Distributive congruence¶

A join-semilattice congruence whose equivalent inputs can be lifted to cover an ordered equivalent target above their join.

Version: v1 · 2026-10-25 · History
Domain-specific #: 14382
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomain: Semilattice Theory → Mathematics

Core Idea¶

A distributive congruence, in Schmidt's 1968 sense, is a congruence on a join-semilattice with a specific lifting property.^[1] If a join x∨y lies below an element u in its own congruence class, the two inputs can be replaced by congruent elements whose join reaches at least u.^[1] In symbols, for a join congruence θ,

x∨y ≤ u and x∨y ≡ u (mod θ) imply that some x₁ ≡ x (mod θ) and y₁ ≡ y (mod θ) satisfy x₁∨y₁ ≥ u.^[1]

Schmidt also observes that the replacements can be chosen with x₁≥x and y₁≥y.^[1] This is a property of a relation on one carrier. The same source calls a semilattice homomorphism distributive when its induced congruence has this property.^[1] The relation is the subject here. The qualifier “in Schmidt's 1968 sense” is essential because his 1981 work formulates a qualified join criterion; no cross-edition equivalence on all semilattices is assumed.^[2]

Structural Signature¶

Join-semilattice carrier: a set H with associative, commutative, idempotent join. Its order is a≤b exactly when a∨b=b. The inequalities in the test use this order.^[1]
Join congruence: an equivalence θ compatible with join. It permits x∨y and u to be compared as members of the same class and makes the quotient join well defined.^[1]
Ordered congruent target: the test begins with x∨y≤u and x∨y≡u.^[1] If either fails, that triple imposes no lifting obligation.
Class-preserving input lifts: representatives x₁≡x and y₁≡y must jointly reach u or higher.^[1] This is the defining demand; merely changing the output representative is not enough.

The quantifier is universal over every x,y,u that satisfies the premise. A single failed triple refutes the classification. The source's optional choice of upward representatives is a strengthened form of the same test, not an extra condition added to the definition.^[1]

What It Is Not¶

Distributivity of the semilattice carrier is a different property. The 1968 paper investigates when a distributive carrier's quotient remains distributive; it defines the congruence through the lift, not simply by inspecting the carrier or quotient.^[1] Its Lemma 4.2 proves that a distributive congruence on a distributive semilattice has a distributive quotient.^[1] The converse does not follow from that statement alone.

A monomial congruence instead has a greatest member in every class. Schmidt defines that class-top condition separately in Definition 4.7.^[1] Neither condition implies the other in general: the finite Boolean-to-M₃ kernel below has class tops but fails the lift, while the infinite prime-filter example passes the lift yet has an unbounded class. These concrete independence witnesses are derived from the definitions, not examples named in the paper.

The 1981 Definition 3 describes a join of weak-distributive monomial congruences, including the qualifier.^[2] Schmidt's 1968 Problem 2 asks about a qualified monomial decomposition.^[1] Neither passage licenses the frozen source page's unqualified claim that joining monomial congruences is an if-and-only-if test.

Scope of Application¶

The test belongs to join-semilattice and lattice congruence theory. It can be applied to a lattice after forgetting meet and retaining its join operation, as Schmidt explicitly does for join congruences.^[1] It matters in representation arguments where a quotient's distributive behavior and congruence constructions are at issue.

It does not require a meet operation, finite classes, or a greatest class member. The ideal-generated and prime-filter-local constructions below live on different semilattices and demonstrate literal reuse of the same four-role test. An arbitrary equivalence partition outside an ordered join structure has no such lifting formula.

Clarity¶

To check a proposed instance, specify the carrier and join, prove that θ is join compatible, and then examine every ordered congruent target above a two-input join. The phrase “distributive congruence” by itself does not say which historical test is intended; here the 1968 lifting formula fixes that ambiguity. Writing the premise and conclusion separately prevents the common mistake of checking only that equivalent output elements exist. They already exist by the premise; the question is whether they can be reached by equivalent replacements of both inputs.

Manages Complexity¶

The property replaces a potentially opaque statement about an entire quotient with a local lifting obligation involving two inputs and one ordered target. For a concrete relation, a proof can classify all ways the premise arises. In the three-element chain there is only one strict nontrivial case. In the infinite prime-filter construction, the proof handles infinitely many cases with one rule: enlarge an input already in the filter by the missing finite elements. This compression works only after join compatibility and the universal lift have both been established.

Abstract Reasoning¶

The positive proof pattern is: take arbitrary x,y,u satisfying the two premises, identify which congruence classes can contain a strict upper target, and construct x₁,y₁ in the input classes. For a negative result, retain a valid join congruence and find one target above a join that cannot be assembled from the input fibers. The Boolean-to-M₃ case below does exactly that.

The test explains why a map's kernel can lose a distributive quotient even when its source carrier is distributive. In Schmidt's Lemma 4.2 the lift is sufficient to preserve semilattice distributivity in the quotient.^[1] The counterexample's quotient is the nondistributive diamond M₃; its direct failed triple identifies the missing lift rather than relying solely on the lemma's contrapositive.

Knowledge Transfer¶

The same test transfers literally from a finite chain to an infinite generalized Boolean lattice because both have a join-induced order, a join congruence, and equivalent input replacements. The proof methods differ: finite case analysis versus a general set-enlargement argument. The transfer is mathematical, not a metaphor about “distributing” work or information. A relation on a different algebraic structure must first supply the required join and order before this name is used.

Examples¶

Lower ideal collapse in a three-element chain¶

Let H={0<a<1} with join max. Take the ideal I={0,a} and its generated congruence θ[I]; its classes are {0,a} and {1}. Schmidt's Example 1 asserts that ideal-generated congruences of distributive lattices are distributive join congruences.^[1] This particular chain is a derived instance, not a named example in the paper.

For a strict premise, the only possibility is x∨y=0<u=a, so x=y=0. Choose x₁=a≡0 and y₁=0≡0; their join is a=u. Every equality premise uses unchanged inputs. Mapped back: the chain supplies join/order; θ[I] supplies compatible classes; 0≤a is the ordered congruent target; a,0 are the required equivalent lifts.

Upper prime-filter collapse in an infinite Boolean setting¶

Let H be the finite subsets of the positive integers, joined by union. It is a distributive relatively complemented lattice without a greatest element. Let P be the sets containing 1, a dual prime ideal, and take the universal congruence on P. Schmidt's Example 2 extends such a congruence to a distributive join congruence on H.^[1] In this derived instance all sets in P form one class and each set outside P remains alone. The particular finite-subset carrier and filter are derived choices.

Join compatibility holds because union with any set keeps two members of P in P. If x∪y⊂u and x∪y≡u, both unions lie in P, so at least one of x,y contains 1. Add the finite set u\(x∪y) to that input. It remains in P, hence equivalent to the original input, and the new union equals u; the other input stays fixed. Equality cases need no change. Mapped back: finite sets supply join/order; the filter class and exterior singletons supply the congruence; u is the ordered congruent target; the enlarged filter input is the lift. The P class has no greatest finite set, so this distributive congruence is not monomial.

Near miss: a Boolean quotient onto the diamond¶

Let H be the subsets of {a,b,c} under union, and let M₃ be the five-element diamond join-semilattice. Send the empty set to 0, the three singleton sets to the three distinct atoms, and every set with at least two elements to 1. This is a join homomorphism, so its kernel θ is a join congruence. Take x={a}, y={b}, and u={a,b,c}. Then x∨y={a,b}⊂u, and both images are 1, so the premise holds. Yet the classes of x and y are each singleton blocks: their only equivalent replacements still join to {a,b}, below u. The lift fails. Mapped back: Boolean union supplies join/order; the kernel supplies a valid congruence; u is the ordered congruent target; the missing input lifts refute distributivity. Since H is finite, each kernel class has a top, making this a monomial but nondistributive congruence. This counterexample is derived, not printed by Schmidt.

Structural Tensions¶

The mathematical definition does not encode a design tradeoff. Its useful diagnostic tension is logical: quotient compatibility alone does not ensure the upward lift, and classwise top elements alone do not ensure it either. Those are separate proof obligations, not competing values. No invented optimization objective is attached to the formal type.

Structural–Framed Character¶

The identity is structural within mathematics: its truth depends on a carrier, join, congruence classes, and universally quantified inequalities. There is no evaluative ranking of congruences in the definition. Choosing a carrier and candidate relation is human practice, but those choices do not relax the test. Its vocabulary travels literally between semilattice examples; outside a join-ordered algebra it travels only as an analogy and should not be treated as another instance. One may construct a congruence to satisfy the lift or recognize the property in an existing relation. Either route is judged by the same quantified condition. Schmidt's 1968 publication supplies the historical naming context; no institution's designation makes a relation satisfy the formula. Its character: a formally testable, evaluatively neutral semilattice type whose literal reuse remains inside join-ordered algebra.

Structural Core vs. Domain Accent¶

Prime Equivalence Relation supplies the partition into reflexive, symmetric, transitive classes. The present entry adds join compatibility, order, and a universal lifting condition. Removing any of those additions leaves a broader relation, not this named formal type. The named type stays below Prime because its defining lift requires semilattice join and the order induced by it; Equivalence Relation carries only the portable skeleton of treating members as equivalent. A proof about generic equivalence classes cannot supply the missing lift, as the Boolean-to-M₃ near miss shows.

This entry is a kind of Equivalence Relation.

Equivalence Relation is the strict genus: every distributive congruence is an equivalence relation, while an arbitrary equivalence relation need not respect join or satisfy the lift. Monomial Congruence is a related formal type with neither subsumption direction; Weakly Distributive Homomorphism is related through semilattice theory but its two-carrier map test is a different identity.

Relationships to Other Abstractions¶

Current abstraction Distributive congruence Domain-specific

Parents (1) — outgoing typed relations

Distributive congruence is a kind of Equivalence Relation Prime

Every distributive congruence is an equivalence relation; join compatibility and Schmidt's 1968 lift make it narrower.

Hierarchy path (1) — routes to 1 parentless root

Distributive congruence → is a kind of Equivalence Relation

Neighborhood in Abstraction Space¶

Distributive congruence sits in a moderately populated region (56^th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Set-Theoretic & Order Structures (54 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-25

Not to Be Confused With¶

Distributive semilattice: a property of the carrier, not automatically of every congruence on it.
Distributive quotient: a consequence under Schmidt's stated hypotheses, not this definition by itself.
Monomial congruence: requires a greatest member in each class; it can pass or fail the lifting test.
Weakly distributive homomorphism: tests a map between semilattices, whereas this entry tests a congruence on one semilattice.
Unqualified monomial-join characterization: not established by Schmidt 1968 Problem 2; the 1981 wording has a weak-distributive qualifier.

References¶

[1] E. Tamás Schmidt, “Zur Charakterisierung der Kongruenzverbände der Verbände,” Matematický časopis 18(1) (1968), 3–20, §4, Definition 4.1 and Examples 1–2 (printed pp. 6–7), Definition 4.7 (p. 9), Problem 2 (p. 10). Full paper: https://dml.cz/bitstream/handle/10338.dmlcz/126757/MathSlov_18-1968-1_1.pdf registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r

[2] E. T. Schmidt, “The ideal lattice of a distributive lattice with 0 is the congruence lattice of a lattice,” Acta Scientiarum Mathematicarum 43 (1981), 153–168, Definitions 2–3 (printed pp. 156–157). Full issue: https://acta.bibl.u-szeged.hu/38622/1/math_043_fasc_001_002.pdf registry ↩a ↩b