Section 3.

1: Direct Proof and Counterexample 1

In this chapter, we introduce the notion of proof in mathematics. A

mathematical proof is valid logical argument in mathematics which
shows that a given conclusion is true under the assumption that the
premisses are true. All major mathematical results you have considered
since you first started studying mathematics have all been derived in
this way e.g. Pythagoras Theorem, Fundamental Theorem of Calculus,
Fundamental Theorem of Algebra. Most of these proofs are long and
complicated and will be considered in further mathematics courses.
In this course, we shall consider more elementary proofs, mainly in
number theory, to start and strengthen our proof writing abilities.

1. Definitions
As stated at the beginning of the course, one of the most important
parts of mathematical proof is knowing and understanding the defini-
tions of what you are trying to prove things about. In this class and
all future classes if you do not learn and understand the definitions
you will not be able to prove things. Again, just for emphasis, defini-
tions are one of the most important parts of a mathematical
proof. For a comparison, you wouldn’t write an essay using words
which you don’t know what they mean, so why would you try to write
a mathematical proof about things you don’t understand?
As remarked above, in this chapter, we shall be considering a number
of different proofs in number theory, so we start by writing formal
definitions. Note that none of the definitions we are going to write
down are new to us, but the formal definition probably is.
Definition 1.1. (Odd and Even Integers) An integer n is even if and
only if n = 2k for some integer k. An integer is odd, if and only if
n = 2k + 1 for some integer k. Symbolically:

∀n ∈ Z, n is even ⇐⇒ ∃k ∈ Z, n = 2k

∀n ∈ Z, n is odd ⇐⇒ ∃k ∈ Z, n = 2k + 1
Definition 1.2. (Prime Numbers) An integer n is prime if and only
if n > 1 and for all positive integers r and s, if r · s = n, then r = 1
or s = 1. An integer n is composite if and only if n > 1 and n = r · s
for some positive integers r and s with r 6= 1 and s 6= 1. In formal
notation:
n is prime ⇐⇒ ∀r ∈ Z+ , ∀s ∈ Z+ , n = r · s → (s = 1 ∨ r = 1)
1
2

n is composite ⇐⇒ ∃r ∈ Z+ , ∃s ∈ Z+ , (n = r · s) ∧ ((s 6= 1) ∧ (r 6= 1))

Notice that definitions are ⇐⇒ statements i.e. quantified bicondi-
tional statements. We consider some examples of how to use these
definitions.
Example 1.3. Use the definitions we have given to answer the follow-
ing:
(i ) Is 5 odd?
5 is odd if 5 can be written in the form 2k + 1 for some
integer k. Let k = 2. Then 5 = 2 · 2 + 1, and thus by definition
5 is odd.
(ii ) Is nm − m composite for n > 2 and m > 1.
nm − m is composite if nm − m = r · s for some integers
r, s > 1. However, nm − m = m(n − 1), so if we choose r = m
and s = n − 1, then nm − m = r · s where r, s > 1, so by
definition, nm − m is composite.
(iii ) Is 2r − 6 + 10s even?
2r − 6 + 10s is even if 2r − 6 + 10s2 · k for some integer k > 0.
If we choose k = (r − 2 + 5s) > 0, then r − 2 + 5s > 1 and
s = n − 1, then nm − m = r · s where r, s > 1, so by definition,
nm − m is composite.

2. Proving and Disproving Existential Statements

Arguably, the easiest statements to prove are existential statements i.e.
statements of the form
∃x ∈ D, Q(x)
or “there exists x such that Q(x) is true”. In order to prove such
statements, we need to exhibit an explicit example of x ∈ D with
property Q (or such that Q(x) is true), or describe a set of directions in
order to find such an x. These methods of proof are called “constructive
proofs”, as opposed to non constructive proofs where the existence of
an element is guaranteed by an axiom, or previous proof, or is proved
by contradiction (by assuming such an x does not exist). We illustrate
with some examples.
Example 2.1. Show that there exists two prime numbers n and m
such that n + m = 18.
Choose n = 7 and m = 11, then n + m = 11 + 7 = 18.
Example 2.2. Suppose that k is an even integer and k > 2. Show
that there exists two prime numbers n and m such that n + m = k.
(Goldbachs conjecture!!!!)
 3

Example
√ √2.3. Show
√ that there exists real numbers a and b such that
a + b = a + b.
Take a = 0 and b = 1. Then
√ √ √ √
0+1= 1= 0+ 1
To disprove an existential statement, we need to prove its negation i.e.
to disprove an existential statement, we need to prove the universal
statement which is the negation of the existential statement. Since we
shall be considering universal statements until later, we shall return to
this problem then.

3. Disproving a Result by Counterexample

To disprove a statement means to show it is false. One typical way to
disprove a universal statement is to present a counterexample to what
is being posed. Formally:
Result 3.1. (Disproof by Counterexample) To disprove the universal
statement
∀x, P (x) → Q(x)
means to find an x such that P (x) is true, but Q(x) is false. In notation,
∃x, P (x)∧ ∼ Q(x).
We call such an x a counterexample.
Example 3.2. Consider the statement “for all real numbers x, if x2
is rational, then x is rational”. Disprove this statement by giving a
counter example.
√ √ 2
Consider
√ the number x = 2. Clearly 2 = 2 is rational. However,
2 is not rational (we shall see why later in the course). Thus we have
exhibited a real number x such that x2 is rational but x is not rational,
and thus the universal statement is not true.

4. Proving Universal Statements

Some of the most difficult statements to try to prove (and usually the
most interesting and useful statements to try to prove) are universal
conditional statements i.e. statements of the form
∀x ∈ D, P (x) → Q(x).
The first obvious way to attempt to prove such a statement is the
following:
Result 4.1. (Method of Exhaustion) When the domain D of x is finite,
to prove the universal statement
∀x ∈ D, P (x) → Q(x)
4

we can simply check for every element in D that if P (x) is true, then
so is Q(x).
It seems that the method of exhaustion is the best way to prove univer-
sal conditionals. However, it has one huge obstacle - it only works for
finite sets! (and in math, we are nearly always considering statements
about infinite sized sets). Therefore, in general, we would try a method
more like the following:
Result 4.2. (Method of Direct Proof) Suppose you are trying to prove
a universal conditional statement. Then we do the following:
(i ) Express the expression in logical form i.e. identify the hypoth-
esis and conclusion of the statement and the domain.
(ii ) Start the proof by supposing that x is a particular, but arbi-
trary chosen element of D for which P (x) is true.
(iii ) Show that the conclusion Q(x) is true by using definitions, pre-
viously established results, and the rules of logical inference.
The last two steps are sometimes called “The method of generalizing
from the generic particular”.
We illustrate these proof techniques with a couple of examples.
Example 4.3. For each integer n with 1 6 n 6 5, n2 − n + 11 is prime.
Since there are only finitely many integers with 1 6 n 6 5, we can use
the method of exhaustion. Specifically, we have
12 −1+11 = 11, 22−2+11 = 13, 32 −3+11 = 17, 42 −4+11 = 23, 52−5+11 = 31.
In each case, the resulting number is prime, so the statement is true.
As remarked before, proving universal statements with infinite domains
is much more difficult. The following three steps will usually help with
such problems:
(i ) (Formal Restatement) Always try to write a formal restate-
ment of the theorem you are trying to prove i.e. transform the
statement from informal language to logic.
(ii ) (Starting Point) Write down the things you are allowed to
assume given the statement of the theorem i.e. to prove a
statement ∀x, P (x) → Q(x), you need to suppose x is an arbi-
trary object which makes P (x) true, and then show that Q(x)
is true. For example, in the last problem we considered, the
starting point was the assumption that x was odd.
(iii ) (The Conclusion) Always keep in mind what the conclusion
is going to be. Sometimes, it may even be useful to have it
written somewhere on the page so you have a “roadmap” of
where you want to go.
 5

Example 4.4. Show that for any integer n, if n is odd, then n2 is odd.
In this case, we cannot use the method of exhaustion since there are
infinitely many different odd integers. Therefore, we must use the
method of generalizing from the generic particular. Specifically, as-
suming that x is an odd integer, we need to show that without any
further assumptions that x2 is odd. We shall follow the steps above.
(i ) (Formal Restatement) ∀x ∈ Z, x is odd → x2 is odd
(ii ) (Starting Point) Our only assumption is that x is some odd
integer. This means that x = 2k + 1 for some integer k.
(iii ) (Body) Since x = 2k + 1, we have x2 = (2k + 1)2 = 4k 2 + 4k + 1
(using standard rules of algebra). Therefore, if m = 2k 2 + k,
then x2 = 2m + 1.
(iv ) (Conclusion) Therefore x2 is an odd integer

5. General Proof Techniques and Common Mistakes

A large portion of this class will be dedicated to learning how to write
mathematical proofs of statements (we usually call such statements
“Theorems”). There is a widely accepted structure to such communi-
cation, so we shall briefly outline how you should present your proofs.
In addition, there are some useful general steps and techniques which
should always be taken when writing a proof, and also some common
pitfalls which people fall into when writing proofs. First, when trying
to prove a theorem, the general structure should be as follows:
(i ) Write the word “Theorem” and then the statement you are
trying to prove after it.
(ii ) On the next line write the word “proof” - it is from this point
onward you shall start to write your proof. Note that this sep-
arates the statement of the theorem from the proof, so should
help avoid any confusion.
(iii ) Clearly mark the end of your proof with “QED”, or some other
such symbol (this is especially important if you are proving
three or four statements.
In addition to the general structure given above, a good (and correct!)
proof will exhibit the following:
(i ) A proof should always be self-contained, meaning all variables
which are used in the proof should be clearly defined
(ii ) You should write a proof in complete sentences. This doesn’t
mean you should not use symbols or abbreviations in a proof,
but rather they should be incorporated into your sentences
(iii ) Provide a reason for each assertion you make in you proof or
each step you take - if you don’t back up the steps you take,
you could end up assuming something that isn’t true.
6

(iv ) Use typical “buzzwords” between statements to make the ar-

gument in your proof more clear. For example, if one state-
ment is a consequence of the previous, we could use the word
“therefore”, or “it follows that” with a brief reason why the
second statement follows from the first at the end of the sen-
tence. When introducing new variables, we use the word “let”
(e.g. let x be an even integer).
As usual, there are certain mistakes which beginners (and indeed ex-
perts) make when writing proofs. Some of the more common ones you
should watch out for are the following:
(i ) Trying to prove a universal statement through examples: re-
member, just because a statement holds for a small number of
examples, does not mean it holds for all examples!
(ii ) Using the same variable to represent more than one thing
(iii ) Jumping to a conclusion i.e. asserting the truth of a statement
without giving a reason
(iv ) Begging the question i.e. assuming what you are trying to
prove
(v ) Misuse of the word “if” i.e. “if” can sometimes be used instead
of the word “because”, so a statement which is meant to be an
assertion can turn out to be conditions because the word “if”
is used instead of because.
We illustrate with a formal proof.
Example 5.1. Prove the statement “The difference of any two odd
integers is even”
Theorem 5.2. The difference of any two odd integers is even or
∀x ∈ Z, ∀y ∈ Z, x is odd ∧ y is odd → x − y is even
Proof. Suppose x and y are (arbitrary but particular) odd integers.
Then there exists an integer k such that x = 2k + 1 and an integer m
such that y = 2m + 1. Taking the difference of x and y, we get
x − y = 2k + 1 − (2m + 1) = 2k + 1 − 2m − 1 = 2k − 2m = 2(k − m)
use the standard rules of arithmetic. In particular, x − y = 2(k − m)
and k − m is an integer. Therefore, x − y is even.

We finish with an example.
Example 5.3. Check to see if the following statement seems true or
false and then prove or disprove accordingly: “There exists an integer
m > 3 such that m2 − 1 is prime” is false.
Looking at a small number of examples, we do not get any primes, so
our gut feeling is that this should be false. In order to show that this
statement is false, we need to show the negation is true.
 7

Theorem 5.4. ∀m ∈ D, m2 − 1 is not prime where D is the set of all

integers greater or equal to 3.
Proof. Suppose m ∈ D. Using simple algebra, m2 −1 = (m−1)(m+1).
Since m is an integer, so are m − 1 and m + 1. Since m > 3, m − 1 > 2
and m + 1 > 2. In particular, m2 − 1 is the product of two positive
integers greater than 1 and hence by definition, it is not prime. 
Homework
(i ) From the book, pages 139-141: Questions: 2, 7, 11, 14, 21, 25,
32, 36, 42, 46, 47