CHAPTER 4

General Vector Spaces

CHAPTER CONTENTS

4.1 Real Vector Spaces 202

4.2 Subspaces 211
4.3 Spanning Sets 220
4.4 Linear Independence 228
4.5 Coordinates and Basis 238
4.6 Dimension 248
4.7 Change of Basis 256
4.8 Row Space, Column Space, and Null Space 263
4.9 Rank, Nullity, and the Fundamental Matrix Spaces 276

Introduction
Recall that we began our study of vectors by viewing them as directed line segments
(arrows). We then extended this idea by introducing rectangular coordinate systems, and
that enabled us to view vectors as ordered pairs and ordered triples of real numbers. As we
developed properties of these vectors we noticed patterns in various formulas that enabled
us to extend the notion of a vector to an n-tuple of real numbers. Although n-tuples took
us outside the realm of our “visual experience,” it gave us a valuable tool for understand-
ing and studying systems of linear equations. In this chapter we will extend the concept
of a vector yet again by using the most important algebraic properties of vectors in 𝑅n as
axioms. These axioms, if satisfied by a set of objects, will enable us to think of those objects
as vectors.

4.1 Real Vector Spaces

In this section we will extend the concept of a vector by using the basic properties of
vectors in 𝑅n as axioms, which if satisfied by a set of objects will guarantee that those
objects behave like familiar vectors.

Vector Space Axioms

The following definition consists of ten axioms, eight of which are properties of vectors
in 𝑅n that were stated in Theorem 3.1.1. It is important to keep in mind that one does
202
 4.1 Real Vector Spaces 203

not prove axioms; rather, they are assumptions that serve as the starting point for proving
theorems.

Definition 1

Let 𝑉 be an arbitrary nonempty set of objects for which two operations are defined: In this text scalars will
addition and multiplication by numbers called scalars. By addition we mean a be either real numbers or
rule for associating with each pair of objects u and v in 𝑉 an object u + v, called complex numbers. Vector
the sum of u and v; by scalar multiplication we mean a rule for associating with spaces with real scalars will
each scalar k and each object u in 𝑉 an object ku, called the scalar multiple of u be called real vector spaces
by k. If the following axioms are satisfied by all objects u, v, w in 𝑉 and all scalars and those with complex
k and m, then we call 𝑉 a vector space and we call the objects in 𝑉 vectors. scalars will be called com-
plex vector spaces. For now
1. If u and v are objects in 𝑉, then u + v is in 𝑉.
we will consider only real
2. u + v = v + u vector spaces.
3. u + (v + w) = (u + v) + w
4. There exists an object in 𝑉, called the zero vector, that is denoted by 0 and
has the property that 0 + u = u + 0 = u for all u in 𝑉.
5. For each u in 𝑉, there is an object −u in 𝑉, called a negative of u, such that
u + (−u) = (−u) + u = 0.
6. If k is any scalar and u is any object in 𝑉, then ku is in 𝑉.
7. k(u + v) = ku + kv
8. (k + m)u = ku + mu
9. k(mu) = (km)(u)
10. 1u = u

Observe that the definition of a vector space does not specify the nature of the vectors
or the operations. Any kind of object can be a vector, and the operations of addition and
scalar multiplication need not have any relationship to those on 𝑅n . The only requirement
is that the ten vector space axioms be satisfied. In the examples that follow we will use four
basic steps to show that a set with two operations is a vector space.

Steps to Show That a Set with Two Operations Is a Vector Space

Step 1. Identify the set 𝑉 of objects that will become vectors.
Step 2. Identify the addition and scalar multiplication operations on 𝑉.
Step 3. Verify Axioms 1 and 6; that is, adding two vectors in 𝑉 produces a vector in 𝑉,
and multiplying a vector in 𝑉 by a scalar also produces a vector in 𝑉.
Axiom 1 is called closure under addition, and Axiom 6 is called closure under
scalar multiplication.
Step 4. Confirm that Axioms 2, 3, 4, 5, 7, 8, 9, and 10 hold.

Our first example is the simplest of all vector spaces in that it contains only one object.
Since Axiom 4 requires that every vector space contain a zero vector, the object will have
to be that vector.
204 C H APT ER 4 General Vector Spaces

EXAMPLE 1 | The Zero Vector Space

Let 𝑉 consist of a single object, which we denote by 0, and define

0+0=0 and k0 = 0
for all scalars k. It is easy to check that all the vector space axioms are satisfied. We call this
the zero vector space.

Our second example is one of the most important of all vector spaces—the familiar
n n
space 𝑅 . It should not be surprising that the operations on 𝑅 satisfy the vector space
n
axioms because those axioms were based on known properties of operations on 𝑅 .

n
EXAMPLE 2 | R Is a Vector Space

n
Let 𝑉 = 𝑅 , and define the vector space operations on 𝑉 to be the usual operations of addi-
tion and scalar multiplication of n-tuples; that is,
u + v = (u1 , u2 , . . . , un ) + (𝑣1 , 𝑣2 , . . . , 𝑣n ) = (u1 + 𝑣1 , u2 + 𝑣2 , . . . , un + 𝑣n )
ku = (ku1 , ku2 , . . . , kun )
n
The set 𝑉 = 𝑅 is closed under addition and scalar multiplication because the foregoing
operations produce n-tuples as their end result, and these operations satisfy Axioms 2, 3, 4,
5, 7, 8, 9, and 10 by virtue of Theorem 3.1.1.

Historical Note
The notion of an “abstract vector space” evolved over many
years and had many contributors. The idea crystallized with the
work of the German mathematician H. G. Grassmann, who pub-
lished a paper in 1862 in which he considered abstract systems of
unspecified elements on which he defined formal operations of
addition and scalar multiplication. Grassmann’s work was con-
troversial, and others, including Augustin Cauchy (p. 136), laid
reasonable claim to the idea.
[Image: © Sueddeutsche Zeitung Photo/The Image Works]

Hermann Günther
Grassmann
(1809–1877)

n
Our next example is a generalization of 𝑅 in which we allow vectors to have infinitely
many components.
 4.1 Real Vector Spaces 205

EXAMPLE 3 | The Vector Space of Infinite Sequences of

Real Numbers

Let 𝑉 consist of objects of the form

u = (u1 , u2 , . . . , un , . . . )
in which u1 , u2 , . . . , un , . . . is an infinite sequence of real numbers. We define two infinite
sequences to be equal if their corresponding components are equal, and we define addition
and scalar multiplication componentwise by
u + v = (u1 , u2 , . . . , un , . . . ) + (v1 , v2 , . . . , vn , . . . )
= (u1 + v1 , u2 + v2 , . . . , un + vn , . . . )
ku = (ku1 , ku2 , . . . , kun , . . . )
In the exercises we ask you to confirm that 𝑉 with these operations is a vector space. We will
denote this vector space by the symbol 𝑅∞ .

E(t)
Vector spaces of the type in Example 3 arise when a transmitted signal of indefinite Voltage
duration is digitized by sampling its values at discrete time intervals (Figure 4.1.1). 1
In the next example our vectors will be matrices. This may be a little confusing at t

first because matrices are composed of rows and columns, which are themselves vectors Time
–1
(row vectors and column vectors). However, from the vector space viewpoint we are not
concerned with the individual rows and columns but rather with the properties of the
matrix operations as they relate to the matrix as a whole. FIGURE 4.1.1 This transmitted
signal continues indefinitely.

EXAMPLE 4 | The Vector Space of 2 × 2 Matrices

Let 𝑉 be the set of 2 × 2 matrices with real entries, and take the vector space operations on
𝑉 to be the usual operations of matrix addition and scalar multiplication; that is,
u11 u12 v11 v12 u11 + v11 u12 + v12 Note that Equation (1)
u+v=[ ]+[ ]=[ ] (1)
u21 u22 v21 v22 u21 + v21 u22 + v22 involves three different addi-
tion operations: the addition
u11 u12 ku ku12
ku = k [ ] = [ 11 ] operation on vectors, the
u21 u22 ku21 ku22
addition operation on
The set 𝑉 is closed under addition and scalar multiplication because the foregoing operations matrices, and the addition
produce 2 × 2 matrices as the end result. Thus, it remains to confirm that Axioms 2, 3, 4, 5, operation on real numbers.
7, 8, 9, and 10 hold. Some of these are standard properties of matrix operations. For example,
Axiom 2 follows from Theorem 1.4.1(a) since
u11 u12 v11 v12 v11 v12 u11 u12
u+v=[ ]+[ ]=[ ]+[ ]=v+u
u21 u22 v21 v22 v21 v22 u21 u22
Similarly, Axioms 3, 7, 8, and 9 follow from parts (b), (h), ( j), and (e), respectively, of that
theorem (verify). This leaves Axioms 4, 5, and 10 that remain to be verified.
To confirm that Axiom 4 is satisfied, we must find a 2 × 2 matrix 0 in 𝑉 for which
u + 0 = 0 + u for all 2 × 2 matrices in 𝑉. We can do this by taking
0 0
0=[ ]
0 0
With this definition,
0 0 u11 u12 u11 u12
0+u=[ ]+[ ]=[ ]=u
0 0 u21 u22 u21 u22
206 C H APT ER 4 General Vector Spaces

and similarly u + 0 = u. To verify that Axiom 5 holds we must show that each object u in
𝑉 has a negative −u in 𝑉 such that u + (−u) = 0 and (−u) + u = 0. This can be done by
defining the negative of u to be
−u11 −u12
−u = [ ]
−u21 −u22
With this definition,
u11 u12 −u11 −u12 0 0
u + (−u) = [ ]+[ ]=[ ]=0
u21 u22 −u21 −u22 0 0
and similarly (−u) + u = 0. Finally, Axiom 10 holds because
u11 u12 u u12
1u = 1 [ ] = [ 11 ]=u
u21 u22 u21 u22

EXAMPLE 5 | The Vector Space of m × n Matrices

Example 4 is a special case of a more general class of vector spaces. You should have no trou-
ble adapting the argument used in that example to show that the set 𝑉 of all m × n matri-
ces with the usual matrix operations of addition and scalar multiplication is a vector space.
We will denote this vector space by the symbol 𝑀mn . Thus, for example, the vector space in
Example 4 is denoted as 𝑀22 .

EXAMPLE 6 | The Vector Space of Real-Valued Functions

Let 𝑉 be the set of real-valued functions that are defined at each x in the interval (−∞, ∞). If
f = 𝑓(x) and g = g(x) are two functions in 𝑉 and if k is any scalar, then define the operations
of addition and scalar multiplication by
(f + g)(x) = 𝑓(x) + g(x) (2)
(kf)(x) = k𝑓(x) (3)
One way to think about these operations is to view the numbers 𝑓(x) and g(x) as “compo-
nents” of f and g at the point x, in which case Equations (2) and (3) state that two functions
are added by adding corresponding components, and a function is multiplied by a scalar by
multiplying each component by that scalar—exactly as in 𝑅n and 𝑅∞ . This idea is illustrated
In Example 6 the functions in parts (a) and (b) of Figure 4.1.2. The set 𝑉 with these operations is denoted by the symbol
𝐹(−∞, ∞). We can prove that this is a vector space as follows:
are defined on the entire
interval (−∞, ∞). However, Axioms 1 and 6: These closure axioms require that if we add two functions that are defined
the arguments used in that at each x in the interval (−∞, ∞), then sums and scalar multiples of those functions must
also be defined at each x in the interval (−∞, ∞). This follows from Formulas (2) and (3).
example apply as well on all
subintervals of (−∞, ∞), Axiom 4: This axiom requires that there exists a function 0 in 𝐹 (−∞, ∞), which when
such as a closed interval added to any other function f in 𝐹(−∞, ∞) produces f back again as the result. The function
whose value at every point x in the interval (−∞, ∞) is zero has this property. Geometrically,
[a, b] or an open interval
the graph of the function 0 is the line that coincides with the x-axis.
(a, b). We will denote the
vector spaces of functions Axiom 5: This axiom requires that for each function f in 𝐹(−∞, ∞) there exists a function
−f in 𝐹(−∞, ∞), which when added to f produces the function 0. The function defined by
on these intervals by F[a, b]
−f(x) = −𝑓(x) has this property. The graph of −f can be obtained by reflecting the graph of
and 𝐹(a, b), respectively. f about the x-axis (Figure 4.1.2c).
 4.1 Real Vector Spaces 207

Axioms 2, 3, 7, 8, 9, 10: The validity of each of these axioms follows from properties of real
numbers. For example, if f and g are functions in 𝐹(−∞, ∞), then Axiom 2 requires that
f + g = g + f. This follows from the computation
(f + g)(x) = 𝑓(x) + g(x) = g(x) + 𝑓(x) = (g + f)(x)
in which the first and last equalities follow from (2), and the middle equality is a property of
real numbers. We will leave the proofs of the remaining parts as exercises.

y y y
f+g
g

f(x) + g(x) f
g(x) kf kf (x) f(x)
x
f f(x) f f(x) 0
x x –f –f(x)
x x

(a) (b) (c)

FIGURE 4.1.2

It is important to recognize that you cannot impose any two operations on any set 𝑉
and expect the vector space axioms to hold. For example, if 𝑉 is the set of n-tuples with
positive components, and if the standard operations from 𝑅n are used, then 𝑉 is not closed
under scalar multiplication because if u is a nonzero n-tuple in 𝑉, then (−1)u has at least
one negative component and hence is not in 𝑉. The following is a less obvious example in
which only one of the ten vector space axioms fails to hold.

EXAMPLE 7 | A Set That Is Not a Vector Space

Let 𝑉 = 𝑅2 and define addition and scalar multiplication operations as follows: If u = (u1 , u2 )
and v = (v1 , v2 ), then define
u + v = (u1 + v1 , u2 + v2 )
and if k is any real number, then define
ku = (ku1 , 0)
For example, if u = (2, 4), v = (−3, 5), and k = 7, then
u + v = (−1, 9)
ku = 7u = (14, 0)
The addition operation is the standard one from 𝑅2 , but the scalar multiplication is not. In
the exercises we will ask you to show that the first nine vector space axioms are satisfied, but
Axiom 10 fails to hold for certain vectors. For example, if u = (u1 , u2 ) is such that u2 ≠ 0,
then
1u = 1(u1 , u2 ) = (u1 , 0) ≠ u
Thus, 𝑉 is not a vector space with the stated operations.

Our final example will be an unusual vector space that we have included to illustrate
how varied vector spaces can be. Since the vectors in this space will be real numbers, it will
be important for you to keep track of which operations are intended as vector operations
and which ones as ordinary operations on real numbers.
208 C H APT ER 4 General Vector Spaces

EXAMPLE 8 | An Unusual Vector Space

Let 𝑉 be the set of positive real numbers, let u = u and v = v be any vectors (i.e., positive
real numbers) in 𝑉, and let k be any scalar. Define the operations on 𝑉 to be
u + v = uv [Vector addition is numerical multiplication.]
ku = uk [Scalar multiplication is numerical exponentiation.]

Thus, for example, 1 + 1 = 1 and (2)(1) = 12 = 1—strange indeed, but nevertheless

the set 𝑉 with these operations satisfies the ten vector space axioms and hence is a vector
space. We will confirm Axioms 4, 5, and 7, and leave the others as exercises.

• Axiom 4—The zero vector in this space is the number 1 (i.e., 0 = 1) since
u+1=u⋅1=u
• Axiom 5—The negative of a vector u is its reciprocal (i.e., −u = 1/u) since
1 1
u+ = u( ) = 1 (= 0)
u u
• Axiom 7—k(u + v) = (uv)k = uk vk = (ku) + (kv).

Some Properties of Vectors

The following is our first theorem about vector spaces. Although the statements in this
theorem closely parallel familiar results in the arithmetic of real numbers, this is no guar-
antee that they are also true in vector arithmetic, so proof of their validity is required. The
proofs are very formal with each step being justified by a vector space axiom or a known
property of real numbers. There will not be many rigidly formal proofs of this type in the
text, but we have included this one to reinforce the idea that the familiar properties of
vectors can all be derived from the vector space axioms.

Theorem 4.1.1

Let 𝑉 be a vector space, u a vector in 𝑉, and k a scalar; then:

(a) 0u = 0
(b) k0 = 0
(c) (−1)u = −u
(d) If ku = 0, then k = 0 or u = 0.

We will prove parts (a) and (c) and leave proofs of the remaining parts as exercises.

Proof (a) We can write

0u + 0u = (0 + 0)u [Axiom 8]
= 0u [Property of the number 0]

By Axiom 5 the vector 0u has a negative, −0u. Adding this negative to both sides above
yields
[0u + 0u] + (−0u) = 0u + (−0u)
or
0u + [0u + (−0u)] = 0u + (−0u) [Axiom 3]
0u + 0 = 0 [Axiom 5]
0u = 0 [Axiom 4]
 4.1 Real Vector Spaces 209

Proof (c) To prove that (−1)u = −u, we must show that u + (−1)u = 0. The proof is as
follows:
u + (−1)u = 1u + (−1)u [Axiom 10]
= (1 + (−1))u [Axiom 8]
= 0u [Property of numbers]
=0 [Part (a) of this theorem]

A Closing Observation
This section of the text is important to the overall plan of linear algebra in that it estab-
lishes a common thread among such diverse mathematical objects as geometric vectors,
vectors in 𝑅n , infinite sequences, matrices, and real-valued functions, to name a few. As
a result, whenever we discover a new theorem about general vector spaces, we will at the
same time be discovering a theorem about geometric vectors, vectors in 𝑅n , sequences,
matrices, real-valued functions, and about any new kinds of vectors that we might dis-
cover.
To illustrate this idea, consider what the rather innocent-looking result in part (a) of
Theorem 4.1.1 says about the vector space in Example 8. Keeping in mind that the vec-
tors in that space are positive real numbers, that scalar multiplication means numerical
exponentiation, and that the zero vector is the number 1, the equation
0u = 0
is really a statement of the familiar fact that if u is a positive real number, then
u0 = 1

Exercise Set 4.1

1. Let 𝑉 be the set of all ordered pairs of real numbers, and con- d. Show that Axiom 5 holds by producing a vector −u such
sider the following addition and scalar multiplication opera- that u + (−u) = 0 for u = (u1 , u2 ).
tions on u = (u1 , u2 ) and v = (v1 , v2 ):
e. Find two vector space axioms that fail to hold.
u + v = (u1 + v1 , u2 + v2 ), ku = (0, ku2 )
In Exercises 3–12, determine whether each set equipped with the
given operations is a vector space. For those that are not vector spaces
a. Compute u + v and ku for u = (−1, 2), v = (3, 4), and identify the vector space axioms that fail.
k = 3.
3. The set of all real numbers with the standard operations of
b. In words, explain why 𝑉 is closed under addition and scalar addition and multiplication.
multiplication.
4. The set of all pairs of real numbers of the form (x, 0) with the
c. Since addition on 𝑉 is the standard addition operation on standard operations on 𝑅2 .
𝑅2 , certain vector space axioms hold for 𝑉 because they are
known to hold for 𝑅2 . Which axioms are they? 5. The set of all pairs of real numbers of the form (x, y), where
x ≥ 0, with the standard operations on 𝑅2 .
d. Show that Axioms 7, 8, and 9 hold.
6. The set of all n-tuples of real numbers that have the form
e. Show that Axiom 10 fails and hence that 𝑉 is not a vector
(x, x, . . . , x) with the standard operations on 𝑅n .
space under the given operations.
7. The set of all triples of real numbers with the standard vector
2. Let 𝑉 be the set of all ordered pairs of real numbers, and con-
addition but with scalar multiplication defined by
sider the following addition and scalar multiplication opera-
tions on u = (u1 , u2 ) and v = (v1 , v2 ): k(x, y, z) = (k2 x, k2 y, k2 z)

u + v = (u1 + v1 + 1, u2 + v2 + 1), ku = (ku1 , ku2 ) 8. The set of all 2 × 2 invertible matrices with the standard matrix
addition and scalar multiplication.
a. Compute u + v and ku for u = (0, 4), v = (1, −3), and 9. The set of all 2 × 2 matrices of the form
k = 2. a 0
[ ]
b. Show that (0, 0) ≠ 0. 0 b
c. Show that (−1, −1) = 0. with the standard matrix addition and scalar multiplication.