1 Binary relations

Let S be a set. A binary relation < on S is any subset of S S: One usually

writes x<y when (x; y) 2 and we say that x is in relation with y. is called
the graph of <.
Intuitively speaking a relation < on a set S is a property that concerns
couples (x; y) of elements from S. Let < be a binary relation on S.

< is said to be re‡exive if for any x 2 S, x<x.

< is said to be symmetric if for any x; y 2 S, x<y then y<x.
< is said to be antisymmetric if for any x; y 2 S; x<y and y<x then x = y.
< is said to be transitive if for any x; y; z 2 S; x<y and y<z then x<z.

Example 1

1. Perpendicularity is a binary relation between the lines of a plane which is

symmetric, not re‡exive and not transitive.
2. Parallelism is a binary relation between the lines of a plane which is re-
‡exive, symmetric and transitive.
3. Equality between the elements of any set is a binary relation which is
re‡exive, symmetric, antisymmetric and transitive.
4. Let S be a set. On P(S) the relation < de…ned by: 8A; B 2 P(S); A<B
if and only if A = B or A = B is re‡exive, symmetric and transitive.
5. On R; the binary relation < de…ned by: 8x; y 2 R; x<y if and only if
x 2y is neither symmetric nor antisymmetric.
6. On R; the relation < de…ned by: x<y if and only if x2 = y 2 is symmetric
but not antisymmetric.

I Note that the notion of symmetric and antisymmetric are not opposite
(are not contrary to each other).

– Equality is in the same time symmetric and antisymmetric.

8
< < is antisymmetric
– We have and =) < non symmetric.
:
9x0 ; y0 2 E; x0 6= y0 ; x0 Ry0

1
1.1 Equivalence relation
Let S be a set and < be a binary relation on S:
< is said to be an equivalence relation if it satis…es the following conditions:

< is re‡exive.
< is symmetric.
< is transitive.

Example 2

1. Equality is an equivalence relation on any set S.

2. Parallelism is an equivalence relation on the set of lines of a plane.
3. On N0 Z ; the relation de…ned by 8 (a; b) ; (c; d) 2 N0 Z : (a; b) <(c; d) ,
a + b = c + d is an equivalence relation.
4. On R; the relation < de…ned as x<y if and only if x2 y2 = x y is an
equivalence relation.

1.1.1 Equivalence classes and quotient set

Let < be an equivalence relation on S and let x 2 S. The set of all elements
y 2 S such that x<y is called the equivalence class of x and is denoted by x or
Cl< (x): That is, x = fy 2 S : y<xg :

Proposition 3 Let < be an equivalence relation on a set S. Every equivalence

class is nonempty, and S is the disjoint union of the equivalence classes of <.

Proof. Let x 2 S. Then the re‡exivity of < implies that x 2 x. Therefore every
equivalence class is nonempty and the union of the equivalence classes is S. To
prove that the union is disjoint, we show that for any x; y 2 S either x \ y = ;
(if x is not in relation with y) or x = y (if x<y). Suppose that x \ y 6= ;.
0 0
Let z 2 x \ y be an element in both equivalence classes. If x 2 x, then x <z
0 0
and z<y, so x <y by the transitivity of the relation <, and therefore x 2 y. It
follows that x y. A similar reasoning implies that y x, and therefore x = y.

The set of all equivalence classes fx; x 2 Sg is called the quotient set asso-
ciated to the relation < and denoted by S=<: That is, S=< = fx; x 2 Sg :
Conversely, let P be a partition of a set S, then there exists a unique equiv-
alence relation <P such that the quotient set is P . Indeed, let P be a partition
of S: We de…ne <P in S as follows

8x; y 2 S; x<p y if and only if 9A 2 P such that x, y 2 A

(in other words, x; y are in relation if and only if they are lying in the same
subset belonging to P ).

2
 <P is re‡exive: Because P is a partition then for any x 2 S there must exists
A 2 P such that x 2 A therefore x<P x.
<P is symmetric: Let x; y 2 S such that x<P y: Then, x; y 2 A for some
A 2 P so y; x 2 A hence y<P x.
<P is transitive: Let x; y; z 2 S such that x<P y and y<P z: This gives 9A1
2 P; x; y 2 A1 and 9A2 2 P; y; z 2 A2 : Here we have y 2 A1 \A2 ; but A1 ; A2 are
elements of the partition P; therefore A1 = A2 and since x 2 A1 and A1 = A2
we have x 2 A2 ; so x; z 2 A2 and x<P z. This shows that <P is an equivalence
relation and x = fy 2 S; x<P yg = A So S=<P = P: It is obvious that <P is the
unique equivalence relation S such that P = S=<P :

Example 4

1. Let n be a positive integer, and consider the equivalence relation < de…ned
on Z by:

8x; y 2 Z; x<y if and only if 9k 2 Z such that x y = nk:

Then, the quotient set is Z=< = 0; 1; :::; n 1 which we denote usually

by Z=nZ.
2. Let S = R2 : The relation < on R2 de…ned by: (x1 ; y1 )<(x2 ; y2 ) if and
only
p if k(x1 ; y1 )k = k(x2 ; y2 )k is an equivalence relation where k(x; y)k =
x2 + y 2 : Here R2 =< = fcircles centred at (0; 0)g :
3. Let S = R2 :The relation < on R2 de…ned by: (x1 ; y1 )<(x2 ; y2 ) if and only
if x1 = x2 is an equivalence relation. Here R2 =< = fVertical linesg :

1.2 Partial and total order

Let S be a set and < be a binary relation on S: < is said to be a partial order
relation if it satis…es the following conditions:

1. < is re‡exive.
2. < is antisymmetric.
3. < is transitive.

A set equipped with a partial order relation is called ordered set and often
we use to denote a partial order relation. A binary relation < on a set S is
said to be a total order if:

1. < is a partial order.

2. Any x; y 2 S are comparable. That is, for any (x; y) 2 S 2 ; x<y or y<x:

Example 5

1. Equality is a total order on any set S.

3
 2. The relation less or equal " " is a total order on R:
3. The inclusion on P(S) is a partial order relation for any set S .
4. The relation < on N given by: a<b if and only if b = 2k a for some integer
k 0 is a partial order.
5. On R R; the relation < de…ned by: (a; b)<(x; y) if and only if a x and
b y is a partial order.

There are only one relation which is in the same time an equivalence relation
and a partial order. This is the equality. Obviously as seen above the equality
is an equivalence and a partial order. Prove that it is the only one. Indeed, We
must prove
8x; y 2 S; x<y () x = y:
Let x; y 2 S. x<y gives y<x because < is symmetric then we have x = y
because < is antisymmetric. Conversely, Suppose x = y. Since < is re‡exive we
have x<x or x<y since y = x:

1.2.1 Remarkable elements in an ordered set

Let X be an ordered set, and let Y X.

An element a of Y is called a maximal element of Y if

(8x 2 Y ); (a x =) x = a):

In other words, there is no element x of Y such that a x and a 6= x:

An element a of Y is called a minimal element of Y if

(8x 2 Y ); (x a =) x = a):

In other words, there is no element x of Y such that x a and a 6= x:

An element a of Y is called greatest element (maximum) of Y if

a 2 Y and 8x 2 Y; x a:

The maximum of Y when exists we denote it by max Y:

An element a of Y is called least element (minimum) of Y if

a 2 Y and 8x 2 Y; a x:

The minimum of Y is denoted min Y:

Remark 6 Let X be an ordered set, and let Y X.

1. Y may have several maximal elements.

4
 2. Y may have several minimal elements.

Example 7 Let X = N and x y if and only if x divides y. It is obvious that

it is a partial order.

1. Let Y1 = f2; 4; 8g Here 8 is a maximum and 2 is a minimum.

2. Let Y2 = f2; 4; 3; 5; 6; 8; 10; 12g : Here Y2 has no maximum and no mini-

mum. 8; 10; 12 are maximal elements and 2; 3; 5 are minimal elements.

Example 8 Let S = fa; b; c; d; e; f g, X = P(S); and denotes the inclusion:

1. Let Y1 = ffag ; fbgg :Then Y1 has no maximal element, no minimal ele-

ment and also no maximum and no minimum.
2. Let Y2 = ffag ; fa; bg ; fa; b; cgg we have fag fa; bg fa; b; cg so, 8x 2
Y2 fag x fa; b; cg : Then Y2 has fa; b; cg as a maximum and fag as
a minimum. Y3 = ffag ; fbg ; fa; bg ; fa; b; cg ; fa; cg ; fa; b; eg ; fa; c; egg :
Here Y3 has no maximum and no minimum. But Y3 has three maximal
elements fa; b; cg ; fa; b; eg ; fa; c; eg and two minimal elements fag and
fbg :

Let X be an ordered set, and let Y X.

An element a of X is called upper bound (majorant) of Y if 8x 2 Y; x a:

An element a of X is called lower bound (minorant) of Y if 8x 2 Y; a x:
An element a of X is called supremum of Y if a is the least element of the
set of upper bounds of Y: The supremum of Y when exists we denote it
by sup Y .
An element a is called in…mum of Y if a is the greatest element of the
set of lower bounds of Y . The in…mum of Y when exists we denote it by
inf Y:

Theorem 9 Let X be an ordered set, and let Y X.

1. The maximum of Y , when exists, is unique.

2. The minimum of Y; when exists, is unique.

Proof. Let us prove 1)

1. Let X be an ordered set and Y X. Let a; a0 being two greatest elements

of Y . As a is an upper bound of Y and that a0 2 Y , we have a a0 . in
the same way a0 a. Therefore a = a0 .

5
 It is obvious that if a is a maximum (respectively in…mum) of Y then a is
also a supremum (respectively in…mum).
Conversely, if sup Y 2 Y (respectively if inf Y 2 Y ) then sup Y is also a
maximum (respectively inf Y is also a minimum).

Theorem 10 Let X be an ordered set, and let Y X.

1. The in…mum element of Y; when exists, is unique.

2. The supremum element of Y; when exists, is unique.

Remark 11 Let X be an ordered set, and let Y X.

1. Let a 2 X. We have the following implications: a maximum of Y ) a

supremum of Y ) a upper bound of Y .
2. Let a 2 X. we have the following implications: a minimum element of Y
) a in…mum of Y ) a lower bound of Y .

Indeed, The second implication is clear. Let us prove the …rst assertion: let
a be the maximum of Y . First, a is an upper bound of Y . Let b another upper
bound. as a 2 Y , we have b a. Therefore a is the smallest of the upper
bounds of Y , in other words, a is the supremum of Y:

Example 12

1. Let E = fa; b; c; d; e; f g X = P(E); the relation is the inclusion ; and

let Y = ffa; bg ; fc; dgg ; then sup Y = fa; b; c; dg and inf Y = ;:
2. Let X = N and is the relations divides: Let Y = f2; 3; 5; 6g : Then
sup Y = 30 and inf Y = 1:

Example 13 Let X = R and Y = [0; 2[. So 2 is the supremum of Y (but Y

has no maximum since 2 2
= Y ).