Theory of Equations

Lesson 1
by

Barry H. Dayton
Northeastern Illinois University
Chicago, IL 60625, USA
www.neiu.edu/bhdayton/theq/

These notes are copyrighted by Barry Dayton, 2002. The PDF files are freely available on
the web and may be copied into your hard drive or other suitable electronic storage devices.
These files may be shared or distributed. Single copies may be printed for personal use but
must list the website www.neiu.edu/bhdayton/theq/ on the title page along with
this paragraph.
Maple and MAPLE represent registered trademarks of Waterloo Maple Inc. TI-85/86
are trademarks of Texas Instruments Inc.

Preface
These notes were first prepared for a course given at Northeastern Illinois University
in the summer of 1989. Theory of Equations was a traditional course in the mathematics
curriculum which dealt with the important topics of the theory and solution of polynomial and linear equations. Recently, thanks to the increased availabilty of computers,
linear algebra has taken a central role in mathematics and there are specialized courses
devoted to that topic. On the other hand, the theory of polynomial equations, while still
important, has been scattered among various courses.
I had two motivations for preparing notes rather than using a standard text. First
of all there have many changes in the way we view the material since the 1940s when
such classics such as the text by Uspensky were written. Computers, and even calculators, have made much of the traditional material, which dealt with efficient hand
computation schemes and separation of roots, obsolete. Different numerical methods
are now prefered. Interest in dynamical systems and chaos has brought new insights
into Newtons method. And there have been exciting new algorithms developed for
factoring polynomials. Thus one motivation for these notes is to give a modern view of
this classical subject.
My other motivation has to do with the number of different areas of mathematics in
which polynomials arise. Thus by using the concept of the polynomial as a common
thread I am able to cover a large area of the mathematical landscape. Generally students
of mathematics are able to see only a small portion of the total of mathematics and
often are not given the motivation, practical and historical, for many of the branches
of modern mathematics. Thus it is my intention to use this course partly as a survey
of a large chunk of pure and applied mathematics from a historical as well as modern
point of view. It is my hope that students will emerge from this course not so much
with specific skills in solving polynomial equations but rather with a better perspective
on what mathematics is, where it came from, and where it is going.
I wish to thank various students who have helped motivate me and make these notes
possible. First of all, my Summer 1989 class for their patience in receiving these notes
a few pages at a time and who pointed out many typographical and other errors. I also
wish to acknowledge my debt to James Moor and Ken Touff whose masters projects
provided the main motivation and much of the material for Chapter 3.
And finally I wish to give special thanks to James Moor who contributed to the
preparation for this edition, especially with careful proofreading and contribution of
new problems for Chapter 1.

Chapter 1
College Algebra
R EAL AND C OMPLEX P OLYNOMIALS
In this chapter we review the standard theory of real and complex polynomials as
might appear in a good high school college algebra course.

1.1

The Real Number System

We start out with an axiomatic approach to the real number system; that is, we list a
number of properties that the system of real numbers possesses. In this course we will
not attempt to construct this system nor justify the axioms. This is left for a course
in real analysis. We will use these axioms as our starting point and base future proofs
and constructions on them. We also have a second motivation: the properties of these
axioms, associative laws, distributive laws, etc., will arise in many other contexts.
The real number system consists of a set of numbers R, two binary operations +, ,
and a relation < which satisfy the following axioms:
1. Closure
C1. Given a, b R, there exists a unique element a + b R.
C2. Given a, b R, there exists a unique element a b R.
2. Ring Axioms
R1. For all a, b R. the commutative law a + b = b + a holds.
1

CHAPTER 1. COLLEGE ALGEBRA

R2. For all a, b R, the commutative law a b = b a holds.
R3. For all a, b, c R, the associative law a + (b + c) = (a + b) + c holds.
R4. For all a, b, c R, the associative law a (b c) = (a b) c holds.
R5. For all a, b, c R, the distributive law a (b + c) = (a b) + (a c) holds.
R6. There is a unique number 0 such that a + 0 = a for all a R. 0 is called
the additive identity.
R7. There is a unique number 1 such that a 1 = a for all a R. 1 is called
the multiplicative identity or unit element.
R8. For each a R, there is a unique (a) R so that a + (a) = 0. The
number (a) is called the additive inverse of a.
3. Integral Domain Axiom
I1. Given a, b R, if a b = 0 then a = 0 or b = 0.
4. Field Axiom
F1. Given a R such that a 6= 0, there exists a unique (a1 ) R such that
a (a1 ) = 1. (a1 ) is called the multiplicative inverse of a.
5. Order Axioms
O1. Given a, b R, exactly one of a < b, a = b, or b < a holds.
O2. If a, b, c R, then a < b and c < d implies a + c < b + d.
O3. If a, b R, then 0 < a and 0 < b implies 0 < a b.
6. Completeness Axiom
O4. If S is a non-empty subset of R and M R satisfies a < M for all a S,
then there exists u R so that a u for all a S, and if some b < u,
then there is an element a S with b < a. The element u is called the least
upper bound of S.

A few comments on these axioms are in order. The closure axioms say that the
operations of addition and multiplication are defined; that is any two real numbers
can be added or multiplied. We will define the operations of subtraction and division

1.2. A LITTLE HISTORY OF REAL AND COMPLEX NUMBERS

in terms of these basic operations. In particular, since axiom R8 says that each real
number b has a unique additive inverse b, we can define subtraction by
a b = a + (b).
In a similar manner we can define division by
a/b = a (b1 ) when b 6= 0.
Division, unlike addition, multiplication, and subtraction does not satisfy closure axioms; division by 0 is not possible. Note also that subtraction and division fail many of
our laws, in particular, the commutative and associative ones.
One should also note that axiom I1 is redundant: it can be proved from axiom F1.
For if a b = 0 but a 6= 0, then multiplying by a1 gives
b = 1 b = (a1 a) b = a1 (a b) = a1 0 = 0.
Axiom I1 is included here for later reference.
The relation < is defined by the order axioms. One can easily define the relations
>, , from this.
Finally, it is the Completeness Axiom that really distinguishes the real number system from other systems such as the rational number
system. It is from the
completeness

property that we can define such numbers as 2, , and e. For example, 2 can be defined as the least upper bound of the set S = {a R|a2 2}. The set S is bounded
above by,
for instance, the number 2, and S is not empty since 1 S, so the least upper
bound 2 exists.

1.2

A little History of Real and Complex Numbers

Although we, as in many algebra and calculus books, take the real numbers as a starting
point, historically the real number system as we know it today appeared rather late in
the game. In fact, only since the end of the last century have the properties of the real
numbers been carefully established and axiomizations such as we have given date from
early in this century.
Although the Pythagoreans started out promisingly around 500 BC with mathematics as the study of numbers, the discovery by Pythagoras, or more likely one of his
followers,
of the incommensurability of the diagonal of a square (that is, the fact that

2 is not a rational number) put a damper on the development of the number system.

CHAPTER 1. COLLEGE ALGEBRA

Euclid, following the precepts of Plato and Aristotle, developed his geometry entirely
without the aid of numbers, using line segments instead. Unfortunately for mathematics
the Euclidean approach was dominant for over 1500 years in Europe.
The Hindu mathematicians did use numbers and it is from them that we get our
current numerals. The ninth century Arab mathematician Al-Khwarizmi did advocate
the use of numbers and is known as the father of algebra (Al-jabr was the name of his
book), but his immediate followers insisted on proving algebra using Euclidean geometry. Fibonacci was instrumental in bringing algebra and numbers to Europe about
1200 but owing to the still powerful influence of Euclid the practice of proving algebraic solutions by geometry remained through the time of Cardano and beyond. It was
Cardano who, in his 1545 book Ars Magna, first suggested the idea that an equation
could have a negative solution. Descartes furthered the use of negative numbers in his
famous La Geometrie in 1637 using negative numbers as coefficients as well as roots,
but still called negative roots false.
With the invention of calculus late in the 17th century numbers once again became
central in mathematics. Yet it was not until the middle 19th century that it was felt
necessary to establish the formal properties of the real number system. These properties
were finally established by the work of mathematicians such as Bolzano, Weierstrass,
Cantor, and Dedekind. The present axiomatic approach is due largely to the influence
of Hilbert in the early part of this century.
Complex numbers are no more recent than real numbers. In fact, complex numbers
made their appearance in the very same chapter of Ars Magna that negative numbers
first appeared. Cardano asserted that these numbers did not really exist; you had to
imagine them. It is from this comment that non-real complex numbers are still called
imaginary. In fact the term real appears to have been used first to mean not imaginary. This usage appears for instance in Descartes. Complex numbers were studied
by Bombelli in 1572 and by Girard in 1629. Leibniz, co-inventor of calculus, was fascinated by complex numbers in the late 17th century only partly because of his belief
that they could be used to give a proof of the existence of the Holy Ghost. De Moivre,
a follower of Newton, studied complex numbers in 1707. Complex numbers played
an important role as Euler, in the years from 1746 through 1777, established the modern terminology of mathematics with such symbols as e, , i, f (x) and introduced the
modern trigonometric functions.
Later in this chapter and in Chapter 3 we will see the importance of complex numbers in solving equations with the Fundamental Theorem of Algebra. The further importance of complex numbers in modern mathematics was firmly established by the
work of Cauchy and Riemann in the 19th century. The Fundamental Theorem of Algebra is only one of many powerful theorems in mathematics that would be lost without

1.3. COMPLEX NUMBERS

the use of complex numbers. In physics, complex numbers are very important in the
study of periodic behavior, such as electromagnetic waves.

1.3

Complex Numbers

We assume the existence of a number i (not in R) with the property that i2 = 1. The
set C of complex numbers consists of all numbers a + bi where a, b are real numbers.
In order to have laws R1 - R5 hold we are forced to define operations +, by
(a + bi) + (c + di) = (a + c) + (b + d)i
and
(a + bi) (c + di) = (ac bd) + (ad + bc)i
Note that 0 = 0 + 0i is an additive identity and 1 = 1 + 0i is a multiplicative identity.
Furthermore (a + bi) = (a) + (b)i so in fact all of the rules R1 - R8 hold.
For a complex number z = a + bi the real number a is called the real part of z,
i.e., a = Re(z), and the real number b is called the imaginary part of z. We write
b = Im(z). z is called real if Im(z) = 0 and imaginary if Im(z) 6= 0.
For a complex number z = a + bi the complex conjugate is z = a bi. Several
important properties of the conjugate should be noted:
z + w = z + w and z w = z w
and if z = a + bi, then
z z = a2 + b2
is a non-negative real number.
In light of the last equation we can
define the modulus
or absolute value of a com
2
2
plex number z= a + bi to be |z| = a + b = z z (as usual in mathematics,
in these notes x will denote the positive square root of the positive real number x).
Again note the identities
|z w| =
| z| =
|
z| =
|z + w|
|z| = 0

|z| |w|
|z|
|z|
|z| + |w|.
if and only if z = 0.

CHAPTER 1. COLLEGE ALGEBRA

z
r

If r is a real number and z = a + bi a complex number it makes sense to write

= ar + rb i. Thus given z = a + bi 6= 0 we have a complex number
w=

z
a
b
= 2
2
i.
2
2
|z|
a +b
a + b2

A simple calculation shows that zw = 1. Thus we can write z 1 = z/|z|2 and we see
axiom F1 holds. As in the case of R we can write (for w 6= 0) z/w = z w1 . Thus
the complex numbers satisfy all the field axioms that the reals do.
We note that it is impossible for us to define a relation < on C satisfying O1 - O4
for by O1 we must have either 0 < i or i < 0 which in the second case would give by
O2 0 < i. But since i2 = (i)2 = 1 < 0, O3 would be violated. Thus we will only
use the relation < when both sides are known to be real.
Until recently doing arithmetic with complex numbers was sort of a pain since
complex numbers were not supported by most calculators and most computer languages
such as BASIC, Pascal etc. (FORTRAN was an exception). However many of the new
generation of graphing calculators do support complex numbers, as do the symbolic
mathematics programs such as Maple and Mathematica. A handy way to do complex
arithmetic today is to use the TI-85/86 calculator. Although we will not use the TI85/86 later in these notes, we will explore the complex number capabilities of these
calculators since this gives insight into the arithmetic of complex numbers.
TI-85/86 Implementation
The complex constant a+bi is expressed in ordered pair notation
(a, b), for instance 3 + 2i is entered as (3, 2). To store as a variable
simply use the [STO] key, i.e., (3,5) [STO] z assigns the variable z the value 3 + 5i. Likewise you can use (0,1) [STO] i to
assign the number i to the variable i. Then you can enter 3 + 4i by 3
+ 4*i, but it will still display as (3,4). Note that complex variables
are a special variable type on the VARS or MEM screens. Complex
constants, expressions or variables can be used as list, vector and
matrix elements.
I find it convenient to set the number of decimal places to 3 or 4
on the MODE screen so I can see the whole output on the screen at
one time. To do arithmetic just type it in as if you were calculating
with real numbers i.e.,
(1,3) + (5,7)

1.3. COMPLEX NUMBERS

(2,1) * (4,-7)
(3,6) / (8,2)
z + w/u
3*(2,5)+8
and so forth. Note that in the last example, when mixing real and
complex numbers the real numbers are evaluated as complex numbers i.e., 3 = 3 + 0i etc. Your TI-85/86 will even interpret the expression (X,Y) correctly if X and Y are real variables or expressions.
(Note: for those unfamiliar with the TI-85/86, negative numbers are
entered with the [(-)] key, not the usual minus key.)
The CPLX menu has functions which calculate the conjugate,
find the real and imaginary parts and the absolute value or modulus of a complex number or variable, for example you can enter
expressions such as
conj (3,5)
real z
abs (z*w)
It is not an accident that the TI-85/86 uses the notation (x,y) for a complex number. Geometrically we can identify the set C with the ordinary Euclidean plane by
associating the complex number a + bi with the point or vector (a, b). One can then
translate some of our algebra to geometry as follows. The x-axis can now be called the
real axis while the y-axis is the imaginary axis. Addition of complex numbers corresponds to vector addition in the plane. The modulus of a complex number is the length
of the corresponding vector. The negative of a complex number corresponds to the
negative of the vector and the conjugate of z = a + bi is the reflection of (a, b) in the
x-axis. The geometrical interpretation of multiplication is more complicated but even
more beautiful and important than the above.
Before discussing multiplication we must discuss the argument or angle of a complex number. Given z = a + bi we measure the angle from the positive real axis to
the point (a, b) in radians (ALWAYS radians). This angle is allowed to be positive or
negative and if one wants one may go all the way around several times. Thus it is
permissible to measure the argument of the complex number i as /2, 3/2, 5/2 or
even 29/2. Unfortunately there is no one correct way to measure the argument, we
do note however that different permissible values for the argument of a given complex
number differ by 2n where n is an integer.

CHAPTER 1. COLLEGE ALGEBRA

Then if a complex number z = a + bi has modulus r = |z| and some argument

then the polar form of z is z = r(cos + i sin ). From the equations
cos( + ) = cos cos sin sin
sin( + ) = sin cos + cos sin
we see that if z = r(cos + i sin ) and w = s(cos + i sin ) then z w = rs(cos( +
) + i sin( + )). Thus multiplying complex numbers corresponds geometrically to
multiplying the moduli and adding the arguments. Note also that for z, w as above
z
r
=
(cos( ) + i sin( ))
w
s
z = r(cos( + ) + i sin( + ))
z = r(cos() + i sin()).
From the above geometry of multiplication we get De Moivres famous rule (given
in 1707) for finding an integral power of a complex number z = r(cos + i sin ),
mainly
z n = rn (cos n + i sin n).
The importance of this rule is that one can go backwards, i.e., one can find the nth
root of any complex number. However we must note carefully that because of the
uncertainty in the argument there are several, in fact n, angles which when multiplied
by n give equivalent arguments. Thus for z 6= 0, and n > 0 an integer the n nth roots
of z = r(cos + i sin ), (r > 0) are
1

z n = r n (cos(

+ 2k
+ 2k
) + i sin(
))
n
n

for k = 0, 1, 2, . . . , n 1.
Of particular importance are the nth roots of unity, that is the n complex nth roots
of 1. These are the numbers
k = cos

2k
2k
+ i sin
n
n

These numbers are useful, for if we can find one nth root w of a complex number z,
then the other roots are 1 w, 2 w, . . . , n1 w.
TI-85/86 Implementation

1.3. COMPLEX NUMBERS

The TI-85/86 has the special notation (r6 ) for the complex number z = r(cos + i sin ). Note that the 6 key is [2nd][ , ] so
6 takes the place of the comma in the ordered pair. For example,
entering (36 3 /2) into the TI-85/86 gives the result (0,-3) for
the complex number 3i, whereas entering (36 /4) gives the result (2.1213, 2.1213). However changing the MODE screen from
RectC to PolarC changes the output of each complex number
to polar form. One can also use the functions Rec and Pol in
the CPLX menu to change from polar to rectangular and back.
Note that the syntax to change rectangular to polar form is (x,y)
[CPLX][MORE][F1] .
To find an integral power of a complex number simply use
the power key, [] For example (3,4)7 gives the result
(76443,16124). With the PolarC mode on, (26 .7) raised to the 5th
power gives the result (326 2.7832). Note that by DeMoivres rule
the argument should be .7 5 = 3.5 but the TI-85/86 prefers to give
an argument between and so gives 3.5 2 = 2.7832.
In PolarC mode it appears that taking the nth root of (r6 ) gives
the one value (r1/n 6 /n) with the angle normalized to the standard
range to . Thus to get all the nth roots you need to add the
angles 2k/n to the argument for k = 0, 1, . . . , n 1. The easiest
way to do this is to multiply the original result by the nth roots of
unity (16 2k/n) for the appropriate values of k.
This works also in RectC mode. Thus to find the 3 cube roots
of 2 + 5i, for example, enter (2,5)(1/3) (or use the special
nth root key on the [MATH][MISC] menu) getting (1.6166, .6773)
which is one cube root. Now multiply this by (16 2 /3) (remember to use the internal value of which is [2nd][] ) to get
(1.3949, 1.0614). Now multiply this last value again by (16 2 /3)
to get the third root (.2217, 1.7387). An alternate approach is to
use STO and the CPLX menu to first do [CPLX][abs] (2,5)
[STO] R to set R to the value |2 + 5i| = 5.3852 and to then do
[CPLX][angle] (2,5) [STO] A to set A to the value 1.1903
which is the argument of 2 + 5i. Then do R(1/3) [STO] S and
the first cube root is (S 6 A/3) which is (1.6166, .6773) as above, the
second is (S 6 A/3 + 2 /3) which gives (1.3949, 1.0614) as above
and the third is then (S 6 A/3 + 4 /3), again the same as above.
Finally note that the nth root functions on the TI-85/86 are not

10

CHAPTER 1. COLLEGE ALGEBRA

continuous. For example (26 3.132) is (.0068, 1.4142) whereas

6
(2 3.153) evaluates to (.0081, 1.4142). This is not a defect of the
TI-85/86 calculator; it can be proven mathematically that there does
not exist a continuous function on C which gives one of the square
roots of each number z. The TI-85/86 is a marvelous little device,
but it occasionally cant do the mathematically impossible!

Exercise 1.3.1 Find (preferably using a TI-85/86) the following in the form a + bi,
using decimal approximations when appropriate.
1. (3 + 4i) (4 3i)
2. (2 3i)/(5 + 6i)
3. (3 7i)4
4. The 5 fifth roots of 3 + 4i
5. The three complex cube roots of 2
6. The four complex fourth roots of 5.
In the future in thess notes we will be using MAPLE as our calculating and computing environment. We take our first look at MAPLE now.
Maple Implementation
The syntax for MAPLE is not that different from the TI-85 with
some important exceptions. Variables which have not previously
been assigned numerical values are treated as indeterminants. Algebraic expressions in these variables are then manipulated algebraically. Also MAPLE treats rational numbers and floating point
numbers differently. The quotient of two rational numbers is a fraction. Roots of rational numbers are not evaluated as floating point
numbers but left in surd form. Numbers with decimal points are
treated as floating point numbers, others are not.
The most important thing to remember about MAPLE are that
you must end your input with a semicolon ; unless you do not want
to see output in which case you use a colon :. Also when you
assign a value to a variable use the assignment operator := as in

1.3. COMPLEX NUMBERS

x := sqrt(2);

to assign the variable x the value 2. If you want a decimal value

but MAPLE returns an exact (symbolic) value use evalf as in
evalf(sqrt(2));
to get a decimal approximation of the square root of 2.
We now turn to complex
numbers. The letter I (capital I) is re
served for the MAPLE 1. Thus a complex number such as 3 + 4i
would be written 3 + 4*I using MAPLE syntax.
To extract the real or imaginary part of a complex number z use
x:=Re(z);
y:=Im(z);
note that to retain consistency with earlier versions of MAPLE Re,
Im violate the usual case rules of MAPLE which usually start built in
MAPLE procedures with small letters. The procedure evalf works
with complex as well as real numbers to convert to floating point. A
procedure evalc is is useful for breaking an imaginary number into
its real and imaginary parts. The modulus can be calculated by r:=
abs(z);, and the argument is given by t := argument(z);.
There is a provision for conversion between polar and rectangular coordinates. polar(r,alpha) is a name given to the
complex number z = r(cos + i sin ) where alpha is the variable name of the numerical value of . This representation is only
known to evalc. Thus evalc(polar(2,Pi/4)) should give
1.41421 + 1.41421 I. Note Pi is the predefined name for the
internal value of . The inverse conversion uses the convert function, i.e., convert(2+3*I, polar); might return
(1/2)
polar(13
, arctan(3/2))
As in the TI-85/86 case, if given w as one, say, third root of z you
can find the others by
w*polar(1, 2*Pi/3)
w*polar(1, 4*Pi/3)
Incidentally, I should mention that MAPLE works in adjustable
accuracy which means that you can have as many digits calculated

11

12

CHAPTER 1. COLLEGE ALGEBRA

and printed out as you wish. If too many digits are being printed out
for your tastes, use Digits:=5; to set the ouput to 5 digits. On
the other hand if you want to see 40 digits, use Digits:= 40; and
you will get them, and they will be mostly all accurate.