 1.2.1, we let kp > 1 be the least element in Ep . (This
means that p lies in the kp th diagonal.) Since p 2, it follows from equation (1) that
c ðkp  1Þ < p  c ðkp Þ ¼ c ðkp  1Þ þ kp :
Let mp :¼ p  c ðkp  1Þ so that 1  mp  kp , and let np :¼ kp  mp þ 1 so that 1 
np  kp and mp þ np  1 ¼ kn . Therefore,
hðmp ; np Þ ¼ cðmp þ np  2Þ þ mp ¼ c ðkp  1Þ þ mp ¼ p:
Thus h is a bijection and N  N is denumerable. Q.E.D.

The next result is crucial in proving Theorems 1.3.9 and 1.3.10.

B.3 Theorem If A  N and A is infinite, there exists a function w : N ! A such that

wðn þ 1Þ > wðnÞ n for all n 2 N. Moreover, w is a bijection of N onto A.

Proof. Since A is infinite, it is not empty. We will use the Well-Ordering Property 1.2.1 of
N to give a recursive definition of w.
Since A 6¼ 0, there is a least element of A, which we define to be wð1Þ; therefore,
wð1Þ 1.
Since A is infinite, the set A1 :¼ Anfwð1Þg is not empty, and we define wð2Þ to be least
element of A1 . Therefore wð2Þ > wð1Þ 1, so that wð2Þ 2.
 APPENDIX B FINITE AND COUNTABLE SETS 359

Suppose that w has been defined to satisfy wðn þ 1Þ > wðnÞ n for n ¼ 1; . . . ; k  1,
whence wðkÞ > wðk  1Þ k  1 so that wðkÞ k. Since the set A is infinite, the set
Ak :¼ Anfwð1Þ; . . . ; wðkÞg
is not empty and we define wðk þ 1Þ to be the least element in Ak . Therefore wðk þ 1Þ >
wðkÞ, and since wðkÞ k, we also have wðk þ 1Þ k þ 1. Therefore, w is defined on all
of N.
We claim that w is an injection. If m > n, then m ¼ n þ r for some r 2 N. If
r ¼ 1, then wðmÞ ¼ wðn þ 1Þ > wðnÞ. Suppose that wðn þ kÞ > wðnÞ; we will show that
wðn þ ðk þ 1ÞÞ > wðnÞ. Indeed, this follows from the fact that wðn þ ðk þ 1ÞÞ ¼ wððnþ
kÞ þ 1Þ > wðn þ kÞ > wðnÞ. Since wðmÞ > wðnÞ whenever m > n, it follows that w is an
injection.
~ :¼ AnwðNÞ is not empty,
We claim that w is a surjection of N onto A. If not, the set A
~
and we let p be the least element in A. We claim that p belongs to the set fwð1Þ; . . . ; wðpÞg.
Indeed, if this is not true, then
p 2 Anfwð1Þ; . . . ; wðpÞg ¼ Ap ;
so that wðp þ 1Þ, being the least element in Ap , must satisfy wðp þ 1Þ  p. But this
~ is empty and w is a surjection
contradicts the fact that wðp þ 1Þ > wðpÞ p. Therefore A
onto A. Q.E.D.

B.4 Theorem If A  N, then A is countable.

Proof. If A is finite, then it is countable, so it suffices to consider the case that A is infinite.
In this case, Theorem B.3 implies that there exists a bijection w of N onto A, so that A is
denumerable and, therefore, countable. Q.E.D.

1.3.9 Theorem Suppose that S and T are sets and that T  S.

(a) If S is a countable set, then T is a countable set.
(b) If T is an uncountable set, then S is an uncountable set.

Proof. (a) If S is a finite set, it follows from Theorem 1.3.5(a) that T is finite, and
therefore countable. If S is denumerable, then there exists a bijection c of S onto N. Since
cðSÞ  N, Theorem B.4 implies that cðSÞ is countable. Since the restriction of c to T is a
bijection onto cðTÞ and cðTÞ  N is countable, it follows that T is also countable.
(b) This assertion is the contrapositive of the assertion in (a). Q.E.D.
 2.5 INTERVALS 49

then for any e > 0, there exists an m 2 N such that 0  h  j  bm  am < e. Since this
holds for all e > 0, it follows from Theorem 2.1.9 that h  j ¼ 0. Therefore, we conclude
that j ¼ h is the only point that belongs to In for every n 2 N. Q.E.D.

The Uncountability of R
The concept of a countable set was discussed in Section 1.3 and the countability of the set
Q of rational numbers was established there. We will now use the Nested Interval Property
to prove that the set R is an uncountable set. The proof was given by Georg Cantor in 1874
in the first of his papers on infinite sets. He later published a proof that used decimal
representations of real numbers, and that proof will be given later in this section.

2.5.4 Theorem The set R of real numbers is not countable.

Proof. We will prove that the unit interval I :¼ ½0; 1 is an uncountable set. This implies
that the set R is an uncountable set, for if R were countable, then the subset I would also be
countable. (See Theorem 1.3.9(a).)
The proof is by contradiction. If we assume that I is countable, then we can enumerate
the set as I ¼ fx1 ; x2 ; . . . ; xn ; . . .g. We first select a closed subinterval I 1 of I such that
x1 2
= I 1 , then select a closed subinterval I2 of I1 such that x2 2 = I 2 , and so on. In this way, we
obtain nonempty closed intervals
I1 I2  In 
such that I n  I and xn 2
= I n for all n. The Nested Intervals Property 2.5.2 implies that there
exists a point j 2 I such that j 2 I n for all n. Therefore j 6¼ xn for all n 2 N, so the
enumeration of I is not a complete listing of the elements of I, as claimed. Hence, I is an
uncountable set. Q.E.D.

The fact that the set R of real numbers is uncountable can be combined with the fact that
the set Q of rational numbers is countable to conclude that the set RnQ of irrational numbers
is uncountable. Indeed, since the union of two countable sets is countable (see 1.3.7(c)), if
RnQ is countable, then since R ¼ Q [ ðRnQ Þ, we conclude that R is also a countable set,
which is a contradiction. Therefore, the set of irrational numbers RnQ is an uncountable set.

Note: The set of real numbers can also be divided into two subsets of numbers called
algebraic numbers and transcendental numbers. A real number is called algebraic if it is a
solution of a polynomial equation PðxÞ ¼ 0 where all the coefficients of the polynomial P
are integers. A real number is called transcendental if it is not an algebraic number. It can
be proved that the set of algebraic numbers is countably infinite, and consequently the set of
transcendental numbers is uncountable. The numbers p and e are transcendental numbers,
but the proofs of these facts are very deep. For an introduction to these topics, we refer the
interested reader to the book by Ivan Niven listed in the References.

y
Binary Representations
We will digress briefly to discuss informally the binary (and decimal) representations of real
numbers. It will suffice to consider real numbers between 0 and 1, since the representations
for other real numbers can then be obtained by adding a positive or negative number.

y
The remainder of this section can be omitted on a first reading.