Professional Documents
Culture Documents
net/publication/339413592
CITATION READS
1 87
1 author:
Saeed Rasouli
Persian Gulf University
34 PUBLICATIONS 194 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Faculty of Intelligent Systems Engineering and Data Science, Persian Gulf University View project
All content following this page was uploaded by Saeed Rasouli on 26 February 2020.
FOUNDATIONS
Abstract
In this paper, the class of quasicomplemented residuated lattices is introduced and investigated, as a subclass of residuated
lattices in which any prime filter not containing any dense element is a minimal prime filter. The notion of a disjunctive
residuated lattice is introduced, and it is observed that a residuated lattice is Boolean if and only if it is disjunctive and
quasicomplemented. Finally, some characterizations for quasicomplemented residuated lattices are given by means of the
new notion of α-filters.
123
S. Rasouli
element a ∈ A is called nilpotent, provided that a n = 0, for Table 2 The Cayley table for 0 a b c d 1
some integers n. The set of nilpotent elements of A shall be “” of A6
denoted by N(A). It is well known that N(A) is an ideal of 0 0 0 0 0 0 0
(A). Idziak (1984) has shown that the class of residuated a a 0 a 0 a
lattices is equational and so it forms a variety. The properties b 0 0 b b
of residuated lattices were presented in Galatos et al. (2007). c a b c
For a survey of residuated lattices, we refer to Jipsen and d d d
Tsinakis (2002). 1 1
r1 x (y ∨ z) = (x y) ∨ (x z);
r2 x ∨ (y z) ≥ (x ∨ y) (x ∨ z).
123
Quasicomplemented residuated lattices
Table 4 The sets of filters of A4 , A6 , and A7 Table 5 The sets of maximal, prime, and minimal prime filters of A4 ,
A6 , and A7
Filters
Prime filters
A4 {1}, {a, 1}, {b, 1}, A4
Maximal filters Minimal prime filters
A6 {1}, {a, c, 1}, {d, 1}, A6
A7 {1}, {b, d, 1}, {e, 1}, {a, b, c, d, e, 1}, A7 A4 {a, 1}, {b, 1} {a, 1}, {b, 1}
A6 {d, 1}, {a, c, 1} {d, 1}, {a, c, 1}
A7 {a, b, c, d, e, 1} {b, d, 1}, {e, 1}
Example 2.4 Consider the residuated lattice A4 from Exam-
ple 2.1, the residuated lattice A6 from Example 2.2, and the
de f .
residuated lattice A7 from Example 2.3. The sets of their only if P c = A\P is a ∨-closed subset of A which it is
filters are presented in Table 4. maximal with respect to the property of not containing 1.
The following proposition has a routine verification. Example 2.8 Consider the residuated lattice A4 from Exam-
Proposition 2.5 Let A be a residuated lattice and F be a filter ple 2.1, the residuated lattice A6 from Example 2.2, and the
of A. The following assertions hold, for any x, y ∈ A: residuated lattice A7 from Example 2.3. The sets of their
maximal, prime, and minimal prime filters are presented in
de f . Table 5.
(1) F(F, x) = F F(x) = {a ∈ A| f x n ≤
a, for some f ∈ F and n ∈ N}; Let (A; ∨, ∧, 0, 1) be a bounded lattice. Recall (Grätzer
(2) x ≤ y implies F(F, y) ⊆ F(F, x); 2011, §I.6.1) that an element x ∈ A is called complemented
(3) F(F, x) ∩ F(F, y) = F(F, x ∨ y); if there is an element y ∈ A such that x ∨y = 1 and x ∧y = 0;
(4) F(x) F(y) = F(x y) = F({x, y}); y is called a complement of x. Complements in a bounded
(5) PF(A) is a sublattice of F(A); lattice are generally not unique unless the lattice is distribu-
(6) if A is finite, then F(A) = PF(A). tive. If y is the unique complement of x, we write y = x c . If
A is a residuated lattice, the set of complemented elements of
A proper filter of a residuated lattice A is called maximal (A) is denoted by B(A) and called the Boolean center of A.
if it is a maximal element in the set of all proper filters. In residuated lattices however, although the underlying lat-
The set of maximal filters of A is denoted by Max(A). A tices need not be distributive, the complements are unique.
proper filter P of A is called prime if x ∨ y ∈ P implies For further study about the Boolean center of a residuated
x ∈ P or y ∈ P, for any x, y ∈ A. The set of prime filters lattice, we refer to Buşneag and Piciu (2006).
of A is denoted by Spec(A). Since F(A) is a distributive
lattice, Max(A) ⊆ Spec(A). By Zorn’s lemma follows that Proposition 2.9 (Georgescu and Mureşan 2014, Lemma 2.5)
any proper filter is contained in a maximal filter and so in a Let A be a residuated lattice. The following assertions hold:
prime filter. A non-empty subset C of A is called ∨-closed
if it is closed under the join operation, i.e., x, y ∈ C implies (1) F(e) = {a ∈ A|e ≤ a}, for any e ∈ B(A);
x ∨ y ∈ C. (2) x ∈ B(A) if and only if x ∨ ¬x = 1, for any x ∈ A;
(3) ec = ¬e, for any e ∈ B(A).
Theorem 2.6 (Rasouli 2019, Theorem 3.18) If C is a ∨-
closed subset of A which does not meet the filter F, then Let A be a residuated lattice. Following by Rasouli (2018),
F is contained in a filter P which is maximal with respect to for any subset X of A, we write X ⊥ = {a ∈ A|a ∨ x =
the property of not meeting C; furthermore, P is prime. 1, ∀x ∈ X } and we set:
Let A be a residuated lattice and X be a subset of A. A
prime filter P of A is called a minimal prime filter belonging • (A) = {X ⊥ |X ⊆ A}.
to X or X -minimal prime filter if P is a minimal element in
the set of prime filters containing X . The set of X -minimal Elements of (A) are called coannihilators of A. For the
prime filters of A is denoted by Min X (A). A prime filter basic facts concerning coannihilators and coannulets of a
P is called minimal prime if P ∈ Min{1} (A). The set of residuated lattice, we refer to Rasouli (2018).
minimal prime filters of A is denoted by Min(A). For the
Proposition 2.10 (Rasouli 2018, Proposition 3.1) Let A be a
basic facts concerning minimal prime filters of a residuated
residuated lattice. The following assertions hold:
lattice belonging to a filter, we refer to Rasouli (2019).
Theorem 2.7 (Rasouli 2019, Theorem 3.24) Let A be a resid- (1) X ⊆ Y ⊥ implies Y ⊆ X ⊥ , for any X , Y ⊆ A;
uated lattice. A subset P of A is a minimal prime filter if and (2) X ∩ X ⊥ ⊆ {1}, for any X ⊆ A;
123
S. Rasouli
(3) X ⊆ X ⊥⊥ , for any X ⊆ A; Corollary 2.13 (Rasouli 2019, Proposition 5.1) Let A be a
(4) (F(X ))⊥ = X ⊥ , for any X ⊆ A; residuated lattice. Then, for any subset X of A, we have
(5) X ⊥ is a filter of A, for any X ⊆ A;
(6) X ⊥⊥⊥ = X ⊥ , for any X ⊆ A; X⊥ = {m ∈ Min(A)| X m}.
⊥
⊥
(7) X ∈X X = ( X ) , for any X ⊆ P(A).
Proposition 2.14 (Rasouli 2018, Proposition 3.15) Let A be
Let A be a ∧-semilattice with zero. Recall (Grätzer 2011, a residuated lattice. The following assertions hold, for any
§I.6.2) that an element a ∗ ∈ A is a pseudocomplement of x, y ∈ A:
a ∈ A if a ∧ a ∗ = 0 and a ∧ x = 0 implies that x ≤
a ∗ . An element can have at most one pseudocomplement. A (1) x ≤ y implies x ⊥ ⊆ y ⊥ ;
is called pseudocomplemented if every element of A has a (2) x ⊥ ∩ y ⊥ = (x y)⊥ ;
pseudocomplement. The set S(A) = {a ∗ | a ∈ A} is called (3) x ⊥ y ⊥ ⊆ x ⊥ ∨ y ⊥ = (x ∨ y)⊥ ;
the skeleton of A, and we have S(A) = {a ∈ A | a = a ∗∗ }. (4) x ⊥⊥ ∩ y ⊥⊥ = (x ∨ y)⊥⊥ ;
(5) x ⊥⊥ ∨ y ⊥⊥ = (x y)⊥⊥ ;
Proposition 2.11 Let A be a residuated lattice. The following (6) e⊥ = F(¬e), for any e ∈ B(A).
assertions hold:
Let A be a residuated lattice. We set:
(1) F ⊥ is the pseudocomplement of F, for any F ∈ F(A);
(2) (F(A); ∩, {1}) is a pseudocomplemented ∧-semilattice; • γ (A) = {x ⊥ |x ∈ A};
(3) S(F(A)) = {F ⊥ | F ∈ F(A)} = {F ∈ F(A) | F = • λ(A) = {x ⊥⊥ |x ∈ A}.
F ⊥⊥ };
(4) (A) = S(F(A)). Elements of γ (A) and λ(A) are called coannulets and dual
coannulets of A, respectively.
Proof (1): Let F be a filter of A. By Proposition 2.10(2)
follows that F ∩ F ⊥ = {1}. Assume that F ∩ G = {1}, Proposition 2.15 Let A be a residuated lattice. Then γ (A)
for some filter G. Let a ∈ G and b ∈ F. Since a, b ≤ and λ(A) are sublattices of (A).
a ∨ b so a ∨ b ∈ F ∩ G = {1}. Thus, a ∈ F ⊥ , and it Proof By Proposition 2.14((2) and (3)) follows that γ (A) is
shows that G ⊆ F ⊥ . a sublattice of (A), and by Proposition 2.14((4) and (5))
(2): It follows by (1). follows that λ(A) is a sublattice of (A).
(3): It is evident.
(4): It is obvious that S(F(A)) ⊆ (A). Let F ∈ (A). Let A be a residuated lattice. A subset X of A is called
Thus, F = X ⊥ , for some X ⊆ A. By Proposition dense if X ⊥ = {1}. The set of all dense elements of A shall
2.10(5) follows that F ∈ F(A), and by Proposi- be denoted by d(A). By Proposition 2.14((1) and (3)) follows
tion 2.10(6) follows that F ⊥⊥ = F. So F ∈ S(F(A)). that d(A) is an ideal of (A). Also, by Proposition 2.10(1)
follows that a filter of A is dense, provided that it contains a
dense element.
By Grätzer (2011, Theorem 100) follows that if A is a
pseudocomplemented complete ∧-semilattice, then S(A) is Proposition 2.16 Let A be a residuated lattice. Then, any
a complete Boolean lattice, where the meet in S(A) is cal- non-dense prime filter of A is a coannulet.
culated in A, the join in S(A) is given by ∨X = (∧{x ∗ |
de f . Proof Let P be a non-dense prime filter of A. So there exists
x ∈ X })∗ , for any X ⊆ S(A), and 1 = 0∗ . So by Proposi- 1 = x ∈ P ⊥ . This implies that P ⊆ P ⊥⊥ ⊆ x ⊥ . Otherwise,
tions 2.10(7) and 2.11follows that ((A); ∩, ∨ , {1}, A) is a y ∈ x ⊥ implies that x ∨ y = 1 ∈ P. But x ∈/ P since x ∈ P
complete Boolean lattice, where for any F ⊆ (A) we have states that x ∈ P ∩ P ⊥ = {1} and this means that x = 1, a
∨ F = (∪F)⊥⊥ . This gives another proof for the following contradiction. So y ∈ P and it shows that P = x ⊥ .
theorem:
Let A be a residuated lattice. For any ideal I of (A), we
Theorem 2.12 (Rasouli 2018, Corollary 3.14) Let A be
write ω(I ) = {a ∈ A|a ∨ x = 1, ∃x ∈ I } and we set:
a residuated lattice. ((A); ∩, ∨ , {1}, A) is a complete
Boolean lattice, where for any F ⊆ (A) we have ∨ F =
• (A) = {ω(I )|I is an ideal of (A)}.
(∪F)⊥⊥ .
The following corollary characterizes coannihilators of a Elements of (A) are called ω-filters of A. By Rasouli and
residuated lattice in terms of minimal prime filters. Kondo (2019, Proposition 3.7) follows that ( (A); ∩, ∨ω ,
{1}, A) is a bounded distributive lattice, where F ∨ω G =
123
Quasicomplemented residuated lattices
ω(I F IG ), for any F, G ∈ (A) (by , we mean the (3)⇒(1): By Corollary 2.13 follows that x ⊥ = d(x) =
join operation in the lattice of ideals of (A)). Also, by d(y) = y ⊥ .
Rasouli and Kondo (2019, Lemma 3.8) follows that γ (A)
is a sublattice of (A). For a prime filter P of A, we write
D(P) = ω(P c ). For the basic facts concerning ω-filters of a 3 Quasicomplemented residuated lattices
residuated lattice, we refer to Rasouli and Kondo (2019).
In this section, the notion of a quasicomplemented residuated
Proposition 2.17 Any proper ω-filter in a residuated lattice
lattice is introduced and studied.
A contains no dense elements.
Definition 3.1 Let A be a residuated lattice. A is called qua-
Proof Let F be a proper ω-filter. Assume that F contains sicomplemented, provided that for any x ∈ A, there exists
a dense element as d. Hence, d ∈ x ⊥ for some x ∈ I F . It y ∈ A such that x ⊥⊥ = y ⊥ (i.e., λ(A) ⊆ γ (A)).
implies that x ∈ d ⊥ and it means 1 ∈ I F . So F = A, a
contradiction.
The following proposition characterizes quasicomple-
mented residuated lattices.
The following corollary characterizes minimal prime fil-
Proposition 3.2 Let A be a residuated lattice. The following
ters belonging to a filter.
assertions are equivalent:
Theorem 2.18 (Rasouli and Kondo 2019, Theorem 3.16) Let
A be a residuated lattice. The following assertions are equiv- (1) A is quasicomplemented;
alent: (2) for any x ∈ A, there exists y ∈ A such that x y ∈ d(A)
and x ∨ y = 1;
(1) P is a minimal prime filter; (3) γ (A) is a Boolean lattice.
(2) P = D(P);
(3) for any x ∈ A, P contains precisely one of x or x ⊥ . Proof (1)⇒(2): Consider x ∈ A. So there exists y ∈ A
such that x ⊥⊥ = y ⊥ . Applying Proposition 2.10(2) and
Corollary 2.19 (Rasouli and Kondo 2019, Corollary 3.23) 2.14(2), it follows that x y is a dense element and
Let A be a residuated lattice and P be a prime filter of A. We Proposition 2.10(2) shows that x ∈ x ⊥⊥ = y ⊥ and so
have x ∨ y = 1.
(2)⇒(3): By applying Proposition 2.14((3) and (4)), it is
D(P) = {m ∈ Min(A)|m ⊆ P}. evident.
(3)⇒(1): Let x ∈ A. So there exists some y ∈ A such
Let A be a residuated
lattice. For a subset M of Min(A), that x ⊥ ∩ y ⊥ = {1} and x ⊥ ∨ y ⊥ = A. By Proposi-
we write k(M) = {m|m ∈ M}, and for a subset X of A, we tion 2.11(1), the former states y ⊥ ⊆ x ⊥⊥ and the latter
write h(X ) = {m|X ⊆ m}. Also, let d(X ) = Min(A)\h(X ). states the reverse inclusion.
If the collection {h(x)|x ∈ A} is taken as a closed basis, the
resulting topology is called the hull-kernel topology, which The following proposition derives a sufficient condition
is denoted by τh , and if the collection {h(x)|x ∈ A} is taken for a residuated lattice to become quasicomplemented.
as an open basis, the resulting topology is called the dual Proposition 3.3 Let A be a residuated lattice. A is quasicom-
hull-kernel topology, which is denoted by τd . For a detailed plemented, provided that in this any coannulet is principal.
discussion of spaces of minimal prime filters in residuated
lattices, we refer to Rasouli and Dehghani (2018). Proof Consider x ∈ A. So there exists y ∈ A such that x ⊥ =
F(y). Using Proposition 2.10(4) follows that x ⊥⊥ = y ⊥ , and
Corollary 2.20 Let A be a residuated lattice. The following
so A is quasicomplemented.
assertions are equivalent for any x, y ∈ A:
Corollary 3.4 Any finite residuated lattice is quasicomple-
(1) x ⊥ = y ⊥ ; mented.
(2) h(x) = h(y);
(3) d(x) = d(y). Proof It follows by Proposition 2.5(6) and 3.3.
123
S. Rasouli
(1) A is quasicomplemented;
(2) any prime filter not containing any dense element is min-
imal prime;
(3) any filter not containing any dense element is contained
in a minimal prime filter.
123
Quasicomplemented residuated lattices
A4
A6 ×
A7 × ×
123
S. Rasouli
Proposition 5.4 Let {F} ∪ {Fi }i∈I be a family of filters in a The following proposition gives some equivalent asser-
residuated lattice A, and x, y ∈ A. The following assertions tions for a filter to be an α-filter.
hold: Proposition 5.6 Let A be a residuated lattice and F be a filter
of A. The following assertions are equivalent:
(1) α(F) = ∪x∈F x ⊥⊥ ;
α F = {a ∈ A|a ∈ ( f · · · f )⊥⊥ , ∃n ∈
(2) ∨i∈I i i1 in (1) F is an α-filter;
N, i 1n ∈ I , f i j ∈ Fi j }; (2) if x ⊥ = y ⊥ and x ∈ F, then y ∈ F for any x, y ∈ A;
de f .
(3) α(F, x) = α(F ∪ {x}) = ∪ f ∈F,n∈N ( f x n )⊥⊥ ; (3) if d(x) = d(y) and x ∈ F, then y ∈ F for any x, y ∈ A;
(4) α(F, x) ∩ α(F, y) = α(F, x ∨ y); (4) if h(x) = h(y) and x ∈ F, then y ∈ F for any x, y ∈ A.
(5) α(F, x) ∨α α(F, y) = α(F, x y).
Proof By Corollary 2.20, it follows that (1), (2), and (3) are
Proof (1): It proves quite in a routine way. equivalent. So we only prove the other cases.
(2): We have ∨i∈I α F = α(
i i∈I Fi ) = {a ∈ A|a ∈
⊥⊥
x , f or some x ∈ i∈I Fi }. Let
= {a ∈ A|a ∈ (1)⇒(2): Let x ⊥ = y ⊥ for some y ∈ A and x ∈ F. Then,
(it=i
n
f )⊥⊥ , ∃n, i 1 , · · · , i n ∈ N f i j ∈ Fi j }. It is obvi- y ∈ y ⊥⊥ = x ⊥⊥ ⊆ F.
1 t
ous that
⊆ ∨i∈I α F . Suppose that a ∈ ∨α F . Thus,
i (2)⇒(1): Let x ∈ F and y ∈ x ⊥⊥ . So x ⊥ ⊆ y ⊥ . By
i∈I i
a ∈ x ⊥⊥ for some x ∈ i∈I Fi . So there exist an Proposition 2.14(4) follows that y ⊥ = x ⊥ ∨ y ⊥ =
integer n, i 1 , · · · , i n ∈ I and f i j ∈ Fi j such that (x ∨ y)⊥ and it states that y ∈ F since x ∨ y ∈ F.
123
Quasicomplemented residuated lattices
Let A = (A; ≤) and B = (B; ) be two posts. We recall In the sequel for any residuated lattice A, we set Specα (A)
that a pair ( f , g) is called an adjunction (or isotone Galois = Spec(A) ∩ α(A).
connection) between posets A and B, where f : A −→ B
Corollary 5.9 Let F be an α-filter of a residuated lattice A
and g : B −→ A are two functions such that for all a ∈ A
and X be a non-empty subset of A. The following assertions
and b ∈ B, f (a) b if and only if a ≤ g(b). It is well known
hold:
that ( f , g) is an adjunction connection if and only if g f is
inflationary, f g is deflationary and f , g are isotone (García-
(1) If X F, then there exists P ∈ Specα (A) such that
Pardo et al. 2013, Theorem 2). It is well known Cg f = g(B),
F ⊆ P and P is maximal with respect to the property
where Cg f is the set of fixed points of the closure operator
X P;
gf .
(2) α(X ) = {P ∈ Specα (A)|X ⊆ P}.
Theorem 5.7 Let A be a residuated lattice. We define
Proof (1): Let x ∈ X − F. By taking C = {x}, it follows by
: F(A) −→ F(γ (A)) : F(γ (A)) −→ F(A) Theorem 5.8.
F −→ {x ⊥ |x ∈ F}, F −→ {x ∈ A|x ⊥ ∈ F}.
(2): Set σ X = {P ∈ Specα (A)|X ⊆ P}. Obviously, we have
Then, the pair (, ) is an adjunction and we have α(X ) ⊆ σ X . Now, let a ∈ / α(X ). By (1) follows that
(F) = α(F) for any F ∈ F(A). In particular, we have there exits an α-prime filter
P containing α(X ) such that
C = α(A). a∈/ P. It shows that a ∈ / σX .
Proof Quite in a routine way, we can show that the pair The following proposition characterizes weakly disjunc-
(, ) forms an adjunction. Let F be a filter of A. We have tive residuated lattices by using of α-filters.
the following formulas: Proposition 5.10 Let A be a residuated lattice. The following
assertions are equivalent:
(F) = ((F))
= {a ∈ A|a ⊥ ∈ (F)} (1) A is weakly disjunctive;
= {a ∈ A|a ⊥ = x ⊥ , ∃x ∈ F} (2) any filter of A is an α-filter;
= {a ∈ A|a ∈ x ⊥⊥ , ∃x ∈ F} = ∪x∈F x ⊥⊥ = α(F). (3) any prime filter of A is an α-filter.
The rest is evident.
Proof (1)⇒(2): Let F be a filter of A. Consider x, y ∈ A
The next theorem should be compared with Theorem 2.6. such that x ⊥ = y ⊥ and x ∈ F. Since A is weakly
disjunctive, F(x) = F(y) and this implies that y ∈
Theorem 5.8 Let C be a ∨-closed subset of A which does not F(y) = F(x) ⊆ F. Hence, F is an α-filter due to
meet the α-filter F. Then, F is contained in an α-filter P Proposition 5.6(2).
which is maximal with respect to the property of not meeting (2)⇒(3): It is trivial.
C; furthermore, P is prime. (3)⇒(1): Let x ⊥ = y ⊥ and F(x) = F(y) for some x, y ∈ A.
Without loss of generality, suppose that F(x)
Proof Let
= {G ∈ α(A)|F ⊆ G, G ∩C = ∅}. It is easy to F(y). Let
= {F ∈ F(A)|y ∈ F, x ∈ / F}.
see that
satisfies the conditions of Zorn’s lemma. Let P be F(y) ∈
and
satisfies the conditions of Zorn’s
a maximal element of
. Assume that x ∨ y ∈ P and neither lemma. Let P be the maximal element of
. Let
x ∈
/ P nor y ∈ / P. By maximality of P, we have α(P, x) ∩ a ∨b ∈ P and a, b ∈ / P. So F(P, a), F(P, b) ∈ /
C = ∅ and α(P, y) ∩ C = ∅. Suppose ax ∈ α(P, x) ∩ C and this states that x ∈ F(P, a) ∩ F(P, b). By
and a y ∈ α(P, y) ∩ C. By Proposition 5.4(3), there exist Proposition 2.5(3) follows that x ∈ F(P, a) ∩
px , p y ∈ P and integers n, m such that ax ∈ ( px x n )⊥⊥ F(P, b) = F(P, a ∨ b) = P and this leads us
and a y ∈ ( p y y m )⊥⊥ . It follows that to a contradiction. Thus, P is a prime filter, and
this implies that x ∈ x ⊥⊥ = y ⊥⊥ ⊆ P, a contra-
ax ∨ a y ∈ ( px x n )⊥⊥ ∩ ( p y y m )⊥⊥ diction.
= (( px x n ) ∨ ( p y y m ))⊥⊥
Remark 5 For the sake of more clarity of Proposition 5.10,
⊆ (( px ∨ p y ) ( px ∨ y m )) (x n ∨ p y ) see Examples 4.5 and 5.2.
(x ∨ y)nm )⊥⊥
In the following proposition, we show that any coannihi-
⊆ P.
lator filter and any ω-filter is an α-filter.
It is a contradiction. So x ∈ P or y ∈ P and it shows that P Proposition 5.11 Let A be a residuated lattice. The following
is a prime filter.
assertions hold:
123
S. Rasouli
(1) Any non-dense prime filter is an α-filter; In the following proposition we give some equivalent con-
(2) any minimal prime filter is an α-filter. ditions for each α-filter to be an ω-filter.
Theorem 5.14 Let A be a residuated lattice. The following
Proof (1): It is an immediate consequence of Proposi- assertions are equivalent:
tion 2.16 and Proposition 5.11(1).
(2): It is an immediate consequence of Theorem 2.18 and (1) A is quasicomplemented;
Proposition 5.11(2).
(2) every α-filter is an ω-filter;
(3) every coannihilator is an ω-filter;
In the following proposition, we give some equivalent con- (4) For any x ∈ A, x ⊥⊥ is an ω-filter.
ditions for each α-filter to be a coannihilator.
Proposition 5.13 Let A be a residuated lattice. The following Proof (1)⇒(2): Let F be an α-filter and set I F = {a ∈
assertions are equivalent: A| f ⊥ ⊆ a ⊥⊥ , f or some f ∈ F}. It is obvious that
0 ∈ I F . Let a1 , a2 ∈ I F . So f 1⊥ ⊆ a1⊥⊥ and f 2⊥ ⊆ a2⊥⊥
for some f 1 , f 2 ∈ F. By Proposition 2.14 follows that
(1) Any proper α-filter is non-dense;
( f 1 f 2 )⊥ ⊆ (a1 ∨ a2 )⊥⊥ and it states that a1 ∨ a2 ∈ I F .
(2) any dense filter contains a dense element;
Now let a1 ≤ a2 and a2 ∈ I F . So there exists f ∈ F
(3) any α-filter is a coannihilator;
such that f ⊥ ⊆ a2⊥⊥ and on the other hand we have
(4) for any proper α-filter F, there exists a proper α-filter G
a2⊥⊥ ≤ a1⊥⊥ . Hence, we have f ⊥ ⊆ a1⊥⊥ and it means
such that F∩G = {1} (i.e., α(A) is semi-complemented);
that a1 ∈ I F . Thus, I F is a lattice ideal of A. Now, let
(5) α(A) has a unique dense element.
x ∈ ω(I F ). Consequently, x ∈ a ⊥ for some a ∈ I F and
so x ∈ f ⊥⊥ for some f ∈ F. Since F is an α-filter,
Moreover, any of the above assertions implies that A is qua- it follows that x ∈ F. Otherwise, let x ∈ F. Since A
sicomplemented. quasicomplemented, it follows that x ⊥⊥ = a ⊥ for some
a ∈ A. It means that a ∈ I F . Therefore, x ∈ x ⊥⊥ =
Proof (1)⇒(2): Let F be a dense filter of A. It implies that
a ⊥ ⊆ ω(I F ) and it shows that F ⊆ ω(I F ).
α(F) is a dense filter and it means that α(F) = A. Therefore,
(2)⇒(3): It is obvious, since any coannihilator is an α-
0 ∈ α(F) and it means that x ⊥ = 1 for some element x ∈ F.
filter.
(2)⇒(3): Let F be an α-filter. F, F ⊥ ⊆ F F ⊥ and so
(3)⇒(4): It is obvious, since for any x ∈ A, x ⊥⊥ is a
(F F ⊥ )⊥ ⊆ F ⊥ ∩ F ⊥⊥ = {1}. So F F ⊥ is a dense
coannihilator.
filter and so it contains a dense element. Assume that x is a
(4)⇒(1): Let x ∈ A. So there exists a lattice ideal I such
dense element in F F ⊥ . So there exist a ∈ F and b ∈ F ⊥
that x ⊥⊥ = ω(I ). So x ∈ ω(I ) and it implies that x ∈ y ⊥
such that a b ≤ x. It implies that a ⊥ ∩ b⊥ ⊆ x ⊥ = {1}
for some y ∈ I . Thus, x ⊥⊥ ⊆ y ⊥ ⊆ ω(I ) = x ⊥⊥ .
and since a ⊥⊥ is the pseudocomplement of a ⊥ , it follows
that b⊥ ⊆ a ⊥⊥ ⊆ F as F is an α-filter. On the other hand, Lemma 5.15 Let F be a filter of a residuated lattice A. Then,
b ∈ F ⊥ implies that F ⊥⊥ ⊆ b⊥ and it states that F ⊥⊥ ⊆ F. α(F) is proper if and only if F contains no dense element.
The other inclusion is evident.
(3)⇒(4): Since the set of all coannihilators forms a Boolean Proof Let F ∩ d(A) = ∅. So there exists a dense element d
lattice, it follows that α(A) is semi-complemented. in F. By Proposition 5.4(1), we get that A = d ⊥⊥ ⊆ α(F)
123
Quasicomplemented residuated lattices
and it implies α(F) = A. Otherwise, if α(F) = A, then 0 ∈ Lemma 5.18 Let A be a residuated lattice, F be an α-filter
α(F). So there exists x ∈ F such that 0 ∈ x ⊥⊥ . Therefore, and a ∈ F. If for any prime α-filter P which contains a we
x ⊥ ⊆ 0⊥ = {1} and it means that x is a dense element. have F ⊆ P, then F = α(a).
Lemma 5.16 Let A be a quasicomplemented residuated lat- Proof It is a direct consequence of Corollary 5.9(2).
tice and P be a prime filter. The following assertions are Proposition 5.19 Let A be a residuated lattice. The following
equivalent: assertions are equivalent:
(1) P is an α-filter; (1) For any P ⊆ Specα (A) and any α-filter F, F ⊆ P
(2) P contains no dense element; implies F ⊆ P for some P ∈ P;
any P ⊆ Specα (A) and any prime α-filter Q, Q ⊆
(3) P is minimal prime; (2) for
(4) P contains precisely one of x, y ∈ A such that x ⊥⊥ = P implies Q ⊆ P for some P ∈ P;
y⊥. (3) any prime α-filter is principal;
(4) any α-filter is principal.
Proof (1)⇒(2) follows by Lemma 5.15 and (2)⇒(3) follows
by Proposition 3.5(2). Proof (1)⇒(2): It is obvious.
(2)⇒(3): Let Q be a prime α-filter which is not principal.
(3)⇒(4): Let x and y be any pair of elements for which By Lemma 5.18, for any a ∈ Q there exists a
x ⊥⊥ = y ⊥ . Since x ∨ y = 1 and P is prime so either prime
α-filter Pa such that Q Pa . We have Q ⊆
x ∈ P or y ∈ P. Also, x ⊥ ∩ y ⊥ = {1} implies that {Pa |a ∈ Q} and so Q ⊆ Pa for some a ∈ Q, a
x ⊥ ⊆ P or y ⊥ ⊆ P. Now we can deduce the result by contradiction.
Theorem 2.18(3). (3)⇒(4): It is obvious that A is a principal α-filter. Set be
(4)⇒(1): Let x ∈ P and y be an element such that x ⊥⊥ = the set of all proper non-principal α-filters of A.
y ⊥ . Since y ∈
/ P, y ⊥ ⊆ P and it shows that P is an α- Let be not empty. In a routine way, we can show
filter.
that satisfies the conditions of Zorn’s lemma.
Let M be a maximal element of . Let a ∨ b ∈ M
and a, b ∈/ M. Thus, there exist x, y ∈ A such that
In the next corollary, we characterize the quasicomple-
α(M, a) = α(x) and α(M, b) = α(y). Applying
mented residuated lattice in terms of α-filters. This result
Proposition 5.4, it follows that M = α(x ∨ y), a
should be compared with Proposition 3.5.
contradiction. Thus, M is a prime α-filter and so it
Corollary 5.17 Let A be a residuated lattice. The following is principal, a contradiction. Hence, = ∅ and it
assertions are equivalent: gets the result.
(4)⇒(1): Let F be an α-filter such that F ⊆ ∪P for some
(1) A is quasicomplemented; P ⊆ Specα (A). By hypothesis, F = α(a) for
(2) any prime α-filter is minimal prime; some a ∈ A. So there exists P ∈ P such that a ∈
(3) any proper α-filter is the intersection of minimal P. Hence, F ⊆ P and it proves the implication.
prime filters; Lemma 5.20 If any coannulet of a residuated lattice is prin-
(4) any proper α-filter is contained in a minimal prime cipal, then any its prime α-filter is a minimal prime filter.
filter.
Proof Let P be a prime α-filter of A. By Proposition 5.6(2),
Proof (1)⇒(2) is followed by Proposition 3.5 and Lemma it is obvious that P has no any dense element. Let x ∈ P.
5.16, (2)⇒(3) is followed by Corollary 5.9(2) and (3)⇒(4) By hypothesis x ⊥ = F(y) for some y ∈ A. So we have
is obvious. (x y)⊥ = x ⊥ ∩ y ⊥ = x ⊥ ∩ x ⊥⊥ = {1} and so x y ∈ / P.
Hence, y ∈ / P and so x ∈ D(P). By Theorem 2.18, the result
holds.
(4)⇒(1): Let F be a filter such that F ∩ d(A) = ∅. By
Lemma 5.15, α(F) is proper and so it is contained in Proposition 5.21 Let A be a residuated lattice. The following
a minimal prime filter. It means that F is contained in assertions are equivalent:
a minimal prime filter. Hence, the result is followed by
Proposition 3.5(3).
(1) Any α-filter is principal;
(2) any ω-filter is principal;
We end this paper by deriving a set of equivalent assertions (3) any coannulet is principal and any minimal prime filter
for any α-filter of a residuated lattice to become principal. is non-dense;
123
S. Rasouli
(4) any prime α-filter is principal. Galatos N, Jipsen P, Kowalski T, Ono H (2007) Residuated Lattices:
An Algebraic Glimpse at Substructural Logics, vol 151. Elsevier,
Amsterdam
Further, any of the above assertions implies that A is quasi- García-Pardo F, Cabrera IP, Cordero P, Ojeda-Aciego M (2013) On
complemented. Galois connections and soft computing. In: International work-
conference on artificial neural networks. Springer, pp 224–235
Proof (1)⇒(2): It is obvious, by Proposition 5.11(2). Georgescu G, Mureşan C (2014) Boolean lifting property for residuated
lattices. Soft Comput 18(11):2075–2089
(2)⇒(3): By γ (A) ⊆ (A) follows that any coannulet is Grätzer G (2011) Lattice Theory: Foundation. Springer, Berlin
principal. Let m be a minimal prime filter. So m is Haveshki M, Mohamadhasani M (2015) On α-filters of BL-algebras. J
an ω-filter and it means that m = F(x) for some Intell Fuzzy Syst 28(1):373–382
x ∈ A. If m is dense, follows that x ⊥ = 1 and it Idziak PM (1984) Lattice operations in BCK-algebras. Math Jpn
29:839–846
contradicts with Proposition 2.17. Jayaram C (1986) Prime α-ideals in a 0-distributive lattice. Indian J
(3)⇒(4): Let P be a prime α-filter. By Lemma 5.20, P is Pure Appl Math 17(3):331–337
a minimal prime filter and so it is non-dense. By Jipsen P, Tsinakis C (2002) A survey of residuated lattices. In: Martinez
proposition 2.16, P is a coannulet and this means J (ed) Ordered Algebraic Structures. Kluwer Academic Publishers,
Dordrecht, pp 19–56
that P is principal. Kist J (1963) Minimal prime ideals in commutative semigroups. Proc
(4)⇒(1): It follows by Proposition 5.19. Lond Math Soc 3(1):31–50
The rest is evident by Proposition 3.3.
Knox M, Levy R, McGovern WW, Shapiro J (2009) Generalizations
of complemented rings with applications to rings of functions. J
Acknowledgements The author is highly grateful to referees for their Algebra App 8(01):17–40
valuable comments and suggestions which were helpful in improving Pawar Y, Khopade S (2010) ‘α-ideals and annihilator ideals in 0-
this paper. distributive lattices’. Acta Universitatis Palackianae Olomucensis
Facultas Rerum Naturalium. Mathematica 49(1):63–74
Pawar Y, Mane D (1993) α-ideals in 0-distributive semilattices and o-
Compliance with ethical standards distributive lattices. Indian J Pure Appl Math 24(7–8):435–443
Rasouli S (2018) Generalized co-annihilators in residuated lattices. Ann
Univ Craiova-Math Comput Sci Ser 45(2):190–207
Conflict of interest The author declares that he has no conflict of inter- Rasouli S (2019) The going-up and going-down theorems in residuated
est. This article does not contain any studies with human participants lattices. Soft Comput 23(17):7621–7635. https://doi.org/10.1007/
or animals performed by any of the authors. s00500-019-03780-3
Rasouli S, Dehghani A (2018) The hull-kernel topology on residuated
lattices. arXiv preprint arXiv:1812.11510
References Rasouli S, Kondo M (2019) n-normal residuated lattices. Soft Comput.
https://doi.org/10.1007/s00500-019-04346-z
Speed T (1969a) Some remarks on a class of distributive lattices. J Aust
Banaschewski B (1964) On lattice-ordered groups. Fundamenta Math- Math Soc 9(3–4):289–296
ematicae 55(2):113–122 Speed T (1969b) Two congruences on distributive lattices. Bulletin de
Büchi JR (1948) Die boolesche partialordnung und die paarung von la Société Royale des Sciences de Liège 38:86–95
gefuegcn. Port Math 7:80–119 Speed T (1974) Spaces of ideals of distributive lattices II. Minimal
Buşneag D, Piciu D (2006) Residuated lattice of fractions relative to a ∧- prime ideals. J Aust Math Soc 18(1):54–72
closed system. Bulletin mathématique de la Société des Sciences Varlet J (1963) Contribution à l’étude des treillis pseudocomplémentés
Mathématiques de Roumanie 49(97)(1):13–24 et des treillis de stone. Mem de la Soc R des Sci de Liege, 5eme
Cornish WH (1972) Normal lattices. J Aust Math Soc 14(2):200–215 series 8(4):1–17
Cornish WH (1973) Annulets and α-ideals in a distributive lattice. J Varlet J (1968) A generalization of the notion of pseudo-
Aust Math Soc 15(1):70–77 complementedness. Bull Soc R Sci Liege 37:149–158
Dong YY, Xin XL (2019) α-filters and prime α-filter spaces in residu-
ated lattices. Soft Comput 23(10):3207–3216
Endo S (1961) On semi-hereditary rings. J Math Soc Jpn 13(2):109–119 Publisher’s Note Springer Nature remains neutral with regard to juris-
Evans M (1972) On commutative pp rings. Pac J Math 41(3):687–697 dictional claims in published maps and institutional affiliations.
123