Lattices

• Lattice
A lattice is a poset (L, ≤) in which every subset {a, b}
consisting of two elements has a least upper bound and
a greatest lower bound. we denote
LUB({a, b}) by a b (the join of a and b)
GLB({a, b}) by a b (the meet of a and b)

1
Lattices

• Example 1
Let S be a set and let L=P(S). As we have seen, ,
containment, is a partial order on L. Let A and B belong to
the poset (L, ). Then
a b =A U B & a b = A ∩ B
Why?
Assuming C is a upper bound of {a, b}, then
A C and B C thus A U B C
Assuming C is a lower bound of {a, b}, then
C A and C B thus C A ∩ B

2
Lattices

• Example 2
Consider the poset (Z+, ≤), where for a and b in Z+, a ≤
b if and only if a | b , then
a b = LCM(a,b)
a b = GCD(a,b)

LCM: least common multiple

GCD: greatest common divisor

3
Lattices

• Example 3
Let n be a positive integer and Dn be the set of all
positive divisors of n. Then Dn is a lattice under the
relation of divisibility. For instance,
D20= {1,2,4,5,10,20} D30= {1,2,3,5,6,10,15,20}
30
20

6 10 15
4 10

2 3 5
2 5

1
1
4
Lattices

• Example 4
Which of the Hasse diagrams represent lattices?
f g e
d d e
c
e
c
b c d d b c d
b
b
a f b c
a a
a
a a
d e c d
b

b c

a a
5
Lattices

• Example 6
Let S be a set and L =P(S). Then (L, ) is a lattice, and
its dual lattice is (L, ), where “ ” is “contained in”,
and “ ” is “contains”. Then, in the poset (L, )
join: A B=A∩B ,
meet: A B=A∪B.

6
Lattices

• Theorem 1
If (L1, ≤ ) and (L2, ≤ ) are lattices, then (L, ≤ ) is a lattices,
where L= L1 ×L2, and the partial order ≤ of L is the product
partial order.
Proof: we denote
the join and meet in is L1by 1 and 1
the join and meet in is L2by 2 and 2
We know that L is a poset (Theorem 1 in p.219)
for (a1,b1) and (a2,b2) in L. then
(a1,b1) (a2,b2) = (a1 1 a2, b1 2 b2 ) in L
(a1,b1) (a2,b2) = (a1 1 a2, b1 2 b2 ) in L
7
Lattices

• Example 7
(I1 , I2)

(I1 ,a) (I1 ,b)

I1
I2 (01 ,I2)

× a b
= (01 ,a)
(I1 ,02 )
(01 ,b )

02
01 (01 ,02 )

L1 L2 L1×L2

8
Lattices

• Sublattice
Let (L, ≤) be a lattice. A nonempty subset S of L is
called a sublattice of L if a b in S and a b in S
whenever a and b in S
For instance
Example 3 is one of sublattices of Example 2

9
 Lattices

• Example 9
I

I c
e f a b
I
e f
a b 0
e f
a sublattice
c Not a lattice
a b
c
a b
0

0
a lattice, not a Sublattice

10
Lattices

• Isomorphic Lattices
If f: L1L2 is an isomorphism form the poset (L1, ≤1 )
to the poset (L2, ≤ 2) , then L1 is a lattice if and only if
L2 is a lattice. In fact, if a and b are elements of L1, then
f(a b)= f(a) f(b) & f (a b)=f(a) f(b).
If two lattices are isomorphic, as posets, we say they are
isomorphic lattices.

11
Lattices

• Example 10 (P.225 Ex.17)

Let A={1,2,3,6} and let ≤ be the relation | .
Let A’= {ф, {a}, {b}, {a, b}} and let ≤’ be set containment, .
If f(1)= ф, f(2)={a}, f(3)={b}, f(6)={a, b}, then f is an
isomorphism. They have the same Hasse diagrams.
6
{a, b}

2 3 {a} {b}

1 ф

12
Lattices

• a b (LUB{a, b})
1. a ≤ a b and b ≤ a b; a b is an upper bound of a and b
2. If a ≤ c and b ≤ c, then a b ≤ c ; a b is the least upper
bound of a and b

• a b (GLB{a, b})
3. a b ≤ a and a b ≤ b; a b is a lower bound of a and b
4. If c≤ a and c ≤ b, then c≤ a b ; a b is the greatest lower
bound of a and b

13
Lattices

• Theorem 2
Let L be a lattice. Then for every a and b in L
(a) a b =b if and only if a ≤ b
(b) a b =a if and only if a ≤ b
(c) a b =a if and only if a b =b
Proof:
(a) if a b =b, since a≤ a b, thus a ≤ b
if a ≤ b, since b ≤ b , thus b is a upper bound of a and b, by
definition of least upper bound we have a b ≤ b. since a b
is an upper bound of a and b, b ≤ a b, so a b =b
(b) Similar to (a); (c) the proof follows from (a) & (b)
14
Lattices

• Example 12
Let L be a linearly ordered set. If a and b in L, then
either a ≤ b or b ≤ a. It follows form Theorem 2 that L is
a lattice, since every pair of elements has a least upper
bound and a greatest lower bound.

15
Lattices

• Theorem 3
Let L be a lattice. Then
1. Idempotent properties: a a =a ; a a =a

2. Commutative properties: a b= b a; a b = b a

3. Associative properties: (a) (a b) c= a (b c )

(b) (a b) c= a (b c)

4. Absorption properties: (a) a (a b) =a

(b) a (a b) =a

16
Lattices

Proof: 3. (a) (a b) c= a (b c)
a≤ a (b c) & b c≤ a (b c)
b≤ b c & c ≤ b c (definition of LUB)
b≤ b c & c ≤ b c & b c≤ a (b c) 
b≤ a (b c) & c ≤ a (b c) (transitivity)
a≤ a (b c) & b≤ a (b c)  a (b c) is a upper of a and b
then we have a b ≤ a (b c) (why?)
a b ≤ a (b c) & c ≤ a (b c) 
a (b c) is a upper of a b and c
then we have (a b) c ≤ a (b c)
Similarly, a (b c) ≤ (a b) c
Therefore (a b) c = a (b c) (why?)

17
Lattices

• (a b) c = a (b c) = a b c

• (a b) c = a (b c) = a b c

• LUB({a1,a2,…,an})= a1 a2 … an

• GLB({a1,a2,…,an}) =a1 a2 … an

18
Lattices

• Theorem 4
Let L be a lattice. Then, for every a, b and c in L
1. If a ≤ b, then
(a) a c ≤ b c
(b) a c ≤ b c
2. a ≤ c and b ≤ c if and only if a b ≤ c
3. c ≤ a and c ≤ b if and only if c ≤ a b
4. If a ≤b and c ≤d, then
(a) a c ≤ b d
(b) a c ≤ b d
19
Lattices

• Proof
1. (a) If a ≤ b, then a c≤b c
c≤b c; b≤ b c (definition of LUB)
a ≤ b ; b≤ b c  a≤ b c (transitivity)
therefore,
b c is a upper bound of a and c , which means
a c≤b c (why? )

The proofs for others left as exercises.

20
Lattices

• Bounded
A lattice L is said to be bounded if it has a greatest
element I and a least element 0
For instance:
Example 15: The lattice P(S) of all subsets of a set S, with
the relation containment is bounded. The greatest element is
S and the least element is empty set.

Example 13: The lattice Z+ under the partial order of

divisibility is not bounded, since it has a least element 1, but
no greatest element.

21
Lattices

• If L is a bounded lattice, then for all a in A

0 ≤ a ≤ I
a 0 = a, a I=I
a 0=0, a I=a

Note: I (0) and a are comparable, for all a in A.

22
Lattices

• Theorem 5
Let L={a1,a2,…,an} be a finite lattice. Then L is
bounded.
Proof:
The greatest element of L is a1 a2 … an, and the
least element of L is a1 a2 … an

23
Lattices

• Distributive
A lattice L is called distributive if for any elements a, b
and c in L we have the following distributive properties:
1. a (b c) = (a b) (a c)
2. a (b c) = (a b) (a c)
If L is not distributive, we say that L is nondistributive.

Note: the distributive property holds when

a. any two of the elements a, b and c are equal or
b. when any one of the elements is 0 or I.
24
Lattices

• Example 16
For a set S, the lattice P(S) is distributive, since union
and intersection each satisfy the distributive property.
• Example 17
The lattice whose Hasse diagram shown as follows is
distributive. I

b d

a c

0
25
Lattices

• Example 18
Show that the lattices as follows are nondistributive.
I I

c a b c

0
0

a ( b c) = a I = a a ( b c) = a I = a
(a b) (a c) = b 0 = b (a b) (a c) = 0 0 = 0

26
Lattices

• Theorem 6
A lattice L is nondistributive if and only if it contains a
sublattice that is isomorphic to one of the lattices whose
Hasse diagrams are as show.
I I

c a b c

0
0

27
Lattices

• Complement
Let L be bounded lattice with greatest element I and
least element 0, and let a in L. An element a’ in L is
called a complement of a if
a a’ = I and a a’ =0
Note that 0’=I and I’=0

28
Lattices

• Example 19
The lattice L=P(S) is such that every element has a
complement, since if A in L, then its set complement A
has the properties A A = S and A A=ф. That is, the
set complement is also the complement in L.
• Example 20
I I

c a b c
b

0
0

29
Lattices

• Example 21
20 30

4 10 6 10 15

2 5 2 3 5

1 1

D20 D30

30
Lattices

• Theorem 7
Let L be a bounded distributive lattice. If a complement exists, it is
unique.
Proof: Let a’ and a’’ be complements of the element a in L, then
a a’ = I, a a’’= I ; a a’ = 0, a a’’ =0
using the distributive laws, we obtain
a’= a’ 0 = a’ (a a’’ ) = (a’ a) (a’ a’’)
= I (a’ a’’) = a’ a’’
Also
a’’= a’’ 0 = a’’ (a a’ ) = (a’’ a) (a’’ a’)
= I (a’ a’’) = a’ a’’
Hence a’=a’’

31
Lattices

• Complemented
A lattice L is called complemented if it is
bounded and if every element in L has a
complement.

32
Lattices

• Example 22
The lattice L=P(S) is complemented. Observe that in
this case each element of L has a unique complement,
which can be seen directly or is implied by Theorem 7.
• Example 23
I I

c a b c

0
0

33
Finite Boolean Algebras

• Theorem 1
If S1={x1,x2,…,xn} and S2={y1,y2,…,yn} are any two finite
sets with n elements, then the lattices (P(S1), ) and (P(S2),
) are isomorphic. Consequently, the Hasse diagrams of
these lattices may be drawn identically.
Arrange the elements in S1 and S2
A

S1: x1 x2 x3 … xn S1: x1 x2 x3 … xn

f(A)

S2: y1 y2 y3 … yn S2: y1 y2 y3 … yn

34
Finite Boolean Algebras

• Example 1:
S={a, b, c} and T={2,3,5}. Consider the Hasse diagrams of
the two lattices (P(S), ) and (P(T), ).
{a,b,c} {2,3,5}

{a,b} {b,c} {2,3} {3,5}

{a,c} {2,5}

{b} {3}
{a} {c} {2} {5}

ф ф

Note : the lattice depends only on the number of elements in set,

not on the elements.

35
Finite Boolean Algebras

• Label the subsets

Let a set S={a1,a2,…,an}, then P(S) has 2n subsets. We
label subsets by sequences of 0’s and 1’s of length n.
For instance,
{a1,a2}  1 1 0 0 …0
{a1,an }  1 0 0 0 …1
ф  0 0 0 0 …0
{a1,a2,…,an} 1 1 1 1 …1
…

36
Finite Boolean Algebras

• Get the{2,3,5}
unique Hasse Diagram

{2,3} {3,5}
{2,5} 111

{3} 110
{2} {5} 011
101
ф
{a,b,c}

010
{a,b} {b,c} 100 001
{a,c}

000
{b}
{a} {c}
ф
37
Finite Boolean Algebras

• Lattice Bn
If the Hasse diagram of the lattice corresponding to a set with
n elements is labeled by sequences of 0’s and 1’s of length n,
the resulting lattice is named Bn. The properties of the partial
order on Bn can be described directly as follows. If x=a1a2…an
and y=b1b2…bn are two element of Bn, then

1. x ≤ y iff ak ≤ bk (as numbers 0 or 1) for k=1,2,…,n

2. x y=c1c2…cn, where ck= min{ak,bk}
3. x y=c1c2…cn, where ck= max{ak,bk}
4. x has a complement x’=z1z2…zn, where zk=1 if xk=0 and
zk=0 if xk=1
38
Finite Boolean Algebras

• Boolean algebra
A finite lattice is called a Boolean algebra if it is
isomorphic with Bn for some nonnegative integer n.
111

11
|Bn| =2n 1 110 011
101
10 01

010
100 001
00
0
000

n=0 n=1 n=2 n=3

39
Finite Boolean Algebras

• (P(S), )
Each x and y in Bn correspond to subsets A and B of S. Then
x ≤ y, x y, x y and x’ correspond to A B, A ∩ B, A U
B and A. Therefore,
(P(S), ) is isomorphic with Bn, where n=|S|
• Example 3
Consider the lattice D6 consisting of all positive integer
divisors of 6 under the partial order of divisibility.
6 11
D6 is a Boolean algebras
10 01
2 3

1 00
40
Finite Boolean Algebras

• Example 4
Consider the lattices D20 and D30 of all positive integer
divisors of 20 and 30, respectively.
20 30

D20 is not a Boolean algebra

4 10 (why? 6 is not 2n )
6 10 15

2 D30 is a Boolean algebra,

5 2 3 5 D30  B3

1 1

41
Finite Boolean Algebras

• Theorem 2
Let n=p1p2…pk, where the pi are distinct primes. The Dn is a
Boolean algebra.
Proof:
Let S={p1 , p2 , … , pk}. If T S and aT is the product of the
primes in T, then aT | n. Any divisor of n must be of the form aT
for some subset T of S (let aф=1) .
If V and T are subsets of S, V T if and only if aV | aT
aV ∩ T = aV aT = GCD(aV,aT) &
aV U T= aV aT = LCM(aV,aT)
Thus, the function f: P(S) Dn given by f(T)=aT is a isomorphism
form P(S) to Dn. Since P(S) is a Boolean algebra, so is Dn.
42
Finite Boolean Algebras

• Example
Let S={2,3,5}, show the Hasse diagrams of (P(S), )
and D30 as follows.
2×3×5=30
{2,3,5} 30

{2,3} 3×5=15
{2,5} {3,5} 6 10 15

{2} {3} {5} 2 3 5

ф 1

43
Finite Boolean Algebras

• Example 5
Since 210=2×3×5×7, 66=2×3×11 and 646=2×17×19, then D210,
D66 D646 are all Boolean algebras.

• Example 9
Since 40=23×5, and 75=3×52, neither D40 and D75 are Boolean
algebras.
Note: If n is positive integer and p2 | n, where p is a prime number, then
Dn is not a Boolean algebra.

44
Finite Boolean Algebras

• Theorem 3 (Substitution rule for Boolean algebra)

Any formula involving U or ∩ that holds for arbitrary subsets of a
set S will continue to hold for arbitrary elements of a Boolean
algebra L if is substituted for ∩ and for U.

Example 6 If L is any Boolean algebra and x,y and z are in L, then the
following three properties hold.
1. (x’)’=x 2. (x y)’ = x’ y’ 3. (x y)’ = x’ y’
This is true by theorem 3,
1. (A)=A 2. (A∩B) =A U B 3. (A U B ) = A ∩ B
hold for arbitrary subsets A and B of a set S.
More properties can be found in p. 247, 1 ~12

45
Finite Boolean Algebras

• Example 7
Show the lattice whose Hasse diagram shown below is
not a Boolean algebra.
I

a and e are both complements of c

a f However, based on the 11.
Every element x has a unique complement x’
e Every element A has a unique complement x
d
Theorem 3 (e.g. properties 1~14) is usually used
b
c to show that a lattice L is not a Boolean algebra.

46
Finite Boolean Algebras

Denote the Boolean algebra B1 simply as B. Thus B

contains only the two elements 0 and 1. It is a fact that any
of the Boolean algebras Bn can be described in terms of B.
The following theorem gives this description.

• Theorem 4
For any n>=1, Bn is the product B×B×…×B of B, n
factors, where B×B×…×B is given the product partial
order.

47