CHAPTER 1

VECTOR SPACES
In this chapter we study vector spaces and their subspaces. And we learn a
criterion of the subspaces, sum of subspaces and direct sum of subspaces. Last, we
learn linear transformations from a vector space to another vector space and the
Fundamental Theorem of Linear Maps.
1.1 Vector Spaces
In this section we recall some definitions and results on vector spaces. Let
and be the sets of all real and of all complex numbers. Let be or , called a
field. Elements of are called scalars. The set of all ordered pairs of real numbers:
2
  x, y  : x, y  .
2
The set is also called a plane. The set of all ordered triples of real numbers:
3
  x, y, z  : x, y, z  .
3
The set is an ordinary space.
To generalize 2 and 3 to higher dimensions, we first need to discuss the
concept of lists.
1.1.1 Definition Suppose n is a nonnegative integer. A list of length n is an ordered
collection of n elements separated by commas and surrounded by parentheses. A list
of length n is
( x1 , , xn ).
Two lists are equal if and only if they have the same length and the same elements in
the same order.
Many mathematicians call a list of length n an n-tuple. ( x1 , x2 , ) is said to
have infinite length, is not a list.
n
1.1.2 Definition is the set of all lists of length n of elements of :
n
  x1 , , xn  : x j  for j  1, , n.
For  x1 , , xn   n
and j  1, , n , we say that x j is the jth coordinate of
 x1 , , xn  .
Every element x in n
is called a vector. If  0, , 0  n
, then it is called a
zero vector; denoted by 0.
n
1.1.3 Definition Addition in is defined by adding corresponding coordinates:
 x1 , , xn    y1 , , yn    x1  y1 , , xn  yn  .
1.1.4 Definition The product of a number  and a vector in n
is computed by
multiplying each coordinate of the vector by  :
  x1 , , xn     x1 , ,  xn 
where   and  x1 , , xn   n
.
 2

1.1.5 Definition An addition on a set V is a function from V V into V defined by

 x, y  x  y for all x, y V .
A scalar multiplication on a set V is a function from V into V defined by
  , x   x for each   and each x V .
1.1.6 Definition A vector space is a set V along with an addition on V and a scalar
multiplication on V such that the following properties hold:
(i)  u  v   w  u   v  w  for all u,v,w V (associativity);

(ii) there exists an element 0 V such that

v  0  v  0  v for all v V (additive identity);
(iii) for every v V , there exists w V such that
v  w  0  w v (additive inverse);
(iv) u  v  v  u for all u,v V (commutativity);
(v) ( ) w   (  w) for all w V and for all scalar  ,  (associativity);
(vi)  (u  v)  u  v and (   )u  u   u for all  ,   and all u, v V
(distributive properties);
(vii) 1v  v for all v V (multiplicative identity).
Elements of a vector space are called vectors or points.
The scalar multiplication in a vector space depends on . We write V is a
vector space instead of V is a vector space over .
For example, n is a vector space over , and n is a vector space over .
A vector space over is called a real vector space and a vector space over
is called a complex vector space.
1.1.7 Example If S is a set, then s denotes the set of functions from S to . For
f , g  s , the sum f  g  s is the function defined by

f  g  x   f  x   g  x 
for all x  S. For   and f  s
, the product  f  s
is the function defined by
  f  x    f  x 
for all x  S. If S is a nonempty set then s is a vector space over . Indeed, the
additive identity of s is the function 0 : S  defined by
0 x  0
for all x  S. For f  s
, the additive inverse of f is the function  f : S  defined
by
  f  x    f  x 
for all x  S. Clearly, associative and commutative properties under addition are
satisfied by definition of addition. Also, associative, distributive properties and
multiplicative identity under scalar multiplication are satisfied by definition of scalar
multiplication.
 3

1.2 Subspaces
We have seen that every set has subsets. In the same way, every vector space
has subspaces. In this section, we study some results on subspaces of vector spaces.
These results are: The sum of n subspaces of a vector space is the smallest vector
space containing given n subspaces, direct sum of n subspaces and a test whether the
sum of n subspaces is a direct sum.
1.2.1 Definition A subset U of V is called a subspace of V if U is also a vector space
under the same addition and scalar multiplication on V.
For example,  0, x , x  : x , x
2 3 2 3   is a subspace of 3
.
The next result gives the easiest way to check whether a subset of a vector
space is a subspace.
1.2.2 Theorem A subset U of V is a subspace of V if and only if U satisfies the
following three conditions:
(i) 0 U (additive identity);
(ii) u, w U implies u  w U (closed under addition);
(iii)   and u U implies  u U (closed under scalar multiplication).
Proof. If U is a subspace of V, then U satisfies the three conditions above by the
definition of vector space.
Conversely, suppose U satisfies the three conditions above. Associativity and
commutativity under addition are satisfied since every element in U is in V. The
additive identity element exists by the first condition. u U , then u is in U by the
third condition above. Hence every element of U has an additive inverse in U.
The axioms under scalar multiplication are satisfied since every element in U
is in V. 
1.2.3 Example The set W of vectors of the form (x,0) where x  is a subspace of
2
because W is a subset of 2 whose vectors are of the form (x, y) where x, y  .
The zero vector (0,0) is in W. For any  x1 , 0 ,  x2 , 0 W ,

 x1 , 0   x2 , 0   x1  x2 , 0
is in W. For any   and  x,0 W ,

  x,0   x,0
is in W.
1.2.4 Example The set W of vectors of the form (x, y) such that x  0 and y  0 is
not a subspace of 2 because it is not closed under scalar multiplication, for instance,
 2, 2 W but not 1 2, 2   2, 2 W .
Now we define the sum of subsets of a vector space V.
 4

1.2.5 Definition Suppose U1 , ,U n are subsets of V. The sum of U1 , ,U n , denoted

U1   U n , is the set of all possible sums of elements of U1 , , U n . More precisely,
U1   U n  u1   un : u1 U1 , , un  U n  .

1.2.6 Example If U   x, 0, 0   3
 and W   0, y, 0  
: x 3
: y  then
U  W   x, y, 0   : x, y   . 3

The next result states that the sum of subspaces is a subspace.

1.2.7 Theorem Suppose U1 , ,U n are subspaces of V. Then U1   U n is the
smallest subspace of V containing U1 , , U n .

Proof. Since each U i 1  i  n is a subspace of V, zero vector, 0 is in each U i . Thus

0 U1   U n . For each ui U i (1  i  n), u1   un , u1   un  U1  Un
and for   ,

u1   un  u1  
 un  u1  u1    
 un  un  U1  Un

and
  u1   un    u1    un U1   Un.

Thus U1   U n is a subspace of V.

Each U i 1  i  n  is contained in U1   U n because for each ui  U i we

write
ui  0   ui  0   0.

Conversely, every subspace of V containing U1 , ,U n contains U1   U n because

subspaces must contain all finite sums of their elements. Thus U1   U n is the
smallest subspace of V containing U1 , ,U n . 

Suppose U1 , ,U n are subspaces of V. Every element of V can be written in

the form
u1   un ,

where each u j is in U j . We will be interested in a case where each vector in

U1   U n can be represented in the form above only one way, called direct sum.

1.2.8 Definition Suppose U1 , ,U n are subspaces of V. The sum U1   U n is

called direct sum if each element of U1   U n can be written in only one way as a
 5

sum u1   un , where each u j is in U j . Denote the direct sum of U1 , ,U n by

U1  Un.

1.2.9 Example Let U   x, y, 0   3

: x, y   and W   0, 0, z   3
:z .
Clearly, U and W are subspaces of 3
. Then 3
 U  W for any  x, y, z   3
we
write
 x, y, z   x, y,0   0,0, z 
so that 3
 U  W . To show uniqueness suppose
 x, y, z    x1 , y1 , 0   0, 0, z1    x2 , y2 , 0   0, 0, z2 
where each component is in . Then  x1 , y1 , z1    x2 , y2 , z2  so that x1  x2 , y1  y2
and z1  z2 . Hence 3
 U W.
The next result shows that when deciding whether a sum of subspaces is a
direct sum, we need only consider whether 0 can be uniquely written as an
appropriate sum.
1.2.10 Theorem Suppose U1 , ,U n are subspaces of V. Then U1   U n is a direct
sum if and only if the only way to write 0 as a sum u1   un , where each u j is in
U j , is by taking each u j equal to 0.

Proof. Suppose U1   U n is a direct sum. Then the definition of direct sum implies
that the only way to write 0 as a sum u1   un , where each u j is in U j , is by
taking each u j equal to 0.

Now suppose that the only way to write 0 as a sum u1   un , where each
u j is in U j , is by taking each u j equal to 0. To show that U1   U n is a direct
sum, let v U1   U n . We can write
v  u1   un

for some u1 U1 , , un U n . To show that this representation is unique, suppose we

also have
v  v1   vn ,

where v1 U1 , , vn U n . Subtracting these two equations above gives

0   u1  v1     un  vn  .

Since u1  v1 U1 , , un  vn U n , the equation above implies that each u j  v j equals

0. Thus u1  v1 , , un  vn . 
 6

The next result gives a simple condition for testing which pairs of subspaces
gives a direct sum.
1.2.11 Theorem Suppose U and W are subspaces of V. Then U  W is a direct sum if
and only if U W  0 .

Proof. Suppose that U  W is a direct sum. If v U W , then 0  v   v , where

v U and v W . By the unique representation of 0 as the sum of a vector in U and
a vector in W, we have v  0. Thus U W  0 .

Conversely, suppose that U W  0 . To prove that U  W is a direct sum,

suppose u U and w W , and
0  u  w.
To complete the proof, we must show that u  0 and w  0 by Theorem 1.2.10. The
equation above implies that u   w W . Thus u U W . Hence u  0 and w  0
by the equation above. The proof is complete. 
1.3 Finite Dimensional Vector Spaces
We have been studied vector spaces and their subspaces. In this section, we
study finite dimensional vector spaces.
1.3.1 Definition A linear combination of a list v1 , , vn of vectors in V is a vector of
the form
1v1    n vn

where 1 , ,n  .

The set of all linear combinations of a list of vectors v1 , , vn in V is called the span
of v1 , , vn , denoted span  v1 , , vn  . In other words,

span  v1 , , vn   1v1    n vn : 1 , ,n  .

If span  v1 , , vn  equals V , we say that v1 , , vn spans V .
1.3.2 Definition A vector space is called finite dimensional if some list of vectors in
it spans the space.
The dimension of a finite dimension vector space, denoted dim 𝑉, is the
length of any basis of the vector space.
A vector space is called infinite dimensional if it is not finite dimensional.
1.3.3 Definition A function p :  is called a polynomial with coefficients in
if there exist  0 , ,  n  such that

p( z)  0  1 z   2 z 2   n z n (1.1)

for all z  . P   is the set of all polynomials with coefficients in .

 7

A polynomial p  P   is said to have degree n if there exist scalars

0 , ,  n  with  n  0 such that (1.1) holds for all z  .

For n a nonnegative integer, Pn   denotes the set of all polynomials with

coefficients in and degree at most n.
1.3.4 Definition A list v1 , , vn of vectors in V is called linearly independent if

1v1    n vn  0  1   n  0

where 1 , ,  n  . A list of vectors in V is called linearly dependent if it is not

linearly independent.
1.3.5 Definition A basis of V is a list of vectors in V that is linearly independent and
spans V .

For example, the list 1, 0, , 0  ,  0,1, 0, , 0 , ,  0, , 0,1 is a basis of n

,
n n
called the standard basis of and dim is n.
The next result is criterion for basis.
1.3.6 Theorem A list v1 , , vn of vectors in V is a basis of V if and only if every
v V can be written uniquely in the form
v  1v1    n vn , (1.2)

where 1 , ,n  .

Proof. Suppose that v1 , , vn is a basis of V. Let v V . Since v1 , , vn spans V, there

exists 1 , ,  n  such that (1.2) holds. To show that the representation in (1.2) is
unique, suppose 1 , ,  n  are such that

v  1v1    n vn .
Subtracting the last equation from (1.2), we get
0  1  1  v1    n   n  vn .

Then
 i  i  0 1  i  n 
since v1 , , vn are linearly independent.
Conversely, suppose that every v V can be written uniquely in the form
given by (1.2). Clearly, this implies that v1 , , vn spans V. To show that v1 , , vn is
linearly independent, suppose
0  1v1    n vn

where  i  1  i  n  . By the uniqueness of the representation and (1.2),

 8

0  v1   0  vn  0  1v1    n vn

so that 1    n  0. Thus v1 , , vn is linearly independent and hence is a basis of

V. 
We now consider obtaining a basis from a spanning list of a vector space.
1, 2  ,  3, 6  ,  4, 7  ,  5,9  is a spanning list in 2 . For instance, 8,14   2 we write
as follows
8,14    1, 2   0  3, 6    4, 7    5,9  .
The given spanning list is not a basis because it is not linearly independent. But we
may reduce a basis of 2 . Since
1
1, 2  3,6  0  4,7   0 5,9  ,
3
we removed 1, 2  from the spanning list. Next, we removed  3, 6  because

 3, 6   3  4, 7   3  5,9  .
Now, we show that  4, 7  ,  5,9  is a basis of 2
. To show  4, 7  and  5, 9  are
linearly independent suppose
  4, 7     5,9    0, 0 

where  ,   . Then     0. Thus  4, 7  ,  5,9  are linearly independent.

span   4, 7  ,  5,9    2
because for  x, y   2
,

 x, y    9 x  5 y  4, 7    4 y  7 x  5,9  .
Hence  4, 7  ,  5,9  is a basis of 2
. This motivates to get the following result.
1.3.7 Theorem Every spanning list in a vector space can be reduced to a basis of the
vector space.
Proof. Suppose v1 , , vn spans V. We want to remove some of the vectors from
v1 , , vn so that the remaining vectors form a basis of V. To do this let B  v1 , , vn .

Step 1. If v1  0, delete v1 from B. If v1  0, leave B unchanged.

Step j. If v j is in span  v1 , , v j 1  , delete v j from B.

If v j is not in span  v1 , , v j 1  , leave B unchanged.
Stop the process after step n, getting a list B. This list B spans V because our original
list spanned V and we have discarded only vectors that were already in the span of the
previous vectors. The process ensures that no vector in B is in the span of the previous
ones. Thus, B is a linearly independent. Hence B is a basis of B. 
 9

1.3.8 Corollary Every finite dimensional vector space has a basis.

Proof. By definition, a finite dimensional vector space has a spanning list. By
Theorem 1.3.7, each spanning list can be reduced to a basis. 
1.3.9 Theorem Every linearly independent list of vectors in a finite dimensional
vector space can be extended to a basis of the vector space.
Proof. Suppose u1 , , un is linearly independent in a finite dimensional vector space
V. Let w1 , , wm be a basis of V.
Thus
u1 , , un , w1 , , wm
spans V. By applying the procedure of the proof of Theorem 1.3.7, we get a basis
containing of the vectors u1 , , un and some of the w’s. 

1.3.10 Example The list  2,3, 4 , 1,0,1 is linearly independent in 3

.

 2,3, 4 , 1,0,1 , 1,0,0 , 0,1,0 , 0,0,1

Let be spanning list of 3
. By
Theorem 1.3.7,  2,3, 4 , 1,0,1 ,  0,0,1 is a basis of 3
.
1.4 Linear Transformations on Vector Spaces
We have been studied vector spaces and their subspaces. In this section, we
study a function from a vector space V into another vector space W, called a linear
map. We assume that V and W are vector spaces over .
1.4.1 Definition A linear map from V to W is a function T : V  W with the
following properties:
(i) T  u  v   Tu  Tv for all u, v V (additivity);

(ii) T  v  Tv for all   and all v V (homogeneity).

1.4.2 Example Define T : 3

 2
by
T  x, y, z    2 x,6 y  z  .
Then T is linear map because
T  x, y, z    x1 , y1 , z1   T  x  x1 , y  y1 , z  z1   T  x, y, z   T  x1 , y1 , z1 

and
T    x, y, z    T   x,  y,  z    2 x,6 y   z     2 x,6 y  z   T  x, y, z 

where  x1 , y1 , z1   3
and   .

1.4.3 Example Define f : 2

 by

f  x, y   x 2 y
 10

for x, y  . Then f is not a linear map because

f 1,1  f  0,1  1  0  1
and
f  1,1   0,1   f 1, 2   2.
1.4.4 Example Let L V ,W  be the set of all linear maps from V to W. Suppose
S , T  L V ,W  and   . The sum S  T and the product T are the linear maps
from V to W defined by
( S  T )v  Sv  Tv and (T )v   (Tv)
for all v V . Then L V ,W  is a vector space under the given operations addition and
scalar multiplication. For S , T  L V ,W  ,

S  T  L V ,W 
by given addition.
(i) Since   S  T   T  v   S  T  v  T v   S  T  T   v
1 1 1 ( v V ),

 S  T   T1  S  T  T1 
where S, T and T1 are in L V ,W  .

(ii) Define 0 :V  W by 0  v  0 for all v V . Then the zero linear map,

0  L V ,W  . For any T  L V ,W  and v V ,

T  0  v  Tv  0(v)  Tv  0  Tv  0(v)  Tv   0  T  v.
Hence the zero linear map, 0 is the additive identity in L V ,W  .

(iii) Define T : V  W by  T  v  Tv for all v V . Then T is a linear

transformation so that T  L V ,W  . For any T  L V ,W  , and for all v V ,

T   T   v  Tv   Tv   0  v     T   T  v
so that
T   T   T  T  0.

Hence T is an additive inverse of T in L V ,W  .

(iv) For T  L V ,W  , and for all v V we have

 S  T  v  Sv  Tv  Tv  Sv  T  S  v
so that
S  T  T  S.
 11

For any T  L V ,W  , T  L V ,W  by given scalar multiplication.

(v) For T  L V ,W  , and for  ,   and v V , we have

   T  v    Tv     Tv       T  v      T   v,
so that
  T     T  .
(vi) For S , T  L V ,W  , u, v V and  ,   , we have

( (T  S ))(v)  ( S  T )(v) and      S   v    S   S  v  .

Thus  (T  S )   S  T and     S   S   S .

(vii) For S  L V ,W  , v V and 1  , we have

1S  v   1Sv  Sv.

Hence 1S  S. Hence, L V ,W  is a vector space under the given operations.

Note that L V  is the set of all linear maps from V to V . The members in
L V  are called operators.

1.4.5 Definition If T  L V ,W  and S  L V ,W  , then the product ST  L V ,W  is

defined by
 ST  u  S Tu
for u U .
In other words, ST is just the usual composition S T of two functions, but
when both functions are linear.
1.4.6 Example Define T  L  P   , P   by

Tp x  x2 p  x
and define D  L  P   , P   by

Dp  p.
Then DT  TD because
 TD  p   x   x p  x 
2
but   DT  p   x   x p  x   2 xp  x  .
2

1.4.7 Proposition Suppose T is a linear map V to W. Then T  0  0.

Proof. By additivity, we have
T  0  T  0  0  T  0  T  0 .
 12

By adding T  0 in both side of the equation above,

0  T  0 . 

1.4.8 Definition For T  L V ,W  , the null space of T, denoted null T, is the subset
of 𝑉 consisting of these vectors that 𝑇 maps to 0:
null T  v  V : Tv  0.

1.4.9 Example (a) If T is the zero map from V to W,

null T   v V : 0  v   0  V

(b) Suppose D  L  P   , P   is the differentiation map defined by Dp  p.

Then
null D   p  P  : Dp  0   p  P  : p is a constant function.

1.4.10 Theorem Suppose T  L V ,W  . Then null 𝑇 is a subspace of V.

Proof. Suppose T is a linear map. Then, T  0  0 by Proposition 1.4.7. Thus

0  null T . Suppose u, v  null T . Then

T  u  v  Tu  Tv  0  0  0
so that u  v  null T . Suppose   and u  null T . Then
T  u   Tu  0.

Hence u  null T . Thus null T is a subspace of V. 

1.4.11 Definition A function T : V  W is called injective if Tu  Tv implies u  v.
1.4.12 Theorem Let T  L V ,W  . Then T is injective if and only if null T  0 .
Proof. Suppose 𝑇 is injective. Then
null T   v V : Tv  0  v V : Tv  T  0   v V : v  0  0 .

Conversely, suppose that null T  0 . To show T is injective suppose

Tu  Tv for u, v V . Then

Tu  Tv  0 ; T  u  v  0 ; u  v  0 ; u  v.

Hence T is injective. 
1.4.13 Definition For T is a function from V to W, the range of T is the subset of W
consisting of those vectors that are of the form Tv for some v V :
range T  Tv : v V  .
1.4.14 Example (a) If T is the zero map from V to W,
 13

range T  0  v  : v V   0 .

(b) Suppose D  L  P   , P   is a differentiation map defined by Dp  p '. Then

range D  Dp : p  P    P   .
1.4.15 Theorem If T  L V ,W  , then range T is a subspace of W.

Proof. Suppose T  L V ,W  . Then T  0  0 by Proposition 1.4.7. Thus 0  range T .

For any w1 , w2  range T , there exist v1 , v2 V such that Tv1  w1 and Tv2  w2 . Then

w1  w2  Tv1  Tv2  T  v1  v2 

is in range T. For any   and w1  range T ,

 w1  Tv1  T   v1 

is in range T. Hence range T is a subspace of W. 

1.4.16 Definition A function T : V  W is called surjective if its range equals W.
1.4.17 Example The differentiation map D  L  P4   , P4    defined by Dp  p
is not surjective, because the polynomial x 4 is not in the range of D. However, the
differentiation map S  L  P3   , P2    defined by Sp  p is surjective, because its
range equals P2  .
The next result is an important result in linear algebra. It is called the
Fundamental Theorem of Linear Maps.
1.4.18 Theorem Suppose V is a finite dimensional and T  L V ,W  . Then range T
is finite dimensional and
dim V  dim null T  dim range T .
Proof. Let u1 , , um be a basis of null T. Thus dim null T  m. The linearly
independent list u1 , , um can be extended to a basis

u1 , , um , v1 , , vn
of V by Theorem 1.3.9. Thus dim V  m  n. To complete the proof, we must show
that the range T is finite dimensional and dim range T  n. We will do this by
proving that Tv1 , , Tvn is a basis of range T.

Let v V . Since u1 , , um , v1 , , vn span V , we can write

v  1u1    mum  1v1    n vn ,
 14

where the terms of the form Tu j disappeared because each u j is in null T. The last
equation implies that Tv1 , , Tvn spans range T. In particular, range T is finite
dimensional.
To show Tv1 , , Tvn is linearly independent, suppose  1 , , n  and

 1Tv1    nTvn  0.
Then
T   1v1    n vn   0.

Hence
 1v1    n vn  null T .

Since u1 , , um spans null T, we can write

 1v1    n vn  1u1    m um ,

where the each  i is in . This equation implies that all the  i ’s and  i ’s are 0
because u1 , , um , v1 , , vn is linearly independent. Thus Tv1 , , Tvn is linearly
independent and hence is a basis of range T. 
Now we show that no linear map from a finite dimensional vector space to a
“smaller” vector space can be injective.
1.4.19 Theorem Suppose V and W are finite dimensional vector spaces such that
dim V > dim W . Then no linear map from V to W is injective.
Proof. Let T  L V ,W  . Then by Theorem 1.4.18,
dim null T  dim V  dim range T
 dim V  dim W
 0.
Since dim null T  0, null T contains vectors other than 0. Thus, T is not injective by
Theorem 1.4.12. 
The next result shows that no linear map from a finite dimensional vector
space to a “bigger” vector space can be surjective.
1.4.20 Theorem Suppose V and W are finite dimensional vector spaces such that
dim V  dim W . Then no linear map from V to W is surjective.
Proof. Let T  L V ,W  . Then by Theorem 1.4.18,
dim range T  dim V  dim null T
 dim V
< dim W .
 15

Since dim range T  dim W , the range T cannot equal W. Thus, T is not surjective. 