Weakly Dicomplemented Lattices (WDLs)

• Weakly dicomplemented lattices are bounded lattices equipped with two dual unary operations that generalize negation from Boolean algebras and formal concept analysis.
• Their axioms enforce order reversal and decomposition properties, offering a robust framework for analyzing filter, congruence, and substructure theories.
• WDLs integrate complex substructures such as skeletons, dual skeletons, and Boolean centers, underpinning advanced representation and classification results.

A weakly dicomplemented lattice (WDL) is a bounded lattice $(L, \wedge, \vee, 0, 1)$ equipped with two unary operations, the weak complementation ${}^\triangle$ and the dual weak complementation ${}^\nabla$, which abstract and generalize notion of negation as it arises in concept algebras and Boolean algebras. WDLs axiomatize the equational theory of concept algebras in Formal Concept Analysis and provide a natural extension of Boolean algebraic structures, with significant variety in their filter, skeleton, and congruence theories.

1. Algebraic Structure and Axioms

A weakly dicomplemented lattice is defined as a structure $(L, \wedge, \vee, {}^\triangle, {}^\nabla, 0, 1)$ where $(L, \wedge, \vee, 0, 1)$ is a bounded lattice and the following axioms hold for all $x, y \in L$:

$$\begin{aligned} \mathbf{(1)} &\quad x \wedge x^\triangle = 0 &\quad \mathbf{(1')}\quad x \vee x^\nabla = 1 \ \mathbf{(2)} &\quad x \leq y \implies y^\triangle \leq x^\triangle &\quad \mathbf{(2')}\quad x \leq y \implies y^\nabla \leq x^\nabla \ \mathbf{(3)} &\quad (x \wedge y) \vee (x \wedge y^\triangle) = x &\quad \mathbf{(3')}\quad (x \vee y) \wedge (x \vee y^\nabla) = x \ \end{aligned}$$

The pair $(x^\triangle, x^\nabla)$ for each $x$ is termed the weak dicomplement of $x$ (Kwuida et al., 2010, Jeufack et al., 6 Oct 2025, Jeufack et al., 17 Jan 2026, Kwuida et al., 2019).

The boundedness axiom in the original definition is superfluous: any nonempty lattice with a unary operation satisfying (1)-(3) is itself bounded under suitable definitions of $0$ and $1$ derived from the operation, ensuring purely equational formulation of WDLs (Kwuida et al., 2010).

2. Connections to Concept Algebras and Boolean Algebras

WDLs are the algebraic core of concept algebras arising in Formal Concept Analysis. For a formal context $\mathbb{K} = (G, M, I)$, the concept lattice $\mathfrak{B}(\mathbb{K})$ is formed of pairs of subsets which are mutually derivable via inclusion relations. Defining

$$(A,B)^\triangle = ((G \setminus A)', (G \setminus A)'') \qquad (A,B)^\nabla = ((M \setminus B)', (M \setminus B)'')$$

turns the concept lattice into a WDL—thereby showing that all concept algebras are WDLs (Kwuida et al., 2010, Jeufack et al., 17 Jan 2026).

Boolean algebras $(B, \wedge, \vee, {}', 0, 1)$ form a special subclass of WDLs under the identification $x^\triangle = x^\nabla = x'$. Conversely, a WDL in which $x^\triangle = x^\nabla$ for all $x$ is necessarily a Boolean algebra, and the common unary operation is its unique complementation (Kwuida et al., 2010, Jeufack et al., 17 Jan 2026).

3. Skeletons, Dual Skeletons, and Boolean Center

Every WDL $L$ possesses canonical substructures:

• Skeleton: $S(L) = \{x \mid x^{\nabla \nabla} = x\}$
• Dual skeleton: $$\overline{S}(L) = \{x \mid x^{\triangle \triangle} = x\}$$
• Boolean center: $B(L) = \{x \mid x^\triangle = x^\nabla\}$

$S(L)$ and $\overline{S}(L)$ form ortholattices, with respective orthocomplementers $\nabla$ and $\triangle$. $B(L)$ is a Boolean subalgebra—the largest such contained in $L$. In general, $B(L) \subseteq S(L) \cap \overline{S}(L)$, though equality may fail in various examples. These skeletal sublattices structure the behavior of filters, congruences, and representations in WDLs (Jeufack et al., 17 Jan 2026, Kwuida et al., 2010, Jeufack et al., 6 Oct 2025).

4. Filters, Ideals, and S-Primary Structures

Normal filters and ideals are essential in analyzing congruences and quotient constructions:

• A normal filter $F$ is a lattice filter closed under the operation $x \mapsto x^{\triangle\nabla}$.
• The complete lattice of normal filters $NF(L)$ is isomorphic to the complete lattice of normal ideals $NI(L)$ via order-reversing bijections: $F \mapsto \{z \mid z \leq u^\nabla,\, u \in F\}$ and $I \mapsto \{z \mid z \geq u^\triangle,\, u \in I\}$.

The dual skeleton $\overline{S}(L)$ supports the notion of S-filters: filters in $L$ whose generators are contained in the dual skeleton. S-primary filters, defined by the property that $x \in F$ or $x^\triangle \in F$ for all $x$, correspond bijectively to prime filters in $\overline{S}(L)$. This establishes a categorical isomorphism between the lattice of S-filters and the lattice of filters on $\overline{S}(L)$ (Jeufack et al., 6 Oct 2025, Jeufack et al., 17 Jan 2026).

5. Congruence Theory and Substructure Classification

Every normal filter $F$ generates a congruence $\theta_F$ given by

$$\theta_F = \{ (x, y) \mid \exists u \in F,\ x \vee u^\triangle = y \vee u^\triangle \}.$$

All normal filter generated congruences in distributive WDLs permute, and the set of all congruences is isomorphic to $NF(L)$ in regular WDLs. The class of regular distributive WDLs satisfies permutability and the congruence extension property (CEP).

Simple WDLs are characterized by $NF(L) = \{\{1\}, L\}$, while subdirectly irreducible WDLs have $NF(L)$ with a unique coatom. In finite distributive cases, $NF(L) \cong F(B(L))$, the filter lattice of the Boolean center (Jeufack et al., 17 Jan 2026, Jeufack et al., 6 Oct 2025).

The table below summarizes key relationships:

Structure	Substructure in WDL	Algebraic Type
Skeleton $S(L)$	$x^{\nabla \nabla}=x$	Ortholattice
Dual skeleton $\overline{S}(L)$	$x^{\triangle\triangle}=x$	Ortholattice
Boolean Center $B(L)$	$x^\triangle=x^\nabla$	Boolean algebra
Normal Filters $NF(L)$		Complete lattice (not a sublattice of all filters)

6. Representation Results and Concept Algebra Isomorphism

A central question concerns whether every (complete) WDL is isomorphic to a concept algebra. The answer is negative: there exist complete WDLs (notably atomfree complete Boolean algebras) that fail to be isomorphic to any concept algebra. The obstruction arises from the interplay of weak negations and the essential crossing behavior in concept lattices, which cannot be realized in atomless Boolean algebras (Kwuida et al., 2010).

Nonetheless, every Boolean algebra embeds as a WDL into a concept algebra, and the canonical context construction via ultrafilters realizes Stone's representation theorem: every Boolean algebra is isomorphic to a field of sets, specifically the powerset algebra on its ultrafilters (Kwuida et al., 2010).

7. Existence, Representability, and Extremal Congruence Cardinalities

The structure theory of WDLs is constrained in terms of existence and extremal values:

• Every bounded lattice has canonical (trivial) weak and dual weak complementations.
• Nontrivial (representable) weak (co)complementations exist when suitable (co)atom structure is present; for instance, finite distributive lattices with two or more coatoms admit nontrivial representable weak complementations (Kwuida et al., 2019).
• Explicit characterization of the possible cardinalities of WDL congruence lattices for finite $n$-element lattices is available. The largest values for $|\Con_\mathrm{WDL}(L)|$ with $n \geq 7$ are $2^{n-2}+1$, $2^{n-3}+1$, $2^{n-4}+1$, and $2^{n-5}+1$, realized by specific chain and sum/horizontal sum structures (e.g., $C_n$, $C_{n-2}\oplus C_2$, $C_{n-4}\oplus (C_2\times C_2)$, $C_{n-3}\oplus N_5$), each with their unique WDL-structure (Kwuida et al., 2019).

Open problems include the classification of infinite congruence cardinalities for WDLs and their subvarieties (Kwuida et al., 2019).

---

The structural depth of WDLs connects algebraic logic, lattice theory, and concept analysis, underpinning both generalization of Boolean structures and the precise algebraic underpinning of formal concept dualities. Foundational results provide a categorical, filter-theoretic, and representational landscape for further algebraic, topological, and computational explorations.