CHAPTER 3

Normed Linear Spaces and Banach Spaces

BY

Prof. M. Saha

Professor of Mathematics
The University of Burdwan
West Bengal, India
E-mail : mantusaha.bu@gmail.com
Chapter 3 Normed Linear Spaces and Banach Spaces

Introduction and Objectives

Particularly useful and important metric spaces are obtained if we take a linear space
and define on it a metric by means of a norm. The resulting space is called a normed
linear spaces. If it is a complete metric space then it is called a Banach space. The
theory of normed linear spaces, in particular Banach spaces and the theory of linear
operators on them are the most highly developed parts of the functional analysis. The
present chapter is concerned with the basic ideas of these theories.
This chapter is divided into four Modules:
The Module 1 is discussed with some basic preliminaries and some results on linear
spaces.
The Module 2 is mainly focussed on the concept of normed linear spaces. Some useful
results on convergence of a sequence are shown.
Some examples of Banach spaces are given in Module 3.
Module 4 is dealt with mostly completeness property of a normed linear space. Char-
acterization of completeness of a normed linear spaces are shown by some important
results.

2
 Module-1: Linear Spaces

Definition 3.1.1: Let R(C) denote the field of reals (complex numbers) that are also
called scalars. A linear space (vector space) V is a collection of objects called vectors
satisfying following conditions:

1. V is additivity an abelian (commutative) group, the identity element is the zero

vector (null vector) of V denoted by θ or 0.
¯
2. For every pair (α, v), α being a scalar and v ∈ V , there is a vector, denoted by
α · v (not vα), called a scalar multiple of v such that

(i) 1 · v = v for all v ∈ V

(ii) α · (u + v) = α · u + α · v, for all scalars α and for all vectors u, v ∈ V

(iii) (α + β) · v = α · v + β · v, for all scalars α and β and for all vectors v ∈ V

(iv) α · (β · v) = (αβ) · v, for all scalars α and β and for all vectors v ∈ V .

We generally denote α.v by αv.

Example 3.1.1: Let Rn be the collection of all n-tuples of reals like x = (x1 , x2 , · · · , xn );
xi ∈ R. Thus Rn becomes a linear space with respect to real scalar field where addition
of vectors and scalar multiplication of vectors are defined as x + y = (x1 , x2 , · · · , xn ) +
(y1 , y2 , · · · , yn ) = (x1 + y1 , x2 + y2 , · · · , xn + yn ) ∈ Rn , where y = (y1 , y2 , · · · , yn ) ∈ Rn
and α · x = α(x1 , x2 , · · · , xn ) = (αx1 , αx2 , · · · , αxn ) ∈ Rn ; x ∈ R and α be any scalar.

3
 Chapter 3 Normed Linear Spaces and Banach Spaces

Here Rn is also Euclidean n-spaces with zero vector 0 = (0, 0, · · · , 0, · · · ) (that is

¯
all coordinates are zero), and it is called a real linear space.

Example 3.1.2: Cn be the collection of all n-tuples of complex scalars like z =

(z1 , z2 , · · · , zn ); zi ∈ C. Thus Cn becomes a linear space with complex scalar field
where addition of vectors and scalar multiplication of vectors are defined as
z + z ′ = (z1 , z2 , · · · , zn ) + (z1′ , z2′ , · · · , zn′ ) = (z1 + z1′ , z2 + z2′ , · · · , zn + zn′ ) where
z ′ = (z1′ , z2′ , · · · , zn′ ) ∈ Cn and
αz = α(z1 , z2 , · · · , zn ) = (αz1 , αz2 , · · · , αzn ) and α be any complex scalar.
Here Cn is also called Euclidean n-space over the field of complex numbers with the
zero vector 0 = (0, 0, · · · , 0).
¯

Example 3.1.3: Let c[a, b] denote the collection of all real valued continuous functions
over a closed interval [a, b]. Then c[a, b] is a real linear space (corresponding scalar field
being that of reals) where vector sum and scalar multiplication are defined as under:

(f + g)(t) = f (t) + g(t), a ≤ t ≤ b and f, g ∈ c[a, b]

and (αf )(t) = αf (t); a ≤ t ≤ b and α any real scalar.

Since the sum of two continuous functions is a continuous functions and so is a scalar
multiple of continuous function, we see that f + g, αf ∈ c[a, b], where f, g ∈ c[a, b] and
α being any scalar. Here the zero vector of c[a, b] is the zero function denoted by Θ,
that is Θ(t) = 0 for all t ∈ [a, b].
There are many other linear spaces like the sequence spaces lp (1 ≤ p ≤ ∞), space of
all bounded sequences l∞ , space of all convergent sequences c, space c0 of all convergent
sequences converging to zero, space of polynomials p[a, b], space of all square integrable
functions L2 [a, b] etc. That we encounter in our discussion to follow.

Definition 3.1.2: (a) If A and B are subsets of a linear space V then A + B =

{a + b : a ∈ A and b ∈ B}
(b) For any scalar λ, λA = {λa : a ∈ A}.
The subset A − B = A + (−1)B; and for scalar λ = 0; 0.A = {0}. Also A + B =
¯
B + A; because vector addition is commutative. However A − B ̸= B − A. In fact
if A = {(1, 0)}, B = {(0, 0)} in R2 , we find A − B = {(1, 0)} and B = {(−1, 0)}.

4
 Chapter 3 Normed Linear Spaces and Banach Spaces

It is to be noted that A + A ̸= 2A. In fact if A = {(1, 0), (0, 1)} ⊂ R2 . Then

A + A = {(2, 0), (0, 2), (1, 1)} and 2A = {(2, 0), (0, 2)}.
Given a fixed member b ∈ V , b + A = {b + a : a ∈ A} is called translate of A.
Let X denote a linear space over reals/complex scalars. Given x1 , x2 , · · · , xn ∈ X
∑
n
and for a given set of scalars α1 , α2 , · · · , αn , the vector αi xi of X is called a linear
i=1
combination of x1 , x2 , · · · , xn .
A subset C1 of X is said to span or generate X if and only if every member of X
is a linear combination of some elements of C1 .
Elements x1 , x2 , · · · , xn of X is said to be linearly dependent if and only if there
are corresponding member of scalars α1 , α2 , · · · , αn not all zero such that
∑
n
αi xi = 0.
i=1
¯

A finite number of elements x1 , x2 , · · · , xn of X are said to be linearly independent if

they are not linearly dependent, that is to say
∑
n
αi xi = 0 implies α1 = α2 = · · · = αn = 0.
i=1
¯

An arbitrary system of elements of X is called linearly independent if every finite subset

of the given system becomes linearly independent.
Note that if a set of vectors in X contains a linearly dependent subset, then whole
set becomes linearly dependent.
But a linearly independent set of vectors does not contains the zero vector.

Definition 3.1.3: A non empty subset E of a linear space X is called a subspace of

X if x + y ∈ E, whenever x and y are both in E and also αx ∈ E, whenever x is in E
and α is any scalar.

Example 3.1.4: Let E be any non empty subset of X. If L is the set of all linear
combinations of elements of E, then L is a subspace of X. The space L is called the
sub space spanned or generated by S.
Note that the subset {0} is a sub-space, called the Null space.
¯
Theorem 3.1.1: Let {x1 , x2 , · · · , xn } be a set of n number of vectors with xi ̸= 0.
Then {x1 , x2 , · · · , xn } is linearly dependent if and only if one of the vectors x2 , x3 , · · · , xn
say xk is in the subspace generated by x1 , x2 , · · · , xk−1 .

5
Chapter 3 Normed Linear Spaces and Banach Spaces

Proof: Suppose the given set of vectors is linearly dependent. Then there is a smallest
k with 2 ≤ k ≤ n such that x1 , x2 , · · · , xn is linearly dependent; and we have α1 x1 +
α2 x2 + · · · + αk xk = 0 with not all α’s are zero scalars. Obviously αk ̸= 0; otherwise
x1 , x2 , · · · , xk would form a linearly independent set. Consequently,
α1 α2 αk−1
xk = − x1 − x2 − · · · − xk−1 .
αk αk αk
That is, xk is in the subspace generated by x1 , x2 , · · · , xk−1 .

Conversely, if one assume that some xk is in the subspace generated by x1 , x2 , · · · , xk−1 ;

then we have,
xk = β1 x1 + β2 x2 + · · · + βk−1 xk−1 .

That is, x1 , x2 , · · · , xk are linearly dependent, which in turn we have the set {x1 , x2 , · · · , xk }
is linearly dependent.

Definition 3.1.4: If in a linear space X, there is a positive integer n such that X

contains a set of n vectors that are linearly independent, while every set of (n + 1)
vectors in X is linearly dependent, then X is called finite dimensional and n is called
dimension of X and we write Dim(X) = n

Remark 3.1.1:

(a) The null space is finite dimensional of dimension 0 (zero).

(b) If X is not finite dimensional it is called infinite dimensional.

Definition 3.1.5: A finite set B in a linear space X is called a basis of X if B is

linearly independent and if the subspace spanned (generated) by B is all of X.

Note: If {x1 , x2 , · · · , xn } is a basis for X every member x ∈ X can be expressed as

∑n
x = αi xi , where scalars αi ’s are uniquely determined; so the representation of x
i=1
interms of basis members in unique.
Let Dim(X) = n, where n ≥ 1. Then X has a basis consisting of n members. For
X certainly contains vectors x1 , x2 , · · · , xn that form a linearly independent set. Now
for any member x ∈ X, the set of vectors x1 , x2 , · · · , xn including x consisting of (n+1)
vectors must be linearly dependent. Now Theorem 3.1.1 applies to conclude that x is
in the subspace generated by x1 , x2 , · · · , xn . Hence x1 , x2 , · · · , xn form a basis of X.