De Morgan ADLs
De Morgan ADLs
Bulletin of
Mathematics
c SEAMS. 2020
Abstract. The purpose of this paper is to define and investigate a new equational class
of algebras which we call De Morgan almost distributive lattices (or shortly De Morgan
ADLs) as a common abstraction of De Morgan algebras and almost Boolean algebras
(relatively complemented ADLs).
1. Introduction
2. Preliminaries
This section contains some necessary definitions and results which will be used
in the text.
Definition 2.1. [9] An almost distributive lattice (or shortly an ADL) is an algebra
(L, ∨, ∧, 0) of type (2, 2, 0) satisfying the following set of axioms:
(1) a ∨ 0 = a,
(2) 0 ∧ a = 0,
(3) (a ∨ b) ∧ c = (a ∧ c) ∨ (b ∧ c),
(4) a ∧ (b ∨ c) = (a ∧ b) ∨ (a ∧ c),
(5) a ∨ (b ∧ c) = (a ∨ b) ∧ (a ∨ c),
(6) (a ∨ b) ∧ b = b
for all a, b, c ∈ A.
Lemma 2.2. [9] Let m ∈ L. Then the following statements are equivalent:
(1) m is maximal
(2) m ∨ x = m for all x ∈ L
(3) m ∧ x = x for all x ∈ L
a ∧ ab = 0 and a ∨ ab = a ∨ b
3. De Morgan ADLs
In this section we define De Morgan ADLs and give several characterization for
them.
It follows from the axioms that x◦◦ ∧ n = x ∧ n for all maximal elements n
of L.
Example 3.2. Let (L, ∨, ∧, 0) be a discrete ADL with at least two elements.
Choose a nonzero element m ∈ L and define a unary operation x 7→ x◦ on L as
follows: (
◦ m if a = 0,
a =
0 otherwise
for all a ∈ L. Then (L, ∨, ∧,◦ , 0, m) is a De Morgan ADL. We denote by L1 , the
discrete De Morgan ADL with exactly three elements.
454 G.M. Addis
Lemma 3.6. For any a and b in a De Morgan ADL L and a maximal element
m of L, the following statements are equivalent:
(1) ha] = hb].
(2) a ∧ m = b ∧ m.
(3) a◦ = b◦ .
Proof. It is proved in [9] that (1) and (2) are equivalent. We proceed to show that
(2) if and only if (3). Suppose that a ∧ m = b ∧ m. Then a◦ = a◦◦◦ = (a ∧ m)◦ =
(b ∧ m)◦ = b◦◦◦ = b◦ . Conversely, if a◦ = b◦ , then a ∧ m = a◦◦ = b◦◦ = b ∧ m.
Therefore the equivalence holds.
Theorem 3.9. Let (L, ∨, ∧,◦ , 0, m) be a De Morgan ADL and X a nonempty set.
If LX denotes the class of all functions from X to L. Then (LX , ∨, ∧,◦ , 0X , mX )
is a De Morgan ADL where ∨, ∧, ◦ , 0X and mX are given as follows:
For each f, g ∈ LX and each x ∈ X,
Remark 3.11. But the converse is not always true; i.e., every De Morgan ADL
need not necessarily be relatively complemented. To verify this, consider the De
Morgan ADL given in Example 3.3. There is no x in L such that a ∧ x = 0 and
a ∨ x = a ∨ b = b. So that, L is not relatively complemented.
Theorem 3.12. Let (L, ∨, ∧,◦ , 0, m) be a De Morgan ADL. Then the following
statements are equivalent:
(1) a ∧ a◦ = 0 for all a ∈ L.
(2) a◦ ∨ a◦◦ = m for all a ∈ L.
(3) L is relatively complemented.
(4) Lm is a Boolean algebra where the complementation is the unary operation
◦
.
Proof. (1) ⇒ (2), (3) ⇒ (4) and (4) ⇒ (1) are straight forward. We proceed
to show (2) ⇒ (3). Assume (2). Then for any a ∈ A, (a◦ ∨ a◦◦ )◦ = 0 so that
a◦◦ ∧ a◦◦◦ = a ∧ a◦ = 0. For any a, b ∈ L, we have a ∧ (a◦ ∧ b) = 0. Also consider:
a ∨ (a◦ ∧ b) = (a ∨ a◦ ) ∧ (a ∨ b)
= m ∧ (a ∨ a◦ ) ∧ (a ∨ b)
= (a ∨ a◦ ) ∧ m ∧ (a ∨ b)
= [(a ∧ m) ∨ (a◦ ∧ m)] ∧ (a ∨ b)
= (a◦◦ ∨ a◦ ) ∧ (a ∨ b)
= m ∧ (a ∨ b)
= a∨b
(a, b ∧ m)◦ = (b ∧ m, a ∧ m)
◦
for all a, b ∈ L, then (L × Ldual
m , ∨, ∧, , (m, 0)) is a De Morgan ADL.
Also
Theorem 3.15. Let (L, ∨, ∧,◦ , 0, m) be a De Morgan ADL. Then the following
statements are equivalent:
(1) L is a De Morgan algebra.
(2) The poset (L, ≤) is directed above.
(3) (L, ∨, ∧) is a distributive lattice.
(4) ∨ is commutative.
(5) ∧ is commutative.
(6) ∨ is right distributive over ∧.
(7) The relation θ = {(a, b) ∈ L × L : b ∧ a = a} is antisymetric.
(8) For each a ∈ L, the relation φa given by:
is a congruence relation on L.
458 G.M. Addis
Theorem 3.16. Let (L, ∨, ∧, 0, m) be an ADL. Then the following statements are
equivalent:
(1) L is a De Morgan ADL.
(2) Ln is a De Morgan algebra for all maximal elements n of L.
(3) Lm is a De Morgan algebra.
Proof. (1) ⇒ (2): Suppose that L is a De Morgan ADL. For any maximal
element n of L, define a unary operation x 7→ x on Ln as follows:
(a ∧ n) = a◦ ∧ n for all a ∈ L
where ◦ is the unary operation on L. It follows from Lemma 3.6 that this unary
operation is well defined. Moreover it can be easily verified that Ln together
with this unary operation is a De Morgan algebra.
(2) ⇒ (3) is clear. To prove (3) ⇒ (1) let us define a unary operation x 7→ x
on L as follows:
a = (a ∧ m)◦ for all a ∈ L
where ◦ is the unary operation on Ln . Then it is routine to verify that L together
with this unary operation is a De Morgan ADL.
L◦ = {a◦ : a ∈ L}.
−
Theorem 3.18. If we define binary operations ∨ and ∧ and a unary operation
on L◦ as follows:
Lemma 3.6 assures that f is well defined and injective. A routine verification
shows that f is a dual isomorphism.
In an ADL L, let us denote the set of all unary operations on L which makes
L a De Morgan ADL by U and the set of all maximal elements of L by M. Then
we have the following theorem.
Theorem 3.20. If both U and M are nonempty, then they are in one-to-one
correspondence.
4. Congruence Relations
(x, y) ∈ θ ⇒ (x◦ , y ◦ ) ∈ θ
and
(x, y) ∈ φa if and only if a ∧ x = a ∧ y and a ∧ x◦ = a ∧ y ◦
De Morgan Almost Distributive Lattices 461
Proof. Clearly both ψ a and φa are congruence relations on an ADL (L, ∨, ∧, 0).
Let (x, y) ∈ ψ a . Then x ∧ a = y ∧ a and x◦ ∧ a = y ◦ ∧ a. It remains to show
that x◦◦ ∧ a = y ◦◦ ∧ a. For, consider
x◦◦ ∧ a = x ∧ m ∧ a
= m∧x∧a
= m∧y∧a
= y∧m∧a
= y ◦◦ ∧ a.
a ∧ x◦◦ = a ∧ x ∧ m
= a∧y∧m
= a ∧ y ◦◦ .
x∧a = y∧x∧a
= x∧y∧a
= y ∧ a.
Theorem 4.8. The quotient L/φa is a De Morgan ADL but not necessarily a De
Morgan algebra. We verify this in the following example.
a ∧ x = a ∧ y, a ∧ x◦ = a ∧ y ◦ , b ∧ x = b ∧ y and b ∧ x◦ = b ∧ y ◦ .
So
(a ∨ b) ∧ x = (a ∧ x) ∨ (b ∧ x)
= (a ∧ y) ∨ (b ∧ y)
= (a ∨ b) ∧ y.
a ∧ x = (a ∧ x) ∧ [(a ∧ x) ∨ (b ∧ x)]
= (a ∧ x) ∧ [(a ∨ b) ∧ x]
= (a ∧ x) ∧ [(a ∨ b) ∧ y]
= a ∧ x ∧ (a ∨ b) ∧ y
= a ∧ (a ∨ b) ∧ x ∧ y
= a∧x∧y
and
a ∧ y = [(a ∧ y) ∨ (b ∧ y)] ∧ (a ∧ y)
= [(a ∨ b) ∧ y] ∧ (a ∧ y)
= [(a ∨ b) ∧ x] ∧ (a ∧ y)
= (a ∨ b) ∧ x ∧ a ∧ y
= (a ∨ b) ∧ a ∧ x ∧ y
= a ∧ x ∧ y.
Lemma 4.11. If a is maximal, then φa = ∆L . But the converse need not neces-
sarily be true.
a ∧ (a ∧ a◦ )◦ = a ∧ (a◦ ∨ a◦◦ )
= a ∧ (a◦◦ ∨ a◦ )
= (a ∧ a◦◦ ) ∨ (a ∧ a◦ )
= (a ∧ a ∧ m) ∨ (a ∧ a◦ )
= (a ∧ m) ∨ (a ∧ a◦ )
= a ∧ (m ∨ a◦ )
= a∧m
= a ∧ (a ∧ m)
= a ∧ a◦◦ .
consider:
a ∨ (a ∨ a◦ )◦ = a ∨ (a◦ ∧ a◦◦ )
= (a ∨ a◦ ) ∧ (a ∨ a◦◦ )
= (a ∨ a◦ ) ∧ a
=a
= a ∨ a◦◦ .
So that (a ∨ a◦ , a◦ ) ∈ θa .
References
[1] R. Balbes and P. Dwinger, Distributive Lattices, University of Missouri Press,
Columbia, Missouri, 1974.
[2] A. Bialynicki-Birula and H. Rasiowa, On the representation of quasi-Boolean al-
gebras, Bull. Acad. Polon. Sci. 5 (1957) 259–261.
[3] A. Bialynicki-Birula, Remarks on quasi-Boolean algebras, Bull. Acad. Polon. Sci.
5 (1957) 615–619.
[4] G. Birkhoff, Lattice theory, Amer. Math. Soc. Colloq Publ. 24 (1967) 204–206.
[5] J. Kalman, Lattices with involution, Trans. Amer. Math. Soc. 87 (1958) 485–491.
[6] G.C. Moisi, Recherches sur l’algebra de la logique, Ann. Sci. Univ. Jassy. Partie
I 22 (1935) 1–117.
[7] Y.S. Pawar, I.A. Shaikh, Congruence relations on almost distributive lattices,
Southeast Asian Bull. Math. 36 (2012) 519–527.
[8] U.M. Swamy, S. Ramesh, Birkhoff center of an almost distributive lattice, Int.
Jour. Algebra 11 (2009) 539–346.
[9] U.M. Swamy, G.C. Rao, Almost distributive lattice, J. Aust. Math. Soc. 5
(Section-A) 31 (1981) 77–91 231–240.