METHODS OF PROOF

1.1. Introduction
Mathematics is a language that consists of symbols and definitions. There are also statements
or assertions, whose validity require to be firmly established, before being used in generating
new facts or other statements alike. These statements include theorems, corollaries, and
lemmas. When a mathematical statement is given, it is imperative that its validity is
established. Therefore an investigation of whether the statement is valid or not, should be
carried out. This process of investigation is what we will refer to as a presenting proof. A proof
of the theorem, or other result, is a series of logical deductions, using the assumptions of the
theorem, the definitions of the terms involved, and previous results that have been proven. In
this unit, we will look at different methods of providing proof that a statement or theorem is
correct: direct proof and indirect proof. We will also discuss proof by applying the principle of
mathematical induction. The most common method of showing that a statement is not true, is
by presenting a case that disputes it. This, too, will be presented.
1.2. Aim
The aim of this unit, is to give you the techniques available to prove the validity of a
mathematical statement and how to dispute its assertion.
1.3. Objectives
At the successful completion of this unit, you are expected to be able to
i. Describe what a direct proof and an indirect proof is
ii. Apply the principle of mathematical induction to prove that a given statement is true
iii. Prove the validity of a statement using contradiction
iv. Show that a statement is not correct using a counterexample
1.4. Time Frame
This unit has 3 major sections, including methods of proof, proof by counterexample and the
principle of mathematical induction. These can be covered 4 hours.
1.5. Topics
1.5.1. Proving mathematical statements
Mathematical statements are usually presented in the form of an implication. That is if 𝐴
then 𝐵. Which means that in an event where you find that A is true, then you can conclude that
𝐵 is true as well. 𝐴 is usually a fact or a set of facts known as the premise or hypothesis, and
𝐵 is the conclusion. It is essential that you try to break any statement you are to prove into the
form if premises then conclusion. For example, look at the following theorem
Theorem 2.1
The sum of two even integers is even.
Before we prove that this statement is true, we will need to understand what facts are assumed
to be available, the premise, and what is to be drawn from these facts, the conclusion. These
need to be clearly identified. Then we can restate it in the form “if PREMISE then
CONCLUSION” or equivalently, “PREMISE implies CONCLUSION”. In this case, we are
assuming that two even integers are given and we wish to prove that their sum is even. With
this information, we can restate our theorem as follows:
Theorem - restated
𝐿𝑒𝑡 𝑎 𝑎𝑛𝑑 𝑏 𝑏𝑒 𝑒𝑣𝑒𝑛 𝑖𝑛𝑡𝑒𝑔𝑒𝑟𝑠, 𝑡ℎ𝑒𝑛 𝑡ℎ𝑒 𝑠𝑢𝑚 𝑜𝑓 𝑎 𝑎𝑛𝑑 𝑏 𝑖𝑠 𝑒𝑣𝑒𝑛
𝑜𝑟
𝐼𝑓 𝑎 𝑎𝑛𝑑 𝑏 𝑎𝑟𝑒 𝑒𝑣𝑒𝑛 𝑖𝑛𝑡𝑒𝑔𝑒𝑟𝑠, 𝑡ℎ𝑒 𝑠𝑢𝑚 𝑜𝑓 𝑎 𝑎𝑛𝑑 𝑏 𝑖𝑠 𝑒𝑣𝑒𝑛.
Before we can prove this theorem, we are going to look at the various methods available, and
see which one(s) is applicable for such a case.
1.5.1.1. Method of exhaustion
Like the name suggest, this method proceeds by enumerating all the possible occurrences of
the conclusion and show that they actually agree with the statement.
Example 2.1
Prove that if 𝑎 positive integer less than 5, then its square is less than 30.
Proof
Restated if 𝑥 𝑖𝑠 𝑎𝑛 𝑖𝑛𝑡𝑒𝑔𝑒𝑟 𝑖𝑠 𝑠𝑢𝑐ℎ 𝑡ℎ𝑎𝑡 0 < 𝑥 < 5, then 𝑥 2 < 30
To prove this we simply show all occurrences of 𝑥 2 , 0 < 𝑥 < 5. If all of them are less than 30,
then the proof is complete.
x x*x x*x<30
1 1 TRUE
2 4 TRUE
3 9 TRUE
4 16 TRUE
Figure 1: Enumeration table showing that x*x <30 for x in {1, 2, 3, 4)

The third column of Figure 1 tells the whole story. Since the statement holds for all possible
values in the premise, the statement is correct.
From here it should be immediately clear to you, what challenges this method of proof holds.
For this case, the set of items to check is small and we had that luxury of easily itemising them.
As the set becomes big, however, say if it had 100 items, and worst still, if the set is infinite,
this method shamefully fails to do the job.
Clearly we cannot use it for our theorem on even numbers above, since we have an infinite
number of evens. While we can show that 2 + 4, 6 + 4, 8 + 2 are all even, this is not
sufficient to conclude that the theorem is correct. Another technique will have to be employed.
1.5.1.2. Direct proof
This method of proof, requires that a statement is expressed in the form of "𝑖𝑓 𝑃 𝑡ℎ𝑒𝑛 𝑄". More
specific, suppose we have properties 𝑃 and 𝑄, the statement should take the form
 “𝑓𝑜𝑟 𝑎𝑙𝑙 𝑜𝑐𝑐𝑢𝑟𝑟𝑒𝑛𝑐𝑒𝑠 𝑜𝑓 𝑥 𝑖𝑛 𝑡ℎ𝑒 𝑑𝑜𝑚𝑎𝑖𝑛 𝐷, 𝑖𝑓 𝑥 𝑠𝑎𝑡𝑖𝑠𝑓𝑖𝑒𝑠 𝑃 𝑡ℎ𝑒𝑛 𝑥 𝑠𝑎𝑡𝑖𝑠𝑓𝑖𝑒𝑠 𝑄 𝑎𝑠 𝑤𝑒𝑙𝑙”
𝑜𝑟 𝑒𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡𝑙𝑦
∀𝑥 ∈ 𝐷, ∈ 𝑃(𝑥) → 𝑄(𝑥) … … … (𝑇)
1.5.1.3. Steps for direct proof
To prove such a statement of the form (T) above, we follow the steps below
STEP 1:. If you have not done that yet, express your statement in the form of (𝑇)
STEP 2:. Take an arbitrary 𝑥 from 𝐷 and assume that it satisfies 𝑃 abbreviated as
Suppose x ∈ D and P(x)
Show that the conclusion 𝑄(𝑥) is true, by applying definitions, previously established
results, and the rules for logical inference. Use the directions for writing proofs given
below.
1.5.1.4. Directions for writing proofs
Below are the best practices you should employ when writing your proof:
 Write the theorem to be proved.
 Clearly mark the beginning of your proof, with the word Proof.
 Make your proof self-contained by clearly identifying each variable, giving comments
on how the subsequent steps are arrived at, e.g. “by definition” or reference to a step
previously met.
Example 2.2
Prove that adding an odd number to itself gives an even number
Proof
We start the process by restating or theorem.
Theorem
∀𝑥 ∈ 𝑍 𝑥 𝑖𝑠 𝑜𝑑𝑑 → 𝑥 + 𝑥 𝑖𝑠 𝑒𝑣𝑒𝑛
Proof
Let 𝑥 be an integer and 𝑥 is an odd number
Then 𝑥 = 2𝑘 + 1, by definition of an odd number … (𝑖)
So
𝑥 + 𝑥 = 2𝑘 + 1 + 2𝑘 + 1, 𝑏𝑦 (𝑖)
= 2𝑘 + 2𝑘 + 1 + 1, 𝑏𝑦 𝑐𝑜𝑚𝑚𝑢𝑡𝑎𝑡𝑖𝑣𝑒 𝑙𝑎𝑤
= 4𝑘 + 2,
= 2(2𝑘 + 1), 𝑓𝑎𝑐𝑡𝑜𝑟𝑖𝑠𝑎𝑡𝑖𝑜𝑛
= 2𝑝, 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑖𝑛𝑡𝑒𝑔𝑒𝑟 𝑝, 𝑠𝑖𝑛𝑐𝑒 2𝑘 + 1 𝑖𝑠 𝑎𝑛 𝑖𝑛𝑡𝑒𝑔𝑒𝑟 𝑓𝑜𝑟 𝑎𝑛𝑦 𝑖𝑛𝑡𝑒𝑔𝑒𝑟 𝑘
We see that
𝑥 + 𝑥 = 2𝑝, 𝑓𝑜𝑟 𝑝 ∈ 𝑍
𝐻𝑒𝑛𝑐𝑒 𝑥 + 𝑥 𝑖𝑠 𝑒𝑣𝑒𝑛, 𝑏𝑦 𝑑𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛. [𝑄. 𝐸. 𝐷]
This is a very trivial fact but it demonstrates the ideas stated in above.
Note that the comments may be omitted, but are usually necsessary in certain parts where the
derivation of new information might not be intuitively presented to the reader of your proof.
Example 2.3
Lets draw our attention to the the theorem stated earlier on the sum of even integers. Note
that we restated it as follows:
∀𝑥, 𝑦 ∈ 𝑍 ∶ 𝑥 𝑎𝑛𝑑 𝑦 𝑎𝑟𝑒 𝑒𝑣𝑒𝑛 → 𝑥 + 𝑦 𝑖𝑠 𝑒𝑣𝑒𝑛
Proof
𝑆𝑢𝑝𝑝𝑜𝑠𝑒 𝑥 𝑎𝑛𝑑 𝑦 𝑎𝑟𝑒 𝑒𝑣𝑒𝑛 𝑖𝑛𝑡𝑒𝑔𝑒𝑟𝑠.
Then
𝑥 = 2𝑝, 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑝 ∈ 𝑍 … (𝑖)
𝑦 = 2𝑞, 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑞 ∈ 𝑍 … (𝑖𝑖)

𝑆𝑜
𝑥 + 𝑦 = 2𝑝 + 2𝑞, 𝑏𝑦 (𝑖) 𝑎𝑛𝑑 (𝑖𝑖)
= 2(𝑝 + 𝑞)
= 2𝑟, 𝑓𝑜𝑟 𝑟 = 𝑝 + 𝑞 ∈ 𝑍, 𝑓𝑜𝑟 𝑎𝑛𝑦 𝑝, 𝑞 ∈ 𝑍
𝐻𝑒𝑛𝑐𝑒 𝑥 + 𝑦 𝑖𝑠 𝑒𝑣𝑒𝑛 𝑏𝑦 𝑑𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛
[Q. E. D]
In some text, you may find the acronym 𝑄. 𝐸. 𝐷 at the end of a proof. Some people will tell you
that 𝑄. 𝐸. 𝐷 stands for “𝑞𝑢𝑖𝑡𝑒 𝑒𝑎𝑠𝑖𝑙𝑦 𝑑𝑜𝑛𝑒”. Sorry, but that’s not it. This is stands for the Latin
words “𝑞𝑢𝑜𝑑 𝑒𝑟𝑎𝑡 𝑑𝑒𝑚𝑜𝑛𝑠𝑡𝑟𝑎𝑛𝑑𝑢𝑚”, which translates to English as
“𝑡ℎ𝑖𝑠 𝑖𝑠 𝑤ℎ𝑎𝑡 𝑤𝑒 𝑛𝑒𝑒𝑑𝑒𝑑 𝑡𝑜 𝑠ℎ𝑜𝑤”
QuickQuiz!
Prove that if the sum of any two integers is even, then so is their difference.
In some instances, the proof comes by way or reorganising the given expression into a form
whose property can be used for a conclusion. The example below dwells on the property that
if the derivative of a function 𝑓 is non-decreasing, in the given domain of 𝑥 , then 𝑓(𝑥) > 0,
in that domain.
Example 2.4
Prove that 𝑙𝑜𝑔(𝑥) ≤ 𝑥 for all 𝑥 ≥ 1
Solution
Take the function 𝑔(𝑥) = 𝑥 – 𝑙𝑜𝑔(𝑥)
1
 𝑔’(𝑥) = 1 − 𝑥
 𝑔′(𝑥) ≥ 0, for all 𝑥 ≥ 1
 𝑔(𝑥) ≥ 0 for all 𝑥 ≥ 1, since 𝑔(0) = 0
 𝑥 – 𝑙𝑜𝑔(𝑥) ≥ 0
 𝑥 ≥ 𝑙𝑜𝑔(𝑥)

Exercise 2.1
1. Let 𝑥 be a positive integer. Prove that 𝑥 + 1/𝑥 ≥ 2. [Hint: reduce this inequality to
the form 𝐴2 ≥ 0, for some expression 𝐴 in 𝑥]
2. Show that 𝑠𝑖𝑛(𝑥) ≤ 𝑥 for all 𝑥 ≥ 0. [Hint: −1 ≤ 𝑐𝑜𝑠(𝑥) ≤ 1]
Before we look at the next example, let us define what a rational number is.
Definition 2.1
A real number 𝑟 is rational, if and only if, 𝑟 = 𝑎/𝑏 for some integers a and b with 𝑏 ≠ 0. A
real number that is not rational is irrational. More formally, if 𝑟 is a real number then
𝑟 𝑖𝑠 𝑟𝑎𝑡𝑖𝑜𝑛𝑎𝑙   𝑖𝑛𝑡𝑒𝑔𝑒𝑟𝑠 𝑎 𝑎𝑛𝑑 𝑏 𝑠𝑢𝑐ℎ 𝑡ℎ𝑎𝑡 𝑟 = 𝑎/𝑏 𝑎𝑛𝑑 𝑏 ≠ 0
Note that the word rational is derived from the word ration, which is another word for
quotient. A rational number is a fraction or ratio of integers.
It should be easy for you to show that every integer is a rational number. However lets prove
the following in the example below.
Example 2.5
Prove that the sum of two rational numbers is rational.
Proof
Let 𝑝 and 𝑞 be rational numbers.
This means that there exists integers 𝑎, 𝑏, 𝑐, and 𝑑, with 𝑏 and 𝑑 not equal to 0 𝑠. 𝑡.
𝑎
𝑝 = … . (𝑖)
𝑏
𝑐
𝑞 = , … . (𝑖𝑖)
𝑑
Then
𝑎 𝑐
𝑝 + 𝑞 = + , 𝑓𝑟𝑜𝑚 (𝑖)𝑎𝑛𝑑 (𝑖𝑖)
𝑏 𝑑
= (𝑎𝑑 + 𝑏𝑐)/𝑏𝑑
 𝑟
= , 𝑟, 𝑠 ∈ 𝑍, 𝑤𝑖𝑡ℎ 𝑠 = 𝑏𝑑 ≠ 0
𝑠
𝐻𝑒𝑛𝑐𝑒 𝑝 + 𝑞 𝑖𝑠 𝑟𝑎𝑡𝑖𝑜𝑛𝑎𝑙 𝑏𝑦 𝑑𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛. [𝑄. 𝐸. 𝐷]
Example 2.5
Show that if an integer is divisible by 2, then its square is divisible by 2
Proof
Let a ∈ 𝑍 𝑎𝑛𝑑 𝑎 𝑏𝑒 𝑑𝑖𝑣𝑖𝑠𝑖𝑏𝑙𝑒 𝑏𝑦 2
Then 𝑎 = 2𝑘, for some 𝑘 ∈ 𝑍
So
𝑎2 = (2𝑘)2
= 2×2× 𝑘 2
= 2𝑝, 𝑝 = 2𝑘 2 ∈ 𝑍
𝐻𝑒𝑛𝑐𝑒 𝑎2 𝑖𝑠 𝑑𝑖𝑣𝑖𝑠𝑖𝑏𝑙𝑒 𝑏𝑦 2. [𝑄. 𝐸. 𝐷]
Exercise 2.1
1. Using direct methods to
a. Prove that if the sum of two integers is odd, so is their difference.
b. Show that if the product of integers a and b is odd then both a and b are odd
2. Let 𝑎, 𝑏 and 𝑐 be sides of a right angled triangle with 𝑐 being the hypotenuse. Prove
2
that if the area of the triangle is 𝑐 ⁄4 then the triangle is isosceles.
[Hint: show that a = b, using the Pythagoras theorem]
1.5.2. Indirect proof
In direct proof we saw that to prove the statement “𝑃 implies 𝑄”, we assumed 𝑃 is true and
use it to draw 𝑄. That is, we started with the hypothesis of a statement and made one deduction
after another until you reach the conclusion. Indirect proofs, are kind of a roundabout. It
provides an approach where to prove a statement 𝑆, we prove some statement 𝑆’ that are related
to 𝑆. The result of which will imply that 𝑆 is correct. Two flavours of indirect proof are
presented here. Proof by contradiction ad proof by contraposition.
1.5.2.1. Proof by contradiction
Proof by contradiction starts by assuming that the given statement, 𝑆, to be proved is not true.
That is our 𝑆’ here is simply 𝑁𝑜𝑡 𝑆. Then we go further to establish that assumption that 𝑆 is
not true, leads to a known contradiction. Therefore, we conclude that our assumption 𝑆’ was
wrong and hence 𝑆 is correct.
This method of proof is also known as 𝑟𝑒𝑑𝑢𝑐𝑡𝑖𝑜 𝑎𝑑 𝑖𝑚𝑝𝑜𝑠𝑠𝑖𝑏𝑖𝑙𝑒 or 𝑟𝑒𝑑𝑢𝑐𝑡𝑖𝑜 𝑎𝑑 𝑎𝑏𝑠𝑢𝑟𝑑𝑢𝑚
because it relies on reducing a given assumption to an impossibility or absurdity.
For instance, suppose you are to prove that you did not commit a given crime that you are
unfortunately charged with, e.g. theft. You can then proceed and ask your accusers to suppose
that you actually committed the crime. Then you ask them what time the crime was committed,
to which they may say sometime between 10 and 11 hours of a given day. At that mentioned
time, you were a 1,000 miles away from the scene attending a workshop and there are people
that can attest to that because they were with you. In fact at that exact time you were in front
presenting. Therefore it is totally absurd to assume that you committed the crime. Therefore
you are innocent. Voila! Sounds simple, but that is exactly the concept behind proof by
contradiction.
1.5.2.1.1. Steps in proving by contradiction
Say you have a statement 𝑆 to prove by contradiction. Follow the steps described below:
Method of Proof by Contradiction

STEP 1:. Make a supposition that the 𝑆 is false, i.e. take 𝑆’ (= 𝑁𝑜𝑡 𝑆) to be true
STEP 2:. Show that this supposition 𝑆’ leads logically to a contradiction
STEP 3:. Conclude, therefore, that the statement 𝑆 is true
The first step, making a supposition is a crucial one. It is important that you get it right first
time, otherwise, it may give wrong results. Note that supposing a statement is false, you simply
negate it.
Example 2.6
All integers are rational would negate to not all integers are rational
Equivalently, there exist integers which are not rational.
𝑆 = ∀𝑥 𝑍(𝑥) → 𝑄(𝑥), 𝑆’ = ∃𝑥 𝑍(𝑥)𝑎𝑛𝑑 𝑛𝑜𝑡 𝑄(𝑥)
There are no clear-cut rules on when to use direct proof or proof by contradiction.
Example. 2.7
Let us prove the sum of two even integers is even
Proof
Let us put some facts ahead of us.
 The sum of two integers is an integer.
 Any integer is either even or odd
𝑆𝑡𝑒𝑝 𝐼: 𝑆𝑢𝑝𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛
There exist two even integers 𝑥 𝑎𝑛𝑑 𝑦, 𝑠. 𝑡. 𝑥 + 𝑦 𝑖𝑠 𝑛𝑜𝑡 𝑒𝑣𝑒𝑛.
Equivalently
There exist two even integers 𝑥 𝑎𝑛𝑑 𝑦, 𝑠. 𝑡. 𝑥 + 𝑦 𝑖𝑠 𝑜𝑑𝑑
We have integers, 𝑝, 𝑞 and 𝑟 𝑠. 𝑡.
𝑥 = 2𝑝, … (𝑖)
𝑦 = 2𝑞, … (𝑖𝑖)
 𝑥 + 𝑦 = 2𝑟 + 1, … (𝑖𝑖𝑖)
𝑆𝑡𝑒𝑝 𝐼𝐼:
From (𝑖) and (𝑖𝑖)
𝑥 + 𝑦 = 2𝑝 + 2𝑞
= 2(𝑝 + 𝑞)
= 2𝑠, 𝑠 𝑍, … (𝑖𝑣)
From (𝑖𝑖) and (𝑖𝑣)

2𝑟 + 1 = 2𝑠, … (𝑣)
However, there does not exist 2 integers, 𝑟 and 𝑠 , for which (𝑣) holds. This is a
contradiction.
𝑆𝑡𝑒𝑝 𝐼𝐼𝐼: 𝐶𝑜𝑛𝑐𝑙𝑢𝑠𝑖𝑜𝑛
Since the assumption that there exists two even integers whose sum is odd leads to a
contradiction, it simply means that the theorem is correct. [𝑄. 𝐸. 𝐷]

Example 2.8
Prove that √2 is irrational
Proof
We use proof by contradiction
If we suppose that the statement is false, we are simply saying that
√2 𝑖𝑠 𝑟𝑎𝑡𝑖𝑜𝑛𝑎𝑙. . . . (𝑖)
I.e.
𝑎
 √2 = 𝑏 , 𝑎, 𝑏 ≠ 0 ∈ 𝑍, 𝑎, 𝑏 ℎ𝑎𝑣𝑒 𝑛𝑜 𝑐𝑜𝑚𝑚𝑜𝑛 𝑓𝑎𝑐𝑡𝑜𝑟𝑠 … . . (𝑖𝑖)
 𝑎 = 𝑏√2, 𝑓𝑟𝑜𝑚 (𝑖𝑖)
 𝑎2 = 2𝑏 2 , … (𝑖𝑖𝑖)
 𝑎2 𝑖𝑠 𝑒𝑣𝑒𝑛
 𝑎 𝑖𝑠 𝑒𝑣𝑒𝑛, 𝑟𝑒𝑓𝑒𝑟 𝑡𝑜 𝑒𝑥𝑎𝑚𝑝𝑙𝑒 2.10 … (𝑖𝑣)
 𝑎 = 2𝑘, 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑘 ∈ 𝑍 … . (𝑣), 𝑐ℎ𝑒𝑐𝑘 𝑡ℎ𝑒 𝑛𝑜𝑡𝑒 𝑏𝑒𝑙𝑜𝑤
 (2𝑘)2 = 2𝑏 2 , 𝑠𝑢𝑏𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑛𝑔 (𝑣)𝑖𝑛𝑡𝑜 (𝑖𝑖𝑖)
 4𝑘 2 = 2𝑏 2
 𝑏 2 = 2𝑘 2
 𝑏 2 𝑖𝑠 𝑒𝑣𝑒𝑛
 𝑏 𝑖𝑠 𝑒𝑣𝑒𝑛 … (𝑣𝑖)
 𝑏 = 2𝑝, 𝑝 ∈ 𝑍, … (𝑣𝑖𝑖)
From (𝑣) and (𝑣𝑖𝑖) we see that 𝑎 and 𝑏 have common factors, contradiction (𝑖𝑖).
Therefore since our supposition leads to a contradiction, we conclude that the theorem is true.
[𝑄. 𝐸. 𝐷]
Note: You may be wondering where conclusions (𝑣) and (𝑣𝑖) are coming from. Your worry is
warranted. But it comes from a proposition which states that for any integer 𝑥, if 𝑥 2 is even,
then 𝑥 is also even. We shall provide proof later, although intuition should suggest to you that
this is the case for now.

Example 2.9
Show that the set of primes is infinite
Proof
Suppose not. Then there exist the greatest prime number 𝑝𝑛 . I.e. the set of primes is the
following finite set in which each 𝑝𝑖 is a prime number {𝑝1 , 𝑝2 , … , 𝑝𝑛−1 , 𝑝𝑛 }.
The fundamental principle of Mathematics state that:
𝑒𝑣𝑒𝑟𝑦 𝑛𝑢𝑚𝑏𝑒𝑟 𝑖𝑠 𝑒𝑖𝑡ℎ𝑒𝑟 𝑎 𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑜𝑓 𝑝𝑟𝑖𝑚𝑒𝑠, 𝑜𝑟 𝑖𝑠 𝑖𝑡𝑠𝑒𝑙𝑓 𝑎 𝑝𝑟𝑖𝑚𝑒, … (𝑖)
Consider the number 𝑁 = 𝑝1 × 𝑝2 , … , 𝑝𝑛−1 × 𝑝𝑛 + 1,
𝑁 is not prime since it is greater than 𝑝𝑛 the greatest prime.
Further 𝑁 is not a product of primes, since each prime 𝑝𝑖 , leaves a remainder of 1 when divided
into 𝑁.
Since our assumption that the set of primes is finite leads to the existence of a number 𝑁, which
is neither prime, nor a product of primes, contradicting (𝑖), we conclude that the theorem is
correct
1.5.2.2. Proof by contraposition
To prove that a statement is true using contraposition, you need to prove that the contraposition
of the statement is true. This sounds like a circular jargon. To restore order, we will first define
what a contraposition of a statement is.
Definition 2.2
Suppose you have a statement 𝑆 which is expressed as 𝑃 → 𝑄. Then the contraposition of 𝑆
is the statement 𝑛𝑜𝑡 𝑄 → 𝑛𝑜𝑡 𝑃
Consider the following statement.
"𝐼𝑓 𝑖𝑡 𝑟𝑎𝑖𝑛𝑠, 𝑡ℎ𝑒 𝑔𝑟𝑎𝑠𝑠 𝑤𝑖𝑙𝑙 𝑏𝑒 𝑤𝑒𝑡"
This statement says that, if you notice that it is raining, then you conclude that the grass is wet.
Equivalently, if you find that the grass is not wet, conclude that it did not rain.
The two statements are actually equivalent. Proving one is as good as proving the other.
1.5.2.2.1. Method of proving by contraposition
To prove that the statement 𝑆 (𝑜𝑓 𝑓𝑜𝑟𝑚 𝑃 → 𝑄 ) to be true using contraposition, follow the
following steps:
STEP 1:. Generate 𝑆’ the contraposition of 𝑆

STEP 2:. Use direct proof on 𝑆’: assume ′𝑛𝑜𝑡 𝑄′ and prove ′𝑛𝑜𝑡 𝑃′

STEP 3:. Conclude that 𝑆 is true

Example 2.10
Prove that for any given integer 𝑥, if 𝑥 2 𝑖𝑠 𝑒𝑣𝑒𝑛, 𝑡ℎ𝑒𝑛 𝑥 𝑖𝑠 𝑒𝑣𝑒𝑛
Proof
Show that 𝑖𝑓 𝑥 𝑖𝑠 𝑜𝑑𝑑, 𝑡ℎ𝑒𝑛 𝑥 2 𝑖𝑠 𝑜𝑑𝑑
Premise
𝑥 𝑖𝑠 𝑜𝑑𝑑 … (𝑖)
 𝑥 = 2𝑝 + 1, 𝑝 ∈ 𝑍, 𝑏𝑦 𝑑𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛 𝑜𝑓 (𝑖)
 𝑥 2 = (2𝑝 + 1)2
 (2𝑝 + 1)(2𝑝 + 1)
 4𝑝2 + 4𝑝 + 1
 2(2𝑝2 + 2𝑝) + 1
 2𝑞 + 1, 𝑞 ∈ 𝑍
∴ 𝑝2 𝑖𝑠 𝑜𝑑𝑑, 𝑏𝑦 𝑑𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛
Since the contrapostion of 𝑆 holds, 𝑆 is also true.
QuickQuiz!
Prove the proposition above using contradiction.
Let us look at the even sum of even integers theorem.
𝑎 is even and 𝑏 is even implies 𝑎 + 𝑏 is even
The contraposition will be that
𝑎 + 𝑏 is odd implies that between 𝑎 and 𝑏, one is odd another is even
Exercise 2.3
1. Use proof by contraposition to show that the of even integers is even also.
2. Using proof by contradiciction and proof by contraposition, show
a. if the sum of two integers is odd, so is their difference.
b. if the product of integers a and b is odd then both a and b are odd
 3. Let 𝑎, 𝑏 and 𝑐 be sides of a right angled triangle with 𝑐 being the hypotenuse. Prove
2
using contradiction that if the area of the triangle is 𝑐 ⁄4, then the triangle is
isosceles.
1.5.3. Disproving by counterexample
Suppose you are given a statement which claims that “All men are strong”. To disprove that
statement just pick one of the men you know is not strong and prove it, then the statement is
rendered invalid. This is called disproving by counterexample and the method can be written
as follows:
To disprove a statement of the form "∀𝑥 ∈ 𝐷, 𝑖𝑓 𝑃(𝑥)𝑡ℎ𝑒𝑛 𝑄(𝑥), find a value 𝑥 in 𝐷 for which
𝑃(𝑥) is true, but 𝑄(𝑥) is false. Such an 𝑥 is called a counterexample.
Example 2.11
Disprove that for any pair of real numbers 𝑎 and 𝑏, 𝑖𝑓 𝑎2 = 𝑏 2 𝑡ℎ𝑒𝑛 𝑎 = 𝑏
Proof
To disprove this statement, find a pair of real number a and b such that 𝑎2 = 𝑏 2 𝑏𝑢𝑡 𝑎 ≠ 𝑏
Let 𝑎 = 1 and 𝑏 = −1, 𝑐𝑙𝑒𝑎𝑟𝑙𝑦 𝑎2 = 𝑏 2 𝑏𝑢𝑡 𝑎 ≠ 𝑏, since (−1)2 = 12 = 1, 𝑏𝑢𝑡 − 1 ≠ 1
1.5.4. The Principle of Mathematical Induction (PMI)
The principle of mathematical induction is a fairly new method. It is usually used to prove that
a defined pattern holds over a sequence of numbers, usually infinite. These statements usually
come in the form of ∀ 𝑛 ∈ 𝑍: 𝑛 > 𝑎, 𝑄(𝑛) ℎ𝑜𝑙𝑑𝑠, where 𝑄(𝑛) is some property and 𝑎 is the
least value for which 𝑄 holds. For example:
The sum of the first 𝑛 odd positive integers is equal to the square of 𝑛.
n Summation Value
1 1 1
2 1+3 4
3 1+3+5 9
4 1+3+5+7 16
:
n 1+3+…2n-1 n^2

This is more like the method of exhaustion, but we just adopted the pattern after the first four
sums of odd numbers. We need to be sure that this is true for each 𝑛 in the domain. The
principle of mathematical induction is the tool for this task.
To proceed with PMI, we will, first of all, show that this assertion works for the first value
of 𝑛. i.e. 𝑛 = 1. For this to be done, lets rewrite the theorem
𝑛

∑(2𝑖 − 1) = 𝑛2 , 𝑓𝑜𝑟 𝑛 = 1, 2, …,
𝑖=1

To show that this is true, we replace 𝑛 = 1 on both the left and right hand side of (𝑖)
𝑙ℎ𝑠 = ∑1𝑖=1(2𝑖 − 1) = 1 𝑟ℎ𝑠 = 12 = 1
Since 𝑙ℎ𝑠 = 𝑟ℎ𝑠, we conclude that (𝑖) holds for 𝑛 = 1
Unlike the method of exhaustion, we will not enumerate all the cases as these are infinite.
However, we will make an assumption that this method works for all values of 𝑛  {1,2, … , 𝑘}.
This means that with this assumption, we can add to the premise
𝑘

∑(2𝑖 − 1) = 𝑘 2 … … … … [𝐼]
𝑖=1

With that done, we now check whether, this information can help in proving that the statement
under investigation (𝑖), holds for the next term at 𝑛 = 𝑘 + 1, i.e. if it holds for the (𝑘 + 1)𝑡ℎ
summation. Note that the (𝑘 + 1)𝑡ℎ summation is
𝑘+1

∑(2𝑖 − 1) = (𝑘 + 1)2
𝑖=1

And this is what we intend to prove using [𝐼]. We attempt to show that using the assumption
above as follows:
𝑘+1

∑ 2𝑖 − 1 = 1 + 3 + ⋯ + 2𝑘 − 1 + 2(𝑘 + 1) − 1 … [𝐼𝐼]
𝑖=1

Note that the term that is bound in the rectangle is simply the sum of the first 𝑘 odd integers.
We substitute [𝐼] into [𝐼𝐼], to get
𝑘+1 𝑘

∑ 2𝑖 − 1 = ∑(2𝑖 − 1) + 2(𝑘 + 1) − 1
𝑖=1 𝑖=1

= 𝑘 2 + 2(𝑘 + 1) − 1
= 𝑘 2 + 2𝑘 + 2 − 1
= 𝑘 2 + 2𝑘 + 1
= 𝑘2 + 𝑘 + 𝑘 + 1
= 𝑘(𝑘 + 1) + 1(𝑘 + 1)
= (𝑘 + 1)(𝑘 + 1)
= (𝑘 + 1)2
[𝑄. 𝐸. 𝐷]
Since the assumption at 𝑘 prove the summation at 𝑘 + 1, and 𝑘 is arbitrary, it means 𝑘 + 1
will prove for 𝑘 + 2, and 𝑘 + 2 for 𝑘 + 3 etc. to infinite. This means that the statement, is
true for all values of 𝑛 = 1, 2, 3, ….
Let us formally state the principle of mathematical induction.
Principle of Mathematical Induction
Let 𝑃(𝑛) be a property that is defined for integers n and let 𝑎 be a fixed integer. Suppose the
following statements are true
𝑃(𝑎) is true
For all integers 𝑘 > 𝑎, if 𝑃(𝑘) is true implies 𝑃(𝑘 + 1) is true
then the statement
For all 𝑛 ≥ 𝑎, 𝑃(𝑛)
is true
It might take a while to simmer down, but we simply formally stated the procedure we
performed earlier.
From above we see that to complete the process of proving by mathematical induction, we go
through 3 major steps.
STEP 1:. Test for the base case when 𝑛 = 𝑎
STEP 2:. Assume that the statement holds for an arbitrary number 𝑛 = 𝑘. This
assumption is referred to as the 𝒊𝒏𝒅𝒖𝒄𝒕𝒊𝒗𝒆 𝒉𝒚𝒑𝒐𝒕𝒉𝒆𝒔𝒊𝒔

STEP 3:. The last step is to use the inductive hypothesis to prove that the statement
works for 𝑛 = 𝑘 + 1, this is referred to as the 𝒊𝒏𝒅𝒖𝒄𝒕𝒊𝒗𝒆 𝒔𝒕𝒆𝒑.

The handling of step 3, showing that the inductive hypothesis works for n = k+1 is a critical
step and different problems require different approaches to achieve this step.
Example 2.12
Prove that if a set 𝐴 has n elements, then the cardinality of the powerset of A is 2𝑛 .
Solution
We can use the principle of mathematical induction for 𝑛 = 1, 2, …
Step I: test for 𝑛 = 1, A is a singleton set.
We saw earlier that a singleton set has 2 subsets. The empty set and itself and we see that
21 = 2
Step II: Inductive hypothesis, assume true for 𝑛 = 𝑘
𝐼𝑓 |𝐴| = 𝑘, |𝑃(𝐴)| = 2𝑘 … … (𝐼)
Step III: Inductive step. Prove for 𝑛 = 𝑘 + 1
Let 𝐵 = 𝐴 ∪ {𝑎} then |𝐵| = 𝑘 + 1
All the 2𝑘 subsets of 𝐴, will be subsets of the new set 𝐵, and to each one of these subsets, the
element 𝑎 is added, to get 2𝑘 more subsets of 𝐵
 𝑆𝑜 |𝑃(𝐵)| = 2𝑘 + 2𝑘
= 2. 2𝑘
= 2𝑘+1
[𝑄. 𝐸. 𝐷]
Example 2.13
Let 𝑎, 𝑏 and 𝑛 be integers with n ≠ 0. If 𝑎 ≡ 𝑏(𝑚𝑜𝑑 𝑛)𝑡ℎ𝑒𝑛
𝑎𝑝 ≡ 𝑏 𝑝 (𝑚𝑜𝑑 𝑛), 𝑓𝑜𝑟 𝑝 = 1, 2, …
Proof
We use the PMI. Since n has been used, the variable of induction is p in this case.
STEP I: test for base case, 𝑝 = 1,
This is a trivial matter as it is given that
𝑎 ≡ 𝑏(𝑚𝑜𝑑 𝑛)
𝑖. 𝑒. 𝑏 – 𝑎 = 𝑛𝑔, 𝑔  𝑍 … [𝐼]
Step II: Inductive hypothesis
𝑎𝑘 ≡ 𝑏 𝑘 (𝑚𝑜𝑑 𝑛)
𝑖. 𝑒. 𝑏 𝑘 − 𝑎𝑘 = 𝑛𝑞, 𝑞 ∈ 𝑍 … … . . [𝐼𝐼]
Step III: Prove for 𝑝 = 𝑘 + 1
Note that we need to show that 𝑏 𝑘+1 − 𝑎𝑘+1 = 𝑛𝑟, 𝑓𝑜𝑟 𝑟 ∈ 𝑍
Note that 𝑏 𝑘 − 𝑎𝑘 = 𝑛𝑞, 𝑞 ∈ 𝑍, 𝑓𝑟𝑜𝑚 [𝐼𝐼]
(𝑏 𝑘 − 𝑎𝑘 )(𝑏 + 𝑎) = 𝑏 𝑘+1 + 𝑏 𝑘 . 𝑎 − 𝑎𝑘 . 𝑏 − 𝑎𝑘+1 … [𝐼𝐼𝐼]
 𝑏 𝑘+1 − 𝑎𝑘+1 = (𝑏 𝑘 − 𝑎𝑘 )(𝑏 + 𝑎) − 𝑏 𝑘 . 𝑎 + 𝑎𝑘 . 𝑏
= (𝑏 𝑘 − 𝑎𝑘 )(𝑏 + 𝑎) − 𝑎𝑏(𝑏 𝑘−1 − 𝑎𝑘−1 ), 𝑏𝑦 [𝐼𝐼]
= 𝑛𝑞(𝑏 + 𝑎) − 𝑎𝑏𝑛𝑔
= 𝑛(𝑞(𝑏 + 𝑎) − 𝑎𝑏𝑔)
= 𝑛𝑡, 𝑡 ∈ 𝑍
[𝑄. 𝐸. 𝐷]
Or
𝑏 𝑘+1 − 𝑎𝑘+1 = 𝑏 𝑘+1 − 𝑎𝑏 𝑘 + 𝑎𝑏 𝑘 − 𝑎𝑘+1
= 𝑏(𝑏 𝑘 − 𝑎𝑘 ) + 𝑎𝑘 (𝑏 − 𝑎)
= 𝑏𝑛𝑞 + 𝑎𝑘 𝑛𝑔, 𝑓𝑟𝑜𝑚 [𝐼]𝑎𝑛𝑑 [𝐼𝐼]
= 𝑛(𝑏𝑞 + 𝑎𝑘 𝑔), 𝑛𝑟, 𝑤ℎ𝑒𝑟𝑒 𝑟 = 𝑏𝑞 + 𝑎𝑘 𝑔
Note: that the hypothesis assumes that the statement holds for all 𝑛 = 1, 2, … , 𝑘. This means
that the statement holds for 𝑛 = 𝑘 – 1 as well hence 𝑏 𝑘−1 − 𝑎𝑘−1 = 𝑛𝑔, 𝑔 ∈ 𝑍 is a valid
argument.
Notice how 𝑠𝑡𝑒𝑝 𝐼𝐼 has not been quite uniform in all the examples we have dealt with. A bit of
practice, will soon let you intuitively know how to use the inductive hypothesis, to head home
with an elegant proof using the principle of mathematical induction.

Example 2.14
Show that the function 𝑇(𝑛) = 5𝑛 + 2 is the solution to the recurrence relation
𝑇(𝑛) = 𝑇(𝑛 − 1) + 5, 𝑤𝑖𝑡ℎ 𝑇(0) = 2
Proof

Step I: test for 𝑛 = 0,

𝑇(0) = 5 × 0 + 2
=0+2
= 2 𝑎𝑠 𝑟𝑒𝑞𝑢𝑖𝑟𝑒𝑑

Step II: Inductive Hypothesis

𝑇(𝑘) = 5𝑘 + 2 … … . [𝐼]

Step III: Prove

𝑇(𝑘 + 1) = 𝑇(𝑘) + 5
= 5𝑘 + 2 + 5, by [𝐼]
= 5𝑘 + 5 + 2
= 5(𝑘 + 1) + 2
[𝑄. 𝐸. 𝐷]
Example 2.15
Prove that 4𝑛 − 1 is divisible by 3 for integers 𝑛 ≥ 1
Proof
Step I: test for base case, n = 1
41 − 1
=4– 1
=3
𝑤ℎ𝑖𝑐ℎ 𝑖𝑠 𝑑𝑖𝑣𝑖𝑠𝑖𝑏𝑙𝑒 𝑏𝑦 3
Step II: Inductive Hypothesis
4𝑘 − 1 = 3𝑟, 𝑟 ∈ 𝑍 … … [𝐼]
Step III:
 (4𝑛 − 1)(4 + 1)
= 4𝑛+1 + 4𝑛 − 4 − 1
= 4𝑛+1 − 1 + 4𝑛 − 4

𝑆𝑜 (4𝑛 − 1)(4 + 1) = 4𝑛+1 − 1 + 4𝑛 − 4

 4𝑛+1 − 1 = (4𝑛 − 1)(4 + 1) − 4𝑛 + 4
= 5(4𝑛 − 1) − 4(4𝑛−1 − 1), 𝑏𝑦 [𝐼]
= 5.3𝑟 − 4.3𝑞
= 3(5𝑟 − 4𝑞)
= 3𝑡, ∈ 𝑍
[𝑄. 𝐸. 𝐷]

Or
4𝑛+1 − 1 = 4𝑛 . 4 − 1
= 4𝑛 (3 + 1) − 1
= 3 × 4𝑛 + (4𝑛 − 1)
= 3 × 4𝑛 + 3𝑟, 𝑏𝑦 [𝐼]
= 3(4𝑛 + 𝑟)
= 3𝑞, 𝑞 ∈ 𝑍
[𝑄. 𝐸. 𝐷]
Exercise 2.4
Use the principle of mathematical induction to show that
𝑛
𝑎𝑛+1 − 1
∑ 𝑎𝑖 =
𝑎−1
𝑖=0

1.6. Summary
In this unit, we discussed the different methods of proving that a Mathematical statement is
true. These include, direct methods in which the premise is used to draw the conclusion and
indirect methods, in which a statement related to the statement to be proved is derived, and
based on the outcome of processing the statement, the original statement is proved. We also
looked at disproving a statement by using a counterexample.
The principle of mathematical induction was also discussed ad the various ways in which in
which it can be applied were also given. Before we wrap up this section, it’s time to check your
comprehension of these concepts, by attempting the exercises given below.
1.7. Unit 2 Exercise Set
1) Prove that if the sum of two numbers is even, then so is their difference.[Hint: Use the fact
that 𝑎 + 𝑏 = 2𝑛, to show that 𝑎 – 𝑏 is also expressed in the form of 2m, for some integer
m]

2) Use the proof by contradiction to show that 1 + 3√2 is irrational. [Hint: Assume that 1 +
3√2 is rational and show that this results in √2 being rational, which, is a contradiction]

3) Prove that the sum of a rational number and an irrational number is irrational

(𝑥+𝑦)
4) Using direct methods show that if 𝑥 and 𝑦 are positive integers, then √𝑥𝑦 ≤ . [Hint:
2
get rid of the radical and rearrange the terms]

5) Give the negation of the following

a) A person who doesn’t share is a witch
b) There is no greatest even integer
c) Both 𝑎 and 𝑏 are even
d) 𝑎 + 𝑏 is odd

6) Prove using contraposition that the sum of two even integers is also even.

7) Prove or give counterexamples

a) The difference of any two odd numbers is odd
b) The product of an even integer and any integer is even
c) The sum of an even integer and any integer is even
d) For all real numbers a and b, if 𝑎 < 𝑏, then 𝑎2 < 𝑏 2

𝑛(𝑛+1)
8) Prove that the sum of the first n positive integers is 2

9) Prove that 𝑛3 − 𝑛 is divisible by 6 for all 𝑛 ≥ 1

10) Prove that for any integer 𝑛 ≥ 1, 𝑥 𝑛 − 𝑦 𝑛 is divisible by 𝑥 – 𝑦 where 𝑥 and 𝑦 are any
integers with 𝑥 ≠ 𝑦. [Borrow some ideas from the congruence method in example 2.13]

11) Prove that 1 + 𝑛𝑥 ≤ (1 − 𝑥)𝑛 for all real numbers 𝑥 ≥ −1 and 𝑛 ≥ 2

12) At a party, as each party goer arrives each shakes hands with all other people present. If n
𝑛(𝑛−1)
people attend the party, prove that [ ] handshakes occur. [refer to exercise 8 above]
2

13) Derive a formula for finding the sum of he first 𝑛 even numbers and use the principle of
mathematical induction to show that your formula is correct. [Generate the fisrt few terms
and draw the pattern]