MMT-006

FUNCTIONAL ANALYSIS
Indira Gandhi National Open
University School of Sciences

Block

1
NORMED LINEAR SPACES
Block Introduction 5
Notations and Symbols 6
UNIT 1
Inner Products and Norms 9
UNIT 2
Continuous Linear Maps and Functionals 55
UNIT 3
Hahn-Banach theorems 81
 Course Design Committee
Dr. A.K Vijayrajan, Faculty Members
Kerala School of Mathematics, Calicut School of Sciences, IGNOU
Prof. Sujatha Varma
Prof. K. Parthasarathy, (Retd.) Prof. Deepika
Ramanujam Institute, University of Madras, Chennai Dr. S. Venkataraman
Prof. Parvin Sinclair
Prof. M.S. Balasubhramanian , (Retd.) Prof. Poornima Mital
University of Calicut, Kerala

Dr. Sachi Srivastava,

Delhi University, Delhi

Dr. Shravan Kumar,

IIT Delhi

Prof. V. Muruganandhan
NISER, Orissa

Block Preparation Team

Prof. K. Parthasarathy (Editor) Prof. Sujatha Varma
Ramanujan Institute, University of Madras, Chennai School of Sciences, IGNOU

Prof. M.S. Balasubramanian,

University of Calicut, Kerala

Course Coordinator: Prof. Sujatha Varma

Acknowledgements: Dr. Suvankar Biswas for his comments on the units. Mr. Santosh Kumar Pal
for word processing the Manuscript and Mr Surender Singh Chauhan for the CRC.

Disclaimer – Any materials adapted from web-based resources in this course are being used for educational purposes
only and not for commercial purpose.

, 2023

© Indira Gandhi National Open University

ISBN-
All right reserved. No part of this work may be reproduced in any form, by mimeograph or any other means, without
permission in writing from the Indira Gandhi National Open University.
Further information on the Indira Gandhi National Open University courses may be obtained from the University’s
Office at Maidan Garhi, New Delhi-110 068.
Printed and published on behalf of the Indira Gandhi National Open University, New Delhi by the Director, School of
2 Sciences.
COURSE INTRODUCTION
Welcome to the course “Functional Analysis”. This course is a 4-credit course. In this
course, we introduce you to a branch of mathematics known as Functional Analysis. Under
this topic we study function spaces which are sets of functions with additional structures.
This provides a major link between mathematics and its applications. The important
notions that are dealt in this course are Banach spaces, Hilbert spaces and linear
functionals on these spaces.

This material is divided into 4 Blocks. In Block 1, we familiarise you to some basic
concepts in Functional analysis. We introduce to the concept of a “norm” which is another
distance measuring concept like a metric. Any linear space having a norm defined on it is
called a normed linear space. Any norm function defines a metric on a normed space and
thereby any normed linear space is a metric space. A detailed study of normed spaces
requires the study of maps from one normed spaces to another normed space. In this block,
we consider maps which are linear and bounded (also called continuous). We define
continuous linear maps from one normed linear space to another. We also familiarise you
to one of the important theorems in Functional Analysis, known as Hahn-Banach theorem.

Block 2 deals with Banach spaces. Normed spaces which are complete with respect to the
metric induced by the corresponding norm are called Banach spaces. The Banach spaces
plays a crucial role in the study of function spaces. Here we consider four important
theorems – Open mapping theorem, closed graph theorem, Uniform boundedness principle
and Bounded inverse theorem. In this block we also deal with space of bounded linear
maps defined from a normed space to the space K, the scalar field of real or complex
numbers. These spaces are called dual spaces. We consider dual of dual spaces which are
called reflexive spaces.

Block 3 introduces to another type of normed spaces known as inner product spaces. An
inner product space which is complete, under the norm induced by the inner product, is
called a Hilbert space. Here we discuss the fundamental properties of inner product spaces
more specifically for Hilbert spaces. The inner product enables us to introduce the concept
of orthogonality. We discuss orthonormal sets. Another important theorem for Hilbert
spaces namely, Riesz representation theorem is studied in this block. Finally we consider
operators on Hilbert spaces. Here we define adjoint of an operator and study four important
classes of operators, namely, self-adjoint, unitary, normal operators and compact. We also
discuss an important subclasse of self-adjoint operators, viz, compact self-adjoint-
operators.

Each Block is divided into 3 or 4 units based on different themes and each unit starts with
listing of contents and objectives. In the list of objectives in each unit we have given the
main concepts and results which we thought you should have achieved after reading this
material. Each unit is divided into sections, which may be divided into sub-sections. These
sections/sub-sections are numbered sequentially, as are the exercises and important
equations, within a unit. Regarding the exercises in each unit, these have been placed at
many places in a unit. They are meant for you to assess if you have understood the
concepts and processes that have been discussed till that point. Therefore, you must
attempt to solve all the exercises in the unit as you go along, before you look up the
solution.

Since the material in the different units is heavily interlinked, there will be cross
referencing. For this we will be using the notation Sec. x.y to mean Section y of Unit x.
3
 Further, the end of an example is shown by the sign *** and the end of a proof of,
theorem/proposition/corollary is shown by the mark .

After reading the unit please go back to the objectives to see that you have not missed any
concept. Each unit also contains worked out examples. At the end, we have recapitulated
the learning points of the unit in the form of summary. This is followed by Hints/Solutions
of the exercises. We advise you to look into the answers only after you have tried them
sincerely.
e
It is important to learn the concepts in each block to understand the concepts covered in the
further blocks. Another compulsory component of this course is its assignment, which
covers the whole course. Your academic counselor at the study centre will evaluate it, and
return it to you with detailed comments. Thus, the assignment is meant to be a teaching as
well as an assessment aid. Further, you will not be allowed to take the examination of this
course till you submit your assignment response. So please submit it well in time.

Now, a word about notation. The notations and symbols used in this course material are
given in a separate list. You are advised to follow this. You may note that different text
books use different notations.

This course assumes the knowledge of linear algebra and basic metric space theory
presented in IGNOU courses on “Linear Algebra” MTE-02 of “Real Analysis” MMT-004,
Block 1. If you have any difficulty with any concept or techniques given in Chapter-1. The
course material that we have sent you is self-sufficient. If you have a problem in
understanding any portion, please ask your academic counselor for help. In addition to
these exercises, you may like to also look up some other books in the library of your study
centre, and try to solve some exercises from these books too.

Some useful books and websites are the following:

1. Kreyszig, Ervin: Introductory functional analysis and applications by John Wiley &
Sons. Inc.
2. G. Backman and L. Narici’s: Functional Analysis, Dover Publications
3. S. Kesavan: Functional Analysis, Hindustan Book Agency (Trim Series)
4. S. Kumaresan and D. Sukumar: Functional Analysis, A first course, Narosa
publications.
5. B.V. Limaye: Functional Analysis, New Age International Publishers.
6. Karen Saxe: Beginning Functional Analysis, Springer Verlag.

Best wishes for an enjoyable time on studying “Functional Analysis.”

The Course Team

4
 NOTATIONS AND SYMBOLS (used in Block 1)

BL( X ) Space of bounded linear maps from X to itself

BL( X , Y ) Space of bounded linear maps from X to Y
C Complex field
C( X ) Space of bounded continuous scalar-valued functions on X
C0 ( X ) Subspace of C (T ) consisting of functions with vanishing at infinity
Cc ( X ) Subspace of C (T ) consisting of functions with compact support
C n [ a, b] Space of scalar-valued functions having continuous n-th derivatives on
[ a, b]
C ∞ [ a, b] Space of infinitely differentiable scalar-valued functions on [a, b]
c Space of convergent sequences of scalars
c0 Subspace of c consisting of sequence converging to 0.
c00 Subspace of c consisting of sequences having at most a finite number of
non-zero terms
d ( x, E ) Distance between x and E
d ( x, y ) Distance between x and y
E0 Interior of a set E
E⊥ Orthogonal complement of E
E⊥F E orthogonal to F
E1 + E2 Sum of sets E1 and E 2
ei The i -th unit vector
T Norm of linear map F
fZ Restriction of f to Z
I Identity operator
K Scalar field which may be the real field or complex field
Kn Space of n -tuples of elements in K
lp Space of p -summable scalar sequences
l∞ Space of bounded scalar sequences
N Set of natural numbers
R Real field
Re Real part
R( F ) Range of F
rσ (T ) Spectral radius of T
span S span of the set S
span S Closure of span of S
σ( A) Spectrum of A
σ a ( A) Set of approximate eigenvalues of A
σ e ( A) Eigenspectrum of A
σ r ( A) Residual spectrum of A
T* Adjoint of T
B X (0,1) Open unit ball
B X ( a, r ) Open ball with centre a and radius r
X′ Dual space of X
X ′′ Dual of X ′
~
X Completion of X 5
 x p
p -norm of x
x∞ Sup norm of x
X ×Y Cartesian product of X and Y
X /Y Quotient space
x ⊥ E, x ⊥ y x orthogonal to E, y
x +Y Quotient norm of x + Y
( x, y ) Inner product of x and y
Y Closure of Y
Z Set of all integers
Z (F ) Zero space of F
{0} Zero-dimensional space consisting of 0 only
0/ Empty set

6
BLOCK 1 NORMED LINEAR SPACES
This is the first block of the course “Functional Analysis” MMT-006. In this block we
introduce you to the concept of normed linear spaces which are linear spaces (vector
spaces) with a new structure formed by a norm function. This structure induces a metric
structure which is compatible with the linear structure. Thus every normed linear space is a
metric space. We shall then discuss linear maps defined on these spaces which are
continuous with respect to the metric defined on it. In fact these continuous maps separates
points on the spaces. This is established by an important theorem known as Hahn Banach
theorem which is covered in this block.

This block is divided into 3 units. In Unit 1 we familiarize you with the metric induced by
the norm. This unit deals with basic properties of normed linear spaces. We shall also
introduce you to inner product spaces. These spaces are normed linear spaces, but not vice-
versa. Inner product space allows us to study the concept of angle, perpendicularly and so
on. You will learn that finite dimensional normed linear spaces have nice properties in
common with Euclidean spaces R n . Many examples of infinite dimensional normed linear
spaces are given in this unit. We call them standard spaces. These examples will be
mentioned throughout all the units in this material.

Unit 2 covers continuous linear maps from one normed linear space to another normed
linear space. We define norm of a continuous linear map and prove that the continuity and
boundedness are the same for normed linear spaces. The continuous/bounded linear map
from a normed linear space to its basic field is known as bounded/continuous linear
functional. We discuss examples of continuous linear maps defined on standard spaces and
also show how to compute their norms.

In Unit 3, we discuss two important theorems in Functional Analysis both known as,
Hahn-Banach Theorem – Hahn-Banach Extension Theorem and Hahn-Banach Separation
Theorem. These theorems are very important as they are used in proving many other
important results.

You may like to view the following Video lectures on certain topics related to the content
covered in this block.

You may like to view these video programmes.

1) https://youtu.be/KwYqKWc-Rqo
2) https://youtu.be/y97GvfAbl3U
3) https://youtu.be/bIYBliUwtTU
4) https://youtu.be/2xWpLmDE52E
5) https://www.youtube.com/watch?v=FkJMfsNg2cM&list=PLyqSpQzTE6M8mjwWBz0vXpmJQwR46JI
dS&index=10&t=471s

7
 UNIT 1

INNER PRODUCTS AND NORMS

NORMS

Structure Page No
1.1 Introduction
Objectives
1.2 Definition, Properties and Examples
1.3 Standard Examples
1.4 Subspaces
1.5 Quotient spaces
1.6 Summary
1.7 Hints/Solutions

1.1 INTRODUCTION
You are already familiar with linear spaces such as the real line R , complex
n
plane C and Euclidean spaces R in general. The importance of these spaces
is that we can define algebraic structures as well as the concept of distance or
length. On the real line, for example the modulus function, is used to measure
the distance and for other Euclidean spaces, there is an analogous distance
formula. Can we have spaces other than Euclidean spaces on which we can
have both an algebraic structure and the concept of a distance? In an attempt
to find an answer to these questions, the concept of “norm” was introduced, a
concept which is, in fact, analogous to the concept of length. Any linear space
having a norm defined on it is called a normed linear space or simply a
normed space. On Euclidean spaces, besides the concept of length, angles
and perpendicularity are also studied. An inner product, helps us in defining
these. In this unit, you will study normed linear spaces and inner product
spaces.

Objectives
The objectives of this unit are:

• To define a normed linear space and an inner product space and give
examples;
• To calculate the norm of an element in standard normed linear spaces;
R n , C n , l p , 1 ≤ p < ∞, c0 , C [ a, b], L p ( R), 1 ≤ p < ∞;
• To state when a subset of a normed linear space is a subspace;
• To give examples of subspaces of standard normed linear spaces;
• To state and prove Riesz lemma;
 Block 1 Normed Linear Spaces
• to define quotient spaces of a normed linear space.
• to explain the convergence in quotient spaces of a normed linear space X
and relate this to convergence in X ;
• to state and prove a necessary and sufficient condition for a quotient
space to be complete.

1.2 DEFINITION, PROPERTIES AND EXAMPLES

In this section we define inner product spaces and normed linear spaces and
look at some examples of these. We also discuss properties of these spaces.
You may like to view the
related video through the In what follows all vector spaces considered are over the field K , which is
following link.
either R or C . Vector spaces and linear spaces mean one and the same.
https://youtu.be/KwYqK
Wc-Rqo We start with a definition.

Definition 1: Let X be a vector space over K . An inner product on X is a

function , from X × X → K such that for all x, y, z in X and α in K , we
have:

i) x, x ≥ 0 and x, x = 0 if and only if x = 0 (Positive definite property)

ii) y, x = x, y (Conjugate Symmetry property)

iii) x + y, z = x, z + y, z and α x, y = α x, y (Linearity in the first

variable)

In the real case, conjugate symmetry is just symmetry. An inner product space
is a vector space with an inner product on it.

We have to always remember that inner product of two elements in the vector
space is a scalar.

Now let us look at some properties which we can deduce from the definition.

Proposition 1: Let , be an inner product on a linear space X .

i) , is conjugate linear in the second variable i.e., for x, y , z ∈ X and

α, β ∈ K ,

x, αy + βz = α x, y + β x, z

In this real case, ‘conjugate linear’ is same as ‘linear’.

ii) Let x ∈ X . Then x, y = 0 for all y ∈ X if and only if x = 0.

iii) Schwarz Inequality: For all x, y ∈ X ,

2
x, y ≤ x, x y , y
10
Unit 1 Inner Products and Norms
Proof: We prove this one by one.

(i) We have to show that <, > is conjugate linear in the second variable.
Let, x, y, z ∈ X and α, β ∈ K

Then by definition, x, αy + β z = αy + β z , x
(by linearity in the
= αy , x + β z , x first variable)

= α y , x + β z , x = α x , y + β x, z

Thus, <, > is conjugate linear in the second variable.

(ii) Let x ∈ X . Suppose x , y = 0 ∀ y ∈ X.

Then in particular x, x = 0  x = 0. (by definition)

Conversely suppose x = 0.

Then 0, y = 0 + 0, y = 0, y + 0 , y .

 0, y = 0.

(iii) Schwarz inequality:

Let x, y ∈ X .

Consider z = y , y x − x , y y .

Then: 0 ≤ z, z = y, y x − x, y y, y, y x − x, y y

2
= y, y x, x + x, y . x, y . y, y − y , y x, y x, y

− x, y y, y y, x

= y, y { x, x y , y − x, y
2
}
2
Since y , y ≥ 0, we get x, x y , y − x, y ≥ 0.

2
 x, y ≤ x, x y, y . for y ≠ 0.

For y = 0, both sides are 0.

***

Another important property is the following which is known as Polarization

Identify.

Proposition 2 (Polarisation Identity): Let , be an inner product on a vector

space X. Then for all x , y ∈ X ,

11
Block 1 Normed Linear Spaces

4 x, y = x + y, x + y − x − y, x − y + i x + iy, x + iy − i x − iy, x − iy in the

complex case. In the real case, only the first two terms are present.

Proof: Let us start with the right hand side (R.H.S) and simplify using linearity
in the first variable and conjugate linearity in the second variable.

In the complex case R.H.S = x + y, x + y − x − y , x − y + i x + i y, x + i y

− i x − i y, x − i y .

= x, x + x , y + y , x + y , y

− x, x + x, y + y , x − y , y

+ i x , x + x, y − y , x − i y , y

− i x, x + x, y − y, x + i y, y

= 4 x, y (after simplification)

The real case is similar and simpler.

***

Proposition 3: Let , be an inner product on a linear space X . If x , y ∈ X

and x, z = y, z for all z ∈ X , then x = y.

Proof: We are given = x , z = y , z for all z ∈ X .

Consider x , z = y , z  x, z − y, z = 0 ∀ z ∈ X .

 x − y, z = 0 ∀ z ∈ X .

 x − y = 0  x = y.

***

An inner product gives rise to the length or the norm of a vector.

Note that in the case of K n , the dot product of x = ( x1 ,..., xn ), y = ( y1 ,.....yn ) is

defined as x. y = x1 y 1 + ... + xn y n .

In the case of R 2 , we have x. y = x1 y1 + x2 y 2 and x.x = x12 + x22 . But x12 + x22 is

the square of the distance of x from the origin. This number is called the norm
of the vector and is denoted by x .

Definition 2: Let X be a linear space over K . A norm on X is a function . ,

from X to R satisfying the following:
12
Unit 1 Inner Products and Norms
For all x , y ∈ X and α ∈ K ,

i) x ≥ 0 and x = 0 if and only if x = 0 (called the positive definite

property).

ii) x + y ≤ x + y (called the triangle inequality)

iii) αx = α x (called the homogeneity property).

A normed linear space is a linear space with a norm on it.

1
2 2 2
If x, y are any two elements in R , [( x1 − y1 ) + ( x2 − y2 ) ] 2 gives the distance
between the x and y and it is nothing but x − y . In short the dot product or
the inner product gives a norm which in turn gives a distance.

Let us see an example.

Proposition 4: Let , be an inner product on a linear space X . For x ∈ X ,

1
define x = x, x 2
. Then . is a norm on X . This norm is called the norm
induced by the inner product.

Proof: For x ∈ X , x is defined as

1
x = x, x 2
.

Let us now check the properties of a norm.

(i) x = 0  x , x = 0  x = 0.

Suppose x = 0. Then x , x = 0  x = 0.

(ii) Let α ∈ K and x ∈ X .

2
Consider α x = αx, αx .

2 2
=α x .

 αx = α x .

(iii) Let x, y ∈ X .

2
Then x+ y = x + y, x + y .

= x, x + x, y + y , x + y, y

2 2
= x + y + x, y + x, y

13
Block 1 Normed Linear Spaces
2 2
= x + y + 2 Re x, y ,

2 2
≤ x + y + 2 x, y

(For a complex number z, we have Re z ≤ z )

2 2 1 1
≤ x + y + 2 x, x 2
y, y 2
(By Schwarz inequality).

2 2
= x + y +2 x y

=( x + y ) 2

Taking positive square roots on both sides, we get

x+ y ≤ x + y .

....
All the above three properties prove that is a norm on X .

***

Example 1: We know that for n = 1, 2, 3... K n is an n-dimensional vector space

over K . For x = ( x1 ,....xn ) and y = ( y1 .... y n ) define the inner product x, y as:

x, y = x1 y1 + x2 y2 + ....xn y n

(If K = R , then x, y is just x1 y1 + .... + xn yn )

Find the norm corresponding to this inner product.

Solution: We shall show that ⋅ 2 satisfies all the conditions of a norm.

Let us check one by one

i) ⋅ 2
≥ 0, since all the terms on the R.H.S. are non-negative.
⋅ 2
= 0  x k = 0 for all k. This means that x = 0 and thus the condition in
(i) is satisfied.

1/ 2
 α 2
∞
αx 2 =   αx k  = α  x k )1 / 2
2
ii)
 k =1  k =1

=α x 2

Therefore the condition (ii) is satisfied.

iii) Let x , y be the two elements in R n such that

x = ( x 1 ,..., x k ), y = ( y1 ,...., y k )
Then x + y = ( x 1 + y1 ,..., x k + y k )
14
Unit 1 Inner Products and Norms

x+y 2 = ( x k + yk )
2 1/ 2

Now we apply the Minkowski’s inequality (which we have stated as

Theorem 3 in Sec. 1.3) with p = 2

∴ x+y 2 = ( x ) + ( y )
k
2 1/ 2
k
2 1/ 2

= x2+ y 2

Therefore the condition (iii) is satisfied.

Hence ⋅ 2 is a norm of R n .
***

Example 2: Show that the Euclidean space R n is a normed space with the
1/ 2
 n 2
norm defined by x 2 =   x k  where x = ( x 1 , x 2 ,..., x n ).
 k =1 

Solution: We can check that , defined as above satisfies all the properties
of an inner product. (See E1)

(
Thus K , ,
n
)is an inner product space.
The corresponding induced norm . is given by
x = x, x
1
2
1
( n
= [ x12 + ... + xn2 ] 2 . Thus K , . is a normed linear space. )
***

We shall look into more examples in Section 1.3.

You should be able to try this exercise.

E1) Define a function ⋅ on R 3 by

[(
x = x12 + x 22 ) 3/ 2
+ x3 ]
3 1/ 3

Show that R 3 is a normed space with respect to ⋅ .

Remark 1: An inner product always induces a norm or equivalently every

inner product space is a normed linear space.

What about the converse?

Let us wait for some time and introduce more tools before we come to this
question.

Another important property that arises naturally is the ‘distance’ or the ‘metric’.
Recall that a metric on a set X is a function d : X × X to R satisfying the
15
Block 1 Normed Linear Spaces
following: For x, y , z ∈ X ,

i) d ( x, y ) ≥ 0 and d ( x, y ) = 0 if and only if x = y

ii) d ( y, x) = d ( x, y ) (called the symmetry property)

iii) d ( x, y ) ≤ d ( x, z ) + d ( z , y ) (called the triangle inequality)

Then we have the following result.

Proposition 5: Let . be a norm on a linear space X . For x, y ∈ X , define

d ( x, y) = x − y . Then d is a metric on X . This metric is called the metric
induced by the norm.

Proof: The proof is left as an exercise for you to verify (See E2).

Remark 2: Thus if we start with an inner product on a linear space, this inner
product induces a norm which in turn induces a metric.

Once we have a metric or a distance on a set, we can talk of convergence of

sequences, Cauchy sequences, completeness properties. Further, a metric
always gives a topology and consequently topological properties can be
studied.

We shall be talking about open sets, closed sets, compact sets, completeness
etc in normed linear spaces/inner product spaces.

All these properties are discussed with respect to the metric induced by the
norm.

Inner Product → Norm → Metric.

Remark 3: If , is an inner product on a linear space X and . is the

induced norm on X , then the Schwarz inequality and the polarization identity
can be written as follows:

2 2 2
Schwarz Inequality: x, y ≤ x y or x, y ≤ x y .

2 2 2 2
Polarization Identity: 4 x, y = x + y − x − y + i x + iy − i x − iy .
(complex case).

Now let us look at some basic properties of inner products and norms. Since
d ( x, y) := x − y is the metric induced by the norm, we shall use only x − y to
denote the metric.

Theorem 1: Inner products and norms are continuous functions.

First let us see what we have to prove. Then they becomes easy.

Remember both inner products and norms are functions. So what we have to
prove is that these functions are continuous. Here we shall use a
characterization of continuity of functions on metric spaces given by the
following: ( X , d X ), (Y , d Y ) are metric spaces. A function f : X → Y is
16
Unit 1 Inner Products and Norms
continuous if and only if xn → x in X  f ( xn ) → f ( x) in Y i.e.
d X ( x n , x ) → 0  d Y ( f ( x n ), f ( x )) → 0 as n → ∞ .

In the case of norm, we can think of x as f (x) and in the case of inner
product we can think of x, y as g ( x, y ), a function of two variables. Thus what
we have to prove is

xn → x in X  xn → x in R and

xn → x & yn → y in X  xn , yn → x, y in K

Now we are ready to prove the theorem

Proof: First we prove that the norm is a continuous function.

Let xn → x in X . We have to prove that x n → x in R

Consider: x, y∈ X .

Then: x = x − y + y  x ≤ x − y + y

 x − y ≤ x−y (1)

Similarly − y = x − y − x.

 y  x−y + −x = x−y + x

 y − x ≤ x− y

 −( x − y )≤ x−y (2)

From Eqn. (1) and (2), we get that x − y ≤ x− y .

Since x n → x in X , we have x n − x → 0 as n → ∞.

Now, we have to show xn → x in R, ie. xn − x → 0 as n → ∞.

By what we have proved above xn − x ≤ xn − x → 0 as n → ∞ .

 . is a continuous function on X .

Now let us prove that the inner product is continuous.

Let x n → x and y n → y in X .

To prove that xn , yn → x, y in R , then we have

17
Block 1 Normed Linear Spaces

x n , y n − x, y = xn , y n − x, y n + x, y n − x, y
= x n − x, y n + x, y n − y ≤ xn − x, y n + x, y n − y
1 1 1 1
≤ xn − x 2
yn 2
+ x 2
yn − y 2
(By Schwarz inequality)

We have that { y n } is a convergent sequence in X and so y n is convergent in

R . Hence it is a bounded sequence there exists k > 0 such that y n ≤ k ∀ n.

Thus we get

1 1 1 1
xn yn − x, y ≤ k 2
xn − x 2
+ x 2
yn − y 2

 x n y n → x, y as n → ∞.

 , is a continuous function on X × X .
***

Another important result is the following.

Proposition 6: (Parallelogram Law): Let , be an inner product on a linear

space X and . , the induced norm. Then for all x, y ∈ X ,

2 2 2 2
x + y + x − y = 2( x + y )

Proof: For x , y ∈ X ,

2 2
consider x + y + x−y

= x + y, x + y + x − y, x − y

= x , x + x , y + y , x + y , y + x, x − x , y − y , x + y , y

= 2 x, x + 2 y , y

(
=2 x + y
2 2
)
***

Remark 4: If we consider x, y as the sides of a parallelogram in the plane,

then the vectors x + y and x − y are the diagonals of the parallelogram. The
above results then reflect a property of a parallelogram, namely, the sum of
the squares of the length of the diagonals of a parallelogram equals the sum of
all the squares of the sides of a parallelogram.

Remark 5: Another important implication of the above result is: If a norm in a

normed linear space does not satisfy the parallelogram law, then the norm
cannot be obtained from an inner product i.e. an inner product always induces
a norm and the converse is true only if the norm satisfies the parallelogram

18
Unit 1 Inner Products and Norms
law. However if the norm does satisfy the parallelogram law, then we can
define an inner product as:

x, y =
1
4
[ 2 2 2
x + y − x − y + i x + iy − i x − iy
2
] in the complex case.
So far we have discussed important properties of inner products and norms.
We have considered the corresponding induced metric also.

Why don’t you try this exercise now?

E2) Let X be a normed linear space. Show that for any x, y , the function
d ( x, y ) = x − y defines a metric on X .

Next we shall see important examples of normed linear spaces and inner
products.

1.3 STANDARD EXAMPLES

In this section we shall give specific examples of normed linear spaces.

Example 3: Let X = c 00 , be the set of all scalar sequences which have only a
finite number of non zero entries. Define a norm on X which makes it a
normed linear space.

Solution: We shall first check if X is a vector space (linear space).

For x = ( x1 , x2 ...), y = ( y1 , y2 ,....) and α ∈ K , define

x + y = ( x1 + y1 , x 2 + y 2 ,....)

α x = (α x1 , α x2 ,....)

Then vector addition and scalar multiplication as defined above gives a vector
space structure in c00.

Also the collection of vectors, (e1 , e2 , e3 ...) where e n is the sequence having 1 at
the nth place and 0 everywhere else, for n = 1, 2.. is a countable collection of
linearly independent vectors in c00. So c00 is an infinite dimensional vector
space.

Now for x = ( x1 , x2 , x3 ....), y = ( y1 , y2 , y3 ,....) define x, y as:

∞
x , y =  xn y n .
n=1

∞
Again, if K = R , x, y is just x
n =1
n yn .

19
Block 1 Normed Linear Spaces
Since x, y have only finitely many non zero terms, the sum is a finite sum and
so x, y is well defined.

The properties of the inner product can be checked as in Example 1.

(
Thus c00 , , ) is an inner product space.
Under the corresponding induced norm . , given by

1 1
1 ∞  2
∞ 2
2
x = x, x 2
=  xn x n  =  xn  ,
 n=1   n=1 

(X , . ) is a normed linear space.

***

Before we see more examples, we prove certain inequalities which are very
useful in the proving many results.

Before we state the inequalities, we prove another inequality which is used for
proving the inequalities.

Theorem 2 (Young’s inequality for products): Let a ≥ 0, b ≥ 0, then

a p bq
ab ≤ + (4)
p q

1 1
where p, q are such that + = 1.
p q

Proof: If b = 0, the result is true.

So suppose b ≠ 0 i.e., b > 0 .

Define the function x (t ) as follows:

1 t
x (t ) = + − t1/ p , t ≥ 0
q p

1 1 1/ p −1
Then x ′( t ) = − .t .
p p

1
1 1 −
= − . t q , t ≥ 0.
p p

−1
q
If t < 1, then x ′(t ) < 0 , as, t > 1.

Also, if t ≥ 1, x′(t ) ≥ 0.

Thus x (t ) has a minimal value at t = 1 and x(1) = 0.

20
Unit 1 Inner Products and Norms
Thus x(t ) ≥ 0 ∀ t ≥ 0.

1
p 1 t
Hence t ≤ + , ∀ t ≥ 0.
q p

ap
Taking t = , we get
bq

ap
a 1 p
q/p
≤ +b
b q p

Multiplying by b q , we get

q
q−
p bq a p
a .b ≤ + .
q p

1 1 q q
But + = 1  +1 = q  q − = 1
p q p p

a p bq
Hence ab ≤ + . This proves the inequality (4).
p q

Next we prove two important inequalities theorem using the result shown
above.

Theorem 3 (Holder’s Inequality for sequences): Let an , bn ∈ K for n = 1, 2.....

For 1 < p < ∞,

1 1
Let q be such that + = 1.
p q

1 1
∞
∞ p
p
∞ q
q
Then  a n bn ≤  a n   bn 
n =1  n=1   n=1 

(Here p and q are called conjugate indices. In case p = 1 we take q = ∞ and

vice versa).

Proof: Let a1 ,..., a n , b1 ,..., bn be scalars.

1 1
 n p 
p
 n q 
q
α =   ai  , β =   bi 
 i =1   i =1 

If α = 0 or β = 0, then each term in either sum is zero and the Holders

inequality is true.

So suppose both α ≠ 0 and β ≠ 0.

21
Block 1 Normed Linear Spaces
ai bi
Taking a = and b = in the inequality (4), we get
α β

p q
ai bi ai bi
. ≤ p
+
α β pα q βq

This is true for i = 1, ...., n.

Taking the sum

n
 1 n
1 n

  ai b
p q
a i bi ≤ αβ p
+ i 
i =1  pα i =1 qα q i =1 

 1 1
= αβ +  = αβ
 p q
1 1
 n p
p
 n q 
q
=   ai    bi 
 i =1   i =1 

This inequality is true for every n.

1 1
 ∞ p 
∞ p
 ∞ q 
q
Letting n → ∞, we get  a n bn ≤   a n    bn 
n =1  n =1   n =1 

Theorem 4 (Minkowski’s inequality for sequences):

Let an , bn ∈ K for n = 1, 2..... For 1 < p < ∞,

We have
1 1 1
∞ p
p
∞ p
p
∞ p
p

 an + bn  ≤  a n  +  bn  .
 n=1   n=1   n=1 

Proof: For p = 1, the result is true since ai + bi ≤ ai + bi for i = 1,...n ,....

Taking the sum we get the Minkowski Inequality.

Now suppose 1 < p < ∞

1 1
Let q be the conjugate index of p, i.e. + = 1. Since 1 < p < ∞, q also
p q
satisfies 1 < q < ∞.

First we consider the finite case. Let a1 ,..., a n and b1 ,..., bn be scalars.

(a ) p
Consider i + bi
i =1

22
Unit 1 Inner Products and Norms
n
=  ( ai + bi ) (a ) p −1
i + bi
i =1

n n
=  ai ( ai + bi ) +  bi ( a i + bi )
p −1 p −1

i =1 i =1

Using Holders inequality we get

1 1
 n p
p
n  q
≤   ai   ( ai + bi ) q ( p −1)

 i =1   i =1 
1 1
 n p
p
n  q
+   bi   ( ai + bi ) q ( p −1)

 i =1   i =1 
1 1
 n p
p
n  q
=   ai   ( ai + bi ) p
 as q ( p − 1) = pq − q = p.
 i =1   i =1 
1 1
 n p
p
n  q
+   bi   ( ai + bi ) p

 i =1   i =1 
1 1 1
1−
 n
 q  n p
p
 n p
p
  ( ai + bi ) ≤   ai  +   bi 
p

 i =1   i =1   i =1 

1 1
But 1 − = . Hence
q p

1 1 1
n p
p
 n p
p
 n p
p

 ( ai + bi )  ≤   ai  +   bi  ' ∀ ' stands for ‘for every’

 i=1   i =1   i =1 

Since ai + bi ≤ ai + bi ∀i and p > 1, we get

1 1 1
 n p
p
 n p
p
 n p
p
  ai + bi  ≤   ai  +   bi 
 i =1   i =1   i =1 

Letting n → ∞, we get

1 1 1
∞ p
p
∞ p
p
∞ p
p
  an + bn  ≤   an  +   bn 
 i =1   n=1   n=1 

Example 4: Let X = l 2 , be the set of all scalar sequences x = ( x1 , x2 ,...)

∞

x
2
satisfying n < ∞. Define a norm on X which makes it a normed linear
n =1
space.

(Sequences of scalars satisfying the above condition are called square

summable sequences.) 23
Block 1 Normed Linear Spaces
2
Solution: For x = ( x1 , x2 ,...xn ...), y = ( y1 , y2 ,.. yn ...) in l and α ∈ K ,define

x + y = ( x1 + x2 , y1 + y2 ,....xn + yn ,...) , α x = (αx1 ,..., αxn ,...)

Here, the first thing we have to do is to check that the vector addition and
scalar multiplication as defined above are meaningful i.e. we have to check
that x + y & α x are in fact elements of l 2 whenever x, y are in l 2 . To prove
this, all we have to do is prove that the sequences x + y & α x satisfy the
convergence condition of elements in l 2 .

∞ ∞
x, y ∈ l 2   xn < ∞ & y
2 2
n < ∞.
n=1 n=1

Now what we have to check is

∞

x
2
n + yn < ∞.
n =1

To prove this we make use of Minkowski’s inequality. Then we have

1 1 1
∞ 2
2
∞ 2
2
∞ 2
2

 x n + y n  ≤  xn  +  y n 
 n=1   n=1   n=1 
∞ ∞

 
2 2
Since both xn and y n are finite, we have
n =1 n =1

1
∞ 2
2

  xn + y n  <∞
 n=1 
∞
  xn + y n < ∞
2

n =1

Thus whenever x, y ∈ l 2 , x + y also lies in l 2 , i.e., vector addition as defined in

∞ ∞

 α xn = α x
2 2 2
l 2 is well defined. Moreover n < ∞.
n =1 n =1

This mean α x ∈ l 2 for all α ∈ K & x ∈ l 2 .

i.e., scalar multiplication is also well defined.

The properties of a vector space can now be checked and so l 2 is a vector

space over K [Please see E3].

Now let us make l 2 an inner product space.

∞
For x = ( x1 , x2 ....xn ...), y = ( y1 , y2 ... yn ) define x, y =  xn y n .
n=1

Here again we have to first prove x, y is well defined i.e. the infinite sum

24
Unit 1 Inner Products and Norms
∞

x
n=1
n y n is a well determined scalar for any x, y ∈ l 2 .

But this follows immediately from Holder’s Inequality (Theorem 2), for by
Holder’s inequality, the right hand side is absolutely convergent and hence
convergent, meaning, x, y is well defined.

Claim: , is an inner product.

We leave it as an exercise for you to verify (See E 4)

Thus l , , ( ) is an inner product space. The corresponding induced norm is

2

usually denoted by . called the ‘two norm’. (l , ) is a normed linear

2
2 2
2
space. In the standard terminology we say l is the space of all square
summable scalar sequences with the two norm . 2
given by

1
∞ 2
2
x =  xn  whenever
 n=1 

x = ( x1 , x2 ,... xn ...) ∈ l 2 .

Hence l 2 is a normed linear space.

***

As in Example 2, the collection {e1 , e2 ....} as defined there lie in l 2 and

forms a countably infinite linearly independent set in l 2 . Thus l 2 is also an
infinite dimensional vector space over K .

[The infinite dimensionality of l 2 can also be obtained by noting c00 is a

subspace of l 2 and c00 is infinite dimensional]

Remark 6: l 2 is usually referred to as the infinite dimensional version of the

Euclidean space K n

Example 5: Let X = C ([ a , b ]), the vector space of all continuous scalar

valued functions in [ a, b ]. Let us define a norm on X which makes it a normed
linear space.

Solution: For f ∈ C ([ a , b]) define f as

f = sup{ f (t ) : t ∈ [a, b ]} (5)

We shall show that the function in (5) defines a norm on C [a, b].

Since f is continuous on a compact set [a, b], f is bounded on [a, b ] and

hence sup f (t ) is a well determined number, i.e. f is
t∈[ a ,b ]

well defined ∀ f ∈ C ([ a , b ]).

25
Block 1 Normed Linear Spaces
Let us check that f satisfies the properties of a norm.

(i) If f = 0, then f (t ) = 0 ∀ t  sup f (t ) = 0.

t∈[ a , b ]

 f = 0.

Now suppose f = 0  sup f (t ) = 0.

t∈[ a , b ]

 f (t ) = 0 ∀ t ∈ [ a , b ]

 f = 0.

Thus f = 0 ⇔ f = 0.

(ii) Let α ∈ K and f ∈ C ([ a , b ]).

Consider α f = sup α f (t )
t∈[ a , b ]

= sup α f (t ) = α sup f (t )
t∈[ a , b ] t ∈[ a , b ]

=α f .

(ii) For f , g ∈ C ([ a , b]),

f + g = sup ( f + g ) (t ) = sup f (t ) + g (t )
t∈[ a ,b ] t∈[ a ,b ]

≤ sup f (t ) + sup g (t )
t∈[ a ,b ] t∈[ a ,b ]

= f + g .

Thus . is a norm on C ([ a , b ]). Hence [a, b] is a normed linear space

under the norm.
***

Remark 7: This norm is called the sup norm on C([ a, b ]). This norm is also
called the uniform norm on C ([ a , b]), since, convergence under this norm
gives uniform convergence i.e.,

Suppose f n → f under the above norm. Then given ∈> 0. ∃ N such that
fn − f < ε ∀ n ≥ N.

 sup ( f n − f )(t ) < ε ∀ n ≥ N

t∈[ a , b ]

= ( f n (t ) − f (t ) ) < ε, ∀ n ≥ N , ∀t ∈ [ a, b]

This is nothing but the uniform convergence of f n to f .

26
Unit 1 Inner Products and Norms
The following example shows another norm on c[a, b].

Example 6: Let X = c[a, b], the linear space of all continuous scalar valued
functions on [a, b] as a linear space already mentioned in the example, above.

Solution: For f , g ∈ C ([ a , b ]) define f , g as:

b
f , g =  f (t ) g (t ) dt
a

Then f , g is an inner product on C ([ a , b ]) . With the corresponding norm on

the linear space.

( )
Thus C ([a, b]), , is an inner product space.

Under the induced norm given by

1

f =   f (t ) dt  ,
b 2 2

 a 

(C ([a, b]), . ) is a normed linear space.

***

Here are some exercises for you to try.

E3) l 2 is a linear space.

∞
E4) Prove that x, y = x
n =1
n y n is an inner product on l 2 .

E5) Is the dimension of the space X = C [a, b ] finite or infinite? Justify.

We have seen the sequence spaces c00 and l 2 . Now we look at some more of
them.

Example 6 (The Space B): B is the set of bounded scalar sequences. Let us
show that B is a normed linear space.

Solution: Since the sum of two bounded sequences and scalar multiples of
bounded sequences are again bounded, B has a vector space structure.

For x = ( x1 , x2 ..., xn ,...) in B, define x ∞

= sup xn (called the sup norm). Since
n

x is a bonded sequence sup xn is a well determined scalar, i.e. x ∞ , is well

n

defined for every x ∈ B.

(The suffix ' ∞' is used to denote the supremum norm).

27
Block 1 Normed Linear Spaces

Claim: . ∞
is a norm. The verification is left as an exercise, see (E 7)

However if we take x = (1, 0, 0,.... 0,..) and y = (1, 0, 0,..0,...) then

x ∞
= 1, y ∞
= 1, x + y ∞ = 2 and x − y ∞ = 0. For this x and y , we get

2 2
x+ y + x− y =4

=2 x( 2
+ y
2
)
Which means . ∞
does not satisfy the parallelogram law and hence this norm
is not induced by an inner product.

Thus B, . ( ∞
) is a normed linear space but is not an inner product space.
***

Example 6 (The Space c): Let us show that the set 'c' , the set of all
convergent scalar sequences, is a normed linear space.

Solution: Since the sum of two convergent sequences and scalar multiples of
convergent sequences are convergent, c has a vector space structure.

Further any convergent scalar sequence is bounded, therefore c is a linear

subspace of B . We can define the sup norm on c and c becomes a normed
linear space and this is also not an inner product space.
***

Example 7 (The space c0 ): Let us check wether the set c0 , is the set of all
scalar sequences converging to 0, is a normed linear space.

Solution: Since the limit of the sum is the same as the sum of the limits (i.e.
lim( xn + yn ) = lim xn + lim yn and for any scalar α, lim(αxn ) = α lim xn , c0 has
n→∞ n n n n
a vector space structure.

This space is a linear subspace of c and hence of B and consequently c0

becomes a normed linear space under the sup norm.

This is also not an inner product space.

***

Example 8 (The l p spaces, 1 ≤ p ≤ ∞) : Let X = l p , 1 ≤ p < ∞, the set of scalar

sequences x = ( x1 , x2 ,....) satisfies


p
xn ≤ ∞.
n =1

(We have already seen the space l 2 . ) Then l p is a normed linear space The
space B , that we have discussed in example 5 is usually denoted by l ∞ .

28
Unit 1 Inner Products and Norms

Solution: Remember, l , . ( 2
2
)is an inner product space and the norm is
given by an inner product.

In the case of l 2 , we have seen that the sequences in l 2 satisfy the

∞

x
2
convergence condition n < ∞. whereas in l p , the sequences satisfies
n =1
∞

=1 x
p
the condition n < ∞ where x = ( x1 , x2 ,...).
n

Elements of l p -space are called p-summable sequences.

In particular l1 is the space of all summable sequences, l 2 − the space of all

square summable sequences.

You may note that since l p -spaces are sets of scalar sequences the sum and
scalar multiplication are defined in the usual way-namely termwise addition
and termwise scalar multiplication.

But the important property that we have to check is that these operations are
well defined in the sense, just as we have checked that the sum of two square
summable is square summable, we have to prove that for all p, 1 ≤ p < ∞,
( p = 2 has already been discussed), sum of two p-summable sequences is
again p-summable.

But this follows directly from the Minkowski’s inequality (why don’t you check it
for yourself?).

Thus all the l p − spaces are linear spaces over K . Remember, if K = R ,

these spaces are real linear spaces and if K = C , they are complex linear
spaces.

The norm on the l p − space is defined as follows:

For x = ( x1 ,..., xn ,...) , the norm x p

called the p-norm is defined as follows:

1
 ∞ p
p
x p
=   xn 
 n=1 
∞

x
p
By definition it follows that the p-norm is well defined since n is a well
n =1
determined non-negative real number. The triangle inequality is just the
Minkowiski inequality.
***

Remark 7: Another important fact is that all these sequences spaces are
infinite dimensional spaces since the collection (e1 , e2 ,....en ,...) with
en = (0,0,...,1,0,0...), where 1 appears at the nth place is a countably infinite
linearly independent set.

Thus we have infinitely many examples of normed linear spaces.

29
Block 1 Normed Linear Spaces
Example 9: Let X be the linear space of all polynomials in one variable t with
scalar coefficients. For p ∈ X with p (t ) = a0 + a1t + ... + a n t n , define
p = sup{ p (t ) : 0 ≤ t ≤ 1}, p 1 = a0 + a1 + ... + an .

Show that ⋅ and ⋅ 1 are norms on X, and p ≤ p 1 for every p∈ X. Also

show that there is no constant α such that p 1 ≤ α p for all p ∈ X .

Solution: It can be seen that ⋅ is the sup norm on C [0,1] and that X is a
subspace of C[0,1] . An element p of X is uniquely determined by the finite
sequence {a j } of its coefficients, and p 1 is just the l1 norm of this sequence.
Thus, we see that ⋅ and ⋅ 1 are norms on X .

If p (t ) = a0 + a1t + ... + a n t n , then for 0 ≤ t ≤ 1,

p (t ) ≤ a0 + a1 t + ... + a n t n ≤ a0 + a1 + ... + an = p 1.

To prove the nonexistence of an α as stated in the problem, let

p(t ) = 1 − t + t 2 − ... − t 2 n−1

Now, for 0 ≤ t ≤ 1, we have

p(t ) = (1 − t ) + t 2 (1 − t ) + ... + t 2 n−2 (1 − t ) ≥ 0,

p(t ) = 1 − t (1 − t ) − ... − t 2 n−3 (1 − t ) − t 2 n−1 ≤ 1.

Since p (0) = 1, we get

p = sup{ p (t ) : 0 ≤ t ≤ 1} = 1.

Since p 1 = 1 + 1 + ... + 1 = 2n we have p 1 > α p if 2n > α . Thus, for any α

we can choose a polynomial p(t ) with p 1
> α p . This completes the proof.

***

Now we shall consider another class of function spaces known as Lebesgue

spaces denoted by Lp which you have leant in the Block 3 of the course MMT-
004 (Real Analysis). To understand these spaces you may recall the contents
of the units on Lebesgue measure and Lebesgue integral from Block 3 of
MMT-004.

Let m denote Lebesgue measure. If 1 ≤ p < ∞, then the set of all measurable


p
functions f with f dm < ∞ is denoted by Lp . These are known as
Lebesgue spaces for 1 ≤ p < ∞. We have seen that Lp is a normed linear
space with the norm ⋅ p
defined by ⋅ p
= ( g p
dm )
1/ p
.

30
Unit 1 Inner Products and Norms
For 1 < p < ∞, the triangle inequality is proved using Holder’s inequality,
namely

 fg dm = ( f p
dm ) ( f
1/ p q
dm )
1/ q

1 1
where + = 1. The case p = q = 2 of the Holder’s inequality is called
p q
Schwarz inequality.

The set of all essentially bounded measurable functions on R is denoted by

L∞ . A measurable function f for which the quantity essential supremum
defined by

ess sup( f ) = inf{ α > 0 : f ≤ α almost everywhere on R }

is finite (i.e. < ∞) is called an essentially bounded function. Then L∞ is a

normed linear space with the norm

f ∞
= ess sup ( f )

In Lp ,1 < p ≤ ∞, we identify two measurable functions f , g if f = g a.e

Remark 8: Consider x = (1,0,0,...) y = (0,1,0,...), then for 1 ≤ p < ∞, p ≠ 2, Recall that ' a . e ' is
1/ p 1/ p the abbreviation for
x p
= 1, y p
= 1, x + y p
=2 and x − y p
=2 , showing that the almost everywhere
parallelogram law is not satisfied. [Refer Block-3,
MMT-004]
Thus, for p ≠ 2 , the l p spaces are just normed linear spaces and not inner
product spaces, meaning, we have an infinite collection of normed linear
spaces which are not inner product spaces.

This is natural since inner product spaces have more structure.

As mentioned earlier, the norm induces a metric and hence we can talk about
completeness of normed linear spaces. We will discuss this in detail in the
next block.

Definition 3: A Banach space is a complete normed linear space, complete

w.r.t the norm defined on it.

Definition 4: A Hilbert space is a complete inner product space, complete with

respect is the metric induced by the inner product.

(Remember, an inner product induces a norm and this norm in turn induces a
metric).

It is interesting to note that spaces having structures similar to inner product

spaces and normed linear spaces owe their origin to the study of solutions of
Integral and differential equations. You may also note that Stefan Banach
formalized the study of normed linear spaces in 1930s. Maurice Frechet gave
the name Banach spaces to complete normed linear spaces and John Von
Neumann called complete inner product spaces, Hilbert spaces. Present day
functional analysis has vast applications in various fields. You will find some of
these applications as we proceed.
31
Block 1 Normed Linear Spaces
Why don’t you try these exercises now.

E6) Is there any relation between the sequences in l p and l q ?

E7) Verify that . ∞

= sup xn is a norm in B.
n
2
E8) Let x = (2, 3) ∈ R . Find x p
when p = 1, 2, 4, ∞.

E9) Let X = Lp [0,1] and x(t ) = t 2 . Find x p

for 1 ≤ p < ∞. Show that
lim x p
= x ∞.
p→ ∞

1/ p
 n 
=   x j
p
E10) Let 0 < p < 1 and x  , x ∈ K n . Show that . is not a
p 
 j 
norm on K n if n ≥ 2.

We have seen some examples of normed linear spaces and inner product
spaces. In the next section you will see many more examples.

1.4 SUBSPACES
Suppose we are asked the following questions:

(i) Given a normed linear space, can we in some way find simpler ways of
studying the properties of the space compared with studying the
properties on the whole space as such.

(ii) Given a normed linear space, can we get new spaces from the space?

It is these types of questions that lead us to the study of subspaces and

quotient spaces.

To begin with now let us look at some important properties of subspaces.

Proposition 7 (Jensen’s Inequality): For 1 ≤ p < r < ∞, . r

< . p.

Proof: Let x be any nonzero element of l p .

x x
Let y = . Then yi = i for all i = 1, 2,...
x p x p

1/ p
 ∞ p
Now, x p =   x j  ≥ xi for all i = 1, 2, ...
 j =1 
x
So, yi = i ≤ 1 for all i = 1, 2, ...
x p
r p
 yi ≤ yi [As,1 ≤ p < r < ∞]

32
Unit 1 Inner Products and Norms
∞ ∞
  yi ≤  yi
r p

i =1 i =1

( y r
) ≤( y )
r
p
p

 x 
 y( r
)
r
≤ (1) p As y p
=
p
= 1
 x p 
x
 y( r
)
r
≤1 y r
≤ 1
x
r
≤1
p

 x r
≤ x p
for 1 ≤ p < r < ∞.

Remark 9: From the above Proposition 7, we can observe that if x ∈ l p for

p ≥ 1, then x p < ∞ .
 x r
≤ x p
< ∞, where 1 ≤ p < r < ∞
 x ∈ lr

Therefore, l p ⊂ l r for all 1 ≤ p < r < ∞.

Remark 10: We can also observe that if xn → x in l p then xn − x p

→ 0 as
n → ∞. But from Proposition 7, we have xn − x r ≤ xn − x p .

So, xn − x r → 0 as n → ∞. Hence, xn → x in l r .

Let us now give some important definitions and prove some more important
theorems related to subspaces.

Definition 5: Let Y be a subspace of a normed linear space X . The norm on

X restricted to Y is called the induced norm on Y .

Now if we look at the sequence spaces c, c0 , c00 , l p ,1 ≤ p < ∞, we have the

following relation:

c00 ⊂ p
+l ⊂ c ⊂ c ⊂ l∞ ,
+ 0 + +
where all the inclusions are proper. Also, the above inclusion says, under the
sup norm each member can be considered as induced normed linear sub
space of the next member.

Now we shall discuss an important theorem in Functional Analysis known as

Riesz Lemma. It specifies the conditions that guarantee that a subspace in a
normed linear space is dense.

Before we start the theorem we recall the following concepts.

Recall that you are familiar with the function called the distance function which
is denoted by ‘dist ( A, B)' -distance between two subsets A and B of a metric
space.

If Y is a subspace of a metric space ( X , d ) then distance between an element

x in X is defined by dist ( x,Y ) = inf{ d ( x, y ) : y ∈ Y } .
33
Block 1 Normed Linear Spaces
In the case of normed spaces, the above definition become

dist ( x, Y ) = inf { x−y : y ∈Y}

In general, if A, B are subsets of a normed linear space X , then:

dist ( A, B ) = inf { a − b : a ∈ A, b ∈ B}

Roughly this means the shortest distance between points of A and B .

Now we are ready to state the theorem.

Theorem 4 (Riesz Lemma): Let X be a normed linear space, Y be a closed

subspace of X and Y ≠ X . Let r be a real number such that 0 < r < 1.

Then there exists some x r such that

xr = 1 and r ≤ dist ( xr , Y ) ≤ 1

Proof: First we observe the following:

If x ∈ X and x ∉ Y , then dist ( x , Y ) > 0. On the contrary we suppose

dist ( x, Y ) = 0.

By definition this means

inf { x − y : y ∈ Y } = 0.

Now if we use the definition of infimum, ∀x ∈ N , we can find an y n ∈ Y such

that

1
x − yn < . (6)
n

Now consider the sequence {y n }. This is a sequence in Y and Eqn. (6)

implies that y n → x as n → ∞.

But it is given that Y is closed and hence x must lie in Y . This is a

contradiction since by choice x ∉ Y . ∴ dist ( x, Y ) > 0 .

Now let x ∈ X , x ∉ Y . Since 0 < r < 1, we claim that ∃ y0 ∈ Y such that

x − y0 ≤ dist ( x, Y ) r .

This is because if

x − y 0 > dist ( x, Y ) / r ∀y ∈ Y ,

Then, inf {x − y : y ∈ Y } ≥ dist ( x, Y ) / r

 1 ≥ 1  r ≥ 1, contradiction!
r
34
Unit 1 Inner Products and Norms
Thus for any x ∈ X , x ∉ Y , we can get a corresponding y1 in Y such
that x − y1 ≤ dist ( x, Y ) r . (7)

x − y1
Set xr = . Then x r = 1.
x − y1

Also dist ( xr , Y ) = inf {x r − y : y ∈ Y } ≤ xr − 0 (Q 0 ∈ Y, being a subspace).

Thus dist ( x r , Y ) ≤ 1 (8)

 x − y1 
Also dist ( x r , Y ) = dist  , Y 
 x− y 
 1 

dist ( x − y1 , Y )
= ( by definition of infimum)
x − y1

1
= inf { (x − y1 ) − y : y ∈ y}
x − y1

1
= inf { x − ( y1 + y ) : y ∈ Y } (since y1 , y ∈ Y and Y is a subspace,
x − y1
y1 + y ∈ Y and also as y varies over all
1
= inf { x − y : y ∈ Y } of Y, y1 + Y , also varies in overall of Y )
x − y1

1
= . dist ( x,Y )
x − y1

≥r by (7) above.

Thus dist ( x r , Y ) ≥ r (9)

From (8) and (9) we get the required result

r ≤ dist ( x r , Y ) ≤ 1.
***

1
Remark 8: Let n ∈ N n ≠ 1, Then 0 < 1 − <1
n

By Riesz Lemma, there exists x n ∈ X with x n = 1, and

1
1− ≤ dist ( x r , Y ) ≤ 1.
n
If now ε > 0 is arbitrary, then we get an x ε ∈ X , such that
1 − ε ≤ dist ( xε , Y ) ≤ 1.
The idea we get is: Whenever we consider a proper closed subspace Y of a
normed space X , then we can always find a unit vector in X (equivalently a

35
Block 1 Normed Linear Spaces
vector on the unit sphere of X) such that the distance between Y and that
vector is close to 1.

Before coming to one of the most important results in normed space, we will
prove the following two results.

Theorem 5: Let X be a normed space and Y a subspace of X. Then for

x ∈ X , y ∈ Y and k ∈ K , we have kx + y ≥ k dist ( x, Y ).

Proof: If k = 0 the kx + y = y ≥ 0 which is always true.

Suppose k ≠ 0 .

y
Then kx + y = k x +
k

≥ k dist ( x, Y ) since y / k ∈ Y .

Theorem 6: Let Y be a subspace of a normed space X . If Y is finite

dimensional, then Y is complete. In particular Y is closed (in X ). (This result
says any finite dimensional normed space is always complete.)

Since a complete subset in a metric space is always closed, it is enough to

prove that finite dimensional subspaces are complete.

Proof: Since Y is finite dimensional, it has a finite basis say {y1 ,..., y n }.

Since we have to prove Y is complete, we have to show that every Cauchy

sequence {xn } in Y converges in Y .

We prove this by using the induction principle on the dimension of Y .

Suppose the dimension is 1.

Then each x n is of the form x n = k n y1 , k n ∈ K.

Then x n − x m = k n y1 − k m y1 = k n − k m y1

 k n − k m = xn − xm y (note that y is a basis vector and hence ≠ 0)

1

But {x n }is Cauchy  xn − xm → 0 as n, m → ∞.

Thus k n − k m → 0 as n, m → ∞ in K.

 {k n }is a Cauchy sequence of scalars and the scalar field K is complete.

Hence {k n } is convergent in K.

Suppose k n → k in K.

Set x = ky 1 . Then x ∈ Y .
36
Unit 1 Inner Products and Norms
Also xn − x = k n y1 − ky1

= k n − k y1 → 0 as n → ∞ .

 {xn } is convergent in Y

Since {x n } was an arbitrary Cauchy sequence in Y , every Cauchy sequence

in Y is convergent in Y . This means Y is complete. This means one
dimensional normed spaces are complete. Now we proceed by induction.

Assume the result to be true for all ( n − 1) dimensional spaces.

Now let the dimension of Y be n . We take {y1 , y 2 ,..., y n } to be a basis of Y .

Again to prove Y is complete, let {xm } be any Cauchy sequence in Y .

Set Z = span {y 2 ,..., y n }. Then Z is an (n − 1) dimensional subspace of X . By

induction assumption Z is complete and hence is closed also.

In terms of the basis vectors, suppose each xm has the representation.

xm = k m y1 + k m2 y 2 + ... + k mn y n , m = 1,2,....

Here k m2 y 2 + ... + k mn y n ∈ Z . So let us write this vector as z m .

Thus x m = k m y1 + z m , m = 1,2,... with k m ∈ K and z m ∈ Z.

Using theorem 5, we get

xm − x p = k m y1 + z m − (k p y1 + z p )

= (k m − k p )y1 + (z m − z p )

≥ km − k p dist ( y1 , Z )

Since y1 ∉ Z and Z is closed in X , dist ( y1 , Z ) > 0.

Hence k m − k p ≤ xm − x p dist ( y1 , Z )

→ 0 as m,p → ∞, as {x n }is Cauchy.

Thus {k m } is a Cauchy sequence in K , and hence is convergent in K ,

Suppose k n → k . (10)

Also z m − z p = (xm − k m y1 ) − (x p − k p y1 )

≤ x m − x p + k m y1 − k p y1
37
Block 1 Normed Linear Spaces

= xm − x p + k m − k p y1

→ 0 as m,p → ∞, as both {x m } & {k m }are Cauchy in X & K

respectively.

Thus {z m } is Cauchy in Z.

But by Induction assumption Z is complete and hence {z m }converges in Z.

Suppose z m → z in Z . (11)

Using (10) and (11) set x = ky1 + z.

Then x ∈ Y .
Also x m − x = k m y1 + z m − ky1 − z

≤ k m y1 − ky1 + z m − z

= km − k y1 + z m − z → as m → ∞.

 {x m } is convergent in Y . Arbitrariness of the Cauchy sequence {x m } in Y

implies that every Cauchy sequence in Y is convergent in Y . Thus Y is
complete.

Now let us prove the important theorem on the characterisation of finite

dimensional normed spaces.

Theorem 7: Let X be a normed space over K . The following are equivalent.

(i) Every closed and bounded subset of X is compact.

(ii) { }
The subset x ∈ X : x ≤ 1 of X is compact (equivalently the closed
unit ball in X is compact).

(iii) X is finite dimensional.

(The equivalence of (ii) and (iii) can be stated as: A normed space X
is finite dimensional if and only if the closed unit ball in X is compact).

Proof: (i)  (ii).

{ }
The subset x ∈ X : x ≤ 1 is both closed and bounded and hence by (i), it is
compact.

(ii)  (iii) Suppose X is not finite dimensional. Then we can find an infinite
linearly independent set in X . Suppose {z1 , z 2 ,..., z n ,...} is an infinite linearly
independent set in X .

Set Z n = span {z1 , z 2 ,..., z n }, n = 1,2,...

38
Unit 1 Inner Products and Norms
Then each Z n is a finite dimensional subspace of X and hence by Theorem 6
is closed in X .

Thus each Z n is a proper closed subspace of X .

Also Z n ≠ Z n +1 and Z n is a closed subspace of Z n +1 .

Applying Riesz Lemma to Z n & Z n +1 , there exists a xn ∈ Z n+1 with x n = 1.

1  1
dist ( xn , Z n ) ≥  taking r = 
2  2

Doing this for every n , we get a sequence {x n } in X with x n = 1, ∀n with

1
x n ∈ Z n+1 and dist ( xn , Z n ) ≥
2

1
 x n − xi ≥ , i = 1,...n − 1 (since x1 ,..., x n−1 ∈ Z n )
2

Thus for m ≠ n, if we assume m < n, then x m ∈ Z m +1 ... ⊂ Z m + 2 ... ⊂ Z n ⊂ Z n +1

1
 xm − xn ≥
2

1
Since x m − x n ≥ , ∀ m, n, m ≠ n, this sequence cannot have a convergent
2
{
subsequence. Since xn = 1 ∀ n, xn ∈ x ∈ X : x ≤ 1 ∀ n and thus in }
{x ∈ X : x ≤ 1}we have produced a sequence which has no convergent
{ }
subsequence. But this means the set x ∈ X : x ≤ 1 is not compact. Thus: X
{ }
is not finite dimensional  x ∈ X : x ≤ 1 is not compact.

Taking the negations we get (ii)  (iii). (Try it by yourself)

Now let us prove (iii)  (i).

Let X be a finite dimensional normed space and suppose E is a closed and

bounded set in X .

To prove that E is compact, we will prove that every sequence in E has a

convergent subsequence in E .

Let {y1 ,..., y n } be a basis for X .

Let {xm }be any sequence in E .

Suppose x m = k m ,1 y1 + ... + k m , n y n with k m ,1 ,..., k m ,n ∈ K m = 1,2,...

Since E is bounded, suppose p ≤ M ∀ p ∈ E ,

39
Block 1 Normed Linear Spaces
Then x m ∈ E  x m ≤ M ∀ m

k m,1 y1 + k m, 2 y 2 + ... + k m,n yn ≤ M ∀ m, n (12)

Now let Z j = span {y1 ,..., y i −1 , y i +1 ,..., y n }

Then Z j is closed and y j ∉ Z j , j = 1,2,.., n.

(
Thus km, j y j + k m,1 y1 + ... + k m, j −1 y j −1 + km, j +1 y j +1 + ... + k m, n yn )
≥ k m, j dist ( y j , Z j ), j = 1....

Since y j ∉ Z j and Z j is closed, dist y j , Z j > 0. ( )

k m,1 y1 + ... + k m,n yn
Thus k m, j ≤
dist ( y j , Z j )

M
≤ , m = 1,2,...
dist ( y j , Z j )

 The sequence k m , j { } ∞
m=1
is a bounded sequence in K for j = 1,2,..., n.

So, by Bolzano-Weierstrass theorem, each of these sequences {k m , j}m =1 ,

∞

j = 1, 2,..., n, has a convergent subsequence in K and by passing to a

subsequence of a subsequence several times, we get a subsequence {k mp , j }
that converges in K for each j = 1, 2,..., n. Suppose k m p , j → k j in K as p → ∞
for j = 1,2,..., n.

Set xm p = k m p ,1 y1 + ... + k m p , n y n .

Then by construction, x m p { } ∞
p =1
is a subsequence of {x m }, and hence lies in E.

Set x = k1 y1 + ... + k n y n .

Now x m p − x = k m p ,1 y1 + ... + k m p , n y n − k1 y1 + ... + k n y n

 k m p ,1 − k1 y1 + ... + k m p , n − k n yn

→ 0 as p → ∞, since k m p , j → k j , i = 1,..., n

 The subsequence x m p { } of {xm } is convergent and since {x m p

} lies in E &
E is closed, the limit x ∈ E.

 The subsequence x m p { } of {x m } is convergent in E  E is compact.

40
Unit 1 Inner Products and Norms
Thus we have proved the equivalence of three properties.

Next we shall look at some special properties of normed linear spaces.

Suppose X is a vector space and E ⊂ X . For any point

a ∈ X , a + X = X + a = {a + x; x ∈ E}. This is nothing but a translate of E,
every vector x in E shifted to the vector a + x or x + a (both are the same).

If we have a straight line in R 2 , say L then, for any a ∈ R 2 , a + L is again a

straight line shifted through the vector a.

Suppose now X is a normed space and a ∈ X . If U (a, r ) is any

neighbourhood of a, then b + U ( a , r ) now becomes a neighbourhood of a + b ,
namely U ( a + b, r ).

In particular if U (0, r ) is any neighbourhood of the zero vector in X , then to

get a neighbourhood of any point x in X , all we have to do is consider
x + U ( 0, r ). Using system of neighbourhoods of the zero vector in X , we can
get neighbourhoods of any point in X .

This is a very big advantage in the study of normed linear spaces.

Proposition 8: Let X be a normed linear space over K . If E 1 ⊂ X is open

and E 2 ⊂ X , then E 1 + E 2 is open.

Proof: Let x ∈ X and x1 ∈ E1 . Since E1 is open, there is some r > 0 such that
U ( x1 , r ) ⊂ E1 . But U ( x1 + x, r ) = U ( x1 , r ) + x ⊂ E1 + x.

Hence, E1 + x is open for every x ∈ X . Since E1 + E2 = ∪{E1 + x2 : x2 ∈ E2 }, it

follows that E1 + E 2 is open.

Proposition 9: Let Y be a subspace of X . Then Y o ≠ φ if and only if Y = X .

Proof: If Y = X , then Y o = X ≠ φ. Conversely, let Y o ≠ φ. Consider a ∈ Y o and

let r > 0 be such that U (a, r ) ⊂ Y . We first prove that
U (0, r ) = U ( a , r ) − a ⊂ Y . Then we shall check whether Y = X .

U (0, r ) = U (a, r ) − a ⊂ Y , where a ∈ Y o .

Let x ∈ U ( o , r )  x ≤ r .

Consider x + a

x + a − a = x ≤ r  x + a ∈ U (a, r )

 x = x + a − a ∈ U ( a , r ) − a.

Thus U (o, r ) ⊂ U (a, r ) − a. (13)

Now let x ∈ U ( a , r ) − a. 41
Block 1 Normed Linear Spaces
 x = y − a, y ∈ U ( a , r ).

y ∈ U (a , r )  y − a ≤ r  x ≤ r  x ∈ U (o, r )

 U (a, r ) − a ⊂ U (o, r ) (14)

(13) & (14)  U (o, r ) = U (a, r ) − a.

Also U (a, r ) ⊂ Y 0 ⊂ Y . (Q a ∈ Y o & Y is a

subspace.)
 U (a , r ) − a ⊂ Y − a = Y .

Now we shall check whether Y = X .

rx
Let x ∈ X . If x = 0, then clearly x ∈ Y . If x ≠ 0, then ∈U (0, r ). Hence x ∈ Y
x
as well. That is, y = x. Hence the result.

Here are some exercises for you.

E11) If E1 and E 2 are two closed sets, then is it necessary that E1 + E 2 is

closed? Justify your answer.

E12) If E1 is compact and E 2 is closed, then show that E1 + E 2 is closed.

Recall, in the beginning of this section, we mentioned that an important

question is: given a normed space can we get new spaces from it.
Construction of quotient spaces is one way of getting another normed linear
space from a given normed space. Next we shall consider this

1.5 QUOTIENT SPACES

In this section we shall discuss quotient spaces for normed linear spaces.

Given a vector space V If X is a non-trivial vector space and a is any non-zero vector in X , then
over a field K , the span 2
[a] – the span of a, is always a subspace of X . In R itself we have an
of a set S of vectors (not
necessarily infinite) is infinite collection of subspaces.
defined to be the
intersection W of all Now suppose X is a normed space and Y is a non-trivial subspace of X. As
subspaces of V that a group X is abelian and so is Y and hence the quotient
contains S . W is X / Y = {x + Y : x ∈ Y } is also an abelian group.
referred to as the
subspace spaned by S Now, for α ∈ K and x + Y ∈ X / Y , if we define α ( x + Y ) = αx + Y , then
or by the vectors in S . α ( x + Y ) ∈ X / Y . This is well defined and X / Y become a vector space over
K.

Now let us see how we can give a normed space structure on X / Y .

Every vector in X / Y is a coset which is in fact a collection of vectors.

42
Unit 1 Inner Products and Norms

We define the norm, denoted by . and call it as the “quotient norm”, as

follows:

For x + Y ∈ X / Y ,

x + Y = inf { x + y : y ∈ Y }. (15)

We shall show that this defines a norm when Y is closed.

Theorem 8: The function . defined on X / Y as in (15) above, where Y is

closed subspace of X satisfies all the properties of a norm.

Proof: Since x + y ≥ 0, ∀ y ∈ Y ,

 x +Y ≥ 0 ∀ x +Y ∈ X /Y

(i) Suppose x + Y = 0 (in X / Y .)

Then x + Y = 0 + Y = Y , (0 + Y is the zero vector in X / Y ).

{
Hence 0 = inf x + y : y ∈ Y . }
since 0 ∈ Y , Y is a
= 0+0 =0
subspace.)
Thus x + Y = 0  x + Y = 0

Conversely suppose

x +Y = 0.

x + Y = 0  inf { x + y : y ∈ Y } = 0

Now, what we have to prove is that

x + Y is the zero vector in X / Y .

But x + Y is zero in X / Y ⇔ x ∈ Y .

Thus to prove x + Y = 0 (in X / Y ), all we have to prove is x ∈ Y .

{
By definition of infimum, inf x + y : y ∈ Y = 0 .}
1
 ∀ n ∈ N , ∃ y n ∈ Y such that x + y n < .
n

Consider the sequence y n . This is a sequence in Y and

1
x + yn <  y n → − x as n → ∞ . Since Y is closed − x ∈ Y . Y is a subspace
n
43
Block 1 Normed Linear Spaces
 x ∈ Y . This is what we have to prove. Thus
x + Y = 0  x + Y = Y = 0(in X / Y )

Hence x + Y = 0 ⇔ x + Y = 0.

(ii) Let α ∈ K and x + Y ∈ X / Y .

Here what we observe is as y varies over the whole of Y , αy also varies over
the whole of Y if α ≠ 0. Hence
[As, Y is a subspace on K and
α(x + Y ) = αx + Y = inf { αx + y : y ∈ Y }
α ∈ K , we can write the
general element as αy where
= inf { αx + αy : y ∈ Y } α ≠ 0 and y ∈ Y ]

= inf {α x + y : y ∈ Y }

= α inf { x + y : y ∈ Y }

= α x+ y

Thus the second property of a norm is true.

(iii) Let us now prove the triangle inequality.

Let x + Y , y + Y ∈ X / Y

( x + Y ) + ( y + Y ) = ( x + y) + Y

= inf { ( x + y ) + z : z ∈ Y } (15)

Let ε > 0.

Again by definition of the infinimum, ∃ y 1 ∈ Y such that,

x + y1 ≤ x + Y + ε / 2 and

∃ y 2 ∈ Y such that

y + y 2 ≤ y + Y + ε / 2.

Hence ( x + Y ) + ( y + Y )

= ( x + y ) + Y = inf { ( x + y ) + z : ε z Y }

≤ ( x + y ) + y1 + y 2

≤ ( x + y1 ) + ( y + y 2 )
44
Unit 1 Inner Products and Norms
≤ x + y1 + y + y 2

≤ x + Y + ε + y + Y + ε / 2.
2

= x + Y + y + Y + ε.

since ԑ is arbitrary.

(x + Y ) + ( y + Y ) ≤ x + Y + y + Y .

Thus the triangle inequality is also satisfied. All these prove the quotient norm
is in fact a norm.

Thus under the quotient norm, X / Y is a normed space over K .

For an example of a quotient space, let us consider the simple question: for
m > n, can we consider the quotient space R m / R n ?

n m n
First of all, how do we consider R as a subspace of R , since R consists
m
of n-tuples and R consists of m-tuples.

n n
By adjoining (m−n) zeroes to the elements of R , we can consider R as a
m
subspace of R .

By definition of the quotient space

{
R m / R n = ( x1 ,..., xm ) + R n : ( x1 ,..., x m) ∈ R m }
Suppose ( x1 ,..., xm ) ∈ R m and ( y1 ,..., yn ) ∈ R n ,

( x1 + y1,..., xn + yn , xn +1,..., xm )

= ( x1 + y1 ,..., xn + yn ,0,0,...,0) + (0,...,0, xn +1,..., xn )

Thus, the coset ( x1 ,..., xm ) + R n can be written as

n
( x1 ,..., xn ,0,...,0) + (0,...0, xn +1 ,..., x m ) + R .
(m−n) zeros n zeros

= ( x1 ,..., xm ,0,..., 0) + R n + (0,... 0, xn +1 ,..., xm ) + R n .

= R n + ( 0,...0, xn +1 ,..., xm ) + R n

Here ( x1 ,..., xn ,0,..., 0) + R n = R n since

( x1 ,..., x n ,0,...,0) is identified with ( x1 ,..., xn ) ∈ R n .

n
But R is the zero element in R m / R n .
45
Block 1 Normed Linear Spaces
n
Ultimately the coset ( x1 ,..., xm ) + R can be identified with the coset
n m -n
(0,...0, x n +1 ,..., x m ) + R . Here ( xn+1 ,..., xm ) varies in R .

n
Hence, identifying the coset (0,...0, xn +1 ,..., xm ) + R with element
( xn +1 ,..., xm ) in R m − n , we can identify the quotient space R m R n with the
space R m − n . Hence for m > n, R m / R n is a quotient space.

Now let us look at another example of a quotient space.

Let B (T ) be collection of all bounded scalar valued functions on a compact

space T . Then under point-wise addition and scalar multiplication of
functions, B (T ) becomes a vector space over K . We have already seen that
B(T ) is a normed linear space with the norm defined as follows:

For f ∈ B(T ), define

f = sup{ f ( x ) : x ∈ T }

This norm is called the sup norm on B (T ) , it satisfies all the properties of a
norm.

Hence B (T ) is a normed space and moreover it is a complete normed space

i.e. Banach space.

Now consider C (T ) , the set of all scalar valued continuous function on T.

Since T is compact, every continuous function is bounded and so C (T ) is a
subspace of B (T ) . Further C (T ) is also complete and so C (T ) becomes a
complete subspace of B (T ) .

Since complete subspaces are closed we can talk about the quotient space
B(T ) C(T ) . Giving the quotient norm on B(T ) C(T ) , B(T ) C(T ) becomes a
normed linear space.

Remember: In general there are bounded functions which are not continuous.

In the case of sequence spaces we have c00 ⊂ c0 ⊂ c ⊂ l ∞ .

+ + +

Can we talk about the quotient spaces:

l ∞ c , l ∞ c0 , l ∞ c00 , c c0 , c c00 , c c00 .. ?

Try to give reasons for your conclusions.

Thus if we are given a normed space X and a closed subspace Y , then we

can talk about the quotient space X / Y . The quotient space X Y is also read
as the quotient space X modulo Y .

Now let us ask: under what conditions the quotient space will also become
complete.

46
Unit 1 Inner Products and Norms
We will answer this question a little later. However we now look at an
important property about convergence of sequences in the quotient space
X Y.

Theorem 9: A sequence {x n + Y } in the quotient space X Y converges to

x + Y in X Y if and only if there is a sequence {y n } in Y such that the
sequence {x n + y n } converges to x in X .

Proof: Suppose the sequence {x n + Y } conveys to x + Y in X Y .

Let ε > 0 .

x n + Y → x + y  for the ε, ∃ N ∈ N

such that (x n + Y ) − ( x + Y ) ≤ ε , ∀ n > N.

2

 (x n − x ) + Y ≤ ε , ∀ n ≥ N.
2

Again by definition of the quotient norm, ∀ n, we can find a y n ∈ Y such that

(xn − x ) + y n ≤ ( xn − x) + Y + ε 2 .

From the above two inequalities, we get

(xn − x ) + y n ≤ ( xn − x) + Y + ε 2 ∀ n

< ε 2 + ε 2 ∀n ≥ N .

Thus we have found a sequence {y n }in Y .

∃ (x n − x ) + y n < ε ∀ n ≥ N .

 the sequence {x n + y n } converges to x in X .

Conversely suppose a sequence { y n } in Y is such that the sequence

x n + y n → x in X .

Let ε > 0 .

∃ n ∈ N such that

(x n + y n ) − x < ε , ∀ n ≥ N.

 ( x n − x ) + y n < ε , ∀ n ≥ N.

By definition of the quotient norm,

(x n − x ) + Y ≤ (x n − x ) + y n ∀ n
47
Block 1 Normed Linear Spaces
< ε ∀n ≥ N .

 (x n + Y ) − ( x + Y ) = ( x n − x ) + Y < ε ∀ n > N .

 x n + Y → x + Y in X Y .

This proves the theorem.

As an immediate consequence of the theorem we get the following result.

Theorem 10: If X is a normed linear space and Y is a closed subspace of X ,

X
then the quotient space is complete if and only if Y is complete.
Y
You may try the following exercise.

E13) Show that if X is a normed space separable and Y is a closed

subspace of X , then the quotient space X / Y is separable.

With this we come to the end of this unit. Let us summarise the points covered
in this unit.

1.6 SUMMARY
In this unit we have covered the following points:

1. Defined normed linear spaces and inner product spaces.

2. Given examples of standard normed linear spaces,

R n , C n , c0 , C[ a , b], l p ,1 ≤ p < ∞, Lp ( R )

3. Defined subspace of a normed linear space.

4. Stated and proved Riesz’s Lemma.

5. Defined quotient space.

6. Stated and proved the result that the quotient space X / Y is complete if
and only if Y is complete.

1.7 SOLUTIONS/ANSWERS
E1) For X = ( x1 , x2 , x3 ) ∈ R3 , define the function . , given by

[
x = ( x12 + x22 ) 3 / 2 + x3 ]
3 1/ 3
, x∈ X.

To show that . defines a norm on X . We have to show that .

satisfies the three condition. We note that

i) x ≥ 0 and x = 0 if and only if x = 0.

48
Unit 1 Inner Products and Norms
ii) We show that x + y ≤ x + y

Where x = ( x1 , x 2 , x 3 ), y = ( y1 , y 2 , y 3 )

We apply Minkowski’s inequality, given by

1/ p 1/ p 1/ p
 n p  n p  n p
  a i + bi  ≤  ai  +   bi 
 i =1   i =1   i=1 

We have x + y = ( x1 + y1 , x 2 + y 2 , x 3 + y 3 ) . Also

[
x + y == (( x1 + y1 ) 2 + ( x2 + y 2 ) 2 )3 / 2 + ( x3 + y3 ) 2 ]1/ 2

1/ 2
 2 
2
((x 1 + y1 ) + ( x 2 + y 2 ) ) 2 1/ 2
=   (x i + yi ) 2 
 i =1 
1/ 2 1/ 2
 2   2 
≤   (x i ) 2  +   ( yi ) 2 
 i=1   i =1 

= ( x 12 + x 22 )1 / 2 + ( y12 + y 22 )1/ 2

Now we apply Minkowski’s inequality for p = 3, given by

1/ 3 1/ 3 1/ 3
 2 3  2 3  2 3
  ai + bi  ≤   ai  +   bi 
 i =1   i =1   i=1 

Apply with a1 = ( x12 + x22 )1 / 2 , b1 = ( y12 + y 22 )1 / 2 , a 2 = x3 , b2 = y3

Then we have

[
x + y ≤ (( x12 + x22 )1 / 2 + ( y12 + y 22 )1 / 2 ) 3 + x3 + y3 ]
3 1/ 3

3 3
≤ ((x 12 + x 22 ) 3 / 2 + x 3 )1 / 3 + (( y12 + y 22 ) 3 / 2 + y 3 )1 / 2

= x + y .

Hence R 3 is a normed linear space.

E2) For x , y ∈ X , d ( x , y ) := x − y

Let us prove d gives a metric in X .

(i) x = y  x − y = 0  d x , y = 0.

Suppose d x , y = 0 .

 x− y =0 x= y

49
Block 1 Normed Linear Spaces
(ii) d ( y, x) = y − x = − ( x − y )

= −1 x − y = x − y

= d ( x, y ), ∀x, y ∈ X .

(iii) For x , y , z ∈ X ,

d ( x, y ) = x − y = x − z + z − y

≤ x−z + z− y

= d ( x, z ) + d ( z , y ).

Therefore d is a metric.

E3) To show that l 2 is closed under addition and scalar multiplication.

x
2
To check for addition, we have to show that n + y n is convergent
n
2
for x = {xn } and y = { y n } in l . We have

x + y n ≤  x n +  y n + 2 Re xn y n
2 2 2
n
n n n

yn 1/ 2
 ∞ 2 
∞
2 (by
≤  xn +  y n + 2  xn    y n 
2 2
Cauchy-
n n  n=1   n=1  Schwarz
2
Hence x + y ∈ l inequality)

To check for scalar multiplication, we have

 αx = α  xn
2 2
n
n

Hence αx ∈ l 2 .

Therefore l 2 is a linear space.

E4) We first have to show that the map < , >: l 2 × l 2 → K, defined by
∞
< x, y > =  xn y n is well defined where x = { x n }n∞=1 and y = { y n }∞n =1 .
n =1
∞
To show that we have to check that the infinite series x
n =1
n yn

converges. For this, it is enough to show that the series is absolutely

∞
convergent, i.e. x n=1
k y k is convergent.

By applying Cauchy-Schwarz inequality in K n for each n, we get that

the partial sums of the series is given by

50
Unit 1 Inner Products and Norms
n n

x
k =1
k yk =  xk yk
k =1

≤ ( x ) ( y ) k
2 y2
k
2 y2
dfd

y2 y2
 ∞ 2 
∞
2
≤   xk    y k 
 k =1   k =1 

<∞ since x, y ∈ l 2 .

Hence we conclude that the series is convergent and therefore the

map is well-defined.

Next we shall verify the three properties for inner product space. We
note that
∞
x, x =  xn xn ≥ 0
n =1

∞
x, y =  x n y n = y , x
n =1

∞
αx + β y , z =  α ( xn + y n )z n where z = {z n }
n =1

∞
=  [α ( xn ) z n + β( y n ) z n ]
n =1

∞ ∞
= α  xn z n + β y n z n
n =1 n =1

= α x, z + β y , z

Hence the result.

E5) We know the polynomial functions f n ( x ) = x n is dense in C [0,1] which

is a linearly independent set in C[0,1] . Hence C[0,1] is infinite
dimensional.

E6) We will show that if x ∈ l r for some r ≥ 1, then x ∈ l p for all p > r.

x
r
Let x ∈ l r , then n converges where x = {xn }. That means xn < 1
n

for all n ≥ n0 , for some n0 .

Then, for p > r , we get

p r
xn ≤ xn for n ≥ n0

51
Block 1 Normed Linear Spaces

x
p
This implies that n < ∞. Hence x ∈ l p .

E7) The properties (i) and (ii) for norm is automatically satisfied. To check
the triangle inequality, note that

x + y = sup x n + y n , x = {xn }, y = { y n }
n

≤ sup x n + sup y n , (follows from the properties of real numbers)

n n

≤ x + y.

Thus the triangle inequality is satisfied. Hence the result.

E8) By definition

x p
(
= x1p + x2p )1/ p
where x = ( x1 , x 2 )

Here x1 = 2, x2 = 3

For p = 1, x 1 = 5

For p = 2, x 2
(
= 2 2 + 322 )
1/ 2
= 13

For p = 4, x 4
= (2 4 + 34 )1 / 2 = 97

For p = ∞, x ∞
= max {2, 3} = 3 .

1/ p 1/ p
1 2 p   1 2p 
E9) x p
=   (t ) dt  =   t dt 
0  0 

1/ p
 t 2 p +1  1 
= 
 2 p + 1 
 0

1/ p
 1 
=  
 2 p +1

1
=
(2 p + 1)1 / p

→ 1 = x ∞ as p → ∞ .

E10) The triangle inequality is not satisfied if n ≥ 2. For example, let

x = (1, 0, 0,...0), y = (0,1, 0,...,0).

Then x + y = (1,1, 0, ...., 0),

52
Unit 1 Inner Products and Norms
p 1/ p p p 1/ p 1/ p
x p
= (1 ) =1= y p, x + y p
= (1 + 1 ) =2

Hence,

xp+ y p
= 2 < 21/ p = x + y p ,

Since 1 / p > 1. This proves the result.

E11) We shall show that it is not necessary. For example Take X = K and
consider the set

 1 
E1 = n + : n ∈ N , E2 = {− n : n ∈ N}.
 n 

These sets have no limit points. Hence they are closed. Since

1  1 
p+ =  n + p +  + (− n) ∈ E1 + E2 , for all p, n ≥ 1
n+ p  n + p 

1
But lim p + = p i.e.
n→∞ n+ p

That means p is a limit point of E1 + E2 .

If p ∈ E1 + E2 . Then p is of the form

1
p = n+ + ( − m)
n

for some n, m ∈ N (by the choice of E1 and E2 )

i.e. np = n 2 + 1 − mn

i.e. np + mn − n 2 = 1

This is possible only if n = 1, m = 2 − p .

So p ∉ E1 + E2 if p ≥ 2 since in that case m ∉ N, which is a

contradiction. Thus we have shown that if p is a limit point of E1 + E2 ,
then p ∉ E1 + E2 . Hence E1 + E2 is not closed.

 1
p =  n +  + (−m),
 n

that is, n( m + p − n) = 1. This is possible only if n = 1, m = 2 − p. So

p ∉ E1 + E2 if p ≥ 2. Thus, E1 + E2 is not closed.

E12) We have to show that E1 + E 2 = E1 + E 2 .

Let z ∈ E1 + E 2 . Then there is a sequence {z n } such that

53
Block 1 Normed Linear Spaces
z n → z, z n = xn + yn , xn ∈ E1 , yn ∈ E2 .

Since E1 is compact the sequence {xn } has a subsequence {xni }

converging to some x in E1 . So, as i → ∞, we get

yni = z ni − xni → z − x.

Hence z − x ∈ E2 = E2 and, therefore, z = x + ( z − x) ∈ E1 + E2 . This

proves that E1 + E2 is closed.

E13) Since X is separable, X has a countable dense subset E. Put

F = { y + Y : y ∈ E} . Then E is a countable subset of Y . To show that
F is dense in X / Y . Consider an element x + y in X / Y . Then given
ε > 0 there exists a y ∈ E such that x − y < ε . Then
( x + y ) − ( y + Y ) = x − y + Y ≤ x − y < ε. This shows that F is dense
in X / Y . Hence X / Y is separable.

54
 UNIT 2

CONTINUOUS LINEAR MAPS

AND FUNCTIONALS

Structure Page No
2.1 Introduction
Objectives
2.2 Characterisation of Continuity of Linear Maps
2.3 Spaces of Bounded Linear Operators
2.4 Summary
2.5 Hints/Solutions

2.1 INTRODUCTION
In applications, it is very important to know whether a given linear operator is
continuous or not. For instance, suppose a physical problem is modelled in the
form of an operator equation
Ax = y,
where A : X → Y is linear operator between normed linear spaces X and Y .
The problem may be to find x ∈ X from the knowledge of y ∈ Y , or to find
y ∈ Y from the knowledge of x ∈ X . The given information is often known as
‘data’ and the information which is to be obtained is known as a ‘solution’. The
questions of existence and uniqueness of a solution are purely set-theoretic or
algebraic ones. But, in applications, exact data are not usually available. Also,
in numerical approximation procedures, one normally considers approximate
data. In such situations, ‘stability’ of the solution is of great importance. It
amounts to knowing whether the operator A or its inverse A−1 is continuous.

In this unit our main study is about continuous linear operators between
normed linear spaces and their various properties. We shall also discuss
briefly a class of discontinuous operators which are closely related to
continuous operators, namely, the class of closed operators.

Objectives
The objectives of this unit are:
• To characterise continuity of linear maps on normed spaces;
Block 1 Normed Linear Spaces
• To identify continuous linear maps on standard spaces;

• To establish the properties of continuous linear maps on finite

dimensional spaces

2.2 CHARACTERISATION OF CONTINUITY OF

LINEAR MAPS
In this we introduce you to the concept of “continuity” of linear maps defined
on normed linear spaces with respect to the metric topology.

Before proceeding to discuss about continuity, let us briefly recall some ideas
on linear maps.

Linear maps or linear transformations are maps between vector spaces or

linear spaces having some special properties. If X and Y are vector spaces
over the same scalar field F , a map T : X → Y is said to be linear if T
satisfies the following property:

For
α, β ∈ F , x1 , x2 ∈ X ,
T (α x1 + β x2 ) = αT ( x1 ) + βT ( x2 ).

α x1 + β x2 indicates the linear structure in X and αT ( x1 ) + βT ( x2 ) indicates

the linear structure in Y , as T ( x1 ), T ( x2 ) ∈ Y . The condition
T (α x1 + β x2 ) = α T ( x1 ) + β T ( x2 ) means the map T : X → Y preserves the
linear structure. Whenever a linear map T : X → Y is a bijection, it is called
an isomorphism. If we have an isomorphism between two vector spaces it
means, the two vector spaces are ‘the same’ except for the elements.

The understanding is that the linear maps can be used to classify vector
spaces. As a simple example, any two n-dimensional spaces are isomorphic.

If we want to be specific, we can call vector space isomorphism just as we

have group isomorphism, ring isomorphism, tropological isomorphism and so
on.

What we are now interested in is the study of linear maps between normed
spaces.

In the case of normed spaces, in addition to being vector spaces, they also
have metric structures. When we have additional structures, it is natural to
investigate how maps behave with respect to all these structures.

Important classes of maps that are usually discussed reflecting the metric or
topological structures are the continuous maps.

This naturally leads to the concept of continuous linear maps or continuous

linear transformations between normed spaces.

Since functional analysis is the infinite dimensional version of linear algebra,

first let us look at the properties of linear maps between finite dimensional
normed spaces and then infinite dimensional normed spaces.

56 Let us start with a theorem.

Unit 2 Continuous Linear Maps and Functionals
Theorem 1: Let X , Y be normed spaces with X finite dimensional. Then
every linear map from X to Y is continuous.

Proof: X is a finite dimensional normed space and let T : X → Y be a linear

map.
If X = {0}, there is nothing to prove. So let X ≠ {0} and let its dimension be
n. Let {e1 , e 2 ,..., e n } be a basis for X . Now to prove the continuity of T . We
use the following characterization of maps between metric spaces.

Let X , Y be metric spaces. We note that map T : X → Y is continuous if and

only if, for every sequence { x n } in x with x n → x in X , T ( xn ) → T ( x) in Y .
Now to prove the continuity of T : X → Y , let us start with a sequence { x m } in
X and suppose x m → x in X .
Since {e1 ,..., e n }, is a basis of X , we can write each xm as:
xm = k m,1e1 + k m, 2 e2 + ... + k m,n en where ki , j ' s lie in K and x as
x = k1e1 + ... + k n en

Then k m, j → k j as m → ∞ in K (as was proved in the last part of Theorem 7

in Unit 1).
Since T is a linear map, we have
T ( xm ) = k m,1T (e1 ) + k m, 2T (e2 ) + ... + k m,nT (en )

and T ( x ) = k1 F (e1 ) + ... + k nT (en ).

Consider T ( x m ) − T ( x ) .

= k m ,1T (e1 ) + k m , 2T (e2 ) + ... + k m , nT (en ) − ( k1T (e1 ) + ... + k nT (en ))

≤ k m ,1 − k1 T (e1 ) + k m , 2 − k 2 T ( e2 ) + ... + k m ,n − k n T (en ) (1)

As m → ∞, k m, j → k j , j = 1,..., n.

Letting m → ∞ in (1) we get,

T ( xm ) − T ( x) → 0 as m → ∞.

i.e. T ( x m ) → T ( x) in Y as m → ∞.
Thus, whenever x m → x in X , T ( x m ) → T ( x ) in Y .

This means T is continuous.

Remark 1: On finite dimensional normed spaces all linear maps are

continuous.

It is natural to ask whether this is true for infinite dimensional spaces. The
following proposition shows that this is not true, in general.

Proposition 1: Let X be an infinite dimensional normed linear space and Y be

any non-trivial normed linear space. Then there exists a discontinuous linear
map from X to Y .

57
Block 1 Normed Linear Spaces
Proof: Since X is infinite dimensional, we can find an infinite linearly
independent subset {e n } of X .

en
Let xn = for n = 1,2,...
n . en

Then the set L = { x1 , x 2 , x3 ,...} is an infinite linearly independent subset of X

1
and xn = , ∀n.
n
In a vector space any linearly independent set can always be extended to a
basis of the vector space.

L can be extended to a basis B of X .

Since Y ≠ {0}, consider a vector b ≠ 0 in Y .
Define T : X → Y as:

b if x∈L
T ( x) = 
0 if x ∈ B , x ∉ L.
Then T is defined on all the basis vectors of X and thus T defines a linear
map on X .
Here the sequence {x n } is such that

xn → 0 as n → ∞.

But T ( x n ) = b ≠ 0 ∀n.

Thus T is not continuous.

What is so special about linear maps which are continuous from one normed
space to another?

The following theorem gives the answer:

Theorem 2: Let X and Y be normed spaces and T : X → Y be a linear map.

Then the following conditions are equivalent.
i) T is bounded on U ( 0, r ) for some r > 0.
ii) T is continuous at 0.
iii) T is continuous on X .
iv) T is uniformly continuous on X .
v) T ( x ) ≤ α x for all x ∈ X and some α > 0.

vi) The zero space Z (T ) of is closed in X and the linear map

~ ~
T :X → Y defined by T ( x + Z ( F ) ) = T ( x), x ∈ X , is continuous.
Z (F )
This theorem shows that continuous linear maps can be studied using any one
of the six conditions!

Proof: First we prove that conditions (i) and (v) are equivalent.
Suppose (i) holds.
58
Unit 2 Continuous Linear Maps and Functionals
T is bounded on U ( 0 , r ) for same r > 0.
Thus ∃ K > 0 such that

T ( x) ≤ K ∀ x ∈U (0, r )

i.e. T ( x) ≤ K ∀ x ∈ X such that x ≤ r .

Let now x ∈ X be arbitrary.

If x = 0 , then T ( x ) = 0 .

rx
If x ≠ 0. Set y = . Then y = r .
x

Hence by assumption T ( y ) ≤ K .

 rx 
 T   ≤ K .
 x 
r
 T ( x) ≤ K .
x

K
 T ( x) ≤ x.
r
Taking α = K / r , we see α > 0 &

T ( x ) ≤ α x , for all x ∈ X

Conversely suppose T ( x ) ≤ α x for all x ∈ X and same α > 0.

Let x ∈ U ( 0,1). Then T ( x ) ≤ α.

 T ( x ) ≤ α ∀ x ∈U (0,1).

 T is bounded on U ( 0,1).
(v)  (iv).
Let x1 , x2 ,∈ X be arbitrary.
Let ε > 0.
By (v) ∃α > 0 = T ( x ) ≤ α x ∀ x ∈ X .

By this,
T ( x1 ) − T ( x2 ) = T ( x1 − x2 ) ≤ α x1 − x 2 .

Taking δ = ε , we see that

α
whenever x1 − x 2 < δ,

T ( x1 ) − T ( x2 ) ≤ α x1 − x2

< α. ε = ε.
α
Hence given ε < 0, ∃ δ > 0 such that
59
Block 1 Normed Linear Spaces
T ( x1 ) − T ( x2 ) ≤ ε whenever x1 − x2 < δ.
Thus T is uniformly continuous.
(iv)  (iii).
Since any uniformly continuous function is continuous, T is continuous on X .
(iii)  (ii).
Since continuity on X  continuity at 0, T is continuous at 0.
(ii)  (i).
Suppose T is continuous at 0.
Then corresponding to 1, ∃ an r0 > 0 such that T ( x ) − T ( 0) < 1 whenever
r0
x − 0 < r0 . Taking r = .
2
We get T ( x ) < 1 whenever x ≤ r .

i.e. x ∈ U (0, r )  T ( x) < 1.

 T is bounded on U ( 0 , r ).

Hence the statements (i) to (v) are equivalent.

Now we will prove (iii) and (vi) are equivalent.

Suppose F is continuous on X .

Then the zero space Z (T ) = { x ∈ X : T ( x ) = 0}

= T −1 ({0}).
But {0} is closed in Y and T is continuous  T −1 ({0}) is closed in X i.e.
Z (T ) is closed in X .

Thus the quotient space X / Z ( F ) is a normed space under the quotient norm.

Now consider the map

~
T : X / Z (T ) → Y given by
~
T ( x + Z (T ) ) = T ( x), x + Z (T ) ∈: X / Z (T ).
~
T is well defined and if x ∈ Z (T ), then
~ ~
T ( x + Z (T ) ) = T ( x + z + Z (T ))

= T ( x + z)

≤α x+z for same α > 0.

since T is continuous.
This is true for every z ∈ Z (T ).
~
Hence T ( x + Z (T ) ) ≤ α x + z ∀z ∈ Z (T )
60
Unit 2 Continuous Linear Maps and Functionals
~
∴ T ( x + Z ( f ) ≤ α.inf { x + z : z ∈ Z (T )}

= α. x + Z (T )
~
Hence, by the above equivalent conditions T is continuous.
~
Conversely suppose Z (T ) is closed and T defined as above is continuous.

Hence ∃ α > 0 such that

~
T ( x) = T x + Z (T ) ≤ α x + Z (T )

≤α x.
This is true ∀ x ∈ X .

(Here we use the fact x + Z (T ) ≤ x ∀ x ∈ X ).

Hence T is continuous.

Recall: If X and Y are metric spaces, then a map f : X → Y is called a

−1
homeomorphism if f : X → Y is injective, continuous and f : R( f ) → X is
also continuous. If there is an onto homeomorphism between X and Y , then
X and Y are said to be homeomorphic.

Proposition 2: A linear map T from a normed space X to a normed space Y

is a homeomorphism if and only if ∃ scalars α, β > 0 such that

β x ≤ F ( x) ≤ α x ∀ x ∈ X .

Proof: Suppose T is a linear homeomorphism. Then T is linear bijective and

T : X → Y and T −1 : R (T ) → x are continuous.
By Theorem 2, ∃ scalars α , β 1 > 0 such that

T ( x) ≤ α x ∀ x ∈ X and

T −1 ( y) ≤ β1 y ∀ y ∈ Y . (2)

Since T is a bijection, every y ∈ R(T ) is of the form y = T ( x), for a unique

x ∈ X.
−1
Hence T ( y) ≤ β1 y , ∀ y ∈ Y can be written as:

T −1 (T ( x)) ≤ β1 T ( x) ∀ x ∈ X .

i.e. x ≤ β1 T ( x) , ∀ x ∈ X . (3)

From (2) & (3) we get

1
x ≤ T ( x) ≤ α x , ∀ x ∈ X .
β1
1
Thus T is a homeomorphism  ∃ α, β = > 0 such that
β1
61
Block 1 Normed Linear Spaces
β x ≤ T ( x) ≤ α x , ∀ x ∈ X .

Conversely, suppose the condition holds, i.e. ∃ scalars α, β > 0 such that

β x ≤ T ( x) ≤ α x , x ∈ X .

i) T is 1−1. (to prove this, it is enough to prove Tx = 0  x = 0, since T

is linear). Suppose Tx = 0.
Since β x ≤ T ( x) , Tx = 0  x ≤ 0.

 x = 0. Thus T is 1−1.
By Theorem 2, F is continuous.
Now consider F −1 : R ( F ) → X .

If y ∈ R( F ), then y = F ( x), for a unique x ∈ X . Hence

F −1 ( y ) = F −1 (F ( x) )

1
= x ≤ F ( x) (by assumption)
β
1
= y.
β
1
i.e. F −1 ( y) ≤ y , ∀ y ∈ R( F ).
β
−1
F is continuous.
Thus F : X → Y is injective and continuous and F −1 : R ( F ) → X is also
continuous and hence F is a homeomorphism.

Using this result, we can prove an important result, given as follows:

Corollary 1: If X and Y are linearly homeomorphic normed linear spaces,

then X is complete if and only Y is complete.

Proof: Suppose X and Y are homeomorphic normed linear spaces.

Hence there exists an homeomorphism T from X onto Y .
Then by Proposition 2, ∃ scalars α, β > 0 such that

β x ≤ T ( x) ≤ α x , x ∈ X . (4)

Now suppose X is complete.

Let { y n } be a Cauchy sequence in Y .

Since T : X → Y is a bijection, ∀ n ∃ a unique x n ∈ X such that T ( xn ) = y n .

1
By (4), x n − x m ≤ T ( xn − xm )
β
1 1
= T ( xn ) − T ( x m ) = y n − y m
β β

62
Unit 2 Continuous Linear Maps and Functionals
→ 0, as n, m → ∞.
 { x n } is cauchy in X .

By assumption X is complete  { x n } converges in X , say

x n → x in X . (5)

Let y = T ( x).

Then, y n − y = T ( x n ) − T ( x )

= T ( xn − x)

≤ α x n − x → 0 as n → ∞ (by (4) and (5))

 the sequence { y n } converges to y in Y .

Hence Y is complete.
Now reversing the roles of X and Y and using the function T −1 and inequality
(4) again, we can prove Y is complete  X is complete. (Try yourself !)

In Theorem 1, we have shown that every linear map from a finite dimensional
normed space to any other normed space is continuous.

Now suppose X is finite dimensional and Y any other normed space.

Suppose there exists a bijective linear map from T from X to Y . Since T is a
bijective linear map, T is a linear isomorphism  Y is also finite dimensional.
Since linear maps on finite dimensional spaces are always continuous, both T
and T −1 are continuous.

Conclusion: A bijective linear map T : X → Y , where X is finite dimensional,

is a homeomorphism.

Consequently if X is an n-dimensional normed space, then X is linearly

n
isomorphic to K and hence, by what we have seen above, X is
n
homeomorphic to K .
n
But there is a little problem in the above observation, namely, on K , there are
infinitely many norms. So the question is: with respect to which norm we have
the continuity property!
n
In fact, the norm we give on K does not matter, because of the following
ideas.

We give a definition.
1
Definition 1: Suppose . , . are norms on a linear space X . Then these
norms are said to be equivalent if ∃ scalars α, β > 0 such that
1
β x ≤ x ≤ α x , ∀ x ∈ X.
Note that equivalent norms induce the same topology, the identity map is a
homeomorphism.

Theorem 3: All norms on a finite dimensional linear space are equivalent.

63
Block 1 Normed Linear Spaces
n n
Thus, in particular, all norms on K are equivalent. Hence all norms are K
n
give the same topology on K .

Proof: Suppose X is a finite dimensional linear space.

′
Suppose . and . are any two norms in X .

Consider the identity linear map

I : ( X , . ) →  X , . .
′
 
This is a bijective linear map and X is a finite dimensional  I and I −1 are
continuous, i.e., I : (X , . ) →  X , .  is a homeomorphism.
′
 
Hence by Proposition 2, ∃ scalars α, β > 0 such that
1
β x ≤ I ( x) ≤ α x , ∀ x ∈ X . (6)

I is the identity map  I ( x) = x.

′ ′
i.e. I ( x) = x .
Hence (6) is nothing but
′
β x ≤ x ≤ α x , ∀x∈ X.
1
This means the norms . and . are equivalent.

Remark: Recall if X is a linear space over a field F , then a nonzero linear map
f : X → F is called a linear functional, i.e. a linear map from a linear space
to its base field is called a linear functional.

Notation: We denote a linear functional either by ' f ' or 'T ' .

Note that if X is a normed space over K and T : X → K is a linear functional,

then the range space of F is K since R (T ) ⊆ K and K is 1-dimensional over
K (recall the range of a linear space is always a subspace of the linear map).

Consequently, a nonzero linear functional is always onto.

Theorem 4: Let X and Y be normed spaces and T : X → Y be a linear map

such that the range R(T ) of T is finite dimensional. Then T is continuous if
and only if the null space Z (T ) is closed in X .

In particular, a linear functional T on X is continuous if and only if Z (T ) is

closed in X .

Proof: If T is continuous then by Theorem 2, Z (T ) is closed. So, we only need

to prove the converse part. If R (T ) = {0}, there is nothing to prove. So let
R (T ) ≠ {0}.

By assumption R(T ) is finite dimensional and hence has a finite basis, say
64 { y1 ,..., y n }. Since yi ∈ R(T ), ∃ x ∈ X such that yi = T ( xi ), i = 1,..., n.
Unit 2 Continuous Linear Maps and Functionals
Claim: Span {x1 + Z (T ),..., xn + Z (T )} = X .
Z (T )

Since span {x1 + Z (T ),..., xn + Z (T )} ⊆ X , it is enough to prove the reverse

Z (T )
inclusion.
So let x ∈ X . Then T ( x) ∈ R (T ) and { y1 ,..., yn } is a basis of R (T ) 

T ( x) = k1 y1 + ... + k n yn , ki ∈ K, i = 1,..., n.

= k1T ( x1 ) + ... + k nT ( xn ).

= T ( k1 x1 + ... + k n xn ).

 T ( x − k1 x1 − ... − k n xn ) = 0  x − (k1 x1 + ... + k n xn ) ∈ Z (T ).

 x + Z (T ) = ( k1 x1 + ... + k n xn ) + Z (T ).

= k1 ( x1 + Z (T ) ) + ... + k n (xn + Z (T ) )

 x + Z (T ) ∈ span{x1 + Z (T ),..., xn + Z (T )}.

i.e. X / Z (T ) = span {x1 + Z (T ),..., xn + Z (T )}.

This means the quotient space X / Z (T ) is finite dimensional.

Now, suppose Z (T ) is closed in X .

Then X / Z (T ) becomes a finite dimensional normed space.

By Theorem 1, any linear map on X / Z (T ) is continuous.

~ ~ ~
So, if we define T : X / Z (T ) → Y as T (x + Z (T ) ) = T ( x), x ∈ X , then T is will
defined and continuous. Hence by Theorem 2, T is continuous.

So whenever R(T ) is finite dimensional, T is continuous if and only if Z (T ) is

closed in X .

Next we shall consider some example.

Example 1: Suppose M = [ k i , j ] is an m × n matrix over K .

n
For x ∈ K , define Mx ∈ K m as:

Mx = ( Mx1 , Mx2 ,..., Mxm ) where

n
Mx i =  k i , j x j , i = 1,..., m.
j =1

n m
Then M is a linear map from K → K .
Since both are finite dimensional, we have two facts: Any linear map is
continuous and all norms are equivalent.
n m
Thus M : K → K is continuous under any norm on K n and K m .
***

Example 2: Consider the space C ([0,1)] with the supremum norm, . ∞.

65
Block 1 Normed Linear Spaces
Define the map,
T : C ([ 0 ,1]) → K as:
For any f ∈ C ([ 0 ,1]),
1
T ( f ) =  f (t )dt. Then F is linear.
0

Then R (T ) = K which is 1-dimensional and hence by last result above, T is

continuous.
Thus the integral is a continuous linear functional on the infinite dimensional
space C ([ 0 ,1]).
***

p p p
Example 3: Let X = l ,1 ≤ p ≤ ∞ , define the mapping T : l → l as follows.
If ( x1 , x2 ,..., x n ,...) ∈ l p , define

T (( x1 , x 2 ,..., xn ,...) ) = (0, x1 , x2 ,..., x n ,...).

Then T is a linear map and also,
T (( x1 , x2 ,..., xn ,...)) p = (0, x1 , x2 ,..., xn ,...) p .

= ( x1 , x2 ,..., xn ,...) p .

i.e. we have
T ( x) p
= x p , ∀ x ∈l p .
p
By Theorem 2, T is a continuous linear map on l . This T is called the right
shift operator.
***

Remark 2: Note that by giving specific value for p , we get a specific example
5
i.e. if we take p = 5, we get a particular example on the space l .

So in fact, the above example gives infinite collection of particular examples.

The following example is similar to the above.

p
Example 4: X = l ,1 ≤ p ≤ ∞ . Define T as:
T ( x1 , x2 ,..., xn ,...) = ( x2 , x3 ,..., xn ,...). This is again a continuous linear map on
p
l . Here T satisfies the relation T ( x) p
≤ x p , ∀ x ∈lp.
p
This T is called the left shift operator in l .
***
1
Example 5: Let X = C ([0,1]) the space of all continuously differentiable
scalar valued functions on [0,1]. This is a linear space.

Let Y = C ([0,1]), the space of continuous scalar valued functions on [0,1].

Under the sup norm, both are normed spaces.
Define T : X → Y as
66
Unit 2 Continuous Linear Maps and Functionals
T ( x) = x′ (derivative of the function x )
By the properties of the derivatives T is linear.
n
Consider the sequence { x n } in X given by xn (t ) = t .

Then xn ∞
= 1 ∀n (. ∞
is the sup norm )
1 n −1
xn (t ) = nt , ∀n,

x1n = n, ∀n
∞

Thus the property (v) in Theorem 2 is violated. Hence T is not continuous.

i.e. the differentiation operator is not continuous.
***

Example 6: Let X = c00 , the space of all scalar sequences having only finitely
many non-zero terms.
Consider the space with the one-norm, i.e. if x = ( x1 , x2 ,...) ∈ c00 , then

x 1 = x(1) + x ( 2) + ...
(note, since there are only finitely many non-zero terms, the sum on the right
side is a finite sum.)
Define T0 : c00 → K as

T0 ( x) = x1 + x2 + ... (where x = ( x1 , x 2 ,...)

(the right hand side is a finite sum.)

Then T0 is linear.

Also T0 ( x) = x1 + x2 + ...
∞
≤  xn = x 1 .
n =1

 T0 is continuous.

Now consider the 2-norm, . 2

on X .

i.e. for x = ( x1 , x2 ,...) ∈ c00 ,

1
 ∞ 2
2
x 2
=   xn  .
 n=1 
(Always remember there are only finitely many non-zero terms).
 1 1 
Consider xn = 1, ,..., ,0,0,...
 2 n 
Then xn ∈ c00 ∀n.
In this case

67
Block 1 Normed Linear Spaces
n ∞ 2
1 1 π
= ≤ 2 = .
2
xn 2 2
j =1 j n =1 n 6
n
1
Now T0 ( xn ) =  .
j =1 j

n
1
T0 ( x n ) =  → ∞ as n → ∞.
j =1 j

Again, Theorem 2 (v) is violated  T0 is not continuous when we give the

. 2 on c00 .

Thus continuity depends on the norm that we choose.

Note here that the space c00 is infinite dimensional.

Compare: Linear maps on finite dimensional spaces!

Now you can try the following exercise.

E1) Consider the space c00 . For x = ( x1 , x 2 ... xn ,...) ∈ c00 define
∞
T ( x ) =  xn . Then show that T is a linear functional which is not
n =1

continuous w.r.t. the norm x = sup x ( n) .

n

E2) Let c be the space of convergent sequence and T in the function

defined on c by T ( x ) = 
x n , x = {x n } . Is T linear? Is T continuous?
Justify.

E3) Let X and Y be normed linear space and F : X → Y be linear. Prove

that F is continuous if and only if for every Cauchy sequence { x n } in
X , the sequence {F ( x n )} is Cauchy in Y . Show that this is not true for
non-linear continuous maps.

In the next section we consider the set of all bounded linear maps from one
normed linear space to another.

2.3 SPACES OF BOUNDED LINEAR MAPS

We have seen that if X and Y are normed spaces, then, a linear map
T : X → Y is continuous if and only if F maps bounded sets in X onto
Notation bounded sets in Y (by Theorem 2). Consequently the terms continuous and
BL( X ,Y ) bounded are used as synonyms. (In the previous sections we have been using
the equivalence of the two concepts). The set of all bounded linear maps from
BL(X )
X to Y will be denoted by BL( X , Y ). A linear map from X to itself will be
X′
called a linear operator on X . We shall denote BL( X , X ) simply as BL( X ).
So BL( X ) is the set of all bounded linear operators on X . We shall denote

68
Unit 2 Continuous Linear Maps and Functionals
the set BL( X , K) the set of all bounded linear functionals on X as X ′,
called the dual space of X .
By Theorem 2, the first condition is equivalent to the fifth condition.

Thus boundedness is equivalent to saying that ∃ α > 0 such that

T ( x) ≤ α x , ∀ x ∈ X .
Many a time, we check this inequality to check the boundedness (= continuity)
of a linear map between normed spaces.

Note the difference between the boundedness of (ordinary) maps and those of
linear maps.

For example, if f is a mapping from K to K , we say f is bounded if ∃ M > 0

such that f (t ) ≤ M ∀ t.

Suppose we give such a definition for a linear map T : X → Y , where X , Y are

normed spaces, i.e. ∃ M > 0 such that

T ( x) ≤ M ∀ x ∈ X .

Then T ≡ 0. For suppose T is not identically 0.

 M +1
So there is an x0 ∈ X with T ( x0 ) = α > 0. Consider the vector   x 0 .
 α 
 M +1  M +1 M +1
T . x0  = . T ( x0 ) = .α = M + 1 ≤ M , an absurdity.
 α  α α

So we cannot demand T ( x) ≤ a constant for all x. Using linearity we have

proved such a thing is not true, unless T ≡ 0.

Definition 2: Let X , Y be normed linear spaces over K . A linear map

T : X → Y is said to be bounded below if ∃ a β > 0 such that β x ≤ T ( x ) for
all x ∈ X .

Remark 3: A bounded linear map T is a homeomorphism if and only if T is

bounded below.

You have seen that the set of all linear maps from a vector space X to a vector
spaces, i.e. L( X , Y ), is again a vector space under addition and scalar

multiplication defined pointwise. i.e.,

For T , S ∈ L( X , Y ) and α∈K
T + S is defined by (T + S ) ( x) = T ( x) + S ( x)
αT is defined by (αT ) ( x) = α . T ( x)

Since normed spaces are basically vector spaces, the set of all linear maps is
also a vector space.

69
Block 1 Normed Linear Spaces
But here what we are interested in is: if X , Y are normed spaces, is BL( X , Y )
also a normed linear space?
Suppose we define addition as above. We have to check fist, whether it is
meaningful i.e., if, T , S ∈ BL( X , Y ), is T + S defined as above also in
BL( X , Y ) .
Similarly for δ ∈ K, T ∈ BL( X , Y ), is δT in BL( X , Y ) .

First suppose T , G ∈ BL( X , Y ).

T , G are bounded  ∃α > 0, β > 0 such that

T ( x) ≤ α x ∀ x ∈ X .

S ( x) ≤ β x ∀ x ∈ X .

Consider (T + S ) x = T ( x) + S ( x)

≤ T ( x ) + S ( x)

≤ α x + β x = (α + β ) x ∀ x ∈ X .

 T + S ∈ BL( X ,Y ).
Similarly if δ ∈ K, T ∈ BL( X , Y ).

(δT ) ( x) = δ(T ( x ) ) = δ T ( x )

≤ δα x ∀x∈ X.

 δT is bounded  δT ∈ BL( X , Y ).
Also ‘+’ gives an abelian group structure on BL( X , Y ) and scalar multiplication
satisfies all the properties of a vector space. Thus BL( X , Y ) is a vector space
over K.
Recall if X is n-dimensional and Y is m-dimensional, then L( X , Y ) is mn-
dimensional. Hence if X is n-dimensional normed linear space and Y is m-
dimensional, then BL( X , Y ) also is an mn-dimensional space. Since
BL ( X , Y ) = L ( X , Y ) in this case.

Now we will see how we can make BL( X , Y ) into a normed linear space.

Again remember bounded linear maps take bounded sets onto bounded sets.

Theorem 5: Let X , Y be normed spaces. For T ∈ BL( X , Y ), define

T = sup{ T ( x) : x ∈ X , x ≤ 1}.

Then . is a norm on BL( X , Y ). This norm is called the operator norm on

BL( X , Y ).
Further T ( x ) ≤ T x ,∀ x ∈ X.

Proof: First we observe that {x ∈ X : x ≤ 1} is a bounded set in X .

70
Unit 2 Continuous Linear Maps and Functionals
Hence {T ( x) : x ∈ X : x ≤ 1}is a bounded set in Y which means

{ T ( x) : x ∈ X , x ≤ 1}
is a bounded set of non-negative reals and hence the
sup{ T ( x) : x ∈ X , x ≤ 1}is well defined.

Thus for every T ∈ BL( X , Y ), T is a well defined non-negative real number.

Now we shall check the properties (i), (ii) and (iii) of a norm.
i) We note that if X = {0}, there is nothing to prove.

So let us suppose X ≠ {0}.

By definition T ≥ 0, ∀ T ∈ BL ( X , Y ). To show that T = 0 if and only if

T = 0.
If T = 0, then T ( x) = 0 ∀x  T = 0.

Conversely suppose T = 0.

Then this  sup { T ( x) : x ∈ X , x ≤ 1} = 0.

 T ( x) = 0, ∀ x ∈ X with x ≤ 1.

T (0) = 0 is always time for linear maps.

x x x
Let x ≠ 0. Consider . Then = = 1.
x x x

 x 
Hence T  
 x  = 0.
 
1
 T ( x ) = 0.
x

 T ( x ) = 0.  T ( x ) = 0

Thus T ( x) = 0 ∀x ∈ X .
 T = 0.
(ii) Let α ∈ K and T ∈ BL( X , Y ).

Consider α T = sup{ α T ( x) : x ∈ X , x ≤ 1}

= sup{ α T ( x) : x ∈ X , x ≤ 1}

= α sup{ T ( x ) : x ∈ X , x ≤ 1}

= α T.

(iii) In T , S ∈ BL( X , Y ), we have

(T + S )( x ) , T ( x) + S ( x) ≤ T ( x) + S ( x )

Hence
71
Block 1 Normed Linear Spaces
T + S = sup { (T + S )( x) : x ∈ X , x ≤ 1}

= sup { T ( x) + S ( x ) : x ∈ X , x ≤ 1}

≤ sup { T ( x) + S ( x ) : x ∈ X , x ≤ 1}

≤ sup { T ( x) : x ∈ X , x ≤ 1}

+ sup { S ( x ) : x ∈ X , x ≤ 1}

= T + S.

Thus . defines a norm on BL( X , Y ). i.e. BL( X , Y ) is a normed linear space

over K.
Now let us prove T ( x) ≤ T , x , ∀ x ∈ X .

Let x ∈ X . If x = 0, then the inequality is true.

x
Suppose x ≠ 0. Then has norm 1.
x

 x 
Hence T  
 x  ≤ sup { T ( y) : y ∈ X , y ≤ 1}.
 
= T.

1
 T ( x) ≤ T .
x

 T ( x) ≤ T x.
Hence the claim.

Remark 4: For bounded linear maps, this inequality above is of great practical
use.

The following result gives equivalent definitions of the operator norm.

Theorem 6: Let X , Y be normed spaces and T ∈ BL( X , Y ). Then

T = sup { T ( x ) : x ∈ X , x = 1}

= sup { T ( x) : x ∈ X , x < 1}

= inf {α ≥ 0 : T ( x) ≤ α x , for all x ∈ X }

Proof: Let α 0 = inf {α ≥ 0 : T ( x ) ≤ α x , for all x ∈ X }

β = sup { T ( x) : x ∈ X , x = 1}

r = sup { T ( x) : x ∈ X , x < 1}

Since {x : x = 1} ⊂ {x : x ≤ 1},

β≤ T (7)
72
Unit 2 Continuous Linear Maps and Functionals
since {x : x < 1} ⊂ {x ∈ x ≤ 1},

r≤ T. (8)

Now consider a non-zero x ∈ X & t be such that 0 < t ≤ 1. We have

 tx  x x
T ( x) = T   ≤ sup{ T ( z ) : z ∈ X , z = t}
 x  t t

If t = 1, then T ( x ) ≤ sup{ T ( z ) : z ∈ X , z = 1} x

=β x.

Thus T ( x) ≤ β x , ∀ x ∈ X .

Hence by definition of α 0 , we have

α0 ≤ β (9)

Now suppose t < 1. Then as above, we get

x
T ( x ) ≤ sup { T ( z ) : z ∈ X , z < 1} .
t
x
= r. .
t
Now letting t → 1, we get

T ( x) ≤ r x , ∀ x ∈ X .

 α 0 ≤ r. (10)

Claim: T ≤ α0.

This claim along with (7) and (9) 

T ≤ α0 ≤ p ≤ T .

Hence T = α 0 = β.

Again this claim along with (8) and (10) 

T ≤ α0 ≤ r ≤ T .

Hence T = α 0 = r .

Thus we get
T = α 0 = β = r.

Proof of Claim: Let α ≥ 0 be such that T ( x ) ≤ α x ∀ x ∈ X .

Taking supremum on both sides of this inequality over all x ∈ X , with x ≤ 1,

we get:
sup{ T ( x ) : x ∈ X , x ≤ 1 }.

≤ sup{α x : x ∈ X , x ≤ 1} = α
73
Block 1 Normed Linear Spaces
 T ≤ α. (11)

Since α 0 is the infimum of all such α ' s we get

T ≤ α0.
This proves the claim.

Remark 5: This formula T = α 0 is very useful. Whenever we get inequality

like T ( x) ≤ K x ∀ x ∈ X , we can immediately conclude T ≤ K , i.e. we can
get a bound for T .

Many times it is not possible to get the exact value of a bounded linear map,
however we can get a bound.

Example 7: Consider the continuous linear map T given in example 2 above,

namely T : C([0,1]) → K, given by
1
T ( f ) =  f (t ) dt.
0

On C ([0,1]), , the norm is the sup norm, . ∞

1
Consider T( f ) = 
0
f (t ) dt

1
≤  sup f (t ) dt
0 t∈[ 0 ,1]

1
= f ∞
.  dt = f ∞
∀t ∈ C ([0,1]).
0

By the above theorem, we get

T ≤ 1. (12)

Consider the function f 0 defined as:

f 0 (t ) = 1 ∀t ∈ [0,1].

Then f 0 = 1.
1
Also T ( f 0 ) = 
0
f (t ) dt = 1.

 T ( f 0 ) = 1.

{
 T = sup T ( f ) : f ∈ C ([ 0,1]) : T ∞
}
=1

≥ T ( f 0 ) = 1.

i.e. T ≥ 1. (13)

(12) & (13)  T = 1.

***

Remark 6: To compute the norm of an operator. It is easy to get an upper

74 bound. To get the lower bound, the technique used is as shown in the
Unit 2 Continuous Linear Maps and Functionals
example, above. That is, we consider an element of norm 1 and see whether
the norm of the map at this element ≥ some constant. If the upper bound = the
constant, then the common number gives the norm.
Thus we have seen examples of bounded linear maps on normed linear
spaces. In fact the map operator in Example 7 is a bounded linear functional.

Now recall that a bounded linear map T from a normed linear space X to K is
called a bounded linear functional. We shall see some more examples of
bounded linear functionals and also discuss some properties.


1
Example 8: Consider the linear functional T : L [0,1] → R , f → tf (t )dt. Find
0

T .
1 1


Solution: We have Tf = tf (t ) dt ≤ tf (t ) dt ≤ f
0

0
1

So that Tf ≤ f 1 for all f ∈ L1 [0,1], and therefore, T ≤ 1.

Now we show that T = 1 . For this, we define a sequence { f n } of functions in

L1[ 0,1] :

0 for 0 ≤ t ≤ 1 − 1 / n
f n (t ) = 
n for 1 − 1 / n < t ≤ 1
1 1
1
Tf n =  tf n (t )dt = n  tdt = 1 − → 1 as n → ∞,
0 1−1 / n
2n
and
1 1
f n 1 =  f n (t ) dt = n  dt = 1.
0 1−1 / n

This shows that T = 1.

***

Next we shall see some properties of bounded linear functional.

Theorem 7: Let T be a linear functional on the normed linear space X . Then

if T is continuous at x0 ∈ X , it must be continuous at every point of X .

Proof: By the hypothesis, if xn → x0 , then T ( xn ) → T ( x0 ). To prove continuity

everywhere, we must show, for any y ∈ X . that if y n → y , then
T ( y n ) → T ( y ). To this end suppose y n → y ∈ X and consider

T ( y n ) = T ( y n − y + x0 + y − x0 )

= f ( y n − y + x0 ) + f ( y ) − f ( x0 ).

Since y n → y + x0 → x0 ,

T ( y n − y + x0 ) → f ( x0 )
and 75
Block 1 Normed Linear Spaces
T ( y n ) → T ( y ).
Before proceeding, a new definition is necessary.
Definition 2: The linear functional T on the normed linear space X is called
bounded if there exists a real constant k , such that, for all x ∈ X ,

T ( x) ≤ k x (14)

It is immediately evident that there may be many such real numbers k

satisfying (14) for a particular T . For example, if k1 works, certainly anything
bigger than k1 will also work.

Our next theorem shows that the condition of boundedness is equivalent to

continuity.

Theorem 8: The linear functional T defined on the normed linear space X is

bounded if and only if it is continuous.

Proof: First we shall show that continuity implies boundedness, and the proof
will be by contradiction; thus, suppose T is continuous but not bounded. The
negation of being bounded is that for any natural number n, however large,
there is some point, xn say, such that

T ( xn ) > n xn .
Consider now the vectors
xn
yn =
n xn
with norms
1
yn = .
n
Clearly, then the sequence
yn → 0
Since any linear functional maps the zero vector into the zero scalar and T is
continuous, this implies
T ( y n ) → f ( 0) = 0
But
1
T ( yn ) = T ( xn )
n xn
and
T ( xn ) > n xn ,
which implies
T ( yn ) > 1

and precludes the possibility of T ( y n ) approaching 0. Since we have arrived at

contradictory results, the assumption (the only one we made) that T was not
bounded must be false when T is continuous.

76
Unit 2 Continuous Linear Maps and Functionals
To prove the converse, that boundedness implies continuously, we need only
note that boundedness certainly implies continuity at the origin and apply the
preceeding theorem.
Now you can try the following exercise

E4) Let f : C 1 [0,1] → K be defined by f ( x) = x′(1), x ∈ X . Consider the

sequence { z n } in C 1 [0,1] where z n (t ) = t − t n / n, t ∈ [0,1]. Prove that
z n → x1 where x1 where x1 is the function such that x1 (t ) = t , t ∈ [0,1].
Prove further that z n ∈ Z ( f ) but x1 ∉ Z ( f ). What conclusion about
Z ( f ) can you draw?
E5) Consider the operators defined on C[a, b] (with supnorm) given by the
following
T0 f ( x) = f ( x0 ) for some x0 ∈ [a, b]

Show that T0 is a bounded linear functional. Also find T0 .

E6) Let X = B, be the normed linear space of all bounded real-valued

function on R with the norm defined by
f = sup{ f ( x) : x ∈ R}, f ∈ B.

For fixed x0 ∈ R, define T : X → X by Tf ( x) = f ( x − x0 ) . Show that T

is bounded linear operator. Find T .

With this we come to the end of this unit. Let us summarise the points covered
in this unit.

2.6 HINTS/SOLUTIONS
∞
E1) Since x ∈ c00 , x
n =1
n contains only a finite number of non-zero terms.

Hence T (x) exists for each x ∈ c00 . Also if x, y ∈ c00 , k ∈ K, then

(kx + y) n = kxn + yn for all n

That is T (kx + y ) =  kx n + yn , contains only finite number of terms as
x and y are in c00 . Hence T is linear.

To check that T is continuous, w note that the norm on c00 is the

supnorm . ∞
. Define xn ∈ c00 , by

1 for 1 ≤ j ≤ n
( xn ) j = 
0 for j > n
Then xn ∞
= 1 and T ( xn ) =  xn ( j ) = n .

77
Block 1 Normed Linear Spaces
Therefore supT ( x) ≥ n for all n. This shows that T cannot bounded on
u (0, r ) for any r > 0. Hence T is not continuous.

E2) T is well defined as lim xn exists.

n→∞

Let x, y ∈ c and k ∈ K, then

T (kx + y) = k lim xn + lim yn

= k T ( x) + T ( y ).
∴ T is linear.

Also for x = {x n } ∈ c, we have

xn ≤ x ∞

Also xn → T (x) by the definition of T . Thus

T ( x) = lim xn
n

≤ x∞.

This shows that T is continuous.

E3) To show the 'if ' part we suppose that T is continuous. Let {xn } be a
Cauchy sequence in X . To show that {T ( xn )} is Cauchy. Let ε > 0 be
given. Then

T ( x n ) − T ( xm ) = T ( x n − x m )
≤ α xn − x m

ε
Given > 0 there exist N such that
α

ε
xn − xm < for all n ≥ N .
α

Thus T ( x n ) − T ( x m ) < ε for all n ≥ N .

This shows that {T ( x n )} is Cauchy.

To show the “the” other part, let us suppose that T is not continuous.
Then we show that there exists a {z n } which is Cauchy but {T ( z n )} is
not Cauchy.

Since T is not continuous, for n = 1, 2,.... there is a {xn } in X such that

xn ≤ 1 and T ( xn ) ≥ n. This follows from equivalent properties (i) of
continuity. Thus xn ≠ 0. Let

78
Unit 2 Continuous Linear Maps and Functionals
xn 1
yn = , zn = yn
xn n

1
Then y n = 1 and z n = → 0. Hence z n → 0, but
n

1 1
T ( zn ) = T ( yn ) = T ( xn ) > n
n n xn

Thus {T ( z n )} is not bounded. Therefore {T ( z n )} is not Cauchy even

though {z n } is Cauchy as it converges to 0. Hence the result.

To show that the result is not true for non-linear map, consider the map
1
f : R → R given by f ( x) = , x ≠ 0 and f (0) = 0. Then check whether
x
1 
f is linear. For the second part consider the sequence   in R. Then
n 
1  1
  is a Cauchy sequence in R . But f   = n and { n } is not Cauchy
n  n
in R .

This shows that the result is not true for non-linear maps.
tn 1
E4) Hint: To prove that z n → x1 , note that z n − x1 = sup = → 0. as
n n
n → ∞.

To prove that z n ∈ Z ( f ) , observe that

n
f ( zn ) = z′n (t ) t =1 = 1 − t n −1 = 0.
n t =1

But f ( x1 ) = 1 ≠ 0 i.e. x1 ∉ Z ( f ) .

This shows that Z ( f ) is not a closed subspace of C ′[0,1]. From this

you can conclude that f is not continuous (by the property (vi) of
continuity).

E5) Tf ( x) = f ( x0 ), for some x0 ∈ [a, b]

Tf ( x ) ≤ f ( x ) ≤ f ∞
.

Consider f such that f ( x) = 1 ∀ x ∈ [a, b]

Then Tf ( x0 ) = 1 = f ∞

Therefore T = 1.

E6) For fixed x0 ∈ X , we get

79
Block 1 Normed Linear Spaces
Tf ( x) = f ( y) where y = x − x0

Then if h = kf + g where k ∈ K , f , g ∈ X , then

h( y) = kf ( y ) + Tg ( y)
Tf ( x) = k Tf ( x) + Tg ( x).

Thus T is linear. Also

Tf ( x ) = f ( y ) ≤ f

∴ T is bounded and T ≤ 1.

To find T , we take g = T ( f ).

g ( x + x0 ) ≤ f ( x) ≤ f i.e. g ≤ f

Also f ( x) = g ( x + x0 )
≤ g

Therefore f ≤ g

Thus f = g . Hence T = 1.

80
 UNIT 3

HAHN
HAHNBANACH THEOREMS

Structure Page No
3.1 Introduction
Objectives
3.2 Hahn-Banach Separation Theorem
3.3 Hahn-Banach Extension Theorem
3.4 Summary
3.5 Solutions/Answers

3.1 INTRODUCTION
There are two main versions of the Hahn Banach theorems, one is called
Hahn-Banach separation theorem which is geometric in nature and the other
is called Hahn-Banach extension theorem which is analytic in nature. Both the
theorems are concerned with the existence of bounded linear functionals with
special properties. In Sec. 3.2 we discuss Hahn Banach separation theorem
which deals with the separation of two disjoint convex subsets of a normed
space by a closed hyper plane. In Sec. 3.3 we consider Hahn-Banach
extension theorem which ensures the existence of a norm preserving linear
extension of a functional on a subspace of a normed linear space.

The Hahn Banach theorem is a central tool in the study of functional analysis.
Hans Hahn, an Austrian mathematician and Stefan Banach of poland, proved
the result independently in the 1920’s. Giving credit to both, the result is called
the Hahn Banach theorem.

Objectives
The objectives of the unit are:
• to state and prove Hahn-Banach separation theorem
• to state and prove Hahn-Banach extension theorem
• to give examples to show that a Hahn-Banach extension is not unique
• to state conditions under which Hahn-Banach extension are unique.

3.2 HAHN-BANACH SEPARATION THEOREM

Stefen Banach
This theorem illustrates how the set of linear functionals is helpful in studying (30th March, 1892
the topological structures of the underlying normed linear space. It gives us to 31st August, 1945
 Block 1 Normed Linear Spaces
some idea about how the elements in the space can be separated using non-
zero linear functionals.

To understand this we need to recall certain results which we have learnt from
the earlier units. You will be introduced to hyperplanes/hyperspaces and the
connection between these and non-zero bounded linear functionals.

Let us start with some preliminary results.

Before we start we make a remark on some notation.

Remark 1: In this unit as well as in other units also, we use the notation ' f '
instead of T for a bounded linear functional.

27th September, 1879 Proposition 1: Let X be a linear space over C . Regarding X as a real linear
to 24th July, 1934 space (i.e., the scalar multiplication is restricted to reals), consider a real linear
functional u : X → R.
Define f : X → C as

f ( x) = u ( x) − iu (ix), x ∈ X
Then f is a complex linear functional on X .

You may find the Remark 2: This result gives a sort of formula for converting a real linear
following link useful:
https://www.youtube.co
functional to a complex linear functional or complexifying a real linear
m/watch?v=FkJMfsNg2 functional.
cM&list=PLyqSpQzTE6
M8mjwWBz0vXpmJQw Proof: First we note that, since X is a linear space over C , if x ∈ X then ix
R46JIdS&index=10&t= also belongs to X . Also, because u is real linear, we cannot write u (ix) as
471s
equal to iu (x).

First suppose α1 , α 2 are reals and x1 , x 2 ∈ X .

Consider
f (α1 x1 + α 2 x 2 ) = u (α1 x1 + α 2 x2 ) − iu (i (α1 x1 + α 2 x2 ) )
= α1u ( x1 ) + α 2u ( x2 ) − i[u (α1 (i x1 ) + α 2 (i x2 ))]
= α 1u ( x1 ) + α 2 u ( x2 ) − i [α1u (i x1 ) + α 2 u (i x2 )]
(since u is real linear and α 1 , α 2 are reals.)

= α1 [u ( x1 ) − iu (ix1 )] + α 2 [u ( x2 ) − iu (ix2 )]
= α1 f ( x1 ) + α 2 f ( x2 )
Thus f is also real linear.
Now to prove f is complex linear, it is enough to prove f (ix) = if ( x), x ∈ X .
So let x ∈ X , we consider

f (ix) = u (ix) − iu (i(ix) )

= u (ix) − iu (− x) = u (ix) + iu ( x)
= i[u ( x) − iu (ix)] = if ( x).
Thus f is a complex linear functional.
Here is an exercise for you.
82
Unit 3 Hahn-Banach Theorems

E1) Prove that f as defined in the above proposition is complex linear i.e. if
α, β ∈ C, then f (αx + βy) = αf ( x) + βf ( y), ∀x, y ∈ X .

Next we give some definitions.

Definition 1: Suppose X is a normed linear space over K. A proper

subspace X 0 is called a hyperspace in X if there is a u ∈ X \ X 0 such that
span {u : X 0 } = X . i.e. X 0 is a maximal proper subspace of X .

[Recall the definition of span of a set S from the course MMT-002.]

Remark 3: If X is one dimensional then the trivial space {0} is the only hyper
space in X .

If X is 2-dimensional, then all 1-dimensional subspaces are hyperspaces and

in X is an n-dimensional space all (n − 1) dimensional subspaces are hyper
spaces.

Definition 2: If Z is a hyperspace in X and a ∈ X , then a + Z is called a

hyperplane in X .

We also observe that a proper subspace is maximal if and only if the span of
Z ∪ {a} equals X for each a ∉ Z . Further if Z is a hyperspace in X , for each
x ∈ X , ∃ unique z ∈ Z and k ∈ K such that
x = z + ka.
In other words
X = Z ⊕ [a] for any a ∉ Z .
Here [ a ] is 1-dimensional. Thus, Z is a hyperspace in X if Z has co-
dimension 1 in the sense that X is the direct sum of Z and an1-dimensional
space or, equivalently, the quotient X / Z has dimension 1.

Proposition 2: Let X be a non-zero linear space. Let f be a linear functional

on X and f ≠ 0. Then Z ( f ) is a hyperspace in X where
Z ( f ) = {x ∈ X : f ( x) = 0} . Further f is completely determined by Z ( f ) and
the value of f at any one point not in Z ( f ).

Proof: Since f ≠ 0, Z ( f ) is a proper subspace of X .

Let a ∈ X , a ∉ Z ( f ). Hence f (a ) ≠ 0. Now let x ∈ X .

f ( x)
Consider z = x − a.
f (a )
Then f ( z ) = 0  z ∈ Z ( f ).
 f ( x) 
Thus x = z +   a ∈ Z ⊕ [a]. s
 f ( a ) 
Thus X ⊆ Z ⊕ [a].
83
Block 1 Normed Linear Spaces
But Z ⊕ [a ] ⊆ X . Hence X = Z ⊕ [a].
This means Z is a maximal proper subspace of X i.e. Z is a hyper space in X
.
Now let g be any linear functional on X such that Z ( g ) = Z ( f ) & f (a) = g (a).
Then
f ( x)
g ( x) = g ( z ) + g (a )
f (a )
= f ( z ) + f ( x) = 0 + f ( x ) = f ( x ).
This means f is completely determined by Z ( f ) and f (a), where a ∉ Z ( f ).

Proposition 3: If Z is a hyperspace in a linear space X , then Z = Z ( f ) for

some linear functional f on X .

Proof: Since Z is a hyperspace in X ,

X = Z ⊕ [a] for some a ∉ Z .

Hence every x ∈ X has a unique representation x = z + ka, z ∈ Z and k ∈ K.
Define f ( x) = k .
Then f is linear and Z ( f ) = Z (why don’t you check it for yourself?)

Remark 4: The two results above characterize hyperspaces.

Thus we have the following result.

Theorem 1: A subspace is a hyperspace if and only it is the zero space or the

null space of a non-zero linear functional.

Proposition 4: Let X be a linear space over K and Y be a subspace of X

which is not a hyperspace in X (i.e. not a maximal subspace). If x1 , x 2 ∈ X but
∉ Y , then ∃ some x ∈ X , such that
tx1 + (1 − t ) x ∉ Y and
tx 2 + (1 − t ) x ∉ Y for all t ∈[0,1].
Before giving the proof we make a remark.
3
Remark 5: Imagine this in R . Any plane passing through the origin is a
3 3
hyperspace in R . Let Y be such a plane, let x1 , x 2 lie in R but not in Y .
For any x ∈ X ,

tx1 + (1 − t ) x & tx2 + (1 − t ) x, t ∈[0,1]

are two straight line segments with x as a common end point. In this situation
we cannot find a point in X such that both the segments
tx1 + (1 − t ) x & tx 2 + (1 − t ) x do not meet Y . One of the segments has to cut
Y , if the lines should have a common end point.

Proof: If tx1 + (1 − t ) x 2 ∉ Y , ∀t ∈ (0,1) then we set x = x 2 , and the required

property is satisfied.
84
Unit 3 Hahn-Banach Theorems
Now suppose sx1 + (1 − s ) x 2 ∈ Y for some s ∈ (0,1).

Claim: span {Y , x1 ) = span {Y , x 2 ).

By supposition sx1 + (1 − s ) x 2 ∈ Y for some s ∈ (0,1). Suppose

sx1 + (1 − s ) x 2 = y 0 , y 0 ∈ Y .
Then, since s ∈ (0,1), s ≠ 0,1.

1 (1 − s)
So x1 = y0 − x2
s s
and
1 s
x2 = y0 − x1 .
(1 − s ) (1 − s )
Suppose x ∈ span {Y , x1 }
 x = y + kx1 , for some y ∈ Y and k ∈ K.

1 1 − s  
= y + k  y 0 −   x 2 
s  s  
 k  1− s 
=  y + y0  −   x2
 s   s 
∈ span {Y , x 2 }.
Thus span {Y , x1 } ⊆ span {Y , x 2 }.

Similarly we can prove span {Y , x 2 } ⊆ span {Y , x1 }.

This proves the claim.
Since Y is not a hyperspace ∃ x ∈ X such that x ∉ span {Y , x1 }.

Suppose now, tx1 + (1 − t ) x ∈ Y for some t ∉ (0,1).

Set tx1 + (1 − t ) x = y, y ∈ Y .

( y − tx1 )
Then x = ∈ span {Y , x1}, a contradiction.
1− t
Hence tx1 + (1 − t ) x ∉ Y ∀ t ∈ (0,1). Since x1 & x ∉ Y , we get

tx1 + (1 − t ) x ∉ Y ∀ t ∈ [0,1].
Similarly since x ∉ span {Y , x2 }, proceeding as above we get tx 2 + (1 − t ) x ∉ Y
for t ∈ [0,1].
This completes the proof.

Remark 6: You may observe that Proposition 4 already gives that Y c is path
connected and hence connected.

Now we prove a ‘big’ theorem which leads to the Hahn Banach Separation
Theorem.

85
Block 1 Normed Linear Spaces
Theorem 1: Let X be a normed space over K and Y be a subspace of X . Let
E be a non-empty open convex subset of X such that E I Y = φ. Then there
exists a closed hyperspace Z in X such that Y ⊂ Z and E I Z = φ.

Consequently there is some f ∈ X ′ such that f ( x ) = 0 ∀ x ∈ Y and

Re f ( x) ≠ 0 ∀ x ∈ E.
Z
b
Before we prove the theorem we make the following remark. p
s
Remark 7: The first part of the theorem says: if there is a subspace Y of X o
which is disjoint from a given convex set E , then we can find a maximal h
subspace of X which contains Y and still disjoint from E or, Y can be enlarged
to a maximum extent which will still be disjoint from E. m

We prove the theorem in different steps. Each step we call as a claim and
each claim has subclaims the proof of which are given in different steps.
We also prove the theorem first by taking K = R which we call the real part
and then for K = C which we call the complex part.

Proof (Real Part): To begin, we assume X is a real normed space i.e., we

take K = R.
First we note that we want a subspace which is maximal, we naturally appeal
to Zorn’s Lemma.
Set W = {W : W is a subspace of X , Y ⊂ W and E I W = φ}.
By hypothesis Y ∈ W. Hence W is non-empty. Partially order W by inclusion,
i.e. W1 ,W2 ∈ W, say W1 ≤ W2 whenever W1 ⊆ W2 . This defines a partial order
on W .
Now let F be a totally ordered subset of W i.e., a chain in W .
Hence two members of F are comparable.
Set P = U F
F∈F

Since members of F are comparable, P is a subspace of X and Y ⊂ P

since Y ⊂ F , ∀F ∈ F .
Also F I E = φ∀F ∈ F  P I E = φ.

 P I E = ( U F ) I E = U ( F I E ) = Φ 
 F∈F F∈F 
Thus P is a subspace of X such that
Y ⊂ P & P I E = φ  P ∈ W.
Also by definition of P, F ≤ P, ∀F ∈ F .
Thus every totally ordered subset in W has an upper bound.
Hence by Zorn’s Lemma, W has a maximal element, say Z .
Since Z ∈ W, Y ⊂ Z and Z I E = φ.
Claim 1: Z is a hyper space.
Here we have to be careful. By appealing to Zorn’s lemma, we have obtained
a maximal element in W in the sense that there exists no proper subspace in
X between Z and X , having the required property.
86
Unit 3 Hahn-Banach Theorems
Now we have to prove it is also a maximal subspace of X having the required
properties.
On the contrary suppose that Z is not a hyperspace, then we arrive at a
contradiction.
For that we proceed as follows:
Set S = Z + ∪{αE : α > 0}.
Since E is open, αE is open for any α > 0 and since an open set + a subset
of X is always open, S is open. [by Proposition 8 in Unit 1]
Similarly − S is also open.
Further both S and − S are non-empty as E is non-empty.
Now we claim the following:
Subclaim 1: S ∩ − S = φ.
We prove this in step 1 by assuming the contrary and arriving at a
contradiction.
Step 1: Suppose that S ∩ − S ≠ φ

Suppose p ∈ S ∩ − S .

Then p = z1 + α1 x1 and p = − z 2 − α 2 x2 for z1 z 2 in Z , x1 x 2 ∈ E and

α1 > 0, α 2 > 0.
Thus z1 + α1 x1 = − z 2 − α 2 x2

( z1 + z 2 ) α1 α2
 − = x1 + x2
α 1 + α 2 α1 + α 2 α1 + α 2
E is convex  right hand side lies in E.
Z is a subspace  left hand side lies in Z .
( z1 + z 2 ) α1 α2
 − = x1 + x2 ∈ E ∩ Z .
α 1 + α 2 α1 + α 2 α1 + α 2
Thus S ∩ − S ≠ φ  E ∩ Z ≠ φ, a contradiction.

Thus S ∩ − S = φ.
Next we claim the following:
Subclaim 2: S ∪ − S ⊂ Z c .
+
We prove this in Step 2 by assuming the contrary and arriving at a
contradiction.
Step 2: Suppose there exists t ∈ S ∪ − S which also lies in Z .
Then t ∈ S U −S  t = z1 + α1 x1 or
t = − z 2 − α 2 x2 , z1 z 2 ∈ Z , x1 x2 ∈ E.
α1 > 0, α 2 > 0.
t also lies in Z .
t − z1 t − z1
If t = z1 + α1 x1 then = x1 ∈ E , ∈ Z.
α1 α1 1
87
Block 1 Normed Linear Spaces
Thus E I Z ≠ φ, a contradiction. Similarly if t = − z 2 − α 2 x2 , even then we get a
contradiction.
Thus no element of S U − S lies in Z , i.e., S U −S ⊂ Z c .

Now if S ∪ −S = Z c then Z c becomes a disjoint union of two non-empty open

c
sets, i.e. Z is not connected.
By Proposition 5, Z c is connected by our assumption on Z (i.e. Z is not a
hyperspace). Thus we get a contradiction under the supposition S U − S = Z c .

Hence S U −S ⊂ Z c .
+
Thus ∃ x0 ∈ Z c ⊂ X such that

x0 ∉ S U −S .

Hence we have the required result that S ∪ − S ⊂ Z c .

Next we claim the following.
Subclaim 3: E I span {Z , x 0 } = φ.

Once we prove this subclaim we get that Z is not maximal in W , which is a

contradiction.
All this happens since we started with supposition Z is not a hyperspace
which cannot happen.
Hence Z should be a hyperspace.
Therefore to end the proof of claim 1 it is enough to prove subclaim 3.

Now let us prove the last subclaim, namely E I span {Z , x 0 } = φ.

We prove this subclaim by assuming the contrary and arriving at a
contradiction.
Step 3: Suppose that there exists x ∈ E I span {Z , x0 }.

 x ∈ E and x = p + kx0 .
where p ∈ Z & k ∈ R.

If k = 0, then x = p  E I Z ≠ φ, a contradiction.

1 1
If k > 0, then x0 = x− p
k k
1 1
=− p + x ∈ S,
k k
a contradiction since x0 ∉ Z .

1 1
Similarly if k < 0, x0 = − p + x ∈ −S.
k k
again a contradiction.
This means E I span {Z , x 0 } = φ.

Thus Z is hyperspace.
88
Unit 3 Hahn-Banach Theorems
Further, in the theorem we need a closed subspace.
Let us now check whether the hyperspace Z is a closed subspace. By
construction, E I Z = φ.

It is given that E is open and so E c is closed.

E I Z = 0/  Z ⊂ E c
 Z ⊂ Ec
 E I Z = 0/
This also shows Z ≠ X .

Thus, Z is also a subspace of X such that Y ⊂ Z ⊂ Z & E I Z = φ.

 Z ∈ W.

But Z ⊂ Z ⊂ X and Z is maximal  Z = Z  Z is closed.

Thus Z has all the properties stated in Theorem 1.
We have proved the result assuming K = R. Next we prove the result for
K = C.
Complex Part: Here we note that if we consider X as a normed space over
R, by what we have proved above, there exists a real hyperspace Z such
that Y ⊂ Z and Z I E = φ.

By Proposition 3, ∃ a real linear functional u on X such that Z = Z (u ), the zero

space of u. Since E I Z = φ, u ( x ) ≠ 0 ∀ x ∈ E. Since
Y ⊂ Z , u ( x) = 0 ∀ x ∈ Y .
Now define f as:

f ( x) = u ( x) − iu (ix ), x ∈ X ,
as in Proposition 1. Then as proved in Proposition 1, f is a complex linear
functional on X .
But by Theorem 4 of Unit 2, since Z = Z (u ) is closed, u is a real continuous
linear functional on X and f is a continuous complex linear functional on X ,
i.e. f ∈ X 1 .

Now if x ∈ Y , ix also lies in Y since Y is a subspace and u (ix ) = 0.

Hence if x ∈ Y , f ( x ) = u ( x) − iu (ix) = 0. Also if x ∈ E , u ( x ) ≠ 0.
 Re f ( x) = u ( x) ≠ 0.
By Proposition 2, Z = Z ( f ) is a closed hyperspace in X . By what we have
proved above
Y ⊂ Z and E I Z = 0/ .
This completes the proof of the theorem.

Remark 8: By the above result we have, if Y is a proper closed subspace of

X , then ∃ a non-zero f ∈ X ′ such that f Y = 0.
In particular if X ≠ {0}, then taking Y = {0}, we get X ′ ≠ {0}, i.e. the dual
space of a non-trivial space is non-trivial.
89
Block 1 Normed Linear Spaces
Proposition 5: Let X be a normed space over K and f a non-zero linear
functional on X . If E is an open set in X , then f (E ) is open in K, i.e., f is
an open map.

Proof: Since f is non-zero, ∃ a 0 ∈ X such that f (a 0 ) ≠ 0. Suppose

a0 f (a0 )
f (a0 ) = α. Consider a = then f (a ) = = 1. (Q f is linear ).
α α
Now let us prove f (E ) is open in K . Let k ∈ f (E ). Then k = f (x ) for some
x ∈ E. But E is given to be open. Hence ∃r > 0 such that U ( x, r ) ⊂ E. Now
r
suppose β ∈ K be such that β < .
a

Consider x − β a. Then ( x − β a) − x = β a < r.

Now consider the neighbourhood

 r 
U  k ,  of k .
 a 
 
 r 
Let δ ∈ U  k , .
 a 
 
r
Then δ−k < .
a
r
= f ( x) − δ < .
a
By what we have proved above
f (x − ( f ( x ) − δ) a ) ∈ f ( E ).
 f ( x ) − ( f ( x ) − δ ) f (a) ∈ f ( E ).
 δ ∈ f (E ), since f (a) = 1.
 r 
Thus U  k ,  ⊂ f ( E ).
 a 
 
Thus k ∈ f (E )  ∃ a neighbourhood of k contained in f (E ), whenever E is
open in X .
Thus f is an open map.

Theorem 2 (Hahn-Banach Separation Theorem: Hahn 1927, Banach

1929): Let X be a normed space over K and E1 , E 2 be non-empty disjoint
convex subsets of X , where E1 is open in X . Then there exists a real
hyperplane in X which separates E1 and E 2 in the following sense:

For some f ∈ X ′ and t ∈ R, we have

Re f ( x1 ) < t < Re f ( x 2 )
for all x1 ∈ E1 and x2 ∈ E2 .
90
Unit 3 Hahn-Banach Theorems
2
Recall: A hyperplane is just a translate of a hyperspace. In R , a hyperplane
is any straight. In R 3 , a hyperplane is any plane.

Geometrically, in R 2 , the above theorem says, given any two non-empty

disjoint convex set in R 2 , we can always find a hyperplane, i.e., an infinite
straight line which separates them.

Proof: Consider the set E1 − E 2 = {x1 − x 2 : x1 ∈ E1 , x 2 ∈ E 2 }. This set is non-

empty since both E1 and E 2 are non-empty.

Since E1 is open, by Proposition 8 in Unit 1, E1 − E 2 is open in X .

Further, both E1 and E 2 are convex  E1 − E 2 is also convex. (easy to check

that the definition of convexity is satisfied by E1 − E 2 ). Also 0 ∉ E1 − E 2 ,
otherwise 0 ∈ E1 − E 2  0 = x1 − x 2 for some x1 ∈ E1 and x 2 ∈ E 2 . But
x1 − x 2 = 0  x1 = x 2  E1 I E 2 ≠ φ, a contradiction. So 0 ∉ E1 − E 2 .

By Theorem 1, ∃ g ∈ X 1 such that Re f ( x1 − x 2 ) ≠ 0 for all x1 ∈ E1 and

x2 ∈ E2
Re f ( x1 ) − Re f ( x 2 ) ≠ 0 ∀ x1 ∈ E1 and x 2 ∈ E 2 .
 Re f ( x1 ) ≠ Re f ( x 2 ), x1 ∈ E1 , x 2 ∈ E 2 .
Claim: Re f ( E1 ) and Re f ( E 2 ) are convex sets in R.

Let p1 , p 2 ∈ Re f ( E1 ).

The p1 = Re f ( x1 ) & p 2 = Re f ( x 2 ) for some x1 , x2 ∈ E1 .

Now for t ∈ (0,1) consider

tp1 + (1 − t ) p 2 = tf ( x1 ) + (1 − t ) f ( x 2 )
= f (tx1 + (1 − t ) x 2 ) (Q f is linear)

∈ f ( E1 ) since E1 is convex.
Thus, p1 , p 2 ∈ Re f ( E1 ) 

tp1 + (1 − t ) p 2 ∈ Re f ( E1 ) ∀ t ∈[0,1].
 Re f ( E1 ) is convex.
Similarly we can show that Re f ( E 2 ) is convex. But Re f ( E1 ) and Re f ( E 2 )
are convex sets in R  they are intervals.
Since Re f ( x1 ) ≠ Re f ( x 2 ), x1 ∈ E1 , x 2 ∈ E 2

Re f ( E1 ) and Re f ( E 2 ) are non-overlapping intervals in R.

We can assume that Re f ( E1 ) lies to the left of Re f ( E 2 ). (If this not the case
we can replace f by − f ).

Thus we have obtained an f ∈ X ′, f ≠ 0, such that Re f ( E1 ) lies to the left of

Re f ( E 2 ). Since E1 is open, by Proposition 6, Re f ( E1 ) is open, i.e. Re f ( E1 )
is an open interval.
91
Block 1 Normed Linear Spaces
Suppose t is the right end point of Re f ( E1 ).

Then we have: Re f ( x1 ) < t ≤ Re f ( x 2 ), for all x1 ∈ E1 and x 2 ∈ E 2 .

Since a linear functional is onto, ∃ a ∈ X such that Re f (a) = t.

Consider the set
{x ∈ X : Re f ( x) = t}
= {x ∈ X : Re f ( x) = Re f (a)}
= {x ∈ X : Re f ( x − a) = 0}.
Putting y = x − a, we get
= { y + a : Re f ( y )} = 0.
= Z (Re f ) + a.
= translate of a hyperspace & hence a hyperplane.
Thus, {x ∈ X : Re f ( x) = t} is the required hyperplane and sets E1 & E 2 lie
on either side of this hyperplane.
Now let us look at some important consequences of the Hahn Banach
Separation theorem.

Corollary 1: Let X be a normed space over K and E be a non-empty convex

subset of X . Suppose a ∈ X but a ∉ E . Then ∃ f ∈ X ′ and t ∈ R such that

Re f ( x ) ≤ t < Re f (a )∀ x ∈ E.

Proof: Since E is convex, its closure E is also convex.

Since a ∉ E , we can find a neighbourhood. U (a, r ) of a , which is non-empty,

open and convex such that U (a, r ) I E = φ .

Now taking E1 = U (a, r ) and E 2 = E in the Hahn Banach Separation theorem,

we get a g ∈ X ′ and a t ∈ R such that

Re g (a) < t ≤ Re g ( x), ∀ x ∈ E.

Taking f = − g and s = −t , we get Re f ( x) < s ≤ Re f (a), ∀ x ∈ E.
Hence the result .

Remark 9: If a ∉ E , the above corollary says, there exists a real hyperplane

which separates E from a.

Corollary 2: Suppose X is a normed space over K and E a non-empty

convex set in X . Suppose E 0 ≠ φ and b belongs to the boundary of E . Then
there is a non-zero f ∈ X ′ such that
Re f ( x ) ≤ Re f (b), for all x ∈ E .

Recall: In a metric space X , if A is a subset of X , then the boundary of A ,

denoted by ∂A is defined as

∂A = A − A0 = A I A c

92
Unit 3 Hahn-Banach Theorems
0
Proof: Since E is convex, the interior E is also convex.
Since, it is given that b ∈ ∂E , b ∉ E 0 .

Take E1 = E 0 and E 2 = {b}. Then E1 and E 2 are non-empty disjoint convex

set with E1 open, by the hypothesis.

By Hahn-Banach separation theorem ∃ non-zero f ∈ X 1 and a t ∈ R, such

that
Re f ( x ) < t ≤ Re f (a ) for x ∈ E1 = E 0 .

Hence Re f ( x ) < Re f (b), ∀ x ∈ E 0 .

But E 0 ≠ 0/ and b belongs to the boundary of E , hence E = E 0 (please

convince yourself).
Hence we have Re f ( x ) ≤ Re f (b), ∀ x ∈ E .

Definition 3: Let X be a normed space over K. A convex subset E of X with

E 0 ≠ φ is called a convex body. If b is a boundary point of a convex body E
and f ∈ X ′ is such that f ≠ 0 and Re f ( x ) ≤ Re f (b), then f is called a
support functional for E at b.
The corresponding hyperplane {x ∈ X : Re f ( x ) = Re f (b)} is called a support
hyperplane for E at b.

Remark 10: The above result says if E is a convex body and b is a boundary
point of E then there always exists a support hyperplane for E at b.
The inequality Re f ( x ) ≤ Re f (b) again means the support hyperplane
separates E 0 and b !

You can try the following exercises.

E2) Let X be a linear space over C and f be a complex linear functional

on X . Prove that
i) Re f is a real linear functional on X
ii) f ( z ) = Re f ( z ) − i Re f (iz ), z ∈ C
iii) If ⋅ is a norm on X , then Re f = f

Now we move on to look at the Extension version of the Hahn Banach

theorem.

3.3 HAHN-BANACH EXTENSION THEOREM

Hahn-Banach Extension Theorem is one of the very important theorems of
Functional Analysis. It is not only a powerful theorem but also has many
striking consequences. For example it tells us that spaces have a rich supply
of continuous linear functionals. This makes it possible to develop an
adequate theory of conjugate spaces i.e. spaces of bounded linear functionals.
Because of the Hahn-Banach Theorem, we can regard any bounded linear 93
Block 1 Normed Linear Spaces
functional as defined on the whole normed linear space. The theorem is first
proved for real spaces and then, with suitable modification, for complex
spaces.

The theorem says that if a ∈ X , then the function f 0 : span {a } → K defined

by f 0 (α a) = α, α ∈ K is a bounded linear functional on span { x0 } . The
question arises now: Does there exists a bounded linear functional g such that
g is an extension of f . The Extension theorem says that the answer is yes.
First we prove the following result.

Proposition 6: Let X be a normed space over K and f ∈ X′. Suppose

f ≠ 0, a ∈ X and f (a) = 1. Let r > 0.
1
Then U (a, r ) I Z ( f ) = φ ⇔ f ≤ .
r

Proof: Suppose U (a, r ) I Z ( f ) = φ.

x
If x ∈ X and f ( x) ≠ 0, consider, a − then
f ( x)

 x  f ( x)
f  a −  = f (a ) −
 f ( x)  f ( x)
= 1 − 1 = 0.
x
Hence a − ∈ Z( f )
f ( x)
x
By hypothesis a − ∉ U (a, r )
f ( x)

x
Hence a − −a ≥ r
f ( x)

x
 ≥r
f ( x)

 x ≥ r f ( x)

1
 f ( x) ≤ ⋅ x r.
r
Thus for x ∈ X , with f ( x) ≠ 0, we have

1
f ( x) ≤ ⋅ x ∀ x ∈ X .
r
If f ( x) = 0, for some x, the above inequality is trivial.

1
Thus f ( x) ≤ ⋅ x , ∀ x∈ X.
r
1
 f ≤ .
r

94 Conversely suppose
Unit 3 Hahn-Banach Theorems
1
f ≤ .
r
Let x ∈ U (a, r ). Then x − a < r .

Consider now
f ( x ) − 1 = f ( x ) − f (a ) = f ( x − a )

1
≤ f x − a < . r = 1.
r
 f ( x ) − 1 < 1.

 1 − f ( x ) ≤ 1 − f ( x) ≤ 1 − f ( x) < 1

 − f ( x ) < 0  f ( x) ≠ 0.

 x ∉ Z ( f ).
Hence U (a, r ) I Z ( f ) = φ.

Theorem 3 (Hahn-Banach Extension Theorem): Let X be a normed space

over K and Y be a subspace of X and g ∈ Y ′. Then there is some f ∈ X ′
such that
f / y = g and f = g.

Remark 11: This extension theorem says, a bounded (= continuous) linear

functional on a subspace of a normed space can be extended to the whole
pace with the preservation of the norm.

Proof: If g = 0 on Y , we can take f = 0. So let us suppose g is a non-zero

bounded functional on Y .
Let a ∈ Y be such that g (a) ≠ 0. Assume g (a) = 1. (If g (a ) ≠ 1, we can
a  a   g (a ) 
replace a by . Then g   =  = 1.
g (a )  g (a )   g (a ) 
1
Taking r = in the above lemma, we get
g
 1 
U Y  a,  I Z ( g) = φ (1)
 g 
 
We have to consider the neighbourhood in Y , since g acts on Y .

 1
Let E = U X  a, . Then E is a non-empty open convex subset of X and it
 g
 
is disjoint from the subspace Z (g ) of X . Hence by the Hahn-Banach
separation theorem, ∃ f ∈ X ′ such that f ( y ) = 0 ∀ y ∈ Z ( g ) and Re f ( x) ≠ 0
for every x ∈ E .
Hence f ( x ) ≠ 0 for every x ∈ E. (2)

Since a ∈ E , f (a) ≠ 0.

95
Block 1 Normed Linear Spaces
f
f / y (a ) = 1 = g (a ). If necessary we an replace and assume f (a) = 1
f (a)
f
(calling also as f ). Now f / y is a mapping from
f (a)
Y → K, f / y (a) = 1 = g (a).
Let y ∈ Y . Consider y − g ( y )a

g ( y − g ( y )a ) = g ( y ) − g ( y ) g (a ) = g ( y ) − g ( y ).1 = 0
 y − g ( y )a ∈ Z ( g ) ⊂ Z ( f / y ).
 f ( y − g ( y ) a ) = 0  f ( y ) − g ( y ) f ( a) = 0
 f ( y ) = g ( y ).
Thus for every y ∈ Y , f ( y ) = g ( y ).
i.e., f / y = g .

Hence g = f /y ≤ f . (3)

 1 
By (2) f ( x) ≠ 0 ∀ x ∈ E = U X  a, 
 g 
 
 1 
 U X  a,  I Z ( f ) = φ.
 g 
 
1
Using Proposition 7, we get f ≤ = g. (4)
1
g

From (3) & (4) we get f = g .

Thus f is the required extension of g , with the preservation of the norm.

Definition 4: For a subspace Y of a normed space X and g ∈ Y ′, a Hahn-

Banach extension of g to X is an element f of X ′ such that f / y = g &
f = g.

Remark 12: The Hahn-Banach extension theorem says, every continuous

(= bounded) linear functional on a subspace always has a Hahn-Banach
extension to the whole space.
Definition 5: Let T be a set and X be the linear space of all scalar valued
functions on T . A linear functional f on X is said to be positive if f ( x ) ≥ 0
whenever x ∈ X and x(t ) ≥ 0 for t ∈ T . The constant function which takes the
value 1 ∀ t ∈ X will be dented by 1.

Proposition 7: Let T be a set and B (T ) be the linear space of all scalar

valued bounded functions on T . Consider B (T ) with the sup norm, . ∞
. Let
X be a subspace of B(T ) such that 1∈ X . Let f be a linear functional on X . If
f is continuous and f = f (1), then f is positive.
96
Unit 3 Hahn-Banach Theorems
Conversely if Re x ∈ X whenever x ∈ X and if f is positive, then f is
continuous and f = f (1).

Proof: Assume f is continuous and f = f (1).

Let x ∈ X and 0 ≤ x (t ) ≤ 1 ∀ t ∈ T .

Let us prove that f ( x) ≥ 0.

Set y = 2 x − 1. Then − 1 ≤ y (t ) ≤ 1 ∀ t ∈ T .

Claim: f ( y ) is real.
If K = R, there is nothing to prove.

Let K = C/ , and suppose f ( y ) = α + iβ, where α, β are real.

Now, for every t ∈ T , and any real a,

2
y (t ) + ia = y (t ) 2 + a 2 ≤ 1 + a 2

(Q − 1 ≤ y(t ) ≤ 1).
2
 sup y (t ) + ia ≤ 1 + a 2
t ∈T

 y + ia ∞
≤1+ a2 (5)

(Here the real a is treated as the constant function taking the value a ).

(
Now consider β + a f ) = (β + af (1) )
2 2

2
≤ α + i (β + af (1)
2
= α + iβ + iaf (1)
2
= f ( y ) + f (ia ) , since f is linear.
2
= f ( y + ia )
2 2
≤ f y + ia ∞
(since f is continuous).
2
≤ f (1 + a 2 ) by (1)

2 2 2
Thus we have β + 2aβ f + a2 f
2 2
≤ f + a2 f .
2 2
 β 2 + 2aβ f ≤ f , for every real a.

1
If β ≠ 0, taking a = , we get
2β
2 2
β2 + f ≤ f

 β 2 ≤ 0, impossible.
Hence β = 0.
97
Block 1 Normed Linear Spaces
Hence f ( y ) = α i.e. f ( y ) is real.
This proves the claim.
Since y ∞
≤ 1, we have

f ( y) = α ≤ f .

 y +1 α + f
Hence f ( x) = f  = ≥ 0.
 2  2
i.e. f ( x) ≥ 0.

(Here we have proved f ( x) ≥ 0 whenever x is such that 0 ≤ x ≤ 1. )

Now suppose x ∈ X be arbitrary with x(t ) ≥ 0 and x ≠ 0 on T .

(If x ≡ 0, there is nothing to prove.)

x
Then set z = . Then 0 ≤ z (t ) ≤ 1, ∀ t.
x∞
Hence by what we have proved above
f ( z) ≥ 0

 x 
 f  ≥ 0  f ( x) ≥ 0.
 x 
 ∞
Thus ∀ x ∈ X with x(t ) ≥ 0, we have

f ( x) ≥ 0.
 f is positive.
Converse part: suppose Re x ∈ X whenever x ∈ X , and f is positive.

Let y be a real valued function in X . We have y ∞

− y (t ) ≥ 0, ∀ t ∈ T .

 ( − y) ≥ 0
f y ∞

 f ( y ) − f ( y) ≥ 0
∞

 f ( y .1) − f ( y ) ≥ 0
∞

 y ∞
f (1) − f ( y ) ≥ 0

 y ∞
f − f ( y) ≥ 0

Here y ∞
f is real  f ( y ) also should be real.

Let x be any arbitrary element of X .

Suppose f ( x) = reis , where r, s are real.

−is
Consider y = e x.

Let y1 = Re y & y 2 = Im y.
Since y ∈ X , by hypothesis Re y ∈ X
98
Unit 3 Hahn-Banach Theorems
i.e. y1 ∈ X .

We have r = f (e − is x) = f ( y1 + iy2 )

= f ( y1 ) + if ( y 2 ).
Since y1 is real, by what we have proved above f ( y1 ) is real.

Hence i f ( y 2 ) = r − f ( y1 ) is real.

 f ( y 2 ) = 0.
We have:

y1 (t ) ≤ y1 ∞
≤ e −is x .
∞

= x ∞.

 − x ∞
≤ y1 (t ) ≤ x ∞
∀ t ∈T

Thus ( )
f ( x ) = f e −is x = f ( y1 )

≤ f x( ∞
) (since f is positive and x ∞
− y (t ) ≥ 0)

= x ∞
f (1)

= x ∞
f

Thus f ( x ) ≤ f x ∞
 f is continuous.

We shall see some examples of Hahn-Banach Extensions.

Example 1: Let T be a metric space and let B (T ) be normed space of all

bounded scalar valued functions on T with the sup norm. Let C (T ) be the
subspace of B (T ) consisting of bounded continuous functions on T .
Let g be a positive linear functional on C (T ). Then g is continuous by the
above result and also g = g (1). Here we note that 1∈ C (T ) (1 is the constant
one function) and Re x ∈ C (T ) whenever x ∈ C (T ).

Then by the Hahn-Banach Extension theorem, ∃ a linear functional f on

B(T ) such that f / C (T ) = g and f = g .

But g = g (1). Hence we have

f = g = g (1) = f (1)

i.e. f = f (1).

By the above result we see that f is also positive.

Thus, in this case the Hahn-Banach Extension f of g is also a positive linear
functional on B (T ).
***

Example 2: Let X = K 2 with the 1-norm, . 1 .

99
Block 1 Normed Linear Spaces
(i.e.: for x = ( x1 , x2 ) ∈ X , x 1 = x1 + x2 )

Let Y = {( x1 , x2 ) ∈ X : x2 = 0}

Define g on Y as g ( x1 , x 2 ) = x1 .

Then g is linear and X is finite dimensional.

 g is continuous.

For any x = ( x1 , x 2 ) ∈ Y ,

g ( x ) = x1 ≤ x1 + x 2 = x 1 .

 g ≤1 (6)

Consider x0 = (1,0), x 0 1 = 1.

Also g ( x 0 ) = 1  g ( x 0 ) = 1.

Hence g ≥ g ( x 0 ) = 1. (7)

From (6) & (7), we get g = 1.

Now a functional f on X will be a Hahn-Banach extension of g if and only if

f is linear on X and f = g = 1. Continuity is already there since K 2 is
finite dimensional.
Now f is linear functional on X = K 2

 f should be of the form.

f ( x1 , x 2 ) = k1 x1 + k 2 x 2 , k1k 2 ∈ K
2
and ( x1 , x2 ) ∈ K .
If f has to be an extension of g , then f should be of the form.

f ( x1 , x 2 ) = k1 x1 + k 2 x 2 for some fixed

k1k 2 ∈ K & all ( x1 , x2 ) ∈ K 2

with f / y = g and f = g = 1.
2
Claim: If f ( x1 , x2 ) = k1 x1 + k 2 x2 for some fixed k1 , k 2 ∈ K and all ( x1 , x2 ) ∈ K ,
then f = max {k 1 , k 2 }. Consider
f ( x1 ,.x2 ≤ k1 x1 + k 2 x2 = k1 x1 + k 2 x2 ≤ max{ k1 , k 2 } x1 + max{ k1 , k 2 } x2

= max { k1 , k 2 }( x1 + x2 )

= max { k1 , k 2 } ( x1 , x2 ) 2

 f ≤ max { k1 , k 2 } (8)

Also f (1, 0) = k1  f (1, 0) = k1  f ≥ k1 and

f (0, 1) = k 2  f (0, 1) = k 2  f ≥ k 2
100
Unit 3 Hahn-Banach Theorems
{
Thus f ≥ max k1 , k 2 } (9)

From (8) and (9) we get the claim

f / y = g  f ( x1 , x2 ) = g ( x1 , x2 ) = x1 , ∀( x1 , x2 ) ∈ X .
 But f ( x1 , x 2 ) = k1 x + k 2 x 2

 k1 = 1.
So f should be of the form.

f ( x1 , x 2 ) = x1 + k 2 x 2 , for some k 2 and

all ( x1 , x 2 ) ∈ K 2

But f = g = 1

i.e. 1 = f = max {1, k 2 }

 k2 ≤ 1

Thus the functionals f ( x1 , x 2 ) = x1 + k 2 x 2 with k 2 ≤ 1, give Hahn Banach

extensions of g . Thus there exists infinitely many Hahn Banach extensions of
g.
***

Remark 13: The above example shows the Hahn-Banach extensions are in
general not unique.

Thus Hahn-Banach extensions always exist, but may not be unique. So the
natural question is: Are there conditions under which Hahn-Banach extensions
are unique. We address this question little later.

Before we discuss conditions for uniqueness of Hahn-Banach extensions we

shall discuss some corollaries/consequences of the Extension Theorem. You
will see in later units that these results are very important and will be used in
proving many other important results.

One of the main consequence of the Hahn-Banach theorem is that it assures

that there is a rich supply of bounded linear functional as proved in the
following corollary.

Corollary 1: Let X be a normed linear space over K. Let a ∈ X , a ≠ 0. Then

there exists a bounded linear functional f on X such that f ( a) = a and
f = 1.

{
Hence a = sup f ( a ) : f ∈ X ′, f ≤ 1 . }
Proof: Take Y = {ka : k ∈ K). Then Y is a closed subspace of X .

Define g on Y as g ( ka) = k a , k ∈ K.

Then g is a continuous linear functional on Y . Hence g has a Hahn Banach

extension say f , i.e. f ∈ X ′, f / y = g and f = g .
101
Block 1 Normed Linear Spaces
By definition of g , we have

g ( a) = a

Also g = sup{ g ( ka ) : ka ∈ Y & ka = 1}

= sup{ ka : ka ∈ Y & ka = 1} = 1

i.e. g = 1.
Hence we have:
f ( a ) = g ( a ) = a & f = g = 1.
Thus we have
a = f ( a) = f ( a) (Q f (a) = a ≥ 0)

≤ sup {h( a) : h ∈ X ′, h ≤ 1}

Also h ∈ X ′  h( a) ≤ h a

Hence sup {h( a ) : h ∈ X ′, h ≤ 1}

≤ sup { h a : h ∈ X ′, h ≤ 1}

= a . sup { h : h ∈ X ′, h ≤ 1}

= a .1 = a .

Thus: a = sup { f (a) : f ∈ X ′, f ≤ 1}.

Remark 14: This theorem gives a formula for the norm of an element of X in
terms of bounded linear functionals.

Corollary 2: Let Y be a subspace of a normed linear space X . Let a ∈ X but

a ∉ Y . Then there exists an f ∈ X ′ such that f is zero every where on Y ,
f = 1 and f (a) = dist (a, Y ).

Proof: We have Y is a closed subspace of X . Consider the quotient space

X X
. Since a ∉ Y , the coset a + Y is a non-zero element in .
Y Y
′
~ X ~
Then by the above theorem, ∃ f ∈   such that f (a + Y ) = a + Y and
Y 
~
f = 1.
~
Define f : X → K as f ( x) = f ( x + Y ) for x ∈ X .

Then f Y = 0 and f (a) = a + Y = dist (a, Y ).

~
Also f = f = 1 since:
~ ~
f ( x) = f ( x + Y ) ≤ f x + Y

102
Unit 3 Hahn-Banach Theorems
~
≤ f x
~
 f ≤ f .

Further, for any y ∈ Y ,

~ ~
f ( x + Y ) = f ( x + y + Y ) = ( f ( x + y)

≤ f x+ y
~
 f ( x + Y ) ≤ f inf { x + y : y ∈ Y } = f x +Y .
~
 f ≤ f .
~
Thus f = f .

Corollary 3: Let {a1 ,..., a m } be a linearly independent set in a normed linear

space X . Then there exists bounded linear funcitonals f1 ,..., f m such that
f j ( ai ) = δ ij , (1 ≤ i, j ≤ m).

Proof: For j = 1,..., m.

Let Y j = span {a i : 1 ≤ i ≤ m, i ≠ j}.

Since Y j is finite dimensional, it is closed in X .

Now ai ∉ Y j (i = 1,..., m, i ≠ j )  by Corollary 2, ∃ g i ∈ X ′ such that g i = 0

on Y j and g i ( a j ) = dist ( a j , Yi ) ≠ 0.

gj
Set f j = . Then f i ∈ X ′ and f j (ai ) = δ ij , 1 ≤ i, j ≤ m.
dist (a j , Yi )

Corollary 4: Let X be a normed linear space and X 0 be a finite dimensional

subspace of X . Then there exists a closed subspace X 1 of X such that

X = X 0 + X 1, X 0 ∩ X1 = { 0 }.

In fact, if {x1 ,..., xn } is a basis of X 0 , and if f1 ,..., f n in X ′ are as in Corollary 3,

then we can take

n
X 1 = I N ( f j ).
j =1

For every x ∈ X , x = y + z, where

n
y =  f j ( x) x j , z = x− y.
j =1

103
Block 1 Normed Linear Spaces
Proof: Define X 1 as given in the theorem. Since each null space is a closed
subspace of X and a finite intersection of closed of subspaces is again a
closed subspace, X 1 is a closed subspace of X .

Now let w ∈ X 0 ∩ X 1

w ∈ X 0  w = α1 x1 + ... + α n xn , α i ∈ K.

w ∈ X 1  f i ( w) = 0, i = 1, ..., n.

Thus we have 0 = f i ( w) = f i (α1 x1 + ... + α n xn ) = α i f i ( x1 ) + ... + α i f i ( xi )

+ ... + α n f i xn = α i (by property of f i ' s)

Thus 0 = αi , i = 1,..., n  w = 0

Now let x ∈ X .

n
Set y =  f ( x) x . Since x ,..., x
i =1
i i 1 n is a basis for X 0 , y ∈ X 0

If we set z = x − y , then

f i ( z) = 0, i = 1,..., n  z ∈ X 1  x = y + z, y ∈ X 0 , z ∈ X 1 .

Thus X = X 0 + X 1 , X 0 I X 1 = {0}.

Hence the claim.

Next we shall state the conditions for uniqueness of Hahn-Banach Extension

theorem.

Now we will address the question whether the extension obtained in the Hahn-
Banach theorem is unique.

The following proposition shows that the uniqueness holds in special cases.
We shall only give the statement of the theorem, the proof is omitted.

The general uniqueness Theorem is a deep result which is beyond the scope
of this course.

Theorem 4(Taylor-Foguel, 1958): Let X be a normed space. For every

subspace Y of X and every g ∈ Y ′, there is a unique Hahn-Banach extension
of g to X if and only if X ′ is strictly convex, that is, for f1 ≠ f 2 in X ′ with
f1 = 1 = f 2 , we have f1 + f 2 < 2 .

There are other special cases where uniqueness of extension exist. For
example you will see later that the uniqueness exist in the following cases.

i) If X is a Hilbert space, then a Hahn-Banach extension of every bounded

linear functional on a subspace is unqiue.

ii) If X 0 is a dense subspace of a normed linear space X , then the Hahn-

104 Banach extension is unique.
Unit 3 Hahn-Banach Theorems
Here are some exercises for you.

E3) Consider R 3 as a subspace of R 4 . Define a linear functional on R 3 .

Obtain its Hahn Banach extension in R 4 .
E4) Let Y be a subspace of a normed linear space X . Prove that an
element a of X lies in Y if and only if f (a) = 0 whenever f is a
bounded linear functional on X and f / y = 0.
With this we come to the end of this unit.

3.4 SUMMARY
In this unit, the following points have been covered.

1. We discussed the following theorems

i) The Hahn-Banach separation theorem

ii) The Hahn-Banach extension thereom

2. We have explained some consequences of Hahn-Banach Extension

theorem as corollaries.

3. We have stated Taylor-Foguel Theorem which gives certain conditions of

the uniqueness of the Hahn-Banach extension.

3.5 SOLUTIONS AND ANSWERS

E1) We have f (ix ) = if ( x ).
We consider elements of the form α1 x + iα 2 y and prove the linearity.

f (α1 x + iα 2 y ) = f (α1 x) + f (iα 2 y )

= f (α1 x ) + if (α 2 y )
= α1 f ( x) + iα 2 f ( y )

Next we shall consider the following:

f (αx + αy ) = f (α1 + iα 2 )x + (α1 + iα 2 ) y

[
= f ( α 1 + iα 2 ) x + α 1 + iα 2 y ]
= f [α1 x + iα 2 x + α1 y + iα 2 y ]
= f [α1 x + α 2 y + iα 2 x + iα 2 y ]
= f [(α1 x + α 2 y ) + i (α 2 x + α 2 y )]
= f (α1 x + f (α 2 y ) + i[ f (α 2 x + α 2 y )]
= f (α1 x) + f (α 2 y ) + if (α 2 x ) + if (α 2 y )
= α1 f ( x ) + α 2 f ( y ) + α 2if ( x ) + α 2if ( y )
= (α1 + iα 2 ) f ( x ) + (α1 + iα 2 ) f ( y )
= αf ( x) + αf ( y ).

This shows that f is complex linear.

105
Block 1 Normed Linear Spaces
E2) i) Since f is complex linear, we write
f ( x) = u ( x ) + iv ( x )
where u (x) and v (x) are real parts of f (x). We have to prove that
u is a real functional. For any real number α 0 , we observe that
u (α 0 x) is the real part of f (α 0 x ). Since f is linear, we get that
u is real linear.

ii) Next we note that

f (ix ) = if ( x) = i (u ( x) + iv ( x)) = −v ( x ) + iu ( x)

Also by definition f (ix ) = u (ix ) + iv (ix )

Comparing the real parts of the two equations, we get that

v ( x ) = −iu (ix ).

iii) We note that for any complex number z = x + iy, we have x ≤ z .

Hence

u( x) ≤ f ( x) ≤ f x.

Thus u ≤ f

To get the otherway inequality, we choose e it such that

e it f ( x) = f ( x) . Also we have

f (e it x ) = e it f ( x)

Since f (e it x ) is real, f (e it x) ≤ u (e it x )

Therefore

e it f ( x ) = f (e it x ) = u (e it x) ≤ u x

i.e. f ( x) ≤ u x for all x ∈ X . This implies that f ≤ u

Hence f = u .

E3) Consider the linear functional on R 3 given by

f ( x1 , x2 , x3 ) = x1 , x = ( x1 , x2 , x3 ) ∈ R 3

Then f is a bounded linear functional on R 3

f ( x) ≤ x .

Also if x = (1, 0, 0), f ( x ) = 1. Hence f = 1

106
Unit 3 Hahn-Banach Theorems
4
Consider g on R such that

g ( x) = x = ( x1 , x 2 , x3 , x4 )
= x1 + ax2 + bx3 + cx4

Then g is an extension of f and g = 1 = f .

E4) Let a ∈ Y . Suppose there is a bounded linear functional such that

f / Y = 0. Then Y ⊂ Z ( f ) . Since Z ( f ) is a closed subspace of X ,
Y ⊂ Z ( f ). Hence f (a ) = 0.

The converse part follows from Corollary 2.

107